JP4730804B2 - Method - Google Patents

Description

【００４９】
第２のプール「オリゴミックスＢ」は、３組のオリゴヌクレオチドからなり、それぞれは、５’末端に存在する３つの停止コドンのうち１つを有し、その直後に６つの縮重した残基が続いている。各オリゴの残りの領域は、３つの組全てで同じであり、２つのＩＩＳ型制限酵素部位（ＢｐｍＩとＢｓｅＲＩ）を含有し、３’末端にプライマー「ＬＭＢ２」に対する相補的な認識配列が続いている。３組のオリゴの配列は以下のとおりである。
【００５０】
【表２】

【００５１】
ＥｘｏＩＩＩミックス＋オリゴミックスＡ又はオリゴミックスＢの何れかを、１６℃で３０分間アニールした後、４００ユニットのＴ４ ＤＮＡリガーゼを加え、その後１６℃で一晩反応液をインキュベートした。標準条件下で、ＱＩＡクイック ＰＣＲクリーンアップキット（Ｑｉａｇｅｎ）を用いて連結産物を精製し、プライマー「ＳＴフォワード」と「ＬＭＢ２」により、Ｐｗｏポリメラーゼを用いる標準的なＰＣＲ反応におけるテンプレートとして使用した。３０サイクルのＰＣＲ（９４℃、１分；５７℃１分；７２℃２分）を実施して、１０００ｂｐに及ぶＰＣＲ産物を生成させた。
【００５２】
オリゴミックスＡとオリゴミックスＢは何れも、前の工程で実施された二重鎖のうちの１つの鎖のＥｘｏＩＩＩによる加水分解によって露出された元のテンプレートの一本鎖ＤＮＡ領域に拮抗的にアニールすることができる。アニーリングに続いて、何れかのオリゴとテンプレートとの間で上手く連結が起こるためには、オリゴがその５’末端で絶対的な相補性をもってアニールし、さらに、前記二重鎖テンプレートの陥凹した３’残基が、特異的にアニールされたオリゴの５’残基と直接隣接していることを要する。新しく連結されたオリゴに結合する第１のプライマーと二重鎖テンプレートの５’末端に結合する第２のプライマーを用いたＰＣＲは、その後、このような連結を受けた二重鎖のみを選択的且つ特異的に増幅する。オリゴミックスＡ中の前記１２の各オリゴの５’末端は、停止コドンに対応するので、図１ｃに示されているように、このミックス中に含まれる１２のオリゴのうち、その５’末端に、ＧＳＴの第１のインフレーム停止コドンと該停止コドンのすぐ３’にある塩基とを含む４つの塩基対認識配列に対して絶対的な相補性を有する１つだけがアニールすることができる。残りの１１のオリゴは、それらの５’末端において、前記欠失物の入れ子状態のセットの内部の別の場所に完全にアニールし得るが、これら他の特異的なアニーリング現象は、ＧＳＴ遺伝子中のアウトオブフレームの停止コドン又は最初のインフレーム停止コドンの下流にある停止コドンでしか起こらない。新しくアニールしたオリゴがどこで、二重鎖テンプレートの３’陥凹した残基に直接隣接していても、連結が起こり得る。この段階のＰＣＲは、それ故、正確な、完全長の遺伝子のみならず、末端切断され、且つ伸長された産物のセットも増幅するであろう。オリゴミックスＢ中のオリゴは、同様に反応すると予想される。前記プール中の前記３組の各オリゴの５’末端は、停止コドンに対応し、効果的にランダムである６つの残基が続く。それ故、前記プールは、停止コドンと次の６つの残基が目的の遺伝子の下流にある残基と完全にマッチする１つの順列を含有するであろう。４と比べて、９つのヌクレオチド上に相補部分が伸長しているので、このオリゴは、１２のオリゴのプールから得られた対応するオリゴより高い特異性で結合するであろう。
【００５３】
オリゴミックスＡとオリゴミックスＢとの理論的な差は、各オリゴの５’末端にある停止コドンの直後にある配列中に存在する。オリゴミックスＡでは、単一の縮重する塩基の後にはナンセンスであるが、決まった単一のＤＮＡ配列が続いているので、この決まった領域は、各オリゴの５’末端での４つのデザインされた塩基対の相互作用を超えて、意図的にではないが塩基対の相補性をさらに与えることによって、何れかの特定の遺伝子内の各「停止」コドン（インフレームであるとアウトオブフレームであるとを問わない）を優先するように、オリゴミックスのアニーリングを偏らせることができる。アニーリング中の何れの偏りも、各停止コドン（インフレームであるとアウトオブフレームであるとを問わない）の特異的な切り出しと置換が起こったクローンに対する偏りの下流に現れ得る。このような異なる停止コドンにおける修飾の頻度の偏りは、望ましくないかもしれないので、オリゴミックスＢは、以下のように、これを回避するためにデザインされている。各遺伝子又はライブラリー内の停止コドンは何れも、そのすぐ後に、確定しているが、未知の配列が続いているであろう。３つの停止コドンは全て、オリゴミックスＢの中に提示されており、遺伝子又はライブラリー内の何れの停止コドンに対しても、前記停止コドンとその正確な下流の配列と正確にマッチするオリゴがオリゴミックスＢの中に１つ存在し、全体で９塩基対の相補性が得られるように、それぞれは、全ての可能なヘキサヌクレオチド配列（すなわち、ランダムヘキサマー配列によって）が直後に続いている。これは全ての停止コドンについて当てはまると思われるので、オリゴミックスＢは、それ故、オリゴミックスＡで起こり得るこの種の偏りからは、何れも解放されるはずである。
【００５４】
我々は、前記オリゴミックスのＥｘｏＩＩＩ消化、アニーリング、及び連結、並びに特異的なＰＣＲ増幅の全プロセスが、高度に再現性を有することを見出した。この操作における対照として、エキソヌクレアーゼＩＩＩ、Ｔ４ ＤＮＡリガーゼ、又は何れかのオリゴミックスのうち何れか１つを省略すると、ＰＣＲ産物が全く得られないことを示した。このことは、このプロセスが高度に選択的であることを示している。
【００５５】
Ｔ４ ＤＮＡリガーゼは、異なる特異性を示し得る多数の異なるＤＮＡリガーゼ、例えば、ＴａｑＤＮＡリガーゼ、又はＴｓｃ ＤＮＡリガーゼによって置き換え得ることも理解できるであろう。
【００５６】
（ｄ）操作の変形
明らかに、上記の操作において、３’陥凹した欠失物の必要な入れ子状態のセットを作成するために、多数の異なる３’から５’へのヌクレアーゼ活性を使用することができる。これらには、エキソヌクレアーゼＩＩＩ、Ｅ．ｃｏｌｉ ＤＮＡポリメラーゼＩ、Ｔ４ ＤＮＡポリメラーゼ、Ｔ７ ＤＮＡポリメラーゼが含まれるが、これらに限定されることはない。
【００５７】
停止コドン又は開始コドンが正確に除去された完全長の挿入物を作成するための前記元の操作に対する明確な変形には、コード配列を包含するＰＣＲ産物中に、５’が陥凹した欠失物の入れ子状態のセットを製造することが含まれる。これは、λエキソヌクレアーゼ、Ｅ．ｃｏｌｉ ＤＮＡポリメラーゼＩ、Ｔａｑ ＤＮＡポリメラーゼ、Ｔ７遺伝子６エキソンクレアーゼのようなヌクレアーゼを用いて行うことができるであろう。このように、例えば、前記元のＰＣＲ産物は、５’から３’方向にｄｓＤＮＡの１つの鎖からヌクレオチドを除去するλエキソヌクレアーゼを用いて消化することができる。前記酵素は、５’リン酸基のみを基質として認識するので、２つの鎖のうち１つは、最初のＰＣＲ増幅において適切なプライマーの末端に５’水酸基を取り込むことによって保護され得る。欠失物の入れ子状態のセットは、このように、５’が陥凹した末端を残すように反対鎖が消化されることを除き、エキソヌクレアーゼＩＩＩについて上記したように作成することができる。続いて、停止コドンの相補部分が前記オリゴの３’末端に存在し、すぐ直前に６つのランダム化された残基が先行し、その前にＩＩＳ型制限部位をコードする所定配列が先行するオリゴの混合物を、前記露出した１本鎖領域にアニールすることができる。前記混合物からの１つのオリゴは、各露出された停止コドンに特異的にアニールし、５’から３’の方向にポリメライズするＥ．ｃｏｌｉのＤＮＡポリメラーゼＩに対するプライマーとした働き、その５’から３’へのエキソヌクレアーゼ活性で、その前にある鎖を消化するであろう。この新しい二重鎖ＤＮＡ断片は、続いて、元のＰＣＲ産物の５’末端に結合するプライマーと新しくアニール、伸長されたオリゴに特異的に結合するプライマーとを用いて、特異的に増幅することができる。この操作では、前記アニールされたオリゴの３’末端は、二重鎖テンプレートの５’が陥凹した残基に直接隣接することを要しない。前記混合物からのオリゴは、ランダムなヘキサマープライミングと類似した態様で、前記露出した１本鎖ＤＮＡテンプレート上の任意の相補領域にアニールすることができるが、重要なことに、鎖の伸長は、３’末端が絶対的な相補性をもってアニールし、前記混合物中のオリゴの３’末端が停止コドンのみに相補的であるようにデザインされている場合だけに起こるであろう。このように、オリゴの３’末端が一致する停止コドンに特異的に結合するときだけ、プライマーの伸長とその後のＰＣＲ増幅が起こるであろう。アウトオブフレームの停止コドンと目的の遺伝子の真の停止コドンの下流にあるコドンへの結合も起こり、正確な完全長の遺伝子のみならず、一群の末端切断され、伸長された産物のＰＣＲ増幅が得られる。
【００５８】
（ｅ）ＰＣＲ産物のクローニングと分析（図１ｄ参照）
ＱＩＡｑｕｉｃｋ ＰＣＲ精製キット（Ｑｉａｇｅｎ）を用いて、サイズが８００〜１０００ｂｐにわたるＰＣＲ産物（約５μｇ）を精製した後、ＩＩＳ型制限酵素ＢｐｍＩを用いて完全に消化した。この酵素は、一方の鎖中に存在するその認識配列から離れた１４の塩基と他方の鎖上の１６ｂｐを遠隔的だが特異的に切断して、３’陥凹した末端を残す。このため、この制限修飾は、前記ＰＣＲ産物から停止コドン（前記「オリゴミックス」から各オリゴの５’末端は、前の工程において、ここにアニールし、適切に連結される）を特異的に切り出す。
【００５９】
続いて、マング・ビーンヌクレアーゼを用いて、標準的な条件下で、前記消化された産物の陥凹した３’末端を除去し、ＰＣＲ産物の３’末端に平滑断端を作出する。標準的な条件下で、ＱＩＡｑｕｉｃｋ ＰＣＲ精製キット（Ｑｉａｇｅｎ）を用いて、ＤＮＡを精製し、続いて、制限酵素ＮｃｏＩを用いて完全に消化した。続いて、ＱＩＡｑｕｉｃｋゲル抽出キット（Ｑｉａｇｅｎ）を用いて、１％アガロース／ＴＢＥゲル上で、８００ｂｐ〜１０００ｂｐにわたる制限ＤＮＡ断片を精製した。制限酵素ＮｃｏＩとＨｐａＩでベクターｐＭＭ１０６Ｈ（３μｇ）を完全に消化し、２８７０ｂｐの骨格断片をゲルで精製した。続いて、標準的な条件下で、ベクターＤＮＡと制限ＰＣＲ産物を互いに連結させ、連結混合物を用いてＥ．ｃｏｌｉ ＤＨ５α細胞を形質転換した後、これを回収し、１００μｇ／ｍＬのカルベニシリンを含有するＬＢプレート上に播種した。
【００６０】
このクローニング操作は、前工程で得られたＰＣＲ産物の完全なセットに対して行った。しかしながら、最初のインフレーム停止コドンへのオリゴの特異的なアニーリングと連結に由来するＰＣＲ産物のみが、この操作を介したクローニング工程後に、ヘキサヒスチジンタグとＧＦＰへのインフレーム融合物を生じることができるはずである。このようにしてクローニングされた他の全てのＰＣＲ産物は、ヘキサヒスチジンタグとＧＦＰへのアウトオブフレーム融合物を誘導するにすぎないはずである。ＰＣＲ産物の平滑断端の、ベクターの平滑断端への連結が、下流ベクターＤＮＡの読み取り枠の翻訳が切り出された停止コドンの元の読み取り枠によって指示される遺伝的融合物もたらすので、このようなことが起こる。もし、停止コドンがＧＳＴ遺伝子に関してアウトオブフレームであれば、新しく付加されたヘキサヒスチジンをコードする配列も、ＧＳＴ遺伝子に対してアウトオブフレームになるものと思われるが、停止コドンがＧＳＴに関してインフレームであれば、新しく付加されたヘキサヒスチジンをコードする配列もＧＳＴ遺伝子に関してインフレームとなるであろう。しかしながら、特異的に切り出された停止コドンがＧＳＴの最初のインフレーム停止コドンであったＰＣＲ産物のみが、前記ＤＮＡが転写、翻訳されるときに、ヘキサヒスチジン（とＧＦＰ）が付加されたＧＳＴ融合タンパク質を生じ得る。上記特異的工程全体から生じ得る唯一のヘキサヒスチジン（とＧＦＰ）が付加されたタンパク質は、それ故、必ず、ポリアスパラギン、ヘキサヒスチジンへの完全長のＧＳＴ融合物であろう。
【００６１】
上述したクローニング操作から得られたコロニーは、緑色の蛍光のコロニーを同定するために、３６５ｎｍで可視化された。９０のコロニー（白と緑の両者）をランダムに選んで、レプリカプレーティングを行い、抗Ｈｉｓタグと抗ＧＳＴ抗体を用いた標準的な条件下でコロニーウェスタンブロットによって分析した。抗Ｈｉｓタグ抗体は、ヘキサヒスチジンが付加されたタンパク質を発現するコロニーに結合するだけであると思われるので、ウェスタンブロットは、ヘキサヒスチジンタグへのインフレーム融合物を発現するコロニーの数についての直接的な情報を与える。これに対して、抗ＧＳＴ抗体は、ＧＳＴタンパク質のＣ末端付近に結合するので、完全長又はほぼ完全長のＧＳＴタンパク質を発現するコロニーのみを認識する。我々は、抗Ｈｉｓタグと抗ＧＳＴ抗体の両者によって陽性に認識されたタンパク質を含有するコロニーを同定した。これらのコロニーからのＤＮＡを増幅、精製、及び配列決定した。配列決定データによって、完全長のＧＳＴへの２つの完全なインフレーム融合物（すなわち、元のＧＳＴ遺伝子の最初のインフレーム停止コドンが特異的に切り出され、インフレームのポリアスパラギン、ヘキサヒスチジンタグによって置き換えられたクローン）の存在が確認された。この操作全体を介して我々が取得した修飾の成功率は、それ故、約２．２％である。両陽性クローンは、十分な量の完全長のＧＳＴ−ヘキサヒスチジン−ＧＦＰ融合物が発現されるために、長波長の紫外光に暴露すると、緑の蛍光を発することが見出された。それ故、さらなる全ての実験において、ウェスタンブロットによるさらなる分析のためには、緑の蛍光を発するコロニーのみを選択した。我々は、約７０〜８０％の緑の蛍光を発するコロニーが、抗Ｈｉｓタグ抗体によって認識されるタンパク質を発現することを見出した。残りの２０〜３０％のケースでは、おそらくは、クローニングされた挿入物によって導入された偽リボソーム結合部位の助けによって、ヘキサヒスチジンタグとは独立に、ＧＦＰ遺伝子の最初のＡＴＧから翻訳が開始している。
【００６２】
我々は、抗Ｈｉｓタグと抗ＧＳＴ抗体の両者によって認識されるタンパク質を発現している緑の蛍光を発するコロニーからプラスミドＤＮＡを増幅、精製、配列決定した。オリゴミックスＡとオリゴミックスＢの両者を用いて調製した挿入物に関しては、修飾の成功率は、全ての緑のコロニーの約２５％であることが分かった。それ故、ＧＦＰ遺伝子をインフレーム融合物に対するマーカーとして使用すると、正しいクローンを検出する効率が約１０倍増加する。
【００６３】
マング・ビーンヌクレアーゼは、異なる特異性を示すかもしれない多数の異なる１本鎖ヌクレアーゼ、例えばＳ１ヌクレアーゼ又はＲＮａｓｅＴによって置き換えることができ、適切な位置に置かれた多数の異なるＩＩＳ型制限酵素、例えばＳａｐＩを、ＢｐｍＩの代わりに使用し得ることも理解できるであろう。
【００６４】
（ｆ）タグ付加されたタンパク質の固定化と機能的な分析（図１ｅ参照）
上記の方法によって作成された完全長のヘキサヒスチジンが付加されたＧＳＴプラスミドのうちの１つで、Ｅ．ｃｏｌｉ ＤＨ５α細胞を形質転換した。１０ｍＬの液体培地中において、中期対数期まで、単一のカルベニシリン耐性コロニーを増殖させた後、１００μＭのＩＰＴＧを補充して、ヘキサヒスチジンが付加されたＧＳＴの発現を誘導する。さらに４時間増殖させた後、細胞を採集し、凍結融解／リゾチームによって溶解した。粗可溶化液のＳＤＳ−ＰＡＧＥは、予想されたサイズ（２７ｋＤａ）の過剰発現したタンパク質（全可溶性タンパク質の約２０％を占めた）とともに、アンバーの抑制によって生じた少量の５４ｋＤａのＧＳＴ−ヘキサヒスチジン−ＧＦＰ融合物を示した。続いて、粗可溶化液（５００μＬ；１００μｇ）をニッケル−ＮＴＡ磁気ビーズ（５０μＬ；結合能１５μｇのヘキサヒスチジンが付加されたタンパク質）と混合し、磁場の下での沈降によって前記ビーズを回収した。上清を捨て、ビーズを洗浄した後、各１ｍＭのグルタチオンと１−クロロ−２，４−ジニトロベンゼンを含有するグルタチオンＳトランスフェラーゼアッセイ緩衝液中に再懸濁した。３４０ｎｍの吸光度を測定することによって、室温で３０分後に終末点のアッセイデータを集めた。この波長は、ＧＳＴによって触媒される反応の産物のλ_ｍａｘに相当する。
【００６５】
対照として、親ベクター（ｐＭＭ１０６Ｈ）又は無関係のＨｉｓ付加タンパク質（アラニンラセマーゼ）をコードするプラスミドの何れかを含有するＤＨ５αの培養物を平行して増殖、誘導、採集、溶解し、アッセイした。ＧＳＴ活性は、Ｈｉｓを付加されたＧＳＴを含有する粗可溶化液と混合されたビーズ上のみに検出され、明らかに、観察されたＧＳＴ活性が、固定化されたＨｉｓ付加ＧＳＴから特異的に得られたものであり、さらに、該タンパク質が特異的に固定化されたときに、活性を保持していることが実証された。
【００６６】
酵素アッセイの完了後に、１００ｍＭのイミダゾールを含有する緩衝液を添加することによって、磁気ビーズからタンパク質を溶出し、ＳＤＳ−ＰＡＧＥによって分析した。これによって、陽性活性のアッセイ結果が得られたサンプルが、グルタチオンＳトランスフェラーゼに対して予想された正確なサイズ（２７ｋＤａ）の単一の固定化されたタンパク質を含有することが示され、このため、前記ビーズ上に観察された活性が、この組み換えＨｉｓ付加タンパク質のみによることが確認された。
【００６７】
例２
（ａ）２つの異なるタグを用いたＧＳＴの修飾、固定化、及びアッセイ
ヘキサヒスチジンタグによるグルタチオン−Ｓ−トランスフェラーゼの修飾について例１で記載した操作に従って、我々は、該操作が、加えたタグの正確な性状の影響を受けないことを実証した。
【００６８】
まず、６７６ｂｐのＮｃｏＩ／ＨｐａＩナンセンスＤＮＡスタッファー断片が、Ｅｓｃｈｅｒｉｃｈｉａ ｃｏｌｉ ｇｄｈＡ遺伝子に由来する３００ｂｐのＮｃｏＩ／ＨｐａＩ断片に置き換えられ、ヘキサヒスチジンタグがＦＬＡＧペプチド（エピトープタグ）とＳｔｒｅｐＩＩタグ（ストレプトアビジンに特異的に、且つ高い親和性で結合する）の何れかによって置き換えられていることを除いて、ｐＭＭ１０６Ｈと同一である２つのさらなるベクターを構築した。これらのベクターは、それぞれ、ｐＭＭ１０４Ｆ及びｐＭＭ１０４Ｓと表記された。ＮｃｏＩとＨｐａＩを用いて、これらのベクター（それぞれ３μｇ）を別々に完全消化し、２８７０ｂｐの骨格断片をゲル精製し、例１に記載したようにオリゴミックスＡから生じたＧＳＴ断片に連結した。抗ＦＬＡＧ及び抗ＧＳＴ抗体の組み合わせを使用した後、配列決定し、ＦＬＡＧタグへの前記完全長のＧＳＴ遺伝子の完全なインフレーム融合物を含有するクローンを同定した。同様に、抗ＧＳＴ抗体とストレプトアビジン−西洋ワサビペルオキシダーゼ接合体の組み合わせを用いた後、配列決定し、ＳｔｒｅｐＩＩタグへの前記完全長のＧＳＴ遺伝子の完全なインフレーム融合物を含有するクローンが同定された。何れの例においても、完全長のインフレーム融合物が見出される頻度は、例１で決定したものと同じ（実験誤差の範囲内）であった。
【００６９】
固定化基材が、それぞれ抗ＦＬＡＧ抗体をコートした９６ウェルのプレートとストレプトアビジンコートされた磁気ビーズであることを除き、実質的に例１に記載されているように、これらのクローンを固定化実験で使用した。例１における如く、我々は、これらのタグを介した融合タンパク質の特異的な固定化を実証することができ、さらに、我々は、ＦＬＡＧ又はＳｔｒｅｐタグの何れかを介して固定化したときに、ＧＳＴ融合物がその活性を保持することを示すことができた。
【００７０】
例３
（ａ）ヘキサヒスチジンタグを用いた別のタンパク質の修飾
グルタチオン−Ｓ−トランスフェラーゼについて例１に記載した操作に従って、我々は、この操作が、扱っている正確な遺伝子の影響を受けないことを実証した。
【００７１】
このように、ヒト転写因子ＮＦ−κＢ ｐ５０をコードするプラスミドから開始し、特に明記されていなければ、例１に記載されている操作（オリゴミックスＡを用いる）に正確に従って、我々は、最初のインフレーム停止コドンが特異的に切り出され、ポリアスパラギン、ヘキサヒスチジンタグをコードするＤＮＡへのインフレーム融合物によって置き換えられているようにＮＦ−κＢ ｐ５０が修飾されることを実証することができた。抗Ｈｉｓタグ抗体を用いたコロニーウェスタンブロットによって、ヘキサヒスチジンが付加されたタンパク質を発現するクローンの同定が可能となった。これらのコロニーからのＤＮＡを増幅、精製、配列決定した。配列データによって、幾つかのクローンは、完全長のＮＦ−κＢへの完全なインフレーム融合物（すなわち、前記停止コドンが特異的に切り出され、インフレームヘキサヒスチジンタグによって置き換えられたクローン）をコードすることが確認された。ＮＦ−κＢ ｐ５０のケースで、完全長のインフレーム融合物が見出される頻度は１．１％であり、ＧＳＴについて例１で決定したものと近く、実験誤差の範囲内にある。
【００７２】
（ｂ）ヘキサヒスチジンが付加されたＮＦ−κＢ ｐ５０の固定化と機能分析 上記方法によって作成した完全長のヘキサヒスチジンが付加されたＮＦ−κＢプラスミドのうちの１つで、Ｅ．Ｃｏｌｉ ＤＨ５α細胞を形質転換した。１０ｍＬの液体培地中において、中期対数期まで、単一のカルベニシリン耐性コロニーを増殖させた後、１００μＭのＩＰＴＧを補充して、ヘキサヒスチジンが付加されたＮＦ−κＢ ｐ５０の発現を誘導する。さらに４時間増殖させた後、細胞を採集し、音波処理によって溶解した。粗可溶化液のＳＤＳ−ＰＡＧＥは、予想されたサイズ（３８ｋＤａ）の過剰発現したタンパク質（全可溶性タンパク質の約５％を占めた）とともに、アンバーの抑制によって生じた少量の６５ｋＤａのＮＦ−κＢ ｐ５０−ヘキサヒスチジン−ＧＦＰ融合物を示した。
【００７３】
【表３】

【００７４】
標準的な条件下で３末端トランスフェラーゼ（Ｂｏｅｈｒｉｎｇｅｒ Ｍａｎｎｈｅｉｍ）を用いて、二重鎖オリゴヌクレオチド「κＢモチーフ」（ＮＦ−κＢ ｐ５０に対するパリンドローム結合部位を含有する）の３’塩基をジゴキシゲニンで標識した。
【００７５】
５ｍＭのβ−メルカプトエタノールを含有するＰＢＳ（リン酸緩衝化された生理的食塩水ｐＨ７．５）中でのリゾチーム／凍結溶解法を用いて、タンパク質溶解物を調製した。各クローンからの可溶性タンパク質溶解物２００μＬをＮｉ−ＮＴＡコートしたマイクロウェルに加え、室温で４５分間インキュベートした。インキュベーション時間の終了時に、ＰＢＳＴ（０．０２％のＴｒｉｔｏｎ Ｘ−１００を含有するＰＢＳ）でウェルを３回洗浄して、結合していないタンパク質を全て除去した。ＤＮＡ結合緩衝液（５ｍＭのβ−メルカプトエタノールを含有する１０ｍＭ Ｔｒｉｓ．ＨＣｌ ｐＨ ７．４、７５ｍＭ ＫＣｌ）を用いて、１分の浸漬時間で、ウェルを３回洗浄した。３’ジゴキシゲニンを標識したκＢモチーフ（２ｐｍｏｌ）を、１μｇのポリ（ｄＩ−ｄＣ）非特異的ＤＮＡを含有するＤＮＡ結合緩衝液２００μＬ中にウェルに加えた。さらに３０分インキュベートした後、０．０２％のＴｒｉｔｏｎ Ｘ−１００を含有する１０ｍＭ Ｔｒｉｓ．ＨＣｌ ｐＨ ７．４、２５ｍＭ ＫＣｌで３回ウェルを洗浄することによって、結合していないＤＮＡを除去した。抗ジゴキシゲニン抗体−アルカリホスファターゼ接合体を、０．２％のウシ血清アルブミンが補充された「抗体希釈緩衝液」（１０ｍＭ Ｔｒｉｓ．ＨＣｌ ｐＨ ７．４、２５ｍＭ ＫＣｌ）の中に、１５０ｍＵ／ｍＬになるように希釈した。続いて、希釈した抗体（２００μＬ）をマイクロウェルに加えた。室温で３０分間放置した後、０．０２％のＴｒｉｔｏｎ Ｘ−１００が補充された「抗体希釈緩衝液」（３×３５０μＬ）でマイクロウェルを洗浄することによって、結合していない抗体を除去した。続いて、アルカリホスファターゼの基質である２５０μＭのｐ−ニトロフェニルリン酸（ｐＮＰＰ）を含有する２００μＬの緩衝液（１００ｍＭ Ｔｒｉｓ．ＨＣｌ ｐＨ９．５、１００ｍＭ ＮａＣｌ、５０ｍＭ ＭｇＣｌ_２）をウェルに加え、室温で一晩反応を進行させ、その後、各ウェル中の黄色の呈色（産物の形成に相当する、ｐ−ニトロフェノール）を４０５ｎｍで定量した。基質ｐＮＰＰの加水分解のバックグラウンド速度は低いので、ウェル中に黄色が出現することから、直ちに、陽性のアッセイ結果が明らかとなった。
このアッセイの対照として、我々は、粗可溶化液、又は標識されたオリゴヌクレオチド、又は抗体の何れかを省略し、又は２０倍過剰の未標識の二重鎖オリゴを加え、又は粗可溶化液を含有するヘキサヒスチジンが付加されたＮＦ−κＢ ｐ５０を、同じベクターのバックグラウンド中でヘキサヒスチジンが付加されたＧＳＴを発現しているＤＨ５α細胞から得た等量の粗細胞可溶化液で置き換えた。
【００７６】
このアッセイでは、ＮＦ−κＢ ｐ５０は、まず、特異的な結合部位を介して標識されたオリゴヌクレオチドに結合する。続いて、ヘキサヒスチジンタグを介して、マイクロウェル中に該タンパク質−ＤＮＡ複合体を固定化し、全ての結合していない標識されたオリゴとともに、（前記標識されたオリゴと粗可溶化液中に存在する他のＤＮＡ結合タンパク質との複合体を含む）他の全てのタンパク質を洗浄除去した。抗体−接合体は前記オリゴ上の標識を認識するが、ヘキサヒスチジンが付加されたタンパク質を認識しないので、タグを介してＮＦ−κＢ ｐ５０を固定化したときに、ＮＦ−κＢ ｐ５０−ＤＮＡ相互作用が維持されれば、アッセイ中の陽性シグナルのみを観察することができる。この相互作用が維持されなければ、前記オリゴは、洗浄工程の間に失われ、色の変化は観察されないであろう。
【００７７】
我々は、黄色の産物は、ヘキサヒスチジンが付加されたＮＦ−κＢ ｐ５０粗可溶化液とジゴキシゲニン標識されたオリゴヌクレオチドを含有するマイクロウェル中のみで検出されことを見出し、これに抗ジゴキシゲニン抗体−アルカリホスファターゼ接合体を添加した。これにより、観察された色の変化は、固定化されたＮＦ−κＢ ｐ５０オリゴヌクレオチド複合体から特異的に得られたものであり、さらに、特異的な固定化のときに、ＮＦ−κＢ ｐ５０が活性を保持していることが実証された。
【００７８】
例４
停止コドンが正確に除去された完全長の遺伝子又はｃＤＮＡを作成するための別の操作は、コード配列を包含するＰＣＲ産物において、５’が陥凹した欠失物の鎖特異的な入れ子状態のセットを作成することが含まれる。これは、λエキソヌクレアーゼ、Ｅ．ｃｏｌｉ ＤＮＡポリメラーゼＩ、Ｔａｑ ＤＮＡポリメラーゼ、又はＴ７遺伝子６エキソンクレアーゼのようなヌクレアーゼを用いて行うことができるであろう。このように、例えば、前記元のＰＣＲ産物は、任意の５’から３’方向へのエキソヌクレアーゼを用いて行うことができるであろう。５’が陥凹した欠失物の入れ子状態のセットが一度作成されると、オリゴミックスは、露出した１本鎖領域にアニールすることができる。該オリゴミックスは、各停止コドンの相補部分が３’末端に存在し、すぐ直前に６つのランダムな残基が先行し、その前にＩＩＳ型制限部位をコードする所定配列とプライマー「ＬＢＭ２」（例１参照）に対する相補性認識配列が先行する１群のオリゴからなる。このため、前記オリゴ群の配列は、以下のとおりである。
【００７９】
【表４】

【００８０】
前記混合物からの１つのオリゴは、センス鎖上に露出された各停止コドンに特異的にアニールし、Ｔａｑポリメラーゼのように、又はＥ．ｃｏｌｉのＤＮＡポリメラーゼＩのような５’から３’へのエキソヌクレアーゼ活性のように何れかの鎖を置き換える活性を有するＤＮＡポリメラーゼに対するプライマーとなるであろう。新しく生じた二重鎖ＤＮＡ断片は、続いて、元のＰＣＲ産物の５’末端に結合するプライマーと新しくアニール、伸長されたオリゴに特異的に結合するプライマーとを用いて、特異的に増幅することができる。この操作では、前記アニールされたオリゴの３’末端は、二重鎖テンプレートの５’が陥凹した残基に直接隣接することを要しない。前記混合物からのオリゴは、ランダムなヘキサマープライミングと類似した態様で、前記露出した１本鎖ＤＮＡテンプレート上の任意の相補領域にアニールすることができるが、重要なことに、鎖の伸長は、３’末端が絶対的な相補性をもってアニールする場合だけに起こるであろう。前記混合物中のオリゴの３’末端は、停止コドンに相補的であるようにデザインされているので、オリゴの３’末端が一致する停止コドンに特異的に結合するときだけ、プライマーの伸長とその後のＰＣＲ増幅が起こるであろう。アウトオブフレームの停止コドンと目的の遺伝子の真の停止コドンの下流にあるコドンへの結合も起こり、正確な完全長の遺伝子のみならず、一群の末端切断され、伸長された産物のＰＣＲ増幅が得られる。しかしながら、これらは、ペプチドタグへのインフレーム融合物を生じないと思われるので、続く工程で容易にスクリーニングすることができる。
【００８１】
記載のごとく５’欠失物の入れ子状態のセットに対してアニーリングと伸長操作を行うことは、全コード領域を包含する完全に１本鎖のＤＮＡ又はＲＮＡ分子に対してアニーリングと伸長操作を行うことに比べて、著しい利点を有する。これは、伸長の前に、プライマーが特異的にアニールすることができる部位の数が、前記欠失物の入れ子状態のセットの２本鎖部分の存在によって大きく制限されることによるものである。欠失物の入れ子状態のセットの１本鎖部分は、主として、前記遺伝子の３’未翻訳領域を包含し、これは、コード配列の外にある停止コドンを優先して、さらにはより長い伸長産物を優先して、前記伸長産物に強く偏らせる効果を有するであろう。これらの因子は、最初のインフレーム停止コドンが操作全体によって特異的に除去される頻度を著しく増加させるように作用するであろう。実際に、我々は、完全に１本鎖のコード領域テンプレートを用いて、以下の工程（ｂ）に記載したオリゴをアニールし、伸長することを試みたが、我々は、その実験からは、正しく修飾された完全長のクローンを何ら同定できなかった。これに対して、以下に詳述されている操作の結果は、アニーリングと伸長工程のためのテンプレートとして５’欠失物の入れ子状態のセットの使用が、コード配列の最初のインフレーム停止コドンの特異的な除去を促進して、ポリペプチドタグへの完全長のインフレーム融合物を得る上で効果的且つ効率的であることを明らかに実証している。
【００８２】
（ａ）タグ付加に先立つ遺伝子のＰＣＲ増幅とエキソヌクレアーゼ消化
このように、我々は、前記増幅において、ここでは「ＳＴリバース」プライマーが５’リン酸化されていることを除いて、正確に、例１の工程（ａ）と（ｂ）に記載されているように、まずＧＳＴ遺伝子の最初のＰＣＲ増幅を行った。続いて、２．５ユニットのλエキソヌクレアーゼ（Ｎｏｖａｇｅｎ Ｓｔｒａｎｄａｓｅキット）を用い、標準的な反応緩衝液（６７ｍＭ グリシン−ＫＯＨ ｐＨ９．４、２．５ｍＭ ＭｇＣｌ_２、５０μｇ／ｍＬ ウシ血清アルブミン）中において、精製されたＰＣＲ産物を３７℃で４０分間消化した。続いて、１５分間７５℃に加熱することによって該酵素を不活性化した。λエキソヌクレアーゼ酵素は、５’リン酸基のみを基質として認識するので、センス鎖は消化から保護され、それ故、消化産物は、前記ＰＣＲ産物のアンチセンス鎖の５’末端から得られる欠失物の入れ子状態のセットである。
【００８３】
（ｂ）タグを導入するための特異的な連結、伸長、及び増幅
２５０μＭのｄＮＴＰと約２５倍モル過剰の下記の縮重オリゴ群の存在下で、Ｅ．ｃｏｌｉ ＤＮＡポリメラーゼＩ反応緩衝液中に、５μＬのλエキソヌクレアーゼ反応ミックスを希釈した。
【００８４】
【表５】

【００８５】
消化された断片とオリゴを３７℃で３０分間アニールした後、５ユニットのＥ．ｃｏｌｉ ＤＮＡポリメラーゼＩを加え、その後３７℃で３時間反応液をインキュベートした。標準的な条件下で、ＱＩＡクイック ＰＣＲクリーンアップキット（Ｑｉａｇｅｎ）を用いて伸長産物を精製し、プライマー「ＳＴフォワード」と「ＬＭＢ２」により、Ｐｗｏポリメラーゼを用いる標準的なＰＣＲ反応におけるテンプレートとして使用した。３０サイクルのＰＣＲ（９４℃、１分；５７℃１分；７２℃２分）を実施して、１０００ｂｐに及ぶＰＣＲ産物を生成させた。
【００８６】
上記操作から生じたＰＣＲ産物を消化し、例１の工程（ｅ）で記載されているベクターｐＭＭ１０６Ｈ中にクローニングした。クローニング操作から得られたコロニーを３６５ｎｍで可視化して、緑の蛍光を発するコロニーを同定した。７３のこのようなコロニーを選択し、レプリカプレーティングを行い、抗Ｈｉｓタグと抗ＧＳＴ抗体を用いた標準的な条件下でコロニーウェスタンブロットによって分析した。緑の蛍光を発するコロニーのうち５８％が、抗Ｈｉｓタグと抗ＧＳＴ抗体の両者によって陽性に認識されるタンパク質を発現していた。抗Ｈｉｓと抗ＧＳＴの両者に陽性であった１５のコロニーからのＤＮＡを増幅、精製、配列決定した。配列決定データによって、完全長のＧＳＴへの１０個の完全なインフレーム融合物（すなわち、元のＧＳＴ遺伝子の最初のインフレーム停止コドンが特異的に切り出され、インフレームのポリアスパラギン、ヘキサヒスチジンタグによって置き換えられたクローン）の存在が確認された。この操作全体を介して我々が取得した修飾の成功率は、それ故、緑の蛍光を発するコロニーの総数の３９％である。
【００８７】
例５
（ａ）１１の遺伝子のプールから得られたあるタンパク質の同定
我々は、下表に列記された１１の異なる遺伝子のプールに特定されていることを除き、正確に例１に記載されているように前記操作を適用した。我々は、アレイ中の各位置が、該操作の結果として付加されたタグを通じて固定化された単一の組換えタンパク質に対応するように、得られた特異的に修飾されたタンパク質のアレイを作成した。続いて、我々は、機能的アッセイによって該アレイをスクリーニングし、前記プールの各タンパク質成分を首尾良く同定した。
【００８８】
【表６】

【００８９】
とりわけ、ｃＤＮＡライブラリーで遭遇する状況を模倣するために、まず、１１の遺伝子を全て、同じｐＴｒｃＨｉｓＡベクター骨格の中にサブクローニングした。例１に記載されたプライマー「ＳＴフォワード」と「ＳＴリバース」は、ｐＴｒｃＨｉｓＡベクター骨格内にコードされた遺伝子を増幅するための万能プライマーとなるようにデザインした。
プライマー「ＳＴフォワード」は、以下のように、多数の制限部位をコードするようにデザインされた。
【００９０】
【表７】

【００９１】
このように、エキソヌクレアーゼを保護するための３’突出部を作成する目的で、制限酵素ＡａｔＩＩかＳｆｉＩのうち何れかを使用することができる。修飾操作の最後に定方向のクローニングを行う目的で、本例では、統計的にはライブラリー中をより高い頻度で切断すると思われるので、ここで使用したタグベクターｐＭＭ１０６Ｈ中のＮｃｏＩクローニング部位と適合する粘着末端を作出し、本プール中の１１の何れの遺伝子の内部も切断しないので、我々はＢｓｐＨＩを使用することを選択した。明らかに、原則的としては、タグベクターがプロモーターの下流にある均等なクローニング部位を含有していれば、制限部位をコードする任意のプライマーを使用することができる。ＳｆｉＩは、８ｂｐの認識配列を有し、このため、ある遺伝子内のＳｆｉＩ部位のランダムな発生頻度が、ＢｓｐＨＩのような６ｂｐ認識配列の頻度（１／４０９６）と比べてずっと低い（１／６．５×１０^４）ので、より大きなライブラリーフォーマットにおいてこの点で著しい利点を有するであろう。
【００９２】
タグベクターｐＭＭ１０６Ｈは「ＡＴＧ」ベクターである、すなわち、５’クローニング部位（ＮｃｏＩ）が、ネイティブタンパク質を発現するためのリボソーム結合部位（ＲＢＳ）の下流に位置するＡＴＧ開始コドンと重なっている。しかしながら、ここに記載した操作では、我々は、開始コドンに共通の制限部位を有するクローニングされた遺伝子には頼らない。代わりに、我々は、単に、クローニングされた遺伝子自体によって翻訳の開始に必要なシグナルが提供されるように、ｍＲＮＡを生成するための転写を開始するベクターにコードされたプロモーターに頼っている。このように、この例では、元のプール中の全ての遺伝子は、ＡＴＧに位置するクローニング部位の存在又は不存在にかかわらず、ＲＢＳが直前に存在する開始コドンを有する。プライマー「ＳＴフォワード」は、１１の最初のクローン全てで、ＲＢＳの上流に結合するので、制限部位をコードする何れかのプライマーを用いたその後の修飾後クローニングは、それらの元のＲＢＳとＡＴＧとともに、タグベクターの中に新しく修飾された遺伝子を導入し、このため、翻訳の開始が確実となるであろう。ｃＤＮＡライブラリーフォーマットでは、全ての完全長のｃＤＮＡが、真核細胞の翻訳開始シグナルを含有するそれ自身の５’未翻訳領域（ＵＴＲ）を有すると思われる点で、同じ状況が当てはまる。続いて、この場合に適切な翻訳の開始を得るためには、転写開始シグナルをこのようにもう一度与える真核細胞ベクターの中に、その５’ＵＴＲとともに、修飾されたｃＤＮＡをクローニングすることのみが必要である。それ故、この例で使用したプライマーと均等な万能ＰＣＲプライマー群を使用することができる。
【００９３】
以下の修飾を用いて、例１で記載したように修飾操作を実施した。プライマー「ＳＴフォワード」と「ＳＴリバース」を用いる最初のＰＣＲ増幅のテンプレートとして、１１の遺伝子全ての等モルのプールを使用し、その後、ＳｆｉＩ酵素は８ｂｐの認識配列を有し、１１の遺伝子のうち何れの遺伝子の内部も切断しないので、ＳｆｉＩで断片を消化して５’末端を保護した。アニーリング工程では、オリゴミックスＡ（例１参照）を使用した。第２のＰＣＲ増幅後に、修飾された断片を２つのポットに分け、これをＢｐｍＩとＳａｐＩで別個に消化した。統計的には、何れのライブラリー中の遺伝子も、僅かな割合で、ＳａｐＩ部位（７ｂｐ認識配列；ランダムな発生確率＝１／１６，３８４）又はＢｐｍＩ部位（６ｂｐ認識配列；ランダムな発生確率＝１／４，０９６）の何れかを含有していると思われるが、両者を含有する割合はずっと少ないであろう（ランダムな発生確率＝１／６．７×１０^７）。２つのＩＩ型制限酵素は、このように、任意のある完全長の遺伝子の特異的な修飾が、その遺伝子内の１又は他の制限部位の存在によって除外されないことを効果的に確実にするために、別個に使用した。
【００９４】
続いて、マング・ビーンヌクレーゼによる処理及びＢｓｐＨＩによる消化のために、前記２つのポットからの消化された断片をプールした。得られた断片を４つの異なるサイズ範囲に、ゲルで精製し、より小さな挿入物が優先的に連結されるのを避けるために、ベクターｐＭＭ１０６Ｈ（それ自身、ＮｃｏＩとＨｐａＩで完全に消化され、ゲルで精製された）に別々に連結した。ＵＶ光（３６５ｎｍ）下で、形質転換された細胞を可視化し、ウェスタンブロットによる分析のために、肉眼で、緑の蛍光を発するコロニーを選別した。形質転換されたコロニーの総数の約３０％が緑の蛍光を発し、そのうち、７３％が、抗Ｈｉｓタグ抗体によって認識されるタンパク質を発現していた。深い９６ウェルのブロック中の液体培地１．５ｍＬの中に、１９０の緑のＨｉｓ陽性コロニーを植え付け、一晩増殖させた。遠心によって細胞を採集し、凍結融解／リゾチームによって溶解した。続いて、ニッケルＮＴＡでコートされた９６ウェルプレートの各ウェルに粗可溶化液を加え、洗浄によって未結合のタンパク質を除去して、Ｈｉｓ付加された組換えタンパク質をウェルに固定化した。続いて、例２に記載したアッセイを用いて固定化したタンパク質のＮＦ−κＢ活性をアッセイし、黄色の呈色の出現によって、陽性の「ヒット」を含有するウェルを同定した。３つのクローンが、陽性の「κＢ―モチーフ」ＤＮＡ結合活性を示した。該陽性のクローンのさらなる性質決定によって、１つのクローンが、予想通り、ヘキサヒスチジンタグへのＮＦ−κＢ ｐ５０遺伝子の正確な完全長インフレーム融合物をコードすることが示された。残りの２つのクローンは、結合親和性はより低いが、ＮＦ−κＢ ｐ５０として同じＤＮＡ結合特異性を共有することが知られている関連するＤＮＡ結合タンパク質をコードすることが分かった。
【００９５】
それ故、本結果は、この操作によって作成されたアレイの機能的な検査により、特異的な相互作用とこれより弱いが特異的且つ生物学的に関連した相互作用の両者を同定できることを実証している。我々は、それ故、マイクロウェルフォーマットで機能的なタンパク質のアレイを作るためにこの操作を使用し、これらのアレイを用いて、我々は、特異的なタンパク質−リガンド相互作用に基づいて、小さなプールから、各タンパク質を同定することに成功した。
【図面の簡単な説明】
【図１ａ】 ベクターｐＭＭ１０６Ｈの構築を示す図。
【図１ｂ】 タグ付加に先立つ、一例の遺伝子（ＧＳＴ）のＰＣＲ増幅とエキソヌクレアーゼによる消化の詳細を示す図。
【図１ｃ】 タグを導入するための特異的な連結とＰＣＲ増幅の詳細を示す図。
【図１ｄ】 ＰＣＲ産物のクローニングの詳細を示す図。
【図１ｅ】 ＧＳＴによって触媒された、グルタチオンと１−クロロ−２，４−ジニトロベンゼンとの間の反応を示す図。[0001]
The present invention relates to a novel method of creating protein expression arrays and the use of such arrays in rapid screening.
[0002]
The genome mapping project is revolutionizing the process of discovering therapeutic targets and drug discovery using them. Once new therapeutic targets are identified, rapid mass screening of existing and combinatorial chemical libraries will present many lead candidates that are active against these targets . It will be clear that it is uneconomical to pursue all lead compounds, even in early laboratory tests. However, there is currently no rapid method for assessing such lead compounds with respect to the activity profile they may have on all proteins present in the organism. If available, such methods allow early assessment of the toxicological profile that all lead compounds may have, and with this information, which compounds can be tracked and which The process of determining whether to exclude certain compounds will be significantly facilitated.
[0003]
The pharmaceutical industry has the complementary demand to identify all the targets of existing drugs (both those already on the market and those still in development) and thereby determine their mechanism of action. To do. It is becoming increasingly clear that regulators now believe that knowledge of the mechanism of action is crucial, so once this information is available, the process of obtaining regulatory approval for new drugs Will be much easier. In addition, this type of information would allow the design of improved second generation drugs. This is because many of the drugs probably have at least slight side effects caused by binding of the drug or its metabolites to undesirable targets. All these target proteins need to be identified in order to establish the criteria necessary to design improved drugs. At present, however, there is no simple way to obtain this information, and many millions of drug candidates have been abandoned due to unsuccessfulness simply because the target of action is unknown.
[0004]
Protein-protein interactions are increasingly recognized as extremely important in governing cellular responses to both internal and external stresses. Therefore, specific protein-protein interactions are potential targets for drug-mediated treatment in infections and other medical conditions. At present, the yeast two-hybrid assay is the only reliable method for assessing protein-protein interactions, but this type of in vivo assay is capable of protein-protein interactions even in non-fast mass formats. It would not be very suitable for identifying specific agonists or antagonists. A functional proteome expression array, or “proteome chip”, would be able to determine the specificity of protein-protein interactions and the specificity of any drug-mediated effect in an in vitro format. Therefore, it can be said that the proteome array has enormous potential to revolutionize this field of research.
[0005]
One way in which a functional proteome array can be created is to clone, express, purify, and immobilize all proteins expressed in a particular proteome separately. However, along with consideration of the availability of sequence data for the entire genome, the first important thing to consider is the absolute size of the target genome. To explain these points, the typical bacterial genome is about 5 Mbp, and so far only a few have been fully sequenced (eg, Helicobacter pylori, Escherichia coli, and Mycobacterium tuberculosis), fungi The genome of is typically about 40 Mbp, the mammalian genome is 3 Gbp, and the plant genome is about 10 Gbp. Currently, it is estimated that the sequencing of the human genome will be completed around 2003, but the extent to which this information is placed in the public domain must be questioned greatly. Obviously, it is completely unrealistic to expect that all genomes other than the representative model organism will be available within a realistic time frame, and from a functional proteomics perspective, The value is limited. Therefore, in principle, it may be possible to design and synthesize primers for cloning about 100,000 genes in the human genome from a cDNA library within the past four years. For some, even if the necessary sequence data is available, this would be extremely expensive (primer costs alone would cost millions of dollars) and would be a very laborious process.
[0006]
But what about pharmacologically suitable organisms that will not have complete sequence data? What could be an alternative if they cannot be completely ignored from functional proteomics? Expression cDNA libraries can in principle be used with nonspecific immobilization to create protein arrays, but nonspecific immobilization usually disrupts protein folding Thus, this technique is severely limited by the fact that it involves loss of function. Furthermore, when the host cell protein is immobilized, at least the S / N ratio is significantly reduced, and in the worst case, a clear result cannot be obtained. Therefore, a functional proteome array in which each protein is specifically immobilized and purified via a common motif or tag without affecting function and without requiring knowledge of the entire genome sequence. If it can be made, it will make tremendous progress in the field of functional proteomics.
[0007]
The present inventors have solved the above problem by providing a method capable of adding a common marker to a predetermined position in each protein in the proteome without requiring prior knowledge of the DNA sequence of the corresponding gene. It led to the development of a new approach to solve. This “tag” can then be used to confer commonality and specificity for subsequent immobilization and purification operations, which can subsequently result in thousands of gains from a given proteome. It is possible to create a spatially ordered array in which proteins are displayed.
[0008]
The important thing to consider here is to place the “tag” in the correct location. If the tag is inserted in-frame at an indeterminate and random position of any gene, the resulting tagged protein will be cleaved in an indeterminate manner. In this case, the correct folding and function will be destroyed. The method described here uses any particular gene so that each full-length tagged protein is correctly folded and thus retains its function when specifically immobilized in the array. This problem is avoided by inserting the tag immediately after the start codon or immediately before the stop codon.
[0009]
Since each protein in the array should be fully functional, the array can then be screened directly to identify drug targets and other biologically relevant molecules. The spatial definition of the array will allow the phenotype of each protein to be directly associated with its genotype to identify “hits”.
[0010]
Thus, in the first aspect, the present invention comprises cloning and expressing one or more proteins as full-length proteins each having a marker moiety added to either the N-terminus or C-terminus. A method of creating a protein array is provided.
[0011]
The marker moiety may be any peptide sequence, such as a hexahistidine tag, an antibody epitope, or a biotin mimic, or indeed a complete protein, or a protein domain, such as a maltose binding protein domain. The marker moiety itself can be post-translationally modified, for example by adding biotin or lipid molecules. In a preferred embodiment, the marker moiety will also allow for the purification of “tagged” proteins.
[0012]
Thus, the method of the present invention allows all members of a cDNA library to be specifically modified in one pot in a manner that does not rely on any knowledge about the sequence of each gene. Instead, the method of the invention is based on a common start or stop codon that is present in all genes. The modification is made in the form of accurately inserting in-frame additional DNA to be added with a known sequence, either immediately after the start codon of each cDNA or just before the stop codon, if necessary. Will.
[0013]
The DNA to be added will encode a known marker moiety that will be present in the same reading frame as the product of each cDNA. Each genetically modified cDNA produced by the method of the present invention thus comprises a “tag”, eg, a polypeptide, which is a common part precisely fused to either its N or C terminus. Will encode each protein it has. Since all members of a cDNA library appear to be modified in exactly the same way, all proteins encoded by the cDNA library will eventually have their N or A common part will be added to any of the C-termini.
[0014]
In general, the protein expressed from the cDNA library will be “tagged” and will be easily identified and isolated. Once purified, the protein can be attached to, for example, a microarray. Attachment can be through the tag itself or through another part that was first attached to the protein.
[0015]
Arrays formed by the methods described herein form a second aspect of the present invention. Such arrays typically comprise an expression library of “tagged” proteins immobilized on a solid support. Those skilled in the art generally use a wide range of supports that can be solid supports in the field of arrays, and any of these “substrates” can be used to produce the arrays of the present invention. You will understand what you can do.
[0016]
As described herein, the method of the present invention allows for a specific tag to be added to either the N-terminus or C-terminus of all proteins in a proteome. Some proteins cannot tolerate an N-terminal extension, and some proteins cannot tolerate a C-terminal extension, but most proteins will tolerate any such extension. However, existing library cloning methods address this challenge because they clone genes as full-length, unmodified cDNA, or as fusions to certain protein partners that are almost unavoidably truncated at random. It cannot be solved at all.
[0017]
Compared to the latter, according to the present method, it is possible to create an accurate, full-length cDNA library, for example, as a fusion to a desired peptide partner. Compared to the former, the method of immobilizing proteins in the arrays described herein allows these specific interactions to occur in each cDNA rather than through non-specific interactions. This is a function of a tag added to the end of the. Furthermore, the methods described herein can be used to screen purified, immobilized proteins expressed in non-bacterial host organisms to help maintain function through correct folding and post-translational modifications. Whereas existing methods such as phage display or λ-cDNA libraries can be used in many eukaryotic cells due to either misfolding or incorrect post-translational modifications. Limited to bacterial hosts where the protein is synthesized in a non-functional form.
[0018]
The method of the present invention has a wide range of in vitro applicability that can be broadly divided into three main regions. These are protein-ligand interaction studies, protein-protein interaction studies, and protein-DNA interaction studies.
[0019]
Protein-ligand interaction
The methods described herein will be able to rapidly delineate the interaction between a new chemical and all proteins in a proteome. This can be accomplished simply by probing an array of appropriate proteomes with NCE at various stringencies that can be considered reverse fast mass screening. Information read from such screening will be immediately useful in many situations, some of which are described below.
[0020]
Rapid mass screening programs that examine libraries of compounds against cells or complete organisms often identify cues that cause phenotypic changes without knowing their target prior to screening. However, subsequent identification of the primary target can be a very laborious process. After creating a functional proteome array for the species, it is likely that the lead compound can be used to screen this array to identify which protein in the proteome is targeted. The inventive method can be applied directly to this type of problem. This massively parallel approach to identifying protein-ligand interactions greatly speeds up and simplifies the determination of NCE primary targets and also identifies weaker secondary interactions that may also be important. Will make it possible. In addition, the method can be directly applied to the problem of species cross-reactivity, for example, allowing rapid assessment of antifungal compound candidates, for example, for interactions with all proteins in the human proteome. It becomes. It will be appreciated that this type of information is extremely useful in any subsequent optimization of the lead compound.
[0021]
Fast mass screening methods now allow rapid identification of small molecules that bind to certain proteins that have previously been identified as potential therapeutic targets. However, these methods do not answer this question, although it may be important to know how selective an interaction is in determining whether to track a lead compound. Absent. A superior would argue that a compound targeting a single protein appears to have fewer side effects than a compound that also encounters many related or unrelated proteins.
[0022]
Although well underway to Phase 3 clinical trials, there are many examples of compounds that cannot be approved by regulatory authorities because their primary mechanism is unknown. The antidepressants Mianserin and Trazadone, and Pfizer's anti-arthritic drug Tenidap are examples of this, each with no return on the billions of dollars invested. If the methods described herein find major targets for such drugs, and then the mechanism of action becomes clear, very expensive clinical trial data is already available for regulatory approval. As it becomes available, it could be used to revive such failed drugs.
[0023]
All existing drugs have more or less side effects, but for this example is clozapine, an otherwise attractive anti-schizophrenic drug. If the molecules from which such side effects are derived can be determined, this will greatly facilitate the design of future drug discovery that combines minimal side effects and optimal main effects. In creating a profile of the interaction of a compound with all proteins in the proteome, abnormal secondary interactions are identified and subsequently evaluated in terms of whether they are associated with known side effects As such, the methods described herein can again be applied directly to such challenges.
[0024]
The methods of the invention can also be used to identify protein families such as serine proteases by screening proteome arrays with common inhibitors. This allows subsequent development of biochips that display, for example, all human serine proteases, or all kinases, or all p450 enzymes, for further focused lead screening. It will be. For example, the p450 biochip is because the hydroxylation mediated by the p450 chip is often the first step in the metabolic process and is considered one of the main causes of inter-patient variation in drug response It may be useful to assess whether the lead compound has been metabolized. In fact, one of the ultimate goals of current drug design is to create compounds that are not first metabolized, and again, the p450 chip could have significant utility.
[0025]
Protein-protein interaction
Protein-protein interactions and multiprotein complexes are particularly important in cell biology. For example, signal transduction pathways are generally initiated by the interaction of cell surface receptors with external ligands, followed by a cascade of protein-protein interactions, ultimately activating specific genes. Bring. Each protein-protein interaction can depend on the presence of a specific ligand or can be blocked by a specific ligand, although some multiprotein complexes are formed only in a ligand-dependent manner. right.
[0026]
Thousands of new protein-protein interactions have been identified using a two-hybrid approach. The methods described herein use each labeled protein to overcome the limitations of such methods and reveal the relative strength of each interaction as well as the interacting partners. Can be used to screen proteome arrays. The method can also be used to identify components of multiprotein complexes and to further elucidate where their collection is ligand dependent.
[0027]
An example of the use of this method to elucidate novel protein-protein interactions is the identification of signaling partners in the cytoplasmic domain of certain cell surface receptors that are thought to be involved in disease states. I will. Since such protein-protein interactions may immediately indicate potential therapeutic targets, the identification of such signaling partners will be directly related to the pharmaceutical perspective.
[0028]
Protein-DNA interaction
Although approximately 10% of all genes in the human genome are estimated to encode transcription factors, only a small percentage of these have been identified so far. Binding of specific transcription factors to DNA enhancer elements (often occurring in response to external stimuli) is essential for the formation of an enhancesome complex, which is Switch on gene expression. There are various points where gene expression can in principle be affected by the administration of a drug. Drugs can block protein or small molecule binding to cell surface receptors, thereby blocking the signal transduction cascade from the beginning; drugs can block protein-protein interactions or be in signal transduction cascades Alternatively, the enzyme activity can be inhibited; alternatively, the drug can block the formation of specific protein-DNA or protein-protein interactions in the enhanceosome complex. As an example here, the transcription factor NF-κB is involved in various cellular processes ranging from immune and inflammatory responses, limb development, septic shock, asthma, and production of HIV propeptides. Most of the intracellular signaling cascades in NF-κB activation are common to all these processes and are therefore not a practical target for treatment. Therefore, the difference between the reactions lies in either the initial ligand-receptor interaction or the formation of a specific enhanceosome complex. NF-κB is a candidate target for therapy because it is known to bind to at least 14 different enhancer elements and enhanceosome complexes. However, visualization of each enhanceosome complex requires knowledge of the number of each DNA binding protein involved and their mutual protein-protein action. The method can be used to directly solve any of these problems. Proteome arrays can be screened with specific DNA probes to identify novel DNA binding proteins. Alternatively, the proteome array can be screened with the transactivation domain of a particular transcription factor to identify other proteins that interact with it. Such screening cross-correlation should be able to identify novel components of specific enhanceosome complexes.
[0029]
The protein array created by the method of the present invention will also allow the selection of molecules that recognize each protein displayed in the array. In a preferred embodiment, the selected molecule is an antibody or antibody-like protein and will be covalently linked to the mRNA displayed or encoded on the phage or ribosome.
[0030]
Thus, the antibody library displayed on the phage is applied to each immobilized protein in the array, and unbound antibodies can be removed by washing. Subsequently, the selected phage can be recovered and used to infect bacteria according to normal procedures. Bacteria infected with phage can subsequently produce phage particles that display the selected antibody for further rounds of selection, or produce soluble antibody fragments for direct use. As used herein, the term “antibody” or “antibody fragment” means a single chain Fvs, FAB fragment, each light chain or heavy chain fragment derived from a mouse, human, camel, or other organism.
[0031]
In a preferred embodiment, the protein array is such that after the selection step, phage particles can be recovered by adding appropriate bacterial cells to each well, where the bacterial cells will be infected with the selected phage. It could be present in a microwell format. Subsequently, while maintaining the physical separation of the antibody fragments selected for each immobilized protein in the array, a growth medium is added to each well to propagate the infected bacteria and antibody Fragments can be expressed. If desired, new phage particles produced by the infected bacteria can be used in subsequent rounds of selection. Such an operation is now a common operation for selecting polyclonal or monoclonal antibody fragments against a single purified and immobilized protein. In fact, the original protein array can then be used to generate polyclonal or monoclonal antibody fragments against thousands of correctly folded proteins in parallel, in large numbers using otherwise standard in vitro antibody selection methods. It will be possible.
[0032]
The selected, soluble and expressed antibody fragments obtained from each well of the original array are fixed at a single fixed position in the original array with the antibody fragments present in each position of the new array. As such, it can be immobilized in a new spatially defined array so that it is selected for the conjugated protein. The array of antibodies thus produced will contain either polyclonal or monoclonal antibody fragments at each position, depending on the number of rounds of selection performed prior to immobilization of the soluble antibody fragments.
[0033]
Such an array of antibodies will have a number of potential uses, including capturing each protein of the crude cell or tissue lysate for differential expression monitoring of the relevant proteome. Alternatively, proteins captured by antibodies can be screened directly for ligand-binding function. In general, one monoclonal antibody can bind to a target protein so as to block the function of the protein, while another monoclonal antibody can bind without blocking function. In a massively parallel approach, it is clearly impractical to separately assess the ability of all monoclonal antibodies to all proteins in the proteome to bind without affecting function. However, a group of polyclonal antibodies against all proteins in the proteome may contain each antibody with the desired ability to bind without affecting function, and even all post-translational modifications of a protein. Each antibody that recognizes. For this reason, in general, an array of polyclonal antibodies would be advantageous for direct screening for the function of the captured protein rather than an array of monoclonal antibodies prepared as described above.
[0034]
Compared to the original protein array, an antibody array made by the method described herein will have the advantage that all proteins immobilized on the array will be stable under similar conditions. Let's go. The protein captured from the crude cell or tissue lysate will be naturally expressed rather than recombinant. Furthermore, the captured proteins can be screened directly for function or ligand binding, etc., after capture from the crude cell or tissue lysate, which should help maintain function.
[0035]
Thus, in a further aspect, the present invention provides:
(1) A method for screening the biological activity of one or more compounds, the step of contacting said one or more compounds with a protein array as described herein, and said protein to said protein in said array Measuring the binding of one or more compounds;
(2) A method of screening for a specific protein-protein interaction of one or more proteins, wherein the one or more proteins, eg, cell surface receptors, are contacted with an array described herein; Measuring the binding of the one or more specific proteins to the protein of the array;
(3) A method of screening for specific protein-nucleic acid interactions of one or more proteins, the step of contacting said one or more nucleic acid probes with an array described herein, and said array of said probes Measuring the binding of said protein to said protein;
(4) use of the arrays described herein in rapid screening of compounds, proteins, or nucleic acids;
(5) Use of the array described herein in screening for molecules that recognize each protein in the array, wherein said molecule is preferably an antibody;
(6) A method of making an antibody array, wherein the protein array described herein is bound to an antibody library such that one or more proteins in the protein array bind to at least one antibody in the antibody library. And removing all unbound antibodies; and immobilizing antibodies bound to proteins in said protein array: and
(7) A method of screening for protein function or abundance, comprising the step of contacting the antibody array described herein with a mixture of one or more proteins.
[0036]
Methods (1), (2), (3), and (6) may first include providing an array according to one or more methods of the present invention.
[0037]
Preferred features of each aspect of the invention apply to each other, mutatis mutandis.
[0038]
The present invention will now be described with reference to the following examples, which should not be construed as limiting the scope of the invention in any way.
[0039]
Example 1
(A) Construction of vector (see FIG. 1a)
We constructed a vector pMM106H derived from pUC19 containing a strong hybrid promoter (Ptrc) to induce gene expression cloned into the NcoI site immediately downstream of the promoter sequence. We inserted a 676 bp nonsense DNA sequence as a stuffer fragment between the NcoI site and the downstream HpaI site. HpaI is a blunt end cleaving enzyme, and if the reading frame is on the first base of the blunt end, it is arranged to excise the vector so that the downstream DNA encodes polyasparagine and hexahistidine peptide Has been. The hexahistidine tag is followed by an amber stop codon (TAG) followed by a gene encoding the green fluorescent protein (GFP) of the jellyfish Aequorea victoria. A gene cloned as an NcoI / blunt end fragment in pMM106H will only yield a fusion to the His tag and GFP when the correct reading frame is created at the HpaI site during cloning. Here, GFP is used as a reporter gene to facilitate the visual screening of clones expressing His tags, but since GFP is active only when folded into the correct conformation, It also gives the correct folding indicator. The amber stop codon will result in a small amount of full-length fusion protein to visualize green colonies, but many of the fusion proteins appear to terminate immediately after the His tag so that they can then be used for immobilization and enzyme assays. Can be used. The construction of pMM106H was confirmed by sequencing.
[0040]
We first prepared the primers “GTSfwd2” (5′-ATG CTG CAG ACG TCA ACA GTA TCC ATG GCC CCT ATA CTA GG-3 ′) and “GSTHindIII” (5′-GCG AGG AAG CTT GTC AAT CAGT GATC A A second vector pGSTN was constructed by PCR amplification of the Glutathione S transferase (GST) gene of Schistosoma japonicum from pGEX-2T (Pharmacia) using CCC G-3 ′) under standard conditions. These primers introduce an NcoI restriction site at the start codon of GST, mutate the second residue of GST from serine to alanine, and within the multiple cloning site 3 'of the GST gene followed by a HindIII restriction site. Introduce a stop codon. Subsequently, the PCR product was cloned as an NcoI / HindIII fragment into pTrcHisA (Invitrogen) previously digested with NcoI / HindIII to generate pGSTN under standard conditions.
[0041]
(B) PCR amplification and exonuclease digestion of gene before tag addition (see FIG. 1b)
We have custom-designed primers “ST forward” (5′-ATG CTG ACG TCA TGA GGC CCA TGG GGC CCG GAT AAC AAT TTC ACA that bind 156 bp upstream of the vector start codon and 84 bp downstream of the stop codon, respectively. The GST gene was amplified from the construct pGSTN using polymerase chain reaction with CAG G-3 ′) and “ST reverse” (5′-GCG GAT CCT TGC GGC CGC CAG GCA AAT TCT GTT T-3 ′). 30 cycles of PCR (94 ° C. 1 min; 57 ° C. 1 min; 72 ° C. 2 min) were performed in 4 separate 100 μL reactions. Each PCR reaction contained approximately 20 ng template DNA, 50 pmol of each primer, and 2.5 units of Pwo polymerase. Each PCR reaction was performed using standard buffer (10 mM Tris.HCl pH 8.8, 25 mM KCl, 5 mM (NH ₄ ) ₂ SO ₄ 2 mM MgSO ₄ 10% DMSO). Subsequently, each of the four PCR reactions also contained a non-standard deoxynucleotide triphosphate mix as follows:
[0042]
Reaction 1) 200 μM dATP, 200 μM dTTP, 200 μM dCTP, 150 μM dGTP, 50 μM α-S-dGTP;
Reaction 2) 200 μM dATP, 200 μM dTTP, 20 OμM dGTP, 150 μM dCTP, 50 μM α-S-dCTP;
Reaction 3) 200 μM dATP, 200 μM dGTP, 200 μM dCTP, 150 μM dTTP, 50 μM α-S-dTTP;
Reaction 4) 200 μM dGTP, 200 μM dTTP, 20 OμM dCTP, 150 μM dATP, 50 μM α-S-dATP.
[0043]
Inclusion of a single α-thiodeoxynucleotide triphosphate in each specific PCR mix allows random but statistical incorporation of the appropriate α-S-dNTP into the specific final PCR product . Subsequently, each of the four PCR mixes was pooled, purified using the QIA quick PCR cleanup kit (Qiagen) under standard conditions, and completely digested with the restriction enzyme AatII. The resulting approximately 1000 bp PCR product was subsequently gel purified.
[0044]
Subsequently, 5 μg of the digested PCR product was incubated with 375 units of exon clease III for 45 minutes at 37 ° C. in a 50 μL reaction. ExoIII digestion was performed using standard reaction buffer (66 mM Tris.HCl pH 8.0, 6.6 mM MgCl ₂ 5 mM DTT, 50 μg / mL bovine serum albumin). These conditions ensure complete digestion with ExoIII. Subsequently, the enzyme is inactivated by heating to 75 ° C. for 15 minutes. The product of ExoIII digestion is a nested set of deletions obtained from the 3 ′ end of the PCR product. Restriction with AatII then leaves a 3 ′ overhang that resists exonuclease III activity, thus protecting the 5 ′ end of the PCR product from digestion.
[0045]
Since exonuclease III is a non-progressive 3 ′ to 5 ′ exonuclease that is not able to hydrolyze α-thio containing nucleotides, in this protocol, every time ExoIII reaches an α-thio-deoxynucleotide base, The cleavage of the recessed 3 ′ end of the progressing PCR product stops. Thus, as a result of the random incorporation of each α-S-dNTP in the initial stage, a final nested set of deletions is obtained. The ratio of α-S-dNTP to dNTP used in the original PCR amplification is such that the nested deletion envelope spans a 400 bp sized window centered about 100 bp shorter than the original full length PCR product. And determined empirically. We confirmed this by taking a portion of the ExoIII mix and treating the nested deletion with Mung Bean nuclease. This process removes the 5 ′ and 3 ′ overhangs to give a blunt end product, followed by the size of the product on a 1% agarose / TBE gel using a 100 bp DNA ladder as a standard. It was determined.
[0046]
Obviously, in the above operations, many different 3 ′ to 5 ′ nuclease activities can be used to create the necessary set of nested 3 ′ recessed deletions; Are exonuclease III, E.I. Coli's DNA polymerase I, T4 DNA polymerase, T7 DNA polymerase, but are not limited thereto.
[0047]
(C) Specific ligation and PCR amplification to introduce a tag (see FIG. 1c)
In the presence of an approximately 25-fold molar excess of “oligomix”, 5 μL of ExoIII reaction mix was diluted in T4 DNA ligase buffer. The “oligomix” consists of any one of two different oligonucleotide pools. The first pool, “Oligomix A”, contains 12 oligonucleotides, each of which 3 possible stop codons are presented at the 5 ′ end, followed immediately by a degenerate base. The remaining region of each oligo is the same in all 12 cases, two 2 ′ followed by a complementary restriction sequence (5′-GTA AAA CGA CGG CCA GT-3 ′) to the primer “LBM2” at the 3 ′ end. Except for the type IIS restriction enzyme sites (SapI and BpmI), the sequence is effectively random. The sequence of 12 oligos is as follows.
[0048]
[Table 1]

[0049]
The second pool “Oligomix B” consists of three sets of oligonucleotides, each having one of the three stop codons present at the 5 ′ end, immediately followed by six degenerate residues. It is continuing. The remaining region of each oligo is the same in all three sets, contains two type IIS restriction enzyme sites (BpmI and BseRI), followed by a complementary recognition sequence for primer “LMB2” at the 3 ′ end. Yes. The sequences of the three sets of oligos are as follows:
[0050]
[Table 2]

[0051]
Either ExoIII mix + Oligomix A or Oligomix B was annealed at 16 ° C. for 30 minutes, then 400 units of T4 DNA ligase was added, and then the reaction was incubated overnight at 16 ° C. Under standard conditions, the ligation product was purified using the QIA quick PCR cleanup kit (Qiagen) and used as a template in a standard PCR reaction using Pwo polymerase with primers “ST forward” and “LMB2”. Thirty cycles of PCR (94 ° C., 1 minute; 57 ° C., 1 minute; 72 ° C., 2 minutes) were performed to generate PCR products spanning 1000 bp.
[0052]
Both Oligomix A and Oligomix B are antagonistically annealed to the single-stranded DNA region of the original template exposed by ExoIII hydrolysis of one of the duplexes performed in the previous step. can do. Following successful annealing, in order for a successful ligation between any oligo and the template to occur, the oligo anneals with absolute complementarity at its 5 'end, and the duplex template is recessed. It requires that the 3 ′ residue is directly adjacent to the 5 ′ residue of the specifically annealed oligo. PCR using a first primer that binds to the newly ligated oligo and a second primer that binds to the 5 ′ end of the duplex template then selects only the duplex that has undergone such ligation. And it amplifies specifically. Since the 5 ′ end of each of the 12 oligos in the oligo mix A corresponds to a stop codon, as shown in FIG. 1c, among the 12 oligos contained in this mix, at the 5 ′ end Only one that has absolute complementarity to the four base pair recognition sequences, including the first in-frame stop codon of GST and the base immediately 3 ′ of the stop codon, can anneal. The remaining eleven oligos can anneal completely at their 5 'end to another location within the nested set of deletions, although these other specific annealing events are found in the GST gene. Occurs only with an out-of-frame stop codon or a stop codon downstream of the first in-frame stop codon. Ligation can occur wherever the newly annealed oligo is directly adjacent to the 3 ′ recessed residue of the duplex template. This stage of PCR will therefore amplify not only the correct, full-length gene, but also the truncated and extended set of products. The oligos in Oligomix B are expected to react similarly. The 5 ′ end of each of the three sets of oligos in the pool corresponds to a stop codon and is followed by 6 residues that are effectively random. Therefore, the pool will contain a stop codon and one permutation in which the next six residues are perfectly matched to residues downstream of the gene of interest. Compared to 4, this oligo will bind with higher specificity than the corresponding oligo from a pool of 12 oligos, since the complementary portion extends over 9 nucleotides.
[0053]
The theoretical difference between Oligomix A and Oligomix B is in the sequence immediately following the stop codon at the 5 'end of each oligo. In Oligomix A, a single degenerate base is nonsense but followed by a fixed single DNA sequence, so this fixed region consists of four designs at the 5 'end of each oligo. Each “stop” codon within any particular gene (out-of-frame as in-frame) by providing additional, but not intentional, base-pair complementarity beyond The annealing of the oligo mix can be biased so as to give priority to (whether or not). Any bias during annealing can appear downstream of the bias for clones where specific excision and replacement of each stop codon (whether in-frame or out-of-frame) has occurred. Since this frequency of modification at different stop codons may not be desirable, Oligomix B is designed to avoid this as follows. Any stop codon in each gene or library is immediately followed, but will be followed by an unknown sequence. All three stop codons are presented in Oligomix B, and for any stop codon in the gene or library, an oligo that exactly matches the stop codon and its exact downstream sequence Each is immediately followed by all possible hexanucleotide sequences (ie, by random hexamer sequences) so that there is one in oligomix B and a total of 9 base pair complementarity is obtained. . Since this seems to be true for all stop codons, Oligomix B should therefore be free from any such bias that can occur with Oligomix A.
[0054]
We have found that the entire process of ExoIII digestion, annealing and ligation and specific PCR amplification of the oligomix is highly reproducible. As a control in this operation, it was shown that if any one of exonuclease III, T4 DNA ligase, or any oligomix was omitted, no PCR product was obtained. This indicates that this process is highly selective.
[0055]
It will also be appreciated that T4 DNA ligase can be replaced by a number of different DNA ligases that may exhibit different specificities, such as Taq DNA ligase, or Tsc DNA ligase.
[0056]
(D) Deformation of operation
Obviously, a number of different 3 ′ to 5 ′ nuclease activities can be used in the above procedure to create the necessary nested set of 3 ′ recessed deletions. These include exonuclease III, E.I. Examples include, but are not limited to, E. coli DNA polymerase I, T4 DNA polymerase, and T7 DNA polymerase.
[0057]
A clear variation on the original procedure to create a full-length insert in which the stop or start codon has been correctly removed is a 5 'recessed deletion in the PCR product containing the coding sequence. Manufacturing a nested set of. This is known as lambda exonuclease, E. coli. It could be performed using nucleases such as E. coli DNA polymerase I, Taq DNA polymerase, T7 gene 6 exon clease. Thus, for example, the original PCR product can be digested with λ exonuclease that removes nucleotides from one strand of dsDNA in the 5 ′ to 3 ′ direction. Since the enzyme recognizes only the 5 ′ phosphate group as a substrate, one of the two strands can be protected by incorporating a 5 ′ hydroxyl group at the end of the appropriate primer in the first PCR amplification. A nested set of deletions can thus be made as described above for exonuclease III, except that the opposite strand is digested to leave a 5 ′ recessed end. Subsequently, the complementary part of the stop codon is present at the 3 ′ end of the oligo, immediately preceded by 6 randomized residues, and preceded by a predetermined sequence encoding a type IIS restriction site. Can be annealed to the exposed single stranded region. One oligo from the mixture specifically anneals to each exposed stop codon and polymerizes in the 5 ′ to 3 ′ direction. It acts as a primer for E. coli DNA polymerase I and its 5 ′ to 3 ′ exonuclease activity will digest the preceding strand. This new double-stranded DNA fragment is then specifically amplified using a primer that binds to the 5 'end of the original PCR product and a primer that specifically binds to the newly annealed and extended oligo. Can do. In this operation, the 3 ′ end of the annealed oligo does not need to be directly adjacent to the residue in which the 5 ′ of the duplex template is recessed. Oligo from the mixture can anneal to any complementary region on the exposed single stranded DNA template in a manner similar to random hexamer priming, but importantly, strand extension is This will only occur if the 3 'end anneals with absolute complementarity and the 3' end of the oligo in the mixture is designed to be complementary only to the stop codon. Thus, primer extension and subsequent PCR amplification will occur only when the 3 ′ end of the oligo specifically binds to the matching stop codon. Coupling of the out-of-frame stop codon to the codon downstream of the true stop codon of the gene of interest also occurs, not only the exact full-length gene, but also PCR amplification of a group of truncated and extended products. can get.
[0058]
(E) Cloning and analysis of PCR products (see FIG. 1d)
A PCR product (approximately 5 μg) ranging in size from 800 to 1000 bp was purified using the QIAquick PCR purification kit (Qiagen) and then completely digested using the type IIS restriction enzyme BpmI. This enzyme remotely but specifically cleaves 14 bases away from its recognition sequence present in one strand and 16 bp on the other strand, leaving a 3 ′ recessed end. For this reason, this restriction modification specifically excises a stop codon from the PCR product (from the “oligomix” the 5 ′ end of each oligo is annealed and appropriately ligated here in the previous step). .
[0059]
Subsequently, Mung Bean Nuclease is used to remove the recessed 3 ′ end of the digested product under standard conditions, creating a blunt end at the 3 ′ end of the PCR product. Under standard conditions, DNA was purified using the QIAquick PCR purification kit (Qiagen) followed by complete digestion with the restriction enzyme NcoI. Subsequently, restriction DNA fragments ranging from 800 bp to 1000 bp were purified on a 1% agarose / TBE gel using the QIAquick gel extraction kit (Qiagen). The vector pMM106H (3 μg) was completely digested with the restriction enzymes NcoI and HpaI, and the 2870 bp backbone fragment was purified by gel. Subsequently, the vector DNA and the restriction PCR product are ligated together under standard conditions and E. coli is used with the ligation mixture. After transformation of E. coli DH5α cells, they were harvested and seeded on LB plates containing 100 μg / mL carbenicillin.
[0060]
This cloning operation was performed on the complete set of PCR products obtained in the previous step. However, only PCR products derived from the specific annealing and ligation of the oligo to the first in-frame stop codon can result in an in-frame fusion to the hexahistidine tag and GFP after the cloning step via this procedure. It should be possible. All other PCR products cloned in this way should only induce an out-of-frame fusion to the hexahistidine tag and GFP. This is because the ligation of the blunt end of the PCR product to the blunt end of the vector results in a genetic fusion directed by the original open reading frame of the stop codon from which the translation of the open reading frame of the downstream vector DNA was excised. Something happens. If the stop codon is out-of-frame with respect to the GST gene, the newly added sequence encoding hexahistidine will also be out-of-frame with respect to the GST gene, but the stop codon is in-frame with respect to GST. If so, the sequence encoding the newly added hexahistidine would also be in-frame with respect to the GST gene. However, only the PCR product in which the specifically excised stop codon was the first in-frame stop codon of GST was added with hexahistidine (and GFP) when the DNA was transcribed and translated. Can produce protein. The only protein with added hexahistidine (and GFP) that can result from the entire specific process will therefore necessarily be a full-length GST fusion to polyasparagine, hexahistidine.
[0061]
Colonies obtained from the cloning operation described above were visualized at 365 nm to identify green fluorescent colonies. Ninety colonies (both white and green) were randomly picked, replica plated, and analyzed by colony western blot under standard conditions using anti-His tag and anti-GST antibody. Since the anti-His tag antibody appears to only bind to colonies that express proteins with added hexahistidine, Western blots are directly related to the number of colonies expressing in-frame fusions to the hexahistidine tag. Information. In contrast, the anti-GST antibody binds near the C-terminus of the GST protein, and therefore recognizes only colonies that express full-length or almost full-length GST protein. We identified colonies containing proteins that were positively recognized by both anti-His tag and anti-GST antibody. DNA from these colonies was amplified, purified, and sequenced. The sequencing data specifically cuts out two complete in-frame fusions to full-length GST (ie, the first in-frame stop codon of the original GST gene, and the in-frame polyasparagine, hexahistidine tag The presence of the replaced clone) was confirmed. The success rate of modification we obtained through this entire operation is therefore about 2.2%. Both positive clones were found to fluoresce green when exposed to long-wavelength ultraviolet light because a sufficient amount of full-length GST-hexahistidine-GFP fusion was expressed. Therefore, in all further experiments, only colonies emitting green fluorescence were selected for further analysis by Western blot. We have found that approximately 70-80% green fluorescent colonies express proteins that are recognized by anti-His tag antibodies. In the remaining 20-30% cases, translation is initiated from the first ATG of the GFP gene, possibly independent of the hexahistidine tag, with the aid of a pseudo-ribosome binding site introduced by the cloned insert. .
[0062]
We have amplified, purified and sequenced plasmid DNA from green fluorescent colonies expressing proteins recognized by both anti-His tag and anti-GST antibody. For inserts prepared using both Oligomix A and Oligomix B, the success rate of modification was found to be about 25% of all green colonies. Therefore, using the GFP gene as a marker for in-frame fusions increases the efficiency of detecting the correct clone by about 10-fold.
[0063]
Mang bean nuclease can be replaced by a number of different single-stranded nucleases that may exhibit different specificities, such as S1 nuclease or RNase T, and a number of different type IIS restriction enzymes, such as SapI, placed in the appropriate positions. It will also be appreciated that can be used in place of BpmI.
[0064]
(F) Immobilization and functional analysis of tagged proteins (see FIG. 1e)
One of the GST plasmids added with full-length hexahistidine prepared by the method described above. E. coli DH5α cells were transformed. After growing a single carbenicillin resistant colony in 10 mL liquid medium to mid-log phase, supplement with 100 μM IPTG to induce the expression of GST with added hexahistidine. After an additional 4 hours of growth, cells were harvested and lysed by freeze-thawing / lysozyme. The crude lysate SDS-PAGE, along with the oversized protein of the expected size (27 kDa) (accounted for about 20% of the total soluble protein), together with the small amount of 54 kDa GST-hexahistidine produced by the suppression of amber. -GFP fusion was shown. Subsequently, the crude solubilized solution (500 μL; 100 μg) was mixed with nickel-NTA magnetic beads (50 μL; protein to which 15 μg of binding ability of hexahistidine was added), and the beads were recovered by sedimentation under a magnetic field. The supernatant was discarded and the beads were washed and then resuspended in glutathione S transferase assay buffer containing 1 mM glutathione and 1-chloro-2,4-dinitrobenzene. End point assay data was collected after 30 minutes at room temperature by measuring absorbance at 340 nm. This wavelength is the λ of the product of the reaction catalyzed by GST. _max It corresponds to.
[0065]
As controls, cultures of DH5α containing either the parent vector (pMM106H) or a plasmid encoding an irrelevant His-added protein (alanine racemase) were grown, induced, harvested, lysed and assayed in parallel. GST activity was detected only on beads mixed with a crude lysate containing His-added GST, and clearly the observed GST activity was specifically obtained from immobilized His-added GST. Furthermore, it was demonstrated that the protein retains its activity when specifically immobilized.
[0066]
After completion of the enzyme assay, proteins were eluted from the magnetic beads by adding a buffer containing 100 mM imidazole and analyzed by SDS-PAGE. This indicates that the sample with a positive activity assay result contains a single immobilized protein of the exact size (27 kDa) expected for glutathione S-transferase, It was confirmed that the activity observed on the beads was solely due to this recombinant His-added protein.
[0067]
Example 2
(A) Modification, immobilization, and assay of GST using two different tags
Following the procedure described in Example 1 for modification of glutathione-S-transferase with a hexahistidine tag, we demonstrated that the procedure was not affected by the exact nature of the tag added.
[0068]
First, the 676 bp NcoI / HpaI nonsense DNA stuffer fragment was replaced with a 300 bp NcoI / HpaI fragment derived from the Escherichia coli gdhA gene, and the hexahistidine tag was specific to the FLAG peptide (epitope tag) and the StrepII tag (streptavidin). Two additional vectors identical to pMM106H were constructed, except that they were replaced by either These vectors were designated pMM104F and pMM104S, respectively. These vectors (3 μg each) were separately digested separately using NcoI and HpaI, the 2870 bp backbone fragment was gel purified and ligated to the GST fragment resulting from Oligomix A as described in Example 1. After using a combination of anti-FLAG and anti-GST antibodies, sequencing was performed and clones were identified that contained a complete in-frame fusion of the full-length GST gene to the FLAG tag. Similarly, after using a combination of an anti-GST antibody and a streptavidin-horseradish peroxidase conjugate, a clone was identified that contained a complete in-frame fusion of the full-length GST gene to a StrepII tag. It was. In all examples, the frequency with which full-length in-frame fusions were found was the same as that determined in Example 1 (within experimental error).
[0069]
Immobilize these clones substantially as described in Example 1 except that the immobilization substrates were 96-well plates each coated with anti-FLAG antibody and streptavidin-coated magnetic beads. Used in experiments. As in Example 1, we can demonstrate the specific immobilization of the fusion protein via these tags, and when we immobilize via either the FLAG or Strep tag, It could be shown that the GST fusion retains its activity.
[0070]
Example 3
(A) Modification of another protein using a hexahistidine tag
Following the procedure described in Example 1 for glutathione-S-transferase, we demonstrated that this procedure was not affected by the exact gene being handled.
[0071]
Thus, starting from a plasmid encoding the human transcription factor NF-κB p50, and unless otherwise specified, following the procedure described in Example 1 (using Oligomix A), we We were able to demonstrate that NF-κB p50 was modified such that the in-frame stop codon was specifically excised and replaced by an in-frame fusion to DNA encoding a polyasparagine, hexahistidine tag. . Colony Western blots using anti-His tag antibodies allowed identification of clones expressing proteins with added hexahistidine. DNA from these colonies was amplified, purified and sequenced. By sequence data, some clones encode complete in-frame fusions to full-length NF-κB (ie, clones in which the stop codon was specifically excised and replaced by an in-frame hexahistidine tag). Confirmed to do. In the case of NF-κB p50, the frequency at which full-length in-frame fusions are found is 1.1%, which is close to that determined in Example 1 for GST and within experimental error.
[0072]
(B) Immobilization and functional analysis of NF-κB p50 to which hexahistidine was added. One of the NF-κB plasmids to which full-length hexahistidine was added prepared by the above-described method. Coli DH5α cells were transformed. Single carbenicillin resistant colonies are grown in 10 mL liquid medium until mid-log phase and then supplemented with 100 μM IPTG to induce the expression of NF-κB p50 supplemented with hexahistidine. After an additional 4 hours of growth, the cells were collected and lysed by sonication. The crude lysate SDS-PAGE showed the oversized protein of the expected size (38 kDa) (accounting for about 5% of the total soluble protein) as well as the small amount of 65 kDa NF-κB p50 produced by amber suppression. -Hexahistidine-GFP fusion was shown.
[0073]
[Table 3]

[0074]
The 3 ′ base of the double-stranded oligonucleotide “κB motif” (containing the palindromic binding site for NF-κB p50) was labeled with digoxigenin using a 3 terminal transferase (Boehringer Mannheim) under standard conditions.
[0075]
Protein lysates were prepared using the lysozyme / freeze lysis method in PBS (phosphate buffered saline pH 7.5) containing 5 mM β-mercaptoethanol. 200 μL of soluble protein lysate from each clone was added to Ni-NTA coated microwells and incubated for 45 minutes at room temperature. At the end of the incubation period, the wells were washed 3 times with PBST (PBS containing 0.02% Triton X-100) to remove any unbound protein. The wells were washed 3 times with a DNA binding buffer (10 mM Tris.HCl pH 7.4, 75 mM KCl containing 5 mM β-mercaptoethanol) with a 1 minute soak time. A κB motif (2 pmol) labeled with 3 ′ digoxigenin was added to the wells in 200 μL of DNA binding buffer containing 1 μg of poly (dI-dC) non-specific DNA. After further incubation for 30 minutes, 10 mM Tris. Containing 0.02% Triton X-100. Unbound DNA was removed by washing the wells 3 times with HCl pH 7.4, 25 mM KCl. The anti-digoxigenin antibody-alkaline phosphatase conjugate is 150 mU / mL in “antibody dilution buffer” (10 mM Tris.HCl pH 7.4, 25 mM KCl) supplemented with 0.2% bovine serum albumin. Diluted. Subsequently, diluted antibody (200 μL) was added to the microwells. After leaving at room temperature for 30 minutes, unbound antibody was removed by washing the microwells with “antibody dilution buffer” (3 × 350 μL) supplemented with 0.02% Triton X-100. Subsequently, 200 μL of buffer (100 mM Tris.HCl pH 9.5, 100 mM NaCl, 50 mM MgCl) containing 250 μM p-nitrophenyl phosphate (pNPP) which is a substrate for alkaline phosphatase. ₂ ) Were added to the wells and the reaction was allowed to proceed overnight at room temperature, after which the yellow color (p-nitrophenol, corresponding to product formation) in each well was quantified at 405 nm. Since the background rate of hydrolysis of the substrate pNPP was low, the appearance of a yellow color in the wells immediately revealed a positive assay result.
As a control for this assay, we omitted either the crude lysate, or labeled oligonucleotide, or antibody, or added a 20-fold excess of unlabeled double-stranded oligo, or crude lysate. NF-κB p50 supplemented with hexahistidine containing HT5 was replaced with an equal amount of crude cell lysate obtained from DH5α cells expressing GST supplemented with hexahistidine in the same vector background .
[0076]
In this assay, NF-κB p50 first binds to the labeled oligonucleotide through a specific binding site. Subsequently, the protein-DNA complex was immobilized in a microwell via a hexahistidine tag, and all unbound labeled oligos were present (in the labeled oligo and the crude lysate). All other proteins (including complexes with other DNA binding proteins) were washed away. The antibody-conjugate recognizes the label on the oligo, but does not recognize the protein to which hexahistidine is added, so when NF-κB p50 is immobilized via the tag, the NF-κB p50-DNA interaction Is maintained, only positive signals in the assay can be observed. If this interaction is not maintained, the oligo will be lost during the washing step and no color change will be observed.
[0077]
We have found that the yellow product is detected only in microwells containing NF-κB p50 crude lysate with hexahistidine added and digoxigenin-labeled oligonucleotide, which contains anti-digoxigenin antibody-alkaline A phosphatase conjugate was added. As a result, the observed color change was specifically obtained from the immobilized NF-κB p50 oligonucleotide complex. Further, when the NF-κB p50 was specifically immobilized, It has been demonstrated to retain activity.
[0078]
Example 4
Another operation to generate a full-length gene or cDNA with the stop codon removed correctly is to set a strand-specific nested state of the 5′-recessed deletion in the PCR product containing the coding sequence. Includes creating. This is known as lambda exonuclease, E. coli. It could be done using a nuclease such as E. coli DNA polymerase I, Taq DNA polymerase, or T7 gene 6 exon clease. Thus, for example, the original PCR product could be performed using any 5 ′ to 3 ′ exonuclease. Once a nested set of deletions that are 5 ′ recessed is created, the oligomix can be annealed to the exposed single-stranded region. In the oligomix, the complementary portion of each stop codon is present at the 3 ′ end, immediately preceded by 6 random residues, and preceded by a predetermined sequence encoding a type IIS restriction site and the primer “LBM2” ( Complementary recognition sequences for (see Example 1) consist of a group of preceding oligos. For this reason, the sequences of the oligo groups are as follows.
[0079]
[Table 4]

[0080]
One oligo from the mixture specifically anneals to each stop codon exposed on the sense strand, like Taq polymerase, or E. coli. It would be a primer for a DNA polymerase that has activity to displace either strand, such as 5 'to 3' exonuclease activity, such as E. coli DNA polymerase I. The newly generated double-stranded DNA fragment is then specifically amplified using a primer that binds to the 5 ′ end of the original PCR product and a primer that specifically binds to the newly annealed and extended oligo. be able to. In this operation, the 3 ′ end of the annealed oligo does not need to be directly adjacent to the residue in which the 5 ′ of the duplex template is recessed. Oligo from the mixture can anneal to any complementary region on the exposed single stranded DNA template in a manner similar to random hexamer priming, but importantly, strand extension is This will only occur if the 3 'end anneals with absolute complementarity. Since the 3 ′ end of the oligo in the mixture is designed to be complementary to the stop codon, primer extension and subsequent only when the 3 ′ end of the oligo specifically binds to the matching stop codon. PCR amplification will occur. Coupling of the out-of-frame stop codon to the codon downstream of the true stop codon of the gene of interest also occurs, not only the exact full-length gene, but also PCR amplification of a group of truncated and extended products. can get. However, they do not appear to produce in-frame fusions to the peptide tag and can therefore be easily screened in subsequent steps.
[0081]
As described, performing annealing and extension operations on a nested set of 5 ′ deletions will perform annealing and extension operations on a completely single-stranded DNA or RNA molecule encompassing the entire coding region. There are significant advantages over This is due to the fact that the number of sites that the primer can specifically anneal before extension is greatly limited by the presence of the double stranded portion of the nested set of deletions. The single stranded portion of the nested set of deletions mainly includes the 3 ′ untranslated region of the gene, which favors stop codons outside the coding sequence and even longer extension products. Will have an effect of strongly biasing the extension product. These factors will act to significantly increase the frequency with which the first in-frame stop codon is specifically removed by the entire procedure. In fact, we tried to anneal and extend the oligo described in step (b) below using a fully single-stranded coding region template, No modified full-length clones could be identified. In contrast, the results of the operations detailed below show that the use of a nested set of 5 'deletions as a template for the annealing and extension steps makes it possible to identify the first in-frame stop codon of the coding sequence. Clearly demonstrates effective and efficient in promoting full removal to obtain full length in-frame fusions to polypeptide tags.
[0082]
(A) PCR amplification and exonuclease digestion of genes prior to tag addition
Thus, we have described exactly in steps (a) and (b) of Example 1 in the amplification, except here the “ST reverse” primer is 5 ′ phosphorylated. First, the first PCR amplification of the GST gene was performed. Subsequently, standard reaction buffer (67 mM glycine-KOH pH 9.4, 2.5 mM MgCl) was used using 2.5 units of λ exonuclease (Novagen Strandase kit). ₂ , 50 μg / mL bovine serum albumin), the purified PCR product was digested at 37 ° C. for 40 minutes. Subsequently, the enzyme was inactivated by heating to 75 ° C. for 15 minutes. Since the λ exonuclease enzyme recognizes only the 5 ′ phosphate group as a substrate, the sense strand is protected from digestion, so the digestion product is a deletion obtained from the 5 ′ end of the antisense strand of the PCR product. Is a set of nested states.
[0083]
(B) Specific ligation, extension, and amplification to introduce a tag
In the presence of 250 μM dNTPs and about 25-fold molar excess of the following degenerate oligos group, 5 μL of λ exonuclease reaction mix was diluted in E. coli DNA polymerase I reaction buffer.
[0084]
[Table 5]

[0085]
After annealing the digested fragments and oligos at 37 ° C. for 30 minutes, 5 units of E. coli. E. coli DNA polymerase I was added, and then the reaction solution was incubated at 37 ° C. for 3 hours. Under standard conditions, the extension product was purified using the QIA Quick PCR Cleanup Kit (Qiagen) and used as a template in a standard PCR reaction using Pwo polymerase with primers “ST Forward” and “LMB2”. . Thirty cycles of PCR (94 ° C., 1 minute; 57 ° C., 1 minute; 72 ° C., 2 minutes) were performed to generate PCR products spanning 1000 bp.
[0086]
The PCR product resulting from the above operation was digested and cloned into the vector pMM106H described in step (e) of Example 1. Colonies obtained from the cloning operation were visualized at 365 nm to identify colonies emitting green fluorescence. 73 such colonies were selected, replica plated, and analyzed by colony western blot under standard conditions using anti-His tag and anti-GST antibody. Of the green fluorescent colonies, 58% expressed a protein that was positively recognized by both anti-His tag and anti-GST antibody. DNA from 15 colonies that were positive for both anti-His and anti-GST was amplified, purified and sequenced. Sequencing data specifically cuts out 10 complete in-frame fusions to full-length GST (ie, the first in-frame stop codon of the original GST gene, in-frame polyasparagine, hexahistidine tag The presence of clones replaced by The success rate of modification we obtained through this entire operation is therefore 39% of the total number of colonies that fluoresce green.
[0087]
Example 5
(A) Identification of a protein obtained from a pool of 11 genes
We applied the above procedure exactly as described in Example 1 except that it was identified in 11 different gene pools listed in the table below. We create the resulting array of specifically modified proteins so that each position in the array corresponds to a single recombinant protein immobilized through a tag added as a result of the manipulation did. Subsequently, we screened the array by functional assay and successfully identified each protein component of the pool.
[0088]
[Table 6]

[0089]
In particular, to mimic the situation encountered in a cDNA library, all 11 genes were first subcloned into the same pTrcHisA vector backbone. The primers “ST forward” and “ST reverse” described in Example 1 were designed to be universal primers for amplifying the gene encoded in the pTrcHisA vector backbone.
The primer “ST forward” was designed to encode a number of restriction sites as follows.
[0090]
[Table 7]

[0091]
Thus, either the restriction enzyme AatII or SfiI can be used for the purpose of creating a 3 ′ overhang for protecting the exonuclease. For the purpose of performing directed cloning at the end of the modification operation, in this example, it seems statistically more likely to cleave in the library, so it matches the NcoI cloning site in the tag vector pMM106H used here. We chose to use BspHI because it produces sticky ends that do not cleave inside any of the 11 genes in this pool. Obviously, in principle, any primer encoding a restriction site can be used provided that the tag vector contains an equivalent cloning site downstream of the promoter. SfiI has an 8 bp recognition sequence, so the random frequency of SfiI sites within a gene is much lower (1/6) compared to the frequency of 6 bp recognition sequences such as BspHI (1/4096). .5x10 ⁴ So will have significant advantages in this regard in larger library formats.
[0092]
Tag vector pMM106H is an “ATG” vector, ie, the 5 ′ cloning site (NcoI) overlaps with the ATG start codon located downstream of the ribosome binding site (RBS) for expressing the native protein. However, in the procedure described here, we do not rely on cloned genes that have a common restriction site at the start codon. Instead, we simply rely on a promoter encoded in the vector that initiates transcription to generate mRNA so that the cloned gene itself provides the signals necessary for initiation of translation. Thus, in this example, all genes in the original pool have a start codon immediately preceding the RBS, regardless of the presence or absence of a cloning site located in the ATG. Primer “ST forward” binds upstream of RBS in all 11 first clones, so subsequent post-modification cloning with any primer encoding a restriction site along with their original RBS and ATG Introduce a newly modified gene into the tag vector, thus ensuring translation initiation. In the cDNA library format, the same situation applies in that all full-length cDNAs appear to have their own 5 ′ untranslated region (UTR) containing the eukaryotic translation initiation signal. Subsequently, in order to obtain an appropriate translation start in this case, it is only necessary to clone the modified cDNA together with its 5′UTR into a eukaryotic vector thus once again giving a transcription initiation signal. is necessary. Therefore, universal PCR primer groups equivalent to the primers used in this example can be used.
[0093]
The modification operation was performed as described in Example 1 using the following modifications. As an initial PCR amplification template using primers “ST forward” and “ST reverse”, an equimolar pool of all 11 genes was used, after which the SfiI enzyme had an 8 bp recognition sequence and Since neither gene was cleaved, the fragment was digested with SfiI to protect the 5 ′ end. In the annealing step, Oligomix A (see Example 1) was used. After the second PCR amplification, the modified fragment was divided into two pots, which were digested separately with BpmI and SapI. Statistically, genes in any library have a small percentage of SapI sites (7 bp recognition sequence; random occurrence probability = 1 / 16,384) or BpmI sites (6 bp recognition sequence; random occurrence probability = 1/4, 096), but the proportion of both will be much less (random probability = 1 / 6.7 × 10 ⁷ ). Two Type II restriction enzymes thus effectively ensure that any specific modification of any full-length gene is not ruled out by the presence of one or other restriction sites within that gene. Used separately.
[0094]
Subsequently, the digested fragments from the two pots were pooled for treatment with Mung Bean Nuclese and digestion with BspHI. The resulting fragment was gel-purified into four different size ranges and the vector pMM106H (by itself, completely digested with NcoI and HpaI to avoid preferential ligation of the smaller insert, Purified separately). Transformed cells were visualized under UV light (365 nm), and green fluorescent colonies were selected with the naked eye for analysis by Western blot. Approximately 30% of the total number of transformed colonies emitted green fluorescence, of which 73% expressed a protein that was recognized by the anti-His tag antibody. 190 green His-positive colonies were planted in 1.5 mL liquid medium in a deep 96-well block and grown overnight. Cells were harvested by centrifugation and lysed by freeze-thawing / lysozyme. Subsequently, a crude lysate was added to each well of a nickel NTA-coated 96-well plate, unbound protein was removed by washing, and the His-added recombinant protein was immobilized on the well. Subsequently, NF-κB activity of the immobilized protein was assayed using the assay described in Example 2 and wells containing positive “hits” were identified by the appearance of a yellow coloration. Three clones showed positive “κB-motif” DNA binding activity. Further characterization of the positive clone showed that one clone, as expected, encoded the correct full-length in-frame fusion of the NF-κB p50 gene to the hexahistidine tag. The remaining two clones were found to encode related DNA binding proteins that have lower binding affinity but are known to share the same DNA binding specificity as NF-κB p50.
[0095]
Therefore, the results demonstrate that functional testing of the array created by this operation can identify both specific interactions and weaker but specific and biologically relevant interactions. ing. We therefore use this manipulation to create functional protein arrays in a microwell format, and with these arrays we use small pools based on specific protein-ligand interactions. From these results, each protein was successfully identified.
[Brief description of the drawings]
FIG. 1a shows the construction of vector pMM106H.
FIG. 1b shows details of PCR amplification and exonuclease digestion of an example gene (GST) prior to tagging.
FIG. 1c shows details of specific ligation and PCR amplification for introducing tags.
FIG. 1d shows details of PCR product cloning.
FIG. 1e shows the reaction between glutathione and 1-chloro-2,4-dinitrobenzene catalyzed by GST.

Claims (11)

(A) providing a plurality of cDNA molecules encoding a protein in which each protein is tagged with a marker moiety at either the N-terminus or C-terminus;
(B) expressing individual tagged proteins in a spatially separated format;
(C) Each tagged protein is purified and immobilized in a spatially defined format to create a protein array on a solid support , where the tag and solid are both in said purification and immobilization. A method of making a protein array comprising using a specific interaction with a support . The marker portion is
(A) a hexahistidine tag;
(B) a complete protein or protein domain;
(C) the epitope of the antibody;
2. The method of claim 1, wherein the peptide sequence is selected from the group consisting of (d) a biotin analog; and (e) a maltose binding protein domain. The method of claim 1 or 2 , wherein the marker moiety is post-translationally modified. 4. The method of claim 3 , wherein the post-translational modification comprises the addition of biotin or lipid molecules. The method according to any one of claims 1 to 4 , wherein the tagged protein is correctly folded and maintains its function. 6. The method of any one of claims 1-5 , wherein the tagged protein is full length. The cDNA molecule encoding the tagged protein inserts an additional known DNA sequence encoding a marker moiety either immediately after the start codon or immediately before the stop codon present in the one or more DNA molecules. The method according to any one of claims 1 to 6 , which is prepared by a method comprising the following steps:
(A) creating a set of nested deletions from the one or more DNA molecules;
(B) annealing and ligating a first or second set of oligonucleotides to a set of nested deletions;
(I) one or more sequences or complementary sequences of the three possible stop codons are presented in a mixture of the first set of oligonucleotides;
(Ii) one or more sequences or complementary sequences of the three major start codons are presented in a mixture of the second set of oligonucleotides;
(C) specifically amplifying the ligated product by a primer extension reaction using oligonucleotide primers selected to be complementary to the first or second set of oligonucleotides. 7. The method according to any one of claims 1-6 , comprising contacting one or more test compounds with the protein array, and binding of the one or more compounds to the protein in the array. Measuring and thereby screening said one or more compounds for biological activity. 7. The method according to any one of claims 1-6 , wherein one or more proteins, for example, cell surface receptors, are contacted with the protein array, and the one or more specific proteins. Measuring the binding of the array to the protein, thereby screening one or more proteins for specific protein-protein interactions. The method according to claim 1 , wherein one or more nucleic acid probes are brought into contact with the protein array, and the binding of the probes to the protein in the array is measured. And thereby screening one or more proteins for specific protein-nucleic acid interactions. 7. The method of any one of claims 1-6 , wherein the protein array is antibody live such that one or more proteins in the protein array bind to at least one antibody in an antibody library. And a step of removing all unbound antibodies; and immobilizing antibodies bound to the proteins in the protein array, thereby producing an antibody array.

Images