CB1 Structural prediction of the adenosine A2A receptor

Min-Wei Liu1 and Jung-Hsin Lin2,1,3

1School of Pharmacy, National Taiwan University, Taipei, Taiwan
2Division of Mechanics, Research Center for Applied Sciences, Academia Sinica, Taipei, Taiwan
3Institute of Biomedical Sciences, Academia Sinica, Taipei, Taiwan


The activation of human adenosine A2A receptor has recently been identified as a candidate target for designing therapeutics for the Huntington disease. Adenosine A2A receptor belongs to the GPCR, rhodopsin-like superfamily. It consists of 7 transmembrane£\-helices, and each is about 25 residues in length. Nowadays, the three-dimensional structure of adenosine A2A receptor obtained from experimental methods is still unavailable. The determination of membrane protein structures via experimental approaches remains a major challenge in the field of structural genomics. An alternative approach to building a molecular model of a protein is from homology modelling procedure. We used bovine rhodopsin crystal structure (PDB code: 1U19, 2.2Å) as homology modelling template to construct the adenosine A2A receptor (1-301) except its c-tail using MODELLER9v1. In the alignment step, we aligned all the rat adenosine receptor family sequences (A1, A2A, A2B, and A3) with rhodopsin sequence to take evolutionary relationship into account and improve the alignment accuracy by using CLUSTALW. The exceptional long c-tail structure of adenosine A2A receptor (302-410) was modeled from best model of TASSER-Lite web server using the fold-recognition approach. After the full-length adenosine A2A receptor structure has been constructed, we put the receptor into lipid-water environment to run 2-ns molecular dynamics simulation refinement using AMBER9. We also made the pharmacophore model of adenosine A2A receptor agonist and antagonist, respectively, using Catalyst®. Then, the most potent agonist and antagonist conformations which fitted its pharmacophore model best were selected. The chosen agonist and antagonist conformations were docked with adenosine A2A receptor using Autodock3. Finally, we made 15-ns molecular dynamics simulations of receptor only, receptor with an agonist, and receptor with an antagonist, to analyze the receptor-ligand binding interactions. The detailed knowledge of the binding locations will help to design more potent inhibitors for this receptor.



CB2 Construction of whole genomic and proteomic trees based on DNA and Protein probes

Chi-Ching Lee (§õ©u«C), Wei-Cheng Lo (ù±©¥¿) and Ping-Chiang Lyu (§f¥­¦¿)

Institute of Bioinformatics and Structural Biology, Department of Life Science, National Tsing Hua University, Hsinchu, Taiwan, ROC


The classification of microorganisms is difficult because of the large variations of morphology and the environmental distributions. Since 1970, people have developed taxonomy systems based on some stable and standard molecular biomarkers; for instance, sequence similarity of 16S rRNA is the first and still wildly used biomarker nowadays for prokaryotes. However, it has been supposed insufficient to classify all kinds of organisms by using one or only a few biomarkers. After the year 2000, the development of genome sequencing techniques has been so rapid that it is now possible to analyze the evolutionary relationships of organisms on the scale of whole genomes.
We have developed a whole genome/proteome clustering approach based on the frequency of biologically meaningful protein signatures and restriction enzyme recognition elements. Such signatures and elements are provided by the Prosite database, which is a collection of annotated motif descriptors from protein families and domains, and REBASE (the Restriction Enzyme dataBASE), which is a collection of restriction enzymes. We compared bacteria, archaea and fungi to build the genomic and proteomic tree by an unsupervised clustering method. The GOLD database (Genome OnLine Database) was consulted as the phenotypic cross-reference.
It was found that our genomic tree grouped together microorganisms with similar GC contents, and proteomic tree clustered bacteria, archaea and fungi into two branches, where the latter two share the same node. Furthermore, the tree built based on Prosite signatures agreed well with the traditional phylogeny at the basal branches while the distal classifications showed different degrees of variations which seemed influenced by specific phenotypes, such as, the parasitism, thermophilicity, methanogenesis and photosynthesis. A web service for genomic and proteomic comparisons has been set up.



CB3 Molecular dynamic simulation and essential dynamic analysis of vigna radiate plant defensin 1

Chan-Lan Shang1, Chao-Sheng Cheng2, Ping-Chiang Lyu3

Institute of Bioinformatics and Structural Biology,Department of Life Science, National Tsing Hua University, Hsinchu


    Vigna radiata plant defensin 1¡]VrD1¡^, a plant defensin, exhibits insecticidal activity. The three-dimension structure of this protein has being determined in our laboratory. This protein consists of three antiparallel £]-Sheet¡Bone £\-helix and one 310 helix to form cysteine-stabilized £\£] motif. As compared with other plant defensins, VrD1 has a unique 310 helix which perhaps stabled by Arg26. In our previous study, the flexibility of 310 helix may be critical for VrD1 biological function, Therefore, we used molecular dynamic¡]MD¡^simulation to study the dynamic property of this protein.
    GROMACS software was used to implement this MD simulation. Simulation system were solvated in a box which full with SPC water model and added Cl- counterions to neutralize system. Consequently, energy minimization and position restraints process were executed to equilibrate overall system, Finally, molecular dynamic simulation was carry out until it reached a length of 20ns. On the other hand, our mutagenesis assay showed that Arg26¡BArg38¡BLys34¡BTrp10 play a significant role in inhibition function. Therefore, we also modeled the structure of these mutations, used MD simulation and essential dynamic to compared their dynamic properties.



CB4 Molecular dynamics and free energy calculations applied to malic enzymes on structural stability

Jia-kai Zhou 1, Sheh-Yi Sheu 2

1 Faculty of Biotechnology and Laboratory Science & Institute of Biotechnology in Medicine, Structural Biology Program, National Yang-Ming University, 155, Li-Nong St., Sec.2, Peitou, Taipei, Taiwan,112, R.O.C.
2 Faculty of Life Sciences and Institute of Genome Sciences, Structural Biology Program, National Yang-Ming University, 155, Li-Nong St., Sec.2, Peitou, Taipei, Taiwan,112, R.O.C.


Malic enzymes are widely distributed in many organisms such as bacteria, plants, and animals in nature. It is a tetrameric protein which the four monomers have identical environment and the dimer interface is much more intimately contacted than tetramer interface. One malic enzyme monomer has about 573 residues and has MW about 66 kDa. Malic enzyme catalyzes the reversible oxidative decarboxylation of L-malate to generate carbon dioxide, pyruvate and NAD(P)H. Trp572 on C-terminal is important for the quaternary structure of human malic enzyme. Previous studies of pigeon liver malic enzyme indicate that malic enzyme is reversibly dissociated and has the tetramer¡Xdimer¡Xmonomer equilibrium when the environmental pH shifts from alkaline to acidic. Here, we performed molecular dynamics simulations and calculated the binding free energy change to study how Trp572 is important in maintaining the quaternary structure of human malic enzyme under different pH values. The preliminary results indicate that substitutions with other amino acids on Trp572 at pH 7.4 or pH4.5 are unfavorable and such substitutions would like to destabilize the tetrameric structures of malic enzymes. In addition, the destabilizations are more remarkable under acidic environment.



CB5 A Docking Study of PDK1 Inhibitor Ligands

Top Lin and Ying-chieh Sun*

Department of Chemistry, National Taiwan Normal University, Taipei, Taiwan


Inhibition of PDK1 kinase activity is one of the targets in developing cancer drug. In the present study, a docking calculation for ligand-PDK1 kinase complexes was carried out using the Autodock program. Experimental binding modes of 9 available lignad-PDK1 structures were reproduced. The binding energies of these complexes were correlated well with experimental IC50 values except 2 ligands. Toward designing better inhibitors, we have carried out docking calculation for a number of derivatives of an inhibitor of low IC50 value (in nM range), UCN. The calculations for several derivatives gave lower binding energy, suggesting that they are better inhibitors. The rationals of the calculated binding energies are discussed. In addition, calculations for a series of celebrex-based compounds were also carried out. Their correlations with experimental available IC50 values will be discussed as well.