HK1149280B - Identification of antigen- or ligand-specific binding proteins - Google Patents

Description

Identification of antigen-or ligand-specific binding proteins

Technical Field

The present invention discloses novel methods for the production, expression and screening of diverse collections of binding proteins in vertebrate cells in vitro, allowing the identification and isolation of ligand-reactive or antigen-reactive binding proteins. In particular, the invention relates to methods for the expression, isolation and identification of at least one nucleotide sequence encoding a binding protein, such as an antibody or fragment thereof specific for a desired antigen or ligand, of a retrovirus.

Background

Display technology plays an important role in the isolation of specific high affinity binding proteins for diagnostic and therapeutic applications in a variety of disorders and diseases. These techniques extend to a broad field of antibody engineering, synthetases, proteomics, and cell-free protein synthesis. Biomolecule display techniques, which allow the construction of large libraries of modularly encoded biomolecules, their display for property selection, and rapid characterization (decoding) of their structures, are particularly useful for large-scale assessment and analysis of the diversity of proteins. In vitro display technologies have recently become attractive due to the isolation of antibodies by phage display, ribosome display and microbial display, which have now become the mainstream platform for antibody and protein engineering. However, microbial expression and display systems suffer from a number of limitations, particularly for expressing large dimeric vertebrate proteins, such as antibodies. This is due to the general inability to express full-length antibodies in such expression systems (which requires the display of engineered antibody fragments), but also due to the lack of glycosylation, the lack of chaperones, the lack of subcellular compartments, and eukaryotic cell-specific protein trafficking, which individually and collectively lead to the illusion of protein folding (artemi) in mammalian proteins expressed by microorganisms. Recently, various in vitro display methods have also been developed using eukaryotic host cells, including yeast, plant, and mammalian cells. Expression systems in yeast and plant cells also suffer from a lack of glycosylation, and specific vertebrate and mammalian cell-specific chaperones, such that the same limitations on protein folding apply to expression of vertebrate proteins in such systems. Expression of large recombinant proteins (like antibodies), proper protein folding, and post-translational modifications can only be expected to occur with reasonable efficiency and quality in vertebrate expression systems, ideally expressing multiple proteins in systems closest to the relevant cells in terms of phylogeny.

Thus, proteins of therapeutic interest (like antibodies from rodents or humans) are ideally expressed in rodent or human cells, and it is not surprising that regulatory agencies only approve expression systems from such species for the production of full-length therapeutic antibodies of clinical grade. However, vertebrate and mammalian cell-based expression systems are laborious, require long time periods to establish stably producing cell lines and clones, and an efficient and controllable genetic modification of such cells is often not trivial and thus makes these systems less attractive for a variety of screening and display methods. For example, DNA transfection methods cannot be used to control the number of DNA constructs that are transiently or stably incorporated into transfected cells, which precludes clonal expression of a protein library and therefore a clean gene (clearene) for phenotypic screening. These alternative viral systems lack the proper control of clonal expression, stable maintenance of these genetic constructs, and/or suffer from the following facts: such systems often elicit a cytopathic effect (e.g., vaccinia virus expression) in the target cells such that protein clones cannot be displayed and/or are sequentially enriched for a particular phenotype, such as, for example, specific binding to an antigen.

It is therefore an object of the present invention to provide a method which significantly overcomes the prokaryotes and eukaryotes of the prior artAll the above mentioned limitations and disadvantages of the expression and selection system. The method according to the invention makes use of stable retroviral expression of binding proteins, such as, in particular, antibodies in mammalian cells, in particular in B-lymphocyte cell lines, such that stable and preferred clonal expression of antibody proteins is achieved when appropriate glycosylation, chaperone proteins and protein trafficking are present, ensuring appropriate protein folding and allowing efficient and repeated screening, if desired, of antibody clones that bind to an antigen. Since preferred embodiments of the method according to the invention are based on retroviral expression of antibodies or fragments thereof in precursor lymphocytes, the technology disclosed herein is referred to as 'reverse transcriptase Display' (collectively referred to as 'reverse transcriptase Display') (all references to 'reverse transcriptase Display')Reverse transcriptionViral precursor B lymphCell display)。

Summary of The Invention

The present invention relates generally to the provision of therapeutic or diagnostic antibodies or fragments thereof. In particular, it relates to the identification and selection of antigen-reactive antibodies with fully human amino acid sequences that are of interest for therapeutic applications. These embodiments of the invention relate to retroviral expression vectors capable of expressing diverse collections of binding proteins, including antibodies or fragments thereof, in vertebrate (preferably mammalian) cells, as well as various methods for the efficient isolation of ligand-reactive or antigen-reactive molecules. The present invention provides novel methods for generating diverse collections of binding proteins (e.g., antibodies or fragments thereof) by three alternative methods. First, chain shuffling of a diverse collection (library) of light or heavy chains by at least one heavy or light chain molecule, (chain shuffling process), or second, diversification of at least one combination of heavy and light chains of an antibody after in situ retroviral transduction into vertebrate cells by somatic mutation of expression constructs transduced by the retrovirus (somatic mutation process), or third, V (D) J recombination of expression constructs transduced by the retrovirus, which constructs contain coding regions for the variable binding domains of the antibody, is referred to as a "quaysia-germline" configuration, e.g. a recombination process that still divides into V, optionally D and J gene segments (V (D) J. The binding protein or antibody or fragment thereof is displayed on the surface of precursor lymphocytes.

In contrast to alternative, plasmid-based or non-integrating virus-based vertebrate expression systems known in the art, the present invention provides, inter alia, methods that allow the transduction of diverse collections of binding proteins (preferably antibodies) stably (and optionally clonally) expressed in vertebrate cells using retroviruses, which significantly facilitate the amplification, isolation, and cloning of genes encoding the binding proteins. As a representative but non-limiting example, the transduction of retroviruses by murine precursor lymphocytes incapable of expressing endogenous antibodies is disclosed, such that only heterologous recombinant antibodies in these host cells are expressed as membrane-bound antibodies. Furthermore, the present invention illustrates how cells expressing a ligand-reactive or antigen-reactive binding protein (e.g., an antibody or fragment thereof) can be isolated and optionally expanded in vitro to iteratively enrich for a population of antigen-reactive conjugate cells, from which genes encoding antigen-reactive or ligand-reactive binding proteins can then be cloned and sequenced by standard molecular biology procedures known in the art (fig. 1).

Although a preferred embodiment of the method according to the invention is directed to retroviral expression of binding proteins, preferably human full length antibodies, it can equally be used to express any fragment thereof (e.g. single chain F of an antibody)_vOr F_abFragments). Disclosed herein are scientific protocols for retroviral transduction that optionally allow for (i) delivery of a construct encoding a single binding protein to a single target cell,in order to ensure clonal expression of the binding protein in these host cells; (ii) shuffling at least one expression construct encoding a first polypeptide chain with at least one expression construct encoding a second polypeptide chain, thereby producing a functional multimeric binding protein (e.g., an antibody molecule); (iii) mutating somatic cells of at least one expression construct encoding at least one binding protein when the vertebrate cells are transduced in situ; and (iv) producing binding protein expression from the at least one expression construct by a mechanism of V (D) J recombination when the in situ retrovirus is transduced into a vertebrate cell.

To achieve somatic mutations in binding proteins encoding constructs in situ, retroviral expression vectors containing cis-regulatory gene elements that target somatic hypermutations (hypermutation) to the protein-encoding sequence, preferably via an activation-induced cytidine deaminase (AID) pathway (papavailiou & Schatz, 2002), or using other enzymes that target somatic mutations to the binding protein-encoding sequence, and their use are disclosed. In order to generate diverse collections of binding proteins, preferably antibodies or fragments thereof, by V (D) J recombination in situ, retroviral vector constructs comprising gene segments arranged in a "quasi-germline" configuration, variable (V), optionally diversity (D), and connecting regions (J), allowing the assembly of coding regions for immunoglobulins or immunoglobulin-like binding proteins by a process known as V (D) J recombination, recombination-activated gene (RAG) -mediated rearrangement of these gene segments, and their use are disclosed (Grawunderet al, 1998).

According to another aspect, the invention further demonstrates how retroviral transduced cells stably expressing diverse collections of recombinant binding proteins are subsequently labeled by binding to at least one ligand or antigen of interest, and how cells bound to the aforementioned ligand or antigen of interest are detected by a suitable secondary reagent (secondary reagent). Methods for specifically labeling ligand-reactive cells or antigen-reactive cells, and their enrichment or isolation, preferably by high-speed Fluorescence Activated Cell Sorting (FACS), are described herein. Due to the stable expression phenotype of the retrovirus transduced cells, it is illustrated how antigen reactive cells can optionally be isolated and expanded again by tissue culture (in tissue culture) so that an optional iterative cycle of antigen labeling, antigen directed enrichment, and amplification of ligand reactive or antigen reactive cells can be performed until subcloning of these cells is performed, allowing identification of the nucleotide coding region for the antigen reactive antibody by a variety of standard PCR cloning methods (figure 1).

The methods disclosed herein allow for the expression of diverse collections of antibody chains or fragments thereof from at least one vector construct that, when transferred and expressed in situ into cells of a vertebrate, may optionally be capable of producing diverse collections of binding proteins. Expression of the antibody chains in vertebrate cells is preferably mediated by retroviral transduction.

Thus, a first aspect of the invention relates to a method for isolating and identifying at least one nucleotide sequence encoding an antibody or fragment thereof specific for a desired antigen or ligand, comprising the steps of:

(a) transducing at least one retroviral expression construct encoding the antibody or fragment thereof into a vertebrate host cell;

(b) expressing the antibody or fragment thereof in the vertebrate host cell;

(c) enriching vertebrate host cells expressing the antibodies or fragments thereof based on their ability to bind to the desired antigen or ligand; and is

(d) Isolating and identifying said at least one nucleotide sequence encoding said antibody or fragment thereof from the retroviral transduced and enriched vertebrate host cells.

In addition to the above mentioned steps, step (d) may be preceded by a step of expanding the enriched vertebrate host cells by tissue culture. Furthermore, step (c) may be followed by a step of expanding the enriched vertebrate host cells by tissue culture, after which step (c) is repeated at least once before step (d) is performed.

To achieve clonal expression of at least one antibody, it is preferred to control such retroviral transduction such that the majority of the cells transduced by the retrovirus are genetically modified by only one recombinant retroviral construct per chain of antibodies integrated into the host cell genome. Thus, in one embodiment of the invention, retroviral transduction is performed at a multiplicity of infection (MOI) equal to or less than 0.1.

An antibody according to a method of the invention is preferably a full length antibody. A fragment of an antibody may be selected from the group consisting of: one heavy chain, one light chain, one V_HSingle domain, a V_LSingle domain, one scFv fragment, one Fab fragment, and one F (ab') 2 fragment. The antibody or one or more fragments thereof can have a naturally occurring amino acid sequence, an artificially engineered amino acid sequence, or a combination thereof.

Although the method of the invention is preferably used to isolate and identify at least one nucleotide sequence encoding an antibody chain, it will be clear to the person skilled in the art that the method of the invention may also be used to isolate and identify at least one nucleotide sequence encoding a cell surface receptor of any one monomer or multimer belonging to the Ig-superfamily and any functional fragment thereof, or a cell surface receptor of one monomer or multimer belonging to the TNF α -receptor superfamily or any fragment thereof.

And wherein the binding protein is a full length antibody selected from the group consisting of: a fully human antibody, a humanized antibody in which the CDR regions of one or more non-human antibodies have been grafted onto the framework of a human antibody, and a chimeric antibody in which the variable region domain from one vertebrate species is bound to the constant region domain of another vertebrate species, in which the constant region domain of the chimeric antibody is preferably derived from one or more human antibodies.

In one embodiment of the methods disclosed herein, the vertebrate host cells can be derived from a group of species including: cartilaginous fish, teleostean, amphibian, reptilian, avian, and mammalian species. This group of mammalian species may include pigs, sheep, cattle, horses and rodents. The group of rodents may further include mice, rats, rabbits, and guinea pigs. In a preferred embodiment of the invention, these vertebrate host cell species are mice (Mus musculus).

The vertebrate host cells used in a method of the invention may be derived from any vertebrate organ, but are preferably derived from cells of a lymphoid cell line. Preferred lymphocytes for use in the present invention are B cell lines because these cells express antibody-specific chaperone proteins and because the helper molecules (like Ig α and Ig β) required to mediate cell surface anchorage of antibodies are expressed in these cells. More preferably, these B cells are precursor B lymphocytes, as pre-B cells can be found that do not express any endogenous antibody chains. Indeed, these preferred lymphocytes as used in the present invention are incapable of expressing endogenous antibody polypeptides comprising a so-called surrogate light chain (surrogate light chain) component encoded by the genes λ -5, VpreB1, and VpreB 2. Thus, these preferred lymphocytes express accessory membrane proteins that promote membrane deposition of antibody molecules, such as B-cell specific Ig α and Ig β molecules, but they lack expression of any endogenous antibody polypeptides or surrogate light chain components. It should be noted, however, that it is possible to ectopically express Ig α and Ig β molecules by methods known in the art (e.g., stably transfected using expression vectors for these proteins). In a preferred embodiment of the invention, the antibody molecule is anchored to the cell membrane of lymphocytes via endogenously expressed Ig α and Ig β proteins which are naturally expressed in murine pre-B lymphocytes.

The methods disclosed herein include procedures that allow for the isolation of cells that exhibit desired binding characteristics to a ligand or antigen of interest, and the isolation of genes encoding a desired binding protein of interest. In contrast to alternative, plasmid-based or non-integrating virus-based vertebrate expression systems known in the art, the preferred methods of retroviral expression of an antibody in vertebrate cells disclosed herein allow for stable and preferably clonal expression of antibodies that significantly facilitate amplification, isolation, and cloning of antibody-encoding genes. The disclosed methods allow for the efficient production of diverse collections of binding proteins in vitro by either:

(i) shuffling at least one expression construct encoding at least one polypeptide chain of a multimeric binding protein, like for example a heavy chain of an antibody, with at least one expression construct encoding at least one matching polypeptide chain, like for example a light chain of an antibody, to produce a functional multimeric binding protein, like for example an antibody;

(ii) subjecting at least one expression vector encoding at least one binding protein to somatic mutation when transferred in situ into a vertebrate cell;

(iii) somatic recombination of gene segments encoding the variable region binding domains of immunoglobulins and immunoglobulin-like molecules comprised in at least one expression vector, optionally D (diversity region), and J (connecting region), when transferred in situ into vertebrate cells by a procedure known as V (D) J recombination; or

(iv) By any combination of operations (i), (ii), and (iii).

According to a preferred embodiment, the at least one nucleotide sequence is a plurality of nucleotide sequences comprising an antibody heavy chain sequence and a plurality of antibody light chain sequences, or in the alternative comprises an antibody light chain sequence and a plurality of antibody heavy chain sequences.

According to another preferred embodiment, the antibody or fragment thereof comprises a variable binding domain encoded by at least one retroviral expression construct capable of undergoing v (d) J recombination such that the coding sequence for a variable binding domain is produced when a retrovirus is transduced, or

In another preferred embodiment, step (b) of the above method is performed under mutagenic conditions, preferably via activation-induced expression of cytidine deaminase (AID), which is endogenously expressed or ectopically expressed, wherein ectopic expression of AID is performed under inducible conditions.

In one aspect of the above method, at least one retroviral expression construct encoding said antibody or fragment thereof comprises a combination of a cis-regulated promoter and enhancer elements that allows targeting of AID-mediated somatic mutations to a variable binding domain encoded by the expression construct, wherein the promoter and enhancer elements are selected from the group consisting of:

(a) an immunoglobulin heavy chain promoter, an intron enhancer (E.mu.H), and a 3' alpha enhancer element,

(b) immunoglobulin kappa light chain promoter, kappa intron enhancer (kappa iE), and 3 'kappa enhancer (3' kappa E) elements,

(c) immunoglobulin lambda light chain promoter, lambda 2-4 and lambda 3-1 enhancer elements, and

(d) any functional combination thereof.

Description of the drawings

FIG. 1: the figure illustrates the principle of "reverse transcription cell display" allowing the identification and isolation of binding proteins, such as antibodies, specific for a desired antigen or ligand. In a first step, at least one retroviral expression construct capable of producing expression of a diverse collection of binding proteins is stably transduced into suitable vertebrate host cells ("selector cells"). This is achieved by transfecting at least one retroviral vector encoding at least one binding protein into retroviral packaging cells, which may constitutively or transiently express retroviral proteins Gag, Pol and Env (step 1). Packaging cells transfected with the at least one retroviral binding protein construct will then produce recombinant retroviral particles containing the at least one retroviral expression construct within 24-72 hours after transfection. The resulting retroviral particles accumulate in the cell culture supernatant of these retroviral packaging cells and can be used to transduce suitable vertebrate host cells ("selection cells") (table 2), which in turn express the binding protein. In this preferred method, the binding proteins (e.g., antibodies or fragments thereof) are expressed on the cell surface of the "selected cells" and the cells are then labeled with a desired antigen or ligand (step 3). The antigen-bound cells or ligand-bound cells are then analyzed, preferably by Fluorescence Activated Cell Sorting (FACS), and cells displaying specific antigen binding are separated from the unbound cell population, preferably by preparative, high speed FACS (step 4). The antigen-reactive or ligand-reactive cells may optionally be expanded again by tissue culture, and the cycle of antigen-directed cell sorting and tissue culture expansion may be repeated until a detectable antigen-reactive or ligand-reactive cell population is obtained due to the stably expressing phenotype of the retrovirus transduced cells. This antigen-reactive or ligand-reactive cell population may be subjected to a final, preferably single cell sorting step, or may be used directly to clone the gene encoding the binding protein on a population basis. In the next step (step 5), the coding regions of the relevant binding regions are cloned from the antigen-or ligand-selected cell banks or cell clones by RT-PCR or genomic PCR using primer pairs that bind to sequences that are specific for this binding protein library and/or specific for other vector sequences by standard methods known in the art. The cloned and sequenced coding region for the binding protein may then be expressed, optionally as a recombinant protein, in any expression system selected for further functional characterization and thereby confirm the specificity of antigen binding or ligand binding (step 6).

FIG. 2: (a) the figure shows schematic structures of antibodies or immunoglobulins, and fragments thereof, which are preferred binding proteins according to the disclosed invention. FIG. 2a) shows the schematic structure (left side) of an IgG antibody, which is characterized by a characteristic Y-type structure and is composed of two identical immunoglobulin (Ig) heavy and light chains, which comprise 4 immunoglobulin domains (V) respectively_H-C_H1-C_H2-C_H3) And 2 immunoglobulin domains (V)_L-C_L). The V domain is a highly variable antigen-binding region of IgH and IgL chains, where C_HAnd C_LThe domains represent constant region domains. The variable region domain of IgH chains is encoded by V, D and J gene segments, while the variable region domain of IgL chains is encoded by V and J gene segments only, which require assembly from germline immunoglobulin loci (FIGS. 2B and 2c) during early B lymphoid tissue production by a process called V (D) J recombination. .

Antibodies IgH and IgL chains are covalently bound together by disulfide bridges at a position close to the flexible hinge region (e.g.at C)_H1 and C_H2 domains) are linked together, however at C_H1 and C_LAdditional disulfide bridges between domains (as depicted) covalently link IgH and IgL chains (fig. 2a left).

Fab fragments are V containing only one natural disulfide bridge_H-C_Hl/V_L-C_LMonovalent fragments of full-length antibodies of domains, which may be obtained from full-length antibodies by enzymatic papain cleavage, or which may be obtained by cleaving C_H2-C_HThe 3-deleted IgH chain is expressed together with the IgL chain to be expressed as a recombinant protein. Additional fully human antibody fragments are single chain variable domain fragments (scFv fragments) which comprise only the variable region domains of IgH and IgL chains, which are connected by a synthetic linker or an artificial disulfide bridge. Expression of full-length antibodies or antibody fragments (Fab and scFv fragments as described) can also be expressed as binding proteins in order to practice the invention.

FIG. 2b) schematically depicts the V (D) J recombination process occurring on one germline IgH chain allele, which leads to antibody V_HAssembly of coding regions for domains. The variable domains of IgH chains in vertebrate species are encoded by multiple V, D and J gene segments that are separated in germline configuration. During V (D) J recombination that occurs during early B lymphoid tissue production, a selected one of the V, D and J gene segments is site-specifically rearranged to generate V directed against an antibody_HUnique coding regions for the domains. V (D) J recombination in IgH chain loci is an ordered process and begins with the rearrangement of a selected D to a selected J gene segment, usually on both IgH chain alleles. Only after the rearrangement of the D to J genes, a selected V region is site-specifically ligated to the already assembled DJ region, thereby generating the gene encoding the V region_HOne V-D-J ORF of the domain. V (D) the J recombination process is dependent on the expression of the precursor lymphocyte-specific Recombinant Activation Genes (RAG)1 and 2.

FIG. 2c) schematically depicts the V (D) J recombination process occurring on one germline IgL chain allele, which leads to antibody V_LAssembly of coding regions for domains. The variable domains of IgL chains in vertebrate species are encoded solely by V and J gene segments, which are separated in germline configuration, similar to the gene segments in IgH chain loci. Antibody V_LDomain generation requires only one site-specific v (d) J rearrangement event, as depicted.

FIG. 3: the figure schematically illustrates the principle of stable genetic modification of target cells for expression of a binding protein of interest (BPOI), e.g., an antibody, alternatively labeled "X", by retroviral transduction.

Fig. 3a) depicts a schematic organization of a wild-type retroviral genome (top left), wherein the genes for the structural and functional proteins Gag, Pol and Env are located between the so-called 5 'and 3' Long Terminal Repeat (LTR) sequences flanking the retroviral genome. The 5' LTR is important for the expression of these retroviral genes, and for the replication of the retroviral genome in infected host cells. Another important region in the retroviral genome is the Psi (Psi) packaging signal required for packaging the retroviral RNA during replication and/or production of retroviral particles.

To produce recombinant retroviral particles, the gag, pol and env genes can be removed from a wild-type retroviral genome, leaving only the 5 'and 3' LTRs, and the Psi (Psi) packaging signal. To construct a recombinant retroviral vector, it is then convenient to introduce a Multiple Cloning Site (MCS) containing several unique and convenient restriction enzyme sites. This design (as depicted above/right) represents the simplest retroviral transfer vector.

For expression of recombinant retroviruses that allow expression of a recombinant protein (e.g., a binding protein of interest (BPOI) "X"), such as an antibody), it is minimally necessary to insert an Open Reading Frame (ORF) of a BPOI into an "empty" retroviral transfer vector because the 5' LTR region has a promoter activity capable of driving expression of any gene located downstream. However, to improve expression levels, expression of a gene of interest (e.g., a BPOI- "X") may optionally be driven by an additional heterologous promoter (Prom), and optionally the addition of a marker gene (e.g., the 5' LTR promoter and downstream of the ψ packaging signal) (as depicted herein) may allow selection and/or tracking of retroviral transduced constructs.

FIG. 3b) schematically shows the manipulation of retroviral transduction of target cells resulting in stable expression of BPOI- "X", e.g.antibodies. To this end, recombinant retroviral constructs containing expression cassettes for a BPOI- "X" are first transiently transfected into a retroviral Packaging Cell Line (PCL) expressing the structural and functional retroviral proteins Gag, Pol and Env of a wild-type retrovirus (left). Retroviral PCLs can be generated by stably or transiently transfecting expression constructs directed against Gag, Pol and Env proteins into an appropriate and easily transfected cell line (e.g., standard human 293HEK cells, or mouse NIH 3T3 fibroblasts). 2 to 3 days after transfection, recombinant retroviral genomes containing the BPOI- "X" gene were packaged into replication incompetent retroviral particles, which accumulated in the cell culture supernatant of PCL. These retroviral particles are replication incompetent because they lack genes for functional retroviral Gag, Pol and Env proteins and therefore they can only deliver their genetic payload once into a target cell, a process known as retroviral transduction or single round infection. In retroviral transduction, the packaging RNA of a recombinant retrovirus is introduced into the target cells where it is reverse transcribed into cDNA, which is then stably integrated into the target cell genome. 2-3 days after retroviral transduction, a gene of interest (e.g., BPOI- "X") is then permanently expressed by the target cells due to integration of the cDNA retroviral construct into the host cell genome.

FIG. 4: this figure shows a schematic design of a preferred type of retroviral expression construct that can be used to carry out the present invention. The figure depicts a schematic design of retroviral vectors contained in the backbone of a standard DNA cloning plasmid (closed black line); the relevant genes and regions for the retroviral genome are highlighted. A preferred vector generation depicted in Panel (panel) (a) (detailed clones thereof are depicted in FIGS. 5 and 6, and provided in example 1) comprises for human Ig γ₁The cDNA coding regions of H and Ig κ L chains, which are driven by a strong constitutive CMV promoter (Prom) and flanked upstream and downstream by Ig κ intron enhancer (κ iE) and 3 'κ enhancer (3' κ E) elements, facilitate somatic hypermutation of the V coding regions of IgH and IgL chains. These retroviral IgH and IgL chain expression constructs additionally contain a corresponding hygromycin B (hygro) marker for antibiotic resistance^R) And puromycin (puro)^R) Allows for the selection of stable integration of IgH and IgL chain constructs that selectively administer the corresponding antibiotic drug to retroviral transduced vertebrate cell cultures. ) In addition, convenient, unique restriction enzyme sites are highlighted, allowing the direct replacement of the V coding region with HindIII and Eco47III, or the replacement of the entire IgH and IgL chain coding regions by using the restriction enzymes HindIII and NotI. Thus, different V-regions and even entire sets of V-regions can be easily cloned into the disclosed expression vectors from an existing IgH or IgL chain expression construct.

(b) In this picture, another preferred class of vectors is described, which carry the replacement of the variable coding region by a DNA segment, wherein the variable coding region is still divided into V, D and J gene segments (for IgH constructs) and V and J gene segments (for IgL chain constructs) in a "quasi-germline" configuration. Although otherwise identical to the retroviral expression vectors provided in (a), these retroviral vectors capable of undergoing V (D) J recombination first need to undergo site-specific rearrangement of the V, optionally D and J gene segments to produce a coding region for the variable binding domain of an IgH or IgL chain. This detailed cloning allowing the expression of IgH chains after V (D) J recombination is depicted in FIG. 11.

One unique feature of these constructs is their ability to generate diverse V domain coding regions in situ in V (d) J recombinant competent cells (e.g., in precursor lymphocytes expressing endogenous RAG1 and RAG2 proteins). Because the process of V (D) J recombination is not precise, a diverse collection of variable coding region sequences can be generated from a single retroviral vector within a given set of V, D and J gene segments (for IgH) or within a given set of V, D and J gene segments (for IgL). The diversity of the joining V, optionally D and J gene segments is due to a combination of exonuclease activity, TdT-mediated N-region addition, and P nucleotide generation, which may contribute individually or collectively to the diversity of coding junctions. Because these V, D and J gene segments have been cloned in a manner that the different V, D and J gene segment family members can be readily replaced by unique restriction enzyme sites, the limited number of constructs generated and introduced into host cells capable of v (d) J recombination can result in a vast diversity of in situ-generated binding protein diversity. Because these vectors contain additional kappa iE and 3' kappa E elements, conferring somatic hypermutation to a V (D) J rearranged V domain coding region, a major diverse set of binding proteins generated in situ can optionally be further mutagenized by an AID-dependent somatic hypermutation method. Thus, the entire process of generating antibody diversity in vivo can be generalized to the use of the disclosed retroviral constructs, and host cells that exhibit v (d) J recombination activity (e.g., precursor lymphocytes), in situ and in vitro, and wherein AID-mediated somatic hypermutation is active, or can be activated.

(c) However, this figure schematically depicts another design of retroviral constructs that can be used to implement the present invention. Here, the expression of these IgH and IgL coding regions is driven by the 5' LTR promoter of the retroviral backbone and the expression of IgH and IgL is coupled to the expression of GFP and YFP autofluorescent markers, respectively, allowing the tracking and isolation of IgH and IgL expressing cells by only analyzing the transduced cells for green and yellow fluorescence. These constructs are very useful for controlling multiple infections of "selected cells" without additional labeling procedures.

The symbol legends used in FIGS. 4a) to c) for the important DNA sequences contained in the constructs are provided herein. For a better understanding of these figures, a subdivision of the IgH and IgL coding regions into a plurality of variable domains (V) is provided_HAnd V_L) They all contain an endogenous leader sequence (L), a hinge region (H), a constant region (C)_H1、C_H2、C_H3、C_L) And the transmembrane coding region (M1/2, since this region is encoded by two exons).

FIGS. 5a-e show the construction of a retroviral IgH (human Ig. gamma₁Isoform) expression vector, which is disclosed in detail in example 1 and provided as a basic design in fig. 4 (a). The cloning of expression constructs for membrane-bound IgG and secreted IgG is depicted here, as detailed in example 1-for general reference purposes, restriction enzyme sites unique in these plasmid maps are provided here. Based on this final retrovirus Ig gamma₁H chain expression constructs, as disclosed herein in FIG. 5e, any other V may be substituted_HDomain coding region, or one diversity V_HThe collection (library) of domain coding regions was introduced into these vectors using unique HindIII and Eco47III restriction endonuclease sites by adding existing V_HZone replacement by any other V_HDomain coding region. FIG. 5a depicts the firstA preparative cloning procedure in which one Eco47III restriction enzyme site (circled) was removed from the commercially available pl hcx vector backbone by site-directed mutagenesis, as illustrated in example 1. This resulted in the retroviral vector backbone pLHCXm1, into which the Eco47III restriction endonuclease site can subsequently be reintroduced for cloning, and V_HReplacement of domain coding regions. The advantage of using Eco47III for this purpose is based on the fact that: eco47III is a substance that can be directly introduced into human V_HAnd Cgamma₁On the borders between coding regions without altering the expressed human Ig gamma₁The only restriction enzyme sites that are composed of the amino acids of the H chain. FIG. 5a Gene further demonstrates how human γ can be engineered using unique HindIII and ClaI restriction endonuclease sites present in the MCS of pLHCXm1₁A fragment of the constant region gene (with or without the transmembrane exon M1/M2) was cloned into the pLHCXm1 backbone. For the purposes of later cloning, these fragments were designed to contain additional flanking Eco47III and NotI restriction endonuclease sites, as detailed in example 1. FIG. 5b) shows no V_HPlasmid map of the cloning intermediates of the domain coding region and shows a specific V flanked by HindIII and Eco47III sites_HHow the coding region was cloned into these constructs. The constructs thus produced are depicted in fig. 5c and are in principle sufficient to provide human Ig γ in any recipient cell line₁Expression of H chain. However, V is treated in an AID-dependent manner_HThe probability of additional mutations in the coding region is an aspect of the present invention, and herein additional two cloning steps are disclosed, wherein the core kappa iE element with additional flanking sequences is cloned into a unique BglII site upstream of the CMV promoter of this expression cassette (FIG. 5c bottom and FIG. 5d), and wherein the 3' kappa E element with some flanking DNA sequences is cloned against human Ig gamma₁In a unique ClaI site downstream of the expression cassette for the H chain. This results in the final human Ig gamma for membrane-bound or secreted forms₁Expression vectors for H chain, for which these plasmid maps are provided in FIG. 5 e. These constructs correspond to those already in FIG. 4aSchematic plasmid maps are disclosed, but precise restriction enzyme maps are used herein and plotted to scale.

FIGS. 6a-d show the detailed cloning strategy provided in example 1 for the construction of retroviral IgL (human Ig kappa L isotype) expression constructs, the basic design of which has been provided in FIG. 4 (b). Any other V can be substituted based on the final retroviral Ig kappa L chain expression construct as disclosed herein in FIG. 6d_LDomain coding region, or multiple V_LThe collection (library) of domain coding regions was introduced into vectors using unique HindIII and Eco47III restriction endonuclease sites by using the existing V_LZone replacement by any one or more of the other V' s_LDomain coding region. The cloning strategy for these retroviral IgL chain expression vectors requires a preparative cloning step, resulting in a modified retroviral vector backbone into which the desired elements can be cloned using convenient restriction enzyme sites as depicted. The first step, as illustrated in example 2, was to remove one unwanted Eco47III site from the ψ (Psi) packaging signal by site-directed mutagenesis from the commercially available plasmid plcpcx, resulting in a modified plasmid plcpcx ml (fig. 6 a). In a second step, a novel pLPCXm2 backbone was generated by ligating a large AscI-BlpI digested fragment from the commercially available plasmid pLHCX with an AscI-NcoI fragment from pLPCXm1 (FIG. 6 b). For both fragments, the incompatible BlpI and NcoI DNA ends need to be filled in with nucleotides using the Klenow fragment, as illustrated in example 1. To the resulting pLPCXM2 backbone, the constant region of a human kappa L chain (Ck) was inserted via HindIII and ClaI, as shown (FIG. 6 b). Similar to the cloning strategy for human IgH chains, these human C.kappa.fragments are further flanked by Eco47III and NotI sites to facilitate additional cloning procedures. After insertion of the human ck fragment, a selected human V κ element was cloned into the construct via the unique HindIII and Eco47III sites (fig. 6C). The construct may in principle be sufficient to provide expression of human Ig κ L chain in any recipient cell line. However, with respect toAs in the case of the IgH chain expression constructs (fig. 5a-E), additional kappa iE and 3' kappa E elements were cloned into the constructs into unique BglII and ClaI sites upstream and downstream of the Ig kappa L chain expression cassette, as was the cloning strategy for these IgH chain constructs (fig. 6c and 6 d). In the final construct, this vk domain coding region can then also be a target for AID-mediated somatic hypermutation. The final construct corresponds to a schematic plasmid map (which is detailed in fig. 4 (b)), but here a fine restriction enzyme map is included and drawn to scale.

FIG. 7: the figure demonstrates the cloning strategy for retroviral expression constructs for activation-induced cytidine deaminase (AID). As depicted, the commercially available plcpcx retroviral vector backbone was used and a specific RT-PCR fragment containing the AID coding region from mouse spleen cDNA was cloned into the unique XhoI restriction enzyme site of plcpcx vector using compatible XhoI restriction enzyme sites inserted into the PCR amplification primers as illustrated in example 2.

FIG. 8: the figure (8a and continuing to 8b) shows a detailed cloning strategy (also provided in example 2) for retroviral reporter constructs with and without the Ig κ L chain enhancer element, allowing identification and quantification of somatic mutations by back-mutating a defined EGFP termination mutation.

FIG. 9: this figure provides an experimental proof-of-concept that the disclosed retroviral vectors allow AID-mediated somatic mutation of sequences like the preferred antibody V coding region, downstream of the cloned V promoter elements. Panel (a) shows AID expression analysis by Western blotting of 5 selected FA-12A-MuLV transformed cell clones that have been stably transfected with a retroviral AID expression construct, the clones of which are depicted in FIG. 7. This protein blot analysis showed an AID-specific signal (approximately 25kD) that could be distinguished in one of FA-12 transfectant clones 1 to 4, but not in transfectant 5, and also in the non-transfected Negative Control (NC). Transfectant 3 was used to further test the retroviral reporter vector for AID-mediated Somatic Hypermutation (SHM), depicted in panel (b): here, the retroviral reporter construct of FIG. 8 (used once and not using the Ig kappa enhancer element once) was transduced retrovirally, correspondingly, into the FA-12 transfectants 3 and 5 expressing AID and not expressing AID. As expected, only when reporter constructs containing these enhancer elements were transduced into FA-12 transfectant clone 3 expressing AID, it was possible to detect a green reversion transductant with a frequency of 0.2% 10 days after transduction. Of these 0.2% green cells, 100 individual cell clones were isolated by single cell sorting, and these clones were again analyzed for green fluorescence by FACS after expansion. Most (95%) of these single cell sorted clones exhibited uniform green fluorescence expression with the same fluorescence intensity as the intermediate green fluorescence signal of the initially sorted 0.2% green cells and similar to the representative GFP expression pattern provided in the lower left panel of fig. 9b, confirming that the initial population of green cells was due to back-mutation of this EGFP termination mutation. 4 clones showed a bimodal green fluorescence pattern, as representatively depicted by the FACS histogram in the middle, and only 1 out of 100 sorted, cloned single cells showed hardly any green fluorescence (right-hand FACS histogram).

FIG. 10: the figure shows the EGFP coding region sequence with an engineered termination mutation for cloning an EGFP reporter construct for quantification of somatic hypermutations. It is known in panel (a) which of these four nucleotides has been mutated in codons 107 and 108 of the EGFP open reading frame, resulting in a stop codon in codon 107 and a lysine to threonine amino acid change in codon 108. These four nucleotide changes additionally led to the introduction of a unique SpeI restriction enzyme site, which as indicated can be used as a diagnostic marker for stop codon back-mutations during somatic hypermutation. The G nucleotide of the TAG stop codon was implanted into a so-called RGYW sequence motif, which is known to be a hot spot for somatic hypermutation. Of the 24 revertant clones analyzed by SpeI restriction enzyme digestion, it could be confirmed that this site became resistant to SpeI digestion (and thus mutated). In 10 of these clones, sequence analysis revealed that the G nucleotide in the initial TAG stop codon had been mutated to a C nucleotide, creating a TAC codon, confirming restriction enzyme analysis and demonstrating that AID-mediated somatic mutations have been targeted to G in the RGYW motif.

(b) This picture shows the complete ORF of the mutated EGFP (which has been cloned into the retroviral Ig. gamma.1H chain construct (as disclosed in FIG. 5 (e)) rather than V_HA domain coding region.

Fig. 11(a) and (b): a detailed cloning strategy for retroviral IgH chain expression vectors capable of V (D) J recombination is shown, as detailed in example 4.

FIG. 12: conceptual evidence that retroviral constructs requiring V, D and J gene segments in a "quasi-germline" configuration for V (D) J recombination can produce significant, rearranged heavy chain expression constructs and Ig + cells when transduced into RAG1/RAG2 positive precursor lymphocytes. Panel (a) contains data showing the generation of surface immunoglobulin positive cells (0.04%, upper right quarter of left FACS plot) after transduction of a retroviral expression vector capable of V (D) J recombination (detailed description of cloning, see FIG. 11) into A-MuLV transformed pre-B cell line 230-238. Expression of the immunoglobulin was coupled to EGFP expression using a construct as schematically shown in fig. 4 c. Thus, immunoglobulin expressing cells can be produced only in this population of green (i.e., stably transduced) cells. Right staining pictures show a re-analysis of surface immunoglobulin expression after one single round of FACS enrichment and 8 days expansion of rare (0.04%) surface immunoglobulins by tissue culture. After this round of enrichment, the binding frequency of immunoglobulin-positive cells had increased to 17.8% (as would be expected to be detectable in these green cell populations (i.e., stably transduced cell populations)), from which PCR amplicons had been obtained and sequenced. (b) As a representative example, the picture shows the DNA sequence obtained from a PCR amplicon derived from surface immunoglobulin cells (clone 225, with amino acid translation above) after a round of enrichment that has transduced retroviral vectors capable of "quasi-germline" v (d) J recombination. As a reference, the coding region sequences of the V, D and J gene segments are provided above in (b), with amino acid translation above these V and J gene segments, since the D segment sequences can be read in three different reading frames, depending on the diversity of the junctions following V (D) J recombination. The insertion between V, D and the J gene segment in the "quasi-germline" configuration is depicted by a dashed line. The sequence of recovered clone 225 clearly represents a true V (D) J rearrangement with the following typical characteristics: nucleotide deletions and TdT catalyzed N-sequence additions (all inserts had been deleted from clone 225) clearly detectable at the coding junction between the assembled V, D and J gene segments. The sequence of clone 225 displayed an open reading frame and contained no additional somatic mutations in the V, D and J sequences except for the above-mentioned changes at the coding nodes.

FIG. 13: data are shown for determining susceptibility to homophilic MLV-derived vector gene transfer for a panel of different a-MuLV transformed murine pre-B cell lines. 1X 10 transduction at an MOI of 0.5 using a vector preparation already coated with the reporter EGFP (containing the transfer vector LEGFP-N1)⁵And (4) cells. Transduction was performed as detailed in example 5. 2 days after transduction, gene transfer was detected by expression of EGFP using FACS. All other A-MuLV transformed pre-B cell lines tested, except pre-B cell line 18/81, were contra-rotating in the frequency range of 40% to 60% under the conditions employedLeads are sensitive and can in principle be used in the present invention. Untreated naive target cells served as negative controls and showed no green fluorescence (not shown).

FIG. 14: characterization of intracellular expression of a panel of murine pre-B cell lines directed to endogenous IgM heavy chains (cy- μ H) to identify cells lacking endogenous murine antibody expression, which can be used as selection cells for reverse transcriptase cell display. The cells were permeable and stained with anti-murine IgM heavy chain antibody conjugated with FITC (FL 1). Untreated cells served as negative controls. The experiment shows that: cell lines FA-12, 1624-5, 1624-6, 18/81-c18-11 and 40E1 actually have undetectable endogenous antibody expression and can therefore be used in the methods of the invention.

FIG. 15: the complexity of retroviral expression vectors following the design disclosed in FIG. 4(c) is demonstrated, as well as the experimental principle for generating a library of IgH and IgL chain shuffled antibodies. (a) Retroviral vector libraries IgH (650) -LIB-IRES-GFP and IgL (245) -LIB-IRES-YFP cover a defined set of coding regions for the Heavy (HC) and Light (LC) chains of fully human antibodies with different fully sequenced clones of 650 and 245 complexity, respectively. Both vectors have a packaging sequence Psi (Psi), flanking Long Terminal Repeats (LTRs) and an Internal Ribosome Entry Sequence (IRES). In parallel with the expression of the antibody polypeptide chain mediated by the viral promoter in the 5' LTR, this IRES was able to correspondingly couple the expression of yfp for the reporter gfp. This allows for convenient detection and enrichment of successfully transduced and immunoglobulin chain expressing cells using FACS when viral genes are transferred into selected cells.

(b) A pool of fully human antibodies was generated in the pre-transformed B cells. To generate transient packaging cells, a library of retroviral transfer vector libraries encoding the heavy chain of a human antibody (IgH (650) -LIB-IRES-GFP) was co-transfected into appropriate recipient cells using a packaging construct (pVPack-GP) and an envelope construct (pVPack-Eco). 2 days after transfection, the resulting library of vector particles that had packaged the corresponding library of transfer vectors was collected and used to transduce pre-selection B cells. Transduced cells expressing these transferred heavy chains and reporter gfp were expanded and enriched using FACS. After expansion, the cells were subjected to secondary transduction. At this point, the IgL (245) -LIB-IRES-YFP library was transferred followed by expansion and enrichment of YFP and human light chain expressing cells using FACS. The resulting population consisted of a fully human antibody displaying a defined library of human antibodies expressed by 1624-5 cells, containing the complexity of the 159' 250 largest clone.

FIG. 16: this figure shows how a two-step transduction with IgH-IRES-GFP and IgL-IRES-YFP libraries is performed under conditions that ensure transduction, which results in clonal expression of the polypeptide chain in most transduced cells. Mixing 1.5X 10⁶1624-5 murine A-MulV transformed pre-B cells, containing recombinant retroviral vectors encoding IgH and IgL chain libraries IgH-LIB-IRES-GFP or IgL-LIB-IRES-YFP, respectively, were suspended in 1ml of tissue culture medium supplemented with different amounts of vector particle supernatants (diluted 1: 1, 1: 5, 1: 20, 1: 50, 1: 100, 1: 200), as illustrated in FIG. 15. To ensure that most of these transduced cells received a single copy of the transfer vector integrated into the host cell genome, cells displaying less than 10% gene transfer efficiency (MOI < 0.1 as detected by expression of conjugated GFP or YFP reporter) were enriched 4 days after infection using FACS sorting. Cells were expanded for 6 days and subjected to a second transduction using vector particles already packaged with the light chain coding region of the antibody, diluted 1: 5 as described above. Here, vector particles transduced with the IgL-LIB-IRES-YFP library were used to infect GFP positive cells selected for heavy chain expression, and vice versa. 4 days after infection, transduced cells expressing GFP and YFP were enriched using FACS. After the second transduction, approximately 20% of the cells showed GFP and YFP expression. To ensure that only a single vector integration per cell occurred, approximately enrichment was performedOne third of these colonies, which showed only low or moderate expression (about 8%) of the transduced reporter gene in the second round.

FIG. 17: IL-15 staining titrations were performed by FACS using a pre-B cell population expressing an anti-IL-15 reference antibody to define optimal conditions that allow optimal IL-15 antigen staining conditions for reverse transcriptase cell display experiments. This staining procedure (as disclosed in detail in example 7) included titrating the IL-15 antigen as indicated in a range of 2.5 μ g/ml to 0.1 μ g/ml with two different concentrations of polyclonal, biotinylated anti-IL-15 secondary antibody, detected by FACS using streptavidin PE conjugate. Surface Ig + cells were counterstained with anti-Ig kappa L chain-APC antibody. As can be seen, the best IL-15 staining was achieved with polyclonal anti-IL-15 antibody at concentrations of 0.1 or 0.5. mu.g/ml IL-15 antigen and using a secondary antibody of 3. mu.g/ml.

FIG. 18: analysis of FACS identification of one anti-IL-15 reference antibody (PC ═ positive control) expressing pre-B cell lines spiked into a diverse antibody library of antibody-expressing pre-B cells at different dilutions, by using optimized IL-15 staining conditions as set out and determined in figure 17. The upper left panel shows the Ig κ L chain-APC/IL-15 double staining of control pre-B cells transduced with a combination of IgH and IgL chain libraries, the generation of which has been shown in fig. 16 (NC ═ negative control). The upper right panel shows Ig κ L chain-APC/IL-15 double staining of pre-B cells transduced with retroviral expression vectors encoding IgH and IgL chains of a reference IL-15 antibody (PC ═ positive control), as disclosed in detail in example 7. FACS plots of these NC cells showed that approximately 50% of the Ab library transduced cells were surface-Ig +, as detected by anti-Ig κ L chain-APC staining. However, cells without surface Ig + exhibited binding to IL-15. In contrast, PC cells (of which more than 90% express surface Ig, a specific IL-15 antigen binding) are cleared by a specific signal on the x-axis. As expected, the higher the expression of surface Ig on PC cells, the more pronounced the shift in this specific IL-15 signal, which resulted in a diagonal staining pattern of surface Ig +/IL-15 bound cells, highlighted by an oval gate (gate), as indicated. The bottom panel (showing dual FACS staining for surface-Ig and IL-15-binding in five different dilutions of PC cells spiked into the cell population expressing the NC random antibody library) shows that PC cells expressing a specific anti-IL-15 reference antibody can be detected at a frequency close to the percentage of PC cells spiked into the NC cell bank.

FIG. 19: conceptual evidence of enrichment of IL-15-reactive cell populations by reverse-transcribed cell display from a diverse antibody library, as disclosed in detail in example 7. The upper panel shows FACS staining for GFP/YFP expression (y-axis) (indicating the frequency of cells transduced by Ig retroviral vectors), and IL-15/anti-IL-15-bio (x-axis) (indicating specific IL-15 staining). The upper left panel shows a two-color FACS analysis of untransduced control pre-B cells (NC ═ negative control), and the upper middle panel shows a two-color FACS analysis of pre-B cells transduced using an anti-IL-15 reference antibody as Positive Control (PC). The upper right panel shows the same two-color FACS staining of a population of cells that have been transduced with a single IgH chain encoding a retroviral vector encoding the IgH chain of a reference anti-IL 15 antibody in combination with a diverse, greater than 7X 10⁴Binding of different Ig κ L chain libraries. Thus, such IgL chain shuffled libraries potentially contain greater than 7X 10⁴And it is expected that even by very narrow gating of antibody-expressing, and IL-15-reactive cells (as indicated in the right upper FACS plot), very few IL-15-reactive cells can be detected (here 2.42%, as the gating is close to the negative population, as indicated). The enriched colonies were amplified by tissue culture and identical staining procedures and FACS sorting were repeated three times as shown in the lower three FACS pictures for the three FACS stains under identical conditions. As can be seen, successive cycles of enrichment/cell expansion produce a population of cells thatExpression was almost 100% for positive antibodies and was even more positive for IL-15 reactivity than the initial PC cell line. These data clearly show that: by repeating FACS sorting and expansion, a population of highly antigen-reactive cells can be successfully enriched from a population of barely detectable antigen-reactive cells into a population of essentially 100% antigen-reactive cells using three successive rounds of reverse transcriptase cell display.

FIG. 20: the figure shows and confirms that: specific IL-15 antigen reactivity of 4 representative cell clones out of 24 individual cell clones established after single cell sorting from a 3-fold IL-15 antigen enriched cell population, as illustrated in figure 19. These 4 selected cell clones were designated clone F, H, V and W, and all clones showed specific IL-15 reactivity on GFP/YFP positive cells, indicating a stably transduced, Ig-encoding retroviral vector. As expected, higher GFP/YFP expressing cells (expressing higher antibody levels) showed higher IL-15 specificity, which produced a characteristic diagonal staining signal in the Ig/IL-15 double staining. All cell clones showed specific IL-15 reactivity, as demonstrated by ignoring the IL-15 antigen in the staining, which resulted in a reduction of IL-15 specific reactivity (not shown). These data provide the following conceptual evidence: reverse transcription cell display is an effective method for obtaining antigen-reactive cell clones at high frequency from an antigen-reactive cell population which has been originally shown to be hardly detectable.

FIG. 21: this figure provides second conceptual evidence for successful retroviral cell display enrichment for antigen-reactive cells by starting with a minimal IL-1 β -reactive cell population in the starting cell population, using three successive rounds of retroviral cell display cell enrichment/tissue culture expansion, demonstrating the successful enrichment of IL-1 β -antigen-reactive cells into a substantially 40% antigen-reactive cell population. Double staining for GFP/YFP expression (indicative of antibody expression) and IL-1 β reactivity are provided herein. The top FACS staining was for non-transducedAs indicated, pre-B selected cells (NC, top left) and cells co-transduced with retroviral vectors (encoding an anti-human IL-1 β specific reference antibody SK48-E26) (PC, top right as positive control) were provided. The bottom panel shows FACS staining of antibody expression and IL-1 β reactivity of an antibody library before (0-fold enrichment) and after, by cells with more than 2 x 10 of reverse transcriptase display enrichment in 1, 2 and 3 rounds⁵A diverse IgL chain library of individual IgL chain clones was shuffled with the IgH chains of SK48-E26 reference antibody, as indicated and as disclosed in detail in example 8. These data provide an independent conceptual evidence for the use of a second antigen, that retroviral cell display expression and enrichment is a powerful tool for enriching an antigen-specific cell population from levels that were barely detectable initially.

FIG. 22: this figure shows confirmation of IL-1 β antigen reactivity of a novel antibody identified by reverse transcriptase cell display, as disclosed in detail in example 8. At this three-fold enriched cell population (shown in fig. 21), 24 individual cell clones have been established by single cell sorting. Of these 24 cell clones, 12 clones had a novel IgL chain, designated LCB24, as disclosed in example 8. When retroviral expression vectors cloned and sequence-characterized IgL and IgH chains were re-transduced into the original selected cell line, the co-expressed novel LCB24 Ig. kappa.L chain IgH chains of the IL-1. beta. specific reference antibody SK48-E26 and the IL-1. beta. specificity of the IgH chains were analyzed by FACS (see example 8). These FACS staining plots show analysis of antibody expression (via GFP/YFP) and IL-1 β reactivity by two-color FACS, as indicated. As expected, no IL-1 β reactivity was detected in non-transduced selected cells (NC ═ negative control, left), whereas a clear IL-1 β specific staining was detected in positive control cells expressing IgH and IgL chains of reference antibody SK48-E26 (middle). A similar IL-1. beta. specific signal was detectable in antibody-expressing selected cells transduced with the SK48-E26 reference antibody IgH chain vector and the novel fully human LCB24 Ig. kappa.L chain cloned from an IL-1. beta. specific retrovirus display cell clone (right).

FIG. 23: confirmation of lack of cross-reactivity with IL-15 of the novel antibody encoded by LCB24IgkL chain/SK 48-E25IgH chain. The two FACS staining plots on the left show negative and positive controls determined for IL-15FACS staining as indicated (NC ═ negative control, untransduced selected cells, PC ═ positive control, selected cells transduced with IgH and IgL chain vectors encoding an anti-IL-15 reference antibody). The two FACS staining patterns on the right show no IL-15 reactivity on antibody-expressing cells encoding novel antibodies consisting of SK48-E26IgH and LCB24IgL chains, or on cells expressing this initial SK48-E26IgH/IgL combination. This proves that: this novel antibody, consisting of SK48-E26IgH and LCB24IgL chains, is not only specific for IL-1 β, but it is also substantially non-cross-reactive (or cohesive) with other proteins like IL-15.

FIG. 24: the figure demonstrates the successful enrichment of streptavidin-APC-Cy 7 antigen reactive cells from an antibody library generated by shuffling a diverse IgH chain library with a diverse IgL as disclosed in example 9, by three consecutive rounds of reverse transcribed cell display cell enrichment/tissue culture expansion. streptavidin-APC-Cy 7 reactive cells were enriched by three consecutive rounds of high-speed cell sorting followed by cell culture expansion, as indicated. The binding specificity of antibody-expressing cells against this streptavidin-APC-Cy 7 antigen was demonstrated by analyzing FACS plots of these sequentially enriched cell populations in the presence (lower panel) and in the absence (upper panel) of the antigen. This provides conceptual evidence for efficient enrichment of specific antigens by reverse transcriptase cell display in the absence of any reference antibody, which can be used in chain shuffling methods.

FIG. 25: the data presented in this figure provide evidence for the specificity of the 3-fold reverse transcribed cell display enriched cell population disclosed in figure 24 for the specific reactivity of the Cy7 fluorescent dye against the streptavidin-APC-Cy 7 tandem dye. To this end, non-transduced selected cells, unenriched cells expressing a IgH/IgL chain library combination, and a tripling streptavidin-APC-Cy 7-enriched cell population were analyzed by FACS for antibody expression (indicated by GFP/YFP fluorescence) and reactivity to different streptavidin-fluorescent dye conjugates, as indicated. This tripling of the enriched cell population of streptavidin-APC-Cy 7 only bound to streptavidin-APC-Cy 7, but not to streptavidin-APC or streptavidin-APC-Cy 5.5, and the non-specific staining of streptavidin-APC-Cy 7 was also undetectable for these selected cells, or selected cells expressing a diverse antibody library. This provides conceptual evidence for efficient and highly specific reverse-transcribed cell display enrichment of specific antibodies from cells expressing a diverse antibody library, without the need for antibody IgH or IgL chains from antigen-specific reference antibodies.

FIG. 26: two novel human antibodies identified by reverse-transcribed cell display sharing the same IgH chain were shown to be specific for the antigen streptavidin-APC-Cy 7. As disclosed in example 9, after three rounds of reverse transcribed cell display enrichment, two different IgH chain sequences (HC49 and HC58) and two different IgL chain sequences (LC4 and LC10) can be identified from a single sorted cell clone. In this figure, all possible pairings of the IgL chains LC4 and LC10 with HC49 and HC58 were examined to determine the reactivity to the target antigen streptavidin-APC-Cy 7. For this purpose, combinations of retroviral expression vectors encoding different IgH and IgL chains are transduced into selected cells, as indicated in example 9 and as disclosed in example 9. As shown, the novel antibodies HC58/LC4 and HC58/LC10 (both sharing the same IgH chain) demonstrated specific binding to the streptavidin-APC-Cy 7 antigen, whereas the antibodies encoded by HC49/LC4 and HC49/LC10 did not show significant binding activity. The specific binding of these two novel antibody clones to the antigen streptavidin-APC-Cy 7 provided conclusive evidence when re-transduced into selected cells: it is possible to use reverse transcriptase cell display as disclosed herein for the identification of rare antibody binders in complex antibody libraries.

Term(s) for

It is convenient to note here that: where the use of "and/or" herein is to be understood as a specific disclosure of each of the two identified features or components (with or without the other). For example, "a and/or B" should not be read as specifically disclosing each of (i) a, (ii) B, and (iii) a and B, as if each were individually listed herein.

Saturation of affinity: the highly regulated immune process of antigen-driven improved antibody binding specificity, produced by antigen-stimulated B lymphocytes, occurs mainly in the germinal center. This process is caused by somatic hypermutations that are primarily targeted to the coding regions of the variable regions of some antibodies that are associated with this selective expansion and survival of B lymphocytes, resulting in higher affinity antibodies.

Antibody: the term describes an immunoglobulin, whether natural or partially or wholly synthetically produced. The term also covers any polypeptide or protein that includes an antibody antigen binding site, such as a heavy chain-only antibody from, for example, a camel or llama. A full-length antibody comprises two identical heavy chains (H) and two identical light chains (L). In its monomeric form, two IgH and two IgL chains assemble into a symmetric Y-disulfide linked antibody molecule with two binding domains formed by the combination of the variable regions of the IgH and IgL chains.

Antibodies may be isolated, or obtained by purification from natural sources, or also obtained by genetic engineering, recombinant expression, or by chemical synthesis, and then they may contain amino acids that are not encoded by germline immunoglobulin genes. A fully human antibody comprises human heavy and light chains, such as variable and constant regions from human species. A chimeric antibody comprises a combination of variable region domains from one vertebrate species and constant region domains from another vertebrate species. The constant domains of a chimeric antibody are typically derived from one or more human antibodies. Humanized antibodies can be generated by grafting the CDRs of a non-human antibody into the framework regions of IgH and IgL variable regions of human origin.

Antibody fragment: it has been shown that a fragment of a whole antibody can perform the function of binding antigen. Examples of binding fragments are (i) a Fab fragment consisting of the VL, VH, CL and CH1 domains; (ii) an Fd fragment consisting of the VH and CH1 domains; (iii) (ii) an Fv fragment consisting of the VL and VH domains of a single antibody; (iv) a dAb fragment which consists of one VH or VL domain; (v) an isolated CDR region; (vi) f (ab') 2 fragments, bivalent fragments including two linked Fab fragments; (vii) single chain Fv molecules (scFv), one V of which_HDomains and a V_LThe domains are linked by a peptide linker that allows the two domains to associate to form an antigen binding site; (viii) bispecific single chain Fv diabodies; and (ix) "diabodies", multivalent or multispecific fragments constructed by gene fusion. The Fv, scFv or diabody molecules may be produced by linking a V_HAnd V_LThe disulfide bridges of the domains are bound and stabilized. Minimal antibodies comprising an scFv conjugated to one CH3 domain can also be prepared. Other examples of binding fragments are Fab ' (which is distinguished from Fab fragments by the addition of several residues at the carboxy terminus of the CH1 domain of the heavy chain, including one or more cysteines from the antibody hinge region), and Fab ' -SH (which is a Fab ' fragment in which one or more cysteine residues of the constant region have a free sulfhydryl group). In some cases, a heavy or light chain may also be considered an antibody fragment. As one of ordinary skill will readily appreciate, all of the above antibody fragments display at least one function of the all-natural antibody from which they are derived, and are therefore referred to as "A functional "fragment.

Antigen: any biological molecule or chemical entity that can bind by variable domains of an immunoglobulin (or antibody).

Binding protein: the term defines a protein of a pair of molecules bound to each other. A binding partner of a binding protein is generally referred to as a ligand. A protein of a binding pair may be naturally derived, or wholly or partially synthetically produced. One protein of the pair of molecules has a region, or a cavity, at its surface that binds to, and is thus complementary to, a particular spatial and polar organization of the other protein of the pair of molecules. Examples of types of binding pairs are antigen-antibody, biotin-avidin, hormone-hormone receptor, receptor-ligand, enzyme-substrate. The invention preferably relates to antigen-antibody type reactions.

Complementarity Determining Region (CDR): the term refers to the highly variable regions of heavy and light chains of an immunoglobulin. CDRs are regions that are in the three-dimensional structure of an immunoglobulin that are in direct contact with an antibody. An antibody typically comprises 3 heavy chain CDRs and 3 light chain CDRs. These CDRs are usually the most diverse parts of antigen receptors.

Domain (b): a structural part of a biomolecule is characterized by a specific three-dimensional structure (e.g., structurally related immunoglobulin variable or constant region domains, such as Ig-like domains, which can be found in many molecules of the immune system, which belong to the so-called Ig-superfamily).

And (3) growing centers: a distinguishable histological structure in peripheral lymphoid organs (e.g., lymph nodes or spleen) in which homologous interactions occur between antigen presenting cells and between different lymphocyte populations, resulting in proliferative expansion of antigen-reactive lymphocytes, along with affinity mutations and class-switch recombination for antibody production by antigen-reactive B lymphocytes.

Embryonic system configuration: the unrearranged configuration of the genes and loci (when they are inherited from a parent and when they are to be passed to another generation by the germ line). DNA recombination events that occur in somatic cells, like, for example, v (d) J recombination in lymphocytes, result in a reorganization or a loss of genetic information at certain loci, and thus in a change in the gene from the germline configuration.

pre-B lymphocytes: Pre-B lymphocytes are characterized by the expression of specific pre-B cell-specific genes (such as, for example,. lamda.5 and V)_preB1And V_preB2Genes), and expresses prodymphoid-specific factors (e.g., RAG-1, RAG-2) involved in V (D) J recombination. Furthermore, pre-B lymphocytes are characterized by DJ on both heavy chain alleles_HOr at least one V on at least one immunoglobulin heavy chain allele_HDJ_HThe presence of rearrangement, although these light chain loci are still in an unrearranged germline configuration, renders these pre-B cells incapable of expressing full antibodies.

Primary lymphoid tissue: in vivo in mice and humans hematopoietic stem cells develop into the organs in which lymphocytes are located, such as bone marrow, thymus, and during the embryo, liver.

"Quasi-germline" ("Quasi-germline") configuration: artificial arrangement of V, optionally D, and J gene segments, wherein flanking recombination signal sequences are cloned from germline immunoglobulin loci into artificial gene constructs, such that the arrangement of V, optionally D, and J gene segments in such artificial gene constructs still allows site-specific recombination of the gene segments into a variable coding region by the V (D) J recombination process.

Somatic mutation: a process in somatic cells that results in the introduction of point mutations into specific regions of the genome. When the process is performed at high frequency (greater than 10 per base pair per cell division)^-4Second mutation) occurs, it is called somatic hypermutation.

V (D) J recombination: this is a method for generating antibodies and T cell receptor diversity, and is a method employed to generate functional antibody genes. It involves the rearrangement of various gene segments that encode the heavy and light chain proteins of immunoglobulins, and it occurs only in lymphocytes.

Transfection/transfection was performed: in the context of eukaryotic cells, this is the process of introducing a nucleic acid sequence into a eukaryotic cell, typically in association with the use of chemical and/or physical methods.

Transformation/transformation was performed: in the context of eukaryotic cells, this is the process of immortalizing cells for establishing a continuously proliferating cell line.

Transduction: a process for delivering DNA into vertebrate cells via the production of recombinant viruses. To this end, a packaging cell line expressing structural proteins for viral particles is transfected with a recombinant viral DNA construct comprising regulatory elements for packaging the viral DNA construct into the viral structural proteins. In this way, recombinant viruses are produced which can be used to infect (mammalian) target cells, which results in the introduction of the cloned genetic information into the recombinant virus genome.

Vector/construct: an artificially generated nucleic acid sequence that can be used to shuttle nucleic acid elements between different organisms and species, and that can further be used to spread, amplify, and maintain genomic information.

Detailed Description_＜0}

Antibodies, or immunoglobulins, are the most common type of binding protein that have proven to be particularly useful for therapeutic and diagnostic applications. Therapeutic antibodies have evolved as the most commercially successful class of biopharmaceuticals, and there is a continuing interest in developing novel and effective methods of antibody-based therapy (Baker, 2005).

Antibodies consist of two identical heavy (H) and light (L) chain glycoproteins, covalently linked via disulfide bonds (fig. 2 a). Each immunoglobulin heavy (IgH) and light (IgL) chain polypeptide comprises an N-terminal variable domain (which is different between different antibodies), and a C-terminal constant domain (which is identical between different antibodies belonging to the same immunoglobulin subtype (isotype)) (fig. 2 a). The combination of IgH and IgL chain variable domains creates an antibody antigen binding pocket and determines its specificity, whereas the constant domains determine an antibody's immune effector function. The variability of immunoglobulins in their variable domains is derived from the fact that: v_HAnd V_LThe domains are encoded by multiple gene segments, referred to as V (variable region), D (diversity region), and J (joining region) gene segments. During B lymphocyte differentiation, one V, one D (present only at IgH chain loci) and one J gene segment are randomly selected per cell and site-specifically rearranged to generate V-directed genes_HAnd V_LA domain coding region. This site-specific gene recombination process occurs only in precursor lymphocytes and is referred to as v (d) J recombination (Grawunder et al, 1998) (see also fig. 2b and 2 c). Rearrangement of gene segments is mediated by the products of Recombinant Activating Genes (RAG) and 2. Due to the large amount of V, D, and the J gene segment, and imprecision in the joining of the gene segments, huge expression profiles of different V-region specificities can be generated daily by the millions of B-lymphocytes generated by the immune system (Grawunder et al, 1998). Because the immunoglobulin heavy chain locus contains V, D and J gene segments, V is directed against an antibody_HThe coding region of the domain requires two consecutive V (D) J rearrangements, however the immunoglobulin loci lack D gene segments and the V_LThe coding region is generated by a V to J rearrangement (fig. 2 c). Thus, the linkage diversity produced by v (d) J recombination in the CDR3 regions of IgH chains is greater than the CDR3 linkage diversity produced by rearrangement events directed to these IgL chains only. Furthermore, in the early B cell differentiation stage(IgH chain gene rearrangements occur at this stage), the expression of the enzyme terminal deoxynucleotidyl transferase (TdT), which is capable of adding non-template nucleotides at the D to J and V to D linkages (so-called N-sequence diversity), additionally diversifies the expression profile of this IgH chain CDR 3. In contrast, the CDR3 expression profile of this IgL chain, which is formed when the V to J gene segments are subsequently joined during B-cell differentiation, is somewhat less complex when TdT expression is predominantly down-regulated (Li et al, 1993). Except at V_HAnd V_LIn addition to creating diversity in CDR3 in the coding region, the chain expression profiles of both IgH and IgL can be further diversified during a T-cell dependent immune response, by somatic hypermutation triggered in mature B-cells (Papavasiliou)&Schatz, 2002). These somatic mutations are specifically targeted to V_HAnd V_LOf the coding region, and is cytosine nucleoside deaminase (abbreviated as AID, see Papavasiliou) induced by B-cell line-specific enzymatic activation&Schatz, 2002). As a result of somatic hypermutations that occur during immunization, cells expressing higher affinity antibody mutants against the immunogen are positively selected during immunization that typically occurs at germinal centers, and result in an enrichment of cells producing higher affinity antibodies. These antibodies now also accumulate mutations in CDR 1 and 2, a process known as affinity mutation of the antibody expression profile. By cis-regulating the presence of gene elements or motifs, particularly located adjacent to rearranged V_HAnd V_LThe presence of enhancer elements at IgH and IgL chain loci in the coding region, this AID-mediated diversity and specific targeting to these V_HAnd V_LThe coding region was significantly increased (Bachl)& Olsson，1999)。

In addition to classical full-length therapeutic antibodies, additional binding protein forms (including fully human antibody fragments, e.g. so-called F_abFragment (FIG. 2a), Single Strand F_vFragment (FIG. 2a), Nanobody, composed of V only_HSingle domain compositions, etc.) are also increasingly being developed as therapeutic and diagnostic testsAnd (3) preparing. However, it will be clear to one of ordinary skill in the art that such functional antibody fragments can be readily derived from full-length antibodies by standard biochemical methods based on proteins, or by conventional molecular biological methods, provided that the coding information for a desired full-length antibody is available.

Traditionally, monoclonal antibodies directed against a molecule or epitope (antigen) of interest are produced by immunizing small laboratory animals, or domestic animals (e.g., mice, rats, rabbits, or goats and donkeys), respectively. After repeated immunizations, the animals were bled for isolation of polyclonal antibodies from their sera or sacrificed for monoclonal antibody production to isolate lymphocytes from secondary lymphoid organs like lymph nodes or spleen. Isolated lymphocytes are fused with immortal myeloma cells for the production of hybridomas which are then subcloned and screened for secretion of monoclonal antibodies exhibiting the desired functional properties, such as, for example, binding to a particular antigen or target.

The first breakthrough from antibody engineering (referred to as "fromTechnical study of hybridomas producing so-called monoclonal antibodies with Milstein: (&Milstein, 1975)) have begun, and monoclonal antibodies have been available for use in the treatment of human diseases for a considerable period of time.

The main reason for the slow entry of antibodies into the clinic was the frustration initially associated with the use of rodent antibodies for the treatment of human patients. If such antibodies are infused into the immune system of a patient, the immune system recognizes the rodent antibodies as a foreign protein and generates an immune response to the antibodies, including the generation of neutralizing antibodies (known as a HAMA ═ human-anti-mouse antibody response). HAMA responses can lead to a significant reduction in half-life and hence the efficacy of the antibodies used, and may even lead to serious side effects if the immune system is over-reactive to the injected non-human protein.

Therefore, there is great medical and commercial interest in developing therapeutic antibodies that are more similar to human antibodies. First, this was achieved by genetic engineering of existing rodent antibodies, leading to the development of chimeric or humanized antibodies (Clark, 2000). Chimeric antibodies are generated by fusing the variable binding domain of a rodent antibody to the constant region of a human antibody using standard genetic engineering and cloning techniques. In contrast, humanized antibodies are generated by transfer of Complementarity Determining Regions (CDRs) from a variable domain of a rodent antibody to the variable region framework of a human antibody, again by standard molecular biology techniques. Although the process of generating chimeric antibodies is straightforward, these antibodies still contain 33% heterologous sequences and have a significant potential for immunogenicity (Clark, 2000). Indeed, the immune response of the mouse part against one chimeric antibody has been well documented and is called the HACA response (HACA ═ human anti-chimeric antibody).

The immunogenic potential of the humanized antibody is further reduced compared to that mentioned above. However, methods of genetically engineering humanized antibodies while maintaining the initial binding affinity and specificity of these rodent antibodies after CDR grafting are not trivial and often require extensive additional optimization through repeated mutagenesis and screening cycles. For the reasons mentioned above, the chimeric method as well as the humanization method have been less considered as one method of choice for developing a therapeutic antibody in recent years.

The development of new and innovative technology platforms for the development of therapeutic antibodies, remote from chimeric and humanized antibodies, has also been driven by the development of innovative technology platforms that allow the development of "fully human" antibodies, whose amino acid sequences are consistent with human serum antibodies. In theory, fully human antibodies are thought to elicit minimal immunogenicity and side effects in human patients.

The two most established "fully human antibody" development platforms are:

A) human immunoglobulin transgenic mouse technology, in which germline human immunoglobulin heavy and light chain loci have been introduced with large numbers of transgenes into the mouse genome (Green & Jakobovits, 1998; jakobovits et al, WO98/24893A 2). To use these transgenic mice for the development of human antibodies, these transgenic mouse strains have been crossed with knockout mouse strains that have functional deletions in their endogenous mouse immunoglobulin heavy chain and kappa light chain loci. Thus, these human immunoglobulin transgenic mice generate a substantial human humoral immune response upon immunization, with the exception that approximately half of these antibody-producing cells still have endogenous mouse λ light chains, which are therefore useless for further therapeutic antibody development and need to be removed.

B) Phage display technology, which is based on highly diverse antibody fragment libraries (e.g., such as, for example, single-chain F)_vOr F_abFragment) on the phage surface of e.coli (Clackson et al, 1991; McCafferty et al, WO 92/01047a 1). To identify specific binders, phage libraries of appropriate complexity, nature, and origin are bound to immobilized antigens ("panning"), thereby enriching for phage clones that bind to the immobilized antigen. After several rounds of panning, the sequence of the selected binding clone was determined. A variant of this method is the completely cell-free ribosome display technique, in which the antibody fragment is not displayed on phage but is expressed by in vitro transcription and translation under conditions in which the translated binder remains "stuck" to the ribosome (Hanes)&Pl ü ckthun, 1997). For phage display or ribosome display, a crucial step is the re-engineering of the bound fragment into a full-length antibody,the full length antibody is then expressed in vertebrate cells. After the phage-selected clone is engineered again into the full-length antibody form and the vertebrate cells are expressed, it needs to be analyzed: whether these antibodies can be sufficiently expressed and whether these initial phage binding characteristics are still retained (in which case this may not be necessary).

Although human immunoglobulin transgenic mice and phage display technology have had a great impact on the development of therapeutic antibodies, both of these technology platforms have advantages and disadvantages associated with them.

One advantage of the transgenic mouse technology is that it is capable of delivering high affinity antibodies due to natural affinity saturation that occurs in these mice when immunised, and it has been demonstrated that the affinity profile of human antibodies for a given antigen from human transgenic mice is comparable to that of wild type mice. However, several disadvantages associated with transgenic animals of these human immunoglobulins are: 1) if these transgenic animals are tolerant to the antigen, most often due to high structural similarity to endogenously expressed host proteins, the production of high affinity antibodies against such "conserved" antigens may become very difficult or even impossible. 2) As in normal wild-type animals, antibodies from human immunoglobulin transgenic animals are preferably raised against strong epitopes, which can make the development of an antibody against a functional, but therapeutically valuable epitope lacking, challenging task. 3) Finally, human immunoglobulin transgenic animals cannot be used for affinity optimization of existing antibodies. The reason for this is that the time period required for the production of transgenic animals (for the optimization of only one given antibody clone) is rather long. This approach would involve the generation of two IgH chain and IgL chain transgenic mouse strains directed against a particular antibody, and would then additionally require genetic backcrossing of the two transgenic strains with at least two knockout animal strains deficient for both endogenous immunoglobulin heavy chains and for kappa light chain expression, a process that would require several breeding generations and an extended period of time.

Similar restrictions as described above apply to a recently described technical platform based on mouse development, where germline variable (V), diverse (D) and linked (J) gene segments of the mouse immunoglobulin heavy and light chain loci have been replaced by (part of) human germline V, D and J gene regions by site-specific gene targeting (Murphy & Yancopoulos, WO 02/066630a 1). In these "immunoglobulin gene knock-in" mice, the murine V, D as well as the J gene segments have been site-specifically replaced in the mouse germline by homologous recombination from those segments of the human immunoglobulin gene heavy and light chain loci. In contrast to human immunoglobulin transgenic mice (which produce full length human antibodies), this mouse strain therefore produces "anti-chimeric" antibodies with human antigen binding regions on a mouse constant region backbone.

The phage display and/or ribosome display method has the understandable advantage of being a very rapid technological platform, since the identification of the first binder from a complex library of binding proteins can be achieved within a few weeks. However, phage display is also associated with a number of significant disadvantages. 1) The identification of high affinity binders from a phage display screen or ribosome display screen is not trivial due to the lack of any affinity saturation in the system. To address this problem, representatives of over 10 have been developed¹²A very complex phage library of individual clones. However, even with these complex libraries, the initial binding clones often have sub-optimal affinity for the antigen, and such binders often still need to be optimized using additional tedious and time-consuming optimization procedures. 2) In phage display, only antibody fragments (e.g., scF) are expressed_vOr F_abFragments) because the phage genome can only accommodate the coding region of molecules of relatively small size. 3) The binding protein must be fused to a carrier protein (e.g., a phage gIII protein). The resulting fusion eggThe leukocytes often exhibit lower antigen reactivity (Hoogenboom) compared to their parent antibodies or binding-activated proteins&Chames, 2000). 4) Phage display cannot readily allow controlled assembly of proteins (which acquire a binding phenotype by forming homo-and hetero-multimeric forms) because, for example, dimeric proteins are forced to assemble by covalently linked molecules. However, in the case of engineered antibodies, properly regulated assembly of immunoglobulin heavy and light chains is necessary, since not every antibody heavy chain is capable of pairing with any one light chain. 5) Bacterial-based or bacteriophage-based systems do not provide adequate post-translational modifications (glycosylation, myristoylation, and the like) of the displayed protein of interest, which often negatively impact the binding characteristics of the expressed protein. 6) Prokaryotic expression results in a different protein folding of the protein compared to vertebrate cells, since this cytoplasmic environment is significantly different (e.g., in terms of redox potential and lack of chaperones) from eukaryotic or vertebrate host cells. 7) The phage display system is then subjected to antigen binding or capture assays to enrich for reactive cells under rather non-physiological "panning" conditions that can lead to the identification of a large percentage of false positive binders that ultimately need to be discarded.

As a result of the above-mentioned disadvantages, many phage display selected antibody fragments have moderate affinities and/or may carry structural artifacts (artifacts). Furthermore, when phage display of selected binders is re-engineered and expressed as full-length antibodies in vertebrate cells, it may happen that the phage-selected antibodies are under-expressed, or not expressed at all, or they display altered binding characteristics.

To address some of the limitations of human immunoglobulin transgene/knock-in mouse technology and phage display, an alternative technology has recently been developed that involves genetic modification of primary murine pre-B cells in vitro (producing cells that express human antibodies) followed by their transplantation into immunodeficient recipient mice that lack a functional B cell compartment (Grawunder & Melchers, WO 03/068819a 1). This results in partial reconstitution of the B cell subpopulation expressing human antibodies in these transplanted mice, which are then immunized with any desired antigen or ligand. This technique can be used to develop novel antibodies or binding proteins, or to optimize existing antibodies in view of their affinity against a defined target (Grawunder & Melchers, WO 03/068819a 1).

The use of retroviral expression systems in this approach is preferred because a single copy of the expression construct's gene can be transferred into individual pre-B cells (Kitamura et al, 1995; Stitz J et al, 2005) and also because there is complete freedom of choice as to which antibody expression constructs are being used in the transplanted mice (e.g., antigen pre-selected antibody libraries, antibody libraries from patients with disease, or individual antibody clones). Thus, one particular advantage of the present technology is its flexibility, i.e. it can be used for de novo development of antibodies, as well as for the optimization of existing therapeutic antibody candidates.

Other rodent-based systems for the development of full-length human antibodies have been described which involve the transplantation of human hematopoietic progenitor cells isolated from human donors into immunodeficient mice (Mosier & Wilson, WO 89/12823a 1). In such human cell-transplanted mice, human B cells can develop to some extent; however, despite recent improvements to this approach (Traggiai et al, 2004), a satisfactory humoral immune response involving affinity saturation of human antibodies has not been achieved in such "humanized mice". Furthermore, existing antibodies cannot be optimized as in the case of human immunoglobulin transgenics or "knock-in" mice.

Any kind of mouse-based antibody technology platform usually requires in vivo immunization, which is still a time-consuming process when compared to in vitro methods.

Thus, in addition to the above mentioned approaches with respect to mice, a number of alternative in vitro techniques have recently been developed. However, there is still a need to demonstrate how these systems will be effective in developing high quality, high affinity antibody products. An in vitro system is based on the isolation of antigen enriched memory B-cells from human patients with a specific disease, which can be isolated and subsequently immortalized by Epstein-Barr Virus (EBV) transformation in vitro, followed by screening against antigen reactive EBV cell lines (Lanzavecchia, WO 04/76677A 2). Conceptually similar, but methodically different, is a process in which antibody-producing plasma cells from patients with an acute disease state are first isolated from peripheral blood and then immortalized by fusion with a non-productive hybridoma, followed by subsequent screening for the desired antibody producer (Lang et al, WO 90/13660A 2). Alternatively, methods have been described which aim to isolate B cells from vaccinated or immunized individuals, followed by isolation and cloning of specific antibody genes from cell populations (Lawson & Lightwood, WO 04/106377A 1; Schrader, WO 92/02551A1) or by single cell PCR (Muraguchi et al, WO 04/051266A 1). However, all of these techniques rely on the availability of the relevant B cell population in human patients and are rather limited in their general application and are therefore mainly used to identify anti-infective therapeutic antibody candidates. Furthermore, with either human B cell-based screening method, affinity saturation, or antigen-directed development of antibodies, or optimization of existing antibodies is not possible.

Thus, additional alternative in vitro methods have recently been described involving the expression and screening of recombinant antibodies in eukaryotic cells using transient expression systems (Zauderer & Smith, WO 02/102855a2 and Beerliet al, WO 08/055795a 1). Although these systems circumvent some of the bottlenecks of transgenic mice, phage display, and human B cell derived technologies, the features of these systems remain a number of limitations. First, known eukaryotic cell-based antibody expression/screening techniques do not confer a stable expression pattern for recombinant antibodies, which precludes repeated enrichment cycles of antibody-expressing cells with the desired binding specificity. Secondly, none of the known methods involving eukaryotic cell-based antibody expression allow for control of the clonal expression of the binding partner clones, which makes identification of a matching IgH chain and IgL chain with the desired antigen or ligand binding activity a challenging task in the case of therapeutic antibody development. Third, the techniques described in Zauderer and Smith (WO 02/102855A2) do not allow any in vitro mutagenesis, or genetic recombination of the expressed antibodies, a purely screening procedure. Thus, the affinity maturation aspect of binding proteins cannot be addressed by this technique. Finally, none of these eukaryotic expression/screening systems are compatible with the in situ generation of diverse antibody expression profiles from individual antibody expression constructs employing the v (d) J recombination mechanisms of immunoglobulin heavy and light chains V, D and J fragments.

An alternative method for identifying biologically active peptides and nucleic acids has been proposed by Jensen et al (EP 1041143A). The preferred method described in EP 1041143A comprises an initial screening procedure in which a large number of retroviral vectors can be introduced into cells so that individual cells can express a large number of different RNAs or peptides. Cells showing a phenotypic change are then isolated and retroviral DNA from the clone can be isolated by PCR. This PCR product can then be used to re-transfect viral packaging cells to produce additional retroviral vectors. These retroviral vectors can then be used to transduce different cells, and finally the active substance can be identified after the second cloning step. This method essentially results in an indirect change in the phenotype of the cell by importing biologically active peptides or nucleic acids. This is in contrast to the method of the invention whereby these retroviral transduced constructs directly encode the binding protein, preferably an antibody, to which the screen is directed. It should also be noted that the peptides and nucleic acids described in EP 1041143 a differ greatly in size from the antibodies or antibody fragments identified by the methods of the invention.

Additional methods for retroviral genome-based screening are set forth in WO 03/083075A2(Bremel et al). The method involves expressing and screening genomic DNA sequences encoding uncharacterized genes and proteins. A process is described in which a cell line is transduced with a retroviral expression construct such that a genomic DNA virus is inserted into the genome of the cell line (as a provirus) and then the expression of polypeptides from the provirus is directly analysed. This method does not provide an opportunity for enrichment of the cell line nor for isolation and identification of the expressed polypeptide is performed prior to analysis, which would be subtracted from the technique of high throughput screening developed by Bremel and co-workers.

A recently published patent application from Beerli and co-workers (WO 08/055795A1) describes a screening platform for the isolation of human antibodies using the Sindbis virus expression system. An essential feature of this platform is the generation of initial libraries in which B cells specific for an antigen of interest are isolated directly from Peripheral Blood Mononuclear Cells (PBMCs) of a human donor. From this B cell pool, recombinant antigen-reactive scFv libraries were generated and screened by mammalian cell surface display using the sindbis virus expression system. Similar to phage display, one of the drawbacks of this system is the need to re-engineer the scFvs of interest and express full-length IgG in vertebrate cells. This process may be associated with a loss of affinity of the antibodies of interest upon transformation, as these antibodies may not be well expressed in vertebrate cells and/or may exhibit altered binding characteristics.

In contrast, the invention disclosed herein comprises a unique and extremely powerful combination of methods for the development and optimization of binding proteins, preferably antibodies or fragments thereof. The main advantages of the invention disclosed herein compared to mouse-based technologies are the thorough flexibility in the optimization and de novo development of antibodies, and the speed of identifying specific binders in a short time. Since all aspects of the invention have been achieved in vitro, there is no limitation to the development of antibodies against antigens that are highly conserved among different species or that may be toxic in experimental animals.

The main advantage of the invention disclosed herein over phage display based techniques is that the binding proteins, in particular antibodies, can be expressed as full length antibodies in vertebrate cells, and preferably in the B lymphocyte environment (i.e., the natural host cell for the antibody), ensuring most natural and proper protein folding, correct post-translational modification, and quality control for heavy and light chain pairing.

The key advantages of the disclosed invention compared to the human B-cell approach are the complete flexibility in view of developing antibodies against any desired target, the possibility of affinity optimization for existing antibodies, the complete freedom to select which type of antibody is expressed in the system (antigen-enriched, artificial, from patients, under conditions of IgH and IgL chain shuffling, etc.).

The key advantages of the disclosed "retroviral cell display" invention over other eukaryotic cell-based expression systems involving plasmid-based expression constructs or non-integrated viral vectors are that stable, sustained and clonal expression can be achieved using retroviral gene transfer techniques. Stable, sustained and clonal expression of recombinant antibodies in these target cells allows for repeated cycles of enrichment of antigen-specific or ligand-specific cells, including the possibility of isolating and expanding monoclonal cells for the identification of these antibody genes. However, the invention disclosed herein additionally allows for the additional generation of genetic diversity when in situ retroviral transduction into vertebrate host cells by either lymphocyte-specific mechanisms employing v (d) J recombination or methods employing somatic hypermutation for further mutagenesis of binding proteins.

Thus, the reverse transcriptase cell display method disclosed herein provides a unique novel and powerful solution to many of the significant limitations of the previously existing technologies, as compared to known technologies known in the art for the development of therapeutic antibodies.

The invention disclosed herein has broad application to expressing, screening and identifying binding proteins that specifically bind to a ligand or antigen of interest. Although the invention can be carried out with any binding protein (including but not limited to monomeric, homo-or heteromultimeric membrane-bound receptors (like T cell receptors, cytokine receptors, or chemokine receptors), but also with other scaffold proteins, preferred binding proteins according to the invention are full-length antibodies, with fully human antibodies being particularly preferred, however, it is understood that any (functional) fragment of an antibody (including but not limited to a single chain Fv fragment (scF)_v) Fab fragment, F (ab') 2, V_HOr V_LSingle domain, single heavy or light chain, or any combination thereof, with any naturally occurring or artificially engineered modification) may be used to practice the invention. For full-length antibodies, the invention is particularly applicable to any kind of artificial engineering treatment or designed modification of antibody binding regions, e.g. by site-directed or region-directed mutagenesis, fusion of naturally occurring sequences from different antibodies, randomization of CDR sequences, DNA shuffling, error-prone PCR, to mention a few methods by way of illustration only.

A preferred method for expressing binding proteins according to the invention is transduction of vertebrate host cells using retroviral vectors.

The use of retroviral vectors has been studied in the field of gene therapy for many years. For example, to engineer adeno-associated virus (AAV) vectors that can target specific cell types, Perabo et al (WO 03/054197a2) have inserted random sequences encoding target peptides into the viral capsid genes at a site critical for binding to primary cell receptors, and generated AAV libraries that display the peptides in the context of such viral capsids. The selective pressure provided by the culture environment drives this selection by virtue of the ability of these viral clones to carry out each step in the infection process (i.e., binding, uptake, uncoating, nuclear translocation, replication, and gene expression). By using this technique, vectors are generated that efficiently transduce leukemic cells. While this technique can be used to generate viral mutants that infect target cells previously resistant to infection by wild-type AAV, it does not provide a diverse collection of binding proteins generated in vitro.

Thus, the methods for expressing binding proteins described herein have several key advantages over any other method known in the art for expressing recombinant proteins in eukaryotic and/or vertebrate host cells.

1) The recombinant retroviral construct is stably integrated into the host cell genome. And thereby provides a stable and sustained expression phenotype of the binding protein. 2) By using appropriate ratios of retroviral particles to target cells (known as the "multiplicity of infection" (MOI)), preferably at an MOI of 0.1 or less, the transduction of such retroviruses can be controlled such that most of the cells transduced by the retrovirus are genetically modified by integration of only one recombinant retroviral construct into the host cell genome resulting in clonal expression of at least one desired binding protein. Since clonal expression of a binding protein significantly facilitates the identification and cloning of individual binding proteins, this aspect represents a preferred embodiment of the present invention. However, in an alternative embodiment, the invention may also be practiced using retroviruses, at MOI greater than 0.1.

Although the advantages of retroviral transduction mentioned above are the basis for retroviral display, expression of recombinant binding proteins in vertebrate host cells can also be achieved by alternative methods such as, for example and without limitation: transient or stable DNA transfection, RNA transfection, or by transfer of DNA-based viral vectors (like adenovirus-or poxvirus-based vectors) -although none of the above mentioned alternative methods allow for a binding protein that can be easily expressed under control, stably and clonally in vertebrate host cells.

Preferred vertebrate host cells for carrying out the invention are cells of a B lymphocyte cell line, in particular precursor B lymphocytes, which usually lack endogenous antibody expression, but which express advantageous auxiliary proteins, like for example chaperones for appropriate protein folding and antibody assembly, or auxiliary membrane proteins which facilitate membrane deposition of antibody molecules, like for example B cell-specific Ig alpha or Ig beta proteins.

The principle of expressing recombinant proteins by retroviral transduction in vertebrate host cells is an established method and involves the construction of recombinant retroviral vectors, which are relatively small (the largest dimension of the recombinant DNA to be added: 8-10kB) and which can be cloned and manipulated by standard cell biology methods as plasmid vectors from which the retroviral RNA genome can be transcribed. The wild-type retroviral genome contains only three genes, gag, pol and env, which correspondingly encode the core proteins (a retroviral integrase, protease, rnase), and a reverse transcriptase, as well as envelope proteins (fig. 3 a). Furthermore, this retroviral genome contains cis-regulatory sequences like the Psi (Psi) sequence required for packaging the retroviral RNA genome into viral particles, a polyA signal for retroviral transcription termination, and finally the so-called 5 '-and 3' -Long Terminal Repeat (LTR) promoter elements and signals for retroviral integration into the host cell genome (fig. 3 a). To construct recombinant retroviruses, the gag, pol and env coding regions of a wild-type retrovirus are replaced with any expression cassette for the gene of interest, including the relevant cis-regulatory elements (like promoters or enhancers) (FIG. 3 a). In order to stably integrate such a recombinant retroviral genome into a host genome, it is necessary to transiently or stably transfect a plasmid vector containing a retroviral genome into a so-called retroviral Packaging Cell Line (PCL) which expresses in a transient or stable manner the viral structural proteins encoded in trans by gag, pol and env and thus allows packaging of the recombinant viral genome (transfer vector) into replication-incompetent retroviral particles (fig. 3 b). These retroviral particles allow a single round of infection (transduction) of the target cells (FIG. 3 b). Entry of retroviral particles into a target cell is mediated by the specific interaction of the Env protein with a specific receptor on the target cell. Thus, the nature of the Env protein determines the tropism of these retroviral particles to the specific host cell expressing the cognate receptor. Ecotropic retroviruses are restricted to rodent cells, amphotropic retroviruses can infect different species (including rodent and human cells), and pantropic retroviruses can infect any replicative cell with a cell membrane, since cell entry occurs via structures present on all eukaryotic cell membranes. Retroviral vector particles with multiple different tropisms can also be produced using heterologous envelope proteins, or even cell membrane proteins, of other viruses (e.g., gibbon ape leukemia virus (GaLV), Vesicular Stomatitis Virus (VSV), or HIV and SIV), to mention a few by way of illustration of a technique known as "pseudotyping". Upon entry into the cell, a retrovirus may deliver the viral genome into the host cell, where the viral proteins mediate the reverse transcription of the genome into cDNA and eventually stably integrate it into the host cell genome, allowing stable expression of the delivered gene (fig. 3 b). In a preferred embodiment of the invention, ecotropic MLV particles are used to mediate gene transfer into murine B cells. However, it will be appreciated by any person of ordinary skill in the art that any infectious retroviral vector pseudotyped with any other envelope or transmembrane protein may be used to practice the present invention provided that it mediates transduction in any appropriate target selected cell (such cells are independent of their parental donor species, cell type) or their expression of a cognate acceptor that mediates entry of the vector cell.

To achieve retroviral vector mediated gene transfer, vector-containing retroviral particles (transcripts containing recombinant retroviral genomes, or transfer vectors) can be collected from the cell culture supernatant of packaging cells for stably or transiently expressed transfer vectors (FIG. 3 b). This can be accomplished within a wide range of scientific protocols and variations thereof known to those of ordinary skill in the art. Preferred embodiments of the present invention include: 1) the supernatant containing the cell-free retroviral particles is prepared by passing through a suitable filter or a centrifugation step to separate the packaging cells from the carrier particles. These retroviral particle preparations are then used to transduce vertebrate host cells by co-incubation for variable periods of time or by performing so-called "spin-infection". Here, the target cell suspension is mixed with a medium containing retroviral particles and subjected to low speed centrifugation (FIG. 3 b). 2) Alternatively, co-culturing (enabling cell-cell contact or separation of the two cell populations) of the target cells with the packaging cells via a membrane (which allows passage of retroviral particles but not the packaging cells) is possible, thereby enabling transduction of the target cells.

As host target cells for retroviral transduction, a preferred embodiment of the method is the use of rodent-derived B-lymphocyte lineage cells which do not express endogenous murine immunoglobulins and which can be transduced with a homophilic host range retrovirus. The cells of this B lymphocyte cell line have the following advantages: they have expressed B-cell specific Ig α and Ig β proteins, which are advantageous for cell surface expression and anchorage of membrane-bound full-length immunoglobulins. In this regard, cells derived from immunoglobulin-negative plasma cells (such as, for example, myeloma cells, as exemplified by, but not limited to, Sp2/0, NSO, X63, and Ag8653) typically lack the auxiliary Ig α and Ig β proteins for membrane immunoglobulin deposition. In these cases and in any other vertebrate host cell in which the Ig α and Ig β proteins are not expressed, the method can still be employed provided that expression of the Ig α and Ig β proteins is conferred upon transfection or transduction of an expression vector for Ig α and Ig β (a standard procedure for those of ordinary skill in the art). Thus, when both proteins Ig α and Ig β are ectopically expressed, the method can be carried out with any vertebrate host cell line, provided that retroviral particles with the appropriate tropism are produced, which are capable of transducing said vertebrate host cell line. For clarity of illustration, the innovations disclosed herein can be implemented with any vertebrate host cell, provided that pantropic retroviral particles (e.g., without limitation, pseudotyped particles of G protein with VSV) are used in conjunction with a host cell that has been modified to ectopically express the immunoglobulin anchoring molecules Ig α and Ig β.

Preferred cells of such B lymphocyte cell lines are for example but not limited to: pre-B lymphocytes, B-leukemia cells, or B-lymphoma cells from any vertebrate species, as well as primary pre-B cells that can be grown in tissue culture for extended periods of time. pre-B lymphocytes may represent an ideal host cell for retroviral expression of immunoglobulins, as most such cell lines do not express endogenous immunoglobulins. In particular, since the pre-murine B cell line can be easily obtained from any mouse strain transformed with Abelson white mouse leukemia virus (A-MuLV). However, primary long-term proliferating pre-B cells, as well as a-MuLV transformed pre-B cells, express the pre-B cell specific proteins VpreB and λ 5, which together form a so-called surrogate light chain that, in the absence of a conventional light chain, can form the pre-B cell receptor complex of an immunoglobulin heavy chain and a surrogate light chain. Because it is desirable to express immunoglobulins comprised of recombinant heavy and light chains, pre-B cells are preferred, which lack expression of surrogate light chain components, including the gene products of the λ 5, or VpreB1, or VpreB2 genes (single, double, or triple gene knockout). Since surrogate light chains are known to bind to heterologous heavy chains, it is expected that surrogate light chain expression may interfere to varying degrees with the screening of IgH/IgL pairs, but the method may also be carried out using wild-type pre-B cells expressing surrogate light chain components, due to the generally low levels of surrogate light chain protein expression in pre-B cells. In summary, any vertebrate cell line that expresses Ig α and Ig β and does not express endogenous immunoglobulins can be used as a target host cell for the present method, where surrogate light chain deficient pre-B cells are preferred host cells for practicing the present invention.

Preferred binding proteins to be expressed, screened and identified are full length antibodies and in terms of amino acid sequence are fully human immunoglobulins. However, it is understood that any binding protein capable of cellular expression in vertebrate cells can be subjected to screening and specific ligand or antibody binding selected according to the disclosed methods. For example, such binding proteins may include fragments of antibodies from any vertebrate species, like for example single chain Fv, Fab fragments (fig. 2a) or V_HOr V_LThe single domain, or one of the heavy or light chains, is preferably expressed in a manner that enables deposition on the cell surface membrane. This can be achieved, for example, by fusion to a membrane-anchoring molecule of other type I transmembrane proteins, using a GPI-anchoring domain, or other methods known in the art. Moreover, the method may also be applied to other membrane-bound proteins such as, but not limited to, monomeric or multimeric cytokine receptors, or dimeric T cell receptors, and the like. Retroviral expression of immunoglobulin heavy and light chains is preferably achieved by sequentially transducing separate retroviral expression constructs for the heavy and light chains. However, the invention can also be achieved by carrying out a target cell co-transduction, wherein separate retroviral constructs for the IgH and IgL chains are used. The separate expression of IgH and IgL chains from different retroviral vectors offers the following advantages: a collection of retroviral vectors encoding a diverse collection of immunoglobulin heavy chains may be combined with a collection of retroviral vectors encoding a diverse collection of immunoglobulinsThe collection of retroviral expression vectors of the light chain collection was randomly combined. This so-called heavy and light chain shuffling can produce a high diversity of different immunoglobulin binding specificities even when the total number of heavy and light chain repertoires is limited (e.g., 10)⁴Heavy chains of different length, random from 10⁴The different light chain combinations theoretically yield 10⁸Species-different antibody specificities). Shuffling of collections of IgH and IgL chain vectors is preferably performed with a single-sided shuffling, i.e. the polypeptide chains of an antibody are separate constructs encoding a single antibody chain.

However, it will be appreciated that retroviral IgH and IgL chain expression may also be achieved, provided that both proteins are encoded on the same retroviral backbone (see below). In its easiest configuration, heavy and light chain expression is conferred by cloning the heavy and light chain cdnas into an empty retroviral vector, where expression is driven by promoter activity of the 5 'LTR and appropriate RNA processing is mediated by the 3' LTR sequence (fig. 3 a). These heavy chain constructs should preferably contain their endogenous transmembrane coding regions, allowing for optimal membrane deposition of these recombinant immunoglobulins. However, it will be clear to one of ordinary skill in the art that the transmembrane regions of other transmembrane proteins may also be fused to the constant region of the antibody, thereby ensuring surface deposition of the expressed modified immunoglobulin. In particular, in the context of expressing antibody fragments, or expressing non-immunoglobulin binding proteins, different transmembrane regions of membrane-bound proteins may be advantageous for cell surface expression of these binders.

However, cell surface expression of antibodies or fragments thereof is a preferred embodiment of the invention, and these biomolecules may alternatively be expressed as soluble, secreted proteins, such that detection of such antibodies is performed in the mobile phase. This form of expression may be advantageous if the screening of individual producer clones and binders involves an assay requiring soluble antibody, or if the assay is performed in a semi-solid medium, where one assay allowsThe expression level and binding specificity on single cell clones were quantified. Expression vectors for recombinant immunoglobulins can be used to encode all known immunoglobulin heavy and light chain isotypes, which in the case of fully human antibodies allow the expression of IgM, IgD, IgG₁、IgG₂、IgG₃、IgG₄、IgA₁、IgA₂And IgE antibodies, which contain an Ig kappa chain or an Ig lambda light chain. In all retroviral expression vectors directed against human heavy and light chains, it is preferred that only one of the variable coding regions of the human heavy and light chains can be replaced with a unique restriction enzyme (such as, for example, but not limited to, HindIII and Eco47III), as depicted in the schematic representation of the retroviral antibody expression vector (4a and 4 b). This will allow for easy cloning and replacement of the variable coding regions (using a V-region library or separate V-region coding regions) in retroviral expression vectors, in-frame with constant coding regions for the immunoglobulin heavy and light chains. Such a scheme of exchanging only the variable region domains (aimed at generating an expression vector encoding a single specificity, or aimed at generating a collection of binding proteins) should be advantageous. In this aspect, full length antibodies can be expressed that contain variable and constant region domains derived from different species (chimeric antibodies).

The simplest retroviral expression vector for a binding protein can be constructed by inserting a cDNA coding region for the binding protein or gene of interest into an "empty" retroviral expression vector backbone (FIG. 3 a). The present invention can be achieved even in the absence of any selection and/or screening markers (e.g., enhanced green fluorescent protein, EGFP) that allow direct detection of transduced cells, since cells stably expressing binding proteins from these retroviral vectors can be identified and isolated based on stable expression (in secreted form or in membrane-bound form) of these binding proteins. However, the different features included in retroviral expression vectors are preferred. The first is a strong constitutive or inducible promoter element driving expression of the recombinant binding proteins, which is located directly upstream of the coding cDNA region. (FIGS. 4a, b). These promoters may be, for example, but are not limited to, constitutive promoters (like immediate early CMV promoter, beta-actin promoter, EF-1. alpha. promoter), or inducible promoters (like tetracycline or any other antibiotic inducible promoter), which may be made up-regulated or down-regulated by the addition or removal of tetracycline or other antibiotics and derivatives thereof (e.g., doxycycline). The inclusion of inducible promoter elements in these retroviral expression constructs is another preferred embodiment, since it is known that in some retroviral expression vector backbones the 5' LTR promoter or even a strong constitutive promoter may be silenced.

In addition to promoter elements, a preferred embodiment is to include in the retroviral expression constructs marker genes which subsequently allow selection and/or monitoring of stably retroviral transduced host cells without the need to detect the recombinant binding proteins (fig. 4a, b) selection and/or screening markers are particularly useful for this preferred two-step retroviral transduction scientific assay protocol involving the sequential transduction of immunoglobulin heavy and light chain retroviral expression vectors. In a two-step transduction scientific protocol, a vertebrate host cell is first transduced with at least one retroviral expression construct encoding one or more first immunoglobulin polypeptide chains and, after stable expression of at least one polypeptide chain for the first time, a second transduction with at least one retroviral expression construct encoding the corresponding other immunoglobulin polypeptide chain or chains is performed, then allowed to produce a complete antibody or collection of antibodies. If a selection or screening marker is used to select or screen for a successful first transduction event, it is very useful to optimize the co-transduction frequency of at least two retroviral expression constructs encoding separate chains of a multimeric binding protein (e.g., antibodies). Therefore, the use of selection and/or screening markers is strongly preferred.

Selection markers for selection of mammalian cells that confer antibiotic resistance include, but are not limited to: for example, genes for resistance to puromycin, neomycin, hygromycin B, mycophenolic acid, histidinol, bleomycin, and phleomycin. For expression of the multi-subunit proteins (e.g. antibodies) encoded by the individual retroviral constructs, the expression of the different polypeptide chains is preferably linked to different selection markers, allowing separate selection of the stably transduced respective expression constructs.

Marker genes that allow monitoring of retroviral transduction into host cells include, but are not limited to: genes that confer autofluorescence to transduced cells, such as, for example, but not limited to, Green Fluorescent Protein (GFP), Enhanced Green Fluorescent Protein (EGFP), Yellow Fluorescent Protein (YFP), Blue Fluorescent Protein (BFP), and Red Fluorescent Protein (RFP). Alternatively, cell surface markers such as CD7 or truncated variants thereof, CD34 or truncated variants thereof, or low affinity nerve growth factor receptors may be used. In a preferred embodiment, the expression of these antibiotic selection markers, fluorescent markers or cell surface markers is coupled to the expression of recombinant binding proteins via so-called Internal Ribosome Entry Sequences (IRES), which allow for the coupled co-expression of two genes from a single promoter element in vertebrate cells (fig. 4 b). However, one of ordinary skill in the art can also practice the invention by expressing a selection and/or marker gene from a separate expression cassette (driven by an additional promoter element) contained in the retroviral construct. For expression of multi-subunit proteins (like immunoglobulins) from separate retroviral vectors, it is preferred to link different binding protein chains to different selection and/or screening markers, thereby allowing stable transduction of these different expression constructs to be monitored separately.

In case the expression of the recombinant binding protein is driven by a separate promoter as outlined above, any selection or screening marker gene can also be cloned downstream of the 5 ' LTR, and downstream of the 5 ' LTR and ψ packaging signals, so that its expression is driven by the 5 ' LTR promoter (see fig. 3 and 4).

As mentioned above, a preferred embodiment of the invention is the expression of the recombinant antibody or fragment thereof in cells of a B lymphocyte cell line, preferably in the pre-B lymph. Thus, it is further preferred that expression of the recombinant antibody is driven by a combination of a promoter and an enhancer, which combination is known to confer selectively high levels of expression in cells of a B cell line. Such promoter/enhancer combinations may be, for example but not limited to: the combination of an immunoglobulin kappa light chain promoter, kappa intron, and 3 'kappa enhancer, or the combination of an immunoglobulin heavy chain, heavy chain intron, and 3' alpha enhancer. The combination of the immunoglobulin kappa light chain promoter, kappa intron and 3' kappa enhancer combination is preferred (fig. 4a, B) because it is known that this combination allows high level expression of immunoglobulin chains in B cell line cells, and because this combination of cis-regulatory gene elements is capable of promoting somatic hypermutation in the coding region of antibodies in a regulated manner (mediated by the activated B cell specific enzyme AID (activation induced cytidine deaminase)), which is an embodiment of the present invention, as further detailed below.

However, one of ordinary skill in the art will recognize that expression of a particular recombinant antibody in a retroviral vector in order to practice the present invention can be effected by any combination of cis-regulated promoter/enhancer elements and coding regions that allow expression of the antibody in a desired vertebrate host cell, either on the cell surface membrane or in secreted form.

Although a preferred embodiment of the invention is the expression of a multi-subunit binding protein (e.g., an antibody) from a single retroviral expression construct (FIGS. 4a and b), the invention can be practiced provided that the expression of the different protein chains of the multi-subunit binding protein are linked on the same retroviral expression construct. In the case of immunoglobulins, this may be achieved by (but is not limited to) expression of the heavy and light chains from one promoter and by the separation of the coding regions for the heavy and light chains by the IRES sequences. In this alternative, it is preferred to clone the heavy chain directly downstream of the promoter and downstream of the light chain of the IRES, since a gene following a known IRES is generally expressed at somewhat lower levels than a gene upstream of the IRES. Since the light chain is a smaller molecule, it is expected that a better stoichiometric expression of the expressed heavy and light chains via IRES linkage is achieved, provided that light chain expression is controlled via IRES.

Alternatively, co-expression of both chains of a dimeric binding protein (e.g., an antibody) can be achieved by cloning two separate expression cassettes into a single retroviral backbone, thus controlling expression of each individual binding protein chain separately. As an alternative to this approach, it is also possible to link the expression of two different binding protein chains into the same vector by using a bidirectional promoter that confers transcriptional activity in opposite directions. The latter option has the potential advantage that no promoter interference occurs (which negatively affects the expression level of the promoter at nearby locations).

It should be emphasized that this method allows the transfer of a single retroviral gene of a binding protein pair into a target cell independently of the detailed genetic organization of a retroviral vector having two binding protein coding regions (e.g., heavy and light chains of an immunoglobulin), which allows for improved control of the clonal expression of a dimeric binding protein and reduced time-limits for generating one population of cells expressing a binding agent compared to the scientific protocol of two-step retroviral transduction.

In addition to cis-regulatory gene elements (like promoters and enhancers), and selectable or screenable marker genes (like anti-biotic resistance markers), and genes encoding autofluorescent proteins, coding regions for immunoglobulin heavy and light chains can be cloned into retroviral expression vectors in different contexts.

In the inventionIn a preferred embodiment, the immunoglobulin heavy and light chain coding regions are cloned into retroviral expression constructs as contiguous cDNA sequences, including appropriate leader sequences required for surface expression and/or secretion. An example of the basic design for such an expression vector with an enhancer element is depicted in fig. 4 a. Preferably, these heavy chains encode human gamma 1 heavy chain isoforms as well as light chain kappa light chain isoforms, however it is to be understood that any other heavy and light chain isoforms of human or other vertebrate species may be used to practice the invention. In such retroviral cDNA expression vectors, it is preferred to include a unique restriction enzyme at the junction between the variable and constant coding regions, which will allow for the replacement of only V_HAnd V_LCoding regions to alter the specificity of the expressed antibody, or which allow the insertion of multiple V_HAnd V_LThe coding region is used to express a diverse library of retroviral antibodies in target cells. In a preferred embodiment, at V_H-C.gamma.1 and V_LThe restriction enzyme site introduced at the C.kappa.border is an Eco47III site (FIG. 4a, b) which does not alter the amino acid composition of the expressed heavy chain and only leads to a conserved threonine to serine amino acid change in the first position of the constant kappa coding region, which does not affect the human IgG expressed by retroviruses₁Binding properties of the molecule.

As an alternative to retroviral constructs containing the coding information for heterologous, preferably fully human antibodies in cDNA configuration, retroviral expression vectors containing coding regions in the genomic structure can be used, in which the typical exon-intron structures for immunoglobulin heavy and light chains are found in germline. Because the retroviral vector will be transcribed into mRNA when the retroviral particle is packaged, this organization of the expression construct requires that the transcriptional organization of the coding regions move in the opposite direction to the transcriptional start site of the 5' LTR of the retroviral genome, since otherwise the retroviral transfer vector will have been spliced, at which time the exon-intron structure will be lost before transduction and stable integration of the recombinant construct into the target cell. However, these constructs provide the functionality that the antibody can be expressed as a membrane-bound or secreted antibody depending on the nature of the target cell used for transduction, and the ability of the target cell to terminate transcription at an internal stop codon for the secreted antibody, or the ability to splice a splice donor upstream of the stop codon for the secreted antibody by alternatively splicing in order to splice the transmembrane exon receptors of the membrane-bound immunoglobulin.

A preferred aspect of the invention is the generation and use of retroviral expression constructs directed against human antibodies or any heterologous antibody or fragment thereof, wherein the variable coding regions of the heavy and/or light chain still need to be assembled from V, optionally D, and J gene segments in "quasi-germline" configuration by the process of V (D) J recombination in the target cells. An illustration of the basic design of such an expression vector is depicted in FIG. 4b, which still has the characteristics of a non-rearrangeable construct, i.e., the "germline" V-D-J or V-J cassette for the heavy and light chains can be replaced by unique restriction enzyme sites, including preferably Eco47III at the 3' edge of the J-element coding region. These V, D and J elements contained in these vectors are flanked by conserved Recombination Signal Sequences (RSS) known as recognition motifs for the Recombinant Activation Genes (RAG)1 and 2. When RAG1 and RAG2 are co-expressed in any vertebrate cell, such vectors will site-specifically recombine V, optionally D and J gene segments to produce V encoding variable domains of antibody heavy and light chains, respectively_HAnd V_LAnd (4) a zone. The expression of the RAG-1 and RAG-2 genes, and thus the V (D) J recombination activity, is generally defined as early precursor lymphocytes. Thus, the invention preferably uses precursor lymphocytes to effect automatically provides activity for v (d) J recombination. However, it is known that over-expression of RAG-1 and RAG-2 may result in any somatic vertebrate cell line suitable for V (D) J recombination, and thus any person of ordinary skill in the art may also use any non-precursor lymphocyte line to achieve this by providing ectopic expression of RAG-1 and RAG-2And (4) an aspect. As an alternative, even RAG-1 or RAG-2 deficient cell lines may be used, wherein the RAG-1 or RAG-2 deficiency is complemented by overexpression of the corresponding RAG gene or a fragment thereof.

Such v (d) J rearrangeable constructs have the following advantages: from a single retroviral expression construct stably transduced into a vertebrate host cell, diverse antibody-specific expression profiles can be generated via RAG-1 and RAG-2 mediated V (D) J recombination.

Although it is known that the joining of V, D and the J gene element involves a large degree of imprecision, this imprecision leads significantly to the presence of V_HAnd V_LThe diverse amino acid sequences found in Complementarity Determining Region (CDR)3, preferably using a collection of V-D-J-C γ 1 and V-J-C κ retroviral constructs representing several V region families, D and J elements, thereby increasing the existing variability at the level of germline gene segment sequences provided. Nevertheless, the use of retroviral constructs (allowing somatic assembly of the V, optionally D, and J gene segments mediated by the V (D) J recombination process) preferably allows for the generation of a large diversity of variable domain binding regions when transduced into precursor lymphocytes in situ, such that a diverse collection of IgH and IgL chains can be generated from a single or limited number of constructs.

The diversity created by imprecise joining of V, D and the J gene segment is significantly increased by the presence of the gene terminal deoxynucleotidyl transferase (TdT) specifically expressed by precursor lymphocytes, the only DNA polymerase that is capable of adding nucleotides to the 3' DNA end without the need for a complementary template DNA strand. To increase the diversity of the ligation, it is preferred to use cells with high endogenous TdT expression levels, or alternatively to ectopically express TdT in target host cells for reverse-transcribed cell display by methods known in the art.

Another embodiment of the invention is the use of V (D) J rearrangeable retroviral constructs containing more than one V, or D or J gene segment, so that different V, D and J gene segments can be used in different rearranged clones from the same construct by the V (D) J recombination process. Incorporation of a variety of different V, D and J gene fragments into such constructs is limited only by the total ability of the retroviral vector to accept DNA, which is reported to be in the range of 8-10kb at the maximum.

Although the use of retroviral constructs capable of V (D) J recombination (FIG. 4a, panel below B) for the expression of heterologous antibodies or fragments thereof is an aspect of the present invention, it is clear that the generation of a diverse expression profile via this method is largely limited to the generation of diversity in the CDR3 regions of the immunoglobulin heavy and light chains, much like the characteristics of an anti-expression profile generated during early B-lymphocyte proliferation.

The adaptive immune system's authentic marker (hallmark) is its ability to affinity saturate the variable domain of an antibody, based on somatic hypermutation of the variable domain coding region. Somatic hypermutations are known to be strongly enhanced by the activation-induced cytidine deaminase (AID) enzyme. High levels of somatic hypermutations are additionally dependent on the presence of cis-regulatory enhancer elements from the immunoglobulin locus, and beneficial effects have been most clearly described for the combination of the Ig kappa intron and the 3' kappa enhancer element. Thus, one aspect of the present invention is the use of retroviruses containing these cis-regulatory elements to retroviral express such immunoglobulin expression constructs in target cells that endogenously or ectopically express the AID enzyme, either constitutively or inducibly by methods known in the art.

The use of "reverse transcriptase display" in the context of retroviral constructs capable of somatic hypermutation and in the context of host cells expressing AID allows for further diversification of an antibody in situ after transduction into host cells expressing AID.

The combination of these aspects of the invention is summarized as all molecular and genetic events that occur in the adaptive immune system (referred to as the generation of a primary antibody expression profile from one or a limited number of constructs containing a limited number of V, D and J gene segments), as well as additional AID-mediated somatic hypermutation of the coding region for the antigen-binding variable domain of an antibody.

Specific selection of higher affinity antibody conjugates to the desired antigen can be achieved by reverse transcribing the cells, sorting the strong antigen conjugates with the selected desired antibody via an increase detected by standard FACS-based techniques, followed by high-speed preparative cell sorting. The strong binders will thus be selectively isolated and the antibody genes encoded by the retroviral vectors can be re-isolated, cloned and sequenced from the selected cells or cell clones by standard molecular biology methods known in the art, including but not limited to genomic PCR and RT-PCR.

In a preferred embodiment, the final cell sorting step is performed as single cell sorting, allowing isolation of such clones and final expansion of antigen-reactive cell clones, which facilitates cloning and sequencing of the coding regions for IgL chain pairs from homologous IgH from the selected binders.

If desired, FACS-rich cells can be expanded by culture, and can again optionally be subjected to antigen binding and the high speed cell sorting of highly reactive cells is repeated, optionally a process that is used repeatedly, until the desired staining intensity and thus the binding specificity for a desired antigen is reached (fig. 1). This selective enrichment and in vitro expansion of antigen-reactive cells mimics the selective results of higher affinity binders that occur in T cell-dependent immune responses.

It should be noted that high speed cell sorter assisted enrichment of antigen reactive cells is only one preferred method of carrying out the invention, but other methods of selecting and isolating cells for antigenic activity (such as, for example, but not limited to, panning) may also be employed, wherein cells are bound to antigens immobilized on a solid support. Furthermore, it is possible to enrich antigen reactive cells by methods of micromanipulation such as, but not limited to: cells are grown in microwell plates or as cell clones in semi-solid media under limiting dilution conditions, which allows staining and/or labeling of specific antigens and their identification by microscope-assisted methods followed by manual and/or manipulator-assisted selection of antigen-reactive clones.

Another embodiment of the invention is to perform repeated cycles of antigen-selection/FAS-sorting/expansion of antigen-reactive cells in the presence of mutagenic conditions, particularly to specifically target the mutations to the coding region of the variable antigen-binding domain. This approach produced higher affinity mutants generated in situ in each round of cell proliferation. Higher affinity mutants can be selectively enriched and expanded when enriched for antibody binding that would show increased upon cell sorting and upon reverse transcription cell display. The hypermutation rate of the targeted antibody variable region is achieved by overexpression of the AID enzyme in cells expressing the antibody, particularly when these expression constructs contain cis-regulated promoters and multiple enhancer elements (including but not limited to immunoglobulin kappa introns and 3' kappa enhancer elements known to provide AID-mediated somatic hypermutation to the antibody variable region (fig. 4a and b)). Although this method can be accomplished using cells that constitutively express AID either endogenously or ectopically, one aspect of the invention uses an AID expression vector in which AID expression can be induced and turned off again using inducible promoters such as, but not limited to, the tetracycline and/or doxycycline inducible promoter systems (Gossen & Bujard, 1992), where one gene expression of interest is controlled by the minimal CMV promoter flanked by tandem repeats of a prokaryotic Tet-operator, and which can be induced or repressed for expression using the HSV-VP 16-Tet-repressor, which fusion protein binds to the Tet-operator under the allosteric control of tetracycline or a tetracycline derivative.

In the following non-limiting examples, the invention is explained in more detail.

Example 1

Retroviral expression vectors for the corresponding cloning of fully human immunoglobulin heavy (IgH) and immunoglobulin light (IgL) chains containing hygromycin B and puromycin antibiotic drug screening markers

As mentioned previously, the present invention can be used with retroviral expression vectors to achieve binding to a differently designed protein (compare, for example, FIGS. 4 a-c). As an example of one of the vector designs that can be used to implement the present invention, detailed cloning strategies for retroviral expression vectors are described herein that allow for the expression of fully human IgG₁The/kappa L antibodies, and antibiotic resistance markers were used to select for these vectors that are stably maintained in the target cells.

a) Construction of retroviral expression vectors for human immunoglobulin heavy chain (IgH) As a starting point for construction of retroviral human immunoglobulin heavy chain expression vectors, the commercially available retroviral vector pLHCX (BD-Clontech, Mountain View, CA) was used (FIG. 5 a). pLHCX contains a hygromycin B resistance marker gene driven by the 5' LTR promoter of the retroviral backbone. In addition, pLHCX contains the CMV immediate early promoter followed by a simple Multiple Cloning Site (MCS) for insertion of the gene of interest to be expressed. In addition, the pLHCX backbone contains a convenient unique BglII restriction site upstream of the CMV promoter (FIG. 5a) into which additional genetic elements can be cloned.

A preferred embodiment of the present invention is the use of Eco47III restriction enzyme for the treatment of human V_HThe coding region was cloned in-frame into these human constant gamma 1 heavy chain coding regions, since this specific restriction enzyme site could be introduced into V_HAnd the C γ 1 coding region without altering the amino acid composition of the expressed IgH chain. However, pLHCX contains an Eco47III restriction endonuclease site in the psi packaging signal (FIG. 5a), which would preclude the direct use of Eco47III for the above-mentioned V_HRegion cloning strategy. To remove this inconvenient Eco47III restriction endonuclease site from the pLHCX vector backbone, one carries outThe following first preliminary cloning step is performed, as detailed in fig. 5 a. Eco47III sites in the psi packaging signal were removed by site-directed mutagenesis using a commercial Quikchange^TMThe kit (Stratagene, La Jolla, CA) replaces the third C nucleotide of the Eco47III recognition sequence AGCGCT with a, using a specific primer pair providing the desired mutation, according to the manufacturer's instructions. The modified vector was named pLHCX-ml, and it was confirmed that this single base pair substitution in the Psi (Psi) packaging signal did not affect the retroviral transduction efficiency of the modified vector pLHCX-m1 (data not shown).

cDNAs encoding human C.gamma.1 constant regions (with or without the transmembrane coding regions M1 and M2) have been cloned in parallel into the pLHCX-ml backbone. C.gamma.1-m and C.gamma.1-s DNA fragments were amplified by RT-PCR using cDNA of human peripheral blood lymphocytes as a template, and forward and reverse primers Seq-ID1, Seq-ID2, Seq-ID3 (see below). For RT-PCR amplification of this membrane bound form of human Ig, a primer combination of Seq-ID1 and Seq-ID2 was used, and for cloning secreted human IgG, a primer combination of Seq-ID1 and Seq-ID3 was used. These forward and reverse PCR amplification primers contained HindIII and ClaI restriction sites, respectively, allowing for the directed cloning of these PCR amplified fragments downstream of the unique HindIII and ClaI sites of the CMV promoter in pLHCX-m1 (FIG. 5 a). The forward PCR amplification primer additionally contains a primer for converting V_HThe domains were fused in frame to an internal Eco47III site of the constant domains without altering the expressed full-length IgGs₁Amino acid composition of the heavy chain. The reverse PCR amplification primers Seq-ID2 and Seq-ID3 contain additional internal NotI sites which allow restriction endonuclease cleavage of such constructs directly downstream of the coding region for general cloning purposes, e.g., exchange of the region encoding the constant region for expression of different Ig isotypes.

Seq-ID1：5′-GATCTCCACCAAGGGCCCATCGGTCTTCCC-3′

HindIII/Eco47III

Primer Seq-ID2 was used as a reverse primer for secretory human IgG₁Together with the PCR amplification of Seq-ID1, and containing a unique NotI site (underlined) for cloning purposes.

Seq-ID2：5′-GATCTCATTTACCCGGAGACAGGGAGAGG-3′

ClaI/NotI

Primer Seq-ID3 was used as a reverse primer for membrane-bound human IgG₁Together with the PCR amplification of Seq-ID1, and containing a unique NotI site (underlined) for cloning purposes.

Seq-ID3：5′-GATCTAGGCCCCCTGCCTGATCATGTTC-3′

ClaI/NotI

The resulting PCR products (about 1.0kb for secreted human C.gamma.1 and about 1.2kb for membrane-bound human C.gamma.1) were digested with HindIII and ClaI restriction enzymes and directionally cloned in parallel into the appropriate restriction sites pLHCX-m1, resulting in plasmids pLHCX-m 1-Cy.1-s and pLHCX-m 1-Cy.1-m, respectively (see also FIG. 5 b). V was then treated with unique restriction enzymes HindIII and Eco47III_HThe chain region is cloned in frame into the coding region of a secreted or membrane-bound human C.gamma.1, flanked by V_HRegion fragments (FIG. 5 b). Combination of these restriction enzymes human V in all 7 human V gene fragment families_HVery little was found in the coding region.

To construct a fully human IgG1 heavy chain expression vector, a pair of NIP-ovalbumin proteins from a previously identified fully human antibody was addedHuman V specific to white_HThe coding region was inserted as a HindIII-Eco47III fragment into constructs pLHCX-m 1-Cy 1s and pLHCX-m 1-Cy 1m, resulting in plasmids pLHCX-m 1-VHCy 1s and pLHCX-m 1-VHCy 1m, respectively (FIG. 5C). V in Seq-ID4 for human antibodies specific for NIP-ovalbumin_HThe coding region, including the leader sequence and 5 '-HindIII and 3' Eco47III cloning sites. It should be noted that two additional C nucleotides have been added upstream of the start-ATG for improved translation (an approximation of the Kozak-consensus sequence):

Seq-ID4：

CCATGGAGTTTGGGCTcAGCTGGGTTTTCCTTGTTGCTCTTTTAA

GAGGTGTCCAGTGTCAGGTGCAGCTGGTGGAGTCTGGGGGAGGCGTGGT

CCAGCCTGGGAGGTCCCTGAGACTCTCCTGTGCAGCCTCTGGATTCACCTT

CAGTAGCTATGCTATGCACTGGGTCCGCCAGGCTCCAGGCAAGGGGCTGG

AGTGGGTGGCAGTTATATCATATGATGGAAGCAATAAATACTACGCAGACT

CCGTGAAGGGCCGATTCACCATCTCCAGAGACAATTCCAAGAACACGCTG

TATCTGCAAATGAACAGCCTGAGAGCTGAGGACACGGCTGTGTATTACTGT

GCGAGAATGGTCGACCACGCGGAAAGCTACTACTACTACTACGGTATGGA

CGTCTGGGGCCAAGGGACAATGGTCACCGTCTCT

HindIII and Eco47III restriction sites used for cloning are underlined in Seq-ID 4. The start ATG of the leader sequence starts at position 9.

Although these human IgGs are depicted in FIG. 5c₁Heavy chainExpression vectors are sufficient for carrying out the invention and for reverse transcription cell display in combination with retroviral IgL chain expression vectors for targeting somatic hypermutations to V_HThe functionality of the coding region requires the presence of certain cis-regulatory enhancer elements of the immunoglobulin light or heavy chain locus. While the kappa intron and 3 'kappa enhancer element of the kappa light chain locus are known to be capable of targeting somatic hypermutations to the V region downstream of an active promoter, these basic retroviral human Ig heavy chain expression vectors pLHCXm1-VHCg1m and pLHCXm1-VHCg1s (FIG. 5c) have been additionally modified in the following manner to contain the kappa intron and 3' kappa enhancer element. The sequence of the murine kappa intron enhancer (kappa iE), which is located in an approximately 2.3kb intergenic region between the J kappa 5 element and the constant kappa coding region, is available from NCBI-Genbank as input V00777. This core kappa iE contains only about 0.5kb in this intergenic region, and its sequence is available from NCBI Genbank import X00268. The total 2.3kb fragment from J.kappa.5 to C.kappa.containing the kappa iE region contains an internal BglII site, precluding the use of the restriction enzyme for cloning a PCR amplified genomic fragment into pLHCX-m1-VHC γ 1-s and pLHCX-m1-VHC γ 1-m vectors. However, there was no internal BamHI restriction endonuclease fragment in this region, thus allowing cloning of a genomic PCR fragment flanked by BamHI sites into the BglII linearized vectors pLHCX-mod-VHC γ 1s and pLHCX-m1-VHC γ 1m (FIG. 5 c). Vectors have been constructed containing two intergenic regions of approximately 2.3kb in length between J.kappa.5 and C.kappa.by PCR amplification of the genomic fragment from mouse genomic DNA using the forward and reverse primers Seq-ID5 and Seq-ID6, both containing additional BamHI restriction sites (underlined) for cloning into the unique BglII restriction sites of pLHCX-m 1-VHCgamma 1 and pLHCX-m 1-VHCgamma 1m, resulting in plasmids pLHCX-m 1-VHCgamma 1 s-kappa E and pLHCX-m 1-VHCgamma 1 m-kappa E respectively (FIG. 5 d).

Seq-ID5：5′-GATCGTACACTTTTCTCATCTTTTTTTATGTG-3′

BamHI

Seg-ID6：5′-GATCCTGAGGAAGGAAGCACAGAGGATGG-3′

BamHI

In addition to the insertion of the complete about 2.3kb kappa iE containing the genomic fragment from the mouse kappa light chain locus, a shorter about 0.8kb genomic PCR fragment containing the core kappa iE (position 3634-4394 of V00777, Seq-ID7) has been cloned into the unique BglII sites of pLHCX-m1-VHC γ 1-s and pLHCX-m1-VHC γ 1-m (not shown). Forward and reverse PCR primers for PCR amplification of this genomic DNA fragment are depicted in Seq-ID8 and Seq-ID 9.

Seq-ID7：

5’GAAAAATGTTTAACTCAGCTACTATAATCCCATAATTTTGAAAACTATTTA

TTAGCTTTTGTGTTTGACCCTTCCCTAGCCAAAGGCAACTATTTAAGGACC

CTTTAAAACTCTTGAAACTACTTTAGAGTCATTAAGTTATTTAACCACTTTT

AATTACTTTAAAATGATGTCAATTCCCTTTTAACTATTAATTTATTTTAAGGG

GGGAAAGGCTGCTCATAATTCTATTGTTTTTCTTGGTAAAGAACTCTCAGT

TTTCGTTTTTACTACCTCTGTCACCCAAGAGTTGGCATCTCAACAGAGGGG

ACTTTCCGAGAGGCCATCTGGCAGTTGCTTAAGATCAGAAGTGAAGTCTG

CCAGTTCCTCCAAGGCAGGTGGCCCAGATTACAGTTGACCTGTTCTGGTG

TGGCTAAAAATTGTCCCATGTGGTTACAAACCATTAGACCAGGGTCTGATG

AATTGCTCAGAATATTTCTGGACACCCAAATACAGACCCTGGCTTAAGGCC

CTGTCCATACAGTAGGTTTAGCTTGGCTACACCAAAGGAAGCCATACAGA

GGCTAATATCAGAGTATTCTTGGAAGAGACAGGAGAAAATGAAAGCCAGT

TTCTGCTCTTACCTTATGTGCTTGTGTTCAGACTCCCAAACATCAGGAGTG

TCAGATAAACTGGTCTGAATCTCTGTCTGAAGCATGGAACTGAAAAGAAT

GTAGTTTCAGGGAAGAAAGGCAATAGAAGGAAGCCTGAGAATATCTTCAA

AGGG-3’

Seq-ID8：5′-GATCGAAAAATGTTTAACTCAGCTAC-3′

BamHI

Seq-ID9：5′-GATCCCCTTTGAAGATATTCTCAGGCTTCC-3′

BamHI

The core kappa iE, which contained an approximately 0.8kb fragment of the genomic PCR fragment, was also cloned as a BamHI digested PCR fragment into the unique BglII restriction sites of both vectors pLHCX-m1-VHC γ 1-s and pLHCX-m1-VHC γ 1-m (not shown here).

The sequence of the stored murine 3' kappa enhancer element can be retrieved under NCBI-Genbank reference number X15878 and is contained in an 808bp gene sequence located approximately 8.7kb downstream of the constant kappa coding region in the mouse genome.

The murine 3 ' kappa enhancer, which does not contain an internal ClaI site and therefore is PCR amplified from murine genomic DNA using the forward and reverse PCR primers Seq-ID10 and Seq-ID11, respectively, contains additional ClaI restriction sites for cloning into the unique ClaI sites of the retroviral vectors pLHCX-m1-VHC γ 1s-3 ' kappa E and pLHCX-m1-VHC γ 1m-3 ' kappa E (FIG. 5 d).

Seq-ID10：5′-GAGAAGCTCAAACCAGCTTAGGCTACAC-3′

ClaI

Seq-ID11：5′-GAGATAGAACGTGTCTGGGCCCCATG-3′

ClaI

This produces the final Ig gamma encoding either Ig heavy chain₁H chain expression vectors pLHCX-m1-VHC γ 1s-3 'kappa E-kappa iE and pLHCX-m1-VHC γ 1 m-3' kappa E-kappa iE (FIG. 5E), which correspondingly result in secreted human IgG chains upon co-expression of IgL chains (FIG. 5E)₁Production of antibodies or resulting in membrane-bound human IgG₁An antibody.

Both vectors are additionally in Ig gamma₁The H chain expression cassette contains kappa iE and 3' kappa E cis-regulatory elements upstream and downstream, and provides somatic hypermutations to these expressed Ig gamma₁V of H chain_HAnd (4) a zone.

b) Cloning of retroviral expression vectors directed against human Ig kappa light chains

As a starting point for the construction of retroviral human immunoglobulin light chain expression vectors that allow antibiotic screening for integration of the retrovirus, the commercially available retroviral vector pLPCX (BD-Clontech, Mountain View, CA) has been used (FIG. 6 a). The vector contains an antibiotic selection marker that provides puromycin resistance driven by the 5' LTR promoter of the retroviral backbone. Although similar in design to the pLHCX backbone (see example 1a), pLPCX contains two Eco47III sites and one MCS with multiple restriction enzyme sites, but lacks a convenient unique BglII site upstream of the CMV promoter (FIG. 6 a).

To remove the Eco47III restriction enzyme from the plcpcx vector backbone and at the same time to introduce a unique BglII restriction enzyme upstream of the CMV promoter, the following preliminary cloning steps were performed: in the first step, the Eco47III site in pLHCX packaging signal was removed by site-directed mutagenesis using a commercial Quikchange^TMThe kit (Stratagene, La Jolla, CA) replaced the third C nucleotide of the Eco47III recognition sequence AGCGCT with a, using a primer pair providing specificity for the desired mutation, according to the manufacturer's instructions (fig. 6 a). It was demonstrated that this single base pair substitution in the Psi (Psi) packaging signal did not affect the retroviral transduction efficiency of these mutated vectors (data not shown). This mutated vector was named pLPCX-m1 (FIG. 6 a). To obtain a vector backbone of plcpcx completely free of Eco47III sites and additionally comprising a unique BglII site located upstream of the CMV promoter, an AscI-NcoI fragment from plcpcx-m 1 (in which the NcoI-digested DNA ends have been filled up by klenow) was cloned into an AscI-BlpI-digested plxcx backbone, in which the BlpI-digested DNA ends have been filled up by klenow (fig. 6B), thereby creating a vector designated plcpcx-m 2 in which essentially only the hygromycin B gene of plxcx has been replaced by the puromycin resistance marker of plcpcx (fig. 6B).

To construct the Ig kappa L chain expression vector, the constant kappa light chain coding region was PCR cloned from human peripheral blood lymphocyte cDNA using forward and reverse primers Seq-ID12 and Seq-ID13 (containing HindIII and ClaI restriction sites, respectively) for directed cloning in pLPCX-m2 (FIG. 6b) as described in section a.) and the forward primer Seq-ID12 additionally contains an Eco47III site allowing V to be cloned into_IIn-frame fusion of the coding region to the constant kappa light chain coding region results in only a conserved threonine at the first position of the human constant kappa light chainAcid to serine amino acid substitutions. The reverse primer contains an additional internal NotI site to facilitate later cloning operations, such as, for example, exchange of constant kappa coding regions.

Seq-ID12：5′-GATCCTGTGGCTGCACCATCTGTCTTCATC-3′

HindIII/Eco47III

Seq-ID13：5′-GATCTAACACTCTCCCCTGTTGAAGCT-3′

ClaI/NotI

This constant kappa light chain coding region flanked at the 5 'end by HindIII/Eco47III sites and at the 3' end by NotI/ClaI sites was inserted into pLPCX-m2 to create plasmid pLPCX-m 2-Ckappa.

To construct a complete human Ig kappa L heavy chain expression vector, a human V kappa coding region specific for NIP-ovalbumin from a previously identified fully human antibody was inserted as a HindIII-Eco47III fragment into construct pLPCX-m 2-Ckappa. (FIG. 6c) the V.kappa.coding region of NIP-ovalbumin-specific human antibodies containing leader sequences and 5 '-HindIII and 3' Eco47III cloning sites is provided in Seq-ID 14. It should be noted that two additional C nucleotides have been added upstream of the start-ATG for improved translation (an approximation of the Kozak-consensus sequence):

Seq-ID14：5’-

CCATGGATATGAGGGTCCCCGCTCAGCTCCTGGGGCTCCTGCTA

CTCTGGCTCCGAGGTGCCAGATGTGACATCCAGATGACCCAGTCTCCATCC

TCCCTGTCTGCATCTGTAGGAGACAGAGTCACCATCACTTGCCGGGCAAG

TCAGAGCATTAGCAGCTATTTAAATTGGTATCAGCAGAAACCAGGGAAAG

CCCCTAAGCTCCTGATCTATGCTGCATCCAGTTTGCAAAGTGGGGTCCCAT

CAAGGTTCAGTGGCAGTGGATCTGGGACAGATTTCACTCTCACCATCAGC

AGTCTGCAACCTGAAGATTTTGCAACTTACTACTGTcAACAGAGTTACAGT

ACCCCCACTTTCGGCCAAGGGACCAAGGTGGAAATCA-3’

HindIII and Eco47III restriction sites used for cloning are underlined in Seq-ID 14. The start ATG of the leader sequence starts at position 9. This HindIII-Eco47III fragment was inserted into HindIII-Eco47III linearized pLPCX-m 2-Ck to give expression construct pLPCX-m 2-Vkappa Ck (FIG. 6C).

While such retroviral kappa light chain expression vectors have been sufficient for the present invention and for retroviral cell display by co-expression with retroviral Ig heavy chain expression vectors, additional vectors also containing kappa iE and 3' kappa E elements have been cloned following the same cloning strategy as for Ig heavy chain expression constructs. Thus, mouse kappa iE was inserted into a unique BglII site in pLPCX-m 2-Vkappa-Ckappa upstream of the CMV promoter as an approximately 2.3kb genomic BamHI digested PCR fragment amplified with the primers Seq-ID5 and Seq-ID6 (see above) or as an approximately 0.8kb genomic BamHI digested PCR fragment amplified with the primer pairs Seq-ID8 and Seq-ID9 (see above). The cloning of this fragment of mouse kappa iE containing a genome of approximately 2.3kb into pLPCX-m 2-Vkappa C kappa is depicted here only, resulting in the plasmid pLPCX-m 2-Vkappa C kappa-iE (FIG. 6 d).

Finally, and in analogy to the construction of IgH chain retroviral expression vectors containing kappa iE and 3 ' kappa E as described in example 1a above, murine 3 ' kappa E was inserted as a ClaI-digested genomic PCR fragment amplified with primer pair Seq-ID10 and Seq-ID11 into a unique ClaI restriction site downstream of the kappa light chain coding region to generate the retroviral expression vector plcpck-kappa iE-3 ' kappa E (fig. 6 d).

Similar to the IgH chain expression vector containing both kappa iE and 3' kappa E elements, this vector now contains all the cis regulatory elements required to provide somatic hypermutation to any one of the vk coding regions cloned into the construct (see below).

Example 2

Generating a cell line overexpressing activation-induced cytidine deaminase (AID)

Activated B-cell specific protein activation-induced cytidine deaminase (AID) has been shown to be a unique transactivating factor that is required and sufficient to provide a somatic hypermutant phenotype to any vertebrate cell line. Somatic hypermutations can be specifically targeted to transcriptionally activated loci in cells expressing AID, provided they are arranged in the correct environment for cis-regulatory enhancer elements, particularly for the kappa iE and 3' kappa E elements of the immunoglobulin kappa light chain locus. To obtain cell lines stably expressing AID, a retroviral expression construct encoding murine AID was first constructed as follows.

Murine AID cDNA was PCR amplified from all mouse spleen cDNA using high fidelity Pfx polymerase (Invitrogen, Carlsbad, Calif.) according to the manufacturer's instructions using forward and reverse PCR primers Seq-ID15 and Seq-ID16 (containing additional XhoI cloning sites for ligation of the PCR amplified fragment into a suitable vector). In addition, the forward primer contained two additional C nucleotides (highlighted in italics) downstream of the XhoI site and upstream of the initiating ATG codon of the murine AID ORF to approximate a Kozak translation initiation sequence and thereby ensure proper translation of the cloned cDNA.

Seq-ID15：5′-AATCCATGGACAGCCTTCTGATGAAGCAAAAG-3′

XhoI

Seq-ID16：5′-AATATCAAAATCCCAACATACGAAATGCATC-3′

XhoI

The resulting 620bp RT-PCR product was digested with XhoI and ligated to pLPCX (from BD-Clontech (Mountain View, CA)) digested with XhoI and treated with alkaline phosphatase. The presence of the ligation product inserted in the correct orientation is determined by enzymatic digestion with a diagnostic restriction enzyme. One clone containing the murine AID cDNA insert in the correct orientation, with the correct restriction enzyme pattern, was verified by DNA sequencing and was named pLPCX-mAID (FIG. 7).

The sequence of the cloned murine AID cDNA exactly corresponds to the published murine AID cDNA ORF (provided at NCBI-Genbank import AF 132979).

Next, 10. mu.g of the PvuI linearized pLPCX-mAID construct was transfected at ambient temperature by electroporation at 300V, 960. mu.F into 5X 10 resuspended in 800. mu.common RPMI medium⁶FA-12Abelson transformed pre-B cells. Transfected cells were resuspended in 20ml growth medium containing FCS and plated at 200 μ l/well in 10 96-well plates. At 48 hours post-transfection, stably transfected cells were selected by adding 2. mu.g/ml puromycin antibiotic to the medium.

After 10 to 14 days post transfection, several tens of puromycin resistant colonies were detectable, and selected clones were transferred to fresh medium containing 2 μ g/ml puromycin. The puromycin antibody clones were further expanded and selected colonies were tested for murine AID protein expression by ECL western blot using a commercial anti-mouse AID antibody as recommended by the manufacturer (see figure 9 a).

As expected, a specific AID protein band was detectable in approximately 80% of the analyzed FA-12-AID transfected cell clones and showed an apparent molecular weight of 25 kD. From this result, it was concluded that: several cell lines constitutively overexpressing the murine AID protein were obtained.

Example 3

Demonstrates that somatic hypermutation targeting one such reporter gene in retroviral human immunoglobulin expression constructs containing cis-regulatory kappa iE and 3' kappa E elements

Next, it was demonstrated that the human antibody variable regions in the retroviral expression constructs (as disclosed in the present invention) are targets for AID-mediated somatic hypermutation. To this end, a reporter construct was created in which the V-region ORF of a human IgH chain was replaced with a mutated EGFPORF, in which a stop codon has been introduced into the background of an RGYW sequence motif known as a hot spot for somatic hypermutation (Bachl & Olsson, 1999).

The stop mutation was introduced at codon 107 of the EGFP ORF, changing a tyrosine codon to a TAG stop codon. In addition, codon 108 was modified to create a novel diagnostic SpeI restriction site in the mutated EGFP sequence, such that when the stop mutation in codon 107 is back-mutated, the SpeI site is disrupted, thereby facilitating the identification and characterization of the back-mutants. The sequence modifications introduced into the EGFP ORF are depicted in fig. 10a, and the entire mutated EGFP ORF is provided in fig. 10 b.

A reporter construct for demonstrating somatic hypermutation was constructed as follows:

from asOne template plasmid pIRES-EGFP (BD-Clontech, Mountain View, CA) was prepared using high fidelity Pfx-polymerase (Invitrogen, Carlsbad, CA), and forward primers Seq-ID17 and Seq-ID18 (each containing additional HindIII and Eco47III restriction sites allowing the replacement of V in pLHCXm1-VHC γ -s- κ iE-3' κ E with one EGFP ORF_H-region) the EGFP ORF was PCR amplified. The forward primer contains two additional C nucleotides (highlighted in italics) upstream of the ATG start codon, which approximate a Kozak translation initiation consensus sequence and ensure proper initiation of translation at the correct ATG start codon.

Seq-ID17：5′-CGCCCATGGTGAGCAAGGGCGAGGAGCTGTTC-3′

HindIII

Seq-ID18：5′-TAGCTTGTACAGCTCGTCCATGCCGAGAGTG-3′

Eco47III

The 737bp Pfx amplified EGFP PCR fragment was cloned directly into the pCR4-Topo vector, which is part of a Zero-Blunt PCR cloning kit (Invitrogen, Carlsbad, Calif.), yielding pCR4-Topo-EGFP vector (FIG. 11). Next, a confirmed sequence clone of pCR4-Topo-EGFP was mutated at codons 107 and 108 of the EGFP ORF, as depicted in FIG. 10, using a Quikchange according to the manufacturer's instructions^TMThe kit (Stratagene, La Jolla, Calif.) was used with specific primer pairs providing these desired mutations, thereby generating plasmid pCR 4-Topo-EGFPmut.

The recovery of the sequence of pCR4-Topo-EGFPmut from this plasmid by double restriction enzyme digestion with the restriction enzymes HindIII and Eco47III was confirmedThe mutant EGFP ORF of (a). The digested fragments were ligated into the HindIII and Eco47III double digested plasmid pLHCXm1-VHC γ -s- κ iE-3' κ E (FIG. 5E) and into the HindIII and Eco47III double digested plasmid pLHCXm1-VHC γ -s (FIG. 5c) which did not contain enhancer elements. Thus, in both vectors these V_HThe coding regions were replaced with mutated EGFP ORFs, which were fused in frame to these C γ s₁In this region, reporter plasmid pLHCXm1-E (mut) -C γ -s- κ iE-3' κ E and control reporter plasmid pLHCXm1-E (mut) -C γ -s were obtained (FIG. 11).

Both plasmids were transduced into anti-puromycin FA-12AID transfectant clone 3 (expressing AID) and 5 (no AID expression) as a control. Transduced cells were cultured under 2mg/ml hygromycin B selection starting 24 hours after transduction and analyzed for the presence of green autofluorescent cells after 6, 8 and 10 days of culture. Only in this experiment, where a mutant EGFP reporter construct was expressed in the background with kappa iE and 3 'kappa E (i.e., using plasmid pLHCXm1-E (mut) -Cgamma-s-kappa iE-3' kappa E) and in the FA-12AID transfectant, where AID expression was detectable by Western blotting (i.e., FA-12AID transfectant clone 3), green autofluorescent cells could be detected after 6, 8 and 10 days in culture when a steady state frequency of approximately 0.2% (FIG. 9b) green cells could be detected with FACS. None of the control experiments (no AID expression, and/or no enhancer element present in the constructs, data not shown) detected green cells over the 10 day duration of the experiment.

After sufficient cells had grown, 192 single cells were sorted from this 0.2% EGFP positive population into each well of two 96-well plates, and 100 clones from these single-sorted clones were analyzed for green fluorescence by FACS. Of the 100 clones analyzed, 95 clones exhibited a homogenous (fluorescent) pattern similar in intensity to the fluorescence detected in the single sorted cells, e.g., at about 10²Logarithmic fluorescence (auto-fluorescence of control cells of FA-12 cells is still at 10)¹Below the logarithmic fluorescence level, by threshold lineOut). 4 clones displayed a different fluorescence pattern, with approximately half of the cells negative for EGFP expression and half positive. Only one of the 100 clones analyzed showed practically no EGFP fluorescence, although this clone was also at an auto-fluorescence level slightly above background. These 5 clones with heterogeneous and negative EGFP patterns could be due to (partially) silent reverse transcription of EGFP expression sites of viral integrants, or these results could be due to the artifacts of single cell sorting. Nevertheless, EGFP expression was clearly detectable in most clones (95%). 24 of these clones were analyzed by PCR using the cloning primers Seq-ID17 and Seq-ID18 to reamplify the EGFP gene from the stably transduced cells.

In contrast to one PCR product from the reporter vector containing the mutated EGFP ORF, none of the 24 PCR products from the expression EGFP clones could be digested with SpeI restriction enzyme, suggesting a TAG termination mutational back mutation in codon 107 of the mutated EGFP ORF (data not shown).

10 of these PCR products have been analyzed by DNA sequencing, confirming that all of these 10 clones contain a G- > C mutation of the G nucleotide in the RGYW motif introduced into the EGFP ORF, as previously described in the literature (Bachl & Olsson, 1999).

This proves that: depending on the presence of cis-regulatory gene elements (like kappa iE and 3' kappa E elements) and depending on AID expression, elevated levels of somatic mutations and thus mutagenesis can be targeted to DNA regions downstream of an active promoter and thus to the V of human antibody chains in the context of the disclosed retroviral expression constructs_HA coding region.

With respect to assessing the level of AID-dependent somatic mutations in these kappa iE and 3' kappa E constructs, the estimated mutation rate was at about 3X 10^-5Minor mutations/bp/generation. This value is still low, since the already reported hypermutation rate of somatic cells in vivo can reach high valuesUp to 10^-4Less than or even 10^-3Ratio of minor/bp/generation. Nevertheless, the detected mutation rate was still significantly higher than the background mutation rate reported in vertebrate cells (the mutation rate was estimated to be at 10^-8Minor mutations/bp/generation). Therefore, it is concluded that: the high somatic mutation rate specifically targets a region downstream of an activated promoter in the disclosed retroviral constructs in an enhancer and AID dependent manner, allowing for display using reverse transcription under in vivo mutagenesis conditions based on somatic hypermutation mediated by AID expression.

Example 4

Demonstration of in situ Generation of human antibody coding regions by Using expression vectors capable of V (D) J recombinant retrovirus

a) A retroviral human heavy chain (IgH) expression vector requiring V (D) J recombination is cloned prior to IgH chain expression.

As an alternative to retroviral expression of heavy (H) and light (L) chains from the cDNA expression vector described in example 1, a different class of retroviral IgH chain vectors has been constructed in which the variable coding region is encoded by the division of the V, D and J segments into a "quasi-germline" configuration which still requires assembly by a v (d) J recombination process prior to expression. V (d) J recombination mediates site-specific but somewhat imprecise assembly of V, D and J gene segments such that multiple V coding regions can be generated from a single expression construct when transduced into V (d) J recombination-competent cells (like, for example, pre-B cells) in situ.

To this end, germline V has been individually isolated from genomic DNA derived from B-depleted human Peripheral Blood Mononuclear Cells (PBMC)_H3.30、D_H1.26 and J_H3 the gene fragment was cloned by PCR. PCR primers selected for amplification of germline V, D and J gene segments were chosen such that flanking DNA sequences including conserved Recombination Signal Sequences (RSSs), as well as additional interfering DNA sequences, allowed for proper assembly of V, and J gene segments,D and J gene segments. All PCR amplicons were generated using the proofreading thermostable DNA polymerase Pfx (Invitrogen, Carlsbad, CA) and originally subcloned into a pSC-BPCR cloning vector (Stratagene, La Jolla, CA), in both cases according to the manufacturer's instructions. The PCR fragment subcloned into pSC-B was confirmed by DNA sequencing and only some fragments could be used for further cloning only, provided that the DNA sequence was sequence confirmed.

For human V of one germ line_H3.30 fragments were PCR amplified using the DNA primers Seq-ID19 and Seq-ID20 containing BamHI and NheI restriction sites as indicated, allowing further subcloning of the DNA fragments of this PCR clone.

Seq-ID19：5′-ATTTCACCATGGAGTTTGGGCTGAGCTGGGTTTTCCTCG-3′

BamHI

Seq-ID20：5′-CCCTCCTGACAGGAAACAGCCTCCATCTGCACCT-3′

NheI

In this way, a PCR amplicon (Seq-ID21) of 623bp length with flanking DNA containing the germline VH3.30 gene segment was obtained, see FIG. 11 a.

Seq-ID21：

5’atttCACCATGGAGTTTGGGCTGAGCTGGGTTTTCCTCGTTGCTC

TTTTAAGAGGTGATTCATGGAGAAATAGAGAGACTGAGTGTGAGTGAACA

TGAGTGAGAAAAACTGGATTTGTGTGGCATTTTCTGATAACGGTGTCCTTC

TGTTTGCAGGTGTCCAGTGTCAGGTGCAGCTGGTGGAGTCTGGGGGAGGC

GTGGTCCAGCCTGGGAGGTCCCTGAGACTCTCCTGTGCAGCCTCTGGATT

CACCTTCAGTAGCTATGCTATGCACTGGGTCCGCCAGGCTCCAGGCAAGG

GGCTAGAGTGGGTGGCAGTTATATCATATGATGGAAGTAATAAATACTACGC

AGACTCCGTGAAGGGCCGATTCACCATCTCCAGAGACAATTCCAAGAACA

CGCTGTATCTGCAAATGAACAGCCTGAGAGCTGAGGACACGGCTGTGTAT

TACTGTGCGAGAGACACAGTGAGGGGAAGTCATTGTGCGCCCAGACACA

AACCTCCCTGCAGGAACGCTGGGGGGAAATCAGCGGCAGGGGGCGCTCA

GGAGCCACTGATCAGAGTCAGCCCTGGAGGCAGGTGCAGATGGAGGCTG

TTTCCTGTCAGGAggg3’

Next, PCR amplification of a human genomic DNA fragment containing the human DH1.26 fragment with flanking genomic DNA was achieved using the primer pair Seq-ID22 and Seq-ID23 (containing restriction endonuclease NheI and XhoI sites, respectively) (FIG. 11 a).

Seq-ID22：5′-GGAGGGCTGCCAGTCCTCACCCCACACCTAAGGT-3′

NheI

Seq-ID23：5′-GGGTCCTCACCATCCAATGGGGACACTGTGGAGC-3′

XhoI

In this way, a PCR amplicon (Seq-ID24) of 336bp length containing the germline DH1.26 gene fragment with flanking DNA was obtained.

Seq-ID24：

5’ggaGGGCTGCCAGTCCTCACCCCACACCTAAGGTGAGCCACAG

CCGCCAGAGCCTCCACAGGAGACCCCACCCAGCAGCCCAGCCCCTACCC

AGGAGGCCCCAGAGCTCAGGGCGCCTGGGTGGATTCTGAACAGCCCCGA

GTCACGGTGGGTATAGTGGGAGCTACTACCACTGTGAGAAAAGCTATGTC

CAAAACTGTCTCCCGGCCACTGCTGGAGGCCCAGCCAGAGAAGGGACCA

GCCGCCCGAACATACGACCTTCCCAGACCTCATGACCCCCAGCACTTGGA

GCTCCACAGTGTCCCCATTGGATGGTGAGGAccc33’

Finally, PCR amplification of a human genomic DNA fragment containing the human JH3 fragment with flanking genomic DNA was achieved using the primer pair Seq-ID25 and Seq-ID26 (containing restriction endonuclease SalI and XbaI/HindIII sites, respectively), as indicated (FIG. 11 a).

Seq-ID25：5′-GGCCCTGCCTGGGTCTCAGCCCGGGGGTCTGTG-3′

SalI

Seq-ID26：5′-TATATATGCCATCTTACCTGAAGAGACGGTGACC-3′

XbaI HindIII

By this method, a PCR amplicon (Seq-ID27) of 239bp in length containing the germline JH3 gene segment with flanking DNA was obtained.

Seq-ID27：

5’ggaCCCTGCCTGGGTCTCAGCCCGGGGGTCTGTGTGGCTGGGG

ACAGGGACGCCGGCTGCCTCTGCTCTGTGCTTGGGCCATGTGACCCATTC

GAGTGTCCTGCACGGGCACAGGTTTGTGTCTGGGCAGGAACAGGGACTG

TGTCCCTGTGTGATGCTTTTGATATCTGGGGCCAAGGGACAATGGTCACCG

TCTCTTCAGGTAAGATGGCTatattata3’

The three DNA fragments Seq-ID21, Seq-ID24 and Seq-ID27 have been cloned sequentially into a shuttle vector containing unique BamHI, NheI, XhoI and XbaI restriction sites so that cassettes containing the gene fragments VH3.30, DH1.26 and JH3 can be assembled by sequential ligation of these DNA fragments via appropriate restriction endonuclease sites. Seq-ID21 was ligated as a BamHI-NheI fragment into a BamHI-NheI linearized shuttle vector, then NheI-XhoI digested fragment Seq-ID24 was ligated into a NheI-XhoI linearized shuttle vector, which already contained Seq-ID21, and finally, SalI-XbaI digested fragment Seq-ID27 was ligated into a XhoI-XbaI linearized shuttle vector, which already contained cloned Seq-ID21 and Seq-ID24, to generate an artificial VH3.30-DH1.26-JH3 cassette in one shuttle vector (FIG. 11 a).

The entire "quasi-germline" cassette containing the artificially assembled VH3.30, DH1.26 and JH3 gene segments was then cloned into the retroviral vector MigR1 (pearl et al 1998), the vector MigR1 already containing the human μ H chain (Seq-ID28, see below) coding region (construct MigR1-muH, fig. 11b) that had been cloned into the unique BglII and HpaI sites of the MigR1 vector. A unique XhoI site (highlighted in bold print in the middle of Seq-ID 28) that separates the VH coding region from the constant μ H chain coding region in MigR1-muH can be used to link the VH3.30-DH1.26-JH3 box to the constant μ H chain coding region in-frame without affecting the amino acid sequence in the transition form JH to this constant coding region.

Seq-ID28：

5’ACCATGGAGTTTGGGCTGAGCTGGGTTTTCCTTGTTGCGATTTT

AGAAGGTGTCCAGTGTGAGGTGCAGCTGGTGGAGTCTGGGGGAGGCTTG

GTACAGCCCGGCAGGTCCCTGAGACTCTCCTGTGCGGCCTCTGGATTCAC

CTTTGATGATTATGCCATGCACTGGGTCCGGCAAGCTCCAGGGAAGGGCCT

GGAATGGGTCTCAGCTATCACTTGGAATAGTGGTCACATAGACTATGCGGA

CTCTGTGGAGGGCCGATTCACCATCTCCAGAGACAACGCCAAGAACTCCC

TGTATCTGCAAATGAACAGTCTGAGAGCTGAGGATACGGCCGTATATTACT

GTGCGAAAGTCTCGTACCTTAGCACCGCGTCCTCCCTTGACTATTGGGGCC

AAGGTACCCTGGTCACCGTCGCTAGTGCATCCGCCCCAACCCTT

TTCCCCCTCGTCTCCTGTGAGAATTCCCCGTCGGATACGAGCAGCGTGGCC

GTTGGCTGCCTCGCACAGGACTTCCTTCCCGACTCCATCACTTTCTCCTGG

AAATACAAGAACAACTCTGACATCAGCAGCACCCGGGGCTTCCCATCAGT

CCTGAGAGGGGGCAAGTACGCAGCCACCTCACAGGTGCTGCTGCCTTCCA

AGGACGTCATGCAGGGCACAGACGAACACGTGGTGTGCAAAGTCCAGCA

CCCCAACGGCAACAAAGAAAAGAACGTGCCTCTTCCAGTGATTGCCGAG

CTGCCTCCCAAAGTGAGCGTCTTCGTCCCACCCCGCGACGGCTTCTTCGG

CAACCCCCGCAAGTCCAAGCTCATCTGCCAGGCCACGGGTTTCAGTCCCC

GGCAGATTCAGGTGTCCTGGCTGCGCGAGGGGAAGCAGGTGGGGTCTGG

CGTCACCACGGACCAGGTGCAGGCTGAGGCCAAAGAGTCTGGGCCCACG

ACCTACAAGGTGACCAGCACACTGACCATCAAAGAGAGCGACTGGCTCA

GCCAGAGCATGTTCACCTGCCGCGTGGATCACAGGGGCCTGACCTTCCAG

CAGAATGCGTCCTCCATGTGTGTCCCCGATCAAGACACAGCCATCCGGGTC

TTCGCCATCCCCCCATCCTTTGCCAGCATCTTCCTCACCAAGTCCACCAAG

TTGACCTGCCTGGTCACAGACCTGACCACCTATGACAGCGTGACCATCTCC

TGGACCCGCCAGAATGGCGAAGCTGTGAAAACCCACACCAACATCTCCGA

GAGCCACCCCAATGCCACTTTCAGCGCCGTGGGTGAGGCCAGCATCTGCG

AGGATGACTGGAATTCCGGGGAGAGGTTCACGTGCACCGTGACCCACACA

GACCTGCCCTCGCCACTGAAGCAGACCATCTCCCGGCCCAAGGGGGTGGC

CCTGCACAGGCCCGATGTCTACTTGCTGCCACCAGCCCGGGAGCAGCTGA

ACCTGCGGGAGTCGGCCACCATCACGTGCCTGGTGACGGGCTTCTCTCCC

GCGGACGTCTTCGTGCAGTGGATGCAGAGGGGGCAGCCCTTGTCCCCGGA

GAAGTATGTGACCAGCGCCCCAATGCCTGAGCCCCAGGCCCCAGGCCGGT

ACTTCGCCCACAGCATCCTGACCGTGTCCGAAGAGGAATGGAACACGGGG

GAGACCTACACCTGCGTGGTGGCCCATGAGGCCCTGCCCAACAGGGTCAC

CGAGAGGACCGTGGACAAGTCCACCGAGGGGGAGGTGAGCGCCGACGA

GGAGGGCTTTGAGAACCTGTGGGCCACCGCCTCCACCTTCATCGTCCTCTT

CCTCCTGAGCCTCTTCTACAGTACCACCGTCACCTTGTTCAAGGTGAAATG

AGCGGCCGCTTTACGC3’

The BglII-HpaI restriction sites at the 5 'and 3' ends of the insert were also correspondingly highlighted in bold print and underlined, and indicated conversion to the MigR1 vector backbone (Pear et al 1998).

To replace the V-coding region in Seq-ID28 contained in the MigR1 retroviral backbone with the VH3.30-DH1.26-JH3 "quasigillary" cassette, this approximately 1.1kb VH3.30-DH1.26-JH3 fragment required re-amplification by PCR using a forward primer containing BglII and a reverse primer containing XhoI (Seq-ID 29 and Seq-ID30, respectively) (FIG. 11 a).

Seq-ID29：5′-GCACCATGGAGTTTG-3′

BglII

Seq-ID30：5′-ATCTTACCTACGGTGA-3′

XhoI

The resulting PCR fragment of approximately 1.1kb was digested with BglII and XhoI and ligated into MigR1 vector containing the linearized. mu.H chain of BglII-XhoI, thereby obtaining retroviral expression vector pVDJ-muH-MigR1 capable of V-D-J recombination (FIG. 11 b).

b) True V (D) J recombination that occurs in retroviral V-D-J vectors producing diverse sequences is demonstrated when transduced into pre-B cells.

As a conceptual proof, to demonstrate that suitable V (D) J recombination could occur in a retroviral vector containing V, D and J gene segments in a "quasi-germline" configuration, the vector pVDJ-muH-MigR1, along with a retroviral IgL chain expression vector, was co-transduced into A-MuLV-transformed pre-B cell line 230-238 as illustrated in example 1B. A full-length human IgM antibody can be expressed on the cell surface of the doubly transduced cells only if a v (d) J recombination event occurs on the pVDJ-muH-MigR1 construct resulting in an in-frame rearrangement of the V, D and J gene fragments. The transduction efficiency of the pVDJ-muH-MigR1 vector can be monitored by co-expression of an IRES-coupled EGFP marker gene. As can be seen in fig. 12(a), a very small 0.04% population of cells transduced with at least the pVDJ-muH-MigR1 construct (transduction efficiency 44.7%) displayed detectable IgM expression on the cell surface as measured by FACS staining with an anti-kappa light chain antibody. Notably, no IgM expressing cells were actually detected in the cell population not transduced with the pVDJ-muH-MigR1 construct (lower right panel in fig. 12 (a)), demonstrating the specificity of this staining.

The rare IgM expressing cells that can be detected in the upper right quadrant of figure 12(a) have been sorted by preparative cell sorting using a FACS-Aria high speed cell sorter (BD, Franklin Lakes, NJ), and have been expanded by tissue culture for 8 days to expand the cells for characterization of the retroviral integrants. FACS plots for EGFP expression (indicating integrated pVDJ-muH-MigR1 construct) and surface IgM after 8 days of expansion showed possibly few clonally expanded cells displaying IgM on the cell surface and containing the pVDJ-muH-MigR1 construct (as measured by green fluorescence, fig. 12 (b)). Cell populations distinguishable in the upper right region of the FACS plot of fig. 12(b) have been sorted, and genomic DNA has been prepared from pooled cell populations.

Genomic DNA was analyzed by diagnostic PCR using primers that bind to the pVDJ-muH-MigR1 construct upstream of the VH and downstream of the JH region. As expected, these diagnostic PCRs produced two separate bands of nearly identical intensity (indicating the unrearranged V, D and J gene fragments (approximately 1.2kb fragments)), and a smaller approximately 0.5kb fragment (indicating the v (d) J recombined gene fragments) (data not shown). Sequencing of the larger PCR bands confirmed: this PCR amplicon represents the unrearranged V-D-J cassette, still in the "quasisotyl" configuration. The unrearranged constructs remain detectable in IgM positive cells, provided that the cells are transduced with more than one construct, wherein not all of the constructs are accessible for v (d) J recombination. When this PCR product was sequenced, this smaller PCR amplicon did not produce a unique sequence and required subcloning into the pSC-BPCR cloning vector for PCR sequence analysis of the individual PCR fragments.

Of the 6 plasmids analyzed, 2 contained identical authentic V (D) J recombination sequences, which showed all the characteristic features of V, D and J gene segments site-directed joined by V (D) J recombination, including nucleotide deletions in the coding regions and the addition of non-template sequences (N-region) catalyzed by the precursor lymphocyte-specific enzyme terminal deoxynucleotidyl transferase (TdT) (FIG. 12 (b)).

The two recovered sequences represented by the sequence of clone 225 (fig. 12B) are already solid evidence for the V, D and J gene segments containing retroviral expression vectors being able to undergo v (d) J recombination in pre-B lymphocytes, since there is no other explanation as to how these sequences could otherwise be generated which show all the signals of a true v (d) J recombination event, even though the efficiency at this point appears to be low. And (5) drawing a conclusion that: when the efficiency of v (d) J recombination is increased in the context of retroviral transduction of precursor lymphocytes, a diverse collection of antibody sequences can be generated from a limited number of retroviral vectors containing V, D and J gene segments in a "quasi-germline" configuration.

Example 5

Identification and characterization of appropriate selection murine B cell lines for reverse transcriptase cell display

a) Expression of endogenous antibodies in B cell lines may potentially hinder their use as selection cells (selector cells) in reverse transcriptase cell display, as pairing of endogenous mouse immunoglobulin chains with recombinant full antibody chains may negatively affect their cell surface display, or may result in expression of mixed human-mouse immunoglobulins with an undefined binding specificity. Thus, a panel of Abelson murine leukemia virus (A-MuLV) transformed murine pre-B cell lines described in the literature was examined for intracellular expression of endogenous IgM heavy chain (. mu.H) using anti-murine IgM heavy chain antibody conjugated to FITC (Southem Biotech). These cells include lineages 40E1, 230-. These cells were permeabilized using the Fix/Perm-kit (Caltech) following the manufacturer's instructions. As depicted in fig. 14, cell lines FA-12, 40E1, and 18-81 subclones 8-11, 1624-5, and 1624-6, showing minimal or no signal for intracellular IgM staining, were referred to as suitable selection cells for reverse transcriptase cell display.

b) Since retroviral cell display is based on retroviral vector-mediated gene transfer, the group of A-MuLV-transformed pre-B cell lines were further examined for their susceptibility to retroviral transduction using ecotropic MLV-derived vector particles containing a Green Fluorescent Protein (GFP) marker gene (FIG. 13). 1X 10 transduction with a vector preparation at an MOI of 0.5⁵Cells, which had been packaged with reporter GFP, transfer vector LEGFP-N1(Clontech) including pre-B cells subcloned by 18-81 by titration prior to limiting dilution 8-11. The carrier particles are produced as described below. Transduction was performed by rotary infection using essentially centrifugation in a 1.5ml-Eppendorf tube at 3.300rpm and 30 ℃ for 3 hours. Two days after transduction, gene transfer was analyzed by determining the frequency of GFP-expressing cells (by FACS analysis). Untreated, native target cells served as negative controls. As shown in FIG. 13, only the initial 18-81 cells showed very low transduction (< 10%) of permissive MLV vectors. All other cell lines showed a gene transfer efficiency of less than or equal to 40%. Notably, the gene transfer efficiency was highest in this experiment using FA-12, 40E1 and 1624-5 cells and reached greater than 50%.

Taken together, FA12, 40E1, and 1624-5 cells were found to be the most suitable for reverse transcriptase cell display, considering two indicators: a) low or no endogenous murine immunoglobulin expression, and b) susceptibility to retroviral transduction. However, because it is desirable to express immunoglobulins composed of recombinant heavy and light chains, B cells are preferred, which also lack surrogate light chain components (which may be expressed from λ 5, or VpreB1, or VpreB2 genes) because those may compete for heavy chains associated with the recombinant light chains in wild-type pre-B cells. Therefore, 1624-5 cells obtained from surrogate light chain triple knockout mice are expected to be the most suitable cells for further reverse transcription cell display technology. It is understood, however, that any other cell line (including those additional cell lines analyzed herein that meet the criteria of no/low endogenous immunoglobulin expression, and retroviral transduction capability) can be used to practice the methods disclosed herein.

Example 6

Selected cells producing a clonally and stably expressed fully human antibody library

In order to generate vector particles already packaged with transfer vectors (encoding fully human antibody chains and libraries thereof) and subsequently use them for transduction of murine B cell lines, infection experiments were performed by the following method. 16 to 24 hours before transfection, human embryonic kidney 293T-HEK cells were plated in 10ml of Dulbeccos Modified Eagle Medium (DMEM) at 2X 10⁶One was plated and 10% Fetal Calf Serum (FCS) and L-glutamine were added to each 10cm tissue culture dish. A mixture of 5. mu.g of the corresponding transfer vectors IgL (245) -LIB-IRES-YFP and IgH (650) -LIB-IRES-GFP (libraries encoding either heavy or light chains linked to GFP or YFP expression by an IRES, FIG. 15), 3. mu.g of pVPack-GP (an expression construct with gag and pol genes of MLV) and 2.5. mu.g of pVPack-Eco (an env gene expression construct including monotropic MLV, all from STRATAGENE) was prepared and incubated with 30. mu.l of DMEM in 1ml serum-free Fugene (Roche) and left to stand at room temperature for 15 to 30 minutes. This Fugene/DNA mixture was then gently added to 293T-HEK cells seeded in 10cm dishes. The transfer vectors encoding the heavy and light chains were stained into separate transient packaging cells.

48 hours after transfection, the supernatant containing the vector particles was collected from the transient packaging cells and centrifuged at 3,000rpm to remove contaminating cells. Mixing 1.5X 10⁶1624-5 murine B cells were suspended in 1ml of medium supplemented with different numbers of carrier particles (1: 1; 1: 5, 1: 20; 1: 50; 1: 100; 1: 200 dilution) which had been packagedCoding regions for the heavy or light chain of the antibody. Transduction was performed by centrifugation at 3.300rpm and 30 ℃ for 3 hours in a 1.5ml-Eppendorf tube. The unused supernatant was stored at-80 ℃ for use at a later time point. To ensure that a single copy of the transfer vector has integrated into the host cell genome, cells showing gene transfer efficiency of less than 10% (detected by expression of GFP or YFP) 4 days after infection were enriched using FACS (fig. 16). The cells were expanded for 6 days and subjected to a second transduction procedure employing previously frozen vector particles which had been packaged with the light chain coding region of the antibody diluted 1: 5 as described above. Here, vector particles transducing the light chain-IRES-YFP library were used to infect GFP positive cells selected for heavy chain expression, and vice versa (FIG. 15)&16). Successfully transduced cells expressing GFP and YFP were enriched 4 days after infection using FACS. After the second transduction, approximately 20% of these cells showed GFP and YFP expression. To ensure that only single vector integration occurred per cell, approximately one third of these populations were enriched, showing only low or moderate expression of the transduced reporter gene in the second round (approximately 8%, see FIG. 16).

Example 7

Detection and enrichment of cells expressing antigen-reactive human antibodies by reverse transcription cell display

a) As a preparation for the conceptual evidence experiments in reverse transcriptase cell display, the optimal staining and detection conditions for retrovirally expressed IL-15 binding antibodies on selected cells were first determined. 1624-5A-MuLV transformed pre-B cells co-expressing a library of human IgH and IgL chains, which express retroviral expression vectors encoding a human anti-IL-15 antibody on the cell surface, were mixed with 1624-5A-MuLV transformed pre-B cells in a ratio of 2: 1. These mixed cell samples were incubated (as indicated) with different concentrations of recombinant human IL-15(0.1 to 2.5. mu.g/ml) and different concentrations of polyclonal, biotinylated anti-human IL-15 antibody (1.0 and 3.0. mu.g/ml), which was finally displayed as streptavidin phycoerythrin (strep-PE). To distinguish antibody-displaying cells from non-immunoglobulin expressing cells, multiple samples were counterstained with an anti-human Ig κ L-APC antibody. As can be seen in the two upper right FACS pictures of FIG. 17, IL-15 responsive cells (20.1% and 20.4%) were most efficiently detected using a combination of 0.1 and 0.5. mu.g/ml recombinant IL-15 as a first reagent and 3.0. mu.g/ml biotinylated anti-human-IL-15 antibody as a second staining reagent.

b) Next, a proof of concept experiment was performed in which a reference antibody specific for human IL-15 was spiked into a cell bank expressing a diverse human antibody library, and the spiked antigen-reactive cells were analyzed by FACS at the time of spiking. In preparation for this experiment, a library of antibodies retrovirally expressed in 1624-5 cells (see example 6) was stained for surface Ig expression and IL-15 binding (NC) using a 1624-5 cell line (PC) expressing a reference antibody specific for human IL-15 antigen. FACS analysis of these NC and PC cell lines is shown in fig. 18 and demonstrates specific IL-15 staining of the anti-IL-15 reference antibody displayed on the surface of these PC cells. To analyze whether this anti-IL-15 Ab expressing reference cell line can still be quantitatively detected by FACS (provided that the PC cells are inserted into NC cell lines expressing a random pool of human antibodies), different dilutions of PC cells in the NC library are analyzed for specific IL-15 binding by FACS using the optimal IL-15 staining conditions determined above. FACS analysis of negative control cells showed that only one population showed IL-15 binding activity. In contrast, over 60% of the positive control population was demonstrated to bind IL-15. When the above two populations were mixed in a ratio of 10%, 12.5%, 25%, 37.5% and 50%, a correlation was observed between the percentage of positive control cells mixed into the antibody library cell bank and the fraction of cells showing binding to IL-15. Therefore, it is concluded that: IL-15-reactive cells can be quantitatively detected by FACS staining in a mixture with cells expressing other non-specific antibodies.

c) Next, a proof of concept experiment was performed in which rare IL-15 responsive cells were enriched by reverse transcriptase cell display. For this purpose, an IL-15IgH chain (coupled with GFP) is transduced by retroviruses and a complex set (complexity of about 7X 10) is co-transduced⁴) The human Ig κ L chain (coupled to YFP) produced a highly diverse pool of human antibodies expressed in 1624-5 pre-B cells. Thus, a reverse-transcribed cell-displayed antibody library was generated by shuffling a diverse pool of human Ig κ L chains for a single IgH chain from a human anti-IL-15 specific antibody. These cells were stained for IL-15 reactivity using the optimal conditions as determined previously. IL-15 responsive cells were enriched by three successive rounds of high speed FACS cell sorting followed by cell culture expansion. After three rounds of reverse transcriptase cell display enrichment, a cell population can be obtained that expresses human antibodies and stains essentially quantitatively for the antigen IL-15 from an initial cell population in which IL-15 responsive cells are barely detectable. (FIG. 19) this experiment demonstrates that: repeated rounds of reverse transcriptase cell display enrichment can effectively enrich for IL-15 binding cells.

d) The IL-15 binding specificity of individual cell clones established from a 3 × IL-15 enriched cell bank was confirmed (see previous examples). Then, in this cell bank that underwent enrichment for IL-15 specific reverse transcriptase cell display three times, 25 individual cell clones have been established by single cell sorting. These 25 cell clones have been characterized for IL-15 specificity by IL-15 staining using the previously optimal staining conditions (see example 7 a). As a control for the specificity of this IL-15 staining, all clones were also incubated with all staining reagents (except the IL-15 antigen). Most of these 25 single cell clones displayed a highly specific staining pattern for IL-15, which disappeared when the IL-15 antigen was removed from the staining reaction. Representative IL-15 specific staining with 4 selected individual cell clones is shown in figure 19. The staining of these clones was representative of all 25 cell clones (named alphabetically A to Y) established from a 3 × IL-15 antigen-enriched cell population, all of which were positive for IL-15 antigen binding detection. Negative and positive controls specific for these stains are provided in fig. 19, as indicated.

Example 8

Methods of chain shuffling or directed evolution: cells expressing antigen-specific human antibodies are detected and repeatedly enriched by high-speed cell sorting, cloning of variable regions encoding regions from antigen-selected cells, and confirmation of antigen specificity.

As described above (example 6), a library of cells was generated which expressed a complexity of approximately 1.2X 10⁵In combination with the heavy chain of a reference antibody SK48-E26 directed against the target antigen human IL-1 β (Young et al, WO 95/07997A 1). The retroviral vector backbone containing these chains is depicted in more detail in FIGS. 4c and 11 (see also example 4). For this purpose, a 3X 10 transfer vector pair containing particles encoding the SK48-E26IgH chain was used⁶A1624-5A-MuLV transformed 1624-5 pre-B cell was transduced with an MOI of less than 0.1. One day after transduction, GFP positive cells were enriched by standard high-speed cell sorting using FACSAria from BD. The sorted cells were expanded by tissue culture in a humidified incubator for five days. After expansion, the cell population was transduced by standard rotational infection as described above (example 5) with particles already packaged with the Ig κ L chain library (MOI of 1.5) and the cells were allowed to recover two days from transduction by tissue culture. Two days after recovery and expansion period, 5 × 10 was enriched using FACSAria from BD using preparative cell sorting⁵A cell co-expressing constructs of GFP and YFP and thus containing at least one heavy chain and one light chain. The now enriched cell population for co-expression of IgH and IgL chain constructs was expanded by tissue culture for an additional four days. After this final expansion, 2. mu.g/ml of recombinant human IL-1. beta. (R) in a volume of 100. mu.l was used&DSystems) express IgH/IgL chains on ice2X 10 of the library⁶Aliquots of individual cells were stained for 30 minutes, followed by two washing steps using Phosphate Buffered Saline (PBS) supplemented with 1% Fetal Calf Serum (FCS). After incubation with polyclonal antibodies directed against IL-1 β and coupled with biotin, the cells were washed two more times and then stained with streptavidin-APC for detection of antigen-bound cells, and their subsequent enrichment, using flow cytometry. After a first round of cell sorting by FACS, the cells were expanded for five days and another round of anti-IL-1 β staining and enrichment of cells staining positive was performed, as explained above. This selection was repeated three times (fig. 21). As explained above, the cell populations obtained after three rounds of enrichment of the retroviral display were re-stained for IL-1 β binding, but this time the reactive cells were not enriched in large colonies as described before, but individual cell clones were sorted into 96-well plates by single cell sorting using FACSAria from BD. After seven days of culture and expansion, individual cell clones were again analyzed for IL-1. beta. antigen specificity using the scientific protocol described, and in addition using all secondary reagents (except for the antigen IL-1. beta.) as a negative control. As expected and confirmed using flow cytometry, some clones showed specific binding to the target antigen as revealed by a specific FACS signal in the presence of the antigen (but not in its absence) (excluding background binding of the clones to any of the second detection reagents). However, some clones showed a staining signal, regardless of the presence or absence of the antigen indicating non-specific binding of these clones to any of the secondary reagents used to detect IL-1. beta. reactivity. In summary, genomic DNA of 24 cell clones was isolated and served as a template for standard genomic PCR using oligonucleotides Seq-ID31 and Seq-ID32 that specifically bind upstream and downstream, respectively, to the variable regions of human light chains encoded in the retroviral light chain library.

Seq-ID31：5′-CCTTGAACCTCCTCGTTCGACCC-3′

Seq-ID32：5′-AGGCACAACAGAGGCAGTTCCAG-3′

PCR amplicons of the expected size were obtained from each analyzed cell clone and were directly subjected to DNA sequencing analysis. 12 of the 24 clones analyzed were shown to have an identical, but novel Ig kappa L chain (designated LCB24), and as expected IgH chains also containing SK48-E26, as determined separately.

As expected, all 12 clones expressing the combination of LCB24IgL chains with SK48-E26IgH chains exhibited specific IL-1 β signals, using flow cytometry, as mentioned above.

A selected PCR amplicon containing the novel LCB24Ig kappa L chain amplified was digested with restriction enzymes HindIII and Eco47III, flanked by variable coding regions (FIG. 4c), and this fragment was cloned into a retroviral Ig kappa L chain expression vector with appropriate restriction sites, allowing LCB24V_LThe coding region is fused in frame with the constant human kappa light chain coding region. Thus, the resulting vector encodes a novel, fully human Ig κ L chain.

The re-cloned and sequence-confirmed retroviral expression vector for LCB24Ig kL chain was transduced into 1624-5 cells along with SK48E26IgH chain of IL-1. beta. reference antibody SK 48-E26. After two days of expansion by tissue culture, GFP +/YFP + cells were enriched by high-speed cell sorting using FACSAria from BD. As illustrated, the resulting Ig expressing cells were first tested for their ability to bind IL-1 β. As expected, their reactivity mediated by display of LCB2 along with the heavy chain of SK48-E26 was confirmed (fig. 22). To exclude that this novel antibody is generally cross-reactive with other antigens or proteins, IL-15 reactivity was determined for cells expressing the LCB24IgL/SK48-E26IgH combination, as previously described. As depicted in FIG. 23, no reactivity to IL-15 could be detected for this novel IgLLCB24/HC SK48-E26 antibody, indicating the target antigen specificity of this novel antibody. Additional controls included a cell line expressing an anti-IL-15 specific reference antibody (as a positive control), and the original SK48-E26IL-1 β antibody. Although such cells expressing anti-IL-15 antibodies showed specific staining for IL-15 as expected, no reactivity was detected for the SK48-E26IL-1 β antibody or for these cells (FIG. 23).

In summary, a novel light chain-mediated antigen-specific reactivity was identified in a screening experiment employing a library of light chains shuffled against the heavy chain of an IL-1 β -specific reference antibody SK 48-E26.

Example 9

The shuffled IgH and IgL chain libraries are subjected to reverse transcription cell display screening. Cells expressing antigen-specific human antibodies are detected and repeatedly enriched by high-speed cell sorting, screening for variable regions encoding regions from antigen-selected cells, and confirming antigen specificity.

As described above (example 6), a library of cells was generated which expressed a library of heavy chains having approximately 6.5X 10 using an MOI of approximately 0.1⁵Complexity (coupled with GFP). The retroviral vector backbone containing these chains is depicted in more detail in FIGS. 4c and 11 (see also example 4). For this purpose, the abovementioned IgH chain library pairs which code for particle-containing transfer vectors are used in a 3X 10 manner⁶A1624-5A-MuLV transformed 1624-5 pre-B cell was transduced with an MOI of less than 0.1. 2 days after transduction, GFP positive cells were enriched by standard high-speed cell sorting using FACSAria from BD. After sorting the GFP + cells, the cells were expanded by tissue culture for two more days. After expansion, the GFP + cell population was transduced with particles already packaged with a light chain library consisting of 245 full sequences characterized by light chains with MOI greater than 1, as described above. Two days after transduction, GFP +/YFP + were double-transduced by high-speed cell sortingThe resulting cells were enriched and the cell population (now containing both IgH and IgL chain libraries in most of these cells) was expanded by tissue culture for a further three days. After this, 2.5X 10 pairs of antigen mixtures (a cocktail antibodies) comprising inter alia SAV (streptavidin) -APC-Cy7 (see example 8) as described above were used⁵Aliquots of individual cells are stained and enriched for detection of the reactivity of the target antigen using flow cytometry. In parallel, cell populations expressing antibody IgH/IgL libraries are stained with anti-IgL kappa-specific antibodies. Approximately 75% of the cells were found to display human antibodies on the cell surface (data not shown). Antigen-reactive cells have been sorted out by high-speed cell sorting using FACSAria from BD, and enriched cells are expanded by tissue culture for seven days. As explained above, the same staining and cell enrichment operations were repeated two more times. After three rounds of reverse-transcribed cell display selection, the resulting cell population was re-stained to assess binding of the target antigen SAV-APC-Cy7 and analyzed using flow cytometry. As depicted in fig. 24, the large population obtained showed binding to SAV-APC-Cy7, indicating successful selection of antibodies with antigen specificity for SAV-APC-Cy 7. To assess the specificity of reactivity against the target antigen SAV-APC-Cy7, cells expressing the triple enriched library were also stained using the antigens SAV-APC and SAV-PerCP-Cy5.5 (fig. 25). These antigens were not bound by these three SAV-APC-Cy7 enriched cells, similar to untransduced cells and cells expressing unselected libraries that served as negative controls. However, the latter cells again showed strong reactivity to SAV-APC-Cy7, indicating that the antigen specificity of these selected cell populations was directed against the Cy7 fluorescent pigment of the SAV-APC-Cy7 tandem dye.

The genomic DNA of the 3 enriched cell populations was isolated and served as a template for standard genomic PCR using oligonucleotides Seq-ID31 (see above) and Seq-ID33 that specifically bind upstream and downstream of the coding regions for the human light and heavy chains encoded in the retroviral libraries.

Seq-ID33：5′-CGGTTCGGGGAAGTAGTCCTT GAC-3′)

PCR amplicons of heavy and light chains of the expected size were obtained and subcloned separately into standard PCR fragments of a cloning vector pSC-B (Stratagene), as recommended by the manufacturer. The pSC-B plasmid with these cloned heavy and light chain regions was isolated from 10 bacterial clones, each from a subclone of the IgH PCR fragment and a subclone of the IgL PCR fragment, which were subjected to DNA sequencing analysis. DNA sequencing revealed two different IgH chain sequences (designated HC49, HC58), and two different IgL chain sequences (designated LC4 and LC 10).

Isolation of clones containing V from sequence confirmation by digestion with HindIII and Eco47III_HAnd V_LDNA fragments coding for the region, since these restriction sites are at the corresponding V of HC49, HC58, LC4 and LC10_HAnd V_LThe variable regions of the region (see FIG. 4 c). V of separated HC49, HC58, LC4 and LC10_HAnd V_LThe regions were cloned into a retroviral receptor vector with a constant region of a human Ig κ 1H chain (expressing IRES coupled to GFP) and a human Ig κ L chain (expressing IRES coupled to YFP) respectively, as described above. Thus, retroviral expression vector constructs were generated which encode novel HC49 and HC58IgH chains, LC4 and LC10IgL chains, full length human IgH and IgL chains. When these vectors were co-transduced into 1624-5 pre-B cells and expanded for 8 days, GFP +/YFP + cells were enriched using flow cytometry as previously described. As illustrated, the resulting cells were first tested for their ability to bind SAV-APC-Cy 7. As depicted in figure 26, the reactivity of cells expressing antibodies HC49/LC4 and HC/LC10 did not show significant binding activity against SAV-APC-Cy 7. In contrast, reactivity mediated by the antibodies HC58/LC4 and HC58/LC10 was readily detected. This was successful in displaying cells by reverse transcription based on diverse collections of IgH and IgL chainsThe identification of novel antigen-specific antibodies without the need for a known IgH or IgL chain from a reference antibody of known antigen specificity provides a conceptual proof.

Reference to the literature

Alt F，Rosenberg N，Lewis S，Thomas E，Baltimore D(1981)“Organizationand reorganization of immunoglobulin genes in A-MULV-transformed cells：rearrangement ofheavy but not light chain genes”Cell 27，381-390.

·Bachl J and Olsson C(1999)“HyPermutation targets a green fluorescentprotein-encoding transgene in the presence of immunoglobulin enhancers”Eur J.Immunol.29，1383-1389.

·Baker M(2005)“Upping the ante on antibodies”Nature Biotechnology 23，1065-1072.

·Boder ET，Midelfort KS，Wittrup KD(2000)“Directed evolution of antibodyfragments with monovalent femtomolar anigen-binding affinity”Proc.Natl.Acad.Sci.USA 97，10701-5.

·Clackson T，Hoogenboom H，Griffiths AD，Winter G(1991)“Making antibodyfragments using phage display libraries”Nature 352，624-628.

·Clark M(2000)“Antibody humanization：A case of the‘Emperor’s newclothes’？”Immunol.Today 21，397-402.

·Dunn IS(1995)“Assembly of functional bacteriophage lambda virionsincorporating C-terminal peptide or protein fusions with the major tail protein”J.Mol.Biol.248，497-506.

·Efimov VP，Nepluev IV，Mesyanzhinov VV(1995)“Bacteriophage T4 as asurface display vector”Virus Genes 10，173-7.

·Gossen M and Bujard H(1992)“Tight control of gene expression in vertebratecells by tetracycline-responsive promoters”Proc.Natl.Acad.Sci USA 89，5547-5551.

·Grawunder U，West RB，Lieber MR(1998)“Antigen receptor generearrangement”Curr.Opin.Immunol.10：172-180.

·Green LL and Jakobovits A(1998)“Regulation of B cell development byvariable gene complexity in mice reconstituted with human immunoglobulin yeastartificial chromosomes”J.Exp.Med.188，483-495.

·Hanes J and Plückthun A(1997)“In vitro selection and evolution of functionalproteins by using ribosome display”Proc.Natl.Acad.Sci.USA 94，4937-4942.

·Hanes J，Schaffitzel C，Knappik A & Pluckthun A(2000)“Picomolar affinityantibodies from a fully syntheticlibrary selected and evolved by ribosomedisplay”Nature Biotech.18，1287-1292.

·Hoogenboom HR and Chames P(2000)“Natural and designer binding sitesmade by phage display technology”Immunol.Today 21，371-378.

·Kitamura T，Onishi T，Kinoshita S，Shibuya A，Miyajima A，Nolan GP(1995)”Efficient screening of retroviral cDNA expression libraries”Proc.Natl.Acad.Sci.USA.92，9146-9150.

·G and Milstein C(1975)“Continuous cultures of fused cells secretingantibody of predefined specificity”Nature 256，495-497.

·Li YS，Hayakawa K，Hardy RR(1993)“The regulated expression of B lineageassociated genes during B cell differentiation in bone marrow and fetal liver”J.Exp.Med.178，951-960.

·Li M(2000)“Applications of display technology in protein analysis”NatureBiotechnology 18，1251-6.

·Lipovsek D & Pluckthun A(2004)“In-vitro protein evolution by ribosomedisplay and mRNA display”J.Immunol.Methods 290，51-67.

·Maruyama IN，Mar uyama HI，Brenner S(1994)“Lambda foo：a lambda phagevector for the expression of foreign proteins”Proc.Natl.Acad.Sci.USA 91，8273-7.

·Papavasiliou FN and Schatz DG(2002)“Somatic hyPermutation ofimmunoglobulin genes：merging mechanisms for genetic diversity”Cell 109，Suppl：S35-S44.

·Pear WS，Miller JP，Xu L，Pui JC，Soffer B，Quackenbush RC，Pendergast AM，Bronson R，Aster JC，Scott ML，Baltimore D(1998)“Efficient and rapid induction ofa chronic myelogenous leukemia-like myeloproliferative disease in mice receivingP210bcr/abl-transduced bone marrow”Blood 92，3780-92.

·Ren ZJ，Lewis GK，Wingfield PT，Locke EG，Steven AC，Black LW(1996)“Phage display of intact domains at high copy number：a system based on SOC，thesmall outer capsid protein of bacteriophage T4”Protein Sci.5，1833-43.

·Rosenberg N and Baltimore D(1978)“The effect of helper virus on Abelsonvirus-induced transformation of lymphoid cells”J.Exp.Med.147，1126-1141.

·Rosenberg A(1996)“T select phage display system：a powerful new proteindisplay system based on the bacteriophage T7”Innovations 6，1-6.

·Santini C，Brennan D，Mennuni C，Hoess RH，Nicosia A，Cortese R，LuzzagoA(1998)“Efficient display of an HCV cDNA expression library as C-terminalfusion to the capsid protein D of bacteriophage lambda”J.Mol.Biol.282，125-35.

·Shimizu T，Mundt C，Licence S，Melchers F，IL(2002)“VpreB1/VpreB2/lambda 5 triple-deficient mice show impaired B cell developmentbut functional allelic exclusion of the IgH locus”J.Immunol.168，6286-6293.

·Shinkai Y，Rathbun G，Lam KP，Oltz EM，Stewart V，Mendelsohn M，CharronJ，Datta M，Young F，Stall AM，et al.(1992)“RAG-2-deficient mice lack maturelymphocytes owing to inability to initiate V(D)J rearrangement”Cell 68，855-867.

·Smith GP(1985)“Filamentous fusion phage：hovel expression vectors thatdisplay cloned antigens on the Viron surface”Science 228，1315-17.

·Stemberg N & Hoess RH(1995)“Display of peptides and proteins on thesurface of bacteriophage lambda”Proc.Natl.Acad.Sci.USA 92(5)，1609-13.

·Stitz J，Krutzik PO，Nolan GP(2005)“Screening of retroviral cDNA librariesfor factors involved in protein phosphorylation in signaling cascades”Nucleic AcidsRes.33，e39.

·Traggiai E，Chicha L，Mazzucchelli L，Bronz L，Piffaretti JC，Lanzavecchia A，Manz MG(2004)“Development of a human adaptive immune system in cord bloodcell-transplanted mice”Science 304，104-107.

·EP 1 041 143 A：Jensen MR，Pedersen FS，Mouritzen S，HinderssonP，Duch M，Soerensen MS，Dalum I，Lund AH“A method for identification ofbiologically active peptides and nucleic acids”.

·WO 89/12823 A1：Mosier DE and Wilson DB“Human immune system innon-human animal”.

·WO 90/13660 A2：Lang AB，Larrick JW，Cryz SJ“Human monoclonalantibodies to sero-specific determinants of gram-negative bacteria”.

·WO 92/01047 A1：McCafferty J，Pope AR，Johnson KS，Hoogenboom HRJM，Griffiths AD，Jackson RH，Holliger，KP，Marks JD，Clackson TP，Chiswell DJ，Winter GP，Bonnert TP“Methods for producing members of specific binding pairs”.

·WO 92/02551 A1：Schrader JW“Methods for the production of proteins with adesired function”.

·WO 95/01997 A1：Young PR，Gross MS，Jonak ZL，Theisen TW，Hurle MR，Jackson JR“Recombinant and humanized IL-1beta antibodies for treatment of IL-1mediated inflammatory disorders in man”.

·WO 98/24893 A2：Jakobovits A，Kucherlapati R，Klapholz S，Mendez M，Green L“Transgenic mammals having human Ig loci including plural V_H and V_kregions and antibodies produced therefrom”.

·WO 02/066630 A1：Murphy AJ and Yancopoulos GD“Methods of modifyingeukaryotic cells”.

·WO 02/102855 A2：Zauderer M and Smith ES“In vitro methods for producingand identifying immunoglobulin molecules in eukaryotic cells”.

·WO 03/017935 A2：van der Winkel JGJ，van Dijk MA，Schuurman J，GerritsenAF，Baadsgaard O“Human antibodies specific for interleukin 15(IL-15)”.

·WO 03/054197A2：Perabo L，Büning H，Enssle J，Reid M，Hallek M“Alibrary of modified structural genes or capsid modified particles useful for theidentification of viral clones with desired cell tropism”.

·WO 03/068819 A1：Grawunder U and Melchers GF“Method for thegeneration of genetically modified vertebrate precursor lymphocytes and use thereoffor the production of heterologous binding proteins”.

·WO 03/083075 A2：Bremel RD，Bleck GT，Imboden M，Eakle K“Retrovirus-based genomic screening”.

·WO 04/051266 A1：Muraguchi A，Kishi H，Tamiya E，Suzuki M“Microwellarray chip for detecting antigen-specific lymphocyte，method of detectingantigen-specific lymphocyte and method of cloning antigen-specific lymphocyteantigen receptor gene”.

·WO 04/076677A2：Lanzavecchia A“Monoclonal antibody production byEBV transformation of B cells”.

·WO 04/106377Al：Lawson ADG and Lightwood DJ“Methods for producingantibodies”.

·WO 08/055795A1：Beerli R，Bachmann M，Bauer M“Selection of humanmonoclonal antibodies by mammalian cell display”.

Claims (15)

1. A method for isolating and identifying at least one nucleotide sequence encoding an antibody or fragment thereof specific for a desired antigen or ligand, said method comprising the steps of:

(a) transducing at least one retroviral expression construct encoding an antibody or fragment thereof into a vertebrate host cell by using replication incompetent retroviral particles, wherein the at least one construct is stably integrated into the host cell genome such that the transduced host cell is capable of expressing and displaying said antibody or fragment thereof on its cell surface, and wherein said vertebrate host cell is a precursor B lymphocyte that endogenously expresses lg α and Ig β molecules that promote membrane deposition of said antibody or fragment thereof and is incapable of expressing endogenous antibody polypeptides and at least one surrogate light chain component;

(b) stably expressing the antibody or fragment thereof in the vertebrate host cell and displaying it on the cell surface of the cell;

(c) selectively isolating strong antibody binders with high affinity for the desired antigen or ligand by separating cells displaying specific antigen binding from an unbound population of cells, enriching vertebrate host cells expressing the antibody or fragment thereof based on their ability to bind to the desired antigen or ligand; and is

(d) Isolating and identifying the at least one nucleotide sequence encoding the antibody or fragment thereof from the retroviral transduced and enriched vertebrate host cells.

2. The method of claim 1, wherein step (d) is preceded by expanding the enriched vertebrate host cells in tissue culture.

3. The method according to claim 1, wherein step (c) is followed by expanding the enriched vertebrate host cells in tissue culture, after which step (c) as defined in claim 1 is repeated at least once.

4. The method according to any one of claims 1 to 3, wherein such retroviral transduction is performed at a multiplicity of infection equal to or less than 0.1.

5. The method according to any one of claims 1 to 3, wherein the antibody is a full length antibody.

6. The method according to any one of claims 1 to 3, wherein the fragment of the antibody is selected from the group consisting of: heavy chain, light chain, V_HSingle domain, V_LSingle domains, scFv fragments, Fab fragments, and F (ab') 2 fragments.

7. A method according to any one of claims 1 to 3, wherein the vertebrate host cells are derived from a group of species consisting of: cartilaginous fish, teleostean, amphibian, reptilian, avian, mammalian species including porcine, ovine, bovine, equine, and rodent species including mouse, rat, rabbit, and guinea pig.

8. The method of claim 7, wherein the preferred vertebrate host cell species is mouse.

9. The method according to claim 5, wherein the full length antibody is selected from the group consisting of: fully human antibodies, humanized antibodies, and chimeric antibodies.

10. The method according to any one of claims 1 to 3, 8 and 9, wherein the at least one nucleotide sequence is a plurality of nucleotide sequences encoding (i) an antibody heavy chain sequence and a plurality of antibody light chain sequences, or (ii) an antibody light chain sequence and a plurality of antibody heavy chain sequences.

11. The method according to any one of claims 1 to 3, 8 and 9, wherein the antibody or fragment thereof comprises a variable binding domain encoded by said at least one retroviral expression construct which enables v (d) J recombination to occur, thereby producing a coding sequence for the variable binding domain upon retroviral transduction.

12. The method according to any one of claims 1 to 3, 8 and 9, wherein expression of said antibody or fragment thereof in the retroviral transduced vertebrate host cells is operably linked to

(a) At least one antibiotic selection marker, wherein the antibiotic selection marker is selected from the group consisting of,

(b) at least one screening marker, and/or linked to

(c) In combination of the above-mentioned components,

and wherein expression of said antibody or fragment thereof is coupled using at least one Internal Ribosome Entry Sequence (IRES).

13. The method according to claim 12, wherein the at least one screening marker is

(i) A fluorescent protein; or

(ii) A cell surface marker.

14. The method of claim 13, wherein the fluorescent protein is selected from the group consisting of Green Fluorescent Protein (GFP), Yellow Fluorescent Protein (YFP), Red Fluorescent Protein (RFP), and Blue Fluorescent Protein (BFP), and the cell surface marker is selected from the group consisting of CD7, CD34, and low affinity nerve growth factor receptor.

15. The method according to any one of claims 1 to 3, 8, 9, 13 and 14, wherein the enriching step (c) is performed by physically separating the cells from the non-binding cell population using the following method:

(i) fluorescence Activated Cell Sorting (FACS);

(ii) carrying out micromanipulation; or

(iii) Panning method for immobilized desired antigen or ligand.