JP2008520208A - Protein backbone and uses thereof - Google Patents

Abstract

The present invention provides thrombospondin, thyroglobulin and trefoil / PD monomer domains, and multimers comprising these monomer domains. Methods, compositions, libraries and cells that express one or more library members, as well as kits and integration systems, are also included in the present invention.

Description

This application claims the benefit of US Provisional Patent Application No. 60 / 628,596, filed on November 16, 2004, and is a US Provisional Patent filed on June 17, 2004. This is a continuation-in-part of application 10/871602, which is a continuation-in-part of US Provisional Application No. 10 / 840,723, filed May 5, 2004, US Provisional Patent Application No. 10 / 693,056 filed on May 24 and Partial Continuation Application of US Provisional Patent Application No. 10 / 693,057 filed on October 24, 2003. Are both continuation-in-part of U.S. Provisional Patent Application No. 10 / 289,660, filed on November 6, 2002, which is a US Provisional Patent Application No. 10 filed on April 26, 2002. This is a continuation-in-part of U.S. Patent No./133,128, which was filed on US Provisional Patent Application No. 60 / 374,107, filed November 26, 2001, filed April 18, 2002. U.S. Provisional Patent Application No. 60 / 333,359, U.S. Provisional Patent Application No. 60 / 337,209 filed on November 19, 2001, and U.S. Provisional Patent Application No. 60 / 286,823 filed on April 26, 2001. All of these applications are incorporated herein by reference in their entirety for all purposes.

BACKGROUND OF THE INVENTION Analysis of protein sequences and three-dimensional structures has revealed that many proteins are composed of several individual monomer domains. Because such proteins are linear mosaics of repetitive building blocks, they are often referred to as “mosaic proteins”. The majority of individual monomer domain proteins are extracellular or constitute the extracellular portion of membrane-bound proteins.

  An important property of an individual monomer domain is its ability to fold independently of other domains within the same protein. Folding these domains may require limited assistance from, for example, chaperonins (eg, receptor binding proteins (RAP)), metal ions, or cofactors. The ability to fold independently prevents misfolding of the domain when inserted into a new protein or new environment. This property has allowed individual monomer domains to be evolutionarily mobile. As a result, individual domains propagate in the course of evolution and are present in other proteins that are not currently relevant. Some domains, including type III fibronectin domains and immunoglobulin-like domains, are present in many proteins, while other domains are found only in a limited number of proteins.

Proteins containing these domains are involved in various processes such as cellular transporters involved in development and neurotransmission, cholesterol transport, signaling and signaling functions. See Herz, (2001) Trends in Neurosciences 24 (4): 193-195 (Non-Patent Document 1); Goldstein and Brown, (2001) Science 292: 1310-1312 (Non-Patent Document 2). The function of individual monomer domains is often specific, but it also contributes to the overall activity of the protein or polypeptide. For example, the LDL receptor class A domain (also called class A module, complementary repeat, or A domain) is a phospholipid membrane that is a γ-carboxyglutamate (Gla) domain found in vitamin K-dependent blood clotting proteins. Involved in ligand binding while being involved in high affinity binding to. Other individual monomer domains include, for example, epidermal growth factor in tissue-type plasminogen activator that mediates binding to hepatocytes and thereby regulates the removal of this fibrinolytic enzyme from circulation (EGF) -like domains and the cytoplasmic tail of the LDL receptor involved in receptor-mediated endocytosis.

Individual proteins can have one or more individual monomer domains. Proteins that contain a large number of repetitive domains are often referred to as mosaic proteins. For example, members of the LDL receptor family include a number of domains belonging to four major families: cysteine-rich A domain repeats, epidermal growth factor precursor-like repeats, transmembrane domains, and cytoplasmic domains. The LDL receptor family includes 1) members that are cell surface receptors; 2) members that recognize extracellular ligands; and 3) members that internalize them for degradation by lysosomes. See Hussain et al., (1999) Annu. Rev. Nutr. 19: 141-72 (Non-Patent Document 3). For example, some members include very low density lipoprotein receptor (VLDL-R), apolipoprotein E receptor 2, LDLR-related protein (LRP), and megalin. Members of the family have the following characteristics: 1) cell surface expression; 2) extracellular domain binding mediated by A domain; 3) calcium requirement for folding and ligand binding; 4) receptor binding protein and Recognition of apolipoprotein (apo) E; 5) Epidermal growth factor (EGF) precursor homology domain including YWTD repeats; 6) Single transmembrane region; and 7) Receptor-mediated endocytosis of various ligands. See Hussain, supra. These families bind several structurally different ligands.

  It would be advantageous to develop a method for creating and optimizing the desired properties of these individual monomer domains. However, although individual monomer domains are often structurally conserved, they are not conserved at the nucleotide or amino acid level, except for certain amino acids, eg, cysteine residues in the A domain. Thus, existing nucleotide recombination methods do not meet the creation and optimization of desirable properties of these individual monomer domains.

  The present invention addresses these and other issues.

Herz, (2001) Trends in Neurosciences 24 (4): 193-195 Goldstein and Brown, (2001) Science 292: 1310-1312 Hussain et al, (1999) Annu. Rev. Nutr. 19: 141-72

SUMMARY OF THE INVENTION The present invention relates to a protein comprising a monomer domain that specifically binds to a target molecule, a polynucleotide encoding the protein, a method of using such a protein, a single quantity for use with such a protein. Methods for identifying body domains and libraries comprising monomer domains are provided.

およびサイログロブリン単量体ドメインは、以下の配列由来のわずか3点の挿入、突然変異、または欠失しか含まない。

式中「x」は任意のアミノ酸である。いくつかの態様において、トロンボスポンジン単量体ドメインは以下の配列を含み：

およびサイログロブリン単量体ドメインは以下の配列を含む。

式中「x」は任意のアミノ酸である。いくつかの態様において、トロンボスポンジン単量体ドメイン配列は、以下の配列由来のわずか3点の挿入、突然変異、または欠失しか含まない。

式中C₁-C₅、C₂-C₆およびC₃-C₄はジスルフィド結合を形成する；サイログロブリン単量体ドメイン配列は、以下の配列由来のわずか3点の挿入、突然変異、または欠失しか含まない。

式中C₁-C₂、C₃-C₄およびC₅-C₆はジスルフィド結合を形成し；αはw、y、f、およびlから選択され；φはd、e、およびnから選択され；ならびに「x」は任意のアミノ酸から選択される。いくつかの態様において、トロンボスポンジン単量体ドメインは以下の配列を含み：

式中C₁-C₅、C₂-C₆およびC₃-C₄はジスルフィド結合を形成する；サイログロブリン単量体ドメインは以下の配列を含む。

式中C₁-C₂、C₃-C₄およびC₅-C₆はジスルフィド結合を形成し；αはw、y、f、およびlから選択され；φはd、e、およびnから選択され；ならびに「x」は任意のアミノ酸から選択される。いくつかの態様において、トロンボスポンジン単量体ドメイン配列は、以下の配列由来のわずか3点の挿入、突然変異、または欠失しか含まず：

サイログロブリン単量体ドメイン配列は、以下の配列由来のわずか3点の挿入、突然変異、または欠失しか含まない。

いくつかの態様において、トロンボスポンジン単量体は以下の配列を含み：

およびサイログロブリン単量体ドメイン配列は以下の配列を含む。

In one embodiment of the invention, a protein comprising a non-natural monomer domain that specifically binds to a target molecule is provided. The monomer domain is 30-100 amino acids in length and is selected from a thrombospondin monomer domain and a thyroglobulin monomer domain. In some embodiments, the monomer domain includes at least one, two, three, or more disulfide bonds. In some embodiments, C ₁ -C ₅ , C ₂ -C ₆ and C ₃ -C ₄ of the thrombospondin monomer domain form a disulfide bond, and C ₁ -C ₂ of the thyroglobulin monomer domain, C ₃ -C ₄ and C ₅ -C ₆ form disulfide bonds. In some embodiments, the thrombospondin monomer domain sequence comprises only 3 insertions, mutations, or deletions from the following sequences:

And the thyroglobulin monomer domain contains only three insertions, mutations or deletions from the following sequences:

In the formula, “x” is any amino acid. In some embodiments, the thrombospondin monomer domain comprises the following sequence:

And the thyroglobulin monomer domain comprises the following sequence:

In the formula, “x” is any amino acid. In some embodiments, the thrombospondin monomer domain sequence contains only 3 insertions, mutations, or deletions from the following sequence:

Where C ₁ -C ₅ , C ₂ -C ₆ and C ₃ -C ₄ form a disulfide bond; the thyroglobulin monomer domain sequence is only 3 insertions, mutations or deletions from the following sequences: Contains only lost.

Where C ₁ -C ₂ , C ₃ -C ₄ and C ₅ -C ₆ form a disulfide bond; α is selected from w, y, f, and l; φ is selected from d, e, and n And “x” is selected from any amino acid. In some embodiments, the thrombospondin monomer domain comprises the following sequence:

Wherein C ₁ -C ₅ , C ₂ -C ₆ and C ₃ -C ₄ form a disulfide bond; the thyroglobulin monomer domain comprises the following sequence:

Where C ₁ -C ₂ , C ₃ -C ₄ and C ₅ -C ₆ form a disulfide bond; α is selected from w, y, f, and l; φ is selected from d, e, and n And “x” is selected from any amino acid. In some embodiments, the thrombospondin monomer domain sequence comprises only 3 insertions, mutations, or deletions from the following sequences:

Thyroglobulin monomer domain sequences contain only 3 insertions, mutations or deletions from the following sequences:

In some embodiments, the thrombospondin monomer comprises the following sequence:

And the thyroglobulin monomer domain sequence comprises the following sequences:

トレフォイル単量体ドメインは以下の配列を含み：
C₁(xx)xxxpxxRxnC₂gx(x)pxitxxxC₃xxxgC₄C₅fdxxx(x)xxxpwC₆f
およびサイログロブリン単量体ドメインは以下の配列を含み：

「x」は任意のアミノ酸である。いくつかの態様において、トロンボスポンジン単量体ドメインのC₁-C₅、C₂-C₆およびC₃-C₄はジスルフィド結合を形成し；ならびにサイログロブリン単量体ドメインのC₁-C₂、C₃-C₄およびC₅-C₆はジスルフィド結合を形成する。いくつかの態様において、トロンボスポンジン単量体ドメインは以下の配列を含み：

式中C₁-C₅、C₂-C₆およびC₃-C₄はジスルフィド結合を形成し；トレフォイル単量体ドメインは以下の配列を含み：

サイログロブリン単量体ドメインは以下の配列を含む。

式中C₁-C₂、C₃-C₄およびC₅-C₆はジスルフィド結合を形成し；αはw、y、f、およびlから選択され；φはd、e、およびnから選択され；ならびに「x」は任意のアミノ酸から選択される。いくつかの態様において、トロンボスポンジン単量体は以下の配列を含み：

トレフォイル単量体ドメインは以下の配列を含み：

およびサイログロブリン単量体は以下の配列を含む。

The present invention also provides a protein comprising a non-natural monomer domain that specifically binds to a target molecule. The target molecule is not bound by a non-natural monomer domain that is at least 75%, 80%, 85%, 90%, 85%, 98%, or 99% identical to the non-natural monomer domain. The monomer domain is selected from a thrombospondin monomer domain, a trefoil monomer domain, and a thyroglobulin monomer domain. In some embodiments, the monomer domain includes at least one, two, three, or more disulfide bonds. In some embodiments, the monomer domain is 30-100 amino acids in length. In some embodiments, the thrombospondin monomer domain comprises the following sequence:

The trefoil monomer domain includes the following sequences:
C ₁ (xx) xxxpxxRxnC ₂ gx (x) pxitxxxC ₃ xxxgC ₄ C ₅ fdxxx (x) xxxpwC ₆ f
And the thyroglobulin monomer domain includes the following sequences:

“X” is any amino acid. In some embodiments, C ₁ -C ₅ , C ₂ -C ₆ and C ₃ -C ₄ of the thrombospondin monomer domain form a disulfide bond; and C ₁ -C ₂ of the thyroglobulin monomer domain , C ₃ -C ₄ and C ₅ -C ₆ form disulfide bonds. In some embodiments, the thrombospondin monomer domain comprises the following sequence:

Wherein C ₁ -C ₅ , C ₂ -C ₆ and C ₃ -C ₄ form a disulfide bond; the trefoil monomer domain comprises the following sequence:

The thyroglobulin monomer domain comprises the following sequence:

Where C ₁ -C ₂ , C ₃ -C ₄ and C ₅ -C ₆ form a disulfide bond; α is selected from w, y, f, and l; φ is selected from d, e, and n And “x” is selected from any amino acid. In some embodiments, the thrombospondin monomer comprises the following sequence:

The trefoil monomer domain includes the following sequences:

And the thyroglobulin monomer comprises the following sequence:

トレフォイル単量体ドメインは以下の配列を含み：

およびサイログロブリン単量体ドメインは以下の配列を含み：

「x」は任意のアミノ酸である。いくつかの態様において、トロンボスポンジン単量体ドメインのC₁-C₅、C₂-C₆およびC₃-C₄はジスルフィド結合を形成し；ならびにサイログロブリン単量体ドメインのC₁-C₂、C₃-C₄およびC₅-C₆はジスルフィド結合を形成する。いくつかの態様において、トロンボスポンジン単量体ドメインは以下の配列を含み：

式中C₁-C₅、C₂-C₆およびC₃-C₄はジスルフィド結合を形成し；トレフォイル単量体ドメインは以下の配列を含み：

サイログロブリン単量体ドメインは以下の配列を含む。

式中C₁-C₂、C₃-C₄およびC₅-C₆はジスルフィド結合を形成し；αはw、y、f、およびlから選択され；φはd、e、およびnから選択され；ならびに「x」は任意のアミノ酸から選択される。いくつかの態様において、トロンボスポンジン単量体は以下の配列を含み：

トレフォイル単量体ドメインは以下の配列を含み：

およびサイログロブリン単量体は以下の配列を含む。

The invention further provides that at least one monomer domain is a non-natural monomer domain and that the monomer domain binds ions and at least one monomer domain is a thrombospondin monomer domain, Compositions comprising trefoil monomer domains and at least two monomer domains selected from thyroglobulin monomer domains are provided. In some embodiments, at least one of the two monomer domains is less than about 50 kD. In some embodiments, the two domains are linked by a peptide linker. In some embodiments, the linker is heterologous to at least one of the monomer domains. In some embodiments, the thrombospondin monomer domain comprises the following sequence:

The trefoil monomer domain includes the following sequences:

And the thyroglobulin monomer domain includes the following sequences:

“X” is any amino acid. In some embodiments, C ₁ -C ₅ , C ₂ -C ₆ and C ₃ -C ₄ of the thrombospondin monomer domain form a disulfide bond; and C ₁ -C ₂ of the thyroglobulin monomer domain , C ₃ -C ₄ and C ₅ -C ₆ form disulfide bonds. In some embodiments, the thrombospondin monomer domain comprises the following sequence:

Wherein C ₁ -C ₅ , C ₂ -C ₆ and C ₃ -C ₄ form a disulfide bond; the trefoil monomer domain comprises the following sequence:

The thyroglobulin monomer domain comprises the following sequence:

Where C ₁ -C ₂ , C ₃ -C ₄ and C ₅ -C ₆ form a disulfide bond; α is selected from w, y, f, and l; φ is selected from d, e, and n And “x” is selected from any amino acid. In some embodiments, the thrombospondin monomer comprises the following sequence:

The trefoil monomer domain includes the following sequences:

And the thyroglobulin monomer comprises the following sequence:

  The present invention further provides isolated polynucleotides encoding the proteins described herein and cells containing the polynucleotides.

トレフォイル単量体ドメインが以下の配列を含み：

およびサイログロブリン単量体ドメインが以下の配列を含み：

「x」が任意のアミノ酸である、トロンボスポンジン単量体ドメイン、トレフォイル単量体ドメイン、およびサイログロブリン単量体ドメインから選択される、非天然の単量体ドメインのライブラリーを提供する段階；(2) 第1の標的分子に対する親和性について単量体ドメインのライブラリーをスクリーニングする段階；ならびに(3) 少なくとも1つの標的分子に結合する少なくとも1つの単量体ドメインを同定する段階により、標的分子に結合する単量体ドメインを同定する方法を提供する。いくつかの態様において、少なくとも1つの単量体ドメインは、非天然の単量体ドメインと少なくとも90%同一である天然の単量体ドメインによって結合されない標的分子に特異的に結合する。いくつかの態様において、トロンボスポンジン単量体ドメインのC₁-C₅、C₂-C₆およびC₃-C₄はジスルフィド結合を形成し；ならびにサイログロブリン単量体ドメインのC₁-C₂、C₃-C₄およびC₅-C₆はジスルフィド結合を形成する。いくつかの態様において、トロンボスポンジン単量体ドメインは以下の配列を含み：

式中C₁-C₅、C₂-C₆およびC₃-C₄はジスルフィド結合を形成し；トレフォイル単量体ドメインは以下の配列を含み：

サイログロブリン単量体ドメインは以下の配列を含む。

式中C₁-C₂、C₃-C₄およびC₅-C₆はジスルフィド結合を形成し；αはw、y、f、およびlから選択され；φはd、e、およびnから選択され；ならびに「x」は任意のアミノ酸から選択される。いくつかの態様において、トロンボスポンジン単量体は以下の配列を含み：

トレフォイル単量体ドメインは以下の配列を含み：

およびサイログロブリン単量体は以下の配列を含む。

いくつかの態様において、この方法はさらに、同定された単量体ドメインを第2の単量体ドメインに連結させて、各多量体が少なくとも2つの単量体ドメインを含む、多量体のライブラリーを形成させる段階；第1の標的分子に結合する能力について多量体のライブラリーをスクリーニングする段階；および第1の標的分子に結合する多量体を同定する段階を含む。選択された多量体の各単量体ドメインは、同じ標的分子にまたは異なる標的分子に結合する。いくつかの態様において、選択された多量体は2つ、3つ、4つ、またはそれ以上の単量体ドメインを含む。いくつかの態様において、この方法はさらに、少なくとも1つの単量体ドメインを突然変異させ、それによって突然変異した単量体ドメインを含むライブラリーを提供する段階を含む。いくつかの態様において、突然変異させる段階は、ポリペプチドドメインをコードする少なくとも1つのポリヌクレオチドの、複数のポリヌクレオチド断片を組換える段階を含む。いくつかの態様において、この方法はさらに、第2の標的分子に対する親和性について単量体ドメインのライブラリーをスクリーニングする段階；第2の標的分子に結合する単量体ドメインを同定する段階；第1の標的分子に対する親和性を有する少なくとも1つの単量体ドメインを第2の標的分子に対する親和性を有する少なくとも1つの単量体ドメインと連結させ、それによって第1および第2の標的分子に対する親和性を有する多量体を形成させる段階を含む。いくつかの態様において、単量体ドメインのライブラリーはファージディスプレイ、リボソームディスプレイまたは細胞表面ディスプレイとして発現される。いくつかの態様において、単量体ドメインのライブラリーはマイクロアレイ上に提示される。 The present invention similarly relates to (1) the monomer domain is
The thrombospondin monomer domain contains the following sequence:

The trefoil monomer domain includes the following sequence:

And the thyroglobulin monomer domain comprises the following sequence:

Providing a library of non-natural monomer domains selected from thrombospondin monomer domains, trefoil monomer domains, and thyroglobulin monomer domains, wherein "x" is any amino acid; (2) screening a library of monomer domains for affinity for a first target molecule; and (3) identifying at least one monomer domain that binds to at least one target molecule. Methods for identifying monomer domains that bind to a molecule are provided. In some embodiments, at least one monomer domain specifically binds to a target molecule that is not bound by a natural monomer domain that is at least 90% identical to a non-natural monomer domain. In some embodiments, C ₁ -C ₅ , C ₂ -C ₆ and C ₃ -C ₄ of the thrombospondin monomer domain form a disulfide bond; and C ₁ -C ₂ of the thyroglobulin monomer domain , C ₃ -C ₄ and C ₅ -C ₆ form disulfide bonds. In some embodiments, the thrombospondin monomer domain comprises the following sequence:

Wherein C ₁ -C ₅ , C ₂ -C ₆ and C ₃ -C ₄ form a disulfide bond; the trefoil monomer domain comprises the following sequence:

The thyroglobulin monomer domain comprises the following sequence:

Where C ₁ -C ₂ , C ₃ -C ₄ and C ₅ -C ₆ form a disulfide bond; α is selected from w, y, f, and l; φ is selected from d, e, and n And “x” is selected from any amino acid. In some embodiments, the thrombospondin monomer comprises the following sequence:

The trefoil monomer domain includes the following sequences:

And the thyroglobulin monomer comprises the following sequence:

In some embodiments, the method further comprises linking the identified monomer domain to a second monomer domain, wherein each multimer comprises at least two monomer domains. Screening a multimeric library for the ability to bind to a first target molecule; and identifying a multimer that binds to the first target molecule. Each monomer domain of the selected multimer binds to the same target molecule or to a different target molecule. In some embodiments, the selected multimer comprises 2, 3, 4, or more monomer domains. In some embodiments, the method further comprises mutating at least one monomer domain, thereby providing a library comprising the mutated monomer domain. In some embodiments, mutating comprises recombining a plurality of polynucleotide fragments of at least one polynucleotide encoding a polypeptide domain. In some embodiments, the method further comprises screening a library of monomer domains for affinity for a second target molecule; identifying a monomer domain that binds to the second target molecule; Linking at least one monomer domain having affinity for one target molecule with at least one monomer domain having affinity for a second target molecule, thereby affinities for the first and second target molecules A step of forming a sex multimer. In some embodiments, the library of monomer domains is expressed as phage display, ribosome display or cell surface display. In some embodiments, a library of monomer domains is presented on the microarray.

トレフォイル単量体ドメインは以下の配列を含み：

およびサイログロブリン単量体ドメインは以下の配列を含み：

「x」は任意のアミノ酸である。いくつかの態様において、多量体の各単量体ドメインは、非天然の単量体ドメインである。いくつかの態様において、ライブラリーは、リンカーによって連結された少なくとも2つの単量体ドメインを含む複数の多量体を含む。いくつかの態様において、ライブラリーは、異なる単量体ドメインを含む少なくとも100種の異なるタンパク質を含む。 The present invention further provides a library of proteins that are non-natural monomer domains comprising a monomer domain selected from a thrombospondin monomer domain, a trefoil monomer domain, and a thyroglobulin monomer domain. including. In some embodiments, the thrombospondin monomer domain comprises the following sequence:

The trefoil monomer domain includes the following sequences:

And the thyroglobulin monomer domain includes the following sequences:

“X” is any amino acid. In some embodiments, each monomer domain of the multimer is a non-natural monomer domain. In some embodiments, the library comprises a plurality of multimers comprising at least two monomer domains connected by a linker. In some embodiments, the library comprises at least 100 different proteins that include different monomer domains.

  The present invention also provides methods for identifying domain monomers and multimers that bind to target molecules. In some embodiments, the method comprises providing a library of monomer domains; screening the library of monomer domains for affinity for the first target molecule; and at least one target molecule Identifying at least one monomer domain that binds. In some embodiments, each monomer domain binds an ion (eg, calcium).

  In some embodiments, the method further comprises linking the identified monomer domain to a second monomer domain, wherein each multimer comprises at least two monomer domains. Screening a library of multimers for the ability to bind to a first target molecule; and identifying a multimer that binds to the first target molecule.

  In some embodiments, each monomer domain of the selected multimer binds to the same target molecule. In some embodiments, the selected multimer comprises three monomer domains. In some embodiments, the selected multimer comprises 4 monomer domains.

  In some embodiments, the monomer domain is selected from a type I thrombospondin domain, a type I thyroglobulin repeat domain, a trefoil (P type) domain, and an EGF-like domain (eg, a laminin-type EGF-like domain).

  In some embodiments, the method includes the further step of mutating at least one monomer domain, thereby providing a library comprising the mutated monomer domain. In some embodiments, mutating comprises recombining a plurality of polynucleotide fragments of at least one polynucleotide encoding a monomer domain. In some embodiments, the mutating step comprises directed evolution; a combination of different loop sequences; site-directed mutagenesis; or site-specific recombination that creates a crossover that results in the generation of a sequence that is identical to a human sequence. .

  In some embodiments, the method further comprises screening a library of monomer domains for affinity for a second target molecule; identifying a monomer domain that binds to the second target molecule; Linking at least one monomer domain having affinity for one target molecule with at least one monomer domain having affinity for a second target molecule, thereby affinities for the first and second target molecules A step of forming a sex multimer.

  In some embodiments, the target molecule is selected from viral antigens, bacterial antigens, fungal antigens, enzymes, cell surface proteins, intracellular proteins, enzyme inhibitors, reporter molecules, serum proteins, and receptors. In some embodiments, the viral antigen is a polypeptide required for viral replication.

  In some embodiments, the library of monomer domains is a phage display, phagemid display, ribosome display, polysome display, or cell surface display (e.g., E. coli cell surface display), yeast cell surface display or Expressed as by display via fusion with a protein that binds to a polynucleotide encoding the protein. In some embodiments, the library of monomer domains is presented on a microarray, including 96-well, 384-well or higher density microtiter plates.

  In some embodiments, the monomer domains are linked by a polypeptide linker. In some embodiments, the polypeptide linker is a linker that is naturally associated with the monomer domain. In some embodiments, the polypeptide linker is a linker that is naturally associated with a family of monomer domains. In some embodiments, the polypeptide linker is a variant of the linker that is naturally associated with the monomer domain. In some embodiments, the linker is a gly-ser linker. In some embodiments, the linking step comprises linking the monomer domains with various linkers of different lengths and compositions.

In some embodiments, these domains form secondary and tertiary structures by formation of disulfide bonds. In some embodiments, the multimer comprises an A domain that is linked to the monomer domain by a polypeptide linker. In some embodiments, the linker is 1-20 amino acids. In some embodiments, the linker is made up of 5-7 amino acids. In some embodiments, the linker is 6 amino acids in length. In some embodiments, the linker comprises the following sequence A ₁ A ₂ A ₃ A ₄ A ₅ A ₆ , wherein A ₁ is selected from amino acids A, P, T, Q, E, and K; A ₂ and A ₃ are any amino acids other than C, F, Y, W, or M; A ₄ is selected from amino acids S, G, and R; A ₅ is selected from amino acids H, P, and R A ₆ is the amino acid T. In some embodiments, the linker comprises a native sequence between the C-terminal cysteine of the first A domain and the N-terminal cysteine of the second A domain. In some embodiments, the linker comprises glycine and serine.

  The present invention is also a method for identifying multimers that bind to at least one target molecule, each multimer comprising at least two monomer domains and each monomer domain binding specificity to a target molecule Providing a library of multimers; and screening the library of multimers for target molecule-bound multimers. In some embodiments, the method further comprises identifying a target molecule binding multimer that has a binding activity for the target molecule that is greater than the binding activity of a single monomer domain for the target molecule. In some embodiments, the one or more multimers comprise a monomer domain that specifically binds to a second target molecule.

  An alternative method of identifying multimers that bind to a target molecule includes providing a library of monomer domains and / or immune domains; monomer domains and / or immune domains for affinity for the first target molecule Screening at least one monomer domain and / or immune domain that binds to at least one target molecule; identifying the identified monomer domain and / or immune domain as a monomer domain And / or linking to a library of immune domains to form a library of multimers, each multimer comprising at least two monomer domains, immune domains or combinations thereof; binding to a first target molecule Screening a multimeric library for the ability to: and a first target molecule The method comprising the step of identifying a multimer that binds contain.

  In some embodiments, each monomer domain binds an ion. In some embodiments, the ions are selected from calcium and zinc.

  In some embodiments, the linker comprises at least 3 amino acid residues. In some embodiments, the linker comprises at least 6 amino acid residues. In some embodiments, the linker comprises at least 10 amino acid residues.

  The invention also provides a polypeptide comprising at least two monomer domains separated by a heterologous linker sequence. In some embodiments, each monomer domain specifically binds to a target molecule; and each monomer domain is a monomer domain of a non-natural protein. In some embodiments, each monomer domain binds an ion.

  In some embodiments, the polypeptide comprises a first monomer domain that binds a first target molecule and a second monomer domain that binds a second target molecule. In some embodiments, the polypeptide includes two monomer domains, each monomer domain having specific binding specificity for a different site on the same target molecule. In some embodiments, the polypeptide further comprises a monomer domain that has binding specificity for a second target molecule.

  In some embodiments, the monomer domains of a library, multimer or polypeptide are typically about 40% identical to each other, usually about 50% identical, and in some cases about 60% identical. And is often at least 70% identical.

  The present invention also provides a polynucleotide encoding the above polypeptide.

  The invention also provides immune domain multimers having binding specificity for a target molecule, as well as methods for generating and screening libraries of such multimers for binding to a desired target molecule. More specifically, the present invention provides a method for identifying a multimer that binds to a target molecule comprising providing a library of immune domains; a library of immune domains for affinity for a first target molecule; Screening; identifying one or more (eg, two or more) immune domains that bind to at least one target molecule; linking the identified monomer domains so that each multimer is at least Forming a multimeric library comprising three immune domains (eg, 4 or more, 5 or more, 6 or more, etc.); abundant in ability to bind to the first target molecule A method comprising screening a library of bodies; and identifying a multimer that binds to a first target molecule. Libraries of multimers of at least two immune domains that are minibodies, single domain antibodies, Fabs, or combinations thereof are also utilized in the practice of the invention. Such libraries can be easily screened for multimers that bind to the desired target molecule by the methods of the invention described herein.

  The invention further provides a method of identifying a heteroimmune multimer that binds to a target molecule. In some embodiments, the method provides a library of immune domains; screening a library of immune domains for affinity for a first target molecule; providing a library of monomer domains Screening a library of monomer domains for affinity for a first target molecule; identifying at least one immune domain that binds to at least one target molecule; at least one binding to at least one target molecule; Identifying one monomer domain; linking the identified immune domains with the identified monomer domains to form a multimeric library, each multimer comprising at least two domains; Screen a library of multimers for the ability to bind to one target molecule Floors; as well as identifying a multimer that binds to the first target molecule.

  The invention also provides a method for identifying a laminin EGF monomer domain, a type I thrombospondin monomer domain, a thyroglobulin monomer domain, or a trefoil monomer domain that binds to a target molecule. In some embodiments, the method provides a library of laminin EGF monomer domain, type I thrombospondin monomer domain, thyroglobulin monomer domain, or trefoil monomer domain; Screening a library of laminin EGF monomer domain, type I thrombospondin monomer domain, thyroglobulin monomer domain, or trefoil monomer domain for affinity; and a single amount of laminin EGF that binds to the target molecule Identifying a body domain, a type I thrombospondin monomer domain, a thyroglobulin monomer domain, or a trefoil monomer domain.

  In some embodiments, the method comprises a library of laminin EGF monomer domains, type I thrombospondin monomer domains, thyroglobulin monomer domains, or trefoil monomer domains in the identified monomer domains. Are linked to form a multimeric library; screening the multimeric library for affinity for the target molecule; and identifying multimers that bind to the target. In some embodiments, the multimer binds to the target with a higher affinity than the monomer. In some embodiments, the method further comprises expressing the library using a display format selected from phage display, ribosome display, polysome display, or cell surface display.

  In some embodiments, the method further mutates at least one monomer domain, thereby mutating a laminin EGF monomer domain, a type I thrombospondin monomer domain, a thyroglobulin monomer domain. Or providing a library comprising a trefoil monomer domain. In some embodiments, the mutating step is directed evolution; site-directed mutagenesis; a combination of different loop sequences, or site-specific recombination that creates crossings that result in the generation of sequences that are identical to human sequences. Including those by.

  The present invention also provides a library of laminin EGF monomer domain, type I thrombospondin monomer domain, thyroglobulin monomer domain, or trefoil monomer domain; laminin for affinity to target molecule Screening a library of EGF monomer domains, type I thrombospondin monomer domains, thyroglobulin monomer domains, or trefoil monomer domains; and laminin EGF monomer domains that bind to target molecules, I A method of producing a polypeptide comprising a multimer identified by a method comprising identifying a type thrombospondin monomer domain, a thyroglobulin monomer domain, or a trefoil monomer domain is provided. In some embodiments, the multimer is produced by recombinant gene expression.

  The present invention also provides a type I thrombospondin monomer domain, thyroglobulin monomer domain, or trefoil monomer domain derived from a type I thrombospondin monomer domain, thyroglobulin monomer domain, or trefoil monomer domain. Methods are provided for generating libraries of monomer domains. In some embodiments, the method comprises each of two different natural variants of a human laminin EGF monomer domain, a type I thrombospondin monomer domain, a thyroglobulin monomer domain, or a trefoil monomer domain. Providing a loop sequence that corresponds to at least one loop from and wherein the loop sequence is a polynucleotide or polypeptide sequence; covalently linking the loop sequences so that each chimeric sequence comprises at least two loops Creating a library of chimeric monomer domain sequences encoding a chimeric type I thrombospondin monomer domain, thyroglobulin monomer domain, or trefoil monomer domain having: phage display, ribosome display, polysome display And cell table Expressing a library of chimeric type I thrombospondin monomer domains, thyroglobulin monomer domains, or trefoil monomer domains using a display format selected from display; chimera for binding to a target molecule Screening an expression library of a type I thrombospondin monomer domain, thyroglobulin monomer domain, or trefoil monomer domain; and a chimeric type I thrombospondin monomer domain, thyroglobulin single Identifying a monomer domain, or a trefoil monomer domain.

  In some embodiments, the method further comprises replacing the identified chimeric type I thrombospondin monomer domain, thyroglobulin monomer domain, or trefoil monomer domain with the chimeric type thrombospondin monomer domain, thyroglobulin. Ligating to each member of a library of monomer domains, or trefoil monomer domains, to form a multimeric library; multimeric for the ability to bind to a first target molecule with higher affinity Screening a library; and identifying multimers of chimeric type I thrombospondin monomer domains, thyroglobulin monomer domains, or trefoil monomer domains that bind to the first target molecule with higher affinity Including stages.

  The invention also relates to a loop corresponding to at least one loop from each of two different natural variants of a human type I thrombospondin monomer domain, thyroglobulin monomer domain, or trefoil monomer domain Providing a loop sequence that is a polynucleotide or a polypeptide sequence; covalently linking the loop sequences so that each chimeric sequence has at least two loops, a chimeric type I thrombospondin monomer Creating a library of chimeric monomer domain sequences encoding domains, thyroglobulin monomer domains, or trefoil monomer domains; a display selected from phage display, ribosome display, polysome display, and cell surface display Format Expressing a library of chimeric type I thrombospondin monomer domains, thyroglobulin monomer domains, or trefoil monomer domains; and chimeric type I thrombospondin monomer domains for binding to a target molecule Screening an expression library of a thyroglobulin monomer domain, or trefoil monomer domain; and a chimeric type I thrombospondin monomer domain, thyroglobulin monomer domain, or trefoil monomer that binds to a target molecule Methods are provided for producing chimeric type I thrombospondin monomer domains, thyroglobulin monomer domains, or trefoil monomer domains identified in a method comprising identifying a domain. In some embodiments, the chimeric type I thrombospondin monomer domain, thyroglobulin monomer domain, or trefoil monomer domain is produced by recombinant gene expression.

  In some embodiments, the monomer domain binds to the target molecule. In some embodiments, the polypeptide is 45 or less amino acids in length. In some embodiments, the heterologous amino acid sequence is an affinity peptide, heterologous type I thrombospondin monomer domain, heterologous thyroglobulin monomer domain, or heterologous trefoil monomer domain, purification tag, enzyme (e.g., horseradish peroxidase Or alkaline phosphatase), and a reporter protein (eg, green fluorescent protein or luciferase). In some embodiments, the target is not an antibody variable or hypervariable region.

  The present invention provides a method of screening a monomer domain or a multimeric library comprising monomer domains for binding affinity to a plurality of ligands. In some embodiments, the method comprises contacting a monomer domain or a library of monomer domain multimers with a plurality of ligands; and a monomer domain or multimer that binds to at least one of the ligands. The step of selecting is included.

  In some embodiments, the method comprises (i.) Contacting a library of monomer domains with a plurality of ligands; (ii.) Selecting a monomer domain that binds to at least one of the ligands. (Iii.) A library of multimers, each comprising a selected monomer domain and a second monomer domain, wherein the selected monomer domains are linked to a library of monomer domains. (Iv.) Contacting the library of multimers with a plurality of ligands to form a plurality of complexes, each complex comprising a multimer and a ligand; and (v.) At least Selecting a complex.

  In some embodiments, the method further comprises linking the multimer of the selected complex to a monomer domain or a library of multimers, wherein the selected multimer and at least a third monomer domain are Forming a second library of multimers, each comprising; contacting the second library of multimers with a plurality of ligands to form a plurality of second complexes; and at least one of Selecting a second complex.

  In some embodiments, the identity of the ligand and multimer is determined. In some embodiments, the library of monomer domains is contacted with multiple ligands. In some embodiments, the multimeric library is contacted with multiple ligands.

  In some embodiments, the plurality of ligands are present in the mixture. In some embodiments, multiple ligands are present in the array. In some embodiments, the plurality of ligands are present in or on the cell or tissue. In some embodiments, the plurality of ligands are immobilized on a solid support.

  In some embodiments, the ligand is a polypeptide. In some embodiments, the polypeptide is expressed on the surface of the phage. In some embodiments, a library of monomer domains or multimers is expressed on the surface of the phage.

  In some embodiments, a multimeric library is expressed on the surface of a phage to form a library expressing phage, a ligand is expressed on the surface of the phage to form a ligand expressing phage, and the method comprises: Contacting an expression phage with a ligand-expressing phage to form a ligand-expressing phage / library-expressing phage pair; removing a ligand-expressing phage that does not bind to the library-expressing phage; or a library-expressing phage that does not bind to the ligand-expressing phage. Removing; and selecting a ligand-expressing phage / library-expressing phage pair. In some embodiments, the method further comprises isolating the polynucleotide from the phage pair and amplifying the polynucleotide to produce a polynucleotide hybrid comprising the polynucleotide from the ligand expressing phage and the library expressing phage. including.

  In some embodiments, the method includes isolating the polynucleotide hybrid from a plurality of phage pairs, thereby forming a mixture of polynucleotide hybrids. In some embodiments, the method comprises contacting a mixture of hybrid polynucleotides with a cDNA library under conditions that permit polynucleotide hybridization, thereby hybridizing the hybrid polynucleotide to cDNA in the cDNA library. And determining the nucleotide sequence of the hybridized hybrid polynucleotide, thereby identifying a monomer domain that specifically binds to the polypeptide encoded by the cDNA. In some embodiments, a library of monomer domains is expressed on the surface of a phage to form a library expressing phage, a ligand is expressed on the surface of the phage to form a ligand expressing phage, and the selected complex Comprises a library-expressing phage bound to a ligand-expressing phage, and in this method, dividing the selected monomer domain or multimer into first and second parts, the monomer domain of the first part or Ligating the multimer to a solid surface and contacting the phage display ligand library with the monomer domain or multimer of the first part to bind to the monomer domain or multimer of the first part Identifying bacteria; infecting bacteria with phage displaying a second part monomer domain or multimer Te, step expressing phage; comprising the step of forming a well by the expression phage contacting the target ligand phage, phage pairs of phage displaying a target ligand phage and monomer domains or multimers.

  In some embodiments, the method further comprises isolating a polynucleotide from each phage of the phage pair, thereby identifying a multimeric or monomeric domain that binds a ligand in the phage pair. In some embodiments, the method further comprises amplifying the polynucleotide to produce a polynucleotide hybrid comprising the polynucleotide from the target ligand phage and the library phage.

  In some embodiments, the method includes isolating and amplifying a polynucleotide hybrid from a plurality of phage pairs, thereby forming a mixture of polynucleotide hybrids. In some embodiments, the method comprises contacting a mixture of hybrid polynucleotides with a cDNA library under conditions that allow for hybridization, thereby hybridizing the hybrid polynucleotide to the cDNA in the cDNA library. And determining the nucleotide sequence of the associated hybrid polynucleotide, thereby identifying a monomer domain that specifically binds to a ligand encoded by the cDNA-associated cDNA.

The present invention is similarly
At least 1%, 2%, 3%, 4%, 5%, 6%, 7%, 8%, 9%, 10%, 11%, 12%, 13%, 14%, 15% of the amino acids in the sequence , 16%, 17%, 18%, 19%, 20% or more is cysteine; and the amino acid sequence is at least 10, 20, 30, 45, 50, 55, 60, 70, 80, 90, 100 or And / or the amino acid sequence is less than 150, 140, 130, 120, 110, 100, 90, 80, 70, 60, 50, or 40 amino acids in length; and / or at least of amino acids 5%, 10%, 15%, 20%, 25%, 30%, 35%, 40%, 45%, 50% or more are unnatural amino acids,
Non-naturally occurring polypeptides comprising amino acid sequences are provided. For example, in some embodiments, the amino acid sequence comprises at least 10% cysteine, and the amino acid sequence is at least 50 amino acids long, or at least 25% of the amino acids are unnatural. In some embodiments, the amino acid sequence is a non-natural A domain.

  In some embodiments, the polypeptides of the invention have at least 10%, 15%, 20%, 25%, 30%, 35%, 40%, 45%, 50% or more unnatural amino acids 1 Contains one, two, three, four, or more monomers. In some embodiments, the one or more monomer domains are at least 10%, 15%, 20%, 25%, 30%, 35%, 40%, 45 that do not appear in that position in the native human protein. Contains 50%, 50% or more amino acids. In some embodiments, the monomer domain is derived from a native human protein sequence. In some embodiments, the polypeptides of the invention are also at least, for example, 1, 2, 3, 4, 5, 10, 20, 30, 40, 50, 60, 70, 80, 90, 100, 150 Have a serum half-life of 200, 250, 400, 500 or more hours.

Definitions Unless otherwise stated, the following definitions are used in place of definitions in the art.

  The terms “monomer domain” or “monomer” are used interchangeably herein to refer to individual regions found in a protein or polypeptide. Monomeric domains form a native three-dimensional structure in solution in the absence of adjacent natural amino acid sequences. The monomer domains of the present invention can be selected to specifically bind to the target molecule. As used herein, the term “monomer domain” does not encompass the complementarity determining region (CDR) of an antibody.

  The term “monomer domain variant” refers to a domain obtained from the artificial manipulation of monomer domain sequences. Examples of artificially manipulated changes include, for example, random mutations, site-directed mutagenesis, recombination, directed evolution, forced crossover events with oligos, and uptake of mutations by direct gene synthesis. The term “monomer domain variant” does not encompass the complementarity determining region (CDR) of the mutated antibody.

  The term “loop” refers to that portion of the monomer domain that is normally exposed to the environment by assembly of the backbone structure of the monomer domain protein, which is involved in target binding. The present invention provides three types of loops identified by specific features such as disulfide bonds, cross-links between secondary protein structures, and the possibility of molecular dynamics (ie flexibility). The three types of loop sequences are a loop sequence defined by cysteine, a loop sequence defined by structure, and a loop sequence defined by factor B.

As used herein, the term “a loop sequence defined by cysteine” is a partial sequence of a sequence encoding a natural monomer domain and comprising at least one other natural monomer of the same family. A sequence whose cysteine residues conserved with respect to a domain are bound at both ends. The loop sequence defined by cysteine is identified by multiple sequence alignment of natural monomer domains followed by sequence analysis to identify conserved cysteine residues. The sequence between each successive pair of conserved cysteine residues is a loop sequence defined by cysteine. The loop sequence defined by cysteine does not include a cysteine residue adjacent to each end. Monomer domains having a loop sequence defined by cysteine include a thrombospondin domain, a thyroglobulin domain, a trefoil / PD domain, and the like. Thus, for example, the thrombospondin domain is represented by the consensus sequence CX ₃ CX ₁₀ CX ₁₆ CX ₁₁ CX ₄ C, where X ₃ , X ₁₀ , X ₁₆ , X ₁₁ , and X ₄ are defined by cysteine Each represents a loop sequence; the trefoil / PD domain is represented by the consensus sequence CX ₁₀ CX ₉ CX ₄ CCX ₁₀ C, where X ₁₀ , X ₉ , X ₄ , and X ₁₀ are loop sequences defined by cysteine Each represented; and the thyroglobulin domain is represented by the consensus sequence CX ₂₆ CX ₁₀ CX ₆ CX ₁ CX ₁₈ C, where X ₂₆ , X ₁₀ , X ₆ , X ₁ , and X ₁₈ are loop sequences defined by cysteine Respectively.

  The term `` multimer '' refers to at least two monomer domains and / or immune domains (e.g., at least two monomer domains, at least two immune domains, or at least one monomer domain and at least one immunity). Is used herein to refer to a polypeptide comprising a domain. Separate monomer domains and / or immune domains in a multimer can be joined together by a linker. Multimers are also known as combinatorial mosaic proteins or recombinant mosaic proteins.

  The terms “family” and “family class” are used interchangeably to refer to proteins that are grouped together based on their amino acid sequence similarity. These similar sequences are generally conserved because they are important for the function of the protein and / or the maintenance of the three-dimensional structure of the protein. Examples of such families include the LDL receptor A domain family and the EGF-like family.

  The term “ligand”, also referred to herein as “target molecule”, encompasses a wide variety of substances and molecules ranging from simple molecules to complex targets. The target molecule can be a protein, nucleic acid, lipid, carbohydrate, or any other molecule that can be recognized by a polypeptide domain. For example, the target molecule is a chemical compound (i.e., an organic molecule, an inorganic molecule, or a non-biological compound such as a molecule having both organic and inorganic atoms, excluding polynucleotides and proteins), chemical compound Mixtures of, spatially localized compounds, biological macromolecules, bacteriophage peptide display libraries, polysome peptide display libraries, bacteria, plants, fungi, or animal (e.g., mammalian) cells or It may include extracts, proteins, toxins, peptide hormones, cells, viruses, etc. made from biological material such as tissue. Other target molecules include, for example, whole cells, whole tissues, a mixture of related or unrelated proteins, a mixture of virus strains or bacterial strains, and the like. Target molecules can also enhance or inhibit specific protein interactions (i.e., selectively bind binding interactions between two given polypeptides) by inclusion in the screening assays described herein. Inhibiting factors).

  As used herein, the term “immune domain” refers to a protein binding domain comprising at least one complementarity determining region (CDR) of an antibody. The immune domain can be a natural (i.e., isolated from the natural state) immunological domain, or an artificial manipulation (e.g., random mutation, site-directed mutagenesis, recombination, etc.) It may be a non-natural immunological domain that has been modified through induction methods as well as by directed evolution methods such as recursive error-prone PCR, recursive recombination, and the like. Different types of immune domains suitable for use in the practice of the present invention include minibodies, single domain antibodies, single chain variable fragments (ScFv), and Fab fragments.

The term “minibody” is used herein to refer to only two complementarity of a natural or non-natural (eg, mutagenized) heavy chain variable domain or light chain variable domain, or combinations thereof. A polypeptide that encodes a determining region (CDR). Examples of minibodies are described in Pessi et al., A designed metal-binding protein with a novel fold, (1993) Nature 362: 367-369.

As used herein, the term “single domain antibody” refers to the heavy chain variable domain (“V _H ”) of an antibody, ie, the heavy chain variable domain without the light chain variable domain. Exemplary single domain antibodies utilized in the practice of the present invention include, for example, Hamers-Casterman, C. et al., Naturally occurring antibodies devoid of light chains (1993) Nature 363: 446-448, and Dumoulin et al. ., Single-domain antibody fragments with high conformational stability (2002) Protein Science 11: 500-515, Camelid heavy chain variable domain (about 118-136 amino acid residues).

The terms “single chain variable fragment” or “ScFv” are used interchangeably herein to refer to the heavy and light chain variable domains of an antibody linked by a peptide linker having at least 12 amino acid residues. Single chain variable fragments contemplated for use in the practice of the present invention include Bird, et al., (1988) Science 242 (4877): 423-426 and Huston et al., (1988) PNAS USA 85 (16 ): Includes those described in 5879-83.

As used herein, the term “Fab fragment” refers to an immune domain having two protein chains, one of which is a light chain composed of two light chain domains (V _L variable domain and C _L constant domain). And a heavy chain consisting of two heavy chain domains (ie, a V _H variable domain and a C _H constant domain). Fab fragments utilized in the practice of the present invention include those having an interchain disulfide bond at the C-terminus of each heavy and light chain component, and those not having such a C-terminal disulfide bond. Each fragment is approximately 47 kD. Fab fragments are described by Pluckthun and Skerra, (1989) Methods Enzymol 178: 497-515.

  The term “linker” is used herein to refer to a moiety or group of moieties that link or connect two or more individual separated monomer domains. A linker can leave separate discrete monomer domains separated when linked together in a multimer. The linker moiety is typically a substantially linear moiety. Suitable linkers include polypeptides, polynucleic acids, peptide nucleic acids and the like. Suitable linkers also include optionally substituted alkylene moieties having one or more oxygen atoms incorporated into the carbon skeleton. Typically, the molecular weight of the linker is less than about 2000 daltons. More typically, the molecular weight of the linker is less than about 1500 daltons and usually less than about 1000 daltons. A linker allows, for example, individual discrete monomer domains to cooperate when each individual discrete monomer domain in the multimer binds to the same target molecule via a separate binding site. Can be made sufficiently small. Exemplary linkers include a polynucleotide encoding a polypeptide, or an amino acid polypeptide or other non-naturally occurring portion. The linker can be a natural sequence, a variant thereof, or part of a synthetic sequence. The linker can include, for example, natural amino acids, unnatural amino acids, or a combination of both.

  The term “isolated” is used herein to refer to the properties of a moiety that remains independent and independent, even when complexed with other moieties, including, for example, other monomer domains. Used in A monomer domain is a domain separated in a protein. This is because monomer domains have independent properties that can be recognized and separated from proteins. For example, the ligand binding ability of the A domain in LDLR is an independent property. Examples of other separations include discrete monomer domains in a multimer that remain separate and independent domains even when complexed or joined together in the multimer by a linker. . Another example of a separated property is a separated binding site in the multimer for the ligand.

  As used herein, “directed evolution” refers to the process by which a polynucleotide variant is created, expressed, and screened for activity (eg, a polypeptide having binding activity) in a recursive process. . One or more candidates are selected in the screen, and then the process is repeated using a polynucleotide encoding the selected candidates to create new variants. Directed evolution involves at least 2 rounds of mutagenesis and can include 3, 4, 5, 10, 20 or more rounds of mutagenesis and selection. Mutations can be made by any method known to those skilled in the art including, for example, error-prone PCR, genetic recombination, chemical mutagenesis, and the like.

  The term “shuffling” is used herein to refer to recombination between non-identical sequences. In some embodiments, shuffling can include crossing via homologous recombination or via non-homologous recombination, eg, via the cre / lox system and / or the flp / frt system. Shuffling includes in vitro and in vivo shuffling formats, in silico shuffling formats, shuffling formats using either double-stranded or single-stranded templates, primer-based shuffling formats, nucleic acid fragmentation-based shuffling formats, and oligonucleotide-mediated shuffling formats. All of these are based on recombination events between non-identical sequences, including other similar recombination in addition to the various different formats described or referred to in more detail herein below. This can be done by using a format based on. As used herein, the term “random” refers to a polynucleotide or amino acid sequence composed of two or more amino acids and constructed by a stochastic or random process. The random polynucleotide or amino acid sequence can include a framework or backbone motif, which can include an invariant sequence.

  As used herein, the term “pseudorandom” is limited in variability, so that the degree of residue variability at a position is limited, but any pseudorandom position is at least some residue. Refers to a series of sequences, polynucleotides or polypeptides that are allowed to vary.

  The terms “polypeptide”, “peptide” and “protein” are used interchangeably herein to refer to an amino acid sequence of two or more amino acids.

  “Conservative amino acid substitution” refers to the interchangeability of residues having similar side chains. For example, the group of amino acids having an aliphatic side chain is glycine, alanine, valine, leucine, and isoleucine; the group of amino acids having an aliphatic hydroxyl side chain is serine and threonine; The group is asparagine and glutamine; the group of amino acids having aromatic side chains is phenylalanine, tyrosine, and tryptophan; the group of amino acids having basic side chains is lysine, arginine, and histidine; and the sulfur-containing side A group of amino acids with chains are cysteine and methionine. Preferred conservative amino acid substitution groups are valine-leucine-isoleucine, phenylalanine-tyrosine, lysine-arginine, alanine-valine, and asparagine-glutamine.

  The phrase “nucleic acid sequence” refers to a single-stranded or double-stranded polymer of deoxyribonucleotide bases or ribonucleotide bases read in the 5 ′ to 3 ′ end direction. This includes chromosomal DNA, self-replicating plasmids and DNA or RNA that plays a primary structural role.

  The term “encode” refers to a polynucleotide sequence that encodes one or more amino acids. The term does not require a start or stop codon. The amino acid sequence can be encoded in any one of six different reading frames donated by the polynucleotide sequence.

  The term “promoter” refers to a region or sequence located upstream and / or downstream from the start of transcription that is involved in the recognition and binding of RNA polymerase and other proteins that initiate transcription.

  A “vector” refers to a polynucleotide that is capable of replicating in a host organism when it is independent of the host chromosome. Examples of vectors include plasmids. Vectors typically have an origin of replication. Vectors can include, for example, transcription and translation terminators, transcription and translation initiation sequences, and promoters useful for the regulation of specific nucleic acid expression.

  The term “recombinant” when used in connection with, for example, a cell, or a nucleic acid, protein, or vector, introduces a heterologous nucleic acid or protein, or a naturally occurring nucleic acid or protein, where the cell, nucleic acid, protein, or vector is Indicates that the cell has been modified, or that the cell is derived from a cell so modified. Thus, for example, a recombinant cell expresses a gene that is not found in a natural (non-recombinant) cell form, or is otherwise abnormally expressed, decreased in expression, or not expressed at all Not expressing the natural gene.

  The phrase “specifically (or selectively) binds” to a polypeptide, when referring to monomers or multimers, refers to the presence of the polypeptide in a heterogeneous population of proteins and other biologics. A binding reaction that can be a determinant. Thus, under standard conditions or assays used in antibody binding assays, a particular monomer or multimer may exceed background (e.g., 2x, 5x, 10x or more background). Binds to a specific target molecule and does not bind in a significant amount to other molecules present in the sample.

  In the context of two or more nucleic acid or polypeptide sequences, the term “identical” or “percent identity” refers to two or more sequences or subsequences that are identical. “Substantially identical” refers to the largest match for a specified region using a comparison window or one of the following sequence comparison algorithms or as determined by manual alignment and visual inspection. When compared and aligned, a certain percentage of identical amino acid residues or nucleotides (i.e., 60% identity, optionally 65%, 70% over a particular region or, if not specified, over the entire sequence) , 75%, 80%, 85%, 90%, or 95% identity). Optionally, identity or substantial identity exists over a region that is at least about 50 nucleotides in length, or more preferably over a region that is 100-500 or 1000 or more nucleotides or amino acids in length. To do.

  A polynucleotide or amino acid sequence is “heterologous” to a second sequence if the two sequences are not linked in the same way as found in the native sequence. For example, a promoter operably linked to a heterologous coding sequence refers to a coding sequence that is different from any naturally occurring allelic variant. The term “heterologous linker” when used in connection with a multimer includes a linker and a monomer that are not found in the same relationship to each other in nature (e.g., they form a fusion protein). )

  An “unnatural amino acid” in a protein sequence refers to a Genbank non-redundant (“nr”) database using BLAST 2.0 where the comparison window is the length of the query monomer domain and is described herein. Any amino acid other than the amino acid that appears at the corresponding position in the alignment with the natural polypeptide that has the lowest total probability when compared.

  The “percent sequence identity” is determined by comparing two sequences that are optimally aligned over the comparison window, where the portion of the polynucleotide sequence within the comparison window is the optimal alignment of the two sequences. Thus, it can contain additions or deletions (ie gaps) compared to a reference sequence (which does not contain additions or deletions). This percentage determines the number of positions where the same nucleobase or amino acid residue appears in both sequences to obtain the number of matched positions, and divides the number of matched positions by the total number of positions in the comparison window. Multiplied by 100 to get the percent sequence identity.

  In the context of two or more nucleic acid or polypeptide sequences, the term “identical” or “percent identity” refers to a comparison window or one of the following sequence comparison algorithms or manual alignment And two, or two, having the same or a certain proportion of identical amino acid residues or nucleotides when compared and aligned for maximum match against a specified region as determined by visual inspection It refers to the above nucleic acid or partial sequence. Such sequences are thus said to be “substantially identical”. This definition also refers to the complement of the test sequence. Optionally, identity exists over a region that is at least about 50 amino acids or nucleotides in length, or more preferably over a region that is 75-100 amino acids or nucleotides in length.

  For sequence comparison, typically one sequence acts as a reference sequence, to which test sequences are compared. When using a sequence comparison algorithm, test and reference sequences are entered into a computer, subsequence coordinates are designated, if necessary, and sequence algorithm program parameters are designated. Default program parameters may be used, or alternative parameters may be specified. The sequence comparison algorithm then calculates the percent sequence identities for the test sequences relative to the reference sequence, based on the program parameters.

  As used herein, a “comparison window” is any number of consecutive positions selected from the group consisting of 20 to 600, usually about 50 to about 200, more usually about 100 to about 150. Includes a reference to a segment that can be compared to a reference sequence of the same number of consecutive positions after the two sequences are optimally aligned. Methods for sequence alignment for comparison are well known in the art. Optimal sequence alignments for comparison are, for example, those of Needleman and Wunsch (1970) J. Mol. Biol. 48: 443 by the local homology algorithm of Smith and Waterman (1970) Adv. Appl. Math. The homology alignment algorithm allows the similarity search of Pearson and Lipman (1988) Proc. Nat'l. Acad. Sci. USA 85: 2444 to be used to compute these algorithms (GAP, BESTFIT, FASTA And by performing TFASTA, Genetics Computer Group, 575 Science Dr., Madison, WI) or by manual alignment and visual inspection (see, for example, Ausubel et al., Current Protocols in Molecular Biology (1995)). )It can be carried out.

  One example of a useful algorithm is the BLAST 2.0 algorithm, which is described in Altschul et al. (1990) J. Mol. Biol. 215: 403-410, respectively. Software for performing BLAST analyzes is publicly available through the National Center for Biotechnology Information (http://www.ncbi.nlm.nih.gov/). This algorithm first calculates whether a short word of length W in the query sequence (which matches or satisfies a certain positive threshold score T when aligned with a word of the same length in the database sequence) Identifying a high-scoring sequence (HSP) pair. T is referred to as the neighborhood word score threshold (Altschul et al, supra). These initial neighborhood word hits act as sources for initiating searches to find longer HSPs containing them. Word hits extend in both directions along each sequence as long as the cumulative alignment score can be increased. The cumulative score is calculated for a nucleotide sequence using parameter M (reward score for matched residue pair; always> 0) and parameter N (penalty score for mismatched residue; always <0). For amino acid sequences, a cumulative score is calculated using a scoring matrix. Word hit extension in each direction is stopped when: The cumulative alignment score decreases by an amount X from its maximum achieved value; due to accumulation of one or more negative score residue alignments If the cumulative score is 0 or less; or if either end of the sequence is reached. The BLAST algorithm parameters W, T, and X determine the sensitivity and speed of the alignment. The BLASTN program (in the case of nucleotide sequences) uses by default 11 word lengths (W), 10 expected values (E), M = 5, N = -4 and a comparison of both strands. For amino acid sequences, the BLASTP program defaults to a word length of 3 and an expected value of 10 (E) and a BLOSUM62 scoring matrix (Henikoff and Henikoff (1989) Proc. Natl. Acad. Sci. USA 89: 10915. See), 50 alignments (B), 10 expected values (E), M = 5, N = -4, and comparison of both strands.

  The BLAST algorithm similarly performs a statistical analysis of the similarity between two sequences (see, eg, Karlin and Altschul (1993) Proc. Natl. Acad. Sci. USA 90: 5873-5787). One measure of similarity provided by the BLAST algorithm is the minimum total probability (P (N)), which provides an indication of the probability that a match between two nucleotide or amino acid sequences will occur by chance. For example, a nucleic acid is considered similar to a reference sequence if the minimum total probability of comparing the test nucleic acid to the reference nucleic acid is less than about 0.2, more preferably less than about 0.01, and most preferably less than about 0.001. It is done.

Detailed Description of the Invention The present invention provides an affinity agent comprising a monomer domain and a multimer of monomer domains. Affinity agents can be selected for their ability to bind to the desired ligand or ligand mixture. Monomeric domains and multimers should be screened to identify those with improved properties, such as improved binding activity or affinity for ligands or ligand mixtures or altered specificity compared to distinct monomer domains Can do. The monomer domains of the present invention include specific variants of laminin EGF-like domain, type 1 thrombospondin domain, trefoil domain, and thyroglobulin domain.

I. Monomer Domains A number of suitable monomer domains can be used in the polypeptides of the invention. Typically, a suitable monomer domain contains three disulfide bonds, 30-100 amino acids, and has a binding site for a divalent metal ion, such as, for example, calcium. In some embodiments, type 1 thrombospondin monomer domains, trefoil monomer domains, or thyroglobulin monomer domains are used in the scaffolds of the invention. In other embodiments, laminin EGF monomer domains are used.

  A monomer domain can have any number of properties. For example, in some embodiments, the monomer domain is less or less immunogenic in animals (eg, humans). Monomer domains can have a small size. In some embodiments, the monomer domains are small enough to penetrate the skin or other tissues. Monomeric domains can have a range of in vivo half-lives or stability. The properties of the monomer domain include the ability to fold independently and form a stable structure.

The monomer domain can be a polypeptide chain of any size. In some embodiments, the monomer domains have about 25 to about 500, about 30 to about 200, about 30 to about 100, about 35 to about 50, about 35 to about 100, about 90 to about 200, about 30. Have from about 250, about 30 to about 60, about 9 to about 150, about 100 to about 150, about 25 to about 50, or about 30 to about 150 amino acids. Similarly, the monomer domains of the present invention comprise, for example, about 30 to about 200 amino acids; about 25 to about 180 amino acids; about 40 to about 150 amino acids; about 50 to about 130 amino acids; or about 75 to about 125 amino acids. Can be included. Monomeric and immune domains are typically capable of maintaining a conformation that is stable in solution and are often heat stable, e.g., at least 10 without loss of binding affinity. Stable at 95 ° C for minutes. Monomer domains are typically about 10 ^-15, 10 ^-14, 10 ^-13, 10 ^-12, 10 ^-11, 10 ^-10, 10 ^-9, 10 ^-8, 10 ^-7, 10 ^- It binds with a K _d of less than ⁶ , 10 ⁻⁵ , 10 ⁻⁴ , 10 ⁻³ , 10 ⁻² , about 0.01 μM, about 0.1 μM, or about 1 μM. Monomeric and immune domains may fold independently into stable conformations. In one embodiment, the stable higher order structure is stabilized by metal ions. A stable conformation may optionally include disulfide bonds (eg, at least 1, 2, or 3 or more disulfide bonds). A disulfide bond may be formed between two cysteine residues. In some embodiments, the monomer domain, or monomer domain variant, is an exemplified sequence (e.g., thrombospondin, trefoil, or thyroglobulin) or any other sequence cited herein. Substantially the same.

  An exemplary monomer domain that is particularly suitable for use in the practice of the present invention is a cysteine-rich domain containing disulfide bonds. Typically, disulfide bonds facilitate domain folding into a three-dimensional structure. Usually, cysteine-rich domains have at least two disulfide bonds, more typically at least three disulfide bonds. Suitable cysteine rich monomer domains include, for example, a type 1 thrombospondin domain, a trefoil domain, or a thyroglobulin domain.

  A monomer domain can similarly have a cluster of negatively charged residues. Monomeric domains can bind ions and maintain their secondary structure. Such monomer domains include, for example, A domain, EGF domain, EF hand (e.g. those present in calmodulin and troponin C), cadherin domain, C-type lectin, C2 domain, annexin, Gla domain, type 3 thrombosis. Zinc fingers, including the sponge domain, all of which bind calcium and zinc (e.g., C2H2 type C3HC4 type (RING finger), integrase zinc binding domain, PHD finger, GATA zinc finger, FYVE zinc finger, B box Zinc finger). Although not intended to limit the present invention, ionic bonds, depending on the primary sequence, stabilize the secondary structure while providing sufficient flexibility to allow for multiple bond conformations. It is thought that

  The structure of the monomer domain is often conserved, but the polynucleotide sequence encoding the monomer need not be conserved. For example, the domain structure can be conserved among members of the domain family, but the nucleic acid sequence of the domain is not conserved. Thus, for example, a monomer domain is classified as a type 1 thrombospondin domain, trefoil domain, or thyroglobulin domain by its affinity to cysteine residues and metal ions (e.g., calcium), depending on its nucleic acid sequence. Are not necessarily classified.

In some embodiments, a suitable monomer domain (e.g., a domain that has the ability to fold independently or with some limited assistance) is a Simple Modular Architecture Research Tool (SMART) (Shultz et al., SMART). : A web-based tool for the study of genetically mobile domains, (2000) Nucleic Acids Research 28 (1): 231-234) or CATH (Pearl et. Al., Assigning genomic sequences to CATH, (2000 ) Nucleic Acids Research 28 (1): 277-282) can be selected from a family of protein domains containing a β sandwich or β barrel three-dimensional structure as defined by computer sequence analysis tools such as:

  In some embodiments, the monomer domain is modified to bind to a substrate to enhance protein function, including, for example, enzymatic activity and / or substrate conversion.

  As described herein, monomer domains can be selected for their ability to bind to targets other than those to which the natural homologous domain can bind. That is, in some embodiments, the invention provides monomer domains (and multimers comprising such monomers) that do not bind to a target or target protein class or family to which a natural homologous domain can bind. .

  Each domain described herein utilizes an exemplary motif (ie, skeleton). Certain positions are marked with x, indicating that any amino acid can occupy that position. These positions include the possibility of several different amino acids, which can allow for sequence diversity, ie affinity for different target molecules. The use of square brackets in the motif indicates possible alternative amino acids within the position (eg, “[ekq]” indicates that any of E, K, or Q may be present at that position). The use of parentheses in the motif indicates that the position within the parentheses may or may not exist (e.g., "([ekq])" does not exist or E, K Or either Q may be present at that position). When multiple “x” are used in parentheses (eg, “(xx)”), each x represents a possible position. That is, “(xx)” indicates that 0, 1 or 2 amino acids can be present at that position, wherein each amino acid is independently selected from any amino acid. α represents an aromatic / hydrophobic amino acid such as, for example, W, Y, F, or L; β represents a hydrophobic amino acid such as, for example, V, I, L, A, M, or F; Represents a smaller polar amino acid such as, for example, G, A, S, or T; δ represents a charged amino acid such as, for example, K, R, E, Q, or D; ε represents, for example, V, Represents a small amino acid such as A, S, or T; and φ represents a negatively charged amino acid such as, for example, D, E, or N.

  Suitable domains include, for example, type 1 thrombospondin domain, trefoil domain, and thyroglobulin domain.

  Type 1 thrombospondin (“TSP1”) domains contain about 30-50 or 30-65 amino acids. In some embodiments, this domain comprises about 35-55 amino acids and optionally about 50 amino acids. Among the 35-55 amino acids there are typically about 4 to about 6 cysteine residues. Of the six cysteines, disulfide bonds are typically found between the following cysteines: C1 and C5, C2 and C6, C3 and C4. The cysteine residues of the domain are disulfide linked to form a small, stable, functionally independent part containing a distorted β chain. These repeat clusters constitute the ligand binding domain, and differential cluster formation can provide specificity for ligand binding.

Exemplary TSP1 domain sequences and consensus sequences are as follows:

  In some embodiments, the type 1 thrombospondin domain variant comprises a sequence that is substantially identical to any of the sequences described above.

  To date, at least 1677 natural thrombospondin domains have been identified based on cDNA sequences. Exemplary proteins containing the natural thrombospondin domain include, for example, complement pathway proteins (e.g., properdin, C6, C7, C8A, C8B, and C9), extracellular matrix proteins (e.g., mindin ), F-spondin, SCO-spondin), sporozoite surface protein 2, and Plasmodium TRAP protein. Type 1 thrombospondin domains are described, for example, by Roszmusz et al., BBRC 296: 156 (2002); Higgins et al., J Immunol. 155: 5777-85 (1995); Schultz-Cherry et al., J. Biol Chem. 270: 7304-7310 (1995); Schultz-Cherry et al., J. Biol Chem. 269: 26783-8 (1994); Bork, FEBS Lett 327: 125-30 (1993); and Leung-Hagesteijn et al., Cell 71: 289-99 (1992).

  Another exemplary monomer domain suitable for use in the practice of the present invention is the trefoil domain. Trefoil monomer domains are typically about 30-50 or 30-65 amino acids. In some embodiments, this domain comprises about 35-55 amino acids and optionally about 45 amino acids. Of the 35-55 amino acids, there are typically about 6 cysteine residues. Of the six cysteines, disulfide bonds are typically found between the following cysteines: C1 and C5, C2 and C4, C3 and C6.

  To date, at least 149 natural trefoil domains have been identified based on cDNA sequences. Exemplary proteins containing the natural trefoil domain include, for example, protein pS2 (TFF1), antispasmodic peptide SP (TFF2), intestinal trefoil factor (TFF3), intestinal sucrase-isomaltase, and protects the epithelium. Proteins that may be involved in defense against microbial infection (eg, Xenopus xP1, xP4, integument mucins A.1 and C.1). Trefoil domains are described, for example, in Sands and Podolsky, Annu. Rev. Physiol. 58: 253-273 (1996); Carr et al., PNAS USA 91: 2206-2210 (1994); DeA et al., PNAS USA 91: 1084-1088 (1994); Hoffman et al., Trends Biochem Sci 18: 239-243 (1993).

Exemplary trefoil domain sequences and consensus sequences are as follows:

  Another exemplary monomer domain suitable for use in the present invention is a thyroglobulin domain. A thyroglobulin monomer domain is typically about 30-85 or 30-80 amino acids. In some embodiments, this domain comprises about 35-75 amino acids and optionally about 65 amino acids. Within 35-75 amino acids there are typically about 6 cysteine residues. Of the six cysteines, disulfide bonds are typically found between the following cysteines: C1 and C2, C3 and C4, C5 and C6.

  To date, at least 251 natural thyroglobulin domains have been identified based on cDNA sequences. The N-terminal part of Tg contains 10 repeats of a domain of about 65 amino acids known as type 1 Tg repeat PUBMED: 3595599, PUBMED: 8797845. Exemplary proteins containing natural thyroglobulin domains include, for example, HLA class II binding invariant chain, human pancreatic cancer marker protein, nidogen (entactin), insulin-like growth factor binding protein (IGFBP), saxiphilin, chum salmon egg cysteine proteinase Inhibitors, and ectatin. Thyr-1 and related domains belong to the MEROPS proteinase inhibitor family I31, clan IX. Thyroglobulin domains are described, for example, by Molina et al., Eur. J. Biochem. 240: 125-133 (1996); Guncar et al., EMBO J 18: 793-803 (1999); Chong and Speicher, DW 276: 5804. -5813 (2001).

Exemplary thyroglobulin domain sequences and consensus sequences are as follows:

  Another exemplary monomer domain that can be used in the present invention is a laminin EGF domain. A laminin EGF domain is typically about 30-85 or 30-80 amino acids. In some embodiments, this domain comprises about 45-65 amino acids and optionally about 50 amino acids. There are typically about 8 cysteine residues in 45-65 amino acids that interact to form four disulfide bonds. Laminin is the main non-collagenous component of the basement membrane that mediates cell adhesion, proliferation migration, and differentiation. They are composed of different but related α, β, and γ chains. The three chains form a cross-shaped molecule consisting of a long arm and three short spherical arms. The long arm consists of a coiled-coil structure contributed by all three chains and bridged by interchain disulfide bonds.

Exemplary laminin EGF domain sequences and consensus sequences are as follows:

  As noted above, the monomer domain can be a natural variant or a non-natural variant. The term “natural” is used herein to refer to the ability of a subject to be found in nature. For example, natural monomer domains include human monomer domains or optionally domains derived from different species or sources such as mammals, primates, rodents, fish, birds, reptiles, plants, etc. be able to. Natural monomer domains can be obtained by several methods, for example, by PCR amplification of genomic DNA or cDNA. The library of monomer domains utilized in the practice of the present invention can include natural monomer domains, non-natural monomer domain variants, or combinations thereof.

  Monomer domain variants can include ancestral domains, random domains, chimeric domains, mutant domains, and the like. For example, ancestral domains can be based on phylogenetic analysis. A random domain is a domain in which one or more regions are randomized. Randomization can be based on overall randomization or, optionally, partial randomization based on the natural distribution of sequence diversity. A chimeric domain is a domain in which one or more regions are replaced by corresponding regions from other domains of the same family. For example, a chimeric domain can be constructed by combining loop sequences from multiple related domains of the same family to form a new domain that may be reduced in immunogenicity. Those skilled in the art recognize the immunological advantages of building modified binding domain monomers by combining loop regions from various related domains of the same family, rather than generating random amino acid sequences. Will. For example, by constructing a mutant domain by combining natural loop sequences or optionally multiple loop sequences in a human type I thrombospondin monomer domain, thyroglobulin monomer domain, or trefoil monomer domain. The resulting domain may contain new binding properties, but since all exposed loops are of human origin, it should not contain immunogenic protein sequences. Combining loop amino acid sequences in an endogenous relationship can be applied to all of the monomer constructs of the present invention.

  Non-natural monomer domains or modified monomer domains can be produced by several methods. Any method of mutagenesis can be used, including site-directed mutagenesis and random mutagenesis (eg, chemical mutagenesis). In some embodiments, error-prone PCR is utilized to generate mutants. Further methods include aligning the natural monomer domains by aligning the conserved amino acids in the natural monomer domains; and maintaining the conserved amino acids and the amino acids around the conserved amino acids. Designing non-natural monomer domains by insertion, deletion or modification to create non-natural monomer domains. In one embodiment, the conserved amino acid comprises cysteine. In another embodiment, the insertion step uses random amino acids, or optionally, the insertion step uses portions of the natural monomer domain. This portion can ideally encode a loop of domains from the same family. Amino acids are inserted or exchanged using synthetic oligonucleotides or by shuffling or by restriction enzyme-based recombination. The human chimeric domains of the invention are useful for therapeutic applications where minimal immunogenicity is desired. The present invention provides a method of generating a library of human chimeric domains.

  The multimeric or monomeric domains of the invention can be produced by any method known in the art. In some embodiments, E. coli containing a plasmid that encodes the polypeptide under the transcriptional control of a bacterial promoter is used to express the protein. After collection, they can be lysed by sonication, heating, or homogenization and clarified by centrifugation. The polypeptide can be purified using a Ni-NTA agarose column (if 6 × His tag is attached) or DEAE Sepharose elution (if not tagged) and refolded by dialysis. Misfolded proteins can be nullified by capping free sulfhydryls with iodoacetic acid. Polypeptides can be purified using Q sepharose elution, butyl sepharose influx, SP sepharose elution, DEAE sepharose elution, and / or CM sepharose elution. An equivalent purification step by anion and / or cation exchange or hydrophobic interaction may be utilized.

  In some embodiments, the monomer or multimer is purified by heat lysis, typically followed by rapid cooling to prevent most of the protein from restoring. Due to the thermal stability of the proteins of the present invention, the desired protein should not be denatured by heating and therefore should allow a high purity purification step (ie, purification to remove contaminating proteins). In some embodiments, a continuous flow heating process is used to purify the monomer or multimer from the bacterial cell culture. For example, the cell suspension can be passed through a stainless steel coil submerged in a water bath set at a temperature that results in bacterial lysis (e.g., about 5, 10, 15, 20, 25, 30, 35 40, 45, 50, 55, or 60 minutes, approximately 55 ° C, 60 ° C, 65 ° C, 70 ° C, 75 ° C, 80 ° C, 85 ° C, 90 ° C, 95 ° C, or 100 ° C). The lysed effluent is routed through a cooling bath to achieve rapid cooling and prevent the restoration of denatured E. coli proteins. E. coli proteins are denatured and prevented from restoring, but monomers or multimers do not denature under these conditions due to the exceptional stability of the backbone. The heating time is controlled by adjusting the flow rate and the coil length. This method yields active protein in high yield and exceptionally high purity (e.g.> 60%,> 65%,> 70%,> 75%, or> 80%) compared to alternative methods, and clinical material And high-throughput (e.g., 96-well or 384-well) production and large-scale (e.g., about 100 μl to about 1) of materials, including materials for screening assays (e.g., in vitro binding and inhibition assays and cell-based activity assays). 2, 5, 10, 15, 20, 50, 75, 100, 500, or 1000 liters) easy to accept production.

  In some embodiments, after the production of the monomer or multimer of the present invention, the polypeptide is treated in a solution containing iodoacetic acid to cap the free-SH moiety of cysteine not forming a disulfide bond. To do. In some embodiments, 0.1-100 mM (eg, 1-10 mM) iodoacetic acid is included in the solution. Typically, iodoacetic acid can be removed prior to administration to an individual.

  Polynucleotides encoding monomer domains (also referred to as nucleic acids) are typically used to create monomer domains via expression. Nucleic acids encoding monomer domains can be derived from a variety of different sources. A library of monomer domains is prepared by expressing multiple different nucleic acids encoding natural monomer domains, modified monomer domains (i.e., monomer domain variants), or combinations thereof can do.

  Nucleic acids encoding fragments of the natural monomer domain and / or immune domain can be similarly mixed and / or recombinantly (e.g., by using chemically or enzymatically produced fragments) to complete Long modified monomer domains and / or immune domains can be generated. This fragment and monomer domain can likewise be recombined by manipulation of the nucleic acid encoding the domain or fragment. For example, a modified monomer domain can be generated using ligation of nucleic acid constructs that encode fragments of the monomer domain.

  Modified monomer domains also provide a group of synthetic oligonucleotides (e.g., overlapping oligonucleotides) that encode conserved, random, pseudorandom, or defined sequences of peptide sequences, and these sequences Then, it can produce by inserting in the predetermined site | part in the polynucleotide which codes a monomer domain by ligation. Similarly, the sequence diversity of one or more monomer domains can be determined using site-directed mutagenesis, random mutation, pseudorandom mutation, defined kernal mutation, codon-based mutation, etc. Can be expanded by mutating the monomer domain. The resulting nucleic acid molecule can be expanded in the host for cloning and amplification. In some embodiments, the nucleic acid is recombined.

  The present invention also provides a method of screening a library obtained for a monomer domain that recombines a plurality of nucleic acid encoding monomer domains and binds to a desired ligand or ligand mixture. The selected monomer domain nucleic acid is also a neutral sequence (i.e., has a substantial Can be backcrossed by recombination with a polynucleotide sequence encoding a non-functional functional sequence) to produce a natural-like functional monomer domain. In general, during backcrossing, subsequent selections are applied to retain properties, eg, binding to the ligand.

  In some embodiments, the monomer library is prepared recombinantly. In such cases, the monomer domains are isolated and recombined to recombine nucleic acid sequences encoding the monomer domains in a combinatorial manner (recombination is between or within the monomer domains, or within the monomer domains, or It can happen both.) The first step involves the identification of monomer domains that have the desired properties, eg, affinity for a particular ligand. Nucleic acid sequences encoding monomer domains can be recombined or recombined and linked to multimers while maintaining conserved amino acids during recombination.

II. Multimers A method for producing multimers (recombinant mosaic proteins or combinatorial mosaic proteins) is a feature of the present invention. Multimers contain at least two monomer domains. For example, the multimer of the present invention has 2 to about 10 monomer domains, 2 to about 8 monomer domains, about 3 to about 10 monomer domains, about 7 monomer domains , About 6 monomer domains, about 5 monomer domains, or about 4 monomer domains. In some embodiments, the multimer comprises at least 3 monomer domains. In view of the possible range of monomer domain sizes, the multimers of the present invention are, for example, 100 kD, 90 kD, 80 kD, 70 kD, 60 kD, 50 kD, 40 kD, 30 kD, 25 kD, 20 kD, 15 kD, 10 kD, 5 kD or more Can be smaller or larger. Typically, the monomer domains are preselected for binding to the target molecule of interest.

In some embodiments, each monomer domain specifically binds to one target molecule. In some of these embodiments, each monomer binds to a different location (similar to an epitope) on the target molecule. Multiple monomer domains and / or immune domains that bind to the same target molecule cause a binding activity effect that results in improved binding activity of the multimer to the target molecule as compared to each individual monomer. In some embodiments, the multimer is at least about 1.5 times, 2 times, 3 times, 4 times, 5 times, 10 times, 20 times, 50 times or 100 times or 1000 times the binding activity of the monomer domain alone. Have a binding activity of Typically, the multimer has a K _d of less than about 10 ⁻¹⁵ , 10 ⁻¹⁴ , 10 ⁻¹³ , 10 ⁻¹² , 10 ⁻¹¹ , 10 ⁻¹⁰ , 10 ⁻⁹ , or 10 ⁻⁸ . In some embodiments, at least one, two, three, four or more (including all) monomers of the multimer bind ions such as calcium or another ion.

  In another embodiment, the multimer comprises monomer domains having specificity for different target molecules. For example, such diverse monomer domain multimers can specifically bind to different components of the viral replication system or to different serotypes of the virus. In some embodiments, at least one monomer domain binds to a toxin and at least one monomer domain binds to a cell surface molecule, thereby acting as a mechanism for targeting the toxin. In some embodiments, the at least two monomer domains and / or immune domains of the multimer bind to different target molecules in the target cell or tissue. Similarly, a therapeutic molecule can be obtained by linking a therapeutic agent to a multimeric monomer that also includes other monomer domains and / or immune domains with cell or tissue binding specificity. Or it can be targeted to tissue. In some embodiments, different monomers bind to signal transduction pathways, different components of metabolic pathways, or components of different metabolic pathways that exert the same additive or synergistic physiological or biological effects.

  Multimers can contain combinations of various monomer domains. For example, in a single multimer, the selected monomer domains may be similar or identical, and may optionally be different or non-identical. Furthermore, the selected monomer domains include various different monomer domains from the same monomer domain family, or various monomer domains from different domain families, or optionally a combination of both. sell.

Multimers produced in the practice of the present invention can be any of the following:
(1) Homo multimers (multimers of the same domain, ie A1-A1-A1-A1);
(2) Heteromultimers of different domains of the same domain class, for example A1-A2-A3-A4. For example, in heteromultimers, A1, A2, A3 and A4 are those in which non-natural different variants of a particular type I thrombospondin monomer domain, thyroglobulin monomer domain, or trefoil monomer domain Or multimers in which some of A1, A2, A3, and A4 are natural variants of the type I thrombospondin monomer domain, thyroglobulin monomer domain, or trefoil monomer domain.
(3) Heteromultimers of domains from different monomer domain classes, eg, A1-B2-A2-B1. For example, A1 and A2 are two different monomer domains derived from type I thrombospondin (either natural or non-natural), and B1 and B2 are two different monomer domains derived from thyroglobulin (natural or If it is non-natural)

  Multimeric libraries utilized in the practice of the present invention are homomultimers, heteromultimers of different monomer domains of the same monomer class (natural or non-natural), or single monomers from different monomer classes. It can include heteromultimers of somatic domains (natural or non-natural), or combinations thereof. Other exemplary multimers include, for example, trimers and higher order multimers (eg, tetramers).

  As described herein, monomer domains are also readily available in immune domain-containing heteromultimers (i.e., multimers having at least one immune domain variant and one monomer domain variant). Used. Accordingly, a multimer of the invention comprises at least one immune domain such as a minibody, single domain antibody, single chain variable fragment (ScFv), or Fab fragment; and, for example, a type I thrombospondin domain, a type I thyroglobulin Repeat domain, trefoil (P-type) domain, EGF-like domain (e.g. laminin-type EGF-like domain), kringle domain, type I fibronectin domain, type II fibronectin domain, type III fibronectin domain, PAN domain, Gla domain, SRCR domain, Kunitz / bovine pancreatic trypsin inhibitor domain, casal type serine protease inhibitor domain, type C von Willebrand factor domain, anaphylatoxin-like domain, CUB domain LDL receptor class A domain, sushi domain, link domain, type 3 thrombospo Njin domain, immunoglobulin-like domain, C type lectin domain, MAM domain, A type von Willebrand factor domain, somatomedin B domain, WAP type 4 disulfide core domain, C type F5 / 8 domain, hemopexin domain, SH2 domain, It can have at least one monomer domain, such as a SH3 domain, an EF hand domain, a cadherin domain, an annexin domain, a zinc finger domain, and a C2 domain, or variants thereof.

  It is not necessary for the domains to be selected before the domains are joined to form a multimer. On the other hand, the domain can be selected for its ability to bind to the target molecule before being linked to the multimer. Thus, for example, a multimer can comprise two domains that bind to one target molecule and a third domain that binds to a second target molecule.

  Typically, the multimer of the present invention is a single discrete polypeptide. A multimer of partial linker-domain-partial linker moieties is a linkage of multiple polypeptides, each corresponding to a partial linker-domain-partial linker moiety.

Accordingly, multimers of the present invention may have the following characteristics: multivalent, multispecific, single stranded, thermostable, in blood and / or prolonged storage half-life. Furthermore, at least one, more than one or all of the monomer domains may bind ions (eg, metal ions or calcium ions), and at least one, more than one or all of the monomer domains may be type I thrombospondin. May be derived from a monomer domain, a thyroglobulin monomer domain, or a trefoil monomer domain, and at least one, multiple, or all of the monomer domains may not be naturally occurring and / or At least one, multiple or all of the monomer domains may contain 1, 2, 3, or 4 disulfide bonds per monomer domain. In some embodiments, the multimer has at least two (or at least three) monomers in which at least one monomer domain is a non-natural monomer domain and the monomer domain binds calcium. Includes domain. In some embodiments, the multimer comprises at least 4 monomer domains, wherein at least one monomer domain does not occur in nature and wherein:
each monomer domain is 30-100 amino acids and each of the monomer domains contains at least one disulfide bond; or
b. Each monomer domain is 30-100 amino acids and is derived from an extracellular protein; or
c. Each monomer domain is 30-100 amino acids and binds to a protein target.

In some embodiments, the multimer comprises at least 4 monomer domains, wherein at least one monomer domain does not occur in nature, and wherein
each monomer domain is 35-100 amino acids; or
b. Each domain contains at least one disulfide bond and is derived from a human protein and / or an extracellular protein.

In some embodiments, the multimer comprises at least two monomer domains, wherein at least one monomer domain does not occur in nature and each domain is:
a. 25-50 amino acids in length and containing at least one disulfide bond; or
b. 25-50 amino acids long and derived from extracellular protein; or
c. binds to 25-50 amino acids and protein target; or
d. 35-50 amino acids long.

In some embodiments, the multimer comprises at least two monomer domains, wherein at least one monomer domain is not naturally occurring and
each monomer domain contains at least one disulfide bond; or
b. at least one monomer domain is derived from an extracellular protein; or
c. At least one monomer domain binds to the target protein.

  In some embodiments, the multimers of the invention bind to the same or other multimer to produce an aggregate. Aggregation is mediated, for example, by the presence of two monomer domains and / or a hydrophobic domain on the immune domain, leading to the formation of non-covalent interactions between the two monomer domains and / or immune domains Can do. Alternatively, aggregation can be facilitated by one or more monomer domains in a multimer that have binding specificity for monomer domains in another multimer. Aggregates can also be caused by the presence of affinity peptides on monomer domains or multimers. Aggregates can contain more target molecule binding domains than a single multimer.

  A multimer with affinity for both a cell surface target and a second target can provide a higher binding activity effect. In some cases, membrane fluidity may be more flexible than protein linkers in optimizing interaction spacing and valency (due to self-assembly). In some cases, a multimer can bind to two different targets, each on a different cell or on a molecule that has multiple binding sites, one on the cell and the other on the other.

III. Linker The selected monomer domains can be linked by a linker to form a single chain multimer. For example, the linker is located between each separate individual monomer domain in the multimer. Typically, the immune domains are also linked to each other or to the monomer domain via a linker moiety. The linker moieties that can be readily used to link immune domain variants together are the same as those described for multimers of monomer domain variants. Exemplary linker moieties suitable for linking immune domain variants to other domains to make multimers are described herein.

  Linking selected monomer domains via a linker can be accomplished using various techniques known in the art. For example, combinatorial assembly of polynucleotides encoding selected monomer domains can be achieved by restriction digestion and religation, by PCR-based, self-priming duplication reactions, or by other recombinant methods. The linker can be attached to the monomer before the monomer is identified for its ability to bind to the target multimer or after the monomer is selected for its ability to bind to the target multimer. .

  The linker can be natural, synthetic, or a combination of both. For example, the synthetic linker can be a random linker (eg, in both sequence and size). In one aspect, the random linker can comprise a fully randomized sequence, or optionally, the random linker can be based on a natural linker sequence. A linker can include, for example, a non-polypeptide moiety, a polynucleotide, a polypeptide, and the like.

  The linker can be rigid, or alternatively, flexible, or a combination of both. Linker flexibility can be a function of the composition of both the linker and the monomer with which the linker interacts. The linker joins the two selected monomer domains and maintains the monomer domains as separate monomer domains. A linker can work together with separate individual monomer domains and still have separate properties such as multiple binding sites in a multimer for the same ligand, or multiple multiple sites in a multimer for different ligands, for example. It may be possible to maintain a separate binding site. In some cases, a disulfide bridge is present between two linked monomer domains or between a linker and a monomer domain. In some embodiments, the monomer domain and / or linker comprises a metal binding center.

  Selection of a suitable linker in certain cases where two or more monomer domains (i.e., polypeptide chains) are linked, e.g., the nature of the monomer domains, polypeptide multimers should bind It may depend on various parameters including the structure and properties of the target and / or the stability of the peptide linker against proteolysis and oxidation.

  The present invention provides a method for optimizing the choice of linker once the desired monomer domain / variant has been identified. In general, multimeric libraries having a composition that is fixed with respect to the composition of the monomer domains but can vary in the composition and length of the linker can be readily prepared and screened as described above. .

  Typically, the linker polypeptide may mainly comprise amino acid residues selected from Gly, Ser, Ala and Thr. For example, the peptide linker is at least 80% of amino acid residues selected from Gly, Ser, Ala and Thr, such as at least 75%, such as at least 85% or at least 90% (in total number of residues present in the peptide linker). Calculated on the basis of The peptide linker can likewise consist only of Gly, Ser, Ala and / or Thr residues. A linker polypeptide has two monomer domains that exhibit the correct conformation with respect to each other so that they retain the desired activity as, for example, an antagonist of a given receptor. It should be long enough to link two monomer domains.

  Suitable lengths for this purpose are at least 1 amino acid residue, and typically less than about 50 amino acid residues (e.g., 2-25 amino acid residues, 5-20 amino acid residues, 5-15 Amino acid residues, 8-12 amino acid residues, or 11 amino acid residues). Similarly, a polypeptide encoding a linker may range in size from, for example, about 2 to about 15 amino acids, about 3 to 15 amino acids, about 4 to 12 amino acids, about 10 amino acids, about 8 amino acids, or about 6 amino acids. sell. In methods and compositions containing nucleic acids (e.g., DNA, RNA, or a combination of both), the polynucleotide comprising the linker sequence is, for example, between about 6 nucleotides to about 45 nucleotides, between about 9 nucleotides to about 45 nucleotides. Between about 12 nucleotides and about 36 nucleotides, about 30 nucleotides, about 24 nucleotides, or about 18 nucleotides. Similarly, the amino acid residues selected for inclusion in the linker polypeptide should exhibit properties that do not significantly interfere with the activity or function of the polypeptide multimer. Thus, it is natural that the peptide linker does not exhibit an overall charge, which is inconsistent with the activity or function of the polypeptide multimer, interferes with intrinsic folding, or is polysylated against the target in question. It binds or forms other interactions with amino acid residues in one or more monomer domains that severely interfere with the binding of peptide multimers.

  In another embodiment of the invention, the peptide linker is selected from a library, wherein amino acid residues in the peptide linker are randomized for a specific set of monomer domains in a specific polypeptide multimer. Is done. With flexible linkers, you can find the right combination of monomer domains and optimize it with this random library of variable linkers to obtain linkers with optimal length and shape Like. The optimal linker is the minimum number of correctly typed amino acid residues that participate in binding to the target and limit movement of the monomer domains relative to each other in the polypeptide multimer when not binding to the target Can be included.

  The use of natural and artificial peptide linkers to link polypeptides to novel linking fusion polypeptides is well known in the literature (Hallewell et al., (1989), J. Biol. Chem. 264, 5260-5268; Alfthan et al., (1995) Protein Eng. 8, 725-731; Robinson and Sauer (1996), Biochemistry 35, 109-116; Khandekar (1997), J. Biol. Chem. 272, 32190-32197; Fares et al (1998), Endocrinology 139, 2459-2464; Smallshaw et al., (1999), Protein Eng. 12, 623-630; US Pat. No. 5,856,456).

One example of the widespread use of peptide linkers is for the production of single chain antibodies in which the variable regions of the light chain (V _L ) and heavy chain (V _H ) are linked via an artificial linker. There are numerous publications in this particular field. A widely used peptide linker is a 15mer consisting of 3 repeats of the Gly-Gly-Gly-Gly-Ser amino acid sequence ((Gly ₄ Ser) ₃ ). Other linkers have been used, and phage display technology, as well as selective infectious phage technology, has been used to diversify and select appropriate linker sequences (Tang et al., (1996), J. Biol. Chem. 271, 15682-15686; Hennecke et al., (1998), Protein Eng. 11, 405-410). Peptide linkers are in heterodimeric and homodimeric proteins such as T cell receptor, λCro repressor, P22 phage Arc repressor, IL-12, TSH, FSH, IL-5, and interferon-γ Has been used to connect individual strands. Peptide linkers have also been used to create fusion polypeptides. Various linkers have been used, and in the case of Arc repressor phage display, phage display has been used to optimize linker length and composition in order to increase the stability of single chain proteins. (Robinson and Sauer (1998), Proc. Natl. Acad. Sci. USA 95, 5929-5934).

  Another type of linker is an intein, ie, a peptide stretch that is expressed with a single chain polypeptide but is post-translationally removed by protein splicing. The use of inteins is reviewed by F.S. Gimble, Chemistry and Biology, 1998, Vol. 5, No. 10, pp. 251-256.

Yet another way to obtain a suitable linker is by optimizing a simple linker, such as (Gly ₄ Ser) _n , through random mutagenesis.

  As mentioned above, it is generally preferred that the peptide linker has at least some flexibility. Thus, in some embodiments, the peptide linker comprises 1-25 glycine residues, 5-20 glycine residues, 5-15 glycine residues, or 8-12 glycine residues. Peptide linkers typically comprise at least 50% glycine residues, for example at least 75% glycine residues. In some embodiments of the invention, the peptide linker comprises only glycine residues.

から選択され、かつx、yおよびzは、各々、1〜5の範囲の整数である。いくつかの態様において、各Xaaは、独立して、Ser、AlaおよびThr、特にSerから選択される。より詳細には、ペプチドリンカーは、アミノ酸配列Gly-Gly-Gly-Xaa-Gly-Gly-Gly-Xaa-Gly-Gly-Glyを有し、ここでXaaは、独立して、

からなる群より選択される。いくつかの態様において、各Xaaは、独立して、Ser、AlaおよびThr、特にSerから選択される。 In addition to glycine residues, the peptide linker comprises other residues, in particular residues selected from Ser, Ala and Thr, in particular Ser. Thus, one example of a specific peptide linker includes a peptide linker having the amino acid sequence Gly _x -Xaa-Gly _y -Xaa-Gly _z , where each Xaa is independently

And x, y and z are each an integer ranging from 1 to 5. In some embodiments, each Xaa is independently selected from Ser, Ala and Thr, particularly Ser. More particularly, the peptide linker has the amino acid sequence Gly-Gly-Gly-Xaa-Gly-Gly-Gly-Xaa-Gly-Gly-Gly, where Xaa is independently

Selected from the group consisting of In some embodiments, each Xaa is independently selected from Ser, Ala and Thr, particularly Ser.

  In some cases, it may be desirable or necessary to provide a peptide linker with some degree of robustness. This can be achieved by including a proline residue in the amino acid sequence of the peptide linker. Accordingly, in another embodiment of the invention, the peptide linker comprises at least one proline residue in the amino acid sequence of the peptide linker. For example, a peptide linker has an amino acid sequence in which at least 25%, such as at least 50%, such as at least 75% of the amino acid sequence is a proline residue. In one particular embodiment of the present invention, the peptide linker contains only proline residues.

  In some embodiments of the invention, the peptide linker is modified in such a way that amino acid residues are introduced that contain a linking group for the non-polypeptide moiety. An example of such an amino acid residue may be a cysteine residue (to which a non-polypeptide moiety is subsequently attached), or the amino acid sequence may be an N-glycosylation site in vivo (thereby providing a sugar moiety ( (In vivo) attached to a peptide linker). A further option is to genetically engineer unnatural amino acids into monomer domains or linkers using evolutionary tRNA and tRNA synthetases (see, for example, US patent application 2003/0082575). . For example, the insertion of keto-tyrosine allows site-specific coupling with the expressed monomer domain or multimer.

  In some embodiments of the invention, the peptide linker comprises at least one cysteine residue, eg, one cysteine residue. Thus, in some embodiments of the invention, the peptide linker comprises an amino acid residue selected from Gly, Ser, Ala, Thr and Cys. In some embodiments, such a peptide linker comprises only one cysteine residue.

In further embodiments, the peptide linker comprises only glycine and cysteine residues, eg, glycine and cysteine residues. Typically, only one cysteine residue is included per peptide linker. Thus, one example of a specific peptide linker comprising a cysteine residue includes a peptide linker having the amino acid sequence Gly _n -Cys-Gly _m , where n and m are each 1 to 12, for example, It is an integer of 3-9, 4-8, or 4-7. More particularly, the peptide linker may have the amino acid sequence GGGGG-C-GGGGG.

This approach (ie, the introduction of an amino acid residue containing a linking group for a non-peptide moiety) can also be used for a more robust proline-containing linker. Thus, a peptide linker can include only proline and cysteine residues, eg, proline and cysteine residues. One example of a specific proline-containing peptide linker comprising a cysteine residue includes a peptide linker having the amino acid sequence Pro _n -Cys-Pro _m , where n and m are each 1 to 12, preferably 3 to 9, for example, an integer of 4-8 or 4-7. More particularly, the peptide linker may have the amino acid sequence PPPPP-C-PPPPP.

  In some embodiments, the purpose of introducing an amino acid residue, such as a cysteine residue, that includes a linking group for a non-polypeptide moiety is to subsequently attach a non-polypeptide moiety to that residue. . For example, a non-polypeptide moiety can improve the serum half-life of a polypeptide multimer. Thus, cysteine residues can be covalently linked to non-polypeptide moieties. Preferred examples of non-polypeptide moieties include PEG or mPEG, particularly polymer molecules such as mPEG, and non-polypeptide therapeutics.

  One skilled in the art will recognize that amino acid residues other than cysteine can be used to attach the non-polypeptide moiety to the peptide linker. One particular example of such other residues includes attaching a non-polypeptide moiety to a lysine residue.

  Another possibility to introduce a site-specific linking group to a non-peptide moiety in a peptide linker is to introduce an in vivo N-glycosylation site, eg, one in vivo N-glycosylation site, into the peptide linker. It is to be. For example, one in vivo N-glycosylation site can be introduced into a peptide linker comprising amino acid residues selected from Gly, Ser, Ala and Thr. In order to ensure that the sugar moiety is actually attached to its in vivo N-glycosylation site, the nucleotide sequence encoding the polypeptide multimer must be inserted into the eukaryotic expression host that performs the glycosylation. It will be understood.

A specific example of a peptide linker containing an N-glycosylation site in vivo is a peptide linker having the amino acid sequence Gly _n -Asn-Xaa-Ser / Thr-Gly _m , preferably Gly _n -Asn-Xaa-Thr-Gly _m Where Xaa is any amino acid residue other than proline, and n and m are each an integer in the range of 1-8, preferably in the range of 2-5.

  Often, the amino acid sequences of all peptide linkers present in a polypeptide multimer are identical. Nevertheless, in certain embodiments, the amino acid sequences of all peptide linkers present in the polypeptide multimer can be different. The latter is particularly relevant when the polypeptide multimer is a trimer or tetramer of a polypeptide, particularly when the peptide linker contains an amino acid residue that contains a linking group for a non-peptide moiety. It is thought that there is.

  Often, to achieve the desired effect (e.g., increased serum half-life), the polypeptide multimer has only a few, typically only one, non-peptide moiety (e.g., mPEG, It is desirable or necessary to attach a sugar moiety or a non-polypeptide therapeutic agent). Obviously, in the case of a polypeptide trimer comprising two polypeptide linkers, only one peptide linker typically needs to be modified, for example by introduction of cysteine residues, On the other hand, modifications of other peptide linkers are typically not necessary. In this case, all (both) peptide linkers of the polypeptide multimer (trimer) are different.

  Thus, in a further embodiment of the invention, the amino acid sequence of all peptide linkers present in the polypeptide multimer is, for example, one or two peptide linkers, except for one, two or three peptide linkers. Except for one peptide linker in particular. The peptide linker has an amino acid sequence that includes amino acid residues that include a linking group for a non-peptide moiety. Preferred examples of such amino acid residues include cysteine residues at N-glycosylation sites in vivo.

The linker can be a natural or synthetic linker sequence. Exemplary natural linkers include, for example, the last cysteine of the first type I thrombospondin monomer domain, the thyroglobulin monomer domain, or the trefoil monomer domain and the second type I thrombospondin monomer. Includes sequences between the monomer domain, the thyroglobulin monomer domain, or the first cysteine of the trefoil monomer domain, which can be used as a linker sequence. Analysis of various domain linkages has shown that natural linkers range from at least 3 amino acids to less than 20 amino acids, e.g., 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16 , 17 or 18 amino acids in length. However, those skilled in the art will recognize that longer or shorter linker sequences may be used. In some embodiments, the linker is a 6mer of the following sequences _{A 1 A 2 A 3 A 4} A 5 A 6, wherein, A ₁ is selected amino acids A, P, T, Q, from E and K A ₂ and A ₃ are any amino acids other than C, F, Y, W or M; A ₄ is selected from amino acids S, G and R; A ₅ is from amino acids H, P and R; And A ₆ is the amino acid T.

  A method of making a multimer from a monomer domain and / or an immune domain can include linking a selected domain with at least one linker to make at least one multimer. For example, a multimer can comprise at least two monomer domains and / or immune domains and a linker. Multimers are then screened for improved binding activity or affinity or altered specificity for the desired ligand or ligand mixture compared to the selected monomer domain. The composition of multimers produced by this method is included in the present invention.

  In other methods, the selected multimer domain is linked to at least one linker to create at least two multimers comprising two or more selected monomer domains and a linker. Two or more multimers are screened for improved binding activity or affinity or altered specificity for the desired ligand or ligand mixture compared to the selected monomer domain. Compositions of two or more multimers produced by the above method are also a feature of the invention.

  Linkers, multimers or selected multimers produced by the methods described above and below are a feature of the present invention. Libraries containing multimers, eg, libraries comprising about 100, 250, 500 or more members produced by or selected by the methods of the invention, are provided. In some embodiments, one or more cells comprising a library member are also included. Recombinant polypeptide libraries, eg, libraries comprising about 100, 250, 500 or more different recombinant polypeptides are also a feature of the invention.

Suitable linkers utilized in the practice of the present invention include absolute heterodimers of partial linker moieties. The term `` absolute heterodimer '' (also referred to as `` affinity peptide '') is used herein to bind two domains together in a specific manner that differs in composition from each other and is non-covalent. A dimer of two partial linker moieties that are linked together. Specific binding is such that two partial linkers bind substantially to each other as compared to binding to other partial linkers. Thus, in contrast to multimers of the invention expressed as a single polypeptide, multimers of domains linked together via heterodimers are discrete, partial linker-monomers- Constructed from partial linker units. Heterodimeric assembly can be achieved, for example, by mixing. Thus, where the partial linker is a polypeptide segment, each partial linker-monomer-partial linker unit can be expressed as a separate peptide prior to multimeric assembly. Disulfide bonds can be added to covalently lock the peptides together after correct non-covalent pairing. Suitable partial linker moieties for forming an absolute heterodimer include, for example, polynucleotides, polypeptides, and the like. For example, if the partial linker is a polypeptide, binding domains are independently produced with their unique linking peptide (ie, partial linker) and later combined to yield a multimer. See, for example, Madden, M., Aldwin, L., Gallop, MA, and Stemmer, WPC (1993) Peptide linkers: Unique self-associative high-affinity peptide linkers. Thirteenth American Peptide Symposium, Edmonton, Canada (abstract) I want. Thus, the spatial order of the binding domains in the multimer is left to the heterodimeric binding specificity of each partial linker. The partial linker may comprise a terminal amino acid sequence that specifically binds to a given heterologous amino acid sequence. Examples of such amino acid sequences are the Hydra neuropeptide head, as described in Bodenmuller et al., The neuropeptide head activator loses its biological activity by dimerization, (1986) EMBO J 5 (8): 1825-1829. Activator. See, for example, US Pat. No. 5,491,074 and WO 94/28173. These partial linkers allow multimers to be produced initially as monomer-partial linker units or partial linker-monomer-partial linker units. These are then mixed together and allowed to build in an ideal order based on the binding specificity of each partial linker. Alternatively, the monomer linked to the partial linker may be brought into contact with the surface of a cell or the like, and in this case, a plurality of monomers are bonded to each other and more binding activity is obtained via the partial linker. High complex can be produced. In some cases, binding can occur via random Brownian motion.

  When the partial linker includes a DNA binding motif, each monomer domain has upstream and downstream partial linkers, including DNA binding proteins that exclusively have a unique DNA binding specificity (i.e., Lp-domains). -Lp, where “Lp” is a representation of a partial linker). These domains may be produced individually and then combined with the appropriate nucleotide sequence (i.e., partial linker DNA binding of the two desired domains to link the domains in the desired order. Specific multimers can be constructed by mixing with DNA fragments containing specific recognition sites for proteins. Furthermore, the same domain can be built into many different multimers by adding DNA sequences containing various combinations of DNA binding protein recognition sites. Further randomization of the combination of DNA binding protein recognition sites in the DNA fragment may allow for the assembly of multimeric libraries. This DNA can be synthesized with analogs of the basic backbone to prevent degradation in vivo.

  In some embodiments, the multimer comprises monomer domains that have specificity for different proteins. Different proteins can be related or unrelated. Examples of related proteins include members of the protein family or different serotypes of the virus. Alternatively, multimeric monomer domains can target different molecules (eg, different blood clotting proteins) in a physiological pathway. In yet other embodiments, the monomer domains bind to proteins in unrelated pathways (e.g., two domains bind to blood factors, two other domains bind to inflammation-related proteins, 5 domains bind to serum albumin). In another embodiment, the multimer is composed of monomer domains that bind to different pathogens or contaminants of interest. Such multimers are useful as a single detection agent that can detect the potential of any of several pathogens or contaminants.

IV. Methods for Identifying Monomer Domains and / or Multimers with Desired Binding Affinity The present invention provides a method for identifying monomer domains that bind to a selected or desired ligand or ligand mixture. provide. In some embodiments, monomer domains and / or immune domains are identified or selected for a desired property (e.g., binding affinity), after which the monomer domains and / or immune domains are formed into multimers. . For those embodiments, any method that results in the selection of domains having the desired properties (eg, specific binding properties) can be used. For example, the method includes providing a plurality of different nucleic acids, each nucleic acid encoding a monomer domain; translating a plurality of different nucleic acids, thereby providing a plurality of different monomer domains; Screening different monomer domains for binding of a desired ligand or ligand mixture; and identifying members of a plurality of different monomer domains that bind the desired ligand or ligand mixture.

  Selection of monomer domains and / or immune domains from a library of domains can be accomplished by a variety of procedures. For example, one method for identifying monomer domains and / or immune domains having desired properties includes translating a plurality of nucleic acids that each encode a monomer domain and / or an immune domain, Screening a polypeptide encoded by the nucleic acid, and, for example, identifying a monomer domain and / or immune domain that binds to the desired ligand or ligand mixture, thereby selecting the monomer domain and / or immune Producing a domain. Monomer domains and / or immune domains expressed by each nucleic acid can be tested for their ability to bind ligand by methods known in the art (i.e. panning, affinity chromatography, FACS analysis). .

As described above, the selection of monomer domains and / or immune domains can be based on binding to a ligand such as a target protein or other target molecule (eg, lipid, carbohydrate, nucleic acid, etc.). Other molecules may optionally be included in the method along with the target, eg, ions such as Ca ²⁺ . The ligand may be a known ligand, eg, a ligand known to bind to one of a plurality of monomer domains, or, for example, the desired ligand is an unknown monomer domain ligand. There may be. Other selections of monomer domains and / or immune domains can be based on, for example, inhibition or enhancement of a particular function or activity of the target protein. Target protein activity includes, for example, endocytosis or internalization, induction of the second messenger system, up- or down-regulation of genes, binding to extracellular matrix, release of molecules, or conformational changes Can be. In this case, the ligand need not be known. Selection can also include the use of high throughput assays.

  When monomer domains and / or immune domains are selected based on their ability to bind to a ligand, the basis for the selection usually includes selection based on slow dissociation rates, where high affinity is expected. sell. The valency of the ligand can also be varied to control the average binding affinity of the selected monomer domain and / or immune domain. The ligand can be bound to the surface or substrate at varying concentrations, for example, by including a competitor compound, by dilution, or by other methods known to those skilled in the art. A high concentration (valency) of a given ligand can be used to enrich for monomer domains with relatively low affinity, while a low concentration (valency) is a higher affinity unit. Can preferentially enrich body domains.

  A variety of reporting display vectors or systems can be used to express nucleic acids encoding the monomer domains, immune domains, and / or multimers of the present invention and to test for the desired activity. For example, the phage display system is a system in which monomer domains are expressed as fusion proteins on the phage surface (Pharmacia, Milwaukee Wis.). Phage display involves the display on the surface of filamentous bacteriophages of polypeptide sequences encoding monomeric domains and / or immune domains, typically as fusions with bacteriophage coat proteins. sell.

  In general, in these methods, each phage particle or cell functions as an individual library member that displays a single species of displayed polypeptide in addition to the native phage or cell protein sequence. . Multiple nucleic acids are cloned into phage DNA at sites that cause transcription of the fusion protein, and a portion of the fusion protein is encoded by the multiple nucleic acids. Phage containing nucleic acid molecules undergo replication and transcription in the cell. The leader sequence of the fusion protein directs the transport of the fusion protein to the tip of the phage particle. Thus, the fusion protein partially encoded by this nucleic acid is displayed on the phage particle for detection and selection by the methods described above or below. For example, a phage library can be incubated with a given (desired) ligand so that a phage particle that provides a fusion protein sequence that binds to the ligand does not provide a polypeptide sequence that binds to the given ligand Differential distribution from phage particles. For example, this separation can be provided by immobilizing a given ligand. The phage particles that bind to the immobilized ligand (i.e., library members) are then recovered and replicated to amplify the selected phage subpopulation for subsequent rounds of affinity enrichment and phage replication. Is done. After several rounds of affinity enrichment and phage replication, the phage library members thus selected are isolated and the nucleotide sequence encoding the displayed polypeptide sequence is determined, thereby providing a predetermined The sequence of the polypeptide that binds the ligand of is identified. Such methods are further described in PCT Publication Nos. 91/17271, 91/18980, 91/19818, and 93/08278.

  Examples of other display systems include ribosome display, nucleotide binding display (see, e.g., U.S. Patent Nos. 6,281,344; 6,194,550, 6,207,446, 6,214,553, and 6,258,558), Examples include polysome display and cell surface display. Cell surface displays include various cells such as E. coli cells, yeast cells and / or mammalian cells. When cells are used as a display, for example, nucleic acids obtained by PCR amplification followed by digestion are introduced into the cells and translated. Optionally, a polypeptide encoding a monomer domain or multimer of the invention may be introduced, for example, by injection into a cell.

  One skilled in the art will recognize that the steps of creating mutations and screening for the desired property can be repeated (ie performed recursively) to optimize the results. For example, in phage display libraries or other similar formats, the first screening of the library can be performed with relatively low stringency, thereby selecting as many particles as possible that bind to the target molecule. The selected particles can then be isolated and the polynucleotide encoding the monomer or multimer can be isolated from the particles. Further mutations can then be made from these sequences and subsequently screened with higher affinity.

  The monomer domain can be selected to bind any type of target molecule, including protein targets. Exemplary targets include, for example, IL-6, Alpha3, cMet, ICOS, IgE, IL-1-R11, BAFF, CD40L, CD28, Her2, TRAIL-R, VEGF, TPO-R, TNFα, LFA-1 , TACI, IL-1b, B7.1, B7.2, or OX40, but are not limited to these. When the target is a receptor for a ligand, the monomer domain can serve as an antagonist or agonist of the receptor.

  If multimers capable of binding relatively large targets are desired, they can be made by a “walking” selection method. As shown in FIG. 3, the method is performed by providing a library of monomer domains and screening the library of monomer domains for affinity for the first target molecule. Once at least one monomer is identified that binds to the target, that particular monomer is covalently linked to each remaining member of the new library or the original library of monomer domains. Each new library member includes one common domain and at least one domain that is different, ie, randomized. Thus, in some embodiments, the present invention provides multimeric libraries generated using a “walking” selection method. This new multimer (eg, dimer, trimer, tetramer, etc.) library is then screened for multimers that bind to the target with higher affinity and binds to the target with higher affinity Multimers can be identified. The “walking” monomer selection method provides a method for assembling multimers composed of monomers that can act additively or optionally synergistically with each other, subject to linker length limitations. This walking technique is very useful when selecting and assembling multimers that can bind large target proteins with high affinity. Repeat the walking method to add more monomers and thereby two, three, four, five, six, seven, eight or more monomers linked together Can result in multimers.

  In some embodiments, the selected multimer comprises more than two domains. Such multimers can be made in stages, for example in this case the addition of each new domain is tested individually and the effects of the domains are tested sequentially. In an alternative embodiment, the domains are joined to form a multimer comprising three or more domains, which is combined with no prior knowledge of which smaller multimer or each domain binds. select.

  The methods of the present invention also include methods for evolving monomers or multimers. As illustrated in FIG. 10, intradomain recombination can be introduced into the monomer by forming the newly recombined unit throughout the monomer or incorporating portions of different monomers . Different monomers can bind the same target or different targets. For example, in some embodiments, portions of different thrombospondin monomers can be recombined. In some embodiments, a portion of a thrombospondin monomer can be combined with a portion of a thyroglobulin monomer and / or a portion of a trefoil / PD monomer. Recombination between domains (eg, recombining different monomers into or between multimers) or recombination (eg, multiple monomers within a multimer) may be achieved. Recombination between libraries is also contemplated.

  FIG. 8 illustrates the process of intradomain optimization by recombination. Shown is a 3 fragment PCR overlap reaction, recombining 3 segments of a single domain against each other. Duplicate reactions of two, three, four, five or more fragments can be utilized in the same manner as illustrated. This recombination process has many uses. One use is to recombine hundreds of large pools of previously selected clones without sequence information. For each overlap to work, it is sufficient to have one known region of a (relatively) constant sequence that is in the same position in each of the clones (positional approach). Intradomain recombination methods involve sequence-related monomers by standard DNA recombination based on random fragmentation (eg, Stemmer, Nature 370: 389-391 (1994)) and reassembly based on DNA sequence homology. It may be done in a pool of domains and does not require a constant overlap site in all of the recombined clones.

  Another use of this process is multiple separate, unprocessed (meaning unpanned) libraries, in which only one of the loops between cysteines is randomized, To create a library in which different loops are randomized in each library. After panning these libraries separately against the target, the selected clones are randomized. From each panning library, only randomized segments are amplified by PCR, and multiple randomized segments are then combined into a single domain that is panned and / or screened for higher potency Create a library. This process can also be used to shuffle a small number of clones of known sequence.

  Any common sequence may be used as an intersection. In the case of cysteine-containing monomers, cysteine residues are a logical place to cross. However, there are other methods for determining the optimal intersection, such as computer modeling. Alternatively, the residue with the maximum entropy, or the minimum number of intramolecular contacts, can likewise be a good site for crossing.

  The method of evolving a monomer or multimer can include, for example, any or all of the following steps: providing a plurality of different nucleic acids, each of which encodes a monomer domain. Translating a plurality of different nucleic acids to provide a plurality of different monomer domains; screening a plurality of different monomer domains for binding of a desired ligand or ligand mixture; binding a desired ligand or ligand mixture Identifying members of a plurality of different monomer domains to provide selected monomer domains; linking the selected monomer domains with at least one linker to form at least two selected single domains Creating at least one multimer comprising a monomer domain and at least one linker; and selected Screening at least one multimer for improved affinity or binding activity or altered specificity for the desired ligand or ligand mixture as compared to the monomer domain.

  Mutations can be introduced into either monomers or multimers. As noted above, examples of improving the monomer are two or more of the monomers (e.g., three, four, five, or more) under conditions that introduce mutations into the resulting amplification product. Intradomain recombination (eg, by shuffling or other recombination methods) in which the portions are separately amplified, thereby synthesizing a library of variants for different portions of the monomer. Combine the resulting “left” and “right” libraries by overlapping PCR by positioning the 5 ′ end of the central primer within the “central” or “overlapping” sequence where both PCR fragments share something in common Thus, new variants of the original monomer pool can be made. These new mutants can then be screened for the desired property, eg, panned against the target or screened for functional effects. The “center” primer can be selected to correspond to any segment of the monomer, typically one or more common amino acids within the backbone or monomer (e.g., those found in the A domain, etc. Cysteine).

  Similarly, multimers can be made by introducing mutations at the monomer level and then recombining the monomeric mutant library. On a larger scale, multimers (single or pool) with the desired properties can be recombined to form longer multimers. In some cases, mutations are introduced into monomers or linkers (typically synthetically) to form libraries. This may involve, for example, using two different multimers that bind to two different targets, thereby ultimately selecting a multimer that has a portion that binds to one target and a portion that binds another target. Can be achieved. For example, see FIG.

  Additional mutations can be introduced by inserting linkers of different lengths and compositions between domains. This allows the selection of the optimal linker between domains. In some embodiments, the optimal length and composition of the linker will allow optimal binding of the domains. In some embodiments, domains with a specific binding affinity are linked by different linkers, and the optimal linker is selected in the binding assay. For example, domains are selected for the desired binding properties and then formed into a library containing various linkers. The library can then be screened to identify the optimal linker. Alternatively, multimeric libraries may be formed in which the domain or linker effects on target molecular binding are not known.

  The methods of the invention also include making one or more selected multimers by providing a plurality of monomer domains and / or immune domains. Multiple monomer domains and / or immune domains are screened for binding of the desired ligand or ligand mixture. A plurality of domain members that bind the desired ligand or ligand mixture are identified, thereby providing a domain with the desired affinity. Ligating the identified domains with at least one linker to create a multimer, each multimer comprising at least two selected domains and at least one linker; and compared to the selected domains Screening multimers for improved affinity or binding activity or altered specificity for the desired ligand or ligand mixture, thereby identifying one or more selected multimers.

A multimeric library, in some embodiments, is a recombinase-based approach to two or more libraries or monomers or multimers, where each library member includes a recombination site (e.g., a lox site). It can produce by combining in. A larger pool of molecularly diverse library members will in principle contain more variants with the desired properties, such as higher target binding affinity and functional activity. When a library is constructed in a phage vector that can be transformed into E. coli, the size of the library (10 ⁹ to 10 ¹⁰ ) is limited by the transformation efficiency of E. coli. Recombinase / recombination site systems (eg, the Cre-loxP system) and in vivo recombination can be used to generate libraries whose size is not limited by the transformation efficiency of E. coli.

For example, a dimer library having diversity of 10 ¹⁰ , 10 ¹¹ , 10 ¹² , 10 ¹³ or more can be prepared using the Cre-loxP system. In some embodiments, E. coli as the host for one untreated monomer library and filamentous phage carrying another monomer library are used. The library size is in this case limited only by the number of infectious phages (holding one library) and the number of susceptible E. coli cells (holding the other library). For example, 10 ¹² By infecting more than phage 10 ¹² E. coli cells (OD600 = 1 in 1 L), could be used to produce dimeric combinations also of 10 ¹² number.

  Multimer selection can be accomplished using a variety of techniques, including those previously described for identifying monomer domains. Other selection methods include, for example, selection based on improved affinity or binding activity for the ligand or altered specificity relative to the selected monomer domain. For example, selection can be based on selective binding with a particular cell type or with a series of related cell or protein types (eg, different viral serotypes). Then, for example, optimization of the properties selected for the binding activity of the ligand can be achieved by recombining the domains as described in the present invention and the amino acid sequence of the individual monomer domain or linker domain or This can be achieved by manipulating the nucleotide sequence encoding the correct domain.

  One method of identifying multimers can be achieved by displaying multimers. As with the monomer domains, multimers are optional and, as described above, various display systems such as phage display, ribosome display, polysome display, nucleotide binding display (eg, US Pat. Nos. 6,281,344; 6,194,550). No. 6,207,446, 6,214,553, and 6,258,558) and / or may be expressed or displayed on a cell surface display. Cell surface displays can include, but are not limited to, E. coli, yeast or mammalian cells. In addition, multimeric display libraries with multiple binding sites can be panned for binding activity or affinity or altered specificity for a ligand or for multiple ligands.

  Monomers or multimers can be screened for target binding activity in yeast cells using a two-hybrid screening assay. In this type of screening, the monomer or multimer library screened is the formation of a fusion protein between each monomer or multimer in the library and the transcriptional activator fragment of yeast (i.e., Gal4). Cloned into a vector that directs The sequence encoding the “target” protein is cloned into a vector that results in the production of a fusion protein between the target and the remainder of the Gal4 protein (DNA binding domain). The third plasmid contains a reporter gene downstream of the DNA sequence of the Gal4 binding site. Monomers that can bind to the target protein together provide a Gal4 activation domain, thus reconstituting a functional Gal4 protein. This functional Gal4 protein bound to the binding site upstream of the reporter gene results in expression of the reporter gene and selection of a monomer or multimer as the target binding protein. (See Chien et al., (1991) Proc. Natl. Acad. Sci. (USA) 88: 9578; Fields S. and Song O. (1989) Nature 340: 245). The use of the two-hybrid system for library screening is further described in US Pat. No. 5,811,238 (also, Silver SC and Hunt SW (1993) Mol. Biol. Rep. 17: 155; Durfee et al. (1993) Genes Devel. 7: 555; Yang et al., (1992) Science 257: 680; Luban et al., (1993) Cell 73: 1067; Hardy et al., (1992) Genes Devel. 6: 801; Bartel et al., (1993) Biotechniques 14: 920; and Vojtek et al., (1993) Cell 74: 205). Another screening system useful for practicing the present invention is the E. coli / BCCP interaction screening system (Germino et al., (1993) Proc. Nat. Acad. Sci. (USA) 90: 993; Guarente L (1993) Proc. Nat. Acad. Sci. (USA) 90: 1639).

  Other variants involve the use of multiple binding compounds so that the monomer domains, multimers or libraries of these molecules can be screened simultaneously for various ligands or compounds with different binding specificities. it can. Multiple predetermined ligands or compounds can be screened simultaneously in a single library, or can be screened sequentially against several monomer domains or multimers. In one variation, multiple ligands or compounds, each encoded on a separate bead (or part of a bead), under appropriate binding conditions, monomer domains, multimers or live of these molecules. Can be mixed and incubated with the rally. Then, recovery of beads containing multiple ligands or compounds can be utilized to isolate selected monomer domains, selected multimers or library members by affinity selection. In general, subsequent rounds of affinity screening can include beads containing the same mixture of beads, portions thereof, or only one or two individual ligands or compounds. This approach provides efficient screening and is compatible with laboratory automation, batch processing, and high-throughput screening methods.

  In another embodiment, multimers can be screened simultaneously for the ability to bind multiple ligands, each containing a different label. For example, each ligand can be labeled with a different fluorescent label and contacted simultaneously with the multimer or multimer library. Multimers with the desired affinity are then identified based on the presence of the label linked to the desired label (eg, by FACS sorting).

  A library of either monomer domains or multimers (referred to as “affinity agents” in the following discussion for convenience) is screened (ie panned) against multiple ligands in several different formats simultaneously be able to. For example, multiple ligands can be screened in a simple mixture, in an array and displayed on a cell or tissue (e.g., a cell or tissue can be bound by a monomer domain or multimer of the invention). And / or can be immobilized. For example, see FIG. A library of affinity agents may optionally be displayed in yeast or phage display systems. Similarly, if desired, a ligand (eg, encoded in a cDNA library) may be displayed on a yeast or phage display system.

  First, the affinity agent library is panned against multiple ligands. Optionally, the resulting “hits” may be panned against the ligand one or more times to enrich the resulting group of affinity agents.

  If desired, the identity of individual affinity agents and / or ligands can be determined. In some embodiments, affinity agents are displayed on phage. The affinity agent identified as binding in the initial screen is divided into first and second parts. The first part is infected with bacteria, giving rise to either plaques or bacterial colonies, depending on the type of phage used. The expression phage is immobilized and then probed with the ligand displayed on the selected phage as described below.

  The second part is coupled to beads or otherwise immobilized, and the immobilized second part is contacted with a phage display library containing at least some of the ligands in the original mixture. The phage that binds to the second portion is subsequently eluted and contacted with the immobilized phage described in the paragraph above. The interaction between the phages can be detected (eg, using a monoclonal antibody specific for the ligand-expressing phage) and the resulting phage polynucleotides can be isolated.

  In some embodiments, the identity of the affinity agent-ligand pair is determined. For example, if both the affinity agent and the ligand are displayed on phage or yeast, the DNA from that pair can be isolated and sequenced. In some embodiments, polynucleotides specific for the ligand and affinity agent are amplified. The amplification primer for each reaction contains a complementary 5 'sequence so that the resulting amplification product is fused, thereby including a polynucleotide encoding at least a portion of the affinity agent and at least a portion of the ligand. Hybrid polynucleotides can be formed. The resulting hybrid can be used to probe an affinity agent or ligand (eg, cDNA encoded) polynucleotide library to identify both the affinity agent and the ligand. For example, see FIG.

  The above method can be easily combined with “walking” to simultaneously create and identify multiple multimers, each of which binds to a ligand in a ligand mixture. In these embodiments, a first library of affinity agents (monomer domains, immune domains or multimers) is panned against a plurality of ligands, and the eluted affinity agent is the first or second affinity agent. Linked to the library, including multimeric affinity agents (e.g., 2, 3, 4, 5, 6, 7, 8, 9, or more monomers or immune domains ), Which is then panned against multiple ligands. This process can be repeated to continue making larger multimeric affinity agents. Increasing the number of monomer domains can result in higher affinity and binding activity for a particular target. Of course, panning may optionally be repeated at each stage to concentrate important binders. In some cases, walking will be facilitated by inserting a recombination site (eg, a lox site) at the end of the monomer and recombining the monomer library with a recombinase-mediated event.

  The selected multimer of the above method is selected, for example, by recombining or shuffling the selected multimer (recombination can occur between and / or within the multimer). Can be further manipulated, such as by mutating. This results in a modified multimer, which can be screened and selected for members having enhanced properties relative to the selected multimer, thereby producing the selected modified multimer.

In view of the description in the present specification, it is apparent that the following process can be continued. Natural or non-natural monomer domains may be recombined or mutants may be formed. Optionally, domains may be selected first or later for sequences that are unlikely to be immunogenic in the host in which they are intended. Optionally, a phage library containing the recombined domains may be panned for the desired affinity. Monomer domains or multimers expressed by the phage may be screened for IC ₅₀ against the target. Heteromeric or homomeric multimers can be selected. The selected polypeptide may be selected for its affinity for any target, including, for example, hetero- or homo-multimeric targets.

  A significant advantage of the present invention is that monomer domains and / or multimers can be selected using known or unknown ligands. Prior information regarding the ligand structure is not required to isolate the monomer domain of interest or the multimer of interest. The identified monomer domains and / or multimers can have biological activity, which means that they contain at least specific binding affinity for the selected or desired ligand, In some cases, it may further include the ability to block the binding of other compounds, stimulate or inhibit metabolic pathways, act as a signal or messenger, stimulate or inhibit cellular activity, and the like. Monomeric domains can be made to function as ligands for receptors for which the natural ligand for the receptor has not been identified (orphan receptor). These orphan ligands can be made to block or activate the receptor tip to which they bind.

  A single ligand may be used to select monomer domains and / or multimers, or optionally various ligands may be used. The monomer domains and / or immune domains of the present invention can bind a single ligand or various ligands. The multimers of the present invention may have multiple distinct binding sites for a single ligand, or optionally multiple binding sites for various ligands.

V. Libraries The present invention also provides a library of monomer domains and a library of nucleic acids encoding monomer domains and / or immune domains. The library can contain, for example, about 10, 100, 250, 500, 1000, or 10,000 or more nucleic acids encoding monomer domains, or the library can contain, for example, monomer domains. It may contain about 10, 100, 250, 500, 1000 or 10,000 or more polypeptides encoding. The library can include monomer domains containing the same cysteine frame, eg, thrombospondin domain, thyroglobulin domain, or trefoil / PD domain.

  In some embodiments, variants are made by recombining two or more different sequences from the same family of monomer domains (eg, LDL receptor class A domains). Alternatively, two or more different monomer domains from different families can be combined to form a multimer. In some embodiments, the multimer is formed from at least one monomer or monomer variant of the following family class: type I thrombospondin domain, type I thyroglobulin repeat domain, trefoil (P type) domain , EGF-like domain (e.g. laminin-type EGF-like domain), kringle domain, type I fibronectin domain, type II fibronectin domain, type III fibronectin domain, PAN domain, Gla domain, SRCR domain, Kunitz / bovine pancreatic trypsin inhibitor domain, Casal Type serine protease inhibitor domain, type C von Willebrand factor domain, anaphylatoxin-like domain, CUB domain LDL receptor class A domain, sushi domain, link domain, type 3 thrombospondin domain, immunoglobulin -Like domain, C-type lectin domain, MAM domain, A-type von Willebrand factor domain, somatomedin B domain, WAP type 4 disulfide core domain, C-type F5 / 8 domain, hemopexin domain, SH2 domain, SH3 domain, EF hand Domains, cadherin domains, annexin domains, zinc finger domains, and C2 domains and derivatives thereof. In another embodiment, the monomer domain and the different monomer domains can comprise one or more domains found in the Pfam database and / or the SMART database. A library produced by the above method, one or more cells comprising one or more members of the library, and one or more displays comprising one or more members of the library are also included in the present invention. included.

  Optionally, a data set of nucleic acid strings encoding monomer domains includes, for example, a first string encoding monomer domains, one or more strings encoding different monomer domains, and It may be created by mixing and thereby providing a dataset of nucleic acid strings that encode monomer domains, including those described herein. In another embodiment, the monomer domain and the different monomer domains can comprise one or more domains found in the Pfam database and / or the SMART database. The method further includes inputting into the computer a first character string encoding a monomer domain and one or more second character strings encoding different monomer domains and the computer Creating a body string or library can be included.

  The library may be screened for a desired property, such as binding of a desired ligand or ligand mixture, or may be exposed to selective conditions in other ways. For example, a member of a library of monomer domains may be displayed for binding to a known or unknown ligand or ligand mixture, prescreened, or incubated in serum to show the effects of serum proteases. Susceptible clones may be removed. The monomer domain sequence may then be mutagenized (e.g., recombined, chemically modified, etc.) or otherwise modified to allow the new monomer domain to be modified with improved affinity. It may be screened again for binding to the ligand or ligand mixture. Selected monomer domains can be combined or linked to form multimers, which can be screened for improved affinity or binding activity or altered specificity for the ligand or ligand mixture. it can. Modified specificity may mean broader specificity, eg, binding of multiple related viruses, or, optionally, altered specificity is reduced specificity, eg, a ligand May mean a bond within a specific region. One skilled in the art will recognize that there are several methods that can be used to calculate binding activity. See, for example, Mammen et al., Angew Chem Int. Ed. 37: 2754-2794 (1998); Muller et al., Anal. Biochem. 261: 149-158 (1998).

  The present invention is also a method for generating a library of chimeric monomer domains derived from human proteins, corresponding to at least one loop from each of at least two different natural variants of the human protein. Providing a loop sequence that is a polynucleotide or polypeptide sequence; and covalently linking the loop sequences so that each chimeric sequence encodes a chimeric monomer domain having at least two loops. A method comprising the step of creating a library of at least two different chimeric sequences. Typically, a chimeric domain has at least 4 loops, and usually at least 6 loops. As mentioned above, the present invention provides three types of loops identified by specific characteristics such as disulfide bonds, cross-links between secondary protein structures, and the possibility of molecular dynamics (ie flexibility). . The three types of loop sequences are a loop sequence defined by cysteine, a loop sequence defined by structure, and a loop sequence defined by factor B.

  Alternatively, a human chimeric domain library can be generated by modifying the natural human monomer domain at the amino acid level in light of the loop level. In order to minimize the possibility of immunogenicity, only the naturally occurring residues in protein sequences from the same family of human monomer domains are utilized to create a chimeric sequence. This provides a sequence alignment of at least two human monomer domains from the same family of monomer domains and identifies corresponding amino acid residues in human monomer domain sequences that differ between human monomer domains And each human chimeric monomer domain sequence is composed of two amino acid residues corresponding to residues from two or more human monomer domains from the same family of monomer domains, in type and position. Alternatively, it can be achieved by creating more human chimeric monomer domains. The library of human chimeric monomer domains is targeted to the target of interest by screening the library of human chimeric monomer domains for binding to the target molecule and identifying human chimeric monomer domains that bind to the target molecule. It can be used to identify human chimeric monomer domains that bind. Suitable natural human monomer domain sequences utilized in the initial sequence alignment step include those corresponding to any of the natural monomer domains described herein.

  The human chimeric domain library of the present invention (whether created by changing the loop or by changing a single amino acid residue) can be prepared by methods known to those skilled in the art. Particularly suitable methods for generating these libraries are the split-pool format and the trinucleotide synthesis format described in WO 01/23401.

VI. Fusion proteins In some embodiments, a monomer or multimer of the invention is linked to another polypeptide to form a fusion protein. Although any polypeptide in the art can be used as a fusion partner, it can be useful when the fusion partner forms a multimer. For example, the monomers or multimers of the invention can be fused, for example, to the following positions or combinations of positions of antibodies:
1. At the N-terminus of the VH1 and / or VL1 domain, optionally immediately after the leader peptide and before the domain begins (skeletal region 1);
2. At the N-terminus of the CH1 or CL1 domain to replace the VH1 or VL1 domain;
3. At the N-terminus of the heavy chain, optionally after the CH1 domain and before the cysteine residue in the hinge (Fc-fusion);
4. At the N-terminus of the CH3 domain;
5. It may be attached to the C-terminus of the CH3 domain to the last amino acid residue optionally via a short linker;
6. At the C-terminus of the CH2 domain to replace the CH3 domain;
7. At the C-terminus of the CL1 or CH1 domain, optionally after a cysteine forming an interchain disulfide; or
8. At the C-terminus of the VH1 or VL1 domain. For example, see FIG.

  In some embodiments, the monomeric or multimeric domain is linked to a pharmaceutically useful molecule (eg, protein, nucleic acid, small organic molecule, etc.). Exemplary pharmaceutical proteins include, for example, cytokines, antibodies, chemokines, growth factors, interleukins, cell surface proteins, extracellular domains, cell surface receptors, cytotoxins and the like. Exemplary small molecule pharmaceuticals include small molecule toxins or therapeutic agents.

  In some embodiments, the monomer or multimer is selected to bind to a tissue-specific or disease-specific target protein. A tissue-specific protein is a protein that is expressed exclusively or at a significantly higher level in one or several specific tissues compared to other tissues in an animal. Similarly, disease-specific proteins are proteins that are expressed exclusively or at significantly higher levels in one or several diseased cells or tissues compared to other non-lesioned cells or tissues in animals. Examples of such diseases include actinic keratosis, arteriosclerosis, atherosclerosis, bursitis, cirrhosis, hepatitis, mixed connective tissue disease (MCTD), myelofibrosis, paroxysmal night Hemoglobinuria, polycythemia vera, psoriasis, primary thrombocythemia, and cancers including adenocarcinoma, leukemia, lymphoma, melanoma, myeloma, sarcoma, teratocarcinoma, specifically adrenal gland, bladder, bone, Bone marrow, brain, chest, neck, gallbladder, ganglion, gastrointestinal tract, heart, kidney, liver, lung, muscle, ovary, pancreas, parathyroid, penis, prostate, salivary gland, skin, spleen, testis, thymus, thyroid, And proliferative diseases such as cancer of the uterus; acquired immune deficiency syndrome (AIDS), Addison's disease, adult respiratory distress syndrome, allergy, ankylosing spondylitis, amyloidosis, anemia, asthma, atherosclerosis, Autoimmune hemolytic anemia, autoimmune thyroiditis Autoimmune multiple endocrine disorder (APECED), bronchitis, cholecystitis, contact dermatitis, Crohn's disease, atopic dermatitis, dermatomyositis, diabetes, emphysema, lymphotoxic transient lymphopenia, fetus Erythroblastosis, erythema nodosum, atrophic gastritis, glomerulonephritis, Goodpasture syndrome, gout, Graves' disease, Hashimoto's thyroiditis, hypereosinophilia, irritable bowel syndrome, multiple sclerosis, myasthenia gravis Disease, myocardial or pericardial inflammation, osteoarthritis, osteoporosis, pancreatitis, polymyositis, psoriasis, Reiter syndrome, rheumatoid arthritis, scleroderma, Sjogren's syndrome, systemic anaphylaxis, systemic lupus erythematosus, systemic sclerosis , Thrombocytopenic purpura, ulcerative colitis, uveitis, Werner syndrome, cancer, hemodialysis, complications of extracorporeal circulation, viral, bacterial, fungal, parasitic, Autoimmune / inflammatory diseases such as infections caused by protozoa and helminths and trauma; congestive heart failure, ischemic heart disease, angina, myocardial infarction, hypertensive heart disease, degenerative valvular heart disease, Calcified aortic stenosis, congenital aortic bicuspid valve, mitral annulus lime, mitral valve prolapse, rheumatic fever and rheumatic heart disease, infective endocarditis, nonbacterial thrombotic endocarditis, Endocarditis associated with systemic lupus erythematosus, carcinoid heart disease, cardiomyopathy, myocarditis, pericarditis, neonatal heart disease, congenital heart disease, heart transplant complications, arteriovenous fistula, atherosclerosis, hypertension , Vasculitis, Raynaud's disease, aneurysm, arterial dissection, varicose vein, thrombophlebitis and venous thrombosis, vascular tumors, thrombolysis, balloon angioplasty, vascular grafting, and coronary artery bypass Heart diseases such as epilepsy, ischemic Cerebrovascular disorder, stroke, brain neoplasm, Alzheimer's disease, Pick's disease, Huntington's disease, dementia, Parkinson's disease and other extrapyramidal diseases, amyotrophic lateral sclerosis and other motor neuron diseases, progressive neuropathy Muscle atrophy, retinitis pigmentosa, hereditary ataxia, multiple sclerosis and other demyelinating diseases, bacterial and viral meningitis, brain abscess, subdural abscess, epidural abscess, purulent skull Internal thrombophlebitis, myelitis and radiculitis, viral central neuropathy, prion diseases including Kull, Creutzfeldt-Jakob disease, and Gerstmann-Streisler-Scheinker syndrome, lethal familial insomnia, nerves Nutritional and metabolic diseases of the system, neurofibromatosis, tuberous sclerosis, cerebelloretinal hemangioblastomatosis, cerebral trigeminal syndrome, intellectual disability and Other developmental disorders of the central nervous system, including Down's syndrome, cerebral palsy, neurological disorders, autonomic nervous system disorders, cranial nerve disorders, spinal cord disorders, muscular dystrophy and other neuromuscular disorders, peripheral nervous system disorders, dermatomyositis and Psychiatric disorders, including polymyositis, hereditary, metabolic, endocrine and addictive myopathy, myasthenia gravis, periodic limb paralysis, mood disorders, anxiety disorders, and schizophrenia, seasonal Affective disorder (SAD), inability to sit still, amnesia, catatonia, diabetic neuropathy, tardive dyskinesia, myopathy, paranoid psychosis, postherpetic neuralgia, touret disorder, progressive supranuclear palsy, cerebral cortex base Nuclear degeneration, and neurological disorders such as familial frontotemporal dementia; and tubular acidosis, anemia, Cushing syndrome, chondrogenic dystrophy, Duchenne and Becker Muscular dystrophy, epilepsy, gonadal dysfunction, WAGR syndrome (Wilms tumor, aniridia, genitourinary abnormalities, and intellectual disability), Smith-Margenis syndrome, myelodysplastic syndrome, hereditary mucosal epithelial dysplasia, hereditary keratopathy Hereditary neuropathies such as Charcot-Marie-Tooth disease and neurofibromatosis, hypothyroidism, hydrocephalus, Sydenham chorea, and paroxysmal diseases such as cerebral palsy, spina bifida, anencephalopathy, cranial spine, congenital Examples include, but are not limited to, developmental disorders such as glaucoma, cataracts, and sensorineural hearing loss. Exemplary diseases or conditions include, for example, MS, SLE, ITP, IDDM, MG, CLL, CD, RA, factor VIII hemophilia, transplantation, arteriosclerosis, Sjogren's syndrome, Kawasaki disease, antiphospholipid Ab , AHA, ulcerative colitis, multiple myeloma, glomerulonephritis, seasonal allergies, and IgA nephropathy.

  In some embodiments, the monomer or multimer that binds to the target protein is a pharmaceutical protein such that the resulting complex or fusion is targeted to the specific tissue or disease-related cell in which the target protein is expressed. Or linked to a small molecule. Monomers or multimers used in such complexes or fusions can be initially selected for binding to the target protein, and binding is reduced or eliminated in other non-target cells or tissues. It may be desirable to continue to select by negative selection on other cells or tissues (eg to avoid targeting bone marrow or other tissues that determine the lower limit of drug toxicity). By moving the drug away from the sensitive tissue, the therapeutic concentration range is increased, so that higher doses can be safely administered. Another method involves in vivo panning in an animal by injecting a library of monomers or multimers into the animal and then isolating the monomer or multimer that binds to the specific tissue or cell of interest. It can be performed.

  The above fusion protein can also include a linker peptide between the pharmaceutical protein and the monomer or multimer. Peptide linker sequences can be utilized, for example, to separate polypeptide components at a distance sufficient to ensure that each polypeptide is folded into its secondary and tertiary structure. Fusion proteins can generally be prepared using standard techniques, including chemical conjugation. The fusion protein can also be expressed as a recombinant protein in an expression system by standard techniques.

  Exemplary tissue specific or disease specific proteins can be found, for example, in Tables I and II of US Patent Application No. 2002/0107215. Exemplary tissues in which the target protein can be specifically expressed are, for example, liver, pancreas, adrenal gland, thyroid, salivary gland, pituitary, brain, spinal cord, lung, heart, chest, skeletal muscle, bone marrow, thymus, spleen, lymph Includes nodal, colorectal, stomach, ovary, small intestine, uterus, placenta, prostate, testis, colon, colon, gastrointestinal, bladder, trachea, kidney, or adipose tissue.

VII. Compositions The present invention also includes compositions produced by the methods of the present invention. For example, the invention includes monomer domains selected or identified from a library containing monomer domains produced by the methods of the invention.

  Nucleic acid and polypeptide compositions are included in the invention. For example, the invention provides a plurality of different nucleic acids, each nucleic acid encoding at least one monomer domain or immune domain. In some embodiments, the at least one monomer domain is an EGF-like domain (e.g., a laminin EGF domain), a trefoil (P-type) domain, a type I thyroglobulin repeat, a type I thrombospondin domain, and one or more of these Selected from a plurality of variants. Suitable monomer domains also include those described in the Pfam database and / or SMART database.

  The invention also relates to recombinant nucleic acids encoding one or more polypeptides comprising a plurality of monomer domains, the monomer domains comprising sequences or sequence modifications relative to the native polypeptide. provide. For example, the natural polypeptide is selected from an EGF-like domain (eg, a laminin EGF domain), a trefoil (P-type) domain, a type I thyroglobulin repeat domain, a type I thrombospondin domain, and one or more variants thereof. can do. In another embodiment, the native polypeptide encodes a monomer domain found in the Pfam database and / or SMART database.

  Compositions produced by the methods of the invention, such as the entire composition of the invention, including monomer domains and multimers and libraries thereof, can optionally be bound to a matrix of affinity material. Examples of affinity materials include beads, columns, solid supports, microarrays, other pools of reagent supports, and the like. In some embodiments, screening in solution uses a biotinylated target. In these embodiments, the target is incubated with a phage library and the target with bound phage is captured by streptavidin beads.

  The compositions of the invention can be bound to a matrix of affinity material, such as a recombinant polypeptide. Examples of affinity materials include beads, columns, solid supports and the like.

VIII Therapeutic and Prophylactic Treatment Methods The present invention also provides one or more of the above-described nucleic acids or polypeptides of the invention (or pharmaceutically acceptable excipients and one or more such A composition comprising a nucleic acid or polypeptide) in vivo or ex vivo, e.g., a mammal including a human, primate, mouse, pig, cow, goat, rabbit, rat, guinea pig, hamster, horse, sheep, or not a mammal It includes methods of treating a disease or disorder therapeutically or prophylactically by administering to a subject, including vertebrates, eg, birds (eg, chickens or ducks), fish, or invertebrates.

  In one aspect of the invention, in an ex vivo method, one or more cells or a population of cells of interest in a subject (e.g., tumor cells, tumor tissue samples, organ cells, blood cells, skin cells, lungs, Heart, muscle, brain, mucous membrane, liver, intestine, spleen, stomach, lymphatic system, cervix, vagina, prostate, mouth, tongue, etc.) are obtained or removed from the subject, causing a disease, disorder, or other condition It is contacted with an amount of selected monomer domains and / or multimers of the present invention effective for prophylactic or therapeutic treatment. The contacted cells are then returned or delivered to the subject at the site from which it was obtained or another site of interest in the subject to be treated, including, for example, those defined above. The If desired, the contacted cells are transplanted to a site of tissue, organ, or system of interest in a subject (including all of the above) using standard and well-known transplant techniques. Or can be delivered to the blood or lymphatic system using, for example, standard delivery or transfusion techniques.

  The invention also relates to the invention wherein one or more cells or a population of cells of interest in a subject are effective to preventively or therapeutically treat a disease, disorder, or other condition. In vivo methods are provided in which a certain amount of selected monomer domains and / or multimers are contacted directly or indirectly. In a direct contact / administration mode, the selected monomer domains and / or multimers typically are treated cells or sites of tissue of interest (e.g., tumor cells, tumor tissue samples, Organ cells, blood cells, skin cells, lungs, heart, muscle, brain, mucous membrane, liver, intestine, spleen, stomach, lymphatic system, cervix, vagina, prostate, mouth, tongue, etc.), local administration, injection (e.g. , By use of a needle or syringe), or by direct delivery or delivery by any of a variety of modes, including vaccine delivery or gene gun delivery that pushes into tissue, organ, or skin tissue. The selected monomer domains and / or multimers can be delivered, for example, intramuscularly, intradermally, subdermally, subcutaneously, orally, intraperitoneally, intrathecally, intravenously, or It can be placed in a body cavity (eg, including during surgery) or delivered by inhalation or vaginal or rectal administration. In some embodiments, the protein of the invention is prepared at a concentration of at least 25 mg / ml, 50 mg / ml, 75 mg / ml, 100 mg / ml, 150 mg / ml or more. Such a concentration is useful, for example, in subcutaneous formulations.

  In an in vivo indirect contact / administration mode, the selected monomer domain and / or multimer is a polypeptide or polypeptide of the invention from which one or more cells or cells from which treatment can be facilitated. By directly contacting or administering to the population, typically the cells to be treated or the tissue of interest, including those described above (e.g. skin cells, organ system, lymphatic system, blood cell system, etc.) It is administered or transported indirectly to the site. For example, tumor cells in a subject's body are directed to selected monomer domains and / or multimeric sites of interest (e.g., tissues, organs or cells of interest, or blood or lymphatic system). Contact of blood or lymphoid cells, skin, or organs with a sufficient amount of selected monomer domains and / or multimers such that delivery of the drug is effected to provide effective prophylactic or therapeutic treatment Can be treated. Such contact, administration, or transport is typically performed using one or more of the routes or modes of administration described above.

  In another aspect, the present invention provides an ex vivo method, wherein one or more cells of interest or a population of cells of interest in a subject (e.g., tumor cells, tumor tissue samples, organ cells, Blood cells, skin cells, lungs, heart, muscle, brain, mucous membrane, liver, intestine, spleen, stomach, lymphatic system, cervix, vagina, prostate, mouth, tongue, etc.) obtained or removed from the subject A biologically active polypeptide of interest (e.g., a selection) effective to preventively or therapeutically treat a disease, disorder, or other condition in one or more cells or populations of cells. By contacting with a polynucleotide construct comprising a nucleic acid sequence of the invention that encodes a monomer domain and / or multimer). The one or more cells or populations of cells are contacted with a sufficient amount of the polynucleotide construct and a promoter that controls expression of the nucleic acid sequence, so that the polynucleotide construct (and promoter) is transferred to the cell. Selected monomer domains and / or effective for uptake, resulting in sufficient expression of the target nucleic acid sequences of the present invention to prophylactically or therapeutically treat a disease, disorder, or condition A certain amount of biologically active polypeptide encoding the multimer is produced. This polynucleotide construct comprises a promoter sequence that controls the expression of a nucleic acid sequence of the invention (e.g., a CMV promoter sequence) and / or, optionally, at least one or more other polypeptides of the invention, It may include one or more additional nucleotide sequences encoding cytokines, adjuvants, or costimulatory molecules, or other polypeptides of interest.

  After transfection, the transformed cells are directed to the subject to the tissue site or system from which they were obtained or to another site that is treated within the subject (e.g., tumor cells, tumor tissue samples, organ cells, blood cells, skin Cell, lung, heart, muscle, brain, mucous membrane, liver, intestine, spleen, stomach, lymphatic system, cervix, vagina, prostate, mouth, tongue, etc.), delivered or transported. If desired, the cells can be transplanted into a tissue, skin, organ, or body system of interest within the subject using standard and well-known transplant techniques, or blood using standard delivery or transfusion techniques. Or it can be delivered to the lymphatic system. Such delivery, administration, or transport of transformed cells is typically performed using one or more of the above administration routes or modes. Expression of the target nucleic acid may occur or be induced naturally (described in further detail below), and the amount of the encoded polypeptide sufficient and effective to treat the disease or condition at the site or tissue system. Expressed.

  In another aspect, the invention provides a biologically active polypeptide of interest (e.g., a selected single molecule) that is effective in prophylactically or therapeutically treating a disease, disorder, or other condition. Contacting a cell or population of cells with a polynucleotide construct comprising a nucleic acid sequence of the invention encoding a multimeric domain and / or a multimer) (or using one or more of the above routes or modes of administration) By administering or transporting to a population of cells) a population of cells or cells of interest or a subject's cells (including the cells and cell lines and subjects described above) in the subject's body In vivo methods of being transformed in are provided.

  The polynucleotide construct can be directly administered or transported to a diseased or disordered cell (eg, by direct contact using one or more of the above routes or modes of administration). Alternatively, the polynucleotide construct can be administered or transported indirectly to cells affected by the disease or disorder. This can be achieved by using a non-disease cell or a cell with another disease using one or more of the above administration routes or modes and a nucleic acid sequence encoding a biologically active polypeptide and This is done by first contacting directly with a sufficient amount of the polynucleotide construct, including a promoter that controls expression. As a result, uptake of the polynucleotide construct (and promoter) into the cell occurs, resulting in sufficient expression of the nucleic acid sequences of the invention, in an amount effective to prevent or therapeutically treat the disease or disorder. Producing a biologically active polypeptide, whereby the polynucleotide construct or the resulting expressed polypeptide is naturally or automatically, in the subject's body, the first delivery site, system, tissue, Or from the organ to the site of the disease, tissue, organ or system of the subject's body (eg, via the blood or lymphatic system). The expression of the target nucleic acid can occur naturally or can be induced (as described in further detail below) so that a certain amount of expressed polypeptide is present at that site or tissue system. It is sufficient and effective to treat a disease or condition. This polynucleotide construct comprises a promoter sequence that controls the expression of a nucleic acid sequence of the invention (e.g., a CMV promoter sequence) and / or, optionally, at least one or more other polypeptides of the invention, It may include one or more additional nucleotide sequences encoding cytokines, adjuvants, or costimulatory molecules, or other polypeptides of interest.

  In each of the in vivo or ex vivo methods of treatment as described above, a composition comprising an excipient and a polypeptide or nucleic acid of the invention can be administered or delivered. In one aspect, a composition comprising a pharmaceutically acceptable excipient and a polypeptide or nucleic acid of the invention is administered to a subject as described above in an amount effective to treat a disease or disorder, or Delivered.

  In another aspect, in each of the above in vivo and ex vivo treatment methods, the amount of polynucleotide administered to the cell or subject is such that uptake of the polynucleotide into one or more cells of the subject occurs An amount of biologically active polypeptide effective to enhance an immune response in a subject, including an immune response induced by an immunogen (e.g., an antigen) when sufficient expression of the nucleic acid sequence occurs. It can be in such an amount that it produces. In another aspect, in each of such methods, the amount of polypeptide administered to the cell or subject includes an immune response in the subject, including one induced by an immunogen (e.g., an antigen). It can be an amount sufficient to enhance.

  In yet another aspect, in an in vivo or in vivo processing method, wherein a polynucleotide construct (or a composition comprising a polynucleotide construct) is used to deliver a physiologically active polypeptide to a subject, Nucleotide construct expression can be induced using inducible on and off gene expression systems. Examples of such on and off gene expression systems include the Tet-On ™ gene expression system and the Tet-Off ™ gene expression system, respectively (eg, a detailed description of each such system. (See Clontech catalog 2000, pages 110-111). Other controllable or inducible on and off gene expression systems are known to those skilled in the art. With such a system, the expression of the target nucleic acid of the polynucleotide construct can be regulated in an accurate, reversible and quantitative manner. Gene expression of the target nucleic acid can be achieved, for example, when a stable transfected cell containing a polynucleotide construct comprising the target nucleic acid is delivered or transported to the tissue site, organ or system of interest, or the tissue site of interest. Can be induced after being made in contact with an organ or system. Such systems are particularly advantageous in methods and modes of treatment where it is advantageous to delay or precisely control the expression of the target nucleic acid. (Eg, to allow time for completion of surgery and / or to allow post-surgical healing; time for a polynucleotide construct comprising the target nucleic acid to reach the site, cell, system, or tissue being treated) To allow time for the implant containing cells transformed with the construct to be taken up on or in the tissue or organ to which the implant has been joined or adhered).

IX. Further Multimer Uses The potential applications of the multimers of the present invention are diverse and include any use where an affinity agent is desired. For example, the present invention can be used in applications for making antagonists. Here, the selected monomer domain or multimer blocks the interaction between the two proteins. Optionally, the present invention may make an agonist. For example, a multimer that binds two different proteins, eg, an enzyme and a substrate, can enhance the function of the protein, including, for example, enzyme activity and / or substrate conversion.

  Other applications include cell targeting. For example, multimers consisting of monomeric domains and / or immune domains that recognize specific cell surface proteins can selectively bind to specific cell types. Applications that include monomer domains and / or immune domains as antiviral agents are also included. For example, multimers that bind to different epitopes on viral particles can be useful as antiviral agents because of their multivalent nature. Other applications can include, but are not limited to, protein purification, protein detection, biosensors, ligand-affinity capture experiments, and the like. Furthermore, domains or multimers can be synthesized in bulk by conventional means for any suitable use, for example as a therapeutic or diagnostic agent.

  In some embodiments, the monomer domain is used for ligand inhibition, ligand clearance or ligand stimulation. Possible ligands in these methods include, for example, cytokines, chemokines, or growth factors.

  If inhibition of ligand binding to the receptor is desired, select a monomer domain that binds to the ligand at the portion of the ligand that contacts the receptor of the ligand, or binds to the receptor at the portion of the receptor that contacts the ligand. Thereby blocking the ligand-receptor interaction. If desired, the monomer domain may optionally be linked to a half-life extending molecule.

  Ligand clearance refers to regulating the half-life of a soluble ligand in body fluids. For example, in the absence of a half-life extending molecule, most monomer domains have a short half-life. That is, binding of the monomer domain to the ligand will reduce the half-life of the ligand, thereby reducing the ligand concentration. The portion of the ligand to which the monomer domain is bound is generally not critical, but it is beneficial to bind the ligand with the portion of the ligand that binds to its receptor, which can further inhibit the effect of the ligand. . This method is useful for reducing the concentration of any molecule in the bloodstream. In some embodiments, the concentration of molecules in the bloodstream is reduced by promoting the renal clearance rate of the molecules. Typically, the monomer domain-molecule complex is less than about 40 KDa, less than about 50 KDa, or less than about 60 KDa.

  Alternatively, using a multimer comprising a first monomer domain that binds to a half-life extending molecule and a second monomer domain that binds to a portion of the ligand that does not bind to the receptor of the ligand, the half-life of the ligand Can be increased.

  The invention further provides monomer domains that bind to blood factors (eg, serum albumin, immunoglobulins, or red blood cells).

  In some embodiments, the monomer domain binds to an immunoglobulin polypeptide or portion thereof.

  Four families of monomer domains that bind immunoglobulin (ie, families 1, 2, 3 and 4) have been identified.

Family 1 sequences are shown below. Dash symbols are included only for spacing.

Family 2 has the following motifs:

An exemplary sequence comprising an IgG family 2 motif is shown below. Dash symbols are included only for spacing.

Family 3 has either of the following two motifs:

An exemplary sequence containing an IgG family 3 motif is shown below. Dash symbols are included only for spacing.

An additional non-natural monomer domain that binds IgG and has the sequence SSGR immediately before the third cysteine in the A domain backbone, based on family 3 alignment. The sequences of these monomer domains are shown below. Dash symbols are included only for spacing.

Monomeric domains that bind to red blood cells (RBC) or serum albumin (CSA) are described in US Patent Application No. 2005/0048512, including, for example:

  The present invention provides a method of extending the serum half-life of a protein, including, for example, a protein of interest in a multimer or animal of the present invention. The protein of interest can be any protein (including another monomer domain or multimer of the present invention) having a therapeutic, prophylactic, or otherwise desirable functionality. This method specifically binds albumin (eg, human serum albumin) or serum proteins such as transferrin, IgG or parts thereof, blood transported molecules such as red blood cells or half-life extending molecules such as cells. Initially providing a monomer domain identified as a binding protein. In some embodiments, the half-life extending molecule-binding monomer can be covalently linked to another monomer domain that has binding affinity for the protein of interest. This multimer, which may optionally bind a protein of interest, is administered to a mammal, where the multimer is associated with a half-life extending molecule (e.g., HSA, transferrin, IgG, erythrocyte, etc.) to form a complex. Could be formed. This complex formation protects the multimer and / or binding protein from proteolysis and / or other removal of the multimer and / or protein, thereby extending the half-life of the protein and / or multimer. (See, eg, Example 3 below). One variation of this application of the invention involves a half-life extending molecule-binding monomer covalently linked to a protein of interest. The protein of interest can include a monomer domain, a multimer of monomer domains, or a synthetic drug. Alternatively, monomers that bind to either immunoglobulins or erythrocytes could be made by the method described above and used to extend half-life.

  Half-life extending molecule-bound multimers typically bind to molecules or cells carried by blood with at least two domains, a chimeric domain, or a mutant domain (i.e. one that binds to the target of interest). One) multimer. Appropriate domains, such as those described herein, can be further screened and selected for binding to a half-life extending molecule. Half-life extending molecule-binding multimers are made, for example, by methods of making multimers described herein using monomer domains prescreened for half-life extending molecule-binding activity. For example, some half-life extending molecule-bound LDL receptor class A domain monomers are described in Example 2 below.

  In some embodiments, the multimer is at least one domain that binds to an HSA, transferrin, IgG, erythrocyte or other half-life extending molecule comprising a trefoil / PD domain motif, thrombosponge provided herein And a multimer is at least another domain that binds a target molecule, the trefoil / PD domain motif provided herein, the thrombospondin domain Contains another motif that contains a motif or thyroglobulin domain motif. The serum half-life of the molecule is, for example, at least 1, 2, 3, 4, 5, 10, 20, 30, 40, 50, 60, 70, 80, 90, 100, 150, 200, 250, 400, 500 It can be extended to be hours or more.

  The present invention also provides a method for suppressing or reducing an immune response in a mammal. This method involves first selecting a monomer domain that binds to an immunosuppressive target. Such an “immunosuppressive target” is defined as any protein that, when bound to another protein, results in an immunosuppressive result in a mammal. The immunosuppressive monomer domain can then be administered directly, or to another protein or to another monomer domain that can result in the desired targeting of the immunosuppressive monomer domain. They may be covalently linked. An immunosuppressive multimer is typically a multimer of at least two domains, a chimeric domain, or a mutant domain. Suitable domains include all of those described herein and are further screened and selected for binding to an immunosuppressive target. Immunosuppressive multimers can be produced by, for example, using the trefoil / PD monomer domain, thrombospondin monomer domain, or thyroglobulin monomer domain, and the method of making a multimer described herein. Produced.

  In another embodiment, a multimer comprising a first monomer domain that binds to a ligand and a second monomer domain that binds to a receptor is used to increase the effective affinity of the ligand for the receptor. Can be made.

  In another embodiment, a multimer comprising at least two monomers that bind to the receptor is used to bring the two receptors closer together by binding the multimer, thereby activating the receptor .

  In some embodiments, multimers with two different monomers can be used to employ a target-driven increase in binding activity. For example, a first monomer can be targeted to a cell surface molecule of a first cell type, and a second monomer can be targeted to a surface molecule of a second cell type. By linking the two monomers to form a multimer, and then adding the multimer to a mixture of the two cell types, binding should occur between cells, with the initial binding event being one multimer. If it happens between the two cells, other multimers will also be able to join them.

  Further examples of potential uses of the invention include drug binding (e.g., binding of radioactive nucleotides for targeting, binding of drugs for extending drug half-life, treatment by overdose and control agents for addiction therapy) Binding), immune function modulation (e.g., blocking immunogenicity by binding receptors such as CTLA-4, enhancing immunogenicity by binding receptors such as CD80, or complement activity by Fc-type binding) And specific delivery (eg, sustained release by linker cleavage, electrotransport domain, dimerization domain, or specific binding to: cell entry domain, clearance receptors such as FcR, for transmucosal transport) Oral delivery receptors such as plgR and blood brain barrier transport receptors such as transferrin R), monomer domains, and multimers thereof.

  In further embodiments, the monomer or multimer can be linked to a detectable label (e.g., Cy3, Cy5, etc.) or a reporter gene product (e.g., CAT, luciferase, horseradish peroxidase, alkaline phosphatase, GFP, etc.) ).

  In some embodiments, the monomers of the invention are selected for their ability to bind antibodies from specific animals, such as goats, rabbits, mice, etc., for use as secondary reagents in detection assays.

In some cases, monomeric or multimeric pairs are selected to bind to the same target (ie, for use in sandwich-based assays). In order to select a matched monomer or multimer pair, two different monomers or multimers are typically capable of binding the target protein simultaneously. One technique for identifying such pairs involves the following steps:
(1) Immobilizing a previously selected phage or protein mixture to bind the target protein
(2) contacting the target protein with the immobilized phage or protein and washing;
(3) contacting the phage or protein mixture with the bound target and washing; and
(4) The step of eluting the bound phage or protein without eluting the immobilized phage or protein.
In some embodiments, different phage populations with different drug markers are used.

  One application of the multimeric or monomeric domains of the invention is to replace antibodies or other affinity agents in detection assays or other affinity-based assays. That is, in some embodiments, monomer domains or multimers are selected for their ability to bind components other than the target in the mixture. A general approach can include performing affinity selection under conditions that closely resemble the conditions of the assay, including mimicking the sample composition during the assay. Thus, the selection step may include contacting the monomer domain or multimer with a mixture that does not include the target ligand and negatively selecting any monomer domain or multimer that binds to the mixture. I can do it. Therefore, blocking the mixture (in the absence of target ligand that can be depleted using antibodies, monomer domains or multimers) corresponding to the sample of the assay (serum, blood, tissue, cells, urine, semen, etc.) It can be used as an agent. Such subtraction is useful, for example, to make pharmaceutical proteins that bind to the target but not to other serum proteins or non-target tissues.

X. Further Manipulation of Monomeric Domains and / or Multimeric Nucleic Acids and Polypeptides As noted above, the polypeptides of the present invention can be modified. A description of a variety of different production means for producing modified or modified nucleic acid sequences encoding these polypeptides is given above and below in the following publications and references cited therein. is described.

ウラシル含有鋳型を用いる突然変異誘発

オリゴヌクレオチド指向性突然変異誘発

ホスホロチオエート修飾DNA突然変異誘発

ギャップを有する二重鎖DNAを用いる突然変異誘発

が挙げられる。 Mutation methods that create diversity include, for example, site-directed mutagenesis

Mutagenesis using uracil-containing templates

Oligonucleotide-directed mutagenesis

Phosphorothioate-modified DNA mutagenesis

Mutagenesis using double-stranded DNA with gaps

Is mentioned.

Further suitable methods include point mismatch repair (Kramer et al., Point Mismatch Repair, (1984) Cell 38: 879-887), mutagenesis using repair-deficient host strains (Carter et al., (1985) Nucl Acids Res. 13: 4431-4443; and Carter, (1987) Methods in Enzymol . 154: 382-403), deletion mutagenesis (Eghtedarzadeh & Henikoff, (1986) Nucl. Acids Res. 14: 5115), Restriction selection and purification (Wells et al., (1986) Phil. Trans. R. Soc. Lond. A 317: 415-423), mutagenesis by total gene synthesis (Nambiar et al., (1984) Science 223: 1299-1301; Sakamar and Khorana, (1988) Nucl. Acids Res. 14: 6361-6372; Wells et al., (1985) Gene 34: 315-323; and Grundstrom et al., (1985) Nucl. Acids Res. 13: 3305-3316),
Double strand break repair (Mandecki, (1986) PNAS USA , 83: 7177-7181; and Arnold, (1993) Curr. Op. Biotech. 4: 450-455). Further details on many of the above methods can be found in Methods in Enzymology, Volume 154, which also describes useful standards for solving problems with various mutagenesis methods .

の中で見出すことができる。 More details on how to create various diversity can be found at

Can be found inside.

の中で見出される。インビトロで増幅された核酸をクローニングする改良法は、Wallaceら、米国特許第5,426,039号に記述されている。PCRによって大きな核酸を増幅する改良法は、Cheng et al., (1994) Nature 369：684-685およびその中で引用されている参考文献に要約されており、その中では40 kbまでのPCRアンプリコンが作製されている。逆転写酵素およびポリメラーゼを用いて、本質的にどんなRNAでも制限消化、PCR伸長および配列決定に適した二本鎖DNAに変換できることを当業者は理解するであろう。Ausubel、SambrookおよびBerger、全て前記を参照されたい。 Another aspect of the invention involves cloning and expression of a monomer domain encoding a nucleic acid, a selected monomer domain, a multimer and / or a selected multimer. Thus, multimeric domains can be synthesized as a single protein using expression systems well known in the art. In addition to much of the text above, vectors, promoters and more relevant for expressing nucleic acids such as monomer domains, selected monomer domains, multimers and / or selected multimers General texts describing molecular biology techniques useful herein, including the use of other topics, include Berger and Kimmel, Guide to Molecular Cloning Techniques, Methods in Enzymology Vol. 152, Academic Press, Inc. , San Diego, CA (Berger); Sambrook et al., Molecular Cloning-A Laboratory Mannual (2nd edition) Volumes 1-3, Cold Spring Harbor Laboratory, Cold Spring Harbor, New York, 1989 ("Sambrook") and Current Protocols in Molecular Biology , FM Ausubel et al. (Ed.), Current Protocols (Joint venture between Greene Publishing Associates, Inc. and John Wiley & Sons, Inc.) (additional since 1999) (`` Ausubel '') included . Identify monomer domains and multimers that encode nucleic acids, including polymerase chain reaction (PCR), ligase chain reaction (LCR), Q-replicase amplification, and other RNA polymerase-mediated techniques (e.g., NASBA). Examples of techniques sufficient to guide one skilled in the art by in vitro amplification methods useful in releasing and cloning are Berger, Sambrook, and Ausubel, and

Found in An improved method for cloning nucleic acids amplified in vitro is described in Wallace et al., US Pat. No. 5,426,039. Improved methods for amplifying large nucleic acids by PCR are summarized in Cheng et al., (1994) Nature 369: 684-685 and references cited therein, in which PCR amplifiers up to 40 kb are included. Recon is being made. One skilled in the art will understand that using reverse transcriptase and polymerase, essentially any RNA can be converted to double stranded DNA suitable for restriction digestion, PCR extension and sequencing. See Ausubel, Sambrook and Berger, all above.

  The present invention also provides for the introduction of the vectors of the present invention into host cells and the monomer domains, selected monomer domains, immune domains, multimers and / or selected macromolecules of the present invention by recombinant techniques. It relates to the production of the body. Host cells are genetically engineered (ie, introduced, transformed or transfected) with the vectors of the invention, which can be, for example, a cloning vector or an expression vector. The vector can be in the form of, for example, a plasmid, viral particle, phage, or the like. Genetically engineered host cells can be selected to activate promoters, select transformants, or monomer domains of interest, selected monomer domains, multimers and / or selected It can be cultured in conventional nutrient media modified to be suitable for amplifying multimeric genes. Culture conditions such as temperature, pH, etc., were previously used with the host cell selected for expression and will be apparent to those of skill in the art, as well as, for example, Freshney (1994) Culture of Animal Cells , a Manual of Basic Technique, 3rd edition, Wiley-Liss, New York and references cited therein, will be apparent from the references cited herein.

  As mentioned above, the polypeptides of the present invention can also be produced in cells other than animals such as plants, yeasts, fungi, bacteria and the like. Indeed, as stated throughout, phage display is a particularly relevant technique for producing such polypeptides. In addition to Sambrook, Berger and Ausubel, more details on cell culture can be found in Payne et al., (1992) Plant Cell and Tissue Culture in Liquid Systems John Wiley & Sons, Inc. New York, NY; Gamborg and Phillips (ed.) 1995) Plant Cell, Tissue and Organ Culture; Fundamental Methods Springer Lab Manual, Springer-Verlag (Berlin Heidelberg New York) and Atlas an Parks (ed) The Handbook of Microbiological Media (1993) CRC Press, Boca Raton, FL Can be found.

  The invention also provides for improving pharmaceutical properties, reducing immunogenicity, or transporting multimeric and / or monomeric domains to cells or tissues (e.g., through the blood brain barrier, or Including modification of monomer domains, immune domains and / or multimers to facilitate (through the skin). These types of alterations include various modifications (e.g., addition of sugar groups or glycosylation), addition of PEG, addition of protein domains that bind certain proteins (e.g., HSA or other serum proteins), to cells. , Including the addition of protein fragments or sequences that move or transport signals from and through the cell. Additional components can be added to the multimer and / or monomer domain to manipulate the properties of the multimer and / or monomer domain. For example, selective cleavage to activate domains that bind known receptors (e.g., Fc-region protein domains that bind Fc receptors), toxins or portions of toxins, multimers or monomer domains Add various components, including possible prodomains, reporter molecules (e.g., green fluorescent protein), components that bind reporter molecules (such as radionuclides for radiotherapy, biotin or avidin) or combinations of modifications You can also

XI. Additional Screening Methods The present invention also provides a method for screening proteins for potential immunogenicity by the following steps:
Providing a candidate protein sequence;
Comparing the candidate protein sequence to a database of human protein sequences;
Identifying a portion of the candidate protein sequence that corresponds to a portion of the human protein sequence from the database; and determining a degree of match between the candidate protein sequence and the human protein sequence from the database.

  Generally, the greater the degree of match between a candidate protein sequence and one or more of the human protein sequences from the database, the more immunogens are compared to the candidate protein that has little match to any of the human protein sequences from the database. The possibility of sex is expected to be low. Limiting or eliminating the number of immunogenic amino acids and / or sequences can also be utilized to reduce the immunogenicity of the monomer domains, for example, before or after screening the library. Immunogenic sequences include, for example, HLA type I or type II sequences or proteasome sites. Various commercially available products and computer programs such as Tepitope (Roche), Parker Matrix, ProPred-I matrix, Biovation, Epivax, Epimatrix can be used to identify these amino acids.

)。この方法は、例えば、キメラ単量体ドメインなどの、キメラタンパク質中の交差配列が免疫原性による事象を引き起こす可能性が高いかどうかを判定する際に特に有用である。交差配列がヒトタンパク質配列のデータベースで見出される配列の部分に当たるなら、交差配列が免疫原性による事象を引き起こす可能性はいっそう低いと考えられる。 A database of human protein sequences suitable for use in carrying out the methods of the invention for screening candidate proteins can be found on the World Wide Web at ncbi.nlm.nih.gov/blast/Blast.cgi ( In addition, you can search for a short, almost complete match using the following website: World Wide Web

). This method is particularly useful in determining whether a crossover sequence in a chimeric protein, such as a chimeric monomer domain, is likely to cause an immunogenic event. If the crossover sequence corresponds to a portion of the sequence found in the human protein sequence database, the crossover sequence is less likely to cause an immunogenic event.

  Human chimeric domain libraries prepared by the methods of the present invention can be screened for potential immunogenicity in addition to binding affinity. Furthermore, a protein library of a human-like chimeric protein can be designed using information on a portion of a human protein sequence derived from a database. Such libraries can be generated using information regarding “crossover sequences” present in natural human proteins. The term “crossover sequence” is used herein to refer to a sequence that is found in its entirety in at least one natural human protein, a portion of which is found in two or more natural proteins. . That is, recombination of the latter two or more natural proteins could produce a chimeric protein in which the chimeric portion of the sequence actually corresponds to the sequence found in another natural protein. The crossover sequence is found in the first and second natural human protein sequences, but not in the third natural human protein sequence, and the first amino acid position by an amino acid residue of the same type and position. Contains a chimeric junction at two consecutive amino acid residue positions occupied. The second amino acid position is occupied by amino acid residues of the same type and position that are found in the second and third natural human protein sequences but not in the first natural human protein sequence . In other words, the “second” native human protein sequence corresponds to the native human protein in which the crossover sequence appears in its entirety as described above.

  According to the present invention, a library of human-like chimeric proteins is generated by the following steps: identifying human protein sequences from a database corresponding to proteins from the same family of proteins; Aligning a human protein sequence with a reference protein sequence; a set of subsequences derived from different human protein sequences of the same family, each of which is at least one other subsequence derived from a different native human protein sequence; Identifying subsequences that share a region of identity; chimera junctions from the first, second, and third subsequences, each subsequence derived from a different native human protein sequence; Common to the first and second natural human protein sequences, but the third natural human Two consecutive amino acid residues that are occupied by an amino acid residue that is not common to the protein sequence, and that the second amino acid position is occupied by an amino acid residue that is common to the second and third natural human protein sequences Identifying a chimeric junction comprising the determined amino acid residue positions; and each sequence corresponding to two or more partial sequences from a set of partial sequences, each comprising a plurality of identified chimeric junctions Creating a human-like chimeric protein molecule comprising one of them.

  Thus, for example, if the first natural human protein sequence is A-B-C, the second is B-C-D-E, and the third is D-E-F, the chimeric junction is CD. Alternatively, if the first natural human protein sequence is D-E-F-G, the second is B-C-D-E-F, and the third is A-B-C-D, the chimeric junction is D-E. Human-like chimeric protein molecules can be produced by various methods. For example, an oligonucleotide comprising a sequence encoding a chimeric junction is recombined with an oligonucleotide whose sequence corresponds to two or more subsequences from the set of subsequences described above to form a human-like chimeric protein, and Libraries can be created. The reference sequence used to align the native human proteins is a sequence derived from the same family of native human proteins, or a chimera or other variant of a protein within the family.

XII. Animal Models Another aspect of the invention is the development of specific non-human animal models for testing the immunogenicity of monomeric or multimeric domains. A method for producing such a non-human animal model is a method for producing a gene encoding a plurality of human proteins derived from the same family of proteins in at least some cells of the recipient non-human animal, Introducing a vector containing a gene that is operably linked to a promoter that is functional in at least some of the cells into which the vector is introduced so that a transgenic non-human animal capable of expressing multiple human proteins is obtained. Including stages.

  Suitable non-human animals utilized in the practice of the present invention include all vertebrates (eg, mice, rats, rabbits, sheep, etc.) except humans. Typically, a plurality of members of a protein family includes at least two members of the family, and usually at least 10 family members. In some embodiments, the plurality includes all known members of the protein family. Exemplary genes that can be used encode monomer domains, such as members of the type I thrombospondin domain family, thyroglobulin domain family, or trefoil domain family, and other domain families described herein. Including what to do.

  The non-human animal model of the present invention can be used to screen for immunogenicity of monomeric or multimeric domains derived from the same family of proteins expressed by the non-human animal model. The present invention is a DNA molecule encoding a non-human animal model made by the above method, as well as a plurality of human proteins (such as the monomer domains described herein) derived from proteins of the same family. DNA molecules that have been introduced into transgenic non-human animals at the embryogenesis stage and that are operably linked to promoters in at least some of the cells into which the DNA molecules have been introduced, somatic cells and reproductive cells Includes transgenic non-human animals that cells contain and express.

  An example of a mouse model useful for screening binding proteins from type I thrombospondin, thyroglobulin, or trefoil domains is described as follows. Gene clusters encoding wild type human type I thrombospondin monomer domain, thyroglobulin monomer domain, or trefoil monomer domain are amplified from human cells by PCR. Next, using these fragments, a transgenic mouse is prepared by the method described above. This transgenic mouse recognizes the human type I thrombospondin domain, thyroglobulin domain, or trefoil domain as “self”, and thus human “self” for the type I thrombospondin domain, thyroglobulin domain, or trefoil domain. Will imitate. Individual type I thrombospondin-derived monomers, thyroglobulin-derived monomers, or trefoil-derived monomers or multimers are monomers of type I thrombospondin-derived monomers or multimers, thyroglobulin-derived monomers. These mice are tested by injecting them with a monomer or multimer, or a monomer or multimer from trefoil, and then analyzing the induced immune response (or lack of response). Mice are tested to determine if they expressed a mouse anti-human response (MAHR). Monomers and multimers that do not cause the development of MAHR are likely to be non-immunogenic when administered to humans.

  Historically, the MAHR test in transgenic mice has been used to test individual proteins in mice that are transgenic for that single protein. On the other hand, the above method is a non-human animal model that recognizes the entire human protein family as “self”, and evaluates a large number of mutant proteins each with significantly altered binding activity and ability to use. Provide a non-human animal model that can be used for

XIII. Kits Components required in the methods of the invention (typically in unmixed form) and kit components (packaging materials, ingredients and / or instructions for using the method, instructions for use) Kits that contain one or more containers (reaction tubes, columns, etc.) for holding are a feature of the invention. The kit of the present invention may comprise a multimeric library or a single type of library. The kit can also include a reagent suitable for facilitating binding of the target molecule, such as a buffer or a reagent that facilitates detection, including detectably labeled molecules. Standards for calibrating ligand binding to monomer domains and the like can also be included in the kits of the invention.

  The present invention also provides commercially valuable binding assays and kits for practicing the assays. In some of the assays of the invention, one or more ligands are utilized to detect monomer domain, immune domain and / or multimer binding. Such an assay detects any binding of the ligand to the monomer domain and / or multimer to detect any known method in the art, such as flow cytometry, fluorescence microscopy, plasmon resonance, etc. based on.

  Assay-based kits are also provided. The kit typically includes a container and one or more ligands. The kit may optionally include instructions for performing the assay, additional detection reagents, buffers, instructions for use of any of these components, and the like. Alternatively, the kit can include cells, vectors (eg, expression vectors, secretion vectors, including polypeptides of the invention) for expression of the monomer domains and / or multimers of the invention.

  In a further aspect, the invention relates to the use of any composition, monomer domain, immune domain, multimer, cell, cell culture, device, device component or kit herein, any composition herein. Method, assay practice, and / or use of any device or kit to practice any assay or method herein and / or cell, cell culture, composition herein as a therapeutic formulation Provide the use of objects or other features. Also provided is the manufacture of all ingredients herein as a therapeutic formulation for the treatments described herein.

XIV. Integration System The present invention relates to a computer comprising a monomer domain, a selected monomer domain, a multimer and / or a selected multimer and a character string corresponding to a nucleic acid encoding such a polypeptide. Computer readable media and integrated systems are provided. These sequences can be manipulated by in silico recombination methods or by standard sequence alignment or text processing software.

  For example, different types of similarities and various stringency considerations and string lengths can be detected and recognized in the integrated systems herein. For example, many homology determination methods have been designed for comparative analysis of biopolymer sequences, for spell checking in text processing, and for retrieving data from various databases. A model that mimics the annealing of complementary homologous polynucleotide strings as well as an understanding of the interaction of a double helix pair complement between four basic nucleobases in a natural polynucleotide is also described herein. As a basis for sequence alignment or other operations typically performed on strings corresponding to sequences in (e.g., text processing operations, construction of diagrams containing strings of sequences or subsequences, output tables, etc.) Can do. One example of a software package that uses GO to calculate sequence similarity is BLAST, which can be adapted to the present invention by entering a string corresponding to the sequences herein.

BLAST is described in Altschul et al., (1990) J. Mol. Biol. 215: 403-410. Software for performing BLAST analyzes is publicly available through the National Center for Biotechnology Information (available at ncbi.nlm.nih.gov on the World Wide Web). This algorithm first calculates whether a short word of length W in the query sequence (which matches or satisfies a certain positive threshold score T when aligned with a word of the same length in the database sequence) Identifying a high-scoring sequence (HCP) pair. T is referred to as the neighborhood word score threshold (Altschul et al., Supra). These initial neighborhood word hits act as sources for initiating searches to find longer HSPs containing them. The word hits then extend in both directions along each sequence as long as the cumulative alignment score can be increased. The cumulative score is calculated for a nucleotide sequence using parameter M (reward score for matched residue pair; always> 0) and parameter N (penalty score for mismatched residue; always <0). For amino acid sequences, a cumulative score is calculated using a scoring matrix. Word hit extension in each direction is stopped when: The cumulative alignment score decreases by an amount X from its maximum achieved value; due to accumulation of one or more negative score residue alignments If the cumulative score is 0 or less; or if either end of the sequence is reached. The BLAST algorithm parameters W, T, and X determine the sensitivity and speed of the alignment. The BLASTN program (for nucleotide sequences) uses 11 word lengths (W), 10 expected values (E), 100 cut-offs, M = 5, N = -4 and a comparison of both strands as default. For amino acid sequences, the BLASTP program defaults to 3 word lengths (W), 10 expected values (E), and BLOSUM62 scoring matrix (Henikoff & Henikoff (1989) Proc. Natl. Acad. Sci. USA 89: 10915).

A further example of a useful sequence alignment algorithm is PILEUP. PILEUP uses progressive, pairwise alignments to create multiple sequence alignments from groups of related sequences. This can similarly plot a dendrogram showing the accumulated associations used to create the alignment. PILEUP uses a simplification of the progressive alignment method of Feng & Doolittle, (1987) J. Mol. Evol. 35: 351-360. The method used is similar to the method described by Higgins & Sharp, (1989) CABIOS 5: 151-153. The program can, for example, align up to 300 sequences with a maximum length of 5000 characters. The multiple alignment procedure begins with the alignment of the two most similar sequence pairs that produce a cluster of two aligned sequences. This cluster can then be aligned to the next most related sequence or cluster of aligned sequences. A cluster of two sequences can be aligned by simple extension of the alignment of two individual sequence pairs. Final alignment is achieved by a series of progressive, pairwise alignments. This program can also be used to plot an accumulated related phylogenetic tree or dendrogram. This program is executed by designating a specific sequence and its amino acid or nucleotide coordinates for the sequence comparison region. For example, to determine conserved amino acids in a monomer domain family or to compare the sequences of monomer domains in a family, align the sequences of the invention, or encoding nucleic acid, to provide structure-function information To do.

  In one aspect, the computer system is used to perform “in silico” sequence recombination or shuffling of strings corresponding to monomer domains. Various such methods were filed on Feb. 5, 1999 (U.S. Patent Application No. 60/118854), Selifonov and Stemmer, `` Methods For Making Character Strings, Polynucleotides & Polypeptides Having Desired Characteristics '' and 1999. It is shown in “Methods For Making Character Strings, Polynucleotides & Polypeptides Having Desired Characteristics” filed on Oct. 12 (US Patent Application No. 09 / 416,375) by Selifonov and Stemmer. Briefly, gene operators are used in genetic algorithms to alter a given sequence by mimicking genetic phenomena such as mutation, recombination, death, and the like. Multidimensional analysis to optimize the sequence can also be performed in a computer system, for example, as described in the '375 application.

  The digital system can also instruct the oligonucleotide synthesizer to synthesize oligonucleotides used, for example, for gene reconstruction or recombination, or to order oligonucleotides from commercial sources (For example, by printing the appropriate order form or linking to the order form on the Internet).

  This digital system also provides an output element for controlling nucleic acid synthesis (e.g., based on sequence or recombination alignment, e.g., based on a recombinant monomer domain, as described herein). That is, the integrated system of the present invention may optionally include an oligonucleotide synthesizer or an oligonucleotide synthesis controller. The system may include other operations that occur downstream of the alignment or other operations that are performed using strings corresponding to the sequences herein, for example, as described above with respect to the assay. Can do.

Examples The following examples are provided for the purpose of illustration and are not intended to limit the invention described in the claims.

Example 1
This example describes the selection of monomer domains and the production of multimers.

  Starting materials and procedures for identifying monomer domains and for producing multimers from selected monomer domains can be derived from any of a variety of human and / or non-human sequences. For example, a family of monomer domains in which one or more monomer domain genes bind to specific ligands to produce selected monomer domains with specific binding to the desired ligand or mixture of ligands Selected from. Nucleic acid sequences encoding one or more monomer domain genes can be obtained by PCR amplification of genomic DNA or cDNA, or optionally can be produced synthetically using overlapping oligonucleotides. .

  Most commonly, these sequences are then cloned into a cell surface display format (ie, bacterial, yeast, or mammalian (COS) cell surface display; phage display) for expression and screening. The recombinant sequences are transfected (transduced or transformed) into appropriate host cells where they are expressed and displayed on the cell surface. For example, cells can be stained with a desired ligand that is labeled (eg, fluorescently labeled). Stained cells are sorted by flow cytometry and selected monomer domains encoding the gene are recovered from positive cells (eg, by plasmid isolation, PCR or propagation and cloning). The staining and fractionation process can be repeated multiple times (eg, using progressively decreasing concentrations of the desired ligand until the desired level of enrichment is obtained). Alternatively, any screening or detection method known in the art that can be used to identify cells that bind the desired ligand or ligand mixture can be utilized.

  A gene encoding a selected monomer domain recovered from a desired ligand or ligand mixture that binds cells may optionally be recombined by any of the methods described herein or in the cited references. . The recombinant sequences produced in this round of diversification are then screened by the same or different methods to identify recombinant genes with improved affinity for the desired or target ligand. The diversification and selection process may optionally be repeated until the desired affinity is obtained.

  The selected monomer domain nucleic acids selected by these methods can be obtained, for example, by combinatorial assembly of nucleic acid sequences encoding the selected monomer domains by DNA ligation or, optionally, by PCR-based self-priming overlap reactions. Can be linked together via a linker sequence to create a multimer. The nucleic acid sequence encoding the multimer is then cloned into a cell surface display format (ie, bacterial, yeast, or mammalian (COS) cell surface display; phage display) for expression and screening. The recombinant sequences are transfected (transduced or transformed) into appropriate host cells where they are expressed and displayed on the cell surface. For example, the cells can be stained with a labeled, eg, fluorescently labeled, desired ligand or ligand mixture. Stained cells are sorted by flow cytometry and selected multimers encoding the gene are recovered from positive cells (eg, by PCR or growth and cloning). Positive cells contain multimers with improved binding activity or affinity or altered specificity for the desired ligand or ligand mixture compared to the selected monomer domain. The staining and fractionation process can be repeated multiple times (eg, using progressively decreasing concentrations of the desired ligand or ligand mixture until the desired level of enrichment is obtained). Alternatively, any screening or detection method known in the art that can be used to identify cells that bind the desired ligand or ligand mixture can be utilized.

  A gene encoding a selected multimer recovered from a desired ligand or ligand mixture that binds cells may optionally be recombined by any of the methods described herein or in the cited references. The recombinant sequences produced in this round of diversification are then the same or different to identify recombinant genes with improved binding activity or affinity or altered specificity for the desired or target ligand. Screen by method. The diversification and selection process may optionally be repeated until the desired binding activity or affinity or altered specificity is obtained.

Example 2
This example describes the selection of monomer domains that can bind to human serum albumin (HSA).

For phage production, E. coli DH10B cells (Invitrogen) were transformed with a phage vector encoding a library of LDL receptor class A domain variants as a fusion with pIII phage protein. To transform these cells, the electroporation system MicroPulser (Bio-Rad) was used with the cuvette provided by this same manufacturer. The DNA solution was mixed with 100 μl of cell suspension, incubated on ice and transferred into a cuvette (electrode gap 1 mm). After pulsing, 2 ml of SOC medium (2% w / v tryptone, 0.5% w / v yeast extract, 10 mM NaCl, 10 mM MgSO ₄ , 10 mM MgCl ₂ ) is added and the transformation mixture is incubated at 37 ° C. for 1 hour. Incubated. Multiple transformants were combined and diluted in 500 ml of 2 × YT medium containing 20 μg / m tetracycline and 2 mM CaCl ₂ . In 10 electroporations using a total of 10 μg of ligated DNA, 1.2 × 10 ⁸ independent clones were obtained.

A 160 ml culture containing cells transformed with a phage vector encoding a library of A domain mutant phages was grown at 22 ° C. and 250 rpm for 24 hours and then transferred to a sterile centrifuge tube. Cells were sedimented by centrifugation (15 minutes, 5000 g, 4 ° C.). The supernatant containing the phage particles was mixed with 1/5 volume of 20% w / v PEG 8000, 15% w / v NaCl and incubated at 4 ° C. for several hours. After centrifugation (20 minutes, 10000 g, 4 ° C.), the precipitated phage particles were dissolved in 2 ml of cold TBS (50 mM Tris, 100 mM NaCl, pH 8.0) containing 2 mM CaCl ₂ . The solution was incubated on ice for 30 minutes and dispensed into two 1.5 ml reaction vessels. After centrifugation (5 minutes, 18500 g, 4 ° C.) to remove undissolved components, the supernatant was transferred to a new reaction vessel. Phage were reprecipitated by adding 1/5 volume of 20% w / v PEG 8000, 15% w / v NaCl and incubation on ice for 60 minutes. After centrifugation (30 minutes, 18500 g, 4 ° C.) and removal of the supernatant, the precipitated phage particles were dissolved in a total of 1 ml of TBS containing 2 mM CaCl ₂ . After 30 minutes incubation on ice, the solution was centrifuged as described above. The supernatant containing the phage particles was used directly for affinity enrichment.

Phage affinity enrichment was performed using 96-well plates (Maxisorp, NUNC, Denmark). Single wells were coated for 12 hours at RT by incubation with 150 μl TBS solution of 100 μg / ml human serum albumin (HSA, Sigma). The binding sites remaining after HSA incubation were saturated by incubation with 250 μl of 2% w / v bovine serum albumin (BSA) TBST (TBS with 0.1% v / v Tween 20) for 2 hours at RT. Thereafter, 40 μl of a phage solution containing approximately 5 × 10 ¹¹ phage particles was mixed with 80 μl of TBST containing 3% BSA and 2 mM CaCl ₂ at RT for 1 hour. To remove unbound phage particles, the wells were washed 5 times for 1 minute with 130 μl TBST containing 2 mM CaCl ₂ .

  Phage bound to the well surface were eluted in a competitive manner by incubation with 130 μl of 0.1 M glycine / HCl pH 2.2 for 15 minutes or by adding 130 μl of 500 μg / ml HSA TBS solution. In the former case, the pH of the elution fraction was immediately neutralized after removal from the well by mixing the eluate with 30 μl of 1 M Tris / HCl pH 8.0.

For phage amplification, the eluate was used to infect E. coli K91BluKan cells (F ⁺ ). 50 μl of the eluted phage solution was mixed with 50 μl of cell preparation and incubated at RT for 10 minutes. Thereafter, 20 ml of LB medium containing 20 μg / ml tetracycline was added, and the infected cells were grown at 22 ° C. and 250 rpm for 36 hours. Cells were then sedimented (10 minutes, 5000 g, 4 ° C.). Phages were recovered from the supernatant by precipitation as described above. For the repeated affinity enrichment of phage particles, the same procedure as described in this example was used. After two subsequent rounds of panning against HSA, random colonies were picked and tested for their binding properties to the target protein used.

  Although this example demonstrates the use of the LDL receptor A domain, those skilled in the art will appreciate that the same techniques can be used to create the desired binding properties in the monomer domains of the present invention.

Example 3
This example describes the determination of the biological activity of a monomer domain that can bind to HSA.

  In order to demonstrate the ability of the HSA binding domain to prolong the serum half-life of proteins in vivo, the following experimental setup was performed. A multimeric A domain consisting of an A domain evolved for HSA binding (see Example 2) and a streptavidin binding A domain was compared with the streptavidin binding A domain itself. The proteins were injected into mice with or without human serum albumin (HSA) (as a control). Blood levels of A domain protein were monitored.

Therefore, the evolved A domain for HSA binding (see Example 1) was fused at the gene level with streptavidin-binding A domain multimers by standard molecular biology methods (see Maniatis et al., thing). The resulting gene constructs encoding A domain multimers and hexahistidine tags and HA tags were used to produce proteins in E. coli. After refolding and purification via affinity tag, the protein was dialyzed several times against 150 mM NaCl, 5 mM Tris pH 8.0, 100 μM CaCl ₂ and sterilized by filtration (0.45 μM).

  Two sets of animal experiments were performed. In the first set, 1 ml of each protein solution prepared at a concentration of 2.5 μM was injected into the tail vein of individual mice, and serum samples were collected at 2, 5 and 10 minutes after injection. In the second set, 50 mg / ml human serum albumin was added to the protein solution. As above, 1 ml of each solution was injected per animal. In the case of injected streptavidin-binding A domain dimers, serum samples are taken at 2, 5 and 10 minutes after injection, whereas in the case of trimers, serum samples are collected at 10, 30, and Collected after 120 minutes. All experiments were performed in duplicate and individual animals were assayed at each time point.

An enzyme linked immunosorbent assay (ELISA) was performed to detect blood levels of A domain in serum samples. Therefore, wells of Maxisorp 96 well microtiter plates (NUNC, Denmark) were coated with 1 μg of each anti-His ₆ antibody in TBS with 2 mM CaCl ₂ for 1 hour at 4 ° C. After remaining binding sites were blocked with casein (Sigma) solution for 1 hour, the wells were washed 3 times with TBS containing 0.1% Tween and 2 mM CaCl ₂ . Serial dilutions of serum samples were prepared and incubated in wells for 2 hours to capture A domain protein. After washing as described above, anti-HA tag antibody (Roche Diagnostics, 25 μg / ml) conjugated with horseradish peroxidase (HRP) was added and incubated for 2 hours. After washing as described above, HRP substrate (Pierce) was added, and the color of the detection reaction was developed according to the manufacturer's instructions. Light absorption reflecting the amount of A domain protein present in the serum sample was measured at a wavelength of 450 nm. The values obtained were normalized and plotted against a time scale.

  Evaluation of the values obtained showed a serum half-life of streptavidin-binding A domain of about 4 minutes in the absence of HSA and 5.2 minutes each when the animals were loaded with HSA. A domain trimer, including the HSA-binding A domain, showed a serum half-life of 6.3 minutes in the absence of HSA, but was significantly increased by 38 minutes when HSA was present in animals A half-life was indicated. This clearly suggests that the HSA binding A domain can be used as a fusion partner to increase the serum half-life of any protein, including protein therapeutics.

Example 4
This example describes an experiment that demonstrates an extension of the half-life of a protein in the blood.

  Individual monomers selected for monkey serum albumin, human serum albumin, human IgG, and human erythrocytes to further demonstrate the ability to extend the blood half-life of proteins using the monomer domains of the present invention Domain proteins were added to aliquots of heparinized human or monkey whole blood.

The following list provides the sequence of the monomer domains analyzed in this example.

  A aliquot of blood containing monomeric protein was then added to individual dialysis bags (25,000 MWCO), sealed, and stirred in 4 L of Tris-buffered saline overnight at room temperature.

  Anti-6xHis antibody was immobilized by hydrophobic interaction with 96 well plate (Nunc). Serial dilutions of serum from each blood sample were incubated with this immobilized antibody for 3 hours. The plate was washed to remove unbound protein and probed with α-HA-HRP to detect the monomer.

  Monomers identified as having a long half-life in dialysis experiments were constructed to contain either HA, FLAG, E-Tag, or myc epitope tags. Four monomers containing one protein for each tag were pooled to create two pools.

  One monkey was injected subcutaneously per pool at a dose of 0.25 mg / kg / monomer in a total volume of 2.5 mL in saline. Blood samples were drawn at 24, 48, 96, and 120 hours. Anti-6xHis antibody was immobilized by hydrophobic interaction with 96 well plate (Nunc). Serial dilutions of serum from each blood sample were incubated with this immobilized antibody for 3 hours. The plate was washed to remove unbound protein and probed separately with α-HA-HRP, α-FLAG-HRP, α-ETag-HRP, and α-myc-HRP to detect the monomer.

The following illustrates a comparison between commercially available antibodies and anti-IgG multimers:

Example 5
This example describes the development of protein-specific monomer domains and dimers by “walking”.

  A library of DNA sequences encoding monomer domains is generated by assembly PCR as described in Stemmer et al., Gene 164: 49-53 (1995).

The PCR fragment was digested with appropriate restriction enzymes (eg, XmaI and SfiI). The digestion product is separated on a 3% agarose gel and the domain fragment is purified from the gel. The DNA fragment was ligated to the corresponding restriction sites of the phage display vector fuse5-HA, a derivative of fuse5 carrying the in-frame HA-epitope. The ligation mixture is electroporated into TransforMax ™ EC100 ™ electrocompetent E. coli cells. Transformed E. coli cells are grown overnight at 37 ° C. in 2 × YT medium containing 20 μg / ml tetracycline and 2 mM CaCl ₂ .

The phage particles are purified from the medium by PEG precipitation. Individual wells of a 96 well microtiter plate (Maxisorp) are coated with 0.1 M NaHCO ₃ solution of target protein (1 μg / well). After blocking the wells with TBS buffer containing 10 mg / ml casein, purified phage is added in a typical number of about 1-3 × 10 ¹¹ . The microtiter plate is incubated for 4 hours at 4 ° C., washed 5 times with wash buffer (TBS / Tween) and the bound particles are eluted by addition of glycine-HCl buffer pH 2.2. Neutralize the eluate by adding 1 M Tris-HCl (pH 9.1). The phage eluate is amplified by E. coli K91BlueKan cells and, after purification, used as input for the second and third round of affinity selection (repeat above steps).

  The phage from the final eluate is used directly as a template for PCR amplification of the domain encoding the DNA sequence without purification.

  The PCR product is purified and then digested with appropriate restriction enzymes (eg, 50% with BpmI and 50% with BsrDI).

  The DNA fragments are used to “walk” the digested monomer fragments into dimers by attaching a library of unprocessed domain fragments. The raw domain sequence is obtained by PCR amplification of an initial domain library (obtained from PEG purification above) using primers suitable for amplifying the domain. The PCR fragment is purified, halved and then digested with an appropriate restriction enzyme (eg, either BpmI or BsrDI).

  Digested products were separated on a 2% agarose gel and domain fragments were purified from the gel. Purified fragments are combined in two separate pools (eg, untreated / BpmI + selected / BsrDI & untreated / BsrDI + selected / BpmI) and ligated overnight at 16 ° C.

The dimer domain fragment is PCR amplified (5 cycles), digested with appropriate restriction enzymes (eg, XmaI and SfiI) and purified from a 2% agarose gel. The screening step is repeated as described above, except for further stringent washing to obtain a high affinity binder. After infection, K91BlueKan cells are plated on 2 × YT agar plates containing 40 μg / ml tetracycline and grown overnight. Single colonies are picked and grown overnight in 2 × YT medium containing 20 μg / ml tetracycline and 2 mM CaCl ₂ . Phage particles are purified from these cultures.

  The binding of individual phage clones to their target protein was analyzed by ELISA. The clone that gave the greatest ELISA signal was sequenced and then recloned into a protein expression vector.

  IPTG induces protein production with an expression vector and the protein is purified by metal chelate affinity chromatography. Protein specific monomers are characterized as follows.

Biacore
250 RU protein is immobilized on a CM5 chip (Biacore) by NHS / EDC coupling. Run 0.5 and 5 μM solutions of monomeric protein over one side of the derivatized chip and analyze the data using a standard Biacore software package.

ELISA
10 nanograms of protein per well is immobilized by hydrophobic interaction with a 96 well plate (Nunc). The plate was blocked with 5 mg / mL casein. A serial dilution of monomeric protein was added to each well and incubated for 3 hours. The plate was washed to remove unbound protein and probed with α-HA-HRP to detect the monomer.

Functional assays Functional assays that determine the biological activity of a monomer can be performed in the same way, such as assays that determine the binding specificity of a monomer, and the metabolic pathway by binding a monomer to its target molecule. Assays that determine whether to antagonize or stimulate can be included.

Example 6
This example describes in vivo intraprotein recombination to create a library of even greater diversity.

  An orthologous loxP site flanked by a monomer-encoding plasmid vector (a vector derived from pCK; see below) was recombined Cre-dependently with the phage vector via its compatible loxP site. Recombinant phage vectors were detected by PCR using primers specific for the recombinant construct. DNA sequencing suggested that the correct recombinant product was produced.

Reagents and experimental procedures
pCK-cre-lox-Mb-loxP
This vector has two particularly relevant characteristics. First, it carries the cre gene encoding the site-specific DNA recombinase Cre under the control of P _lac . Cre was PCR amplified from p705-cre (from GeneBridges) using a cre specific primer incorporating XbaI (5 ′) and SfiI (3 ′) at the end of the PCR product. This product was digested with XbaI and SfiI and cloned into the same site of pCK110919-HC-Bla (pACYCori) bla ⁻ and Cm ^R derivative pCK to obtain pCK-cre.

  The second feature is the untreated A-domain live flanked by two orthologous loxP sites, loxP (wild type) and loxP (FAS), which are required for site-specific DNA recombination catalyzed by Cre. Rally. See, for example, Siegel, R. W., et al., FEBS Letters 505: 467-473 (2001). These sites rarely recombine with others. loxP sites were sequentially incorporated into pCK-cre. Both 5 'phosphorylated oligonucleotides loxP (K) and loxP (K_rc) hybridize with loxP (WT) and EcoRI and HinDIII compatible overhangs that allow ligation to digested EcoRI and HinDIII digested pCK Soybeans were ligated to pCK-cre with a standard ligation reaction (T4 ligase; overnight at 16 ° C.).

  The resulting plasmid was digested with EcoRI and SphI and ligated to hybridized 5 ′ phosphorylated oligos loxP (L) and loxP (L_rc) carrying loxP (FAS) and EcoRI and SphI compatible overhangs. To prepare for library construction, large-scale purification (Qiagen MAXI prep) of pCK-cre-lox-P (wt) -loxP (FAS) was performed according to the Qiagen protocol. The Qiagen purified plasmid was subjected to CsCl gradient centrifugation for further purification. This construct was then digested with SphI and BglII and ligated to the digested untreated A domain library insert obtained via PCR amplification of the existing A domain library pool. By design, the loxP site and Mb are in-frame, which creates an Mb with a loxP-encoded linker. This library was utilized in in vivo recombination procedures as detailed below.

fUSE5HA-Mb-lox-lox vector This vector is a derivative of fUSE5 from George Smith's laboratory (University of Missouri). This was then modified to carry a HA tag for immunodetection assays. The loxP sites were sequentially incorporated into fUSE5HA. The 5 'phosphorylated oligonucleotides loxP (I) and loxP (I) _rc, which carry loxP (WT), a stop codon chain and an XmaI and SfiI compatible overhang are hybridized together and a standard ligation reaction ( Ligated to XmaI and SfiI digested fUSE5HA with New England Biolabs T4 ligase (overnight at 16 ° C.).

  The resulting phage vector was then digested with XmaI and SphI and ligated to hybridized oligos loxP (J) and loxP (J) _rc carrying loxP (FAS) and overhangs compatible with XmaI and SphI. This construct was digested with XmaI / SfiI and then ligated to a pre-cut (XmaI / SfiI) untreated A domain library insert (PCR product). A stop codon is located between the loxP sites and prevents the expression of gIII and thus the production of infectious phage.

  The ligated vector / library was then transformed into an E. coli host with a gIII expression plasmid that allowed fUSE5HA-Mb-lox-lox phage rescue as detailed below.

pCK-gIII
This plasmid carries gIII under the control of its native promoter. This was constructed by PCR amplification of gIII and its promoter from VCSM13 helper phage (Stratagene) using primers gIII promoter_EcoRI and gIII promoter_HinDIII. This product was digested with EcoRI and HinDIII and cloned into the same site of pCK110919-HC-Bla. Since gIII is under the control of its own promoter, gIII expression is presumed to be constitutive. pCK-gIII was transformed into E. coli EC100 (Epicentre).

In Vivo Recombination Procedure In summary, this procedure includes the following important steps: a) Generation of an infectious agent of fUSE5HA-Mb-lox-lox library in an E. coli host expressing plasmid-derived gIII (ie, rescue) B) Cloning of the secondary library (pCK) and transformation into F ⁺ TG1 E. coli; c) Culture carrying the secondary library in the rescued fUSE5HA-Mb-lox-lox phage library Infection of things.

a. Rescue of phage vectors
Electrocompetent cells carrying pCK-gIII were prepared by standard protocols. These cells have a transformation rate of 4 × 10 ⁸ / μg DNA and are electroporated with a large ligation of the fUSE5HA-lox-lox vector and the untreated A domain library insert (approximately 5 μg of vector DNA) It was done. Following individual electroporation (DNA 100 ng / electroporation) with approximately 70 μL of cell fluid / cuvette, 930 μL of warm SOC medium was added and cells were allowed to recover by shaking at 37 ° C. for 1 hour. Next, tetracycline was added to a final concentration of 0.2 μg / mL and the cells were shaken at 37 ° C. for about 45 minutes. An aliquot of this culture was removed, serially diluted 10-fold, and placed in a culture dish to determine the resulting library size (1.8 × 10 ⁷ ). Dilute the remaining culture in 2 x 500 mL 2 x YT (with 20 μg / mL chloramphenicol and 20 μg / mL tetracycline to select vectors based on pCK-gIII and fUSE5HA, respectively) And grown overnight at 30 ° C.

Rescued phage were recovered using a standard PEG / NaCl precipitation protocol. The titer was approximately 1 × 10 ¹² / mL transduction unit.

b. Cloning of secondary library and transformation into E. coli host Electroporate the ligated pCK / untreated A domain library into a bacterial F ⁺ host. The result is expected to be approximately 10 ⁸ library size. After a 1 hour recovery time at 37 ° C. with shaking, the electroporated cells were diluted to 2 × YT (plus 20 μg / mL chloramphenicol) to an OD _{600 of} approximately 0.05 and fUSEHA-Mb-lox- Grow to mid-log growth at 37 ° C before infection with lox.

c. Infection of culture harboring secondary library with rescued fUSE5HA-Mb-lox-lox phage library High infection rate of E. coli during culture to maximize the production of recombinants (> 50%) is preferred. Infectivity of E. coli depends on several factors, including F pili expression and growth conditions. E. coli background TG1 (carrying F ′) and K91 (Hfr strain) were in the recombinant hosts.

Oligonucleotide:

Example 7
This example describes the optimization of multimers by optimizing monomers and / or linkers for target binding.

  FIG. 8 illustrates an approach for optimizing multimer binding to a target, exemplified by trimeric multimers. In the figure, initially a library of monomers is panned for binding to a target (eg, BAFF). However, some monomers can bind to locations on the target that are far from each other, so that domains that bind to these sites cannot be linked by a linker peptide. Therefore, it is useful to generate and screen large libraries of homo- or heterotrimers derived from these monomers prior to monomer optimization. These trimer libraries may be screened, for example, on phage (typically in the case of heterotrimers made from a large pool of monomers) or may be made and assayed separately. Good (eg in the case of homotrimers). This method identifies the best trimer. Assays can include determination of multimeric agonist or antagonist potency in target binding assays or functional protein or cell based assays.

  The best single trimer monomer domain is then optimized as a second step. Homomultimers are easiest to optimize because there is only one domain sequence, but heteromultimers may be synthesized. In the case of homomultimers, the increase in binding by the multimer compared to the monomer is a binding activity effect.

  After optimization of the domain sequence itself (eg, by recombination or NNK randomization) and phage panning, dimers are constructed with linker libraries using improved monomers. A linker library can be formed, for example, from linkers having NNK composition and / or variable sequence length.

  After panning this linker library, the best clones (e.g., determined by inhibition potency or other functional assay) are multiple (e.g. 2, 3, 4, 5, 6, 7). , 8 etc.) to a multimer consisting of a sequence optimization domain and a length and sequence optimization linker.

  To demonstrate this method, the multimer is optimized for binding to BAFF. The BAFF binding clone anti-BAFF2 binds to BAFF with approximately equal affinity as a trimer or as a monomer. The linker sequence separating the monomers in the trimer is 4 amino acids in length, which is unusually short. It has been proposed that linker length extension between monomers allows multiple binding contacts for each monomer within the trimer and can greatly increase the affinity of the trimer compared to the monomer molecule. .

  To test this, a library of linker sequences is created between the two monomers to create a potentially higher affinity dimer molecule. The identified optimal linker motif is then used to create a potentially higher affinity trimer BAFF binding molecule.

These libraries consist of random codons NNK that vary in length from 4 to 18 amino acids. The linker oligonucleotides for these libraries are:

を用いる。PCR産物をQiagen Qiaquickカラムで精製し、次にBsrDIで消化する。親の抗BAFF2クローンをBpmIで消化する。これらの消化物をQiagen Qiaquickカラムで精製し、ともにライゲーションする。このライゲーション物をSfiIプライマーおよびプライマーBpmI

を用いた10サイクルのPCRによって増幅させる。Qiagen Qiaquickカラムを用いた精製の後、DNAをXmaIおよびSfiIで消化する。消化産物を3%アガロースゲルで分離し、二量体BAFFドメイン断片をゲルから精製する。DNA断片をファージディスプレイベクターfuse5-HA、つまりインフレームのHA-エピトープを保有するfuse5の派生体の対応制限部位にライゲーションする。ライゲーション混合物をTransforMax(商標) EC100(商標)エレクトロコンピテント大腸菌細胞にエレクトロポレーションする。形質転換された大腸菌細胞を20μg/mlのテトラサイクリンの入った2×YT培地中37℃で終夜増殖させる。ファージ粒子をPEG沈殿により培地から精製し、パニングに使用する。 A library of these sequences is generated by PCR. General purpose primer SfiI with linker oligonucleotides in PCR using clone anti-BAFF2 as template

Is used. The PCR product is purified on a Qiagen Qiaquick column and then digested with BsrDI. The parental anti-BAFF2 clone is digested with BpmI. These digests are purified on a Qiagen Qiaquick column and ligated together. This ligation product is converted to SfiI primer and primer BpmI.

Amplify by 10 cycles of PCR using After purification using a Qiagen Qiaquick column, the DNA is digested with XmaI and SfiI. The digestion products are separated on a 3% agarose gel and the dimeric BAFF domain fragment is purified from the gel. The DNA fragment is ligated to the corresponding restriction sites of the phage display vector fuse5-HA, a derivative of fuse5 carrying the in-frame HA-epitope. The ligation mixture is electroporated into TransforMax ™ EC100 ™ electrocompetent E. coli cells. Transformed E. coli cells are grown overnight at 37 ° C. in 2 × YT medium containing 20 μg / ml tetracycline. The phage particles are purified from the medium by PEG precipitation and used for panning.

Example 8
This example describes intradomain recombination to identify monomer domains with improved function.

Monomer sequences were generated by several steps of panning and one step of recombination to identify monomers that bind to either CD40 ligand or human serum albumin. CD40L and HSA were panned against three different A domain phage libraries. After two rounds of panning, the eluted phage pool was PCR amplified with two sets of oligonucleotides to produce two overlapping fragments. The two fragments were then fused together and cloned into the phagemid vector pID to fuse the two fragment recombination products. The recombinant library (10 ¹⁰ size each) was then panned twice against CD40L and HSA targets by panning in solution and streptavidin magnetic bead capture.

  The selected phagemid pool was then recloned into protein expression vector pET, a T7 polymerase driven vector, for high protein expression. Approximately 1400 clones were screened for anti-CD40L binding monomer by standard ELISA and approximately 2000 clones were screened for HSA. All clones were unique sequences.

  ELISA plate wells were coated with 0.2 μg CD40L or 0.5 μg HAS, and 5 μl of monomer expressing clone lysate was applied to each well. The bound monomer (which was produced as a hemagglutinin (HA) fusion) was then detected with an anti-HA-HRP-conjugated antibody, colored by horseradish peroxidase enzyme activity, and measured at an OD of 450 nm. Positive clones were selected by comparing ELISA measurements with existing trimeric anti-CD40L 2.2 and sequenced using T7 primers.

  In the case of anti-CD40L samples, two anti-CD40L 2.2Ig clones were grown on the same plate as the selected monomer clone and processed in parallel as a positive control. Two transformed empty pET vector clones were grown and treated as negative controls. The ELISA measurements at OD450 and the corresponding clone sequences are shown.

  The same selection and screening process applies to HSA. Existing anti-HSA monomers and trimers were used as positive controls and empty pET vectors were used as negative controls. Positive binders were selected as having an ELISA signal equal to or better than anti-HSA trimers.

The positive rate of clones with OD ₄₅₀ greater than or equal to anti-CD40L2.2Ig binding was approximately 0.7% for CD40L and 0.4% for HSA.

The identified sequences are listed below.

  Although this example demonstrates the use of the LDL receptor A domain, those skilled in the art will appreciate that the same techniques can be used to create the desired binding properties with the monomer domains of the present invention.

Example 9
This example describes an exemplary method for the design and analysis of libraries containing monomers that contain only residues found in the natural domain at a given sequence position. For this purpose, a sequence alignment of all natural domains of a given family is constructed. Since cysteine residues are likely to be the most conserved feature of the alignment, these residues are used as a guide for further design. Consider each stretch of sequence between two cysteines separately to account for structural variability due to changes in length. A length histogram is constructed for the sequence between each cysteine. Lengths found in the known domains with a frequency of approximately 10% or more are considered for use in the library design. Separate sequence alignments are constructed for each length, and amino acids that appear at a given position in the partial alignment at frequencies greater than approximately 5% are allowed in the final library design of that length. This process is repeated for the sequence segment between each cysteine to create the final library design. Oligonucleotides with degenerate codons designed to optimally express the desired protein diversity are then synthesized and assembled using standard methods to create the final library.

  Typically, 4 sets of overlapping oligonucleotides are designed for PCR assembly with 9 base overlap between set 1 and 2, set 2 and 3, and set 3 and 4. In some cases, two sets of overlapping oligonucleotides are designed with a 9 base overlap between the two sets. Build the library with the following protocol.

Oligonucleotides Prepare a 10 μM diluted standard solution of each oligonucleotide. Mix equimolar amounts of oligos for each set (sets 1, 2, 3 and 4). Assemble the oligonucleotides in two PCR assembly steps: Assemble sets 1 and 2 and set 3 and 4 in the first round of PCR, and use the first round PCR product in the second round of PCR. Assemble the full length of the rally.

PCR assembly-Round 1
Separate PCR reactions are performed using the following pairs of oligos: Set 1 vs. each Pool Set 2 oligo; Set 2 vs. each Pool Set 1 oligo; Set 3 vs. each Pool Set 4 oligo; Set 4 vs. Pool Set 3 each oligo. The PCR reaction mixture is 50 μL in volume and contains 5 μL of 10 × PCR buffer, 8 μL of 2.5 mM dNTPs, 5 μL of each oligo and its paired oligo pool, 0.5 μL of LA Taq polymerase and 26.5 μL of water. PCR reaction conditions are as follows: [94 ° C / 10 seconds, 25 ° C / 30 seconds, 72 ° C / 30 seconds] 18 cycles and [94 ° C / 30 seconds, 25 ° C / 30 seconds, 72 ° C / 1 Min] 2 cycles. Run 5 μL of each PCR reaction on a 3% low melting point agarose gel in TBE buffer to verify the presence of the expected PCR product.

PCR assembly-Round 2
Pool all round 1 PCR products with 5 μL from each PCR reaction. Full length product of each library backbone using 4 μL of 10 × PCR buffer, 8 μL of 2.5 mM dNTPs, 10 μL of pooled round 1 PCR product, 50 μL reaction volume containing 0.5 μL LA Taq and 27.5 μL water and the following reaction conditions: Assemble by PCR: 8 cycles of [94 ° C / 10 sec, 25 ° C / 30 sec, 72 ° C / 30 sec] and 2 [94 ° C / 30 sec, 25 ° C / 30 sec, 72 ° C / 1 min] cycle.

Rescue PCR and Sfi digestion A fully assembled library backbone is amplified by PCR to create sufficient material for library production. Perform four separate 50 μL PCR reactions. Each reaction mixture contains 2.5 μL of 10 × PCR buffer, 8 μL of 2.5 mM dNTPs, 25 μL of round 2 PCR product, LA Taq 0.5 μL, 10 μM 5 ′ and 3 ′ rescue PCR primers (Table 2) 5 μL, and water 4 μL . The reaction conditions are as follows: [94 ° C / 10 sec, 25 ° C / 30 sec, 72 ° C / 30 sec] 8 cycles and [94 ° C / 30 sec, 45 ° C / 30 sec, 72 ° C / 1 min ] For 2 cycles. Run 5 μL of the reaction mixture on a 3% low melting point agarose gel in TBE buffer to confirm that the amplification product is the correct size. The amplified product is then purified by QIAGEN QIAquick column, eluted in EB buffer and digested with Sfi restriction enzyme for cloning into Sfi digested ARI 2 vector. Digest 20 μg of the assembled library backbone with 200 units of Sfi restriction enzyme in a total volume of 1,000 μL at 50 ° C. for 3 hours. The digested DNA is purified on a QIAGEN QIAquick column and eluted in water.

Test ligation To determine the optimal library insert / vector ratio for ligation, 1 μL of each dilution series (1/1, 1/5, 1/25, 1/125 and 1/625) of Sfi digested library inserts Used for ligation with 1 μL of digested ARI 2 vector, 1 μL of T4 DNA ligase, 1 μL of 10 × ligase buffer and 7 μL of water. The ligation reaction mixture is incubated at room temperature for 2 hours to produce a ligation product. 1 μL of ligation product is mixed with 40 μL of EC100 cells in a 0.1 cm cuvette, incubated on ice for 5 minutes, electroporated and allowed to recover for 1 hour at 37 ° C. in 1 mL of SOC. For each electroporation, spot 5 μL of each dilution series (1/1, 1/10, 1/100, 1 / 1,000) on agar plates with tetracycline to determine the optimal insert / vector ratio. In addition, 50 μL of each dilution is plated on a culture dish to grow a single colony for library QC.

Sequence analysis and protein expression Individual clones are picked and grown overnight in 0.4 mL of 2 × YT containing 20 μg / mL tetracycline in a 96-well plate. The cells grown overnight are spun down and the library insert is amplified with 5 ′ and 3 ′ rescue primers for sequencing using 0.5 μL of 1/5 dilution of the supernatant. DNA sequence analysis is used to verify the presence of the expected library insert. To examine protein expression, the library insert is transferred to the pEVE expression vector. Pool supernatants of selected clones from overnight cultures, 0.5 μL, are amplified by a pair of PCR primers with an N-terminal HA epitope and a C-terminal His8 tag and an Sfi restriction site in frame. The PCR reaction mixture contains: 0.5 μL of phage (32 supernatant pool), 5 μL of 10 × LA Taq buffer, 8 μL of 2.5 mM dNTPs, 5 μL each of 10 μM EGF Eve 5 and 10 μM 3Sfi N primer, and LA Taq polymerase 0.5 μL. PCR reaction conditions are as follows: [94 ° C / 10 sec, 45 ° C / 30 sec, 72 ° C / 30 sec] 23 cycles and [94 ° C / sec, 45 ° C / 30 sec, 72 ° C / 1 min ] For 2 cycles. The amplification product is purified by QIAquick column, digested with Sfi enzyme and ligated for 2 hours at room temperature according to Sfi digested pEVE vector and manufacturer's specifications. 1 μL of the ligation product is transformed into 40 μL of BL21 cells by electroporation, plated on kanamycin plates and grown overnight in a 37 ° C. incubator. Pick colonies and incubate overnight in 0.5 mL of 2X YT medium. The next day, 50 μL of overnight culture is inoculated into 1 mL of 2 × YT medium and allowed to grow for approximately 2.5 hours until OD600 reaches approximately 0.8, at which point IPTG is added to a final concentration of 1 mM for protein expression. Cells are spun down at 3,600 rpm for 15 minutes, the pellet is suspended in 100 μL of TBS / 2 mM Ca ⁺⁺ , heated at 65 ° C. for 5 minutes to release the protein, and spun down at 3,600 rpm for 15 minutes. The supernatant from each clone is run with or without 4-12% NuPAGE gel, 10 μL each of reducing agent (Invitrogen). Movement of the band position between the reduced and non-reduced samples suggests that the expressed protein is likely to be properly folded.

Library scale-up The complete library is ligated to ARI 2, transformed into EC100 cells, and then amplified in K91 cells. Ligation is performed overnight at room temperature in a final volume of 2.5 mL with 25 μg of Sfi digested vector, 2.5 μg of Sfi digested library insert, 5 μL of T4 DNA ligase, and 250 μL of 10 × DNA ligase buffer. The ligation product is precipitated with sodium acetate and ethanol, suspended in 400 μL of water, reprecipitated with NaAc / EtOH, and resuspended in 50 μL of H 2 O. The library is electroporated in a container containing 10 μL of DNA and 200 μL of EC100 cells, transferred to 50 mL of SOC medium, and grown at 37 ° C. for 1 hour at 300 rpm. Remove 5 μL aliquots, (1) serially dilute to determine library size, and (2) plate out for sequence verification. Aliquot the transformed EC100 in 50 mL SOC and add to 6 500 mL cultures of K91 cells with an OD600 of 0.5 and incubate for 30 minutes at 37 ° C. without shaking. Tetracycline is added to a concentration of 0.2 μg / mL and the culture is grown for 30 minutes at 37 ° C. at 300 rpm. Finally, tetracycline is added to a final concentration of 20 μg / mL and the culture is grown overnight at 37 ° C. at 300 rpm. Centrifuge the cells at 8,000 rpm for 10 minutes. Precipitate the phage in the supernatant by addition of 40 g PEG and 30 g / 1000 mL NaCl and centrifugation at 8,000 rpm for 10 minutes. Resuspend the phage in 50 mL of TBS / 2 mM Ca ^{++ and} centrifuge at 5,000 rpm for 10 minutes to remove cell debris. Add 20% final concentration of PEG and 1.5 M NaCl to the supernatant, place it on ice for 40 minutes, centrifuge the phage at 5,000 rpm for 10 minutes, and resuspend in 10 mL of TBS / 2 mM Ca ⁺⁺ . The phage titer is determined by serial dilution.

Example 10
This example describes the design and analysis of a trefoil / PD domain derived library utilizing the method described in Example 9 above.

Based on the sequence alignment of the natural trefoil / PD domain, a panel of degenerate oligonucleotides encoding the trefoil / PD domain was designed that contained amino acids at each position found only in the natural trefoil / PD domain. The design of the trefoil / PD library is shown below.

Degenerate oligonucleotide sequences are shown in the table below.

  N represents A, T, G, or C; B represents G, C, or T; D represents G, A, or T; H represents A, T, or C; K represents G or M represents A or C; R represents A or G; S represents G or C; V represents G, A, or C; W represents A or T; and Y represents T or C is represented.

Individual 32 phage from each library were amplified by PCR and the amplification products were sequenced. Sequencing results confirmed that the phage contained inserts of the expected size and sequence for the library. The library contained 2.31 × 10 ⁹ monomer domains containing 57, 58, 61, or 62 amino acids. The results of sequencing are shown in the table below. Clones 5 and 6 were identified as clones containing no domain insert, but rather represent an empty vector background from the transformation.

  Clones from the trefoil / PD library were tested for their ability to produce folded proteins. SDS-PAGE verified that these clones produced full-length soluble protein after heat lysis.

Example 11
This example describes the design and analysis of a thrombospondin domain derived library using the method described in Example 9 above.

Based on the sequence alignment of the native thrombospondin domain, a panel of degenerate oligonucleotides encoding the thrombospondin domain containing amino acids at each position found only in the native thrombospondin domain was designed. The design of the thrombospondin library is shown below.

Degenerate oligonucleotide sequences are shown in the table below.

  N represents A, T, G, or C; B represents G, C, or T; D represents G, A, or T; H represents A, T, or C; K represents G or M represents A or C; R represents A or G; S represents G or C; V represents G, A, or C; W represents A or T; and Y represents T or C is represented.

Individual 32 phage from the library were amplified by PCR and the amplification products were sequenced. Sequencing results confirmed that the phage contained inserts of the expected size and sequence for the library. The library contained 1.98 × 10 ⁹ monomer domains containing 60-70 amino acids. The results of sequencing are shown in the table below. Clones 1, 4, 8, 11, 12, 22, 26, and 30 were identified as clones that did not contain a domain insert, but rather an empty vector background from the transformation.

  Clones from the thrombospondin library were tested for their ability to produce folded proteins. SDS-PAGE verified that these clones produced full-length soluble protein after heat lysis.

Example 12
This example describes an exemplary method for generating a library of proteins with randomized cysteine loops. In this example, in contrast to the separate loop, separate library approach described above, loops between multiple cysteines are simultaneously randomized within the same library.

ここで、
C1〜C6：システイン；
X(n)：各位置に任意の残基を有するnアミノ酸の配列；
E1〜E3：グルタミン；
F：フェニルアラニン；
R1〜R2：アルギニン；
A：アラニン；
G1〜G4：グリシン；
I：イソロイシン；
P：プロリン；
S1〜S3：セリン；
W：トリプトファン；
V：バリン；
D1〜D5：アスパラギン酸；ならびに
C1-C3、C2-C5およびC4-C6はジスルフィドを形成する。 A domain NNK library encoding a protein domain of 39-45 amino acids with the following pattern was constructed.

here,
C1-C6: cysteine;
X (n): a sequence of n amino acids with any residue at each position;
E1-E3: Glutamine;
F: phenylalanine;
R1-R2: Arginine;
A: Alanine;
G1-G4: glycine;
I: isoleucine;
P: proline;
S1-S3: Serine;
W: Tryptophan;
V: Valine;
D1-D5: aspartic acid; and
C1-C3, C2-C5 and C4-C6 form disulfides.

ここでR=A/G、Y=C/T、M=A/C、K=G/T、S=C/G、W=A/T、B=C/G/T、D=A/G/T、H=A/C/T、V=A/C/G、およびN=A/C/G/T。 Libraries are generated by assembly PCR as described in Stemmer et al., Gene 164: 49-53 (1995) to create a library of DNA sequences containing tyrosine codons (TAT) or variable non-conserved codons (NNK). Built by Compared to the design used to build the natural A-domain backbone and library A1 (described above), this approach 1) More than instead of randomizing these potentially important residues 2) all 20 amino acids (NNK codons) so that the average number of residues between cysteines extends beyond that of the natural A domain or A1 library. Insert a variable length amino acid chain. Tyrosine was found to occupy a large proportion at the antibody binding site, probably due to the many different contacts that tyrosine can make, so the inclusion of tyrosine codons in the oligonucleotide increased the proportion of tyrosine residues. . The oligonucleotides used in this PCR reaction are:

Where R = A / G, Y = C / T, M = A / C, K = G / T, S = C / G, W = A / T, B = C / G / T, D = A / G / T, H = A / C / T, V = A / C / G, and N = A / C / G / T.

The library was constructed through an initial round of 10 cycles of PCR amplification using a mixture of 4 pools of oligonucleotides, each pool containing 400 pmol of DNA. Pool 1 contained oligonucleotides 1-9, pool 2 contained 10-17, pool 3 contained only 18 and pool 4 contained 19-27. A fully assembled library was obtained through 8 additional cycles of PCR using pools 1 and 4. The library fragment was digested with XmaI and SfiI. The DNA fragment was ligated to the corresponding restriction sites of the phage display vector fuse5-HA, a derivative of fuse5 carrying the in-frame HA-epitope. The ligation mixture was electroporated into TransforMax ™ EC100 ™ electrocompetent E. coli cells, resulting in a library of 2 × 10 ⁹ individual clones. Transformed E. coli cells were grown overnight at 37 ° C. in 2 × YT medium containing 20 μg / ml tetracycline. Phage particles were purified from the medium by PEG precipitation and a titer of 1.1 × 10 ¹³ / ml was measured. The sequence of 24 clones was determined and was consistent with library design expectations.

  Although the foregoing invention has been described in some detail for purposes of clarity and understanding, various changes in form and detail may be made from reading this disclosure without departing from the true scope of the invention. It will be apparent to those skilled in the art. For example, all the techniques, methods, compositions, devices, and systems described above can be used in various combinations. All publications, patents, patent applications, or other documents cited in this application are incorporated by reference for each purpose as if each individual publication, patent, patent application, or other document was included. Are incorporated by reference in their entirety for all purposes to the same extent as are individually indicated.

Identify monomer domains that bind to ligands, isolate selected monomer domains, multimers of selected monomer domains by linking selected monomer domains in various combinations And schematically screen a multimer to identify a multimer comprising multiple monomers that bind to a ligand. Fig. 4 is a schematic representation of another selection strategy (guide selection) Monomeric domains with appropriate binding properties are identified from a library of monomer domains. The identified monomer domains are then linked to monomer domains from another library of monomer domains to form a multimeric library. A multimeric library is screened to identify a pair of monomer domains that simultaneously bind to the target. This process can then be repeated until optimal binding properties are obtained with the multimer. Illustrates walking selection to create multimers that bind targets with higher affinity. Figure 4 illustrates the screening of a library of monomer domains for multiple ligands displayed on cells. Illustrates embodiments of monomer domains and multimers for increased binding activity. While this figure illustrates specific gene products and binding affinities, it should be understood that these are only examples and that other binding targets can be used in the same or similar conformation. Illustrates embodiments of monomer domains and multimers for increased binding activity. While this figure illustrates specific gene products and binding affinities, it should be understood that these are only examples and that other binding targets can be used in the same or similar conformation. Figure 3 illustrates the higher order structure of various possible antibody-monomers or multimers of the present invention. In some embodiments, the monomer or multimer replaces the Fab fragment of the antibody. Illustrates the method of monomer intradomain optimization. Illustrates the sequence of possible multimer optimization steps, where the optimal monomer followed by the multimer is selected, followed by monomer optimization, linker optimization and then multimer optimization Is done. Illustrates four exemplary methods of recombining monomeric and / or multimeric libraries to introduce new mutations. FIG. 10A illustrates one exemplary embodiment of monomeric intradomain recombination in which portions of different monomers are recombined to yield a new monomer. FIG. 10B illustrates a second embodiment of intradomain recombination in which the portion of the recombined monomer as described in FIG. 10A is further recombined, resulting in newer monomers. FIG. 10C illustrates one embodiment of interdomain recombination in which different recombined monomers are linked together, ie, produce a multimer. FIG. 10D shows a new linked conjugate that binds to the same target molecule, i.e. a multimer is linked to another recombinant monomer that recognizes a different target molecule and binds to different target molecules simultaneously. 1 illustrates one embodiment of intermodule recombination that yields a multimer. 2 depicts a possible conformation of a multimer of the invention comprising at least one monomer domain that binds to a half-life extending molecule and other monomer domains that bind to two other different molecules. In this figure, two monomer domains bind to the first target molecule and another monomer domain binds to the second target molecule.

Claims (28)

A method for identifying a monomer domain that binds to a target molecule comprising the following steps:
a) providing a library of non-natural monomer domains selected from the group consisting of thrombospondin monomer domains, trefoil monomer domains, and thyroglobulin monomer domains;
Where the thrombospondin monomer domain includes the following sequences:

The trefoil monomer domain includes the following sequences:

And the thyroglobulin monomer domain includes the following sequences:

Wherein “x” is any amino acid;
b) screening a library of monomer domains for affinity for the first target molecule; and
c) identifying at least one monomer domain that binds to at least one target molecule.   2. The method of claim 1, wherein the at least one monomer domain specifically binds to a target molecule that is not bound by a natural monomer domain that is at least 90% identical to a non-natural monomer domain. Thrombospondin monomer domains C ₁ -C ₅ , C ₂ -C ₆ and C ₃ -C ₄ form disulfide bonds; and thyroglobulin monomer domains C ₁ -C ₂ , C ₃ -C The method of claim 1, wherein ₄ and C ₅ -C ₆ form a disulfide bond. The thrombospondin monomer domain contains the following sequence:

Wherein C ₁ -C ₅ , C ₂ -C ₆ and C ₃ -C ₄ form a disulfide bond;
The trefoil monomer domain includes the following sequence:

The thyroglobulin monomer domain contains the following sequence:

Wherein C ₁ -C ₂ , C ₃ -C ₄ and C ₅ -C ₆ form a disulfide bond; and wherein α is selected from the group consisting of w, y, f, and l; φ is d, 2. The method of claim 1, wherein the method is selected from the group consisting of e and n; and "x" is selected from any amino acid. The thrombospondin monomer contains the following sequence:

The trefoil monomer domain includes the following sequence:

And the thyroglobulin monomer comprises the following sequence:

The method of claim 1. The method of claim 1, further comprising the following steps:
Linking the identified monomer domains to a second monomer domain to form a multimeric library, each containing at least two monomer domains;
Screening a library of multimers for the ability to bind to a first target molecule; and identifying multimers that bind to the first target molecule.   7. The method of claim 6, wherein each monomer domain of the selected multimer binds to the same target molecule.   7. The method of claim 6, wherein the selected multimer comprises three monomer domains.   7. The method of claim 6, wherein the selected multimer comprises 4 monomer domains.   2. The method of claim 1, further comprising mutating at least one monomer domain thereby providing a library comprising the mutated monomer domains.   11. The method of claim 10, wherein mutating comprises recombining a plurality of polynucleotide fragments of at least one polynucleotide encoding a polypeptide domain. The method of claim 1 further comprising the following steps:
Screening a library of monomer domains for affinity for a second target molecule;
Identifying a monomer domain that binds to a second target molecule;
Linking at least one monomer domain having affinity for the first target molecule with at least one monomer domain having affinity for the second target molecule, thereby causing the first and second target molecules Forming a multimer having an affinity for.   2. The method of claim 1, wherein the library of monomer domains is expressed as phage display, ribosome display or cell surface display.   2. The method of claim 1, wherein the library of monomer domains is displayed on a microarray. The target molecule is not bound by a natural monomer domain that is at least 90% identical to the non-natural monomer domain;
The non-natural monomer domain is selected from the group consisting of a thrombospondin monomer domain, a trefoil monomer domain, and a thyroglobulin monomer domain;
A protein comprising a non-natural monomer domain that specifically binds to a target molecule.   16. A protein according to claim 15, wherein the monomer domain comprises at least one disulfide bond.   16. The protein of claim 15, wherein the monomer domain comprises at least 3 disulfide bonds.   16. The protein of claim 15, wherein the monomer domain is 30 to 100 amino acids long. The thrombospondin monomer domain contains the following sequence:

The trefoil monomer domain includes the following sequence:

And the thyroglobulin monomer domain comprises the following sequence:

Where `` x '' is any amino acid,
16. The protein according to claim 15. Thrombospondin monomer domains C ₁ -C ₅ , C ₂ -C ₆ and C ₃ -C ₄ form disulfide bonds; and thyroglobulin monomer domains C ₁ -C ₂ , C ₃ -C ₄ and C ₅ -C ₆ forms a disulfide bond, claim 19 of the protein. The thrombospondin monomer domain contains the following sequence:

Wherein C ₁ -C ₅ , C ₂ -C ₆ and C ₃ -C ₄ form a disulfide bond;
The trefoil monomer domain includes the following sequence:

The thyroglobulin monomer domain contains the following sequence:

Wherein C ₁ -C ₂ , C ₃ -C ₄ and C ₅ -C ₆ form a disulfide bond; and wherein α is selected from the group consisting of w, y, f, and l; φ is d, selected from the group consisting of e, and n; and "x" is selected from any amino acid;
16. The protein according to claim 15. The thrombospondin monomer contains the following sequence:

The trefoil monomer domain includes the following sequence:

And the thyroglobulin monomer comprises the following sequence:

16. The protein according to claim 15.   An isolated polynucleotide encoding the protein of claim 15. The monomer domain is selected from the group consisting of a thrombospondin monomer domain, a trefoil monomer domain, and a thyroglobulin monomer domain;
Where the thrombospondin monomer domain comprises the following sequence:

The trefoil monomer domain includes the following sequence:

And the thyroglobulin monomer domain comprises the following sequence:

Where `` x '' is any amino acid,
A library of proteins containing non-natural monomer domains.   25. The library of claim 24, wherein each monomer domain of the multimer is a non-natural monomer domain.   25. The library of claim 24, wherein the library comprises a plurality of multimers comprising at least two monomer domains linked by a linker.   25. A library according to claim 24, wherein the library comprises at least 100 different proteins comprising different monomer domains.   A library of polynucleotides encoding the library of proteins of claim 24.

Images