Lithocholic Acid Amides as Potent Vitamin D Receptor Agonists

1α,25-Dihydroxyvitamin D3 [1α,25(OH)2D3, 1] is an active form of vitamin D3 and regulates various biological phenomena, including calcium and phosphate homeostasis, bone metabolism, and immune response via binding to and activation of vitamin D receptor (VDR). Lithocholic acid (LCA, 2) was identified as a second endogenous agonist of VDR, though its potency is very low. However, the lithocholic acid derivative 3 (Dcha-20) is a more potent agonist than 1α,25(OH)2D3, (1), and its carboxyl group has similar interactions to the 1,3-dihydroxyl groups of 1 with amino acid residues in the VDR ligand-binding pocket. Here, we designed and synthesized amide derivatives of 3 in order to clarify the role of the carboxyl group. The synthesized amide derivatives showed HL-60 cell differentiation-inducing activity with potency that depended upon the substituent on the amide nitrogen atom. Among them, the N-cyanoamide 6 is more active than either 1 or 3.


Introduction
Vitamin D receptor (VDR) is a ligand-dependent transcriptional factor belonging to the nuclear receptor superfamily [1,2], and mediates most of the biological functions of vitamin D 3 , including calcium and phosphate homeostasis, bone metabolism, and immune regulation. The endogenous agonist of VDR is 1α,25-dihydroxyvitamin D 3 [1α,25(OH) 2 D 3 , 1], an active metabolite of vitamin D 3 (Figure 1), which induces increased expression of target genes. Various vitamin D derivatives have been synthesized as candidate drugs for skin and bone diseases [3][4][5], though most of them have the same secosteroid structure as 1α,25(OH) 2 D 3 (1). [9]. In HL-60 cell differentiation-inducing assay, 3 was more potent than al. recently reported that 3 shows potent vitamin D activity with a lower than 1 [10]. Analysis of the crystal structure of VDR ligand-binding domain ( (Dcha-20) showed that the 3-substituent forms direct hydrogen bonds w dine residues His301 and His393, like the 25-hydroxyl group of 1α,25(O ferent from the indirect interaction, via a water molecule, of the 3-hydro (2) with the same amino acid residues of the VDR [8,10]. The carboxyl hydrogen bonds with Tyr143 and Ser274, as in the case of 1α,25(OH)2D The carboxyl group of LCA (2) also formed indirect hydrogen bonds Ser233 via a water molecule, while the corresponding interaction was no case of 3. The 3-hydroxyl group of 1α,25(OH)2D3 (1) forms direct hydr these amino acid residues in the crystal. Further, our preliminary result cokinetics of 3 indicated that this compound is eliminated very quickly f mice (data not shown), and the carboxyl group is one of the target func improvement of this undesirable feature. Therefore, in this study, we d thesized several lithocholic acid amide derivatives 4-8 with various N-s der to clarify the role of the carboxyl group in the VDR binding and vita 3 ( Figure 2). Lithocholic acid (LCA, 2, Figure 1) is a secondary bile acid formed from chenodeoxycholate, and was identified as a second endogenous agonist of VDR [6][7][8][9]. However, its VDR binding affinity and potency are very low, compared with those of 1. We recently developed a potent lichocholic acid derivative 3 (Dcha-20) that has a 2-hydroxy-2-methylprop-1-yl moiety instead of the 3-hydroxyl group at the 3 α position of 2 ( Figure 1) [10]. In HL-60 cell differentiation-inducing assay, 3 was more potent than 1. Gaikwad, S. et al. recently reported that 3 shows potent vitamin D activity with a lower calcemic activity than 1 [11].
Analysis of the crystal structure of VDR ligand-binding domain (LBD) bound to 3 (Dcha-20) showed that the 3-substituent forms direct hydrogen bonds with the two histidine residues His301 and His393, like the 25-hydroxyl group of 1α,25(OH) 2 D 3 (1), but different from the indirect interaction, via a water molecule, of the 3-hydroxyl group of LCA (2) with the same amino acid residues of the VDR [8,11]. The carboxyl group of 3 forms hydrogen bonds with Tyr143 and Ser274, as in the case of 1α,25(OH) 2 D 3 (1) or LCA (2). The carboxyl group of LCA (2) also formed indirect hydrogen bonds with Arg270 and Ser233 via a water molecule, while the corresponding interaction was not observed in the case of 3. The 3-hydroxyl group of 1α,25(OH) 2 D 3 (1) forms direct hydrogen bonds with these amino acid residues in the crystal. Further, our preliminary results on the pharmacokinetics of 3 indicated that this compound is eliminated very quickly from the serum in mice (data not shown), and the carboxyl group is one of the target functional groups for improvement of this undesirable feature. Therefore, in this study, we designed and synthesized several lithocholic acid amide derivatives 4-8 with various N-substituents in order to clarify the role of the carboxyl group in the VDR binding and vitamin D activity of 3 ( Figure 2).
Compound 3 was synthesized from aldehyde 15a in 7 steps according to our reported method (Scheme 2). Briefly, 15a was reduced to the 3-hydroxymethyl compound 16, followed by tosylation and reaction with sodium cyanide, to afford the nitrile 18. Two-step methylation of 18 gave compound 20 with a 2-hydroxy-2-methylprop-1-yl moiety at the 3 position. The terminal polar group in the side chain of 20 was converted to a carboxyl group in 2 steps to afford 3 (Dcha-20). The amide derivatives 4-8 were synthesized from 3 by the method shown in Scheme 3.

Biological Evaluation
The vitamin D activity of the synthesized lithocholic acid amide derivatives was evaluated in terms of cell differentiation-inducing activity toward human acute promyelocytic leukemia cell line HL-60 [12]. HL-60 cell differentiation was evaluated in terms of the ratio of nitroblue tetrazolium (NBT)-positive cells ( Figure 3 and Table 1). All the amide derivatives examined induced dose-dependent differentiation of HL-60 cells. In this assay, the unsubstituted amide 4a exhibited more potent activity (EC 50 : 0.44 nM), compared with that of the carboxylic acid derivative 3 (Dcha-20, EC 50 : 1.01 nM) or 1α,25(OH) 2 D 3 (1, EC 50 : 0.74 nM). Interestingly, N-monomethylation of compound 4a, yielding compound 4b, diminished the activity. N-Methyl group would disturb the hydrogen bond formation of amide group with the amino acid residues of VDR. The introduction of an N-hydroxyl (compound 5a, EC 50 : 1.18 nM) or N-methoxyl group (compound 5b, EC 50 : 1.45 nM) slightly decreased the activity, though these compounds still showed activity comparable to that of 3. Interestingly, compound 6 bearing an N-cyano group (EC 50 : 0.32 nM) was the most active among the synthesized amide derivatives, being more potent than 3 or 1.
(EC50: 0.45 nM) and 7c (EC50: 0.64 nM) with longer alkyl chains were more active than 7a. A similar tendency was observed for the compounds bearing an N-sulfoalkyl group. Thus, compound 8b (EC50: 6.56 nM) was more active than compound 8a (EC50: 18.5 nM). Terminal polar groups (carboxyl for 7 and sulfo for 8) adjacent to the amide group appear to have a negative effect, possibly blocking hydrogen bond formation of the amide with amino acid residues of VDR, whereas groups more distant from the amide group might have positive effects such as formation of additional hydrogen bond(s). Next, we examined the VDR trasactivation ability for selected compounds ( Figure 4 and Table 1), according to the method reported in our previous study [13]. 1α,25(OH)2D3 (1, EC50: 0.058 nM) and compound 3 (Dcha-20, EC50: 0.083 nM) showed potent transactivation activity at the concentrations above 0.1 nM, while LCA (2) did not show the activity at the concentration below 1 μ M. All the amide derivatives examined showed dose-dependent transactivation activity, which was well correlated with the activity in HL-60 cell assay. Among them compounds 6 (EC50: 0.10 nM) and 7c (EC50: 0.081 nM) were as potent as 1α,25(OH)2D3 (1) and compound 3 (Dcha-20). The results indicated that the differentiation-inducing activity of the lithocholic acid amide derivatives would be mediated by VDR.  Among the three derivatives bearing an N-carboxyalkyl group, compound 7a with one methylene group between the amide and carboxyl groups (EC 50 : 2.03 nM) showed lower activity than the parent amide compound 4a (EC 50 : 0.44 nM), while compounds 7b (EC 50 : 0.45 nM) and 7c (EC 50 : 0.64 nM) with longer alkyl chains were more active than 7a. A similar tendency was observed for the compounds bearing an N-sulfoalkyl group. Thus, compound 8b (EC 50 : 6.56 nM) was more active than compound 8a (EC 50 : 18.5 nM). Terminal polar groups (carboxyl for 7 and sulfo for 8) adjacent to the amide group appear to have a negative effect, possibly blocking hydrogen bond formation of the amide with amino acid residues of VDR, whereas groups more distant from the amide group might have positive effects such as formation of additional hydrogen bond(s).
Next, we examined the VDR trasactivation ability for selected compounds ( Figure 4 and Table 1), according to the method reported in our previous study [13]. 1α,25(OH) 2 D 3 (1, EC 50 : 0.058 nM) and compound 3 (Dcha-20, EC 50 : 0.083 nM) showed potent transactivation activity at the concentrations above 0.1 nM, while LCA (2) did not show the activity at the concentration below 1 µ M. All the amide derivatives examined showed dose-dependent transactivation activity, which was well correlated with the activity in HL-60 cell assay. Among them compounds 6 (EC 50 : 0.10 nM) and 7c (EC 50 : 0.081 nM) were as potent as 1α,25(OH) 2 D 3 (1) and compound 3 (Dcha-20). The results indicated that the differentiationinducing activity of the lithocholic acid amide derivatives would be mediated by VDR.

X-ray Crystallographic Analysis
We next attempted X-ray crystallographic analysis of the complex of rat VDR LB (residues 116-423, Δ165-211) with several of the lithocholic acid amide derivatives. A cording to the method reported in our previous study [9,13], a synthetic peptide conta ing the target sequence of the coactivator MED1 (mediator of RNA polymerase II tra scription subunit 1, also known as ARC205 or DRIP205) was included in the crystallizati solution of VDR LBD and the test compound. However, the analysis was successful on for the complex of compound 7b ( Table 2). The electron density map clearly shows t VDR LBD, the coactivator peptide, the ligand and a relatively low number of water m ecules. Figure 5a shows the overall structure of the VDR LBD complex with 7b; it is simi to those previously reported for VDR LBD complexes with other lithocholic acid deriv

X-ray Crystallographic Analysis
We next attempted X-ray crystallographic analysis of the complex of rat VDR LBD (residues 116-423, ∆165-211) with several of the lithocholic acid amide derivatives. According to the method reported in our previous study [10,14], a synthetic peptide containing the target sequence of the coactivator MED1 (mediator of RNA polymerase II transcription subunit 1, also known as ARC205 or DRIP205) was included in the crystallization solution of VDR LBD and the test compound. However, the analysis was successful only for the complex of compound 7b ( Table 2). The electron density map clearly shows the VDR LBD, the coactivator peptide, the ligand and a relatively low number of water molecules. Figure 5a shows the overall structure of the VDR LBD complex with 7b; it is similar to those previously reported for VDR LBD complexes with other lithocholic acid derivatives [10,14,15]. The interactions of compound 7b with amino acid residues of the VDR LBD (Figure 5b) are compared with those of 3 (Dcha-20) in Figure 5c. The hydroxyl group in the 3-substituent of 7b forms direct hydrogen bonds with two histidine residues, His301 (O···N distance: 2.79 Å) and His393 (O···N distance: 2.68 Å). This is the same as in the case of 3 (Dcha-20), in which the O···N distances were 2.80 Å for His301 and 2.66 Å for His393, whereas LCA (2) forms indirect hydrogen bonds with these amino acid residues via a water molecule. The direct interactions of the hydroxy group in the 3-substituent with two histidines may contribute to the potent activity of 7b and 3. The carboxyl group of compound 3 (Dcha-20) formed hydrogen bonds with the phenolic hydroxyl group of Tyr143 (O··O distance: 2.81 Å) in helix 1 and the hydroxymethyl group of Ser274 (O··O distance: 3.13 Å) in helices 4/5, whereas the amide group of 7b did not form a hydrogen bond with any amino acid residue. Instead, the terminal carboxyl group of 7b formed hydrogen bonds with Arg270 (O···N distance: 2.82 Å) and the backbone amide bond of Tyr143 (O···N distance: 2.91 Å). Similar hydrogen bond formation with these amino acid residues of the VDR LBD was observed in secosteroid derivatives bearing a hydroxylated substituent at the 2-position of the cyclohexane ring. the 2-position of the cyclohexane ring.

Conclusions
We

Conclusions
We designed and synthesized several lithocholic acid amide derivatives of 3 (Dcha-20) as a lead compound. The carboxyl group of compound 3 (Dcha-20) can be replaced with various amide bonds without decreasing the activity. Among the synthesized amide derivatives, compounds 4a, 6 with an N-cyano group and 7b with an N-2-carboxyethyl group showed the most potent activity. Crystallographic analysis of the complex of the VDR LBD with 7b showed that the terminal carboxyl group, but not the amide group, forms hydrogen bonds with amino acid residues of the VDR LBD. The hydroxyl group in the 3-substituent also forms direct hydrogen bonds with two histidine residues, His301 and His393. Recently, compound 3 (Dcha-20) was reported to have lower calcemic activity than 1α,25(OH) 2 D 3 (1) [11], which would be favorable for clinical application, but preliminary studies on the pharmacokinetics of 3 (Dcha-20) indicated that it is eliminated very quickly in mice. The carboxyl group of 3 (Dcha-20) appears to be important for both the potent vitamin D activity and the pharmacokinetic properties. The novel amide derivatives of compound 3 also showed potent vitamin D activities, and studies on their pharamacokinetic properties are now on going. Our results suggest that it may be possible to develop lithocholic acid derivatives having potent activity and drug-like pharmacokinetic properties by chemical modification at the terminal polar group in the side chain.

General
1 H and 13 C NMR spectra were recorded on JNM-ECS 400, JNM-ECS 500, and Bruker Avance 600 spectrometers. The 1 H NMR chemical shifts are reported in parts per million (ppm) relative to the centerline of the singlet signal of the solvent molecule (7.26 ppm for chloroform); coupling constants are given in hertz (Hz). The 13 C NMR chemical shifts are reported in ppm relative to the centerline of the triplet at 77.16 ppm for CDCl 3 . Mass spectral data were obtained on a Bruker Daltonics micro TOF-2focus in the positive and negative ion detection modes.

Synthesis
Synthesis of compound 9: Acetic anhydride (6.174 g, 60.48 mmol) and 4-dimethylaminopyridine (97 mg, 0.80 mmol) were added to a solution of lithocholic acid (1.522 g, 4.04 mmol) in dry pyridine (40 mL). The mixture was stirred for 20 h at room temperature, then quenched with water and extracted with a mixture of ethyl acetate and n-hexane (1:1). The organic layer was washed with 2 M hydrochloric acid and brine, dried over sodium sulfate, and filtered. The filtrate was concentrated to give 9 (1.757 g, quant.) as a yellow solid. 1  Synthesis of compound 10: Triethylamine (533 mg, 5.27 mmol) and ethyl chloroformate (616 mg, 5.680 mmol) were added to a solution of 9 (1.757 g, 4.04 mmol) in distilled THF (40 mL). The mixture was stirred for 2 h at room temperature, then cooled to 0 • C, and sodium borohydride (737 mg, 19.5 mmol) and dry methanol (20 mL) were added to it. The reaction mixture was stirred for 2 h 15 min at 0 • C and then quenched with water. After removal of the solvent in vacuo, the residue was extracted with ethyl acetate. The organic layer was washed with brine, dried over sodium sulfate, filtered, and concentrated. The residue was purified by silica gel column chromatography (ethyl acetate/n-hexane = 1:3) to give 10 (1.672 g, quant.) as a colorless oil. 1 (1.344 g, 2.72 mmol) in dry methanol (30 mL) and distilled THF (5 mL). The mixture was stirred for 6 h 20 min at room temperature under an argon atmosphere and then quenched with acetic acid. After removal of the solvent in vacuo, the residue was extracted with ethyl acetate. The organic layer was washed with water and brine, dried over sodium sulfate, filtered, and concentrated. The residue was purified by silica gel column chromatography (ethyl acetate/n-hexane = 1:3) to give 12 (1.190 g, 97%) as a colorless oil. 1 138.65, 128.32, 127.61, 172.43 Synthesis of compound 13: Sulfuric acid (0.23 mL) was added to a cooled solution of chromium (VI) oxide (267 mg) in water (0.77 mL) just prior to use. An aliquot (0.7 mL) of this Jones reagent was added to a solution of 12 (1.173 g, 2.59 mmol) in dry acetone (30 mL). The mixture was stirred for 30 min at room temperature, then quenched with 2-propanol, and the solvent was removed in vacuo. The extract was extracted with diethyl ether. The organic layer was washed with brine, dried over sodium sulfate, filtered, and concentrated. The residue was purified by silica gel column chromatography (ethyl acetate/n-hexane = 1:4) to give 13 (1.148 g, 98%) as a colorless oil. 1   Synthesis of compound 14: A mixture of (methoxylmethyl)triphenylphosphonium chloride (18.111 g, 52.8 mmol) and potassium tert-butoxide (6.188 g, 46.7 mmol) in distilled THF (112 mL) was stirred for 45 min at 0 • C under an argon atmosphere. A solution of 13 (6.812 g, 15.1 mmol) in distilled THF (18 mL) was added to it. The resulting mixture was allowed to warm to room temperature, stirred for 2 h, quenched with water, and extracted with ethyl acetate. The organic layer was washed with water and brine, dried over sodium sulfate, filtered, and concentrated. The residue was purified by silica gel column chromatography (chloroform/n-hexane = 3:2) to give 14 (7.280 g, quant.) as a mixture of geometrical isomers. Each isomer was isolated in a small amount to determine the structure.  138.75, 138.63, 128.32, 127.61, 127 Synthesis of compound 15: 6 M hydrochloric acid (30 mL) was added to a solution of 14 (6.975 g, 14.6 mmol) in distilled THF (80 mL), and the mixture was stirred for 5 h at room temperature. The reaction mixture was quenched with water, and extracted with ethyl acetate. The organic layer was washed with water and brine, dried over sodium sulfate, filtered, and concentrated. The residue was purified by silica gel column flash chromatography (dichloromethane/n-hexane = 1:1) to give a mixture of 15a and 15b (6.664 g, 98%, 15a:15b = 8:2) as a colorless oil. 15a: 1  The ratio of 15a in the mixture of the epimers was increased by treatment of the mixture with K 2 CO 3 , MeOH, THF (66%), and the isolated 15a was converted to compound 3 (Dcha-20), according to our reported method.
Synthesis of compound 20: Acetyl chloride (0.01 mL, 0.140 mmol) was added to a cooled solution of 3 (64 mg, 0.15 mmol) in methanol (7 mL) at 0 • C under an argon atmosphere. The mixture was stirred for 4 h at room temperature, then quenched with water at 0 • C, and the precipitate was collected to obtain 20 (72 mg, quant.) as a colorless solid. 1 ) were successively added to a solution of 3 (49 mg, 0.11 mmol) in dry dichloromethane (10 mL) at room temperature. After 10 min, N,N'-dicyclohexylcarbodiimide (29 mg, 0.14 mmol) was added to it. The mixture was stirred for 23 h at room temperature, and then filtered. The filtrate was washed with 5% hydrochloric acid, and brine, dried over sodium sulfate, filtered, and concentrated. The residue was purified by silica gel column chromatography (ethyl acetate/n-hexane = 1:2) to give 21 (68 mg, quant.) as a colorless solid. 1  Synthesis of compound 5b: Compound 3 (20 mg, 0.047 mmol) in DMF (0.5 mL) was added to a solution of triethylamine (12 mg, 0.12 mmol) in DMF (0.5 mL). The mixture was stirred at 0 • C for 10 min, and then ethyl chloroformate (6 mg, 0.055 mmol) in DMF (0.5 mL) was added to it. The resulting mixture was stirred at 0 • C for 45 min, and a mixture of methoxyamine hydrochloride (4 mg, 0.052 mmol) and triethylamine (12 mg, 0.12 mmol) in DMF (1.0 mL) was added to it. Stirring was continued at room temperature for 4 h, then the solvent was removed in vacuo, and the residue was extracted with ethyl acetate. The organic layer was washed with brine, dried over sodium sulfate, filtered, and concentrated. The residue was purified by GPC (chloroform) to give 5b (16 mg, 73%) as a colorless solid. 1  Synthesis of compound 5a: Palladium hydroxide (13 mg) was added to a solution of 21 (68 mg, 0.11 mmol) in dry methanol (15 mL). The mixture was stirred for 24 h at room temperature under a hydrogen atmosphere, then filtered, and the filtrate was concentrated. The residue was purified by silica gel column chromatography (ethyl acetate/n-hexane = 2:1, ethyl acetate, then, ethyl acetate/methanol = 20:1) to give 5a (31 mg, 55%) as a colorless solid. 1 13  Synthesis of compound 6: 4-Dimethylaminopyridine (23 mg, 0.19 mmol), cyanamide (19 mg, 0.45 mmol), 1-(3-dimethylaminopropyl)-3-ethylcarbodiimide hydrochloride (42 mg, 0.22 mmol) and N,N-diisopropylethylamine (39 mg, 0.30 mmol) were successively added to a solution of 3 (59 mg, 0.14 mmol) in dry dichloromethane (5 mL). The mixture was stirred for 18 h at room temperature under an argon atmosphere, then diluted with dichloromethane, washed with 2 M hydrochloric acid and brine, dried over sodium sulfate, filtered, and concentrated. The residue was purified by silica gel flash column chromatography (chloroform/methanol = 10:1) to give 6 (5 mg, 89%) as a colorless oil. 1  Synthesis of compound 22: L-Glycine methyl ester hydrochloride (8 mg, 0.06 mmol) and N-methylmorpholine (13 mg, 0.12 mmol) were added to a solution of 3 (20 mg, 0.047 mmol) in dry dichloromethane (8 mL). 1-(3-Dimethylaminopropyl)-3-ethylcarbodiimide hydrochloride (12 mg, 0.06 mmol) was added to the mixture under an argon atmosphere. The resulting mixture was stirred for 24 h at room temperature, then diluted with dichloromethane, washed with 2 M hydrochloric acid, and brine, dried over sodium sulfate, filtered, and concentrated. The residue was purified by silica gel column chromatography (dichloromethane/methanol = 19:1) to give 22a (19 mg, 83%) as a colorless solid. 1