The mechanisms of skeletal muscle atrophy in response to transient knockdown of the vitamin D receptor in vivo

Key points Reduced vitamin D receptor (VDR) expression prompts skeletal muscle atrophy. Atrophy occurs through catabolic processes, namely the induction of autophagy, while anabolism remains unchanged. In response to VDR‐knockdown mitochondrial function and related gene‐set expression is impaired. In vitro VDR knockdown induces myogenic dysregulation occurring through impaired differentiation. These results highlight the autonomous role the VDR has within skeletal muscle mass regulation. Abstract Vitamin D deficiency is estimated to affect ∼40% of the world's population and has been associated with impaired muscle maintenance. Vitamin D exerts its actions through the vitamin D receptor (VDR), the expression of which was recently confirmed in skeletal muscle, and its down‐regulation is linked to reduced muscle mass and functional decline. To identify potential mechanisms underlying muscle atrophy, we studied the impact of VDR knockdown (KD) on mature skeletal muscle in vivo, and myogenic regulation in vitro in C2C12 cells. Male Wistar rats underwent in vivo electrotransfer (IVE) to knock down the VDR in hind‐limb tibialis anterior (TA) muscle for 10 days. Comprehensive metabolic and physiological analysis was undertaken to define the influence loss of the VDR on muscle fibre composition, protein synthesis, anabolic and catabolic signalling, mitochondrial phenotype and gene expression. Finally, in vitro lentiviral transfection was used to induce sustained VDR‐KD in C2C12 cells to analyse myogenic regulation. Muscle VDR‐KD elicited atrophy through a reduction in total protein content, resulting in lower myofibre area. Activation of autophagic processes was observed, with no effect upon muscle protein synthesis or anabolic signalling. Furthermore, RNA‐sequencing analysis identified systematic down‐regulation of multiple mitochondrial respiration‐related protein and genesets. Finally, in vitro VDR‐knockdown impaired myogenesis (cell cycling, differentiation and myotube formation). Together, these data indicate a fundamental regulatory role of the VDR in the regulation of myogenesis and muscle mass, whereby it acts to maintain muscle mitochondrial function and limit autophagy.


Introduction
Vitamin D deficiency (25-hydroxyvitamin D <50 nmol/L) is globally widespread, affecting ∼40% of all adults (Forrest & Stuhldreher, 2011). The classical function of vitamin D is to regulate calcium (Ca 2+ ) and phosphate (P i ) homeostasis to maintain bone health and prevent rickets (Arthur W. Ham, 1934), osteomalacia (Bhan et al. 2010) and osteoporosis (Rizzoli et al. 2008). However, vitamin D deficiency has also been linked to deterioration of wider musculoskeletal health, including a reduction in skeletal muscle mass/function (Dhanwal et al. 2013), a higher risk of falls in old age (Bischoff-Ferrari et al. 2004b), and the exacerbation of cachexia (Dev et al. 2011) and type 2 diabetes (Pittas et al. 2007). Furthermore, chronic vitamin D deficiency in humans is accompanied by signs of myopathy that can be rescued by vitamin D supplementation (Prabhala et al. 2000). Similarly, vitamin D supplementation has been shown to enhance muscle function and myofibre cross-sectional area (CSA) in the elderly (Sato et al. 2005;Ceglia et al. 2013) and in athletes (Close et al. 2013;Wyon et al. 2016), although not consistently (Halfon et al. 2015). Nonetheless, in summary, there are multi-faceted links between vitamin D status and musculoskeletal health. This has led to efforts to elucidate mechanistic links between vitamin D and muscle mass and function.
Consistent with a pro-myogenic role of vitamin D, exogenous vitamin D promoted the differentiation of muscle cells in vitro (Okuno et al. 2012;Girgis et al. 2014a). Vitamin D exerts its genomic functions through the ubiquitously expressed vitamin D receptor (VDR) (Norman, 2008). Studies have confirmed expression of the VDR (Pike, 2014) and CYP27B1 [inactive 25-hydroxyvitamin D 3 > active 1α,25-dihydroxyvitamin D 3 (1α,25(OH) 2 D 3 )] in fully differentiated skeletal muscle of both rodents and humans (Srikuea et al. 2012). Exogenous vitamin D up-regulates VDR mRNA and vitamin D metabolism in skeletal muscle (van der Meijden et al. 2016). Furthermore, the VDR is required for vitamin D-induced anti-proliferative, pro-differentiation effects (Irazoqui et al. 2014), while acute VDR-knockdown (KD) [i.e. small interfering RNA (siRNA)] silencing results in impaired myogenic differentiation (Tanaka et al. 2014). However, acute siRNA of the VDR is inadequate for the characterization of the role of VDR throughout proliferation and terminal myogenesis. Nevertheless, the available in vitro data are consistent with a role of VDR expression in the regulation of skeletal muscle metabolism, in the absence of the manipulation of vitamin D bioavailability.
In vivo down-regulation of the VDR has been linked to multiple muscle catabolic disease states, while its transgenic deletion also induces muscle pathology. For instance, clinical studies have shown that VDR expression in muscle decreases with age (Bischoff-Ferrari et al. 2004a;M. Scimeca, F. Centofanti, M. Celi, E. Gasbarra, G. Novelli, A. Botta, 2018), as have in vivo experiments, in which whole-body VDR knockout (VDR-KO) mice have been shown to exhibit reductions in muscle fibre size and grip strength (Endo et al. 2003;Girgis et al. 2015). Furthermore, aberrant expression of myogenic regulatory factors (myf5 and myogenin) has been shown in VDR-KO muscle (Endo et al. 2003). However, whole-body VDR-KO impacts Ca 2+ /Pi homeostasis and bone metabolism, such that mice require a rescue diet to normalize blood mineral ion levels (Amling et al. 1999), which are notoriously difficult to control. Thus, developmental dysregulation and hypocalcaemia in VDR-KO models preclude the identification of a post-natal role of the VDR in skeletal muscle. While regulation of skeletal muscle mass and myogenic development by the VDR has been suggested, the mechanisms of this are ill-defined. Here, we report a comprehensive study into the autonomous role of the VDR in skeletal muscle. First, we demonstrate that a reduction of function (knockdown) of the VDR in a pre-clinical model in vivo results in myofibre atrophy and an associated up-regulation of autophagy-related transcriptional and post-transcriptional pathways, rather than down-regulation of protein synthesis. Second, we show that sustained in vitro VDR-KD impairs myogenesis (proliferation and myotube development).

Ethical approval
All animal experimental procedures were undertaken and approved by the Royal Veterinary College's Ethics and Welfare committee and were carried out under UK Home Office licence to comply with the Animals (Scientific Procedures) Act 1986.

Animal handling
Eight-week-old male Wistar rats were housed at 22 ± 0.5°C under a 12 h day/12 h night cycle and acclimatized to their new surroundings for 1 week. Animals were provided with water and a standard chow diet ad libitum (Special Diet Services, LBS Biotechnology, London, UK) [contained cholecalciferol (D 3 ) at 621.7 IU/kg]. Seven rats underwent in vivo electrotransfer (IVE), whereby after 10 days rats were fasted overnight and killed by injection of pentobarbitone. Muscles were rapidly dissected thereafter; transverse sections were mounted on cork tiles in optimum cutting temperature (OCT) medium and snap-frozen in liquid nitrogen-cooled isopentane. The remaining muscle was snap frozen by freeze-clamping and stored at −80°C.
In vivo electrotransfer IVE procedures were undertaken as previously described (Cleasby et al. 2007). Animals were operated upon under surgical depth anaesthesia, induced and maintained using isofluorane (2.5%), and their hind limbs shaved and prepared with ethanol. Tibialis anterior muscles received six spaced intramuscular injections of 50 μl aliquots of lenti shRNA particles prepared in endotoxin-free sterile saline at 0.5 mg/ml. Each shRNA cassette contained a U6 promoter, the target hairpin and a termination sequence. VDR-KD groups received VDR shRNA (Origene, Rockville, MD, USA) (Table 1) into the right tibialis muscle and scramble shRNA into the left tibialis muscle. Immediately following this, one 900 V/cm, 100 μs pulse and four 90 V/cm, 100 ms pulses were administered across the distal limb via tweezer-electrodes attached to an ECM-830 electroporator (BTX, Holliston, MA, USA). Animals subsequently received a subcutaneous injection of carprofen (50 mg/kg), before recovery and monitoring from anaesthesia.

Cell culture
C2C12 murine myoblasts (passage 6−10, ECACC, Salisbury, UK) were grown in Dulbecco's-modified Eagle's medium (DMEM, Invitrogen, Paisley, UK) supplemented with 10% (v/v) heat-inactivated fetal bovine serum (FBS, Sigma Aldrich, Poole, UK), penicillin (100 U/ml), streptomycin (100 μg/ml), amphotericin B (250 ng/ml), l-glutamine (2 mM; all Sigma-Aldrich) at 37°C in a 5% CO 2 atmosphere (growth media). Myoblasts were seeded in six-well plates (Nunclon Delta; Thermo Scientific, Loughborough, UK) and grown until ∼95% confluent (2−3 days); differentiation was induced by switching to medium containing 2% FBS (v/v) (differentiation media). Medium was changed every 48 h; however, in experiments in which signalling was assessed, medium was changed 24 h in advance to avoid associated acute perturbations. The lentiviral plasmid used was based on pLKO.1 Clone ID: RMM3981-201757375 and targeted the (3' untranslated region) mouse sequence 5 -TTA AATGTGATTGATCTCAGG-3 of the mouse VDR gene; a non-targeting scrambled (SCR) shRNA sequence was used as a negative control, with the hairpin sequence CCTAAGGTTAAGTCGCCCTCGCTCTAGCGAGGGC GACTTAACCTTAGG (Addgene, Cambridge, MA, USA). Oligonucleotides were obtained from ITDDNA USA and suspended, annealed and cloned into pLKO.1 at the EcoRI and AgeI sites as per the pLKO.1 protocol from Addgene. DH5α cells were transformed with the resultant plasmids for amplification and isolation. HEK293FT cells (Invitrogen, Carlsbad, CA, USA) were grown in DMEM, the 80−85% confluent plates were rinsed once with Opti-MEM (Invitrogen) and then incubated with Opti-MEM for 4 h before transfections. psPAX2 and pMD2.G, along with either scramble or pLKO.1 clones targeting mouse VDR (three clones), were added after mixing with Lipofectamine 2000, as per the manufacturer's instructions (Invitrogen). Opti-MEM was changed after overnight incubation with DMEM containing 10% FBS without antibiotics to allow cells to take up the plasmids and recover. Culture media was collected 36 and 72 h post-transfection to obtain viral particles. The viral particles present in the supernatant were harvested after 15 min of centrifugation at 1500 g to remove cellular debris. The supernatant was further filtered using a 0.45 μm syringe filter. Supernatant containing virus was either stored at −80°C for long-term storage or at 4°C for immediate use. C2C12 cells at 60% confluence were infected twice overnight with 3 ml of viral supernatant containing 8 μg/ml polybrene in serum-free, antibiotic-free DMEM. Fresh DMEM medium containing 10% FBS, antibiotics and 2 μg/ml puromycin (Sigma) were added the next day. Cells that survived under puromycin selection were either harvested as stable clones and stored or studied following differentiation.

Cell counts and BrdU assay
Myoblasts were seeded in six-well plates and counted by trypsinization of adherent cells into 1 ml DMEM. Measurements were made 96 h after seeding and each well was counted in quadruplicate using a haemocytometer. DNA synthesis of myoblasts was measured via the incorporation of BrdU in place of thymidine, and detected by a colorimetric ELISA assay (no. 6813, Cell Signalling Technology, Danvers, MA, USA) following the manufacturer's recommendations. Cells were seeded in a 96-well plate (Nunclon Delta; Thermo Scientific) with the addition of BrdU at a final 1× concentration 24 h before measurement.

Cell cycle analysis
Myoblasts were seeded in six-well plates and allowed to proliferate for 24 h before being synchronized to G1/early S phase by incubation with 2 mmol/l hydroxyurea (Sigma-Aldrich) for 15 h. The cell cycle lock was then released by changing the medium. After 12 h the cells were collected by trypsinization and centrifuged, before being re-suspended in PBS/1% FBS and fixed in ice cold 80% ethanol for at least 1 h. The cells were pelleted by centrifugation at 300 g and the ethanol-decanted pellet was washed in PBS. The cells were then incubated in propidium iodide (PI) staining solution [0.1% Triton X-100, 10 μg/ml PI (Sigma-Aldrich), 100 μg/ml RNase in PBS) for 30 min in the dark, before flow cytometry analysis on a Coulter FC 500 device (Beckman Coulter).

Myotube morphology of cultured cells
The medium was aspirated from the differentiated cells (Day 6) which were washed twice in 2 ml PBS before fixation in 2 ml PBS and methanol/acetone (2:1:1) for 5 min at room temperature. Fluoroshield mounting medium with DAPI (Sigma Aldrich) was applied to allow visualization of myonuclei. Normal light and fluorescent images were obtained and merged, and the myotube diameters were then measured using ImageJ software (National Institutes of Health, Frederick, MD, USA). Myotubes were defined as cells containing three or more myonuclei and measured at three equidistant points along their length. Five random fields of view were chosen per well from four wells, with five myotubes per view being measured on blinded images. Myotubes were stained with phalloidin (Life Technologies, Manchester, UK) after fixing in a 1:5 dilution of 2% horse serum/PBS before visualization.

Protein/DNA/RNA measurements
Total cellular alkaline soluble protein, DNA and RNA were analysed spectrophotometrically, as previously described (Crossland et al. 2013). For rat muscle samples, frozen whole muscle (∼10 mg) was desiccated under vacuum, before the addition of 0.3 mol/l NaOH and homogenization in a bead beater. For cell culture experiments, cells were scraped into 0.3 mol/l NaOH and incubated at 37°C for 30 min before an aliquot was taken for the measurement of total protein using a Nanodrop (Thermo Scientific). Thereafter, 1 M perchloric acid (PCA) was added to the remaining sample, which was incubated at 4°C for 10 min, before centrifugation at 3000 g for 10 min, washing in 0.2 mol/l PCA, and collection of the supernatant for RNA measurement. The resulting pellet was incubated at 70°C for 1 h in 2 M PCA, before centrifugation at 5000 g, washing in 2 M PCA, and collection of the supernatant for DNA measurement. RNA and DNA were quantified by measuring the absorbances at 260 and 275 nm, and 268 and 284 nm, respectively.

Measurement of muscle protein synthesis (MPS)
Cells. Measurements of MPS were made in differentiated cells using the stable isotope tracer deuterium oxide (D 2 O) (Crossland et al. 2013). The medium was changed 24 h before the addition of D 2 O (70 at.%) to an enrichment of 5%. Cells were incubated for 2 h under normal conditions, then 1 ml of medium was collected and cells were scraped into cold homogenization buffer [50 mM Tris·HCl (pH 7.4), 50 mM NaF, 10 mM β-glycerophosphate disodium salt, 1 mM EDTA, 1 mM EGTA and 1 mM activated Na 3 VO 4 ].
Rats. Seven days after IVE, the animals were administered a D 2 O bolus by oral gavage (7. 2 ml/kg, 70 at.%). For basal and maximal D 2 O body water enrichment, two animals were killed prior to and 2 h after oral gavage (respectively) and blood was collected in pre-chilled tubes containing lithium heparin. These were subsequently cold-centrifuged at 1750 g, and plasma aliquots were frozen at −80°C. Ten days after IVE, the animals were overnight fasted and killed, before blood and muscle was collected.
Samples were prepared as previously described (Wilkinson et al. 2014). Briefly, ∼50 mg of muscle was homogenized in ice-cold homogenization buffer, before continuous vortexing for 10 min, and centrifugation at 13,000 g for 5 min at 4°C. For cells, scraped lysates were homogenized on ice, before centrifugation as previously mentioned. Pellets were washed in 70% ethanol, before being hydrolysed overnight at 110°C in 1 ml of 0.1 mol/l HCl and 1 ml of H + dowex resin.
The hydrolysed amino acids were eluted into 2 mol/l NH 4 OH, then evaporated to dryness. The deuterium labelling of protein-bound alanine was determined though conversion to its tert-butyldimethysilyl derivative and assessed by single ion monitoring (SIM) of m/z 260 and 261 by gas chromatography-mass spectrometry. D 2 O precursor enrichment of both plasma and cell culture medium was measured using a modified acetone exchange method (Yang et al. 1998) that was previously used to assess MPS in cell culture (Crossland et al. 2013). Two microlitres of 10 M NaOH and 1 μl acetone were added to 100 μl of sample medium and vortex-mixed for 15 s, before incubation at room temperature for 24 h, allowing the exchange of hydrogen atoms for deuterium. Acetone was extracted into n-heptane and injected into a gas chromatograph-mass spectrophotometer, and D 2 O enrichment measured via SIM of m/z 58 and 59. For rats, basal plasma samples were used to determine the natural D 2 O enrichment, with maximal enrichment occurring 2 h after D 2 O ingestion and the final plasma sample used to average temporal precursor enrichment . All samples were referenced to a standard curve of known D 2 O enrichments, and fractional synthetic rate (FSR) was calculated using the following equation: where MPE Ala represents protein bound alanine enrichment, MPE MW represents precursor enrichment and t signifies time in hours (Gasier et al. 2010); 3.7 represents the average number of deuteriums labelled in the precursor alanine.

qRT-PCR
RNA from cultured cells and skeletal muscle was extracted in TRizol reagent (Invitrogen 15596026) and reverse transcribed using a High-Capacity cDNA Reverse Transcription kit (Applied Biosystems 4368814). Quantitative reverse transcriptase PCR (qRT-PCR) was performed on a ViiA7 Real-Time PCR system (Life Technologies) with SYBR Select Master Mix (Applied Biosystems 4472908) and primers designed in-house using Primer Express (Table 2). Quantification was performed using the 2 − CT method and normalized to GAPDH.

Western blotting
To quantify signalling molecules, rat muscles were homogenized in ice-cold homogenization buffer using clean sharp scissors. Samples were centrifuged at 11,000 g for 10 min at 4°C, and the supernatant was removed and quantified by a Nanodrop. Extraction of rat VDR proteins required homogenization and preparation in a hyperosmolar lysis buffer (HLB) [urea 6.7 M, glycerol 10%, Tris-HCl 10 mM, SDS 1%, DTT 1 mM, PMSF 1 mM, and Protease Inhibitor Cocktail tablet (Roche, West Sussex, UK)], as previously described (Girgis et al. 2014b). Cell lysates were homogenized and centrifuged as previously stated, before protein content was quantified by a Nanodrop. All samples were diluted in homogenization buffer and Laemmli loading buffer to the same concentration.
Samples were loaded onto Criterion XT Bis-Tris 12% SDS-PAGE gels (Bio-Rad, Hemel Hempstead, UK) for electrophoresis for 1 h at 200 V. Separated proteins were transferred onto a PVDF membrane for 45 min at 100 V, then blocked in 5% low-fat milk in Tris-buffered saline and 0.1% Tween-20 (TBST) for 1 h at room temperature. Membranes were then incubated at 4°C overnight in 5% milk in TBST primary antibody solutions. Afterwards, membranes were washed three times for 5 min with J Physiol 599.3

Gene
Accession number Primer sequence TBST and incubated for 1 h at room temperature in their respective HRP-conjugated secondary antibody, anti-rabbit (Cell Signalling Technologies) 1:2000 5% low-fat milk in TBST, anti-mouse (Cell Signalling Technologies) or 1:2000 5% low-fat milk in TBST.
Membranes were washed three times for 5 min in TBST, incubated for 5 min in enhanced chemiluminescence reagent (Millipore, Watford, UK) and visualized using a Chemidoc XRS. Bands were quantified using ImageLab software and normalized to total loaded protein visualized by Coomassie brilliant blue or Ponceau staining (Bass et al. 2017). Mitochondrial complex expression was undertaken as previously described (Ashcroft et al. 2020). MitoProfile OXPHOS antibody cocktail (110413) was purchased from Abcam and citrate synthase (SAB2701077) was from Sigma Aldrich.

Immunofluorescence
Muscle cross-sections 5 μm thick were cut at −22°C using a Cryostat, before mounting on glass slides and air-drying at room temperature. Sections were fixed in acetone/ethanol (3:1) for 5 min, and washed three times in PBS. Fibre CSA and VDR expression analysis was under-taken at the University of Birmingham. For CSA analysis, primary antibodies towards VDR (Rabbit, Ab109234, Abcam) and dystrophin [Mouse, MANDYS1(3B7), Developmental Studies Hybridoma Bank, Iowa City, IA, USA] were diluted in 5% goat serum in PBS at 1:50 and 1:200, respectively. Antibody solutions were applied to each section before incubation for 2 h in a humidity chamber at room temperature, then washed in PBS three times. Secondary fluorescent anti-rabbit (Alexa Fluor 594, A11012, Invitrogen) and anti-mouse (Alexa Fluor 488, A21121, Invitrogen) antibodies were diluted in PBS 1:200 and sections were incubated for 30 min as before. Slides were then washed three times in PBS, and 1:1000 DAPI stain (Invitrogen) was applied for 5 min before three PBS washes. Mounting medium (Invitrogen) was applied to each section and dried in darkness overnight. Additional sections were probed using anti-MHC IIa (SC-71) or anti-MHC IIb (BF-F3) diluted 1:50 in PBS. All sections were imaged in a blinded fashion using a Nikon Eclipse E600 and analysed using Image Pro3D capture software. Three random fields of view were measured per section.

RNA-sequencing analysis
RNA was extracted from snap-frozen muscle using an RNeasy mini kit (Qiagen, Valencia, CA, USA), following the manufacturer's recommendations. All RNA samples had RNA integrity number (RIN) (Schroeder et al. 2006) scores of greater than 8. RNA was prepared using the Tru-Seq RNA library preparation kit (Illumina) and RNA-sequencing (RNA-seq) was carried out by Edinburgh Genomics using the Illumina HiSeq 4000 platform, which generated 75 bp paired-end reads. Following quality control and base-calling, tag data were examined with Fast QC and adaptor sequences were trimmed where necessary using Trimmomatic (Bolger et al. 2014;Andrews, 2016). Unpaired reads were found to be of low quality and were dropped from the analysis, with no set of paired reads failing quality control. Alignment and feature counts were created using the Rsubread package in R and the edgeR package to examine differential expression Liao et al. 2013). The count data were filtered as recommended by the authors of edgeR by identifying the CPM at a count of 10 (Lun et al. 2016). Subsequent normalization was with the trimmed mean of M-values method (Robinson & Oshlack, 2010). Differential expression was analysed using the glmFit function of edgeR with design matrices to take account of biological pairing between treated and control limbs. Subsequent geneset testing was carried out using the GSEABase library (Morgan et al. 2017) in R, using genesets from the Molecular Signatures Database, maintained by the Broad Institute (Liberzon et al. 2011).
PathVisio v 3.3.0 (van Iersel et al. 2008;Kutmon et al. 2015) was used to construct all pathway analysis, with pathways from the WikiPathways repository (Kutmon et al. 2016). The Rattus norvegicus Derby Ensembl 91 database was used for the identity-mapping of genes, with log fold changes (</>0.26) of each gene mapped to pathway nodes and differently expressed genes were visualized.

Statistical analysis
The results are displayed as mean ± SD. The primary measure was the assessment of MPS, requiring a sample size of n = 7 animals for paired contralateral design (Power = 0.8, d = 1.64, α = 0.05). All analysis was performed by an unpaired (or paired where appropriate) t-test for two group comparisons or ANOVA with Bonferroni or Tukey post hoc analysis between multiple groups on GraphPad Prism7. A P value < 0.05 was considered to represent statistical significance.

In vivo VDR-KD induces skeletal muscle atrophy
To avoid the confounding effects of VDR-mediated developmental and hypocalcaemia-related dysregulation inherent to germline genetic manipulation, we modified VDR expression in post-natal skeletal muscle. Employing established IVE (Cleasby et al. 2005) of vectors expressing Vdr shRNA (or scrambled sequences in the contralateral limb, Fig. 1A), we confirmed Vdr knockdown (VDR-KD) in rat tibialis anterior (TA) by qRT-PCR (VDR-KD: −59 ± 20%) and immunoblotting (VDR-KD: −52 ± 29%) ( Fig. 1B and C). Functional protein suppression by IVE is persistent and acute inflammation abates after 7 days in skeletal muscle (Cleasby et al. 2005). To determine the effects of VDR-KD on fibre CSA, we immunostained fibres for dystrophin ( Fig. 2A), and showed that VDR-KD fibres were smaller (Fig. 2B) than contralateral control limb myofibres. Because tibialis muscle mainly consists of fast myosin isoforms, cryosections were immunostained for myosin IIa and IIb, which revealed marked reductions in type IIb fibre CSA following VDR-KD ( Fig. 2C and D). To verify myofibre atrophy, we also quantified muscle alkaline soluble protein (ASP) content and observed decreases in VDR-KD muscles (Fig. 2E). Analyses of total RNA and DNA content demonstrated no difference between VDR-KD and control muscle (Fig. 2F and G).

Anabolic signalling is unaffected, whereas autophagic processes are upregulated by VDR-KD
Because fibre atrophy can result from an impairment in MPS, we quantified MPS using the stable isotope tracer deuterium oxide (D 2 O) and found that global (i.e. across multiple muscle fractions, including myofibrillar, collagenous and sarcoplasmic) MPS rates were not affected by VDR-KD (VDR-KD 13.2 ± 4%d vs. Control 11.4 ± 2.7%d, Fig. 3A). Furthermore, neither protein abundance nor activity (phosphorylation) of indispensable muscle anabolic signalling pathways (AKT/mTORc1) were significantly reduced (e.g. p-mTOR VDR-KD 116 ± 78% vs. Control 100 ± 67%, Fig. 3B). These results suggest that the VDR-KD-induced myofibre atrophy and associated net loss of muscle protein were not due to changes in MPS, but instead due to greater proteolysis.

Discussion
Although recent work has demonstrated impaired regulation of skeletal muscle mass in the presence of vitamin D deficiency or low VDR expression (Dhanwal et al. 2013;Girgis et al. 2019), the function of the VDR in relation to skeletal muscle mass regulation remained poorly defined. Here, IVE transfection techniques were used to achieve short-term but sustained muscle VDR-KD in an internally controlled (i.e. contralateral limb) model, which also overcame the confounders of developmental dysregulation and dietary confounders (Amling et al. 1999;Endo et al. 2003). Using a combination of targeted and untargeted approaches, we show VDR loss-of-function rapidly induces muscle atrophy, and that this is specifically associated with an upregulation of autophagic processes.
First, we found a reduction of function in the VDR in vivo resulted in muscle atrophy at the level of individual fibres, along with an explanatory decrease in protein content. This is consistent with the findings of previous studies which showed that deletion of the VDR results in lower whole-body lean mass and function (Girgis et al. 2019), which occurs secondary to myofibre atrophy (Endo et al. 2003;Girgis et al. 2015). Muscle atrophy may occur as a result of a reduction in MPS, an increase in MPB or a combination of the two, culminating in a negative net protein balance (Rudrappa et al. 2016). In the present study, no reductions in mRNA translation-related signalling (i.e. AKT/mTORc1) were observed, nor were there detectable changes in mRNA translational efficiency/capacity processes (i.e. ribosomal content or directly quantified global/fraction-specific rates of MPS; Brook et al. 2016). Collectively, these data demonstrate that while the vitamin D/VDR augments AKT/mTORc1 signalling (Salles et al. 2013), no loss occurs during acute VDR deficiency. Weaker anabolic signalling (via RPS6 and P70S6K) has been observed in response Figure 7. In vitro VDR-KD impairs myogenesis and differentiation A, representative images of myotubes 5 days after differentiation, stained with phalloidin (red) and DAPI (blue). B, quantification of myotube population (n = 4). C, myotube diameter quantification (n = 4). D, quantification of the number of myonuclei per myotube (n = 4). E, total alkaline soluble protein, DNA and RNA content throughout differentiation (n = 6). F, western blot analysis of myosin IIx expression throughout differentiation (n = 6). G, myofibrillar protein synthesis rates 5 days after differentiation (n = 6). Scale bars represent 100 μm. For all grouped plots non-pairwise comparisons were made using two-sided t-tests. For time-course measurements, ANOVA with multiple comparison by Tukey analysis were used. Data are individual values with mean ± SD. * P < 0.05, * * * * P < 0.0001 between the groups indicated. † † P < 0.01, † † † P < 0.001 between VDR-KD and shRNA controls at that time point; § P < 0.05, § § P < 0.01 versus the initial time point. [Colour figure can be viewed at wileyonlinelibrary.com] to vitamin D deficiency (18 weeks) (Gogulothu et al. 2020); however, this is probably due to systemic chronic responses to vitamin D deficiency. Interestingly, it was previously proposed that vitamin D deficiency induces muscle atrophy through greater proteasome-mediated degradation (Bhat et al. 2013). However, we observed no modulation of proteasomal pathways; instead, previous findings may be related to Ca 2+ dysregulation, as evidenced by the expression of proteasomal markers and partial rescue of atrophy by Ca 2+ supplementation (Bhat et al. 2013).
In contrast, targeted measurement of proteolytic markers in VDR-KD muscle indicated an induction of autophagosome formation (phosphorylated AMPK; Kim et al. 2011), in addition to an upregulation of lysosomal enzymes (at the mRNA and protein level). Crucially, there were increases in the expression of Beclin 1 and LC3B-II, vital contributors to autophagosome formation, which demonstrates active autophagy induction (Wu et al. 2006). A role for the vitamin D-VDR axis in the regulation of autophagy has been previously reported. Incubation of vitamin D with neurones, macrophages and MCF-7 cells induces autophagy, while basal levels of autophagy are much higher in VDR-KO MCF-7 cells and mouse mammary glands (Yuk et al. 2009;Tavera-Mendoza et al. 2017), which suggests that LC3B is constitutively repressed by the VDR (Tavera-Mendoza et al. 2017). Vitamin D and subsequent VDR induction attenuates the dysfunction of autophagy that occurs in response to traumatic brain injury (Cui et al. 2017). Moreover, IL6/STAT3, which are essential regulators of autophagy, are upregulated in atrophied VDR-KO muscle, and this is partially rescued through STAT3 inhibition (Gopinath, 2017). Recent investigations have suggested direct VDR induction of IL6 and STAT3 phosphorylation in dendritic cells (Català-Moll et al. 2020), which further suggests a key role for the VDR in autophagy regulation. The present findings extend this to skeletal muscle and illustrate that in vivo ablation of the VDR may not actively induce autophagy; rather, its expression is inhibitory, preventing muscle protein degradation. Consistent with preliminary findings, follow-up global RNA-seq analysis demonstrated upregulation of established muscle atrophy-related genesets (lysosomal, including vesicle biogenesis and NF-kappaB signalling) (Jackman et al. 2013). Furthermore, VDR is a well-established TF, and ChIP-Seq datasets indicate that it has wide effects on gene expression (Satoh & Tabunoki, 2013) and regulates autophagy-related genes. These data build on and help to explain previous observations that genetic loss of the VDR results in a reduction in fibre size (through poorly defined mechanisms; Endo et al. 2003;Girgis et al. 2015), and perhaps the clinical links between differences in VDR expression and muscle atrophy in disease (Bischoff-Ferrari et al. 2004a;Punzi et al. 2012).
Indeed, in agreement with our findings, it has been suggested that VDR-KO-mediated atrophy may occur through unexplained reductions in protein content, rather than necrosis (i.e. protein degradation, not apoptosis) (Endo et al. 2003).
Analysis of the RNA-seq data revealed down-regulation of numerous mitochondrial and energy metabolism-related processes that have not previously been described in response to VDR-KD. Consistent with this, TF analysis of the RNA-seq data revealed that MEF2 TFs, which are essential for ATP production, complex I activity (She et al. 2011), and the maintenance of mitochondrial integrity and gene expression (e.g. PGC1-α) (Naya et al. 2002), were markedly down-regulated by VDR-KD, which implies that VDR is required for normal mitochondrial function. The importance of VDR expression has previously been suggested in other cell types: its expression has been shown to be vital in keratinocytes, MCF-7 cells and fibroblasts for normal mitochondrial regulation (Ricca et al. 2018), and there is some evidence that it may localize to the mitochondria (Silvagno et al. 2010). Indeed, cultured myoblasts exhibit lower mitochondrial respiration in response to VDR-KD, potentially through impaired mitochondrial organization, membrane permeability or calcium handling (Ashcroft et al. 2020), which implies that VDR is required for normal mitochondrial function in skeletal muscle. Interestingly, vitamin D may have additional regulatory functions. Recent work (Ryan et al. 2016) has demonstrated that the incubation of skeletal muscle cells with vitamin D increases mitochondrial oxygen consumption and fusion/fission dynamics, with ∼2000 associated mRNA gene expression changes that are dependent upon the VDR. Furthermore, vitamin D appears to have a physiological role, because supplementation in humans reduces muscle phosphocreatine recovery time (Sinha et al. 2013), which is an established index of mitochondrial oxidative capacity (Arnold et al. 1984). Here, the downregulation of multiple mitochondrial/TCA cycle genesets by VDR-KD reaffirms that the VDR is required for the maintenance of mitochondrial energetics (Ricca et al. 2018) and highlights it as a potential therapeutic target. It may also be that the maintenance of mitochondria is upstream of muscle mass preservation (Romanello & Sandri, 2016;Coen et al. 2019).
Given the putative role of the VDR in myogenesis, we investigated the potential mechanisms further by developing a sustained VDR-KD system in vitro, to study the impacts on proliferation and differentiation in skeletal muscle cells, somewhat reflecting the processes involved in development, although our short-term post-natal in vivo approach was designed not to capture these. Vitamin D administration provokes G 0 -G 1 cell cycle arrest, with induction of cyclin D3, p21 and p27, which are J Physiol 599.3 important regulators of cell cycle withdrawal (Irazoqui et al. 2014). Moreover, it has been demonstrated that they are induced by VDR-dependent genomic mechanisms (Irazoqui et al. 2014), with p21 being a vitamin D-target gene, containing a VDR promoter region binding site (Liu et al. 1996). Our findings extend and help to explain these observations, revealing the VDR is required for both cell cycle progression (G 0 -G 1 transition) and transition to differentiation. Importantly, MPS in established muscles remain unchanged (i.e. in vivo), demonstrating the decreased in vitro MPS is due to lack of myosin induction, highlighting the VDR requirement for successful differentiation and development of muscle fibres. Indeed, this clarifies why muscle developmental dysregulation occurs in in vivo VDR-KO models prior to study, which preclude the physiological characterization of a post-natal role of the VDR (Endo et al. 2003;Girgis et al. 2019). This may be evidenced by the lack of overall differentiation in VDR-KD cells (reflecting ongoing DNA synthesis) and the creation of very few, large myotubes. In contrast, mature skeletal muscle is post-mitotic and has limited capacity for satellite cell DNA synthesis (Drake et al. 2015). Previous in vitro studies have demonstrated that exogenous vitamin D positively mediates myofibre hypertrophy (Salles et al. 2013) and increases myosin expression (Garcia et al. 2011;van der Meijden et al. 2016) in a VDR-dependent fashion (Buitrago et al. 2013). Thus, concordant suppression of myofibrillar MPS and a lack of myosin induction in VDR-KD myotubes may explain the reductions in muscle mass and fibre size in VDR-null mice (Endo et al. 2003;Girgis et al. 2015), highlighting the key autonomous role of the VDR in myoblast differentiation, which does not require vitamin D stimulation.
Finally, it is important to note the limitations of this investigation. While these data demonstrate robust muscle fibre atrophy 10 days after IVE, earlier measures would allow for temporal CSA comparisons to be made. This would clarify whether VDR-KD-induced fibre atrophy was as a result of an arrest in fibre growth or overall reduction in CSA. Similarly, later measures (i.e. >10 days) may have permitted reductions in gene expression to manifest in protein expression. Furthermore, we did not directly measure the impact VDR-KD would have upon muscle function (i.e. strength). Moreover, while anabolic signalling remained unchanged, loss of the VDR may impact stimulation of these pathways by exercise or diet. Future studies may be able to clarify such responses and whether a reduction in VDR expression (as seen in aged muscle) may impair muscle mass gains.

Conclusion
The current study establishes an autonomous role of the VDR in skeletal muscle mass regulation, with lower expression eliciting myofibre atrophy through an induction of autophagy-related pathways, and myogenic dysregulation. Furthermore, the VDR plays a fundamental modulatory role in skeletal muscle mitochondrial function. These data suggest that modulation of VDR expression or VDR-targeting therapies may positively regulate skeletal muscle mass.