Human trials: impact of vitamin D status (observational study) and vitamin D supplementation (interventional study)

The observational study protocol was approved by the local Ethical Review Board of the General Clinical Research Center (GCRC) at Clermont-Ferrand, France, in accordance with the Declaration of Helsinki. Each participant gave written informed consent after being explained the purposes, methodology and potential risks of the study. This trial was registered at clinicaltrials.gov as NCT01066091. Thirty-one healthy male subjects, including 15 young (from 20–35 years old) and 16 older individuals (over 60 years old), with a body mass index (BMI) between 21.4 and 34.2 kg/m² were included in the study, as previously reported^29,30. Each subject had a normal blood biochemical profile and was sedentary. Participants were not taking any medications liable to affect outcome parameters, i.e., corticosteroids, b-adrenergic blockers, lipid-lowering agents, or anticoagulants. On the day of the experiment, subjects were studied after an overnight fast of at least 10 h. Indirect calorimetry was performed when subjects were awake, in supine position in a quiet room (Deltatrac, Datex, Geneva, Switzerland). Gas samples were collected every minute for 1 h. The data of the first 10 min were not used because this period corresponds to the stabilization of the metabolism of the subjects under the canopy. The means of the O₂ consumption and CO₂ production values obtained over the last 50 min period were used in the analysis. REE was calculated using the Weir equation³¹. Blood samples were then collected in EDTA-coated tubes. Plasma was prepared and stored at −80 °C until 25(OH)D concentration measurement using an ELISA kit (Promokine, Heidelberg, Germany).

The interventional study protocol was approved by the Ethics Committee of the institutional review board at Saint Charles Hospital, Beirut, Lebanon, in accordance with the Declaration of Helsinki. This trial was registered at clinicaltrials.gov as NCT02942732. One hundred and fifteen elderly subjects (mean age: 73.31 ± 2.05 years) with vitamin deficiency (25(OH)D < 20 ng/mL as per Institute of Medicine (IOM) recommendations) who were pre-sarcopenic (i.e., skeletal muscle mass/height² = 7.26 kg/m² for males and 5.45 kg/m² for females) were asked to join a 6-month randomized, controlled, double-blind intervention study^27,28. A written informed consent was signed by all the subjects which was approved by the Ethics Committee of the hospital’s review board. Participants were randomized into two groups: a vitamin D group (33 men and 27 women) received a supplement of 10,000 IU vitamin D3, i.e., cholecalciferol (Euro-Pharm International, Canada) three times per week, and a placebo group (26 men and 29 women) who received three times per week a similar-size and similar-color tablet containing microcrystalline cellulose (66.3%), starch (33.2%), and magnesium stearate (0.5%), per serving. The vitamin D and the placebo were presented in the same container. The interventions were led for a period of 6 months. Biochemical analyses and muscle assessments were performed at baseline and at six months. Fasting blood concentrations of vitamin D [25(OH)D] were measured by radioimmunoassay (DiaSorin, Stillwater, MN). Appendicular skeletal muscle mass was determined from bioimpedance analysis measurements (Tanita BC-418 Segmental Body Composition Analyzer, IL) and was expressed in kg^27,28. It should be noted that this method of measuring body composition is not the most accurate compared to magnetic resonance imaging or dual-energy X-ray absorptiometry. Handgrip strength was measured at 9:00 am in the dominant hand with a Martin vigorimeter (Martin; Elmed, Addison, IL), and the force was expressed in kg.

Rat experiment: long-term vitamin D depletion

All animal procedures were approved by the institution’s (INRAE) animal welfare committee to comply to the Ethical license of the national competent authority, securing full compliance with EU Directive 2010/63/EU on the use of animals for scientific purposes. Twenty 15-month-old male rattus norvegicus strain Wistar rats were purchased from the Janvier breeding center (Le Genest-St-Isle, France). The rats were housed at one rat per cage in opaque regulatory-sized cages. Sawdust was used as bedding for the animals and was renewed once a week. Room temperature and humidity were regulated at approximately 21 °C and 50%, respectively. Rats were held on a reversed 12-h light/dark cycle. Lights were off from 6:00am till 6:00 pm. Fat and lean body mass (g) were determined non-invasively using an echoMRI system as previously described¹⁵. The rats were then randomly divided into two groups according to body weight, fat mass and lean mass, and assigned (n = 10 per group, for a total of 20 rats) to either the AIN-93 M maintenance diet containing 1 IU vitamin D3/g diet (control rats) or to a modified AIN-93 M diet with no vitamin D for 9 months (vitamin D-depleted rats) (TestDiet, MO). During the experiment, 4 rats died or were euthanized because they developed a tumor. The results presented above are based on analyses conducted with data of the rats without any signs of underlying illness. Dioxygen consumption (VO₂), carbon dioxide production (VCO₂), food and drink intakes, and daily activity of control and vitamin D-depleted rats were measured using a PhenoMaster/LabMaster four-cage TSE System (Bad Homburg, Germany) at the end of the 9-month experimental period. Rats were housed individually in cages for indirect calorimetry. Energy expenditure was calculated using Weir’s equation(31). Food and drink consumption was measured using a weight sensor. Spontaneous activity was assessed using a three-dimensional meshing of light beams.

On the day of sacrifice, the remaining control (n = 7) and vitamin D-depleted (n = 9) rats were weighed and sacrificed by exsanguination under anesthesia by inhalation with a mix of isoflurane and oxygen. Blood was collected from the aorta artery. Liver, heart, adipose tissues and hindlimb skeletal muscles were removed, weighed, snap-frozen in liquid nitrogen, and stored at −80 °C until later analysis. Plasma 25(OH)-vitamin D concentration was measured by ELISA kit (Promokine, Heidelberg, Germany) according to the manufacturer’s instructions. ELISA kit was also used to determine parathormone (PTH) (Immunotopics, San Clemente, CA, USA). Plasma concentrations of calcium and phosphorus were assessed using a Konelab 20 analyzer (Thermo Electron, Waltham, MA, USA).

Mouse experiment 1: long-term vitamin D depletion

Procedures were conducted according to national and institutional guidelines on the care and use of animals and were reviewed and approved by the local Committee for the Care and Use of Laboratory Animals at Wageningen University to comply with EU Directive 2010/63/EU on the use of animals for scientific purposes. Thirty male 5-month-old C57BL/6j RJ mice were purchased from Janvier Laboratories (Le-Genest-Saint-Isle, France). Mice were housed in Makrolon® type-II cages individually to avoid dominance among pairs and to monitor dietary intake and prevent large differences in body weight within groups. The cages were cleaned every two weeks and contained standard woodchip bedding material. Room temperature and humidity were regulated at approximately 21 °C and 50%, respectively. Mice were held on a reversed 12-h light/dark cycle. Lights were off from 6:00am till 6:00 pm. All mice received a control diet (AIN-93M diet) until 10 months of age, and were then randomized according to body weight and fasting glucose concentration and subsequently subdivided to either a group (n = 15) that received the control diet or a group (n = 15) that received the same diet with no vitamin D for 14 months. Mice were weighed once every two weeks. Forearm grip strength was measured at 23 months of age, using a calibrated grip strength device (Panlab, Cornella, Spain). Body composition was measured by DEXA scan using a PIXImus imager (GE Lunar, Madison, WI). Six mice were removed during the study or showed pathologies at necropsy. The results presented below are based on the analyses conducted with data of the mice without any signs of underlying illness. At the end of the study, the remaining control (n = 10) and vitamin D-depleted (n = 9) mice were sacrificed under anesthesia by inhalation with a mix of isoflurane and oxygen, and blood was collected. Hindlimb skeletal muscles were removed, weighed, snap-frozen in liquid nitrogen, and stored at −80 °C until later analysis.

Mouse experiment 2: generation of human skeletal actin-MCM-VDRfl/fl transgenic mice

All animal procedures were approved by the institution’s (INRAE) animal welfare committee to comply with the Ethical license of the national competent authority, securing full compliance with EU Directive 2010/63/EU on the use of animals for scientific purposes. Mice with flox sites flanking exon 2 of the VDR gene (referred to as VDRfl/fl mice, gifted by Dr Geert Carmeliet, Katholieke Universiteit Leuven, Leuven, Belgium) and mice containing a skeletal muscle tissue-specific tamoxifen-inducible Cre recombinase (MER-Cre-MER) (referred to as HSA-MCM mice, gifted by Dr. Karyn Esser, University of Kentucky, Lexington, KY) were backcrossed to C57Bl6/J background mice ten times^32,33. Then, VDRfl/fl mice and HSA-MCM mice were bred to obtain HSA-MCM–VDRfl/fl transgenic mice. Ten adult HSA-MCM-VDRfl/fl mice received oral gavage of tamoxifen dissolved in corn oil (0.2 mg per gram of body weight) or vehicle solution for five consecutive days the first week, two consecutive days two weeks later, and two consecutive days each following month. After 15 weeks of tamoxifen or vehicle treatment, mice were fasted overnight and then euthanized under anesthesia by inhalation with a mix of isoflurane and oxygen. Liver, kidney, small intestine and hindlimb skeletal muscles were removed, weighed, snap-frozen in liquid nitrogen and stored at −80 °C until later analysis. The genotype of the mice was determined by genomic PCR at birth and confirmed at the beginning of the muscle analyses.

Cell culture

C2C12 myoblasts (ATCC, Manassas, VA) were grown to 80–90% confluence in DMEM supplemented with 10% fetal calf serum at 37 °C in a 5% CO₂-humidified atmosphere. Cells were then induced to differentiate into myotubes by switching to DMEM containing 2% heat-inactivated horse serum^15,34. Human primary skeletal muscle myoblasts (Lot#: SLSK002) were purchased from Zenbio inc. (New York City, NY) and cultured following the company’s instructions. After 5 days of differentiation, C2C12 myotubes or human skeletal myotubes were cultured for 3 days with 0, 1, or 10 nM of 1,25(OH)2 vitamin D3 dissolved in ethanol^15,34.

Western blot analysis

C2C12 myotubes and mice gastrocnemius muscles were homogenized in an ice-cold lysis buffer (50 mM HEPES pH 7.4, 150 mM NaCl, 10 mM EDTA, 10 mM NaPPi, 25 mM β-glycerophosphate, 100 mM NaF, 2 mM Na orthovanadate, 10% glycerol, 1% Triton X-100) containing a protease-inhibitor cocktail (1%) (Sigma # P8340) as we previously described^15,34,35. Denatured proteins were separated by SDS-PAGE on 4–15% precast polyacrylamide gel (BIO-RAD, Marnes-la-Coquette, France) and transferred to a polyvinylidene membrane (Millipore, Molsheim, France). Immunoblots were incubated in a blocking buffer and then probed with primary antibodies: total OXPHOS Rodent WB Antibody Cocktail (Abcam, Paris, France), Complex IV subunit IV antibody (ThermoFisher Scientific, Courtaboeuf, France), NRF1 antibody (Genetex, Euromedex, Souffelweyersheim, France), NRF2 antibody (Genetex, Euromedex, Souffelweyersheim, France) or VDAC/Porin antibody (Biovision, Clinisciences, Nanterre, France). Immunoblots were then incubated with the corresponding horseradish peroxidase-conjugated secondary antibody (DAKO, Trappes, France). Luminescent visualization was done using ECL Western Blotting Substrate (Pierce, Thermo Fisher Scientific, Courtaboeuf, France) and an MF-ChemiBIS 2.0 imaging system (F.S.V.T., Courbevoie, France). The density of the bands was quantified using MultiGauge 3.2 software (Fujifilm Corporation). Ponceau S staining was used to normalize protein loading between samples. Data are expressed as fold change vs. control group value.

RNA and DNA analysis

Total RNAs and nuclear/mitochondrial DNA were extracted using Tri-Reagent (Euromedex, Mundolsheim, France) according to the manufacturer’s instructions and as reported in our previous publication³⁵. Total RNAs and mitochondrial and genomic DNA concentrations were quantified by measuring optical density at 260 nm^15,35. Concentrations of mRNAs corresponding to genes of interest were measured by reverse transcription followed by PCR using a Rotor-Gene Q system (Qiagen, Courtaboeuf, France). Five micrograms of total RNA was reverse-transcribed using SuperScript III reverse transcriptase and a combination of random hexamer and oligo dT primers (Invitrogen, Life Technologies, Saint- Aubin, France). cDNAs were diluted 1:60 before PCR analysis. PCR amplification was performed in a 10 µL total reaction volume. The PCR mixture contained 2.5 µL diluted cDNA template, 5 µL 2× Rotor-Gene SYBR Green PCR master mix (Qiagen, Courtaboeuf, France), and 0.5 µM forward and reverse primers. The amplification profile was initiated by a 5-min incubation at 95 °C to activate HotStarTaq Plus DNA Polymerase, followed by 40 cycles of two steps: 95 °C for 5 s (denaturation step) and 60 °C for 10 s (annealing/extension step). Relative mRNA concentrations were analyzed using Rotor-Gene software. Sevenfold serial dilutions from a mix of all undiluted cDNA were used for each target gene to construct linear standard curves from which the concentrations of the test sample were calculated. mRNA levels were normalized to the corresponding housekeeping genes depending on species and tissues as indicated in legends of Tables 1 and 2, Figs. 3 and 5, and Supplementary Fig. 3. Nuclear/mitochondrial DNA was isolated using Tri-Reagent followed by back extraction with 4 M guanidine thiocyanate, 50 mM sodium citrate, and 1 M Tris, and an alcohol precipitation as described previously by us³⁵. Mitochondrial DNA (mtDNA) content was determined by quantitative real-time PCR analysis using a Rotor-Gene Q system (Qiagen, Courtaboeuf, France). To this end, the levels of NADH dehydrogenase subunit 1 (ND1) (mitochondrial DNA) were normalized to the levels of beta-actin (genomic DNA). Data are expressed as fold change vs. control group value. The list of the primers used for real-time PCR amplification is reported in Supplementary Table 5.

Mitochondrial enzymatic assays

Fifty milligrams of frozen rat plantaris muscle, mice gastrocnemius muscle, C2C12 myotubes or human primary myotubes were homogenized with a glass–glass Potter in 9 volumes of homogenization buffer (225 mM mannitol, 75 mM sucrose, 10 mM Tris-HCl, 10 m EDTA, pH 7.2) and spun down at 650 g during 20 min at 4 °C. The supernatant was kept and the pellet was suspended in 9 volumes of homogenization buffer and submitted to the same procedure. Both supernatants were pooled and used for the assay as previously described^35,36,37,38. After protein quantification, activities of citrate synthase (CS), 3-hydroxyacyl-CoA dehydrogenase (HAD) and activities of complexes I–IV of the respiratory chain were spectrophotometrically assayed^{35,36,37,38,39}.

CS activity was measured by following, at 412 nm, the chemical reduction of 5,5-dithiobis(2-nitrobenzoicacid) (DTNB) by CoASH. The reaction mixture contained 200 mM Tris-HCl (pH 8.0), 300 µM acetyl-CoA (Sigma # A2181), 150 µM DTNB (Sigma # D8130), and 20 µg of supernatant proteins at 37 °C. The reaction was initiated by adding 500 µM oxaloacetate (Sigma # O4126).

HAD activity was assessed by measuring the disappearance of NADH at 340 nm in presence of acetoacetyl-CoA. The reaction mixture contained 100 mM triethanolamine-HCl (pH 7.0), 5 mM EDTA, 120 µM acetoacetyl-CoA (Sigma # A1625) and 60 µg of supernatant proteins at 26 °C. The reaction was initiated by adding 500 µM NADH (Sigma # N8129).

Complex I activity was measured by following the oxidation of NADH at 340 nm in presence of decylubiquinone. The reaction mixture contained 50 mM KH2PO4 (pH 7.5), 3.75 mg/ml bovine serum albumin (BSA) (Sigma # A2153), 75 µM decylubiquinone (Sigma # D7911), 0 or 5 µg/ml rotenone (Sigma # R8875) and 50 µg of supernatant proteins at 37 °C. The reaction was initiated by adding 100 µM NADH (Sigma # N8129). Complex I activity was determined by calculating the difference between the activities measured in the absence and in the presence of rotenone.

Measurement of complex II activity was performed by following the decrease in absorbance at 600 nm resulting from the reduction of 2,6-dichlorophenolindo-phenol (DCPIP). The reaction mixture contained 50 mM KH2PO4 (pH 7.5), 1 mM potassium cyanide (KCN) (Sigma # 207810), 2 µg/mL rotenone (Sigma # R8875), 20 mM succinate (Sigma # S7501), 100 µM ATP (Sigma # A7699), 100 µM DCPIP (Sigma # 33125) and 40 µg supernatant proteins at 37 °C. The reaction was initiated by adding 100 µM decylubiquinone (Sigma # D7911).

Complex III activity was determined after measuring, at 550 nm, ubiquinol cytochrome c oxidoreductase activity in the presence and in the absence of antimycin, a specific complex III inhibitor. The reaction mixture contained 80 mM KH2PO4 (pH 7.5), 2.5 mg/ml BSA (Sigma # A2153), 50 µM cytochrome c (Sigma # C7752), 1 mM potassium cyanide (KCN) (Sigma # 207810), 250 µM EDTA, 0 or 12.5 µg/ml antimycin A (Sigma # A8674) and 20 µg of supernatant proteins at 37 °C. The reaction was initiated by adding 200 µM decylubiquinol. Decylubiquinol was prepared by mixing decylubiquinone (Sigma # D7911) and sodium borohydride (Sigma # 452882) in ethanol that leads to a fast decyluquiquinone reduction. The complex III activity is calculated by subtracting the antimycin insensitive activity from the total activity.

Complex IV activity (cytochrome c-oxidase activity) was determined by monitoring at 550 nm the oxidation of cytochrome c at 37 °C. The reaction mixture contained 50 mM KH2PO4 (pH 7.0) and 100 µM reduced cytochrome c. The reaction was initiated by adding 20 µg of supernatant proteins. Reduced cytochrome c was prepared by adding sodium dithionite (Sigma # 157953) to a solution that contained 50 mM KH2PO4 (pH 7.0) and oxidized cytochrome c (Sigma # C7752).

All activities are expressed as fold change vs. control group value.

Respiration measurements

Rat plantaris muscles were harvested and muscle fibers were permeabilized in a solution containing 10 mM EGTA, 3 mM Mg2+, 20 mM taurine, 0.5 mM DTT, 20 mM imidazole, 0.1 M K+ 2-[N-morpholino]ethane sulfonic acid, pH 7.0, 5 mM ATP, 15 mM phosphocreatine and 50 µg/mL saponin⁴⁰. Mitochondrial oxygen consumption was monitored at 30 °C in a 1-mL thermostat-controlled chamber equipped with a Clark oxygen electrode. Respiratory substrates used were: pyruvate 10 mM with malate 10 mM, or succinate 25 mM in presence of rotenone 0.4 µg/mL, or ascorbate 3 mM and N,N,N′,N′-tetramethyl-p-phenylenediamine (TMPD) 0.5 mM. State 3 was obtained by adding 2 mM ADP. After each respiration experiment, the fibers were dried and weighed. Activities are expressed in natom O/min/mg of fiber^38,40.

Microarray analysis

RNA profiling in rat plantaris muscle and C2C12 myotubes was performed using a Rat GE 4x44K v3 Microarray Kit and a Mouse GE 4x44K v2 Microarray Kit, respectively (Agilent Technologies, Massy, France). Briefly, 200 ng of total RNA isolated from frozen muscle samples or C2C12 cells was labeled using the Low Input Quick Amp Labeling kit (CY3) (Agilent Technologies, Massy, France), and microarrays were hybridized and scanned following the manufacturer’s instructions. Data extraction was performed using Agilent Feature Extraction Software 11.5.1.1. Data quality analysis and processing was performed with bioconductor using Agi4x44PreProcess and Limma packages (Smyth GK, 2004). Background subtraction was performed using the Normexp correction with offset = 50. Quantile normalization was then applied. Features flagged in Feature Extraction as control, non-uniform outliers, saturating or too weak were excluded. Statistical analysis was performed on 26,902 probes for plantaris muscle samples and on 30,094 probes for C2C12 samples with the Limma package. Probes selected for further analysis had a p-value < 0.05 for plantaris muscle experiments and a Benjamini–Hochberg adjusted p-value < 0.05 for C2C12 experiments. Functional enrichment analysis was performed with g Profiler using a list of downregulated genes differentially expressed in plantaris muscles of vitamin D-depleted old rats and upregulated genes differentially expressed in C2C12 myotubes treated with 10 nM of 1,25(OH)2 vitamin D3. Functional enrichment results were filtered and only the top 10 annotations for Gene Ontology BP, CC, KEGG, and Reactome were kept for functional enrichment visualization using Cytoscape (v3.8) with EnrichmentMap and AutoAnnotate plugins^21,41. Datasets are available from the GEO database (GSE67274 and GSE64803).

Statistics and reproducibility

All data are presented as means ± SEM.

For the human vitamin D supplementation study (interventional study), the sample size was calculated to detect the difference of appendicular skeletal muscle mass at 6 months (primary efficacy endpoint) between the two treatment arms with a two-sided type-I error of 0.05 and a power of 80%.

Statistical analysis for the human interventional study, the long-term vitamin D depletion rat experiment, the HSA-MCM-VDRfl/fl transgenic mice study and the cell culture studies with two conditions was performed using a two-tailed unpaired t test. In the long-term vitamin D depletion mouse experiment, the Mann–Whitney U-test was used to calculate between-group differences. In cell culture studies comparing more than two group means, one-way analysis of variance (ANOVA) was performed to test the effect of the experimental conditions. When a significant effect was detected, a post hoc Fisher test was applied to locate pairwise differences between conditions. Cell culture data are combined from at least two independent experiments. Linear correlations between variables were evaluated using Pearson’s correlation analysis. Statistical analysis was performed using StatView (version 4.02; Abacus Concepts, Berkeley, CA). Significance was set at p < 0.05.

Reporting summary

Further information on research design is available in the Nature Research Reporting Summary linked to this article.