Closing in on Vitamin D Action in Skeletal Muscle: Early Activity in Muscle Stem Cells?

It is now widely known that the vitamin D hormone 1,25-dihydroxyvitamin D₃ (1,25(OH)₂D₃) exerts its actions in multiple cell types not only to integrate regulatory control of mineral homeostasis through its effects on intestine, kidney and bone, but to modulate key functional activities in cells of the immune system, the skin, the cardiovascular system, and certainly other tissues as well (1, 2). These actions are mediated by the vitamin D receptor (VDR), which is a member of the steroid receptor transcription factor family that controls the expression of genes in a temporally sensitive and cell type-specific manner. Although early studies defined key components of this activity at single genes, more recent studies have both confirmed many of these fundamental findings on a genome-wide scale and provided a wealth of new functional insight (3, 4). A primary discovery, consistent with that of other transcription factors, is the finding that the VDR frequently acts at multiple sites within a gene locus and that these sites can be located many kilobases distal to the gene's transcriptional start site. This complexity has significant negative implications for the use of traditional methods that have been used to study transcription where gene segments are evaluated out of endogenous gene context as well as in the absence of organized chromatin architecture (5).

Vitamin D status has long been linked to definable biological activities in individual tissues. Thus, vitamin D deficiency can result in rickets and in a myriad of other distinct abnormalities that can arise as a result of changes in blood calcium and phosphorus levels (6). The need to distinguish between indirect activities that are secondary to the direct actions of the VDR in well-established target tissues is extremely important, because mechanisms responsible for pathophysiological alterations drive the selection of drugs for therapeutic treatment. As a consequence, it is not surprising that the presence of the VDR in tissues has become an appropriate hallmark determinant for distinguishing between direct and indirect actions. Interestingly, although the biological actions of vitamin D and its mechanisms in a number of tissues fall into this category, no tissues have been the topic of as much investigation as those related to cardiac and skeletal muscle (7). This is because as summarized in multiple reviews, there seems to be no doubt that normal vitamin D levels are necessary for the appropriate functioning of both cardiac and skeletal muscle tissue and that vitamin D deficiency can lead to distinct pathologies (8).

There have been a number of recent attempts to demonstrate the presence of the VDR in muscle and other tissues not considered to be traditional targets. The problem here is that these attempts are largely based upon detection of VDR transcripts by RT-PCR analysis, an exquisitely sensitive methodology that defines specific RNA transcript levels but leaves open the question of VDR protein product level. These analyses have established, however, that transcripts for the VDR in some tissues are strikingly lower than those seen in traditional vitamin D target cells. For example, VDR transcript levels in muscle in vivo are 3–4 logs lower than levels measured in the intestine and kidney, and relative concentrations in other tissues such as liver and brain are similarly low (9). These results almost ensure that the detection of the VDR protein by immunological means will be difficult, because the latter methodology is far less sensitive than RT-PCR analysis and in many cases may be conceptually unable to detect existing VDR protein levels. This prediction has proven correct as both earlier studies of VDR expression in muscle tissue as well as more recent studies using either Western blot analysis, immunohistochemistry or both have been highly problematic (10). At the very least, it highlights the necessity for key experimental controls, preferably those than can take advantage of both genetic VDR deletion in the identical tissues being examined (11, 12). The simultaneous and convincing measurement of the VDR in tissues known to be positive for VDR expression is also essential. Interestingly, recent studies have suggested that particular commercial antibodies may be artifact free relative to others and therefore of preferential usage (11). This conclusion needs to be viewed with caution, as commercial lots of antibodies vary and conditions under which these antibodies are exploited from laboratory to laboratory also vary. Accordingly, an antibody that is useful in one researcher's hands may not be as useful in another. Thus, the value of the essential controls outlined above takes on even greater significance. It is also worth noting that the quality of commercial antibody preparations varies considerably and may have a striking effect on the experimental outcome. For example, our own independent preparation of the monoclonal antibody 9A7 results in a VDR detection reagent that is far superior to any commercial antibody that we have examined (my unpublished data). It is also particularly sensitive for the analysis of human VDR. Despite this, the successful studies of Wang and DeLuca are extremely valuable (9). They indicate that the VDR cannot be detected directly in adult mouse muscle tissue even under conditions in which the appearance of the VDR in traditional tissues such as intestine and kidney is highly robust while not detectable in those same tissues derived from a VDR-null mouse. Given the discrepancy in sensitivity between RT-PCR analysis and immunohistochemistry, however, the only conclusion that can really be reached in such studies is that the level of VDR expression in muscle tissue is below the level of sensitivity of the assay.

More recent studies by Girgis et al (10), discussed in an earlier review (13), suggest an alternative explanation; the VDR is indeed detectable in muscle tissue. Although the evidence put forth by these authors is not particularly robust and lacks the rigorous controls mentioned above, it does support the idea that although low, the VDR is detectable in the skeletal muscle of adult mice by Western blot analysis. Perhaps of greatest interest, the authors also show that the level of the VDR in muscle tissue of young mice is considerably higher, although this observation has not yet been pursued from a mechanistic perspective. Accompanying this analysis is an experiment that demonstrates the relative ease with which the VDR can be detected in extracts from the mouse C2C12 muscle precursor cell line. This analysis follows numerous previous studies of the VDR in this cell line but falls victim to the fact that cultured stem cell lines frequently up-regulate or in some cases down-regulate expression of the VDR gene (as they do other genes) during differentiation in vitro and therefore cannot be used to establish whether the VDR is expressed in muscle in vivo. Nevertheless, despite these findings the cumulative weight of evidence derived from the studies of Girgis et al (10) and earlier by others seems to support the idea that the VDR is indeed expressed in muscle in vivo, albeit at strikingly low levels.

Low levels of VDR expression can be interpreted from several vantage points: 1) low levels of expression in all cell types within a tissue, 2) higher levels of expression in a smaller subset of cells within a tissue, and 3) significant levels of expression in cells that reside within a given tissue but that are of a distal origin, for example expression of the VDR in macrophages that are known to reside in almost all tissues. It is also important to note that despite its absence in normal tissues, VDR expression can be up-regulated as a function of disease, either directly in the tissue or as a result of an increase in the presence of invading cell types such as macrophages that frequently accompany a pathological process. Mass action suggests that receptor levels below a few molecules per cell are unlikely to be able to mediate biological response to 1,25(OH)₂D₃. On the other hand, a significant level of expression of the VDR in a selected cell population in a tissue can be fully functional. A similar conclusion can be drawn for the detection of the VDR in invading cells, whether in health or disease. Although past studies indicate that VDR expression is extremely low in liver, for example, recent experiments indicate that the VDR is actually expressed in hepatic stellate cells, a minor cell population in the liver, and that the biology associated with these cells during pathological prompting is highly relevant in the development of cirrhosis of the liver (14, 15). The inability to detect the VDR via immunohistochemistry in these focal pockets of stellate cells that comprise 10%–15% of the liver remains puzzling, however. Perhaps stellate cell activation prompts an up-regulation of the VDR in these cells during the disease process.

The recent paper by Olsson et al highlights the potential for the selectivity of VDR expression in muscle (16). In this study, the authors explored the presence of the VDR and the activity of 1,25(OH)₂D₃ in myoblasts derived from human skeletal tissue and in subsequently differentiated myotubes. Importantly, myoblasts represent muscle satellite stem cells that serve to replenish active muscle and may also function in the regeneration of deteriorated muscle as well. Although the present studies involve primary cell culture, a potential caveat to the final conclusions, the authors identify reasonable levels of VDR RNA and protein by RT-PCR and Western blot analyses, respectively, in cells derived directly from human muscle tissue and show that the levels of the VDR are highest in myoblasts and recede during differentiation and fusion into myotubes. Not surprisingly, the VDR was not detected directly in the human skeletal muscle biopsy. These changes in VDR expression as a function of differentiation are seen in other cell types in vivo, including those of hematopoietic cells. Accompanying the presence of the VDR in these studies is the observation that 1,25(OH)₂D₃ prompts the transcriptional regulation of a number of genes including not only CYP24A1, a quintessential VDR target that serves to curb the activity of the hormone, but additional genes as well that likely play a role in controlling the proliferative expansion of the myoblasts and curbing their differentiation and fusion into myotubes. The overall inhibitory action of these processes is somewhat surprising biologically, but the fact that VDR function was correlated to the regulation of gene expression was reassuring. Perhaps isolated muscle cells require paracrine or juxtacrine input from adjacent cell types within functional muscle in vivo, a hypothesis that will require further experimentation. Although not all the pieces of the muscle puzzle fit together perfectly, the idea that vitamin D action might be focused on a small subset of cells that are important to normal muscle structure, function and perhaps regeneration is an attractive one worthy of further exploration.

Regardless of the levels of the VDR detected in muscle, a routine approach over the past several decades has been to delete the product of a specific gene in mice and to explore the consequence of that deletion on the phenotype that emerges. These types of studies are prevalent with respect to deletion of the VDR gene (17). With regard to this gene, it is interesting that mutations in the VDR gene that affect its activity or cause it to be absent in the human syndrome of hereditary 1,25(OH)₂D₃ resistance and to promote rickets and other skeletal deformities have not been reported to lead to an aberrant muscle phenotype (18). Both cardiac and skeletal muscle defects do appear in global VDR-null mice (19), however, although aberrant hormonal status and mineral homeostasis that are seen in these mice prevent the definitive conclusion that the phenotype is due to the loss of VDR expression in those tissues. The application of a rescue diet could serve to ameliorate these concerns, although rescued mice cannot be considered phenotypically normal, and in some cases the gene expression patterns remain altered as well. The VDR has been conditionally deleted in cardiac myocytes by crossing the VDR floxed mouse with a cardiomyocyte-specific myosin light chain-CRE driver, revealing specific defects in heart structure and function that are exaggerated during the onset of tissue fibrosis (20, 21). Importantly, a recent study has now emerged in which the VDR has been selectively deleted from skeletal muscle using myosin light chain-1f driven CRE (22). Unfortunately, rather than using this model to explore the role of the VDR in muscle, the authors explore the idea that the VDR influences insulin-resistance. Indeed, using C2C12 cells, an appropriate use of the cell line for this purpose, the authors follow up mechanistically and show that the VDR functions to retard the actions of Forkhead transcripton factor O1, which plays a role in controlling insulin sensitivity. Accordingly, the loss of the opposing actions of the VDR leaves Forkhead transcription factor O1 activity unchecked thereby promoting resistance. We might hope to see a more thorough examination of the effects of deleting the VDR on normal muscle physiology in this adult mouse and perhaps during aging in the future. Nevertheless, studies using this as well as additional mouse models could resolve the issue of VDR action in skeletal muscle and lead to a better understanding of the receptor's role in muscle, the identity of its target genes, and the biology that ensues.

The field is clearly closing in on the molecular mechanisms through which vitamin D is likely to influence skeletal muscle. The evidence suggests that a small population of extremely important cells, perhaps stem cells, in which the VDR is strongly expressed and capable of controlling muscle growth and perhaps regeneration may underlie vitamin D activity. Although this idea is an intriguing one, it seems reasonable that other vitamin D contributions may be identified as well. As with many emerging concepts, continued investigation to iron out potential confounding issues will be essential to a final understanding of the biology itself.