25-hydroxyvitamin D upregulates L6 muscle cell differentiation induced by mononuclear cells via the Notch signaling pathway

During muscle regeneration, myoblasts engage in cross-talk with immune cells to achieve optimal proliferation and differentiation. In this process, cytokines secreted by immune cells are described to modulate the kinetic of muscle differentiation. Taking into account that immune and muscle cells are both targets of vitamin D, we investigated in vitro the impact of 25-hydroxyvitamin D (25(OH)D) on the transcriptional response of muscle cells in presence of mononuclear cells. To address this objective, an in vitro model of co-culture using L6 myogenic cell line and peripheral blood mononuclear cell (PBMC) isolated from rat was used and compared to L6 cultured alone. Cells were treated with 25(OH)D (125 nM) during the 6 days of differentiation. Gene expression of 25(OH)D metabolism actors, muscle differentiation and metabolism markers, and of Notch signaling pathway effectors were studied in L6 cells by qPCR. In mono-cultured L6 cells, a 25(OH)D treatment induced a 3-fold increase (p < 0.05) in VDR mRNA expression at 24 h while no change in mRNA expression of the muscle differentiation markers i.e. Myog, Myh2 and Des was observed. In the presence of PBMCs, the mRNA expression of these markers was enhanced (27.5 times for myogenin, p < 0.05) resulting in an overexpression of the Notch pathway effectors (Dll: 6.8-fold and Hes1: x3.8-fold, p < 0.05). The 25(OH)D counteracted these effects of the PBMCs on L6 gene expression with the exception of the interleukin 6 transcript and protein. In the present study, our in vitro approach demonstrates the importance of immune cells in stimulating muscle cell differentiation. Taken as a whole, the data show that 25(OH)D attenuates in vitro the Notch pathway-dependent effects of immune cells on muscle cell differentiation and energy metabolism.


Introduction
Adult skeletal muscle has a remarkable ability to regenerate following trauma. Because adult myo bers are terminally differentiated, the regeneration of skeletal muscle is largely dependent on a small population of resident cells termed satellite cells. The contribution of immune cells (i.e. macrophages and lymphocytes) in regulating satellite cell migration, proliferation and differentiation is a major process of muscle regeneration (1).
Muscle regeneration is a tightly coordinated process composed of four consecutive interlinked phases: (i) necrosis, (ii) in ammation, (iii) activation and differentiation of satellite cells i.e., muscle stem cells, in myocytes and (iv) fusion of myocytes and maturation of newly formed myo bers (4). Alternatively, after activation and proliferation, satellite cells return to their quiescent state until the next regeneration process (5). The signaling pathways and the transcription factors orchestrating muscle regeneration have been studied extensively. In sum, myogenesis is controlled by the sequential action of lineage determination markers (i.e., Pax3/Pax7) that act together with Six and with Myocyte enhancer factor-2 (Mef2) proteins to regulate muscle gene expression. Pax7 and Pax3 are thought to be the principal regulators of muscle cell speci cation and tissue (6,7). Satellite cells can be activated by numerous signals from the regenerative microenvironment, including those mediated by adhesion molecules or by growth factors as well as cytokines produced by neighboring cells such as resident immune cells (7). Vitamin D seems a likely candidate to stimulate muscle recovery and performance, as muscle and immune cells are preferred targets of this nutrient (8). The infusion of vitamin D in vivo led to an increase in muscle regeneration in different experimental models (9). Moreover, it is known to shift the T-cell response from a T helper 1 (Th1) to a Th2-mediated one, which reduces in ammation and promotes an immunosuppressive state (10) by decreasing the production of type 1 cytokines (IL-6, Interferon-γ (IFN-γ)) and increasing the production of type 2 cytokines (IL-4, IL-10) (11).
A recent clinical trial has failed to support the effectiveness of vitamin D supplementation on physical performance and infection rates in older adults (12). In contrast, epidemiologic studies have shown that circulating 25(OH)D level and muscle strength/function are positively correlated suggesting that a target of vitamin D is the skeletal muscle (13). Indeed, skeletal L6 muscle cells have been demonstrated to express the 1 α-hydroxylase enzyme (CYP27B1) and therefore are able to metabolize 25(OH)D in 1,25 dihydroxyvitamin D (1,25(OH) 2 D or calcitriol) which interacts with VDR (14).
In vitro studies have established that 25(OH) 2 D positively controls muscle anabolism and inhibits muscle cell proliferation, but stimulates myogenesis (15).
Furthermore, PBMCs including monocytes, T and B cells, express VDR and CYP27B1 enzyme and most likely contribute to the majority of the 1,25(OH) 2 D formed locally in the tissues (16,17). 1,25(OH) 2 D plays numerous roles through both genomic and non-genomic pathways (8,18). The genomic effects of 1,25(OH)2D are mediated by an interaction with a cytoplasmic nuclear vitamin D receptor (VDR) from the superfamily of ligand-activated transcription factors. The 1,25(OH)2D-VDR forms an heterodimeric complex with the Retinoid-X-Receptor (RXR) and regulates the expression of target genes with a vitamin D response element (VDRE) in their promoter. The nongenomic effects, still poorly understood, are initiated by the binding of 1,25(OH) 2 D to a distinct membrane receptor (mVDR) (19). This complex, after internalization, induces the entry of calcium via activation of calcium channels and thus activate the protein kinase C (PKC). Subsequently, this stimulates the activation of the Mitogen-Activated Protein Kinase (MAPK) and Extracellular-Regulated Protein Kinase (ERK) pathways (8,20).
There is evidence of VDR expression and a direct effect of vitamin D on precursor (15,21) and mature skeletal muscle cells (22), which provides a rationale for a role of vitamin D in muscle function. Our team have demonstrated that, in old rats, vitamin D de ciency down-regulates the Notch pathway, known to play a leading role in muscle regeneration (23). Furthermore, mice lacking VDRs show an abnormal skeletal muscle phenotype with smaller, variable muscle bers and the persistence of immature muscle gene expression during adult life, suggesting a role of vitamin D in muscle development (22).
Taking into account these data, we planned to characterize the impact of 25(OH)D on the transcriptional response of muscle cells in presence of mononuclear cells. For this, we assessed the in uence of 25(OH)D on L6 myogenic cell co-cultured with fresh mononuclear cells isolated from rat's blood by evaluating (i) the muscle differentiation and metabolism markers by transcriptomic analysis and (ii) cytokine production.

Page 4/16
The rat skeletal muscle-derived cell line L6 myoblasts from the American Type Culture Collection were handled according to the guidelines of the European Center for the Validation of Alternative Methods Task force based on the Guidance on Good Cell Culture Practices (24). Cells were grown on plates (2.10 4 cells/mL) in a proliferation medium composed of DMEM (PAN Biotech, France) supplemented with 10% fetal bovine serum (Biowest, France), 4 mM glutamine (Sigma, France), 1% of penicillin-streptomycin (PAN Biotech, France) and 1X non-essential amino acids (Sigma, France) in an atmosphere of 5% CO 2 at 37°C. After 24 h of proliferation, the culture medium was replaced by a differentiation medium containing 2% horse serum (Dominic Dutscher, France) instead of the fetal bovine serum. Then, the L6 cells were cultured for six days in the presence or not Quantitative RT-PCR analysis Total RNA was extracted from the L6 mono-cultures (n = 6) and from the co-cultures (n = 6) with Tri-Reagent ® (Sigma, France) according to the manufacturer's instructions. RNA concentrations were determined using a Nanodrop ND-8000 spectrophotometer (Thermo Scienti c, France) and reverse transcription was performed according to the manufacturer's protocol. A quantitative RT-PCR array was carried out on a Rotor Gene Real-Time PCR system (Qiagen, France). For the mono-cultured L6 RNA, the primers used are listed in Table 1 and for the cocultured RNA, PCR plates with 36 genes were used (rat RT 2 pro ler TM , Qiagen, Table 2). Data were analyzed using the comparative CT method, based on the formula 2 −ΔΔCT (26). Each transcript level was normalized to the Hprt1 housekeeping gene and compared to the transcript expression in the D1 control sample without 25(OH)D treatment.

Cytokine quanti cation
Co-cultured L6-PBMC supernatants were collected at D1 and D3 for quanti cation of IL-6 and IL-10 concentrations using Milliplex kit (map rat cytokine, Millipore, France) according to the manufacturer's instructions. The uorescence intensity was determined with Luminex System (Bio-Rad Laboratories, Germany).

Statistical analysis
Data analyses were performed using GraphPad Prism® 5.03 for Windows (GraphPad Software Inc., San Diego, CA, USA). The experimental design required a two-way ANOVA to discriminate between the time and the treatment effects followed by Newman-Keüls post-hoc test. The results are expressed as mean ± SEM. Differences could be considered statistically signi cant when the P value was less than 0.05. For transcriptome analysis, we considered a fold change lower than 0.5 or greater than 2 as signi cant.

25(OH)D induces an overexpression of VDR mRNA in L6 cells without effect on muscle differentiation gene expression
We assessed the impact of vitamin D on the expression of VDR mRNA in L6 cells. At D1, VDR expression was 3fold higher in the presence of 25(OH)D than in the control (1.0 vs 3.1 ± 0.5; p < 0.001; Fig. 1-A). A signi cant lower overexpression compared to the control was also observed at D3 (1.3-fold, p < 0.05) and D6 (2-fold, p < 0.05). VDR protein expression was determined by immuno uorescence staining ( Fig. 1-B). The VDR protein expression was more pronounced at D1 and D3 in 25(OH)D conditions than in the control. Moreover, this protein localization was in the perinuclear space at D1 and in the cytoplasm at D3 according to the histomorphologic changes of the cell during the differentiation. We also studied the expression of the 1α-hydroxylase gene (Cyp27b1), an enzyme

25(OH)D counteracts the effects of PBMCs on L6 gene expression except for IL-6 transcript
We determined the L6 cells gene expression co-cultured with PBMCs in presence or not (control) of 25(OH)D treatment for 3 days (125 nM) ( Table 2). Firstly, we considered mRNA levels in D3 controls against D1 ones to characterize the effect of PBMCs alone on L6 cells gene expression. At D3, a signi cant overexpression of myogenesis marker mRNA was observed for: Des (3.3-fold); Myog (27.5-fold) and Myh2 (3.2-fold). This upregulation was in accordance with Notch pathway Delta-1 (Dll) as well as Hairy and enhancer of split 1 (Hes1) overexpression (D3 vs D1: 6.8-fold and 3.8-fold, respectively). The Bone morphogenetic protein 4 (Bmp4) mRNA, a cell proliferation factor induced by the Notch pathway, was overexpressed 2.9-fold at D3. Interleukin-6 (Il-6) mRNA expression was also increased signi cantly (9.1-fold). The mRNA expression of F-box protein 32 (Fbxo32), one of the four subunits of the ubiquitin protein ligase complex, was upregulated at D3 in the presence of PBMCs (10.2-fold). Two metabolic marker transcripts were overexpressed i.e. Solute carrier family 2 member 4 (Slc2a4 or Glut 4) (2.6-fold) and ATPase sarco/endoplasmic reticulum Ca 2+ transporting 1 (Atp2a1), a Ca 2+ -ATPase pump gene (42.1fold).
Secondly, we assessed the impact of 25(OH)D on L6 co-cultured with PBMCs considering mRNA levels in the 25(OH)D group against the control for each D1 and D3 (Table 2) Table 3). Comparison of IL-6 and IL-10 concentrations at D3 vs D1 in the 25(OH)D treated cells showed an increase but not signi cant according to the high inter-individual variability (IL-6: 2124 ± 1072 vs 545 ± 474 pg/mL; IL-10: 21.6 ± 9.9 vs 72.6 ± 27.3 pg/mL; Table 3). Notably, the increase of the IL-6 level appeared more pronounced (12.2 ± 5.0) than the IL-10 one (5.8 ± 2.1). No signi cant increase in the IL-6/IL-10 ratio was observed between D3 and D1 (Table 3).

Discussion
This study aimed to evaluate, in vitro, how muscle cell differentiation is in uenced by a 25(OH)D treatment in the presence of immune cells. The dose of 125 nM of 25(OH)D was chosen to mimic blood physiological condition (27,28).
The 25(OH)D treatment on mono-cultured L6 cells induced an upregulation of VDR mRNA and protein expressions with a decreasing effect during differentiation kinetics (from D1 to D6). This suggests that 25(OH)D is converted into active 1,25(OH) 2 D which auto-regulates the expression of the VDR gene through intronic and upstream enhancers as previously described (29,30). The 25(OH)D to 1,25(OH) 2 D conversion is performed by the CYP27B1 enzyme expressed in muscle cells. As expected, the mRNA of this enzyme was underexpressed at D1 25(OH)D treatment due to the repressive effect of VDR on gene transcription (31). After that, the Cyp27b1 mRNA expression was upregulated simultaneously to the decrease of the VDR mRNA and protein expressions associated to the migration of VDR protein from the perinuclear space to the cytoplasm. In the light of these ndings, the enhancement of the Cyp27b1 expression seems to be a major key to the effects of 25(OH)D on muscle cell metabolism. Similarly, in a recent study, Sustova et al showed that 25(OH)D on C2C12 muscle cells induced an overexpression of Cyp27b1 either directly or upon IL-6 stimulation (32).
Among the muscle differentiation markers determined, only Myog mRNA expression was overexpressed from 72 h without effects of 25(OH)D treatment. This is in agreement with the van der Meijden's study which reported in C2C12 mouse myoblasts that the expression of MyoD and ki67 were not signi cantly affected by both 25(OH)D or 1,25(OH) 2 D3 (33) Considering that, in vivo, the proliferation and the differentiation of muscle cells are facilitated by the surrounding immune cells (34), we used an in vitro model of L6 muscle cell and PBMCs insert co-culture in the presence of 125nM of 25(OH)D or not. We focused on a 24 h and a 72 h time lapse for differentiation when gene expression is most liable to be modulated (15,35).
In the presence of PBMCs without 25(OH)D, eleven L6 muscle cell genes from myogenesis, Notch signaling, cell proliferation and metabolism clusters were upregulated over time. In fact, PBMCs induced a signi cant overexpression of Des, Myog, Myh2 and Musk mRNA in the L6 cells in conjunction with Dll1 and Hes1, effectors of Notch signaling. Once the pathway is activated, the Notch receptor is cleaved and its intracellular domain acts as a transcription factor to induce Hes1 gene expression. (23,36). This highlight, for the rst time to our knowledge, the importance of the cross-talk between immune and muscle cells to promote the expression of muscle differentiation genes without a direct cell-cell contact. An overexpression of both Bmp4, a key gene in the regulation of cell proliferation (37,38), and Il-6, a myokine described as a muscular proliferative factor (39), was observed. This follows with Serrano's study showing that IL-6 promotes murine satellite cell proliferation via regulation of the cell-cycle-associated genes cyclin D1 and c-myc (40). Finally, an overexpression of genes implicated in some energetic metabolism pathways was observed. This included proteolysis (Fbxo32) as described before (41), glucose transport (Slc2a4) as shown in Broydell's study (42) and calcium in ux (Atp2a1) which plays an essential role in the regulation of intracellular Ca 2+ level and in the skeletal muscle differentiation (43). Taken together, these observations con rm the hypothesis of the crosstalk between immune and muscle cells to promote metabolic changes and muscle differentiation.
The treatment with 25(OH)D reduced the L6 gene overexpression induced by PBMCs for 5 genes: Myog, Myh2, Hes1, Fbxo32 and Atp2a1. Concerning Myog, this nding is coherent with the study undertaken by Endo which demonstrated an upregulation of myogenin mRNA expression in VDR null mice (22). For Myh2, the data are consistent with a previous study on L6 cell line showing a downregulation by a VDR knockdown, reversed after vitamin D treatment (44). The decreased level of Hes1 mRNA appears to be consistent with the overexpression of the other Notch signaling factors such as Dll1 and Bmp4 (23). Hence, Hes1 protein could contribute to the negative feedback regulation of Notch signaling (45). The underexpression of Fbxo32 (or Atrogin 1) in C2C12 muscle cells was previously described in Chen's study (46) involving the FOXO1 signaling pathway as clari ed recently by Dzik (47). Moreover, the Il-6 mRNA overexpression in L6 cells induced by PBMCs is majored by 2-fold in the presence of 25(OH)D in link to the increase of IL-6 level in culture medium (12.2-fold). That can reinforce the promotion of L6 cells differentiation (48). The larger increase observed for IL-6 as opposed to IL-10 may be due to its double origin from both muscle and immune cells whereas IL-10 is exclusively produced by immune cells. Unexpectedly, vitamin D is described as having adverse effects on IL-6 production from these two cell types: i.e. inhibition on PBMCs (49,50) and activation on L6 cells (48).

Conclusions
In the present study, our in vitro approach con rmed the importance of 25(OH)D and immune cells in stimulating muscle cell differentiation. Taken as a whole, the data highlight that 25-hydroxyvitamin D attenuates the Notch pathway-dependent effects induced by immune cells on muscle differentiation and cell energy metabolism. . At the end of the experiment, the rats were sacri ced by decapitation after iso urane anaesthesia and all efforts were made to minimize animal suffering.

Consent for publication
All authors support the submission to this journal.

Availability of data and materials
All data generated or analyzed during this study are included in this manuscript or are available from the corresponding author on reasonable request.

Competing interests
The authors declare no con ict of interest.  Each transcript level was normalized to the Hprt1 housekeeping gene and compared to the transcript expression in the day 1 control sample without the 25(OH)D treatment by the 2 -ΔΔCT method. Data are means of fold change ± SEM (n = 6); D1: day 1; D3: day 3; Statistical analysis was performed using a two-way ANOVA to discriminate between the time and the treatment effects (p < 0.05). When the ANOVA indicated significant interactions, the Newman-Keüls post-hoc test was used. Superscript letters (a, b, c, d) indicate significant differences (p < 0.05); ns: not significant. Data are means ± SEM (n = 6); D1: day 1; D3: day 3.