H. STRUCTURAL BIOLOGY CENTER
H-d. Biomolecular Structure Laboratory - Yasuo Shirakihara Group

RESEARCH ACTIVITIES

(1) Crystallographic Study of ATP synthase

Yasuo Shirakihara and Aya Shiratori

--ATP synthase is responsible for ATP production in living cells, and is a membrane protein located in the energy conversion membrane. ATP synthase consists of a channel Fo portion (about 100,000 dalton, subunit composition of ab₂c_8-12) and a large soluble catalytic F1 portion (380,000 dalton, α₃β₃γδε). The unique rotational catalysis mechanism of F1 includes rotation of the rod-like γ subunit which is thought to control the conformations of the three catalytic β-subunits in a cyclic manner by its rotation. Starting from elucidation of the α₃β₃ sub-assembly structure of the thermophilic F1, we have been moving up to the higher sub-assembly. The α₃β₃γ sub-assembly was difficult to crystallize, but the α₃β₃γε sub-assembly gave crystals that allowed to see a novel conformation of F1. We are now dealing with the holo-enzyme, ATP synthase.
--The membrane protein ATP synthase is still a challenging target for a structural study, in view of relatively few solved structures of the membrane proteins so far. After an initial trial for preparation and crystallization of the protein made last year, we have continued examination of both aspects extensively. Detergents such as decy-maltoside, undecyl-maltoside, dodecyl-maltoside, tridecyl-malltoside were good in preparation, but others like alkyl glucosides were not. Among various columns examined, useful ones were Q-sepharose high performance, Superdex prep grade 200 columns. An array of Q-sepharose high performance, Superdex prep grade 200 and the second Q-sepharose high performance columns yielded a preparation of crystallization grade, but a smaller array of Superdex prep grade 200 and Q-sepharose columns only was indistinguishable from the first array. Initial tiny crystals are now replaced by crystals with typical dimensions of 0.2mm, 0.2mm, 0.01mm after extensive and systematic crystallization condition search. Among the detergents mentioned above, only decy-maltoside and dodecyl-maltoside gave diffracting crystals, though all of the four detergents gave similar-looking crystals. ADP is the most favorite ligand, as others like AMPPNP, an ATP analogue, gave crystals that diffracted poorly. Other factors such as temperature in crystallization, kind and concentration of monovalent salts and those of divalent salts have been optimized. ATP synthase in our crystals contains all the 8 subunits, in contrast to yeast ATP synthase that lacks a and b subunits in crystals. Our crystals diffract to a resolution of about 7 A. Further refinement of the crystallization conditions to get better diffracting crystals is in progress.
--The preparation and crystallization study was done in collaboration with Satoshi MURAKAMI at Institute of scientific and industrial research, Osaka University, Toshiharu SUZUKI and Masasuke YOSHIDA, at Research Laboratory of Resource Utilization, Tokyo Institute of technology.

(2) Comprehensive Crystallographic Study of Transcription factors and Genome-partitioning Factors from E.coli

Yasuo Shirakihara and Aya Shiratori

--In E.coli, more than 160 transcription factors control transcription of their target gene(s) by binding to both their specific DNA sequence and the transcription apparatus. Sixty-five such transcription factors had been purified in Ishihama laboratory. In Niki laboratory, a number of novel proteins, that are judged to play roles in the genome and plasmid partitioning, have been prepared. Setting these transcription and genome partitioning factors as targets for a structural study, we are doing comprehensive structure determination in the Protein 3000 project (the sub-field of 'transcription and translation'). In the previous first two years, we examined 65 proteins for their crystals but identified only three proteins that gave crystals of diffraction quality. This year, we have concentrated to refine the crystals obtained so far.
--Mlc, an E.coli transcription factor, was found to diffract to about 3 A resolution. The pattern was recorded after reducing the aggregating tendency of the Mlc crystals. Phasing was hampered by either unavailability of appropriate heavy atom derivatives or inability of a Se-methionine substituted form of the protein to form crystals.
--YmcB, a protein believed to be involved in the genome and plasmid partitioning, was found to form extremely good crystals, diffracting beyond 2 A even with the laboratory beam source, after a sustained crystallization condition search. We did some expression experiments on YmcB, as the delivered preparation ran out and the subsequent expression in our hands turned out to be difficult. By changing expression conditions from the previous one (BL21(pLys), 30° incubation, IPTG induction) to new ones (BL21, 25° or 20° incubation, no induction), we got two preparations: from 25° incubation, a preparation of slightly shorter fragment of the molecule: from 20° incubation, a preparation of the intact molecule. The initial delivered preparation was a mix of the two, and the species responsible for crystals was found to be the fragment. Collection of diffraction data and analysis of the data is in progress.
--This work has been done in collaboration with Akira ISHIHAMA, Emi KANDA (Nippon Institute for Biological Science), Hironori NIKI, Rie INABA, Katsynori YATA, Yasushi OGATA (isotope center).

(3) Comprehensive structure-based functional studies on transcription factors

Hiroshi Itou and Yasuo Shirakihara

--Genomic-DNA information on number of organisms is now available. Using the information, comprehensive structure analysis of transcription factors for their structure-based functional understanding is in progress. Our research targets are transcription factors (including putative ones) from hyper-thermophilic archaeon Pyrococcus horikoshii, mesophilic bacteria Corynebacterium glutamicum and vertebrates Homo sapience. 162 genes were selected as target, and 146 of 162 were cloned, over-expressed (125 of 142) and purified (49 of 125). These purified proteins were tried crystallization for X-ray crystal structure analysis, and ten initial crystals were obtained. We had already succeeded in structure analysis of three of crystals, PH1161 and PH1932 from P.horikoshii and CGL2612 from C.glutamicum. PH1161 protein is a homologue of bacterial transcriptional activator TenA, and PH1932 and CGL2612 are homologue of transcriptional repressor protein ArsR and QacR, respectively. As a functional approach, target DNA sequences for these proteins were determined using SELEX (Systematic Evolution of Ligand by EXponential enrichment) method. This led to a tentative identification of the target genes for these transcription factors. This combination of tertiary structure analysis and binding DNA sequence determination by SELEX method is a powerful tool that provides us with key information on biological functions of transcription factors newly discovered by genomic analysis.
--This work has been done in collaboration with Ui OKADA and Isao TANAKA at Division of Biological Sciences, Graduate School of Science, Hokkaido University.

(4) Crystallographic Study of Transcription factors from Pseudomonas aeruginosa

Yasuo Shirakihara and Aya Shiratori

--Pseudomonas aeruginosa is the well-known opportunistic bacterial pathogen, and a number of transcription factors responsible for the pathogenicity have been identified. Among those, we are doing structural study of PtxR, PtxS, PhzR, and PA3547.
--Last year, we analyzed the MAD data from PhzR crystals and got an electron density map which showed clear helical densities in some part. However, inspired by a lesson from a failing expression experiment of a plant cell-death related protein SAG12 (molecular mass of about 30k) where SAG12 was not expressed but lactamase from a vector was instead expressed, C terminus amino acid sequence of PhzR (26K) expressed in our hands was examined. The sequence corresponded to that of C terminus residues of lactamase from Staphylococcus aureus, not to that of PhzR. There is a strong suspect that our PhzR crystals were not genuine. Diffraction data have been re-examined in this regard.
--This work has been done in collaboration with Hironori ARAMAKI (Daiichi Pharmaceutical College).

(5) Crystallization of Kid, a chromosome mover

Yasuo Shirakihara and Aya Shiratori

--Kid is involved in spindle formation and chromosome movements in mitosis/meiosis. Kid consists of three domains: an N-terminal kinesin-like motor domain (35-370), a C-terminal chromosome-binding domain (594-647) and a connecting domain (371-593) including a coiled-coil region. Our Previous trials to crystallize the N-terminal motor and the C-terminal chromosome-binding domains had failed, partly because both preparations contained a GST-tag that was attached for expression and easy preparation. We tried this year to crystallize a longer N-terminal fragment (1-462), and got small crystals from jeffamine or isopropanol. The crystals were formed only in the presence of MgADP. Further refinement of the crystallization conditions is in progress.
--We also tried to express the full-length Kid using our favorite E. coli expression system described last year. B834(DE3), JM109(DE3) and Rosetta produced the protein, though each in slightly different conditions. Expression was checked by a Western blot analysis. Efforts are being made to get a pure preparation in 10mg scale.
--The study was done in collaboration with Noriko TOKAI and Jun-ichiro INOUE at the Institute of medical science, the university of Tokyo.

PUBLICATIONS

Papers
1. Itou, H., Yao, M., Watanabe, N. and Tanaka, I. (2004). Structure analysis of PH1161 protein, a transcriptional activator TenA homologue from the hyperthermophilic archaeon Pyrococcus horikoshii. Acta Cryst. D60, 1094-1100.

POSTER PRESENTATIONS

1. Itou, H., Okada, U., Yao, M., Watanabe, N. and Tanaka, I. Comprehensive structure-based functional analysis of transcription factors. 1^st pacific-rim international conference on protein science, Yokohoma, 2004.
2. 白木原康雄、白鳥綾、村上聡、鈴木俊治、吉田賢右『ATP合成酵素の結晶化』、日本生物物理学会第42回年会、京都市、2004年12月
3. 岡田有意、伊藤啓、湯通堂紀子、木村誠、田中勲「X線結晶構造解析法とSELEX法の組み合わせによる、転写因子の機能同定に向けた研究」日本分子生物学会第27回大会、神戸市、2004年12月