No effect of 25-hydroxyvitamin D supplementation on the skeletal muscle transcriptome in vitamin D deficient frail older adults

Abstract

Objective: Vitamin D deficiency is common among older adults and has been linked to muscle weakness. Vitamin D supplementation has been proposed as a strategy to improve muscle function in older adults. The aim of this study was to investigate the effect of calcifediol (25-hydroxycholecalciferol) on whole genome gene expression in skeletal muscle of vitamin D deficient frail older adults.

Methods: A double-blind placebo-controlled trial was conducted in vitamin D deficient frail older adults (aged above 65), characterized by blood 25-hydroxycholecalciferol concentrations between 20 and 50 nmol/L. Subjects were randomized across the placebo group and the calcifediol group (10 μg per day). Muscle biopsies were obtained before and after 6 months of calcifediol (n = 10) or placebo (n = 12) supplementation and subjected to whole genome gene expression profiling using Affymetrix HuGene 2.1ST arrays.

Results: Expression of the vitamin D receptor gene was virtually undetectable in human skeletal muscle biopsies, with Ct values exceeding 30. Blood 25-hydroxycholecalciferol levels were significantly higher after calcifediol supplementation (87.3 ± 20.6 nmol/L) than after placebo (43.8 ± 14.1 nmol/L). No significant difference between treatment groups was observed on strength outcomes. The whole transcriptome effects of calcifediol and placebo were very weak, as indicated by the fact that correcting for multiple testing using false discovery rate did not yield any differentially expressed genes using any reasonable cut-offs (all q-values ~ 1). P-values were uniformly distributed across all genes, suggesting that low p-values are likely to be false positives. Partial least squares-discriminant analysis and principle component analysis was unable to separate treatment groups.

Conclusion: Calcifediol supplementation did not significantly affect the skeletal muscle transcriptome in frail older adults. Our findings indicate that vitamin D supplementation has no effects on skeletal muscle gene expression, suggesting that skeletal muscle may not be a direct target of vitamin D in older adults.

Trial registration: This study was registered at clinicaltrials.gov as NCT02349282 on January 28, 2015.

Keywords: 25-hydroxyvitamin D; Vitamin D; older adults; skeletal muscle; transcriptomics.

Conflict of interest statement

LdG declares to have filed a patent related to vitamin D and cognitive executive function. All other authors have nothing to declare.

Figures

Fig. 1

Very low expression of VDR in human skeletal muscle. qPCR Amplification curves of VDR in muscle biopsies. cDNA obtained from human muscle biopsies was PCR amplified using primers against human VDR. The different lines represent different subjects

Fig. 2

Minimal effect of calcifediol and placebo on skeletal muscle gene expression in vitamin D deficient frail older adults. Volcano plot showing the relation between signal log ratio (log2[fold-change], x-axis) and the -log10 of the IBMT P-value (y-axis) for the effect of calcifediol and placebo on the skeletal muscle transcriptome. The dotted lines show the threshold for fold-change of 2, and P-value of 0.001

Fig. 3

Minimal effect of calcifediol and placebo on skeletal muscle gene expression in vitamin D deficient frail older adults. Top left (a): P-values for the change in gene expression for all genes after filtering; before and after calcifediol supplementation. Uniform distribution (i.e. no increased frequency of genes for lower p-values) indicates an absence of an effect. Top right (b): values for the change in gene expression; before and after in the placebo group. Bottom left (c): P-values for the interaction effect (change in the calcifediol group vs. change in placebo group). Bottom right (d): receiver operator characteristic curve for the PLS-DA model

Fig. 4

Multilevel principle component analysis (PCA) of gene expression data could not distinguish the four experimental groups. Sample plot from an unsupervised multilevel PCA. In multilevel PCA the paired structure of the data, i.e. measurements performed before and after intervention on the same subject, is taken into account to eliminate the inherent between-subject variation. Results revealed no separation between the experimental groups

Fig. 5

Lack of effect of calcifediol supplementation on expression of putative VDR-dependent genes. Heatmap of gene expression changes (signal log ratio, SLR) of putative genomic targets of VDR

Fig. 6

Partially overlapping effect of calcifediol and placebo on muscle transcriptome at the pathway level. Gene set enrichment analysis was performed separately for the effect of calcifediol treatment and placebo treatment. Venn diagram showing overlap in upregulated gene sets (FDR q-value