Expression of the Vitamin D Receptor in Skeletal Muscle: Are We There Yet?

The diverse biological actions of the vitamin D hormone 1,25-dihydroxyvitamin D₃ (1,25(OH)₂D₃) that are beyond its contribution to the maintenance of mineral metabolism are now well recognized (1). These actions include significant roles in skin maturation, protection, and function, the immune system, cardiovascular activity, neuromuscular function, bile acid metabolism, xenobiotic detoxification, muscle activity, and hepatic function (2). They also include broad cellular growth control mechanisms that include blockade of proliferation, prodifferentiation, induced apoptosis, and other fundamental cellular processes that may be of therapeutic relevance in cancer (3). Many of these activities have been described over a span of several decades both in cultured cells of specific lineage origin and in vivo. In the latter case, these studies have made frequent use of genetic strains of mice that are either globally or tissue specifically deficient in the expression of the vitamin D receptor (VDR), the central mediator of vitamin D action in all tissues (2, 4, 5). Perhaps most importantly, the beneficial biological effects of vitamin D in many of these systems appear to have translational and clinical components as well, because clinical pathologies associated with these systems frequently correlate with vitamin D deficiency, and at least, a subset have been shown to respond positively to increased vitamin D intake (6). Accordingly, the actions of 1,25(OH)₂D₃ in many of these tissues in humans are generally not in dispute.

The mechanism through which vitamin D acts in tissues centers on the presence and activity of the VDR, a transcription factor that is activated by hormonal vitamin D and functions at the level of the genome in a cell-specific manner to regulate the transcription of genes (5). This receptor is highly expressed in tissues, such as intestinal epithelial cells and proximal and distal tubules of the kidney, and mesenchymal lineage cells, such as chondrocytes, early osteoblast precursors, mature osteoblasts, and osteocytes, and in many other cell types as well (7). Molecularly cloned and the protein product studied over several decades, many of the principles of the VDR's modulatory actions to regulate the expression of specific genes are now well established, most recently at the genome-wide level (8–11). Although not fully understood, our advanced appreciation of the mechanisms through which the VDR operates at the genetic level stands in stark contrast to those that have been proposed to account for the so called rapid, nongenomic actions of the hormone, where the VDR or other “receptors” have been suggested to regulate specific membrane-associated activities in cells (12). Despite suggestions to the contrary, both the mechanisms that underlie these latter activities and their relevance in vivo remain to be determined, although considerable mechanistic precedent has been established for nongenomic activity through the study of nuclear receptors for estrogen, the androgens, and progesterone.

With these issues in mind, it is not surprising that studies aimed at understanding the actions of the vitamin D hormone in such tissues as the nervous system, the cardiovascular system, and the skeletal muscle system have focused upon detection of the VDR as an initial prerequisite for defining a direct mechanism of action of the hormone in these systems. Although this may appear to the outsider to be a relatively easy task, the VDR protein is expressed at extremely low levels and thus is not easily detected even in bone fide vitamin D-sensitive target tissues. Therefore, it is even more difficult to identify the receptor in atypical tissues where VDR abundance may be restricted to rare cell subtypes or is generally low for unknown reasons. Accordingly, these features lead to more fundamental questions: Is the VDR expressed in a particular tissue above cellular background, can its activity(s) be identified in unequivocal terms, is its activity localized to a specific cell type and not to a contaminating or invading cell type, and is it possible that the VDR manifests biological roles in tissues that are completely independent of those that involve transcription? These questions need to be answered before excluding the possibility that in certain tissues vitamin D's actions are indirect, perhaps through the regulation of local or systemic modulators, such as peptide or steroid-like hormones from other tissues or through the maintenance of extracellular calcium and phosphorus levels, which are known to impact muscle, bone, and nerve cell function. What also must be considered is the possibility that VDR expression can be selectively regulated in cell types that are temporally unique to tissues and organs, such as those that occur during development, growth, differentiation, and/or during physiologic states, such as pregnancy, lactation, and aging. Alternatively, VDR expression may appear in a specific cell type(s) as a consequence of disease. A final challenge is to interpret the expression of the VDR in cell lines and in primary cells in culture, where lineage relationships to the originating tissue are often times uncertain and frequently problematic. Efforts to identify the VDR in liver, cardiovascular tissues, skeletal muscle cells, and neurons of the central nervous system are particularly prone to these latter issues, because the VDR is generally found in cell lines derived from these tissues but has been difficult to detect in the primary tissues.

In this current volume, Girgis et al (13) examine whether the VDR is present in skeletal muscle both in vitro and in vivo and assess whether the receptor is capable of mediating biological responses. In view of the numerous and often conflicting studies that have attempted to address this question over the years, as summarized, in part, in table 1 of the manuscript, one would have hoped that the current study would resolve these issues experimentally once and for all. Although the observations are interesting, however, this is really not the case. In short, some of the major issues still remain major points of controversy. Experiments in this study are initiated using the mouse C2C12 cell line, and early mesenchymal lineage cell has been employed by muscle biologists for decades to study myoblast to myotube differentiation and by many previous investigators to support the concept that muscle cells contain the VDR (14–17). As in these earlier studies, the present authors show that the VDR is indeed readily expressed in this cell line (at the mRNA level and protein levels) and regulated by 1,25(OH)₂D₃. Because the VDR can be dynamically up-regulated in many cell lines passaged in culture as well as during differentiation, however, these observations documented by Girgis et al (13) do not really resolve the question of whether the VDR is expressed in skeletal muscle targets in vivo.

Thus, is this question answered in the current study? An attempt is made. The authors show using traditional semiquantitative RT-PCR analysis that VDR transcripts are detected, however, at concentrations 3–4 logs lower than that found in the intestine. Unfortunately, this technique does not measure the VDR protein itself and perhaps more importantly, is many logs more sensitive than those that do. Thus, it is not surprising that these investigators and many before them have struggled to identify the VDR using Western blot analysis or through immunocytochemistry (see Ref. 13 and table therein). This seems to be highlighted by the Western blot analyses in the current work, where it is unclear whether the immunoband detected in normal adult skeletal muscle is in fact the VDR, given the requirement for an unusual hyperosmolar (denaturing) extraction buffer (18), the presence of additional bands in a size range that is similar to the VDR, and the fact that virtually all of the bands that are detected are absent in the VDR-null control extract. With regard to the use of the hyperosmolar buffer, although the authors suggest that it may remove the VDR more efficiently from DNA, most the VDR extracted here is not likely to be bound to the genome, because normal levels of gene occupancy were shown years ago to be surprisingly low (∼15%) in vitamin D-sufficient animals (19). In addition, most nuclear receptors are known to be extracted efficiently in high salt buffers. In the case of the ligand-free VDR, concentrations of 0.15M NaCl are generally sufficient (20, 21), although higher levels (∼0.3M) are necessary when the VDR is bound to DNA via activation by 1,25(OH)₂D₃. In the current case, the addition of detergents, such as Triton X-100 and NP-40 (very similar nonionic detergents), only increase VDR sensitivity to extraction. One has to wonder whether it is technically feasible using Western blot analysis to detect VDR protein in crude extracts of muscle such as these when VDR transcripts are present at 1/4284th the concentration of those observed in the duodenum.

Interestingly, a recent series of well-controlled immunocytochemical studies has been reported wherein VDR was detected in a highly robust fashion in intestinal tissue and other VDR-positive tissues (22–24). These experiments, however, failed to detect the VDR in skeletal muscle, numerous cardiovascular sites, including the heart and liver, and in other cell types as well, suggesting its absence in these tissues. Interestingly, these reports have been widely used to call into question the presence of the VDR in muscle. Ironically, however, they simply reiterate the historical inability of investigators to detect the VDR in muscle in studies using sucrose gradient analysis, hormone-binding assays, scintillation autoradiography, the earliest Western blot analyses, and finally various forms of mRNA analysis, culminating in highly sensitive reverse transcriptase, real time PCR approaches that have spanned almost 4 decades. Perhaps the most important lesson to be learned from the studies of DeLuca and coworkers (22–24) is the absolute requirement for highly robust positive and negative controls when assessing VDR levels in test tissues such as the muscle. In the current study, this does not appear to be the case, because the fluorescence intensity of signals generated from VDR-positive tissue such as the intestine are not particularly remarkable and the cellular and subcellular sources of the signals are unclear. Even in the most recent studies (22–24), however, the question of whether low levels of VDR in muscle and other tissues would be detectable by immunological means should still be raised. Is it possible that mRNA levels and VDR proteins levels might be strikingly discordant, and thus the levels of receptor protein in muscle tissue are higher than would be predicted as a result of low transcript levels? The answer is of course yes; but if so, such a relationship needs to be demonstrated. What additional observations are made regarding the receptor in this study? Perhaps the most important is the finding that the concentrations of the VDR in neonatal mouse muscle is much higher than that found in adults. Although further exploration is necessary, it is possible that this interesting discovery could provide an important entre into defining the mechanism through which vitamin D exerts its actions on muscle in adults as well.

The VDR functions in a traditional sense to regulate the expression of genes by virtue of its ability to bind to sites on DNA, to recruit coregulatory complexes that modify chromatin structure and function, and to interact with other transcription factors in a DNA-independent manner (5). In the current study, the authors show that both the VDR and Cyp24a1 genes are modestly up-regulated by 1,25(OH)₂D₃, thereby establishing that this traditional function is apparently intact in passaged primary myotubes. Unfortunately, the regulation of these or other genes by 1,25(OH)₂D₃ is not followed up by in vivo studies. Rather, the authors explore the possibility that 1,25(OH)₂D₃ might stimulate the uptake of 25OHD₃ into isolated muscle myofibers, an interaction that has been previously described in muscle, which occurs between the vitamin D-binding protein-bound vitamin D metabolite and membrane-bound endocytotic receptors megalin and cubilin, and which may mediate storage of 25OHD₃ (25). Accordingly, the authors show that there is a modest involvement of the VDR in this process in that pretreatment of both myofibers as well as C2C12 cells with low levels of 1,25(OH)₂D₃ are indeed stimulatory for uptake. Unfortunately, the nature of this stimulatory effect is entirely obscure, because despite the suggestion that this phenomenon may represent a rapid nongenomic activity, the requirement for a 3-hour hormone pretreatment of the myofibers (and cells) is hardly the stuff of a nongenomic action. Because neither megalin nor cubilin are up-regulated, it would appear that this activity is not dependent upon transcription either. Clearly, further studies will be necessary to facilitate an understanding of this proposed role for the VDR in muscle tissue.

The present studies highlight the difficulties of using the presence of the VDR in cells and tissues as a centerpiece for defining the mechanism that underlies the biological actions of vitamin D in nontraditional vitamin D-responsive tissues. At this stage, hard experimental evidence will be necessary to define a regulatory mechanism in muscle tissue and to conclude that 1,25(OH)₂D₃ plays a direct rather than an indirect biological role in this tissue, perhaps through the regulation of extracellular phosphate levels. These effects must also be distinguished from the more subtle yet indirect mechanisms that have already been shown to occur. For example, 1,25(OH)₂D₃ affects bile acid metabolism in the liver not via a direct effect but rather by inducing intestinal expression of fibroblast growth factor 15, an endocrine fibroblast growth factor that is rapidly transported via the portal system to the liver, where its effects on bile acid metabolism are actually observed (26). They must also be distinguished from mechanisms related to distinct VDR up-regulation during a disease process. Accordingly, a recent study suggests a direct action of 1,25(OH)₂D₃ in the liver centered on the hormone's ability to counter TGFβ-induced liver fibrosis (27). However, TGFβ activates normally quiescent liver stellate cells. Thus, the possibility exists that the VDR is up-regulated during activation, and that the protein's presence is actually representative of a disease process rather than a physiological model of normal hepatic tissue response. Finally, studies must also be undertaken to quantitate the levels of the VDR in specific cell types as well. For example, although VDR-specific peptides have recently been demonstrated in whole rat brain using mass spectrometry, the starting material was derived from 20 separate mice (28). Clearly, additional details similar to those generated recently will be necessary to interpret this interesting result (29). Studies of this nature will be necessary to understand 1,25(OH)₂D₃ action in muscle tissues as well. Perhaps careful genetic deletion of the VDR gene product in adult myocytes and/or their precursors will be revealing, as has been accomplished in cardiac tissue (30). These latter studies have provided supportive evidence for the direct effects of vitamin D on the heart.

There is no doubt that vitamin D is beneficial for the maintenance of normal muscle form and function based upon both animal models as well as overwhelming positive clinical experience (31). Unfortunately, the biological effects observed in mice and men do not really speak to the mechanism through which the vitamin D hormone acts in this tissue. Thus, resolution will likely be achieved using creative basic approaches both in vitro and in vivo. Fortunately, the biochemical, molecular biological, genetic, and genomic techniques are now available such that this issue should be resolved in the next few years.

Response to J.W. Pike by C.M. Girgis, N. Mokbel, K.M. Cha, P.J. Houweling, M. Abboud, D.R. Fraser, R.S. Mason, R.J. Clifton-Bligh, and J.E. Gunton

We thank Professor Pike for his comments above in response to our work (13) and Professor Bouillon for his comments in the “News and Views” (32). We agree that clear detection of VDR in skeletal muscle has been controversial due to a number of technical factors, including protein extraction methods, variability in different muscle models, past problematic antibodies, and of course, the low level of VDR that is present in mature muscle at baseline. Our work shows clear absence of VDR in VDR knockout mice by immunohistochemistry and Western blotting, and perhaps more importantly, conclusive proof of functional presence comes with the demonstration of a novel physiological function, specifically the VDR-mediated uptake of 25-hydroxyvitamin D in muscle fibers (ie, nongenomic as indicated by inhibition using the chloride-channel blocker 4,4′-diisothiocyanatostilbene-2,2′-disulfonic acid). We agree that the substantially higher levels of VDR in muscle of younger mice and immature muscle cells is intriguing; this suggests the possibility for a pleiotropic role for VDR in muscle and its potential activation after muscle injury from relatively low baseline levels of expression. We hope that the findings of our study bring some closure to this controversial field and may assist in future work examining roles of VDR in muscle development, regeneration, and 25-OHD uptake, ultimately justifying the generation of a skeletal muscle-specific VDR knockout model.