METHODS AND MATERIALS FOR TARGETED AFFINITY ENHANCEMENT

- XOMA TECHNOLOGY LTD.

The present disclosure relates to materials and methods for enhancing the binding affinity of antibodies by introducing a number of targeted amino acid changes in antibody variable domains. The present disclosure provides novel methods of flowcharted amino acid change cycles, of identifying residues as members of unique proximity groups and of determining the degree of conspicuousness of amino acid residues in antibody variable regions. The disclosed methods are useful for the selection of variable region amino acid residues that are candidates for change to produce antibody variable domains with enhanced affinity for their binding partners.

Skip to: Description  ·  Claims  · Patent History  ·  Patent History
Description
CROSS-REFERENCE TO RELATED APPLICATIONS

This application claims the benefit of International Patent Application No. PCT/US 2008/088631, filed on Dec. 31, 2008; U.S. Provisional Application No. 61/018,113, filed Dec. 31, 2007; U.S. Provisional Application No. 61/018,105, filed Dec. 31, 2007; and U.S. Provisional Application No. 61/018,101, filed Dec. 31, 2007, the disclosures of which are herein incorporated by reference in their entirety.

FIELD

The present disclosure relates to methods and materials for enhancing the binding affinity of an antibody. Modified antibody variable domains obtained by novel targeted affinity enhancement methods demonstrate an increased binding affinity compared to the binding affinity exhibited by the unmodified (e.g., parent) variable domain.

BACKGROUND

Affinity enhancement of a monoclonal antibody (beyond the ordinary nanomolar affinity which is typically achieved in an animal system) is desirable when producing a therapeutic agent, regardless of how the antibody was originally generated (e.g., by transgenic mice, by phage display, by yeast display, or by ordinary murine hybridoma methods). Extremely high affinity antibodies (e.g., scFv, Fab or IgG) are advantageous because they can be administered with equivalent efficacy in much lower doses, thereby decreasing the cost of producing the drug and/or diminishing its adverse side-effects.

Although natural immunological systems typically yield antibodies of nanomolar (10−9 M) affinity, greater affinities may be desirable. However, since astronomical numbers of different antibody combining sites are possible, it has been difficult to design a method for choosing a few key mutations in an antibody variable domain which might lead to greater binding affinity, particularly in the absence of reliable structural data (e.g., three-dimensional x-ray crystallographic information between the antibody and its antigen).

SUMMARY

The present disclosure relates to methods and materials for enhancing the binding affinity of an antibody by introducing a number of targeted amino acid changes (e.g., mutations) in an antibody variable domain. Such methods may minimize the total number of mutational steps for improved (e.g., enhanced) affinity and/or avoid the total unpredictability which is typical of randomly engineered mutations.

The present disclosure also relates to methods and materials for changing an amino acid residue at one or more positions in an antibody variable region to enhance affinity. A series of changes (e.g., mutations) are designed to probe the chemical environment at the interface between an antibody and its binding partner (e.g., antigen) for example, by making or breaking noncovalent intermolecular bonds. Observed changes in binding affinity due to any change (e.g., mutation) inform the kinds of chemical-functional groups present on the binding partner (e.g., antigen) to cause the observed changes. Methods are disclosed with exemplary mutation cycles, flow charts of changes and/or decision trees, which allow selection of amino acids changes that enhance affinity. In some embodiments, antibody variable region sequences may be aligned according to a standard numbering system such as Kabat.

Methods are provided for enhancing the affinity of an antibody by carrying out one or more steps that introduce changes in the amino acid sequence of an antibody variable region, including, for example: testing the effect of either removing a charge or adding a charge at one or more amino acid positions, and then collectively retaining those amino acid changes which have increased binding affinity; replacing an unfavorable charge with the opposite charge at one or more amino acid residues individually, for example, if introducing the charge had previously caused a loss of affinity, or if removing the charge had caused a gain of affinity, and then collectively retaining improvements; testing one or more charged residues individually by changing sidechains of the same charge but different sizes, and retaining improvements; switching one or more uncharged residues individually between polar and nonpolar to determine which choice yields better affinity, and retaining improvements; introducing a large amino acid residue (e.g., tyrosine (Y)) individually at one or more positions, if the switch between polar and nonpolar had produced no change in affinity and retaining improvements; and testing each uncharged residue individually by changing sidechains of the same polar or nonpolar functionality but of different sizes, and retaining amino acid changes that result in enhanced binding affinity. The steps individually or in a cycle (e.g., mutation cycle) may then be repeated after selecting a new group of positions comprising the next stage until a desired level of affinity is reached. Optionally or additionally, at the end of a step, any or all the amino acid changes which have led to higher affinity may be combined collectively into a single optimized construct, which may then form the basis for each of the individualized changes in a next step. Any of these steps may be carried out individually and/or may be combined in a series of steps in varying order as described herein.

Methods are disclosed which minimize the total number of amino acid changes necessary for enhancement of an antibody's affinity. Such methods may make a series of amino acid changes at an original amino acid position. Amino acid residues in an antibody variable domain, including a heavy chain variable region and/or light chain variable region, may be selected for change by employing novel methods which assign each amino acid in the variable region to one of the following unique groups: contacting (C), peripheral (P), supporting (S), interfacial (I), or distant (D). These so-called proximity groups permit the selection of amino acid residues that are candidates for change. Additionally or alternatively, positions for amino acid changes may be based upon a novel method of determining the degree to which the original amino acid residue differs from the corresponding consensus or germline residue in terms of charge, size or chemical functionality to calculate “conspicuousness.” For example, the methods provided by the disclosure may include utilization of tables of numerical components, which can be added together to identify “conspicuous” amino-acid changes.

Methods are also disclosed for enhancing the affinity of a variable region of an antibody by identifying the proximity assigned to amino acid positions in the variable region of the antibody and changing one or more contacting (C), peripheral (P), supporting (S) interfacial (I) and/or distant (D) amino acid residues with other amino acid residues.

An exemplary method for targeted affinity enhancement may be described as a flowchart and/or as a mutation cycle (see, e.g., FIG. 5 or FIG. 6) and may include the following steps: (1) selecting one or more contacting (C), supporting (S), peripheral (P) interfacial (I) and/or distant (D) amino acid residues to be changed; (2) testing the effect of either removing a charge or adding a charge at each selected residue individually, and then collectively retaining those amino acid changes which have increased binding affinity; (3) replacing an unfavorable charge with the opposite charge at each selected residue, either if introducing the charge had previously caused a loss of affinity, or if removing the charge had caused a gain of affinity, and then collectively retaining improvements; (4) testing each selected charged residue individually by changing sidechains of the same charge but different sizes, and retaining improvements; (5) switching each selected uncharged residue individually between polar and nonpolar to determine which choice yields better affinity, and retaining improvements; (6) introducing a large amino acid residue (e.g., tyrosine (Y)) individually at each selected position, if the switch between polar and nonpolar had produced no change in affinity, and retaining improvements; and (7) testing each selected uncharged residue individually by changing sidechains of the same polar or nonpolar functionality but of different sizes, and retaining amino acid changes that result in enhanced binding affinity. Any of these steps may be carried out individually and/or may be combined in a series of steps in varying order as described herein. An exemplary flowchart of amino acid changes or mutation cycle is shown in FIG. 5 and/or FIG. 6.

Methods are also provided for enhancing the binding affinity of an antibody variable domain (e.g., a heavy chain variable region and/or light chain variable region) to obtain a modified antibody variable domain with enhanced affinity compared to an unmodified (parent) antibody variable domain by changing residues at preferably one or more contacting (C), peripheral (P), supporting (S) and/or interfacial (I) amino acid residues to other amino acids, thereby generating a modified variable domain; testing the binding affinity of the modified antibody variable domain; and obtaining a modified antibody variable domain with enhanced binding affinity. Less preferably one or more distant (D) amino acid residues may additionally or alternatively be changed in the methods.

Methods are also provided for enhancing the binding affinity of an antibody variable domain (e.g., a heavy chain variable region and/or light chain variable region) compared to an unmodified (parent) antibody variable domain by changing amino acid residues at preferably one or more contacting (C), peripheral (P), supporting (S) and/or interfacial (I) amino acid residues to other amino acids, wherein charged residues are changed to uncharged or uncharged residues are changed to charged or both charged residues are changed to uncharged and uncharged residues are changed to charged, thereby generating a modified antibody variable domain; testing the binding affinity of the modified antibody variable domain; and obtaining a modified antibody variable domain with enhanced binding affinity. Less preferably one or more distant (D) amino acid residues may additionally or alternatively be changed in the methods.

Methods are provided for enhancing the binding affinity of an antibody variable domain (e.g., a heavy chain variable region and/or light chain variable region) compared to an unmodified (parent) antibody variable domain by changing amino acid residues at preferably one or more contacting (C), peripheral (P), supporting (S) and/or interfacial (I) amino acid residues to other amino acids, wherein residues are changed to a different size within the same amino acid class, that is, positively charged, negatively charged, polar or nonpolar, thereby generating a modified antibody variable domain; testing the binding affinity of the modified antibody variable domain; and obtaining a modified antibody variable domain with enhanced binding affinity. Less preferably one or more distant (D) amino acid residues may additionally or alternatively be changed in the methods.

Methods are also provided for enhancing the binding affinity of an antibody variable domain (e.g., a heavy chain variable region and/or light chain variable region) compared to an unmodified (parent) antibody variable domain by changing amino acid residues at preferably one or more contacting (C), peripheral (P), supporting (S) and/or interfacial (I) amino acid residues to other amino acids, wherein charged residues are changed to opposite charge residues, that is, from positively charged to negatively charged or from negatively charged to positively charged, thereby generating a modified antibody variable domain; testing the binding affinity of the modified antibody variable domain; and obtaining a modified antibody variable domain with enhanced binding affinity. Less preferably one or more distant (D) amino acid residues may additionally or alternatively be changed in the disclosed methods.

Methods are also provided for enhancing the binding affinity of an antibody variable domain (e.g., a heavy chain and/or light chain variable region) compared to an unmodified (parent) antibody variable domain by changing amino acid residues at preferably one or more contacting (C), peripheral (P), supporting (S) and/or interfacial (I) amino acid residues to other amino acids, wherein polar residues are changed to nonpolar residues or nonpolar residues are changed to polar residues or both polar residues are changed to nonpolar residues and nonpolar residues are changed to polar residues, thereby generating a modified antibody variable domain; testing the binding affinity of the modified antibody variable domain; and obtaining a modified antibody variable domain with enhanced binding affinity. Less preferably one or more distant (D) amino acid residues may additionally or alternatively be changed in the disclosed methods.

An exemplary method for targeted affinity enhancement may be described as a flowchart and/or as a mutation cycle (see, e.g., FIG. 5 and/or FIG. 6) and may include (a) firstly, changing amino acid residues at one or more contacting (c) residues in an antibody variable domain (e.g., a heavy chain variable region and/or light chain variable region) to other amino acid residues, wherein charged residues are changed to uncharged residues thereby removing a charge and uncharged residues are changed to charged residues thereby adding a charge; (b) testing the effect on affinity to a binding partner by removing or adding a charged amino acid residue at each position individually changed in step (a); (c) obtaining a first modified antibody variable domain comprising the changed amino acid residues from step (a) that resulted in an increase in affinity as individually tested in step (b); (d) secondly, changing amino acid residues that were firstly changed in step (a) from charged to uncharged or from uncharged to charged to amino acid residues of the opposite charge, if removing the charged residue in step (a) had caused a gain in affinity when tested in step (b) or if adding the charged residue in step (a) caused a gain in affinity when tested in step (b); (e) testing the effect on affinity to the binding partner by changing the amino acid residue to the opposite charged residue at each position individually changed in step (d); (f) obtaining a second modified antibody variable domain comprising the changed residues from step (d) that resulted in an increase in affinity as individually tested in step (e); (g) thirdly, changing charged amino acid residues in the second modified antibody variable domain of step (f) to an amino acid residue of a different size and the same charge; (h) testing the effect on affinity to the binding partner by changing the size of the charged residue at each position individually changed in step (g); (i) obtaining a third modified antibody variable domain comprising the changed amino acid residues from step (g) that resulted in an increase in affinity as individually tested in step (h); (j) fourthly, changing uncharged amino acid residues in the third modified antibody variable domain of step (i) to other amino acid residues, wherein polar uncharged residues are changed to nonpolar uncharged residues and wherein nonpolar uncharged residues are changed to polar uncharged residues; (k) testing the effect on affinity to the binding partner by changing polar to nonpolar and nonpolar to polar at each position individually changed in step (j); (l) obtaining a fourth modified antibody variable domain comprising the changed amino acid residues from step (j) that resulted in an increase in affinity as individually tested in step (k); (m) fifthly, changing uncharged residues that were fourthly changed in step (j) from polar to nonpolar or from nonpolar to polar to a large amino acid residue (e.g. tyrosine), if changing from polar to nonpolar or from nonpolar to polar in step (j) had not caused a gain or loss in affinity when tested in step (k); (n) testing the effect on affinity to the binding partner by changing to a large amino acid residue (e.g., tyrosine (Y)) at each position individually changed in step (m); (o) obtaining a fifth modified antibody variable domain comprising the changed large amino acid residues (e.g., tyrosine (Y)) from step (m) that resulted in an increase in affinity as individually tested in step (n); (p) sixthly, changing uncharged polar or nonpolar amino acid residues in the fifth modified antibody variable domain of step (o) to polar or nonpolar residues, respectively, of a different size; (q) testing the effect on affinity to the binding partner by changing the size of the uncharged residue at each position individually changed in step (p); and (r) obtaining a sixth modified antibody variable domain comprising the changed amino acid residues from step (p) that resulted in an increase in affinity as individually tested in step (q). An exemplary flowchart of amino acid changes or mutation cycle is shown in FIG. 5 and/or FIG. 6.

In some embodiments, the contacting (C) residue may be in complementarity determining region-1 (CDR1) in a light chain variable region. In certain embodiments, the contacting (C) residue may be at a position corresponding to position 28, 30 or 31 in CDR1. In other embodiments, the contacting (C) residue may be in CDR2 in a light chain variable region. In certain embodiments, the contacting (C) residue may be at a position corresponding to position 50, 51 or 53 in CDR2. In other embodiments, the contacting (C) residue may be in CDR3 in a light chain variable region. In some embodiments, the contacting (C) residue may be in CDR1 in a heavy chain variable region. In certain embodiments, the contacting (C) residue may be at a position corresponding to position 32 or 33 in CDR1. In some embodiments, the contacting (C) residue may be in CDR2 in a heavy chain variable region. In certain embodiments, the contacting (C) residue may be at a position corresponding to position 50, 52, 53, 54, 56, or 58 in CDR2. In some embodiments, the contacting (C) residue may be in CDR3 in a heavy chain variable region.

In some embodiments, changing one or more contacting (C), peripheral (P), supporting (S), interfacial (I), and/or distant (D) amino acid residues may refer to changing all the amino acid residues of a proximity group in a variable region, including, for example, all the contacting (C), peripheral (P), supporting (S), interfacial (I), and/or distant (D) amino acid residues.

In some embodiments, contacting (C), peripheral (P), supporting (S), interfacial (I), and/or distant (D) amino acid residues may be changed to either alanine (Ala, A), arginine (Arg, R), asparagine (Asn, N), aspartic acid (Asp, D), cysteine (Cys, C), glutamine (Gln, Q), glutamic acid (Glu, E), glycine (Gly, G), histidine (H is, H), isoleucine (Ile, I), leucine (Leu, L), lysine (Lys, K), methionine (Met, M), phenylalanine (Phe, F), proline (Pro, P), serine (Ser, S), threonine (Thr, T), tryptophan (Trp, W), tyrosine (Tyr, Y) or valine (Val, V). In some embodiments, amino acid residues are not changed to cysteine and/or proline.

In some embodiments, the variable domain is from a chimeric antibody. In other embodiments, the variable domain is from a human engineered or humanized antibody. In some embodiments, the variable domain is from a human antibody.

In some embodiments, binding affinity is determined by measuring Koff. In some embodiments, binding affinity may be determined by Biacore (e.g., Biacore 2000 or A100).

In some embodiments, enhanced binding affinity is determined by contacting a parent variable domain (e.g., a heavy chain variable region and/or light chain variable region) with the binding partner under conditions that permit binding; contacting the modified variable domains with the binding partner under conditions that permit binding; and determining binding affinity of the modified variable domains and the parent variable domain for the binding partner, wherein modified variable domains that have a binding affinity greater than the binding affinity of the parent variable domain for the binding partner are identified as having enhanced binding affinity.

BRIEF DESCRIPTION OF THE FIGURES

FIG. 1 is a generalized schematic map of an exemplary antibody combining site as described herein, looking downward onto the “top” surface of a variable domain comprising a light chain variable region and a heavy chain variable region. It shows the six CDR loops (L1, L2, L3, H1, H2, H3) which are shown spatially located directly above the three-dimensional structure of the evolutionarily-conserved framework underneath. As shown and as described herein, this map provides roughly approximate higher-order structural information, which is not available from the linear primary sequence alone, such as the identity of potential nearest neighbors in the space-filling model of a generic variable domain. Specific features of the murine ING-1 monoclonal antibody have been added to this map, so that it can also call attention to localized domains of the antibody's combining site containing clusters of high-conspicuousness positions as described herein which are likely to be in contact with sidechains on the binding partner. In particular, each amino-acid position in the murine ING-1 antibody is represented on this map by a white rectangle containing a group of symbols. The letter and number at the bottom-left of each rectangle (e.g., “H 98” in CDR-loop H3) is the Kabat-position number of the amino-acid residue on the antibody molecule within either chain (L=light, H=heavy). The small upper-case letter (e.g., “C”) at the bottom-right is a designation for the residue's proximity as described herein (C=Contacting, P=Peripheral, S=Supporting, I=Interfacial) relative to the antibody's binding site (shown on “prox” line in FIGS. 2A and 2B). The large upper-case letter (e.g., “A”) at the upper-left is the amino-acid code for the residue's sidechain (line “murING1” in FIGS. 2A and 2B). The large single digit at the upper right (e.g., “3”) in some rectangles is the non-zero conspicuousness-value, as described herein of affinity enhancement for the sidechain (line “cspc” in FIGS. 2A and 2B), calculated in reference to the appropriate human consensus sequence for light chain (hK2) or heavy chain (hH1). Rectangles with no such value reflect a conspicuousness of zero.

FIGS. 2A and 2B are alignments of sequences in the light chain and heavy chain, with lines (e.g., “prox”, “cspc”) relating to affinity enhancement and lines relating to human engineering (e.g., risk). In each set of lines, the top ones apply the present disclosure to the murine ING1 antibody, and the bottom ones relate the present disclosure to the general principles of human engineering (Studnicka et al., Protein Engineering, 7(6):805-814 (1994); U.S. Pat. No. 5,766,886). Each set of lines shows the Kabat position numbers (pos), the general classification of proximity groups as described herein for each position of every antibody (“prox”), the murine ING1 monoclonal antibody sequence to be affinity-enhanced (murINGI), the conspicuousness value as described herein of each position for affinity-enhancement when the murine ING1 antibody is compared to several murine consensus sequences (cspc), the murine consensus sequences to which ING1 is compared (mK2 or mH2a), the human ING1 residues which are introduced during the human engineering process (humING1), the degree of disconnection of the sidechain from the antibody's combining site (disc) as described herein, the degree of outward-orientation of the sidechain on the antibody's surface (outw) as described herein, the degree of risk for human engineering (risk), and the Kabat position numbers (pos).

FIGS. 3A and 3B are mutual alignments of consensus sequences [Kabat et al. (eds), Sequences of Proteins of Immunological Interest, 4th ed, (1987)] for major murine and human subgroups of the light chain and heavy chain. Each alignment relates them to the proximity groups as described herein for each position (“prox”), and the Kabat position numbers (pos) (see, e.g., Table 1 where contacting is abbreviated as (C), peripheral is abbreviated as (P), supporting is abbreviated as (S), interfacial is abbreviated as (I), and distant is abbreviated as ).

FIG. 4 shows a chart of the numerical components which can be added together to calculate each amino acid's affinity-enhancement conspicuousness value as described herein, including the components for changes in class-and-charge, for changes in physical size due to somatic mutation, and for repeated identical mutations as described herein, at the same position in multiple homologous antibodies.

FIG. 5 shows a pictorial diagram illustrating an exemplary flowchart of amino acid changes and/or an exemplary amino acid change cycle (e.g., mutation cycle), not including the details of the SIZE section or the SPECIAL section. Labels with thick solid borders indicate temporary exploratory amino acid changes designed to probe the chemical microenvironment. Labels without borders indicate the conceptual purpose of each exploratory amino acid change. Labels with dashed borders indicate the current status of the residue at each point in the process, including beneficial changes which are kept, as well as reversions to previous sidechains after detrimental changes. Small circles containing a directional arrow indicate affinity changes as the result of a change, either increasing or decreasing or remaining constant. The SIZE label indicates that the SIZE section of the flowchart is to be carried out with the current residue (shown in the dashed label above it) as its starting point. The SPECIAL label indicates the entry point to the main section of the amino acid change cycle, which is followed after previously retaining changes (e.g., a valine or a serine or a glycine) from the SPECIAL section if they have removed certain special conditions (e.g., a proline or a cysteine or a deletion).

FIG. 6 shows a pictorial diagram illustrating an exemplary flowchart of amino acid changes and/or an exemplary amino acid change cycle (e.g., mutation cycle). Labels with thick solid borders indicate generated (e.g., exploratory) changes designed to probe the chemical microenvironment, by switching either charged vs. uncharged or polar vs. nonpolar. Labels with dashed borders indicate the type of binding partner sidechain (amino acids are listed directly underneath) which is consistent with the observed experimental results. Small circles containing a directional arrow indicate affinity changes as the result of a change, either increasing or decreasing or remaining constant. Symbols (“+,−, p,n”) next to the arrow-circles indicate the direction of the mutational switch at the antibody residue (more positive or negative or polar or nonpolar), so that “+” means making the antibody residue more positive (from uncharged to positive or from negative to uncharged) than it was before the switch.

FIG. 7 relates the present disclosure to changes selected in a directed-evolution experiment using yeast display [Boder et al., Proc. Natl. Acad. Sci. USA, 97:10701-10705 (2000)]. Each set of lines shows the Kabat position numbers (pos), the unmutated amino acids of the Boder experiment (BoderU), the affinity-enhancement conspicuousness value as described herein for each position of the Boder experiment (BoderC), the mutated amino acids of the Boder experiment with strongly-selected residues in upper case and less-frequent residues in lower case (BoderM), and the positions marked by asterisks which would be reverted to conserved consensus during low-risk human engineering of the Boder experiment (BoderH).

FIGS. 8A-8E apply the present disclosure to a variety of antibodies. Each set of lines shows the Kabat position numbers (pos), the general classification of proximity groups as described herein, for each position of every antibody (“prox”), the antibody sequence to be affinity enhanced (e.g., humH65 in 8A, humING1 in 8B, mur4A2 in 8C, murIND1 in 8D, murIND2 in 8E), the conspicuousness number as described herein of each position for affinity enhancement when the antibody is compared to the binding partner sequences (cspc), the binding partner consensus sequences to which the antibody sequence is compared (hKI, hH3, hK2, hHI, mK3, mH3d, mKI, mH3c, mK5, mH2a), and (in cases of human engineered antibodies) the original murine residues which were replaced during the human engineering process (murH65, murING1).

FIG. 9 is a symmetric matrix, which gives the conspicuousness value as described herein, for possible amino acid changes from one amino acid to another, including nonconserved (X) and gap (−) positions. One-letter amino-acid codes are shown along each edge of the matrix, and the appropriate conspicuousness value as described herein is located at the intersection of the row and column corresponding to the substituted and unsubstituted amino acid.

FIG. 10A is an alignment of variable region sequences in the light chain and heavy chain for affinity enhancement of human antibodies A8.2 and C5A. Each set of lines shows the Kabat position numbers (pos), the general classification of proximity groups as described herein, for each position of every antibody (“prox”), the conspicuousness value as described herein, of each position for affinity enhancement when the human A8.2 antibody is compared to the human consensus sequences hKI or hH3 (cspcA), the human A8.2 antibody's amino-acid sequence (A8.2), the human consensus sequences for light or heavy chain (hKI or hH3), the human C5A antibody's amino-acid sequence (C5A), the conspicuousness value as described herein, of each position for affinity-enhancement when the human C5A antibody is compared to the human consensus sequences hKI or hH3 (cspcC), and the Kabat position numbers (pos).

FIGS. 10B and 10C apply the present disclosure to human antibodies (A8.2, C5A). Each drawing is a generalized schematic map of an exemplary antibody combining site, similar to FIG. 1.

FIG. 11 relates the proximity groups of the present disclosure to the crystallographic contact-analysis of 26 antibody-binding partner complexes [MacCallum et al., Journal of Molecular Biology, 262:732-745 (1996)]. Each set of lines shows the Kabat position numbers (pos), the alternative system of Chothia position numbers (posCho), a sequence of single-digit codes (La, first digit of two-digit percentage) representing the approximate fraction of antibodies making contact at each position according to the MacCallum analysis (MacC), the general classification of proximity groups as described herein, for each position of every antibody (“prox”), the degree of disconnection as described herein, of the sidechain from the antibody's combining site (disc), the degree of outward-orientation as described herein, of the sidechain on the antibody's surface (outw), the degree of risk for human engineering (risk), and the Kabat position numbers (pos).

FIGS. 12A-12E apply the present disclosure to a variety of antibodies (humH65 in 12A, humINGI in 12B, mur4A2 in 12C, murIND1 in 12 D, murIND2 in 12 E). Each drawing is a generalized schematic map of an exemplary antibody combining site, similar to FIG. 1.

FIG. 13 is a generalized schematic map of an exemplary antibody combining site, similar to FIGS. 1, 10B, 10C and 12. It is a “blank” form, showing data common to antibodies, and lacking data specific to any individual antibody.

 DETAILED DESCRIPTION

The present disclosure provides methods for enhancing the binding affinity of an antibody by introducing targeted amino acid changes (e.g., mutations) at positions in an antibody variable domain (e.g., an ING-1 heavy chain variable region as encoded in SEQ ID NO: 1, an ING-1 light chain variable region as encoded in SEQ ID NO: 2, an XPA-23 heavy chain variable region as encoded in SEQ ID NO: 3 or an XPA-23 light chain variable region as encoded in SEQ ID NO: 4). These methods for targeted affinity enhancement may be utilized even in the complete absence of any detailed information about the interaction between the antibody and its binding partner. The methods do not require any three-dimensional x-ray crystallographic structures of the chosen antibody's combining site with its binding partner (e.g., an epitope on an antigen's surface) and/or any type of energy-minimization algorithm. For methods of the present disclosure, in some embodiments, antibody variable region sequences may be aligned according to a standard numbering system such as Kabat.

The present disclosure provides methods for enhancing the affinity of an antibody heavy and/or light chain variable region by changing one or more amino acid residues to other amino acid residues. One (preferably first) step can test the effect of either removing a charge or adding a charge at each position selected for change individually. Another (preferably second) step can replace an unfavorable charge with the opposite charge at each position selected for change individually, either if introducing a charge had previously caused a loss of affinity, or if removing a charge had caused a gain of affinity. Another (preferably third) step can test each charged residue selected for change individually by changing sidechains of the same charge but different sizes. Another (preferably fourth) step can test the effect of switching each uncharged residue selected for change individually between polar and nonpolar to determine which choice yields better affinity. Another (preferably fifth) step can introduce a large amino acid residue (e.g., tyrosine) to test each position selected for change individually, if the switch between polar and nonpolar had produced no change in affinity. Another (preferably sixth) step can test each uncharged residue selected for change individually by changing sidechains of the same polar or nonpolar functionality but of different sizes. The steps individually or in a cycle (e.g., mutation cycle) may then be repeated after selecting a new group of positions comprising the next stage until a desired level of affinity is reached. At the end of a step, any or all the amino acid changes which have led to higher affinity may be combined collectively into a single optimized construct, which may then form the basis for each of the individualized changes in a next step. Any of these steps may be carried out individually and/or may be combined in a series of steps in varying order. An exemplary flowchart of a series of steps or mutation cycle is shown in FIGS. 5 and/or 6.

The present disclosure also provides methods for enhancing the affinity of an antibody variable region by identifying the proximity assigned to amino acid positions in the variable domain using the “prox” line as shown in FIG. 3A or 3B (e.g., by aligning a variable region sequence, including one or more light chain variable region sequences and/or one or more heavy chain variable region sequence's, according to a standard numbering system such as Kabat) and preferably changing one or more contacting (C), supporting (S), peripheral (P) and/or interfacial (I) amino acid residues with other amino acid residues. Less preferably one or more distant (D) amino acid residues may additionally or alternatively be changed in the methods.

In an exemplary method to accomplish targeted affinity enhancement, amino acid residues may be selected for change by aligning (e.g., according to a standard numbering system such as Kabat) an antibody light chain and/or heavy chain variable region sequence and comparing the sequence with any variable region sequence (e.g., a homologous consensus sequence for the light and heavy chain subgroups to which it is similar, and/or with its own precursor germline sequence if it is available). The “prox” line shown in FIGS. 3A and/or 3B may be used to identify the proximity assigned to amino acid positions in the variable region of a light chain and/or heavy chain as contacting (C), peripheral (P), supporting (S), interfacial (I) and/or distant (D), amino acid residues and amino acid residues may be selected for change according to the disclosed methods. For example, using the “prox” line as shown in FIG. 3A or 3B, sequences may be aligned according to a standard numbering system such as Kabat.

The present disclosure provides methods for enhancing the binding affinity of an antibody (e.g., an ING-1 heavy chain variable region as encoded in SEQ ID NO: 1, an ING-1 light chain variable region as encoded in SEQ ID NO: 2, an XPA-23 heavy chain variable region as encoded in SEQ ID NO: 3 or an XPA-23 light chain variable region as encoded in SEQ ID NO: 4) by means of changing one or more amino acid residues in the antibody's variable domain (e.g., a heavy chain variable region and/or light chain variable region) at one or more residues (e.g., contacting (C), peripheral (P), supporting (S), interfacial (I) and/or distant (D)) where amino acid changes are likely to produce enhanced affinity.

The present disclosure provides methods to enhance the binding affinity of an antibody variable domain (e.g., a heavy chain variable region and/or light chain variable region) to a binding partner, to obtain a modified antibody variable domain with enhanced affinity by identifying the proximity assigned to amino acid positions in the antibody variable domain using the “prox” line as shown in FIG. 3A or 3B; changing amino acid residues preferably at one or more contacting (C) proximity positions in the domain to other amino acid residues, thereby generating a modified variable domain; testing the binding affinity of the modified antibody variable domain for the binding partner; and thereby obtaining a modified antibody variable domain with enhanced binding affinity. In some embodiments, all the identified contacting (C) proximity positions in a variable domain are changed.

The present disclosure provides methods to enhance the binding affinity of an antibody variable domain (e.g., a heavy chain variable region and/or light chain variable region) to a binding partner, to obtain a modified antibody variable domain with enhanced affinity by identifying the proximity assigned to amino acid positions in the antibody variable domain using the “prox” line as shown in FIG. 3A or 3B; changing amino acid residues preferably at one or more peripheral (P) proximity positions in the domain to other amino acid residues, thereby generating a modified variable domain; testing the binding affinity of the modified antibody variable domain for the binding partner; and thereby obtaining a modified antibody variable domain with enhanced binding affinity.

The present disclosure provides methods to enhance the binding affinity of an antibody variable domain (e.g., a heavy chain variable region and/or light chain variable region) to a binding partner, to obtain a modified antibody variable domain with enhanced affinity by identifying the proximity assigned to amino acid positions in the antibody variable domain using the “prox” line as shown in FIG. 3A or 3B; changing amino acid residues preferably at one or more supporting (S) proximity positions in the domain to other amino acid residues, thereby generating a modified variable domain; testing the binding affinity of the modified antibody variable domain for the binding partner; and thereby obtaining a modified antibody variable domain with enhanced binding affinity.

The present disclosure provides methods to enhance the binding affinity of an antibody variable domain (e.g., a heavy chain variable region and/or light chain variable region) to a binding partner, to obtain a modified antibody variable domain with enhanced affinity by identifying the proximity assigned to amino acid positions in the antibody variable domain using the “prox” line as shown in FIG. 3A or 3B; changing amino acid residues preferably at one or more interfacial (I) proximity positions in the domain to other amino acid residues, thereby generating a modified variable domain; testing the binding affinity of the modified antibody variable domain for the binding partner; and thereby obtaining a modified antibody variable domain with enhanced binding affinity.

The present disclosure provides methods to enhance the binding affinity of an antibody variable domain (e.g., a heavy chain variable region and/or light chain variable region) to a binding partner, to obtain a modified antibody variable domain with enhanced affinity by identifying the proximity assigned to amino acid positions in the antibody variable domain using the “prox” line as shown in FIG. 3A or 3B; changing amino acid residues at less preferably one or more distant (D) proximity positions in the domain to other amino acid residues, thereby generating a modified variable domain; testing the binding affinity of the modified antibody variable domain for the binding partner; and thereby obtaining a modified antibody variable domain with enhanced binding affinity.

Additionally or alternatively, the primary amino acid sequence may be characterized to identify amino acid residues that are “conspicuous” (e.g., by calculations as described herein) and that may be candidates for change. Residues differing markedly in charge or size or chemical functionality from the corresponding residues in the selected sequence, including, for example, the consensus or the germline, may confer specific affinity for antigen upon the antibody.

In an exemplary method to accomplish targeted affinity enhancement one or more amino acid residues in an antibody heavy and/or light chain variable region selected for change (e.g., by using the “prox” line as shown in FIGS. 3A and/or 3B and for example, including aligning a variable region sequence with the “prox” line according to a standard numbering system such as Kabat) may be changed to other amino acid residues. A first step can test the effect of either removing a charge or adding a charge at each position selected for change individually. A second step can replace an unfavorable charge with the opposite charge at each position selected for change individually, either if introducing the charge had previously caused a loss of affinity, or if removing the charge had caused a gain of affinity. A third step can test each charged residue selected for change individually by changing sidechains of the same charge but different sizes. A fourth step can test the effect of switching each uncharged residue selected for change individually between polar and nonpolar to determine which choice yields better affinity. A fifth step can introduce a large amino acid residue (e.g., tyrosine) at each position selected for change individually, if a switch between polar and nonpolar had produced no change in affinity. A sixth step can test each uncharged residue selected for change individually by changing sidechains of the same polar or nonpolar functionality but of different sizes to find a better fit. The steps individually or in a cycle may then be repeated after selecting a new group of positions comprising the next stage until a desired level of affinity is reached. At the end of a step, any or all the amino acid changes which have led to higher affinity may be combined collectively into a single optimized construct, which may then form the basis for each of the individualized changes in a next step. Any of these steps may be carried out individually and/or may be combined in a series of steps in varying order. An exemplary flow chart of amino acid changes or mutation cycle is shown in FIG. 5 and/or FIG. 6.

Although any methods and materials similar or equivalent to those described herein can be used in the practice or testing of the present disclosure, exemplary methods and materials are described.

Methods for Affinity Enhancement

The present disclosure provides methods for the change (e.g., mutation) of one or more amino acid residues at a position in an antibody with other amino acid residues to identify an amino acid change which results in the antibody having enhanced binding affinity (see, e.g., FIG. 5 and/or FIG. 6). Enhanced binding affinity refers to a modified variable domain that binds to a binding partner (e.g., antigen) with a higher equilibrium constant of association (KA) or lower equilibrium constant of dissociation (KD) than the parent variable domain when the amounts of modified and parent variable domains in the binding assay are essentially the same. For example, the modified variable domain with improved affinity for its binding partner may display at least a 10%, at least a 15%, at least a 25%, at least a 50%, at least a 75%, at least a 100% (or two-fold), at least a 5-fold, at least an 8-fold, at least a 10-fold, at least a 50-fold, at least a 100-fold, or more, higher affinity to a binding partner than the corresponding parent variable domain. A binding partner for a variable domain, including a modified variable domain, includes an antigen (e.g., an epitope on an antigen) recognized by an antibody or a molecular target of an antibody.

Changes at an original amino acid position on an antibody are made according to novel strategies and methods which may take into account characteristics of the original amino acid residue. For example, amino acid residues may be categorized as charged, uncharged, polar, non-polar or special (e.g., special amino acids are those that are either proline, cysteine or empty spaces). Strategies for changing an original amino acid residue are described below, including residues from various proximity groups and/or residues with certain conspicuousness values as described herein. Methods are provided for probing an antibody combining site by making a series of amino acid changes in a variable region amino acid sequence as shown, for example, in FIG. 5 and/or FIG. 6.

1. Special Amino Acid Residues

Amino acid residues classified as special (e.g., a proline (P), cysteine (C) or deletion) may be changed with an amino acid residue with minimal change to size or chemical function to test whether the special condition is necessary for maintaining affinity (see, e.g., FIG. 5). For example, proline (P) may be replaced with valine (V), cysteine (C) may be replaced with serine (S), and a deletion may be augmented by inserting a single glycine (G).

Without being bound by a theory of the invention, a decrease in the affinity of the modified variable domain for its binding partner after the special change suggests that the residue may be relevant to or critical for binding affinity. The original configuration (e.g., proline (P), cysteine (C) or deletion) may be kept. Further amino acid changes may not be necessary or desired at the position.

Without being bound by a theory of the invention, an increase or similar level in the affinity of the modified variable domain for its binding partner after the special change suggests that the original residue may not relevant to or critical for binding. The new residue (e.g., valine (V) or serine (S) or glycine (G)) may be kept at the position and the methods outlined for “uncharged” amino acid residues as described below may be followed.

2. Charged Amino Acid Residues

Amino acid residues classified as charged may be changed with an uncharged amino acid (e.g., asparagine (N)) to test whether the binding partner position contains an undetected charged residue (e.g., whether positive or negative) which had previously been either attracted or repelled by the charge being removed (see, e.g., FIG. 5).

Without being bound by a theory of the invention, a decrease in the affinity of the modified variable domain for its binding partner after the original position is changed with an uncharged amino acid (e.g., asparagine (N)) suggests that the binding partner position may already have the opposite charge. The original charged residue may be kept at the amino acid position and other sizes of the same charge may be tested as described in the “size” section below.

Without being bound by a theory of the invention, an increase in the affinity of the modified variable domain for its binding partner after the original position is changed with an uncharged amino acid (e.g., asparagine (N)) suggests that the binding partner residue may have the same charge as the antibody position. The amino acid at the antibody position may be changed with an amino acid residue with a charge opposite to that of the original. If affinity increases again with the opposite charge, the new opposite charged amino acid may be kept and other sizes of this new opposite charged amino acid may be tested as described in the “size” section below. Alternatively, if affinity decreases or remains constant with the opposite charge, the uncharged amino acid (e.g., asparagine (N)) may be kept at the original position and further changes are carried out as described in the “polarity” section below. Alternatively, without being bound by a theory of the invention, if affinity remains constant after the charge is removed, it suggests that the binding partner position may be uncharged, or may have an empty space in which contact with antibody is not occurring. The uncharged amino acid (e.g., asparagine (N)) may be kept at the original amino acid position and further changes may be tested as described in the “polarity” section below.

3. Uncharged Amino Acid Residues

Amino acid residues classified as uncharged (e.g., polar or nonpolar residues) may be changed to a negatively charged amino acid (e.g., aspartic acid (D)) to test whether the binding partner position contains an undetected charged residue (e.g., whether positive or negative) which may be either attracted or repelled by the newly-introduced residue (see, e.g., FIG. 5).

Without being bound by a theory of the invention, an increase in the affinity of the modified variable domain for its binding partner after introducing a negatively charged amino acid (e.g., aspartic acid (D)) suggests that the binding partner position may have a positive charge which attracts the negatively charged residue (e.g., aspartic acid (D)). The negatively charged residue (e.g., aspartic acid (D)) may be kept and other sizes of negatively charged residues may be tested at the position as described in the “size” section below.

Without being bound by a theory of the invention, a decrease in the affinity of the modified variable domain for its binding partner after introducing a negatively charged amino acid (e.g., aspartic acid (D)) suggests that the binding partner position may also have a negative charge which repels the negatively charged amino acid (e.g., aspartic acid (D)), or that it may be sterically hindered if the original uncharged residue was smaller than the negatively charged amino acid (e.g., aspartic acid (D), valine (V), tyrosine (T), serine (S), cysteine (C), alanine (A), or glycine (G)). The negatively charged amino acid (e.g., aspartic acid (D)) may be replaced with a positively charged residue (e.g., lysine (K)) to specifically probe for a negative charge on the binding partner. If affinity increases after introducing the positively charged amino acid, the positively charged residue (e.g., lysine (K)) may be kept at the original position and other sizes of positively charged amino acid resides may be tested at the original position as described in the “size” section below. Alternatively, if affinity remains constant or decreases after introducing a positively charged amino acid (e.g., lysine (K)), the original uncharged residue may be kept and further amino acid changes may be tested as described in the “polarity” section below.

Without being bound by a theory of the invention, if affinity remains constant after introducing a negatively charged amino acid (e.g., aspartic acid (D)), it suggests that the binding partner position may be uncharged, or may have an empty space in which contact with antibody is not occurring. The original uncharged residue may be kept and other amino acid changes may be tested as described in the “polarity” section below.

4. Polar and Non-Polar Amino Acid Residues

Amino acid residues classified as polar may be changed with a new residue of equal size or slightly smaller (see, e.g., FIGS. 4 and/or 5). Without being bound by a theory of the invention, the polar versus nonpolar quality may be switched to test whether changing the chemical function of the sidechain while keeping the size nearly constant may make or break an intermolecular bond. For example, a polar amino acid (e.g., asparagine (N)) may be replaced with a nonpolar amino acid residue of similar size (e.g., leucine (L)), while a nonpolar amino acid (e.g., valine (V)) may be replaced with a polar amino acid of similar size (e.g., threonine (T)), whereas a polar amino acid residue (e.g., tyrosine (Y)) may be replaced with a slightly smaller nonpolar amino acid (e.g., phenylalanine (F)).

Without being bound by a theory of the invention, an increase in binding affinity with the new uncharged sidechain suggests that the switch between polar and nonpolar was beneficial, so the new residue may be kept at the position and different sizes of the new type may be tested as described in the “size” section below.

Without being bound by a theory of the invention, a decrease in binding affinity with the new uncharged sidechain suggests that the switch between polar and nonpolar was detrimental. The original residue may be kept at the position and different sizes of the original type may be tested as described in the “size” section below.

Without being bound by a theory of the invention, if affinity remains constant with the new uncharged sidechain, it may suggest that the binding partner has an empty solvent-filled space in which contact with antibody is not occurring. To test whether any form of contact may be made, a large amino acid residue (e.g., tyrosine (Y)) may be introduced at the position. If affinity increases after introducing the large amino acid (e.g., tyrosine (Y)), the large amino acid residue may be kept at the position and different sizes of polar residues may be tested at the position as described in the “size” section below. Alternatively, if affinity decreases after introducing a large amino acid residue (e.g., tyrosine (Y)), contact may be achieved and the large amino acid residue (e.g., tyrosine (Y)) may not form a favorable bond. The previous polar residue (either original or new) may be kept at the position and different sizes of polar resides may be tested at the site as described in the “size “section below. Alternatively, if affinity remains constant again after introducing a large amino acid (e.g., tyrosine (Y)), it is hypothesized that contact between antibody and the binding partner may not be achieved. The previous polar residue may be kept. Further amino acid changes may not be necessary or desired at the position.

5. Size

A positively charged amino acid residue (e.g., lysine (K), histidine (H) or arginine (R)) may be changed with other positively charged residues to determine which residue when present in an antibody variable domain provides enhanced binding to its binding partner (see, e.g., FIGS. 4 and/or 5). The residue which yields the greatest increase in binding affinity may be kept. Further amino acid changes may not be necessary or desired at the position.

Alternatively, a negatively charged amino acid residue (e.g., aspartic acid (D) or glutamic acid (E)) may be changed with other negatively charged residues to determine which residue when present in an antibody variable domain provides enhanced binding to its binding partner. The residue which yields the greatest increase in binding affinity may be kept. Further amino acid changes may not be necessary or desired at the position.

Further, an uncharged amino acid residue may be changed with the next-smaller sized sidechain in the same class, for example, a polar or a nonpolar amino acid (see, e.g., FIG. 4) to determine which residue when present in an antibody variable domain provides enhanced binding to its binding partner. If affinity increases, smaller residues in the same class may be changed at the position until no more improvement is achieved. The residue which yields the greatest increase in binding affinity may be kept at the position. Further amino acid changes may not be necessary or desired at the position. Alternatively, if affinity remains constant or decreases, the next-larger sized sidechain in the same class may be changed at the position until no further improvement is achieved. The residue which yields the greatest increase in binding affinity may be kept. Further amino acid changes may not be necessary or desired at the position.

Characterization of Amino Acid Residues in an Antibody Variable Domain

The present disclosure provides novel methods which assign each amino acid in an antibody heavy or light chain variable domain to one of the following unique groups, which includes, contacting (C), peripheral (P), supporting (S), interfacial (I), or distant (D) residues, as shown, for example, on the “prox” lines of FIGS. 2A, 2B, 3A, and/or 3B. For example, each of the more-than-200 amino-acid positions in an antibody's light chain variable region and heavy chain variable region has been designated as a member of one of these five novel groups. The “prox” line as shown in FIG. 3A or 3B is useful for any variable region sequence, irrespective of the specific amino acid sequence, such that residues can be selected as candidates for change (e.g., any and/or all contacting (C) residues). Additionally or alternatively, methods are provided that identify the presence of “conspicuous” amino-acid residues which may be candidates for change. Conspicuous amino acid changes may differ in charge or size or chemical functionality from the corresponding residues in the selected sequence (e.g., consensus or germline sequence) and represent positions where amino acid changes may enhance affinity.

Exemplary methods for characterization of amino acid residues in an antibody binding domain may include: a determination of each amino acid residue's proximity group as designated on the “prox” line of FIGS. 2A, 2B, 3A and/or 3B and additionally or alternatively a determination of each amino acid residue's conspicuousness as calculated by the methods provided in the present disclosure. In some embodiments, antibody variable region sequences may be aligned according to a standard numbering system such as Kabat.

A. Determination of Proximity Groups

The characterization process may determine the proximity group for each amino-acid position simply by inspecting the corresponding symbol (“CPSI.:”) on the “prox” lines as shown, for example, in FIGS. 2A and/or 2B. In some embodiments, the antibody's light-chain and/or heavy-chain sequences are aligned with appropriate sequences (e.g., such as consensus or germline sequences) and also with the “prox” lines of the present methods (see, e.g., FIGS. 2A and/or 2B).

Each position in the light chain and heavy chain has been assigned to one of five novel groups designated as contacting (C), peripheral (P), supporting (S), interfacial (I), or distant (D) on the “prox” lines, for example, of FIGS. 2A, 2B, 3A, and/or 3B according to the methods disclosed herein. These Figures (e.g., 2A, 2B, 3A and/or 3B) contain a disc line to reflect disconnection from any significant effect upon an antibody's binding site, and an outw line to reflect outward-orientation on an antibody's surface.

Table 1 shows five proximity groups, as well as a novel designation of disconnection (as shown on a “disc” line, for example, in FIGS. 2A, 2B, 3A and/or 3B) and outward-orientation (shown as an “outw” line, for example, in FIGS. 2A, 2B, 3A and/or 3B) as defined for each group. The number of positions of each type of proximity group for an exemplary antibody (e.g., ING-1, as described herein) in a light chain, a heavy chain, and both chains together are shown in Table 2.

TABLE 1 Proximity Abbr Disc/Outw Contacting C −/+ −/◯ Peripheral P ◯/+ ◯/◯ Supporting S −/− ◯/− Interfacial I −/= ◯/= +/= Distant • +/+ +/◯ +/− p c

TABLE 2 Proximity L H L + H Contacting 16 21 37 Peripheral 3 8 11 Supporting 14 16 30 Interfacial 9 10 19 Distant 70 63 133

Without being bound by a theory of the invention, it has been hypothesized that amino acid residues designated as contacting (C) are located within the combining site (see, e.g., “−” on the “disc” line of FIGS. 2A and/or 2B), and their sidechains are mostly outward-oriented (see, e.g., “+” or “◯” on the outw line). It has been further hypothesized that these are generally surface-exposed residues in the CDR loops themselves, so their sidechains are very favorably situated for making direct contact with corresponding residues on a binding partner.

Without being bound by a theory of the invention, it has been hypothesized that amino acid residues designated as peripheral (P) are slightly disconnected from the binding site (see, e.g., “◯” on the “disc” line), and their sidechains are mostly outward-oriented (see, e.g., “+” or “◯” on the outw line). It has been further hypothesized that many of these are framework residues with variable orientation, which are located at curves or twists in the polypeptide chain not too far from CDR loops. It has been further hypothesized that although they may normally make direct contact with its binding partner, they may possibly make contact if a particular binding partner is bound preferentially toward one side of the binding site instead of being centered.

Without being bound by a theory of the invention, it has been hypothesized that amino acid residues designated as supporting (S) are either directly within or close to the combining site (see, e.g., “−” or “◯” on the “disc” line), and their sidechains are inward-oriented (see, e.g., “−” on the outw line). It has been further hypothesized that many of these residues are buried in the vernier-zone platform directly underneath a combining site, so that their nonpolar sidechains are able to act as conformation-stabilizing “anchors” for binding partner CDR loops which rest on top of them.

Without being bound by a theory of the invention, it has been hypothesized that amino acid residues designated as interfacial (I) may be located anywhere in relation to the binding site (see, e.g., “+” or “◯” or “−” on the “disc” line), but their sidechains form the interface between the light and heavy subunits of the variable domain (see, e.g., “=” on the outw line). It has been further hypothesized that amino acid changes of these residues may cause the two subunits to pivot or rotate relative to one another along their shared hydrophobic interfacial surface, producing strong allosteric effects upon an entire binding site, for example, all six CDR loops may be forced to change their conformation in response.

Without being bound by a theory of the invention, it has been hypothesized that amino acid residues designated as distant (D) are of two different types, with those of the first type being disconnected from a combining site and its targeted epitope (see, e.g., “+” on the “disc” line), and their sidechains may have any orientation except interfacial (see, e.g., “+” or “◯” or “−” but not “=” on the outw line). Distant (D) proximity positions of the second type are the positions (thirteen reclassified from seven peripheral (P), four supporting (S), one interfacial (I) and one distant (D)) containing the conserved prolines or cysteines indicated as “:” on the prox line. It is further hypothesized that amino acid changes at these positions generally will have little or no effect on enhanced affinity to a binding partner.

B. Determination of Conspicuousness

In some embodiments, alternatively or additionally with determination of the proximity groups by inspection of the “prox” lines, the characterization process may involve a calculation of the conspicuousness value for each amino-acid position. The conspicuousness value of a sidechain at a particular antibody position is hypothesized to represent the degree to which it appears strikingly different or unusually outstanding in comparison with selected sequences (e.g., a consensus or germline sequence). Without being bound by a theory of the invention, this value indicates the likelihood that this particular residue may be a somatic mutation which was necessary to confer binding partner-specific affinity upon an antibody. Consequently, the conspicuousness value also correlates with the hypothesis that a new engineered amino acid change at or near this position could possibly lead to forming or strengthening a bond with a residue on a binding partner surface.

Conspicuousness values are calculated by comparing each sidechain of a candidate antibody with the corresponding sidechain of an appropriate consensus or germline sequence, for example, from a mutual alignment. For example, numerical values for conspicuousness can be calculated readily for each amino-acid position in a given antibody, according to the following formula: add 1 point for each three units of difference in size (e.g., divide the absolute value of the size-difference by 3 and drop the decimal without rounding); add 1 point for a shift from one sidechain class to another; add 1 point for each unit (absolute value) of difference in charge, and add 1 point for nonidentity (see, e.g., FIG. 4).

For example, where a single antibody sequence is aligned or compared with a single consensus or germline sequence, there is one “pair” of sequences being compared. The conspicuousness value for each amino-acid position in the alignment or comparison is the sum of the points for chemical function and physical size and nonidentity at that position. Where more than two sequences are aligned or compared together at the same time, each of the antibody sequences may form a separate “pair” with each of the consensus or germline sequences. The conspicuousness values are calculated as described (e.g., sum of function and size and nonidentity) for each pair of sequences being aligned or compared, and then the overall conspicuousness value for each amino acid position in the whole alignment is the sum of the values obtained from each pair at that position, while also adding in a value for repeated identical mutations.

It is hypothesized that nonidentity simply marks an amino-acid position as minimally conspicuous if it displays any kind of difference when compared with a corresponding consensus or germline position. Even a conservative mutation (e.g., from leucine (L) to isoleucine (I) or valine (V)) may suggest a possible bond with a binding partner, especially if a slight change of size or shape was necessary to fine-tune steric relationships between the two molecules.

An exemplary calculation of conspicuousness is illustrated as follows. Four monoclonal antibodies to the same epitope were isolated, and portions of their heavy chains were mutually aligned with a germline sequence, between Kabat positions 25 and 57 [Mendez et al., Nature Genetics, 15:146-152 (1997)] (see, Table 3). Since this alignment contains more than two sequences, each of the four antibody sequences can separately form a “pair” with the one germline sequence. Thus, conspicuousness values are calculated separately for each of the four pairs, and then totaled at each amino-acid position, while also adding in the additional values for repeated identical mutations.

TABLE 3 prox:PSSCSCCCCSISI.I.:...I.ISSCSCCCCCC  pos:    30        40        50 germ:GSISSGGYYWSWIRQHPGKGLEWIGYIYYSGST mAbl:   N  D                  S     N mAb2:      D   T                    N mAb3:  v   D        p         HL    N mAb4:   N  D              DC

Three repetitions are shown in Table 3, at positions 28 and 31 and 56. In each of these cases, an identical amino acid (asparagine (N) or aspartic acid (D)) has appeared at the same location in more than one independently isolated antibody. Accordingly, as described herein, these positions are given very high conspicuousness in the affinity enhancement process. An additional 2 points are added for each repetition of an identical amino acid at a given position (e.g., four aspartic acids amount to three repetitions of the first aspartic acid, so it is worth 3×2=6 points).

In an example, at position 50, the first pair (germ:mAb 1) gets 3 points (Y to S=2 for size+0 for class+0 for charge+1 for nonidentity), the second pair (germ:mAb2) gets 0 points (unmutated Y=0+0+0+0), the third pair (germ:mAb3) gets 3 points (tyrosine (Y) to histidine (H)=0 for size+1 for class+1 for charge+1 for nonidentity), and the fourth pair (germ:mAb4) gets 0 points (unmutated Y=0+0+0+0). The total conspicuousness for position 50 is the sum (3+0+3+0) of these, plus 0 extra points for no repeated identical mutations, which finally gives 6.

In another example, at position 28, the first pair gets 1 point (serine (S) to asparagine (N)=0+0+0+1), the second and third pairs get 0 points, and the fourth pair gets I point. Since the somatic mutation asparagine appears at position 28 twice, it is repeated once, and thus gets 2 extra points. The total conspicuousness for position 28 is the sum (1+0+0+1), plus 2 points for one repetition, which finally gives 4.

In another example, at position 31, each of the four pairs gets 4 points (G to D=1+1+1+1). Since the somatic mutation aspartic acid (D) appears at position 31 four times, it is repeated three times, and thus gets 3×2=6 extra points. The total conspicuousness for position 28 is the sum (4+4+4+4), plus 6 points for three repetitions, which finally gives 22.

The conspicuousness points can be calculated (one pair at a time and then summed) for positions 28, 31, and 50 in the antibody sequence provided in Table 3.

Methods for Targeted Affinity Enhancement

The present disclosure provides methods for the targeted change (e.g., mutation) of one or more selected amino acid residues at one or more positions in an antibody variable domain (e.g., a heavy chain variable region and/or light chain variable region) with other amino acid residues to enhance its binding affinity. Amino acid residues on the antibody's variable domain where targeted amino acid changes are most likely to increase affinity may be selected for change by employing novel methods which assign each amino acid in a variable region of a heavy or light chain to one of the following unique proximity groups: contacting (C), peripheral (P), supporting (S), interfacial (I), or distant (D).

Methods are provided for enhancing the binding affinity of an antibody variable domain (e.g., a heavy chain variable region and/or light chain variable region) to its binding partner, to obtain a modified antibody variable domain with enhanced affinity compared to an unmodified antibody variable domain by identifying the proximity assigned to amino acid positions in the antibody variable domain using the “prox” line as shown in FIG. 3A or 3B; changing amino acid residues at preferably a contacting (C), peripheral (P), supporting (S) and/or interfacial (I), amino acid residue to other amino acid residues, thereby generating a modified variable domain; testing the binding affinity of the modified antibody variable domain; and obtaining a modified antibody variable domain with enhanced binding affinity compared to the binding affinity exhibited by the unmodified antibody variable domain. Less preferably one or more distant (D) amino acid residues may additionally or alternatively be changed according to the disclosed methods.

In some embodiments, modified variable domains (e.g., a heavy chain and/or light chain variable region) may have amino acid changes at preferably one or more contacting (C), peripheral (P), supporting (S) and/or interfacial (I) amino acid residues identified from the “prox” line as shown in FIG. 3A or 3B. Less preferably one or more distant (D) amino acid residues may additionally or alternatively be changed. In some embodiments, the modified variable domains may have 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17 or 18 amino acid changes at preferably a contacting (C), peripheral (P), supporting (S) and/or interfacial (I) amino acid residue identified from the “prox” line as shown in FIG. 3A or 3B. Less preferably multiple distant (D) amino acid residues may additionally or alternatively be changed according to the disclosed methods.

In some embodiments, amino acid residues at preferably one or more contacting (C), peripheral (P), supporting (S) and/or interfacial (I) amino acid residues as identified from the “prox” line as shown in FIG. 3A or 3B, may be changed with at least one of the following amino acid residues: alanine, arginine, asparagine, aspartic acid, cysteine, glutamine, glutamic acid, glycine, histidine, isoleucine, leucine, lysine, methionine, phenylalanine, proline, serine, threonine, tryptophan, tyrosine and valine. In some embodiments, amino acid residues are not changed to cysteine and/or proline. Alternatively, in other embodiments, amino acid residues at preferably a contacting (C), peripheral (P), supporting (S) and/or interfacial (I) amino acid residue as identified from the “prox” line as shown in FIG. 3A or 3B may be changed with each of the above-mentioned amino acid residues. Less preferably one or more distant (D) amino acid residues may additionally or alternatively be changed according to the disclosed methods.

In some embodiments, a modified variable domain may have amino acid changes at preferably two or more contacting (C), peripheral (P), supporting (S) and/or interfacial (I) amino acid residues identified from the “prox” line as shown in FIG. 3A or 3B. Less preferably two or more distant (D) amino acid residues may additionally or alternatively be changed according to the disclosed methods.

In some embodiments, a modified variable domain exhibits enhanced binding affinity compared to the binding affinity exhibited by the parent variable domain. In some embodiments, a modified variable domain exhibits at least a 10%, at least a 15%, at least a 25%, at least a 50%, at least a 75%, at least a 100% (or a two-fold), at least a 5-fold, at least a 10-fold, at least a 50-fold, at least a 100-fold, or more, higher affinity than the corresponding parent variable domain.

Serialization and Exemplary Amino Acid Change Cycle

In some embodiments, methods for affinity enhancement may focus upon contacting (C) residues, and/or may include any of the other proximity groups as candidates for change. For example, one simple approach involves changing only contacting (C) residues while leaving all the others unchanged. Alternatively, all four non-Distant proximity groups could be subjected to an amino acid change cycle to extract a maximized affinity enhancement from a candidate antibody.

In some embodiments, methods for affinity enhancement may be carried out either “vertically” or “horizontally”. These variations may use similar methods, but may process the changes in antibody variable domain residues in a different order. For example, a “vertical” serialization may consist of a series of many sequential stages, each of which operates on a few positions concurrently, or even on only one residue at a time. In contrast, for example, a “horizontal” serialization may consist of a series of one or a few sequential stages, each of which operates on many positions concurrently, or even on many or all of the residues hypothesized to be in or near the combining site at once. In any case, many amino acid changes are applied to each stage in the series before moving on to the next sequential stage. For example, a vertical serialization may carry out multiple steps of a cycle (e.g., 6 steps as shown in FIG. 5) at each individual residue one after the other, whereas a “horizontal” serialization carries out the first step of a cycle on multiple residues at the same time and then proceeds with a second step and subsequent steps.

In some embodiments, serialization may consist of four components: first designating some or all of the proximity groups for change, next partitioning each proximity group into one or more conveniently-sized subgroups, then arranging the amino-acid positions of each subgroup to be processed in sequential order from high to low conspicuousness, and finally scheduling the steps of amino acid changes to be applied to each stage of the series. For example, it may be convenient to process amino acid changes hypothesized to be in an antibody combining site at once. Alternatively, light and/or heavy chains may be processed separately or each of the six CDRs by itself one after another, or even one residue at a time.

In some embodiments, when it is not practical or desirable to change all the residues in a subgroup or proximity group at once, it may be preferable to subdivide a larger group into a series of smaller stages, starting with the highest range of conspicuousness values and proceeding eventually toward lower ranges.

In some embodiments, the total number of amino acid changes and assays in an affinity-enhancement process may be reduced, for example, where positions whose conspicuousness is below some designated cutoff value may not be changed. Such a lower threshold may be important for positions which are not hypothesized to be outward-oriented. For example, one approach might change all the degrees of conspicuousness for the contacting (C) amino acid residues, while changing only a few of the most conspicuous interfacial (I) or supporting (S) or peripheral (P) proximity positions.

In some embodiments, the amino acid changes may be terminated after one or more steps by omitting later steps, especially if it is desirable to minimize the total number of amino acid changes. In other embodiments, a more successful result may be obtained by first carrying out the beginning steps of the cycle on certain amino-acid positions at an early stage of the series, and then completing the rest of the changes on those same positions at a later stage of the series, after other amino-acid positions have already been optimized as desired as a result of the changes of intermediate stages.

One exemplary approach of an amino acid change process is designated herein as epitope mapping. In some embodiments, only the contacting (C) residues may be changed. These contacting (C) positions may be changed to exploratory amino acid changes together in one stage, as one large “horizontal” domain spanning what is hypothesized to be the combining site, and including degrees of conspicuousness at the same time.

Epitope mapping may further include dividing what is hypothesized to be the combining site into mutually-exclusive non-overlapping groups, so that each group may be separately subjected to the amino acid changes.

In some embodiments, a more comprehensive serialization of amino acid changes may span preferred proximity groups (e.g., contacting (C), peripheral (P), supporting (S) and/or interfacial (I) amino acid residues) and involve a larger number of changes to achieve maximum affinity enhancement for the antibody. If necessary, such a project can be broken into a series of stages of more convenient size by partitioning each proximity group into smaller domains and then processing each domain's residues sequentially from high to low conspicuousness.

In some embodiments, the first stage of a serialization may handle any special conditions that are designated (e.g., prolines (P) or cysteines (C) or deletions). The next stage may carry out steps 1, 2, 4, and 5 of a flowcharted amino acid change cycle, as shown for example in FIG. 5, for many or all of the contacting (C) residues, to choose the sidechain's chemical function without yet proceeding to steps 3 or 6 where the sidechain's size may be fine-tuned. Without being bound by a theory of the invention, these initial steps (e.g., adding or removing a charge followed by testing opposite charges and then switching between polar and nonpolar) may generate a comprehensive map of the binding partner's sidechains (e.g., suggesting likely chemical functions and intermolecular bonds) at the interface of the antibody-binding partner complex.

In some embodiments, exploratory amino acid changes may be made according to a step-by-step “decision tree” embodied within an exemplary flowcharted amino acid change cycle, for example, as shown in FIG. 5 and/or FIG. 6. Such a flowchart may be used for each position, regardless of proximity or conspicuousness. Further, depending upon the chemical properties of the amino acid's sidechain, a flowcharted amino acid change cycle may be entered at one of several (e.g., three) alternative starting points. For example, using FIG. 5, one of six starting points may be entered.

In some embodiments, steps in the exemplary flowcharted amino acid change cycle can be skipped or modified. Without being bound by any theory of the invention it has been hypothesized that many supporting (S) proximity positions are evolutionarily conserved as inward-oriented (“−” on the outw line of FIG. 2A or 2B) and buried in the vernier-zone platform directly underneath the combining site, such that their nonpolar sidechains are able to act as conformation-stabilizing anchors for the binding partner's CDR loops which rest on top of them. Without being bound by any theory of the invention it has been hypothesized that interfacial (I) proximity positions (“=” on the outw line) are sometimes nonpolar residues buried along the hydrophobic interfacial surface between light and heavy subunits. Mutational operations on these kinds of positions might advantageously skip the “uncharged” and “polarity” sections, and instead begin immediately with the “size” section.

Methods of Making Antibody Variable Domains with Enhanced Affinity to a Binding Partner

Methods are provided for enhancing the binding affinity of an antibody by means of producing targeted amino acid changes in the antibody's variable domain (e.g., a heavy chain variable region and/or light chain variable region). For example, changes may be introduced at one or more amino acid residues by using an exemplary flowcharted amino acid change cycle, as described herein. Additionally or alternatively, amino acid residues may be selected for change by using the novel methods for assigning the residues in an antibody variable region to one of five unique groups, for example, contacting (C), peripheral (P), supporting (S), interfacial (I) amino and/or distant (D) residues, where changes are likely to produce enhanced affinity. Additionally or alternatively, positions for amino acid changes may be based upon a novel method of determining the degree to which the original amino acid residue differs from the corresponding consensus or germline residue in terms of charge, size or chemical functionality to calculate “conspicuousness.” In some embodiments, antibody variable region sequences may be aligned according to a standard numbering system such as Kabat.

In an exemplary method, amino acid changes are engineered at preferably one or more contacting (C), peripheral (P), supporting (S) and/or interfacial (I) amino acid residue as identified from the “prox” lines of FIGS. 2A, 2B, 3A, and/or 3B. Less preferably one or more distant (D) amino acid residues may additionally or alternatively be changed according to the disclosed methods.

For example, methods are provided for making a modified variable domain (e.g., a heavy chain and/or light chain variable region) of an antibody with enhanced binding affinity to a binding partner by modifying the nucleotide sequence of an antibody variable domain at a sites that preferably encode one or more contacting (C), peripheral (P), supporting (S) and/or interfacial (I) amino acid residues identified from the “prox” line as shown in FIGS. 3A and/or 3B, thereby generating a modified antibody variable domain; and selecting a modified variable domain that has enhanced binding affinity to its binding partner. Less preferably one or more distant (D) amino acid residues may additionally or alternatively be changed according to the disclosed methods.

In some embodiments, one or more contacting (C) residues to be changed may be in complementarity determining region-1 (CDR1) in a light chain variable region. In certain embodiments, the contacting (C) residues may be at a position corresponding to position 28, 30 and/or 31 in CDR1. In other embodiments, one or more contacting (C) residues to be changed may be in CDR2 in a light chain variable region. In certain embodiments, the contacting (C) residues may be at a position corresponding to position 50, 51 and/or 53 in CDR2. In other embodiments, the contacting (C) residue may be in CDR3 in a light chain variable region. In some embodiments, one or more contacting (C) residues to be changed may be in CDR1 in a heavy chain variable region. In certain embodiments, the contacting (C) residues may be at a position corresponding to position 32 and/or 33 in CDR1. In some embodiments, one or more contacting (C) residues to be changed may be in CDR2 in a heavy chain variable region. In certain embodiments, the contacting (C) residues may be at a position corresponding to position 50, 52, 53, 54, 56, and/or 58 in CDR2. In some embodiments, the contacting (C) residue may be in CDR3 in a heavy chain variable region. The CDRs (e.g., LCDR1, LCDR2 and LCDR3 for the light chain and HCDR1, HCDR2 and HCDR3 for the heavy chain) may be defined according to any known method in the art including, for example, Kabat, Chothia or IMGT. According to Kabat, LCDR1 comprises amino acid residues 24 to 34, LCDR2 comprises amino acid residues 50 to 56, LCDR3 comprises amino acid residues 89 to 97, HCDR1 comprises amino acid residues 31 to 35b, HCDR2 comprises amino acid residues 50 to 65 and HCDR3 comprises amino acid residues 95 to 102. According to Chothia, LCDR1 comprises amino acid residues 24 to 34, LCDR2 comprises amino acid residues 50 to 56, LCDR3 comprises amino acid residues 89 to 97, HCDR1 comprises amino acid residues 26 to 32, HCDR2 comprises amino acid residues 52 to 56 and HCDR3 comprises amino acid residues 95 to 102. According to IMGT, LCDR1 comprises amino acid residues 27 to 32, LCDR2 comprises amino acid residues 50 to 52, LCDR3 comprises amino acid residues 89 to 97, HCDR1 comprises amino acid residues 26 to 33, HCDR2 comprises amino acid residues 51 to 57 and HCDR3 comprises amino cid residues 93 to 102. Residues numbers for the Kabat, Chothia and IMGT CDRs are given as Kabat position numbers.

Modified variable domains are synthesized by modifying the nucleic acid of a parent variable domain (e.g., a heavy chain variable region and/or light chain variable region), inserting the modified nucleic acid into an appropriate cloning vector and expressing the modified nucleic acid to produce modified variable domains. Exemplary protocols are described below.

1. Making Modified Variable Domain Nucleic Acid

A modified variable domain(s) comprises one or more amino acid sequence changes (e.g., changes) relative to a parent variable domain sequence to provide for enhanced binding affinity to a binding partner compared to the parent variable domain.

One or more amino acid residues to be changed may be identified by the novel flowcharted amino acid change cycle, as described herein. Additionally or alternatively, amino acid residues may be selected for change by using the novel methods for assigning the residues in an antibody variable region to one of five unique groups, for example, contacting (C), peripheral (P), supporting (S), interfacial (I) amino and/or distant (D) residues, where changes are likely to produce enhanced affinity. Additionally or alternatively, positions for amino acid changes may be based upon a novel method of determining the degree to which the original amino acid residue differs from the corresponding consensus or germline residue in terms of charge, size or chemical functionality to calculate “conspicuousness.”

In some embodiments, a modified variable domain may have amino acid changes at preferably one or more contacting (C), peripheral (P), supporting (S) and/or interfacial (I) amino acid residues identified from the “prox” line as shown in FIGS. 3A and/or 3B. In some embodiments, modified variable domains may be constructed comprising amino acid changes at preferably 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19 or 20 contacting (C), peripheral (P), supporting (S) and/or interfacial (I) amino acid residues identified from the “prox” line as shown in FIGS. 3A and/or 3B. Less preferably 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17 or 18 distant (D) amino acid residues (as shown in FIGS. 3A and/or 3B as “” or the “:” on the prox line) may additionally or alternatively be changed according to the disclosed methods. The twenty amino acids may include alanine (Ala, A), arginine (Arg, R), asparagine (Asn, N), aspartic acid (Asp, D), cysteine (Cys, C), glutamine (Gln, Q), glutamic acid (Glu, E), glycine (Gly, G), histidine (H is, H), isoleucine (Ile, I), leucine (Leu, L), lysine (Lys, K), methionine (Met, M), phenylalanine (Phe, F), proline (Pro, P), serine (Ser, S), threonine (Thr, T), tryptophan (Trp, W), tyrosine (Tyr, Y) or valine (Val, V). In some embodiments, cysteine and/or proline are not included.

In some embodiments, amino acid residues at preferably one or more contacting (C), peripheral (P), supporting (S) and/or interfacial (I) amino acid residue, as identified from the “prox” line as shown in FIGS. 3A and/or 3B, may be changed with one or more of the following preferred amino acid residues: alanine, arginine, asparagine, aspartic acid, glutamine, glutamic acid, glycine, histidine, isoleucine, leucine, lysine, methionine, phenylalanine, serine, threonine, tryptophan, tyrosine and valine. In other embodiments, amino acid residues at preferably one or more contacting (C), peripheral (P), supporting (S) and/or interfacial (I) amino acid residues as identified from the “prox” line as shown in FIGS. 3A and/or 3B may be changed with alanine, arginine, asparagine, aspartic acid, glutamine, glutamic acid, glycine, histidine, isoleucine, leucine, lysine, methionine, phenylalanine, serine, threonine, tryptophan, tyrosine or valine. Less preferably one or more distant (D) amino acid residues may additionally or alternatively be changed according to the disclosed methods.

In some embodiments, a modified variable domain may have amino acid changes at preferably two or more contacting (C), peripheral (P), supporting (S) and/or interfacial (I) amino residues identified from the “prox” line as shown in FIGS. 3A and/or 3B. Less preferably one or more distant (D) amino acid residues may additionally or alternatively be changed according to the disclosed methods.

In some embodiments, a modified variable domain exhibits enhanced binding affinity to a binding partner compared to the binding affinity exhibited by the parent variable domain. In some embodiments, a modified variable domain exhibits at least a 10%, at least a 15%, at least a 25%, at least a 50%, at least a 75%, at least a 100% (or a two-fold), at least a 5-fold, at least an 8-fold, at least a 10-fold, at least a 50-fold, at least a 100-fold, or more, higher affinity to a binding partner than the corresponding parent variable domain.

Modified variable domains may be generated which contain amino acid changes at preferably multiple contacting (C), peripheral (P), supporting (S) and/or interfacial (I) amino acid residues as designated on the “prox” line of FIGS. 2A, 2B, 3A and/or 3B). Less preferably one or more distant (D) amino acid residues may additionally or alternatively be changed according to the disclosed methods.

In some embodiments, degenerate primers are used to perform one PCR reaction. In other embodiments, each degenerate primer may be used in a separate PCR reaction. Any combination of PCR primers may be used in a PCR reaction.

DNA encoding modified variable domains may be prepared by a variety of methods known in the art. These methods include, but are not limited to, preparation by oligonucleotide-mediated (or site-directed) mutagenesis, PCR mutagenesis, Dpn1 mutagenesis, Kunkel mutagenesis and cassette mutagenesis of an earlier prepared modified variable domain or parent variable domain. These techniques may utilize antibody nucleic acid (DNA or RNA), or nucleic acid complementary to the antibody nucleic acid. Oligonucleotide-mediated mutagenesis (e.g., PCR mutagenesis) may be used for preparing substitution, deletion, and insertion variants of antibody DNA (see, e.g., Adelman et al. (1983) DNA, 2: 183). Briefly, the antibody DNA is altered by hybridizing an oligonucleotide encoding the desired mutation to a DNA template, where the template is the single-stranded form of a plasmid or bacteriophage containing the unaltered or native DNA sequence of the binding partner polypeptide. After hybridization, a DNA polymerase is used to synthesize an entire second complementary strand of the template that will thus incorporate the oligonucleotide primer, and will code for the selected alteration in the binding partner polypeptide.

DNA encoding a modified variable domain with one or more amino acid residues to be changed may be generated in one of several ways. If the amino acids are located close together in the polypeptide chain, they may be mutated simultaneously using one oligonucleotide that codes for all of the desired amino acid changes. If, however, the amino acids are located some distance from each other (separated by more than about ten amino acids), it is more difficult to generate a single oligonucleotide that encodes all of the desired changes. Instead, one of two alternative methods may be employed.

In the first method, a separate oligonucleotide is generated for each amino acid to be changed. The oligonucleotides are then annealed to the single-stranded template DNA simultaneously, and the second strand of DNA that is synthesized from the template will encode all of the desired amino acid changes.

The alternative method involves two or more rounds of mutagenesis to produce the desired mutant antibody. The first round is as described for the modified variable domain which comprise one amino acid change: wild-type DNA is used for the template, an oligonucleotide encoding the first desired amino acid change(s) is annealed to this template, and the heteroduplex DNA molecule is then generated. The second round of mutagenesis utilizes the mutated DNA produced in the first round of mutagenesis as the template. Thus, this template already contains one or more mutations. The oligonucleotide encoding the additional desired amino acid change(s) is then annealed to this template, and the resulting strand of DNA now encodes mutations from both the first and second rounds of mutagenesis. This resultant DNA can be used as a template in a third round of mutagenesis, and so on.

2. Insertion of DNA into a Cloning and/or Expression Vehicle

The cDNA or genomic DNA encoding the modified antibody variable domain may be inserted into a replicable vector for further cloning (amplification of the DNA) or for expression. Many vectors are available, and selection of the appropriate vector will depend on 1) whether it is to be used for DNA amplification or for DNA expression, 2) the size of the DNA to be inserted into the vector, and 3) the host cell to be transformed with the vector. Each vector contains various components depending on its function (amplification of DNA or expression of DNA) and the host cell for which it is compatible. The vector components generally include, but are not limited to, one or more of the following: a signal sequence, an origin of replication, one or more marker genes, an enhancer element, a promoter, and a transcription termination sequence.

For example, the cDNA or genomic DNA encoding the modified variable domain may be inserted into a modified phage vector (i.e. phagemid) for expression. Often, filamentous phage such as M13, f1 or fd are used. Filamentous phage contain single-stranded DNA surrounded by multiple copies of genes encoding major and minor coat proteins, e.g., pill. Coat proteins are displayed on the capsid's outer surface. DNA sequences inserted in-frame with capsid protein genes are co-transcribed to generate fusion proteins or protein fragments displayed on the phage surface. The peptides expressed on phage can then bind target molecules, i.e., they can specifically interact with binding partner molecules such as antibodies (Petersen (1995) Mol. Gen. Genet. 249:425-31), cell surface receptors (Kay (1993) Gene 128:59-65), and extracellular and intracellular proteins (Gram (1993) J. Immunol. Methods 161:169-76).

In addition to phage or phagemid expression of antibody variable domains, the methods of the disclosure can also use yeast surface expression (see, e.g., Boder (1997) Nat. Biotechnol. 15:553-557), using such vectors as the pYD1 yeast expression vector. Other potential expression systems include mammalian display vectors and E. coli libraries, and/or in vitro transcription/translation systems. Thus, any expression and/or display system, such as ribosome expression/display, phage expression/display, bacterial expression/display, yeast expression/display, arrayed expression/display or any other suitable expression/display system known in the art may be used in methods of the disclosure.

An antibody or antibody fragment, e.g., a scFv, Fab or Fv may be expressed on the surface of a phage. Exemplary antibody phage methods are known to those skilled in the art and are described, e.g., in Hoogenboom, Overview of Antibody Phage-Display Technology and Its Applications, from Methods in Molecular Biology: Antibody Phage Display: Methods and Protocols (2002) 178:1-37 (O'Brien and Aitken, eds., Human Press, Totowa, N.J.). For example, antibodies or antibody fragments (e.g., scFvs, Fabs, Fvs with an engineered intermolecular disulfide bond to stabilize the VH-VL pair, and diabodies) can be expressed on the surface of a filamentous phage, such as the nonlytic filamentous phage fd or M13. Antibodies or antibody fragments with the desired binding specificity can then be selected.

In some embodiments, cDNA is cloned into a phagemid vector (e.g., pCES 1, pXOMA Fab or pXOMA Fab-gIII). In certain embodiments, cDNA encoding both heavy and light chains may be present on the same vector. In some embodiments, cDNA encoding scFvs are cloned in frame with all or a portion of gene III, which encodes the minor phage coat protein pill. The phagemid directs the expression of the scFv-pIII fusion on the phage surface. In other embodiments, cDNA encoding heavy chain (or light chain) may be cloned in frame with all or a portion of gene III, and cDNA encoding light chain (or heavy chain) is cloned downstream of a signal sequence in the same vector. The signal sequence directs expression of the light chain (or heavy chain) into the periplasm of the host cell, where the heavy and light chains assemble into Fab fragments. Alternatively, in certain embodiments, cDNA encoding heavy chain and cDNA encoding light chain may be present on separate vectors. In certain embodiments, heavy chain and light chain cDNA may be cloned separately, one into a phagemid and the other into a phage vector, which both contain signals for in vivo recombination in the host cell.

General recombinant DNA techniques, e.g., manipulation of nucleic acids, epitopes, antibodies, and vectors of interest, subcloning into expression vectors, labeling probes, sequencing DNA, DNA hybridization are described in the scientific and patent literature, see e.g., Sambrook and Russell, eds., Molecular Cloning: a Laboratory Manual (3rd), Vols. 1-3, Cold Spring Harbor Laboratory Press, (2001); Current Protocols in Molecular Biology, Ausubel, ed. John Wiley & Sons, Inc., New York (1997-2001) (“Ausubel”); and, Laboratory Techniques in Biochemistry and Molecular Biology Hybridization with Nucleic Acid Probes, Part I. Theory and Nucleic Acid Preparation, Tijssen, ed. Elsevier, N.Y. (1993). Sequencing methods typically use dideoxy sequencing, however, other methodologies are available and well known to those of skill in the art.

3. Transformation of Host Cells

Suitable host cells for cloning or expressing the vectors herein may include prokaryote, yeast, or higher eukaryote cells. Suitable prokaryotes include eubacteria, such as Gram-negative or Gram-positive organisms, for example, E. coli, Bacilli such as B. subtilis, Pseudomonas species such as P. aeruginosa, Salmonella typhimurium, or Serratia marcescens.

For example, recombinant phagemid or phage vectors may be introduced into a suitable bacterial host, such as E. coli. In some embodiments using phagemid, the host may be infected with helper phage to supply phage structural proteins, thereby allowing expression of phage particles carrying the antibody-pill fusion protein on the phage surface.

Methods for Identifying an Antibody Variable Domain Having Enhanced Affinity for a Binding Partner

Methods are provided for identifying a modified variable domain of an antibody having enhanced binding affinity for a binding partner by contacting a parent variable domain with the binding partner under conditions that permit binding; contacting modified variable domains made by the methods of the present disclosure with the binding partner under conditions that permit binding; and determining binding affinity of the modified variable domains and the parent variable domain for the binding partner, wherein modified variable domains that have a binding affinity for the binding partner greater than the binding affinity of the parent variable domain for the binding partner are identified as having enhanced binding affinity for the binding partner.

Each binding assay needs to detect only whether the antibody's affinity has increased or decreased or remained the same. In addition, there is no need to sequence the entire antibody-fragment gene to ensure that additional spurious mutations have not been inadvertently introduced. Instead, rapid DNA hybridization to detect the desired mutant can be followed by pooling of multiple independent isolates of the same mutant, so that any unwanted aberrations in sequence are sufficiently diluted out in the final mixture and do not appreciably affect the results of the binding assay.

Isolated variable domains may exhibit binding affinity as single chains, in the absence of assembly into a heteromeric structure with their respective VH or VL subunits. As such, populations of VH and VL altered variable domains can be expressed alone and screened for binding affinity having substantially the same or greater binding affinity compared to the parent VH or VL variable domain.

Alternatively, populations of VH and VL altered variable domains polypeptides can be co-expressed so that they self-assemble into heteromeric altered variable domain binding fragments. The heteromeric binding fragment population can then be screened for species exhibiting enhanced binding affinity to a binding partner compared to the binding affinity of the parent variable domain to the binding partner.

The expressed population of modified variable domains can be screened for the identification of one or more altered variable domain species exhibiting enhanced binding affinity to a binding partner as compared with the parent variable domain. Screening can be accomplished using various methods well known in the art for determining the binding affinity of a polypeptide or compound. Additionally, methods based on determining the relative affinity of binding molecules to their partner by comparing the amount of binding between the modified variable domain and the binding partner can similarly be used for the identification of species exhibiting binding affinity substantially the same or greater than the parent variable domain to the binding partner. These methods can be performed, for example, in solution or in solid phase. Various formats of binding assays are well known in the art and include, for example, immobilization to filters such as nylon or nitrocellulose; two-dimensional arrays, enzyme linked immunosorbant assay (ELISA), radioimmuno-assay (RIA), panning and plasmon resonance (see, e.g., Sambrook et al., supra, and Ansubel et al., supra).

Modified variable domains produced by the methods of the disclosure may be screened by immobilization of the modified variable domains to filters or other solid substrates. Such filter lifts allow for the identification of modified variable domains that exhibit enhanced binding affinity compared to the parent variable domain to the binding partner.

Alternatively, the modified variable domains may be expressed on the surface of a cell or bacteriophage, for example, panning on immobilized binding partner can be used to efficiently screen for the relative binding affinity of species within the population of modified variable domains and for those which exhibit enhanced binding affinity to the binding partner than the parent variable domain.

Another affinity method for screening populations of modified variable domains is a capture lift assay that is useful for identifying a binding molecule having selective affinity for a ligand. This method employs the selective immobilization of modified variable domains to a solid support and then screening of the selectively immobilized modified variable domains for selective binding interactions against the binding partner. Selective immobilization functions to increase the sensitivity of the binding interaction being measured since initial immobilization of a population of modified variable domains onto a solid support reduces non-specific binding interactions with irrelevant molecules or contaminants which can be present in the reaction.

Another method for screening populations or for measuring the affinity of individual modified variable domains is through surface plasmon resonance (SPR). This method is based on the phenomenon which occurs when surface plasmon waves are excited at a metal/liquid interface. Light is directed at, and reflected from, the side of the surface not in contact with sample, and SPR causes a reduction in the reflected light intensity at a specific combination of angle and wavelength. Biomolecular binding events cause changes in the refractive index at the surface layer, which are detected as changes in the SPR signal. The binding event can be either binding association or disassociation between a receptor-ligand pair. The changes in refractive index can be measured essentially instantaneously and therefore allows for determination of the individual components of an affinity constant. More specifically, the method enables accurate measurements of association rates (kon) and disassociation rates (koff).

If desired, kinetic binding assays (using a Biacore instrument such as a Biacore 2000 or Biacore A100) can optionally be incorporated into the process to separately optimize the on-rate and the off-rate of the antibody's total affinity constant, since certain mutations will have much more effect upon one rate than the other. The off-rate is a function of resistance to dissociation, which can be improved in many different ways by targeted engineered mutations. The on-rate is a function of two different components: activation and diffusion. The activation component can be improved mutationally by relieving conformational stresses which interfere with the initial steps of the binding reaction. However, the diffusion component cannot be significantly changed by mutations, since it is determined mostly by the molecular weight of the antibody fragment and by the viscosity of the liquid solution. Kinetic binding assays make it possible to identify mutations which make the on-rate faster or the off-rate slower, as well as those which improve both rates.

Measurements of kon and koff values can be advantageous because they can identify modified variable domains with enhanced binding affinity for a binding partner. For example, an modified variable domain can be more efficacious because it has, for example, a higher kon valued compared to the parent variable domain. Increased efficacy is conferred because molecules with higher kon values can specifically bind and inhibit their binding partner at a faster rate. Similarly, a modified variable domain can be more efficacious because it exhibits a lower koff value compared to molecules having similar binding affinity. Increased efficacy observed with molecules having lower koff rates can be observed because, once bound, the molecules are slower to dissociate from their binding partner.

Methods for measuring the affinity, including association and disassociation rates using surface plasmon resonance are well known in the arts and can be found described in, for example, Jonsson and Malmquist, Advances in Biosensors, 2:291-336 (1992) and Wu et al. Proc. Natl. Acad. Sci. USA, 95:6037-6042 (1998). Moreover, one apparatus well known in the art for measuring binding interactions is a Biacore 2000 instrument and another is a Biacore A100 instrument which are commercially available through GE Healthcare (Biacore, Inc., Piscataway, N.J. 08854).

Using any of the above described screening methods, as well as others well known in the art, a modified variable domain having binding affinity substantially the same or greater than the parent variable domain is identified by detecting the binding of at least one altered variable domain within the population to its binding partner.

Detection methods for identification of species within the population of modified variable domains can be direct or indirect and can include, for example, the measurement of light emission, radioisotopes, calorimetric dyes and fluorochromes. Direct detection includes methods that operate without intermediates or secondary measuring procedures to assess the amount of the binding partner bound by the modified variable domain. Such methods generally employ ligands that are themselves labeled by, for example, radioactive, light emitting or fluorescent moieties. In contrast, indirect detection includes methods that operate through an intermediate or secondary measuring procedure. These methods generally employ molecules that specifically react with the binding partner and can themselves be directly labeled or detected by a secondary reagent. For example, a modified variable domain specific for a binding partner can be detected using an antibody capable of interacting with the modified variable domain, again using the detection methods described above for direct detection. Indirect methods can additionally employ detection by enzymatic labels. Moreover, for the specific example of screening for catalytic antibodies, the disappearance of a substrate or the appearance of a product can be used as an indirect measure of binding affinity or catalytic activity.

In some embodiments, the modified variable domain has a binding affinity for the binding partner greater than the binding affinity of the parent variable domain for the binding partner and thus is identified as having enhanced binding affinity for the binding partner.

In some embodiments, a modified variable domain exhibits enhanced binding affinity to a binding partner compared to the binding affinity between the parent variable domain and the binding partner. In some embodiments, a modified variable domain exhibits an at least 10%, at least 15%, at least 25%, at least 50%, at least 75%, at least 100% (or two-fold), at least 5-fold, at least 8-fold, at least 10-fold, at least 50-fold, at least 100-fold, or more, higher affinity than the corresponding parent variable domain.

In other embodiments, the modified variable domain has a binding affinity for the binding partner less than the binding affinity of the parent variable domain for the binding partner and thus is identified as having reduced binding affinity for the binding partner.

In some embodiments, where the candidate antibody has specificity for a family of structurally similar binding partners, the antibody's binding affinity may be directed toward a related (but not identical) binding partner simply by carrying out the binding assays using the new binding partner instead of the original binding partner.

Example 1 ING-1 Mutation Analysis

Arrays of modified antibody variable domains (e.g., modified ING-1 or C5A (XPA23) variable domains comprising a heavy chain variable region and a light chain variable region) with amino acids changes at desired positions (e.g., contacting (C) residues) may be generated and tested for enhanced binding affinity compared to the parent variable domain (e.g., ING-1 or C5A (XPA23)).

In exemplary experiments, each contacting (C) residue in the heavy and light chain variable regions of ING-1 or C5A (XPA23) is separately changed (e.g., by PCR mutagenesis) with alanine, arginine, asparagine, aspartic acid, glutamine, glutamine acid, glycine, histidine, isoleucine, leucine, lysine, phenylalanine, proline, serine, threonine, tryptophan, tyrosine and valine to generate modified ING-1 variable domains. DNAs encoding the modified ING-1 or C5A (XPA23) variable domains are then inserted into a vector and used to transform electrocompetent cells. The clones are plated on plates (e.g., Teknova) and the plates filled with growth media in each well (e.g., Teknova). Each well is inoculated with a single colony comprising a single amino acid change at a contacting (C) residue. The colonies are grown by incubating the plates. After the incubation, the plates are duplicated to sequencing plates by filling new deep-well culture plates (e.g., Thomson) with media in each well from the grown cultures. A 96-pin replicator is used to transfer cells from the master plate to the new sequencing plates. The sequencing plates are grown. After the incubation, the sequencing plate is spun down and the supernatant is discarded. Samples from the plate are sequenced (e.g., samples may be submitted for automated miniprep and automated sequencing. After the incubation, Master Plates are made by adding glycerol to the wells on the glycerol plate and storing the plates at −80° C. The unique clones and their well position in the master plate are identified after sequencing results are returned.

Eighteen different clones, each containing an amino acid change at a contacting (C) residue in ING-1 or C5A (XPA23), are identified (typically 96 sequenced clones yield all eighteen clones). Unique clones from the master plates are rearrayed to a new 96-well master plate containing growth media by transferring an aliquot of glycerol stock from the master plate to the rearrayed master plate. Alternatively, automation, is used to transfer the glycerol stock containing the unique clones to the new master plate. The new rearrayed glycerol master plates are replicated into new expression plates to perform Biacore analysis. Data from the exemplary ING-1 experiment are shown in Table 4 and Table 5. Data from the exemplary C5A (XPA23) experiment are shown in Table 6 and Table 7.

TABLE 4 Biacore Analysis of Modified ING-1 Light Chain Variable Regions1, 2 3 NP Aromatic Neg Pos Polar D E R K H Y W F Q N S CDR1 K 27 1.26 −1.00 1.26 ? 1.06 nd −1.00 1.62 1.52 −1.00 −1.00 S 28 1.63 1.02 2.78 2.32 1.90 2.02 nd 2.38 1.99 −1.00 1.08 L 29 −1.00 −1.00 −1.00 nd nd −1.00 −1.00 −1.00 −1.00 1.85 2.03 L 30 1.47 −1.00 −1.00 −1.00 1.45 1.53 −1.00 1.56 1.60 1.45 1.43 H 31 0.71 0.68 0.06 0.05 0.95 2.16 1.66 nd 0.57 0.50 0.82 S 32 0.94 nd 1.79 1.32 1.13 1.37 1.27 1.64 1.10 −1.00 ? N 33 0.49 0.65 0.71 0.70 0.73 0.80 1.04 0.93 0.73 1.38 0.48 I 35 0.19 0.16 0.92 0.61 0.59 0.51 0.34 0.66 0.41 0.50 0.55 T 36 0.05 1.60 −1.00 1.15 0.79 nd 1.30 nd 1.04 0.74 1.10 Y 37 nd 0.01 nd 0.02 4.07 0.95 0.85 0.63 0.02 0.06 0.09 CDR2 Y 54 0.03 0.05 −1.00 3.62 −1.00 0.92 0.96 0.94 1.23 −1.00 0.90 Q 55 0.05 0.05 5.31 0.46 3.82 nd 4.11 0.86 0.95 0.36 0.56 M 56 1.36 0.71 0.92 0.98 1.32 1.21 1.29 1.40 1.12 0.99 1.05 S 57 0.95 0.93 1.17 1.54 1.01 −1.00 2.34 0.96 −1.00 1.17 1.00 N 58 nd 0.97 1.77 1.40 1.16 1.43 1.99 1.03 1.55 0.95 1.65 CDR3 L 97 −1.00 0.75 −1.00 0.61 0.42 0.98 1.59 0.93 0.91 0.48 0.79 E 98 1.62 0.98 3.08 2.22 1.23 1.23 1.10 1.43 1.41 −1.00 −1.00 L 99 0.02 0.01 0.04 0.02 0.05 1.00 0.89 0.43 0.02 2.00 0.04 P 100 0.02 0.06 0.05 0.03 0.05 1.94 1.51 1.65 0.04 0.05 0.06 R 101 −1.00 −1.00 0.93 0.04 −1.00 −1.00 −1.00 −1.00 −1.00 −1.00 −1.00 NP Aliphatic Small NP T V I L A C G P M CDR1 K 27 −1.00 nd 1.15 1.39 1.10 nd 1.11 1.29 nd S 28 1.19 2.15 2.35 2.60 1.53 nd nd 2.42 nd L 29 nd nd −1.00 ? 1.97 nd 1.53 −1.00 nd L 30 1.89 1.69 0.96 0.97 1.27 nd 1.64 0.82 nd H 31 1.94 1.17 1.14 0.59 1.26 nd 0.73 1.19 nd S 32 nd 0.93 1.17 1.12 1.35 nd 1.39 0.83 nd N 33 0.60 −1.00 nd nd 0.76 nd 0.72 0.67 nd I 35 0.87 1.08 nd 0.60 0.46 nd 0.39 0.70 nd T 36 0.94 0.98 0.76 1.64 1.02 nd 1.09 0.67 nd Y 37 −1.00 −1.00 0.90 nd 0.04 nd 1.30 nd nd CDR2 Y 54 0.61 1.32 −1.00 3.44 0.08 nd 1.85 0.86 nd Q 55 0.66 1.53 1.42 0.71 0.64 nd 0.70 0.95 nd M 56 0.80 −1.00 1.37 0.74 0.86 nd 0.80 1.38 nd S 57 0.86 0.89 0.98 1.38 1.15 nd −1.00 nd nd N 58 1.42 2.84 2.51 1.47 1.79 nd 1.87 3.47 nd CDR3 L 97 0.54 1.44 2.62 0.93 0.50 nd −1.00 0.95 nd E 98 4.90 nd nd 2.82 1.35 nd 1.63 −1.00 nd L 99 0.09 1.04 2.07 0.93 1.43 nd 0.02 0.01 nd P 100 0.08 0.14 0.14 0.03 0.05 nd 1.62 1.01 nd R 101 −1.00 −1.00 −1.00 1.33 −1.00 nd −1.00 −1.00 nd 1A value of −1 indicates no binding 2Bolded values indicate the highest affinity (as measured by how many “fold” differences in affinity the mutant is in comparison to original, e.g., 2.0 as twice as strong and 0.5 as half as strong) obtained for an amino acid change at the position 3nd indicates that binding affinity was not determined

TABLE 5 Biacore Analysis of Modified ING-1 Heavy Chain Variable Regions1, 2, 3 NP Aromatic Neg Pos Polar D E R K H Y W F Q N S CDR1 T 28 0.98 1.23 1.23 1.77 1.15 1.86 1.08 1.28 −1.00 0.69 1.07 T 30 0.63 0.73 1.39 0.94 nd 2.00 1.26 nd nd 0.73 0.91 K 31 0.66 0.54 0.76 0.96 0.78 1.00 1.02 nd nd −1.00 0.49 Y 32 nd 0.08 0.43 0.08 0.60 0.84 nd 1.05 0.10 nd 0.01 G 33 −1.00 −1.00 0.03 −1.00 0.02 6.16 −1.00 7.19 0.06 −1.00 nd CDR2 W 50 3.27 −1 0.10 0.04 0.02 0.04 0.97 0.09 0.01 0.03 0.02 N 52 0.02 −1 −1 −1 0.02 −1 −1 −1 −1 0.98 −1 T 53 −1 −1 −1 1.79 −1 −1 −1 −1 −1 −1 0.17 Y 54 0.05 0.07 3.72 3.62 1.00 0.92 0.96 0.65 0.66 2.11 0.49 T 55 0.03 −1 0.14 0.45 0.05 0.03 0.10 0.03 0.03 0.17 0.42 E 56 0.81 0.95 1.34 1.27 1.74 1.04 1.17 0.78 1.23 1.01 1.46 E 57 1.17 1.07 1.71 nd 1.16 1.37 1.39 1.06 −1.00 1.57 1.41 P 58 0.54 0.44 nd 1.14 nd 0.99 1.11 0.98 1.11 0.90 1.07 T 59 0.87 0.51 1.22 1.43 0.40 nd 2.24 0.43 −1.00 0.96 nd CDR3 G 100 −1 −1 7.51 1.59 −1 −1 −1 −1 −1.00 1.55 1.68 S 101 0.21 0.76 nd 2.20 1.35 1.79 1.22 1.16 2.18 0.97 nd A 102 0.28 0.51 2.18 1.48 2.40 3.01 3.13 2.97 1.01 0.94 0.94 D 104 nd 0.14 −1 −1 −1 −1 −1 −1 −1 0.73 1.87 Y 105 −1 −1 0.66 −1 0.94 nd 0.84 0.91 0.87 nd 0.09 NP Aliphatic Small NP T V I L A C G P M CDR1 T 28 nd 2.08 2.45 nd 1.56 nd 0.92 2.16 nd T 30 0.93 1.30 nd 1.26 0.89 nd −1.00 0.93 nd K 31 nd 1.41 1.17 0.60 0.39 nd 1.02 nd nd Y 32 nd 0.11 0.03 0.05 0.01 nd 0.02 0.01 nd G 33 0.01 0.04 0.55 2.27 0.06 nd nd 6.31 nd CDR2 W 50 0.03 0.02 0.07 0.04 0.03 nd 0.03 −1 nd N 52 −1 −1 −1 −1 −1 nd −1 −1 nd T 53 1.19 2.44 11.40 nd 9.03 nd −1 −1 nd Y 54 0.36 1.32 0.28 1.80 0.47 nd 3.72 0.86 nd T 55 nd 0.28 nd 0.95 nd nd 0.02 nd nd E 56 1.21 0.86 0.85 0.64 1.67 nd 1.37 0.01 nd E 57 1.44 nd 1.34 −1 1.65 nd 1.45 −1.00 nd P 58 1.00 1.01 nd 0.64 1.03 nd 1.31 1.06 nd T 59 0.99 1.03 0.76 0.95 nd nd 0.35 −1.00 nd CDR3 G 100 0.61 2.17 nd 0.65 1.99 nd nd −1 nd S 101 0.84 1.92 3.53 −1 1.15 nd 3.31 nd nd A 102 0.68 1.20 0.79 −1 nd nd 3.58 0.87 nd D 104 −1 −1 nd −1 −1 nd 0.41 −1 nd Y 105 0.12 0.18 0.23 0.22 0.08 nd −1.00 −1 nd 1A value of −1 indicates no binding 2Bolded values indicate the highest affinity (as measured by how many “fold” differences in affinity the mutant is in comparison to original, e.g., 2.0 as twice as strong and 0.5 as half as strong)obtained for an amino acid change at the position 3nd indicates that binding affinity was not determined

TABLE 6 Biacore Analysis of Modified C5A (XPA23) Light Chain Variable Regions1, 2, 3 NP Aromatic Neg Pos Polar D E R K H Y W F Q N S CDR1 Q27 1.07 1.08 0.89 −1.00 1.09 nd −1.00 2.96 1.06 −1.00 3.30 D28 1.00 0.74 0.94 0.82 1.23 1.45 3.81 1.43 nd 1.00 5.25 N30 0.81 0.64 0.74 0.61 1.00 1.40 1.06 1.59 0.60 1.08 0.72 R31 11.11 −1.00 1.06 1.18 −1.00 −1.00 0.43 0.92 0.27 0.38 0.44 W32 −1.00 −1.00 −1.00 −1.00 −1.00 −1.00 0.99 −1.00 −1.00 −1.00 −1.00 CDR2 H49 0.11 −1.00 0.21 0.10 1.10 0.52 0.21 0.50 −1.00 0.31 0.14 S50 −1.00 0.02 −1.00 0.10 0.05 0.05 0.02 −1.00 0.21 0.25 1.04 A51 0.13 0.29 0.18 nd 0.61 0.45 0.24 −1.00 0.24 0.30 0.95 T52 0.72 0.61 3.37 3.23 0.91 1.01 0.87 1.05 0.83 0.83 0.78 S53 −1.00 1.13 3.29 4.07 nd 1.23 1.11 1.24 1.09 1.23 1.08 CDR3 A91 1.08 0.10 −1.00 nd 1.12 −1.00 −1.00 −1.00 −1.00 0.10 0.90 D92 0.83 0.99 −1.00 0.23 0.63 −1.00 0.12 0.26 0.56 0.67 6.59 S93 4.59 3.71 0.91 0.86 1.08 1.45 5.49 1.32 1.54 3.76 1.20 F94 −1.00 −1.00 −1.00 −1.00 −1.00 −1.00 −1.00 1.14 −1.00 −1.00 −1.00 P95 1.32 0.88 −1.00 −1.00 −1.00 −1.00 −1.00 0.98 −1.00 −1.00 4.05 L96 −1.00 −1.00 −1.00 −1.00 0.07 −1.00 4.49 0.28 0.21 0.06 0.23 NP Aliphatic Small NP T V I L A C G P M CDR1 Q27 0.64 −1.00 0.82 0.83 1.47 nd −1.00 −1.00 nd D28 1.19 0.85 −1.00 9.64 0.88 nd 1.21 1.17 nd N30 0.97 0.89 0.65 0.68 0.61 nd 0.71 0.69 nd R31 0.24 0.44 0.68 0.51 8.48 nd 0.92 −1.00 nd W32 −1.00 −1.00 −1.00 −1.00 −1.00 nd −1.00 −1.00 nd CDR2 H49 −1.00 −1.00 −1.00 0.07 0.23 nd −1.00 0.86 nd S50 0.87 0.21 0.25 0.49 0.13 nd 0.14 −1.00 nd A51 0.76 0.31 −1.00 0.16 1.09 nd 3.84 0.13 nd T52 1.02 0.82 0.80 0.78 0.66 nd −1.00 1.08 nd S53 0.95 1.42 1.42 0.99 1.07 nd 1.03 4.92 nd CDR3 A91 0.67 0.51 −1.00 0.39 −1.00 nd −1.00 −1.00 nd D92 0.34 −1.00 0.33 0.14 0.40 nd 0.26 −1.00 nd S93 1.47 3.81 1.35 1.08 1.16 nd 0.75 −1.00 nd F94 −1.00 0.20 0.88 0.55 −1.00 nd −1.00 0.17 nd P95 −1.00 −1.00 −1.00 −1.00 3.83 nd −1.00 −1.00 nd L96 0.06 nd 0.67 1.16 0.26 nd −1.00 −1.00 nd 1A value of −1 indicates no binding 2Bolded values indicate the highest affinity (as measured by how many “fold” differences in affinity. The mutant is in comparison to original, e.g., 2.0 as twice as strong and 0.5 as half as strong) obtained for an amino acid change at the position 3nd indicates that binding affinity was not determined

TABLE 7 Biacore Analysis of Modified C5A (XPA23) Heavy Chain Variable Regions1, 2, 3 NP Aromatic Neg Pos Polar D E R K H Y W F Q N S CDR1 T28 nd −1.00 −1.00 −1.00 −1.00 nd −1.00 nd −1.00 nd nd S30 −1.00 −1.00 0.10 −1.00 −1.00 −1.00 0.77 −1.00 −1.00 0.15 0.85 K31 0.04 −1.00 0.90 1.30 1.11 0.04 0.84 0.05 −1.00 −1.00 −1.00 Y32 0.68 0.12 nd nd 0.62 nd nd 0.75 −1.00 0.04 1.33 F33 0.92 0.86 −1.00 0.85 −1.00 0.77 0.79 −1.00 0.78 0.97 0.73 F35 0.06 −1.00 −1.00 0.68 −1.00 0.85 −1.00 −1.00 −1.00 0.86 0.09 CDR2 V50 −1.00 −1.00 −1.00 −1.00 −1.00 −1.00 −1.00 −1.00 −1.00 0.03 −1.00 I51 0.07 0.73 0.10 0.11 0.08 1.75 −1.00 0.90 0.83 0.68 0.69 S52 −1.00 −1.00 −1.00 −1.00 −1.00 −1.00 −1.00 −1.00 −1.00 −1.00 nd P53 −1.00 0.03 nd 0.04 −1.00 0.03 0.05 0.05 0.03 0.06 0.03 S54 0.12 0.03 1.43 1.53 nd 0.99 1.07 0.87 1.01 0.08 nd G55 −1.00 0.11 0.95 0.11 nd 0.08 nd nd 0.90 0.10 1.02 G56 0.05 1.21 2.11 2.05 0.82 1.27 1.51 1.41 1.31 1.06 nd M57 −1.00 −1.00 0.06 0.04 0.08 0.86 nd 1.03 0.02 0.03 0.04 T58 0.12 −1.00 1.14 −1.00 0.98 1.01 0.93 1.04 −1.00 0.90 0.82 R59 −1.00 −1.00 −1.00 0.94 −1.00 −1.00 0.09 0.92 0.86 0.86 0.13 CDR3 V99 nd −1.00 nd −1.00 nd nd nd 0.03 −1.00 −1.00 −1.00 G100 −1.00 nd −1.00 −1.00 0.06 0.09 0.05 −1.00 −1.00 nd 2.80 Y101 −1.00 0.03 0.85 nd 0.04 1.00 0.81 0.85 0.06 0.06 nd G102 nd −1.00 −1.00 −1.00 −1.00 −1.00 −1.00 nd −1.00 1.17 nd G103 −1.00 −1.00 −1.00 −1.00 nd nd −1.00 −1.00 −1.00 nd nd N104 −1.00 −1.00 −1.00 −1.00 0.07 −1.00 −1.00 −1.00 0.03 1.02 1.02 S105 nd 1.17 −1.00 nd 0.05 −1.00 −1.00 1.25 0.98 0.85 1.09 D106 0.98 0.04 0.06 0.04 0.02 −1.00 −1.00 −1.00 0.02 0.07 0.03 Y107 0.85 0.90 0.85 0.82 −1.00 −1.00 0.89 0.90 0.87 −1.00 0.89 NP Aliphatic Small NP T V I L A C G P M CDR1 T28 −1.00 nd nd nd nd nd −1.00 nd nd S30 0.06 −1.00 0.08 0.91 0.77 nd 0.13 −1.00 nd K31 −1.00 0.81 0.74 1.08 nd nd nd −1.00 nd Y32 0.80 1.07 nd 0.75 −1.00 nd 1.00 0.68 nd F33 0.78 0.76 0.75 0.88 0.75 nd 0.76 0.91 nd F35 0.74 0.89 0.85 0.78 0.07 nd 0.04 0.06 nd CDR2 V50 0.03 0.19 0.10 0.09 0.09 nd 0.03 −1.00 nd I51 0.86 0.99 0.95 0.79 1.04 nd 0.94 −1.00 nd S52 0.04 0.04 −1.00 −1.00 0.05 nd 0.03 0.05 nd P53 nd 1.10 0.08 0.04 0.05 nd 0.02 0.94 nd S54 0.14 1.00 0.86 0.91 1.00 nd 1.43 0.06 nd G55 nd nd 0.86 0.08 nd nd 0.99 nd nd G56 1.11 nd 1.71 1.41 1.21 nd nd 0.85 nd M57 0.06 0.03 0.03 0.06 0.03 nd nd −1.00 nd T58 1.04 0.99 0.95 0.95 0.14 nd 0.95 1.81 nd R59 0.76 −1.00 −1.00 0.08 −1.00 nd 1.04 0.07 nd CDR3 V99 0.10 nd nd −1.00 0.03 nd nd −1.00 nd G100 0.12 nd nd −1.00 nd nd 1.02 −1.00 nd Y101 0.06 0.09 0.83 1.91 0.04 nd −1.00 0.04 nd G102 nd −1.00 nd −1.00 −1.00 nd nd −1.00 nd G103 nd 2.56 nd nd 0.03 nd 0.93 nd nd N104 0.95 1.27 nd 1.19 1.61 nd 1.21 nd nd S105 0.11 0.08 0.07 −1.00 0.84 nd 0.12 1.41 nd D106 −1.00 −1.00 −1.00 −1.00 −1.00 nd 0.03 −1.00 nd Y107 0.83 0.93 0.96 0.86 0.83 nd 0.10 0.04 nd 1A value of −1 indicates no binding 2Bolded values indicate the highest affinity o affinity (as measured by how many “fold” differences in affinity. The mutant is in comparison to original, e.g., 2.0 as twice as strong and 0.5 as half as strong) obtained for an amino acid change at the position 3nd indicates that binding affinity was not determined

1. Exemplary ING-1 Experiment

The dataset obtained from the exemplary ING1 experiment provides an empirical test for methods of targeted affinity enhancement. The data of Tables 4 and 5 are reorganized and displayed in Table 8. In Table 8 (broken horizontally into parts a and b), each row represents a mutated antibody position in either light chain (top half) or heavy chain (bottom half). Amino acid positions are numbered at the extreme left, first according to Kabat, and then (in the adjacent column to the right) according to the linear peptide sequence accompanying the mutational dataset. Each column in the main part of Table 8 represents a possible mutant amino acid. The columns are grouped according to the chemical functionality of the sidechain (e.g., negative, positive, polar, nonpolar, special), and sidechains within each group are arranged in order of size (from smallest on the left to largest on the right). Each cell in Table 8 corresponds to one of the 702 possible mutations (18 changes at 39 positions), and it contains a number indicating how many “fold” difference in affinity the mutant is, in comparison to the original (2.00 is twice as strong and 0.50 is half as strong). The number −1.00 means that no dissociation curve could be measured. The entry “nd” means “no data”, for cases in which a particular mutant was not obtained or tested in the exemplary experiment.

The sensitivity and reproducibility of the experiment can be estimated mathematically, simply by collecting together as internal controls (in column “ctrl” of Table 8B) all “mutants” which are identical to the original. For the 26 controls with measured affinities, the mean is 0.99 and the standard deviation is 0.10, with the single data point 1.38 as an outlier. Therefore, affinity measurements differing from 1.00 by two standard deviations (0.20) can be considered statistically significant. These are indicated in Table 8, which includes cells with an increase (>1.20 fold) of affinity, cells with a decrease (<0.80 fold) of affinity, cells with no binding (−1.00), cells with “no data” (nd), and cells with unchanged or not-significantly-different (0.80 to 1.20 fold) affinity.

Of the 635 mutations with results in this exemplary experiment (not including the 67 “nd” cells in Table 8), 122 (19%) showed no binding, 190 (30%) decreased affinity, 145 (10%) made no significant change, and 178 (28%) increased affinity. Thus, more than one-fourth of all possible mutations led to significant enhancements of affinity. These 178 beneficial mutations can be subdivided into 122 (19% of the total 635) showing an increase of 1.2 to 2.0 fold of affinity, 32 (5%) showing an increase of 2.0 to 3.0 fold, and 24 (4%) showing an increase of more than 3.0 fold. Thus, nearly one-tenth of the possible mutations are able to increase affinity by more than two fold.

Using the data from this exemplary experiment, the steps of the flowchart/mutation cycle as shown on FIGS. 5 and 6 can be reconstructed as described below except in cases when the “no data” (nd) condition aborts the cycle before it can be completed. However, the collection of beneficial mutations at the end of each individual step of the flowchart mutation cycle cannot be reconstructed. Each contacting (C) amino-acid position in the ING-1 antibody's combining site is analyzed below, with the “fold” of enhancement (from Table 8a-8b) indicated in parentheses next to the mutation. Since the mutation dataset was prepared with linear amino-acid numbering, each residue is identified by its sidechain, its Kabat position number, and its linear sequence number. Thus, for example, S 27a [28] in the light chain is the serine located at position 27a according to Kabat, but at position 28 in the linear peptide sequence.

TABLE 8a Affinities of ING-1 Mutations Negative Positive Polar D E K H R S T N Q Y Light K 27 27 1.26 −1.00 ? 1.06 1.26 −1.00 −1.00 −1.00 1.52 nd S  27a 28 1.63 1.02 2.32 1.90 2.78 1.08 1.19 −1.00 1.99 2.02 L  27b 29 −1.00 −1.00 nd nd −1.00 2.03 nd 1.85 −1.00 −1.00 L  27c 30 1.47 −1.00 −1.00 1.45 −1.00 1.43 1.89 1.45 1.60 1.53 H  27d 31 0.71 0.68 0.05 0.95 0.06 0.82 1.94 0.50 0.57 2.16 S  27e 32 0.94 nd 1.32 1.13 1.79 ? nd −1.00 1.10 1.37 N 28 33 0.49 0.65 0.70 0.73 0.71 0.48 0.60 1.38 0.73 0.80 I 30 35 0.19 0.16 0.61 0.59 0.92 0.55 0.87 0.50 0.41 0.51 T 31 36 0.05 1.60 1.15 0.79 −1.00 1.10 0.94 0.74 1.04 nd Y 32 37 nd 0.01 0.02 4.07 nd 0.09 −1.00 0.06 0.02 0.95 Y 49 54 0.03 0.05 3.62 −1.00 −1.00 0.90 0.61 −1.00 1.23 0.92 Q 50 55 0.05 0.05 0.46 3.82 5.31 0.56 0.66 0.36 0.95 nd M 51 56 1.36 0.71 0.98 1.32 0.92 1.05 0.80 0.99 1.12 1.21 S 52 57 0.95 0.93 1.54 1.01 1.17 1.00 0.86 1.17 −1.00 −1.00 N 53 58 nd 0.97 1.40 1.16 1.77 1.65 1.42 0.95 1.55 1.43 L 92 97 −1.00 0.75 0.61 0.42 −1.00 0.79 0.54 0.48 0.91 0.98 E 93 98 1.62 0.98 2.22 1.23 3.08 −1.00 4.90 −1.00 1.41 1.23 L 94 99 0.02 0.01 0.02 0.05 0.04 0.04 0.09 2.00 0.02 1.00 P 95 100 0.02 0.06 0.03 0.05 0.05 0.06 0.08 0.05 0.04 1.94 R  95a 101 −1.00 −1.00 0.04 −1.00 0.93 −1.00 −1.00 −1.00 −1.00 −1.00 Heavy T 28 28 0.98 1.23 1.77 1.15 1.23 1.07 nd 0.69 −1.00 1.86 T 30 30 0.63 0.73 0.94 nd 1.39 0.91 0.93 0.73 nd 2.00 K 31 31 0.66 0.54 0.96 0.78 0.76 0.49 nd −1.00 nd 1.00 Y 32 32 nd 0.08 0.08 0.60 0.43 0.01 nd nd 0.10 0.84 G 33 33 −1.00 −1.00 −1.00 0.02 0.03 nd 0.01 −1.00 0.06 6.16 W 50 50 3.27 −1.00 0.04 0.02 0.10 0.02 0.03 0.03 0.01 0.04 N 52 52 0.02 −1.00 −1.00 0.02 −1.00 −1.00 −1.00 0.98 −1.00 −1.00 T  52a 53 −1.00 −1.00 1.79 −1.00 −1.00 0.17 1.19 −1.00 −1.00 −1.00 Y  52b 54 0.05 0.07 3.62 1.00 3.72 0.49 0.36 2.11 0.66 0.92 T 54 55 0.03 −1.00 0.45 0.05 0.14 0.42 nd 0.17 0.03 0.03 E 55 56 0.81 0.95 1.27 1.74 1.34 1.46 1.21 1.01 1.23 1.04 E 56 57 1.17 1.07 nd 1.16 1.71 1.41 1.44 1.57 −1.00 1.37 P 57 58 0.54 0.44 1.14 nd nd 1.07 1.00 0.90 1.11 0.99 T 58 59 0.87 0.51 1.43 0.40 1.22 nd 0.99 0.96 −1.00 nd G 96 100 −1.00 −1.00 1.59 −1.00 7.51 1.68 0.61 1.55 −1.00 −1.00 S 97 101 0.21 0.76 2.20 1.35 nd nd 0.84 0.97 2.18 1.79 A 98 102 0.28 0.51 1.48 2.40 2.18 0.94 0.68 0.94 1.01 3.01 D 101  104 nd 0.14 −1.00 −1.00 −1.00 1.87 −1.00 0.73 −1.00 −1.00 Y 102  105 −1.00 −1.00 −1.00 0.94 0.66 0.09 0.12 nd 0.87 nd

TABLE 8b Affinities of ING-1 Mutations Nonpolar Sp G A V I L F W P ctrl Light K 27 27 1.11 1.10 nd 1.15 1.39 1.62 −1.00 1.29 ? S  27a 28 nd 1.53 2.15 2.35 2.60 2.38 nd 2.42 1.08 L  27b 29 1.53 1.97 nd −1.00 ? −1.00 −1.00 −1.00 ? L  27c 30 1.64 1.27 1.69 0.96 0.97 1.56 −1.00 0.82 0.97 H  27d 31 0.73 1.26 1.17 1.14 0.59 nd 1.66 1.19 0.95 S  27e 32 1.39 1.35 0.93 1.17 1.12 1.64 1.27 0.83 ? N 28 33 0.72 0.76 −1.00 nd nd 0.93 1.04 0.67 1.38 I 30 35 0.39 0.46 1.08 nd 0.60 0.66 0.34 0.70 nd T 31 36 1.09 1.02 0.98 0.76 1.64 nd 1.30 0.67 0.94 Y 32 37 1.30 0.04 −1.00 0.90 nd 0.63 0.85 nd 0.95 Y 49 54 1.85 0.08 1.32 −1.00 3.44 0.94 0.96 0.86 0.92 Q 50 55 0.70 0.64 1.53 1.42 0.71 0.86 4.11 0.95 0.95 M 51 56 0.80 0.86 −1.00 1.37 0.74 1.40 1.29 1.38 nd S 52 57 −1.00 1.15 0.89 0.98 1.38 0.96 2.34 nd 1.00 N 53 58 1.87 1.79 2.84 2.51 1.47 1.03 1.99 3.47 0.95 L 92 97 −1.00 0.50 1.44 2.62 0.93 0.93 1.59 0.95 0.93 E 93 98 1.63 1.35 nd nd 2.82 1.43 1.10 −1.00 0.98 L 94 99 0.02 1.43 1.04 2.07 0.93 0.43 0.89 0.01 0.93 P 95 100 1.62 0.05 0.14 0.14 0.03 1.65 1.51 1.01 1.01 R  95a 101 −1.00 −1.00 −1.00 −1.00 1.33 −1.00 −1.00 −1.00 0.93 Heavy T 28 28 0.92 1.56 2.08 2.45 nd 1.28 1.08 2.16 nd T 30 30 −1.00 0.89 1.30 nd 1.26 nd 1.26 0.93 0.93 K 31 31 1.02 0.39 1.41 1.17 0.60 nd 1.02 nd 0.96 Y 32 32 0.02 0.01 0.11 0.03 0.05 1.05 nd 0.01 0.84 G 33 33 nd 0.06 0.04 0.55 2.27 7.19 −1.00 6.31 nd W 50 50 0.03 0.03 0.02 0.07 0.04 0.09 0.97 −1.00 0.97 N 52 52 −1.00 −1.00 −1.00 −1.00 −1.00 −1.00 −1.00 −1.00 0.98 T  52a 53 −1.00 9.03 2.44 11.40 nd −1.00 −1.00 −1.00 1.19 Y  52b 54 3.72 0.47 1.32 0.28 1.80 0.65 0.96 0.86 0.92 T 54 55 0.02 nd 0.28 nd 0.95 0.03 0.10 nd nd E 55 56 1.37 1.67 0.86 0.85 0.64 0.78 1.17 0.01 0.95 E 56 57 1.45 1.65 nd 1.34 −1.00 1.06 1.39 −1.00 1.07 P 57 58 1.31 1.03 1.01 nd 0.64 0.98 1.11 1.06 1.06 T 58 59 0.35 nd 1.03 0.76 0.95 0.43 2.24 −1.00 0.99 G 96 100 nd 1.99 2.17 nd 0.65 −1.00 −1.00 −1.00 nd S 97 101 3.31 1.15 1.92 3.53 −1.00 1.16 1.22 nd nd A 98 102 3.58 nd 1.20 0.79 −1.00 2.97 3.13 0.87 nd D 101  104 0.41 −1.00 −1.00 nd −1.00 −1.00 −1.00 −1.00 nd Y 102  105 −1.00 0.08 0.18 0.23 0.22 0.91 0.84 −1.00 nd

A. Analysis of ING-1 Light Chain Mutations

K 27 [27]. Charged. Remove the charge, with N (−1.00). Since an affinity is not measured, return to K. Size. Try other positive charges, with H (1.06), R (1.26). Keep R (1.26), for a total of 3 mutations. In this exemplary experiment, the mutation cycle does not identify Q (1.52), L (1.39), F (1.62), P (1.29).

S 27a [28]. Uncharged. Introduce a charge, with D (1.63). Since affinity increases, keep D. Size. Try other negative charges, with E (1.02). Keep D (1.63), for a total of 2 mutations. In this exemplary experiment, the mutation cycle does not identify K (2.32), H (1.90), R (2.78), Q (1.99), Y (2.02), V (2.15), I (2.35), L (2.60), F (2.38), P (2.42).

L 27b [29]. Uncharged. Introduce a charge, with D (−1.00). Since an affinity is not measured, return to L. Try an opposite charge, with K (nd). No data. Since data for this mutation was not experimentally obtained, the mutation cycle cannot be further evaluated at this position. Keep L (original), for a total of 1 mutation so far. Possible improvements may be S (2.03), N (1.85), G (1.53), A (1.97).

L 27c [30]. Uncharged. Introduce a charge, with D (1.47). Since affinity is increased, keep D. Size. Try other negative charges, with E (−1.00). Keep D (1.47), for a total of 2 mutations. In this exemplary experiment, the mutation cycle does not identify T (1.89), Q (1.60), Y (1.53), G (1.64), V (1.69), F (1.56).

H 27d [31]. Charged. Remove the charge, with N (0.50). Since affinity is decreased, return to H. Size. Try other positive charges, with K (0.05), R (0.06). Keep H (original), for a total of 3 mutations. In this exemplary experiment, the mutation cycle does not identify T (1.94), Y (2.16), A (1.26), W (1.66).

S 27e [32]. Uncharged. Introduce a charge, with D (0.94). Since affinity is not significantly affected, return to S. Polarity. Switch polar S with smaller nonpolar A (1.35). Since affinity increases, keep A. Size. Test smaller nonpolar sidechains, with G (1.39). Since affinity increases, keep G. Test larger nonpolar sidechains, but none exist. Keep G (1.39), for a total of 3 mutations. In this exemplary experiment, the mutation cycle does not identify R (1.79), F (1.64).

N 28 [33]. Uncharged. Introduce a charge, with D (0.49). Since affinity decreases, return to N. Try an opposite charge, with K (0.70). Since affinity decreases again, return to N. Polarity. Switch polar N with nonpolar L (nd). No data. Since this data for mutation was not experimentally obtained, the mutation cycle cannot be further evaluated at this position. Keep N (original), for a total of 2 mutations so far. Possible improvements may be N (1.38).

I 30 [35]. Uncharged. Introduce a charge, with D (0.19). Since affinity decreases, return to I. Try an opposite charge, with K (0.61). Since affinity decreases again, return to I. Polarity. Switch nonpolar I with polar N (0.50). Since affinity decreases, return to I. Size. Try a smaller nonpolar sidechain, with V (1.08). Since affinity does not change significantly, return to I. Try a larger nonpolar sidechain, with F (0.66). Since affinity does not change significantly, return to I. Keep I (original), for a total of 5 mutations. In this exemplary experiment, the mutation cycle finds the optimal sidechain at this position.

T 31 [36]. Uncharged. Introduce a charge, with D (0.05). Since affinity decreases, return to T. Try an opposite charge, with K (1.15). Since affinity is not significantly changed, return to T. Polarity. Switch polar T with nonpolar V (0.98). Since affinity is not significantly changed, return to T. Size. Try a smaller polar sidechain, with S (1.10). Since affinity is not significantly changed, return to T. Try a larger polar sidechain, with N (0.74). Since affinity decreases, return to T. Keep T (original), for a total of 5 mutations. In this exemplary experiment, the mutation cycle does not identify E (1.60), L (1.64), W (1.30).

Y 32 [37]. Uncharged. Introduce a charge, with D (nd). No data. Since data for this mutation was not experimentally obtained, the mutation cycle cannot be further evaluated at this position. Keep Y (original), for a total of 0 mutations so far. Possible improvements may be H (4.07), G (1.30).

Y 49 [54]. Uncharged. Introduce a charge, with D (0.03). Since affinity decreases, return to Y. Try an opposite charge, with K (3.62). Since affinity increases, keep K. Size. Try other positive charges, with H (−1.00) and R (−1.00). Keep K (3.62), for a total of 4 mutations. In this exemplary experiment, the mutation cycle finds the optimal sidechain at this amino-acid position.

Q 50 [55]. Uncharged. Introduce a charge, with D (0.05). Since affinity decreases, return to Q. Try an opposite charge, with K (0.46). Since affinity decreases again, return to Q. Polarity. Switch polar Q with a [smaller] nonpolar sidechain, either L (0.71) or I (1.42). Since these produce opposite effects, each must be followed separately. On the one hand, the case of L (0.71). Since affinity decreases, return to Q. Size. Try smaller polar sidechains, with N (0.36). Since affinity decreases, return to Q, Try larger polar sidechains, with Y (nd). No data. Since data for this mutation was not experimentally obtained, the mutation cycle cannot be further evaluated at this position. Keep Q (original), for a total of 4 mutations so far. Possible improvements may be H (3.82), R (5.31), V (1.53), I (1.42), W (4.11). On the other hand, in the case of I (1.42). Since affinity increases, keep I. Size. Try a smaller nonpolar sidechain, with V (1.53). Since affinity increases, keep V. Try the next smaller sidechain A (0.64). Since affinity decreases, return to V. Keep V (1.53), for a total of 5 mutations. In this exemplary experiment, the mutation cycle does not identify H (3.82), R (5.31), W (4.11).

M 51 [56]. Uncharged. Introduce a charge, with D (1.36). Since affinity increases, keep D. Size. Try other negative charges, with E (0.71). Keep D (1.36), for a total of 2 mutations. In this exemplary experiment, the mutation cycle does not identify I (1.37), F (1.40), P (1.38).

S 52 [57]. Uncharged. Introduce a charge, with D (0.95). Since affinity is not significantly changed, return to S. Polarity. Switch polar S with [smaller] nonpolar A (1.15). Since affinity is not significantly changed, return to S. Size. Try a smaller polar sidechain, but none exist. Try a larger polar sidechain, with T (0.86). Since affinity decreases, return to S. Keep S (original), for a total of 3 mutations. In this exemplary experiment, the mutation cycle does not identify K (1.54), L (1.38), W (2.34).

N 53 [58]. Uncharged. Introduce a charge, with D (nd). No data. Since data for this mutation was not experimentally obtained, the mutation cycle cannot be further evaluated at this position. Keep N (original), for a total of 0 mutations so far. Possible improvements may be K (1.40), R (1.77), S (1.65), T (1.42), Q (1.55), Y (1.43), G (1.87), A (1.79), V (2.84), I (2.51), L (1.47), W (1.99), P (3.47).

L 92 [97]. Uncharged. Introduce a charge, with D (−1.00). Since affinity is abolished, return to L. Try an opposite charge, with K (0.61). Since affinity decreases, return to L. Polarity. Switch nonpolar L with polar N (0.48). Since affinity decreases, return to L. Size. Try a smaller nonpolar sidechain, with V (1.44). Since affinity increases, keep V. Try the next smaller nonpolar sidechain, with A (0.50). Since affinity decreases, return to V. Keep V (1.44), for a total of 5 mutations. In this exemplary experiment. In this exemplary experiment, the mutation cycle does not identify I (2.62), W (1.59).

E 93 [98]. Charged. Remove the charge, with N (−1.00). Since affinity is abolished, return to E. Size. Try other negative charges, with D (1.62). Keep D (1.62), for a total of 2 mutations. In this exemplary experiment, the mutation cycle does not identify K (2.22), R (3.08), T (4.90), G (1.63), L (2.82).

L 94 [99]. Uncharged. Introduce a charge, with D (0.02). Since affinity decreases, return to L. Try an opposite charge, with K (0.02). Since affinity decreases, return to L. Polarity. Switch nonpolar L with polar N (2.00). Since affinity increases, keep N. Size. Try a smaller polar sidechain, with T (0.09). Since affinity decreases, return to N. Try a larger polar sidechain, with Q (0.02). Since affinity decreases, return to N. Keep N (2.00), for a total of 5 mutations. In this exemplary experiment, the mutation cycle does not identify I (2.07).

P 95 [100]. Special. Test the special condition by replacing P with V (0.14). Since affinity decreases, return to P. Keep P (original), for a total of 1 mutation. In this exemplary experiment, the mutation cycle does not identify Y (1.94), G (1.62), F (1.65), W (1.51).

R 95a [101]. Charged. Remove the charge, with N (−1.00). Since affinity is abolished, return to R. Size. Try other positive charges, with K (0.04) and H (−1.00). Keep R (original), for a total of 3 mutations. In this exemplary experiment, the mutation cycle does not identify L (1.33).

B. Analysis of ING-1 Heavy Chain Mutations

T 28 [28]. Uncharged. Introduce a charge, with D (0.98). Since affinity is not significantly changed, return to T. Polarity. Switch polar T with nonpolar V (2.08). Since affinity increases, keep V. Size. Try a smaller nonpolar sidechain, with A (1.56). Since affinity decreases to 1.56 in comparison with 2.08, return to V. Try a larger nonpolar sidechain, with either L (nd) or I (2.45). Since these produce opposite results, each must be followed separately. On the one hand, the case of L (nd). No data. Since this mutation was not experimentally obtained, the mutation cycle cannot be further evaluated at this position. Keep V (2.08), for a total of 4 mutations so far. Possible improvements would be I (2.45), P (2.16). On the other hand, the case of I (2.45). Since affinity increases, keep I. Try the next larger nonpolar sidechain, with F (1.28). Since affinity decreases to 1.28 in comparison with 2.45, return to I. Keep I (2.45), for a total of 5 mutations. In this exemplary experiment, the mutation cycle finds the optimal sidechain at this position.

T 30 [30]. Uncharged. Introduce a charge, with D (0.63). Since affinity decreases, return to T. Try an opposite charge, with K (0.94). Since affinity decreases, return to T. Polarity. Switch polar T with nonpolar V (1.30). Since affinity increases, keep V. Size. Try a smaller nonpolar sidechain, with A (0.89). Since affinity decreases, return to V. Try a larger nonpolar sidechain, with either L (1.26) or I (nd). Since L is not better than V, return to V. Keep V (1.30), for a total of 5 mutations. In this exemplary experiment, the mutation cycle does not identify R (1.39), Y (2.00).

K 31 [31]. Charged. Remove the charge, with N (−1.00). Since affinity is abolished, return to K. Size. Try other positive charges, with H (0.78) and R (0.76). Keep K (original), for a total of 3 mutations. In this exemplary experiment, the mutation cycle does not identify V (1.41).

Y 32 [32]. Uncharged. Introduce a charge, with D (nd). No data. Since this mutation was not experimentally obtained, the mutation cycle cannot be further evaluated at this position. Keep Y (original), for a total of 0 mutations so far. In this exemplary experiment, since none of the possible mutations can significantly increase affinity, the mutation cycle would find that the original residue Y is the optimal sidechain at this position.

G 33 [33]. Uncharged. Introduce a charge, with D (−1.00). Since affinity is abolished, return to G. Try an opposite charge, with K (−1.00). Since affinity decreases, return to G. Polarity. Switch nonpolar G with [larger] polar S (nd). No data. Since data for this mutation was not experimentally obtained, the mutation cycle cannot be further evaluated at this position. Keep G (original), for a total of 2 mutations so far. Possible improvements may be Y (6.16), L (2.27), F (7.19), P (6.31).

W 50 [50]. Uncharged. Introduce a charge, with D (3.27). Since affinity increases, keep D. Size. Try other negative charges, with E (−1.00). Keep D (3.27), for a total of 2 mutations. In this exemplary experiment, the mutation cycle identifies the optimal mutation at this position.

N 52 [52]. Uncharged. Introduce a charge, with D (0.02). Since affinity decreases, return to N. Try an opposite charge, with K (−1.00). Since affinity decreases, return to N. Polarity. Switch polar N with nonpolar L (−1.00) or I (−1.00). Since affinity decreases in either case, return to N. Size. Try a smaller polar sidechain, with T (−1.00). Since affinity decreases, return to N. Try a larger polar sidechain, with Q (−1.00). Since affinity decreases, return to N. Keep N (original), for a total of 5 mutations. In this exemplary experiment, the mutation cycle finds the optimal sidechain at this position.

T 52a [53]. Uncharged. Introduce a charge, with D (−1.00). Since affinity is abolished, return to T. Try an opposite charge, with K (1.79). Since affinity increases, keep K. Size. Try other positive charges, with H (−1.00) and R (−1.00). Keep K (1.79), for a total of 4 mutations. In this exemplary experiment, the mutation cycle does not identify A (9.03), V (2.44), I (11.40).

Y 52b [54]. Uncharged. Introduce a charge, with D (0.05). Since affinity decreases, return to Y. Try an opposite charge, with K (3.62). Since affinity increases, keep K. Size. Try other positive charges, with H (1.00) and R (3.72). Keep R (3.72), for a total of 4 mutations. In this exemplary experiment, the mutation cycle finds the optimal sidechain at this position.

T 54 [55]. Uncharged. Introduce a charge, with D (0.03). Since affinity decreases, return to T. Try an opposite charge, with K (0.45). Since affinity decreases, return to T. Polarity. Switch polar T with nonpolar V (0.28). Since affinity decreases, return to T. Try a smaller polar sidechain, with S (0.42). Since affinity decreases, return to T. Try a larger polar sidechain, with N (0.17). Since affinity decreases, return to T. Keep T (original), for a total of 5 mutations. In this exemplary experiment, the mutation cycle finds the optimal sidechain at this amino-acid position.

E 55 [56]. Charged. Remove the charge, with N (1.01). Since affinity is not significantly changed, keep uncharged N. Polarity. Switch polar N with nonpolar L (0.64) or I (0.85). Since affinity decreases in either case, return to N. Try a smaller polar sidechain, with T (1.21). Since affinity increases, keep T. Try the next smaller polar sidechain, with S (1.46). Since affinity increases again, keep S. Try the next smaller polar sidechain, but none exist. Keep S (1.46), for a total of 4 mutations. In this exemplary experiment, the mutation cycle does not identify H (1.74), A (1.67).

E 56 [57]. Charged. Remove the charge, with N (1.57). Since affinity increases, keep N. Try an opposite charge, with K (nd). No data. Since this mutation was not experimentally obtained, the mutation cycle cannot be further evaluated at this position. Keep N (1.57), for a total of 1 mutation so far. Possible improvements may be R (1.71), A (1.65).

P 57 [58]. Special. Test the special condition by replacing P with V (1.01). Since affinity is not significantly changed, keep V. Uncharged. Introduce a charge, with D (0.54). Since affinity decreases, return to V. Try an opposite charge, with K (1.14). Since affinity is not significantly changed, return to V. Polarity. Switch nonpolar V with polar T (1.00). Since affinity is not significantly changed, keep polar T to improve solubility. Size. Try a smaller polar sidechain, with S (1.07). Since affinity is not significantly changed, return to T. Try a larger polar sidechain, with N (0.90). Since affinity is not significantly changed, return to T. Keep T (1.00), for a total of 6 mutations. In this exemplary experiment, the mutation cycle does not identify G (1.31).

T 58 [59]. Uncharged. Introduce a charge, with D (0.87). Since affinity is not significantly changed, return to T. Polarity. Switch polar T with nonpolar V (1.03). Since affinity is not significantly changed, return to T. Size. Try a smaller polar sidechain, with S (nd). No data. Since this mutation was not experimentally obtained, the mutation cycle cannot be further evaluated at this position. Keep T (original), for a total of 2 mutations so far. Possible improvements may be K (1.43), R (1.22), W (2.24).

G 96 [100]. Uncharged. Introduce a charge, with D (−1.00). Since affinity is abolished, return to G. Try an opposite charge, with K (1.59). Since affinity increases, keep K. Size. Try other positive charge, with H (−1.00) and R (7.51). Keep R (7.51), for a total of 4 mutations. In this exemplary experiment, the mutation cycle finds the optimal sidechain at this amino-acid position.

S 97 [101]. Uncharged. Introduce a charge, with D (0.21). Since affinity decreases, return to S. Try an opposite charge, with K (2.20). Since affinity increases, keep K. Size. Try other positive charges, with H (1.35) and R (nd). Keep K (2.20), for a total of 3 mutations. In this exemplary experiment, the mutation cycle does not identify G (3.31), I (3.53).

A 98 [102]. Uncharged. Introduce a charge, with D (0.28). Since affinity decreases, return to A. Try an opposite charge, with K (1.48). Since affinity increases, keep K. Try other positive charges, with H (2.40) and R (2.18). Keep H (2.40), for a total of 4 mutations. In this exemplary experiment, the mutation cycle does not identify Y (3.01), G (3.58), F (2.97), W (3.13).

D 101 [104]. Charged. Remove the charge, with N (0.73). Since affinity decreases, return to D. Size. Try other negative sidechains, with E (0.14). Keep D (original), for a total of 2 mutations. In this exemplary experiment, the mutation cycle does not identify S (1.87).

Y 102 [105]. Uncharged. Introduce a charge, with D (−1.00). Since affinity is abolished, return to Y. Try an opposite charge, with K (−1.00). Since affinity is abolished, return to Y. Polarity. Switch polar Y with [smaller] nonpolar F (0.91). Since affinity is not significantly changed, return to Y. Size. Test a smaller polar sidechain, with Q (0.87). Since affinity is not significantly changed, return to Y. Test a larger polar sidechain, but none exist. Keep Y (original), for a total of 4 mutations. In this exemplary experiment, the mutation cycle finds the optimal sidechain at this amino-acid position.

2. Exemplary C5A (XPA23) Experiment

The dataset obtained from the exemplary C5A (XPA23) experiment provides an empirical test for the methods of targeted affinity enhancement. The data of Tables 6 and 7 are reorganized and displayed in Table 9. In Table 9 (broken horizontally into parts a and b), each row represents a mutated antibody position in either the light chain variable region (Q27-L96) or heavy chain variable region. Amino acid positions are numbered at the extreme left first according to Kabat. Each column in the main part of the table represents a possible mutant amino acid. The columns are grouped according to the chemical functionality of the sidechain (negative, positive, polar, nonpolar, special), and sidechains within each group are arranged in order of size (from smallest on the left to largest on the right). Each cell in the table corresponds to one of the 779 possible mutations (19 changes at 41 positions), and it contains a number indicating how many “fold” difference in affinity the mutant is (using K-off measurements), in comparison to the original (2.00 is twice as strong and 0.50 is half as strong). The entry “nd” means “no data”, for cases in which a particular mutant was not obtained in the experiment. These same “fold” values are presented in another display format in Tables 6 and 7.

The sensitivity and reproducibility of the experiment can be estimated mathematically, simply by collecting together as internal controls all “mutants” which are identical to the original. For the 27 controls with measured affinities, the mean is 1.03 and the standard deviation is 0.10, with the single data point 0.19 eliminated as an extreme outlier. Therefore, affinity measurements differing from 1.00 by two standard deviations (0.20) can be considered statistically significant. These are indicated in Table 9, which includes cells with an increase (>1.20 fold) of affinity, cells with a decrease (<0.80 fold) of affinity, cells with no binding (−1.00) affinity, cells with “no data”, and cells with unchanged or not-significantly-different (0.80 to 1.20 fold) affinity.

Using the data from this exemplary experiment, the steps of the flowchart/mutation cycle as shown on FIGS. 5 and 6 can be reconstructed as described below, except in cases when the “no data” (nd) condition aborts the cycle before it can be completed. However, the collection of beneficial mutations at the end of each individual step cannot be reconstructed. Each constructing (C) amino-acid position in the C5A (XPA23) antibody combining site is analyzed below, with the “fold” of enhancement (from Table 9a-9b) indicated in parentheses next to the mutation. Since the mutation dataset was prepared with linear amino-acid numbering, each residue is identified by its sidechain, its Kabat number position, and its linear sequence number. Thus, for example, L95a [96] in the light chain variable region is the leucine located at position 95a according to Kabat, but at position 96 in the linear peptide sequence.

TABLE 9a Affinities of CA5 (XPA23) Mutations D E K H R S T N Q Y Light Q 27 27 1.07 1.08 −1.00 1.09 0.89 3.30 0.64 −1.00 1.06 nd D 28 28 1.00 0.74 0.82 1.23 0.94 5.25 1.19 1.00 nd 1.45 N 30 30 0.81 0.64 0.61 1.00 0.74 0.72 0.97 1.08 0.60 1.40 R 31 31 11.11 −1.00 1.18 −1.00 1.06 0.44 0.24 0.38 0.27 −1.00 W 32 32 −1.00 −1.00 −1.00 −1.00 −1.00 −1.00 −1.00 −1.00 −1.00 −1.00 H 49 49 0.11 −1.00 0.10 1.10 0.21 0.14 −1.00 0.31 −1.00 0.52 S 50 50 −1.00 0.02 0.10 0.05 −1.00 1.04 0.87 0.25 0.21 0.05 A 51 51 0.13 0.29 nd 0.61 0.18 0.95 0.76 0.30 0.24 0.45 T 52 52 0.72 0.61 3.23 0.91 3.37 0.78 1.02 0.83 0.83 1.01 S 53 53 −1.00 1.13 4.07 nd 3.29 1.08 0.95 1.23 1.09 1.23 A 91 91 1.08 0.10 nd 1.12 −1.00 0.90 0.67 0.10 −1.00 −1.00 D 92 92 0.83 0.99 0.23 0.63 −1.00 6.59 0.34 0.67 0.56 −1.00 S 93 93 4.59 3.71 0.86 1.08 0.91 1.20 1.47 3.76 1.54 1.45 F 94 94 −1.00 −1.00 −1.00 −1.00 −1.00 −1.00 −1.00 −1.00 −1.00 −1.00 P 95 95 1.32 0.88 −1.00 −1.00 −1.00 4.05 −1.00 −1.00 −1.00 −1.00 L  95a 98 −1.00 −1.00 −1.00 0.07 −1.00 0.23 0.06 0.06 0.21 −1.00 Heavy T 28 28 nd −1.00 −1.00 −1.00 −1.00 nd −1.00 nd −1.00 nd S 30 30 −1.00 −1.00 −1.00 −1.00 0.10 0.85 0.06 0.15 −1.00 −1.00 K 31 31 0.04 −1.00 1.30 1.11 0.90 −1.00 −1.00 −1.00 −1.00 0.04 Y 32 32 0.68 0.12 nd 0.62 nd 1.33 0.80 0.04 −1.00 nd F 33 33 0.92 0.86 0.85 −1.00 −1.00 0.73 0.78 0.97 0.78 0.77 F 35 35 0.06 −1.00 0.68 −1.00 −1.00 0.09 0.74 0.86 −1.00 0.85 V 50 50 −1.00 −1.00 −1.00 −1.00 −1.00 −1.00 0.03 0.03 −1.00 −1.00 I 51 51 0.07 0.73 0.11 0.08 0.10 0.69 0.86 0.68 0.83 1.75 S 52 52 −1.00 −1.00 −1.00 −1.00 −1.00 nd 0.04 −1.00 −1.00 −1.00 P  52a 53 −1.00 0.03 0.04 −1.00 nd 0.03 nd 0.06 0.03 0.03 S 53 54 0.12 0.03 1.53 nd 1.43 nd 0.14 0.08 1.01 0.99 G 54 55 −1.00 0.11 0.11 nd 0.95 1.02 nd 0.10 0.90 0.08 G 55 56 0.05 1.21 2.05 0.82 2.11 nd 1.11 1.06 1.31 1.27 M 56 57 −1.00 −1.00 0.04 0.08 0.06 0.04 0.06 0.03 0.02 0.86 T 57 58 0.12 −1.00 −1.00 0.98 1.14 0.82 1.04 0.90 −1.00 1.01 R 58 59 −1.00 −1.00 0.94 −1.00 −1.00 0.13 0.76 0.86 0.86 −1.00 V 95 99 nd −1.00 −1.00 nd nd −1.00 0.10 −1.00 −1.00 nd G 96 100 −1.00 nd −1.00 0.06 −1.00 2.80 0.12 nd −1.00 0.09 Y 97 101 −1.00 0.03 nd 0.04 0.85 nd 0.06 0.06 0.06 1.00 G 98 102 nd −1.00 −1.00 −1.00 −1.00 nd nd 1.17 −1.00 −1.00 G 99 103 −1.00 −1.00 −1.00 nd −1.00 nd nd nd −1.00 nd N 100  104 −1.00 −1.00 −1.00 0.07 −1.00 1.02 0.95 1.02 0.03 −1.00 S 100a 105 nd 1.17 nd 0.05 −1.00 1.09 0.11 0.85 0.98 −1.00 D 101  106 0.98 0.04 0.04 0.02 0.06 0.03 −1.00 0.07 0.02 −1.00 Y 102  107 0.85 0.90 0.82 −1.00 0.85 0.89 0.83 −1.00 0.87 −1.00

TABLE 9b Affinities of CA5 (XPA23) Mutations G A V I L F W P M Light Q 27 27 −1.00 1.47 −1.00 0.82 0.83 2.96 −1.00 −1.00 nd D 28 28 1.21 0.88 0.85 −1.00 9.64 1.43 3.81 1.17 nd N 30 30 0.71 0.61 0.89 0.65 0.68 1.59 1.06 0.69 nd R 31 31 0.92 8.48 0.44 0.68 0.51 0.92 0.43 −1.00 nd W 32 32 −1.00 −1.00 −1.00 −1.00 −1.00 −1.00 0.99 −1.00 nd H 49 49 −1.00 0.23 −1.00 −1.00 0.07 0.50 0.21 0.86 nd S 50 50 0.14 0.13 0.21 0.25 0.49 −1.00 0.02 −1.00 nd A 51 51 3.84 1.09 0.31 −1.00 0.16 −1.00 0.24 0.13 nd T 52 52 −1.00 0.66 0.82 0.80 0.78 1.05 0.87 1.08 nd S 53 53 1.03 1.07 1.42 1.42 0.99 1.24 1.11 4.92 nd A 91 91 −1.00 −1.00 0.51 −1.00 0.39 −1.00 −1.00 −1.00 nd D 92 92 0.26 0.40 −1.00 0.33 0.14 0.26 0.12 −1.00 nd S 93 93 0.75 1.16 3.81 1.35 1.08 1.32 5.49 −1.00 nd F 94 94 −1.00 −1.00 0.20 0.88 0.55 1.14 −1.00 0.17 nd P 95 95 −1.00 3.83 −1.00 −1.00 −1.00 0.98 −1.00 −1.00 nd L  95a 98 −1.00 0.26 nd 0.67 1.16 0.28 4.49 −1.00 nd Heavy T 28 28 −1.00 nd nd nd nd nd −1.00 nd nd S 30 30 0.13 0.77 −1.00 0.08 0.91 −1.00 0.77 −1.00 nd K 31 31 nd nd 0.81 0.74 1.08 0.05 0.84 −1.00 nd Y 32 32 1.00 −1.00 1.07 nd 0.75 0.75 nd 0.68 nd F 33 33 0.76 0.75 0.76 0.75 0.88 −1.00 0.79 0.91 nd F 35 35 0.04 0.07 0.89 0.85 0.78 −1.00 −1.00 0.06 nd V 50 50 0.03 0.09 0.19 0.10 0.09 −1.00 −1.00 −1.00 nd I 51 51 0.94 1.04 0.99 0.95 0.79 0.90 −1.00 −1.00 nd S 52 52 0.03 0.05 0.04 −1.00 −1.00 −1.00 −1.00 0.05 nd P  52a 53 0.02 0.05 1.10 0.08 0.04 0.05 0.05 0.94 nd S 53 54 1.43 1.00 1.00 0.86 0.91 0.87 1.07 0.06 nd G 54 55 0.99 nd nd 0.86 0.08 nd nd nd nd G 55 56 nd 1.21 nd 1.71 1.41 1.41 1.51 0.85 nd M 56 57 nd 0.03 0.03 0.03 0.06 1.03 nd −1.00 nd T 57 58 0.95 0.14 0.99 0.95 0.95 1.04 0.93 1.81 nd R 58 59 1.04 −1.00 −1.00 −1.00 0.08 0.92 0.09 0.07 nd V 95 99 nd 0.03 nd nd −1.00 0.03 nd −1.00 nd G 96 100 1.02 nd nd nd −1.00 −1.00 0.05 −1.00 nd Y 97 101 −1.00 0.04 0.09 0.83 1.91 0.85 0.81 0.04 nd G 98 102 nd −1.00 −1.00 nd −1.00 nd −1.00 −1.00 nd G 99 103 0.93 0.03 2.56 nd nd −1.00 −1.00 nd nd N 100  104 1.21 1.61 1.27 nd 1.19 −1.00 −1.00 nd nd S 100a 105 0.12 0.84 0.08 0.07 −1.00 1.25 −1.00 1.41 nd D 101  106 0.03 −1.00 −1.00 −1.00 −1.00 −1.00 −1.00 −1.00 nd Y 102  107 0.10 0.83 0.93 0.96 0.86 0.90 0.89 0.04 nd

A. Analysis of C5A (XPA23) Light Chain Mutations

Q 27 [27]. Uncharged. Introduce a charge, with D (1.07). Since affinity is not significantly changed, return to Q. Polarity. Switch polar Q with a [smaller] nonpolar sidechain, either L (0.83) or I (0.82). Since affinity is still not significantly changed in either case, keep polar Q. Test for an empty solvent-filled space by introducing tyrosine Y (nd). No data. Since data for this mutation was not experimentally obtained, the mutation cycle cannot be further evaluated at this position. Keep Q (original), for a total of 2 mutations so far. Possible improvements may be S (3.30), A (1.47), F (2.96).

D 28 [28]. Charged. Remove the charge, with N (1.00). Since affinity is not significantly changed, keep N. Polarity. Switch polar N with nonpolar L (9.64) or I (−1.00). Since these produce opposite effects, each is followed separately. On the one hand, the case of L (9.64). Since affinity increases, keep L. Size. Try a smaller nonpolar sidechain, with V (0.85). Since affinity decreases relative to L, return to L. Try a larger nonpolar sidechain, with F (1.43). Since affinity decreases again relative to L, return to L. Keep L (9.64), for a total of 4 mutations. The mutation cycle finds the optimal sidechain at this amino-acid position. On the other hand, the case of I (−1.00). Since affinity is abolished, return to N. Size. Try a smaller polar sidechain, with T (1.19). Since affinity is not significantly changed, keep N. Try a larger polar sidechain, with Q (nd). No data. Since data for this mutation was not experimentally obtained, the mutation cycle cannot be further evaluated at this position. Keep N (1.00), for a total of 3 mutations so far. Possible improvements may be H (1.23), S (5.25), Y (1.45), G (1.21), L (9.64), F (1.43), W (3.81).

N 30 [30]. Uncharged. Introduce a charge, with D (0.81). Since affinity is not significantly changed, keep N. Polarity. Switch polar N with nonpolar L (0.68) or I (0.65). Since affinity decreases in either case, return to N. Size. Try a smaller polar sidechain, with T (0.97). Since affinity is not significantly changed, return to N. Try a larger polar sidechain, with Q (0.60). Since affinity decreases, return to N. Keep N (original), for a total of 4 mutations. In this exemplary experiment, the mutation cycle does not identify Y (1.40), F (1.59).

R 31 [31]. Charged. Remove the charge, with N (0.38). Since affinity is decreased, return to R. Size. Try other positive charges, with K (1.18), H (−1.00). Keep R (original), for a total of 3 mutations. In this exemplary experiment, the mutation cycle does not identify D (11.11), A (8.48).

W 32 [32]. Uncharged. Introduce a charge, with D (−1.00). Since affinity is abolished, return to W. Try an opposite charge, with K (−1.00). Since affinity is abolished again, return to W. Polarity. Switch nonpolar W with [smaller] polar Y (−1.00). Since affinity is abolished again, return to W. Size. Try a smaller nonpolar sidechain, with F (−1.00). Since affinity is abolished again, return to W. Try a larger nonpolar sidechain, but none exist. Keep W (original), for a total of 4 mutations. In this exemplary experiment, the mutation cycle finds the optimal sidechain at this amino-acid position.

H 49 [49]. Charged. Remove the charge, with N (0.31). Since affinity is decreased, return to H. Size. Try other positive charges, with K (0.10), R (0.21). Keep H (original), for a total of 3 mutations. In this exemplary experiment, the mutation cycle finds the optimal sidechain at this amino-acid position.

S 50 [50]. Uncharged. Introduce a charge, with D (−1.00). Since affinity is abolished, return to S. Try an opposite charge, with K (0.10). Since affinity decreases, return to S. Polarity. Switch polar S with [smaller] nonpolar A (0.13). Since affinity decreases, return to S. Size. Try a smaller polar sidechain, but none exist. Try a larger polar sidechain, with T (0.87). Since affinity is not significantly changed, return to S. Keep S (original), for a total of 4 mutations. In this exemplary experiment, the mutation cycle finds the optimal sidechain at this amino-acid position.

A 51 [51]. Uncharged. Introduce a charge, with D (0.13). Since affinity decreases, return to A. Try an opposite charge, with K (nd). No data. Since data for this mutation was not experimentally obtained, the mutation cycle cannot be further evaluated at this position. Keep A (original), for a total of 1 mutation so far. Possible improvements may be G (3.84).

T 52 [52]. Uncharged. Introduce a charge, with D (0.72). Since affinity decreases, return to T. Try an opposite charge, with K (3.23). Since affinity increases, keep K. Size. Try other positive charges, with H (0.91) and R (3.37). Keep R (3.37), for a total of 4 mutations. In this exemplary experiment, the mutation cycle finds the optimal sidechain at this amino-acid position.

S 53 [53]. Uncharged. Introduce a charge, with D (−1.00). Since affinity is abolished, return to S. Try an opposite charge, with K (4.07). Since affinity increases, keep K. Try other positive charges, with H (nd) and R (3.29). Keep K (4.07), for a total of 4 mutations. In this exemplary experiment, the mutation cycle does not identify P (4.92).

A 91 [91]. Uncharged. Introduce a charge, with D (1.08). Since affinity is not significantly changed, keep A. Polarity. Switch nonpolar A with [larger] polar S (0.90). Since affinity is still not significantly changed, keep polar S. Test for an empty solvent-filled space by introducing tyrosine Y (−1.00). Since affinity is abolished, return to S. Size. Try a smaller polar sidechain, but none exist. Try a larger polar sidechain, with T (0.67). Since affinity decreases, return to S. Keep S (0.90), for a total of 4 mutations.

D 92 [92]. Charged. Remove the charge, with N (0.67). Since affinity decreases, return to D. Size. Try other negative charges, with E (0.99). Keep D (original), for a total of 2 mutations. In this exemplary experiment, the mutation cycle does not identify S (6.59).

S 93 [93]. Uncharged. Introduce a charge, with D (4.59). Since affinity increases, keep D. Size. Try other negative charges, with E (3.71). Keep D (4.59), for a total of 2 mutations. In this exemplary experiment, the mutation cycle does not identify W (5.49).

F 94 [94]. Uncharged. Introduce a charge, with D (−1.00). Since affinity is abolished, return to F. Try an opposite charge, with K (−1.00). Since affinity is abolished again, return to F. Polarity. Switch nonpolar F with [smaller] polar Q (−1.00). Since affinity is abolished again, return to F. Size. Try a smaller nonpolar sidechain, with L (0.55) or I (0.88). Since affinity decreases or is not significantly changed, return to F. Try a larger nonpolar sidechain, with W (−1.00). Since affinity is abolished, return to F. Keep F (original), for a total of 5 mutations. In this exemplary experiment, the mutation cycle finds the optimal sidechain at this amino-acid position.

P 95 [95]. Special. Test the special condition by replacing P with V (−1.00). Since affinity is abolished, return to P. Keep P (original), for a total of 1 mutation. In this exemplary experiment, the mutation cycle does not identify D (1.32), S (4.05), A (3.83).

L 96 [96]. Uncharged. Introduce a charge, with D (−1.00). Since affinity is abolished, return to L. Try an opposite charge, with K (−1.00). Since affinity is abolished again, return to L. Polarity. Switch nonpolar L with polar N (0.06). Since affinity decreases, return to L. Size. Try a smaller nonpolar sidechain, with V (nd). No data. Since this mutation was not experimentally obtained, the mutation cycle cannot be further evaluated at this position. Keep L (original), for a total of 3 mutations so far. Possible improvements would be W (4.49).

B. Analysis of XPA23 (C5A) Heavy Chain Mutations

T 28 [28]. Uncharged. Introduce a charge, with D (nd). No data. Since this mutation was not experimentally obtained, the mutation cycle cannot be further evaluated at this position. Keep T (original), for a total of 0 mutations so far.

S 30 [30]. Uncharged. Introduce a charge, with D (−1.00). Since affinity is abolished, return to S. Try an opposite charge, with K (−1.00). Since affinity is abolished again, return to S. Polarity. Switch polar S with [smaller] nonpolar A (0.77). Since affinity decreases, return to S. Size. Try a smaller polar sidechain, but none exist. Try a larger polar sidechain, with T (0.06). Since affinity decreases, return to S. Keep S (original), for a total of 4 mutations. In this exemplary experiment, the mutation cycle finds the optimal sidechain at this amino-acid position.

K 31 [31]. Charged. Remove the charge, with N (−1.00). Since affinity is abolished, return to K. Try other positive charges, with H (1.11) and H (0.90). Since affinity is not significantly affected, return to K. Keep K (original), for a total of 3 mutations. In this exemplary experiment, the mutation cycle finds the optimal sidechain at this amino-acid position.

Y 32 [32]. Uncharged. Introduce a charge, with D (0.68). Since affinity decreases, return to Y. Try an opposite charge, with K (nd). No data. Since this mutation was not experimentally obtained, the mutation cycle cannot be further evaluated at this position. Keep Y (original), for a total of 1 mutation so far. Possible improvements may be S (1.33).

F 33 [33]. Uncharged. Introduce a charge, with D (0.92). Since affinity is not significantly changed, return to F. Polarity. Switch nonpolar F with [smaller] polar Q (0.78). Since affinity decreases, return to F. Size. Try a smaller nonpolar sidechain, with L (0.88) or I (0.75). Since affinity is not significantly affected or decreases, try a larger nonpolar sidechain, with W (0.79). Since affinity decreases, return to F. Keep F (original), for a total of 4 mutations. In this exemplary experiment, the mutation cycle finds the optimal sidechain at this amino-acid position.

F 35 [35]. Uncharged. Introduce a charge, with D (0.06). Since affinity decreases, return to F. Try an opposite charge, with K (0.68). Since affinity decreases again, return to F. Polarity. Switch nonpolar F with [smaller] polar Q (−1.00). Since affinity is abolished, return to F. Try a smaller nonpolar sidechain, with L (0.78) or I (0.85). Since affinity decreases or is not significantly changed, return to F. Try a larger nonpolar sidechain, with W (−1.00). Since affinity is abolished, return to F. Keep F (original), for a total of 5 mutations. In this exemplary experiment, the mutation cycle finds the optimal sidechain at this amino-acid position.

V 50 [50]. Uncharged. Introduce a charge, with D (−1.00). Since affinity is abolished, return to V. Try an opposite charge, with K (−1.00). Since affinity is abolished again, return to V. Polarity. Switch nonpolar V with polar T (0.03). Since affinity decreases, return to V. Size. Try a smaller nonpolar sidechain, with A (0.09). Since affinity decreases, return to V. Try a larger nonpolar sidechain, with L (0.09) or I (0.10). Since affinity decreases in either case, return to V. Keep V (original), for a total of 5 mutations. In this exemplary experiment, the mutation cycle finds the optimal sidechain at this amino-acid position.

I 51 [51]. Uncharged. Introduce a charge, with D (0.07). Since affinity decreases, return to I. Try an opposite charge, with K (0.11). Since affinity decreases again, return to I. Polarity. Switch nonpolar I with polar N (0.68). Since affinity decreases, return to I. Size. Try a smaller nonpolar sidechain, with V (0.99). Since affinity does not change significantly, return to I. Try a larger nonpolar sidechain, with F (0.90). Since affinity is not significantly changed, return to I. Keep I (original), for a total of 4 mutations so far. In this exemplary experiment, the mutation cycle does not identify Y (1.75).

S 52 [52]. Uncharged. Introduce a charge, with D (−1.00). Since affinity is abolished, return to S. Try an opposite charge, with K (−1.00). Since affinity is abolished again, return to S. Polarity. Switch polar S with [smaller] nonpolar A (0.05). Since affinity decreases, return to S. Size. Try a smaller polar sidechain, but none exist. Try a larger polar sidechain, with T (0.04). Since affinity decreases, return to S. Keep S (original), for a total of 4 mutations. In this exemplary experiment, the mutation cycle finds the optimal sidechain at this amino-acid position.

P 52a [53]. Special. Test the special condition by replacing P with V (1.10). Since affinity is not significantly changed, keep V. Uncharged. Introduce a charge, with D (−1.00). Since affinity is abolished, return to V. Try an opposite charge, with K (0.04). Since affinity decreases, return to V. Polarity. Switch nonpolar V with polar T (nd). No data. Since this mutation was not experimentally obtained, the mutation cycle cannot be further evaluated at this position. Keep V (1.10), for a total of 3 mutations so far.

S 53 [54]. Uncharged. Introduce a charge, with D (0.12). Since affinity decreases, return to S. Try an opposite charge, with K (1.53). Since affinity increases, keep K. Try other positive charges, with H (nd) and R (1.43). Keep K (1.53), for a total of 2 mutations. In this exemplary experiment, the mutation cycle finds the optimal sidechain at this amino-acid position.

G 54 [55]. Uncharged. Introduce a charge, with D (−1.00). Since affinity is abolished, return to G. Try an opposite charge, with K (0.11). Since affinity decreases, return to G. Polarity. Switch nonpolar G with [larger] polar S (1.02). Since affinity is not significantly changed, keep polar S. Test for an empty solvent-filled space by introducing tyrosine Y (0.08). Since affinity decreases, return to S. Size. Try a smaller polar sidechain, but none exist. Try a larger polar sidechain, with T (nd). No data. Since this mutation was not experimentally obtained, the mutation cycle cannot be further evaluated at this position. Keep S (1.02), for a total of 4 mutations so far.

G 55 [56]. Uncharged. Introduce a charge, with D (0.05). Since affinity decreases, return to G. Try an opposite charge, with K (2.05). Since affinity increases, keep K. Try other positive charges, with H (0.82) and R (2.11). Keep R (2.11). In this exemplary experiment, the mutation cycle finds the optimal sidechain at this amino-acid position.

M 56 [57]. Uncharged. Introduce a charge, with D (−1.00). Since affinity is abolished, return to M. Try an opposite charge, with K (0.04). Since affinity decreases, return to M. Polarity. Switch nonpolar M with polar N (0.03). Since affinity decreases, return to M. Size. Try a smaller nonpolar sidechain, with V (0.03). Since affinity decreases, return to M. Try a larger nonpolar sidechain, with F (1.03). Since affinity is not significantly changed, return to M. Keep M (original), for a total of 5 mutations. In this exemplary experiment, the mutation cycle finds the optimal sidechain at this amino-acid position.

T 57 [58]. Uncharged. Introduce a charge, with D (0.12). Since affinity decreases, return to T. Try an opposite charge, with K (−1.00). Since affinity is abolished, return to T. Polarity. Switch polar T with nonpolar V (0.99). Since affinity is not significantly changed, return to T. Size. Try a smaller polar sidechain, with S (0.82). Since affinity is not significantly changed, return to T. Try a larger polar sidechain, with N (0.90). Since affinity is not significantly changed, return to T. Keep T (original), for a total of 5 mutations. In this exemplary experiment, the mutation cycle does not identify P (1.81).

R 58 [59]. Charged. Remove the charge, with N (0.86). Since affinity is not significantly changed, keep uncharged N. Polarity. Switch polar N with nonpolar L (0.08) or I (−1.00). Since affinity decreases or is abolished, return to N. Try a smaller polar sidechain, with T (0.76). Since affinity decreases, return to N. Try the next larger polar sidechain, with Q (0.86). Since affinity is not significantly changed, return to N. Keep N (0.86), for a total of 4 mutations. In this exemplary experiment, the mutation cycle finds the optimal sidechain at this amino-acid position.

V 95 [99]. Uncharged. Introduce a charge, with D (nd). No data. Since this mutation was not experimentally obtained, the mutation cycle cannot be further evaluated at this position. Keep V (original), for a total of 0 mutations so far.

G 96 [100]. Uncharged. Introduce a charge, with D (−1.00). Since affinity is abolished, return to G. Try an opposite charge, with K (−1.00). Since affinity is abolished again, return to G. Polarity. Switch nonpolar G with [larger] polar S (2.80). Since affinity increases, keep S. Size. Try a smaller polar sidechain, but none exist. Try a larger polar sidechain, with T (0.12). Since affinity decreases, return to S. Keep S (2.80), for a total of 4 mutations. In this exemplary experiment, the mutation cycle finds the optimal sidechain at this amino-acid position.

Y 97 [101]. Uncharged. Introduce a charge, with D (−1.00). Since affinity is abolished, return to Y. Try an opposite charge, with K (nd). No data. Since this mutation was not experimentally obtained, the mutation cycle cannot be further evaluated at this position. Keep Y (original), for a total of 0 mutations so far. Possible improvements may be L (1.91).

G 98 [102]. Uncharged. Introduce a charge, with D (nd). No data. Since this mutation was not experimentally obtained, the mutation cycle cannot be further evaluated at this position. Keep G (original), for a total of 0 mutations so far.

G 99 [103]. Uncharged. Introduce a charge, with D (−1.00). Since affinity is abolished, return to G. Try an opposite charge, with K (−1.00). Since affinity is abolished again, return to G. Polarity. Switch nonpolar G with [larger] polar S (nd). No data. Since this mutation was not experimentally obtained, the mutation cycle cannot be further evaluated at this position. Keep G (original), for a total of 2 mutations so far. Possible improvements may be V (2.56).

N 100 [104]. Uncharged. Introduce a charge, with D (−1.00). Since affinity is abolished, return to N. Try an opposite charge, with K (−1.00). Since affinity is abolished again, return to N. Polarity. Switch polar N with nonpolar L (1.19) or I (nd). Since affinity is not significantly changed, return to N. Size. Try a smaller polar sidechain, with T (0.95). Since affinity is not significantly changed, return to N. Try a larger polar sidechain, with Q (0.03). Since affinity decreases, return to N. Keep N (original), for a total of 5 mutations. In this exemplary experiment, the mutation cycle does not identify G (1.21), A (1.61), V (1.27).

S 100a [105]. Uncharged. Introduce a charge, with D (nd). No data. Since this mutation was not experimentally obtained, the mutation cycle cannot be further evaluated at this position. Keep S (original), for a total of 0 mutations so far. Possible improvements may be F (1.25).

D 101 [106]. Charged. Remove the charge, with N (0.07). Since affinity decreases, return to D. Size. Try other negative sidechains, with E (0.04). Keep D (original), for a total of 2 mutations. In this exemplary experiment, the mutation cycle finds the optimal sidechain at this amino-acid position.

Y 102 [107]. Uncharged. Introduce a charge, with D (0.85). Since affinity is not significantly changed, return to Y. Polarity. Switch polar Y with [smaller] nonpolar F (0.90). Since affinity is not significantly changed, return to Y. Size. Try a smaller polar sidechain, with Q (0.87). Since affinity is not significantly changed, return to Y. Try a larger polar sidechain, but none exist, so return to Y. Keep Y (0.90), for a total of 3 mutations. In this exemplary experiment, the mutation cycle finds the optimal sidechain at this amino-acid position.

EMBODIMENTS

1. A method for enhancing the binding affinity of an antibody variable domain to a binding partner, to obtain a modified antibody variable domain with enhanced affinity to the binding partner compared to an unmodified antibody variable domain, the method comprising:

    • (a) identifying the proximity assigned to amino acid positions in the antibody variable domain using the “prox” line as shown in FIG. 3A or 3B;
    • (b) changing amino acid residues at contacting (C) proximity positions in the domain to another amino acid, thereby generating a modified variable domain;
    • (c) testing the binding affinity of the modified antibody variable domain for the binding partner; and
    • (d) obtaining a modified antibody variable domain with enhanced binding affinity for the binding partner compared to the binding affinity exhibited by the unmodified antibody variable domain for the binding partner.

2. A method for enhancing the binding affinity of an antibody variable domain to a binding partner, to obtain a modified antibody variable domain with enhanced affinity to the binding partner compared to an unmodified antibody variable domain, the method comprising:

    • (a) identifying the proximity assigned to amino acid positions in the antibody variable domain using the “prox” line as shown in FIG. 3A or 3B;
    • (b) changing amino acid residues at peripheral (P) proximity positions in the domain to another amino acid, thereby generating a modified variable domain;
    • (c) testing the binding affinity of the modified antibody variable domain for the binding partner; and
    • (d) obtaining a modified antibody variable domain with enhanced binding affinity for the binding partner compared to the binding affinity exhibited by the unmodified antibody variable domain for the binding partner.

3. A method for enhancing the binding affinity of an antibody variable domain to a binding partner, to obtain a modified antibody variable domain with enhanced affinity to the binding partner compared to an unmodified antibody variable domain, the method comprising:

    • (a) identifying the proximity assigned to amino acid positions in the antibody variable domain using the “prox” line as shown in FIG. 3A or 3B;
    • (b) changing amino acid residues at supporting (S) proximity positions in the domain to another amino acid, thereby generating a modified variable domain;
    • (c) testing the binding affinity of the modified antibody variable domain for the binding partner; and
    • (d) obtaining a modified antibody variable domain with enhanced binding affinity for the binding partner compared to the binding affinity exhibited by the unmodified antibody variable domain for the binding partner.

4. A method for enhancing the binding affinity of an antibody variable domain to a binding partner, to obtain a modified antibody variable domain with enhanced affinity to the binding partner compared to an unmodified antibody variable domain, the method comprising:

    • (a) identifying the proximity assigned to amino acid positions in the antibody variable domain using the “prox” line as shown in FIG. 3A or 3B;
    • (b) changing amino acid residues at interfacial (I) proximity positions in the domain to another amino acid, thereby generating a modified variable domain;
    • (c) testing the binding affinity of the modified antibody variable domain for the binding partner; and
    • (d) obtaining a modified antibody variable domain with enhanced binding affinity for the binding partner compared to the binding affinity exhibited by the unmodified antibody variable domain for the binding partner.

5. The method of embodiment 1, wherein each contacting (C) residue is changed.

6. The method of embodiment 2, wherein each peripheral (P) residue is changed.

7. The method of embodiment 3, wherein each supporting (S) residue is changed.

8. The method of embodiment 4, wherein each interfacial (I) residue is changed.

9. The method of embodiment 1, wherein the contacting (C) residue is in CDR1 in a heavy chain variable domain.

10. The method of embodiment 9, wherein the contacting (C) residue is at position 32 or 33 in CDR1.

11. The method of embodiment 1, wherein the contacting (C) residue is in CDR2 in a heavy chain variable domain.

12. The method of embodiment 11, wherein the contacting (C) residue is at position 50, 52, 53, 54, 56, or 58 in CDR2.

13. The method of any one of embodiments 1 to 4, wherein the amino acid residue is selected from the group consisting of: alanine, arginine, asparagine, aspartic acid, glutamine, glutamic acid, glycine, histidine, isoleucine, leucine, lysine, methionine, phenylalanine, serine, threonine, tryptophan, tyrosine and valine.

14. The method of any one of embodiments 1 to 4, wherein the variable domain is from a chimeric antibody.

15. The method of any one of embodiments 1 to 4, wherein the variable domain is from a humanized antibody.

16. The method of any one of embodiments 1 to 4, wherein the variable domain is from a human antibody.

17. The method of any one of embodiments 1 to 4, wherein binding affinity is determined by measuring Koff.

18. The method of any one of embodiments 1 to 4, wherein step (d) comprises:

    • (a) contacting a parent variable domain with the binding partner under conditions that permit binding;
    • (b) contacting the modified variable domains with the binding partner under conditions that permit binding; and
    • (c) determining binding affinity of the modified variable domains and the parent variable domain for the binding partner, wherein modified variable domains that have a binding affinity for the binding partner greater than the binding affinity of the parent variable domain for the binding partner are identified as having enhanced binding affinity for the binding partner.

19. A method for enhancing the binding affinity of an antibody variable domain to a binding partner compared to an unmodified antibody variable domain, the method comprising:

    • (a) identifying the proximity assigned to amino acid positions in the antibody variable domain using the “prox” line as shown in FIG. 3A or 3B;
    • (b) changing amino acid residues at contacting (C) proximity positions in the domain to another amino acid, wherein charged residues are changed to uncharged or uncharged residues are changed to charged or both charged residues are changed to uncharged and uncharged residues are changed to charged, thereby generating a modified antibody variable domain;
    • (c) testing the binding affinity of the modified antibody variable domain for the binding partner; and
    • (d) obtaining a modified antibody variable domain with enhanced binding affinity to the binding partner compared to the binding affinity exhibited by the unmodified antibody variable domain for the binding partner.

20. A method for enhancing the binding affinity of an antibody variable domain to a binding partner compared to an unmodified antibody variable domain, the method comprising:

    • (a) identifying the proximity assigned to amino acid positions in the antibody variable domain using the “prox” line as shown in FIG. 3A or 3B;
    • (b) changing amino acid residues at peripheral (P) proximity positions in the domain to another amino acid, wherein charged residues are changed to uncharged or uncharged residues are changed to charged or both charged residues are changed to uncharged and uncharged residues are changed to charged, thereby generating a modified antibody variable domain;
    • (c) testing the binding affinity of the modified antibody variable domain for the binding partner; and
    • (d) obtaining a modified antibody variable domain with enhanced binding affinity to the binding partner compared to the binding affinity exhibited by the unmodified antibody variable domain for the binding partner.

21. A method for enhancing the binding affinity of an antibody variable domain to a binding partner compared to an unmodified antibody variable domain, the method comprising:

    • (a) identifying the proximity assigned to amino acid positions in the antibody variable domain using the “prox” line as shown in FIG. 3A or 3B;
    • (b) changing amino acid residues at supporting (S) proximity positions in the domain to another amino acid, wherein charged residues are changed to uncharged or uncharged residues are changed to charged or both charged residues are changed to uncharged and uncharged residues are changed to charged, thereby generating a modified antibody variable domain;
    • (c) testing the binding affinity of the modified antibody variable domain for the binding partner; and
    • (d) obtaining a modified antibody variable domain with enhanced binding affinity to the binding partner compared to the binding affinity exhibited by the unmodified antibody variable domain for the binding partner.

22. A method for enhancing the binding affinity of an antibody variable domain to a binding partner compared to an unmodified antibody variable domain, the method comprising:

    • (a) identifying the proximity assigned to amino acid positions in the antibody variable domain using the “prox” line as shown in FIG. 3A or 3B;
    • (b) changing amino acid residues at interfacial (I) proximity positions in the domain to another amino acid, wherein charged residues are changed to uncharged or uncharged residues are changed to charged or both charged residues are changed to uncharged and uncharged residues are changed to charged, thereby generating a modified antibody variable domain;
    • (c) testing the binding affinity of the modified antibody variable domain for the binding partner; and
    • (d) obtaining a modified antibody variable domain with enhanced binding affinity to the binding partner compared to the binding affinity exhibited by the unmodified antibody variable domain for the binding partner.

23. The method of embodiment 19, wherein each contacting (C) residue is changed.

24. The method of embodiment 20, wherein each peripheral (P) residue is changed.

25. The method of embodiment 21, wherein each interfacial (I) residue is changed.

26. The method of embodiment 19, wherein the contacting (C) residue is in CDR1 in a heavy chain variable domain.

27. The method of embodiment 26, wherein the contacting (C) residue is at position 32 or 33 in CDR1.

28. The method of embodiment 19, wherein the contacting (C) residue is in CDR2 in a heavy chain variable domain.

29. The method of embodiment 28, wherein the contacting (C) residue is at position 50, 52, 53, 54, 56, or 58 in CDR2.

30. The method of any one of embodiments 19 to 22, wherein the amino acid residue is selected from the group consisting of: alanine, arginine, asparagine, aspartic acid, glutamine, glutamic acid, glycine, histidine, isoleucine, leucine, lysine, methionine, phenylalanine, serine, threonine, tryptophan, tyrosine and valine.

31. The method of any one of embodiments 19 to 22, wherein the variable domain is from a chimeric antibody.

32. The method of any one of embodiments 19 to 22, wherein the variable domain is from a humanized antibody.

33. The method of any one of embodiments 19 to 22, wherein the variable domain is from a human antibody.

34. The method of any one of embodiments 19 to 22, wherein binding affinity is determined by measuring Koff.

35. The method of any one of embodiments 19 to 22, wherein step (d) comprises:

    • (a) contacting a parent variable domain with the binding partner under conditions that permit binding;
    • (b) contacting the modified variable domains with the binding partner under conditions that permit binding; and
    • (c) determining binding affinity of the modified variable domains and the parent variable domain for the binding partner, wherein modified variable domains that have a binding affinity for the binding partner greater than the binding affinity of the parent variable domain for the binding partner are identified as having enhanced binding affinity for the binding partner.

36. A method for enhancing the binding affinity of an antibody variable domain to a binding partner compared to an unmodified antibody variable domain, the method comprising:

    • (a) identifying the proximity assigned to amino acid positions in the antibody variable domain using the “prox” line as shown in FIG. 3A or 3B;
    • (b) changing amino acid residues at contacting (C) proximity positions in the domain to another amino acid, wherein residues are changed to a different size within the same amino acid class, that is, positively charged, negatively charged, polar or nonpolar, thereby generating a modified antibody variable domain;
    • (c) testing the binding affinity of the modified antibody variable domain for the binding partner; and
    • (d) obtaining a modified antibody variable domain with enhanced binding affinity for the binding partner compared to the binding affinity exhibited by the unmodified antibody variable domain for the binding partner.

37. A method for enhancing the binding affinity of an antibody variable domain to a binding partner compared to an unmodified antibody variable domain, the method comprising:

    • (a) identifying the proximity assigned to amino acid positions in the antibody variable domain using the “prox” line as shown in FIG. 3A or 3B;
    • (b) changing amino acid residues at peripheral (P) proximity positions in the domain to another amino acid, wherein residues are changed to a different size within the same amino acid class, that is, positively charged, negatively charged, polar or nonpolar, thereby generating a modified antibody variable domain;
    • (c) testing the binding affinity of the modified antibody variable domain for the binding partner; and
    • (d) obtaining a modified antibody variable domain with enhanced binding affinity for the binding partner compared to the binding affinity exhibited by the unmodified antibody variable domain for the binding partner.

38. A method for enhancing the binding affinity of an antibody variable domain to a binding partner compared to an unmodified antibody variable domain, the method comprising:

    • (a) identifying the proximity assigned to amino acid positions in the antibody variable domain using the “prox” line as shown in FIG. 3A or 3B;
    • (b) changing amino acid residues at supporting (S) proximity positions in the domain to another amino acid, wherein residues are changed to a different size within the same amino acid class, that is, positively charged, negatively charged, polar or nonpolar, thereby generating a modified antibody variable domain;
    • (c) testing the binding affinity of the modified antibody variable domain for the binding partner; and
    • (d) obtaining a modified antibody variable domain with enhanced binding affinity for the binding partner compared to the binding affinity exhibited by the unmodified antibody variable domain for the binding partner.

39. A method for enhancing the binding affinity of an antibody variable domain to a binding partner compared to an unmodified antibody variable domain, the method comprising:

    • (a) identifying the proximity assigned to amino acid positions in the antibody variable domain using the “prox” line as shown in FIG. 3A or 3B;
    • (b) changing amino acid residues at interfacial (I) proximity positions in the domain to another amino acid, wherein residues are changed to a different size within the same amino acid class, that is, positively charged, negatively charged, polar or nonpolar, thereby generating a modified antibody variable domain;
    • (c) testing the binding affinity of the modified antibody variable domain for the binding partner; and
    • (d) obtaining a modified antibody variable domain with enhanced binding affinity for the binding partner compared to the binding affinity exhibited by the unmodified antibody variable domain for the binding partner.

40. The method of embodiment 36, wherein each contacting (C) residue is changed.

41. The method of embodiment 37, wherein each peripheral (P) residue is changed.

42. The method of embodiment 38, wherein each supporting (S) residue is changed.

43. The method of embodiment 39, wherein each interfacial (I) residue is changed.

44. The method of embodiment 36, wherein the contacting (C) residue is in CDR1 in a heavy chain variable domain.

45. The method of embodiment 44, wherein the contacting (C) residue is at position 32 or 33 in CDR1.

46. The method of embodiment 36, wherein the contacting (C) residue is in CDR2 in a heavy chain variable domain.

47. The method of embodiment 46, wherein the contacting (C) residue is at position 50, 52, 53, 54, 56, or 58 in CDR2.

48. The method of any one of embodiments 36 to 39, wherein the amino acid residue is selected from the group consisting of: alanine, arginine, asparagine, aspartic acid, glutamine, glutamic acid, glycine, histidine, isoleucine, leucine, lysine, methionine, phenylalanine, serine, threonine, tryptophan, tyrosine and valine.

49. The method of any one of embodiments 36 to 39, wherein the variable domain is from a chimeric antibody.

50. The method of any one of embodiments 36 to 39, wherein the variable domain is from a humanized antibody.

51. The method of any one of embodiments 36 to 39, wherein the variable domain is from a human antibody.

52. The method of any one of embodiments 36 to 39, wherein binding affinity is determined by measuring Koff.

53. The method of any one of embodiments 36 to 39, wherein step (d) comprises:

    • (a) contacting a parent variable domain with the binding partner under conditions that permit binding;
    • (b) contacting the modified variable domains with the binding partner under conditions that permit binding; and
    • (c) determining binding affinity of the modified variable domains and the parent variable domain for the binding partner, wherein modified variable domains that have a binding affinity for the binding partner greater than the binding affinity of the parent variable domain for the binding partner are identified as having enhanced binding affinity for the binding partner.

54. A method for enhancing the binding affinity of an antibody variable domain to a binding partner compared to an unmodified antibody variable domain, the method comprising:

    • (a) identifying the proximity assigned to amino acid positions in the antibody variable domain using the “prox” line as shown in FIG. 3A or 3B;
    • (b) changing amino acid residues at contacting (C) proximity positions in the domain to another amino acid, wherein charged residues are changed to opposite charge residues, that is, from positively charged to negatively charged or from negatively charged to positively charged, thereby generating a modified antibody variable domain;
    • (c) testing the binding affinity of the modified antibody variable domain for the binding partner; and
    • (d) obtaining a modified antibody variable domain with enhanced binding affinity to the binding partner compared to the binding affinity exhibited by the unmodified antibody variable domain for the binding partner.

55. A method for enhancing the binding affinity of an antibody variable domain to a binding partner compared to an unmodified antibody variable domain, the method comprising:

    • (a) identifying the proximity assigned to amino acid positions in the antibody variable domain using the “prox” line as shown in FIG. 3A or 3B;
    • (b) changing amino acid residues at peripheral (P) proximity positions in the domain to another amino acid, wherein charged residues are changed to opposite charge residues, that is, from positively charged to negatively charged or from negatively charged to positively charged, thereby generating a modified antibody variable domain;
    • (c) testing the binding affinity of the modified antibody variable domain for the binding partner; and
    • (d) obtaining a modified antibody variable domain with enhanced binding affinity to the binding partner compared to the binding affinity exhibited by the unmodified antibody variable domain for the binding partner.

56. A method for enhancing the binding affinity of an antibody variable domain to a binding partner compared to an unmodified antibody variable domain, the method comprising:

    • (a) identifying the proximity assigned to amino acid positions in the antibody variable domain using the “prox” line as shown in FIG. 3A or 3B;
    • (b) changing amino acid residues at supporting (S) proximity positions in the domain to another amino acid, wherein charged residues are changed to opposite charge residues, that is, from positively charged to negatively charged or from negatively charged to positively charged, thereby generating a modified antibody variable domain;
    • (c) testing the binding affinity of the modified antibody variable domain for the binding partner; and
    • (d) obtaining a modified antibody variable domain with enhanced binding affinity to the binding partner compared to the binding affinity exhibited by the unmodified antibody variable domain for the binding partner.

57. A method for enhancing the binding affinity of an antibody variable domain to a binding partner compared to an unmodified antibody variable domain, the method comprising:

    • (a) identifying the proximity assigned to amino acid positions in the antibody variable domain using the “prox” line as shown in FIG. 3A or 3B;
    • (b) changing amino acid residues at interfacial (I) proximity positions in the domain to another amino acid, wherein charged residues are changed to opposite charge residues, that is, from positively charged to negatively charged or from negatively charged to positively charged, thereby generating a modified antibody variable domain;
    • (c) testing the binding affinity of the modified antibody variable domain for the binding partner; and
    • (d) obtaining a modified antibody variable domain with enhanced binding affinity to the binding partner compared to the binding affinity exhibited by the unmodified antibody variable domain for the binding partner.

58. The method of embodiment 54, wherein the contacting (C) residue is in complementarity determining domain-1 (CDR1) in a light chain variable domain.

59. The method of embodiment 58, wherein the contacting (C) residue is at position 28, 30 or 31 in CDR1.

60. The method of embodiment 54, wherein the contacting (C) residue is in CDR2 in a light chain variable domain.

61. The method of embodiment 60, wherein the contacting (C) residue is at position 50, 51 or 53 in CDR2.

62. The method of embodiment 54, wherein the contacting (C) residue is in CDR1 in a heavy chain variable domain.

63. The method of embodiment 62, wherein the contacting (C) residue is at position 32 or 33 in CDR1.

64. The method of embodiment 54, wherein the contacting (C) residue is in CDR2 in a heavy chain variable domain.

65. The method of embodiment 64, wherein the contacting (C) residue is at position 50, 52, 53, 54, 56, or 58 in CDR2.

66. The method of any one of embodiments 54 to 57, wherein the amino acid residue is selected from the group consisting of: alanine, arginine, asparagine, aspartic acid, glutamine, glutamic acid, glycine, histidine, isoleucine, leucine, lysine, methionine, phenylalanine, serine, threonine, tryptophan, tyrosine and valine.

67. The method of any one of embodiments 54 to 57, wherein the variable domain is from a chimeric antibody.

68. The method of any one of embodiments 54 to 57, wherein the variable domain is from a humanized antibody.

69. The method of any one of embodiments 54 to 57, wherein the variable domain is from a human antibody.

70. The method of any one of embodiments 54 to 57, wherein binding affinity is determined by measuring Koff.

71. The method of any one of embodiments 54 to 57, wherein step (d) comprises:

    • (a) contacting a parent variable domain with the binding partner under conditions that permit binding;
    • (b) contacting the modified variable domains with the binding partner under conditions that permit binding; and
    • (c) determining binding affinity of the modified variable domains and the parent variable domain for the binding partner, wherein modified variable domains that have a binding affinity for the binding partner greater than the binding affinity of the parent variable domain for the binding partner are identified as having enhanced binding affinity for the binding partner.

72. A method for enhancing the binding affinity of an antibody variable domain to a binding partner compared to an unmodified antibody variable domain, the method comprising:

    • (a) identifying the proximity assigned to amino acid positions in the antibody variable domain using the “prox” line as shown in FIG. 3A or 3B;
    • (b) changing amino acid residues at contacting (C) proximity positions in the domain to another amino acid, wherein polar residues are changed to nonpolar residues or nonpolar residues are changed to polar residues or both polar residues are changed to nonpolar residues and nonpolar residues are changed to polar residues, thereby generating a modified antibody variable domain;
    • (c) testing the binding affinity of the modified antibody variable domain for the binding partner; and
    • (d) obtaining a modified antibody variable domain with enhanced binding affinity to the binding partner compared to the binding affinity exhibited by the unmodified antibody variable domain for the binding partner.

73. A method for enhancing the binding affinity of an antibody variable domain to a binding partner compared to an unmodified antibody variable domain, the method comprising:

    • (a) identifying the proximity assigned to amino acid positions in the antibody variable domain using the “prox” line as shown in FIG. 3A or 3B;
    • (b) changing amino acid residues at peripheral (P) proximity positions in the domain to another amino acid, wherein polar residues are changed to nonpolar residues or nonpolar residues are changed to polar residues or both polar residues are changed to nonpolar residues and nonpolar residues are changed to polar residues, thereby generating a modified antibody variable domain;
    • (c) testing the binding affinity of the modified antibody variable domain for the binding partner; and
    • (d) obtaining a modified antibody variable domain with enhanced binding affinity to the binding partner compared to the binding affinity exhibited by the unmodified antibody variable domain for the binding partner.

74. A method for enhancing the binding affinity of an antibody variable domain to a binding partner compared to an unmodified antibody variable domain, the method comprising:

    • (a) identifying the proximity assigned to amino acid positions in the antibody variable domain using the “prox” line as shown in FIG. 3A or 3B;
    • (b) changing amino acid residues at supporting (S) proximity positions in the domain to another amino acid, wherein polar residues are changed to nonpolar residues or nonpolar residues are changed to polar residues or both polar residues are changed to nonpolar residues and nonpolar residues are changed to polar residues, thereby generating a modified antibody variable domain;
    • (c) testing the binding affinity of the modified antibody variable domain for the binding partner; and
    • (d) obtaining a modified antibody variable domain with enhanced binding affinity to the binding partner compared to the binding affinity exhibited by the unmodified antibody variable domain for the binding partner.

75. A method for enhancing the binding affinity of an antibody variable domain to a binding partner compared to an unmodified antibody variable domain, the method comprising:

    • (a) identifying the proximity assigned to amino acid positions in the antibody variable domain using the “prox” line as shown in FIG. 3A or 3B;
    • (b) changing amino acid residues at interfacial (I) proximity positions in the domain to another amino acid, wherein polar residues are changed to nonpolar residues or nonpolar residues are changed to polar residues or both polar residues are changed to nonpolar residues and nonpolar residues are changed to polar residues, thereby generating a modified antibody variable domain;
    • (c) testing the binding affinity of the modified antibody variable domain for the binding partner; and
    • (d) obtaining a modified antibody variable domain with enhanced binding affinity to the binding partner compared to the binding affinity exhibited by the unmodified antibody variable domain for the binding partner.

76. The method of embodiment 72, wherein the contacting (C) residue is in complementarity determining domain-1 (CDR1) in a light chain variable domain.

77. The method of embodiment 76, wherein the contacting (C) residue is at position 28, 30 or 31 in CDR1.

78. The method of embodiment 72, wherein the contacting (C) residue is in CDR2 in a light chain variable domain.

79. The method of embodiment 78, wherein the contacting (C) residue is at position 50, 51 or 53 in CDR2.

80. The method of embodiment 72, wherein the contacting (C) residue is in CDR1 in a heavy chain variable domain.

81. The method of embodiment 80, wherein the contacting (C) residue is at position 32 or 33 in CDR1.

82. The method of embodiment 72, wherein the contacting (C) residue is in CDR2 in a heavy chain variable domain.

83. The method of embodiment 82, wherein the contacting (C) residue is at position 50, 52, 53, 54, 56, or 58 in CDR2.

84. The method of any one of embodiments 72 to 75, wherein the amino acid residue is selected from the group consisting of: alanine, arginine, asparagine, aspartic acid, glutamine, glutamic acid, glycine, histidine, isoleucine, leucine, lysine, methionine, phenylalanine, serine, threonine, tryptophan, tyrosine and valine.

85. The method of any one of embodiments 72 to 75, wherein the variable domain is from a chimeric antibody.

86. The method of any one of embodiments 72 to 75, wherein the variable domain is from a humanized antibody.

87. The method of any one of embodiments 72 to 75, wherein the variable domain is from a human antibody.

88. The method of any one of embodiments 72 to 75, wherein binding affinity is determined by measuring Koff.

89. The method of any one of embodiments 72 to 75, wherein step (d) comprises:

    • (a) contacting a parent variable domain with the binding partner under conditions that permit binding;
    • (b) contacting the modified variable domains with the binding partner under conditions that permit binding; and
    • (c) determining binding affinity of the modified variable domains and the parent variable domain for the binding partner, wherein modified variable domains that have a binding affinity for the binding partner greater than the binding affinity of the parent variable domain for the binding partner are identified as having enhanced binding affinity for the binding partner.

90. A method for enhancing the binding affinity of an antibody variable domain to a binding partner, said method comprising:

    • (a) firstly, changing amino acid residues in the domain to other amino acid residues, wherein charged residues are changed to uncharged residues thereby removing a charge and uncharged residues are changed to charged residues thereby adding a charge;
    • (b) testing the effect on affinity of the binding partner by removing or adding a charged amino acid residue at each position individually changed in step (a);
    • (c) obtaining a first modified antibody variable domain comprising the changed amino acid residues from step (a) that resulted in an increase in affinity as individually tested in step (b);
    • (d) secondly, changing amino acid residues that were firstly changed in step (a) from charged to uncharged or from uncharged to charged to amino acid residues of the opposite charge, if removing the charged residue in step (a) had caused a gain in affinity when tested in step (b) or if adding the charged residue in step (a) caused a loss in affinity when tested in step (b);
    • (e) testing the effect on affinity of the binding partner by changing the amino acid residue to the opposite charged residue at each position individually changed in step (d);
    • (f) obtaining a second modified antibody variable domain comprising the changed residues from step (d) that resulted in an increase in affinity as individually tested in step (e);
    • (g) thirdly, changing charged amino acid residues in the second modified antibody variable domain of step (f) to an amino acid residue of a different size and the same charge;
    • (h) testing the effect on affinity of the binding partner by changing the size of the charged residue at each position individually changed in step (g);
    • (i) obtaining a third modified antibody variable domain comprising the changed amino acid residues from step (g) that resulted in an increase in affinity as individually tested in step (h);
    • (j) fourthly, changing uncharged amino acid residues in the third modified antibody variable domain of step (i) to other amino acid residues, wherein polar uncharged residues are changed to nonpolar uncharged residues and wherein nonpolar uncharged residues are changed to polar uncharged residues;
    • (k) testing the effect on affinity of the binding partner by changing polar to nonpolar and nonpolar to polar at each position individually changed in step (j);
    • (l) obtaining a fourth modified antibody variable domain comprising the changed amino acid residues from step (j) that resulted in an increase in affinity as individually tested in step (k);
    • (m) fifthly, changing uncharged residues that were fourthly changed in step (j) from polar to nonpolar or from nonpolar to polar to tyrosine, if changing from polar to nonpolar or from nonpolar to polar in step (j) had not caused a gain or loss in affinity when tested in step (k);
    • (n) testing the effect on affinity of the binding partner by changing to tyrosine at each position individually changed in step (m);
    • (o) obtaining a fifth modified antibody variable domain comprising the changed tyrosine amino acid residues from step (m that resulted in an increase in affinity as individually tested in step (n);
    • (p) sixthly, changing uncharged polar or nonpolar amino acid residues in the fifth modified antibody variable domain of step (o) to polar or nonpolar residues, respectively, of a different size;
    • (q) testing the effect on affinity of the binding partner by changing the size of the uncharged residue at each position individually changed in step (p);
    • (r) obtaining a sixth modified antibody variable domain comprising the changed amino acid residues from step (p) that resulted in an increase in affinity as individually tested in step (q).

91. The method of embodiment 90, wherein said method further comprises prior to step (a) identifying the proximity assigned to amino acid positions in an antibody variable domain using the “prox” line as shown in FIG. 3A or 3B.

92. The method of embodiment 91, wherein contacting (C) amino acid residues are changed in step (a).

93. A method for enhancing the binding affinity of an antibody variable domain to a binding partner, to obtain a modified antibody variable domain with enhanced affinity to the binding partner compared to an unmodified antibody variable domain, the method comprising:

    • (a) identifying the proximity assigned to amino acid positions in the antibody variable domain using the “prox” line as shown in FIG. 3A or 3B;
    • (b) changing amino acid residues at one or more contacting (C), peripheral (P), supporting (S) and interfacial (I) proximity positions in the domain to another amino acid, thereby generating a modified variable domain;
    • (c) testing the binding affinity of the modified antibody variable domain for the binding partner; and
    • (d) obtaining a modified antibody variable domain with enhanced binding affinity for the binding partner compared to the binding affinity exhibited by the unmodified antibody variable domain for the binding partner.

94. The method of embodiment 92, wherein each contacting (C), peripheral (P), supporting (S) and interfacial (I) residue in the antibody variable domain is separately substituted.

95. A method for enhancing the binding affinity of an antibody variable domain to a binding partner compared to an unmodified antibody variable domain, the method comprising:

    • (a) identifying the proximity assigned to amino acid positions in the antibody variable domain using the “prox” line as shown in FIG. 3A or 3B;
    • (b) changing amino acid residues at one or more contacting (C), peripheral (P), supporting (S) and interfacial (I) proximity positions in the domain to another amino acid, wherein charged residues are changed to uncharged or uncharged residues are changed to charged or both charged residues are changed to uncharged and uncharged residues are changed to charged, thereby generating a modified antibody variable domain;
    • (c) testing the binding affinity of the modified antibody variable domain for the binding partner; and
    • (d) obtaining a modified antibody variable domain with enhanced binding affinity to the binding partner compared to the binding affinity exhibited by the unmodified antibody variable domain for the binding partner.

96. The method of embodiment 94, wherein each contacting (C), peripheral (P), supporting (S) and interfacial (I) residue in the antibody variable domain is separately substituted.

97. A method for enhancing the binding affinity of an antibody variable domain to a binding partner compared to an unmodified antibody variable domain, the method comprising:

    • (a) identifying the proximity assigned to amino acid positions in the antibody variable domain using the “prox” line as shown in FIG. 3A or 3B;
    • (b) changing amino acid residues at one or more contacting (C), peripheral (P), supporting (S) and interfacial (I) proximity positions in the domain to another amino acid, wherein residues are changed to a different size within the same amino acid class, that is, positively charged, negatively charged, polar or nonpolar, thereby generating a modified antibody variable domain;
    • (c) testing the binding affinity of the modified antibody variable domain for the binding partner; and
    • (d) obtaining a modified antibody variable domain with enhanced binding affinity for the binding partner compared to the binding affinity exhibited by the unmodified antibody variable domain for the binding partner.

98. The method of embodiment 96, wherein each contacting (C), peripheral (P), supporting (S) and interfacial (I) residue in the antibody variable domain is separately substituted.

99. A method for enhancing the binding affinity of an antibody variable domain to a binding partner compared to an unmodified antibody variable domain, the method comprising:

    • (a) identifying the proximity assigned to amino acid positions in the antibody variable domain using the “prox” line as shown in FIG. 3A or 3B;
    • (b) changing amino acid residues at one or more contacting (C), peripheral (P), supporting (S) and interfacial (I) proximity positions in the domain to another amino acid, wherein polar residues are changed to nonpolar residues or nonpolar residues are changed to polar residues or both polar residues are changed to nonpolar residues and nonpolar residues are changed to polar residues, thereby generating a modified antibody variable domain;
    • (c) testing the binding affinity of the modified antibody variable domain for the binding partner; and
    • (d) obtaining a modified antibody variable domain with enhanced binding affinity to the binding partner compared to the binding affinity exhibited by the unmodified antibody variable domain for the binding partner.

100. The method of embodiment 98, wherein each contacting (C), peripheral (P), supporting (S) and interfacial (I) residue in the antibody variable domain is separately substituted.

101. A method for enhancing the binding affinity of an antibody variable domain to a binding partner compared to an unmodified antibody variable domain, the method comprising:

    • (a) identifying the proximity assigned to amino acid positions in the antibody variable domain using the “prox” line as shown in FIG. 3A or 3B;
    • (b) changing amino acid residues at contacting (C), peripheral (P), supporting (S) and interfacial (I) proximity positions in the domain to another amino acid, wherein charged residues are changed to opposite charge residues, that is, from positively charged to negatively charged or from negatively charged to positively charged, thereby generating a modified antibody variable domain;
    • (c) testing the binding affinity of the modified antibody variable domain for the binding partner; and
    • (d) obtaining a modified antibody variable domain with enhanced binding affinity to the binding partner compared to the binding affinity exhibited by the unmodified antibody variable domain for the binding partner.

102. The method of embodiment 100, wherein each contacting (C), peripheral (P), supporting (S) and interfacial (I) residue in the antibody variable domain is separately substituted.

While the present disclosure has been described and illustrated herein by references to various specific materials, procedures and examples, it is understood that the disclosure is not restricted to the particular combinations of material and procedures selected for that purpose. Numerous variations of such details can be implied as will be appreciated by those skilled in the art. It is intended that the specification and examples be considered as exemplary, only, with the true scope and spirit of the disclosure being indicated by the following claims. All references, patents, and patent applications referred to in this application are herein incorporated by reference in their entirety.

Claims

1. A method for enhancing the binding affinity of an antibody variable domain to a binding partner, to obtain a modified antibody variable domain with enhanced affinity to the binding partner compared to an unmodified antibody variable domain, the method comprising:

(a) identifying the proximity assigned to amino acid positions in the antibody variable domain using the “prox” line as shown in FIG. 3A or 3B;
(b) changing amino acid residues at one or more contacting (C), peripheral (P), supporting (S) or interfacial (I) proximity positions in the domain to another amino acid, thereby generating a modified variable domain;
(c) testing the binding affinity of the modified antibody variable domain for the binding partner; and
(d) obtaining a modified antibody variable domain with enhanced binding affinity for the binding partner compared to the binding affinity exhibited by the unmodified antibody variable domain for the binding partner.

2-19. (canceled)

20. A method for enhancing the binding affinity of an antibody variable domain to a binding partner compared to an unmodified antibody variable domain, the method comprising:

(a) identifying the proximity assigned to amino acid positions in the antibody variable domain using the “prox” line as shown in FIG. 3A or 3B;
(b) changing amino acid residues at one or more contacting (C), peripheral (P), supporting (S) or interfacial (I) proximity positions in the domain to another amino acid, wherein charged residues are changed to uncharged or uncharged residues are changed to charged or both charged residues are changed to uncharged and uncharged residues are changed to charged, thereby generating a modified antibody variable domain;
(c) testing the binding affinity of the modified antibody variable domain for the binding partner; and
(d) obtaining a modified antibody variable domain with enhanced binding affinity to the binding partner compared to the binding affinity exhibited by the unmodified antibody variable domain for the binding partner.

21-38. (canceled)

39. A method for enhancing the binding affinity of an antibody variable domain to a binding partner compared to an unmodified antibody variable domain, the method comprising:

(a) identifying the proximity assigned to amino acid positions in the antibody variable domain using the “prox” line as shown in FIG. 3A or 3B;
(b) changing amino acid residues at one or more contacting (C), peripheral (P), supporting (S) or interfacial (I) proximity positions in the domain to another amino acid, wherein residues are changed to a different size within the same amino acid class, that is, positively charged, negatively charged, polar or nonpolar, thereby generating a modified antibody variable domain;
(c) testing the binding affinity of the modified antibody variable domain for the binding partner; and
(d) obtaining a modified antibody variable domain with enhanced binding affinity for the binding partner compared to the binding affinity exhibited by the unmodified antibody variable domain for the binding partner.

40-57. (canceled)

58. A method for enhancing the binding affinity of an antibody variable domain to a binding partner compared to an unmodified antibody variable domain, the method comprising:

(a) identifying the proximity assigned to amino acid positions in the antibody variable domain using the “prox” line as shown in FIG. 3A or 3B;
(b) changing amino acid residues at contacting (C), peripheral (P), supporting (S) or interfacial (I) proximity positions in the domain to another amino acid, wherein charged residues are changed to opposite charge residues, that is, from positively charged to negatively charged or from negatively charged to positively charged, thereby generating a modified antibody variable domain;
(c) testing the binding affinity of the modified antibody variable domain for the binding partner; and
(d) obtaining a modified antibody variable domain with enhanced binding affinity to the binding partner compared to the binding affinity exhibited by the unmodified antibody variable domain for the binding partner.

59-76. (canceled)

77. A method for enhancing the binding affinity of an antibody variable domain to a binding partner compared to an unmodified antibody variable domain, the method comprising:

(a) identifying the proximity assigned to amino acid positions in the antibody variable domain using the “prox” line as shown in FIG. 3A or 3B;
(b) changing amino acid residues at one or more contacting (C), peripheral (P), supporting (S) or interfacial (I) proximity positions in the domain to another amino acid, wherein polar residues are changed to nonpolar residues or nonpolar residues are changed to polar residues or both polar residues are changed to nonpolar residues and nonpolar residues are changed to polar residues, thereby generating a modified antibody variable domain;
(c) testing the binding affinity of the modified antibody variable domain for the binding partner; and
(d) obtaining a modified antibody variable domain with enhanced binding affinity to the binding partner compared to the binding affinity exhibited by the unmodified antibody variable domain for the binding partner.

78-95. (canceled)

96. A method for enhancing the binding affinity of an antibody variable domain to a binding partner, said method comprising:

(a) firstly, changing amino acid residues in the antibody variable domain to other amino acid residues, wherein charged residues are changed to uncharged residues thereby removing a charge and uncharged residues are changed to charged residues thereby adding a charge;
(b) testing the effect on affinity of the binding partner by removing or adding a charged amino acid residue at each position individually changed in step (a);
(c) obtaining a first modified antibody variable domain comprising the changed amino acid residues from step (a) that resulted in an increase in affinity as individually tested in step (b);
(d) secondly, changing amino acid residues that were firstly changed in step (a) from charged to uncharged or from uncharged to charged to amino acid residues of the opposite charge, if removing the charged residue in step (a) had caused a gain in affinity when tested in step (b) or if adding the charged residue in step (a) caused a loss in affinity when tested in step (b);
(e) testing the effect on affinity of the binding partner by changing the amino acid residue to the opposite charged residue at each position individually changed in step (d);
(f) obtaining a second modified antibody variable domain comprising the changed residues from step (d) that resulted in an increase in affinity as individually tested in step (e);
(g) thirdly, changing charged amino acid residues in the second modified antibody variable domain of step (f) to an amino acid residue of a different size and the same charge;
(h) testing the effect on affinity of the binding partner by changing the size of the charged residue at each position individually changed in step (g);
(i) obtaining a third modified antibody variable domain comprising the changed amino acid residues from step (g) that resulted in an increase in affinity as individually tested in step (h);
(j) fourthly, changing uncharged amino acid residues in the third modified antibody variable domain of step (i) to other amino acid residues, wherein polar uncharged residues are changed to nonpolar uncharged residues and wherein nonpolar uncharged residues are changed to polar uncharged residues;
(k) testing the effect on affinity of the binding partner by changing polar to nonpolar and nonpolar to polar at each position individually changed in step (j);
(l) obtaining a fourth modified antibody variable domain comprising the changed amino acid residues from step (j) that resulted in an increase in affinity as individually tested in step (k);
(m) fifthly, changing uncharged residues that were fourthly changed in step (j) from polar to nonpolar or from nonpolar to polar to tyrosine, if changing from polar to nonpolar or from nonpolar to polar in step (j) had not caused a gain or loss in affinity when tested in step (k);
(n) testing the effect on affinity of the binding partner by changing to tyrosine at each position individually changed in step (m);
(o) obtaining a fifth modified antibody variable domain comprising the changed tyrosine amino acid residues from step (m that resulted in an increase in affinity as individually tested in step (n);
(p) sixthly, changing uncharged polar or nonpolar amino acid residues in the fifth modified antibody variable domain of step (o) to polar or nonpolar residues, respectively, of a different size;
(q) testing the effect on affinity of the binding partner by changing the size of the uncharged residue at each position individually changed in step (p);
(r) obtaining a sixth modified antibody variable domain comprising the changed amino acid residues from step (p) that resulted in an increase in affinity as individually tested in step (q).

97-113. (canceled)

Patent History
Publication number: 20110118447
Type: Application
Filed: Dec 31, 2008
Publication Date: May 19, 2011
Applicant: XOMA TECHNOLOGY LTD. (Berkeley, CA)
Inventor: Gary Studnicka (Berkeley, CA)
Application Number: 12/811,334
Classifications
Current U.S. Class: Polyclonal Antibody Or Immunogloblin Of Identified Binding Specificity (530/389.1)
International Classification: C07K 1/107 (20060101);