WO2006068646A1

WO2006068646A1 - Methods for the identification and the isolation of epitope specific antibodies

Info

Publication number: WO2006068646A1
Application number: PCT/US2004/043248
Authority: WO
Inventors: Xiaoming Chen
Original assignee: Individual
Current assignee: Individual
Priority date: 2004-12-20
Filing date: 2004-12-20
Publication date: 2006-06-29
Anticipated expiration: 2007-06-20

Abstract

The present invention relates to methods for the identification of an antibody that is specific to an epitope of interest. In certain preferred embodiments, methods of the invention comprise an enrichment procedure and/or a subtractive selection that facilitates the rapid identification of an epitope specific antibody. Methods of the invention also facilitate the identification of an antibody derived from a desired species that is specific for the epitope recognized by another antibody derived from a different species. Methods of the invention further facilitate the identification of an antibody specific to an epitope found in two different antigens, and an antibody capable of binding an epitope of interest but not a homologue of the epitope.

Description

2004/043248

METHODS FOR THE roENTπ^ICATION AND THE ISOLATION OF EPITOPE SPECBFIC ANTIBODIES

BACKGROUND OF THE INVENTION Antibodies are a part of the immune system in many higher organisms. When an organism is invaded by a substance foreign to that organism (also referred to as an antigen), it typically results in the immune system generating antibodies capable of binding the antigen. These antibodies' ability to bind the antigen typically is elevated when compared to their ability to bind molecules other than the antigen. Antibodies have been recognized as useful in diagnostics, therapeutics and research, especially in view of their potential to bind a target of interest with high affinity when compared to their binding of other targets. Methods were devised to raise antibodies against a target of interest. These methods may give rise to antibodies of different kinds, for example, polyclonal antibodies or monoclonal antibodies. Polyclonal antibodies comprise a mixture of antibodies that can bind different sites or epitopes on the target of interest. Monoclonal antibodies are derived from a single cell clone and typically bind to a more limited part of a target molecule, for example, an epitope. Antibodies can be derived from different sources, for example, from mouse, rat, rabbit, hamster or human, or they may be obtained through genetic engineering.

Antibodies that can bind an epitope of interest with high affinity, while having a substantially lesser affinity for other epitopes, are of particular interest, for example, in diagnostics, therapy, or research. Moreover, antibodies with a molecular make-up of the organism in which they are used, are highly useful, especially in diagnostic and therapeutic applications. Obtaining such antibodies with high specificity, and possibly of human makeup, can be difficult as identifying and isolating such antibodies requires considerable screening efforts. Methods to more easily identify antibodies, and other molecules, that can bind epitopes of interest with high affinity when compared to other epitopes would be highly useful. The current invention provides such methods.

SUMMARY OF THE INVENTION The current invention in certain embodiments relates to methods for the identification of an antibody that is specific for an epitope of interest (epitope specific antibody or target antibody). In certain embodiments of the invention, a target antibody is identified in a mixture of antibodies by enriching the mixture for the target antibody (enrichment procedure). An enrichment procedure according to certain embodiments comprises removing undesired antibody molecules from the antibody mixture, for example, by exposing the antibody mixture to an antigen that includes the target epitope (target antigen), and then removing antibody molecules that do not bind the target antigen.

Complexes comprising the target antigen and an antibody (antigen-antibody complexes) can be separated from other components including antibody molecules, in certain embodiments, by presenting the target antigen attached to a support, for example, a virus, a phage, a cell, E. coli, yeast, a bead, a magnetic bead, a membrane, a matrix. To assist in identifying antigen-antibody complexes, the target antigen is labeled in certain embodiments, for example, with a fluorescent dye.

In certain other embodiments of the invention, the target antibody is further identified and/or isolated by separating it from antibody molecules capable of forming an antigen complex but that are not specific to the target epitope (non-target antibodies). In certain embodiments of the invention, non-target antibody molecules are separated from target antibodies by a subtractive selection procedure (subtractive selection). A subtractive selection in certain embodiments comprises exposing a target antigen to a mixture comprising target antibody molecules and non-target antibody molecules. In certain preferred embodiments, the subtractive selection procedure further comprises a marker that can bind the target epitope (marker). A marker according to certain preferred embodiments comprises any entity or molecule capable of binding the target epitope. In a subtractive selection, according to certain embodiments, the target antigen is labeled and exposed to a mixture of target and non-target antibody molecules and a labeled marker, with the label of the antigen molecules (antigen label) being distinguishable from the label of the marker (marker label). In certain embodiments, an antibody mixture and a marker are exposed to the target antigen in a subtractive selection under conditions facilitating binding of antibodies and marker to the target antigen. Under these conditions in a subtractive selection, in certain embodiments, binding of a target antibody to the target antigen will inhibit binding of the marker to the target antigen because a target antibody and the marker are both specific for the same site on the antigen, i.e., the target epitope. In certain preferred embodiments, an antibody carrying both antigen label and marker label is a non- target antibody; whereas an antibody carrying only antigen label, and no marker label, is a 2004/043248

target antibody. In another certain embodiment, antibody molecules are labeled in a way that is distinguishable from marker label and antigens may remain unlabeled so that in a subtractive selection a target antigen carrying antibody label and no marker label identifies a target antibody. In another certain embodiment, an antibody is identified that is capable of binding the antigen and the marker, for example, where the antigen and the marker both include the epitope of interest. In certain other embodiments, an antibody is identified that is capable of binding an epitope of interest but not a homologue of the epitope of interest. In certain embodiments, a homologue of an epitope can be the same epitope from a different species. In these embodiments, antibody molecules can be displayed on any support capable of displaying a plurality of antibody molecules of the same epitope specificity, for example, on cells with each cell presenting a plurality of antibody molecules, for example, antibody molecules derived from the same clone. In certain preferred embodiments, a support carrying antigen and marker is identified. In certain embodiments of the invention, a marker can be used that is a target antibody, for example, an antibody with an undesirable molecular make-up, an antibody from an undesirable species. In certain embodiments of the invention, a marker can be used that is a ligand of some type to the target epitope, for example, a growth factor, a transcription factor, a co-factor, a hormone, a lectin, a peptide. A method of the current invention, in certain embodiments, can be used to identify an antibody that modulates interactions between entities or molecules, for example, receptor-ligand interactions, protein-protein interactions during complex formation, co- factor-enzyme interactions, protein-DNA interactions, virus-cell surface receptor interactions, transcription factor-transcription factor interactions. In certain other embodiments, a method of the invention is useful to identify an epitope specific molecule (target molecule) which is capable of specifically binding a target epitope. In certain preferred embodiments, a target molecule may comprise, for example, a protein, a peptide, a small molecule, a glycoprotein.

In certain embodiments, a target antibody identified using a method of the invention can be used for diagnostics or therapeutics, for example, to detect or treat a disease caused by an interaction between entities or molecules wherein interaction can be modulated by the target antibody. BRIEF DESCRIPTION OF THE DRAWINGS

FIGURE 1. Figure 1 shows an example of a subtractive selection for identifying an epitope-specific antibody from a library. The black circles with different letters represent different yeast cells; the different Y shapes on the yeast cells represent different antibodies in the library displayed by different yeast cells; the molecule with numbered arrows represents an antigen of interest that has 4 different epitopes (binding sites); the Y shapes that are not on yeast cells (marker) represent a murine antibody that recognizes epitope 1 of the antigen. FIGURE 2. Figure 2 shows an example of a subtractive selection for identifying an antibody that inhibits receptor-ligand binding. The ligand of interest (L) (marker) is labeled and is able to interact with the receptor at binding site 1. Yeast cells displaying a library of fully human antibodies are shown as circles with a different letter for each different cell. A fully human antibody displayed by yeast cell A is able to bind to the receptor at binding site 1, therefore its binding to the receptor prevents the binding of the ligand to the receptor. Yeast cells B, C, and D display different antibodies that recognize binding site 2, 3, and 4. The number 1 binding sites in these complexes are still available for the ligand binding; therefore, a heterotrimeric complex (yeast cell-antibody-receptor-ligand) is formed that carries both antigen label and marker label (dual labeled).

FIGURE 3. Figure 3 shows an example of a subtractive selection for identifying an antibody that recognizes the same epitope shared by two antigens; or antibody specific for one antigen, but not for another homologous antigen. A human antigen of interest (#1) is labeled and is mixed with the rat counterpart of the antigen (#2) that is labeled differently. The labeled antigens are added into a suspension of yeast cells displaying a library of antibodies of any origin. Yeast cell (A) displays an antibody that recognizes the same epitope shared by both the human and rat antigen and yeast cell (A) therefore carries both antigen labels. Yeast cell (C) displaying an antibody that recognizes an epitope specific for the human antigen carries only #1 antigen label. Yeast cell (D) displays an antibody that recognizes an eptiope specific for the rat antigen carries only #2 antigen label. Sorting of dual labeled, #1 labeled or #2 labeled only will allow one to select antibodies that recognize both the human and rat antigens, or antibodies that recognize the human antigen specifically (cell C), but not the rat antigen; or antibodies that recognize the rat antigen specifically (cell D), but not the human antigen. FIGURE 4. Figure 4 shows another example of a subtractive selection for identifying a fully human antibody for an epitope of interest. The epitope of interest is grafted from a mouse antigen into the corresponding region of the human counterpart to create a human-mouse chimera antigen. The human antigen is labeled in with a red dye, the human-mouse chimera is labeled in green. Yeast cell (A) displays an antibody that only recognizes the human antigen, but not the human-mouse antigen, so the cell is red only. Yeast cell (C) displays an antibody that recognizes both human and human-mouse chimera, therefore, it is dual-color labeled. Yeast cell (D) displays an antibody that only recognizes the human-mouse antigen, but not the human antigen, therefore, this cell is green only. Antibody displayed by yeast cell (A) only recognizes the eptiope that is replaced by the mouse sequence, which is the same eptiope recognized by the murine antibody.

DETAILED DESCRIPTION OF THE INVENTION

The current invention relates to methods for the identification of an antibody. In certain embodiments, methods according to the invention facilitate the rapid identification of an antibody through a screening method of the invention. In certain preferred embodiments, an antibody that was identified using a method of the invention is capable of binding a target epitope with higher affinity than other epitopes.

Methods to Identify an Epitope-Specific Antibody.

Methods of the current invention facilitate the identification of an antibody that is specific for an epitope of interest (target epitope) (epitope specific antibody or target antibody). In certain embodiments, a method of the invention is capable of identifying an epitope specific antibody from a mixture of antibodies that are specific to different epitopes (target antibodies and non-target antibodies). A method of the current invention, in certain embodiments, identifies a target antibody by separating it from non-target antibodies. In certain other embodiments, a method of the invention identifies a target antibody by exposing a mixture of antibodies to an antigen that comprises the epitope of interest (target antigen) and by separating antibodies that are not capable of binding the antigen (enrichment procedure). Antibodies capable of binding the target antigen can be separated from non- binding antibodies, in certain embodiments, by labeling the target antigen in a way that 004/043248 facilitates the separation, for example, through a fluorescent dye, a magnetic bead, or any other means.

In certain other embodiments, a method of the invention identifies a target antibody by exposing a mixture of antibodies to a target antigen in the presence of an entity or molecule that is capable of binding the target epitope (marker) (subtractive selection). In certain preferred embodiments, the antigen molecules in the mixture of antibodies are labeled (antigen label) and the marker is also labeled (marker label), preferably, the antigen label is distinguishable from the marker label. In a subtractive selection according to certain embodiments of the invention, the marker binds to the target epitope unless a target antibody binds to the target epitope. When examining for the presence of labeling in a subtractive selection, in certain embodiments, complexes comprising the target antigen and other components can be detected by examining for antigen label and marker label. In certain preferred embodiments, the presence of antibody label and marker label in a complex suggests the presence of a non-target antibody and the marker in that complex. In certain other preferred embodiments, the presence of antigen label but not the marker label in a complex, suggests the presence of a target antibody in the complex. In certain preferred embodiments, an antibody from a complex comprising antigen label but no marker label is further isolated and characterized. In certain other embodiments, an antigen capable of binding the antigen and the marker are identified, for example, by identifying a support with a plurality of anibody molecules with the same, or substantially the same, epitope specificity that carries antigen label and marker label.

A method of the current invention, in certain embodiments, comprises 1, 2, 3, 4, 5, 6, 7, 8, 9, or 10 rounds of an enrichment procedure. Where a plurality of rounds of an enrichment procedure are applied, one or more rounds of enrichment procedure may differ in their execution from another one or more rounds. A method of the current invention, in certain other embodiments, comprises 1, 2, 3, 4, 5, 6, 7, 8, 9, or 10 rounds of a subtractive selection. Where a plurality of rounds of a subtractive selection are applied, one or more rounds of subtractive selection may differ in their execution from another one or more rounds. In certain other embodiments, a method of the current invention comprises an enrichment procedure and a subtractive selection. Certain embodiments of the methods of the invention comprise one or more rounds of an enrichment procedures and one or more rounds of a subtractive selection. In certain embodiments, a method of the invention comprises I₃ 2, 3, 4, 5, 1-3, 2-4, or 1-5 rounds of an enrichment procedure and I₃ 2, 3, 4, 5, 1-3, 2-4, or 1-5 rounds of a subtractive selection. In certain other embodiments, a method of the invention comprises 1, 2, or 3 rounds of an enrichment procedure, followed by 1, 2, or 3 rounds of a subtractive selection, which may be followed by 1, 2, or 3 rounds of an enrichment procedure, which may be followed by 1, 2, or 3 rounds of a subtractive selection. In certain preferred embodiments, a method of the invention comprises 1 or 2 rounds of an enrichment procedure followed by 1 or 2 rounds of a subtractive selection.

Enrichment Procedures A target antibody or molecule can be identified using a method of the invention, in certain embodiments, by enriching a mixture of antibodies for a target antibody. An antibody mixture contains antibodies specific to a plurality of epitopes and an enrichment procedure, in certain embodiments, increases the fraction of antibody molecules in the mixture that are specific to the epitope of interest. In certain preferred embodiments, an enrichment procedure operates by decreasing the number of non-target antibody molecules in the mixture, or by increasing the number of target antibody molecules in the mixture, or by both.

An enrichment procedure, in certain embodiments, separates target and non-target antibody molecules through the target antibodies' ability to bind to an antigen that comprises the epitope of interest (target antigen). Separating target and non-target antibodies can result, in certain embodiments, in an increase in the fraction of target antibodies in the antibody mixture by a percentage that is useful for the identification of target antibodies, for example, an increase of 10 percent or more, 25 percent or more, 50 percent or more, 100 percent or more, 200 percent or more, 300 percent or more, 500 percent or more, 10 to 300 percent, 25 to 200 percent, or 50 to 100 percent. The target antigen can be used to enrich for target antibodies, in certain embodiments, by binding, reversibly or irreversibly, the target antigen or the antibody molecules in the antibody mixture to a solid support. In certain other embodiments, antibody molecules capable of binding the target antigen are isolated by removing unbound antibody molecules or by removing complexes of target antigen and antibody molecules. In certain embodiments, magnetic beads are used as a solid support in combination with a magnet to remove the component carrying the beads through capture by a magnet. In a preferred embodiment, the antigen carries a magnetic bead and an antibody bound to the antigen carrying the magnetic bead will be captured by a magnet. Or, for example, the antigen can be labeled, for example with a fluorescent dye, and exposed to the antibody. Antigen-antibody complexes may be identified, for example, by FACS, or by binding the antigen molecules to a solid support so that only antibody molecules capable of binding the antigen are retained on the solid support. Variations for enriching target antibodies in an antibody mixture through binding to a target antigen can easily designed be in view of the description provided herein and are within the scope of this invention.

An enrichment procedure, in certain embodiments, comprises exposing an antibody mixture to a target antigen. When the antibody mixture and the target antigen are combined, in certain embodiments, the concentration of both components is preferably such that the number of antibody molecules and antigen molecules is about equal. In certain other embodiments, the number of antigen molecules is higher than the number of antibody molecules, for example, 25 percent higher or up to 25 percent higher, 50 percent higher or up to 50 percent higher, 100 percent higher or up to 100 percent higher, 200 percent higher or up to 200 percent higher, 500 percent higher or up to 500 percent higher, or 1000 percent higher or up to 1000 percent higher. In certain other embodiments, the number of antigen molecules is lower than the number of antibody molecules, for example, 25 percent lower or up to 25 percent lower, 50 percent lower or up to 50 percent lower, 100 percent lower or up to 100 percent lower, 200 percent lower or up to 200 percent lower, 500 percent lower or up to 500 percent lower, or 1000 percent lower or up to 1000 percent lower. The concentration of the antigen relative to the antibody, in certain embodiments, is chosen based on how the binding reaction should be driven. For example, where an antibody mixture likely has a high concentration of antibody molecules that can bind the antigen with high affinity, then it would be desirable to use a lower antigen concentration in the enrichment procedure so that only those antibody molecules are identified that bind the antigen with the highest affinity, and vice versa. Whether an antibody mixture is likely to include a high concentration of antibody molecules that can bind the antigen depends on how related the origins of antigen and antibodies are, how close to a naturally occurring molecule the antigen is, how large the antigen is, and whether the antigen would be expected to evoke a strong immune response. Exposing an antibody mixture to a target antigen, in certain embodiments, is carried out by combining both in a solvent that facilitates the binding of antibodies to an antigen, for example, a buffer solution, a medium, water, a salt solution, an inorganic solvent, an organic solvent, or any other suitable solvent. When the antibody mixture and antigen are combined in the solvent, the conditions of protein concentration, salt concentration, pH, temperature, and other conditions, are selected to facilitate the binding of an antibody to an antigen, for example, conditions resembling physiological conditions under which antibodies bind an antigen, or any conditions known to facilitate such binding.

A target antigen that is combined with an antibody mixture, in certain preferred embodiments, is labeled or marked in a way that facilitates the identification of binding complexes comprising the antigen and an antibody. Examples of such labeling or marking are a fluorescent dye, a magnetic bead, a tag, an antibody tag, or any other label or mark that facilitates the identification of the binding complexes of the antigen and an antibody. In certain embodiments, the complexes of the antigen and an antibody molecule can be separated or enriched by any method capable of recognizing the chosen label, for example, by fluorescence activated cell sorting (FACS) when using a label that is useful for FACS, or by employing a magnet when using a magnetic bead label. In certain other embodiments, the antigen is attached to a solid support and the antigen is exposed to the antibody mixture while bound to the support. Preferably, the antigen is removed from the solution in which the antigen is exposed to the antibody mixture so that the antigen and antibody molecules bound to the antigen are separated from antibody molecules that did not bind the target antigen. In certain other embodiments, the target antigen is combined with a mixture of labeled antibodies and target antigen molecules bound to an antibody are further isolated.

Subtractive Selection Methods

A target antibody can be identified, in certain embodiments, through a subtractive selection. A subtractive selection, in certain embodiments, facilitates the identification of a target antibody or an antibody mixture with a high fraction of target antibody molecules. A subtraction selection comprises, in certain embodiments, exposing an antibody mixture to a target antigen and to a molecule capable of selectively binding the target epitope on the target antigen (marker). In certain preferred embodiments, a target antibody in the antibody mixture competes with the marker for binding to the target epitope. In certain other preferred embodiments, a target antibody will bind the target epitope only if its affinity for the target epitope is sufficient to prevent binding of the marker to, or displace bound marker from, the target epitope, and preferably to a degree so that a sufficient number of target antibody molecules bind the target epitope with sufficient strength to facilitate their identification. For example, where the affinity of a target antibody and the marker for the target epitope are about equal, each would bind about half of the target epitopes in the subtractive selection, provided no other component would compete for such binding and provided that an about equal number of antigen and marker molecules were present. The desired affinity of an identified target antibody when compared to a marker used in the subtractive selection can be scaled, preferably by adjusting the quantity of the marker in the subtractive selection. For example, if one desires to identify a target antibody with higher affinity for the target epitope, a higher amount of marker would preferably be included in the subtractive selection and vice versa. Also, in antibody mixtures, according to certain embodiments, target antibody molecules with different molecular characteristics can be found, for example, target antibody molecules with different affinities for the target epitope and different affinities for a non-target epitope. By using a high concentration of the marker in a subtractive selection, in preferred embodiments, target antibodies with high specificity and high affinity for the target epitope can be identified.

A subtractive selection, in certain embodiments, comprises combining about an equal number of target antigen molecules, marker molecules and antibody molecules in an antibody mixture. When the antigen, marker and antibody mixture are combined, in certain embodiments, the number of antigen and marker molecules separately is higher than the number of antibody molecules, for example, 25 percent higher or up to 25 percent higher, 50 percent higher or up to 50 percent higher, 100 percent higher or up to 100 percent higher, 200 percent higher or up to 200 percent higher, 500 percent higher or up to 500 percent higher, or 1000 percent higher or up to 1000 percent higher. In certain other embodiments, the number of antigen and marker molecules separately is lower than the number of antibody molecules, for example, 25 percent lower or up to 25 percent lower, 50 percent lower or up to 50 percent lower, 100 percent lower or up to 100 percent lower, 200 percent lower or up to 200 percent lower, 500 percent lower or up to 500 percent lower, or 1000 percent lower or up to 1000 percent lower. When the antigen, marker and antibody mixture are combined, in certain other embodiments, the number of antigen molecules is higher than the number of marker molecules, for example, 25 percent higher or up to 25 percent higher, 50 percent higher or up to 50 percent higher, 100 percent higher or up to 100 percent higher, 200 percent higher or up to 200 percent higher, 500 percent higher or up to 500 percent higher, or 1000 percent higher or up to 1000 percent higher. In certain other embodiments, the number of antigen molecules is lower than the number of marker molecules, for example, 25 percent lower or up to 25 percent lower, 50 percent lower or up to 50 percent lower, 100 percent lower or up to 100 percent lower, 200 percent lower or up to 200 percent lower, 500 percent lower or up to 500 percent lower, or 1000 percent lower or up to 1000 percent lower. A subtractive selection, in certain embodiments, comprises combining labeled antigen in an antibody mixture. When the antigen and antibody mixture are combined, in certain embodiments, the number of antigen is higher than the number of antibody molecules, for example, 25 percent higher or up to 25 percent higher, 50 percent higher or up to 50 percent higher, 100 percent higher or up to 100 percent higher, 200 percent higher or up to 200 percent higher, 500 percent higher or up to 500 percent higher, or 1000 percent higher or up to 1000 percent higher. In certain other embodiments, the number of antigen is lower than the number of antibody molecules, for example, 25 percent lower or up to 25 percent lower, 50 percent lower or up to 50 percent lower, 100 percent lower or up to 100 percent lower, 200 percent lower or up to 200 percent lower, 500 percent lower or up to 500 percent lower, or 1000 percent lower or up to 1000 percent lower. In certain embodiments, after incubation of the labeled antigen with an antibody mixture in a solvent the facilitates antigen-antibody interaction for a period of time, such as 10 minutes, 20 minutes, 30 minutes, 1 hour, 2 hours, 4 hours, up to overnight, unbound antigens are washed off with a wash buffer which can be the same as the binding buffer or another buffer. In certain other embodiments, the bound antigen-antibody complexes are preferably kept on ice to prevent the dissociation of the bound antigen from the antibody. To these antigen-antibody complexes, in certain embodiments, a marker is added. Preferably the marker is labeled and preferably the marker label is distinguishable from a label used on, for example, the antigen. In certain embodiments, the number of marker molecules or entities added is higher than the number of antigen molecules, for example, 25 percent higher or up to 25 percent higher, 50 percent higher or up to 50 percent higher, 100 percent higher or up to 100 percent higher, 200 percent higher or up to 200 percent higher, 500 percent higher or up to 500 percent higher, or 1000 percent higher or up to 1000 percent higher. In certain embodiments, the marker is incubated with the antigen-antibody mixture in a solvent that facilitates marker- antigen interaction for a period of time, for example, for 10 minutes, 20 minutes, 30 minutes, 1 hour, 2 hours, 3 hours, 5 hours, and preferably up to 1 hour on ice. In certain other embodiments, marker molecules that are not bound to an antigen molecule are washed off with a wash buffer which can be the same as the binding buffer or another buffer.

A subtractive selection, in certain embodiments, comprises exposing an antibody mixture to a target antigen and a marker. Exposing an antibody mixture to a target antigen and a marker, in certain embodiments, is carried out by combining those components in a solvent that facilitates the binding of antibodies and of the marker to an antigen, for example, a buffer solution, a medium, water, a salt solution, an inorganic solvent, an organic solvent, or any other solvent for antibody binding assays. Examples of buffers are a PBS, a phosphate buffer, a citrate buffer, a carbonate buffer, or any other buffer suitable for antibody binding assays. Examples of media are any medium useful for growing or maintaining cells, bacteria, viruses, or phages. Examples of a salt solution are a solution comprising any salt suitable for antibody binding assays (e.g., sodium salts, calcium salts, magnesium salts, potassium salts, salts comprising chloride, sulfate, nitrate). When the antibody mixture, the marker, and the antigen are combined in the solvent, the conditions of protein concentration, salt concentration, pH, temperature, and other conditions, are selected to facilitate the binding of an antibody and the marker to an antigen, for example, conditions resembling physiological conditions under which antibodies bind an antigen or, if applicable, conditions under which the marker binds the antigen, or any other conditions known to facilitate such binding.

A target antigen and a marker that are combined with an antibody mixture, in certain preferred embodiments, are labeled or marked in a way that facilitates the identification of binding complexes comprising the antigen and an antibody, and the identification of binding complexes comprising the antigen, an antibody and the marker. The labeling or marking further facilitates, in certain preferred embodiments, distinguishing between antigen- antibody complexes and antigen-antibody-marker complexes. In certain preferred embodiments, two labels are used in a subtraction selection, one preferably for the antigen (antigen label) and another for the marker (marker label). In certain other embodiments, the antibody molecules in the antibody mixture are labeled (antibody label) and the marker is labeled, but not the antigen. Examples of labeling or marking useful for subtraction selection are a fluorescent dye, a magnetic bead, a tag, an antibody tag, or any other label or mark that facilitates the identification of the different binding complexes. In certain embodiments, the antigen-antibody complexes that do not include the marker can be separated or enriched by any method capable of recognizing the labels chosen for the subtraction selection, for example, by fluorescence activated cell sorting (FACS), for example, by using two different fluorescent labels useful for FACS and that facilitate distinguishing complexes with and without the marker.

In certain embodiments, a target antibody is identified in a subtraction selection by presenting the antibody mixture bound to a solid support and by identifying antibody- antigen complexes on the support. A support suitable for subtraction selection, in certain embodiments, is a cell, a bacterium, a virus, a phage, a membrane, a magnetic bead, agar, or any other support that facilitate the identification and isolation of a target antibody. In certain other embodiments, the antigen is attached to a solid support and the antigen is exposed to the antibody mixture and the marker while bound to the support. Preferably, the antigen is removed from the solution in which the antigen is exposed to the antibody mixture and the marker so that antibody-antigen complexes are separated from remaining components. In certain other embodiments, the target antigen is combined with a mixture of labeled antibodies and differently labeled marker, and target antigen molecules bound to an antibody, but not the marker, are further isolated.

Identifying An Epitope Specific Antibody.

In certain embodiments, an antibody specific for an antigen of interest can be identified using the methods of the invention. For example, a molecule (A) that contains an eptiope of interest is labeled with a color, such as fluorescent dye in red. A 2^nd molecule (B) that interacts with molecule (A) is labeled with a different color, such as green. A library is constructed to display antibodies, proteins, peptides, or small molecules. The vehicle used to display these molecules can be phage, virus, E. coli, yeast, cells, beads, or any matrix. Small displaying vehicle can in turn linked to particles that are suitable for magnetic separation and FACS. Yeast cells displayed antibody library will be used as an example in the description below. Dye labeled (red), or magnetic bead conjugated molecule (A) is incubated with the library to be screened in a binding buffer. After washing off unbound molecule (A), yeast cells displaying antibodies that are able to bind to the labeled molecule A can be selected and enriched from the rest of library by means such as fluorescent-activated cell sorting (FACS), since these yeast cells upon binding to molecule (A) would be red; or by magnetic bead separation using a magnet. The selected yeast cells can be grown up, induced to express antibodies, and used for the next round of selection.

Magnetic beads separation can be used to quickly separate molecule (A) bound cells from a very large library, so it is preferred method for the first 1-2 rounds selection, if screening of a very large library is required. After the first 1-2 rounds of enrichment, a subtractive selection is incorporated into the selection process. After mixing the labeled molecule (A) with the selected yeast cells, and washing away unbound molecule (A), a different color (i.e. green) labeled molecule (B) is added to the mixture. If an antibody displayed by some of the selected yeast cells binds to the same epitope recognized by molecule (B), molecule (B) will not bind to molecule (A) that already bound to the antibody on the yeast, since the epitope on molecule (A) for molecule (B) is already occupied by the displayed antibody, the resulted yeast cell-antibody-molecule (A) complex should be red only.

On the other hand, if an antibody binds to molecule (A) at a different epitope from that recognized by molecule (B), molecule (B) will be able to bind molecule (A), even molecule (A) has formed a complex with the antibody, since the eptiope on molecule (A) for molecule (B) is not occupied by the antibody displayed by the yeast cells. The resulting antibody-molecule (A)-molecule (B) complex will be dual colored (i.e. green and red). Yeast cells in red only [i.e. yeast cell-antibody-molecule (A)] can be separated from those that are dual colors [i.e. yeast cell-antibody-molecule (A)-molecule (B)] by FACS.

Antibody isolated from the yeast cells in red only therefore is able to bind to molecule (A) at the same or overlapping binding site, or epitope, recognized by molecule (B). The isolated antibody can then be used as diagnostic or therapeutic to detect or treat diseases that are caused by the interaction between molecule (A) and molecule (B). The described method can also be used to select for a fully human antibody that binds to an epitope recognized by an antibody with a different specie origin (i.e, murine antibody) from a fully human antibody library. In this case, the antigen for the murine antibody is molecule (A) and the murine antibody is molecule (B) as described above. One such example is described in more detail below (for example, Example 1) and is illustrated in Figure 1.

Identifying An Antibody That Inhibits Molecule-Molecule Interactions.

In certain embodiments, an antibody can be identified that can inhibit an interaction between molecules of interest using the methods of the invention. For example, the current invention provides methods for selecting antibodies, proteins, peptides, or small molecules that inhibit protein-protein, ligand-receptor interactions (for example, Example 2 and as illustrated in Figure 2). In certain embodiments, such protein-protein interactions include interactions of any proteins found in nature, genetically engineered proteins, animal proteins, plant proteins, primate proteins, cytosolic proteins, nuclear proteins, membrane proteins, receptors, growth factors, enzymes, transcription factors, DNA binding proteins, regulatory proteins, bacterial proteins, viral proteins, soluble proteins.

Identifying An Antibody That Recognizes An Epitope Found In Another Protein.

In certain embodiments, the invention also describes methods for selecting antibodies, proteins, peptides, or small molecules that recognize an epitope also found in another protein, for example a homologous protein from a different species, and for selecting antibodies, proteins, peptides, or small molecules that only recognize a molecule from one specie, but not from another, or recognize a particular member molecule in a protein family, but not another member in the same family (for example, Example 3 and as illustrated in Figure 3).

Most antibodies derived from animals are directed against antigens from other hosts and do not recognize antigens from the animals from which the antibodies are produced, in other words, the antibodies do not recognize a self-antigen. The species-specificity of antibodies has been a limitation on the evaluation of antibodies in animals with established disease models. In many cases, surrogate antibodies are used to evaluate the efficacy in the animal models. However, these surrogate antibodies may or may not recognize the same epitope that is recognized by antibodies to be developed as therapeutics. Therefore predicting the actual efficacy for the antibodies in development based on these preclinical data is difficult. Obtaining antibodies that are able to recognize antigens from human and animals with established disease models, therefore, will be very useful. The methods of the current invention thus facilitate using the same antibodies in human trials that are evaluated in animal models.

Identifying An Antibody That Recognizes A Known Epitope In An Antigen.

The current invention also describes methods for selecting antibodies that recognize a known epitope present in an antigen (for example, Example 4 and as illustrated in Figure 4).

Antibodies

In certain embodiments, an antibody identified using a method of the invention is antigen specific. An antigen specific antibody, in certain preferred embodiments, is capable of specifically binding the antigen, for example, by binding the antigen with a certain high affinity or higher affinity, or, for example, by not significantly cross-reacting with a epitopes that are related to the antigen of interest. A certain high affinity, in certain embodiments, is a binding affinity (K_a) of 10⁵ M^"1 or greater, preferably 5 x 10⁵ M^"1 or greater, preferably 10⁶ M^"1 or greater, preferably 5 x 10⁶ M^"1 or greater, preferably 10⁷ M^"1 or greater, preferably 5 x 10⁷ M^"1 or greater, more preferably 10⁸ M^"1 or greater, preferably 5 x 10⁸ M^"1 or greater, preferably 10⁹ M^"1 or greater, and preferably 5 x 10⁹ M^'1 or greater. The binding affinity of an antibody can be determined through routine analysis, for example, by Scatchard analysis described in Munson et ah, Anal Biochem., 107:220 (1980). Antibodies from any origin may be identified using a method of the invention, for example, from horse, cow, dog, chicken, rat, mouse, rabbit, guinea pig, goat, or sheep.

A method of the invention can be used to identify a target antibody of any kind, including, but not limited to, polyclonal, monoclonal, multispecific, human, humanized or chimeric antibodies, anti-idiotype antibodies, bispecific, trispecific, or multispecific antibodies, immuno-conjugates, single chain antibodies, Fab fragments, F(ab') fragments, fragments produced by a Fab expression library, F(ab')2 fragments, Fd fragments, single- chain Fvs (scFv), disulfide-linked Fvs (sdFv), fragments comprising either a VL or VH domain, anti-idiotypic (anti-Id) antibodies (including, e.g., anti-Id antibodies to antibodies of the invention), and epitope-binding fragments of any of the above. In certain embodiments, an antibody identified using a method of the invention is an immunoglobulin molecule or an immunologically active portion of an immunoglobulin molecule with an antigen binding site that is capable of specifically binding a suitable antigen. The immunoglobulin molecules of the invention can be of any type (e.g., IgG, IgE, IgM, IgD, IgA and IgY)₅ class (e.g., IgGl, IgG2, gG3, gG4, gAl and IgA2) or subclass of immunoglobulin molecule. Methods of the invention are useful for identifying an epitope specific antibody from a mixture of antibodies that includes antibody molecules specific to a plurality of epitopes. A mixture of antibodies, in certain embodiments, may be generated using a method known in the art. For example, polyclonal antibodies may be prepared using methods well-known to those of skill in the art. See, for example, Green et al., "Production of Polyclonal Antisera," in Immunochemical Protocols (Manson, ed.), pages 1-5 (Humana Press 1992), and Williams et al., "Expression of foreign proteins in E. coli using plasmid vectors and purification of specific polyclonal antibodies," in DNA Cloning 2: Expression Systems, 2nd Edition, Glover et al. (eds.), page 15 (Oxford University Press 1995). The immunogenicity of an antigen can be increased through the use of an adjuvant, such as alum (aluminum hydroxide) or Freund's complete or incomplete adjuvant.

Or, for example, monoclonal antibodies may be raised as known in the art (see, for example, Kohler et al., Nature 256:495 (1975), Coligan et al. (eds.), Current Protocols in Immunology, Vol. 1, pages 2.5.1-2.6.7 (John Wiley & Sons 1991) (Coligan), Picksley et al., "Production of monoclonal antibodies against proteins expressed in E. coli," in DNA Cloning 2: Expression Systems, 2nd Edition, Glover et al. (eds.), page 93 (Oxford

University Press 1995)). Monoclonal antibodies can also be obtained from hybridoma cultures by a variety of established techniques. Suitable isolation techniques include affinity chromatography with Protein-A Sepharose, size-exclusion chromatography, and ion- exchange chromatography (see, for example, Coligan at pages 2.7.1-2.7.12 and pages 2.9.1- 2.9.3; Baines et al., "Purification of Immunoglobulin G (IgG)," in Methods in Molecular Biology, Vol. 10, pages 79-104 (The Humana Press, Inc. 1992)).

Antibody fragments, for example, may be generated, for example, by proteolytic hydrolysis of the antibody. Antibody fragments can be obtained by pepsin or papain digestion of whole antibodies by conventional methods. As an illustration, antibody fragments can be produced by enzymatic cleavage of antibodies with pepsin to provide a 5 S fragment denoted F(ab')₂. This fragment can be further cleaved using a thiol reducing agent to produce 3.5 S Fab' monovalent fragments. Optionally, the cleavage reaction can be performed using a blocking group for the sulfhydryl groups that result from cleavage of disulfide linkages. As an alternative, an enzymatic cleavage using pepsin produces two monovalent Fab fragments and an Fc fragment directly. These methods are described, for example, by Goldenberg, U.S. Pat. No. 4,331,647, Nisonoff et al., Arch Biochem. Biophys. 89:230 (1960), Porter, Biochem. J. 73: 119 (1959), Edelman et al., in Methods in

Enzymology Vol. 1, page 422 (Academic Press 1967), and by Coligan at pages 2.8.1-2.8.10 and 2.10.-2.10.4.

A peptide coding for a single complementarity-determining region (CDR) is another example of an antibody fragment. CDR peptides (minimal recognition units) can be obtained, for example, by constructing genes encoding the CDR of an antibody of interest. Such genes are prepared, for example, by using the polymerase chain reaction to synthesize the variable region from RNA of antibody-producing cells (see, for example, Larrick et al., Methods: A Companion to Methods in Enzymology 2: 106 (1991), Courtenay-Luck, "Genetic Manipulation of Monoclonal Antibodies," in Monoclonal Antibodies: Production, Engineering and Clinical Application, Ritter et al. (eds.), page 166 (Cambridge University Press 1995), and Ward et al., "Genetic Manipulation and Expression of Antibodies," in Monoclonal Antibodies: Principles and Applications, Birch et al., (eds.), page 137 (Wiley- Liss, Inc. 1995)).

Fv fragments comprise an association of VH and VL chains and may be prepared, for example, as known in the art. These single-chain antigen binding proteins (scFv) can be generated, for example, by constructing a mixture of structural genes comprising DNA sequences encoding the VH and VL domains that may be connected by an oligonucleotide. The structural genes may be inserted into an expression vector, which may be introduced into a host cell, such as E. coli. Preferably, each recombinant host cell synthesizes a single polypeptide chain with a linker peptide bridging the two V domains. Methods for producing scFvs are described, for example, by Whitlow et al., Methods: A Companion to Methods in Enzymology 2:97 (1991) {also see, Bird et al., Science 242:423 (1988), Ladner et al., U.S. Pat. No. 4,946,778, Pack et al., Bio/Technology 11: 1271 (1993), and Sandhu, supra). Polyclonal anti-idiotype antibodies, for example, can be prepared by immunizing animals with antibodies or antibody fragments specific for an antigen or epitope of interest as known in the art. See, for example, Green et al., "Production of Polyclonal Antisera," in Methods In Molecular Biology: Immunochemical Protocols, Manson (ed.), pages 1-12 (Humana Press 1992). Also, see Coligan at pages 2.4.1-2.4.7. Methods for producing antiidiotype antibodies are described, for example, by Irie, U.S. Pat. No. 5,208,146, Greene, et. al., U.S. Pat. No. 5,637,677, and Varthakavi and Minocha, J. Gen. Virol. 77:1875 (1996).

Antibody Libraries

In certain embodiments, an antibody library is used in a method of the invention. An antibody library is a collection of antibodies, preferably one from which one may identify a target antibody using the disclosed methods. In certain preferred embodiments, an antibody library facilitates the identification of a target antibody in the disclosed methods, for example, by presenting the antibodies on a support. Examples of a support useful for presenting an antibody library are cells, a gel, bacteria, viruses, phages, a filter, a glass plate, glas beads, a matrix, or any other support, or any technology useful to display antibodies and that facilitates presenting antibodies for identification of a target antibody using the disclosed methods.

Examples of libraries include, but are not limited to, antibody libraries constructed using cDNA amplified from mRNA from human spleen or B cells whom have been immunized, non-immunized, or with auto-immune diseases by using immunoglobulin heavy and light chain specific primers. Or, for example, antibody libraries can also be constructed using synthetic heavy and light chains with randomized CDRs. Antibody libraries can be expressed, for examle, as scFv, Fab, hAb (E.g. , half antibody, the two cystines in the hinge region of the heavy chain are mutated so that the heavy chain is not able to pair with another heavy chain to form a full IgG antibody, only one heavy chain and one light chain complex will be displayed. This complex is only half of the full IgG antibody.), or full IgG molecules. Or, for example, libraries can also be comprised of cDNAs from cells of any species, such human, mouse, rat, hamster, rabbit or any other species expressing antibodies. Libraries can also be comprised, for example, of random peptides with or without disulfide bond linkage, or of protein fragments, or of small molecules, or of molecules, or combinations of different type of molecules. In certain preferred embodiments, libraries are displayed on the surface of a support, such as virus, phage, E. CoIi, yeast, mammalian cell, or any type of matrix. In certain embodiments, an antibody library is expressed in an expression system, including, but not limited to, microorganisms such as bacteria (e.g., E. coli, B. subtilis) transformed with recombinant bacteriophage DNA, plasmid DNA or cosmid DNA expression vectors containing antibody coding sequences; yeast (e.g., Saccharomyces cerrevision) transformed with recombinant yeast expression vectors containing antibody coding sequences; insect cell systems infected with recombinant virus expression vectors (e.g., baculovirus) containing antibody coding sequences; plant cell systems infected with recombinant virus expression vectors (e.g., cauliflower mosaic virus, CaMV; tobacco mosaic virus, TMV) or transformed with recombinant plasmid expression vectors (e.g., Ti plasmid) containing antibody coding sequences; or mammalian cell systems (e.g., COS, CHO, BHK, 293, 3T3 cells) harboring recombinant expression constructs containing promoters derived from the genome of mammalian cells (e.g., metallothionein promoter) or from mammalian viruses (e.g., the adenovirus late promoter; the vaccinia virus 7.5K promoter). In certain preferred embodiments, bacterial cells such as Escherichia coli, and more preferably, eukaryotic cells, especially for the expression of whole recombinant antibody molecule, are used for the expression of antibodies. For example, mammalian cells such as Chinese hamster ovary cells (CHO), in conjunction with a vector such as the major intermediate early gene promoter element from human cytomegalovirus is an effective expression system for antibodies (Foecking et al., Gene 45:101 (1986); Cockett et al., Bio/Technology 8:2 (1990)).

Where antibodies are expressed in cells, in certain embodiments, they may be grown as a collection or library of different cell clones. The different clones in the library may be screened for a target antibody using the methods of the invention. Where a clone of interest is identified, the antibodies secreted by the cells may be isolated from the culture medium, ascites fluid, or serum by conventional antibody purification procedures, for example, by affinity chromatography (e.g., using protein A or protein G-Sepharose) or ion-exchange chromatography, hydroxylapatite chromatography, gel electrophoresis, dialysis, etc. Also, DNA encoding the antibodies is readily isolated and sequenced using conventional procedures (e.g., by using oligonucleotide probes that are capable of binding specifically to genes encoding the heavy and light chains of murine antibodies). The cells expressing antibodies, in certain embodiments, serve as a preferred source of such DNA. Once isolated, the DNA may be placed into expression vectors, which are then transformed into host cells such as E. coli cells, or are transfected into simian COS cells, Chinese Hamster Ovary (CHO) cells, or myeloma cells that do not otherwise produce antibody protein, to obtain the synthesis of antibodies in the recombinant host cells. Review articles on recombinant expression in bacteria of DNA encoding the antibody include Skerra et al., Curr. Opinion in Immunol., 5:256-262 (1993) and Pluckthun, Immunol. Revs., 130:151-188 (1992).

In a further embodiment, antibodies or antibody fragments can be isolated from antibody phage libraries generated using the techniques described in McCafferty et al., Nature, 348:552-554 (1990). Clackson et al., Nature, 352:624-628 (1991) and Marks et al., J. MoI. Biol., 222:581-597 (1991) describe the isolation of murine and human antibodies, respectively, using phage libraries. Subsequent publications describe the production of high affinity (nM range) human antibodies by chain shuffling (Marks et al., Bio/Technology, 10:779-783 (1992)), as well as combinatorial infection and in vivo recombination as a strategy for constructing very large phage libraries (Waterhouse et al., Nuc. Acids. Res., 21:2265-2266 (1993)). The DNA that encodes the antibody, in certain embodiments, may be modified to produce chimeric or fusion antibody polypeptides, for example, by substituting human heavy chain and light chain constant domain (C_H and C_L) sequences for the homologous murine sequences (U.S. Pat. No. 4,816,567; and Morrison, et al., Proc. Natl. Acad. Sci. USA, 81:6851 (1984)), or by fusing the immunoglobulin coding sequence with all or part of the coding sequence for a non-immunoglobulin polypeptide (heterologous polypeptide). The non-immunoglobulin polypeptide sequences can substitute for the constant domains of an antibody, or they are substituted for the variable domains of one antigen-combining site of an antibody to create a chimeric bivalent antibody comprising one antigen-combining site having specificity for an antigen and another antigen-combining site having specificity for a different antigen.

Antibody molecules in an antibody mixture screened using a method of the invention, in certain embodiments, may be attached to a polypeptide tag (or tag) to facilitate identification of a target antibody and to provide sites for attachment of the antibody molecules to a support. In principal, any peptide or protein for which an antibody or other specific binding agent is available can be used as a tag. Affinity tags include a poly-histidine tract, protein A (Nilsson et al., EMBO J. 4:1075 (1985); Nilsson et al; Methods Enzymol. 198:3 (1991)), glutathione S transferase (Smith and Johnson, Gene 67:31 (1988)), Glu-Glu affinity tag (Grussenmeyer et al., Proc. Natl. Acad. Sci. USA 82:7952 (1985)), substance P, FLAG peptide (Hopp et al., Biotechnology 6:1204 (1988)), streptavidin binding peptide, or other antigenic epitope or binding domain. See, in general, Ford et al. , Protein Expression and Purification 2:95 (1991). DNAs encoding affinity tags are available from commercial suppliers (e.g., Pharmacia Biotech, Piscataway, N. J.).

Epitope Specific Molecules and Libraries thereof.

In certain embodiments, molecules with epitope specific binding ability, other than antibodies, can also be identified using the methods of the invention. For example, peptides and proteins that are not classified as an antibody or antibody fragment of any kind, but that do exhibit epitope specific binding, can be identified using the methods of the invention. For example, a library of peptides or proteins can be presented in an expression system or display system known in the art or by using a support known in the art, and an epitope specific peptide or protein can be identified using the methods of the invention. Similarly, small molecules that exhibit epitope specific binding can be identified using the methods of the invention. For example, a library of small molecules can be presented in a display system known in the art or by synthesized onto a solid support which will facilitate the separation of epitope-bound molecules from those unable to bind the epitope of interest.

Epitopes and Antigens

Methods of the current invention, in certain embodiments, are useful to identify an antibody or molecule capable of recognizing an epitope of interest or target epitope. An epitope specific antibody can be identified, using the methods of the invention, for any epitope that is capable of specifically binding a suitable antibody. Such epitopes include, but are not limited to, proteins, peptides, nucleic acids, DNA, RNA, glycoproteins, lectins, receptors, viruses, pathogens, hormones, growth factors, growth factor receptors, cytokines, transcription factors, transcription cofactors, steroids, peptide hormones, and any other molecule or atom that has the ability to function as an epitope or antigen.

Markers

In certain embodiments, the methods of the invention use a marker that is capable of binding to the epitope of interest (target epitope). In certain preferred embodiments, the marker is capable of binding the target epitope with sufficient affinity so the marker is useful for the methods of the invention. In certain other embodiments, the lower the affinity of the marker for the target epitope, the more marker molecules may be needed in a subtractive selection according to the invention. Markers useful for the methods of the invention include, but are not limited to, any molecule or atom capable of binding the target epitope, for example, a protein, a peptide, a ligand, a co-factor, DNA, KNTA, an antibody, a glycoprotein, a lectin, a receptor, a hormone, a pathogen, a growth factor, a growth factor receptor, a cytokine, a transcription factor, a transcription co-factor, a steroid, an enzyme, small molecules, or any type of molecules of interest that can be labeled with a dye or any type of fluorescence, or can be conjugated to any separable matrix. In certain preferred embodiments, a marker binds only to a single epitope on the antigen used in the methods of the invention, in certain other embodiments, a marker is a small molecule and/or a molecule that is not capable of binding to a large number of antibodies that are specific for distinct epitopes.

Labels

In certain embodiments, a label is used in the methods of the invention, for example, to label an antigen and/or to label a marker and/or to label an antibody. In certain embodiments, a label is used in a method of the invention to aid in the identification of a target antibody. A label useful for the invention includes any molecule or atom which can be conjugated, bound, or attached to an antigen, a marker, or an antibody and which can facilitate the identification of a target antibody using the methods of the invention. Examples of such labels include, but are not limited to, a chelator, a photoactive agent, a radioisotope, a fluorescent agent (for example, Alexa-488, FITC, phycoerythrin (PE) Cy 5, Cy 3, Rhodamine or fluorescent quantum dots that can be conjugated onto an antigen, an antibody, or a marker), a paramagnetic ion, or other marker moieties.

Antibodies Identified Using a Method of the Invention and their Use.

Any type of antibody or antibody fragment may be identified using the methods of the invention, as can non-antibody molecules. Antibodies identified with the disclosed methods are useful for diagnostics, therapeutics, or research. Particularly suitable uses for a particular antibody are, for example, the detection of the target epitope, the inhibition of binding reactions involving the target epitope, diagnosis of a condition involving the presence of the target epitope, therapy of a condition requiring the availability of the target epitope, analysis of the binding functions of the target epitope. In may disease conditions, molecular interactions play critical roles for the manifestation of those conditions and detecting the interactions often yields diagnostic information concerning the presence of a condition. Also, modulating those interactions can result in therapeutically desirable changes.

The present invention is further illustrated by the following examples, which are not intended to be limiting in any way whatsoever.

EXAMPLES

Example 1: Isolation of a fully human antibody that binds to an epitope recognized by a different antibody, such as a murine antibody. First, a fully human antibody library is generated using synthetic variable regions with randomized amino acid sequences in the CDRs (complementarity-determining region) and human heavy and light chain frameworks, germline frameworks are preferred. Fully human antibody library can also be generated by using cDNAs specifically synthesized from human immunoglobulin mRNA isolated from humans, humans immunized with an antigen, or humans with certain autoimmune diseases. These variable region fragments with diversified CDRs are then cloned into display vectors, such as yeast display vectors. The cloned antibodies are displayed on the surface of yeast cells.

Instead of displaying scFv known to those skilled in the art, in this invention, antibody is displayed as Fab or hAb, preferably hAb. The advantage of displaying hAb is the antigen-binding site on the antibody is further away from the yeast cell surface, therefore, upon binding to the antigen or molecule (A) as described above, only the binding site on the antigen for the antibody is blocked by the antibody displayed by the yeast cell, the remainder epitopes on the antigen should be well-exposed. Therefore, in the subtractive selection step, if yeast displayed antibodies bind to epitopes other than that for molecule B, the epitope for molecule (B) on the antibody bound antigen is recognized by molecule (B), and the resulting complexes (yeast cell-antibody-antigen-molecule (B)) can be removed by FACS, since the complexes are dual colored. In this way, antibodies that bind to the eptiope of interest only can be easily enriched.

Shown in Figure 1, an antigen of interest is labeled with fluorescent dye in red (represented by the black antibody). A murine antibody that recognizes this antigen is labeled with a fluorescent dye in green (represented by the antibody without fill-in). The labeled antigen is incubated with the yeast cells expressing a library of Fab or hAb in a binding buffer, after a period of time, such as 1 hour; unbound antigens are washed away with a wash buffer. Yeast cells displaying antibodies that recognize the antigen form a complex with the labeled antigen and should be red; these yeast cells can be selected, grown up, and re-selected in the next round of selection.

After the first 1-2 rounds of enrichment, the labeled murine antibody (in green) is added to the mixture containing the yeast cell-antigen complexes in a binding buffer and incubated for a period of time, such as 1 hour. The unbound labeled antibodies are washed away with a wash buffer. The washed yeast cells are subjected to FACS sorting. Yeast cells that bind to labeled antigen only are red as shown in Figure 1 (yeast cell A). Yeast cells with bound antigen that in turn bind to the labeled murine antibody are red and green such as yeast cells B, C, and D in Figure 1. FACS can be used to selectively sort out the yeast cells with red only from the pool of other cells. The selected yeast cells (red only) are those displaying fully human antibodies that are able to bind to the antigen in such a manner that prevents the binding of the murine antibody to the antigen, indicating that the binding site for the isolated human antibody on the antigen is the same or overlapping for the murine antibody.

After FACS sorting, the selected yeast cells (red only) can be grown up, induced to express the encoded human antibodies, and used for the next round selection as described above. The selection can be performed for multiple rounds until the majority of yeast cells are red only, even in the presence of excess amount of the murine antibody in green. Yeast cells that are detected the brightest red are selected for further analysis.

The sequences of the selected human antibodies can be determined by sequencing the variable regions in the display vectors isolated from individual yeast cell. The antibodies can be expressed as soluble form and used in competition assays, such as ELISA'(enzyme- linked immunosorbent assay). Since the binding site on the antigen for the selected human antibodies is the same or overlapping with the binding site for the murine antibody, many, if not all of the selected antibodies, are able to compete for binding of the murine antibody in this type of assay. The best human antibody with the highest affinity for the antigen most effectively competes with the murine antibody binding and can be further evaluated for inhibition of the interaction between the antigen and its cognate interacting partner in binding assays.

The fully human antibody finally selected has the same binding specificity as the murine antibody, and can be used as a diagnostic or therapeutic antibody. Affinity maturation, if necessary can be performed by generating a subset of variants and selected again with labeled antigen only. A library is generated to have each of the amino acid residue in the CDR regions represented by all 20 different amino acids while the remainder amino acids in the CDR regions are kept unchanged in the same variant. This library is used for affinity maturation with the labeled antigen. The brightest yeast cells are selected.

The described method therefore can substitute current methods for humanizing murine antibodies.

Example 2: Isolation of an antibody that inhibits receptor-ligand interaction.

The method described by this invention can also be used to isolate an antibody, protein, peptide, or other type of molecule that inhibits protein-protein or receptor-ligand interaction. In this example, the invented method can be used to isolate antibodies that inhibit the interaction between a receptor and its cognate ligand. Shown in Figure 2, a receptor of interest is labeled in red (represented by the black molecule with 4 arrows representing 4 different epitopes), the labeled receptor is added to a suspension of yeast cells displaying a library of fully human antibodies. After incubation in binding buffer for 1 hour, unbound receptor is washed away. Labeled yeast cells can be selected and enriched. After a few rounds of enrichment for yeast cells displaying antibodies that are able to recognize the labeled receptor, labeled receptor is added to the suspension of these yeast cells and incubated for a period of time, such as 1 hour, unbound labeled receptor is washed away with a wash buffer, then excess amount of the cognate ligand (for example, 100 fold higher molar ratio of the ligand over the receptor) labeled in green (represented by the black molecule with a L in it) is added to the yeast cells bound with labeled receptor, and to saturate the binding site still available on the receptor for the ligand, such as the binding site on the receptor bound to yeast cell B, C, D. After incubation for 1 hour, unbound ligand is wash away with a wash buffer.

The washed yeast cells are subjected for FACS sorting. Yeast cells with red only are selected, grown, and used for the next round of sorting. Multiple rounds of sorting can be performed until the majority of yeast cells are red, even in the presence of excess amount of labeled ligand in green. Yeast cells that are the brightest red are selected.

The yeast cells finally selected contain vectors encoding antibodies that are able to bind to the receptor of interest in such a manner that blocks the interaction between the receptor and its cognate ligand. The isolated antibodies can be expressed in soluble form and evaluated in receptor-ligand binding assay. The inhibitory activity of these soluble antibodies can be determined in competition binding assay.

Similarly, this approach can be used to isolate antibodies that recognize the ligand and inhibit the ligand' s ability to bind its receptor. In this case, the labeled ligand is added to the suspension of yeast cells displaying an antibody library first. After enriching yeast cells displaying antibodies for the ligand, excess amount of labeled receptor is added to the suspension of yeast cell-ligand complexes. For those receptors that are difficult to purify or for which separation of the receptor from cell membrane changes its conformation, labeled cells or labeled membrane preparations bearing the receptor can be used in the place of purified labeled receptor. After incubating and washing, the yeast cells are subjected for FAC sorting.

Only those yeast cells with ligand labeled in green attached are selected. These cells therefore display antibodies that are able to bind the ligand and inhibit its receptor from binding to the ligand. The selected yeast cells are grown and used for the next round of sorting until the majority of the cells are green. Antibodies displayed by yeast cells with the brightest green are selected for further analysis.

Again, these antibodies are expressed as soluble antibodies and evaluated in receptor-ligand binding assay. The isolated fully human antibodies can be used as a diagnostic or therapeutic antibody. Yeast cell A in Figure 2 can be easily isolated from the pool of yeast without labeling or labeled with dual colors. Isolation of yeast cells with red (yeast cell A) only facilitates the isolation of an antibody that recognizes the receptor and inhibits the receptor's interaction with its cognate ligand. Example 3: Isolation of an antibody that recognizes the same epitope shared by two or multiple different antigens.

It is often ideal to identify an antibody that recognizes an epitope on antigens from different species, so that the identified antibody can be evaluated in animal models of interest and used in humans. The methods of this invention can also be used to achieve this goal, since each yeast cell is able to display hundreds to thousands copies of the same antibody on its surface.

In this example, a human antigen of interest is labeled in red (no#l in Figure 3); the rat counterpart of this antigen is labeled in green (no#2 in Figure 3). These two labeled antigens can be added into a suspension of yeast cells displaying a library of antibodies at the same time. After incubation and washing, the yeast cells are subjected to FAC sorting. Yeast cells (such as yeast A in Figure 3) that are double colored are selected, grown, and used for the next round of selection. Multiple rounds of selections can be performed until the majority of cells are double colored. Yeast cells with the brightest red and green are selected for further analysis.

The final antibody selected is able to recognize the antigen from both human and rat. Therefore, this type of antibody can be evaluated in rat for its efficacy before it is tested in humans. Since each yeast cell only expresses one type of antibody, it is highly likely that the selected antibody recognizes the same epitope present in the human and rat antigen. Other antibodies displayed by different yeast cells selected in this way may be able to recognize other epitopes shared by human and rat antigen. Using the method of the invention described in the Example 1, one can further isolate an antibody that recognizes both human and rat antigen, and binds to the epitope of interest. Similarly, this method can be used to isolate antibodies that recognize an epitope that is specific for one antigen member but not for the other in the same protein family, or specific for an antigen from one specie, but from another.

As shown in Figure 3, when a human antigen labeled in red (represented by the molecule with number 1 in it) and the rat counterpart labeled in green (represented by the molecule with number 2 in it) are incubated with a suspension of yeast cells displaying a library of antibodies, yeast cells binding to the human antigen only are red only, whereas yeast cells that bind to rat antigen only is green only. Sorting of yeast cells with red or green antigen bound only allows one to isolate antibody that recognizes the human or rat antigen specifically.

This method is particularly useful for developing antibodies that are specific for a particular member in a large protein family, and are not reactive to other members in the same family. To develop such an antibody to be used as a research reagent, a diagnostic reagent, or a therapeutic, it is often desirable to have an antibody that recognizes only one antigen but not other related molecules that belong to the same protein family and share significant homology to the antigen of interest.

Example 4: Isolation of an antibody that recognizes a known epitope.

The current invention can also be used to isolate an antibody that recognizes a known epitope. Many antigens from different species are highly homologous to each other. Alignment of antigens from different species allows one to locate epitopes that are different among species. For example, a murine antibody recognizes a human antigen, but does not recognize the mouse counterpart of this antigen. Aligning the human and mouse sequences, one can identify the different epitopes. One of the different epitopes will be that one recognized by the murine antibody.

If such an epitope is known, this murine eptiope can be used to replace the corresponding region in the human antigen. The resulting human-mouse chimera often retains the original conformation of the antigen. The difference between the human and human-mouse chimera is the epitope recognized by the murine antibody. For selecting a fully human antibody that recognizes this epitope, the human antigen is labeled in red (represented by the molecule in complete black), and the human-mouse chimera is labeled in green (represented by the molecule in black with a small circle in white). The labeled antigens can then be used for selection. Yeast cells displaying antibodies that recognize epitopes shared by the human and human-mouse chimera are dual labeled (yeast cell C in Figure 4)

Yeast cells displaying antibodies that recognize the epitope presented by the human antigen, but not by the human-mouse chimera will be red only (yeast cell A). This epitope is the one replaced by mouse sequences in the human-mouse antigen, and is the epitope recognized by the murine antibody. The selected yeast cells in red only, therefore, display antibodies that are able to bind the same epitope recognized by the mouse antibody. Yeast cells displaying an antibody that recognizes the epitope presented by the human-mouse chimera, but not the human antigen will be in green. The displayed antibody therefore recognizes the mouse epitope (yeast cell D).

The present invention is not to be limited in scope by the specific embodiments described herein, which are intended as single illustrations of individual aspects of the invention, and functionally equivalent methods and components are within the scope of the invention. Indeed, various modifications of the invention, in addition to those shown and described herein, will become apparent to those skilled in the art from the foregoing description. Such modifications are intended to fall within the scope of the appended claims. All cited publications, patents, and patent applications are herein incorporated by reference in their entirety.

Claims

What is claimed is:

1. A method for obtaining an epitope specific antibody comprising (a) selecting an epitope of interest; (b) exposing a mixture of antibody molecules to an antigen, said antigen comprising the epitope of interest;

(c) separating antibody molecules capable of binding the antigen from the mixture of antibody molecules;

(d) exposing antibody molecules capable of binding the antigen to the antigen and to a marker, said marker capable of binding the epitope of interest;

(e) selecting antibody molecules capable of binding the antigen and capable of inhibiting binding of the marker to the antigen.

2. A method for obtaining an antibody capable of binding an epitope found in two antigens comprising:

(a) selecting two antigens, said antigens being labeled in a way that facilitates detecting each antigen separately;

(b) exposing a mixture of antibodies to said antigens;

(c) selecting antibodies capable of binding said two antigens; wherein said mixture of antibodies is displayed on a support in a way that facilitates identifying antibody molecules capable of binding said two antigens.

3. A method for obtaining an antibody capable of binding an epitope of interest, but not a homologue of the epitope, comprising: (a) selecting two antigens, said antigens being labeled in a way that facilitates detecting each antigen separately, and wherein one antigen comprises the epitope of interest and antoher antigen comprises the homologue of the epitope;

(b) exposing a mixture of antibodies to said antigens;

(c) selecting antibodies capable of binding said antigen comprising the epitope of interest, but not capable of binding said antigen comprising the homologue of the epitope; wherein said mixture of antibodies is displayed on a support in a way that facilitates identifying antibody molecules capable of binding said antigen comprising the epitope of interest, but not capable of binding said antigen comprising the homologue of the epitope^.