Dongkyu Shin

Molecular mechanism for the regulation of human ACC2 through phosphorylation by AMPK

Acetyl-CoA carboxylases (ACCs) have been highlighted as therapeutic targets for obesity and diabetes, as they play crucial roles in fatty acid metabolism. ACC activity is regulated through the short-term mechanism of inactivation by reversible phosphorylation. Here, we report the crystal structures of the biotin carboxylase (BC) domain of human ACC2 phosphorylated by AMP-activated protein kinase (AMPK). The phosphorylated Ser222 binds to the putative dimer interface of BC, disrupting polymerization and providing the molecular mechanism of inactivation by AMPK. We also determined the structure of the human BC domain in complex with soraphen A, a macrocyclic polyketide natural product. This structure shows that the compound binds to the binding site of phosphorylated Ser222, implying that its inhibition mechanism is the same as that of phosphorylation by AMPK.

Structural basis of triclosan resistance

Triclosan (5-chloro-2-(2,4-dichloro-phenoxy)-phenol, TCL) is a well known inhibitor against enoyl-acyl carrier protein reductase (ENR), an enzyme critical for cell-wall synthesis of bacteria. The inhibitory concentration at 50% inhibition (IC(50)) of TCL against the Escherichia coli ENR is 150nM for wild type (WT), 380, 470 and 68,500nM for Ala, Ser and Val mutants, respectively. To understand this high TCL resistance in the G93V mutant, we obtained the crystal structures of mutated ENRs complexed with TCL and NAD(+). The X-ray structural analysis along with the ab initio calculations and molecular dynamics simulations explains the serious consequence in the G93V mutant complex. The major interactions around TCL due to the aromatic(cation)-aromatic and hydrogen bonding interactions are found to be conserved both in WT and mutant complexes. Thus, the overall structural change of protein is minimal except that a flexible α-helical turn around TCL is slightly pushed away due to the presence of the bulky valine group. However, TCL shows substantial edge-to-face aromatic (π)-interactions with both the flexible R192-F203 region and the residues in the close vicinity of G93. The weakening of some edge-to-face aromatic interactions around TCL in the G93V mutant results in serious resistance to TCL. This understanding is beneficial to design new generation of antibiotics which will effectively act on the mutant ENRs.

A Newly Identified CG301269 Improves Lipid and Glucose Metabolism Without Body Weight Gain Through Activation of Peroxisome Proliferator–Activated Receptor α and γ

OBJECTIVE Peroxisome proliferator–activated receptor (PPAR)-α/γ dual agonists have been developed to alleviate metabolic disorders. However, several PPARα/γ dual agonists are accompanied with unwanted side effects, including body weight gain, edema, and tissue failure. This study investigated the effects of a novel PPARα/γ dual agonist, CG301269, on metabolic disorders both in vitro and in vivo. RESEARCH DESIGN AND METHODS Function of CG301269 as a PPARα/γ dual agonist was assessed in vitro by luciferase reporter assay, mammalian one-hybrid assay, and analyses of PPAR target genes. In vitro profiles on fatty acid oxidation and inflammatory responses were acquired by fatty acid oxidation assay and quantitative (q)RT-PCR of proinflammatory genes. In vivo effect of CG301269 was examined in db/db mice. Total body weight and various tissue weights were measured, and hepatic lipid profiles were analyzed. Systemic glucose and insulin tolerance were measured, and the in vivo effect of CG301269 on metabolic genes and proinflammatory genes was examined by qRT-PCR. RESULTS CG301269 selectively stimulated the transcriptional activities of PPARα and PPARγ. CG301269 enhanced fatty acid oxidation in vitro and ameliorated insulin resistance and hyperlipidemia in vivo. In db/db mice, CG301269 reduced inflammatory responses and fatty liver, without body weight gain. CONCLUSIONS We demonstrate that CG301269 exhibits beneficial effects on glucose and lipid metabolism by simultaneous activation of both PPARα and PPARγ. Our data suggest that CG301269 would be a potential lead compound against obesity and related metabolic disorders.

Structure-based virtual screening approach to the discovery of novel inhibitors of factor-inhibiting HIF-1: Identification of new chelating groups for the active-site ferrous ion

The inhibitors of factor-inhibiting HIF-1 (FIH1) have been shown to be useful as therapeutics for the treatment of anemia. We have been able to identify eight novel FIH1 inhibitors with IC(50) values ranging from 30 to 80microM by means of the virtual screening with docking simulations under consideration of the effects of ligand solvation in the scoring function. The newly identified inhibitors are structurally diverse and have various chelating groups for the active-site ferrous ion including sulfonamide, carboxylate, N-benzo[1,2,5]oxadiazol-4-yl amide, and 2-[1,2,4]triazolo[3,4-b]][1,3,4]thiadiazol-3-yl-quinoline moieties. Each of these four structural classes has not been reported as FIH1 inhibitor, and therefore can be considered for further development by structure-activity relationship or denovo design methods. The interactions with the amino acid residues responsible for the stabilizations of the inhibitors in the active site are addressed in detail.

A nonthiazolidinedione peroxisome proliferator-activated receptor α/γ dual agonist CG301360 alleviates insulin resistance and lipid dysregulation in db/db mice.

Activation of peroxisome proliferator-activated receptors (PPARs) have been implicated in the treatment of metabolic disorders with different mechanisms; PPARα agonists promote fatty acid oxidation and reduce hyperlipidemia, whereas PPARγ agonists regulate lipid redistribution from visceral fat to subcutaneous fat and enhance insulin sensitivity. To achieve combined benefits from activated PPARs on lipid metabolism and insulin sensitivity, a number of PPARα/γ dual agonists have been developed. However, several adverse effects such as weight gain and organ failure of PPARα/γ dual agonists have been reported. By use of virtual ligand screening, we identified and characterized a novel PPARα/γ dual agonist, (R)-1-(4-(2-(5-methyl-2-p-tolyloxazol-4-yl)ethoxy)benzyl)piperidine-2-carboxylic acid (CG301360), exhibiting the improvement in insulin sensitivity and lipid metabolism. CG301360 selectively stimulated transcriptional activities of PPARα and PPARγ and induced expression of their target genes in a PPARα- and PPARγ-dependent manner. In cultured cells, CG301360 enhanced fatty acid oxidation and glucose uptake and it reduced pro-inflammatory gene expression. In db/db mice, CG301360 also restored insulin sensitivity and lipid homeostasis. Collectively, these data suggest that CG301360 would be a novel PPARα/γ agonist, which might be a potential lead compound to develop against insulin resistance and hyperlipidemia.

Structural chemoproteomics and drug discovery

Our laboratories have developed several technologies to accelerate drug discovery process on the basis of structural chemoproteomics. They include SPS™ technology for the efficient determination of protein structures, SCP™ technology for the rapid lead generation and SDF™ technology for the productive lead optimization. Using these technologies, we could determine many 3D structures of target proteins bound with biologically active chemicals including the structure of phosphodiesterase 5/Viagra complex and obtain highly potent compounds in animal models of obesity, diabetes, cancer and inflammation. In this paper, we will discuss concepts and applications of structural chemoproteomics for drug discovery.

Synthesis and Structural Analysis of 6-Aminobicyclo[2.2.1]heptane-2- carboxylic Acid as a onformationally Constrained γ-Turn Mimic

An efficient asymmetric synthesis of 6-aminobicyclo[2.2.1]heptane-2-carboxylic acid as a novel γ-turn mimic has been achieved. Structural analysis of the γ-amino acid derivative was carried out using 1H NMR spectroscopy and intramolecular hydrogen bonding between side chain amides confirmed the turn structure, which had been predicted by Ab initio computational study.

Torsion angle based design of a βI turn adopting dipeptide (Ac-Aib-AzGly-NH2)

Among many peptide conformations, -turns are at the center of interest as it has been identified in bioactive conformations of many peptides. We have attempted to design a model template (acetyl-dipeptide-amide) that can adopt specifically typical torsion angles of a -turn. We believe if and torsion angles of the two residues in the model peptide are fixed by the typical values for i + 1 and i + 2 residues of a specific -turn, the resulting peptide can adopt a -turn to form a hydrogen bond between acetyl and terminal groups.

A Computational Study of Conformational Properties for N-Methylated Azapeptide Models

The modification of peptide structures is a general strategy in drug design to increase the resistance to physiological degradation and decrease the conformational flexibility. Among these modifications, azaamino acids are promising peptidomimetic compounds, which are formed by the replacement of an α-carbon of amino acids with a nitrogen atom. Azaamino acids also could be prepared relatively easily with retention of the side chain in the normal amino acid. Recently, our laboratories reported torsion angle based design, solid phase synthesis and conformational study of a β-I turn adopting peptidomimetic template [1]. During these studies, we found that besides Na, two other nitrogens of an azaamino acid can be alkylated. To understand the effects of such alkylation on the conformation of azaamino acid is critical for the design of conformation directed combinatorial libraries. Thus, we carried out computational studies of model azapeptides in which methyl groups are systematically incorporated, Ac-Sar-NHMe (1), Ac-(NMe)AzAla-NHMe (2), Ac-Sar-NMe2(3) and Ac-(NMe)AzAla-NMe2(4).

Preferred Conformations of Azaamino Acids

Azaamino acids are formed by the replacement of the α-carbon of amino acids with a nitrogen atom. They contain no extra functional group(s) unlike other constrained amino acids, and we believe that the azaamino acid is highly constrained because the Nα-C(O) bond has a double bond character in the urea-type structure. Moreover, when the azaamino acid is incorporated into peptides, the N-Nα bond of azaamino acids is between amide and urea. Since the amide and urea are quite planar, rotation of this bond is restrained. Thus, azaamino acids in peptides have unique conformational preference and their preferred (o, ψ) torsion angles are different from natural amino acids. In spite of wide usage and interesting conformational characteristics of azaamino acids, their conformational studies have not been extensively carried out. Aubry, Boussard and Marraud have made major contributions to this area carrying out crystal-lographic and spectroscopic studies. We also have studied the conformational preference of azaamino acids using computational methods and NMR spectroscopy. In this paper, we report the results of ab initio studies of model peptides containing 6 representative azaamino acids: Ac-AzGly-NHMe (acetyl azaglycine-N-methylamide), Ac-AzAla-NHMe (acetyl azaalanine-N-methylamide), Ac-AzLeu-NHMe (acetyl aza-leucine-N-methylamide), Ac-AzPhe-NHMe (acetyl azaphenylalanine-N-methylamide), Ac-AzAsn-NHMe (acetyl azaasparagine-N-methylamide) and Ac-AzPro-NHMe (acetyl azaproline-N-methylamide).