Classical and non‐classical metabolic transformation of vitamin D in dermal fibroblasts

Until recently, it was widely believed that vitamin D3 (D3) was solely activated through the sequential hydroxylations by CYP27A1 or CYP2R1 at C25, and by CYP27B1 at C1 1, 2: [D3→25(OH)D3→1,25(OH)2D3]. This dogma has recently been broken by the discovery that CYP11A1 hydroxylates D3 producing 20-hydroxyvitamin D3 (20(OH)D3, which can undergo further sequential hydroxylations of the side chain without its cleavage (3-6s2): [D3→20(OH)D3→(OH)nD3]. CYP11A1 can also hydroxylate D3 at C22 to produce 22(OH)D3. Further hydroxylation of 22(OH)D3 or 20(OH)D3 by CYP11A1 produces 20,22(OH)2D3 and 20,23(OH)2D3 6: D3 → 20(OH)D3 + 22(OH)D3 → 20,23(OH)2D3 + 20,22(OH)2D3. The main products of this pathway are biologically active and can be further hydroxylated by CYP27B1 [reviewed in 7]. Furthermore, ex-vivo and ex-utero incubations of D3 with fragments of adrenal glands and human placenta, respectively, demonstrate that these CYP11A1-catalysed pathways of D3 metabolism occur in steroidogenic tissues with 20(OH)D3 being the major metabolite 8. CYP11A1-dependent metabolism of D3 has also been detected in epithelial cells such as epidermal keratinocytes 8 and colonic Caco-2 cells 7 which have a relatively low expression of this enzyme, and 20(OH)D3 has been detected in the human epidermis 9. However, it is unknown whether similar metabolism can occur in cells of mesenchymal origin such as dermal fibroblasts. As human dermal fibroblasts do express CYP11A1 10, we investigated whether they can transform vitamin D3 into novel CYP11A1-dependent hydroxyderivatives with possible 1α-hydroxylation of the major product, 20(OH)D3. The detailed description is provided in the Supplemental file. Briefly, human dermal fibroblasts from the foreskin of African American donors, passage #2, at a concentration of 3 × 106 cells/ml, were incubated in the presence of 0, 50 or 500 μm D3 at 37°C for 18 h and the suspension then extracted and processed as described previously 8. The resulting extracts were directly analysed by liquid chromatography–mass spectrometry (LC-MS) using a Xevo™ G2-S qTOF tandem mass spectrometer and ESI source with separations carried on a Waters Atlantis dC18 column (100 × 4.6 mm, 5 μm particle size) using a methanol in water gradient (Fig. 1). To further confirm the identity of products, extracts were initially separated on long a C18 columns (250 × 4.6 mm, 5 μm particle size) and purified fractions with retention times corresponding to those of appropriate mono- or di-hydroxyvitamin D3 standards were further analysed on a Waters Acquity™ Ultra Performance Liquid Chromatography (UPLC) system (Milford, MA, USA) equipped with a Waters Xevo™ G2-S qTOF MS and an ESI source (see Supporting information). Figure 1 shows dose-dependent transformation of D3 to mono- and di-hydroxyvitamin D3 metabolites for extracts directly analysed by LC-MS and monitored from the extracted mass at m/z = 383.3 [M+H-H2O]+ for mono-hydroxyvitamin D3 and m/z = 399.3 [M+H-H2O]+ for di-hydroxyvitamin D3. Species with RT corresponding to 25(OH)D3, 20(OH)D3, 1,25(OH)2D3, as well as for both 20,22(OH)2D3 and 20,23(OH)2D3 (20,22/23(OH)2D3) which did not separate under the HPLC conditions used, were detectable at 50 μm substrate and were dramatically higher with 500 μm D3 (Fig. 1a–f). The signal for 22(OH)D3 was buried in the descending shoulder of the peak at RT = 11.6 min (Fig. 1b,c), while that for 1,20(OH)2D3 was in the peak preceding 1,25(OH)2D3 (Fig. 1e,f). Therefore, to confirm that these additional expected products were present, and to further confirm the identity of the other major products, species with RT corresponding to 25(OH)D3, 20(OH)D3 22(OH)D3, 1,25(OH)2D3, 1,20(OH)2D3, 20,22(OH)2D3 and 20,23(OH)2D3 were separated and collected using a 250 mm C18 columns with an acetonitrile in water gradient. The individual fractions were then analysed by UPLC-qTOF MS using extracted masses as listed in Fig. 2. These data clearly show the production of the expected mono- and di-hydroxyvitamin D3 metabolites with detectable masses at [M+H]+ and [M+H-H2O]+ for the corresponding D3 derivatives (inserts in Fig. 2). In this study, we show for the first time that human dermal fibroblasts can transform D3, not only to 25(OH)D3 and 1,25(OH)2D3 (classical metabolites), but also to the newly discovered CYP11A1-dependent hydroxyderivatives of vitamin D3 including 20(OH)D3, 22(OH)D3, 20,22(OH)2D3, 20,23(OH)2D3 and 1,20(OH)2D3. These novel metabolic transformations are consistent with our previous detection of CYP11A1 enzyme in human dermal fibroblasts 10. Of note, there are similar levels of production of 1,25(OH)2D3 vs 20,22/23(OH)2D3 but higher production of 20(OH)D3 vs. 25(OH)D3 (Fig 1) which is consistent with greater metabolism of D3 by CYP11A1 than by the classical pathway [D3→25(OH)D3→1,25(OH)2D3] in tissues or cells expressing CYP11A1 6, 8. These findings allow to propose a novel scheme of vitamin D3 hydroxylations in dermal fibroblasts that includes both generation of 25(OH)D3 by 25-hydroxylase with subsequent 1α-hydroxylation, and generation of 20(OH)D3 or 22(OH)D3 by CYP11A1 followed by hydroxylation to 20,22(OH)2D3 or 20,23(OH)2D3. 20(OH)D3 can also be 1α-hydroxylated by CYP27B1 (Scheme S1). These findings are potentially of great biological significance, because novel CYP11A1-generated secosteroids can regulate the functions of epidermal keratinocytes and dermal fibroblasts (7, s5, s6, s11). Their generation in dermal fibroblast suggests an intra- or paracrine mechanism of action in the dermal compartment with possible effects on adnexal structures or immune cells. This extended panel of classical and novel vitamin D3 hydroxyderivatives we have documented may offer an explanation for the pleiotropic effects of vitamin D in the skin so far assigned to only one molecule, 1,25(OH)2D3. In summary, we report for the first time that human dermal fibroblasts, in addition, to producing classical 25(OH)D3 and 1,25(OH)2D3 can metabolize D3 to novel CYP11A1-derived mono- and di-hydroxymetabolites, of which the major product, 20(OH)D3, can be further 1α-hydroxylated to 1,20(OH)2D3. TK performed the experiments, analysed the data and contributed to writing of the manuscript; AS designed the study, analysed the data and wrote the manuscript; WL and RCT provided materials and contributed to the analysis of the data and writing of the manuscript. This work was supported by the grants from the National Institutes of Health (2R01AR052190-A6, R21AR066505-01A1 and 1R01AR056666-01A2) to AS. The authors have no conflict of interest to declare. Figure S1. Proposed metabolism of vitamin D3 in human dermal fibroblasts. Data S1. Materials and Methods. Data S2. Supplemental References. Please note: The publisher is not responsible for the content or functionality of any supporting information supplied by the authors. Any queries (other than missing content) should be directed to the corresponding author for the article.