PP1 Crystal structure and catalytic mechanism of 1,3-fucosyltransferase from Helicobacter pylori

Han-Yu Suna.b, Tzu-Ping Kob, Sheng-Wei Lina,b, Chia-Ling Liua,b, Andrew H.-J. Wang*a,b and Chun-Hung Lin*a,b

aInstitute of Biochemical Sciences, National Taiwan University, Taipei 106, Taiwan
bInstitute of Biological Chemistry, Academia Sinica, Taipei 115, Taiwan


1,3-Fucosyltransferase (FucT) from H. pylori transfers the fucosyl group from the donor GDP--fucose in £\-linkage to the acceptor £]-Gal-1,4-£]-GlcNAc (LacNAc) to produce the Lewis X, which is expressed on the bacterial surface and is structurally similar to tumor-associated carbohydrate antigen found in the host. The enzyme contains a C-terminal leucine zipper and functions as a dimer. With the last 115 residues truncated, the crystal structure of FucT showed two Rossmann-like fold domains typical of the GT-B family of glycosyltransferases. Specific interactions with GDP and GDP-fucose bound to the active site induced conformational changes in the C-terminal domain. Structural comparison with other GT-B members suggested that Glu95 in the N-terminal domain plays the role of general base in catalysis, as confirmed by site-directed mutagenesis. Other mutants at Arg195, Asn240, Glu249 and Lys250 also showed significant decrease in the enzymatic activity. EDTA treatment showed that FucT does not require metal ion. Based on these observations, a catalytic mechanism was proposed. Besides, the truncated FucT formed a dimer in crystal, which may characterize the full-length enzyme.




Effect of metal-binding site geometry on protein structure and function

Dr. Jon D. Wright and Prof. Carmay Lim

IBMS, Academia Sinica, Taipei 11529, Taiwan


We have evaluated the use of a modified CHARMM forcefield with the addition of locally derived charge transfer and polarization (CT+POL) terms between the metal ion and the ion binding residues against the original CHARMM forcefield for proteins. Molecular dynamics simulations of the p53-CD protein, both complexed with DNA and DNA free, using the CT+POL forcefield terms retain the tetrahedral coordination of the Zn2+ binding site in contrast to simulations performed with the original CHARMM forcefield. When comparing the binding interactions of the wild-type p53-CD with DNA and two mutant p53-CDs with DNA, the use of the CT+POL forcefield has been demonstrated to be important in removing artifacts in the free energy decomposition calculations arising due to the change from 4 coordinate with 4 protein residues to 6 coordinate with 4 protein residues and 2 water molecules involved in the Zn2+ binding site.



PP3 Crystal Structures of dTDP-4-keto-2,3,6-trideoxyhexose Reductase (DnmV) from Streptomyces peucetius: Implications for the Inhibition and Catalytic Mechanisms

Yi-Wei Changa, b, Chih-Chien Wuc, Yuh-Ju Sunb, Hsien-Tai Chiuc, and Chwan-Deng Hsiaoa.

aInstitute of Molecular Biology, Academia Sinica, Taipei, 115, Taiwan, ROC; bDepartment of Life Sciences and Institute of Bioinformatics and Structural Biology, National Tsinghua University, Hsinchu, 300, Taiwan, ROC; cInstitute of BioMedical Science, National Chiao Tung University, Hsinchu, 300, Taiwan, ROC


Daunorubicin (DNR) and its C-14-hydroxylated derivative doxorubicin (DXR) which produced by Streptomyces peucetius are clinically important anti-tumor agents in the treatment of a number of malignancies including leukemia, non-Hodgkin lymphoma, and breast cancer. Like many microbial secondary metabolites, DNR and DXR require a deoxyhexose component for their biological activity. The biologically important deoxy-sugar in this case is the 2,3,6-trideoxy-3-aminohexose daunosamine. According to the well characterized biosynthetic pathway of these two anti-tumor components, the last step of daunosamine biosynthesis is performed by the thymidine diphospho-4-keto-2,3,6-trideoxy-3- aminohexose reductase (DnmV). Here, two complex crystal structures of DnmV were determined. One is a binary complex structure in which DnmV coupled with cofactor NADP, and another is a ternary complex structure in which DnmV coupled with cofactor NADP and inhibitor thimidine diphosphate. These two complex structures provide insights into the conformation of DnmV and help us to elucidate the detail inhibition mechanism of TDP. Furthermore, modeling of the saccharide moieties of the non-commercial substrates in the active site based on the DnmV/NADP/TDP ternary complex structure allows us to propose a detailed catalytic mechanism for DnmV. These studies should lead to facilitate efforts to engineer strains that produce larger quantities of more capable and more valuable microbial metabolites.




A DFT/CDM Study of MetãlCarboxylate Interactions in Metalloproteins: Factors Governing the Maximum Number of Metal-Bound Carboxylates

Todor Dudeva and Carmay Limab

aInstitute of Biomedical Sciences, Academia Sinica, Taipei 115, Taiwan
bDepartment of Chemistry, National Tsing Hua University, Hsinchu 300, Taiwan
E-mails: todor@ibms.sinica.edu.tw; carmay@gate.sinica.edu.tw


The number of negatively charged metal-bound Asp/Glu residues determines the net charge of the carboxylate-rich metal-binding site, which has been found to play a role in enhancing the affinity and/or selectivity of a protein cavity for a given metal co-factor. Therefore, it is of interest to know the maximum number of carboxylates that could bind to a given metal (Mq+) of charge q, and the key factors determining this upper limit in protein cavities, which are usually relatively buried. Using density functional theory combined with the continuum dielectric method to compute the H2O¡÷CH3COÕ exchange free energies, the maximum number of carboxylates bound to Mq+ in a relatively buried metal-binding site is found to depend on (i) the metal charge, q, (ii) the carboxylate-binding mode, and (iii) the first-shell carboxylate second-shell ligand interactions. The maximum number of carboxylates bound to Mq+ in a fully/partially solvent inaccessible protein cavity would not likely exceed q+2 if (a) the metal-bound Asp/Glu side chains are hydrogen bonded to a Lys/Arg side chain or several peptide backbone amides/Asn/Gln side chains in the metal¡¦s second coordination shell, or (b) at least one acidic residue binds bidentately, as opposed to monodentately, to the metal cofactor. This number is reduced to q+1 in the absence of stabilizing interactions from outer-shell ligand(s), and if all the carboxylates are bound monodentately to the metal co-factor in a buried cavity. The computational results are consistent with findings from a PDB survey of uni-, di- and trivalent metal-binding sites containing Asp/Glu residues.




Structure-based Functional Analysis of Type-III Geranylgeranyl Pyrophosphate Synthase from Saccharomyces cerevisiae

Tao-Hsin Changa, Rey-Ting Guob, Tzu-Ping Kob, Andrew H.-J. Wanga,b and Po-Huang Lianga,b

aInstitute of Biochemical Sciences, National Taiwan University, Taipei, Taiwan.
bInstitute of Biological Chemistry, Academia Sinica, Taipei, Taiwan.


Geranylgeranyl pyrophosphate synthase (GGPPs) catalyzes the condensation reaction of farnesyl pyrophosphate with isopentenyl pyrophosphate to generate C20 geranylgeranyl pyrophosphate, which is a precursor for carotenoids, chlorophylls, geranylgeranylated proteins, and archaeal ether linked lipid. For short-chain trans-prenyltransferases synthesizing C10¡VC25 products, bulky residues generally occupy the 4th or 5th position upstream from the first DDXXD motif to block further elongation of the final products. However, type-III GGPPs in eukaryotes lack any large residue at these positions. In this study, the first structure of homodimeric type-III GGPPs from Saccharomyces cerevisiae has been determined to 1.98-Å resolution. Each subunit is composed of 15£\-helices joined by connecting loops and is arranged with£\-helices around a large central cavity. An elongated hydrophobic crevice surrounded by D, F, G, H, and I£\-helices contains two DDXXD motifs at the top for substrate binding with one Mg2+ coordinated by Asp75, Asp79, and four water molecules. Distinct from other known structures of trans-prenyltransferases, the N-terminal 17 amino acids (9-amino acid helix A and the following loop) of this GGPPs protrude from the helix core into the other subunit and contribute to the tight dimer formation. Deletion of the first 9 or 17 amino acids caused the dissociation of dimer into monomer and these two mutants of £G(1¡V9) and £G(1¡V17) showed 300-fold decreased and abolished enzyme activity, respectively. Moreover, the Met111 side chains at the interface of two subunits are engaged. Replacement of Met111 with Glu shifted the dimer to monomer with decreasing catalytic activity. The active site is sealed at the bottom with three large residues of Tyr107, Phe108, and His139. Compared to the major product C30 synthesized by mutant H139A, the products generated by mutant Y107A and F108A are predominantly C40 and C30, respectively, suggesting the most important role of Tyr107 in determining the product chain length. Therefore, the new structure of type-III GGPPs, together with other known structures of trans-prenyltransferases, significantly enhances our understanding on the biosynthesis of polyprenyl molecules.




Solution Structure of a Telomere-Binding Domain of Arabidopsis thaliana: New Fold of Telomere-Binding Domain with an Additional Helix in C-terminus

Chih-Pin B. Chung 1, Kuang-Lung Hsueh 1, Hsin-Hao Hsiao 1, Shih-Che Sue 1, Chung Mong Chen 2, Tai-huang Huang 1

1 Institute of Biomedical Sciences, Academia Sinica, Nankang, Taipei, Taiwan, R.O.C.,
2 Institute of Botany, Academia Sinica, Nankang, Taipei, Taiwan, R.O.C.


AtTRP1 is a double stranded telomeric repeat-binding protein (TRP) isolated from Arabidopsis thaliana. In this study we defined the C-terminal 95 amino acids, Gln464 to Val560 as the minimal DNA-binding motif and a C-terminal 40 amino acid extension. To explain the structure-functional relationship, we have applied NMR technique to determine the solution structure of Gln464-Val560. The domain is constituted by four helices instead of three found in other TRPs that the C-terminal extension contains additional fourth helix. The fourth helix stabilizes the overall structure through hydrophobic interaction, thus deletion in this region will abolish its DNA-binding ability. Gln464-Val560 binds to DNA as a monomer in complex with the cognate 13mer duplex DNA fragment. The results of DNA binding-induced chemical shift perturbation suggest that the Myb-like motif in AtTRP1 interacts with DNA via the N-terminal positively charged residues and helix 3. Furthermore, the loop between helix 3 and helix 4 also found to play additional role in the DNA-binding, suggesting that difference DNA recognition mechanism involving the extended c-terminus is present in plants.




The effect of lens crystallin of rat by feeding Natto

T. C. Cioua, W. Chena *, C. Y. Zengb

a NCYU. Department of Applied Chemistry
b NCYU. Department of Food Science & Graduate Institute of Food Science


Glutathione (GSH) is a thiol tripetide, which exhibits significant function in enzyme activity, synthesis of protein and nucleic acid, anti-oxidation and metabolism, and plays an important role in keeping eye lens transparency. The content of GSH in eye lens is extraordinary high and will decrease obviously if lens encounters various insults resulting in lens opacity. We report here the effect of GSH content in rat lens as treated with Natto extracts, which has been well-known for consisting of several nutritional element, such as lecithin, unsaturated fatty acid, and vitamin E. Based on HPLC result, it clearly indicated that the amount of GSH in eye lenses increased for those rats dieted with Natto extracts. And high dosage of Natto extracts would effectively inhibit the aggregation of high molecular weight protein. Besides, the effect of Natto extracts on the structural change of lens protein was discussed according to the Raman spectroscopic results. Preliminary result indicated that the secondary structure of lens protein was mostly in anti-parallel £]-sheet.




Crystal Structure of the Left-Handed Archaeal RadA filament: A Rotational Motif Involved in Homologous DNA Recombination

Li-Tzu Chena,b, Tzu-Ping Kob, Yuan-Chih Changc, Chia-Seng Changc, Ting-Fang Wanga,b and Andrew H.-J.Wanga,b

aInstitute of Biochemical Sciences, National Taiwan University, Taipei 106, Taiwan
bInstitute of Biological Chemistry and cInstitute of Physics, Academia Sinica, Taipei 115, Taiwan Academia Sinica


Homologous recombination mediated by RecA family proteins is an evolutionarily conserved DNA maintenance pathway that protects chromosomes against damage affecting both DNA strands, such as double strand breaks (DSBs) or interstrand crosslinks. The Sulfolobus solfataricus (Sso) RadA, a RecA/Rad51/Dmc1 homolog, is able to promote pairing and exchange of DNA strands with homologous sequences. The RecA family proteins were thought to function as either closed-rings or right-handed helical filaments. We discover that both SsoRadA and yeast Dmc1 proteins in solution could self-polymerize into left-handed helical filaments. Left-handed filament was never seen before in any homologous recombinase. Here we also present the crystal structure of the left-handed SsoRadA filament. In this structure, like those of other RecA family proteins, each monomer is linked to its neighbor via interactions of a £]-strand polymerization motif (PM) with the neighboring ATPase domain. Immediately after PM, we identified a subunit rotation motif (SRM) (residues 83-90) in which axial rotation accompanied with the structural transitions from a left-handed filament to a closed ring, and then a right-handed filament. Comparative analysis of the different recombinase structures in conjunction with SRM point mutant analysis, we conclude that the subunit rotation plays crucial roles in coupling RadA¡¦s different activities, including ATPase, DNA binding, and strand assimilation between homologous sequences. Some key amino acid residues in SRM are conserved from archaeal RadA and Rad51 to eukaryotic Rad51 and Dmc1, suggesting that the subunit rotation is a general mechanism for these recombinases.




Novel mass spectrometry strategies for facile mapping of protein S-nitrosylation sites in relation to identification of controlled cysteine oxidation

Kuan-Ting Pan1; Yi-Yun Chen1; Shu-Yu Lin1; Tzu-Ching Meng2; Kay-Hooi Khoo2

1Core Facilities for Proteomics Research, Academia Sinica, Taipei, Taiwan; 2Institute of Biological Chemistry, Academia Sinica, Taipei, Taiwan


The facile but incomplete loss of the SNO moiety under electrospray conditions without collisional activation readily affords pair of signals related by an m/z value defined by the labile functional groups. Thus the nitrosylated and denitrosylated parent ions can be recognized and automatically selected for MS/MS sequencing during an online LC run if their mass difference was pre-programmed as the criteria for information dependent acquisition. Application to defining the S-nitrosylation sites of phosphatase is reported herein in conjunction with efforts in delineating its catalytic site cysteine modification. As an example, such analysis of PTP1B led to identification of 3 S-nitrosylation sites, including the active site cysteine and two other cysteine residues which were exposed on the surface. Other deeply buried cysteine, as inferred from crystal structure, were also detected and shown not to be modified under the same conditions. These site-specific S-nitrosylations which occurs more readily and to more completeness than oxidation of cysteine to sulfenic, sulfinic and sulfonic acids adversely affects its susceptibility to reversible and irreversible oxidation as mediated by reactive oxygen species. The mass difference method proves to be powerful in etermining S-nitrosylation as the alternative MALDI methods are known to be only indirect since the SNO moiety would be lost and an S-nitrosylated peptide cannot be detected. Online mass difference-triggered MS/MS affords not only selectivity in direct sequencing of modified peptide but also allows better resolution of complex peptide mixtures through LC separation, as required in proteomics applications. Additionally, we demonstrate that the method can be extended to other post-translational modifications especially phosphorylation where an incomplete neutral loss of the phosphate moiety can be readily induced with slight elevation of collision energy. Natural heterogeneity can also be utilized but often impeded by chromatography separation of the related pairs into different time scans.




Mutational study in protease domain of Lon protease from Brevibacillus thermoruber

Jiun-Ly Chir, Dr. and Shih-Hsiung Wu, Dr.

703R, Institute of Biological Chemistry, Academia Sinica, Taipei 115, Taiwan, R.O.C


Cloned protease Lon from Brevibacillus thermoruber had been identified and characterized, but the putative catalytic type (dyad or triad) is still vague. This study is focused on the identification of catalytic type of Bt-Lon with comparison of highly conserved protease domain (P domain) among ATP-dependent Lon proteases and mutations at specific residues within P domain. Mutant enzymes containing substitutions E635A, H664Y, S678A, S678C, S678D, S678E, S678K, K721A and K721Q were obtained by site-directed mutagenesis to check the effect of these P domain mutations on ATPase and protease activity. Unexpectedly, we found only S678D among all mutants almost retained about 90% proteolytic activity of Bt-Lon, whereas other putative S-K dyad mutants (S678A, S678C, S678E, S678K, S678T, K721A, K721Q and K721R) had no proteolytic activity existence. Two proteolytic-related mutants E635A and H664Y indeed presented a little residual protease activity as compared with wild type. Furthermore, both Bt-Lon and S678D were demonstrated by comparison of function with protease activity assay on pH-dependence and inhibition studies and by analysis of quaternary structures with sedimentation velocity and equilibrium, respectively. The pH activity profile displayed a bell-shaped curve for Bt-Lon and S678D under measured casein with apparent pK1 and pK2 value: 6.99¡Ó0.18 and 10.06¡Ó0.18; 6.9¡Ó0.19 and 10.32¡Ó0.2, respectively and they also were strongly inhibited by phenylmethanesulfonyl fluoride (PMSF) with IC50 5 mM. A mixture of monomer-tetramer-hexamer multiple components exists in solution under the maximal solubility but hexamer is a major form for both enzymes. Taken together above results brought us to the following conclusions. Firstly, Bt-Lon probably uses S-K dyad catalytic pair in active site to hydrolyze protein substrates and its ability to the nucleophilicity on oxygen atom at Ser-678 could be enhanced by interacting with general base or other residues from the surroundings upon binding of protein substrates. Secondly, S678D shows the identical active site and mechanism involved in hydrolysis of protein substrates as those identified in Bt-Lon.




Comparative Studies of Protein Folding-Unfolding Induced by Different Denaturants

Y. J. Shiua, U-Ser Jengb, Catherlene Sua, M. Hayashic, S. H. Lina

a Institute of Atomic and Molecular Sciences, Academia Sinica, P. O. Box 23-166, Taipei106, Taiwan R.O.C.
b National Synchrotron Radiation Research Center, Hsinchu 30076, Taiwan.
c Center for Condensed Matter Sciences, National Taiwan University, Taipei 106, Taiwan ROC


In this work, we employed Ising model and the mean field approximation [1] to deduce the thermodynamics properties and microscopic information for the local and global unfolded regular protein. The systematically measurements of the denatured protein, oxidized cytochrome c, with optical spectra probe the local unfolded information, and with the small angle x-ray scattering method probe the protein volume to provide the global unfolded information.

The application of time resolved anisotropy fluorescence measurements identify the unfolded orientation in the cytochrome c. We adopt an ellipsoid model to describe the shape of cytochrome c according to the SAXS results. The analysis indicates the angle between unfolded axis (polar axis) and the vector of emission transition dipole moment is about 55 degree in cytochrome c.

[1] K. K. Liang, M. Hayashi, Y. J. Shiu, Y. Mo, J. Shao, Y.-J. Yan and S. H. Lin, Phys. Chem. Chem. Phys., 5 (2003) 5300¡V5308.




Glycosphingolipid-facilitated membrane insertion and internalization of cobra cardiotoxin: crystal structure of the cardiotoxin/sulfatide complex

Jyung-Hurng Liua, c, Chia-Hui Wangb, c, Shao-Chen Leeb, Wen-guey Wub and Chwan-Deng Hsiaoa.

aInstitute of Molecular Biology, Academia Sinica, Taipei, Taiwan 115 and bDepartment of Life Sciences and Institute of Bioinformatics and Structural Biology, National Tsinghua University, Hsinchu, Taiwan 300, Republic of China. cThese authors contributed equally to this work.


Cobra cardiotoxins, a family of basic polypeptides having lipid- and heparin-binding capacities, induce severe tissue necrosis and systolic heart arrest in snakebite victims. Recent studies showed that CTX A3, the major cardiotoxin from Taiwan cobra venom, binds sulfatide in the outer leaflet of the plasma membrane, and consequently sulfatide mediates CTX A3-induced membrane leakage and CTX A3 internalization into mitochondria. Sulfatide is a glycosphingolipid with 3'-sulfated galactose headgroup. Here we describe the crystal structure of a CTX A3/sulfatide complex in a membrane-like environment at 2.3 Å resolution. CTX A3 recognizes both the headgroup and the ceramide interfacial region of sulfatide and induces a lipid conformational change that may play a key role in CTX A3 oligomerization and cellular internalization.




Interfacial binding intermediates in substrate diffusion of cobra phospholipase A2

Yi-Hung Yeh*,†, Wei-Ning Huang, Yuh-Ju Sun, Hsieh Yi-Heui, Gerard Lambeau¡±,Wen-Guey Wu†,¶ and Chwan-Deng Hsiao†,¶

*Institute of Molecular Biology, Academia Sinica, Taipei, Taiwan; Institute of Bioinformatics and Structural biology, National Tsing Hua University, Hsinchu, Taiwan; Department of Biotechnology, Yuanpei Institute of Science and Technology, Hsinchu, Taiwan; and ¡±Institut de Pharmacologie Moleculaire et Cellulaire, Centre National de la Recherche Scientifique, Sophia Antipolis, France


Phospholipase A2 (PLA2) exhibit high catalytic activities on aggregated substrates via interfacial activation. Recent biophysical characterizations of PLA2 bound to a membrane surface have suggested that both cooperative binding of anionic amphiphiles to the interface and conformational changes of the regulatory N-terminal helix are involved in interfacial activation, but their specific role in either facilitating substrate diffusion and/or a conformational change at the catalytic site remains to be clarified. Herein, we present crystal structures of cobra (Naja atra) PLA2 in complex with anionic sulfate amphiphiles bound at the interfacial and/or catalytic site; we also determine its orientation against phospholipid membranes using the FTIR method. The results suggest that PLA2 bindings to phospholipid membrane induce a tilting of the hydrocarbon chain of phospholipids with an uneven depth of penetration of aromatic residues of the enzyme into the surface. The interfacial amphiphiles are also shown to diffuse from the interfacial binding site to the active site via binding-induced conformational changes as evidenced by a time-dependent change in the crystal form. Thus, interfacial activation of cobra PLA2 may involve a structural intermediate of the enzyme with interfacially bound amphiphiles to facilitate the diffusion of substrates. The intermediate is suggested to behave similarly to the pre-micellar aggregate of the enzyme and involve the binding of amphiphile to anionic binding cluster region of cobra PLA2 with the hydrocarbon tail of the lipid interacting directly with the hydrophobic amino acid residues located near the N-terminal and the pore region.




Crystal Structures of ColE7 in complex DNA/Zn2+ and Im7/Ni2+ Show How a Transition Metal Ion Bound ColE7 Binds and Cleaves DNA

L. G. Doudevaa, H. Huang, Z. Shia, C.-L. Lia, W.-C. Chub and H. S. Yuana

aInstitute of Molecular Biology, Academia Sinica, Taipei, Taiwan, ROC. bInstitutes of Biomedical Engineering, National Yang Ming University, Taiwan, ROC.


The nuclease domain of ColE7 (N-ColE7) contains an H-N-H motif which folds in a metal finger topology. Here we report the crystal structures of a Zn2+-bound N-ColE7 (H545E mutant) in complex with a 12-bp duplex DNA (5¡¦-CGGGATATCCCG-3¡¦) and a Ni2+-bound N-ColE7 in complex with the inhibitor Im7 at a resolution of 2.5 Å and 2.0 Å, respectively. Metal-dependent cleavage assays and site-directed mutagenesis showed that N-ColE7 cleaves double-stranded DNA with a single metal ion cofactor, Ni2+, Mg2+, Mn2+ and Zn2+. In the crystal structure of N-ColE7-DNA complex, the zinc ion is directly coordinated to three histidines and the DNA scissile phosphate in a tetrahedral geometry. In contrast, Ni2+ is bound in N-ColE7 in two different modes, to four ligands (three histidines and one phosphate ion), or to five ligands with an additional water molecule. These data suggest that the divalent metal ion in the His-metal finger motif can be coordinated to six ligands, such as Mg2+ in I-PpoI, Serratia nuclease and Vvn, five ligands or four ligands, such as Ni2+ or Zn2+ in ColE7. Universally, the metal ion in the His-metal finger motif is bound to the DNA scissile phosphate and serves three roles during hydrolysis: polarization of the P-O bond for nucleophilic attack, stabilization of the phosphoanion transition state and stabilization of the cleaved product.




Structural Basis for Sequence-Dependent DNA Cleavage by the Non-Specific Endonuclease

W. Yang, Y. Wang, L G. Doudeva, C.-L. Li and H. S. Yuan

Institute of Molecular Biology Academia Sinica


Non-specific endonucleases hydrolyze DNA without sequence specificity but with sequence preference that they cleave at some sites more efficiently than others. However, the structural basis for sequence-dependent cleavage by the non-specific endonucleases remained elusive. Here we use the non-specific endonucleases ColE7 to dissect this problem. DNA foot printing assays showed that the nuclease domain of ColE7 (N-ColE7) cleaves DNA with a preference for making nicks after (at 3¡¦O-side) thymine bases. The crystal structure of N-ColE7 (H545Q mutant) in complex with an 18-bp DNA was determined at a resolution of 2.8 Å. In the N-ColE7-DNA structure, a preferred thymine residue is located right before the scissile phosphate and the structure of this ¡§preferred¡¨ complex was compared with the previously determined ¡§non-preferred¡¨ complexes in which a guanine is located before the scissile phosphate. The structural comparison shows that the phosphate backbone in the ¡§preferred¡¨ complex is distorted the most, leading to a shorter distance between the zinc ion and the scissile phosphate. This result suggests a general structural basis for the sequence-dependent DNA cleavage that the enzyme-induced DNA backbone conformational change is the local determinant for non-specific endonucleases to decide whether to cleave or not to cleave a DNA.




Structural Hierarchy of Severe Acute Respiratory Syndrome-associated Coronavirus Nucleocapsid Protein

Chung-ke Changa, Shih-Che Suea, Cheng-Kun Tsaib, Yen-Chieh Chiangc, Hsin-I Baia, Lilianty Rianga, Wen-Jin Wudd and Tai-huang Huanga,b,d

aInstitute of Biomedical Sciences, Academia Sinica
bDepartment of Physics, National Taiwan Normal University
cDepartment of Life Sciences, National Tsing Hua University
dHigh Field Nuclear Magnetic Resonance Center, National Research Program for Genomic Medicine


Severe acute respiratory syndrome (SARS) is a novel human disease caused by a new coronavirus (SARS-CoV). We investigated the domain organization of the nucleocapsid protein of SARS-CoV in solution and found that it contains two structural domains sandwiched between three disordered regions. Secondary structure alignment and order-disorder analysis indicate that the domain organization is common to coronavirus nucleocapsid proteins. The N-terminal domain is located between residues 45-181, which has been previously characterized as a putative RNA-binding domain. The C-terminal domain consists of residues 248-365 and is responsible for the dimerization of the nucleocapsid protein. Nuclear magnetic resonance experiments show that the topology of the dimer interface resembles that of the porcine reproductive and respiratory syndrome virus nucleocapsid protein. Deletion studies show that residues 281-365 are sufficient for dimer formation, although the overall structure is stable only when the full C-terminal domain is included. The C-terminal domain also contains RNA-binding activity, which can be mapped to its N-terminus. This region is highly conserved among coronaviruses and contains a high density of positively charged residues. Our results indicate that a common structural hierarchy of coronavirus nucleocapsid proteins exists, and is closely related to the function of the protein.




Inhibition of b-amyloid Peptide Aggregation by Rational Mutation

Ying-En Chena,b, Chi-Jen Loa,b, Chih Ching Wanga,b, Yi-Chen Chenc, Chia-li Sud, Hun-Gen Changd, Hsien-Bin Huangd and Ta-Hsien Lina,b,e

aInstitute of Biochemistry and Molecular Biology, bStructural Biology Program,
National Yang-Ming University, Taipei,112 Taiwan;
cInstitute of Medical Science, Tzu Chi University, Hualien 970, Taiwan;
dInstitute of Molecular Biology, National Chung Cheng University, Chia-Yi 621, Taiwan;
eDepartment of Medical Research & Education, Taipei Veterans General Hospital, Taipei 112, Taiwan


Alzheimer¡¦s disease (AD) is a neurodegenerative disease, which leads to progressive dementia and neuronal death. Currently, the pathogenesis of this disease is still not yet clear. The main histopathological hallmarks of AD are the senile plaques within the cerebral cortex and the neurofibrillary tangles within the nerve cells. The primary component of senile plaques isƒn £]-amyloid peptide (A£])ƒwƒw, which is derived from proteolysis of a much larger membrane-spanning protein known as £]-amyloid precursor protein (APP). A£] has two major isoforms, A£]40 and A£]42, containing 40 and 42 residues, respectively. A£] is a soluble peptide, however, it will polymerize at a very slow rate under certain environmental condition. The aggregation process converts monomeric, soluble A£] to insoluble fibrils that eventually precipitate as amyloid plaques. Recent studies have suggested that A£] has neurotoxic properties in an aggregated state. Studying the unfolding process of A£]ƒn may help us understand the molecular mechanisms of the aggregation process of A£] and facilitate the design and development of fibrillogenesis inhibitors. By applying urea-induced unfolding to probe the conformational properties of A£] in conjunction with NMR analysis, the structural stability of A£]ƒÒ in aqueous sodium dodecyl sulphate (SDS) micelles was characterized at single-residue resolution. The results suggest that residues L17 and F19 are less stable than other residues. Replacement of these two residues by alanine could inhibit A£] aggregation. Structural characterization of the mutant will be presented as well.




Structure of the SARS coronavirus nucleocapsid protein dimerization domain suggests a mechanism for helical packaging of viral RNA

Chun-Yuan Chena,b, Chung-ke Changc,d, Yi-Wei Changa, Shih-che Suec, Hsin-I Baic, Lilianty Riangc, Tai-huang Huangc, Chwan-Deng Hsiaoa.

aInstitute of Molecular Biology, Academia Sinica, Taipei, 115, Taiwan, ROC; bGraduate Institute of Cell and Molecular Biology, Taipei Medical University, Taipei, 110, Taiwan, ROC; cInstitute of Biomedical Sciences, Academia Sinica, Taipei, 115, Taiwan, ROC


The nucleocapsid protein of the severe acute respiratory syndrome (SARS) coronavirus contains two structural domains: the N-terminal putative RNA-binding domain (RBD, residues 45-181) and the C-terminal dimerization domain (DD, residues 248-365), flanked by disordered regions. Here we show that DD binds to ssRNA or ssDNA through its N-terminal positively charged region and with an affinity higher than that of RBD. However, the two domains show synergistic effect in binding to ssRNA and ssDNA. We further report the crystal structure of the DD to a 2.5Å resolution. In the crystal, the DD exists as dimers and four dimer molecules form a ring-like octameric structure of 90 Å in diameter and with a central cavity of 30 Å in diameter in an asymmetric unit. Packaging of the octamers in the crystal created two parallel, positively charged grooves due to the ten N-terminal positively charged residues in the grooves. The grooves wind around the octamers as two left-handed helices. The pattern suggests a novel mechanism for fast and efficient helical viral RNA packaging. Stabilization of the helical RNP requires the synergistic interaction of both RBD and DD, where the DD stabilizes the negatively charged phosphate backbone through non-specific binding and the RBD stabilizes the exposed bases through base-stacking between the several conserved aromatic groups and the RNA bases. The structure thus formed conforms to known features of coronavirus ribonucleoproteins and is supported by extensive experimental data.




Crystal structure of the human FOXK1a/DNA complex and its implications on the diverse binding specificity of winged helix/forkhead proteins

Kuang-Lei Tsaia,b, Cheng-Yang Huanga, Chia-Hao Changc, Yuh-Ju Sunb, Woei-Jer Chuangc,* and Chwan-Deng Hsiaoa,*

From the aInstitute of Molecular Biology, Academia Sinica, Taipei, 115, Taiwan, bInstitute of Bioinformatics and Structural Biology, National Tsing Hua University, Hsinchu, 300, Taiwan, and cDepartment of Biochemistry, National Cheng Kung University, College of Medicine, Tainan 701, Taiwan, Republic of China.


Interleukin enhancer binding factor (ILF) is a human transcription factor and a new member of the winged helix/forkhead family. ILF can bind to purine-rich regulatory motifs such as the human T-cell leukemia virus long terminal region (HILV-1 LTR) and the interleukin-2 (IL-2) promoter. Here we report the 2.4 Å crystal structure of two DNA-binding domains (DBDs) of ILF (FOXK1a) binding to a 16-base pair DNA duplex containing a promoter sequence. Electrophoretic mobility shift assay (EMSA) studies demonstrate that two ILF-DBD molecules cooperatively bind to DNA. In addition to the recognition helix recognizing the core sequences through the major groove, the structure shows that wing 1 interacts with the minor groove of DNA, and the H2-H3 loop region makes ionic bonds to the phosphate group, which permits the recognition of DNA. The structure also reveals that the presence of the C-terminal £\-helix in place of a typical wing 2 in a member of this family alters the orientation of the C-terminal basic residues (RKRRPR) when binding to DNA outside the core sequence. These results provide a new insight into how the DNA-binding specificities of winged helix/forkhead proteins may be regulated by their less conserved regions.




Prolyl Dipeptidase DPP8 and its Dimeric Inactivating Mutations

Hong-Jen Leea, Yuan-Shou Chena, Chi-Yuan Choub, Chia-Hui Chiena, Chun-Hung Linc, Gu-Gang Changb, and Xin Chena#

From aDivision of Biotechnology and Pharmaceutical Research, National Health Research Institutes, Zhu Nan town, Miaoli 350, bFaculty of Life Science, National Yang-Ming University, Taipei 112, cInstitute of Biological Chemistry, Academia Sinica, Taipei 115, Taiwan, ROC


DPP8 belongs to the family of prolyl dipeptidases, which are capable of cleaving the peptide bond after a penultimate proline residue. Because of their unique proline-cleaving activity and their potential as drug targets, the prolyl dipeptidases, including DPP8, DPP-IV, and FAP, have attracted intensive investigation in recent years. However, little information is available on the structure and function of DPP8, and data on the quaternary structure of DPP8 have been conflicting. In this study, we demonstrate, both in vivo and in vitro, that DPP8 protein is dimeric, with small quantities of tetramers present. We have discovered that single-site mutations in the C-terminal loop inactivate DPP8 enzymatic activity without disrupting its quaternary structure, which is significantly different from the homologous enzyme DPP-IV. Three of the DPP8 mutations (F822A, V833A, and Y844A) affect both kcat and Km dramatically, whereas H859A affects only Km. In contrast, the reduced activity of monomeric DPP-IV is largely a kcat effect, independent of the mutation site. Our study shows that the C-terminal loop is critical for the enzymatic activity of DPP8, and that dimerization itself is not sufficient for its optimal enzymatic activity. Finally, we have investigated the substrate preference of DPP8 at both P1 and P2 sites using a position-scanning dipeptide library. These data extend the functional importance of the C-terminal loop to the whole family of prolyl dipeptidases. Despite their sequence homology and the conservation of their dimeric structure, there are fundamental differences in the regulation of the enzymatic activities of the prolyl dipeptidases.




Structural Study of KP3573, a Polymyxin B Resistant Protein (pmrD) from Klebsiella pneumoniae

Shih-Chi Luoa, b, Yuan-Chou Loub, and Chinpan Chena, b

aTaiwan International Graduate Program, bInstitute of Biomedical Science, Academia Sinica, Taipei, Taiwan.


Klebsiella pneumoniae (KP) is one of the top five pathogenic bacteria that cause community and hospital infection of pneumonia in Taiwan. In some immuno-deficient patients such as diabetic, KP forms opportunistic infection and causes septicemia combining with Pyogenic liver abscess. Our laboratory, a member of the KP functional and structural genomics team of Academia Sinica, has selected several target proteins that are related to virulence and drug-resistance factors for structural/functional study. KP3573, annotated as a polymixin b resistant protein (PmrD), is one of the chosen target proteins.
It is known that PmrD is a small protein which can mediate the activation of a PmrA-PmrB two-component system by another PhoP-PhoQ two-component system that responds to the magnesium levels in the environment. The PmrA-PmrB two¡Vcomponent system in Salmonella enterica serovar typhimurium is required for resistance to polymyxin B and to other antimicrobial compounds. Consequently, PmrD may be a promising candidate related to bacterial drug resistance and is deserved for further structural study.
To date, we have expressed and purified this target protein in a large quantity and carried out a variety of biophysical studies, especially high-resolution NMR experiments. The chemical shifts index (CSI) indicates that KP3573 is mainly composed of six beta-strands and the final refinement of 3D structure is ongoing by using the XPLOR program. The NMR structural study as well as other biophysical studies on KP3573 will be discussed in this poster presentation.




Human Pancreatitis-associated Protein Forms Fibrillar Aggregates with A Native-like Conformation

Meng-Ru Ho1,2, Yuan-Chao Lou1, Wen-Chang Lin1, Ping-Chiang Lyu2, Wei-Ning Huang3, and Chinpan Chen1*

From the 1Institute of Biomedical Sciences, Academia Sinica, Taipei 115, Taiwan, ROC; 2Institute of Bioinformatics and Structural Biology, College of Life Sciences, National Tsing Hua University, Hsinchu 300, Taiwan, ROC; 3Department of Biotechnology, Yuanpei Institute of Science and Technology, Hsinchu 300, Taiwan, ROC.


Human pancreatitis-associated protein (hPAP) was identified in pathognomonic lesions of Alzheimer¡¦s disease, a disease characterized by the presence of filamentous protein aggregates. Here, we showed that, at physiological pH, hPAP forms non-Congo red-binding, proteinase K-resistant fibrillar aggregates with diameters from 6 up to as large as 68 nm. Interestingly, CD and FTIR spectra showed that, unlike typical amyloid fibrils, which have a cross-£]-sheet structure, these aggregates have a very similar secondary structure to that of the native protein, which is composed of two a-helices and eight £]-strands. Surface structure analysis showed that the positively-charged and negatively-charged residues were clustered on opposite sides, and strong electrostatic interactions between molecules were therefore very likely, which was confirmed by cross-linking experiments. In addition, several hydrophobic residues were found to constitute a continuous hydrophobic surface. These results and protein aggregation prediction using the TANGO algorithm led us to synthesize peptide Thr84 to Ser116, which, very interestingly, was found to form amyloid-like fibrils with a cross-£] structure. Thus, it seems that hPAP fibrillization is initiated by protein aggregation primarily due to electrostatic interactions, followed by conformational rearrangement, especially of the exposed hydrophobic loop, which is converted into a £]-sheet structure, then the hPAP fibril with a native-like conformation grows by the stacking of this short hydrophobic loop on top of the cross-£] spine.




Effects of site-directed mutagenesis on the lyase activity of allophycocyanin alpha subunit

Hui-Fen Lin , I-Chen Hu, and Ping- Chiang Lyu

Department of Life Science, National Tsing Hua University, Hsinchu


Allophycocyanin (APC), a protein carrying tetrapyrrole phycocyanobilin (PCB), performs distinctive absorption and emission in visible light range. Recombinant APC alpha subunit (ApcA) showed to be a monomeric protein by analytic ultracentifugation and size-exclusion chromatography. ApcA was an autocatalytic bilin:protein lyase to facilitate the biosynthesis of functional chromophorylated ApcA (chromo-ApcA). In order to investigate the critical residues in the lyase active site of ApcA, fourteen residues around PCB were mutated by site-directed mutagenesis. C81A, C81S, R83A and D84A abolished lyase activity although there was no difference in secondary structure between mutants and wild-type ApcA. Lower fluorescence intensity of other mutants was observed in comparison with wild type despite these mutants still kept the ability of PCB attachment to protein. In summary, Cys81 is the most critical residue because PCB assembly to ApcA is through thioether bond. Complete loss of lyase activity in two mutants perhaps resulted from the interference of hydrogen bonding between PCB and Arg83 or Asp84. In addition, residues which are around the active site and with hydrophobic interaction with PCB affect the fluorescence intensity of chromophorylated proteins.




Post-Translationally Pyroglutamate Formation on Proteins: Protein production, crystal structures, catalysis mechanism, inhibition, and pathophysiology of human glutaminyl cyclase

Kai-Fa Huanga, b, c, Yi-Liang Liub, Wei-Ju Chengd, Tzu-Ping Koa, c, and Andrew H.-J. Wanga, b, c, d

aInstitute of Biological Chemistry, Academia Sinica, bInstitute of Biochemical Sciences, National Taiwan University, cNational Core Facility of High-Throughput Protein Crystallography, dDepartment of Pharmacology, School of Medicine, National Yang-Ming University, Taipei, Taiwan


N-terminal pyroglutamate (pGlu) formation from its glutaminyl (or glutamyl) precursor is required in the maturation of numerous hormones and bioactive peptides. The aberrant formation of pGlu may be related to several pathological processes, such as osteoporosis, amyloidotic diseases, and rheumatoid arthritis. This N-terminal cyclization reaction, once thought to proceed spontaneously, was found to be catalyzed by the enzyme glutaminyl cyclase (QC). In this poster, the expression and characterization of functional human QC are present. We demonstrated that the enzyme is abundant in the brain tissues of normal adults and some patients with Alzheimer¡¦s disease. This finding suggests that QC is responsible for the formation of N-terminal pGlu on several amyloid-related peptides, strengthening the possibility that human QC can be a drug target for treatment of several amyloidotic disorders. Furthermore, the first high-resolution crystal structures of human QC in free form and bound to a substrate and three imidazole-derived inhibitors are reported. The structure reveals an £\/£] scaffold akin to that of two-zinc exopeptidases but with several insertions and deletions, particularly in the active-site region. The relatively closed active site of human QC displays alternate conformations primarily due to the different indole orientations of Trp207, resulting in two substrate (glutamine t-butyl ester)-binding modes. The single zinc ion in the active site is coordinated to three conserved residues and one water molecule, which is replaced by an imidazole nitrogen upon binding of the inhibitors. Together with structural and kinetic analyses of several active-site-mutant enzymes, a catalysis mechanism of the formation of protein N-terminal pGlu is proposed. These results provide a structural basis for the rational design of potent inhibitors of human QC.




Structural and Functional Studies of a Putative Flagellar Regulatory Protein from Helicobacter pylori

Shao-Wen, Chou, Jia-Yin, Tsai and Yuh-Ju Sun

Institute of Bioinformatics and Structural Biology, National Tsing Hua University, Hsinchu 300, Taiwan, Republic of China


Helicobacter pylori require flagella-dependent motility to survive and multiply in this particular ecological niche. About 30 % protein sequences in the whole genome of H. pylori are annotated as ¡§hypothetical proteins¡¨, including HP0367, which does not associated with any other known function proteins based on its amino acid sequence. Recently, HP0367 was reported as a putative flagellar regulatory protein (HpFRP) according to the DNA microarray data. The crystal structure of HpFRP from the human gastric pathogen, H. pylori, has been studied by X-ray crystallography at 2.8 Å. The structure phase was determined by multiwavelength anomalous dispersion (MAD) using the Se-Met derivative. Crystallographic analysis revealed that the overall folding of HpFRP consists of coiled-coil structure. The structure is similar to chemotoxis, such as cheZ and other flagella-related protein thus we suggest that HpFRP may have a similar function as cheZ or other flagellar regulatory functions. Meanwhile the pull-down assay and the transmission electron microscopy experiments were used to investigate the flagellar function of H. pylori.




Application of Alkylhydroperoxide Reductase as a Protein Marker for the Diagnosis of Tissue Damages in Patients of Gastro-Duodenal Diseases Infected by Helicobacter pylori.

Ming-Hong Chuang,a Ming-Shiang Wu,b Wan-Lin Lo,a and Shyh-Horng Chioua

aInstitute of Biochemical Sciences, National Taiwan University, Taipei 106, Taiwan; and Institute of Biological Chemistry, Academia Sinica, Taipei 115, Taiwan
bDivision of Gastroenterology, Department of Internal Medicine, National Taiwan University Hospital, Taipei 100, Taiwan


Alkylhydroperoxide reductase (AhpC) is an abundant and important antioxidant protein present in Helicobacter pylori (Hp). Like another ubiquitous protein in mammals, peroxiredoxin (Prx), AhpC protein can switch from a peroxide reductase to a molecular chaperone after long-term oxidative stress. Native AhpC exists as a dimer of a single subunit and comprises a-helix and b-sheet domains and a low surface hydrophobicity as revealed by gel-filtration, circular dichroism (CD), and fluorescence spectroscopic studies. We have further shown that in the presence of thioredoxin (Trx) the protein could be induced to polymerize to high-molecular-weight (HMW) complexes with a size ranging from 700 kDa to 2000 kDa under oxidative stress. It is also found that AhpC from clinical strains isolated from patients of gastric cancer is more susceptible than that of the same protein isolated from patients of duodenal-ulcer to form HMW complexes with chaperone activity under oxidative stress. AhpC from gastric-cancer strains is found to overexpress inside H. pylori cells under short-term oxidative shock, whereas AhpC from duodenal-ulcer strains expresses and concentrates mainly on the outside membranes of bacterial cells under similar oxidative treatments. Based on the significant difference between AhpC isolated from strains of gastric cancer and duodenal ulcer, we have applied ELISA and Western blotting to study the cross-reactivity of AhpC to antisera of H. pylori-infected patients with different pathological outcomes. The immunoassay revealed that antisera from tissues of patients with more extensive inflammation like that in patients of gastric ulcer or cancer react strongly with the HMW complexes of AhpC, whereas those from less extensive inflammation like patients of gastritis react weakly with the LMW oligomeric form of AhpC. Therefore it is conceivable that the antioxidant protein AhpC of H. pylori may prove to be useful as a diagnostic protein marker to monitor extents of tissue damages from patients of gastro-duodenal diseases infected by H. pylori. .




Calorimetric measurement on binding of antimicrobial peptides to lipid vesicles

Yu-San Hunga , Fang-Yu Chena

aDepartment of Physics and Institute of Biophysics, Center of Complex Systems, National Central University, Chung-Li, Taiwan 32054


We employed the technique of isothermal titration calorimetry (ITC) to measure the reaction heat of antimicrobial peptides binding to charged lipid vesicles. At low peptide concentration, the binding was found to be an exothermic process in which the magnitude of heat release is decreased with increase of peptide concentration. Such concentration dependence was interpreted as due to two types of many-body effect. The first effect is the membrane thinning. The membrane bound peptides can stretch membrane surface, because of their embedding in lipid headgroup region. A tensed membrane thus costs more energy for a free peptide to bind. The second effect is the membrane electric charging. Antimicrobial peptides usually carry positive charges. When they are bound to membrane, they will create an electrostatic repulsion to against a free peptide to bind. Quantitative analysis on these two types of many-body effect was made. The result showed that the membrane thinning is a major effect. Furthermore, a correlation of membrane thinning to pore formation occurring in high peptide concentration was discussed.




Structural Genomics Studies of the Plant Pathogen Xanthomonas campestris pv campestris Using X-ray Crystallography and NMR Techniques

Ko-Hsin China, Zhi-Le Tua, Chao-Yu Yanga, Tso-Ning Lia, Shu-Ju Liaoa, Yi-Che Sua, Ying-Der Tsaia, Chien-Hong Lina, Kuo-Cheng Chiua, Wei-Ting Kuoa, Chao-Wei Huanga, Sz-Kai Ruana, Wen-Ting Loa, Tzu-Huei Kaoa, Yu-Sheng Fanga, Ping-Jiang Lyub, Peter Tsaic, Andrew H.J. Wangd, Shan-Ho Choua,*

a Institute of Biochemistry, National Chung-Hsing University, Taichung, 40227, Taiwan, ROC
b Dept. of Life Science, National Tsing-Hwa University, Shin-Chu, Taiwan, ROC
c Dept. of Life Science, National Yang-Ming University, Pei-Tou, Taiwan, ROC
d Institute of Biological Chemistry, Academia Sinica, Nankang, Taipei, Taiwan, ROC


The flood of sequence information available from the various genome projects coupled with the recent advances in molecular and structural biology has led to the concept of structural genomics on a genome-wide scale. Determining three-dimensional structures of proteins is crucial for understanding their biological functions. In this respect, structural genomics is expected to pave way for understanding the intricate interactions among proteins in a whole organism, and is emerging as a powerful approach of functional annotation. We have initiated a coordinated program to study the structural genomics of Xanthomonas campestris pv. campestris, a gram-negative bacterium that is phytopathogenic to cruciferous plants and causes worldwide agricultural loss. Although pathogenic, it, however, also produces exopolysaccharide (xanthan gum) that is of great industrial importance. Besides, this bacterium is unique in several aspects; 1) it is the prototype of secretion system of gram-negative bacteria; 2) its transcription is carried out in single gene, unlike those of poly gene approach widely adopted in E. coli and other bacteria; 3) it lacks cyclic AMP (cAMP) to transduce signals, but utilizes a Clp (cAMP receptor protein-like protein) system to help coordinate regulation. Thus, due to the academic importance, immense economic impact, and simplicity of its genetic constitution (lacking introns) of this organism, we have endeavored to identify and characterize the structures and functions of proteins encoded in Xcc using X-ray crystallography and high resolution NMR techniques.
Until now, approximately 500 target genes of this bacterium have been successfully constructed in expression vectors. Among them, about 50% gene products are soluble, and 200 proteins have been purified. By screening crystallization condition using a robot system, we obtained about 100 crystals, of which ~30% were found to be suitable for structure determination. So far, ten structures have been solved using this structural genomics approach, many of which provide clues to biochemical functions that can be experimentally confirmed. Several novel fold or subfamilies of well-known folds have been identified. The efficiency by which function can be deduced from structure will be further improved by integration with the bioinformatics tools and experiments.




The crystal structure of XC1739: a multiple antibiotic-resistance repressor (MarR) from Xanthomonas campestris at 1.8 Å

Zhi-Le Tua, Juo-Ning Lia, Ko-Hsin China, Chia-Cheng Choub,c, Andrew H.-J. Wangb,c, & Shan-Ho Choua,*

a Institute of Biochemistry, National Chung-Hsing University, Taichung, 40227, Taiwan, ROC
b Institute of Biological Chemistry, Academia Sinica, Nankang, Taipei, Taiwan, ROC
c Core Facility for Protein Crystallography, Academia Sinica, Nankang, Taipei, Taiwan, ROC


The emergence of bacterial resistance to multiple drugs poses a serious and growing health concern. Understanding and deciphering these multiple drug resistance mechanisms are important endeavors. The E. coli transcriptional regulator MarR is believed to be a key factor in regulating marRAB operon, which is responsible for the mar phenotype that is resistant to a wide variety of structurally different and medically important antibiotics.
The crystal structure of a salicylated-bound E. coli MarR was recently determined to a resolution of 2.3 Å. However, the free MarR crystals grown without salicylate were unstable and badly disordered to preclude structural determination. We have successfully crystallized a putative ligand-free MarR family protein (XC1739) from a plant pathogen Xanthomonas campestris and determined its structure to a resolution of 1.8 Å. It shares a 32 % identity (55.6 % similarity) with the E. coli MarR protein. A promoter sequence of 24 bases upstream of the XC1739 gene was also identified from the EMSA (Electrophoretic Mobility Shift Assay) and DNase I footprinting methods. These data, along with the result that the determined structure of XC1739 is analogous to that of the E. coli MarR, suggest that XC1739 is very likely a transcriptional regulator belonging to the MarR family.
Although the overall architecture of XC1739 is similar to that of E. coli MarR, the putative DNA-binding domain of XC1739 is significantly shifted relative to that of the liganded E. coli MarR. When the two MarR proteins were superimposed using the C£\ atoms of the left monomers, the right monomers exhibit significant shifting, especially for the £\4 helices. The only salt bridges detected between Asp67 and Arg73¡¦ and the reciprocal pair in the salicylate-binding E. coli MarR are found disrupted in the ligand-free XC1739 (larger than 9 Å for XC1739 Asp76 and Arg82¡¦). This may allow the DNA-binding lobes of ligand-free XC1739 to act independently for binding to XC1739 marO.
A structural homology search by the DALI program returns with two other protein structures of the MarR superfamily with good matching scores. The first one is a MexR responsible for the repression of MexAB-OprM operon that encodes a multidrug efflux system in P. aeruginosa, resulting in increased resistance to multiple antimicrobials. The second one is an SlyA protein that has been shown to up-regulate the expression of molecular chaperones, acid-resistance proteins, and cytolysin etc. in E. faecalis. Besides, the coordinates of three other MarR family proteins have also been deposited in PDB (1s3j, 1a61, and 1z7u) recently. They also adopt similar triangular architecture despite the very low sequence identities (~ 20%). Since the relative distances between their putative DNA-binding lobes vary a lot, the conformations of MarR family proteins are thus flexible and may adopt different binding modes with their cognate DNA promoters.




Crystal structure of XC6422 from Xanthomonas campestris at 1.6 Å resolution: a small a/b serine hydrolase

Chao-Yu Yanga, Ko-Hsin China, Chia-Cheng Choub,c, Andrew H.-J. Wangb,c & Shan-Ho Choua,*

a Institute of Biochemistry, National Chung-Hsing University, Taichung, 40227, Taiwan, ROC
b Institute of Biological Chemistry, Academia Sinica, Nankang, Taipei, Taiwan, ROC
c Core Facility for Protein Crystallography, Academia Sinica, Nankang, Taipei, Taiwan, ROC


XC6422 is a conserved hypothetical protein from Xanthomonas campestris pathovar campestris (Xcc), a Gram-negative, yellow-pigmented pathogenic bacterium that causes black rot, one of the major worldwide diseases of cruciferous crops. It is a protein consisting of 220 amino acids, which was selected as a target protein for this structural genomics project because no sequence identity higher than 20% could be found in PDB. In this poster we describe the crystal structure determination of XC6422 using the MAD method. The crystals diffracted to a good resolution of 1.6 Å, and the solved structure reveals that the protein belongs to the small serine a/b hydrolase superfamily that lacks a lid, leaving the active site fully accessible to solvent. It adopts the conserved Ser-His-Asp catalytic triad (Ser119-His195-Asp166) common to this superfamily, with Ser119 positioned at the very sharp turn between strand b5 and helix aC, the so-called nucleophile elbow. A structural homology search conducted using DALI method revealed more than 20 protein homologs with Z-scores larger than 15, including a variety of enzymes with esterase, lipase, and peptidase activities. This is consistent with the notion that a/b hydrolase fold family is one of the most versatile and widespread observed to date. Modeling studies with a serine esterase inhibitor E600 indicate that XC6422 adopts a preformed oxyanion hole conformation, and is poised for catalysis without the requirement of further structural rearrangement. From these detailed structural studies, we suggest that XC6422 is most likely an esterase, active on a soluble ester, or a lipase requiring no interfacial activation to act on a lipid or triacylglycerol substrate.
XC6422 formed high-quality crystals that grew to 1.5 mm in a few days and diffracted to a good resolution of 1.6 Å. Detailed subunit interaction studies between XC6422 monomers revealed that the extra strand preceding the first b strand of the canonical a/b hydrolase fold made considerable contribution to the crystal packing. The XC6422 crystal packs along the 21 axis and is mediated by extensive intermolecular hydrogen bonding interaction, including Ser2 HN ¡V Leu187 C=O, Leu5 HN ¡V Gln184 side chain C=O, Leu5 C=O ¡V Gln184 side chain amide, Ala12 C=O ¡V Gln184HN, Ala12 HN ¡V Glu182 C=O, and Thr8 HN ¡V Glu179 C=O. Altogether, excluding possible hydrophobic contributions, a single Gln184 residue contribute half (3) of the total hydrogen bonds to this inter-protein interaction energy. Gln184 is also the only amino acid residue that uses its side amide and carbonyl groups to interact with the backbone atoms of the neighboring subunit. This amino acid residue thus plays important roles in stabilizing the XC6422 crystals.




The crystal structure of a putative acyl-CoA thioesterase from Xanthomonas campestris (XC229) adopts a tetrameric hotdog fold of eg mode

Ko-Hsin China, Chia-Cheng Choub,c, Andrew H.-J. Wangb,c, & Shan-Ho Choua,*

a Institute of Biochemistry, National Chung-Hsing University, Taichung, 40227, Taiwan, ROC
b Institute of Biological Chemistry, Academia Sinica, Nankang, Taipei, Taiwan, ROC
c Core Facility for Protein Crystallography, Academia Sinica, Nankang, Taipei, Taiwan, ROC


Thioesterases are widespread in eukaryotes, bacteria, and archaea and are involved in the thioester hydrolysis in fatty acid metabolism to degradation of environmental pollutant 4-chlorobenzoate. These acyl-CoA binding enzymes are found to adopt a core ¡§hot dog¡¨ domain, comprising mainly a long central a-helix and an enclosing antiparallel b-sheet, in various oligomeric states. Recently, several crystal structures of such enzymes in tetrameric state were solved, including 4-hydroxybenzoyl-CoA thioesterases from Pseudomonas sp. Strain CBS-3 (PsHTE) and Arthrobacter sp. Strain SU (ArHTE), thioesterase II from Escherichia coli (EcTEII), and phenylacetate degradation protein PaaI from Thermus thermophilus HB8 (TtPaaaI). Although all these enzymes form quaternary structures of dimer of g dimers, with active site residues located in a deep crevice formed between the subunit interface, the orientation of the two g dimers in the tetramer can adopt opposite direction; they can either associate on the b-sheet side (so-called bg mode), as those observed in EcTEII, ArHTE, TtPaaaI, or associate on the helical side (so-called eg mode), as that observed in PsHTE. Interestingly, these enzymes involved in the degradation of environmental pollutants contain a unique nucleophilic Asp residue in the active site, contrary to those using active site Ser/Cys residues commonly observed in the thioesterase family of fatty acid metabolism.
XC229 (gi|21112185) from a soil-dwelling plant pathogen Xanthomonas campestris pv. Campestris strain 17 (Xcc) is classified as a conserved hypothetical protein from a BLAST database search (http://xcc.life.nthu.edu.tw/). It contains 134 amino acids, and is annotated as a putative thioesterase in the Pfam database. We have determined the crystal structure of XC229 to a resolution of 1.8 Å, and found it adopts a unique tetramer of eg mode with extensive subunit H-bonding/salt bridge interactions. A sequence alignment and model study with a 4-hydroxybenzoyl-CoA ligand suggest that XC229 very likely belongs to the novel type thioesterase, and is the second member belonging to such a thioesterase superfamily observed to date.




The Crystal Structure of XC847 from Xanthomonas campestris: a 3¡¦-5¡¦ Oligoribonuclease of DnaQ fold family with a Novel Opposingly-Shifted Helix

Ko-Hsin China, Chao-Yu Yanga, Chia-Cheng Choub,c, Andrew H.-J. Wangb,c, & Shan-Ho Choua,*

a Institute of Biochemistry, National Chung-Hsing University, Taichung, 40227, Taiwan, ROC
b Institute of Biological Chemistry, Academia Sinica, Nankang, Taipei, Taiwan, ROC
c Core Facility for Protein Crystallography, Academia Sinica, Nankang, Taipei, Taiwan, ROC


RNA metabolism has been found to play important roles in cell viability and dysregulation of mRNA life-time is implicated in several human diseases, including cancer, inflammation, and Alzheimer's disease. However, RNA metabolism is complicated and requires a wide variety of discrete ribonucleases (RNases) with both endo- and exo-cleaving activities to complete the turnover of aged RNA molecules. Six exo-ribonuclease superfamilies, as well as various subfamilies, were recently identified through extensive data mining. Oligoribonuclease (Orn) was found to belong to the DEDDh type exo-ribonuclease superfamily, and is exclusively responsible for degrading small oligoribonucleotides of 2-5 residues in length to mononucleotides. It is different from other known exo-ribonucleases of E. coli in that knockout of this gene (orn) can lead to cellular lethality.
XC847 from the plant pathogen Xanthomonas campestris pv. campestris str. 17 (Xcc) is classified as belonging to the oligoribonuclease family in the Pfam database. It contains 194 amino acids, and shares a high degree of identity and similarity with other identified oligoribonuclease throughout the DEDDh domains including motifs I, II, and III that are unique to oligoribonucleases. Until now, Orns are less studied structurally. Only the crystal structure of Haemophilus influenzae Orn has been deposited in the PDB (1J9A), but its structural details have not yet been reported. Due to the importance of this type of RNase, we have solved the Xanthomonas campestri Orn (XC847) structure to a resolution of 2.1 Å. Although the overall architecture of this Orn is similar to other reported 3¡¦-5¡¦ DNases, with the active DEDD residues also located in a similar environment, there is one significant difference between them; the helix H in XC847 is oriented opposingly from the similar helices in all reported 3¡¦-5¡¦ DNases discovered so far, possibly to prevent the steric hindrance of accommodating oligoribonucleotide substrates.
Based on extensive analysis, four well conserved motifs for Orn have been proposed. However, the functional significance of these motifs remain unknown. After detailed structural studies of XC847, it becomes clear that these reserved residues may play important roles in stabilizing the Orn dimer or in interacting with the U5 substrate, in addition to the classical role of the Asp12, Glu14, Asp103, Asp167, and His161 residues in forming an active center.




Crystal structure of XC5357 from Xanthomonas campestris: A putative tetracenomycin polyketide synthesis protein adopting a novel cupin subfamily structure

Ko-Hsin China, Chia-Cheng Choub,c, Andrew H.-J. Wangb,c, & Shan-Ho Choua,*

a Institute of Biochemistry, National Chung-Hsing University, Taichung, 40227, Taiwan, ROC
b Institute of Biological Chemistry, Academia Sinica, Nankang, Taipei, Taiwan, ROC
c Core Facility for Protein Crystallography, Academia Sinica, Nankang, Taipei, Taiwan, ROC


Members of the genus Xanthomonas represent a major group of phytopathogenic bacteria infecting most economically important crop plants. Extensive genome analyses and comparative genomics have been used to identify novel gene products of biological importance for these phytopathogens in recent years. Xanthomonas campestris pathovar campestris (Xcc) is a Gram-negative, yellow pigmented bacterium causing black rot of cruciferous crops. A local Xanthomonas campestris str. 17 from Taiwan has been chosen for structural genomics studies. Its genome has been sequenced and ORFs annotated using a bioinformatics approach. XC5357 is one of these gene products consisting of 113 amino acids with a MW of 12.2 Da. A BLAST search against UniProt database has recognized significant sequence identity to the tetracenomycin polyketide synthesis protein from X. campestris ATTC33913 (Q4URX6), X. axonopodis (Q8PN81), and H. marismortui (Q5VOH1, tcmJ), with percentage identities of 99.1%, 82%, and 51%, respectively. InterPro scan for sequence motif against PROSITE and Pfam databases returned with a cupin motif, and sequence matches to existing PDB structures identified several oxalate decarboxylases (1o4t, 1l3j, 1j58, and 1uw8), albeit with low sequence identities (ranging from 25-30%). Here we report the crystal structure of XC5357 from Xcc and use this information to deduce its possible biological function. The determined structure was indeed found to belong to the cupin superfamily. However, it also possesses distinct structural feature and belongs to a novel cupin subfamily.
A ProFunc search has been performed to deduce the possible function of XC5357. No hit was obtained from any of the 189 enzyme active site templates, and neither from any of the 13374 ligand-binding templates. Cleft analysis did show a possible binding pocket of 1632 Å3 close to the metal ion binding site in the 1VJ2, 1O4T, and 1J58 structures. Since no metal ion is present in XC5357, enzymatic reactions requiring metal ions can be eliminated for XC5357. From this combined analyses, XC5357 is likely a tetracenomycin polyketide synthesis enzyme with a novel substrate binding cavity requiring no metal ion for activity.




Solution Structure ofFamily 21 Carbohydrate-binding Module from Rhizopus oryzae Glucoamylase

Yu-Nan Liu, Yen-Ting Lai, Wei-I Chou, Wei-Ting Liu, Margaret Dah-Thyr Chang and Ping-Chaing Lyu

From the Institute of Bioinformatics and Structural Biology & Department of Life Science, National Tsing Hua University, Taiwan, Republic of China
Address correspondence to: Ping Chaing Lyu, Department of LS, National Tsing Hua University, No. 101, Sec. 2, Kuang Fu Rd., Hsinchu, Taiwan 30013, Republic of China


Carbohydrate-binding modules (CBMs) function independently to assist carbohydrate active enzymes. Family 21 CBM is a family of modules with approximate 100 amino acids. Starch-binding functions of some members in this family were evidenced and glycogen binding activities of the other members remain putatively. We report here the first structure of family 21 CBM from Rhizopus oryzae glucoamylase (RoCBM21) determined by nuclear magnetic resonance spectroscopy. Its structure folds into a symmetric b-sandwich fold with an immunoglobulin-like structure. Ligand-binding properties of RoCBM21 were analyzed by chemical shift perturbation and AutoDOCK. Structural and ligand-binding comparisons to previous reported starch-binding CBMs were made. Although the amino acid sequence similarities are low (~20%), the three dimensional structure of RoCBM21 is resemble to structures of family 34 CBMs from Thermoactinomyces vulgris a-amylases (TvCBM34), and shows a ¡§similar-but-shift-one-strand¡¨ topology to the family 20 CBM from Aspergillus niger glucoamlyase (AnCBM20) and families 25-26 tandem CBMs from Bacillus halodurans maltohexaose-forming amylase (BhCBM25 and BhCBM26). Even if the global fold is analogous, some vital ligand-binding residues vary between RoCBM21 and TvCBM34s. Additionally, some residues critical to ligand-binding are conserved between RoCBM21 and AnCBM20, but chemical sift perturbation of ligand-binding shows certain diversity between their ligand-binding activities. The structure of RoCBM21 helps us characterize the structural features of starch-binding CBMs and understand the protein-carbohydrate recognitions therefore shed more lights on engineering of novel carbohydrate active enzymes.




Structural and functional variations in human apoE2, apoE3, and apoE4 72-166 fragments

Yi-Hui Hsieh, Chi-Yuan Chou, and Gu-Gang Chang

Faculty of Life Science and Institute of Genome Science, National Yang-Ming University, Taipei 112, Taiwan


There are three major apolipoprotein E (apoE) isoforms. Whereas APOE-£`3 is considered as a longevity gene, APOE-£`4 is a dual risk factor to atherosclerosis and Alzheimer disease. We have expressed N- and C-terminal truncated apoE2, apoE3 and apoE4 72-166 fragments tailored to eliminate helix- and domain interactions to unveil structural and functional disturbances. In aqueous system without DHPC, using CD spectropolarimetry analysis revealed that 39%, 31% and 53% a-helix for apoE272-166, apoE372-166 and apoE472-166, respectively. The a-helical content of both apoE272-166 and apoE372-166 fragments was increased in the presence of DHPC. ANS titration experiment also suggested that apoE472-166 maintained the lowest Kmand highest maximal ANS binding capacity with or without DHPC. In the presence of DHPC, the apoE472-166 showed more complicated or aggregated species than those of the corresponding apoE272-166 and apoE372-166 counterparts by analytical ultracentrifugation. As illustrated by the DMPC turbidity clearance assay, the apoE472-166 showed the higher lipid-binding ability than that of apoE372-166 and apoE272-166 proteins. The LDL receptor binding ability, determined by a competition binding assay of 3H-LDL to LDL-receptor of HepG2 cells, showed that apoE472-166 maintained greater receptor-binding abilities than their apoE272-166 and apoE372-166 counterparts. The b-VLDL binding ability of the three apoE72-166 proteins was also proven by lipoprotein electrophoresis. The structural preference of truncated apoE4 to remain functional in solution may explain the enhanced opportunity of apoE4 isoform to display its pathophysiologic functions. The apoE472-166 may be a plausible target for the design of a peptide drug to alleviate the aggregation propensity of apoE4.




Sited-Directed Mutagenesis Studies of Mung Bean Nonspecific Lipid Transfer Protein 1

Ya-Fen Yang, Ku-Feng Lin, Yen-Ting Lai and Ping-Chiang Lyu

Department of Life Science, National Tsing Hua University, Hsinchu, Taiwan


Plant nonspecific lipid transfer proteins (nsLTPs) are well known for their ability to bind and transfer lipids in vitro. However, the detailed transport mechanism is still unclear. In our previous study, mung bean (MB) nsLTP1 purified from seed has been identified and analyzed. We found that some hydrophobic residues are probably essential for lipid interaction. Therefore, we have constructed MB nsLTP1 and use site-directed mutagenesis to study its biochemical and biophysical properties. The residues around the lipid binding site and cysteine residues (Cys48, Cys48/Cys87, Cys3/Cys50, Cys13/Cys27 and Cys28/Cys73) were chosen to mutate. The results demonstrated that the influence of hydrophobic effect on protein stability and lipid transfer activity are significant. Particularly, the lipid transfer activity of Cys48Ser was increased enormously. The same phenomenon of lipid transfer activity was also observed in Cys48Ala/Cys87Ala. On the basis of our molecular dynamics simulation, these two mutants probably have higher opportunity to interact with lipid molecules by loosing the protein structure. All together, these observations revealed both the hydrophobic interaction and structural plasticity may play important roles in nsLTP1 biological function.




Identification of critical amino-acid residues in Vigna radiata plant defensin 1 involved in inhibiting Tenebrio molitor £\-amylase

Ping-Hsing, Tsai

Institute of Bioinformatics and structural biology


Vigna radiata plant defensin 1 (VrD1) is a cysteine-rich and basic small peptide of 46 amino acids. In previous study, VrD1 was reported to exhibit insecticidal activity, and three dimensional structure of VrD1 have been determined by nuclear magnetic resonance (NMR) spectroscopy. However, the insecticidal mechanism of VrD1 is still indistinct. Our preliminary data showed that VrD1, which was purified from mung bean, inhibited Tenebrio molitor £\-amylase. To elucidate the £\-amylase inhibition mechanism of VrD1, recombinant VrD1 was constructed, expressed and purified from Escherichia coli. According to amino acid sequence analysis and protein structure comparison, specific residues involved in £\-amylase inhibition were identified by site-directed mutagenesis. Twelve mutants were totally obtained and analyzed by circular dichroism (CD) for secondary structure and £\-amylase activity assay for the inhibition function. The CD spectra showed that all recombinant VrD1 proteins have similar secondary structures and thermal stabilities. The results of £\-amylase inhibition assay show that three mutants, K6A, R26E and R38A, significantly decrease in £\-amylase inhibition. These three residues play important roles in inhibitory function in VrD1.




Ligand Specificity and Structural Requirement of Tachypleus Plasma Lectins for Bacterial Trapping

Tun-Hsun Kuo, Shiao-Cheng Chuang, Sing-Yang Chang,¡± and Po-Huang Liang†¡±

¡±Institute of Biological Chemistry, Academia Sinica, Taipei 115, Taiwan
Institute of Biochemical Sciences, National Taiwan University, Taipei 106, Taiwan


TPL-1 and -2 (Tachypleus Plasma Lectin-1 and -2) from the hemolymph of Tachypleus tridentatus, were previously purified by using Sepharose and LPS affinity column, respectively [Chiou, S.-T., Chen, Y.-W., Chen, S.-C., Chao, C.-F. and Liu, T.-Y. (2000) J. Biol. Chem. 275, 1630-1634] and their encoding genes were cloned [Chen, S.-C., Yen, C.-H., Yeh, M.-S., Huang, C.-J. and Liu, T.-Y., (2001) J. Biol. Chem. 276, 9631-9639]. In the present study, TPL-1 and -2 were produced using yeast and their recombinant proteins secreted into media were purified and characterized. The proteins show specific PGN (peptidoglycan) and LPS (lipopolysaccharide) binding activity, suggesting their role of trapping gram-positive and gram-negative bacteria, respectively, in innate immunity. Using BIAcore, the dissociation constant of the complex of TPL-1 and PGN unit was measured to be 8 x 10-8 M. Substitution of Asn74, the N-glycosylation site of TPL-1, with Asp abolishes the PGN binding affinity, whereas non-glycosylated N3D mutant of TPL-2 retains LPS binding activity. DTT treatment to break disulfide linkages removes TPL-2 activity but does not interfere with TPL-1 function. Cys4 in TPL-2 may form an inter-molecular disulfide bond, which is essential for activity. As a result, the TPL-2 mutant C4S is inactive and eluted as a monomer on a non-reducing gel. C6S is active by forming a non-covalently linked dimer. A model describing TPL-2 binding with LPS is proposed. These two plasma lectins with different ligand specificity can be used for detection and discrimination of bacteria and removal of endotoxin.




A Novel Oligonucleotieds Binding Mode of Single-Stranded DNA Binding Protein from Helicobacter pylori

Chia-Hung Wang and Yuh-Ju Sun

Institute of Bioinformatics and Structural Biology, National Tsing Hua University, Hsinchu 300, Taiwan, Republic of China


Single-stranded DNA binding protein (SSB) plays an important role in DNA metabolism, such as DNA replication, repair, and recombination. SSB of Helicobacter pylori (HpSSB, 179 residues) is encoded by the ssb gene. The crystal structure of truncated HpSSB134 protein (residue 1-134) complexed with dT(pT)34 was determined at 3.1 Å resolution. The N-terminal domain (residue 1-115) contains an OB-fold (oligonucleotides binding fold), which is similar with other species like E. coli, to function as an ssDNA binding. However, the ssDNA binding mode of HpSSB134 exhibits considerable variability with comparison to that of E. coli. The ssDNA wraps on the OB-fold with mainly electrostatic and stacking interactions. Several basic residues, Arg10, Arg35, Arg36, and Lsy108, on the surface of HpSSB134 form a significant patch corresponding to the ssDNA binding. Furthermore, two aromatic residues, Phe50 and Trp84, interact with thymidine by stacking arrangement. The structure of residues 116-134 were unable to be determined because of its flexibility, and many evidences reveal that the C-terminus tail of SSB participates in the protein-protein interaction, which may trigger the activity of partner proteins in DNA metabolism.




Direct interaction between focal adhesion kinase (FAK) and beta4 integrin in tumorigenesis

Hui-Yuan Tseng*, Mossaad Abdel-Ghany#, Bendicht U. Pauli#, Tang-Long Shen*

*Department of Plant Pathology and Microbiology, National Taiwan University, Taipei, Taiwan,
#Cancer Biology Laboratories, Department of Molecular Medicine, Cornell University, Ithaca, New York 14853 U.S.A.


Integrin alpha6beta4 complex is a member of integrin family of adhesion receptors, Previous studies have shown that beta4 subunit involved in tumor progression. Focal adhesion kinase (FAK), a non-receptor tyrosine kinase, is a pivotal mediator in integrin-mediated signaling, which implicated its role in tumor development as well. In the present study, we have found a direct interaction between these two proteins. To further resolve this interaction, several deletion mutants of FAK were generated. In a co-immunoprecipitation study, an 11-amino-acids motif nearby the amino-terminus of kinase domain was defined as the binding motif with beta4 integrin. Furthermore, an alanine-scanning approach was taking to identify key residues responsible for binding to beta4 integrin within FAK. Our result shows, in the first time, that a physical interaction between FAK and integrin beta subunit, thereby suggesting that FAK takes place in tumorigenesis. Subsequently, beta4 integrin binding deficient mutant of FAK will be generated to facilitate the investigation of downstream signaling events and cellular functions mediated by the beta4 integrin-FAK dependent manner. These findings will shed a light on better strategies for cancer therapies.




Empirical Force-Fields for Biologically Active Divalent Metal Ions: Application to Metallothioneins

C. Satheesan Babu and Carmay Lim

Institute of Biomedical Sciences, Academia Sinica, Taipei 115, Taiwan R. O. C.


In a recent paper [1], we derived a set of force-field (FF) parameters for the interaction of biologically important divalent ions with water, which reproduce all available experimental hydration free energies, ion-water coordination numbers and the solution structure. These parameters were derived using a theoretical procedure coupling thermodynamic integration and solvation properties of end point ions Be2+ (the smallest divalent ion) and Ba2+ (the biggest divalent ion). Here, we extend the procedure to derive FF models for the binding of Cd2+ and Zn2+ to rat liver metallothionein (MT) as a first step to model protein-metal interactions in solution. Results from long molecular dynamics simulations of Cd2+ and Zn2+ bound MT using the new FF models will be presented.

Empirical Force Field For Biologically Active Divalent Metal Cations in Water. C. Satheesan Babu and Carmay Lim, J. Phys. Chem. B (2006) 110: 691-699.




Crystal Structure of A Bifunctional Deaminase and Reductase Involved in Riboflavin Biosynthesis

Shen-Chia Chena, b, Yuan-Chih Changa, b, Chao-Hsiung Lin a, c, Chun-Hung Lind, and Shwu-Huey Liawa, c, e

aStructural Biology Program, bInstitute of Biochemistry, and cFaculty of Life Science, National Yang-Ming University, Taipei 11221, Taiwan. dInstitute of Biological Chemistry, Academia Sinica, Taipei 11529, Taiwan. eDepartment of Medical Research and Education, Taipei Veterans General Hospital, Taipei 11217, Taiwan.


Bacterial RibG is an attractive candidate for development of antimicrobial drugs because of its involvement in the riboflavin biosynthesis. The crystal structure of Bacillus subtilis RibG at 2.41-Å resolution revealed a tetrameric ring-like structure with an extensive interface of ~2400 Å2 per monomer. The N-terminal deaminase and C-terminal reductase domains share high structural homology to cytosine deaminase and dihydrofolate reductase, respectively. A detailed comparison illustrates both structural conservation as well as divergence compared to a variety of nucleotide deaminases and reductases. A structure-based sequence alignment of the nucleotide deaminases reveals not only the unique signatures in each family member for gene annotation, but also putative substrate-interacting residues for RNA-editing deaminases. Together with the binding sites of the essential cofactors, zinc ion and NADPH, the structural comparison also assists the modeling of transition-state analogs into the active-site cavities. The active sites of the two enzyme domains are ~40 Å apart, with no structural evidence of a substrate transport channel. Instead, the present structure suggests that the two domains fold independently and are combined to create the high intrinsic stability through a tetramer formation.




Functional Roles of the Double Covalent Linkages of FAD in Glucooligosaccharide Oxidase from Acremonium strictum

Chun-Hsiang Huanga,b, Wen-Lin Laib, Chia-Lin Chena,b, Ying-Chieh Tsaib, and Shwu-Huey Liawa-c

aStructural Biology Program, bInstitute of Biochemistry and Molecular Biology, and cFaculty of Life Sciences and Institute of Genome Sciences, National Yang-Ming University, Taipei, Taiwan.


Glucooligosaccharide Oxidase (GOOX) from Acremonium strictum has been screened for potential application in oligosaccharide acid production and alternative carbohydrate detection, because it catalyzes the oxidation of a variety of carbohydrates to the corresponding lactones with high selectivity for cello- and maltooligosaccharides. The crystal structure of GOOX demonstrates the first known double attachment flavinylation, 6-S-cysteinyl, 8a-N1-histidyl FAD, which is cross-linked with Cys130 and His70, respectively. Single mutants at H70A or C130A retain the covalent FAD linkage, indicating that flavinylation at these two sites are independent. However, the double H70A/C130A replacement abolishes the FAD binding and eliminates the oxidation activity. Circular dichroism spectra of the H70A/C130A mutant overlay nearly with that of the wild-type enzyme, suggesting that the mutations and the FAD loss do not affect the global structure. The urea-induced denaturation experiment also shows that FAD is not involved in structural stabilization. In addition, kinetic measurements display that H70 and C130 mutants exhibit 2-order lower of magnitude in kcat than the wild-type enzyme, but have little effect on Km. Flavin absorption spectra reveal the maximum wavelengths of 385 nm shifted to 400 nm and 375 nm in the H70A and C130A mutants, respectively. Therefore, the double covalent linkages may modulate the redox potential of FAD in GOOX.




Kinetic and structural properties of triosephosphate isomerase from Helicobacter pylori

Chen-Hsi, Chu, Yi-Ju, Lai and Yuh-Ju Sun

Institute of Bioinformatics and Structural Biology, National Tsing Hua University, Hsinchu 300, Taiwan, Republic of China


The glycolysis enzyme triosephosphate isomerase (TIM) has been cloned from Helicobacter pylori. Biochemical characterizations of the recombinant expressed protein (HpTIM) revealed a high degree of similarity to other species TIMs and its enzymatic activity towards the substrate D-glyceraldehyde-3-phosphate was determined (Km= 3.46¡Ó0.23 mM, kcat=8.8¡Ñ104 min-1, and Vmax= 17.70¡Ó0.9 £gM min-1). The HpTIM crystal structure was determined by molecular replacement method at 2.3Å resolution. The overall structure is (£]/£\)£](£]/£\)6, resemble to the common TIM barrel folding with (£]/£\)8. From the conformation of loop 6, connecting £]F and £\8, and binding of phosphate ion, the HpTIM was recognized as the ¡§closed¡¨ state. The conserved salt bridge between Arg and Asp among other TIMs was absence in HpTIM, the corresponding residues were Lys183 and Ser211. However, HpTIM has a unique salt bridge between Lys183 and Asp213. For clarification of the significance of this salt bridge to TIM, the enzymatic activity of salt bridge mutants of HpTIM was determined. The results suggested that the conserved salt bridge might not be necessary for TIM enzymatic activity but contribute to its conformational stability.




Structure and functional studies of an archaeal deaminase MM3105 from Methanosarcina mazei

Ming-Fen Huanga,b, Sheng-Chia Chena,b, Mei-Chin Laid, Shwu-Huey Liawa-c

aStructural Biology Program, bInstitute of Biochemistry, cFaculty of Life Sciences and Institute of Genome Sciences, National Yang-Ming University, Taipei, Taiwan, and dDepartment of Life Sciences, National Chung-Hsing University, Taichung, Taiwan.


The cytidine deaminase (CDA) superfamily with the consensus zinc-binding signatures, H(C)XE and PCX2-9C, consists of the mononucleotide deaminases involved in nucleotide metabolism, and the RNA (DNA)-editing deaminases involved in gene diversity and in anti-virus defense. To search for the putative ancestral members, several diverse archaeal CDA members were identified including a Mathanosarcina-specific protein, such as MM3105 from Mathanosarcina mazei. The recombinant protein has been expressed in E. coli and a functional assay using spectrophotometry have screened several nucleotides but with no significant activity. In addition, the crystal structure has been solved at 1.85-Å resolution, and is also the first structure of an archaeal CDA member. Analytical ultracentrifugation experiments demonstrate that MM3105 exists as a dimer in solution as well as in the crystal. And the monomeric structure consists of a central five-stranded b-sheet (b1-b5) sandwiched by a four-helix bundle (aA¡¦, aA¡¨, aA and aE) on one side, and by three helices (aB, aC and aD) on the other side. Sequence analysis suggests that another member, APOBEC2 (apolipoprotein B mRNA-editing catalytic subunit 2) may contain a similar four-helix bundle. A structure-based evolutionary analysis of the CDA members has been constructed to illustrate the protein evolution of structural plasticity and functional versatility.




The HATH Domain of Human Hepatoma-derived Groth Factor can form a Domanin-swapped Dimer with muchHigher Affinity for Heparin

Wei-Tin Lee, Shih-Che Sue, Stanley C.C. Huang, Jiun-Guo Yu and Tai-huang Huang,

Institute of Biomedical Sciences, Academia Sinica, Taipei, Taiwan, R.O.C.


Hepatoma-derived growth factor (HDGF)-related proteins (HRPs) comprise a new protein family that has been implicated in nephrogenesis, tumorigenesis, vascular development, cell proliferation, and transcriptional activation. All HRPs share a conserved N-terminal HATH domain (homologous to the amino terminus of HDGF), but vary significantly in the C-terminal region. Here, we show that in solution the N- and C-termini of human HDGF (hHDGF) form two structurally independent domains. The 100 amino acid N-terminal HATH domain is well-structured while the 140 amino acid C-terminal domain is disordered. We determined the solution structure of the HATH domain by NMR. The core structure of the HATH domain is a five-stranded beta-barrel followed by two alpha-helices, similar to those of PWWP domains of known structures. Surface plasmon resonance results further showed that the HATH domain is primarily responsible for heparin binding. On the basis of the chemical shift perturbation induced by binding of heparin-derived hexasaccharide, we identified a prominent, highly positively charged region as the putative heparin-binding site. Sequence comparison and structure prediction suggest that all HRPs are likely to adapt a similar modular structure. Furthermore, we found that under physiological conditions both the HATH domain and hHDGF can form dimers. Surface plasmon resonance (SPR) studies reveal that HATH-dimer binds to heparin with two orders of magnitude higher affinity (Kd ~ 13 nM) than that of the HATH-monomer (Kd ~ 1.2 £gM). The monomer-dimer equilibrium properties and NMR structural data together suggest that the HATH-dimer is formed through a domain-swapping mechanism. A structure of the domain-swapped HATH-dimer consistent with all experimental observations was calculated. In this structure the two HATH-monomers swap their C segments to form a structure with the putative heparin binding sites of the two monomers line up to become a long positively charged stripe for enhancing heparin binding affinity. This is the first report of a molecule which employs domain swapping as a mechanism to enhance its heparin binding affinity.




Solution structure of Ferrous Iron Transporter Protein A (FeoA) of Klebsiella Pneumoniae

Kuo-Wei Hunga, Chun-Chia Chenga, Tsan-Hung Yua, Shi-Huan Wangb, Chi-Fon Changc, Shih-Feng Tsaid, and Tai-Huang Huanga

aInstitute of Biomedical Sciences, Academia Sinica, Taiwan
bFu-Jen Catholic University, Taiwan
cHF-NMR center, National Research Program for Genomic Medicine, Academia Sinica, Taiwan
dDivision of Molecular and Genomic Medicine, National Health Research Institute, Taiwan


Klebsiella Pneumoniae (KP) is an enteric gram-negative bacillus causing liver abscess and metastatic meningitis, the most common community-acquired bacterial diseases in Taiwan. Iron acquisition system is one of the most important pathogenic factors. In the present study for the first time we report the cloning, expression, and characterization of ferrous iron transporter protein (FeoA) of KP. The target protein has been expressed successfully as a his-tag protein in E. coli in high yield (12 mg/L). The FeoA protein extracted from inclusion bodies was purified by nickel affinity chromatography. The refolded protein is further rpurified to homogeneity using size-exclusion chromatography. Far-UV CD spectra and 1H-15N HSQC spectra of FeoA indicate that the target protein is highly structured at neutral pH conditions. The chemical shift perturbation experiments reveal that recombinant FeoA is in a biologically active conformation. Using a variety of triple resonance NMR experiments, complete assignment of 1H, 15N, and 13C resonances in FeoA has been accomplished (BioMagResBank accession number 7062). Based on consensus CSI estimation the protein was found to consist of 5£]-strands and 2£\-helixes. The three-dimentional structure of FeoA is calculated using distance-geometry followed by simulated annealing techniques with ARIA 1.2. The root-mean-square distribution for the backbone atoms in the structured regions of 20 best conformers is 0.46 Å (Proyein Data Bank accession number 2GCX). Based on NMR chemical shift perturbation the ferrous iron bonding site in FeoA of KP has been mapped and the ion specificity has been screened.




Crystal Structures of Bacillus subtilis RibG in Complex with Substrates

Yu-Hsin Lin a, b, Sheng-Chia Chen a, c, Shwu-Huey Liaw a, d

aStructural Biology Program, bInstitution of Biotechnology in Medicine, cInstitute of Biochemistry, and dFaculty of Life Sciences and Institution of Genome Sciences, National Yang-Ming University, Taipei, Taiwan


Bacterial RibG contains an N-terminal deaminase domain and a C-terminal reductase domain involved in the second and third steps in the riboflavin biosynthesis, and hence it is an attractive candidate for development of antimicrobial drug. Here we have solved the complex structure of the reductase domain of Bacillus subtilis RibG with its substrate, 5-amino-6-ribosylamino-2,4(1H,3H)-pyrimidinedione 5¡¦-phosphate, at 2.56-Å resolution. Analytic ultracentrifugation experiments demonstrated that BsRibG exists as a tetramer in solution as well as in crystal form, where the enzyme forms a tetrameric ring-like structure. The deaminase domain belongs to the cytidine deaminase superfamily. A structure based sequence alignment of a variety of nucleotide deaminases reveals not only the unique signatures in each family member for gene annotation but also putative substrate-interacting residues for RNA-editing deaminases. The strong structural conservation between the reductase domain and the pharmaceutically important dihydrofolate reductase suggests that these two reductases involved in the riboflavin and folate biosyntheses have evolved from a single ancestral gene. The complex structure exhibits that the O-2 atom of the pyrimidine ring interacts with Lys151N5 and the NH-3 and O-4 with Thr172 O£^1 Thr195 and Asp199 make close contacts with the two hydroxyl groups of the ribose. Arg176, Arg183, and Arg206 form salt bridges with the phosphate group. These detailed enzyme-substrate interactions may provide useful information for rational drug design.




Crystal structure of infectious bursal disease virus VP2 subviral particle at 2.6 Å resolution: implications in virion assembly and immunogenicity

Cheng-Chung Lee, Tzu-Ping Ko, CC Chou, M. Yoshimura, S. R. Doong, M. Y. Wang, Andrew H.-J. Wang

Structural Biology Program, Institute of Biochemistry and Molecular Biology, National Yang-Ming University, Taipei 11221,
Institute of Biological Chemistry and National Core Facility of High-Throughput Protein Crystallography, Academia Sinica, Taipei 11529, Taiwan


The structural protein VP2 of infectious bursal disease virus (IBDV) spontaneously forms a dodecahedral T=1 subviral particle (SVP), and is a primary immunogen of the virus. In this study, the structure of IBDV SVP was determined in a cubic crystal and refined to 2.6 Å resolution. It contains 20 independent VP2 subunits in a crystallographic asymmetric unit. Each subunit is folded mainly into a shell domain and a protrusion domain, both with the Swiss-roll topology, plus a small helical base domain. Three VP2 subunits constitute a tight trimer, which is the building block of IBDV (sub)viral particles. The structure revealed a calcium ion bound to three pairs of symmetry-related Asp31 and Asp174 to stabilize the VP2 trimer. Our results of treatment of SVP with EGTA, a Ca2+-chelating reagent, indicated that the metal ion may be important not only in maintaining highly stable quaternary structure but also in regulating the swelling and dissociation of the icosahedral particles. A Ca2+-dependent assembly pathway was thus proposed, which involves further interactions between the trimers. The 20 independent subunits showed conformational variations, with the surface loops of the protrusion domain being the most diverse. These loops are targets of the neutralizing antibodies. Several common interactions between the surface loops were clearly observed, suggesting a possible major conformation of the immunogenic epitopes.




Studies of Protein Domains in the Transacylase Component of Human Mitochondrial Branched-Chain Ketoacid Dehydrogenase

Chi-Fon Changa,b, Tsun-Ai Yu,b, Hui-Ting Chouc, David T. Chuangd, Tai-Huang Huangb,c

aGemomics Research Center, Academia Sinica, bHFNMRC, NRPGM, cInstitute of Biomedical Sciences Academia Sinica, Taipei, Taiwan, dDept of Biochemistry, U. of Texas Southwestern Medical Center, Dallas, Texas, U.S.A.


The transacylase (E2) subunit of the branched-chain ƒÑ-ketoacid dehydrogenase (BCKD) complex carries three independently folded domains which are linked together by flexible loops: The amino-terminal lipoyl-bearing domain (hbLBD, 1-84), the interim E1/E3-binding domain (hbSBD,111-152), and the carboxy-terminal inner-core domain. The hbLBD and hbSBD play central role in substrate channeling and substrate recognition. We have employed multidimensional heteronuclear NMR techniques to determine the structure and dynamics of three truncated fragments of the human BCKD complex: the hbLDB (a.a. 1-84), hbSBD (a.a. 104-152) and a di-domain (hbDD) comprising residues 1 ¡V 168 of the E2 component. Analysis of backbone 15N-T1, 15N-T2 and 1H-15N-NOE relaxation data of hbDD showed the linker region is highly flexible. High resolution solution structures of hbLBD and hbSBD have been determined previously. In this poster, the domain orientation of hbLBD and hbSBD in E2 subunit, using residual dipolar coupling, will be discussed.




High-Field Nuclear Magnetic Resonance Center ( HFNMRC)

Chi-Fon Chang, Tai-Huang Huang

HFNMRC, NRPGM, Institute of Biomedical Science, Academia Sinica, Taipei, Taiwan


The High-Field NMR Center funded by National Research Program for Genomic Medicine (NRPGM). The Center operates and maintains five spectrometers at 500 MHz (1), 600 MHz (3) and 800 MHz (1). Three of these spectrometers are also equipped with cryoprobes. All of the spectrometers are in full operation and open to service for several years and were used in full capacities. Our objectives are: (i) Provide a readily accessible, state-of-the-art NMR facility to support NRPGM projects; (ii) Provide competent technical support to assist researchers carrying out advanced NMR research in structural genomics; and (iii) Develop new techniques and methodologies for structural Genomics related NMR research.

In addition to general spectrometer service, facility members also assist users on data acquisition, experimental set up, technical consultation, and technical development. Facility members also collaborate with several users in research projects. The facility homepage has been well established and maintained as well (http://www.nmr.sinica.edu.tw ). Through the webpage, users can reserve the machine time, review spectrometer technical information, get easy to follow instructions on usage of NMR related software, download training courses materials, and so on. Outside users can submit the sample by mail and download their data through internet ftp directly.
The HF-NMRC has continued to provide open, efficient and advanced technical service to the NMR community in Taiwan. Please contact us for details.




Molecular Dynamics Simulations of Plant Non-specific Lipid Transfer Protein

Yen-Ting Lai

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


Plant nonspecific lipid transfer proteins (nsLTPs) are small and positively-charged proteins that can transfer various lipid molecules in vitro. Molecular dynamics (MD) simulations were performed in this study in order to characterize different dynamical properties of nsLTP in free or complex form. Large-amplitude conformational changes have been observed in the four-helix bundle. Unlike the four-helix bundle, the C-terminal loop revealed random conformational change. Essential dynamics analysis was performed to isolate principal movement. Principal movement of the four-helix bundle changes upon lipid binding. Importance of one of the four disulfide bonds was also evaluated via MD simulations. Different dynamical behaviors of the C-terminal loop have been observed when the disulfide bond was modified. The result showed that the effect of the disulfide bond on the structure and the lipid transfer ability can be replaced by an intensive hydrogen bonding network. This simulation work provides a possible explanation of the lipid binding process.




CBS GIAO-NMR predicted protonation states in metalloproteins

Teobald Kupka1 and Carmay Lim1,2.

1 - IBMS, Academia Sinica, Taipei 115, Taiwan, and Department of Chemistry, National Tsing Hua University, Hsinchu, 300, Taiwan


Protonation states of proteins, small organic ligands, and water in the first coordination sphere of the central metal ion is essential for understanding the structure and mechanism of enzyme action. Experimental techniques used for determination of protein protonation states include X-ray, neutron diffraction and NMR. Independent theoretical calculations on model systems offer a valuable and complementary way of studying enzyme active sites. In this work we systematically investigated the imidazole (Im) nitrogen¡¦s isotropic shielding and shielding anisotropy changes as a result of its different protonation states in isolation, and in ZnIm, H2OZnImx and OHZnImx complexes (x = 1 or 3) using an accurate CBS GIAO-NMR [1] approach with Jensen¡¦s polarization-consistent basis sets and a ¡§locally dense¡¨ basis set model. The predicted NMR parameters, often observed in solid-state NMR studies of proteins, are extremely sensitive to nitrogen hybridization, presence of directly attached or hydrogen-bonded hydrogen atom, lone electron pair or metal ion, and show significantly more pronounced changes then the interatomic distances revealed by diffraction techniques. This sensitivity allows one to unambiguously assign protonation states in proteins by comparing theoretical and experimental results.

1. Kupka T., Ruscic B., Botto R. E., Toward Hartree-Fock- and Density Functional Complete Basis-Set-predicted NMR parameters, J. Phys. Chem. A, 2002, 106, 10396-10407




Investigating the Requirement of Structural Changes for Gating and Triggering the Ligand Release in Antibiotic Chromoprotein, Neocarzinostatin

Parameswaran Hariharana,b, Shan-Ho Choub, Der-Hang China*

aDepartment of Chemistry, bGraduate Institute of Biochemistry, National Chung Hsing University, Taichung, Taiwan, R.O.C.


Antitumor antibiotic chromoproteins such as enediyne containing neocarzinostatin involve a stable carrier-protein that steers the release of the tightly bound labile toxin. Contrary to the prevalent concept of ligand release, we established that the toxin (ligand) release requires no major backbone conformational changes, and subtle changes in the side chain of a specific amino acid residue is adequate to gate the release.

The wild-type aponeocarzinostatin and its variants mutated around the opening of chromophore binding cleft are employed to demonstrate that it is possible to modulate the temporal velocity of chromophore release by means of subtle changes on the aponeocarzinostatin protein. Kinetic studies of chromophore release monitored by time-course fluorescence and quantitative HPLC analyses show that the ligand release rate is significantly enhanced only in F78 mutants, while the mutants possess similar structural characteristics like wild-type as revealed by circular-dichroism and NMR experiments.

In order to speculate the validity of the above phenomena in the target cells, a novel attempt is employed to demonstrate a biologically relevant triggering mechanism of chromophore release from the chromoprotein complex using a preliminary model system of phospholipids-containing unilammelar vesicles simulating cellular membranes. A fluorescence-based release kinetics study reveals a lipid concentration dependent increase in antibiotic release rate, which correlates with quantitative HPLC data. The structural evaluations on neocarzinostatin in the presence of liposome using circular dichroism clearly depict the phenomenon as a protein conformational changes-independent fashion. The release of the enediyne chromophore is prompted by a hydrophobic-sensing mechanism operating in vicinity of cell membranes by involving aforementioned subtle structural changes in its carrier protein component. After its release the cellular lipid bilayers act as intracellular carriers of the enediyne chromophore. By showing the potential of the lipid-bilayer to replace the aponeocarzinostatin to function as a protector and carrier of the chromophore, the present data speculate an interesting chromophore transport/release model.




Stability of Aponeocarzinostatin Measured by Hydrogen Deuterium Exchange Studies

Shanmuganathan Aranganathana,b,Thallapuranam Krishnaswamy Suresh Kumar c,Chin Yub* and Der-Hang China*

aDepartment of Chemistry, National Chung Hsing University, Taiwan, ROC,
bDepartment of Chemistry, National Tsing Hua University, Taiwan, ROC,
cDepartment of Chemistry and Biochemistry, University of Arkansas, USA


Neocarzinostatin is a potent anticancer antibiotic isolated from the culture filtrate of Streptomyces carzinostaticus. It is a chromoprotein complex composed of an enediyne chromophore and a 113 amino acid apoprotein with a molecular weight 11kDa, Since Neocarzinostatin protein dynamics and stability in solution are fundamental properties to its functions as a protein carrier for the chromophore, these aspects were investigated by NMR hydrogen-deuterium exchange studies. The H/D exchange rate for amino acid residues in the aponeocarzinostatin, and the protection factors, which reflect the protection of residues against the exchange, were calculated. Based on these parameters, ÄGex, which provides site specific contribution of residues to the protein stability, was calculated. The highly protected residues, which are mostly buried from the solvent-accessible surface, constitute a slow-exchange core. Most likely, these exchanges belong to a cooperative unit by global unfolding. Thus, the rigid slow-exchange core probably corresponds to the folding core of the protein. Understandably, the residues with intermediate protection factors are situated further outward from the slow-exchange core, while the residues with smallest protection factors are situated peripherally or at the end of £]-strands. The £]-strands III and XI, which take part in the formation of the binding pocket, show highest protection factors. In addition, the £]-strands II and IX also show considerably higher protection factors. These data suggests the importance of these strands for the structural integrity of the binding pocket and hence for the stability of the apoprotein-bound chromophore. The loops between the strands VII & VIII and IX & X in the opening of the binding pocket show small protection factors, suggesting their flexibility and their probable role in the association and dissociation of the chromophore. Based on the present ÄGex calculations for the individual residues, further decisive experiments on neocarzinostatin protein dynamics and stability can be carried out by using apoprotein mutants of crucial amino acid residues.




Structural and Functional Studies of the Intercellular Adhesin Locus Regulator (IcaR) Protein of Staphylococcus epidermidis

Wen-Yih Jeng and Andrew H. -J. Wang

Institute of Biological Chemistry, Academia Sinica, Taipei, Taiwan


Biofilm formation has been considered an important factor that results in the virulence of pathogenic bacteria causing chronic infections during the past decade. Bacteria in biofilm persist in the body by a strategy that might be characterised as tenacious survival as opposed to aggressive virulence. The process of biofilm formation involves many proteins; accordingly, these proteins would be potential targets for the drug design to block the biofilm formation and resolve biofilm infection. Staphylococcus epidermidis, Gram- positive bacteria, is a commensal microorganism of human skin and the most frequent cause of hospital-acquired infections. S. epidermidis produces an extracellular polysaccharide named polysaccharide intercellular adhesion (PIA). PIA is an essential factor for staphylococcal biofilm formation and causes haem-agglutination and bacterial aggregation. The genes encoding PIA production are organized in the ica (intercellular adhesion) operon. IcaR is a transcriptional repressor involved in environmental regulation of ica operon expression and biofilm formation in S. epidermidis.

The structure of IcaR was determined at 1.33 Å resolution and solved by multiple wavelength anomalous dispersion (MAD) method. IcaR is entirely helical and comprises ten helices. The first three helices form a three-helix bundle DNA binding domain, which contains a helix-turn-helix (HTH) motif. IcaR forms homodimer through helices 4 to 10 and binds DNA cooperatively as a pairs of dimers. The crystal structures of IcaR and its complex with double strand DNA will provide the structural insight into cooperative DNA interactions and gene regulation of biofilm formatiom.




Novel mass spectrometry strategies for facile mapping of protein S-nitrosylation sites in relation to identification of controlled cysteine oxidation

Kuan-Ting Pan1; Yi-Yun Chen1; Shu-Yu Lin1; Tzu-Ching Meng2; Kay-Hooi Khoo2

1Core Facilities for Proteomics Research, Academia Sinica, Taipei, Taiwan; 2Institute of Biological Chemistry, Academia Sinica, Taipei, Taiwan


The facile but incomplete loss of the SNO moiety under electrospray conditions without collisional activation readily affords pair of signals related by an m/z value defined by the labile functional groups. Thus the nitrosylated and denitrosylated parent ions can be recognized and automatically selected for MS/MS sequencing during an online LC run if their mass difference was pre-programmed as the criteria for information dependent acquisition. Application to defining the S-nitrosylation sites of phosphatase is reported herein in conjunction with efforts in delineating its catalytic site cysteine modification. As an example, such analysis of PTP1B led to identification of 3 S-nitrosylation sites, including the active site cysteine and two other cysteine residues which were exposed on the surface. Other deeply buried cysteine, as inferred from crystal structure, were also detected and shown not to be modified under the same conditions. These site-specific S-nitrosylations which occurs more readily and to more completeness than oxidation of cysteine to sulfenic, sulfinic and sulfonic acids adversely affects its susceptibility to reversible and irreversible oxidation as mediated by reactive oxygen species. The mass difference method proves to be powerful in determining S-nitrosylation as the alternative MALDI methods are known to be only indirect since the SNO moiety would be lost and an S-nitrosylated peptide cannot be detected. Online mass difference-triggered MS/MS affords not only selectivity in direct sequencing of modified peptide but also allows better resolution of complex peptide mixtures through LC separation, as required in proteomics applications. Additionally, we demonstrate that the method can be extended to other post-translational modifications especially phosphorylation where an incomplete neutral loss of the phosphate moiety can be readily induced with slight elevation of collision energy. Natural heterogeneity can also be utilized but often impeded by chromatography separation of the related pairs into different time scans.




Role of disulfide bonds on the protein-directed chemical pathway of neocarzinostatin chromophore

Hung-Wen Chi, Chin-Jui Tseng, Der-Hang Chin*

Department of Chemistry, National Chung Hsing University, Taichung, Taiwan, Republic of China


The protein component of the antitumor chromoprotein neocarzinostatin is an all £]-protein with a well-defined hydrophobic cleft for accommodating a non-protein chromophore with a highly reactive dienediyne core. The apoprotein possesses two disulfide bonds which are important for the function of the drug. By a thiol-nucleophilic attack, the dienediyne core of the released chromophore is cycloaromatized into a diadical species which can abstract hydrogens from DNA to form product 1. However, the apoprotein-bound chromophore undergoes another distinct cycloaromatization pathway to form the product 2. How the protein directs the chemical pathway of the chromophore is an interesting question to be addressed.

Earlier, we screened more than fifty thiols for their ability to cycloaromatize the apoprotein-bound chromophore. The results showed that the thiols induce 2 as the major product, along with 1 as the minor one. We demonstrated that 1 was produced under normal protein-bound situation and was not suggested that disulfide bonds, which maintain the steric rigidity of the binding site, may play an influence of disulfide bonds, we desined experiments with apoproteins variants with their cysteine residue substituted by serine residues. The reconstituted chromoprotein from the mutant proteins were subjected to the reactions by the thiol ethyl 2-mercaptoacetate and the products were analyzed by HPLC. The HPLC profiles show the higher level of formation of product 1 from NCS variants mutated at the disulfide bond Cys37-Cys47, which takes part in the structural rigidity of the binding site, than from the wild-type protein. However, the ratio of the product 1 and 2 obtained from the variant mutated at the disulfide bond Cys88-Cys93, which is away from the binding site, is similar to that from wild-type protein. These results are suggestive of the Cys37-Cys47 disulfide bond¡¦s involvement in providing steric rigidity to the binding site and hence influencing the chromophore cycloaromatization pathways.




Engineering of a novel insect alpha-amylase inhibitor, VrD2 chimera, by structural-based rational design

Ku-Feng Lina, Tian-Ren Leea, Ming-Bin Hsub, Ching-San Chenb, and Ping-Chiang Lyua

a Institute of Bioinformatics and Structural Biology, National Tsing Hua University, HsinChu, Taiwan, Republic of China
b Institute of Plant and Microbial Biology, Academia Sinica, Taipei, Taiwan, Republic of China


Plant defensin was reported to have insecticidal activity against bruchid by its ability to inhibit alpha-amylase. In the absence of kinetics analysis and enzyme-inhibitor complexes determination, we have proposed recently the mode of alpha-amylase inhibition of plant defensin by structural analyses and computation docking. Here, we use another novel plant defensin, VrD2, to evaluate our proposed model. VrD2 is an anti-microbial protein of 47 amino acids containing eight cysteine residues that form four disulfide bridges. It shows a similar global fold to the reported Tenebrio molitor alpha-amylase (TMA) inhibitor, VrD1, but does not exhibit the ability to inhibit TMA. Several approaches have shown that small disulfide-rich structures are competent structural scaffolds for protein engineering studies. Thus, a VrD2 chimera was tentatively designed basis on the proposed model. Interestingly, the VrD2 chimera showed an obviously anti-alpha-amylase activity. On the other hand, another mutant which includes an alanine insertion does not recruit the alpha-amylase inhibitory activity. All of these data confirmed the proposed mode of alpha -amylase inhibition of plant defensin. The patterns identified from surface charge distribution among these plant defensins also suggest that the alpha-amylase inhibitory activity is correlated to electrostatic interaction. Besides, the work present here also demonstrates a possible approach to engineer novel alpha-amylase inhibitors.




Structural Characterization of Calmodulin in Complex with the Calcineurin Peptide

D. Irenea, Chia-Yen Liua, Chih-Ching Wangb, T.-H. Linb, and C.-L. Chyana

aDepartment of Chemistry, National Dong Hwa University, Hualien, Taiwan
b Institute of Biochemistry and Structural Biology Program, National Yang-Ming University, Shih-pai, Taipei 112, Taiwan


Calmodulin (CaM) is a ubiquitous protein which regulates a large number of key enzymes and controls a wide spectrum of important biological responses. Unraveling its diversity in activation mechanisms and target recognition has received extensive attention. Calcineurin (CaN), also known as protein phosphatase 2B (PP2B), is a Ca2+/CaM-dependent Ser/Thr protein phosphatase. CaN is critically involved in T cell activation. The immunosuppressive drugs, cyclosporin A (CsA) and FK506 form complexes with cyclophilin and FK506-binding protein (FKBP), respectively. The complexes inhibit CaN activity and prevent the transcription of genes response for T lymphocyte activation. The crystal structures of catalytic domain of CaN bound with FKBP12-FK506, and with CsA-cyclophilin have been solved, however, lacking structural information on the Ca2+/CaM dependent regulatory domain of CaN. In order to elucidate the regulatory role of CaM on CaN, and to gain more insight into the interactions between CaM and CaN from structural point of view, we have applied multidimensional heteronuclear NMR techniques to study the structure of CaM bound with the CaM-binding domain of CaN (CaNp). The full resonance assignments and structure of CaNp when associated with of CaM have been obtained. CaNp forms an £\-helical structure in association with CaM. Based on the analysis of the intermolecular NOEs, the predominant interactions included mostly hydrophobic forces between the N- and C-terminal portions of CaNp and the hydrophobic pockets of the N- and C-terminal globular domains of CaM, respectively. The binding orientation of the £\-helical CaNp with respect to the two domains of CaM is opposite to that of the peptides in the 1-14 and 1-10 binding mode, however, similar to the CaMKK peptide in the 1-16 binding mode.




Structural and Functional Studies of the Glyco-Enzymes from Escherichia coli and Klebsiella pneumoniae

Ying-Yin Chen(³¯¬Õ¿o)1, Li-ping Lo(ùÄRµÓ)2, Chun-Hung Lin(ªL«T§»)2, Andrew HJ Wang(¤ý´f¶v)2

1. Institute of Biochemical Sciences, National Taiwan University, Taipei, Taiwan
2. Institute of Biological Chemistry, Academia Sinica, Nankang, Taipei, Taiwan


We have obtained the first well-ordered single crystals of eubacterial gluconate-6-phosphate dehydrogenase(6-Pgdh), an enzyme of the oxidative branch of the pentose phosphate (PPP) pathway from E. coli. The E. coli 6-Pgdh has 95% identity with the Klebsiella pneumoniae sequence. The crystals are orthorhombic symmetry with unit cell dimensions a =68.52 Å b = 118.53 Å, c = 119.68 Å and belong to the space groups of P212121. X-ray diffraction to 1.56 Å resolution better than 2.4 Å has been recorded at the PF. Knowledge of the structure will assist in assessing the potential use of rationally designed compounds to inhibit this enzyme specifically.

The crystals of Gnd has been obtained in solution consisted of ammonium acetate, sodium citrate dihydrate and polyethene glycerol 4000. The crystals diffract to 1.56 Å resolution at the PF and belong to the space group P212121.

The resulting high resolution three-dimensional structures will provide valuable information to design potent enzyme inhibitors as potential new drug candidates.




Dual Functional Property of non-heme iron containing ferritin Pfr in Helicobacter pylori under acid stress

I-Ju Yeh and Shyh-Horng Chiou

Institute of Biochemical Sciences, National Taiwan University, Taipei 10617, Taiwan


Helicobacter pylori is a spiral, slow-growing Gram-negative microaerophilic (about 5% O2) bacterium. It is unique and distinctive among various bacterial pathogens for its ability to persist in the extreme acidic environment of the human stomach. To identify H. pylori genes in response to low pH, we have used a proteomic approach to study the protein expression of H. pylori under neutral (pH 7) and acidic (pH 5) environments. Global protein-expression profiles were analyzed by high-resolution two-dimensional gel electrophoresis (2-DE) followed by LC-MS/MS plus bioinformatics databases search/peptide sequence and mass comparison. The results revealed that more than ten proteins were differentially expressed under different pH conditions, with decreased amounts in seven proteins and increased in six proteins. Time-course RT-PCR study reveals that mRNA expression level of a non-heme iron containing ferritin Pfr (prokaryotic ferritins) is enhanced 1.2 folds whereas another electron-acceptor protein flavodoxin is decreased 2 folds at pH 5 as compared with that at pH 7. Moreover, non-heme iron containing ferritin Pfr was found to lose its ferrioxidase activity at pH 5, and it was found to bind DNA binding more strongly at pH 5 than at pH 7 as revealed by gel-retardation assay. Fluorescence spectra revealed the tertiary structure difference between pH 5 and pH 7. Therefore in addition to the predominant role of urease in the protection of H. pylori under low pH, non-heme iron containing ferritin Pfr may have a dual function for H.pylori to adapt to the acidic environment of gastric lumen.




Enzyme-substrate interactions revealed by the crystal structures of the archaeal Sulfolobus PTP-fold phosphatase and its phosphopeptide complexes

Hsing-Mao Chu1,2 and Andrew H.-J. Wang1,2,*

1. Institute of Biochemical Sciences, National Taiwan University, Taipei, 106 Taiwan,
2. Institute of Biological Chemistry, Academia Sinica, Taipei 115, Taiwan


The P-loop-containing protein phosphatases are important regulators in signal transduction. These enzymes have structural and functional similarity with a conserved sequence of Dx(25~41)HCxxGxxR(T/S) essential for catalysis. The singular protein tyrosine phosphatase (PTP) from archaeal Sulfolobus solfataricus is one of smallest known PTPs with extreme thermostability. Here, we report the crystal structure of this phosphatase and its complexes with two tyrosyl phosphopeptides A-(p)Y-R and N-K-(p)Y-G-N. The structure suggests the minimal structural motif of the PTP family, having two variable sequences inserted between the £]2-£]3 andƒn£]3-£]4 strands, respectively. The phosphate of both phosphopeptide substrates is bound to the P-loop using several hydrogen bonds. Comparison of several phosphatase_substrate complexes revealed that Gln135 on the Q-loop has different modes of recognition toward phosphopeptides. The substrate specificity of SsoPTP is primarily localized at the phosphotyrosine, suggesting that this phosphatase may be a prototypical PTP.




Distinct Early Folding and Aggregation Properties of Alzheimer Amyloid-ß Peptide Aß40 and Aß42: Stable Tetramer Formation by Aß42

Yun-Ru Chen (³¯Ãý¦p)*1,2 and Charles G. Glabe2

1. Genomics Research Center, Academia Sinica, Taiwan
2. Department of Molecular Biology and Biochemistry, University of California at Irvine, Irvine, California, 92697


The amyloidogenic protein, amyloid s peptide (Aß) composed of 40 or 42 amino acids is a critical component in the etiology of the neurodegenerative Alzheimer¡¦s disease. Aß is prone to aggregate and forms amyloid fibrils progressively both in vitro and in vivo. To understand the process of amyloidogenesis, it is pivotal to examine the initial stages of the folding process. We examined the equilibrium folding properties, assembly states and stabilities of the early folding stages of Aß40 and 42. We found that Aß40 and 42 have different conformations and assembly states upon refolding from their unfolded ensembles. Aß40 is predominantly an unstable and collapsed monomeric species, whereas, Aß42 populates a stable structured tetrameric species at concentrations above approximately 12.5 nM. Thermodynamic analysis showed that the free energy of Aß40 monomer and 42 tetramer are ~1.1 and ~22 kcal/mol, respectively. The early stages of Aß40ƒnand 42 contain different solvent-exposed hydrophobic surfaces, which are located at the flanking sequences of its protease resistant segment. The amyloidogenic folded structure of Aß is important for the formation of spherical ß oligomeric species. However, ß oligomers are not an obligatory intermediate in the process of fibril formation because oligomerization is inhibited at concentrations of urea that have no effect on fibril formation. The distinct initial folding properties of Aß40 and 42 may play an important role in the higher aggregation potential and pathological significance of Aß42. The work was supported by NIH NS31320 and a grant from Larry L. Hillblom foundation from Glabe laboratory and the Genomics Research Center, Academia Sinica, Taiwan.




Calorimetric measurement on binding of antimicrobial peptides to lipid vesicles

Yu-San Hunga ,Fang-Yu Chena

aDepartment of Physics and Institute of Biophysics, Center of Complex Systems, National Central University, Chung-Li, Taiwan 32054


We employed the technique of isothermal titration calorimetry (ITC) to measure the reaction heat of antimicrobial peptides binding to charged lipid vesicles. At low peptide concentration, the binding was found to be an exothermic process in which the magnitude of heat release is decreased with increase of peptide concentration. Such concentration dependence was interpreted as due to two types of many-body effect. The first effect is the membrane thinning. The membrane bound peptides can stretch membrane surface, because of their embedding in lipid headgroup region. A tensed membrane thus costs more energy for a free peptide to bind. The second effect is the membrane electric charging. Antimicrobial peptides usually carry positive charges. When they are bound to membrane, they will create an electrostatic repulsion to against a free peptide to bind. Quantitative analysis on these two types of many-body effect was made. The result showed that the membrane thinning is a major effect. Furthermore, a correlation of membrane thinning to pore formation occurring in high peptide concentration was discussed.




Imaging tissue engineering scaffolds and the monitoring of human bone marrow stem cell morphogenesis in PGA using two-photon fluorescence and second harmonic generation microscopy

1Yen Sun, 2Hsuan-Shu Lee, 1Shu-Wen Teng, 2Hsiao-Ching Chen, 1Wen Lo, 1Tze-Yu Lin, 2Ling-Ling Chiou, 3Ching-Chuan Jiang, 1Chen-Yuan Dong

1 Department of Physics, National Taiwan University, Taipei 106, Taiwan, R. O. C.
Department of 2Internal Medicine, 3Orthopedics, National Taiwan University Hospital and National Taiwan University, College of Medicine, Taipei 100, Taiwan, R. O. C.


In the first part of this work we combined two-photon autofluorescence and second harmonic generation ¡]SHG¡^ imaging to investigate the microstructure and non-linear optical properties of tissue engineering scaffolds. We focused on five different kinds of scaffold commonly used in tissue engineering, including OPLA (D,D-L,L polylactic acid), PGA (polyglycolide), collagen composite scaffold, collagraft bone graft matrix strip and nylon. Using a multi-photon microscope in combination with a motorized stage, we obtained spectrally resolved structural information over a large area or in three-dimensions. Our results show that the non-linear optical properties of the scaffolds will enable us to distinguish the different types of scaffold materials used for tissue engineering. We envision this imaging modality to be useful in understanding the interplay between cultured cells and scaffold materials under in vivo conditions. In the second part of this work, we apply the same technique to monitor the changes to the structure of human mesenchymal stem cells seeded in polyglycolic acid (PGA) scaffold under the induction of the chondrogenic transforming growth factor-£]3 in different time periods. Our results show that this approach may enable us to monitor the physiological processes associated with tissue engineering in real time.





Chia-Hsun Liua, S.-L. Chenc, Min-Feng Hsua, Steve S.-F. Yub, Sunny I. Chanb, Andrew H.-J. Wanga, W.-S. Fannc, and Rita P.-Y. Chena

a Institute of Biological Chemistry, Academia Sinica, Taipei, Taiwan
b Institute of Chemistry, Academia Sinica, Taipei, Taiwan
c Institute of Atomic and Molecular Sciences, Academia Sinica, Taipei, Taiwan


How do proteins fold? No one can completely answer the problem. In order to understand the folding process of proteins, we propose a photo-triggered caging-strategy. Compared with the traditional methods, this strategy is proceeded as soon as tens of nanosecond at room temperature without any denaturants involved. A small protein RD1 was chosen as our target protein. RD1 is an antifreeze protein which stops ice crystals from growing. It is small (7-kDa), compact, and consists of many secondary structures. In order to create the unfolded state of the protein, we need to synthesize cage and add it to the residue in the center of the protein. The pHP group of the photolabile 4-hydroxyphenacyl bromide (HPB) cage was recently found to be a new and fast-release trigger in aqueous solution. The side products are generally biologically inert and transparent at wavelengths longer than 300 nm. Taking advantage of the hydrophilic effect between HPB and water, HPB can escape from the interior cavity surrounded with several hydrophobic residues to unfold the protein. From the CD and NMR data, we confirmed that the caged protein is indeed unfolded. After irradiation, the cage is released from the protein and the refolding of the protein toward its native state is initiated. We are using photoacoustic calorimetry (PAC) and photothermal beam deflection (PBD) to monitor the whole process.