Lysophosphatidic acid (LPA) is a chemical mediator with a very simple glycerophospholipid structure. 1-Acyl LPA and 2-acyl LPA are biosynthesized in vivo. Unlike 1-acyl LPA, the biological function of 2-acyl LPA has been hardly elucidated and even organic synthesis of 2-acyl LPA had not been established. We suppressed acyl migration by formation of a salt with a phosphate group in order to synthesize 2-acyl LPA condensed with docosahexaenoic acid.

Site-specific labeling is an important methodology to elucidate the biological function of a target protein. Here, we report a strategy for site-specific chemical labeling, termed the “on-site reaction”. We designed and readily synthesized a bifunctional ligand possessing two reaction sites, an enone and an azide moiety. This strategy involves an on-site conjugate addition reaction with protein followed by a Hüisgen cycloaddition reaction. We demonstrate this strategy by using fluorescein as a probe and peroxisome proliferator activated receptor γ (PPARγ) as a target protein. The reactions were evaluated by ESI-mass analysis and the binding site and modes of binding were revealed by X-ray crystallization analysis. The proposed methodology can easily convert a covalent ligand into chemical tool for protein functional analysis and the identification of drug targets.Download high-res image (306KB)Download full-size image

AbstractIodosylbenzene with triflimide (Tf2NH) efficiently promotes a three-component regioselective synthesis of tetrahydrofuro[2,3-d]oxazoles from homopropargyl alcohols, benzonitrile, and O atoms via oxidative [2+2+1] annulations induced by iodocyclization of the homopropargyl alcohols. In addition, the resulting bicyclic compounds can be converted into 2,4,5-trisubstituted oxazoles bearing hydroxy groups quantitatively; these oxazole derivatives are cores of PPAR agonist analogues.

Vitamin D receptor (VDR) controls the expression of numerous genes through the conformational change caused by binding 1α,25-dihydroxyvitamin D3. Helix 12 in the ligand-binding domain (LBD) is key to regulating VDR activation. The structures of apo VDR-LBD and the VDR-LBD/antagonist complex are unclear. Here, we reveal their unprecedented structures in solution using a hybrid method combining small-angle X-ray scattering and molecular dynamics simulations. In apo rat VDR-LBD, helix 12 is partially unraveled, and it is positioned around the canonical active position and fluctuates. Helix 11 greatly bends toward the outside at Q396, creating a kink. In the rat VDR-LBD/antagonist complex, helix 12 does not generate the activation function 2 surface, and loop 11–12 is remarkably flexible compared to that in the apo rat VDR-LBD. On the basis of these structural insights, we propose a “folding-door model” to describe the mechanism of agonism/antagonism of VDR-LBD.

To develop strong vitamin D receptor (VDR) antagonists and reveal their antagonistic mechanism, we designed and synthesized vitamin D analogues with bulky side chains based on the “active antagonist” concept in which antagonist prevents helix 12 (H12) folding. Of the synthesized analogues, compounds 3a and 3b showed strong antagonistic activity. Dynamic hydrogen/deuterium exchange coupled with mass spectrometry (HDX-MS) and static X-ray crystal structure analyses indicated that compound 3a stabilizes H11–H12 but displaces H6–H7 so that 3a is a novel rather than “active” or “passive” type of antagonist. We classified 3a as a third type of antagonist and called it “H11–H12 stabilization antagonist”. HDX-MS analysis indicated that antagonist 3b is an “active” antagonist. To date there are no reports relating to nuclear receptor antagonist that strongly stabilizes H12. In this study, we found first VDR antagonist that stabilizes H12 and we showed that antagonistic mechanism is diverse depending on each antagonist structure. Additionally, HDX-MS was proven to be very useful for investigations of protein structure alterations resulting from ligand binding.

17-Hydroxy docosahexaenoic acid (17-HDHA) is an oxidized form of docosahexaenoic acid (DHA) and known as a specialized proresolving mediator. We found that a further oxidized product, 17-oxodocosahexaenoic acid (17-oxoDHA), activates peroxisome proliferator-activated receptors γ (PPARγ) and PPARα in transcriptional assays and thus can be classified as an α/γ dual agonist. ESI mass spectroscopy and X-ray crystallographic analysis showed that 17-oxoDHA binds to PPARγ and PPARα covalently, making 17-oxoDHA the first of a novel class of PPAR agonists, the PPARα/γ dual covalent agonist. Furthermore, the covalent binding sites were identified as Cys285 for PPARγ and Cys275 for PPARα.

17(S)-Hydroxy docosahexaenoic acid (17(S)-HDHA) is a specialized pro-resolving mediator. The oxidation of docosahexaenoic acid (DHA) to 17(S)-HDHA using soybean lipoxygenase was accomplished in the presence of the reducing agent TCEP in high yield and high enantio excess. We demonstrated application of this strategy to the synthesis of other fatty acids and to gram scale synthesis of 17(S)-HDHA.

Covalent modification of proteins is important for normal cellular regulation. Here, we report on the covalent modification of peroxisome proliferator-activated receptor γ (PPARγ), an important drug target, by oxo-fatty acids. In this study, ESI mass spectroscopy showed that the reactivities of oxo-fatty acids with PPARγ are different from one another and that these behaviors are related to the structure of the fatty acids. X-ray crystallography showed that three oxo-fatty acids all bound to the same residue of PPARγ (Cys285), but displayed different hydrogen bonding modes. Moreover, fatty acids formed covalent bonds with both PPARγ moieties in the homodimer, one in an active conformation and the other in an alternative conformation. These two conformations may explain why covalently bound fatty acids show partial rather than full agonist activity.

1α,25-Dihydroxyvitamin D3 exerts its actions by binding to vitamin D receptor (VDR). We are continuing the study related to the alteration of pocket structure of VDR by 22-alkyl substituent of ligands and the relationships between the alteration and agonistic/antagonistic activity. Previously we reported that compounds 2 (22-H), 3 (22S-Et), and 4 (22S-Bu) are VDR agonist, partial agonist and antagonist, respectively. Here, we describe the synthesis and biological evaluation of 22S-hexyl analog 5 (22S-Hex), which was designed to be a stronger VDR antagonist than 4. Unexpectedly, 5 showed partial agonistic but not antagonistic activity when bound to VDR, indicating that it is not necessarily true that the bulkier the side chain is, the stronger the antagonistic activity will be. X-ray crystallographic analysis of the VDR–ligand-binding domain (VDR–LBD) accommodating compound 5 indicated that the partial agonist activity of 5 is dependent on the mixed population of the agonistic and antagonistic conformations. Binding of compound 5 may not bring the complex into the only antagonistic conformation due to the large conformational change of the VDR–LBD. From this study it was found that fine tuning of agonistic/antagonistic activity for VDR is possible by 22-alkyl chain length of ligands.Figure optionsDownload full-size imageDownload as PowerPoint slide

We are continuing to study the structural basis of vitamin D receptor (VDR) agonism and antagonism by using 22S-alkyl vitamin D analogues. Here we report the synthesis and biological evaluation of 22R-alkyl analogues and the X-ray crystallographic analysis of vitamin D receptor ligand-binding domain (VDR-LBD) complexed with a 22R-analogue. VDR-LBD complexed with the partial agonist 8a showed that 8a binds to VDR-LBD with two conformations, one of which is the antagonist/VDR-LBD complex structure and the other is the agonist/VDR-LBD complex structure. The results indicate that the partial agonist activity of 8a depends on the sum of antagonistic and agonistic activities caused by the antagonist and agonist binding conformers, respectively. The structural basis observed here must be applicable to the partial agonism of other ligand-dependent nuclear receptors. This is the first report describing the trapping of a conformational subset of the ligand and the nuclear receptor in a single crystal.

Activated thrombin-activatable fibrinolysis inhibitor (TAFIa) is a zinc-containing carboxypeptidase and significantly inhibits fibrinolysis. TAFIa inhibitors are thus expected to act as profibrinolytic agents. We recently reported the design and synthesis of selenium-containing inhibitors of TAFIa and their inhibitory activity. Here we report the crystal structures of potent selenium-, sulfur-, and phosphinic acid-containing inhibitors bound to porcine pancreatic carboxypeptidase B (ppCPB). ppCPB is a TAFIa homologue and is surrogate TAFIa for crystallographic analysis. Crystal structures of ppCPB complexed with selenium compound 1a, its sulfur analogue 2, and phosphinic acid derivative EF6265 were determined at 1.70, 2.15, and 1.90 Å resolution, respectively. Each inhibitor binds to the active site of ppCPB in a similar manner to that observed for previously reported inhibitors. Thus, in complexes, selenium, sulfur, and phosphinic acid oxygen coordinate to zinc in ppCPB. This is the first observation and report of selenium coordinating to zinc in CPB.

Available therapies for thromboembolic disorders include thrombolytics, anticoagulants, and antiplatelets, but these are associated with complications such as bleeding. To develop an alternative drug which is clinically safe, we focused on activated thrombin-activatable fibrinolysis inhibitor (TAFIa) as the target molecule. TAFIa is a zinc-containing carboxypeptidase that significantly inhibits fibrinolysis. Here we designed and synthesized selenium-containing compounds 5–13 to discover novel TAFIa inhibitors having a superior zinc-coordinating group. Compounds 5–13 significantly inhibited TAFIa activity (IC50 2.2 × 10–12 M – 2.6 × 10–6 M). We found that selenol is a better functional group than thiol for coordinating to zinc at the active site of TAFIa. Furthermore, compound 12, which has an amino-chloro-pyridine ring, was found to be a potent and selective TAFIa inhibitor that lacks carboxypeptidase N inhibitory activity. Therefore, compound 12 is a promising candidate for the treatment of thromboembolic disorders. This is the first report of a selenium-containing inhibitor for TAFIa.

Previously, we reported that 22S-butyl-25,26,27-trinor-1α,24-dihydroxyvitamin D32 represents a new class of antagonist for the vitamin D receptor (VDR). The crystal structure of the ligand-binding domain (LBD) of VDR complexed with 2 showed the formation of a butyl pocket to accommodate the 22-butyl group and insufficient interactions between ligand 2 and the C-terminus of VDR. Here, we designed and synthesized new analogues 5a–c and evaluated their biological activities to probe whether agonistic activity is recovered when the analogue restores interactions with the C-terminus of VDR. Analogues 5a–c exhibited full agonistic activity in transactivation. Interestingly, 5c, which bears a 24-diethyl group, completely recovered agonistic activity, although 3c and 4c act as an antagonist and a weak agonist, respectively. The crystal structures of VDR-LBD complexed with 3a, 4a, 5a, and 5c were solved, and the results confirmed that butyl pocket formation in VDR strongly affects the agonistic or antagonistic behaviors of ligands.

Tetracosahexaenoic acid (C24:6n-3, THA, 3) is an essential biosynthetic precursor in mammals of docosahexaenoic acid (C22:6n-3, DHA, 1), the end-product of the metabolism of n-3 fatty acids. THA 3 is present in commercially valuable fishes, such as flathead flounder. Tricosahexaenoic acid (C23:6n-3, TrHA, 2), an odd-numbered-chain fatty acid, has been identified from marine organisms such as the dinoflagellate, Amphidinium carterae. To date, few studies have examined THA 3 and TrHA 2 due to difficulties in detecting and identifying these compounds, so their chemical and biological properties remain poorly characterized. Only one methodology for the chemical synthesis of THA 3 has been presented, and no method for the synthesis of TrHA 2 has been reported. We report here the efficient synthesis of THA 3 in four steps in 56% overall yield, and the synthesis of TrHA 2 in six steps in 48% overall yield. We also present the synthesis of Δ2-THA 4, an intermediate of β-oxidation of THA 3 to DHA 1, in three steps in 73% overall yield.

We previously reported that 22S-butyl-25,26,27-trinor-1α,24-dihydroxyvitamin D3 2 was a potent VDR antagonist. The X-ray crystal structure of the ligand binding domain of VDR complexed with 2 indicated that this ligand induces an extra cavity within the ligand-binding pocket. The structure also showed that the ligand forms only poor hydrophobic interactions with helix 12 of the protein. Here, to study the effects of the induction of the extra cavity and of insufficient interactions with helix 12 on antagonism, we designed and synthesized a series of vitamin D3 analogues with or without a 22-alkyl substituent and evaluated their biological potency. We found that the 22-butyl analogues 3c and 5c act as full antagonists, the 22-ethyl analogues 3b, 4b, 5b, and 22-butyl analogue 4c act as partial agonists, and the others (3a, 4a, 5a, 6a, 6b, and 6c) act as full agonists for VDR. It is intriguing that 6c is a potent agonist for VDR, whereas its 26,27-dinor analogue 5c is a potent antagonist. Analogue 6c recruited coactivator SRC-1 well, but 5c did not. These results indicate that a combination of induction of the extra cavity and insufficient hydrophobic interactions with helix 12 is important for VDR antagonism in this class of ligands.

CYP1A1 and CYP1A2 exhibit catalytic activity predominantly for the 2-hydroxylation of estradiol, whereas CYP1B1 exhibits catalytic activity predominantly for 4-hydroxylation of estradiol. To understand why CYP1B1 predominantly hydroxylates the 4-position of estradiol, we constructed three-dimensional structures of CYP1A1 and CYP1B1 by homology modeling, using the crystal structure of CYP1A2, and studied the docking mode of estradiol with CYP1A1, CYP1A2, and CYP1B1. The results demonstrated that two particular amino acid residues for each CYP, namely Thr124 and Phe260 of CYP1A2, Ser122 and Phe258 of CYP1A1, and Ala133 and Asn265 of CYP1B1, play an important role in estradiol recognition.

To identify novel vitamin D receptor (VDR) ligands that induce a novel architecture within the ligand-binding pocket (LBP), we have investigated eight 22-butyl-1α,24-dihydroxyvitamin D3 derivatives (3−10), all having a butyl group as the branched alkyl side chain. We found that the 22S-butyl-20-epi-25,26,27-trinorvitamin D derivative 5 was a potent VDR agonist, whereas the corresponding compound 4 with the natural configuration at C(20) was a potent VDR antagonist. Analogues with the full vitamin D3 side chain were less potent agonist, and whether they were agonists or antagonists depended on the 24-configuration. X-ray crystal structures demonstrated that the VDR-LBD accommodating the potent agonist 5 has an architecture wherein the lower side and the helix 11 side of the LBP is simply expanded relative to the canonical active-VDR situation; in contrast, the potent antagonist 4 induces an extra cavity to accommodate the branched moiety. This is the first report of a VDR antagonist that generates a new cavity to alter the canonical pocket structure of the ligand occupied VDR.

Erymelanthine 1 and 8-oxoerymelanthine 2 are unique erythrina alkaloids containing a pyridine ring. We synthesized (±)-8-oxoerymelanthine 2 in 2.0% overall yield using the following key reactions. The characteristic 6-5-6-6-membered ring system was constructed by the stereoselective intermolecular Diels−Alder reaction. Oxidative cleavage of the aromatic D-ring was conducted chemo- and regioselectively by ozonolysis in the presence of BF3-etherate. This cleavage site is identical to the site cleaved during the biosynthesis of erymelanthine 1. Nitrogen incorporation was achieved by aminolysis. Conversion of the D-ring pyridone to the corresponding pyridine was efficiently accomplished by palladium-catalyzed reduction of aryl triflate 21. This is not only the first total synthesis of (±)-8-oxoerymelanthine 2 (where the D-ring is pyridine) but also, more importantly, a biomimetic total synthesis of an erythrinan D-aza alkaloid.

To investigate the molecular mechanism of vitamin D receptor (VDR) antagonists having no structurally bulky group interfering with helix 12 of the ligand-binding domain of the VDR, we have synthesized four diastereomers at C(20) and C(23) of 19-nor-1α-hydroxyvitamin D3 25-methylene-26,23-lactone bearing a 2MD-type A-ring. All four analogs showed significant VDR affinity. Transactivation was tested by using Cos7 cells and HEK293 cells. In both types of cells, LAC67a showed little transactivation potency and inhibited the activation induced by the natural hormone concentration-dependently, indicating that LAC67a works as an antagonist for the VDR in these cells. LAC67b, LAC82a and LAC82b similarly acted as VDR antagonists in Cos7 cells, but in HEK293 cells they behaved as potent VDR agonists. Docking of four lactones into the VDR–LBD, followed by structural analysis, demonstrated that each lactone lacks the hydrophobic interaction with helix12 necessary for maintaining the active conformation of the VDR, indicating that these lactones are passive-type antagonists. Furthermore, each docking structure explained the characteristic transactivation profiles of the four lactones. On the basis of our present findings, we suggest that the ligand acts as an agonist if there are appropriate coactivators in the cells to bind to the looser VDR–ligand complex, and as an antagonist if there are no such appropriate coactivators. The molecular basis of the passive antagonism is discussed in detail.

We recently reported that 22S-butyl-1α,24R-dihydroxyvitamin D3 3 recovers the agonistic activity for vitamin D receptor (VDR), although its 25,26,27-trinor analog 2 is a potent VDR antagonist. To investigate the structural features involved in the recovery of agonism, we crystallized the ternary complex of VDR-ligand-binding domain, ligand 3 and coactivator peptide, and conducted X-ray crystallographic analysis of the complex. Compared with the complex with 2, the complex with 3 recovered the following structural features: a pincer-type hydrogen bond between the 24-hydroxyl group and VDR, the conformation of Leu305, the positioning of His301 and His393, the stability of the complex, and intimate hydrophobic interactions between the ligand and helix 12. In addition, we evaluated the potency of both compounds for recruiting RXR and coactivator. The results indicate that the complex with 3 generates a suitable surface for coactivator recruitment. These studies suggest that the action of 2 as an antagonist is caused by the generation of a surface not suitable for coactivator recruitment due to the lack of hydrophobic interactions with helix 12 as well as insufficient hydrogen bond formation between the 24-hydroxyl group and VDR. We concluded that the action of 3 as an agonist is based on the elimination of these structural defects in the complex with 2.

We recently reported that 22S-butyl-1α,24R-dihydroxyvitamin D3 3 recovers the agonistic activity for vitamin D receptor (VDR), although its 25,26,27-trinor analog 2 is a potent VDR antagonist. To investigate the structural features involved in the recovery of agonism, we crystallized the ternary complex of VDR-ligand-binding domain, ligand 3 and coactivator peptide, and conducted X-ray crystallographic analysis of the complex. Compared with the complex with 2, the complex with 3 recovered the following structural features: a pincer-type hydrogen bond between the 24-hydroxyl group and VDR, the conformation of Leu305, the positioning of His301 and His393, the stability of the complex, and intimate hydrophobic interactions between the ligand and helix 12. In addition, we evaluated the potency of both compounds for recruiting RXR and coactivator. The results indicate that the complex with 3 generates a suitable surface for coactivator recruitment. These studies suggest that the action of 2 as an antagonist is caused by the generation of a surface not suitable for coactivator recruitment due to the lack of hydrophobic interactions with helix 12 as well as insufficient hydrogen bond formation between the 24-hydroxyl group and VDR. We concluded that the action of 3 as an agonist is based on the elimination of these structural defects in the complex with 2.