Oral Intake of Lumisterol Aﬀects the Metabolism of Vitamin D

Scope: The treatment of food with ultraviolet-B (UV-B) light to increase the vitamin D content is accompanied by the formation of photoisomers, such as lumisterol 2 . The physiological impact of photoisomers is largely unknown. Methods and Results: Three groups of C57Bl/6 mice are fed diets containing 50 µg kg -1 deuterated vitamin D 3 with 0, 50 (moderate-dose) or 2000 µg kg -1 (high-dose) lumisterol 2 for four weeks. Considerable quantities of lumisterol 2 and vitamin D 2 are found in the plasma and tissues of mice fed with 2000 µg kg -1 lumisterol 2 but not in those fed 0 or 50 µg kg -1 lumisterol 2 . Mice fed with 2000 µg kg -1 lumisterol 2 showed strongly reduced deuterated 25-hydroxyvitamin D 3 (–50%) and calcitriol (–80%) levels in plasma, accompanied by downregulated mRNA abundance of c ytochrom P450 (Cyp ) 27b1 and upregulated Cyp24a1 in the kidneys. Increased tissue levels of vitamin D 2 were also seen in mice in a second study that are kept on a diet with 0.2% UV-B exposed yeast versus those fed 0.2% untreated yeast containing iso-amounts of vitamin D 2 . Conclusion: High doses of lumisterol 2 can enter the body, induce the formation of vitamin D 2 , reduce the levels of 25(OH)D 3 and calcitriol and strongly impact the expression of genes involved in the degradation and synthesis of bioactive vitamin D. and G.I.S. designed the mouse experiments. J.K. per- formed the mouse experiments. J.K. analyzed the data. J.K. conducted the statistical analysis. J.K. and G.I.S. wrote the manuscript. J.K. and A.C.B. critically reviewed the manuscript. Data available on request from the authors.


Introduction
Many people worldwide are affected by vitamin D deficiency or insufficiency. [1] Because natural food sources, except oily fish, are usually low in vitamin D, the fortification of food with vitamin D has been considered a viable option to improve vitamin D intake in humans. [2] One strategy to enrich food with D 3 incorporation into micelles and in turn reduce the apical uptake of vitamin D 3 in Caco-2 cells. [17] It is also conceivable that vitamin D and photoisomers compete for enzymes that catalyze the synthesis of bioactive and inactive vitamin D metabolites. Structural vitamin D analogs are known to be selective inhibitors of hydroxylases in vivo. [22,23] There is evidence from in vivo and in vitro studies that lumisterol 3 can function as a substrate for enzymes involved in vitamin D metabolism. [13,24] Slominski et al. showed that lumisterol 3 , which is hydroxylated by the cytochrome P450 (CYP) 11A1 enzyme, can also serve as a substrate for CYP27B1, CYP27A1 and CYP24A1, enzymes that usually mediate the synthesis of active and inactive vitamin D metabolites. [25,26] While renal CYP27B1 catalyzes the formation of bioactive 1,25-dihydroxyvitamin D (1,25(OH) 2 D), [27] CYP27A1 and CYP2R1 are important liver enzymes that are involved in the hydroxylation of vitamin D to 25-hydroxyvitamin D (25(OH)D), the primary biomarker of vitamin D status. [28,29] In contrast, CYP24A1 is responsible for the degradation of hydroxylated vitamin D metabolites. [30] To our knowledge, no study has investigated the intestinal uptake and impact of orally administered vitamin D photoisomers on vitamin D metabolism in a mouse model. Therefore, we conducted two studies: the first study addressed the question of whether oral lumisterol 2 , the main photoisomer formed in UV-B exposed baker's yeast, can enter the body and affect vitamin D metabolism; the second study was conducted to investigate the impact of UV-B exposed baker's yeast containing a mixture of lumisterol 2 and tachysterol 2 on vitamin D status in mice.

Animals and Diets
Mice in both studies were cared for and handled according to the guidelines established by the US National Research Council. [31] The experimental procedures were approved by the committee for animal welfare of the Martin Luther University Halle-Wittenberg (approval number of study 1: H1-4/T4-18; approval number of study 2: H1-4/T2-18). All mice were housed in Macrolon cages in pairs with a 12-h light/12-h dark cycle in a UV-free room at a controlled temperature (22 ± 2°C) and relative humidity (50-60%).
In both studies, mice received a semi-synthetic basal diet ( Table 1), supplemented with adequate vitamins and minerals according to the recommendations of the US National Research Council. [32] The diets were kept refrigerated until they were fed to the mice. Food and water were provided ad libitum in both studies.
Study 1 aimed to investigate the uptake of oral lumisterol 2 and its effect on orally administered vitamin D. For this purpose, 36 4-week-old male C57Bl/6NCrl mice (Charles River, Sulzfeld, Germany) with an initial body weight of 21.7 ± 1.05 g were randomly allocated to three groups (n = 12) and were fed basal diets supplemented with 50 µg kg -1 triple-deuterated vitamin D 3 (vitamin D 3 -d 3 , Cambridge Isotope Laboratories, Andover, USA) with either 0 (control), 50 (moderate), or 2000 µg kg −1 (high) lumisterol 2 (Toronto Research Chemicals Inc., North York, ON, Canada) for four weeks. The analyzed concentrations of lumisterol 2 in the two lumisterol 2 diets were 56 and 1871 µg kg −1 . The chosen lumisterol 2 concentrations represent moderate levels (5.45 ± 0.24 µg kg −1 body weight) and high levels of lumisterol 2 (223 ± 6.12 µg kg −1 body weight), respectively. As lumisterol 2 can probably be reconverted to vitamin D 2 by a thermal reaction, the study additionally analyzed vitamin D 2 in the diet and found 15 µg kg −1 vitamin D 2 in the high-lumisterol 2 diet and no detectable vitamin D 2 (limit of quantification 0.84 µg kg −1 ) in the moderate-lumisterol 2 and control diets. The analyzed concentrations of lumisterol 2 and vitamin D 2 in the diets were examined in triplicate. Study 2 investigated the impact of UV-B-exposed versus unexposed baker's yeast containing the same amounts of vitamin D 2 on the plasma and tissue levels of D-vitamers in mice. The study was conducted with 24 4-week-old male C57BL/6NCrl mice (Charles River, Sulzfeld, Germany) with an initial body weight of 19.8 ± 1.47 g. The mice were randomly allotted into two groups (n = 12) and were fed the basal diets with either 2 g kg −1 unexposed baker's yeast (Yeast-UVB) or 2 g kg −1 UV-B irradiated baker's yeast (Yeast+UVB) for 5 weeks. The latter was produced by exposure of commercially available baker's yeast to UV-B light with an intensity of 1150 µW cm -² (UV-8M, Herolab GmbH, Wiesloch, Germany) for 30 min. The analyzed concentrations of vitamin D 2 , lumisterol 2 , and tachysterol 2 in 2 g of UV-B exposed baker's yeast used for one kg diet were 36, 38, and 1.2 µg, respectively. The control diet, which contained the unexposed baker's yeast, was supplemented with 36 µg kg −1 vitamin D 2 to ensure iso-amounts of vitamin D 2 in both diets.
Individual body weights of mice from both studies were recorded weekly. The food intake per cage was recorded daily. After the experimental periods, the mice were deprived of food for four h, anesthetized with diethyl ether and exsanguinated by decapitation. Blood was collected in heparinized or serum tubes (Sarstedt, Nümbrecht, Germany), centrifuged to isolate plasma or serum and stored at −20°C until analysis. Tissue samples were harvested, immediately snap frozen in liquid nitrogen and stored at −80°C until further analysis.
Further, the study examined the temperature-dependent conversion of lumisterol 2 to vitamin D 2 . Therefore, standardized quantities of lumisterol 2 , dissolved in ethanol, were thermally treated at 20°C and 60°C, respectively, for two, four, and eight h. Additionally, one aliquot of the lumisterol 2 solution was preexposed to UV-B light (intensity of 1150 µW cm − ²; UV-8M, Herolab GmbH, Wiesloch, Germany) for 30 min and then analyzed for subsequent changes in the lumisterol 2 and vitamin D 2 concentrations. Finally, the diet containing the UV-B exposed yeast was thermally treated at 37°C (to simulate the body temperature) for 24, 48, and 72 h, respectively, to determine vitamin D 2 formation.

Analysis of Circulating 1,25(OH) 2 D, Parathyroid Hormone and Cholesterol in Mice
The plasma concentration of 1,25(OH) 2 D was analyzed by a commercial enzyme-linked immunoassay (Immunodiagnostic Systems, Frankfurt am Main, Germany), and parathyroid hormone (PTH) was measured by a two-site enzyme-linked immunosorbent assay (Immunotopics, San Clemente, USA). A commercial photometric assay was used to quantify plasma concentrations of cholesterol (Cholesterol FS, Diagnostic Systems GmbH, Holzheim, Germany). All analyses were performed by following the procedures given by the manufacturers with modifications. [36]

Analysis of the Relative mRNA Abundance of Genes Involved in the Uptake and Metabolism of Vitamin D
The relative mRNA abundance of vitamin D target genes and genes involved in vitamin D and sterol metabolism in the  tissues of mice was analyzed by real-time RT-PCR. Total RNA was isolated from the liver, kidney and intestinal mucosa with the peqGOLD TriFast Kit (PEQLAB Biotechnologie GmbH, Erlangen, Germany) according to the manufacturer´s protocol. The RNA concentration was calculated using a NanoDrop Spectrophotometer (Thermo Fisher Scientific GmbH, Schwerte, Germany), and RNA purity was confirmed by agarose gel electrophoresis. cDNA was synthesized using M-MLV Reverse Transcriptase (Promega, Madison, WI, USA). A total of 1 µL of cDNA template was amplified as described elsewhere [37] using GoTaq Flexi DNA-Polymerase (Promega) and the Rotorgene 6000 system (Corbett Research, Mortlake, Australia). According to the RT-PCR protocol, initial denaturation was performed at 95°C for 3 min, followed by 20-30 cycles of denaturation at 95°C, annealing at primer-specific temperatures ( Table 3) and elongation at 72°C. The amplification of the single and specific mRNAs was confirmed by agarose gel electrophoresis. The calculation of relative mRNA concentration was based on the method of Pfaffl. [38] Beta-2-microglobulin (B2m), hypoxanthine guanine phosphoribosyl transferase (Hprt) and the ribosomal protein, large, P0 (Rplp0) were used as appropriate reference genes. The relative mRNA abundance of target genes was expressed as the fold change in relation to the control group. The target and reference genes are shown in Table 3.

Statistical Analysis
Data are expressed as the mean ± standard derivation (SD). Statistical analyses were performed using SPSS version 25.0 (IBM, Armonk, USA). Data from study 1 including three groups of mice were subjected to the Shapiro-Wilk normality and homoscedasticity of variance test (Levene's test). For normally distributed parameters, the groups were compared by one-way analysis of variance (ANOVA). When ANOVA revealed significant differences, data with equally distributed variances were compared with the Tukey-HSD post hoc test. The Games-Howell test was used for data with unequally distributed variances. In the case of non-normally distributed parameters, data were subjected to the nonparametric Kruskal-Wallis test. If the Kruskal-Wallis test revealed significant differences, groups were compared by the Mann-Whitney U test for post hoc comparisons. A Bonferroni correction for multiple testing was used. If the metabolite concentration in a sample within a group was below the LOQ, a randomly generated value (between 0 and the appropriate LOQ) was used for statistical treatment analysis. In cases in which all mice within a group had metabolite levels below the LOQ, the data were not included in the statistical analyses. If two of the three groups had metabolite levels above the LOQ, the means of the two groups were analyzed by Student´s t-test or the nonparametric Mann-Whitney U test. The data were considered significantly different at p < 0.05. The results of study 2 were subjected to a normality test using the Shapiro-Wilk test. If the data followed a normal distribution, differences between the groups were analyzed by Student´s t-test. If not, the nonparametric Mann-Whitney U test was used. Differences were considered to be significant at p < 0.05. The data are presented as the mean ± standard deviation (n = 12). Quantifiable data from two groups were compared with Student's t-test. * Indicates statistically significant differences (p < 0.05). The limit of quantification (LOQ) of lumisterol 2 was 1.2 nmol L −1 for plasma and 7.3 ng g −1 for tissue samples.

Oral Lumisterol 2 Can Enter the Body
To determine whether oral lumisterol 2 can enter the body, we analyzed the levels of lumisterol 2 in the plasma, intestinal mucosa, liver, retroperitoneal adipose tissue and kidney of mice. Detectable concentrations of lumisterol 2 were observed in plasma and all tissues, except kidney, of mice that received 2000 µg kg −1 lumisterol 2 (Figure 1A-D). In mice fed 50 µg kg −1 lumisterol 2 , detectable levels of lumisterol 2 were found only in the retroperitoneal adipose tissue ( Figure 1D), whereas the lumisterol 2 con-centrations in the plasma, intestinal mucosa, liver and kidney were below the LOQ ( Figure 1A-C). Mice receiving the control diet had no detectable lumisterol 2 in plasma or any tissues analyzed.

Orally Administered Lumisterol 2 was Partly Converted to Vitamin D 2 in Mice
As photoisomers can be reconverted to vitamin D 2 via thermoconversion, we analyzed the concentrations of vitamin D 2 and hydroxylated D 2 -vitamers in the plasma and tissues of mice fed lumisterol 2 . Considerable quantities of vitamin D 2 in plasma and all tissues analyzed were observed in mice fed 2000 µg kg −1 lumisterol 2 (Figure 2A-E). In mice fed 50 µg kg −1 lumisterol 2 , low levels of vitamin D 2 were found in the retroperitoneal adipose tissue, whereas in the plasma, intestinal mucosa, liver and kidney, the vitamin D 2 concentrations were below the LOQ. Mice receiving the control diet had no detectable vitamin D 2 in plasma or any tissues analyzed.
To determine whether vitamin D 2 formed in the highlumisterol 2  The data are presented as the mean ± standard deviation (n = 12). Quantifiable data from two groups were compared with Student's t-test. * Indicates statistically significant differences (p < 0.05). The limit of quantification (LOQ) of vitamin D 2 was 0.5 nmol L −1 for plasma and 1.0 ng g −1 for tissue samples. lumisterol 2 but not in the groups fed 0 or 50 µg kg −1 lumisterol 2 (Figure 3A,B).

Oral Lumisterol 2 Slightly Increases Vitamin D 3 -d 3 levels in Tissues
To investigate the impact of oral lumisterol 2 on the uptake and tissue levels of labeled vitamin D 3 , we determined the levels of vitamin D 3 -d 3 in the plasma and tissues of mice. The concentrations of vitamin D 3 -d 3 in the plasma of both groups of mice that received lumisterol 2 were comparable to those in the controls, although the concentration of vitamin D 3 -d 3 was higher in the plasma of mice fed diets with 2000 µg kg −1 lumisterol 2 than in mice fed 50 µg kg −1 lumisterol 2 ( Figure 4A). Interestingly, we found that the vitamin D 3 -d 3 levels were higher in the intestinal mucosa of mice fed diets with 50 and 2000 µg kg −1 lumisterol 2  than in the controls ( Figure 4B). In the liver and retroperitoneal adipose tissue, increased levels of vitamin D 3 -d 3 were only found in the group fed 2000 µg kg −1 lumisterol 2 , whereas no differences were observed between the group fed 50 µg kg −1 lumisterol 2 and the controls ( Figure 4C,E). The levels of vitamin D 3 -d 3 in the kidney remained unaffected by the treatments ( Figure 4D).
To examine the impact of dietary lumisterol 2 on vitamin D transporters, we analyzed ATP binding cassette transporters G5 and G8 (Abcg5/g8) and the Cd36, Npc1l1 and Scarb1 mRNA abundance in the intestinal mucosa. The data indicate that the Npc1l1 mRNA abundance was increased in the group fed 2000 µg kg −1 lumisterol 2 compared to the other two groups ( Table 4). Differences in the mRNA abundance of Abcg5 were only seen between mice fed 2000 and 50 µg kg −1 lumisterol 2 but not between both lumisterol 2 groups and the control (Table 4). No differences in the mRNA abundance of Abcg8, Cd36 and Scarb1 were found among the three groups of mice (Table 4).

Oral Lumisterol 2 Affects the Formation of 25(OH)D 3 -d 3
The data demonstrate that the concentration of 25(OH)D 3 -d 3 was strongly reduced in the plasma, liver and kidney in mice fed 2000 µg kg −1 lumisterol 2 compared to mice fed 0 and 50 µg kg −1 lumisterol 2 (Figure 5A-C). The concentration of 25(OH)D 3d 3 did not differ between the groups fed 0 and 50 µg kg −1 lumisterol 2 .
To provide possible causes for the alterations in the levels of 25(OH)D 3 -d 3 , we analyzed the mRNA abundance of the most important hepatic genes that are involved in the hydroxylation of vitamin D but found no differences in the mRNA abundance of Cyp27a1, Cyp2r1, Cyp2j3, and Cyp3a11 in the livers of mice from the three groups (Table 4).

Oral Lumisterol 2 Affects the Synthesis of the Bioactive Vitamin D Hormone
Next, we quantified circulating levels of 1,25(OH) 2 D in plasma and found that the group fed 2000 µg kg −1 lumisterol 2 had considerably lower concentrations of 1,25(OH) 2 D than the other two groups (Figure 6A). No differences in 1,25(OH) 2 D levels were observed between the groups fed 0 and 50 µg kg −1 lumisterol 2 . To determine whether the reduced level of 1,25(OH) 2 D in mice fed 2000 µg kg −1 lumisterol 2 represents a vitamin D-deficient state, we analyzed the plasma concentrations of PTH in the mice but did not find any differences among the three groups of mice (Figure 6B).
To ascertain whether the reduced plasma level of bioactive vitamin D hormone was associated with impaired vitamin D action, we analyzed the mRNA abundance of the vitamin D receptor (Vdr) and classic vitamin D receptor target genes, including calbindin D 9k (S100g), claudin-2 (Cldn2), and transient receptor potential cation channel, subfamily V, member 6 (Trpv6), in the intestinal mucosa but did not find differences among the three groups of mice (Table 4).
Gene expression data demonstrated a marked reduction in Cyp27b1 mRNA abundance and a strong increase in Cyp24a1 mRNA abundance in the group fed 2000 µg kg −1 lumisterol 2 compared to the other two groups ( Figure 6C,D).

Oral Intake of UV-B-Exposed Yeast did not Influence the Body Weight or Food Intake of Mice
The final body weights (Yeast-UVB, 24.9 ± 1.51 g; Yeast+UVB, 25.2 ± 2.27 g) and mean daily food intake assessed for two mice per cage (Yeast-UVB, 2.64 ± 0.31 g; Yeast+UVB, 2.69 ± 0.43 g) did not differ between the two groups of mice.

UV-B-Exposed Yeast Increased the Levels of Vitamin D 2 in the Intestinal Mucosa and Liver
This study compared the impact of diets containing either UV-B-exposed yeast or unexposed yeast containing iso-amounts of vitamin D 2 on D-vitamer levels in mice. First, we measured the concentrations of the photoisomers lumisterol 2 and tachysterol 2 in the serum and tissues of mice but found no detectable quantities of either photoisomer in the serum, intestinal mucosa, liver or kidney in either group. The serum concentrations of vitamin D 2 , 25(OH)D 2 and 1,25(OH) 2 D did not differ between the two groups ( Figure 7A-C). However, we found approximately 1.3times higher vitamin D 2 concentrations in the intestinal mucosa  and liver of mice fed UV-B exposed yeast than in those of mice that received unexposed yeast ( Figure 7D,E). The concentration of vitamin D 2 in the kidney did not differ between the two groups ( Figure 7F).

Temperature-and UV-B Light-Induced Conversion of Lumisterol 2 to Vitamin D 2
To investigate the temperature-dependent conversion of lumisterol 2 to vitamin D 2 , ethanolic solutions of lumisterol 2 were subjected to 20°C and 60°C for eight h, respectively. Additionally, to determine whether the pretreatment of lumisterol 2 with UV-B light can also stimulate the subsequent vitamin D 2 formation, lumisterol 2 was exposed to UV-B light for 30 min. All treatments resulted in a reduction of lumisterol 2 ( Figure 8A) and an increase of vitamin D 2 ( Figure 8B). The strongest increase of vitamin D 2 was found when lumisterol 2 was pretreated with UV-B light, followed by the 60°C and 20°C treatments ( Figure 8B). Interestingly, the formation of vitamin D 2 did not reach a plateau within 8 h after the lumisterol 2 pretreatment with UV-B light.
To determine whether the higher concentrations of vitamin D 2 in the gut and liver of mice fed UV-B exposed yeast were caused by subsequent conversion of photoisomers to vitamin D 2 in the bodies of the mice, we analyzed the formation of vitamin D 2 at body temperature (37°C). Here, we found that the treatment of the UV-B-exposed yeast diet at 37°C for 24, 48, and 72 h increased the concentrations of vitamin D 2 by 8%, 43% and 63%, respectively.

Discussion
Exposure of food to UV-B light is a novel approach to enrich food with vitamin D. However, this strategy is accompanied by the formation of photoisomers, such as lumisterol, the most abundant photoisomer in UV-B treated baker's yeast and mushrooms. [6,7,11] To elucidate the impact of this photoisomer on the fate of orally administered vitamin D and vitamin D metabolism, we first conducted a study with mice that received 50 µg kg −1 deuteriumlabeled vitamin D 3 with 0, 50 or 2000 µg kg −1 lumisterol 2 . Interestingly, we found that high doses of lumisterol 2 can enter the body and exert pronounced effects on vitamin D metabolism. , liver (E) and kidney (F) of mice receiving diets containing UV-B exposed yeast (Yeast+UVB) or unexposed yeast containing iso-amounts of vitamin D 2 (Yeast-UVB) for five weeks. The data are presented as the mean ± standard deviation (n = 12) and were compared by Student's t-test. **Indicates statistically significant differences (p < 0.01). Most noticeably, there was a strong increase in vitamin D 2 content in the plasma and tissues of mice fed the high-lumisterol 2 diet. Vitamin D 2 can be formed from lumisterol 2 by a thermal reaction. [39,40] Analysis of the diet enriched with 2000 µg kg −1 lumisterol 2 revealed vitamin D 2 concentrations of 15 µg kg −1 , indicating that small amounts of lumisterol 2 were converted to vitamin D 2 during the preparation of the diet. This may explain the increase in vitamin D 2 levels in mice fed the high-lumisterol 2 diet. While the vitamin D 2 concentrations in the intestinal mucosa, liver and kidney were lower than the corresponding levels of vitamin D 3 -d 3 in these tissues, the vitamin D 2 concentrations in the adipose tissue and plasma were markedly higher than those of vitamin D 3 . More importantly, mice fed the diet with 2000 µg kg −1 lumisterol 2 had higher plasma concentrations of 25(OH)D 2 than of 25(OH)D 3 -d 3 . We therefore hypothesize that a subsequent temperature-dependent isomerization of lumisterol 2 to vitamin D 2 took place in the bodies of these mice, which caused an increase in vitamin D 2 levels in plasma and tissues. The detection of vitamin D 2 in the adipose tissue of the group fed the low-lumisterol 2 diet corroborates this hypothesis because no detectable vitamin D 2 was found in the diet containing 50 µg kg −1 lumisterol 2 . In addition, the conversion studies with lumisterol 2 indicate a distinct temperature-dependent formation of vitamin D 2 from lumisterol 2 .
To determine whether photoisomers formed in UV-B-exposed baker's yeast may also contribute to a rise in vitamin D 2 levels in mice, we compared in study 2 the serum and tissue levels of D 2 in mice that were fed identical vitamin D 2 amounts via UV-B exposed yeast and vitamin D 2 -enriched unexposed yeast. The increased concentrations of vitamin D 2 in the intestinal mucosa and liver observed in the mice fed the UV-B exposed yeast are indicative of such stimulated vitamin D 2 formation in the bodies of these mice. The experimental data which showed a pronounced conversion of lumisterol 2 to vitamin D 2 , particularly in the case of prior UV-B irradiation, support the assumption that lumisterol 2 can be converted to vitamin D 2 several hours after the UV-B exposure. Additionally, within a day, we observed an increase in dietary vitamin D 2 content when the environmental temperature was switched from 22°C to 37°C.
In addition to vitamin D 2 , mice from study 1 that were fed the high-lumisterol 2 diet had higher levels of vitamin D 3 -d 3 in tissues but lower levels of 25(OH)D 3 -d 3 and the bioactive vitamin D hormone 1,25(OH) 2 D than mice fed no lumisterol 2 . The observed increase in tissue levels of vitamin D 3 -d 3 was surprising because we hypothesized that lumisterol 2 and vitamin D 3 may compete for intestinal uptake due to their similar chemical structures. It is important to note that recent data from our group also showed an increase in tissue levels of vitamin D 3 when feeding mice 7-DHC [35] or ergosterol, [36] which have structures closely related to that of vitamin D. To determine whether intestinal uptake may explain the increased tissue levels of vitamin D 3 -d 3 in mice fed the high dose of lumisterol 2 , we analyzed the transcription levels of potential vitamin D transporters. These included the cholesterol transporters NPC1L1 [19,20] ; CD36, which facilitates fatty acid uptake [41,42] and chylomicron formation [43] ; SRB1 [20] ; and the reverse sterol transporters ABCG5/G8, which mediate the export of xenosterols, such as phytosterols. [44] Recently, we demonstrated a strong decrease in the tissue levels of vitamin D 3 -d 3 after treatment of mice with the NPC1L1 inhibitor ezetimibe, suggesting that NPC1L1 is the most important vitamin D transporter in the gut. [19] As we found a higher mRNA abundance of NPC1L1 in the intestinal mucosa of mice treated with the high lumisterol 2 dose, we hypothesize that lumisterol 2 or the converted vitamin D 2 improved the intestinal availability of vitamin D by influencing NPC1L1 at the transcriptional level.
Alternatively, the increased tissue levels of vitamin D 3 -d 3 in the high-lumisterol 2 group could also result from impaired conversion of vitamin D 3 -d 3 to 25(OH)D 3 -d 3 by lumisterol 2 or vitamin D 2 , the latter being found to induce a disproportionately strong reduction in 25(OH)D 3 in mice. [45][46][47][48] The liver is the main site for the 25-hydroxylation of vitamin D because it includes a variety of enzymes, including CYP27A1, CYP2R1, CYP3A11 and CYP2J3, which are involved in 25(OH)D synthesis. [46] However, the current data do not indicate any changes in the transcription levels of these enzymes following lumisterol 2 treatment. Thus, we hypothesize that liver hydroxylation may not explain the higher vitamin D 3 -d 3 and the lower 25(OH)D 3 -d 3 levels observed in these mice. The increased mRNA expression of Cyp24a1, which we observed in the mice treated with 2000 µg kg −1 lumisterol 2 , was likely responsible for the reduction in 25(OH)D 3 -d 3. Consistent with this finding is the concurrent reduction in 1,25(OH) 2 D because CYP24A1 catalyzes not only the degradation of 25(OH)D but also that of 1,25(OH) 2 D. [30] The observed impact of oral lumisterol 2 on CYP24A1 probably explains the results from human and rat studies that failed to show an increase in 25(OH)D plasma levels after the consumption of bread containing UV-B exposed vitamin D-rich yeast compared to vitamin D supplements. [47,48] The current finding that the circulating level of 25(OH)D 2 in mice fed UV-B exposed yeast did not differ from that of unexposed yeast, although mice fed UV-B exposed yeast had higher levels of vitamin D 2 in the intestinal mucosa and liver, fits well with this hypothesis.
Since we also found a reduced expression of Cyp27b1, the key enzyme in the synthesis of 1,25(OH) 2 D from 25(OH)D, [27] we suggest that the low plasma levels of the bioactive vitamin D hormone 1,25(OH) 2 D in mice fed the high lumisterol 2 dose resulted from increased degradation and reduced synthesis. The bioactive vitamin D hormone 1,25(OH) 2 D normally binds to the vitamin D receptor, which then forms a dimer with the retinoic acid receptor and stimulates the transcription of proteins, for example, those that are involved in the uptake of calcium. [49] A lack of 1,25(OH) 2 D usually leads to the synthesis and secretion of PTH. [50,51] Despite having low plasma levels of 1,25(OH) 2 D, mice fed the high lumisterol 2 dose did not show increased PTH levels. Three possible explanations can be considered for this finding. First, it is possible that the reduction in 25(OH)D and 1,25(OH) 2 D was not severe enough to induce secondary parathyroidism. This hypothesis is supported by recent data from Mallya and coworkers, who found no changes in PTH levels despite inducing low levels of 25(OH)D and 1,25(OH) 2 D in mice by feeding with vitamin D-deficient diets. [52] Second, it is likely that the hydroxylated lumisterol 2 metabolites exerted an active vitamin D effect. This assumption is corroborated by the finding that hydroxylated lumisterol 3 metabolites that are formed in keratinocytes have been shown to be able to stimulate the vitamin D receptor. [12] Third, the additionally formed D 2 -vitamers likely induced moderate hypervitaminosis D in the mice, which in turn activated the degradation of hydroxylated D-vitamers via CYP24A1 or reduced renal formation of 1,25(OH) 2 D via CYP27B1.
To ascertain whether the marked decline in circulating 1,25(OH) 2 D in mice fed the high lumisterol 2 dose was accompanied by impaired vitamin D action, we analyzed the mRNA abundance of four classic vitamin D receptor target genes but found no differences among the groups. Thus, the data do not indicate that mice treated with the high lumisterol 2 dose had developed vitamin D deficiency.
To estimate the relevance of these studies for humans, the following model calculation was conducted. The presently used doses of lumisterol 2 (50 and 2000 µg kg −1 diet) equate to 5.45 and 223 µg lumisterol 2 per kg body weight of mice. Using the dose translation formula of Reagan-Shaw et al. [53] which is based on the body surface area normalization method, these animal doses correspond to 0.5 and 18 µg lumisterol 2 per kg human body weight or the intake of 35 and 1.26 mg lumisterol 2 , respectively, in a 70 kg person. The question can therefore be posed whether it is possible to achieve such a high intake of lumisterol 2 via the consumption of UV-B treated foods. Data on the quantities of photoisomers in UV-B exposed food for human nutrition are scarce, and photoisomer concentrations can be modified by changing the time of treatment and the radiant intensity used for the production of UV-B treated food sources. [54,55] For example, powder of UV-B-treated Agaricus bisporus mushrooms whose safety has been assessed by the EFSA contained 206 µg g −1 lumisterol 2 and 111 µg g −1 tachsterol 2 . [9] Wittig et al., who exposed oyster mushrooms (Pleurotus ostreatus) to UV-B light, found lumisterol 2 at a concentration of 41.1 µg g −1 dry matter after an irradiation period of 60 min. [11] Provided, that an adult consumes daily 100 g of these UV-B exposed oyster mushrooms (dry matter content of 10%), the estimated lumisterol 2 dose is 0.4 mg kg −1 body weight. This corresponds to a value which lies between the two dosages we used in the current study. However, ascending dose studies are necessary to investigate safety and tolerability of multiple doses of lumisterol 2 and to specify the cut-off level. Because UV-B exposure of foods becomes increasingly important for improving vitamin D supply, it is important to keep attention to possible side effects of photoisomers, not at least due to the potential of photoisomers to be converted to vitamin D 2 .
In conclusion, high lumisterol 2 doses can modulate vitamin D metabolism. The modulations include an increase in the tissue levels of not only vitamin D 2 but also orally administered vitamin D 3 and a strong decline in 25(OH)D and the bioactive compound calcitriol. However, these distinct alterations in vitamin D metabolism are seen only when feeding high levels of lumisterol 2 . Studies 1 and 2 further showed that moderate levels of photoisomers in the food or diet have only a minor impact on vitamin D metabolism. Thus, in the future, the health impact of high photoisomer amounts must be considered when using the UV-B irradiation approach to fortify food with vitamin D.