JPWO2003048203A1 - Three-dimensional structure of DNA replication regulatory protein and use thereof - Google Patents

Abstract

The present invention provides a crystal of a complex of SeqA protein, which is a DNA replication regulatory protein, and hemimethylated DNA, a production method thereof, a three-dimensional structure of the complex by X-ray crystal structure analysis, and thus obtained Provided is a method for screening a compound that inhibits the formation of the complex using three-dimensional structure information. By using the method of the present invention, a drug that inhibits the growth of bacteria having a replication regulation mechanism by hemimethylation can be developed.

Description

（上記式中、Ａはシクロヘキセン、ジメチルシクロヘキセン、ビシクロ［２，２，１］ヘプタン、ビシクロ［２，２，１］ヘプテン、オキサ−ビシクロ［２，２，１］ヘプタンからなる群から選択される何れか１種の環を、Ｒ_１及びＲ_２は水素原子又はメチル基を、Ｒ_３は水素原子又は炭素数１〜４のアルキル基を表す）。
（１９）有効成分として、上記一般式（Ｉ）で示される化合物又はその塩を含むことを特徴とする抗菌性組成物。
である。
発明の実施の形態
本発明において、「ＤＮＡ複製調節タンパク質」とは、複製直後の染色体ＤＮＡの複製開始領域に結合して細胞周期が同調するまで新たな複製開始を抑止することによってＤＮＡ複製を調節するタンパク質をいう。大腸菌におけるこのような調節は、ｓｅｑＡ遺伝子の発現産物であるＳｅｑＡタンパク質により行われることが知られており、すでにその塩基配列やタンパク質の一次構造が明らかにされている（Ｌｕ，Ｍ．，ｅｔ ａｌ．，１９９４，Ｃｅｌｌ，７７，４１３−４２６，ＧｅｎＢａｎｋ Ａｃｃｅｓｓｉｏｎ Ｎｏ．Ｕ０７６５１）。ＳｅｑＡタンパク質は大腸菌をはじめ、Ｄａｍメチラーゼを有する種々の細菌に存在し、例えば、エシェリヒア属、ヘモフィルス属、パスツレラ属又はビブリオ属等の細菌であり、特に、ペスト菌、コレラ菌、赤痢菌、ネズミチフス菌、大腸菌Ｏ１５７等の病原性細菌を含む。更に、少なくとそれらの一部であって、ＤＮＡ複製調節機能を有するタンパク質も本発明における「ＤＮＡ複製調節タンパク質」の中に実質的に含まれる。
また、本発明において「ヘミメチル化ＤＮＡ」とは、ＧＡＴＣ配列を有する二本鎖ＤＮＡにおいて、一方の鎖のアデニンのみがＮ^６位にメチル化を受けたＤＮＡをいい、生体由来ものであっても化学合成したものであってもどちらでも良い。生体由来のものとしては、例えば、大腸菌の複製開始領域であるｏｒｉＣがある。２４５ｂｐからなるｏｒｉＣは、１１個のＧＡＴＣ配列が存在し、半保存的な複製直後には、新生鎖がＤａｍメチラーゼによりメチル化されるまで片方の鎖のみがメチル化されたヘミメチル化状態として存在する。一方、化学合成により任意の長さの二本鎖オリゴヌクレオチドを設計し、一方の鎖のＧＡＴＣ配列のアデニンのみがＮ^６メチル化されたオリゴヌクレオチドを合成し、これを相補鎖とアニールさせることによっても調製することができる。
本発明において、ＤＮＡ複製調節タンパク質の「変異体」とは、配列番号２に示したアミノ酸配列のアミノ酸が少なくとも１つ欠失、付加、挿入若しくは置換又は修飾されたアミノ酸配列を有し、ヘミメチル化ＤＮＡとの結合活性の少なくとも一部を保持する改変されたタンパク質をいう。
（ＳｅｑＡタンパク質の機能ドメインの探索）
図１（ａ）は、ＳｅｑＡタンパク質のＮ末端側及びＣ末端側から種々の長さのアミノ酸配列を欠失した２０種類の変異体ＳｅｑＡタンパク質を示す。これらをコードする変異体ｓｅｑＡ遺伝子を有するプラスミドを構築し、大腸菌に導入してＧＳＴ融合タンパク質を合成することにより、全長ＳｅｑＡタンパク質とＮ末端欠失変異ＳｅｑＡタンパク質を発現するプラスミド５種類において、それらのプラスミドのコピー数が極端に減少していることが分かった。Ｃ末端欠失変異ＳｅｑＡタンパク質を発現するプラスミドに関しては、Ｃ末端を５アミノ酸残基削っただけで、ｓｅｑＡ遺伝子が入っていないベクターのみのプラスミドと同等のコピー数に回復した。このことから、ＳｅｑＡタンパク質は大腸菌の染色体ｏｒｉＣだけでなくコリシンＥ１プラスミドの複製にも関与していることが示唆される。また、ＳｅｑＡタンパク質はＣ末端側にＤＮＡ結合ドメインを持つことも明らかになった（図１（ｂ））。
（ＤＮＡ複製調節タンパク質とヘミメチル化ＤＮＡとの複合体結晶の調製）
タンパク質の立体構造を明らかにする手法として、最も一般的に行われているのは、Ｘ線結晶構造解析の手法である。即ち、タンパク質を結晶化し、その結晶に単色化されたＸ線をあて、得られたＸ線の回折像を基に、該タンパク質の立体構造を明らかにしていくものである（Ｂｌｕｎｄｅｌｌ，Ｔ．Ｌ．及びＪｏｈｎｓｏｎ，Ｌ．Ｎ．，ＰＲＯＴＥＩＮ ＣＲＹＳＴＡＬＬＯＧＲＡＰＨＹ，ｐ．１−５６５，（１９７６）Ａｃａｄｅｍｉｃ Ｐｒｅｓｓ，Ｎｅｗ Ｙｏｒｋ）。
結晶化は、目的のタンパク質溶液に沈殿剤を添加し、または溶媒量を蒸発等によって減少させる等の操作によって、タンパク質が溶液状態から非溶液状態になる場合において、ある特定の条件において結晶として析出する性質を利用している。結晶化には高純度に精製されたタンパク質が必要である。また、結晶化する条件として、タンパク質濃度、塩濃度、水素イオン濃度（ｐＨ）、添加する沈殿剤の種類、温度等の物理的及び化学的な因子が関与している。更には、沈殿剤添加や溶媒量の調整方法として、バッチ法、透析法、蒸気拡散法などの結晶化の手法が数多く存在している（Ｂｌｕｎｄｅｌｌ，Ｔ．Ｌ．及びＪｏｈｎｓｏｎ，Ｌ．Ｎ．，ＰＲＯＴＥＩＮ ＣＲＹＳＴＡＬＬＯＧＲＡＰＨＹ，ｐ．５９−８２，（１９７６）Ａｃａｄｅｍｉｃ Ｐｒｅｓｓ，Ｎｅｗ Ｙｏｒｋ）。例えば、蒸気拡散法は、気相を通じて拡散によって揮発性の溶媒や沈殿剤が移動するようなしくみである。タンパク質溶液が沈殿剤溶液とのあいだで蒸気拡散することにより、タンパク質濃度と沈殿剤濃度が共に高まって過飽和の領域で結晶化が起こる。タンパク質溶液の液滴をどうつくるかによって、ハンギングドロップ法、シッティングドロップ法、サンドイッチドロップ法等の方法がある。このようにＸ線結晶解析に適したタンパク質の結晶を得るには、それぞれのタンパク質において、高純度のタンパク質を得ることと、結晶化における種々の因子や結晶化法を最適化する検討が必要になる。特に高純度のタンパク質の調製は結晶化を成功させるための重要な因子である。このために、クローン化した遺伝子を効率的な発現系で発現させることにより迅速かつ均一なタンパク質を調製することができる。
本発明のＤＮＡ複製調節タンパク質とヘミメチル化ＤＮＡとの複合体の結晶は、以下のようにして調製される。大腸菌ＳｅｑＡタンパク質をコードする塩基配列は既に公知であり、組換えＤＮＡ技術を用いて本来の宿主である大腸菌やその他の種々の宿主において容易に発現させることができる。
本発明の一つの実施形態において、組換え大腸菌を用いた系により大腸菌ＳｅｑＡタンパク質を大量発現させることができる。組換え大腸菌において遺伝子を大量に発現させるための種々のプラスミドベクター等が市販されている。例えば、強力なＴ７ファージプロモーターの下流に発現させたい遺伝子を導入し、宿主大腸菌により供給されるＴ７ＲＮＡポリメラーゼによって転写されるｐＥＴシステム（Ｎｏｖａｇｅｎ社）等を用いることができる。このシステムでは、Ｔ７ＲＮＡポリメラーゼ遺伝子は宿主大腸菌染色体に組み込まれており、ｌａｃＵＶ５プロモーター下流にあるためＩＰＴＧ添加による誘導で発現をコントロールすることができる。また、大腸菌ＳｅｑＡタンパク質のＮ末端側やＣ末端側に種々のタグ配列を導入しておくことによって、金属キレート樹脂（ヒスチジンタグ）や抗体（Ｔ７タグ、ＣＢＤタグ等）等を使って簡単に精製することができる。精製された融合タンパク質は、エンテロキナーゼやトロンビン等のプロテアーゼを用いて更に大腸菌ＳｅｑＡタンパク質部分のみを回収することが可能である。
また、大腸菌ＳｅｑＡタンパク質を発現する組換え大腸菌を、重金属原子であるセレンを含有したセレノメチオニンあるいはセレノシステインを含む培地で生育、培養することにより大腸菌ＳｅｑＡタンパク質中のメチオニンあるいはシステイン残基がセレノメチオニン又はセレノシステインに置換された変異体タンパク質が取得される。
更に結晶化に適するように、これらのタンパク質を高純度に精製する。精製法としては、カラムクロマトグラフィー（アフィニティー、疎水、イオン交換、ゲルろ過等）、塩析、遠心分離、電気泳動等、当業者においてタンパク質を精製する方法として一般的に行われている方法を単独で又は組み合わせて用いることができる。
次いで、ＤＮＡ複製調節タンパク質とヘミメチル化ＤＮＡとの複合体の結晶を調製する。ＳｅｑＡタンパク質のようなＤＮＡ結合タンパク質の機能を解析する場合には特定のオリゴヌクレオチドと複合体を形成したようなモデル化した系で結晶化することが必須である。例えば、大腸菌のＳｅｑＡタンパク質と１０〜５０塩基対程度のヘミメチル化二本鎖ＤＮＡとを１：１〜２：１のモル比であらかじめ混合しておくことが好ましい。
上記セレン含有培地で培養することによる重金属原子のタンパク質中への導入方法の他に、天然型のタンパク質結晶を、金、白金、イリジウム、オスミウム、水銀、鉛、ウラン、サマリウム等を含む金属塩又は有機金属化合物の溶液に浸漬することによっても重原子誘導体結晶を作製することができ、重原子同型置換法及び多波長異常分散法を適用する場合に利用される。
配列番号２に示したアミノ酸配列の７１位から１８１位までを有するＳｅｑＡタンパク質と１０塩基対のヘミメチル化ＤＮＡとの複合体の結晶は、六方晶系の空間群Ｐ６２２に属するもので、単位格子としてａ軸、ｂ軸の方向に１５２．２３８Å、ｃ軸の方向に１１９．３３８Åの大きさを有する。
（ＤＮＡ複製調節タンパク質とヘミメチル化ＤＮＡとの複合体結晶のＸ線結晶構造解析）
このようにして得たＤＮＡ複製調節タンパク質とヘミメチル化ＤＮＡとの複合体の結晶の構造を当業者に知られたＸ線結晶構造解析技術を用いて解析する。類縁タンパク質の立体構造が未知であり、その立体構造を利用した分子置換法による構造解析が不可能な本発明のような立体構造決定においては、重原子置換法又は多波長異常分散法（ＭＡＤ法）が適用され得る。本発明の一つの実施態様に於いて、Ｘ線結晶解析は、ＭＡＤ法であり、これに用いられるＸ線は、シンクロトロン放射光から得られる０．３〜３．０Åの波長である。ＭＡＤ法は、入射Ｘ線エネルギー近傍の吸収端をもつ重原子同型置換結晶から、複数の波長によるＸ線回折データを得る方法である。Ｘ線と該重原子の電子殻との共鳴でＸ線の散乱に差を生じ、これによってタンパク質の構造解析での位相問題を解決することができる。この方法の原理はすでに古くから知られていたが、波長可変であるシンクロトロン放射光の出現によって始めてタンパク質の構造決定に使用されるようになり、Ｈｅｎｄｒｉｃｋｓｏｎらがこの方法によって初めてタンパク質の構造解析に成功した（Ｈｅｎｄｒｉｃｋｓｏｎ Ｗ．Ａ．ｅｔ ａｌ．，Ｐｒｏｔｅｉｎｓ ４，７７−８８，１９８８及び、Ｈｅｎｄｒｉｃｋｓｏｎ Ｗ．Ａ．ｅｔ ａｌ．，ＥＭＢＯ Ｊ．５，１６６５−１６７２，１９９０参照）。
シンクロトロン放射光とは、蓄積リングに打ち込まれた電子を磁場によって加速して得られる放射光であり、通常のＸ線発生装置よりも極めて強いエネルギーを有する。理研播磨研究所の第三世代大型放射光設備ＳＰｒｉｎｇ−８（８ＧｅＶ）等が使用可能である。
測定した回折画像データは、プログラムＤＥＮＺＯ（マックサイエンス）又は同様な画像処理プログラムを用いて、反射強度データに処理される。天然型結晶の反射強度データ、複数波長での重原子置換結晶の反射強度データから、差パターソン図を利用して結晶中のタンパク質に結合した重原子の位置を求めた後、プログラムＰＨＡＳＥＳ若しくはＣＣＰ４又は同様の回折データ解析プログラムを用いて重原子位置パラメータを精密化することにより初期位相が決定される。決定された初期位相は、プログラムＤＭ（ＣＣＰ４パッケージ）又は同様な位相改良プログラム（電子密度改良プログラム）を用いた溶媒平滑化法に従って、信頼性の高い位相へと改良される。
ＤＮＡ複製調節タンパク質とヘミメチル化ＤＮＡとの複合体の立体構造モデルは、プログラムＯにより３次元グラフィックス上に表示した電子密度図から、以下の手順で構築し得る。まず、特徴的なアミノ酸配列を有する複数の領域を電子密度図上で探し出す。次に、見出した領域を起点にしてアミノ酸配列を参照しながら、電子密度に適合するアミノ酸残基の部分構造をプログラムＯを用いて三次元グラフィックス上で構築する。
順次この作業を繰り返すことにより、ＤＮＡ複製調節タンパク質とヘミメチル化ＤＮＡとの複合体のすべてのアミノ酸残基を相当する電子密度に適合させ、複合体全体の初期立体構造モデルを構築する。構築された立体構造モデルは、それを出発モデル構造として、構造精密化プログラムであるＸＰＬＯＲの精密化プロトコルに従って、立体構造を記述する三次元座標が精密化される。精密化の進行状況や分子モデルのチェックは、以下の指標により確認することができる。信頼度因子（Ｒ因子）は、回折強度測定で得られた結晶構造因子の絶対値と分子モデルから計算した結晶構造因子の絶対値の一致度を示す尺度であり、精密化が正常に進行すると小さな値となる。また、タンパク質の主鎖の二面角を計算しそれらをプロットしてそれらの値が許容される領域にあるかどうかを確認する方法（ラマチャンドランプロット）も用いられる。
このようにして本発明であるＤＮＡ複製調節タンパク質とヘミメチル化ＤＮＡとの複合体の３次元原子座標（各原子の空間的な位置関係を示す値）が初めて得られる。得られた原子座標を当業者にとって一般的に用いられている表記法であるプロテイン・データ・バンクのテキスト形式に従って表１に示す。

なお、表１に示した原子座標は、大腸菌のＳｅｑＡタンパク質（７１−１８１）と１０塩基対のヘミメチル化ＤＮＡ（ＳｅｑＡ ｂｓＣ−１０）との複合体の結晶構造であるが、この情報を用いることによって大腸菌と類縁の他の細菌が有するＳｅｑＡタンパク質の立体構造も容易に決定することができる。大腸菌と類縁の細菌としては、例えば、エシェリヒア属、ヘモフィルス属、パスツレラ属及びビブリオ属等が挙げられ、図４に示したように、Ｈａｅｍｏｐｈｉｌｕｓ ｉｎｆｕｅｎｚａｅ Ｒｄ、Ｐａｓｔｅｕｒｅｌｌａ ｍｕｌｔｏｃｉｄａ ＰＭ７０、及びＶｉｂｒｉｏ ｃｈｏｌｅｒａｅのＳｅｑＡタンパク質は、大腸菌ＳｅｑＡタンパク質とアミノ酸配列の相同性が非常に高い。このような相同性の高いタンパク質の立体構造を決定する方法として、分子置換法がある。本発明と同様な方法で上記類縁菌のＳｅｑＡタンパク質とヘミメチル化ＤＮＡとの複合体の結晶を作製し、Ｘ線回折データを取得した後、分子置換法により上記類縁タンパク質の立体構造を決定することができ、これらも本発明の一形態として含まれる。
（ＤＮＡ複製調節タンパク質とヘミメチル化ＤＮＡとの複合体の立体構造モデルを表示するコンピュータシステム）
本発明によるＤＮＡ複製調節タンパク質又は該タンパク質とヘミメチル化ＤＮＡとの複合体の結晶から得られる原子座標を、分子の３次元構造を表現するコンピュータ・プログラムが動作するコンピュータ又はそのコンピュータの記憶媒体に入力することで、ＤＮＡ複製調節タンパク質とヘミメチル化ＤＮＡとの複合体の３次元的な構造を詳細に表現することが可能になる。
図２はこれらのコンピュータシステムの一実施形態を表した概略図である。図２を参照すると、本発明のコンピュータシステムとして、記憶手段と、導出手段と、出力手段とが示されている。記憶手段としては、ＤＮＡ複製調節タンパク質のヘミメチル化ＤＮＡとの結合部位に対応するアミノ酸残基の原子座標を含むコンピュータ読み取り可能なデータを、導出手段によって読み出し可能なように記憶保持するものであれば特に限定されるものではない。例えば、フレキシブルディスク、ハードディスク、ＣＤ−ＲＯＭ、光ディスク、光磁気ディスク、磁気テープ等の記憶媒体に該データを記憶保持させることで実現される。
ＤＮＡ複製調節タンパク質のヘミメチル化ＤＮＡとの結合部位に対応するアミノ酸残基の原子座標を含むコンピュータ読み取り可能なデータを、記憶手段へ入力する方法は、種々の方法により行うことができる。例えば、フレキシブルディスクドライブやＣＤ−ＲＯＭドライブからこれらの記録媒体に記録されたデータを読み込むことができる。あるいはまた、ネットワークを介して例えばＰＤＢのようなデータベースから所望のデータを受け取ることもできる。
上記記憶手段から読み出したデータに基づいて、導出手段はＤＮＡ複製調節タンパク質とヘミメチル化ＤＮＡとの複合体の少なくとも一部を表す三次元立体構造モデルを導出する。導出手段の構成は、特に限定するものではないが、ＱＵＡＮＴＡ（Ｍｏｌｅｃｕｌａｒ Ｓｉｍｕｌａｔｉｏｎｓ社）、ＳＹＢＹＬ（Ｍｏｌｅｃｕｌａｒ Ｍｏｄｅｌｉｎｇ Ｓｏｆｔｗａｒｅ社）、ＡＭＢＥＲ（Ｗｅｉｎｅｒ，Ｓ．Ｊ．，ｅｔ ａｌ．，（１９８４）Ｊ．Ａｍ．Ｃｈｅｍ．Ｓｏｃ．，１０６，７６５−７８４）或は、ＣＨＡＲＭＭ（Ｂｒｏｏｋｓ，Ｂ．Ｒ．ｅｔ ａｌ．，（１９８３）Ｊ．Ｃｏｍｐ．Ｃｈｅｍ．４，１８７−２１７）等のプログラムを所望のコンピュータにインストールすることで実現される。更に、これらの分子モデルを生成するステップに続いて、該複合体の配座エネルギーを計算してエネルギーを最小化するプログラムとしては、上記ＣＨＡＲＭＭやＡＭＢＥＲ等が用いられる。
上記導出手段によって導出された結果を受けて、ＤＮＡ複製調節タンパク質のヘミメチル化ＤＮＡとの結合部位の三次元立体構造モデルを出力する出力手段は、従来より用いられている種々の装置を用いることができる．例えば、ＣＲＴと呼ばれるディスプレイ装置に、例えばＱＵＡＮＴＡのようなプログラムを用いて画像的に表示することもできる。このような画像を紙に表示するプリンターや他の記憶媒体に保存するためのディスクドライブも含まれる。
図３は、このようにして３．０Åの分解能で得られた大腸菌ＳｅｑＡタンパク質とヘミメチル化ＤＮＡとの複合体の立体構造をリボンモデルで示したものである。図３より明らかなように、得られた結晶は、ＳｅｑＡ（７１−１８１）タンパク質と１０塩基対のヘミメチル化ＤＮＡとの複合体が２回回転軸で対称な２量体の結晶である。立体構造から確認されたＳｅｑＡ（７１−１８１）タンパク質のドメイン構造を図４に示した。図４に示したように、ＳｅｑＡ（７１−１８１）タンパク質は７ヶ所のα−ヘリックス構造と２ヶ所のβ−シート構造を有しており、１１５位のグリシンから１１８位のアルギニンを含むループ領域、及び１４９位のスレオニンから１５３位のスレオニンを含むループ領域で、ヘミメチル化ＤＮＡのＧ（^ｍ６Ａ）ＴＣ配列部分と結合していることが分かる（図３参照）。
（立体構造の使用）
本発明が提供するＤＮＡ複製調節タンパク質とヘミメチル化ＤＮＡとの複合体の原子座標のすべて又は一部を、分子の立体構造を表現するコンピュータ・プログラムが動作するコンピュータ又はその記憶媒体に入力することで、ＤＮＡ複製調節タンパク質のヘミメチル化ＤＮＡとの結合部位に結合し、該複合体の形成を阻害する化合物、又はＤＮＡ複製調節タンパク質のヘミメチル化ＤＮＡとの結合部位と類似した立体構造を有し、ヘミメチル化ＤＮＡと強く結合する化合物を同定、検索、評価又は設計することが可能となる。化合物は、天然物化合物、合成化合物の何れでもよく、高分子化合物、低分子化合物の何れでも良い。
このような方法は、まず候補化合物とＤＮＡ複製調節タンパク質とをコンピュータモデルにより表示し、視覚的観察によりＤＮＡ複製調節タンパク質のヘミメチル化ＤＮＡとの結合部位に良く適合する形状又は化学構造をもつ候補化合物（リード化合物）を見つけ出す段階が含まれる。分子モデルを作成、表示するためには、前記ＱＵＡＮＴＡやＳＹＢＹＬ等のプログラムが利用できる。更に、ＤＮＡ複製調節タンパク質のヘミメチル化ＤＮＡとの結合部位との親和性、反発力、及び立体的障害等を推定するための種々のコンピュータプログラムが使用できる。これらのプログラムの例としては、例えば、生体高分子についてエネルギー的に好適な結合部位を決定するためのＧＲＩＤ（Ｇｏｏｄｆｏｒｄ，Ｐ．Ｊ．，（１９８５）Ｊ．Ｍｅｄ．Ｃｈｅｍ．，２８，ｐ．８４９−８５７）、機能的結合部位を探索するためのＭＣＳＳ（Ｍｉｒａｎｋｅｒ，Ａ．，ｅｔ ａｌ．，（１９９１）Ｐｒｏｔｅｉｎｓ：Ｓｔｒｕｃｔｕｒｅ，Ｆｕｎｃｔｉｏｎ ａｎｄ Ｇｅｎｅｔｉｃｓ，１１，ｐ．２９−３４）、あるいは高分子とリガンドとの相互作用を幾何学的に解析し、自動的にドッキングさせるＡＵＴＯＤＯＣＫ（Ｇｏｏｄｓｅｌｌ，Ｄ．Ｓ．，ｅｔ ａｌ．，（１９９０）ＰＲＯＴＥＩＮＳ：Ｓｔｒｕｃｔｕｒｅ，Ｆｕｎｃｔｉｏｎ ａｎｄ Ｇｅｎｅｔｉｃｓ，８，１９５−２０２）、及びＤＯＣＫ（Ｋｕｎｔｚ，Ｉ．Ｄ．，ｅｔ ａｌ．，（１９８２）Ｊ．Ｍｏｌ．Ｂｉｏｌ．１６１，２６９−２８８）等が挙げられる。
これらの候補化合物（リード化合物）が見つけ出されると、続いて、ＤＮＡ複製調節タンパク質のヘミメチル化ＤＮＡとの結合部位を表示したコンピュータ画面上で候補化合物との立体的な相互作用を視覚的に検討し、更に最適化することができる。前述したＱＵＡＮＴＡやＳＹＢＹＬ等のプログラムを用いて候補化合物を最適化するためのモデルを構築することができる。これらの他に、分子モデリングの最適化のために有用な種々のプログラムが利用でき、例えば、ＣＡＶＥＡＴ（Ｌａｕｒｉ，Ｇ．，ａｎｄ Ｂａｒｔｌｅｔｔ，Ｐ．Ａ．，Ｊ．（１９９４）Ｃｏｍｐｕｔ．Ａｉｄｅｄ Ｍｏｌ．Ｄｅｓ．８，５１−６６）やＨＯＯＫ（Ｅｉｓｅｎ，Ｍ．Ｂ．，ｅｔ ａｌ．（１９９４）Ｐｒｏｔｅｉｎｓ：Ｓｔｒｕｃｔ．，Ｆｕｎｃｔ．，Ｇｅｎｅｔ．，１９，１９９−２２１）等が挙げられる。
更に、上記のプログラムの他にＤＮＡ複製調節タンパク質のヘミメチル化ＤＮＡとの結合部位の立体構造から直接（ｄｅ ｎｏｖｏ）、自動的に候補化合物を構築し、その構造を出力することも可能である。これらのためには、ＬＵＤＩ（Ｂｏｈｍ，Ｈ．Ｊ．，（１９９２）Ｊ．Ｃｏｍｐ．Ａｉｄ．Ｍｏｌｅｃ．Ｄｅｓｉｇｎ．）やＬＥＧＥＮＤ（Ｎｉｓｈｉｂａｔａ，Ｙ．，ｅｔ ａｌ．，（１９９１）Ｔｅｔｒａｈｅｄｒｏｎ ４７，８９８５）等のプログラムも利用できる。
続いて、上記候補化合物又は最適化された化合物が、ＤＮＡ複製調節タンパク質とヘミメチル化ＤＮＡとの結合を阻害するかどうかを種々の方法で確認する。例えば、両者を水溶液中で混合する等の方法で接触させ、その後、電気泳動等を行うことにより、ＤＮＡ複製調節タンパク質がヘミメチル化ＤＮＡと結合できるか否かを確認する。上記候補化合物又は最適化された化合物がＤＮＡ複製調節タンパク質とヘミメチル化ＤＮＡとの結合を阻害することができれば、ＤＮＡ複製調節タンパク質による複製調節能もまた阻害されるものと考えられる。
あるいは、上記候補化合物とＤＮＡ複製調節タンパク質との相互作用を、ＮＭＲや質量分析法等を用いて分光学的に検出することもできる。一般的に、ＮＭＲシグナルは分子の局所的環境変化（リガンドとの結合、構造変化等）を敏感に反映するから、リガンドと相互作用するタンパク質を混合してＮＭＲスペクトルを測定した場合、リガンドシグナルのブロード化や消失、化学シフト値の変化などが確認される。反対に、リガンドと結合した状態のタンパク質の構造を特異的に検出することもできる。^１５Ｎで均一にラベルしたタンパク質を用いた^１５Ｎ−^１ＨＨＳＱＣスペクトルを利用すれば、タンパク質のＮＭＲシグナルのみを観測することができ、リガンドと相互作用するアミノ酸残基等の局所構造を見出すことができる。
次に、ヘミメチル化ＤＮＡと特異的に結合する化合物を同定するためには、ＤＮＡ複製調節タンパク質のヘミメチル化ＤＮＡとの結合に関与する領域の立体構造が重要な手がかりとなる。上述したように、配列番号２に示したＳｅｑＡタンパク質の１１５位のグリシンから１１８位のアルギニンまでの領域、及び１４９位のスレオニンから１５３位のスレオニンまでの領域は、表１に示した原子座標によって規定され、図３に示した立体構造を形成することによってヘミメチル化ＤＮＡのＧ（^ｍ６Ａ）ＴＣ配列部分と相互作用していることが分かる。通常の二本鎖ＤＮＡにおいては、塩基部分は相補鎖と水素結合しているため二重らせん構造の内部に存在するが、ヘミメチル化ＤＮＡではアデニンの６位のアミノ基がメチル化されることによって、アデニンとチミンとの水素結合が弱まり、二重らせん構造が部分的に開いた構造をとるものと考えられる。この微妙な構造のゆらぎをＳｅｑＡタンパク質のＤＮＡ結合部位が認識し、これと特異的に結合することが分かる。このヘミメチル化ＤＮＡとの結合に関与する立体構造情報はヘミメチル化ＤＮＡと結合する化合物を同定するための重要な手がかりとなる。
このような候補化合物の一つとして、変異体ＤＮＡ複製調節タンパク質が挙げられる。例えば、ヘミメチル化ＤＮＡと結合する上記２つのループ領域（１１５位のグリシンから１１８位のアルギニンまでの領域、及び１４９位のスレオニンから１５３位のスレオニンまでの領域）を含み、その結合部分の立体構造を保持するような任意のタンパク質又はペプチドである。該結合部分の立体構造は、表１に示す原子座標により実質的に規定される。或は、ＳｅｑＡタンパク質の上記ＤＮＡ結合部位を形成するアミノ酸残基又はその近傍領域のアミノ酸を他のアミノ酸に置換又は化学修飾を行って、ヘミメチル化ＤＮＡとの結合能力が向上した変異体タンパク質を作製することができる。ここに、「その近傍領域」とは、該アミノ酸残基に対して、静電相互作用、疎水性相互作用、ファンデアワールス相互作用、水素結合などに関与する領域、具体的にはおよそ５Å以内にある領域をいう。
このようにして設計された変異体タンパク質は、多くの方法によって調製され得る。例えば、本発明の変異体をコードするＤＮＡを部位特異的変異導入法やＰＣＲ、化学合成等の当業者において公知の方法により、又はこれらを組み合わせて調製することができる。これらの変異体タンパク質をコードするＤＮＡを適当な発現ベクターに導入し、該発現ベクターを更に適当な細胞に導入して得られた形質転換体を培養し、変異体タンパク質を発現させることが可能である。
以上の手法において設計された阻害剤については、その化合物に応じて、一般的に用いられている化学合成の手法を用いることで得ることができる。また、その阻害剤をＤａｍメチラーゼを有する病原性細菌の増殖阻害剤として用いる場合には、一般に製薬的に許容される担体と混合することによって、医薬組成物を調製することができる。医薬組成物用担体としては、特に制限はなく、例えば，乳糖、ブドウ糖、Ｄ−マンノース、澱粉、結晶セルロース、炭酸カルシウム、カオリン、ゼラチン、ヒドロキシプロピルアルコール、ポリビニルピロリドン、エタノール、カルボキシメチルセルロース、カルボキシメチルセルロースナトリウム塩、ステアリン酸マグネシウム、タルク、アセチルセルロース、白糖，酸化チタン、安息香酸、パラオキシ安息香酸エアステル、デヒドロ酢酸ナトリウム、アラビアゴム、トラガント、メチルセルロース，卵黄、界面活性剤、単シロップ、クエン酸、蒸留水、グリセリン、プロピレングリコール、マクロゴール、リン酸一水素ナトリウム、リン酸二水素ナトリウム、リン酸ナトリウム、塩化ナトリウム等があり、目的とする製剤、例えば、注射薬や錠剤等の形態に応じて、本発明の方法によって同定された阻害剤と混合して使用される。
実施例
（実施例１）大腸菌ＳｅｑＡタンパク質の種々の欠失変異体の調製
配列番号１に示した大腸菌ｓｅｑＡ遺伝子（Ｆｕｌｌ ｌｅｎｇｔｈ ５４３ｂｐ）が、プラスミドｐＥＴ１５ｂ（Ｎｏｖａｇｅｎ社）に挿入されたプラスミドｐＭＹ９０３を鋳型としてポリメラーゼ連鎖反応（ＰＣＲ）を行い、グルタチオンＳトランスフェラーゼ（ＧＳＴ）との融合タンパク質を発現する種々の欠失変異体プラスミドを作製した。
まず、プラスミドｐＭＹ９０３を鋳型とし、図１（ａ）に示したＳｅｑＡタンパク質の種々の欠失変異体をコードするＤＮＡ断片をＰＣＲにより増幅した。各ＤＮＡ断片がコードするタンパク質のＮ末端側にＢａｍＨＩ、Ｃ末端側にＥｃｏＲＩの制限酵素部位を付加したそれぞれ２種類のプライマーと組換えＴａｑポリメラーゼ（ＴＯＹＯＢＯ社）とを用いて各ＤＮＡ断片を増幅した後、ＰＣＲ ｐｕｒｉｆｉｃａｔｉｏｎ Ｋｉｔ（ＱＩＡＧＥＮ社）を用いて増幅したＤＮＡ断片を精製した。この増幅された二本鎖ＤＮＡ断片の各３’末端には、鋳型配列とは関係なく一塩基のデオキシアデノシン（Ａ）が付加されることから、これと相補的な一塩基のチミジン（Ｔ）の突出末端を有するプラスミドｐＧＥＭ−ＴＥａｓｙ（登録商標、ＰＲＯＭＥＧＡ社）にＴＡクローニングした。このクローン化されたＤＮＡの塩基配列を確認した後、ＥｃｏＲＩ及びＢａｍＨＩで切り出した各ＤＮＡ断片をＧＳＴ融合タンパク質発現用ベクターｐＧＥＸ６Ｐ−１（アマシャムファルマシアバイオテク社）のＥｃｏＲＩ−ＢａｍＨＩ部位にクローン化してＳｅｑＡタンパク質の種々の欠失変異体のＮ末端にＧＳＴが融合して発現する種々のプラスミドを構築した。
完全長ＳｅｑＡ及び上記方法により作製した種々の欠失変異体遺伝子を有する大腸菌（ＸＬ１−ｂｌｕｅ）をＬＢ培地で培養し、対数増殖期に最終濃度で１ｍＭとなるようにＩＰＴＧを加えてタンパク質の発現を誘導した。その後、４〜６時間培養して菌体を集め、５０ｍＭ Ｔｒｉｓ−ＨＣｌ、０．５Ｍ ＮａＣｌ、１０％グリセロールを含む緩衝液（ｐＨ８．０）中、氷冷下で超音波により細胞を破砕して発現したタンパク質を回収した。一方、同じ培地で一晩培養した培養液３ｍｌからプラスミドＤＮＡを抽出し、ｍｉｎｉ−ｐｒｅｐ（ＱＩＡＧＥＮ社）を用いて精製した。このようにして得られたプラスミドを制限酵素ＢａｍＨＩ及びＥｃｏＲＩで消化したのち、アガロースゲル電気泳動してプラスミドＤＮＡの収量を求めた。菌体破砕液から可溶性タンパク質が回収されたクローンは、そのタンパク質をコードするプラスミドの収率が著しく悪いことが分かった。この実験に用いたプラスミドはコリシンＥ１系（ＣｏｌＥ１系）のプラスミドであり、大腸菌のｏｒｉＣとは異なり、ＲＮＡＩ、ＲＮＡＩＩで制御される別の複製機構が存在するが、上記結果からは、ＳｅｑＡタンパク質もこの複製機構に関与していることが示唆された。
（実施例２）ＳｅｑＡ欠失変異体の精製
実施例１で作製した欠失変異体のうち、ＳｅｑＡ（１−５９）、ＳｅｑＡ（７−１８１）、ＳｅｑＡ（１８−１８１）、ＳｅｑＡ（４２−１８１）、ＳｅｑＡ（６０−１８１）、ＳｅｑＡ（７１−１８１）の６種類の変異体を選択してそれらのタンパク質を精製した。
上記各変異体ＳｅｑＡタンパク質を発現する大腸菌（ＸＬ１−ｂｌｕｅ）を２．５リットルのＬＢ培地で培養し、実施例１と同様の方法で発現させて菌体破砕液を調製した。これを５ｍｌのＧＳＴｒａｐＦＦカラム（アマシャムファルマシアバイオテク社）に添加し吸着させ、開始バッファーでカラムを洗浄した後、０〜２０ｍＭのグルタチオンの濃度勾配を有する５０ｍＭ Ｔｒｉｓ−ＨＣｌ、０．５Ｍ ＮａＣｌ、１０％グリセロールを含む緩衝液（ｐＨ８．０）を用いてＧＳＴ融合タンパク質を溶出した。この溶出画分を透析してグルタチオンを除去した後、ＧＳＴタグを切断するためにタンパク質１００μｇ当たり１ユニット（Ｕ）のＰｒｅＳｃｉｓｓｉｏｎ ｐｒｏｔｅａｓｅ（アマシャムファルマシアバイオテク社）と、４℃で１２〜１８時間反応させた。続いて、ＧＳＴタグを除去するために、反応液を再び５ｍｌのＧＳＴｒａｐＦＦカラムに添加して、非吸着画分（Ｆｌｏｗ ｔｈｒｏｕｇｈ画分）を分取した。更に、Ｓｕｐｅｒｄｅｘ７５カラムにより６種類のタンパク質を精製した。
対照として発現させた完全長ＳｅｑＡタンパク質は、上記方法によりＧＳＴタグを切るとほとんど凝集してしまい、微量しか精製できなかった。ＳｅｑＡ（７−１８１）、ＳｅｑＡ（１８−１８１）はＧＳＴタグを切ると分解する傾向があった。ＳｅｑＡ（１−５９）は多量体として存在することが示唆された。これらの結果より、ＳｅｑＡタンパク質の二量体化はＮ末端領域を介して起こることが分かった。ＣＤスペクトルの測定を行ったところ、何れのタンパク質も高次構造を保持していることが示された。
（実施例３）ゲルシフトアッセイによるヘミメチル化ＤＮＡとの結合活性の測定
実施例２で精製した種々の欠失ＳｅｑＡタンパク質がヘミメチル化ＤＮＡに結合する活性をゲル電気泳動による移動度の変化（ゲルシフトアッセイ）により測定した。４μｇの変異体ＳｅｑＡを含む２０μｌの結合溶液（１０ｍＭ Ｔｒｉｓ−ＨＣｌ（ｐｈ７．５）、１ｍＭ ＥＤＴＡ、１０％グリセロール）に種々のヘミメチル化、又はフルメチル化ＤＮＡを加え、２５℃で３０分以上静置した後に１×ＴＢＥ、１２％アクリルアミドゲルで５０Ｖ定電圧で電気泳動を行い、ゲルをエチジウムブロマイド染色した。
まず、これまでにＳｅｑＡタンパク質は２個以上のヘミメチル化Ｇ（^ｍ６Ａ）ＴＣ配列に結合することが分かっていたので、配列番号３〜８に示した６種類の一本鎖オリゴヌクレオチドを合成し、それぞれ相補鎖をアニールして以下の３種類の二本鎖ＤＮＡを合成した。

フルメチル化、及びヘミメチル化ＤＮＡとしては、上記ＧＡＴＣ配列のアデニンのＮ６位をメチル化したオリゴヌクレオチドを合成し、アニーリングしたものを使用した。例えば、配列番号３のＧＡＴＣ配列（枠で囲んだ部分）のアデニンをメチル化した一本鎖オリゴヌクレオチドと配列番号４の非メチル化オリゴヌクレオチドとアニーリングさせるとヘミメチル化ＤＮＡ（ＳｅｑＡ ｂｓＡ−３７（Ｍ^＋／Ｍ⁻）と表す。）が調製される。一方、配列番号６の一本鎖オリゴヌクレオチドをメチル化し、配列番号５の非メチル化オリゴヌクレオチドとアニーリングさせるとヘミメチル化ＤＮＡ（ＳｅｑＡ ｂｓＢ−２６（Ｍ⁻／Ｍ^＋）と表す。）を調製することができる。ＳｅｑＡ ｂｓＣ−１４と種々の欠失変異体ＳｅｑＡタンパク質を用いて行ったゲルシフトアッセイの結果を図５に示した。図中、１〜４の各レーンに用いたＤＮＡは、１：Ｍ^＋／Ｍ^＋、２：Ｍ^＋／Ｍ⁻、３：Ｍ⁻／Ｍ^＋、４：Ｍ⁻／Ｍ⁻のフルメチル、ヘミメチル又は非メチル化ＤＮＡを添加したことを、又各電気泳動パターンの下に示した数字は変異体ＳｅｑＡタンパク質を構成するアミノ酸残基を示す。図５より、ＳｅｑＡタンパク質のＮ末端領域のみ（１〜５９位）ではＤＮＡ結合能がないことが分かる。一方、Ｎ末端から７個〜７０個のアミノ酸が欠失した変異体でもヘミメチル化ＤＮＡとの結合能を有する。即ち、ＳｅｑＡタンパク質のＤＮＡ結合領域はＮ末端から７１番目以降のＣ末端領域であることが分かった。また、ヘミメチル化部位が１ヶ所でもＳｅｑＡはヘミメチル化ＤＮＡと結合できることが分かった。
（実施例４）結晶化とＸ線結晶構造解析データの収集
実施例３の結果より、Ｎ末端部分を最も大きく欠失した変異体ＳｅｑＡ（７１−１８１）タンパク質をＸ線結晶構造解析のために精製し、ヘミメチル化ＤＮＡとの複合体の結晶を作製した。
実施例２と同様の方法により、５リットルの培養液からＧＳＴと融合した変異体ＳｅｑＡ（７１−１８１）タンパク質を回収し、５ｍｌのＧＳＴｒａｐＦＦカラムに吸着させ、０〜２０ｍＭのグルタチオン濃度勾配で溶出した後、透析によるグルタチオンの除去、及びＰｒｅＳｃｉｓｓｉｏｎ ｐｒｏｔｅａｓｅによるＧＳＴタグの除去を行った。続いて５ｍｌのＧＳＴｒａｐＦＦカラムに通してＧＳＴタグを除去した後、更にＳｕｐｅｒｄｅｘ７５カラムにより変異体ＳｅｑＡ（７１−１８１）タンパク質を精製した。
このタンパク質を種々の長さの二本鎖ヘミメチル化ＤＮＡと１．５：１のモル比で混合し、濃縮して結晶化を行ったところ、以下の１０塩基対からなるヘミメチル化二本鎖ＤＮＡとの複合体が六角柱状の結晶を作ることが分かった。

上記ＳｅｑＡ（７１−１８１）タンパク質とＳｅｑＡ ｂｓｃ−１０（Ｍ^＋／Ｍ⁻）との複合体は、ハンギングドロップ法により、０．１Ｍ ＨＥＰＥＳ−Ｎａ（ｐＨ７．５）、１．４Ｍクエン酸３ナトリウム水和物を含む溶液にて結晶化した。得られた結晶は、ＳｅｑＡ（７１−１８１）タンパク質と１０塩基対のヘミメチル化ＤＮＡとの複合体が２回回転軸で対称な２量体であり、対称性の高い六方晶系のＰ６２２であった。なお、格子定数は、ａ＝ｂ＝１５２．２３８Å、ｃ＝１１９．３３８Å、単位格子はα＝β＝９０°、γ＝１２０°であった。
天然型及びセレノメチオニン置換体結晶の両方のデータセットを、理研播磨研究所の８ＧｅＶの大型放射光施設（ＳＰｉｎｇ−８）のビームラインＩ（ＢＬ４１ＸＵ）で収集した。
セレノメチオニンによる重原子置換体結晶から得られたデータを用いて、多波長異常散乱法（ＭＡＤ法）により位相決定を行い、天然型結晶のデータを用いて拡張（Ｐｈａｓｅ ｅｘｔｅｎｔｉｏｎ）を行い、最終的に３．０Åの電子密度マップが得られた。立体構造のグラフィック表示はプログラムＭＯＬＳＣＲＩＰＴ及びＳＰＯＣＫを用いて行った。
（実施例５）ＳｅｑＡタンパク質のヘミメチル化ＤＮＡとの結合部位に結合する化合物のインシリコスクリーニング
〈データベース最適化〉
低分子化合物データベースの対照データベースとしては、ＳＰＥＣＳ社から提供されている低分子化合物カタログデータベースを用いた。１エントリーに複数分子含むものは１分子ごとに分割した上で、重複を除いたライブラリをスクリーニングの母集団（１５２３２３分子）とした。次に、”Ｌｉｐｉｎｓｋｉ’ｓ Ｒｕｌｅ ｏｆ ５”に基づき、上記低分子化合物データベースについてのターゲット非依存最適化を行った。ここで用いた絞込条件は以下の通りである。
１．分子量１００以上５００以下
２．計算ＬｏｇＰ値（ｏ／ｗ）５以下（ＸＬＯＧＰ−１アルゴリズム使用）、
３．水素結合アクセプター原子数（低分子化合物に含まれるＮとＯの数）１０以下
４．水素結合ドナー原子数（低分子化合物に含まれるＮＨとＯＨの数）５以下
更にドッキング計算を適切に行うために以下のような分子を除外した。
１．回転可能な単結合数が２１以上のもの
２．ラジカルを含むもの
３．Ｈ，Ｃ，Ｎ，Ｏ，Ｆ，Ｓ，Ｐ，Ｃｌ，Ｂｒ，Ｉ以外の元素を含むもの。
絞込み後の分子数は１０３７７３となった。
〈結合部位予測〉
大腸菌のＳｅｑＡタンパク質の中で、図３に示したヘミメチル化ＤＮＡとの結合に本質的に関わっていると推察される残基群（配列番号２に示したアミノ酸配列において１１５位のグリシンから１１８位のアルギニン、及び１４９位のスレオニンから１５３位のスレオニンを含むループ領域）を結合部位とした。
〈スクリーニング〉
（１）１次スクリーニング
Ｄｏｃｋ４．０を用いて、上記結合サイトについて、先に最適化を行った低分子化合物ライブラリー全体に対してドッキングを行った。この時、Ｄｏｃｋのエネルギースコア（ｅｎｅｒｇｙ ｓｃｏｒｅ）に基づいて低分子化合物をランク付けした。
（２）２次スクリーニング
１次スクリーニングの結果得られた上位１００００分子とデータベース内に含まれるその類似化合物（Ｓｙｂｙｌ Ｆｉｎｇｅｒ Ｐｒｉｎｔ ｔａｎｉｍｏｔｏ ｓｉｍｉｌａｒｉｔｙ０．８以下）の１１１７１分子について、ＡｕｔｏＤｏｃｋ３．０．５を用いて詳細なドッキングを行った。その後、ドッキングの際に使用したスフィアーから４Å以内にある化合物に絞込み、更に計算が異常終了した分子を除去した。この結果、最終的に９９６１分子についてドッキング構造を得た。得られた化合物のうち、上述の結合部位との結合性が高い上位１００化合物をリガンド候補化合物とした。このリガンド候補化合物のΔＧ（結合自由エネルギー変化）は、−９．８ｋｃａｌ／ｍｏｌ以上であった。
（実施例６）ＮＭＲによるＳｅｑＡタンパク質のヘミメチル化ＤＮＡとの結合部位に結合する化合物のスクリーニング
実施例５で得られた各リガンド候補化合物をＤＭＳＯ−ｄ６に溶解し、１ｍＭに調整した。実施例２で調製したＳｅｑＡ（７１−１８１）タンパク質は緩衝液Ａ（２０ｍＭ Ｔｒｉｓ−ｄ１１（ｐＨ７．５）／０．５Ｍ ＮａＣｌとなるように重水に溶解したもの）を用いて１０３．７μＭの濃度に溶解した。これらを用いて、リガンド候補化合物単独と、そこにＳｅｑＡ（７１−１８１）タンパク質を１：１の比で加えたもの、との２種類の試料を用意した。

リガンド候補化合物の終濃度は５０μＭとなる。

リガンド候補化合物及びＳｅｑＡタンパク質の終濃度はそれぞれ５０μＭとなる。
上記それぞれの試料について、重水素中、２９８ＫでＢｒｕｋｅｒＤＲＸ５００ＮＭＲ分光計を用いて１次元^１Ｈ−ＮＭＲを測定し、ＳｅｑＡタンパク質を添加する前後でスペクトルに変化があるかどうか観察した。その結果、リガンド候補化合物として、試料ＩＤ：ＡＫ９６８−４１１７０３３５、化合物名称：ｃｉｓ−６−［４−（４−Ｉｓｏｐｒｏｐｙｌ−ｐｈｅｎｙｌ）−５−ｍｅｔｈｙｌ−ｔｈｉａｚｏｌ−２−ｙｌｃａｒｂａｍｏｙｌ］−３，４−ｄｉｍｅｔｈｙｌ−ｃｙｃｌｏｈｅｘ−３−ｅｎｅｃａｒｂｏｘｙｌｉｃ ａｃｉｄ（化合物Ａ、−１０．４７ｋｃａｌ／ｍｏｌ）を用いた場合のＮＭＲスペクトル（７．５ｐｐｍ付近のシグナルを拡大したもの）を図６に示した。図６（ａ）は化合物Ａ単独、図６（ｂ）は化合物ＡとＳｅｑＡタンパク質とを混合したもの、図６（ｃ）は（ａ）と（ｂ）のシグナルを重ねてを比較したものである。これらの結果より、ＳｅｑＡとの相互作用により化合物Ａのベンゼン環の４個の水素に由来するシグナルが消失していることが認められ、該化合物がＳｅｑＡに結合していることが示唆された。
このような方法を用いて、実施例５で得られた１００種類のリガンド候補化合物の中からＤＭＳＯへの溶解度及びＮＭＲによるシグナルの検出可能なものについてＮＭＲスペクトルを測定した結果、ＳｅｑＡと相互作用が検出された化合物を以下の表４に示した。

表４に示した化合物のうち、化合物Ａ，Ｂ，Ｃ，Ｄ，Ｅ，Ｇ，Ｈ，及びＪの化学構造は互いに類似することから、これらの化合物に共通の下記一般式（Ｉ）に示す化合物は、ＳｅｑＡのヘミメチル化ＤＮＡ結合部位に特異的に結合することが推測される。

但し、Ａはシクロヘキセン、ジメチルシクロヘキセン、ビシクロ［２，２，１］ヘプタン、ビシクロ［２，２，１］ヘプテン、オキサ−ビシクロ［２，２，１］ヘプタンからなる群から選択される何れか１種の環を、Ｒ_１及びＲ_２は水素原子又はメチル基を、Ｒ_３は水素原子又は炭素数１〜４のアルキル基を表す。これらの化合物は、細胞内の生理的条件下においてＳｅｑＡタンパク質と結合することから塩の形態（ナトリウム塩、カリウム塩、アンモニウム塩等）であってもよく、分子内に不斉炭素原子を有する場合には光学活性体又はラセミ体であってもよい。
また、上記表４において、化合物Ａとは比較的化学構造の異なる化合物Ｆ，Ｉを用いた場合のＮＭＲスペクトルの測定をそれぞれ行い、図７に化合物Ｆを用いた場合の結果を示した。図７（ａ）は化合物Ｆ単独、図７（ｂ）は化合物ＦとＳｅｑＡタンパク質とを混合したもの、図７（ｃ）は（ａ）と（ｂ）のシグナルを重ねてを比較したものである。化合物Ａの場合と同様に、ＳｅｑＡタンパク質と混合することによって化合物Ｆの特徴的なシグナルが消失していることが認められ、化合物ＦについてもＳｅｑＡタンパク質との相互作用が示唆された。
更に、化合物Ｉを用いた場合にも上述と同様の特徴をもつＮＭＲスペクトルが得られ、ＳｅｑＡタンパク質との相互作用が示唆された。
産業上の利用可能性
本発明によれば、ＤＮＡ複製調節タンパク質とヘミメチル化ＤＮＡとの複合体の立体構造モデルが提供される。本複合体の立体構造は、ヘミメチル化ＤＮＡとのタンパク質の結合様式、更に、該複合体による染色体ＤＮＡの複製の調節機構を理解する上で極めて有用である。ＤａｍメチラーゼによるＧＡＴＣ配列のメチル化は、大腸菌近縁の特殊なグループの細菌のみが有する構造であることから、本発明のＤＮＡ複製調節タンパク質による複製制御を阻害することで、上記細菌の増殖を特異的に調節できる可能性がある。本発明により明らかとなったＤＮＡ複製調節タンパク質の立体構造は、これらの薬剤を開発するためのスクリーニング方法として利用できる。
【配列表】

【図面の簡単な説明】
図１は、本発明において合成した種々の変異体ＳｅｑＡタンパク質の構造を模式的に示した図である。（ａ）は、２０種類の変異体ＳｅｑＡタンパク質の一次構造を示したものであり、（ｂ）は、ＳｅｑＡタンパク質の機能ドメインを示したものである。
図２は、ＤＮＡ複製調節タンパク質とヘミメチル化ＤＮＡとの複合体の立体構造モデルを表示するコンピュータシステムの一実施形態を表した概略図である。
図３は、変異体ＳｅｑＡタンパク質（７１−１８１）と１０塩基対のヘミメチル化ＤＮＡとの複合体結晶の立体構造を表したモデル図である。
図４は、大腸菌を含む種々の細菌が有するＳｅｑＡタンパク質の部分構造を、一次構造の相同性を考慮して並列して表したものである。図中、らせん構造はα−ヘリックスを矢印はβ−シートを形成する領域であることを示す。
図５は、１４塩基対のオリゴヌクレオチド（ＳｅｑＡ ｂｓＣ−１４）と種々の変異体ＳｅｑＡタンパク質とのゲルシフトアッセイの結果を示したものである。（ａ）〜（ｆ）の各実験に用いた変異体ＳｅｑＡタンパク質は、それぞれ（ａ）：１−５９、（ｂ）：７−１８１、（ｃ）：１８−１８１、（ｄ）：４２−１８１、（ｅ）：６０−１８１、（ｆ）：７１−１８１のアミノ酸残基から構成されている。１〜４の各レーンは、それぞれ、１：Ｍ^＋／Ｍ^＋、２：Ｍ^＋／Ｍ⁻、３：Ｍ⁻／Ｍ^＋、４：Ｍ⁻／Ｍ⁻のフルメチル、ヘミメチル又は非メチル化ＤＮＡを添加したことを示す。
図６はリガンド候補化合物ＡＫ−９６８−４１１７０３３５（化合物Ａ）単独（ａ）、及び化合物ＡとＳｅｑＡタンパク質との１：１の混合物（ｂ）のプロトンＮＭＲスペクトル（７．５ｐｐｍ付近のシグナル）である。なお（ｃ）は（ａ）と（ｂ）のシグナルを重ねて比較したものである。
図７はリガンド候補化合物ＡＮ−９８９−１４０７２０１８（化合物Ｆ）単独（ａ）、及び化合物ＦとＳｅｑＡタンパク質との１：１の混合物（ｂ）のプロトンＮＭＲスペクトル（７．５ｐｐｍ付近のシグナル）である。なお（ｃ）は（ａ）と（ｂ）のシグナルを重ねて比較したものである。 Technical field
The present invention relates to crystallization of a DNA replication regulatory protein, determination of a three-dimensional structure of a DNA replication regulatory protein by X-ray crystal structure analysis using the crystal, and use thereof, and in particular, to the protein based on the three-dimensional structure. It relates to rational screening of binding compounds or compounds that specifically bind to hemimethylated DNA.
Background art
There are about 1900 GATC sequences in the chromosomal DNA of E. coli. This GATC sequence is usually methylated by Dam methylase. This Dam methylase transfers a methyl group from S-adenosylmethionine, which is a methyl group donor, to an amino group at the 6-position of an adenine base. The E. coli Dam methylase is encoded by the dam (DNA Adenine Methyltransferase) gene and is a monomer enzyme having a molecular weight of 31 kDa (Boye, E., et al., 1992, J. Bacteriol., 174, 1682-1685).
When replicating chromosomal DNA, the nascent strand is in a hemimethylated state until it is methylated by Dam methylase. That is, in the DNA immediately after replication, the nascent-strand GATC sequence is not yet methylated, but its complementary strand (old strand) is in a methylated state (semi-methylated state). When a Dam methylase acts on the DNA, it becomes a fully methylated DNA in which both strands are methylated. SeqA protein specifically recognizes and binds to GATC sequences that have become hemimethylated immediately after replication (Brendler et al., 1995, EMBO J., 14, 4083-4089, and Slater et al., 1995, Cell). 82, 927-936). There are 11 GATC sequences in the oriC region, which is the replication origin of Escherichia coli. SeqA binds to the oriC region in the hemimethylated state, thereby delaying methylation by Dam methylase and at the same time, the hemimethylated oriC. It is also involved in membrane binding. In addition, the SeqA protein plays a role in synchronizing the initiation of replication from oriC, and is considered to negatively control replication.
Although other factors other than SeqA are involved in this negative regulation of replication (or sequestration of the origin of replication), SeqA protein alone has an activity to bind strongly to hemimethylated DNA in vitro, and this SeqA It has been reported that protein binding requires the presence of at least two hemimethylated GATC sequences (Brendler, T., et al., 1999, EMBO J., 18, 2304-2310).
Such Dam methylase and SeqA protein are not present in mammals including humans, and are a special group of Gram-negative bacteria including pathogenic bacteria such as plague, cholera, shigella, Salmonella typhimurium and Escherichia coli O157. (See Hiraga et al., 2000, Genes to Cells, 5, 327-341). Therefore, by inhibiting the binding mechanism of DNA replication, that is, binding of Dam methylase or SeqA protein to hemimethylated DNA, the growth of these bacteria may be specifically suppressed. It has also been reported that Salmonella typhimurium deficient in Dam methylase works as a live vaccine in mice (see Hiraga et al., 2000, Genes to Cells, 5, 327-341), which is specific for hemimethylated DNA. Development of a drug that binds to and inhibits the action of Dam methylase may reduce the virulence of the bacterium and promote the production of antibodies against the bacterium without causing the infected host to become ill.
However, in order to do so, detailed information on the mode of binding between SeqA protein and hemimethylated DNA and the mechanism of binding and dissociation of these complexes is required. However, the three-dimensional structure of SeqA protein, in particular, hemimethylated DNA and The knowledge of the three-dimensional structure about the bond of is not obtained.
The present invention obtains a single crystal of a complex of E. coli SeqA protein, which is one of DNA replication regulatory proteins, and hemimethylated DNA, reveals its three-dimensional structure by X-ray crystal structure analysis, and regulates replication by SeqA protein. The purpose is to clarify the mechanism.
The present invention also provides a method for screening a compound that inhibits the formation of a complex or a compound that specifically binds to hemimethylated DNA based on the structure of the complex of SeqA protein and hemimethylated DNA. For the purpose.
Disclosure of the invention
In order to clarify the functional domain of the SeqA protein, the present inventors constructed seqA genes having various deletion mutations and expressed them in Escherichia coli so that various physical properties of the mutant protein and hemimethylated DNA As a result, the functional domain involved in the binding to hemimethylated DNA was revealed on the C-terminal side. In addition, the present invention was completed by crystallizing a complex of these functional domains and hemimethylated DNA, analyzing the X-ray crystal structure using the crystal, and clarifying the three-dimensional structure of the complex for the first time. did.
That is, the present invention relates to a crystal of SeqA protein that is a DNA replication regulatory protein, a method for producing the same, analysis of a three-dimensional structure using the same, and the complex using the three-dimensional structure thus obtained. More specifically, the present invention relates to a method for screening a compound that inhibits the formation of, or a compound that specifically binds to hemimethylated DNA.
(1) A polypeptide fragment comprising the amino acid sequence from positions 71 to 181 of the amino acid sequence shown in SEQ ID NO: 2, from glycine at position 115 to arginine at position 118, and threonine at position 149 A polypeptide fragment, wherein amino acid residues up to threonine form a DNA binding site.
(2) A polypeptide fragment comprising the amino acid sequence from positions 71 to 181 of the amino acid sequence shown in SEQ ID NO: 2, from glycine at position 115 to arginine at position 118 and threonine at position 149 A polypeptide fragment, wherein amino acid residues up to threonine form a binding site with hemimethylated DNA.
(3) A crystal of a complex of a polypeptide fragment consisting of positions 71 to 181 of the amino acid sequence shown in SEQ ID NO: 2 and hemimethylated DNA, a hexagonal space group P622, a lattice constant a = b = Crystals of the complex having 152.2 ± 0.2 ｃ, c = 119.3 ± 0.2 Å.
(4) A method for producing a crystal of a complex of a DNA replication regulatory protein consisting of the amino acid sequence shown in SEQ ID NO: 2 and hemimethylated DNA,
A solution prepared by mixing a DNA replication regulatory protein having a concentration of 4 to 6 mg / ml and hemimethylated DNA in a molar ratio of 1: 1 to 2: 1 contains 1-2 M trisodium citrate hydrate, A method of crystallizing by a vapor diffusion method using a buffer solution having a pH of 7.0 to 8.0.
(5) Using the polypeptide fragment described in (1), a compound that binds to the polypeptide fragment is designed or selected, and a compound that inhibits DNA replication regulation ability and / or a compound that specifically binds to hemimethylated DNA is screened. how to.
(6) A method for determining a three-dimensional structure of a DNA replication regulatory protein or a complex of the protein and hemimethylated DNA,
(A) producing a crystal of a DNA replication regulatory protein or a complex of the protein and hemimethylated DNA;
(B) a step of obtaining X-ray diffraction data by the heavy atom isomorphous substitution method or the multiwavelength anomalous scattering method using the crystal; and
(C) determining three-dimensional structure coordinates based on the X-ray diffraction data;
A method for determining a three-dimensional structure of a DNA replication regulatory protein or a complex of said protein and hemimethylated DNA,
(7) The method according to (6), wherein the DNA replication regulatory protein is a SeqA protein of a bacterium of the genus Escherichia, Haemophilus, Pasteurella, or Vibrio.
(8) A method for identifying a compound that inhibits the ability to regulate DNA replication,
(A) constructing a DNA replication regulatory protein or a complex of the protein with hemimethylated DNA, or a three-dimensional structure of the polypeptide fragment of (1) or its DNA binding site using a computer model;
(B) designing or selecting a candidate compound that inhibits the ability to regulate DNA replication based on the three-dimensional structure;
(C) synthesizing or obtaining the designed or selected candidate compound;
(D) contacting the candidate compound with the DNA replication regulatory protein in order to examine the inhibitory activity on the ability of the candidate compound to regulate DNA replication;
A method for identifying a compound that inhibits the ability to regulate DNA replication.
(9) The method according to (8), wherein the DNA replication regulatory protein is a SeqA protein of a bacterium of the genus Escherichia, Haemophilus, Pasteurella or Vibrio.
(10) The method according to (8) or (9), wherein the inhibitory activity on the ability to regulate DNA replication in the step (d) is detected by gel shift assay.
(11) The method according to any one of (8) to (10), wherein the three-dimensional structure model is substantially defined by atomic coordinates shown in Table 1 of the present specification.
(12) A compound that inhibits the ability to regulate DNA replication, characterized by being identified by the method according to any one of (5) or (8) to (11).
(13) A method for producing an antibacterial composition containing a compound that inhibits the ability to regulate DNA replication, wherein the ability to regulate DNA replication by the method according to any one of (5) or (8) to (11) above A method for producing an antibacterial composition, comprising the steps of: identifying a compound that inhibits the activity; and mixing the identified compound with a stabilizer or carrier.
(14) A method for identifying a compound that specifically binds to hemimethylated DNA,
A step of constructing a three-dimensional structure of a complex of (a) a DNA replication regulatory protein or the polypeptide fragment according to (2) and hemimethylated DNA using a computer model;
(B) designing or selecting a candidate compound that binds to hemimethylated DNA based on the three-dimensional structure;
(C) synthesizing or obtaining the designed or selected candidate compound;
(D) contacting the candidate compound with hemimethylated DNA in order to examine the binding ability of the candidate compound to hemimethylated DNA;
A method for identifying a compound that specifically binds to hemimethylated DNA, comprising:
(15) The method according to (14), wherein the candidate compound is a mutant DNA replication regulatory protein.
(16) A computer system that outputs a three-dimensional structure of a DNA replication regulatory protein comprising the amino acid sequence shown in SEQ ID NO: 2 or a complex of the protein and hemimethylated DNA,
(A) storage means for storing and storing computer-readable data including atomic coordinates of amino acid residues corresponding to the binding site of the DNA replication regulatory protein with hemimethylated DNA;
(B) Deriving means for reading out the data held in the storage means and deriving a three-dimensional structure model;
And (c) an output unit that receives the derivation result and outputs the three-dimensional structure model.
(17) The binding site of the DNA replication regulatory protein to hemimethylated DNA is from glycine at position 115 to arginine at position 118 and threonine at position 149 to threonine at position 153 of the amino acid sequence shown in SEQ ID NO: 2. (14) The computer system according to (16), characterized in that it consists of a region.
(18) A DNA replication regulation inhibitor comprising a compound represented by the following general formula (I) or a salt thereof as an active ingredient.

Wherein A is selected from the group consisting of cyclohexene, dimethylcyclohexene, bicyclo [2,2,1] heptane, bicyclo [2,2,1] heptene, oxa-bicyclo [2,2,1] heptane. Any one ring is R₁And R₂Is a hydrogen atom or a methyl group, R₃Represents a hydrogen atom or an alkyl group having 1 to 4 carbon atoms).
(19) An antibacterial composition comprising a compound represented by the above general formula (I) or a salt thereof as an active ingredient.
It is.
BEST MODE FOR CARRYING OUT THE INVENTION
In the present invention, “DNA replication regulatory protein” refers to a protein that regulates DNA replication by binding to the replication initiation region of chromosomal DNA immediately after replication and inhibiting new replication initiation until the cell cycle is synchronized. Such regulation in E. coli is known to be performed by the SeqA protein, which is an expression product of the seqA gene, and the nucleotide sequence and the primary structure of the protein have already been clarified (Lu, M., et al. , 1994, Cell, 77, 413-426, GenBank Accession No. U07651). SeqA protein is present in various bacteria having Dam methylase including E. coli, for example, bacteria of the genus Escherichia, Haemophilus, Pasteurella or Vibrio. And pathogenic bacteria such as Escherichia coli O157. Furthermore, at least a part of the protein having a DNA replication regulation function is substantially included in the “DNA replication regulation protein” in the present invention.
In the present invention, “hemimethylated DNA” refers to double-stranded DNA having a GATC sequence in which only adenine in one strand is N⁶DNA that has undergone methylation at the position may be either biologically derived or chemically synthesized. Examples of the organism-derived material include oriC, which is a replication initiation region of E. coli. OriC consisting of 245 bp has 11 GATC sequences, and immediately after semi-conservative replication, it exists as a hemimethylated state in which only one strand is methylated until the nascent strand is methylated by Dam methylase. . On the other hand, a double-stranded oligonucleotide of any length is designed by chemical synthesis, and only the adenine of the GATC sequence of one strand is N⁶It can also be prepared by synthesizing a methylated oligonucleotide and annealing it with a complementary strand.
In the present invention, a “mutant” of a DNA replication regulatory protein has an amino acid sequence in which at least one amino acid of the amino acid sequence shown in SEQ ID NO: 2 has been deleted, added, inserted, substituted or modified, and hemimethylated A modified protein that retains at least a part of the binding activity with DNA.
(Search for functional domain of SeqA protein)
FIG. 1 (a) shows 20 mutant SeqA proteins from which amino acid sequences of various lengths have been deleted from the N-terminal side and C-terminal side of the SeqA protein. By constructing a plasmid having a mutant seqA gene encoding these and introducing it into E. coli to synthesize a GST fusion protein, in 5 types of plasmids expressing full-length SeqA protein and N-terminal deletion mutant SeqA protein, It was found that the copy number of the plasmid was extremely reduced. With respect to the plasmid expressing the C-terminal deletion mutant SeqA protein, the copy number was restored to the same copy number as that of the vector-only plasmid without the seqA gene only by deleting the C-terminal by 5 amino acid residues. This suggests that the SeqA protein is involved not only in the Escherichia coli chromosome oriC but also in the replication of the colicin E1 plasmid. It was also revealed that SeqA protein has a DNA binding domain on the C-terminal side (FIG. 1 (b)).
(Preparation of complex crystal of DNA replication regulatory protein and hemimethylated DNA)
The most commonly used technique for clarifying the three-dimensional structure of a protein is an X-ray crystal structure analysis technique. That is, the protein is crystallized, the monochromatic X-ray is applied to the crystal, and the three-dimensional structure of the protein is clarified based on the obtained X-ray diffraction image (Blundell, TL). And Johnson, L.N., PROTEIN CRYSTALLOGRAPY, p.1-565, (1976) Academic Press, New York).
Crystallization is performed by adding a precipitant to the target protein solution or reducing the amount of the solvent by evaporation, etc., so that the protein precipitates as a crystal under certain conditions when the protein changes from a solution state to a non-solution state. The property to be used is used. Crystallization requires highly purified protein. In addition, physical and chemical factors such as protein concentration, salt concentration, hydrogen ion concentration (pH), type of precipitating agent to be added, and temperature are involved as conditions for crystallization. Furthermore, there are many crystallization methods such as batch method, dialysis method, vapor diffusion method and the like as methods for adding a precipitant and adjusting the amount of solvent (Blundell, TL and Johnson, L.N.,). PROTEIN CRYSTALLOGRAPY, p. 59-82, (1976) Academic Press, New York). For example, the vapor diffusion method is a mechanism in which a volatile solvent or a precipitant moves by diffusion through a gas phase. As the protein solution vapor diffuses with the precipitant solution, both the protein concentration and the precipitant concentration increase, and crystallization occurs in the supersaturated region. There are methods such as a hanging drop method, a sitting drop method, and a sandwich drop method, depending on how the protein solution droplet is formed. In order to obtain protein crystals suitable for X-ray crystallography in this way, it is necessary to obtain high-purity proteins for each protein and to study various factors in crystallization and optimization of crystallization methods. Become. In particular, the preparation of highly pure protein is an important factor for successful crystallization. For this reason, rapid and uniform protein can be prepared by expressing the cloned gene in an efficient expression system.
A crystal of a complex of the DNA replication regulatory protein of the present invention and hemimethylated DNA is prepared as follows. The base sequence encoding the E. coli SeqA protein is already known, and can be easily expressed in the original host E. coli and various other hosts using recombinant DNA technology.
In one embodiment of the present invention, E. coli SeqA protein can be expressed in a large amount by a system using recombinant E. coli. Various plasmid vectors and the like for expressing a large amount of genes in recombinant E. coli are commercially available. For example, a pET system (Novagen) or the like in which a gene to be expressed can be introduced downstream of a strong T7 phage promoter and transcribed by T7 RNA polymerase supplied by the host E. coli can be used. In this system, the T7 RNA polymerase gene is integrated into the host E. coli chromosome and is downstream of the lacUV5 promoter, so that expression can be controlled by induction by addition of IPTG. In addition, by introducing various tag sequences into the N-terminal side or C-terminal side of the E. coli SeqA protein, purification can be easily performed using a metal chelate resin (histidine tag), antibody (T7 tag, CBD tag, etc.), etc. can do. The purified fusion protein can further recover only the E. coli SeqA protein portion using a protease such as enterokinase or thrombin.
Further, by growing and culturing a recombinant E. coli expressing E. coli SeqA protein in a medium containing selenomethionine or selenocysteine containing selenium as a heavy metal atom, the methionine or cysteine residue in the E. coli SeqA protein is selenomethionine or A mutant protein substituted with selenocysteine is obtained.
Further, these proteins are purified to a high purity so as to be suitable for crystallization. As purification methods, methods commonly used as methods for purifying proteins by those skilled in the art such as column chromatography (affinity, hydrophobicity, ion exchange, gel filtration, etc.), salting out, centrifugation, electrophoresis, etc. are used alone. Or in combination.
Next, crystals of a complex of DNA replication regulatory protein and hemimethylated DNA are prepared. When analyzing the function of a DNA binding protein such as SeqA protein, it is essential to crystallize in a modeled system in which a complex is formed with a specific oligonucleotide. For example, it is preferable to previously mix SeqA protein of E. coli and hemimethylated double-stranded DNA of about 10 to 50 base pairs in a molar ratio of 1: 1 to 2: 1.
In addition to the method of introducing heavy metal atoms into protein by culturing in the above selenium-containing medium, natural protein crystals are converted into metal salts containing gold, platinum, iridium, osmium, mercury, lead, uranium, samarium, etc. A heavy atom derivative crystal can also be prepared by immersing in a solution of an organometallic compound, and is used when applying the heavy atom isomorphous substitution method and the multiwavelength anomalous dispersion method.
A complex crystal of a SeqA protein having positions 71 to 181 of the amino acid sequence shown in SEQ ID NO: 2 and 10 base pairs of hemimethylated DNA belongs to the hexagonal space group P622, and is used as a unit cell. It has a size of 152.238 mm in the a-axis and b-axis directions and 119.338 mm in the c-axis direction.
(X-ray crystal structure analysis of complex crystal of DNA replication regulatory protein and hemimethylated DNA)
The crystal structure of the complex of the DNA replication regulatory protein and hemimethylated DNA thus obtained is analyzed using an X-ray crystal structure analysis technique known to those skilled in the art. In the determination of the three-dimensional structure as in the present invention in which the three-dimensional structure of the related protein is unknown and the structure cannot be analyzed by the molecular substitution method using the three-dimensional structure, the heavy atom substitution method or the multi-wavelength anomalous dispersion method (MAD method) is used. ) Can be applied. In one embodiment of the present invention, the X-ray crystallographic analysis is a MAD method, and the X-ray used for this is a wavelength of 0.3 to 3.0 mm obtained from synchrotron radiation. The MAD method is a method for obtaining X-ray diffraction data at a plurality of wavelengths from a heavy atom isomorphous substituted crystal having an absorption edge near incident X-ray energy. The resonance between the X-ray and the electron shell of the heavy atom causes a difference in X-ray scattering, which can solve the phase problem in protein structural analysis. Although the principle of this method has been known for a long time, it was first used for protein structure determination with the advent of synchrotron radiation, which is tunable in wavelength, and Hendrickson et al. Successful (see Hendrickson WA et al., Proteins 4, 77-88, 1988 and Hendrickson WA et al., EMBO J. 5, 1665-1672, 1990).
The synchrotron radiation light is radiation light obtained by accelerating electrons injected into the storage ring by a magnetic field, and has much stronger energy than a normal X-ray generator. The third generation large-scale synchrotron radiation facility SPring-8 (8 GeV) of RIKEN Harima Institute can be used.
The measured diffraction image data is processed into reflection intensity data using a program DENZO (Mac Science) or a similar image processing program. After obtaining the position of the heavy atom bonded to the protein in the crystal using the difference Patterson diagram from the reflection intensity data of the natural crystal and the reflection intensity data of the heavy atom substitution crystal at a plurality of wavelengths, the program PHASES or CCP4 or The initial phase is determined by refining the heavy atom position parameters using a similar diffraction data analysis program. The determined initial phase is improved to a reliable phase according to a solvent smoothing method using program DM (CCP4 package) or similar phase improvement program (electron density improvement program).
A three-dimensional structure model of a complex of a DNA replication regulatory protein and hemimethylated DNA can be constructed from the electron density diagram displayed on the three-dimensional graphics by the program O by the following procedure. First, a plurality of regions having characteristic amino acid sequences are searched for on an electron density map. Next, a partial structure of amino acid residues suitable for the electron density is constructed on the three-dimensional graphics using the program O while referring to the amino acid sequence starting from the found region.
By sequentially repeating this operation, all amino acid residues of the complex of DNA replication regulatory protein and hemimethylated DNA are adapted to the corresponding electron density, and an initial three-dimensional structure model of the entire complex is constructed. The constructed three-dimensional structure model is used as a starting model structure, and the three-dimensional coordinates describing the three-dimensional structure are refined according to the refinement protocol of XPLOR which is a structure refinement program. The progress of refinement and the check of the molecular model can be confirmed by the following indicators. The reliability factor (R factor) is a measure of the degree of coincidence between the absolute value of the crystal structure factor obtained from the diffraction intensity measurement and the absolute value of the crystal structure factor calculated from the molecular model. Small value. In addition, a method (Ramachandran plot) of calculating dihedral angles of the protein main chain and plotting them to check whether or not those values are in an allowable region is also used.
In this way, the three-dimensional atomic coordinates of the complex of the DNA replication regulatory protein of the present invention and hemimethylated DNA (value indicating the spatial positional relationship of each atom) are obtained for the first time. The obtained atomic coordinates are shown in Table 1 according to the protein data bank text format, which is a notation commonly used by those skilled in the art.

The atomic coordinates shown in Table 1 are the crystal structure of a complex of E. coli SeqA protein (71-181) and 10 base pair hemimethylated DNA (SeqA bsC-10). Thus, the three-dimensional structure of the SeqA protein possessed by other bacteria related to E. coli can be easily determined. Examples of bacteria related to Escherichia coli include the genus Escherichia, the genus Haemophilus, the genus Pasteurella, and the genus Vibrio. As shown in FIG. The homology between the E. coli SeqA protein and the amino acid sequence is very high. As a method for determining the three-dimensional structure of such highly homologous proteins, there is a molecular replacement method. After preparing a crystal of a complex of the SeqA protein of the related bacterium and hemimethylated DNA by the same method as the present invention, obtaining X-ray diffraction data, determining the three-dimensional structure of the related protein by molecular replacement method These are also included as one embodiment of the present invention.
(Computer system that displays a three-dimensional model of a complex of DNA replication regulatory protein and hemimethylated DNA)
Atomic coordinates obtained from a crystal of a DNA replication regulatory protein according to the present invention or a complex of the protein and hemimethylated DNA are input to a computer on which a computer program that expresses a three-dimensional structure of a molecule operates or a storage medium of the computer By doing so, it becomes possible to express in detail the three-dimensional structure of a complex of a DNA replication regulatory protein and hemimethylated DNA.
FIG. 2 is a schematic diagram illustrating one embodiment of these computer systems. Referring to FIG. 2, a storage means, a derivation means, and an output means are shown as the computer system of the present invention. As the storage means, any computer-readable data including the atomic coordinates of amino acid residues corresponding to the binding site of DNA replication regulatory protein with hemimethylated DNA can be read and stored by the derivation means. It is not particularly limited. For example, it is realized by storing the data in a storage medium such as a flexible disk, a hard disk, a CD-ROM, an optical disk, a magneto-optical disk, or a magnetic tape.
Various methods can be used for inputting computer-readable data including the atomic coordinates of amino acid residues corresponding to the binding site of DNA replication regulatory protein to hemimethylated DNA into the storage means. For example, data recorded on these recording media can be read from a flexible disk drive or a CD-ROM drive. Alternatively, desired data can be received from a database such as a PDB over a network.
Based on the data read from the storage means, the derivation means derives a three-dimensional structure model representing at least a part of the complex of DNA replication regulatory protein and hemimethylated DNA. The configuration of the deriving means is not particularly limited, but QUANTA (Molecular Simulations), SYBYL (Molecular Modeling Software), AMBER (Weiner, SJ, et al., (1984) J. Am. Chem. Soc., 106, 765-784) or a program such as CHARMM (Brooks, BR et al., (1983) J. Comp. Chem. 4, 187-217) is installed on a desired computer. This is realized. Further, following the step of generating these molecular models, the above CHARMM, AMBER, etc. are used as a program for calculating the conformational energy of the complex and minimizing the energy.
In response to the result derived by the deriving means, the output means for outputting the three-dimensional structure model of the binding site of the DNA replication regulatory protein with the hemimethylated DNA may use various conventionally used devices. it can. For example, the image can be displayed on a display device called a CRT using a program such as QUANTA. Also included are printers that display such images on paper and disk drives for storing in other storage media.
FIG. 3 shows the three-dimensional structure of the complex of Escherichia coli SeqA protein and hemimethylated DNA obtained in this manner with a resolution of 3.0 mm in a ribbon model. As is clear from FIG. 3, the obtained crystal is a dimeric crystal in which a complex of SeqA (71-181) protein and 10 base pair hemimethylated DNA is symmetric about the two-fold rotation axis. The domain structure of SeqA (71-181) protein confirmed from the three-dimensional structure is shown in FIG. As shown in FIG. 4, the SeqA (71-181) protein has 7 α-helical structures and 2 β-sheet structures, and includes a loop region including glycine at position 115 to arginine at position 118. , And a loop region containing threonine at position 149 to threonine at position 153, in the hemimethylated DNA G (^m6A) It turns out that it couple | bonds with the TC arrangement | sequence part (refer FIG. 3).
(Use of 3D structure)
By inputting all or part of the atomic coordinates of the complex of DNA replication regulatory protein and hemimethylated DNA provided by the present invention into a computer or a storage medium thereof, on which a computer program that expresses the three-dimensional structure of the molecule operates A compound that binds to the binding site of DNA replication regulatory protein with hemimethylated DNA and inhibits the formation of the complex, or has a three-dimensional structure similar to the binding site of DNA replication regulatory protein with hemimethylated DNA, It becomes possible to identify, search, evaluate, or design a compound that binds strongly to the immobilized DNA. The compound may be either a natural product compound or a synthetic compound, and may be either a high molecular compound or a low molecular compound.
In such a method, a candidate compound and a DNA replication regulatory protein are first displayed by a computer model, and a candidate compound having a shape or chemical structure that fits well with the binding site of the DNA replication regulatory protein to hemimethylated DNA by visual observation. The step of finding (lead compound) is included. In order to create and display a molecular model, programs such as QUANTA and SYBYL can be used. Furthermore, various computer programs for estimating affinity, repulsive force, steric hindrance, and the like of the DNA replication regulatory protein with the binding site with hemimethylated DNA can be used. Examples of these programs include, for example, GRID (Goodford, PJ, (1985) J. Med. Chem., 28, p. 849 for determining an energetically suitable binding site for a biopolymer. -857), MCSS for searching for functional binding sites (Miranker, A., et al., (1991) Proteins: Structure, Function and Genetics, 11, p. 29-34), or macromolecules and ligands. AUTODOCK (Goodsell, DS, et al., (1990) PROTEINS: Structure, Function and Genetics, 8, 195-202), and DOCK ( untz, I.D., et al., include (1982) J.Mol.Biol.161,269-288) or the like.
Once these candidate compounds (lead compounds) are found, the three-dimensional interaction with the candidate compounds is then visually examined on a computer screen displaying the binding sites of DNA replication regulatory proteins with hemimethylated DNA. Can be further optimized. A model for optimizing candidate compounds can be constructed using a program such as QUANTA or SYBYL described above. In addition to these, various useful programs for the optimization of molecular modeling are available, such as CAVEAT (Lauri, G., and Bartlett, PA, J. (1994) Compute. Aided Mol. Des. 8, 51-66) and HOOK (Eisen, MB, et al. (1994) Proteins: Struct., Funct., Genet., 19, 199-221).
Further, in addition to the above program, it is also possible to automatically construct a candidate compound directly from the three-dimensional structure of the binding site of DNA replication regulatory protein to hemimethylated DNA (de novo) and output the structure. For these, LUDI (Bohm, HJ, (1992) J. Comp. Aid. Molec. Design.), LEGEND (Nishibata, Y., et al., (1991) Tetrahedron 47, 8985), etc. Other programs are also available.
Subsequently, whether or not the candidate compound or the optimized compound inhibits the binding between the DNA replication regulatory protein and the hemimethylated DNA is confirmed by various methods. For example, it is confirmed whether or not the DNA replication regulatory protein can bind to hemimethylated DNA by bringing the two into contact with each other in an aqueous solution, followed by electrophoresis or the like. If the candidate compound or the optimized compound can inhibit the binding between the DNA replication regulatory protein and the hemimethylated DNA, it is considered that the replication regulatory ability by the DNA replication regulatory protein is also inhibited.
Alternatively, the interaction between the candidate compound and the DNA replication regulatory protein can be detected spectroscopically using NMR or mass spectrometry. In general, NMR signals sensitively reflect changes in the local environment of the molecule (ligand binding, structural changes, etc.), so when NMR spectra are measured by mixing proteins that interact with ligands, Broading and disappearance, changes in chemical shift values, etc. are confirmed. On the contrary, the structure of a protein in a state bound to a ligand can also be specifically detected.¹⁵Using proteins uniformly labeled with N¹⁵N-¹By using the HHSQC spectrum, only the NMR signal of the protein can be observed, and local structures such as amino acid residues that interact with the ligand can be found.
Next, in order to identify a compound that specifically binds to hemimethylated DNA, the three-dimensional structure of the region involved in the binding of DNA replication regulatory protein to hemimethylated DNA is an important clue. As described above, the region from the glycine at position 115 to the arginine at position 118 and the region from threonine at position 149 to threonine at position 153 of the SeqA protein shown in SEQ ID NO: 2 are based on the atomic coordinates shown in Table 1. The G of hemimethylated DNA by forming the three-dimensional structure shown in FIG.^m6A) It turns out that it is interacting with the TC arrangement | sequence part. In normal double-stranded DNA, the base moiety is hydrogen bonded to the complementary strand, so it exists inside the double helix structure. However, in hemimethylated DNA, the amino group at position 6 of adenine is methylated. It is considered that the hydrogen bond between adenine and thymine is weakened and the double helix structure is partially opened. It can be seen that the DNA binding site of the SeqA protein recognizes this subtle structural fluctuation and specifically binds to it. The three-dimensional structure information relating to the binding to hemimethylated DNA is an important clue for identifying a compound that binds to hemimethylated DNA.
One such candidate compound is a mutant DNA replication regulatory protein. For example, the two loop regions that bind to hemimethylated DNA (the region from glycine at position 115 to arginine at position 118, and the region from threonine at position 149 to threonine at position 153), and the three-dimensional structure of the binding portion Any protein or peptide that retains The three-dimensional structure of the bonding portion is substantially defined by the atomic coordinates shown in Table 1. Alternatively, the amino acid residue forming the DNA binding site of the SeqA protein or an amino acid in the vicinity thereof is substituted with another amino acid or chemically modified to produce a mutant protein having improved binding ability to hemimethylated DNA. can do. Here, the “vicinity region” refers to a region involved in electrostatic interaction, hydrophobic interaction, van der Waals interaction, hydrogen bonding, etc., specifically within about 5 mm with respect to the amino acid residue. An area in
Mutant proteins designed in this way can be prepared by a number of methods. For example, DNA encoding the mutant of the present invention can be prepared by methods known to those skilled in the art such as site-directed mutagenesis, PCR, chemical synthesis, or a combination thereof. It is possible to introduce a DNA encoding these mutant proteins into an appropriate expression vector, and further culture the transformant obtained by introducing the expression vector into an appropriate cell to express the mutant protein. is there.
The inhibitor designed in the above method can be obtained by using a commonly used chemical synthesis method according to the compound. In addition, when the inhibitor is used as a growth inhibitor of pathogenic bacteria having Dam methylase, a pharmaceutical composition can be generally prepared by mixing with a pharmaceutically acceptable carrier. The carrier for the pharmaceutical composition is not particularly limited. For example, lactose, glucose, D-mannose, starch, crystalline cellulose, calcium carbonate, kaolin, gelatin, hydroxypropyl alcohol, polyvinylpyrrolidone, ethanol, carboxymethylcellulose, carboxymethylcellulose sodium Salt, magnesium stearate, talc, acetylcellulose, sucrose, titanium oxide, benzoic acid, parahydroxybenzoate air steal, sodium dehydroacetate, gum arabic, tragacanth, methylcellulose, egg yolk, surfactant, simple syrup, citric acid, distilled water, There are glycerin, propylene glycol, macrogol, sodium monohydrogen phosphate, sodium dihydrogen phosphate, sodium phosphate, sodium chloride, etc. Depending on the form of and tablets such as are used in admixture with inhibitors identified by the methods of the present invention.
Example
Example 1 Preparation of various deletion mutants of E. coli SeqA protein
E. coli seqA gene (Full length 543 bp) shown in SEQ ID NO: 1 is subjected to polymerase chain reaction (PCR) using plasmid pMY903 inserted into plasmid pET15b (Novagen) as a template, and a fusion protein with glutathione S transferase (GST) A variety of deletion mutant plasmids that express the above were prepared.
First, DNA fragments encoding various deletion mutants of the SeqA protein shown in FIG. 1A were amplified by PCR using the plasmid pMY903 as a template. Each DNA fragment was amplified using two types of primers each having a BamHI restriction enzyme site added to the N-terminal side of the protein encoded by each DNA fragment and an EcoRI restriction enzyme site added to the C-terminal side, and recombinant Taq polymerase (TOYOBO). Thereafter, the amplified DNA fragment was purified using a PCR purification kit (QIAGEN). A single-base deoxyadenosine (A) is added to each 3 ′ end of the amplified double-stranded DNA fragment regardless of the template sequence, so that a single-base thymidine (T) complementary to this is added. TA was cloned into the plasmid pGEM-TEasy (registered trademark, PROMEGA) having a protruding end of After confirming the base sequence of this cloned DNA, each DNA fragment excised with EcoRI and BamHI was cloned into the EcoRI-BamHI site of GST fusion protein expression vector pGEX6P-1 (Amersham Pharmacia Biotech) and SeqA protein Various plasmids were constructed in which GST was fused and expressed at the N-terminus of various deletion mutants.
Escherichia coli (XL1-blue) having full-length SeqA and various deletion mutant genes prepared by the above method is cultured in LB medium, and IPTG is added to a final concentration of 1 mM in the logarithmic growth phase to express the protein. Induced. Thereafter, the cells were collected by culturing for 4 to 6 hours, and the cells were disrupted by sonication in a buffer solution (pH 8.0) containing 50 mM Tris-HCl, 0.5 M NaCl, and 10% glycerol under ice-cooling. The expressed protein was recovered. On the other hand, plasmid DNA was extracted from 3 ml of a culture solution cultured overnight in the same medium and purified using mini-prep (QIAGEN). The plasmid thus obtained was digested with restriction enzymes BamHI and EcoRI and then subjected to agarose gel electrophoresis to determine the yield of plasmid DNA. It was found that the clone from which the soluble protein was recovered from the cell disruption solution had a significantly poor yield of the plasmid encoding the protein. The plasmid used in this experiment is a colicin E1 (ColE1) plasmid, and unlike E. coli EriC, there is another replication mechanism controlled by RNAI and RNAII. From the above results, SeqA protein is also It was suggested to be involved in this replication mechanism.
(Example 2) Purification of SeqA deletion mutant
Among the deletion mutants prepared in Example 1, SeqA (1-59), SeqA (7-181), SeqA (18-181), SeqA (42-181), SeqA (60-181), SeqA ( 71-181) 6 mutants were selected and their proteins were purified.
Escherichia coli (XL1-blue) expressing each mutant SeqA protein was cultured in 2.5 liters of LB medium and expressed in the same manner as in Example 1 to prepare a cell disruption solution. This was added to a 5 ml GSTrapFF column (Amersham Pharmacia Biotech), adsorbed, washed with the starting buffer, and then 50 mM Tris-HCl, 0.5 M NaCl, 10% glycerol having a gradient of 0-20 mM glutathione. The GST fusion protein was eluted using a buffer solution (pH 8.0) containing. The eluted fraction was dialyzed to remove glutathione, and then reacted with 1 unit (U) of PreScission protease (Amersham Pharmacia Biotech) per 100 μg of protein at 4 ° C. for 12 to 18 hours to cleave the GST tag. . Subsequently, in order to remove the GST tag, the reaction solution was again added to a 5 ml GSTrapFF column, and a non-adsorbed fraction (Flow through fraction) was collected. Furthermore, 6 kinds of proteins were purified by Superdex 75 column.
The full-length SeqA protein expressed as a control almost aggregated when the GST tag was cut by the above method, and only a trace amount could be purified. SeqA (7-181) and SeqA (18-181) tended to decompose when the GST tag was cut. It was suggested that SeqA (1-59) exists as a multimer. From these results, it was found that dimerization of the SeqA protein occurs via the N-terminal region. When the CD spectrum was measured, it was shown that all proteins retained higher-order structures.
(Example 3) Measurement of binding activity to hemimethylated DNA by gel shift assay
The activity of the various deleted SeqA proteins purified in Example 2 to bind to hemimethylated DNA was measured by a change in mobility (gel shift assay) by gel electrophoresis. Various hemimethylated or fully methylated DNAs were added to 20 μl of a binding solution (10 mM Tris-HCl (ph7.5), 1 mM EDTA, 10% glycerol) containing 4 μg of mutant SeqA, and left at 25 ° C. for 30 minutes or more. After that, electrophoresis was performed with 1 × TBE, 12% acrylamide gel at a constant voltage of 50 V, and the gel was stained with ethidium bromide.
First, to date, SeqA protein has more than two hemimethylated G (^m6A) Since it was known to bind to the TC sequence, the six types of single-stranded oligonucleotides shown in SEQ ID NOs: 3 to 8 were synthesized, and the complementary strands were annealed, respectively. Was synthesized.

As the fully methylated and hemimethylated DNA, oligonucleotides obtained by synthesizing and annealing an oligonucleotide in which the N6 position of adenine of the GATC sequence was methylated were used. For example, when the adenine of the GATC sequence of SEQ ID NO: 3 (the part enclosed by a frame) is annealed with a methylated single-stranded oligonucleotide and the unmethylated oligonucleotide of SEQ ID NO: 4, hemimethylated DNA (SeqA bsA-37 (M⁺/ M⁻). ) Is prepared. On the other hand, when the single-stranded oligonucleotide of SEQ ID NO: 6 is methylated and annealed with the unmethylated oligonucleotide of SEQ ID NO: 5, hemimethylated DNA (SeqA bsB-26 (M⁻/ M⁺). ) Can be prepared. The results of the gel shift assay performed using SeqA bsC-14 and various deletion mutant SeqA proteins are shown in FIG. In the figure, the DNA used in each lane 1-4 is 1: M⁺/ M⁺2: M⁺/ M⁻3: M⁻/ M⁺4: M⁻/ M⁻The numbers shown below for each of the electrophoresis patterns indicate the amino acid residues constituting the mutant SeqA protein. FIG. 5 shows that there is no DNA binding ability only in the N-terminal region (positions 1 to 59) of the SeqA protein. On the other hand, even a mutant lacking 7 to 70 amino acids from the N-terminus has the ability to bind hemimethylated DNA. That is, it was found that the DNA binding region of the SeqA protein is the 71st and subsequent C-terminal regions from the N-terminus. It was also found that SeqA can bind to hemimethylated DNA even at one hemimethylated site.
(Example 4) Crystallization and collection of X-ray crystal structure analysis data
From the results of Example 3, mutant SeqA (71-181) protein lacking the largest N-terminal portion was purified for X-ray crystal structure analysis, and a complex crystal with hemimethylated DNA was prepared.
In the same manner as in Example 2, mutant SeqA (71-181) protein fused with GST was recovered from 5 liters of the culture solution, adsorbed on a 5 ml GSTrapFF column, and eluted with a 0 to 20 mM glutathione concentration gradient. Thereafter, glutathione was removed by dialysis and the GST tag was removed by PreScience protease. Subsequently, the GST tag was removed through a 5 ml GSTrapFF column, and then the mutant SeqA (71-181) protein was further purified by a Superdex75 column.
When this protein was mixed with double-stranded hemimethylated DNA of various lengths at a molar ratio of 1.5: 1, concentrated and crystallized, hemimethylated double-stranded DNA consisting of the following 10 base pairs: It has been found that the complex with the above forms a hexagonal columnar crystal.

The SeqA (71-181) protein and SeqA bsc-10 (M⁺/ M⁻) Was crystallized by a hanging drop method in a solution containing 0.1 M HEPES-Na (pH 7.5) and 1.4 M trisodium citrate hydrate. The obtained crystal was a dimer in which a complex of SeqA (71-181) protein and 10-base-pair hemimethylated DNA was symmetric about the 2-fold rotation axis, and was a highly symmetrical hexagonal P622. It was. The lattice constants were a = b = 152.238 Å, c = 119.338 Å, and the unit cell was α = β = 90 ° and γ = 120 °.
Data sets of both natural and selenomethionine substituted crystals were collected at the beamline I (BL41XU) of the 8 GeV large synchrotron radiation facility (SPing-8) at RIKEN Harima Institute.
Using data obtained from selenomethionine heavy atom substitution crystal, phase determination is performed by multi-wavelength anomalous scattering method (MAD method), and extension (Phase extension) is performed using natural crystal data. An electron density map of 3.0 mm was obtained. The graphic display of the three-dimensional structure was performed using the programs MOLSCRIPT and SPOCK.
(Example 5) In silico screening of compounds that bind to the binding site of SeqA protein to hemimethylated DNA
<Database optimization>
As a low molecular compound database, a low molecular compound catalog database provided by SPECS was used. Those containing multiple molecules in one entry were divided into molecules, and the library excluding duplicates was used as the screening population (152323 molecules). Next, based on “Lipinski's Rule of 5”, target-independent optimization for the low molecular compound database was performed. The narrowing conditions used here are as follows.
1. Molecular weight 100 to 500
2. Calculated LogP value (o / w) 5 or less (using XLOGP-1 algorithm),
3. Number of hydrogen bond acceptor atoms (number of N and O contained in low molecular weight compound) 10 or less
4). Number of hydrogen bond donor atoms (number of NH and OH in low molecular weight compound) 5 or less
Furthermore, the following molecules were excluded in order to perform docking calculation appropriately.
1. The number of rotatable single bonds is 21 or more
2. Containing radicals
3. Containing elements other than H, C, N, O, F, S, P, Cl, Br, and I.
The number of molecules after narrowing was 103773.
<Binding site prediction>
Among the SeqA proteins of E. coli, a residue group presumed to be essentially involved in the binding to hemimethylated DNA shown in FIG. 3 (from glycine at position 115 to position 118 in the amino acid sequence shown in SEQ ID NO: 2). Arginine and a loop region comprising threonine at position 149 to threonine at position 153) were used as binding sites.
<screening>
(1) Primary screening
Dock 4.0 was used to dock the entire low-molecular compound library previously optimized for the binding site. At this time, the low molecular weight compounds were ranked based on Dock's energy score.
(2) Secondary screening
Detailed docking was performed using AutoDock 3.0.5 on the 11171 molecules of the top 10,000 molecules obtained as a result of the primary screening and the similar compounds contained in the database (Sybyl Finger Print tanmoto similiarity 0.8 or less). Thereafter, the compounds within 4 mm of the sphere used for docking were narrowed down, and the molecules for which calculation was abnormally terminated were removed. As a result, a docking structure was finally obtained for 9961 molecules. Among the obtained compounds, the top 100 compounds having high binding properties to the above-mentioned binding sites were used as ligand candidate compounds. This ligand candidate compound had a ΔG (change in binding free energy) of −9.8 kcal / mol or more.
(Example 6) Screening of compounds that bind to the binding site of SeqA protein to hemimethylated DNA by NMR
Each ligand candidate compound obtained in Example 5 was dissolved in DMSO-d6 and adjusted to 1 mM. The SeqA (71-181) protein prepared in Example 2 was 103.7 μM in concentration using Buffer A (20 mM Tris-d11 (pH 7.5) /0.5 M NaCl dissolved in heavy water). Dissolved in. Using these, two types of samples were prepared: a ligand candidate compound alone and a SeqA (71-181) protein added thereto at a ratio of 1: 1.

The final concentration of the ligand candidate compound is 50 μM.

The final concentrations of the ligand candidate compound and the SeqA protein are 50 μM, respectively.
For each of the above samples, one-dimensional using a Bruker DRX500 NMR spectrometer at 298K in deuterium.¹H-NMR was measured, and it was observed whether there was any change in the spectrum before and after the addition of SeqA protein. As a result, as a ligand candidate compound, sample ID: AK968-41170335, compound name: cis-6- [4- (4-Isopropyl-phenyl) -5-methyl-thiazol-2-ylcarbamoyl] -3,4-dimethyl- FIG. 6 shows an NMR spectrum (enlarged signal around 7.5 ppm) in the case of using cyclohex-3-enecarboxylic acid (compound A, 10.47 kcal / mol). FIG. 6 (a) is Compound A alone, FIG. 6 (b) is a mixture of Compound A and SeqA protein, and FIG. 6 (c) is a comparison of overlapping signals (a) and (b). is there. From these results, it was recognized that signals derived from the four hydrogens of the benzene ring of compound A disappeared due to the interaction with SeqA, suggesting that the compound was bound to SeqA.
Using such a method, the NMR spectrum of the 100 kinds of ligand candidate compounds obtained in Example 5 with respect to solubility in DMSO and detectable signals by NMR was measured. As a result, interaction with SeqA was found. The detected compounds are shown in Table 4 below.

Among the compounds shown in Table 4, the chemical structures of the compounds A, B, C, D, E, G, H, and J are similar to each other, and therefore, they are represented by the following general formula (I) common to these compounds. It is speculated that the compound specifically binds to the hemimethylated DNA binding site of SeqA.

A is any one selected from the group consisting of cyclohexene, dimethylcyclohexene, bicyclo [2,2,1] heptane, bicyclo [2,2,1] heptene, and oxa-bicyclo [2,2,1] heptane. The seed ring is R₁And R₂Is a hydrogen atom or a methyl group, R₃Represents a hydrogen atom or an alkyl group having 1 to 4 carbon atoms. These compounds may be in the form of a salt (sodium salt, potassium salt, ammonium salt, etc.) because they bind to SeqA protein under intracellular physiological conditions, and have an asymmetric carbon atom in the molecule May be optically active or racemic.
Further, in Table 4 above, NMR spectra were measured when compounds F and I having relatively different chemical structures from compound A were used. FIG. 7 shows the results when compound F was used. FIG. 7 (a) is Compound F alone, FIG. 7 (b) is a mixture of Compound F and SeqA protein, and FIG. 7 (c) is a comparison of overlapping signals (a) and (b). is there. As in the case of Compound A, the characteristic signal of Compound F disappeared by mixing with SeqA protein, suggesting that Compound F also interacts with SeqA protein.
Furthermore, when Compound I was used, an NMR spectrum having the same characteristics as described above was obtained, suggesting an interaction with SeqA protein.
Industrial applicability
According to the present invention, a three-dimensional structure model of a complex of a DNA replication regulatory protein and hemimethylated DNA is provided. The three-dimensional structure of this complex is extremely useful for understanding the mode of protein binding to hemimethylated DNA, and further, the regulatory mechanism of chromosomal DNA replication by the complex. Since methylation of GATC sequences by Dam methylase is a structure possessed only by a special group of bacteria closely related to Escherichia coli, the growth of the bacteria is specifically identified by inhibiting the replication control by the DNA replication regulatory protein of the present invention. May be adjustable. The three-dimensional structure of the DNA replication regulatory protein revealed by the present invention can be used as a screening method for developing these drugs.
[Sequence Listing]

[Brief description of the drawings]
FIG. 1 is a diagram schematically showing the structures of various mutant SeqA proteins synthesized in the present invention. (A) shows the primary structure of 20 mutant SeqA proteins, and (b) shows the functional domain of the SeqA protein.
FIG. 2 is a schematic diagram showing an embodiment of a computer system that displays a three-dimensional structure model of a complex of a DNA replication regulatory protein and hemimethylated DNA.
FIG. 3 is a model diagram showing a three-dimensional structure of a complex crystal of a mutant SeqA protein (71-181) and 10 base pair hemimethylated DNA.
FIG. 4 shows the partial structure of the SeqA protein possessed by various bacteria including E. coli in parallel in consideration of the homology of the primary structure. In the figure, the helical structure indicates an α-helix, and the arrow indicates a region forming a β-sheet.
FIG. 5 shows the results of a gel shift assay of 14 base pair oligonucleotide (SeqA bsC-14) and various mutant SeqA proteins. The mutant SeqA proteins used in the experiments (a) to (f) are (a): 1-59, (b): 7-181, (c): 18-181, (d): 42-, respectively. 181, (e): 60-181, (f): 71-181 amino acid residues. Each lane of 1-4 is 1: M⁺/ M⁺2: M⁺/ M⁻3: M⁻/ M⁺4: M⁻/ M⁻Of full methyl, hemimethyl or unmethylated DNA.
FIG. 6 is a proton NMR spectrum (signal around 7.5 ppm) of a ligand candidate compound AK-968-41170335 (compound A) alone (a) and a 1: 1 mixture (b) of compound A and SeqA protein. . Note that (c) is a comparison of the signals (a) and (b).
FIG. 7 is a proton NMR spectrum (signal around 7.5 ppm) of the ligand candidate compound AN-989-14072018 (compound F) alone (a) and a 1: 1 mixture (b) of compound F and SeqA protein. . Note that (c) is a comparison of the signals (a) and (b).

Claims (19)

Of the amino acid sequence shown in SEQ ID NO: 2, a polypeptide fragment consisting of the amino acid sequence from positions 71 to 181 and from glycine at position 115 to arginine at position 118 and from threonine at position 149 to threonine at position 153 A polypeptide fragment, characterized in that the amino acid residues form a DNA binding site. Of the amino acid sequence shown in SEQ ID NO: 2, a polypeptide fragment consisting of the amino acid sequence from positions 71 to 181 and from glycine at position 115 to arginine at position 118 and from threonine at position 149 to threonine at position 153 A polypeptide fragment characterized in that the amino acid residue forms a binding site with hemimethylated DNA. A crystal of a complex of a polypeptide fragment consisting of positions 71 to 181 of the amino acid sequence shown in SEQ ID NO: 2 and hemimethylated DNA, which is a hexagonal space group P622, lattice constant a = b = 152. Crystals of the complex having 2 ± 0.2Å, c = 119.3 ± 0.2Å. A method for producing a crystal of a complex of a DNA replication regulatory protein consisting of the amino acid sequence shown in SEQ ID NO: 2 and hemimethylated DNA,
A solution prepared by mixing a DNA replication regulatory protein having a concentration of 4 to 6 mg / ml and hemimethylated DNA in a molar ratio of 1: 1 to 2: 1 contains 1-2 M trisodium citrate hydrate, A method of crystallizing by a vapor diffusion method using a buffer solution having a pH of 7.0 to 8.0. A compound that binds to the polypeptide fragment according to claim 1 is designed or selected, and a compound that inhibits DNA replication regulation ability and / or a compound that specifically binds to hemimethylated DNA is screened. how to. A method for determining the three-dimensional structure of a DNA replication regulatory protein or a complex of said protein and hemimethylated DNA,
(A) producing a crystal of a DNA replication regulatory protein or a complex of the protein and hemimethylated DNA;
(B) obtaining X-ray diffraction data by the heavy atom isomorphous substitution method or multi-wavelength anomalous scattering method using the crystal, and (c) determining three-dimensional structure coordinates based on the X-ray diffraction data,
A method for determining a three-dimensional structure of a DNA replication regulatory protein or a complex of said protein and hemimethylated DNA, The method according to claim 6, wherein the DNA replication regulatory protein is a SeqA protein of a bacterium of the genus Escherichia, Haemophilus, Pasteurella or Vibrio. A method for identifying a compound that inhibits the ability to regulate DNA replication, comprising:
(A) constructing a DNA replication regulatory protein or a complex of the protein with hemimethylated DNA, or a three-dimensional structure of the polypeptide fragment according to claim 1 or a DNA binding site thereof using a computer model;
(B) designing or selecting a candidate compound that inhibits the ability to regulate DNA replication based on the three-dimensional structure;
(C) synthesizing or obtaining the designed or selected candidate compound;
(D) contacting the candidate compound with the DNA replication regulatory protein in order to examine the inhibitory activity on the ability of the candidate compound to regulate DNA replication;
A method for identifying a compound that inhibits the ability to regulate DNA replication. The method according to claim 8, wherein the DNA replication regulatory protein is a SeqA protein of a bacterium of the genus Escherichia, Haemophilus, Pasteurella or Vibrio. The method according to claim 8 or 9, wherein the inhibitory activity on the ability to regulate DNA replication in the step (d) is detected by gel shift assay. The method according to any one of claims 8 to 10, wherein the three-dimensional structure model is substantially defined by atomic coordinates shown in Table 1 of the present specification. A compound that inhibits the ability to regulate DNA replication, which is identified by the method according to any one of claims 5 to 8 or 11. A method for producing an antibacterial composition comprising a compound that inhibits the ability to regulate DNA replication, wherein the ability to regulate DNA replication by the method according to any one of claims 5 or 8 to 11. A method for producing an antibacterial composition, comprising the steps of: identifying a compound that inhibits the activity; and mixing the identified compound with a stabilizer or carrier. A method for identifying a compound that specifically binds to hemimethylated DNA, comprising:
(A) a step of constructing, using a computer model, a three-dimensional structure of a complex of a DNA replication regulatory protein or the polypeptide fragment of claim 2 and hemimethylated DNA;
(B) designing or selecting a candidate compound that binds to hemimethylated DNA based on the three-dimensional structure;
(C) synthesizing or obtaining the designed or selected candidate compound;
(D) contacting the candidate compound with hemimethylated DNA in order to examine the binding ability of the candidate compound to hemimethylated DNA;
A method for identifying a compound that specifically binds to hemimethylated DNA, comprising: The method according to claim 14, wherein the candidate compound is a mutant DNA replication regulatory protein. A computer system that outputs a three-dimensional structure of a DNA replication regulatory protein comprising the amino acid sequence shown in SEQ ID NO: 2 or a complex of the protein and hemimethylated DNA,
(A) storage means for storing and storing computer-readable data including atomic coordinates of amino acid residues corresponding to the binding site of the DNA replication regulatory protein with hemimethylated DNA;
(B) Deriving means for reading out the data held in the storage means and deriving a three-dimensional structure model;
And (c) an output unit that receives the derivation result and outputs the three-dimensional structure model. The binding site of the DNA replication regulatory protein to hemimethylated DNA consists of the region from glycine at position 115 to arginine at position 118 and threonine at position 149 to threonine at position 153 in the amino acid sequence shown in SEQ ID NO: 2. The computer system according to claim 16, wherein: A DNA replication regulation inhibitor comprising a compound represented by the following general formula (I) or a salt thereof as an active ingredient.

Wherein A is selected from the group consisting of cyclohexene, dimethylcyclohexene, bicyclo [2,2,1] heptane, bicyclo [2,2,1] heptene, oxa-bicyclo [2,2,1] heptane. Any one of the rings, R ₁ and R ₂ each independently represent a hydrogen atom or a methyl group, and R ₃ represents a hydrogen atom or an alkyl group having 1 to 4 carbon atoms). The antibacterial composition characterized by including the compound or its salt shown by the following general formula (I) as an active ingredient.

Wherein A is selected from the group consisting of cyclohexene, dimethylcyclohexene, bicyclo [2,2,1] heptane, bicyclo [2,2,1] heptene, oxa-bicyclo [2,2,1] heptane. Any one of the rings, R ₁ and R ₂ each independently represent a hydrogen atom or a methyl group, and R ₃ represents a hydrogen atom or an alkyl group having 1 to 4 carbon atoms).

Images