List of Research

1.Structure-function analysis of mammalian vitamin D hydroxylases

The vitamin D3 25-hydroxylases (CYP27A1 and CYP2R1), 25-hydroxyvitamin D3 1 α-hydroxylase (CYP27B1) and 1α,25-dihydroxyvitamin D3 24-hydroxylase (CYP24A1) are members of the cytochrome P450 superfamily, and key enzymes of vitamin D3 metabolism. Using the heterologous expression in E. coli(mitochondrial CYP27A1, CYP27B1, and CYP24A1) or S. cerevisiae (microsomal CYP2R1), enzymatic properties of the P450s were examined in detail. Upon analyses of the metabolites of vitamin D3 by the reconstituted system, CYP2R1 catalyzed 25-hydroxylation of vitamin D3, but CYP27A1 produced at least seven minor metabolites including 1 α,25(OH)2D3 in addition to the major metabolite 25(OH)D3. These results indicated that human CYP27A1 catalyzes multiple reactions involved in the vitamin D3 metabolism. In contrast, CYP27B1 only catalyzes the hydroxylation at C-1 α position of 25(OH)D3 and 24R,25(OH)2D3. Enzymatic studies on substrate specificity of CYP27B1 suggest that the 1 α-hydroxylase activity of CYP27B1 requires the presence of 25-hydroxyl group of vitamin D3 and is enhanced by 24-hydroxyl group while the presence of 23-hydroxyl group greatly reduced the activity. Eight types of missense mutations in the CYP27B1 gene found in vitamin D-dependent rickets type I (VDDR-I) patients completely abolished the 1α-hydroxylase activity. A three-dimensional model of CYP27B1 structure simulated on the basis of the crystal structure of rabbit CYP2C5 supports the experimental data from mutagenesis study of CYP27B1 that the mutated amino acid residues may be involved in protein folding, heme-propionate binding or activation of molecular oxygen. CYP24A1 expressed in E. coli showed a remarkable metabolic processes of 25(OH)D3 and 1α,25(OH)2D3. Rat CYP24A1 catalyzed six sequential monooxygenation reactions that convert 1α,25(OH)2D3 into calcitroic acid, a known final metabolite of C-24 oxidation pathway In addition to the C-24 oxidation pathway, human CYP24A1 catalyzed also C-23 oxidation pathway to produce 1α,25(OH)2D3-26,23-lactone. Surprisingly, more than 70 % of the vitamin D metabolites observed in a living body were found to be the products formed by the activities of CYP27A1, CYP27B1, and CYP24A1.


2. Metabolism of vitamin D analogues

Vitamin D analogs are potentially useful for clinical treatments of type I rickets, osteoporosis, renal osteodystrophy, psoriasis, leukemia, and breast cancer. On the use of vitamin D analogs, the information of metabolism in target tissues such as kidney, small intestine, and bones is pharmacologically essential. CYP24A1 is considered to be associated with the major metabolic pathways of the vitamin D analogs in these tissues. Species-based difference on the metabolism of vitamin D3 analogs appears to be originated from the difference of CYP24A1-dependent reactions. Since human kidney specimen is not easily obtained and substrate specificities of CYP24A1 vary among species, the development of an in vitro system containing human CYP24A1 was essential for the prediction of the drug metabolism in human kidney. We examined CYP24A1-dependent metabolism of such vitamin D analogs as A-ring diastereomers and 20-epimer of 1α,25(OH)2D3, 2α-propoxy- 1α,25(OH)2D3 2α-(3-hydroxypropoxy)- 1α,25(OH)2D3, and 26,26,26,27,27,27- F6-1α,25(OH)2D3. Good correlation was observed between increasing effect on serum calcium level in rats and kcat/Km value of CYP24A1 for each of the vitamin D analogs. In addition, species-based difference was also observed in CYP24A1-dependent metabolism of vitamin D analogs between humans and rats. These results strongly suggest the usefulness of the recombinant systems harboring CYP24A1 for predicting the metabolism and efficacy of vitamin D analogs before clinical trials.。

3. Structure-function analysis of mammalian cytochromes P450

(i) Prediction of dioxin metabolism in human liver

Metabolism of polychlorinated dibenzo-p-dioxins by cytochrome P450 (P450) and UDP-glucuronosyltransferase (UGT) was examined using a recombinant enzyme system and human liver microsomes. A good correlation (R = 0.92) was observed between 2,3,7-triCDD 8-hydroxylation and phenacetin O-deethylation in human liver microsomes, suggesting that CYP1A2 is responsible for 2,3,7-triCDD 8-hydroxylation in human livers. On the other hand, multiple human UGT isozymes showed glucuronidation activity toward 8-hydroxy-2,3,7-triCDD (8-OH-2,3,7-triCDD). Of these UGTs, UGT1A1, 1A9, and 2B7, which are constitutively expressed in human livers, showed remarkable activity toward 8-OH-2,3,7-triCDD. These results strongly suggest that 2,3,7-triCDD would be firstly converted to 8-OH-2,3,7-triCDD by CYP1A2, then further converted to its glucuronide by UGT1A1, 1A9, and 2B7 in human liver.

(ii) Generation of 2,3,7,8-TCDD-metabolizing enzyme by site-directed mutagenesis of rat CYP1A1

Our previous studies revealed that mono-, di-, and tri-chloro-dibenzo-p-dioxins were good substrates of rat CYP1A1. However, rat CYP1A1 showed no activity toward 2,3,7,8-TCDD. Based on these results, we assumed that enlarging the space of substrate-binding pocket of rat CYP1A1 might generate the catalytic activity toward 2,3,7,8-TetraCDD. Large-sized amino acid residues located at putative substrate-binding sites of rat CYP1A1 were substituted for alanine by site-directed mutagenesis. Among the mutants examined, F240A showed a conversion of 2,3,7,8-TCDD to 8-hydroxy-2,3,7-triCDD. To our best knowledge, the F240A mutant of rat CYP1A1 is the first enzyme to be verified as a 2,3,7,8-TCDD-metabolizing enzyme. In addition, we successfully expressed N-terminal truncated F240A mutant (ΔF240A) in E. colicells. These results suggest possible application of fungal or prokaryotic cells expressing F240A or ΔF240A to the bioremediation of PCDD- contaminated soil.

(iii) Prediction of sesamin metabolism in human liver

We examined metabolism of sesamin by cytochrom P450 (P450) using yeast expression system and human liver microsomes, and found that CYP2C9 was the most important P450 isoform for sesamin catecolization in human liver. We also found a weak mechanism-based inhibition of CYP2C9 by sesamin. Next step, we focused on the metabolism of sesamin monocatechol that was further converted into the dicatechol form by cytochrome P450 or the glucuronide by UDP-glucuronosyltransferase (UGT). Catecholization of sesamin monocatechol enhances its anti-oxidant activity, whereas glucuronidation strongly reduces its anti-oxidant activity. In human liver microsomes, the glucuronidation activity was much higher than the catecholization activity toward sesamin monocatechol. Kinetic studies using recombinant human UGT isoforms identified UGT2B7 as the most important UGT isoform for glucuronidation of sesamin monocatechol.

We also observed the methylation activity toward sesamin monocatechol by catechol O-methyl transferase (COMT) in human liver cytosol. Based on these results, we concluded that CYP2C9, UGT2B7, and COMT played essential roles in the metabolism of sesamin in the human liver

4. Protein engineering ofSt. griseolusCYP105A1 for production of 1α,25-dihydrovitamin D3

CYP105A1 from Streptomyces griseolus has the capability of converting vitamin D3 (VD3) to its active form, 1α,25-dihydroxyvitamin D3 (1α,25(OH)2D3) by a two-step hydroxylation reaction. Our crystal structure analysis has suggested that Arg73 and Arg84 are key residues for the activities of CYP105A1. Thus, we prepared a series of single and double mutants by site-directed mutagenesis focusing on these two residues of CYP105A1 to obtain the hyperactive vitamin D3 hydroxylase. The double mutant R73V/R84A exhibited 435- and 110-fold higher kcat/Km values for the 25-hydroxylation of 1α-hydroxyvitamin D3 and 1α-hydroxylation of 25-hydroxyvitamin D3, respectively, compared with the wild-type enzyme. These values notably exceed those of CYP27A1, which is the physiologically essential VD3 hydroxylase. Thus, we successfully generated useful enzymes of altered substrate preference and hyperactivity. Structural and kinetic analyses of single and double mutants suggest that the amino acid residues at 73 and 84 positions affect the location and conformation of the bound compound in the reaction site, and those in the transient binding site, respectively.

5. PMolecular mechanism of vitamin D actions

(i) Anti-proliferative activity of 25-hydroxyvitamin D3 in human prostate cells


1α-Hydroxylation of 25-hydroxyvitamin D3 is believed to be essential for its biological effects. In this study, we attempted to evaluate the biological activity of 25(OH)D3 itself comparing with the effect of cell-derived 1,25-dihydroxyvitamin D3 (1,25(OH)2D3). First, we measured the cell-derived 1,25(OH)2D3 level in human immortalized prostate cell (PZ-HPV-7) using [3H]-25(OH)D3. The effects of the cell-derived 1,25(OH)2D3on vitamin D3 24-hydroxylase (CYP24A1) mRNA level and the cell growth inhibition were significantly lower than the effects of 25(OH)D3 itself added to cell culture. 25-Hydroxyvitamin D3 1-hydroxylase (CYP27B1) gene knockdown gave no significant effects on the 25(OH)D3-dependent effects, whereas vitamin D receptor (VDR) gene knockdown resulted in a significant decrease of the 25(OH)D3-dependent effects. These results strongly suggest that 25(OH)D3 could directly bind to VDR and exerts its biological functions. DNA microarray and real-time RT-PCR analyses suggested that semaphorin 3B, cystatin E/M, and cystatin D were involved in the antiproliferation activity of 25(OH)D3.

(ii) Cyp27b1-independent biosynthesis of 1α,25-dihydroxyvitamin D3in Cyp27b1 knockout mice

Recently, we revealed that anti-proliferation activity of 25-hydroxyvitamin D3 [25(OH)D3] in human prostate PZ-HPV-7 cells depends on the direct action of 25(OH)D3 through VDR.  Then, we attempted to confirm the direct action of 25(OH)D3 in vivo using Cyp27b1 KO mice. Daily administration of 25(OH)D3 at 250 μg/kg bw/day rescued rachitic conditions of Cyp27b1 KO mice. Plasma levels of Ca, P, and parathyroid hormone, bone mineral density, and female sexual cycle were all normalized by the 25(OH)D3 administration. First, we assumed that it was the direct action of 25(OH)D3. However, to our surprise, normal or higher than normal levels of 1,25(OH)2D3 were detected in the plasma of the Cyp27b1 KO mice. Liver mitochondrial fraction prepared from Cyp27b1 KO mice converted 25(OH)D3 to 1,25(OH)2D3. Because mitochondrial Cyp27A1 has low but detectable 1α-hydroxylase activity towards 25(OH)D3,. we assumed that Cyp27A1 produced enough amount of 1,25(OH)2D3 in the Cyp27b1 KO mice with 25(OH)D3 administration. Our findings suggest that 25(OH)D3 might be useful for the treatment and prevention of osteoporosis in patients with chronic kidney disease.

(iii) Generation and characterization of genetically modified rats using CRISPRA/Cas9 system to reveal vitamin D action mechanism

We tried to generate Cyp27b1 KO rats by CRISPR/Cas9 system, because a body size and blood volume of mice are too small to study metabolism of native vitamin D and its analogs in detail. As expected, Cyp27b1 KO rats showed rickets, and daily administration of 25(OH)D3 rescued rachitic conditions of Cyp27b1 KO rats. Cyp24a1 KO rats were also generated, and vitamin D metabolites in its plasma were analyzed. Plasma 25(OH)D3 level of Cyp24a1 KO rats was significantly elevated, but 24,25(OH)2D3 was not detected therein. We also successfully generated vitamin D receptor (VDR) KO rats and genetically modified rats with one amino acid-substituted VDR. These genetically modified rats may play important roles in the elucidation of metabolism and molecular mechanism of native vitamin D and its analogs.

6. Development of novel bioluminescent sensor to detect and discriminate between vitamin D receptor agonists and antagonists in living cells.

Active forms of vitamin D regulate the expression of multiple genes that play essential roles in calcium and phosphate homeostasis, cell differentiation, and the immune system via the vitamin D receptor (VDR). Many vitamin D analogs have been synthesized for clinical use in the treatment of type I rickets, osteoporosis, renal osteodystrophy, psoriasis, leukemia, and breast cancer. We have constructed two fusion proteins containing split-luciferase and the ligand binding domain (LBD) of the VDR designated as LucN-LBD-LucC and LucC-LBD-LucN. Remarkably, the LucC-LBD-LucN, which has the C-terminal domain of luciferase at the N-terminus of the fusion protein, was a significantly better biosensor than LucN-LBD-LucC. Addition of the VDR agonists to COS-7 cells expressing LucC-LBD-LucN dramatically reduced luciferase activity. In contrast, the VDR antagonist significantly increased the chimeric luciferase activity in a dose- and time-dependent manner. Our results on chimeric luciferases containing the LBDs of mutant VDRs derived from patients with vitamin D-dependent type II rickets indicated that our system could detect a conformational change of the LBD of the VDR likely based on a positional change of the helix 12, which occurs upon ligand binding. This novel system to detect and discriminate between VDR agonists and antagonists could be useful for the screening and identification of chemical compounds that bind to normal or mutant VDRs with high affinity.