Novel Vitamin D Analogs as Potential Therapeutics: The Metabolism, Toxicity Profiling, and Antiproliferative Activity

Abstract

Aim

To discover novel [20(OH)D₃] analogs as antiproliferative therapeutics.

Materials and Methods

We studied in vitro liver microsome stability, in vivo toxicity using mice, vitamin D receptor (VDR) translocation, in vitro antiproliferative effect, CYP enzyme metabolism.

Results

20S- and 20R(OH)D₃ had reasonable half-lives of 50 min and 30 min (average) respectively in liver microsomes. They were non-hypercalcemic at a high dose of 60 μg/kg. Three new 20(OH)D₃ analogs were designed, synthesized and tested. They showed higher or comparable potency for inhibition of proliferation of normal keratinocytes and in the induction of VDR translocation from cytoplasm to nucleus, compared to 1,25(OH)₂D₃. These new analogs demonstrated different degrees of metabolism by a range of vitamin D-metabolizing CYP enzymes. Conclusion: Their lack of calcemic toxicity at high doses and their high biological activity suggest that this novel 20(OH)D₃ scaffold may represent a promising platform for further development of therapeutically useful agents.

The active form of vitamin D₃, 1,25-dihydroxy vitamin D₃ [1,25(OH)₂D₃, “D₃” stands for “vitamin D₃” and “D₂” stands for “vitamin D₂” hereafter], functions to regulate cell growth, differentiation, proliferation, apoptosis, and immune responses upon binding to the vitamin D receptor (VDR) (1- 3). However, clinical application of vitamin D₃ therapy is largely limited due to hypercalcemic side-effects at pharmacological doses. Hypercalcemia is a disease state where the concentration of calcium in serum is elevated to a level (<10.5 mg/dl) that leads to calcium deposition in soft tissues such as the heart or kidney. The search for vitamin D₃ analogs with minimal or no hypercalcemia-inducing effect has been intensive. As a result, several vitamin D analogs that exhibit reduced hypercalcemia are in clinical trials and are promising as potential therapeutic agents for the treatment of cancer and other inflammatory diseases (1, 2).

We have previously described novel pathways of vitamin D activation started by the action of cytochrome P450scc 11A1 (CYP11A1) on 7-dehydrocholesterol, with subsequent Ultraviolet B (UVB) induced production of secosteroids (3- 11), or initiated directly by CYP11A1 acting on vitamin D₃ (12-15) or D₂ (16, 17). These latter pathways can operate in vivo (18, 19) and commence with the production of 20S-hydroxyvitamin D3 (20S(OH)D₃) or 20S(OH)D₂, the major products of the pathways, although some subsequent metabolism produces other di- and trihydroxy derivatives. Both 20(OH)D₃ and 20(OH)D₂ are noncalcemic in rats (20, 21) and mice (7, 22) show antiproliferative, pro-differentiation and anti-inflammatory properties in epidermal keratinocytes and immune cells (12, 21, 23-25), antiproliferative and tumorostatic activities against solid tumors (6, 21, 22, 26) and leukemia (20), and antifibrotic activity both in vitro and in vivo (7, 27, 28). These secosteroids act as partial agonists on the vitamin D receptor (VDR) (6, 12, 21, 22, 25). In addition, we have established chemical routes of synthesis of 20S(OH)D₃ and 20R(OH)D₃ (6, 29, 30), with the latter showing lower antiproliferative activity than the former (30).

In the current study, we explored the drug-like properties of these two potent lead compounds. We first performed an in vitro metabolic stability study on them with liver microsomes. Both 20S(OH)D₃ and 20R(OH)D₃ showed good metabolic stability. Furthermore, toxicity studies using mice demonstrated that 20S- and 20R(OH)D₃ were nonhypercalcemic at a high dose of 60 μg/kg. Encouraged by these results and to further gain insight into the structure−activity relationships (SAR) of 20-hydroxy-derivatives of vitamin D₃, we designed and synthesized a set of new 20(OH)D₃ analogs using an efficient synthetic route that we established (29, 31). We herein report the characterization of these new 20(OH)D₃ analogs as potential therapeutic agents.

Materials and Methods

Chemistry

1,25(OH)₂D₃ and 25(OH)D₃, were obtained from Sigma-Aldrich, Co. LLC. (St. Louis, MO, USA). All other secosteroids studied were synthesized following a reported procedure (31). 20S(OH)D₃ (29) and 20R(OH)D₃ (30) have been analyzed previously. 20(OH)-20(CH₃)D₃, 20S(OH)-20(hexyl)D₃ and 20S(OH)-20(phenyl)D₃ are new compounds; we have already reported the kinetics of their formation (31) but did not purify and characterize them by nuclear magnetic resonance (NMR) and electrospray ionization mass spectrometry (ESI-MS). The structure and synthesis of these analogs is shown in Figure 1A and B (internal standard). NMR and mass data: 20S(OH)-20(Hexyl)D₃: ¹H NMR (500 MHz, CD₃OD (methanol-D4)): δ 6.21 (d, J=10.0 Hz, 1 H), 6.03 (d, 1 H, J=10.0 Hz), 5.03 (s, 1 H), 4.75 (s, 1 H), 3.72-3.68 (m, 1 H), 2.87-2.39 (m, 3 H), 2.1-1.30 (m, 25 H), 1.26 (s, 3 H), 0.90 (t, J=6.0 Hz, 3 H), 0.75 (s, 3H). ESI-MS: calculated for C₂₇H₄₄O₃, 400.3, found 423.2 [M+Na]⁺. 20(OH)-20(CH₃)D₃: 1H NMR (500 MHz, CD₃OD): δ 6.23 (d, J=10.0 Hz, 1 H), 6.01 (d, 1 H, J=10.0 Hz), 5.00 (s, 1 H), 4.74 (s, 1 H), 3.56-3.60 (m, 1 H), 2.83-2.41 (m, 3 H), 1.31-2.00 (m, 15 H), 1.30 (s, 3 H), 1.25 (s, 3 H), 0.73 (s, 3 H). ESI-MS: calculated for C₂₂H₃₄O₂, 330.3, found 353.3 [M+Na]⁺. 20S(OH)-20(Phenyl)D₃: 1H NMR (500 MHz, CD₃OD): δ 7.45 (d, J=8.0 Hz, 2 H), 7.28 (t, J=8.0 Hz, 2 H), 7.19 (t, J=7.6 Hz, 1 H), 6.21 (d, J=10.0 Hz, 1 H), 6.02 (d, 1 H, J=10.0 Hz), 5.04 (s, 1 H), 4.76 (s, 1 H), 3.54–3.60 (m, 1 H), 2.30-2.42 (m, 2 H), 1.24-2.03 (m, 19 H), 0.75 (s, 3 H). ESI-MS: calculated for C₂₇H₃₆O₂, 392.3, found 415.3 [M+Na]⁺. Deuterated-20S(OH)₇DHC (used as internal standard for metabolic stability study): 1H NMR (500 MHz, CD₃OD): δ 5.54- 5.57 (m, 1 H), 5.39-5.43 (m, 1 H), 3.60-3.66 (m, 1 H), 2.39-2.44 (m, 1 H), 2.25-2.31 (m, 1 H), 2.17-2.23 (m, 1 H), 1.29-2.03 (m, 18 H), 1.28 (s, 3 H), 1.13-1.23 (m, 3 H), 0.94 (s, 3 H), 0.90 (d, J=5.0 Hz, 6 H), 0.79 (s, 3H). ESI-MS: calculated for C₂₇H₄₄O₃, 400.3, found 423.3 [M+Na]⁺. Deuterated-20S(OH)D₃: 1H NMR (500 MHz, CD₃OD): δ 6.22 (d, J=9.8 Hz, 1 H), 6.02(d, 1 H, J=10.0 Hz), 5.04 (s, 1 H), 4.75 (s, 1 H), 3.74−3.78 (m, 1 H), 2.85-2.38 (m, 3 H), 2.20-1.32 (m, 22 H), 1.23 (s, 3 H), 0.89 (d, J=5.2 Hz, 6 H), 0.70 (s, 3H). ESI-MS: calculated for C₂₇H₄₄O₃, 400.3, found 423.2 [M+Na]⁺.

Figure 1.

Synthesis of new 20(OH)D₃ analogs (A) and internal standard (B).

In vitro metabolic stability in liver microsomes.

Metabolic stability studies were performed by incubating the test compounds (0.5 μM, added from 50 μM of stock solution in acetonitrile) in a total reaction volume of 1.2 ml containing 1 mg/ml microsomal protein in reaction buffer [0.2 M of phosphate buffer solution (pH 7.4), 1.3 mM NADP⁺, 3.3 mM glucose-6-phosphate, and 0.4 U/ml glucose-6-phosphate dehydrogenase] at 37°C in a shaking incubator (32). Pooled liver microsomes (Xenotech LLC, Lenexa, KS, USA) from human, mouse, rat, dog, and monkey were utilized to examine metabolic stability. The NADPH regenerating system (solution A and B) was obtained from Xenotech, LLC. Aliquots (100 μl) of the reaction mixtures to determine metabolic stability were taken at 5, 10, 20, 30, 60, 90 and 120 min. Acetonitrile (200 μl) containing 200 nM of the internal standard (Deuterated-20S(OH)D₃, Figure 1B) was added to quench the reaction and to precipitate the proteins. Samples were then centrifuged at 10,000 rpm (8,000 × g) for 15 min at room temperature and the supernatant was analyzed directly by liquid chromatography-mass spectrometry (LC-MS/MS).

In vivo toxicity study.

The potential toxicity of 20S(OH)D₃ and 20R(OH)D₃ was evaluated using male C57/BL6 mice (Charles River Laboratories International, Inc., Wilmington, MA, USA). Mice weighed approximately 25-26 g and were seven weeks of age. Two positive control compounds, 1,25(OH)₂D₃ and 25(OH)D₃, as well as 20S(OH)D₃ and 20R(OH)D₃, were dissolved in autoclaved sesame oil (Sigma-Aldrich) and administered by intraperitoneal (i.p.) injection (50 μl/mouse) once daily for three consecutive weeks to groups of five mice. 20S(OH)D₃ or 20R(OH)D₃ were administered at a dose of 60 μg/kg body weight. For the positive control group, 1,25(OH)₂D₃ or 25(OH)D₃, was given at a dose of 2 μg/kg body weight. Another group of five mice were injected with autoclaved sesame oil, 50 μl/mouse daily, which served as a vehicle control. Clinical signs of toxicity and body weight were assessed twice a week throughout the experimental period. Terminal blood samples (800-1000 μl/mouse) were collected by cardiac puncture at the end of the three-week treatment. The serum (~300 μl/mouse) was immediately separated using BD Microtainers© tubes (Becton, Dickinson and Company, Franklin Lakes, NJ, USA) and stored at −20°C. All animals were sacrificed by cervical dislocation immediately after the blood collection and the main organs (heart, lung, liver, spleen, kidney, adrenal and one piece of skin) of each mouse were collected and stored separately in 10% buffered formalin phosphate solution (Fisher Scientific, Fair Lawn, NJ, USA) for subsequent pathological analysis.

Blood chemistry and clinical pathological analysis

All the blood and serum samples were shipped to Charles River Research Animal Diagnostic Laboratory (Wilmington, MA, USA) within 24 h after collection. The complete pathology, chemistry and hematology [complete blood count (CBC) with differential] profile was assayed and results were provided by Charles River. For pathological analysis, the organs were formalin-fixed and further processed to paraffin blocks. After sectioning, the slides were stained with hematoxylin and eosin. The slides were scanned to create a digital replica of entire tissues on a glass microscopic slide using ScanScope®XT (Aperio Technologies, Inc., Vista, CA, USA) at 0.25 pixel/μm. The scanning process allowed the tissue images to be displayed and analyzed at different magnifications, closely emulating traditional viewing of tissues with a conventional microscope. All the images were stored on a 20TB server that can be accessed through the internet.

Cell proliferation assay (inhibition of DNA synthesis).

We studied the inhibitory effects of 20(OH)D₃ analogs on cell DNA synthesis using similar methods as reported previously (21, 22). Immortalized human epidermal keratinocytes (HaCaT) were cultured in Dulbecco's modified Eagle's medium (DMEM) supplemented with 5% charcoal-treated fetal bovine serum (FBS) (Hyclone, Logan, UT, USA) and 1% antibiotics (penicillin/streptomycin/amphotericin; Sigma–Aldrich). All cultures were performed at 37°C in 5% CO₂. Incorporation of [³H]-thymidine into DNA was used as a measure of cell proliferation. Cells were added to 24-well plates at 5000 cells/well. After overnight incubation at 37°C, the cultures were placed in serum-free media to synchronize cells at the G₀/G₁ phase of the cell cycle. After 24 h, 20(OH)D₃ analogs (dissolved in ethanol and diluted in culture medium) were added with fresh media containing growth supplements and incubated for an additional 72 h. [³H]-Thymidine (specific activity 88.0 Ci/mmol; GE Healthcare, Piscataway, NJ, USA) was added to a final concentration of 0.5 μCi/ml in medium. After 4 h of incubation at 37°C, media were discarded, cells precipitated with 10% trichloroacetic acid (TCA) in phosphate-buffered saline (PBS) for 30 min, washed twice with 1 ml PBS and then incubated with 1 N NaOH/1% sodium dodecyl sulfate (SDS) (250 μl/well) for 30 min at 37°C. The extracts were collected in scintillation vials and 5 ml of scintillation cocktail were added. ³H-Radioactivity incorporated into DNA was measured with a beta-counter (Direct Beta-Counter Matrix 9600; Packard Instruments, Downers Grove, IL, USA).

VDR translocation.

The effects of the new D₃ analogs on VDR translocation from the cytoplasm to the nucleus were tested using SKMEL-188 cells (33), stably transduced with pLenti-CMV-VDREGFP-pgk-puro (VDR and EGFP expressed as fusion protein) (21, 22). Cells were treated with 10^–10-10^–7 M secosteroids overnight for 16 h followed by fixing with 4% paraformaldehyde (PFA). Fixed cells were mounted with fluorescent mounting media (Dako, Carpinteria, CA, USA) and analyzed with a fluorescence microscope. Translocation to the nucleus was determined by counting cells with a fluorescent nucleus and the results are presented as the percentage of the total cells that displayed nuclear staining, as described previously (21). The data were obtained from at least two separate experiments, with images taken randomly from at least six different fields and counted as described elsewhere (21, 22).

Metabolism of vitamin D analogs by cytochrome P450 (CYP)

Activity assays on the new vitamin D₃ analogs were carried out with purified recombinant P450s in a reconstituted system containing adrenodoxin (15 μM), adrenodoxin reductase (4 μM) and either human CYP11A1, human CYP27A1, mouse CYP27B1 or rat CYP24A1, at concentrations of 0.8 to 1.0 μM. Substrates were solubilized in phospholipid vesicles for CYP27B1 and CYP24A1, or in 0.45% 2-hydroxypropyl-β-cyclodextrin for CYP11A1 and CYP27A1. Products were extracted with dichloromethane and analyzed by reverse-phase HPLC. Further details of the P450 expression, purification and activity assays have been described in detail before for CYP11A1 (14, 32), CYP27A1 (18), CYP27B1 (34, 35) and CYP24A1 (36).

Results

20S(OH)D₃ and 20R)OH)D_{3 have good in vitro metabolic}stability with liver microsomes. 20S(OH)D₃ and 20R(OH)D₃ were examined for in vitro metabolic stability using liver microsomes obtained from human, rat, mouse, monkey, and dog. Both of the secosteroids showed species differences in their metabolic stability. Generally, 20S(OH)D₃ was superior to 20R(OH)D₃ with respect to metabolic stability, but both had reasonable half-lives, 50 min for 20S(OH)D₃ and 30 min for 20R(OH)D₃ (average of five different species). Both 20S(OH)D₃ and 20R(OH)D₃ demonstrated the longest halflife (60 and 34 min, respectively) with human liver microsomes, and shortest half-life (37 and 15 min, respectively) with monkey liver microsomes

20S(OH)D₃ and 20R(OH)D₃ are not hypercalcemic at a dose up to 60 μg/kg

We have previously reported that 20S(OH)D₃ had no effect on systemic calcium levels at a dose of 3 μg/kg for seven consecutive days in a rats (20) and at up to 30 μg/kg for three weeks in a mice (22). In the current study, we further assessed any potential toxicity of 20S(OH)D₃ and 20R(OH)D₃ by administering them to C57BL/6 mice with a very high dose of 60 μg/kg. For positive controls, we used 25(OH)D₃ and 1,25(OH)₂D₃ because they are reported to generate hypercalcemia at a concentration ≥1 μg/kg in mice (37) or rats (20). As indicated in Figure 2, 20S(OH)D₃ and 20R(OH)D₃ did not cause hypercalcemia (calcium=8.8±0.66, 8.6±0.77, and 9.3±0.54 mg/dl for 20S(OH)D₃, 20R(OH)D₃, and vehicle control respectively) at the dose of 60 μg/kg for the three-week i.p. treatment. In contrast, 1,25(OH)₂D₃ at a dose of only 2 μg/kg caused an expected significant rise of calcium to 14.6±0.48 mg/dl. 25(OH)D₃ at a dose of 2 μg/kg had a mild hypercalcemic effect, elevating the serum calcium level to 11.5±0.45 mg/dl, beyond the 10.5 mg/dl upper limit for normal serum calcium.

Figure 2.

20S(OH)D₃ and 20R(OH)D₃ do not cause hypercalcemia at a high dose of 60 μg/kg. Mice were treated with i.p. (intraperitoneal) injections of 20S(OH)D₃ or 20R(OH)D₃ for 3 weeks (five mice per treatment group). Data are presented as the mean value of serum calcium concentration (mg/l) for each group. There were no significant differences between the vehicle control group and the 20S/R(OH)D₃ treatment groups at any dose (p>0.05). There were significant differences between the positive control groups and vehicle control group, and between the positive control groups and the 20S/R(OH)D₃ treatment groups (p<0.01). Dotted line: Upper range for normal calcium level in serum (10.5 mg/dl).

Blood chemistry and clinical pathological analysis.

Mice in the vehicle control and all 20S(OH)D₃ or 20R(OH)D₃ treatment groups maintained a healthy weight (25±1 g), whereas those in both 25(OH)D₃ and 1,25(OH)₂D₃ treated groups lost body weight at the dose (2 μg/kg) tested (22±1 g and 21±1g, respectively), which correlated well with their hypercalcemic state. Hematological testing and other clinical chemistry parameters (Figure 3), e.g. alanine aminotransferase, alkaline phosphatase, aspartate aminotransferase, and cholesterol, which are indicators of liver and kidney functions or lipid metabolism, did not reveal any significant difference between groups after three weeks of treatment. Furthermore the hematological profile was similar for the control and all treatment groups (Figure 3A).

Figure 3.

20S(OH)D₃ and 20R(OH)D₃ did not show toxicity in vivo at 60 μg/kg. A: Analysis of complete clinical pathology chemistry profiles testing liver/kidney functions. B: Complete blood count (CBC) with differential results for all treatment groups in the toxicity study. All data are well within the normal physiological range for mice and there are no statistically significant differences between mice in the vehicle control group and all treatment groups. ALT: Alanine aminotransferase; ALK: alkaline phosphatase; AST: aspartate aminotransferase; TBIL: total bilirubin; ALB: albumin; TPR: total protein; BUN: blood urea nitrogen; CREAT: creatine; PHOS: inorganic phosphate; WBC: white blood cell count; BA: basophil granulocytes; NE: neutrophil granulocytes; LY: lymphocytes; MO: monocytes; EO: eosinophil granulocytes; RBC: red blood cell count; Hb: hemoglobin; HCT%: hematocrit percentage; MCV: mean corpuscular volume; MCH: mean corpuscular hemoglobin; MCHC: mean corpuscular hemoglobin concentration; RDW: red cell distribution width; PLT: platelet count; MPV: mean platelet volume.K:103; M:106.

Figure 4 shows representative images of sections of organs from the vehicle control and 20S/R(OH)D₃-treated groups at the dose of 60 μg/kg. The histopathological analysis of the major organs (liver, heart, kidney) from all of the mice treated with 20R(OH)D₃ and six out of the seven mice treated with 20S(OH)D₃ show unremarkable histology. Calcification was not present in any of the mice treated with 60 μg/kg 20S/R(OH)D₃. One of the mice treated with 20S(OH)D₃ was found to have focal patchy liver necrosis centered on hepatic venules (Figure 4G). This type of necrosis is not characteristic of liver damage due to drug toxicity (with the exception of acetaminophen and cocaine) (38), but rather, the central necrosis is likely secondary to passive congestion around the time of death. This is further supported by lack of any sign of toxicity in heart or kidney (Figure 4H). In conclusion, 60 μg/kg of 20S(OH)D₃ is probably the highest possible dose that does not cause noticeable cytotoxic effects.

Figure 4.

Histopathological analyses of representative organs taken from mice in the vehicle control group and mice receiving 60 μg/kg of 20S/R(OH)D₃. No calcification or other abnormalities were observed in any of the organs. Organs were processed and slides were digitized and analyzed. Photographs of various organs were taken at ×10 (left column) and ×100 (right column). Liver (panel A and D) showing normal mouse liver architecture and small uniform hepatocytes. Kidney (panel B and E) with well-developed renal cortex and unremarkable glomeruli, tubules, and interlobular vessels. Heart (panel C and F) sections showing normal myocardium. Liver (panel G) in one of the seven mouse treated with 20S(OH)D₃ showing focal patchy necrosis centered on hepatic venules with minimal inflammatory cells. Features of drug-induced liver disease, e.g. cholestasis, steatosis, fibrosis and cirrhosis, are not present. No toxicities were observed in the kidney (panel H) and heart (not shown) in this mouse. The focal necrosis in this specimen may be attributed to agonal necrosis.

Effects of new vitamin D₃ analogs on cell proliferation

The antiproliferative activities of 20S(OH)D₃ and its metabolite, 20,23(OH)₂D₃, are already well documented, being similar to 1,25(OH)₂D₃ (6, 7, 20, 23-26). In this study we tested new analogs of 20(OH)D₃ synthesized to explore structure−function relationships in the 20(OH)D₃ scaffold, with the aim of improving efficacy. As shown in Figure 5, all the new analogs had higher or comparable antiproliferative activity, as measured by the inhibition of DNA synthesis, to that of 1,25(OH)₂D₃, and at all three concentrations tested. The results suggest that the new 20(OH)D₃ analogs have potent inhibitory activity against proliferation of normal keratinocytes.

Figure 5.

20(OH)D₃ analogs inhibit DNA synthesis in HaCaT keratinocytes and have similar effects to 1,25(OH)₂D₃. HaCaT cells were incubated with the secosteroids for 72 h in DMEM (Dulbecco's Modification of Eagle's Medium) containing 5% charcoal-treated FBS. [³H]-Thymidine was added for the last 4 h of incubation. DNA synthesis was measured by counting the radioactivity incorporated into TCA precipitable material. Data are presented as means±S.E.M. (n=4).

Effects of new vitamin D₃ analogs on VDR translocation to the nucleus

In order to study the effect of novel vitamin D₃ analogs possessing a 20-hydroxyl group on VDR translocation, SKMEL-188 melanoma cells transduced with pLenti-CMVVDR-EGFP-pgk-puro (21) were treated with 10^–1010^–7 M vitamin D₃ analogs, and the percentage of cells with a fluorescent nucleus was determined. 20,23(OH)₂D₃ and 1,25(OH)₂D₃ were used as positive controls. As shown in Figure 6, most of these new vitamin D₃ analogs are more potent than the control compounds: 20,23(OH)₂D₃ and 1,25(OH)₂D₃. The most potent compounds were 20(OH)-20(CH₃)D₃ and 20R(OH)D₃ with half maximal effective concentration (EC₅₀) values of 1.89×10^–12 and 2.41×10^–11 M, respectively. 1,25(OH)₂D₃ was the least active compound in this assay. These result indicate that the new 20(OH)D₃ analogs are generally more potent than 1,25(OH)₂D₃, the active form of vitamin D₃, in the induction of VDR translocation from the cytoplasm to the nucleus.

Figure 6.

The effect of new vitamin D₃ analogs on vitamin D receptor (VDR) translocation from cytoplasm to nucleus. Data are shown as mean±SEM (n≥6). The dose-dependent stimulation of VDR translocation was analyzed by one-way ANOVA with ♯p<0.05 and ♯♯p<0.01. The differences between control and treatment were analyzed with Student's t-test, where *p<0.05, **p<0.01, ***p<0.001, and ****p<0.001.

Metabolism of the new vitamin D analogs by CYP450

The ability of the new analogs to be metabolized by the major mitochondrial P450s involved in vitamin D metabolism was tested (Table I). CYP11A1 which catalyzes 20-hydroxylation of vitamin D₃ (among other reactions) was able to metabolize all the analogs except for 20(OH)-20(CH₃)-D₃ which only has a short side chain. For the other substrates, as many as eight different products were observed. Even the substrate with the bulky phenyl group on the side chain [20S(OH)-20(phenyl)D₃] was well metabolized. 20R(OH)D₃ was metabolized to a greater extent than the well-characterized 20S(OH)D₃ by CYP11A1, where the products are 20,22-dihydroxyD₃, 20,23-dihydroxyD₃ and 17,20-dihydroxyD3 (14, 39).

Table I.

In vitro metabolic stability of 20S(OH)D₃ and 20R(OH)D₃. The percentage of secosteroid remaining was determined by LC-MS/MS. Metabolic stability is presented as half-lives (min) in liver microsomes. Incubations were conducted with 0.5 μM 20(OH)D₃ and 1 mg/ml liver microsomal proteins in the presence of NADPH at 37°C (n=3, mean±S.D.).

Compound	Species
	Rat	Human	Monkey	Dog	Mouse
20S(OH)D3	43±1	60±4	37±1	47±1	42±2
20R(OH)D3	19±3	34±2	15±1	21±2	25±1

CYP27A1, which hydroxylates cholesterol at C26 and vitamin D₃ at C25 (18), was able to metabolize all the analogs except for 20(OH)-20(CH₃)D₃ and 20S(OH)-20(phenyl)D₃. Where metabolism was observed, the number of products ranged from 3 to 5. 20R(OH)D₃ was a particularly good substrate for human CYP27A1, with 90% of the substrate being consumed under the conditions employed. The known products of metabolism of 20S(OH)D₃ by CYP27A1 are 20,25- and 26,26-dihydroxyD₃ (18).

CYP27B1 is the enzyme catalyzing the final step of vitamin D₃ activation, the 1-hydroxylation of 25-hydroxyD₃ to produce 1,25-dihydroxyD₃, the hormonally active form of vitamin D (35). Mouse CYP27B1 was able to act on all the metabolites although metabolism was low for 20(OH)-20(CH₃)D₃. 20S(OH)-20(hexyl)D₃ showed the greatest metabolism and 20S(OH)D₃ was metabolized better than 20R(OH)D₃. In all cases, only a single product was observed, known to be 1,20S-dihydroxyD₃ in the case of 20S(OH)D₃ (34, 35).

CYP24A1 is the enzyme that inactivates 1,25(OH)₂D₃, initially hydroxylating it at C24 or C23 prior to further oxidations to excretory products such as calcitroic acid. Rat CYP24A1 was able to metabolize all the analogs to multiple products, with the short side chain analog, 20(OH)- 20(CH₃)D₃, and the bulky side chain analog, 20S(OH)-20(phenyl)D₃, showing the greatest metabolism. Equivalent metabolism of 20S(OH)D₃ and 20R(OH)D₃ was observed, although one more product was observed with the latter. For 20S(OH)D₃, two out of the three metabolites have been previously identified as 20,24-dihydroxyD₃ and 20,25-dihydroxyD₃ (36).

Discussion

Following previous characterization of 20S(OH)D₃ as a noncalcemic (20, 22) endogenously-produced secosteroid (18) and an excellent candidate for treatment of hyperproliferative disorders (6, 7, 12, 20, 23, 24, 26), we investigated the stability of 20S- and 20R(OH)D3 with liver microsomes, and tested their in vivo toxicity in mice (hypercalcemia) using a high dose of 60 μg/kg. Both 20S- and 20R-(OH)D₃ had reasonable half-lives with 20S(OH)D₃ being more stable (t1/2=50 min, average of fire different species) than 20R(OH)D₃ (t_1/2=30 min). Our previous study showed that 20S(OH)D₃ is not calcemic or toxic at dose up to 30 μg/kg in mice (22). The present study extends this therapeutic window up to 60 μg/kg for 20S(OH)D₃ and also reveals that 20R(OH)D₃ does not cause hypercalcemia at 60 μg/kg. The current study also confirms the previously reported antiproliferative activity of 20R(OH)D₃ (30) and further shows that neither 20S(OH)D₃ or 20R(OH)D₃ alter the blood chemistry of mice at the dose of 60 μg/kg.

Encouraged by the above results with 20R(OH)D₃ and 20S(OH)D₃, we further explored the structure−activity relationship of vitamin D analogs containing a 20-hydroxyl group, with the aim of discovering more potent vitamin D analogs. A focused set of three new 20(OH)D₃ analogs was synthesized and tested for the antiproliferative activity against normal keratinocytes. 20(OH)-20(CH₃)-D₃ was designed to test the effect of having only a short side chain present and remove the chiral center at the C20 position, 20S(OH)-20(hexyl)D₃ was designed to test the effect of having a straight rather than branched side chain, and 20S(OH)-20(phenyl)D₃ was designed to test the effect of an electron-rich, bulky aromatic side-chain replacing the acyclic alkyl chain. These new 20(OH)D₃ analogs showed similar or higher antiproliferative activity on keratinocytes, a known target of vitamin D action, than 1,25(OH)₂D₃. We have previously established that the antiproliferative activities of 20(OH)D₃/D₂ analogs are, at least in part, mediated through the interaction with VDR (6, 12, 21-23, 27). Consistent with this, the new analogs caused translocation of the VDR from the cytoplasm to the nucleus, and their potency was higher than that of 1, 25(OH)₂D₃. Therefore, the new vitamin D analogs have good potential for the treatment of various hyperproliferative diseases, such as cancer, where VDR plays an important role (6). These include malignant melanoma (26, 40) for which therapy is still very limited (41) and leukemia (20).

Vitamin D and its analogs can be metabolized by a number of cytochrome P450s, particularly mitochondrial P450s (11). Metabolism by these enzymes can cause activation or inactivation, depending on the sites of hydroxylation/oxidation (11, 12). 20R(OH)D₃ was metabolized by CYP11A1 more than 20S(OH)D₃. Whether the major metabolites retain biological activity like the products of 20S(OH)D₃ metabolism, remains to be established. All three of the new analogs were metabolized by CYP24A1, the enzyme that inactivates 1,25(OH)₂D₃ but which further activates 20(OH)D₃ (36). CYP27A1 did not metabolize the short side chain derivative, 20(OH)-20(CH₃)D₃, or the bulky 20S(OH)-20(phenyl)D₃. The short side-chain derivative, 20(OH)-20(CH₃)D₃, was also poorly metabolized by CYP27B1. This might be important in the maintenance of low calcemic activity in vivo since 1-hydroxylation of 20(OH)D₃ by CYP27B1 confers calcemic activity on this secosteroid, as it does for 25(OH)D₃ (20). 20(OH)-20(CH₃)D₃ thus looks particularly attractive as a vitamin D therapeutic as it also has the highest potency for VDR translocation and inhibited keratinocyte proliferation at the lowest concentration tested (10^–10 M). It appears more potent in VDR translocation and antiproliferative effects than other short side-chain derivatives of vitamin D previously synthesized (7, 20, 22). In conclusion, the new analogs tested that have a 20-hydroxyl group display high antiproliferative activity and high potency for the induction of VDR translocation to the nucleus, illustrating the potential of the 20(OH)D₃ scaffold for producing therapeutics for the treatment of cancer.

Table II.

Metabolism of new secosteroids by cytochrome P450s. Substrates were solubilized in 0.45% cyclodextrin to a final concentration of 50 μM for incubation with human CYP11A1 (1 μM) and human CYP27A1 (1 μM). For incubations with mouse CYP27B1 (0.8 μM) and rat CYP24A1 (1 μM) substrates were incorporated into phospholipid vesicles at a concentration of 0.025 mol secosteroid/mol phospholipid. Incubations were for 1 h except for 20(OH)-20(CH₃)D₃, 20S(OH)-20(hexyl)D₃ and 20S(OH)-20(phenyl)D₃ with CYP11A1 and CYP27A1, where incubations were for 2 and 3 h, respectively. Products were analyzed by reverse-phase high performance liquid chromatography (HPLC). % Refers to the percentage conversion of substrate to products and No. refers to the number of different products detected representing at least 2% of total secosteroids.

Substrate	CYP11A1		CYP27A1		CYP27B1		CYP24A1
	%	No.	%	No.	%	No.	%	No.
20(OH)-20(CH3)D3	0	0	0	0	2	1	64	2
20S(OH)-20(Hexyl)D3	91	7	18	5	42	1	20	2
20S(OH)-20(Phenyl)D3	83	4	0	0	27	1	63	3
20R(OH)D3	68	8	90	3	13	1	33	4
20S(OH)D3	21	3	41	3	35	1	33	3