Major histocompatibility complex I‐induced endoplasmic reticulum stress mediates the secretion of pro‐inflammatory muscle‐derived cytokines

Abstract Major histocompatibility complex (MHC) I is an important component of intracellular antigen presentation. However, improper expression of MHC I upon the cell surface has been associated with several autoimmune diseases. Myositis is a rare acquired autoimmune disease which targets skeletal muscle, and MHC I overexpression on the surface of muscle fibres and immune cell infiltration are clinical hallmarks. MHC I overexpression may have an important pathogenic role, mediated by the activation of the endoplasmic reticulum (ER) stress response. Given the evidence that muscle is a diverse source of cytokines, we aimed to investigate whether MHC I overexpression can modify the profile of muscle‐derived cytokines and what role the ER stress pathway may play. Using C2C12 myoblasts we overexpressed MHC I with a H‐2kb vector in the presence or absence of salubrinal an ER stress pathway modifying compound. MHC I overexpression induced ER stress pathway activation and elevated cytokine gene expression. MHC I overexpression caused significant release of cytokines and chemokines, which was attenuated in the presence of salubrinal. Conditioned media from MHC I overexpressing cells induced in vitro T‐cell chemotaxis, atrophy of healthy myotubes and modified mitochondrial function, features which were attenuated in the presence of salubrinal. Collectively, these data suggest that MHC I overexpression can induce pro‐inflammatory cytokine/chemokine release from C2C12 myoblasts, a process which appears to be mediated in‐part by the ER stress pathway.


| 6033
THOMA et al. and infiltration of CD4 + and CD8 + T cells into perifascicular and endomysial regions around non-necrotic fibres. 1 The mechanisms underlying myositis remain poorly understood, though immune cell mediated and immune cell independent processes (e.g. ER stress) likely both play a role. [2][3][4] Overexpression of MHC I on and within muscle cells in myositis patients are associated with activation of the ER stress pathway within muscle cells, in both murine myositis models and in human disease. 5 Markers of ER stress pathway activation (e.g. Grp75 and Grp94) co-localize within muscle fibres staining positive for MHC I in myositis patient biopsies. 6 Moreover, activation of the ER stress response is known to activate inflammatory pathways, such as nuclear factor kappa B (NF-κB), a pleiotropic transcription factor capable of activating many downstream pathways to initiate and propagate inflammation. 7 Skeletal muscle is increasingly recognized as an endocrine organ capable of secreting a diverse range of peptides and proteins, components of the secretome. 8 The issue of muscle-derived cytokine production and release has mostly been examined in the context of exercise. 9 Studies of diagnostic muscle biopsies from myositis patients have demonstrated elevated gene expression of several cytokines and chemokines including: CCL3, CCL4, CCL5, IL-6, IL-15 and CCL2. 10,11 Specifically, an early study by Sugiura et al. (2000) demonstrated increased CD40 expression on both skeletal muscle cells and infiltrating mononuclear cells in patients with PM and DM that led to the elevated IL-6, IL-8 and IL-15 production by the muscle cells. 10 Further to this study, it was found that IL-15 was highly expression on the cytoplasm of PM-and DMderived myoblasts and only slightly expressed in infiltrating mononuclear cells or in healthy controls. 12 Muscle tissue homogenates from patients with DM also showed elevated IFNγ and IL-4 levels, which seemed to be negatively correlated with muscle strength grades. 13 A recent study has also suggested growth differentiation factor-15 as a myositis biomarker, which has been observed to be strongly expressed in both myositis sera and muscle tissue, specifically co-localizing with protein aggregates in sporadic IBM (sIBM) samples. 14 Overall, there has been increased interest in better understanding the potential (patho-) physiological roles of musclederived cytokines (myokines) in the inflammatory myopathies. 15 However, the source of biopsy-localized cytokines remains poorly understood, where muscle and infiltrating immune cells will both likely harbour these factors.
Given the established involvement of MHC I upregulation in myositis and association with ER stress, the ER overload response (EOR) and nuclear factor kappa B (NF-kB) activation 5 16 were grown in culture under standard conditions in Dulbecco's modified eagles' medium (DMEM) growth media supplemented with 10% foetal calf serum, 2 mM l-glutamine, then differentiated into mature myotubes over 7 days in media supplemented with 2% horse serum (HS). 17 To upregulate MHC I levels, myoblasts were transfected with 1 g of the MHC I (H2-k b ) overexpression plasmid using Lipofectamine™ 2000 reagent for a period of 6 h (Invitrogen), both without and with the addition of the ER stress blocking agent, salubrinal (1 M), with a green fluorescent protein (GFP) plasmid used as an additional control, transfection efficiency was between 40% and 70% for each construct. 18 Cells and conditioned media were harvested 18 h later in and stored at −80°C pending subsequent analyses. The plasmid used (MDH1-PGK-H2-Kb 2 ) was a gift from Chang-Zheng Chen (Addgene plasmid #17852). 19 Pharmacological induction of ER stress in C2C12 myotubes was achieved by treating cells with Tunicamycin (0.1 g/ ml) for 24 h. 20

| Conditioned media experiments and microscopy
Normal C2C12 myotubes were grown in culture as previously described, and following which, were treated for 5 days with conditioned media harvested from C2C12 myotubes, transfected with H2-k b , GFP with or without the presence of salubrinal. After 5 days of treatment with conditioned media, cells were imaged to assess myotube morphology/diameter. Images were acquired using a Nikon Eclipse TE 2000 microscope (Nikon Instruments B.V), and analysed using ImageJ software. Following imaging, cells were harvested in ice-cold PBS and stored at −80°C for gene expression analyses by qPCR.

| RNA extraction, cDNA synthesis and qPCR
RNA was extracted from myotubes using the TRIzol method, purified using Genejet RNA kit (Thermo Fisher Scientific) and cDNA synthesized using the iScript first strand kit (Bio-Rad). Quantitative PCR was performed with SYBR green (Roche Diagnostics) and analysed by the 2 −ΔΔC t method, with ribosomal protein S29 used at the housekeeping gene. 21 The primer sequences used are detailed in (Table 1).
The choice of the analysed ER stress pathway genes (i.e. PERK, IRE1 and ATF6) was made because these represent the three main branches of the ER stress pathway thus permitting analysis of the whole pathway. 22 In addition to examining Grp78, which has been previously examined in myositis. 5,23

| Multiplex cytokine analysis
Conditioned media from transfected cells were analysed for cytokine content by multiplex bead analysis with antibodies specific to IL-6, CXCL1, TNFα, CCL2, CCL4 and CCL5 using a Bioplex-200 platform in accordance with the manufacturers protocol (Bio-Rad). This choice of cytokine and chemokine targets was not designed to be comprehensive but based instead on the results from our previous work, which identified several specific cytokines/chemokines released from C2C12 myotubes. 17

| Chemotaxis assay
The murine CD4 + T lymphocyte cell line (BW5147.3) used in the chemotaxis assay studies was obtained from the American Tissue Culture Collection (ATCC), and cultured in suspension, in growth media comprising DMEM containing 10% HS, 2 mM l-glutamine.
The ability of conditioned media from H2-K b transfected cells (±salubrinal) to chemoattract murine CD4 + T lymphocytes was assessed using a CytoSelect 96-well Cell Migration Assay, with a lymphocyte specific 5 M pore filter (Cell Biolabs Inc), in accordance with the manufacturers protocol.

| Mitochondrial function analyses
C2C12 cells were cultured and exposed to conditioned media as previously described, in 8-well seahorse XFp plates. Real time oxygen consumption in cells was measured using a Seahorse XFp extracellular flux analyser, using the XFp mito stress test (Agilent Technologies, Manchester), performed in accordance with manufacturer's instructions. 20

| Statistical analysis
Data are presented as mean ± SEM; statistical analyses of data were undertaken using analysis of variance (anova) with Tukey post hoc test and Kruskal-Wallis test, following Shapiro-Wilk assessment of normality, using GraphPad Prism 8.

| MHC I overexpression resulted in increased cytokine gene expression and ER stress-induced release of chemotactic myokines
The overexpression of MHC I in C2C12 myotubes resulted in significant increases in cytokine gene expression ( Figure 2A) and cytokine release into the culture media, including of: IL-6, CXCL-1, CCL2, CCL4 and CCL5. Salubrinal treatment resulted in a reduced release of IL-6, CCL2, CCL4 and CCL5, but not of CXCL-1 ( Figure 2B).
In contrast, TNFα mRNA gene expression was undetectable, and cytokine release not increased in response to MHC I overexpression ( Figure 2B). Conditioned media from MHC I overexpressing cells induced significant chemotaxis of CD4 + T lymphocytes, an effect significantly reduced by salubrinal ( Figure 3). Treatment of C2C12 myotubes with tunicamycin induced the upregulation in the gene expression ( Figure 4A) and protein release of IL-6, CXCL1, CCL2 and CCL5 from normal C2C12 cells ( Figure 4B). Many cytokines have been detected in muscle tissue from myositis patients, and the current data suggest that muscle may also be a source of some of the chemokines and cytokines detectable in muscle biopsies from myositis patients. [25][26][27] The cytokines released by myotubes in our study do not reflect the full spectrum of cytokines detected in myositis patient muscle biopsies. Equally, infiltrating immune or vascular endothelial cells may also be significant sources for these and other cytokines detected in myositis patient biopsies. 25 cell model, given that this cytokine has so frequently and specifically been implicated in contributing directly to muscle dysfunction in a considerably variety of systemic diseases. 7 TNFα gene expression has, for instance, been detected in the sarcopenic skeletal muscle of older individuals, and co-localized to type I muscle fibres and in myositis patient biopsies. [29][30][31] However, there is little evidence to suggest that skeletal muscle secretes TNFα, though transcript levels are detectable in some cellular models, albeit at very low levels. 32 The expression of both CXCL and CCL chemokines in our model is consistent with similar in vitro studies, following TNFα stimulation. 17 CCL2 (MCP-1) has been reported to be associated with Tcells in non-necrotic muscle fibres of patients with PM and sIBM. 33 Moreover, the expression of CCL4 and CCL5, along with their requisite receptors has been reported in biopsies from patients with inflammatory myopathies. 34 The function of these potential myokines remains poorly understood.

| Conditioned media from MHC I overexpressing cells induced atrophy and mitochondrial dysfunction in untreated C2C12 cells
The release of IL-6 downstream of MHC I is an interesting observation, there is a significant juxtaposition towards the function of IL-6 in muscle. Studies have reported IL-6 can induce muscle atrophy 35 ; however, IL-6 has been shown to be protective in a murine model of myositis. 36 Despite these somewhat paradoxical observations, the pursuit of anti-IL-6 therapies in inflammatory myopathies (and wider autoimmune diseases) is highly prevalent. 37,38 The suppressed cytokine release in the presence of salubrinal is likely a consequence of its role in translational attenuation, through preventing the de-phosphorylation of eiF2-alpha. 18 Studies in Parkinson's disease have reported similar attenuation of the circulating levels of cytokines IL-1b, IL-6 and TNF-a in LPS treated mice. 39 Our findings support the hypothesis that, in myositis, musclederived cytokines may play an important local influential role to effect neighbouring fibres, and in creating a milieu, which could contribute to disease pathogenesis. 40 Mitochondrial dysfunction has been described in both experimental and human models of myositis; moreover, mitochondrial dysfunction has been associated with ER stress. 42,43 The mechanisms responsible are overall poorly characterized, however, exogenous factors such as type I interferon have been heavily implicated. 42 Muscle precursor cells have been reported as a potential source of type I interferon in myositis. 44 Moreover, recent data has shown primary cells from myositis patients display an intrinsic deficit in mitochondrial function. 45 Our data showed a reduction in overall OCR and evidence of mitochondrial dysfunction through increased proton leak (a marker of mitochondrial damage) in cells treated with conditioned media from MHC I overexpressing cells, where media from salubrinal treated cells ameliorated these changes.

F I G U R E 3 T-cell chemotaxis in
Furthermore, the decline in non-mitochondrial respiration suggests perhaps a deficiency in broader metabolic processes. Collectively, our data suggest that secreted factors from disease phenotype cells may affect function in neighbouring cells/fibres through augmenting the local milieu and compound the intrinsic mitochondrial deficits that have previously been reported. 45 The ER stress pathway is a cellular mechanism highly conserved across a wide range of eukaryotes. 46

ACK N OWLED G EM ENTS
The authors are deeply grateful to Professor Robert G Cooper for his invaluable input into the design and development of this study.

FU N D I N G I N FO R M ATI O N
This study was funded by the University of Liverpool, Myositis UK and The Manchester Metropolitan University.

CO N FLI C T O F I NTE R E S T
The authors have no competing interests to declare.

DATA AVA I L A B I L I T Y S TAT E M E N T
The data that support the findings of this study are available from the corresponding author upon reasonable request.