Antibody libraries

This application is a divisional of U.S. application Ser. No. 13/810,570, filed on Apr. 1, 2013, which is a national phase entry under 35 U.S.C. § 371 of International Application No. PCT/US2011/044063, filed on Jul. 14, 2011, which claims priority to U.S. Provisional Application No. 61/365,194, filed on Jul. 16, 2010. The disclosure of each of U.S. application Ser. No. 13/810,570 and U.S. Provisional Application No. 61/365,194 are hereby incorporated herein by reference in their entirety.

In accordance with 37 CFR 1.52(e)(5), a Sequence Listing in the form of a text file (entitled “Sequence Listing.txt” created on May 10, 2016 and 2,451 kilo bytes in size) is incorporated by reference in its entirety.

Antibodies have profound relevance as research tools and in diagnostic and therapeutic applications. However, the identification of useful antibodies is difficult and once identified, antibodies often require considerable redesign or “humanization” before they are suitable for therapeutic applications in humans.

Many methods for identifying antibodies involve display of libraries of antibodies derived by amplification of nucleic acids from B cells or tissues. Some of these methods have utilized synthetic libraries. However, many of these approaches have limitations. For example, most human antibody libraries known in the art contain only the antibody sequence diversity that can be experimentally captured or cloned from a biological source (e.g., B cells). Accordingly, such libraries may over-represent some sequences, while completely lacking or under-representing other sequences, particularly those binding human antigens. Most synthetic libraries known in the art have other limitations, such as the occurrence of unnatural (i.e., non-human) amino acid sequence motifs that have the potential to be immunogenic.

Accordingly, a need exists for diverse antibody libraries that contain candidate antibodies that are non-immunogenic (i.e., are human) and have desired properties (e.g., the ability to recognize a broad variety of antigens). However, obtaining such libraries requires balancing the competing objectives of generating diverse libraries while still maintaining the human character of the sequences within the library. The current invention provides antibody libraries that have these and other desirable features, and methods of making and using such libraries.

The invention provides, among other things, improvements in the design and production of synthetic libraries that mimic the diversity of the natural human repertoire of CDRH3, CDRL3, heavy chain, light chain, and/or full-length (intact) antibody sequences. In some embodiments the invention defines and provides methods of generating theoretical segment pools of TN1, DH, N2, and H3-JH segments to consider for inclusion in a physical manifestation of a library (e.g., polynucleotide or polypeptide) comprising or encoding CDRH3 sequences (e.g., an antibody library). In certain embodiments the invention defines and provides methods of matching the individual members of these theoretical segment pools to a reference set of CDRH3 sequences, to determine the frequency of occurrence (or segment usage weight) of each of the segments in the theoretical segment pool in the reference set. While any set of CDRH3 sequences may be used as a reference set, the invention also defines and provides methods of generating particular reference sets or subsets of interest. For example, among other things, the present invention provides methods for filtering an original reference set to obtain a provided reference set with a preimmune character. Also provided are methods to define and/or identify segments that occur within the CDRH3 sequences in the reference set but not in the theoretical segment pool. Such segments can be added to a theoretical segment pool, for example in order to be considered for inclusion in a physical library. Although the frequency of occurrence of a particular segment in a reference set is useful to select segments for inclusion in a physical library, the invention also provides a number of physicochemical and biological properties that can be used (alone or together with any other criterion or criteria) to select segments for inclusion in a physical library.

In some embodiments the invention provides libraries that differ from certain other libraries known in the art in that they are not sitewise-stochastic in composition or sequence, and are therefore intrinsically less random than these certain other libraries of the art (see e.g., Example 14 of US Pub. No. 2009/0181855, incorporated by reference in its entirety, for a discussion of information content and randomness). In some embodiments, degenerate oligonucleotides may be used to further increase the diversity of the members of a library while further improving matching with a reference set of sequences (e.g., CDRH3, CDRL3, heavy chain, light chain, and/or full-length (intact) antibody sequences).

The invention also provides libraries whose members have sequences that are related to one another in that they would be selected for inclusion in a physical library by performing the analyses described herein, for example by generating a CDRH3 reference set as in Example 3; generating theoretical segment pools as in Examples 5-7; matching the members of a theoretical segment pool to the reference set as in Examples 4 and 8; and selecting members of the theoretical segment pool for inclusion in a physical library as in Examples 8-9. Also provided are methods of further increasing diversity in certain sequences by utilizing degenerate oligonucleotides as in Examples 12-16.

In some embodiments, the present invention provides polynucleotide and polypeptide libraries comprising CDRH3, CDRL3, heavy chain, light chain, and/or full-length (intact) antibody sequences, and methods of making and using such libraries.

In some embodiments, the invention provides libraries comprising, consisting essentially of, or consisting of any of the libraries or theoretical segment pools described herein.

In some embodiments, the present invention recognizes that by mimicking the in vivo activity of the enzyme TdT computationally, theoretical segment pools can be generated and subsequently matched to large reference datasets of CDR sequences to choose, for inclusion in a library, those theoretical segments that best recapitulate the CDR sequences in the reference data sets.

In certain embodiments, the invention provides libraries of polynucleotides comprising at least about 10 4 polynucleotides encoding CDRH3 polypeptides with the structure: [TN1]-[DH]-[N2]-[H3-JH], wherein: TN1 is a polypeptide corresponding to any of the TN1 polypeptides of Tables 9-10 and 18-26, or a polypeptide produced by translation of any of the TN1 polynucleotides of Tables 25-26; DH is a polypeptide corresponding to any of the DH polypeptides of Tables 9, 11, 17-25 and 28, or a polypeptide produced by translation of any of the DH-encoding polynucleotides of Tables 16, 25 and 27; N2 is a polypeptide corresponding to any of the N2 polypeptides of Tables 9, 12, 18-25 and 30, or a polypeptide produced by translation of any of the N2-encoding polynucleotides of Tables 25 and 29; and H3-JH is a polypeptide corresponding to any of the H3-JH polypeptides of Tables 9, 13, 15, 18-25 and 32, or a polypeptide produced by translation of any of the H3-JH-encoding polynucleotides of Tables 14, 25 and 31.

In some embodiments, the invention provides libraries wherein at least about 1%, 5%, or 10% of the sequences in the library have the structure provided above, or that of any of the libraries provided herein.

In certain embodiments, the invention provides libraries comprising polynucleotides encoding CDRH3 polypeptides produced by the sets of TN1, DH, N2, and H3-JH polypeptides provided in any one of Tables 23-25.

In some embodiments, the invention provides libraries comprising polynucleotides encoding CDRH3 polypeptides produced by the set of TN1 polypeptides provided in Table 26, the set of DH polypeptides provided in Table 28, the set of N2 polypeptides provided in Table 30 and the set of H3-JH polypeptides provided in Table 32.

In certain embodiments, the invention provides libraries whose members show (or encode polypeptides that show) at least a certain percent identity with the polypeptides described above, for example, a library comprising at least about 10 4 polynucleotides encoding CDRH3 polypeptides with the structure: [TN1]-[DH]-[N2]-[H3-JH], wherein: TN1 is a polypeptide at least about 80%, 90%, or 95% identical to any of the TN1 polypeptides of Tables 9-10 and 18-26, or a polypeptide at least about 80%, 90%, or 95% identical to a polypeptide produced by translation of any of the TN1 polynucleotides of Tables 25-26; DH is a polypeptide at least about 80%, 90%, or 95% identical to any of the DH polypeptides of Tables 9, 11, 17-25 and 28, or a polypeptide at least about 80%, 90%, or 95% identical to a polypeptide produced by translation of any of the DH-encoding polynucleotides of Tables 16, 25 and 27; N2 is a polypeptide at least about 80%, 90%, or 95% identical to any of the N2 polypeptides of Tables 9, 12, 18-25 and 30, or a polypeptide at least about 80%, 90%, or 95% identical to a polypeptide produced by translation of any of the N2-encoding polynucleotides of Tables 25 and 29; and H3-JH is a polypeptide at least about 80%, 90%, or 95% identical to any of the H3-JH polypeptides of Tables 9, 13, 15, 18-25 and 32, or a polypeptide at least about 80%, 90%, or 95% identical to a polypeptide produced by translation of any of the H3-JH-encoding polynucleotides of Tables 14, 25 and 31.

In some embodiments, the invention provides libraries comprising polynucleotides encoding light chain variable regions, wherein the light chain variable regions are selected from the group consisting of: (a) a VK1-05 sequence varied at one or more of positions 4, 49, and 46; (b) a VK1-12 sequence varied at one or more of positions 4, 49, 46, and 66; (c) a VK1-33 sequence varied at one or more of positions 4, 49, and 66; (d) a VK1-39 sequence varied at one or more of positions 4, 49, and 46; (e) a VK2-28 sequence varied at one or more of positions 2, 4, 46, and 49; (f) a VK3-11 sequence varied at one or more of positions 2, 4, 36, and 49; (g) a VK3-15 sequence varied at one or more of positions 2, 4, 48, and 49; (h) a VK3-20 sequence varied at one or more of positions 2, 4, 48, and 49; and/or (i) a VK4-1 sequence varied at one or more of positions 4, 46, 49, and 66.

In certain embodiments, the invention provides libraries comprising polynucleotides encoding light chain variable regions that comprise polypeptide sequences at least about 80%, 90%, or 95% identical to two or more of the light chain polypeptide sequences provided in Table 3.

In some embodiments, the invention provides libraries wherein the light chain variable regions comprise the polypeptide sequences provided in Table 3.

In certain embodiments, the invention provides libraries comprising polynucleotides encoding light chain variable regions, wherein the L3-VL polypeptide sequences of the light chain variable regions are varied at two or three residues between positions 89 to 94, inclusive, in comparison to an L3-VL germline sequence. In some embodiments, libraries containing a single light chain germline sequence and its variants are provided. In certain embodiments, variants produced from different light chain germline sequences can be combined to produce libraries encoding multiple light chain germline sequences and their variants. Any of the light chain L3-VL germline sequences provided herein may be varied at two or three residues between positions 89 to 94, inclusive, and one of ordinary skill in the art will recognize that any other L3-VL sequence can also be varied according to the principles described herein to produce libraries provided by the invention. In some embodiments, the present invention comprises libraries containing polynucleotides that encode antibody light chain variable regions, wherein the antibody light chain variable regions comprise one or more of the following L3-VL sequences: (i) an amino acid sequence that is identical to an L3-VL germline sequence (e.g., see Table 1); (ii) an amino acid sequence that contains two substitutions between residues 89-94, inclusive, in comparison to an L3-VL germline sequence; and (iii) an amino acid sequence that contains three substitutions between residues 89-94, inclusive, in comparison to an L3-VL germline sequence. In some embodiments, each antibody light chain variable region on a library includes one or more of the above L3-VL sequences. In some embodiments, such a library is combined with one or more sets of other nucleic acids that may or may not encode antibody light chain variable regions, and may or may not contain such L3-VL sequences. In some embodiments, the present invention comprises libraries containing polynucleotides that encode an antibody light chain variable region having an amino acid sequence as set forth in Table 4, or a polynucleotide sequence as set forth in one or more of Tables 5-7, wherein two or three residues at positions 89-94, inclusive, are varied.

In some embodiments, the present invention comprises libraries containing polynucleotides that encode an antibody light chain variable region, wherein, across the library, all encoded antibody light chain variable regions are identical to one another except for substitutions of residues at positions between residue 89 and residue 94, inclusive and further wherein, across the library, sequences of any two encoded antibody light chain variable regions differ from one another at not more than 3 positions.

In some embodiments, the invention provides libraries comprising polynucleotides encoding light chain variable regions comprising polypeptide sequences at least about 80%, 90%, or 95% identical to polypeptides produced by translation of two or more of the polynucleotide sequences provided in Tables 5-7. In certain embodiments all members of the library are at least about 80%, 90%, or 95% identical to polypeptides produced by translation of two or more of the polynucleotide sequences provided in Tables 5-7.

In certain embodiments, the invention provides a library comprising light chain variable regions that comprise the polypeptides produced by translation of the polynucleotide sequences provided in Tables 5-7. In certain embodiments, all members of the library comprise the polypeptides produced by translation of the polynucleotide sequences provided in Tables 5-7.

In some embodiments, any of the libraries described herein as containing or encoding CDRL3 and/or light chain variable regions, contains or encodes such CDRL3 and/or light chain variable regions in the context of complete light chains. Furthermore, in some embodiments, such libraries (and/or complete light chain libraries) further contain or encode one or more heavy chain CDRH3, variable domains, or intact heavy chains. In some embodiments, provided libraries include or encode intact antibodies such as, for example, intact intact IgGs.

In some embodiments, provided libraries include or encode human antibodies or antibody fragments; in some such embodiments, provided libraries include or encode intact human antibodies.

In certain embodiments, the invention provides libraries that comprise nucleic acid vectors containing library nucleic acids described herein. In many embodiments, each such library member comprises the same vector.

In some embodiments, the invention provides host cells containing one or more provided libraries, for example including a vector. In some embodiments the host cell is a yeast, and in certain embodiments the yeast is Saccharomyces cerevisiae.

In some embodiments, the invention provides antibodies isolated from the libraries described herein.

In certain embodiments, the invention provides kits containing any of the libraries described herein.

In some embodiments, the invention provides representations of libraries and/or theoretical segment pools in a computer readable format, for example, the TN1 polypeptides of Tables 10, 23-25 and 26; the DH polypeptides of Tables 11, 23-25 and 28; the N2 polypeptides of Tables 12, 23-25 and 30; the H3-JH polypeptides of Tables 13, 15, 17, 23-25 and 32; the TN1 polynucleotides of Tables 25-26; the DH polynucleotides of Tables 25 and 27; the N2 polynucleotides of Tables 25 and 29; and/or the H3-JH polynucleotides of Tables 25 and 31.

In certain embodiments, the invention provides a representation of the polynucleotide sequences of the Human Preimmune Set (Appendix A), or the polypeptide expression products thereof, in a computer readable format.

In some embodiments, the invention provides a method of making synthetic polynucleotides encoding a CDRH3 library, comprising: (a) providing a theoretical segment pool containing TN1, DH, N2, and H3-JH segments; (b) providing a reference set of CDRH3 sequences; (c) utilizing the theoretical segment pool of (a) to identify the closest match(es) to each CDRH3 sequence in the reference set of (b); (d) selecting segments from the theoretical segment pool for inclusion in a synthetic library; and (e) synthesizing the synthetic CDRH3 library. In certain embodiments, the invention provides libraries made by this method. In some embodiments, the segments selected for inclusion in the synthetic library are selected according to their segment usage weight in the reference set of CDRH3 sequences.

In certain embodiments, the invention provides a method of making synthetic polynucleotides encoding a CDRL3 library, comprising: (i) obtaining a reference set of light chain sequences, wherein the reference set contains light chain sequences with VL segments originating from the same IGVL germline gene and/or its allelic variants; (ii) determining which amino acids occur at each of the CDRL3 positions in the reference set that are encoded by the IGVL gene; (iii) synthesizing light chain variable domain encoding sequences wherein two positions between positions 89 and 94, inclusive, contain degenerate codons encoding two or more of the five most frequently occurring amino acid residues at the corresponding positions in the reference set; and (iv) synthesizing the polynucleotides encoding the CDRL3 library. In certain embodiments, the invention provides libraries made by this method.

In some embodiments, the invention provides a method of using any of the libraries of the invention to isolate an antibody binding an antigen, comprising contacting the polypeptide expression products of said libraries with an antigen and isolating polypeptide expression products that bind to the antigen.

In certain embodiments, the number of N-linked glycosylation sites, deamidation motifs, and/or Cys residues in the libraries of the invention are reduced or eliminated in comparison to libraries produced by amplification of a repertoire from a biological source.

The invention provides a number of polynucleotide and polypeptide sequences and segments that can be used to build larger polynucleotide and polypeptide sequences (e.g., TN1, DH, N2, and H3-JH segments that can be used to build CDRH3). One of ordinary skill in the art will readily recognize that in some instances these sequences can be more succinctly represented by providing consensus sequences after alignment of the sequences provided by the invention, and that these consensus sequences fall within the scope of the invention and may be used to more succinctly represent any of the sequences provided herein.

FIG. 1 shows that Vernier residues 4 and 49 (starred) in VK1-39 have a diversity index comparable to or greater than the diversity indices of the CDR positions (i.e., at or above 0.07, in this example).

FIG. 2 shows that clinically validated CDRL3 sequences deviate little from germline-like sequences (n=35).

FIG. 3 shows the percent of sequences in the jumping dimer CDRL3 libraries of the invention and a previous CDRL3 library, VK-v1.0, with X or fewer mutations from germline. Here, FX is the percentage of sequences in a library with X or fewer mutations from germline.

FIG. 4 shows the application of a provided method used to generate nucleotide sequences (SEQ ID NOS 8748-8759, respectively, in order of appearance) encoding the parent H3-JH segments.

FIG. 5 shows the general schematic of an approach used to select segments from a theoretical segment pool for inclusion in a theoretical and/or synthetic library.

FIG. 6 shows the frequency of “Good” and “Poor” expressing CDRH3 sequences isolated from yeast-based libraries described in US 2009/0181855, and their comparison to the sequences contained in the library design described therein (“Design”), as a function of the DH segment hydrophobicity (increasing to the right).

FIG. 7 shows the percentage of CDRH3 sequences in the LUA-141 library and Exemplary Library Design 3 (ELD-3) that match CDRH3 sequences from Lee-666 and Boyd-3000 with zero, one, two, three, or more than three amino acid mismatches.

FIG. 8 shows that Exemplary Library Design 3 (ELD-3) and the Extended Diversity Library Design both return better matches to clinically relevant CDRH3 sequences than the LUA-141 library.

FIG. 9 shows that the combinatorial efficiency of Exemplary Library Design 3 (ELD-3) is greater than that of the LUA-141 library. Specifically, the ELD-3 segments are more likely to yield a unique CDRH3 than the LUA-141 library segments.

FIG. 10 shows the amino acid compositions of the Kabat-CDRH3s of LUA-141, Exemplary Library Design 3 (ELD-3), and Human CDRH3 sequences from the HPS (Human H3).

FIG. 11 shows the Kabat-CDRH3 length distribution of LUA-141, Exemplary Library Design 3 (ELD-3), and Human CDRH3 sequences from the HPS (Human H3).

FIG. 12 shows the percentage of CDRH3 sequences in the Extended Diversity library that match CDRH3 sequences from Boyd et al. with zero to thirty-two amino acid mismatches

FIG. 13 shows the Kabat-CDRH3 length distribution of Exemplary Library Design 3 (“ELD-3”), the Extended Diversity Library Design (“Extended Diversity”) and human CDRH3 sequences from the Boyd et al. data set (“Boyd 2009”).

FIG. 14 shows the amino acid compositions of the Kabat-CDRH3s of the Extended Diversity Library Design (“Extended Diversity”) and human CDRH3 sequences from the Boyd et al. dataset (“Boyd 2009”).

FIG. 15 shows the combinatorial efficiency of the Extended Diversity Library Design by matching 20,000 randomly selected sequences from the (same) design. About 65% of the sequences appear only once in the design and about 17% appear twice.

The present invention provides, among other things, polynucleotide and polypeptide libraries, methods of producing and using the libraries, kits containing the libraries, and computer readable forms of representations of libraries and/or theoretical segment pools disclosed herein. Libraries taught in this application can be described, at least in part, in terms of components (e.g., polynucleotide or polypeptide “segments”) from which they are assembled. Among other things, the present invention specifically provides and contemplates these polynucleotide or polypeptide segments, methods of producing and using such segments, and kits and computer readable forms of representations that include such library segments.

In certain embodiments, the invention provides antibody libraries specifically designed based on sequences and CDR length distribution in a naturally occurring human antibody repertoire. It is estimated that, even in the absence of antigenic stimulation, an individual human makes at least about 10 7 different antibody molecules (Boyd et al., Science Translational Medicine, 2009, 1:1). The antigen-binding sites of many antibodies can cross-react with a variety of related but different epitopes. In addition, the human antibody repertoire is large enough to ensure that there is an antigen-binding site to fit almost any potential epitope, albeit potentially with low affinity.

The mammalian immune system has evolved unique genetic mechanisms that enable it to generate an almost unlimited number of different light and heavy chains in a remarkably economical way, by combinatorially joining chromosomally separated gene segments prior to transcription. Each type of immunoglobulin (Ig) chain (i.e., kappa light, lambda light, and heavy) is synthesized by combinatorial assembly of DNA sequences, selected from two or more families of gene segments, to produce a single polypeptide chain. Specifically, the heavy chains and light chains each consist of a variable region and a constant (C) region. The variable regions of the heavy chains are encoded by DNA sequences assembled from three families of gene sequences: variable (IGHV), diversity (IGHD), and joining (IGHJ). The variable regions of light chains are encoded by DNA sequences assembled from two families of gene sequences for each of the kappa and lambda light chains: variable (IGLV) and joining (IGLJ). Each variable region (heavy and light) is also recombined with a constant region, to produce a full-length immunoglobulin chain.

While combinatorial assembly of the V, D and J gene segments make a substantial contribution to antibody variable region diversity, further diversity is introduced in vivo, at the pre-B cell stage, via imprecise joining of these gene segments and the introduction of non-templated nucleotides at the junctions between the gene segments (see e.g., U.S. Pub. No. 2009/0181855, which is incorporated by reference in its entirety, for more information).

After a B cell recognizes an antigen, it is induced to proliferate. During proliferation, the B cell receptor locus undergoes an extremely high rate of somatic mutation that is far greater than the normal rate of genomic mutation. The mutations that occur are primarily localized to the Ig variable regions and comprise substitutions, insertions and deletions. This somatic hypermutation enables the production of B cells that express antibodies possessing enhanced affinity toward an antigen. Such antigen-driven somatic hypermutation fine-tunes antibody responses to a given antigen.

Synthetic antibody libraries of the instant invention have the potential to recognize any antigen, including antigens of human origin. The ability to recognize antigens of human origin may not be present in other antibody libraries, such as antibody libraries prepared from human biological sources (e.g., from human cDNA), because self-reactive antibodies are removed by the donor's immune system via negative selection.

Still further, the present invention provides strategies that streamline and/or simplify certain aspects of library development and/or screening. For example, in some embodiments, the present invention permits use of cell sorting technologies (e.g., fluorescence activated cell sorting, FACS) to identify positive clones, and therefore bypasses or obviates the need for the standard and tedious methodology of generating a hybridoma library and supernatant screening.

Yet further, in some embodiments, the present invention provides libraries and/or sublibraries that accommodate multiple screening passes. For example, in some embodiments, provided libraries and/or sublibraries can be screened multiple times. In some such embodiments, individual provided libraries and/or sublibraries can be used to discover additional antibodies against many targets.

Before further description of the invention, certain terms are defined.

Unless defined otherwise, all technical and scientific terms used herein have the meaning commonly understood by one of ordinary skill in the art relevant to the invention. Unless otherwise specified, the Kabat numbering system is used throughout the application. The definitions below supplement those in the art and are directed to the embodiments described in the current application.

The term “amino acid” or “amino acid residue,” as would be understood by one of ordinary skill in the art, typically refers to an amino acid having its art recognized definition such as an amino acid selected from the group consisting of: alanine (Ala or A); arginine (Arg or R); asparagine (Asn or N); aspartic acid (Asp or D); cysteine (Cys or C); glutamine (Gln or Q); glutamic acid (Glu or E); glycine (Gly or G); histidine (His or H); isoleucine (Ile or I): leucine (Leu or L); lysine (Lys or K); methionine (Met or M); phenylalanine (Phe or F); proline (Pro or P); serine (Ser or S); threonine (Thr or T); tryptophan (Trp or W); tyrosine (Tyr or Y); and valine (Val or V), although modified, synthetic, or rare amino acids may be used as desired. Generally, amino acids can be grouped as having a non-polar side chain (e.g., Ala, Cys, Ile, Leu, Met, Phe, Pro, Val); a negatively charged side chain (e.g., Asp, Glu); a positively charged side chain (e.g., Arg, His, Lys); or an uncharged polar side chain (e.g., Asn, Cys, Gln, Gly, His, Met, Phe, Ser, Thr, Trp, and Tyr).

As would be understood by those of ordinary skill in the art, the term “antibody” is used herein in the broadest sense and specifically encompasses at least monoclonal antibodies, polyclonal antibodies, multi-specific antibodies (e.g., bispecific antibodies), chimeric antibodies, humanized antibodies, human antibodies, and antibody fragments. An antibody is a protein comprising one or more polypeptides substantially or partially encoded by immunoglobulin genes or fragments of immunoglobulin genes. The recognized immunoglobulin genes include the kappa, lambda, alpha, gamma, delta, epsilon and mu constant region genes, as well as myriad immunoglobulin variable region genes.

The term “antibody binding region” refers to one or more portions of an immunoglobulin or antibody variable region capable of binding an antigen(s). Typically, the antibody binding region is, for example, an antibody light chain (or variable region or one or more CDRs thereof), an antibody heavy chain (or variable region or one or more CDRs thereof), a heavy chain Fd region, a combined antibody light and heavy chain (or variable regions thereof) such as a Fab, F(ab) 2 , single domain, or single chain antibodies (scFv), or any region of a full length antibody that recognizes an antigen, for example, an IgG (e.g., an IgG1, IgG2, IgG3, or IgG4 subtype), IgA1, IgA2, IgD, IgE, or IgM antibody.

“Antibody fragments” comprise a portion of an intact antibody, for example, one or more portions of the antigen-binding region thereof. Examples of antibody fragments include Fab, Fab′, F(ab′) 2 , and Fv fragments, diabodies, linear antibodies, single-chain antibodies, and multi-specific antibodies formed from intact antibodies and antibody fragments.

The term “antibody of interest” refers to an antibody that has a property of interest that is identified and/or isolated from a library of the invention. Exemplary properties of interest include, for example, but are not limited to, binding to a particular antigen or epitope, binding with a certain affinity, cross-reactivity, blocking a binding interaction between two molecules, and/or eliciting a certain biological effect.

The term “canonical structure,” as understood by those of ordinary skill in the art, refers to the main chain conformation that is adopted by the antigen binding (CDR) loops. From comparative structural studies, it has been found that five of the six antigen binding loops have only a limited repertoire of available conformations. Each canonical structure can be characterized by the torsion angles of the polypeptide backbone. Correspondent loops between antibodies may, therefore, have very similar three dimensional structures, despite high amino acid sequence variability in most parts of the loops (Chothia and Lesk, J. Mol. Biol., 1987, 196: 901; Chothia et al., Nature, 1989, 342: 877; Martin and Thornton, J. Mol. Biol., 1996, 263: 800, each of which is incorporated by reference in its entirety). Furthermore, there is a relationship between the adopted loop structure and the amino acid sequences surrounding it. As is known in the art, the conformation of a particular canonical class is determined by the length of the loop and the amino acid residues residing at key positions within the loop, as well as within the conserved framework (i.e., outside of the loop). Assignment to a particular canonical class can therefore be made based on the presence of these key amino acid residues. The term “canonical structure” may also include considerations as to the linear sequence of the antibody, for example, as catalogued by Kabat (Kabat et al., in “Sequences of Proteins of Immunological Interest,” 5 th Edition, U.S. Department of Heath and Human Services, 1992). The Kabat numbering scheme is a widely adopted standard for numbering the amino acid residues of an antibody variable domain in a consistent manner, and is used herein unless indicated otherwise. Additional structural considerations can also be used to determine the canonical structure of an antibody. For example, those differences not fully reflected by Kabat numbering can be described by the numbering system of Chothia et al. and/or revealed by other techniques, for example, crystallography and two or three-dimensional computational modeling. Accordingly, a given antibody sequence may be placed into a canonical class which allows for, among other things, identifying appropriate chassis sequences (e.g., based on a desire to include a variety of canonical structures in a library). Kabat numbering of antibody amino acid sequences and structural considerations as described by Chothia et al., and their implications for construing canonical aspects of antibody structure, are described in the literature.

The terms “CDR”, and its plural “CDRs”, refer to a complementarity determining region (CDR) of which three make up the binding character of a light chain variable region (CDRL1, CDRL2 and CDRL3) and three make up the binding character of a heavy chain variable region (CDRH1, CDRH2 and CDRH3). CDRs contribute to the functional activity of an antibody molecule and are separated by amino acid sequences that comprise framework regions. The exact definitional CDR boundaries and lengths are subject to different classification and numbering systems. CDRs may therefore be referred to by Kabat, Chothia, contact or other boundary definitions, including for example the CDRH3 numbering system described below. Despite differing boundaries, each of these systems has some degree of overlap in what constitutes the so called “hypervariable regions” within the variable region. CDR definitions according to these systems may therefore differ in length and boundary areas with respect to the adjacent framework region. See, for example Kabat et al., in “Sequences of Proteins of Immunological Interest,” 5 th Edition, U.S. Department of Health and Human Services, 1992; Chothia et al., J. Mol. Biol., 1987, 196: 901; and MacCallum et al., J. Mol. Biol., 1996, 262: 732, each of which is incorporated by reference in its entirety.

The “CDRH3 numbering system” used herein defines the first amino acid of CDRH3 as being at position 95 and the last amino acid of CDRH3 as position 102. Note that this is a custom numbering system that is not according to Kabat. The amino acid segment, beginning at position 95 is called “TN1” and, when present, is assigned numbers 95, 96, 96A, 96B, etc. Note that the nomenclature used in the current application is slightly different from that used in U.S. Publication Nos. 2009/0181855 and 2010/0056386, and WO/2009/036379. In those applications, position 95 was designated a “Tail” residue, while here, the Tail (T) has been combined with the N1 segment, to produce one segment, designated “TN1.” The TN1 segment is followed by the “DH” segment, which is assigned numbers 97, 97A, 97B, 97C, etc. The DH segment is followed by the “N2” segment, which, when present, is numbered 98, 98A, 98B, etc. Finally, the most C-terminal amino acid residue of the “H3-JH” segment is designated as number 102. The residue directly before (N-terminal) it, when present, is 101, and the one before (if present) is 100. The rest of the H3-JH amino acids are numbered in reverse order, beginning with 99 for the amino acid just N-terminal to 100, 99A for the residue N-terminal to 99, and so forth for 99B, 99C, etc. Examples of CDRH3 sequence residue numbers may therefore include the following:

“Chassis” of the invention are portions of the antibody heavy chain variable (IGHV) or light chain variable (IGLV) domains that are not part of CDRH3 or CDRL3, respectively. A chassis of the invention is defined as the portion of the variable region of an antibody beginning with the first amino acid of FRM1 and ending with the last amino acid of FRM3. In the case of the heavy chain, the chassis includes the amino acids including from position 1 to position 94. In the case of the light chains (kappa and lambda), the chassis are defined as including from position 1 to position 88. The chassis of the invention may contain certain modifications relative to the corresponding germline variable domain sequences. These modifications may be engineered (e.g., to remove N-linked glycosylation sites) or naturally occurring (e.g., to account for naturally occurring allelic variation). For example, it is known in the art that the immunoglobulin gene repertoire is polymorphic (Wang et al., Immunol. Cell. Biol., 2008, 86: 111; Collins et al., Immunogenetics, 2008, 60: 669, each incorporated by reference in its entirety); chassis, CDRs and constant regions representative of these allelic variants are also encompassed by the invention. In some embodiments, the allelic variant(s) used in a particular embodiment of the invention may be selected based on the allelic variation present in different patient populations, for example, to identify antibodies that are non-immunogenic in these patient populations. In certain embodiments, the immunogenicity of an antibody of the invention may depend on allelic variation in the major histocompatibility complex (MHC) genes of a patient population. Such allelic variation may also be considered in the design of libraries of the invention. In certain embodiments of the invention, the chassis and constant regions are contained in a vector, and a CDR3 region is introduced between them via homologous recombination.

As used herein, a sequence designed with “directed diversity” has been specifically designed to contain both sequence diversity and length diversity. Directed diversity is not stochastic.

As used herein, the term “diversity” refers to a variety or a noticeable heterogeneity. The term “sequence diversity” refers to a variety of sequences which are collectively representative of several possibilities of sequences, for example, those found in natural human antibodies. For example, CDRH3 sequence diversity may refer to a variety of possibilities of combining the known human TN1, DH, N2, and H3-JH segments to form CDRH3 sequences. The CDRL3 sequence diversity (kappa or lambda) may refer to a variety of possibilities of combining the naturally occurring light chain variable region contributing to CDRL3 (i.e., “L3-VL”) and joining (i.e., “L3-JL”) segments, to form CDRL3 sequences. As used herein, “H3-JH” refers to the portion of the IGHJ gene contributing to CDRH3. As used herein, “L3-VL” and “L3-JL” refer to the portions of the IGLV and IGLJ genes (kappa or lambda) contributing to CDRL3, respectively.

As used herein, the term “expression” refers to steps involved in the production of a polypeptide including, but not limited to, transcription, post-transcriptional modification, translation, post-translational modification, and secretion.

The term “framework region” refers to the art-recognized portions of an antibody variable region that exist between the more divergent (i.e., hypervariable) CDRs. Such framework regions are typically referred to as frameworks 1 through 4 (FRM1, FRM2, FRM3, and FRM4) and provide a scaffold for the presentation of the six CDRs (three from the heavy chain and three from the light chain) in three dimensional space, to form an antigen-binding surface.

The term “full-length heavy chain” refers to an immunoglobulin heavy chain that contains each of the canonical structural domains of an immunoglobulin heavy chain, including the four framework regions, the three CDRs, and the constant region.

The term “full-length light chain” refers to an immunoglobulin light chain that contains each of the canonical structural domains of an immunoglobulin light chain, including the four framework regions, the three CDRs, and the constant region.

The term “germline-like,” when used with respect to the CDRL3 sequences of the light chains of the invention, means those sequences consisting of combinations of: (i) the first six wild-type residues contributed to CDRL3 by the IGVL germline gene (i.e., positions 89 to 94 in the Kabat numbering system; “L” is kappa or lambda); and (ii) one of several amino acid sequences, two one to four amino acids in length, largely, but not exclusively, derived from the JL segment (“L,” again is kappa or lambda). For kappa CDRL3 sequences of the most common lengths (i.e., 8, 9, and 10 residues), the sequences of (ii) number twenty and are: FT, LT, IT, RT, WT, YT, [X]T, [X]PT, [X]FT, [X]LT, [X]IT, [X]RT, [X]WT, [X]YT, [X]PFT, [X]PLT, [X]PIT, [X]PRT, [X]PWT and [X]PYT, where [X] corresponds to the amino acid residue found at position 95 (Kabat) in the respective VK germline sequence. X is most commonly P, but may also be S or any other amino acid residue found at position 95 of a VK germline sequence. For eight exemplified VK chassis exemplified herein, the corresponding 160 germline-like sequences, (i.e., 20 sequences of two to four amino acids in length combined with positions 89 to 94 of each of eight VK germline sequences) are provided in Table 1. A similar approach is applied to define germline-like CDRL3 sequences for lambda light chains. As for the kappa sequences described above, the intact, un-mutated portions of CDRL3 encoded by the IGVL genes (in this case, IGVλ) would be combined with the sequences largely, but not exclusively, derived from the Jλ segment. Here, the latter sequences (corresponding to (ii), above), number five and are: YV, VV, WV, AV or V. In addition, and as described in Example 7 of US 2009/0818155, one could further allow for variation at the last position of the Vλ-gene-encoded portion of CDRL3 by considering partial codons, while still considering the resulting sequences “germline-like.” More specifically, the entire “minimalist library” of Example 7 in US 2009/0818155 would be defined as “germline-like.” One of ordinary skill in the art will readily recognize that these methods can be extended to other VK and Vλ sequences.

The term “genotype-phenotype linkage,” as understood by those of ordinary skill in the art, refers to the fact that the nucleic acid (genotype) encoding a protein with a particular phenotype (e.g., binding an antigen) can be isolated from a library. For the purposes of illustration, an antibody fragment expressed on the surface of a phage can be isolated based on its binding to an antigen (e.g., U.S. Pat. No. 5,837,500). The binding of the antibody to the antigen simultaneously enables the isolation of the phage containing the nucleic acid encoding the antibody fragment. Thus, the phenotype (antigen-binding characteristics of the antibody fragment) has been “linked” to the genotype (nucleic acid encoding the antibody fragment). Other methods of maintaining a genotype-phenotype linkage include those of Wittrup et al. (U.S. Pat. Nos. 6,300,065, 6,331,391, 6,423,538, 6,696,251, 6,699,658, and U.S. Pub. No. 20040146976, each of which is incorporated by reference in its entirety), Miltenyi (U.S. Pat. No. 7,166,423, incorporated by reference in its entirety), Fandl (U.S. Pat. No. 6,919,183, US Pub No. 20060234311, each incorporated by reference in its entirety), Clausell-Tormos et al. (Chem. Biol., 2008, 15: 427, incorporated by reference in its entirety), Love et al. (Nat. Biotechnol., 2006, 24: 703, incorporated by reference in its entirety), and Kelly et al. (Chem. Commun., 2007, 14: 1773, incorporated by reference in its entirety). The term can be used to refer to any method which localizes an antibody protein together with the gene encoding the antibody protein, in a way in which they can both be recovered while the linkage between them is maintained.

The term “heterologous moiety” is used herein to indicate the addition of a moiety to an antibody wherein the moiety is not part of a naturally-occurring antibody. Exemplary heterologous moieties include drugs, toxins, imaging agents, and any other compositions which might provide an activity that is not inherent in the antibody itself.

As used herein, the term “host cell” is intended to refer to a cell comprising a polynucleotide of the invention. It should be understood that such terms refer not only to the particular subject cell but to the progeny or potential progeny of such a cell. Because certain modifications may occur in succeeding generations due to either mutation or environmental influences, such progeny may not, in fact, be identical to the parent cell, but are still included within the scope of the term as used herein.

As used herein, the term “human antibody CDR library” includes at least a polynucleotide or polypeptide library which has been designed to represent the sequence diversity and length diversity of naturally occurring CDRs in human antibodies (e.g., the term “CDR” in “human antibody CDR library” may be substituted with “CDRL1,” “CDRL2,” “CDRL3,” “CDRH1,” “CDRH2,” and/or “CDRH3”). Known human CDR sequences are represented in various data sets, including Jackson et al., J. Immunol Methods, 2007, 324: 26; Martin, Proteins, 1996, 25: 130; Lee et al., Immunogenetics, 2006, 57: 917, Boyd et al., Science Translational Medicine, 2009, 1: 1, and WO/2009/036379, each of which is incorporated by reference in its entirety, and the HPS, which is provided in Appendix A.

The term “Human Preimmune Set,” or “HPS,” refers to a reference set of 3,571 curated human preimmune heavy chain sequences corresponding to the GI Nos. provided in Appendix A.

An “intact antibody” is one comprising full-length heavy- and light-chains (i.e., four frameworks, three CDRs, and a constant region for each of the heavy and light chains). An intact antibody is also referred to as a “full-length” antibody.

The term “length diversity” refers to a variety in the length of a family of nucleotide or amino acid sequence. For example, in naturally occurring human antibodies, the heavy chain CDR3 sequence varies in length, for example, from about 2 amino acids to over about 35 amino acids, and the light chain CDR3 sequence varies in length, for example, from about 5 to about 16 amino acids.

The term “library” refers to a set of entities comprising two or more entities having diversity as described herein, and/or designed according to the methods of the invention. For example, a “library of polynucleotides” refers to a set of polynucleotides comprising two or more polynucleotides having diversity as described herein, and/or designed according to the methods of the invention. A “library of polypeptides” refers to a set of polypeptides comprising two or more polypeptides having diversity as described herein, and/or designed according to the methods of the invention. A “library of synthetic polynucleotides” refers to a set of polynucleotides comprising two or more synthetic polynucleotides having diversity as described herein, and/or designed according to the methods of the invention. Libraries where all members are synthetic are also encompassed by the invention. A “human antibody library” refers to a set of polypeptides comprising two or more polypeptides having diversity as described herein, and/or designed according to the methods of the invention, for example a library designed to represent the sequence diversity and length diversity of naturally occurring human antibodies. In some embodiments, the term “library” may refer to a set of entities sharing similar structural or sequence characteristics, for example, a “heavy chain library,” “light chain library,” “antibody library,” and/or “CDRH3 library.”

The term “physical realization” refers to a portion of a theoretical (e.g., computer-based) or synthetic (e.g., oligonucleotide-based) diversity that can actually be physically sampled, for example, by any display methodology. Exemplary display methodology include: phage display, ribosomal display, and yeast display. For synthetic sequences, the size of the physical realization of a library depends on (1) the fraction of the theoretical diversity that can actually be synthesized, and (2) the limitations of the particular screening method. Exemplary limitations of screening methods include the number of variants that can be screened in a particular assay (e.g., ribosome display, phage display, yeast display) and the transformation efficiency of the host cells (e.g., yeast, mammalian cells, bacteria) which are used in a screening assay. For the purposes of illustration, given a library with a theoretical diversity of 10 12 members, an exemplary physical realization of the library (e.g., in yeast, bacterial cells, or ribosome display) that can maximally include 10 11 members will, therefore, sample about 10% of the theoretical diversity of the library. However, if fewer than 10 11 members of the library with a theoretical diversity of 10 12 are synthesized, and the physical realization of the library can maximally include 10 11 members, less than 10% of the theoretical diversity of the library is sampled in the physical realization of the library. Similarly, a physical realization of the library that can maximally include more than 10 12 members would “oversample” the theoretical diversity, meaning that each member may be present more than once (assuming that the entire 10 12 theoretical diversity is synthesized).

The term “polynucleotide(s)” refers to nucleic acids such as DNA molecules and RNA molecules and analogs thereof (e.g., DNA or RNA generated using nucleotide analogs or using nucleic acid chemistry). As desired, the polynucleotides may be made synthetically, e.g., using art-recognized nucleic acid chemistry or enzymatically using, e.g., a polymerase, and, if desired, be modified. Typical modifications include methylation, biotinylation, and other art-known modifications. In addition, the nucleic acid molecule can be single-stranded or double-stranded and, where desired, linked to a detectable moiety. The representation of nucleotide bases herein follows International Union of Pure and Applied Chemistry (IUPAC) nomenclature (see U.S. Pub. No. 2009/0181855, incorporated by reference in its entirety).

“Preimmune” antibody libraries have sequence diversities and length diversities similar to naturally occurring human antibody sequences before these sequences have undergone negative selection and/or somatic hypermutation. For example, the set of sequences described in Lee et al. (Immunogenetics, 2006, 57: 917, incorporated by reference in its entirety) and the Human Preimmune Set (HPS) described herein (see Appendix A) are believed to represent sequences from the preimmune repertoire. In certain embodiments of the invention, the sequences of the invention will be similar to these sequences (e.g., in terms of composition and length).

As used herein, the term “sitewise stochastic” describes a process of generating a sequence of amino acids, where only the amino acid occurrences at the individual positions are considered, and higher order motifs (e.g., pair-wise correlations) are not accounted for (e.g., see Knappik, et al., J Mol Biol, 2000, 296: 57, and analysis provided in U.S. Publication No. 2009/0181855, each incorporated by reference in its entirety).

The term “split-pool synthesis” refers to a procedure in which the products of a plurality of individual first reactions are combined (pooled) and then separated (split) before participating in a plurality of second reactions. For example, U.S. Publication No. 2009/0181855 (incorporated by reference in its entirety) describes the synthesis of 278 DH segments (products), each in a separate reaction. After synthesis, these 278 segments are combined (pooled) and then distributed (split) amongst 141 columns for the synthesis of the N2 segments. This enables the pairing of each of the 278 DH segments with each of 141 N2 segments.

As used herein, “stochastic” describes a process of generating a random sequence of nucleotides or amino acids, which is considered as a sample of one element from a probability distribution (e.g., see U.S. Pat. No. 5,723,323).

As used herein, the term “synthetic polynucleotide” refers to a molecule formed through a chemical process, as opposed to molecules of natural origin, or molecules derived via template-based amplification of molecules of natural origin (e.g., immunoglobulin chains cloned from populations of B cells via PCR amplification are not “synthetic” as used herein). In some instances, for example, when referring to libraries of the invention that comprise multiple segments (e.g., TN1, DH, N2, and/or H3-JH), the invention encompasses libraries in which at least one, two, three, or four of the aforementioned components is synthetic. By way of illustration, a library in which certain components are synthetic, while other components are of natural origin or derived via template-based amplification of molecules of natural origin, would be encompassed by the invention. Libraries that are fully synthetic would, of course, also be encompassed by the invention.

The term “theoretical diversity” refers to the maximum number of variants in a library design. For example, given an amino acid sequence of three residues, where residues one and three may each be any one of five amino acid types and residue two may be any one of 20 amino acid types, the theoretical diversity is 5×20×5=500 possible sequences. Similarly if sequence X is constructed by combination of 4 amino acid segments, where segment 1 has 100 possible sequences, segment 2 has 75 possible sequences, segment 3 has 250 possible sequences, and segment 4 has 30 possible sequences, the theoretical diversity of fragment X would be 100×75×200×30, or 5.6×10 5 possible sequences.

The term “theoretical segment pool” refers to a set of polynucleotide or polypeptide segments that can be used as building blocks to assemble a larger polynucleotide or polypeptide. For example, a theoretical segment pool containing TN1, DH, N2, and H3-JH segments can be used to assemble a library of CDRH3 sequences by concatenating them combinatorially to form a sequence represented by [TN1]-[DH]-[N2]-[H3-JH], and synthesizing the corresponding oligonucleotide(s). The term “theoretical segment pool” can apply to any set of polynucleotide or polypeptide segments. Thus, while a set of TN1, DH, N2, and H3-JH segments are collectively considered a theoretical segment pool, each of the individual sets of segments also comprise a theoretical segment pool, specifically a TN1 theoretical segment pool, a DH theoretical segment pool, an N2 theoretical segment pool, and an H3-JH theoretical segment pool. Any subsets of these theoretical segment pools containing two or more sequences can also be considered theoretical segment pools.

The term “unique,” as used herein, refers to a sequence that is different (e.g., has a different chemical structure) from every other sequence within the designed set (e.g., the theoretical diversity). It should be understood that there are likely to be more than one copy of many unique sequences from the theoretical diversity in a particular physical realization. For example, a library comprising three unique sequences at the theoretical level may comprise nine total members if each sequence occurs three times in the physical realization of the library. However, in certain embodiments, each unique sequence may occur only once, less than once, or more than once.

The term “variable” refers to the portions of the immunoglobulin domains that exhibit variability in their sequence and that are involved in determining the specificity and binding affinity of a particular antibody (i.e., the “variable domain(s)”). Variability is not evenly distributed throughout the variable domains of antibodies; it is concentrated in sub-domains of each of the heavy and light chain variable regions. These sub-domains are called “hypervariable” regions or “complementarity determining regions” (CDRs). The more conserved (i.e., non-hypervariable) portions of the variable domains are called the “framework” regions (FRM). The variable domains of naturally occurring heavy and light chains each comprise four FRM regions, largely adopting a β-sheet configuration, connected by three hypervariable regions, which form loops connecting, and in some cases forming part of, the β-sheet structure. The hypervariable regions in each chain are held together in close proximity by the FRM and, with the hypervariable regions from the other chain, contribute to the formation of the antigen-binding site (see Kabat et al. Sequences of Proteins of Immunological Interest, 5th Ed. Public Health Service, National Institutes of Health, Bethesda, Md., 1991, incorporated by reference in its entirety). The constant domains are not directly involved in antigen binding, but exhibit various effector functions, such as, for example, antibody-dependent, cell-mediated cytotoxicity and complement activation.

Libraries of the invention containing “VKCDR3” sequences and “VλCDR3” sequences refer to the kappa and lambda sub-sets of the light chain CDR3 (CDRL3) sequences, respectively. Such libraries may be designed with directed diversity, to collectively represent the length and sequence diversity of the human antibody CDRL3 repertoire. “Preimmune” versions of these libraries have similar sequence diversities and length diversities as naturally occurring human antibody CDRL3 sequences before these sequences undergo negative selection and/or somatic hypermutation. Known human CDRL3 sequences are represented in various data sets, including the NCBI database, WO/2009/036379, and Martin, Proteins, 1996, 25: 130 each incorporated by reference in its entirety.

General Design of Libraries

Antibody libraries provided by the present invention may be designed to reflect certain aspects of the preimmune repertoire as created by the human immune system. Certain libraries of the invention are based on rational design informed by collections of human V, D, and J genes, and large databases of human heavy and light chain sequences (e.g., publicly known germline sequences and sequences from Jackson et al., J. Immunol Methods, 2007, 324: 26; Lee et al., Immunogenetics, 2006, 57: 917; Boyd et al., Science Translational Medicine, 2009, 1: 1-8, each incorporated by reference in its entirety; and sequences compiled from rearranged VK and Vλ sequences (see WO/2009/036379, also incorporated by reference in its entirety). Additional information may be found, for example, in Scaviner et al., Exp. Clin. Immunogenet., 1999, 16: 234; Tomlinson et al., J. Mol. Biol., 1992, 227: 799; and Matsuda et al., J. Exp. Med., 1998, 188: 2151, each incorporated by reference in its entirety.

In certain embodiments of the invention, segments representing the possible V, D, and J diversity found in the human repertoire, as well as junctional diversity (i.e., TN1 and N2), are synthesized de novo as single or double-stranded DNA oligonucleotides. In certain embodiments of the invention, oligonucleotides encoding CDR sequences are introduced into yeast along with one or more acceptor vectors containing heavy or light chain chassis sequences and constant domains. No primer-based PCR amplification or template-directed cloning steps from mammalian cDNA or mRNA are employed. Through standard homologous recombination, the recipient yeast recombines the CDR segments with the acceptor vectors containing the chassis sequences and constant regions, to create a properly ordered synthetic, full-length human heavy chain and/or light chain immunoglobulin library that can be genetically propagated, expressed, presented, and screened. One of ordinary skill in the art will readily recognize that the acceptor vector can be designed so as to produce constructs other than full-length human heavy chains and/or light chains. For example, in certain embodiments of the invention, the chassis may be designed to encode portions of a polypeptide encoding an antibody fragment or subunit of an antibody fragment, so that a sequence encoding an antibody fragment, or subunit thereof, is produced when the oligonucleotide cassette containing the CDR is recombined with the acceptor vector.

Thus, in certain embodiments, the invention provides a synthetic, preimmune human antibody repertoire the repertoire comprising:

(a) one or more selected human antibody heavy chain chassis (i.e., amino acids 1 to 94 of the heavy chain variable region, using Kabat's definition);

(b) a CDRH3 repertoire (described more fully below), designed based on the human IGHD and IGHJ germline sequences, and the extraction of TN1 and N2 sequences from reference sets of human CDRH3 sequences, the CDRH3 repertoire comprising (i) a TN1 segment; (ii) a DH segment; (iii) an N2 segment; (iv) an H3-JH segment.

(c) one or more selected human antibody kappa and/or lambda light chain chassis; and

(d) a CDRL3 repertoire designed based on the human IGLV and IGLJ germline sequences, wherein “L” may be a kappa or lambda light chain.

The instant invention also provides methods for producing and using such libraries, as well as libraries comprising one or more immunoglobulin domains or antibody fragments. Design and synthesis of each component of the antibody libraries of the invention is provided in more detail below.

Design of Antibody Library Chassis Sequences

In certain embodiments, provided libraries are constructed from selected chassis sequences that are based on naturally occurring variable domain sequences (e.g., IGHV and IGLV genes). The selection of such chassis sequences can be done arbitrarily, or through the definition of certain pre-determined criteria. For example, the Kabat database, an electronic database containing non-redundant rearranged antibody sequences, can be queried for those heavy and light chain germline sequences that are most frequently represented. An algorithm such as BLAST, or a more specialized tool such as SoDA (Volpe et al., Bioinformatics, 2006, 22: 438-44, incorporated by reference in its entirety), can be used to compare rearranged antibody sequences with germline sequences (e.g., using the V BASE2 database; see, for example, Retter et al., Nucleic Acids Res., 2005, 33: D671-D674, incorporated by reference in its entirety), or similar collections of human V, D, and J genes, to identify germline families that are most frequently used to generate functional antibodies.

Several criteria can be utilized for the selection of chassis for inclusion in the libraries of the invention. For example, sequences that are known (or have been determined) to express poorly in yeast, or other organisms used in the invention (e.g., bacteria, mammalian cells, fungi, or plants) can be excluded from the libraries. Chassis may also be chosen based on the representation of their corresponding germline genes in the peripheral blood of humans. In certain embodiments of the invention, it may be desirable to select chassis that correspond to germline sequences that are highly represented in the peripheral blood of humans. In some embodiments, it may be desirable to select chassis that correspond to germline sequences that are less frequently represented, for example, to increase the canonical diversity of the library. Therefore, chassis may be selected to produce libraries that represent the largest and most structurally diverse group of functional human antibodies.

In certain embodiments of the invention, less diverse chassis may be utilized, for example, if it is desirable to produce a smaller, more focused library with less chassis variability and greater CDR variability. In some embodiments of the invention, chassis may be selected based on both their expression in a cell of the invention (e.g., a yeast cell) and the diversity of canonical structures represented by the selected sequences. One may therefore produce a library with a diversity of canonical structures that express well in a cell of the invention.

Design of Heavy Chain Chassis Sequences

The design and selection of heavy chain chassis sequences that can be used in the current invention is described in detail in U.S. Publication Nos. 2009/0181855 and 2010/0056386, and WO/2009/036379, each of which is incorporated by reference in its entirety, and is therefore described only briefly here.

In general, VH domains of the library comprise three components: (1) a VH “chassis,” which includes amino acids 1 to 94 (using Kabat numbering), (2) the CDRH3, which is defined herein to include the Kabat CDRH3 proper (positions 95-102), and (3) the FRM4 region, including amino acids 103 to 113 (Kabat numbering). The overall VH domain structure may therefore be depicted schematically (not to scale) as:

In certain embodiments of the invention, the VH chassis of the libraries may comprise from about Kabat residue 1 to about Kabat residue 94 of one or more of the following IGHV germline sequences: IGHV1-2, IGHV1-3, IGHV1-8, IGHV1-18, IGHV1-24, IGHV1-45, IGHV1-46, IGHV1-58, IGHV1-69, IGH8, IGH56, IGH100, IGHV3-7, IGHV3-9, IGHV3-11, IGHV3-13, IGHV3-15, IGHV3-20, IGHV3-21, IGHV3-23, IGHV3-30, IGHV3-33, IGHV3-43, IGHV3-48, IGHV3-49, IGHV3-53, IGHV3-64, IGHV3-66, IGHV3-72, IGHV3-73, IGHV3-74, IGHV4-4, IGHV4-28, IGHV4-31, IGHV4-34, IGHV4-39, IGHV4-59, IGHV4-61, IGHV4-B, IGHV5-51, IGHV6-1, and/or IGHV7-4-1. In some embodiments of the invention, a library may contain one or more of these sequences, one or more allelic variants of these sequences, or encode an amino acid sequence at least about 99.9%, 99.5%, 99%, 98.5%, 98%, 97.5%, 97%, 96.5%, 96%, 95.5%, 95%, 94.5%, 94%, 93.5%, 93%, 92.5%, 92%, 91.5%, 91%, 90.5%, 90%, 89%, 88%, 87%, 86%, 85%, 84%, 83%, 82%, 81%, 80%, 77.5%, 75%, 73.5%, 70%, 65%, 60%, 55%, or 50% identical to one or more of these sequences. One of ordinary skill in the art will recognize that given the chassis definition provided above, any IGHV-encoding sequence can be adapted for use as a chassis of the invention. As exemplified in U.S. Publication Nos. 2009/0181855 and 2010/0056386, and WO/2009/036379 (each incorporated by reference in its entirety), these chassis can also be varied, particularly by altering the amino acid residues in the CDRH1 and CDRH2 regions, further increasing the diversity of the library.

Design of Light Chain Chassis Sequences

The design and selection of light chain chassis sequences that can be used in the current invention is described in detail in U.S. Publication Nos. 2009/0181855 and 2010/0056386, and WO/2009/036379, each of which is incorporated by reference in its entirety, and is therefore described only briefly here. The light chain chassis of the invention may be based on kappa and/or lambda light chain sequences.

The VL domains of the library comprise three primary components: (1) a VL “chassis”, which includes amino acids 1 to 88 (using Kabat numbering), (2) the CDRL3, which is defined herein to include the Kabat CDRL3 proper (positions 89-97), and (3) the FRM4 region, including amino acids 98 to 107 (Kabat numbering). The overall VL domain structure may therefore be depicted schematically (not to scale) as:

In certain embodiments of the invention, the VL chassis of the libraries include one or more chassis based on IGKV germline sequences. In certain embodiments of the invention, the VL chassis of the libraries may comprise from about Kabat residue 1 to about Kabat residue 88 of one or more of the following IGKV germline sequences: IGKV1-05, IGKV1-06, IGKV1-08, IGKV1-09, IGKV1-12, IGKV1-13, IGKV1-16, IGKV1-17, IGKV1-27, IGKV1-33, IGKV1-37, IGKV1-39, IGKV1D-16, IGKV1D-17, IGKV1D-43, IGKV1D-8, IGK54, IGK58, IGK59, IGK60, IGK70, IGKV2D-26, IGKV2D-29, IGKV2D-30, IGKV3-11, IGKV3-15, IGKV3-20, IGKV3D-07, IGKV3D-11, IGKV3D-20, IGKV4-1, IGKV5-2, IGKV6-21, and/or IGKV6D-41. In some embodiments of the invention, a library may contain one or more of these sequences, one or more allelic variants of these sequences, or encode an amino acid sequence at least about 99.9%, 99.5%, 99%, 98.5%, 98%, 97.5%, 97%, 96.5%, 96%, 95.5%, 95%, 94.5%, 94%, 93.5%, 93%, 92.5%, 92%, 91.5%, 91%, 90.5%, 90%, 89%, 88%, 87%, 86%, 85%, 84%, 83%, 82%, 81%, 80%, 77.5%, 75%, 73.5%, 70%, 65%, 60%, 55%, or 50% identical to one or more of these sequences.

In certain embodiments of the invention, the VL chassis of the libraries include one or more chassis based on IGλV germline sequences. In certain embodiments of the invention, the VL chassis of the libraries may comprise from about Kabat residue 1 to about Kabat residue 88 of one or more of the following IGλV germline sequences: IGλV3-1, IGλV3-21, IGλ44, IGλV1-40, IGλV3-19, IGλV1-51, IGλV1-44, IGλV6-57, IGλ11, IGλV3-25, IGλ53, IGλV3-10, IGλV4-69, IGλV1-47, IGλ41, IGλV7-43, IGλV7-46, IGλV5-45, IGλV4-60, IGλV10-54, IGλV8-61, IGλV3-9, IGλV1-36, IGλ48, IGλV3-16, IGλV3-27, IGλV4-3, IGλV5-39, IGλV9-49, and/or IGλV3-12. In some embodiments of the invention, a library may contain one or more of these sequences, one or more allelic variants of these sequences, or encode an amino acid sequence at least about 99.9%, 99.5%, 99%, 98%, 97%, 96%, 95%, 94%, 93%, 92%, 91%, 90%, 85%, 80%, 75%, 70%, 65%, 60%, 55%, or 50% identical to one or more of these sequences.

One of ordinary skill in the art will recognize that given the chassis definition provided above, any IGKV- or IGλV-encoding sequence can be adapted for use as a chassis of the invention.

Design and Selection of TN1, DH, N2, and H3-JH Segments

The human germline repertoire contains at least six IGHJ genes (IGHJ1, IGHJ2, IGHJ3, IGHJ4, IGHJ5, and IGHJ6; included in Table 14, where the primary allele is designated “01,” and selected allelic variants are designated “02” or “03”), and at least 27 IGHD genes (Table 16, including allelic variants). In some embodiments, the invention comprises a library of CDRH3 polypeptide sequences, or polynucleotide sequences encoding CDRH3 sequences, the library comprising members of any of the theoretical segment pools disclosed herein.

A person of ordinary skill in the art will recognize that not every segment in a theoretical segment pool provided herein is necessary to produce a functional CDRH3 library of the invention. Therefore, in certain embodiments, a CDRH3 library of the invention will contain a subset of the segments of any of the theoretical segment pools described herein. For example, in certain embodiments of the invention, at least about 15, 30, 45, 60, 75, 90, 100, 105, 120, 135, 150, 165, 180, 195, 200, 210, 225, 240, 255, 270, 285, 300, 320, 340, 360, 380, 400, 420, 440, 460, 480, 500, 520, 540, 560, 580, 600, 620, 640, or 643 of the H3-JH segments of any of the theoretical segment pools provided herein, or generated by the methods described herein, are included in a library. In some embodiments of the invention, at least about 15, 30, 45, 60, 75, 90, 100, 105, 120, 135, 150, 165, 180, 195, 200, 250, 300, 350, 400, 450, 500, 550, 600, 650, 700, 750, 800, 850, 900, 950, 1000, 1050, 1100, 1111, 2000, 3000, 4000, 5000, 6000, 7000, 14000, 21000, 28000, 35000, 42000, 49000, 56000, 63000, or 68374 of the DH segments of any of the theoretical segment pools provided herein, or generated by the methods described herein, are included in a library. In some embodiments of the invention, at least about 10, 20, 30, 40, 50, 60, 70, 80, 90, 100, 110, 120, 130, 140, 141, 150, 160, 170, 180, 190, or 200, 220, 240, 260, 280, 300, 320, 340, 360, 380, 400, 420, 424, 440, 460, 480, 500, 550, 600, 650, 700, 727, 750, 800, 850, 900, 950, or 1000 of the TN1 and/or N2 segments of any of the theoretical segment pools provided herein, or generated by the methods described herein, are included in a library. In certain embodiments, a library of the invention may contain less than a particular number of polynucleotide or polypeptide segments, where the number of segments is defined using any one of the integers provided above for the respective segment. In some embodiments of the invention, a particular numerical range is defined, using any two of the integers provided above as lower and upper boundaries of the range, inclusive or exclusive. All combinations of the integers provided, which define an upper and lower boundary, are contemplated.

In certain embodiments, the invention provides CDRH3 libraries comprising at least about 1%, 2.5%, 5%, 10%, 15%, 20%, 25%, 30%, 40%, 45%, 50%, 55%, 60%, 65%, 70%, 75%, 80%, 85%, 90%, 95%, or 99% of the segments from any of the theoretical segment pools provided herein. For example, the invention provides libraries comprising at least about 1%, 2.5%, 5%, 10%, 15%, 20%, 25%, 30%, 40%, 45%, 50%, 55%, 60%, 65%, 70%, 75%, 80%, 85%, 90%, 95%, or 99% of the TN1, DH, N2, and/or H3-JH segments from any of the theoretical segment pools provided herein. In some embodiments of the invention, a particular percentage range is defined, using any two of the percentages provided above as lower and upper boundaries of the range, inclusive or exclusive. All combinations of the percentages provided, which define an upper and lower boundary, are contemplated.

In some embodiments of the invention, at least about 1%, 2.5%, 5%, 10%, 15%, 20%, 25%, 30%, 40%, 45%, 50%, 55%, 60%, 65%, 70%, 75%, 80%, 85%, 90%, 95%, or 99% of the H3-JH, DH, TN1, and/or N2 segments in a CDRH3 library are H3-JH, DH, TN1, and/or N2 segments of any of the theoretical segment pools provided herein, or generated by the methods described herein. In some embodiments of the invention, at least about 1%, 2.5%, 5%, 10%, 15%, 20%, 25%, 30%, 40%, 45%, 50%, 55%, 60%, 65%, 70%, 75%, 80%, 85%, 90%, 95%, or 99% of the H3-JH, DH, TN1, and/or N2 segments of antibodies isolated from a CDRH3 library (e.g., by binding to a particular antigen and/or generic ligand through one or more rounds of selection) are H3-JH, DH, TN1, and/or N2 segments of any of the theoretical segment pools provided herein, or generated by methods described herein. In certain embodiments, a CDRH3 library of the invention may contain less than a particular percentage of H3-JH, DH, TN1, and/or N2 segments of any of the theoretical segment pools provided herein, or generated by the methods described herein, where the number of segments is defined using any one of the percentages provided above for the respective segment. In some embodiments of the invention, a particular percentage range is defined, using any two of the percentages provided above as lower and upper boundaries of the range, inclusive or exclusive. All combinations of the percentages provided, which define an upper and lower boundary, are contemplated.

One of ordinary skill in the art will appreciate, upon reading the disclosure herein, that given the TN1, DH, N2, and/or H3-JH segments of any of the theoretical segment pools provided herein, or generated by the methods described herein, similar TN1, DH, N2, and/or H3-JH segments, and corresponding CDRH3 libraries, could be produced which, while not 100% identical to those provided in terms of their sequences, may be functionally very similar. Such theoretical segment pools and CDRH3 libraries also fall within the scope of the invention. A variety of techniques well-known in the art could be used to obtain these additional sequences, including the mutagenesis techniques provided herein. Therefore, each of the explicitly enumerated embodiments of the invention can also be practiced using segments that share a particular percent identity to any of the segments of any of the theoretical segment pools provided herein, or generated by the methods described herein. For example, each of the previously described embodiments of the invention can be practiced using TN1, DH, N2, and/or H3-JH segments that are at least about 50%, 55%, 60%, 65%, 70%, 75%, 80%, 85%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98%, 99%, 99.5%, or 99.9% identical to the TN1, DH, N2, and/or H3-JH segments of any of the theoretical segment pools provided herein, or generated by the methods described herein.

In some embodiments, the invention provides libraries produced from one or more VH chassis sequences combined with one or more TN1 segments, one or more DH segments, one or more N2 segments, and one or more H3-JH segments. In certain embodiments at least 1, 2, 5, 10, 20, 50, 75, or 100, of each chassis, TN1, DH, N2, or H3-JH segments are included in a library of the invention.

In some embodiments, the invention provides a method of selecting TN1, DH, N2, and H3-JH segments from a theoretical segment pool for inclusion in a synthetic CDRH3 library, comprising:

(i) providing a theoretical segment pool containing one or more of TN1, DH, N2, and H3-JH segments; (ii) providing a reference set of CDRH3 sequences; (iii) utilizing the theoretical segment pool of (i) to identify the closest match(es) to each CDRH3 sequence in the reference set of (ii); and (iv) selecting segments from the theoretical segment pool for inclusion in a synthetic library.

In some embodiments, the selection process of (iv) can optionally involve any number of additional criteria, including the frequency of occurrence of the segments of (i) in the reference set of (ii); the corresponding segmental usage weights; and any physicochemical properties (see all numerical indices on the world wide web at genome.jp/aaindex/) of the segments (e.g., hydrophobicity, alpha-helical propensity, and/or isoelectric point). Optionally, TN1 and/or N2 segments that do not occur in the theoretical segment pool of (i) but that are found in the reference set of (ii) may be identified and added to prospective theoretical segment pools to produce theoretical segment pools with increased TN1 and/or N2 diversity in the prospective theoretical segment pools and/or synthetic libraries of the invention.

Any characteristic or set of characteristics of the segments can be used to choose them for inclusion in the library, including for example one or more biological properties (e.g., immunogenicity, stability, half-life) and/or one or more physicochemical properties such as the numerical indices provided above. In some embodiments, at least 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15 or more such properties is used to select segments for inclusion in a library of the invention. Physiochemical properties included in the index provided above can include, for example, ANDN920101 alpha-CH chemical shifts (Andersen et al., 1992); ARGP820101 Hydrophobicity index (Argos et al., 1982); ARGP820102 Signal sequence helical potential (Argos et al., 1982); ARGP820103 Membrane-buried preference parameters (Argos et al., 1982); BEGF750101 Conformational parameter of inner helix (Beghin-Dirkx, 1975); BEGF750102 Conformational parameter of beta-structure (Beghin-Dirkx, 1975); BEGF750103 Conformational parameter of beta-turn (Beghin-Dirkx, 1975); BHAR880101 Average flexibility indices (Bhaskaran-Ponnuswamy, 1988); BIGC670101 Residue volume (Bigelow, 1967); BIOV880101 Information value for accessibility; average fraction 35% (Biou et al., 1988); BIOV880102 Information value for accessibility; average fraction 23% (Biou et al., 1988); BROC820101 Retention coefficient in TFA (Browne et al., 1982); BROC820102 Retention coefficient in HFBA (Browne et al., 1982); BULH740101 Transfer free energy to surface (Bull-Breese, 1974); BULH740102 Apparent partial specific volume (Bull-Breese, 1974); BUNA790101 alpha-NH chemical shifts (Bundi-Wuthrich, 1979); BUNA790102 alpha-CH chemical shifts (Bundi-Wuthrich, 1979); BUNA790103 Spin-spin coupling constants 3JHalpha-NH (Bundi-Wuthrich, 1979); BURA740101 Normalized frequency of alpha-helix (Burgess et al., 1974); BURA740102 Normalized frequency of extended structure (Burgess et al., 1974); CHAM810101 Steric parameter (Charton, 1981); CHAM820101 Polarizability parameter (Charton-Charton, 1982); CHAM820102 Free energy of solution in water, kcal/mole (Charton-Charton, 1982); CHAM830101 The Chou-Fasman parameter of the coil conformation (Charton-Charton, 1983); CHAM830102 A parameter defined from the residuals obtained from the best correlation of the Chou-Fasman parameter of beta-sheet (Charton-Charton, 1983); CHAM830103 The number of atoms in the side chain labelled 1+1 (Charton-Charton, 1983); CHAM830104 The number of atoms in the side chain labelled 2+1 (Charton-Charton, 1983); CHAM830105 The number of atoms in the side chain labelled 3+1 (Charton-Charton, 1983); CHAM830106 The number of bonds in the longest chain (Charton-Charton, 1983); CHAM830107 A parameter of charge transfer capability (Charton-Charton, 1983); CHAM830108 A parameter of charge transfer donor capability (Charton-Charton, 1983); CHOC750101 Average volume of buried residue (Chothia, 1975); CHOC760101 Residue accessible surface area in tripeptide (Chothia, 1976); CHOC760102 Residue accessible surface area in folded protein (Chothia, 1976); CHOC760103 Proportion of residues 95% buried (Chothia, 1976); CHOC760104 Proportion of residues 100% buried (Chothia, 1976); CHOP780101 Normalized frequency of beta-turn (Chou-Fasman, 1978a); CHOP780201 Normalized frequency of alpha-helix (Chou-Fasman, 1978b); CHOP780202 Normalized frequency of beta-sheet (Chou-Fasman, 1978b); CHOP780203 Normalized frequency of beta-turn (Chou-Fasman, 1978b); CHOP780204 Normalized frequency of N-terminal helix (Chou-Fasman, 1978b); CHOP780205 Normalized frequency of C-terminal helix (Chou-Fasman, 1978b); CHOP780206 Normalized frequency of N-terminal non helical region (Chou-Fasman, 1978b); CHOP780207 Normalized frequency of C-terminal non helical region (Chou-Fasman, 1978b); CHOP780208 Normalized frequency of N-terminal beta-sheet (Chou-Fasman, 1978b); CHOP780209 Normalized frequency of C-terminal beta-sheet (Chou-Fasman, 1978b); CHOP780210 Normalized frequency of N-terminal non beta region (Chou-Fasman, 1978b); CHOP780211 Normalized frequency of C-terminal non beta region (Chou-Fasman, 1978b); CHOP780212 Frequency of the 1st residue in turn (Chou-Fasman, 1978b); CHOP780213 Frequency of the 2nd residue in turn (Chou-Fasman, 1978b); CHOP780214 Frequency of the 3rd residue in turn (Chou-Fasman, 1978b); CHOP780215 Frequency of the 4th residue in turn (Chou-Fasman, 1978b); CHOP780216 Normalized frequency of the 2nd and 3rd residues in turn (Chou-Fasman, 1978b); CIDH920101 Normalized hydrophobicity scales for alpha-proteins (Cid et al., 1992); CIDH920102 Normalized hydrophobicity scales for beta-proteins (Cid et al., 1992); CIDH920103 Normalized hydrophobicity scales for alpha+beta-proteins (Cid et al., 1992); CIDH920104 Normalized hydrophobicity scales for alpha/beta-proteins (Cid et al., 1992); CIDH920105 Normalized average hydrophobicity scales (Cid et al., 1992); COHE430101 Partial specific volume (Cohn-Edsall, 1943); CRAJ730101 Normalized frequency of middle helix (Crawford et al., 1973); CRAJ730102 Normalized frequency of beta-sheet (Crawford et al., 1973); CRAJ730103 Normalized frequency of turn (Crawford et al., 1973); DAWD720101 Size (Dawson, 1972); DAYM780101 Amino acid composition (Dayhoff et al., 1978a); DAYM780201 Relative mutability (Dayhoff et al., 1978b); DESM900101 Membrane preference for cytochrome b: MPH89 (Degli Esposti et al., 1990); DESM900102 Average membrane preference: AMP07 (Degli Esposti et al., 1990); EISD840101 Consensus normalized hydrophobicity scale (Eisenberg, 1984); EISD860101 Solvation free energy (Eisenberg-McLachlan, 1986); EISD860102 Atom-based hydrophobic moment (Eisenberg-McLachlan, 1986); EISD860103 Direction of hydrophobic moment (Eisenberg-McLachlan, 1986); FASG760101 Molecular weight (Fasman, 1976); FASG760102 Melting point (Fasman, 1976); FASG760103 Optical rotation (Fasman, 1976); FASG760104 pK-N (Fasman, 1976); FASG760105 pK-C (Fasman, 1976); FAUJ830101 Hydrophobic parameter pi (Fauchere-Pliska, 1983); FAUJ880101 Graph shape index (Fauchere et al., 1988); FAUJ880102 Smoothed upsilon steric parameter (Fauchere et al., 1988); FAUJ880103 Normalized van der Waals volume (Fauchere et al., 1988); FAUJ880104 STERIMOL length of the side chain (Fauchere et al., 1988); FAUJ880105 STERIMOL minimum width of the side chain (Fauchere et al., 1988); FAUJ880106 STERIMOL maximum width of the side chain (Fauchere et al., 1988); FAUJ880107 N.m.r. chemical shift of alpha-carbon (Fauchere et al., 1988); FAUJ880108 Localized electrical effect (Fauchere et al., 1988); FAUJ880109 Number of hydrogen bond donors (Fauchere et al., 1988); FAUJ880110 Number of full nonbonding orbitals (Fauchere et al., 1988); FAUJ880111 Positive charge (Fauchere et al., 1988); FAUJ880112 Negative charge (Fauchere et al., 1988); FAUJ880113 pK-a(RCOOH) (Fauchere et al., 1988); FINA770101 Helix-coil equilibrium constant (Finkelstein-Ptitsyn, 1977); FINA910101 Helix initiation parameter at posision i−1 (Finkelstein et al., 1991); FINA910102 Helix initiation parameter at posision 0+0+2 (Finkelstein et al., 1991); FINA910103 Helix termination parameter at posision j−2,j−1,j (Finkelstein et al., 1991); FINA910104 Helix termination parameter at posision j+1 (Finkelstein et al., 1991); GARJ730101 Partition coefficient (Garel et al., 1973); GEIM800101 Alpha-helix indices (Geisow-Roberts, 1980); GEIM800102 Alpha-helix indices for alpha-proteins (Geisow-Roberts, 1980); GEIM800103 Alpha-helix indices for beta-proteins (Geisow-Roberts, 1980); GEIM800104 Alpha-helix indices for alpha/beta-proteins (Geisow-Roberts, 1980); GEIM800105 Beta-strand indices (Geisow-Roberts, 1980); GEIM800106 Beta-strand indices for beta-proteins (Geisow-Roberts, 1980); GEIM800107 Beta-strand indices for alpha/beta-proteins (Geisow-Roberts, 1980) GEIM800108 Aperiodic indices (Geisow-Roberts, 1980); GEI M800109 Aperiodic indices for alpha-proteins (Geisow-Roberts, 1980); GEIM800110 Aperiodic indices for beta-proteins (Geisow-Roberts, 1980); GEIM800111 Aperiodic indices for alpha/beta-proteins (Geisow-Roberts, 1980); GOLD730101 Hydrophobicity factor (Goldsack-Chalifoux, 1973); GOLD730102 Residue volume (Goldsack-Chalifoux, 1973); GRAR740101 Composition (Grantham, 1974); GRAR740102 Polarity (Grantham, 1974) GRAR740103 Volume (Grantham, 1974); GUYH850101 Partition energy (Guy, 1985); HOPA770101 Hydration number (Hopfinger, 1971), Cited by Charton-Charton (1982)

HOPT810101 Hydrophilicity value (Hopp-Woods, 1981); HUTJ700101 Heat capacity (Hutchens, 1970); HUTJ700102 Absolute entropy (Hutchens, 1970); HUTJ700103 Entropy of formation (Hutchens, 1970); ISOY800101 Normalized relative frequency of alpha-helix (Isogai et al., 1980); ISOY800102 Normalized relative frequency of extended structure (Isogai et al., 1980); ISOY800103 Normalized relative frequency of bend (Isogai et al., 1980); ISOY800104 Normalized relative frequency of bend R (Isogai et al., 1980); ISOY800105 Normalized relative frequency of bend S (Isogai et al., 1980); ISOY800106 Normalized relative frequency of helix end (Isogai et al., 1980); ISOY800107 Normalized relative frequency of double bend (Isogai et al., 1980); ISOY800108 Normalized relative frequency of coil (Isogai et al., 1980); JANJ780101 Average accessible surface area (Janin et al., 1978); JANJ780102 Percentage of buried residues (Janin et al., 1978); JANJ780103 Percentage of exposed residues (Janin et al., 1978); JANJ790101 Ratio of buried and accessible molar fractions (Janin, 1979); JANJ790102 Transfer free energy (Janin, 1979); JOND750101 Hydrophobicity (Jones, 1975); JOND750102 pK (—COOH) (Jones, 1975); JOND920101 Relative frequency of occurrence (Jones et al., 1992); JOND920102 Relative mutability (Jones et al., 1992)

JUKT750101 Amino acid distribution (Jukes et al., 1975); JUNJ780101 Sequence frequency (Jungck, 1978); KANM800101 Average relative probability of helix (Kanehisa-Tsong, 1980); KANM800102 Average relative probability of beta-sheet (Kanehisa-Tsong, 1980); KANM800103 Average relative probability of inner helix (Kanehisa-Tsong, 1980); KANM800104 Average relative probability of inner beta-sheet (Kanehisa-Tsong, 1980); KARP850101 Flexibility parameter for no rigid neighbors (Karplus-Schulz, 1985); KARP850102 Flexibility parameter for one rigid neighbor (Karplus-Schulz, 1985); KARP850103 Flexibility parameter for two rigid neighbors (Karplus-Schulz, 1985); KHAG800101 The Kerr-constant increments (Khanarian-Moore, 1980); KLEP840101 Net charge (Klein et al., 1984); KRIW710101 Side chain interaction parameter (Krigbaum-Rubin, 1971); KRIW790101 Side chain interaction parameter (Krigbaum-Komoriya, 1979); KRIW790102 Fraction of site occupied by water (Krigbaum-Komoriya, 1979); KRIW790103 Side chain volume (Krigbaum-Komoriya, 1979); KYTJ820101 Hydropathy index (Kyte-Doolittle, 1982); LAWE840101 Transfer free energy, CHP/water (Lawson et al., 1984); LEVM760101 Hydrophobic parameter (Levitt, 1976); LEVM760102 Distance between C-alpha and centroid of side chain (Levitt, 1976); LEVM760103 Side chain angle theta(AAR) (Levitt, 1976); LEVM760104 Side chain torsion angle phi(AAAR) (Levitt, 1976); LEVM760105 Radius of gyration of side chain (Levitt, 1976); LEVM760106 van der Waals parameter RO (Levitt, 1976)

LEVM760107 van der Waals parameter epsilon (Levitt, 1976); LEVM780101 Normalized frequency of alpha-helix, with weights (Levitt, 1978); LEVM780102 Normalized frequency of beta-sheet, with weights (Levitt, 1978); LEVM780103 Normalized frequency of reverse turn, with weights (Levitt, 1978); LEVM780104 Normalized frequency of alpha-helix, unweighted (Levitt, 1978); LEVM780105 Normalized frequency of beta-sheet, unweighted (Levitt, 1978); LEVM780106 Normalized frequency of reverse turn, unweighted (Levitt, 1978); LEWP710101 Frequency of occurrence in beta-bends (Lewis et al., 1971); LIFS790101 Conformational preference for all beta-strands (Lifson-Sander, 1979); LIFS790102 Conformational preference for parallel beta-strands (Lifson-Sander, 1979); LIFS790103 Conformational preference for antiparallel beta-strands (Lifson-Sander, 1979); MANP780101 Average surrounding hydrophobicity (Manavalan-Ponnuswamy, 1978); MAXF760101 Normalized frequency of alpha-helix (Maxfield-Scheraga, 1976); MAXF760102 Normalized frequency of extended structure (Maxfield-Scheraga, 1976); MAXF760103 Normalized frequency of zeta R (Maxfield-Scheraga, 1976); MAXF760104 Normalized frequency of left-handed alpha-helix (Maxfield-Scheraga, 1976); MAXF760105 Normalized frequency of zeta L (Maxfield-Scheraga, 1976); MAXF760106 Normalized frequency of alpha region (Maxfield-Scheraga, 1976); MCMT640101 Refractivity (McMeekin et al., 1964), Cited by Jones (1975); MEEJ800101 Retention coefficient in HPLC, pH7.4 (Meek, 1980); MEEJ800102 Retention coefficient in HPLC, pH2.1 (Meek, 1980); MEEJ810101 Retention coefficient in NaClO4 (Meek-Rossetti, 1981); MEEJ810102 Retention coefficient in NaH2PO4 (Meek-Rossetti, 1981); MEIH800101 Average reduced distance for C-alpha (Meirovitch et al., 1980); MEIH800102 Average reduced distance for side chain (Meirovitch et al., 1980); MEIH800103 Average side chain orientation angle (Meirovitch et al., 1980); MIYS850101 Effective partition energy (Miyazawa-Jernigan, 1985); NAGK730101 Normalized frequency of alpha-helix (Nagano, 1973); NAGK730102 Normalized frequency of bata-structure (Nagano, 1973)

NAGK730103 Normalized frequency of coil (Nagano, 1973); NAKH900101 AA composition of total proteins (Nakashima et al., 1990); NAKH900102 SD of AA composition of total proteins (Nakashima et al., 1990); NAKH900103 AA composition of mt-proteins (Nakashima et al., 1990); NAKH900104 Normalized composition of mt-proteins (Nakashima et al., 1990); NAKH900105 AA composition of mt-proteins from animal (Nakashima et al., 1990); NAKH900106 Normalized composition from animal (Nakashima et al., 1990); NAKH900107 AA composition of mt-proteins from fungi and plant (Nakashima et al., 1990); NAKH900108 Normalized composition from fungi and plant (Nakashima et al., 1990); NAKH900109 AA composition of membrane proteins (Nakashima et al., 1990); NAKH900110 Normalized composition of membrane proteins (Nakashima et al., 1990); NAKH900111 Transmembrane regions of non-mt-proteins (Nakashima et al., 1990); NAKH900112 Transmembrane regions of mt-proteins (Nakashima et al., 1990); NAKH900113 Ratio of average and computed composition (Nakashima et al., 1990); NAKH920101 AA composition of CYT of single-spanning proteins (Nakashima-Nishikawa, 1992); NAKH920102 AA composition of CYT2 of single-spanning proteins (Nakashima-Nishikawa, 1992); NAKH920103 AA composition of EXT of single-spanning proteins (Nakashima-Nishikawa, 1992); NAKH920104 AA composition of EXT2 of single-spanning proteins (Nakashima-Nishikawa, 1992); NAKH920105 AA composition of MEM of single-spanning proteins (Nakashima-Nishikawa, 1992); NAKH920106 AA composition of CYT of multi-spanning proteins (Nakashima-Nishikawa, 1992); NAKH920107 AA composition of EXT of multi-spanning proteins (Nakashima-Nishikawa, 1992); NAKH920108 AA composition of MEM of multi-spanning proteins (Nakashima-Nishikawa, 1992); NISK800101 8 A contact number (Nishikawa-Ooi, 1980); NISK860101 14 A contact number (Nishikawa-Ooi, 1986); NOZY710101 Transfer energy, organic solvent/water (Nozaki-Tanford, 1971); OOBM770101 Average non-bonded energy per atom (Oobatake-Ooi, 1977); OOBM770102 Short and medium range non-bonded energy per atom (Oobatake-Ooi, 1977); OOBM770103 Long range non-bonded energy per atom (Oobatake-Ooi, 1977)

OOBM770104 Average non-bonded energy per residue (Oobatake-Ooi, 1977); OOBM770105 Short and medium range non-bonded energy per residue (Oobatake-Ooi, 1977); OOBM850101 Optimized beta-structure-coil equilibrium constant (Oobatake et al., 1985); OOBM850102 Optimized propensity to form reverse turn (Oobatake et al., 1985); OOBM850103 Optimized transfer energy parameter (Oobatake et al., 1985); OOBM850104 Optimized average non-bonded energy per atom (Oobatake et al., 1985); OOBM850105 Optimized side chain interaction parameter (Oobatake et al., 1985); PALJ810101 Normalized frequency of alpha-helix from LG (Palau et al., 1981); PALJ810102 Normalized frequency of alpha-helix from CF (Palau et al., 1981); PALJ810103 Normalized frequency of beta-sheet from LG (Palau et al., 1981); PALJ810104 Normalized frequency of beta-sheet from CF (Palau et al., 1981); PALJ810105 Normalized frequency of turn from LG (Palau et al., 1981); PALJ810106 Normalized frequency of turn from CF (Palau et al., 1981); PALJ810107 Normalized frequency of alpha-helix in all-alpha class (Palau et al., 1981); PALJ810108 Normalized frequency of alpha-helix in alpha+beta class (Palau et al., 1981); PALJ810109 Normalized frequency of alpha-helix in alpha/beta class (Palau et al., 1981); PALJ810110 Normalized frequency of beta-sheet in all-beta class (Palau et al., 1981); PALJ810111 Normalized frequency of beta-sheet in alpha+beta class (Palau et al., 1981); PALJ810112 Normalized frequency of beta-sheet in alpha/beta class (Palau et al., 1981); PALJ810113 Normalized frequency of turn in all-alpha class (Palau et al., 1981); PALJ810114 Normalized frequency of turn in all-beta class (Palau et al., 1981); PALJ810115 Normalized frequency of turn in alpha+beta class (Palau et al., 1981); PALJ810116 Normalized frequency of turn in alpha/beta class (Palau et al., 1981); PARJ860101 HPLC parameter (Parker et al., 1986); PLIV810101 Partition coefficient (Pliska et al., 1981); PONP800101 Surrounding hydrophobicity in folded form (Ponnuswamy et al., 1980); PONP800102 Average gain in surrounding hydrophobicity (Ponnuswamy et al., 1980); PONP800103 Average gain ratio in surrounding hydrophobicity (Ponnuswamy et al., 1980); PONP800104 Surrounding hydrophobicity in alpha-helix (Ponnuswamy et al., 1980); PONP800105 Surrounding hydrophobicity in beta-sheet (Ponnuswamy et al., 1980); PONP800106 Surrounding hydrophobicity in turn (Ponnuswamy et al., 1980); PONP800107 Accessibility reduction ratio (Ponnuswamy et al., 1980); PONP800108 Average number of surrounding residues (Ponnuswamy et al., 1980); PRAM820101 Intercept in regression analysis (Prabhakaran-Ponnuswamy, 1982); PRAM820102 Slope in regression analysis×1.0E1 (Prabhakaran-Ponnuswamy, 1982); PRAM820103 Correlation coefficient in regression analysis (Prabhakaran-Ponnuswamy, 1982); PRAM900101 Hydrophobicity (Prabhakaran, 1990); PRAM900102 Relative frequency in alpha-helix (Prabhakaran, 1990); PRAM900103 Relative frequency in beta-sheet (Prabhakaran, 1990); PRAM900104 Relative frequency in reverse-turn (Prabhakaran, 1990); PTI0830101 Helix-coil equilibrium constant (Ptitsyn-Finkelstein, 1983); PTI0830102 Beta-coil equilibrium constant (Ptitsyn-Finkelstein, 1983); QIAN880101 Weights for alpha-helix at the window position of −6 (Qian-Sejnowski, 1988); QIAN880102 Weights for alpha-helix at the window position of −5 (Qian-Sejnowski, 1988); QIAN880103 Weights for alpha-helix at the window position of −4 (Qian-Sejnowski, 1988); QIAN880104 Weights for alpha-helix at the window position of −3 (Qian-Sejnowski, 1988); QIAN880105 Weights for alpha-helix at the window position of −2 (Qian-Sejnowski, 1988); QIAN880106 Weights for alpha-helix at the window position of −1 (Qian-Sejnowski, 1988); QIAN880107 Weights for alpha-helix at the window position of 0 (Qian-Sejnowski, 1988); QIAN880108 Weights for alpha-helix at the window position of 1 (Qian-Sejnowski, 1988); QIAN880109 Weights for alpha-helix at the window position of 2 (Qian-Sejnowski, 1988); QIAN880110 Weights for alpha-helix at the window position of 3 (Qian-Sejnowski, 1988); QIAN880111 Weights for alpha-helix at the window position of 4 (Qian-Sejnowski, 1988); QIAN880112 Weights for alpha-helix at the window position of 5 (Qian-Sejnowski, 1988); QIAN880113 Weights for alpha-helix at the window position of 6 (Qian-Sejnowski, 1988); QIAN880114 Weights for beta-sheet at the window position of −6 (Qian-Sejnowski, 1988); QIAN880115 Weights for beta-sheet at the window position of −5 (Qian-Sejnowski, 1988); QIAN880116 Weights for beta-sheet at the window position of −4 (Qian-Sejnowski, 1988); QIAN880117 Weights for beta-sheet at the window position of −3 (Qian-Sejnowski, 1988); QIAN880118 Weights for beta-sheet at the window position of −2 (Qian-Sejnowski, 1988); QIAN880119 Weights for beta-sheet at the window position of −1 (Qian-Sejnowski, 1988); QIAN880120 Weights for beta-sheet at the window position of 0 (Qian-Sejnowski, 1988); QIAN880121 Weights for beta-sheet at the window position of 1 (Qian-Sejnowski, 1988); QIAN880122 Weights for beta-sheet at the window position of 2 (Qian-Sejnowski, 1988); QIAN880123 Weights for beta-sheet at the window position of 3 (Qian-Sejnowski, 1988); QIAN880124 Weights for beta-sheet at the window position of 4 (Qian-Sejnowski, 1988); QIAN880125 Weights for beta-sheet at the window position of 5 (Qian-Sejnowski, 1988); QIAN880126 Weights for beta-sheet at the window position of 6 (Qian-Sejnowski, 1988); QIAN880127 Weights for coil at the window position of −6 (Qian-Sejnowski, 1988); QIAN880128 Weights for coil at the window position of −5 (Qian-Sejnowski, 1988); QIAN880129 Weights for coil at the window position of −4 (Qian-Sejnowski, 1988); QIAN880130 Weights for coil at the window position of −3 (Qian-Sejnowski, 1988); QIAN880131 Weights for coil at the window position of −2 (Qian-Sejnowski, 1988); QIAN880132 Weights for coil at the window position of −1 (Qian-Sejnowski, 1988); QIAN880133 Weights for coil at the window position of 0 (Qian-Sejnowski, 1988); QIAN880134 Weights for coil at the window position of 1 (Qian-Sejnowski, 1988); QIAN880135 Weights for coil at the window position of 2 (Qian-Sejnowski, 1988); QIAN880136 Weights for coil at the window position of 3 (Qian-Sejnowski, 1988); QIAN880137 Weights for coil at the window position of 4 (Qian-Sejnowski, 1988); QIAN880138 Weights for coil at the window position of 5 (Qian-Sejnowski, 1988); QIAN880139 Weights for coil at the window position of 6 (Qian-Sejnowski, 1988); RACS770101 Average reduced distance for C-alpha (Rackovsky-Scheraga, 1977); RACS770102 Average reduced distance for side chain (Rackovsky-Scheraga, 1977); RACS770103 Side chain orientational preference (Rackovsky-Scheraga, 1977); RACS820101 Average relative fractional occurrence in A0(i) (Rackovsky-Scheraga, 1982); RACS820102 Average relative fractional occurrence in AR(i) (Rackovsky-Scheraga, 1982); RACS820103 Average relative fractional occurrence in AL(i) (Rackovsky-Scheraga, 1982); RACS820104 Average relative fractional occurrence in EL(i) (Rackovsky-Scheraga, 1982); RACS820105 Average relative fractional occurrence in E0(i) (Rackovsky-Scheraga, 1982); RACS820106 Average relative fractional occurrence in ER(i) (Rackovsky-Scheraga, 1982); RACS820107 Average relative fractional occurrence in A0(i-1) (Rackovsky-Scheraga, 1982); RACS820108 Average relative fractional occurrence in AR(i-1) (Rackovsky-Scheraga, 1982); RACS820109 Average relative fractional occurrence in AL(i-1) (Rackovsky-Scheraga, 1982); RACS820110 Average relative fractional occurrence in EL(i-1) (Rackovsky-Scheraga, 1982); RACS820111 Average relative fractional occurrence in E0(i-1) (Rackovsky-Scheraga, 1982); RACS820112 Average relative fractional occurrence in ER(i-1) (Rackovsky-Scheraga, 1982); RACS820113 Value of theta(i) (Rackovsky-Scheraga, 1982); RACS820114 Value of theta(i-1) (Rackovsky-Scheraga, 1982); RADA880101 Transfer free energy from chx to wat (Radzicka-Wolfenden, 1988); RADA880102 Transfer free energy from oct to wat (Radzicka-Wolfenden, 1988); RADA880103 Transfer free energy from yap to chx (Radzicka-Wolfenden, 1988); RADA880104 Transfer free energy from chx to oct (Radzicka-Wolfenden, 1988); RADA880105 Transfer free energy from yap to oct (Radzicka-Wolfenden, 1988); RADA880106 Accessible surface area (Radzicka-Wolfenden, 1988); RADA880107 Energy transfer from out to in (95% buried) (Radzicka-Wolfenden, 1988); RADA880108 Mean polarity (Radzicka-Wolfenden, 1988); RICJ880101 Relative preference value at N″ (Richardson-Richardson, 1988); RICJ880102 Relative preference value at N′ (Richardson-Richardson, 1988); RICJ880103 Relative preference value at N-cap (Richardson-Richardson, 1988); RICJ880104 Relative preference value at N1 (Richardson-Richardson, 1988); RICJ880105 Relative preference value at N2 (Richardson-Richardson, 1988); RICJ880106 Relative preference value at N3 (Richardson-Richardson, 1988); RICJ880107 Relative preference value at N4 (Richardson-Richardson, 1988); RICJ880108 Relative preference value at N5 (Richardson-Richardson, 1988); RICJ880109 Relative preference value at Mid (Richardson-Richardson, 1988); RICJ880110 Relative preference value at C5 (Richardson-Richardson, 1988); RICJ880111 Relative preference value at C4 (Richardson-Richardson, 1988); RICJ880112 Relative preference value at C3 (Richardson-Richardson, 1988); RICJ880113 Relative preference value at C2 (Richardson-Richardson, 1988); RICJ880114 Relative preference value at C1 (Richardson-Richardson, 1988); RICJ880115 Relative preference value at C-cap (Richardson-Richardson, 1988); RICJ880116 Relative preference value at C′ (Richardson-Richardson, 1988); RICJ880117 Relative preference value at C″ (Richardson-Richardson, 1988); ROBB760101 Information measure for alpha-helix (Robson-Suzuki, 1976); ROBB760102 Information measure for N-terminal helix (Robson-Suzuki, 1976); ROBB760103 Information measure for middle helix (Robson-Suzuki, 1976); ROBB760104 Information measure for C-terminal helix (Robson-Suzuki, 1976); ROBB760105 Information measure for extended (Robson-Suzuki, 1976); ROBB760106 Information measure for pleated-sheet (Robson-Suzuki, 1976); ROBB760107 Information measure for extended without H-bond (Robson-Suzuki, 1976); ROBB760108 Information measure for turn (Robson-Suzuki, 1976); ROBB760109 Information measure for N-terminal turn (Robson-Suzuki, 1976); ROBB760110 Information measure for middle turn (Robson-Suzuki, 1976); ROBB760111 Information measure for C-terminal turn (Robson-Suzuki, 1976); ROBB760112 Information measure for coil (Robson-Suzuki, 1976); ROBB760113 Information measure for loop (Robson-Suzuki, 1976); ROBB790101 Hydration free energy (Robson-Osguthorpe, 1979); ROSG850101 Mean area buried on transfer (Rose et al., 1985); ROSG850102 Mean fractional area loss (Rose et al., 1985); ROSM880101 Side chain hydropathy, uncorrected for solvation (Roseman, 1988); ROSM880102 Side chain hydropathy, corrected for solvation (Roseman, 1988); ROSM880103 Loss of Side chain hydropathy by helix formation (Roseman, 1988); SIMZ760101 Transfer free energy (Simon, 1976), Cited by Charton-Charton (1982); SNEP660101 Principal component I (Sneath, 1966); SNEP660102 Principal component II (Sneath, 1966); SNEP660103 Principal component III (Sneath, 1966); SNEP660104 Principal component IV (Sneath, 1966); SUEM840101 Zimm-Bragg parameter s at 20 C (Sueki et al., 1984); SUEM840102 Zimm-Bragg parameter sigma×1.0E4 (Sueki et al., 1984); SWER830101 Optimal matching hydrophobicity (Sweet-Eisenberg, 1983); TANS770101 Normalized frequency of alpha-helix (Tanaka-Scheraga, 1977); TANS770102 Normalized frequency of isolated helix (Tanaka-Scheraga, 1977); TANS770103 Normalized frequency of extended structure (Tanaka-Scheraga, 1977); TANS770104 Normalized frequency of chain reversal R (Tanaka-Scheraga, 1977); TANS770105 Normalized frequency of chain reversal S (Tanaka-Scheraga, 1977); TANS770106 Normalized frequency of chain reversal D (Tanaka-Scheraga, 1977); TANS770107 Normalized frequency of left-handed helix (Tanaka-Scheraga, 1977); TANS770108 Normalized frequency of zeta R (Tanaka-Scheraga, 1977); TANS770109 Normalized frequency of coil (Tanaka-Scheraga, 1977) TANS770110 Normalized frequency of chain reversal (Tanaka-Scheraga, 1977); VASM830101 Relative population of conformational state A (Vasquez et al., 1983); VASM830102 Relative population of conformational state C (Vasquez et al., 1983); VASM830103 Relative population of conformational state E (Vasquez et al., 1983); VELV850101 Electron-ion interaction potential (Veljkovic et al., 1985); VENT840101 Bitterness (Venanzi, 1984); VHEG790101 Transfer free energy to lipophilic phase (von Heijne-Blomberg, 1979); WARP780101 Average interactions per side chain atom (Warme-Morgan, 1978); WEBA780101 RF value in high salt chromatography (Weber-Lacey, 1978); WERD780101 Propensity to be buried inside (Wertz-Scheraga, 1978); WERD780102 Free energy change of epsilon(i) to epsilon(ex) (Wertz-Scheraga, 1978); WERD780103 Free energy change of alpha(Ri) to alpha(Rh) (Wertz-Scheraga, 1978); WERD780104 Free energy change of epsilon(i) to alpha(Rh) (Wertz-Scheraga, 1978); WOEC730101 Polar requirement (Woese, 1973); WOLR810101 Hydration potential (Wolfenden et al., 1981); WOLS870101 Principal property value z1 (Wold et al., 1987); WOLS870102 Principal property value z2 (Wold et al., 1987); WOLS870103 Principal property value z3 (Wold et al., 1987); YUTK870101 Unfolding Gibbs energy in water, pH7.0 (Yutani et al., 1987); YUTK870102 Unfolding Gibbs energy in water, pH9.0 (Yutani et al., 1987); YUTK870103 Activation Gibbs energy of unfolding, pH7.0 (Yutani et al., 1987); YUTK870104 Activation Gibbs energy of unfolding, pH9.0 (Yutani et al., 1987); ZASB820101 Dependence of partition coefficient on ionic strength (Zaslaysky et al., 1982); ZIMJ680101 Hydrophobicity (Zimmerman et al., 1968); ZIMJ680102 Bulkiness (Zimmerman et al., 1968); ZIMJ680103 Polarity (Zimmerman et al., 1968); ZIMJ680104 Isoelectric point (Zimmerman et al., 1968); ZIMJ680105 RF rank (Zimmerman et al., 1968); AURR980101 Normalized positional residue frequency at helix termini N4′(Aurora-Rose, 1998); AURR980102 Normalized positional residue frequency at helix termini N′″ (Aurora-Rose, 1998); AURR980103 Normalized positional residue frequency at helix termini N″ (Aurora-Rose, 1998); AURR980104 Normalized positional residue frequency at helix termini N′(Aurora-Rose, 1998); AURR980105 Normalized positional residue frequency at helix termini Nc (Aurora-Rose, 1998); AURR980106 Normalized positional residue frequency at helix termini N1 (Aurora-Rose, 1998); AURR980107 Normalized positional residue frequency at helix termini N2 (Aurora-Rose, 1998); AURR980108 Normalized positional residue frequency at helix termini N3 (Aurora-Rose, 1998); AURR980109 Normalized positional residue frequency at helix termini N4 (Aurora-Rose, 1998); AURR980110 Normalized positional residue frequency at helix termini N5 (Aurora-Rose, 1998); AURR980111 Normalized positional residue frequency at helix termini C5 (Aurora-Rose, 1998); AURR980112 Normalized positional residue frequency at helix termini C4 (Aurora-Rose, 1998); AURR980113 Normalized positional residue frequency at helix termini C3 (Aurora-Rose, 1998); AURR980114 Normalized positional residue frequency at helix termini C2 (Aurora-Rose, 1998); AURR980115 Normalized positional residue frequency at helix termini C1 (Aurora-Rose, 1998); AURR980116 Normalized positional residue frequency at helix termini Cc (Aurora-Rose, 1998); AURR980117 Normalized positional residue frequency at helix termini C′ (Aurora-Rose, 1998); AURR980118 Normalized positional residue frequency at helix termini C″ (Aurora-Rose, 1998); AURR980119 Normalized positional residue frequency at helix termini C′″ (Aurora-Rose, 1998); AURR980120 Normalized positional residue frequency at helix termini C4′ (Aurora-Rose, 1998); ONEK900101 Delta G values for the peptides extrapolated to 0 M urea (O'Neil-DeGrado, 1990); ONEK900102 Helix formation parameters (delta delta G) (O'Neil-DeGrado, 1990); VINM940101 Normalized flexibility parameters (B-values), average (Vihinen et al., 1994); VINM940102 Normalized flexibility parameters (B-values) for each residue surrounded by none rigid neighbours (Vihinen et al., 1994); VINM940103 Normalized flexibility parameters (B-values) for each residue surrounded by one rigid neighbours (Vihinen et al., 1994); VINM940104 Normalized flexibility parameters (B-values) for each residue surrounded by two rigid neighbours (Vihinen et al., 1994); MUNV940101 Free energy in alpha-helical conformation (Munoz-Serrano, 1994); MUNV940102 Free energy in alpha-helical region (Munoz-Serrano, 1994); MUNV940103 Free energy in beta-strand conformation (Munoz-Serrano, 1994); MUNV940104 Free energy in beta-strand region (Munoz-Serrano, 1994); MUNV940105 Free energy in beta-strand region (Munoz-Serrano, 1994) WIMW960101 Free energies of transfer of AcW1-X-LL peptides from bilayer interface to water (Wimley-White, 1996); KIMC930101 Thermodynamic beta sheet propensity (Kim-Berg, 1993); MONM990101 Turn propensity scale for transmembrane helices (Monne et al., 1999); BLAM930101 Alpha helix propensity of position 44 in T4 lysozyme (Blaber et al., 1993); PARS000101 p-Values of mesophilic proteins based on the distributions of B values (Parthasarathy-Murthy, 2000); PARS000102 p-Values of thermophilic proteins based on the distributions of B values (Parthasarathy-Murthy, 2000); KUMS000101 Distribution of amino acid residues in the 18 non-redundant families of thermophilic proteins (Kumar et al., 2000); KUMS000102 Distribution of amino acid residues in the 18 non-redundant families of mesophilic proteins (Kumar et al., 2000); KUMS000103 Distribution of amino acid residues in the alpha-helices in thermophilic proteins (Kumar et al., 2000); KUMS000104 Distribution of amino acid residues in the alpha-helices in mesophilic proteins (Kumar et al., 2000); TAKK010101 Side-chain contribution to protein stability (kJ/mol) (Takano-Yutani, 2001); FODM020101 Propensity of amino acids within pi-helices (Fodje-Al-Karadaghi, 2002); NADH010101 Hydropathy scale based on self-information values in the two-state model (5% accessibility) (Naderi-Manesh et al., 2001); NADH010102 Hydropathy scale based on self-information values in the two-state model (9% accessibility) (Naderi-Manesh et al., 2001); NADH010103 Hydropathy scale based on self-information values in the two-state model (16% accessibility) (Naderi-Manesh et al., 2001); NADH010104 Hydropathy scale based on self-information values in the two-state model (20% accessibility) (Naderi-Manesh et al., 2001); NADH010105 Hydropathy scale based on self-information values in the two-state model (25% accessibility) (Naderi-Manesh et al., 2001); NADH010106 Hydropathy scale based on self-information values in the two-state model (36% accessibility) (Naderi-Manesh et al., 2001); NADH010107 Hydropathy scale based on self-information values in the two-state model (50% accessibility) (Naderi-Manesh et al., 2001); MONM990201 Averaged turn propensities in a transmembrane helix (Monne et al., 1999); KOEP990101 Alpha-helix propensity derived from designed sequences (Koehl-Levitt, 1999); KOEP990102 Beta-sheet propensity derived from designed sequences (Koehl-Levitt, 1999); CEDJ970101 Composition of amino acids in extracellular proteins (percent) (Cedano et al., 1997); CEDJ970102 Composition of amino acids in anchored proteins (percent) (Cedano et al., 1997); CEDJ970103 Composition of amino acids in membrane proteins (percent) (Cedano et al., 1997); CEDJ970104 Composition of amino acids in intracellular proteins (percent) (Cedano et al., 1997); CEDJ970105 Composition of amino acids in nuclear proteins (percent) (Cedano et al., 1997); FUKS010101 Surface composition of amino acids in intracellular proteins of thermophiles (percent) (Fukuchi-Nishikawa, 2001); FUKS010102 Surface composition of amino acids in intracellular proteins of mesophiles (percent) (Fukuchi-Nishikawa, 2001); FUKS010103 Surface composition of amino acids in extracellular proteins of mesophiles (percent) (Fukuchi-Nishikawa, 2001); FUKS010104 Surface composition of amino acids in nuclear proteins (percent) (Fukuchi-Nishikawa, 2001); FUKS010105 Interior composition of amino acids in intracellular proteins of thermophiles (percent) (Fukuchi-Nishikawa, 2001); FUKS010106 Interior composition of amino acids in intracellular proteins of mesophiles (percent) (Fukuchi-Nishikawa, 2001); FUKS010107 Interior composition of amino acids in extracellular proteins of mesophiles (percent) (Fukuchi-Nishikawa, 2001); FUKS010108 Interior composition of amino acids in nuclear proteins (percent) (Fukuchi-Nishikawa, 2001); FUKS010109 Entire chain composition of amino acids in intracellular proteins of thermophiles (percent) (Fukuchi-Nishikawa, 2001); FUKS010110 Entire chain composition of amino acids in intracellular proteins of mesophiles (percent) (Fukuchi-Nishikawa, 2001); FUKS010111 Entire chain composition of amino acids in extracellular proteins of mesophiles (percent) (Fukuchi-Nishikawa, 2001); FUKS010112 Entire chain composition of amino acids in nuclear proteins (percent) (Fukuchi-Nishikawa, 2001); AVBF000101 Screening coefficients gamma, local (Avbelj, 2000); AVBF000102 Screening coefficients gamma, non-local (Avbelj, 2000); AVBF000103 Slopes tripeptide, FDPB VFF neutral (Avbelj, 2000); AVBF000104 Slopes tripeptides, LD VFF neutral (Avbelj, 2000); AVBF000105 Slopes tripeptide, FDPB VFF noside (Avbelj, 2000); AVBF000106 Slopes tripeptide FDPB VFF all (Avbelj, 2000); AVBF000107 Slopes tripeptide FDPB PARSE neutral (Avbelj, 2000); AVBF000108 Slopes dekapeptide, FDPB VFF neutral (Avbelj, 2000); AVBF000109 Slopes proteins, FDPB VFF neutral (Avbelj, 2000); YANJ020101 Side-chain conformation by gaussian evolutionary method (Yang et al., 2002); MITS020101 Amphiphilicity index (Mitaku et al., 2002); TSAJ990101 Volumes including the crystallographic waters using the ProtOr (Tsai et al., 1999); TSAJ990102 Volumes not including the crystallographic waters using the ProtOr (Tsai et al., 1999); COSI940101 Electron-ion interaction potential values (Cosic, 1994); PONP930101 Hydrophobicity scales (Ponnuswamy, 1993); WILM950101 Hydrophobicity coefficient in RP-HPLC, C18 with 0.1% TFA/MeCN/H2O (Wilce et al. 1995); WILM950102 Hydrophobicity coefficient in RP-HPLC, C8 with 0.1% TFA/MeCN/H2O (Wilce et al. 1995); WILM950103 Hydrophobicity coefficient in RP-HPLC, C4 with 0.1% TFA/MeCN/H2O (Wilce et al. 1995); WILM950104 Hydrophobicity coefficient in RP-HPLC, C18 with 0.1% TFA/2-PrOH/MeCN/H2O (Wilce et al. 1995); KUHL950101 Hydrophilicity scale (Kuhn et al., 1995); GUOD860101 Retention coefficient at pH 2 (Guo et al., 1986); JURD980101 Modified Kyte-Doolittle hydrophobicity scale (Juretic et al., 1998); BASU050101 Interactivity scale obtained from the contact matrix (Bastolla et al., 2005); BASU050102 Interactivity scale obtained by maximizing the mean of correlation coefficient over single-domain globular proteins (Bastolla et al., 2005); BASU050103 Interactivity scale obtained by maximizing the mean of correlation coefficient over pairs of sequences sharing the TIM barrel fold (Bastolla et al., 2005); SUYM030101 Linker propensity index (Suyama-Ohara, 2003); PUNT030101 Knowledge-based membrane-propensity scale from 1D_Helix in MPtopo databases (Punta-Maritan, 2003); PUNT030102 Knowledge-based membrane-propensity scale from 3D_Helix in MPtopo databases (Punta-Maritan, 2003); GEOR030101 Linker propensity from all dataset (George-Heringa, 2003); GEOR030102 Linker propensity from 1-linker dataset (George-Heringa, 2003); GEOR030103 Linker propensity from 2-linker dataset (George-Heringa, 2003); GEOR030104 Linker propensity from 3-linker dataset (George-Heringa, 2003); GEOR030105 Linker propensity from small dataset (linker length is less than six residues) (George-Heringa, 2003); GEOR030106 Linker propensity from medium dataset (linker length is between six and 14 residues) (George-Heringa, 2003); GEOR030107 Linker propensity from long dataset (linker length is greater than 14 residues) (George-Heringa, 2003); GEOR030108 Linker propensity from helical (annotated by DSSP) dataset (George-Heringa, 2003); GEOR030109 Linker propensity from non-helical (annotated by DSSP) dataset (George-Heringa, 2003); ZHOH040101 The stability scale from the knowledge-based atom-atom potential (Zhou-Zhou, 2004); ZHOH040102 The relative stability scale extracted from mutation experiments (Zhou-Zhou, 2004); ZHOH040103 Buriability (Zhou-Zhou, 2004); BAEK050101 Linker index (Bae et al., 2005); HARY940101 Mean volumes of residues buried in protein interiors (Harpaz et al., 1994); PONJ960101 Average volumes of residues (Pontius et al., 1996); DIGM050101 Hydrostatic pressure asymmetry index, PAI (Di Giulio, 2005); WOLR790101 Hydrophobicity index (Wolfenden et al., 1979); OLSK800101 Average internal preferences (Olsen, 1980); KIDA850101 Hydrophobicity-related index (Kidera et al., 1985); GUYH850102 Apparent partition energies calculated from Wertz-Scheraga index (Guy, 1985); GUYH850103 Apparent partition energies calculated from Robson-Osguthorpe index (Guy, 1985); GUYH850104 Apparent partition energies calculated from Janin index (Guy, 1985); GUYH850105 Apparent partition energies calculated from Chothia index (Guy, 1985); ROSM880104 Hydropathies of amino acid side chains, neutral form (Roseman, 1988); ROSM880105 Hydropathies of amino acid side chains, pi-values in pH 7.0 (Roseman, 1988); JACR890101 Weights from the IFH scale (Jacobs-White, 1989); COWR900101 Hydrophobicity index, 3.0 pH (Cowan-Whittaker, 1990) BLAS910101 Scaled side chain hydrophobicity values (Black-Mould, 1991); CASG920101 Hydrophobicity scale from native protein structures (Casari-Sippl, 1992); CORJ870101 NNEIG index (Cornette et al., 1987); CORJ870102 SWEIG index (Cornette et al., 1987); CORJ870103 PRIFT index (Cornette et al., 1987); CORJ870104 PRILS index (Cornette et al., 1987); CORJ870105 ALTFT index (Cornette et al., 1987) CORJ870106 ALTLS index (Cornette et al., 1987); CORJ870107 TOTFT index (Cornette et al., 1987); CORJ870108 TOTLS index (Cornette et al., 1987); MIYS990101 Relative partition energies derived by the Bethe approximation (Miyazawa-Jernigan, 1999); MIYS990102 Optimized relative partition energies −method A (Miyazawa-Jernigan, 1999); MIYS990103 Optimized relative partition energies—method B (Miyazawa-Jernigan, 1999); MIYS990104 Optimized relative partition energies—method C (Miyazawa-Jernigan, 1999); MIYS990105 Optimized relative partition energies—method D (Miyazawa-Jernigan, 1999); ENGD860101 Hydrophobicity index (Engelman et al., 1986); and FASG890101 Hydrophobicity index (Fasman, 1989)

In some embodiments of the invention, degenerate oligonucleotides are used to synthesize one or more of the TN1, DH, N2, and/or H3-JH segments of the invention. In certain embodiments of the invention, the codon at or near the 5′ end of the oligonucleotide encoding the H3-JH segment is a degenerate codon. Such degenerate codons may be the first codon from the 5′ end, the second codon from the 5′ end, the third codon from the 5′ end, the fourth codon from the 5′ end, the fifth codon from the 5′ end, and/or any combination of the above. In some embodiments of the invention, one or more of the codons at or near the 5′ and/or 3′ ends of the DH segment are degenerate. Such degenerate codons may be the first codon from the 5′ and/or 3′ end(s), the second codon from the 5′ and/or 3′ end(s), the third codons from the 5′ and/or 3′ end(s), the fourth codon from the 5′ and/or 3′ end(s), the fifth codon from the 5′ and/or 3′ end(s), and/or any combination of the above. Degenerate codons used in each of the oligonucleotides encoding the segments may be selected for their ability to optimally recapitulate sequences in a theoretical segment pool and/or CDRH3 reference set.

In some embodiments, the invention provides methods of producing a theoretical segment pool of H3-JH segments, as described in the Examples. Theoretical segment pools generated utilizing NNN triplets, instead of or in addition to the NN doublets described in Example 5 also fall within the scope of the invention, as do synthetic libraries incorporating segments from these theoretical segment pools.

In some embodiments, the invention provides methods of producing a theoretical segment pool of DH segments, as described in the Examples. In particular, for example, the invention provides methods of producing a theoretical segment pool of DH segments described by the PYTHON program of Example 6. Example 6 describes the application of this program to produce the 68K theoretical segment pool (minimum length of DNA sequences after progressive deletions=4 bases; and minimum length of peptide sequences for inclusion in the theoretical segment pool=2). An alternative example is provided wherein the minimum length of the DNA sequences after progressive deletions was one base and the minimum length of the peptide sequence is one amino acid. It is also contemplated that other values could be used for these parameters. For example, the minimum length of the DNA sequences after progressive deletions could be set as 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, or 15, and the minimum length of the peptide sequences in the theoretical segment pool could be set as 1, 2, 3, 4, or 5.

Design of CDRH3 Libraries Using the TN1, DH, N2, and H3-JH Segments

The CDRH3 libraries of the invention comprise TN1, DH, N2, and H3-JH segments. Thus, in certain embodiments of the invention, the overall design of the CDRH3 libraries can be represented by the following formula:

[TN1]-[DH]-[N2]-[H3-JH].

In certain embodiments of the invention, a synthetic CDRH3 repertoire is combined with selected VH chassis sequences and heavy chain constant regions, via homologous recombination. Therefore, in certain embodiments of the invention, it may be desirable to include DNA sequences flanking the 5′ and 3′ ends of the synthetic CDRH3 libraries, to facilitate homologous recombination between the synthetic CDRH3 libraries and vectors containing the selected chassis and constant regions. In certain embodiments, the vectors also contain a sequence encoding at least a portion of the non-truncated region of the IGHJ gene (i.e., FRM4-JH). Thus, a polynucleotide encoding an N-terminal sequence (e.g., CA(K/R/T)) may be added to the synthetic CDRH3 sequences, wherein the N-terminal polynucleotide is homologous with FRM3 of the chassis, while a polynucleotide encoding a C-terminal sequence (e.g., WG(Q/R/K)G) may be added to the synthetic CDRH3, wherein the C-terminal polynucleotide is homologous with FRM4-JH. Although the sequence WG(Q/R)G is presented in this exemplary embodiment, additional amino acids, C-terminal to this sequence in FRM4-JH may also be included in the polynucleotide encoding the C-terminal sequence. The purpose of the polynucleotides encoding the N-terminal and C-terminal sequences, in this case, is to facilitate homologous recombination, and one of ordinary skill in the art would recognize that these sequences may be longer or shorter than depicted below. Accordingly, in certain embodiments of the invention, the overall design of the CDRH3 repertoire, including the sequences required to facilitate homologous recombination with the selected chassis, can be represented by the following formula (regions homologous with vector underlined):

(SEQ ID NO. 8762, SEQ ID NO. 8763, and

SEQ ID NO. 8764, respectively)

CA[R/K/T] -[TN1]-[DH]-[N2]-[H3-JH]- [WG(Q/R/K)G] .

In some embodiments of the invention, the CDRH3 repertoire can be represented by the following formula, which excludes the T residue presented in the schematic above:

References describing collections of V, D, and J genes include Scaviner et al., Exp. Clin, Immunogenet., 1999, 16: 243 and Ruiz et al., Exp. Clin. Immunogenet, 1999, 16: 173, each incorporated by reference in its entirety.

Although homologous recombination is one method of producing the libraries of the invention, a person of ordinary skill in the art will readily recognize that other methods of DNA assembly, such as ligation or site-specific recombination, and/or DNA synthesis, can also be used to produce the libraries of the invention.

CDRH3 Lengths

The lengths of the segments may also be varied, for example, to produce libraries with a particular distribution of CDRH3 lengths. In one embodiment of the invention, the H3-JH segments are about 0 to about 10 amino acids in length, the DH segments are about 0 to about 12 amino acids in length, the TN1 segments are about 0 to about 4 amino acids in length, and the N2 segments are about 0 to about 4 amino acids in length. In certain embodiments, the H3-JH segments are at least about 0, 1, 2, 3, 4, 5, 6, 7, 8, 9, and/or 10 amino acids in length. In some embodiments, the DH segments are at least about 0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, and/or 12 amino acids in length. In certain embodiments, the TN1 segments are at least about 0, 1, 2, 3, or 4 amino acids in length. In some embodiments, the N2 amino acids are at least about 0, 1, 2, 3, or 4 amino acids in length. In certain embodiments of the invention, the CDRH3 is about 2 to about 35, about 2 to about 28, or about 5 to about 26 amino acids in length. In some embodiments, the CDRH3 is at least about 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25, 26, 27, 28, 29, 30, 31, 32, 33, 34, and/or 35 amino acids in length. In some embodiments, the length of any of the segments or CDRH3s of the invention may be less than a particular number of amino acids, where the number of amino acids is defined using any one of the integers provided above for the respective segment or CDRH3. In certain embodiments of the invention, a particular numerical range is defined, using any two of the integers provided above as lower and upper boundaries of the range, inclusive or exclusive. All combinations of the integers provided, which define an upper and lower boundary, are contemplated.

Design of CDRL3 Libraries

The design of CDRL3 libraries, and light chain sequences, is described in detail in U.S. Publication Nos. 2009/0181855 and 2010/0056386, and WO/2009/036379, each of which is incorporated by reference in its entirety, and is therefore only described briefly herein. Libraries described herein are designed according to similar principles, with three important differences, namely that the libraries of the current invention contain (1) variability in CDRL1 and CDRL2; (2) variability in the framework regions; and/or (3) variability in CDRL3 that is designed to produce light chain libraries with CDRL3s that closely resemble human germline-like CDRL3 sequences, as defined above (Table 1).

A CDRL3 library of the invention may be a VKCDR3 library and/or a VλCDR3 library. In certain embodiments of the invention, patterns of occurrence of particular amino acids at defined positions within VL sequences are determined by analyzing data available in public or other databases, for example, the NCBI database (see, for example, WO/2009/036379). In certain embodiments of the invention, these sequences are compared on the basis of identity and assigned to families on the basis of the germline genes from which they are derived. The amino acid composition at each position of the sequence, in each germline family, may then be determined. This process is illustrated in the Examples provided herein.

Light Chains with Framework Variability

In some embodiments, the invention provides a library of light chain variable domains wherein the light chain variable domains are varied at one or more of framework positions 2, 4, 36, 46, 48, 49, and 66. In some embodiments, the invention provides a library of light chain variable domains comprising at least a plurality of light chain variable domains whose amino acid sequences are identical to one another except for substitutions at one or more of positions 2, 4, 36, 46, 48, 49, and 66. In certain embodiments, the invention provides a library of light chain variable domains comprising at least a plurality of light chain variable domains whose amino acid sequences are at least about 70%, 75%, 80%, 85%, 90%, 95%, 96%, 97%, 98%, 99%, and/or 99.5% to any of the light chain variable domain sequences disclosed herein, and further have substitutions at one or more of positions 2, 4, 36, 46, 48, 49, and 66. In some embodiments, the amino acids selected for inclusion in these positions are selected from amongst about the most 2, 3, 4, 5, 6, 7, 8, 9, and/or 10 most frequently occurring amino acids at the corresponding position in a reference set of light chain variable domains.

In some embodiments, the invention provides systems and methods of selecting framework positions to be varied in a light chain variable domain, comprising:

(i) obtaining a reference set of light chain sequences, wherein the reference set contains light chain sequences with VL segments selected from the group consisting of sequences found in, or encoded by, a single IGVL germline gene and/or sequences found in, or encoded by, allelic variants of the single IGVL germline gene; (ii) determining which framework positions within the reference set have a degree of variability that is similar to the degree of variability occurring in one more CDR positions of the sequences in the reference set (e.g., the variability in a framework position is at least about 70%, 80%, 90%, or 95%, 100%, or more of the variability found in a CDR position of the sequences in the reference set); (iii) determining the frequency of occurrence of amino acid residues for each of the framework positions identified in (ii); (iv) synthesizing light chain variable domain encoding sequences wherein the framework positions identified in (ii) are varied to include the 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, or 20 most frequently occurring amino acid residues (identified in (iii)) at the corresponding position.

One of ordinary skill in the art, reading the present disclosure will appreciate that the present invention provides analogous methods for developing framework variants of heavy chain sequences.

Light Chains with CDR1 and/or CDR2 Variability

In some embodiments, the invention provides a library of light chain variable domains wherein the light chain variable domains are varied at one or more of CDRL1 positions 28, 29, 30, 30A, 30B, 30E, 31, and 32 (Chothia-Lesk numbering scheme; Chothia and Lesk, J. Mol. Biol., 1987, 196: 901). In some embodiments, the invention provides a library of light chain variable domains wherein the light chain variable domains are varied at one or more of CDRL2 positions 50, 51, 53, and 55. In some embodiments, the amino acids selected for inclusion in these CDRL1 and/or CDRL2 positions are selected from amongst about the most 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, and/or 20 most frequently occurring amino acids at the corresponding position in a reference set of light chain variable domains.

In some embodiments, the invention provides systems and methods for selecting CDRL1 and/or CDRL2 positions to be varied in a light chain variable domain, comprising:

(i) obtaining a reference set of light chain sequences, wherein the reference set contains light chain sequences with VL segments selected from the group consisting of sequences found in, or encoded by, a single IGVL germline gene and sequences found in, or encoded by, allelic variants of the single IGVL germline gene; (ii) determining which CDRL1 and/or CDRL2 positions are variable within the reference set; (iii) synthesizing light chain variable domain encoding sequences wherein the CDRL1 and/or CDRL2 positions identified in (ii) are varied to include the 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, or 20 most frequently occurring amino acid residues at the corresponding position.

One of ordinary skill in the art, reading the present disclosure will appreciate that the present invention provides analogous methods for developing CDRH2 and/or CDRH2 variants of heavy chain sequences.

Light Chain Sequences

In some embodiments, the invention provides a light chain library comprising one or more of any of the light chain sequences provided herein, for example, the polypeptide sequences of Table 3 and/or Table 4 and/or the polynucleotide sequences of Table 5, Table 6, and/or Table 7. A person of ordinary skill in the art will recognize that not every light chain sequence provided herein is necessary to produce a functional light chain library of the invention. Therefore, in certain embodiments, a light chain library of the invention will contain a subset of the sequences described above. For example, in certain embodiments of the invention, at least about 10, 100, 200, 300, 400, 500, 600, 700, 800, 900, 10 3 , 10 4 , and/or 10 5 of the light chain polynucleotide and/or polypeptide sequences provided herein are included in a library. In some embodiments, a library of the invention may contain less than a particular number of polynucleotide or polypeptide segments, where the number of segments is defined using any one of the integers provided above for the respective segment. In certain embodiments of the invention, a particular numerical range is defined, using any two of the integers provided above as lower and upper boundaries of the range, inclusive or exclusive. All combinations of the integers provided, which define an upper and lower boundary, are contemplated.

In certain embodiments, the invention provides light chain libraries comprising at least about 1%, 2.5%, 5%, 10%, 15%, 20%, 25%, 30%, 40%, 45%, 50%, 55%, 60%, 65%, 70%, 75%, 80%, 85%, 90%, 95%, or 99% of the sequences from any of the sets of light chain sequences provided herein. For example, the invention provides libraries comprising at least about 1%, 2.5%, 5%, 10%, 15%, 20%, 25%, 30%, 40%, 45%, 50%, 55%, 60%, 65%, 70%, 75%, 80%, 85%, 90%, 95%, or 99% of the light chain sequences provided in Table 3, Table 4, Table 5, Table 6, and/or Table 7. In some embodiments of the invention, a particular percentage range is defined, using any two of the percentages provided above as lower and upper boundaries of the range, inclusive or exclusive. All combinations of the percentages provided, which define an upper and lower boundary, are contemplated.

In some embodiments of the invention, at least about 1%, 2.5%, 5%, 10%, 15%, 20%, 25%, 30%, 40%, 45%, 50%, 55%, 60%, 65%, 70%, 75%, 80%, 85%, 90%, 95%, or 99% of the light chain sequences in a library are light chain sequences provided herein. In certain embodiments of the invention, at least about 1%, 2.5%, 5%, 10%, 15%, 20%, 25%, 30%, 40%, 45%, 50%, 55%, 60%, 65%, 70%, 75%, 80%, 85%, 90%, 95%, or 99% of the light chain sequences isolated from a light chain library (e.g., by binding to a particular antigen and/or generic ligand) are light chain sequences provided herein. In some embodiments, a light chain library of the invention may contain less than a particular percentage of light chain sequences provided herein, where the percentage of light chain sequences is defined using any one of the percentages provided above. In certain embodiments of the invention, a particular percentage range is defined, using any two of the percentages provided above as lower and upper boundaries of the range, inclusive or exclusive. All combinations of the percentages provided, which define an upper and lower boundary, are contemplated.

One of ordinary skill in the art will further recognize that given the light chain sequences provided herein, similar light chain sequences could be produced which share a designated level of overall sequence identity and/or one or more characteristic sequence elements described herein, which overall degree of sequence identity and/or characteristic sequence elements may confer common functional attributes. Those of ordinary skill in the art will be well familiar with a variety of techniques for preparing such related sequences, including the mutagenesis techniques provided herein. Therefore, each of the explicitly enumerated embodiments of the invention can also be practiced using light chain sequences that share a particular percent identity to any of the light chain sequences provided herein. For example, each of the previously described embodiments of the invention can be practiced using light chain sequences that are at least about 70%, 75%, 80%, 85%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98%, 99%, 99.5%, or 99.9% identical to the light chain sequences provided herein. For example, in some embodiments, light chain libraries provided by the invention comprise light chain variable domains at least about 70%, 75%, 80%, 85%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98%, 99%, 99.5%, or 99.9% identical to the light chain sequences provided herein, with substitutions in one or more of framework positions 2, 4, 36, 46, 48, 49, and 66, CDRL1 positions 28, 29, 30, 30A, 30B, 30E, 31, and 32 (Chothia-Lesk numbering scheme), and/or CDRL2 positions 50, 51, 53, and 55.

In some embodiments, the invention provides systems and methods for varying positions within the portion of CDRL3s encoded by a particular IGVL germline gene, comprising:

(i) obtaining a reference set of light chain sequences, wherein the reference set contains light chain sequences with VL segments originating from the same IGVL germline gene and/or its allelic variants; (ii) determining which amino acids occur at each of the CDRL3 positions in the reference set that are encoded by the IGVL gene (i.e., positions 89-94, inclusive); (iii) synthesizing light chain variable domain encoding sequences wherein two positions in each light chain variable domain encoding sequence contain degenerate codons encoding the 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, or 20 most frequently occurring amino acid residues at the corresponding positions in the reference set.

As described in the examples, the degenerate codons of (iii) can be chosen to best reproduce the amino acid diversity contained in the reference set for each of the two positions varied in each light chain. Finally, while the methods and systems described above are described with respect to CDRL3, one of ordinary skill in the art will readily recognize that the same principles can be applied to CDRH1 and/or CDRH2 of the heavy chain, which are encoded entirely by the IGHV gene.

CDRL3 Lengths

In some embodiments, as an alternative or in addition to other features described herein, the present invention provides libraries in which lengths of CDRL3s may be varied. The present invention therefore provides, among other things, libraries with a particular distribution of CDRL3 lengths. Although CDRL3 libraries of lengths 8, 9, and 10 are exemplified, one of ordinary skill in the art will readily recognize that the methods described herein can be applied to produce light chains with CDRL3s of different lengths (e.g., about 5, 6, 7, 11, 12, 13, 14, 15, and/or 16) that also fall within the scope of the invention. In some embodiments, the length of any of the CDRL3s of the invention may be less than a particular number of amino acids, where the number of amino acids is defined using any one of the integers provided above. In some embodiments of the invention, a particular numerical range is defined, using any two of the integers provided above as lower and upper boundaries of the range, inclusive or exclusive. All combinations of the integers provided, which define an upper and lower boundary, are contemplated.

Synthetic Antibody Libraries

In some embodiments of the invention, provided libraries include one or more synthetic polynucleotides. In some embodiments, provided libraries may comprise synthetic polynucleotides selected from (a) heavy chain chassis polynucleotides; (b) light chain chassis polynucleotides; (c) CDR3 polynucleotides; (d) constant domain polynucleotides; and (e) combinations thereof. Those of ordinary skill in the art will appreciate that such synthetic polynucleotides may be linked to other synthetic or non-synthetic polynucleotides in provided libraries.

Synthetic polynucleotides provided herein may be prepared by any available method. For example, in some embodiments, synthetic polynucleotides can be synthesized by split pool DNA synthesis as described in Feldhaus et al., Nucleic Acids Research, 2000, 28: 534; Omstein et al., Biopolymers, 1978, 17: 2341; Brenner and Lerner, PNAS, 1992, 87: 6378, U.S. Publication Nos. 2009/0181855 and 2010/0056386, and WO/2009/036379 (each incorporated by reference in its entirety).

In some embodiments of the invention, segments representing the possible TN1, DH, N2, and JH diversity found in the human repertoire are synthesized de novo either as double-stranded DNA oligonucleotides, single-stranded DNA oligonucleotides representative of the coding strand, or single-stranded DNA oligonucleotides representative of the non-coding strand. Such sequences can then be introduced into a host cell along with an acceptor vector containing a chassis sequence and, in some cases a portion of FRM4 and a constant region. No primer-based PCR amplification from mammalian cDNA or mRNA or template-directed cloning steps from mammalian cDNA or mRNA need be employed.

Construction of Libraries by Yeast Homologous Recombination

In certain embodiments, the invention exploits the inherent ability of yeast cells to facilitate homologous recombination at high efficiency. The mechanism of homologous recombination in yeast and its applications are briefly described below (also see e.g., U.S. Pat. Nos. 6,406,863; 6,410,246; 6,410,271; 6,610,472; and 7,700,302, each of which is incorporated by reference in its entirety).

As an illustrative embodiment, homologous recombination can be carried out in, for example, Saccharomyces cerevisiae , which has genetic machinery designed to carry out homologous recombination with high efficiency. Exemplary S. cerevisiae strains include EM93, CEN.PK2, RM11-1a, YJM789, and BJ5465. This mechanism is believed to have evolved for the purpose of chromosomal repair, and is also called “gap repair” or “gap filling”. By exploiting this mechanism, mutations can be introduced into specific loci of the yeast genome. For example, a vector carrying a mutant gene can contain two sequence segments that are homologous to the 5′ and 3′ open reading frame (ORF) sequences of a gene that is intended to be interrupted or mutated. The vector may also encode a positive selection marker, such as a nutritional enzyme allele (e.g., URA3) and/or an antibiotic resistant marker (e.g., Geneticin/G418), flanked by the two homologous DNA segments. Other selection markers and antibiotic resistance markers are known to one of ordinary skill in the art.

In some embodiments of the invention, this vector (e.g., a plasmid) is linearized and transformed into the yeast cells. Through homologous recombination between the plasmid and the yeast genome, at the two homologous recombination sites, a reciprocal exchange of the DNA content occurs between the wild type gene in the yeast genome and the mutant gene (including the selection marker gene(s)) that is flanked by the two homologous sequence segments. By selecting for the one or more selection markers, the surviving yeast cells will be those cells in which the wild-type gene has been replaced by the mutant gene (Pearson et al., Yeast, 1998, 14: 391, incorporated by reference in its entirety). This mechanism has been used to make systematic mutations in all 6,000 yeast genes, or open reading frames (ORFs), for functional genomics studies. Because the exchange is reciprocal, a similar approach has also been used successfully to clone yeast genomic DNA fragments into a plasmid vector (Iwasaki et al., Gene, 1991, 109: 81, incorporated by reference in its entirety).

By utilizing the endogenous homologous recombination machinery present in yeast, gene fragments or synthetic oligonucleotides can also be cloned into a plasmid vector without a ligation step. In this application of homologous recombination, a target gene fragment (i.e., the fragment to be inserted into a plasmid vector, e.g., a CDR3) is obtained (e.g., by oligonucleotides synthesis, PCR amplification, restriction digestion out of another vector, etc.). DNA sequences that are homologous to selected regions of the plasmid vector are added to the 5′ and 3′ ends of the target gene fragment. These homologous regions may be fully synthetic, or added via PCR amplification of a target gene fragment with primers that incorporate the homologous sequences. The plasmid vector may include a positive selection marker, such as a nutritional enzyme allele (e.g., URA3), or an antibiotic resistance marker (e.g., Geneticin/G418). The plasmid vector is then linearized by a unique restriction cut located in-between the regions of sequence homology shared with the target gene fragment, thereby creating an artificial gap at the cleavage site. The linearized plasmid vector and the target gene fragment flanked by sequences homologous to the plasmid vector are co-transformed into a yeast host strain. The yeast is then able to recognize the two stretches of sequence homology between the vector and target gene fragment and facilitate a reciprocal exchange of DNA content through homologous recombination at the gap. As a consequence, the target gene fragment is inserted into the vector without ligation.

The method described above has also been demonstrated to work when the target gene fragments are in the form of single stranded DNA, for example, as a circular M13 phage derived form, or as single stranded oligonucleotides (Simon and Moore, Mol. Cell Biol., 1987, 7: 2329; Ivanov et al., Genetics, 1996, 142: 693; and DeMarini et al., 2001, 30: 520, each incorporated by reference in its entirety). Thus, the form of the target that can be recombined into the gapped vector can be double stranded or single stranded, and derived from chemical synthesis, PCR, restriction digestion, or other methods.

Several factors may influence the efficiency of homologous recombination in yeast. For example, the efficiency of the gap repair is correlated with the length of the homologous sequences flanking both the linearized vector and the target gene. In certain embodiments, about 20 or more base pairs may be used for the length of the homologous sequence, and about 80 base pairs may give a near-optimized result (Hua et al., Plasmid, 1997, 38: 91; Raymond et al., Genome Res., 2002, 12: 190, each incorporated by reference in its entirety). In certain embodiments of the invention, at least about 5, 10, 15, 20, 21, 22, 23, 24, 25, 26, 27, 28, 29, 30, 31, 32, 33, 34 35, 36, 37, 38, 39, 40, 45, 50, 55, 60, 65, 70, 75, 80, 85, 90, 95, 100, 110, 120, 130, 140, 150, 160, 170, 180, 187, 190, or 200 homologous base pairs may be used to facilitate recombination. In certain embodiments, between about 20 and about 40 base pairs are utilized. In addition, the reciprocal exchange between the vector and gene fragment is strictly sequence-dependent, i.e. it does not cause a frame shift. Therefore, gap-repair cloning assures the insertion of gene fragments with both high efficiency and precision. The high efficiency makes it possible to clone two, three, or more targeted gene fragments simultaneously into the same vector in one transformation attempt (Raymond et al., Biotechniques, 1999, 26: 134, incorporated by reference in its entirety). Moreover, the nature of precision sequence conservation through homologous recombination makes it possible to clone selected genes or gene fragments into expression or fusion vectors for direct functional examination (El-Deiry et al., Nature Genetics, 1992, 1: 4549; Ishioka et al., PNAS, 1997, 94: 2449, each incorporated by reference in its entirety).

Libraries of gene fragments have also been constructed in yeast using homologous recombination. For example, a human brain cDNA library was constructed as a two-hybrid fusion library in vector pJG4-5 (Guidotti and Zervos, Yeast, 1999, 15: 715, incorporated by reference in its entirety). It has also been reported that a total of 6,000 pairs of PCR primers were used for amplification of 6,000 known yeast ORFs for a study of yeast genomic protein interactions (Hudson et al., Genome Res., 1997, 7: 1169, incorporated by reference in its entirety). In 2000, Uetz et al. conducted a comprehensive analysis-of protein-protein interactions in Saccharomyces cerevisiae (Uetz et al., Nature, 2000, 403: 623, incorporated by reference in its entirety). The protein-protein interaction map of the budding yeast was studied by using a comprehensive system to examine two-hybrid interactions in all possible combinations between the yeast proteins (Ito et al., PNAS, 2000, 97: 1143, incorporated by reference in its entirety), and the genomic protein linkage map of Vaccinia virus was studied using this system (McCraith et al., PNAS, 2000, 97: 4879, incorporated by reference in its entirety).

In certain embodiments of the invention, a synthetic CDR3 (heavy or light chain) may be joined by homologous recombination with a vector encoding a heavy or light chain chassis, a portion of FRM4, and a constant region, to form a full-length heavy or light chain. In certain embodiments of the invention, the homologous recombination is performed directly in yeast cells. In some embodiments, such a method comprises:

(a) transforming into yeast cells:

(i) a linearized vector encoding a heavy or light chain chassis, a portion of FRM4, and a constant region, wherein the site of linearization is between the end of FRM3 of the chassis and the beginning of the constant region; and (ii) a library of CDR3 insert nucleotide sequences that are linear and double stranded, wherein each of the CDR3 insert sequences comprises a nucleotide sequence encoding CDR3 and 5′- and 3′-flanking sequences that are sufficiently homologous to the termini of the vector of (i) at the site of linearization to enable homologous recombination to occur between the vector and the library of CDR3 insert sequences; and

(b) allowing homologous recombination to occur between the vector and the CDR3 insert sequences in the transformed yeast cells, such that the CDR3 insert sequences are incorporated into the vector, to produce a vector encoding full-length heavy chain or light chain.

As specified above, CDR3 inserts may have a 5′ flanking sequence and a 3′ flanking sequence that are homologous to the termini of the linearized vector. When the CDR3 inserts and the linearized vectors are introduced into a host cell, for example, a yeast cell, the “gap” (the linearization site) created by linearization of the vector is filled by the CDR3 fragment insert through recombination of the homologous sequences at the 5′ and 3′ termini of these two linear double-stranded DNAs (i.e., the vector and the insert). Through this event of homologous recombination, libraries of circular vectors encoding full-length heavy or light chains comprising variable CDR3 inserts is generated. Particular instances of these methods are presented in the Examples.

Subsequent analysis may be carried out to determine, for example, the efficiency of homologous recombination that results in correct insertion of the CDR3 sequences into the vectors. For example, PCR amplification of the CDR3 inserts directly from selected yeast clones may reveal how many clones are recombinant. In certain embodiments, libraries with minimum of about 90% recombinant clones are utilized. In certain embodiments libraries with a minimum of about 1%, 5% 10%, 15%, 20%, 25%, 30%, 35%, 40%, 45%, 50%, 55%, 60%, 65%, 70%, 75%, 80%, 85%, 86%, 87%, 88%, 89%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98%, or 99% recombinant clones are utilized. The same PCR amplification of selected clones may also reveal the insert size.

To verify the sequence diversity of the inserts in the selected clones, a PCR amplification product with the correct size of insert may be “fingerprinted” with restriction enzymes known to cut or not cut within the amplified region. From a gel electrophoresis pattern, it may be determined whether the clones analyzed are of the same identity or of the distinct or diversified identity. The PCR products may also be sequenced directly to reveal the identity of inserts and the fidelity of the cloning procedure, and to prove the independence and diversity of the clones.

Expression and Screening Systems

Libraries of polynucleotides generated by any of the techniques described herein, or other suitable techniques, can be expressed and screened to identify antibodies having desired structure and/or activity. Expression of the antibodies can be carried out, for example, using cell-free extracts (and e.g., ribosome display), phage display, prokaryotic cells (e.g., bacterial display), or eukaryotic cells (e.g., yeast display). In certain embodiments of the invention, the antibody libraries are expressed in yeast.

In some embodiments, polynucleotides are engineered to serve as templates that can be expressed in a cell-free extract. Vectors and extracts as described, for example in U.S. Pat. Nos. 5,324,637; 5,492,817; 5,665,563, (each incorporated by reference in its entirety) can be used and many are commercially available. Ribosome display and other cell-free techniques for linking a polynucleotide (i.e., a genotype) to a polypeptide (i.e., a phenotype) can be used, e.g., Profusion™ (see, e.g., U.S. Pat. Nos. 6,348,315; 6,261,804; 6,258,558; and 6,214,553, each incorporated by reference in its entirety).

Alternatively or additionally, polynucleotides of the invention can be expressed in an E. coli expression system, such as that described by Pluckthun and Skerra. (Meth. Enzymol., 1989, 178: 476; Biotechnology, 1991, 9: 273, each incorporated by reference in its entirety). Mutant proteins can be expressed for secretion in the medium and/or in the cytoplasm of the bacteria, as described by Better and Horwitz, Meth. Enzymol., 1989, 178: 476, incorporated by reference in its entirety. In some embodiments, the single domains encoding VH and VL are each attached to the 3′ end of a sequence encoding a signal sequence, such as the ompA, phoA or pelB signal sequence (Lei et al., J. Bacteriol., 1987, 169: 4379, incorporated by reference in its entirety). These gene fusions are assembled in a dicistronic construct, so that they can be expressed from a single vector, and secreted into the periplasmic space of E. coli where they will refold and can be recovered in active form. (Skerra et al., Biotechnology, 1991, 9: 273, incorporated by reference in its entirety). For example, antibody heavy chain genes can be concurrently expressed with antibody light chain genes to produce antibodies or antibody fragments.

In some embodiments of the invention, antibody sequences are expressed on the membrane surface of a prokaryote, e.g., E. coli , using a secretion signal and lipidation moiety as described, e.g., in US2004/0072740; US2003/0100023; and US2003/0036092 (each incorporated by reference in its entirety).

Higher eukaryotic cells, such as mammalian cells, for example myeloma cells (e.g., NS/0 cells), hybridoma cells, Chinese hamster ovary (CHO), and human embryonic kidney (HEK) cells, can also be used for expression of the antibodies of the invention. Typically, antibodies expressed in mammalian cells are designed to be secreted into the culture medium, or expressed on the surface of the cell. Antibody or antibody fragments can be produced, for example, as intact antibody molecules or as individual VH and VL fragments, Fab fragments, single domains, or as single chains (scFv) (Huston et al., PNAS, 1988, 85: 5879, incorporated by reference in its entirety).

Alternatively or additionally, antibodies can be expressed and screened by anchored periplasmic expression (APEx 2-hybrid surface display), as described, for example, in Jeong et al., PNAS, 2007, 104: 8247 (incorporated by reference in its entirety) or by other anchoring methods as described, for example, in Mazor et al., Nature Biotechnology, 2007, 25: 563 (incorporated by reference in its entirety).

In some embodiments of the invention, antibodies can be selected using mammalian cell display (Ho et al., PNAS, 2006, 103: 9637, incorporated by reference in its entirety).

Screening of the antibodies derived from the libraries of the invention can be carried out by any appropriate means. For example, binding activity can be evaluated by standard immunoassay and/or affinity chromatography. Screening of antibodies of the invention for catalytic function, e.g., proteolytic function can be accomplished using a standard assays, e.g., the hemoglobin plaque assay as described in U.S. Pat. No. 5,798,208 (incorporated by reference in its entirety). Determining the ability of candidate antibodies to bind therapeutic targets can be assayed in vitro using, e.g., a BIACORE™ instrument, which measures binding rates of an antibody to a given target or antigen based on surface plasmon resonance. In vivo assays can be conducted using any of a number of animal models and then subsequently tested, as appropriate, in humans. Cell-based biological assays are also contemplated.

One feature of the instant invention is the speed at which the antibodies of the library can be expressed and screened. In certain embodiments of the invention, the antibody library can be expressed in yeast, which have a doubling time of less than about 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, or 24 hours. In some embodiments, the doubling times are about 1 to about 3 hours, about 2 to about 4, about 3 to about 8 hours, about 3 to about 24, about 5 to about 24, about 4 to about 6 about 5 to about 22, about 6 to about 8, about 7 to about 22, about 8 to about 10 hours, about 7 to about 20, about 9 to about 20, about 9 to about 18, about 11 to about 18, about 11 to about 16, about 13 to about 16, about 16 to about 20, or about 20 to about 30 hours. In certain embodiments of the invention, an antibody library is expressed in yeast with a doubling time of about 16 to about 20 hours, about 8 to about 16 hours, or about 4 to about 8 hours. Thus, an antibody library of the instant invention can be expressed and screened in a matter of hours, as compared to previously known techniques which take several days to express and screen antibody libraries. A limiting step in the throughput of such screening processes in mammalian cells is typically the time required to iteratively regrow populations of isolated cells, which, in some cases, have doubling times greater than the doubling times of the yeast used in the current invention.

In certain embodiments of the invention, the composition of a library may be defined after one or more enrichment steps (for example by screening for antigen binding, binding to a generic ligand, or other properties). For example, a library with a composition comprising about x % sequences or libraries of the invention may be enriched to contain about 2x %, 3x %, 4x %, 5x %, 6x %, 7x %, 8x %, 9x %, 10x %, 20x %, 25x %, 40x %, 50x %, 60x % 75x %, 80x %, 90x %, 95x %, or 99x % sequences or libraries of the invention, after one or more screening steps. In some embodiments of the invention, the sequences or libraries of the invention may be enriched about 2-fold, 3-fold, 4-fold, 5-fold, 6-fold, 7-fold, 8-fold, 9-fold, 10-fold, 100-fold, 1,000-fold, or more, relative to their occurrence prior to the one or more enrichment steps. In certain embodiments of the invention, a library may contain at least a certain number of a particular type of sequence(s), such as CDRH3s, CDRL3s, heavy chains, light chains, or whole antibodies (e.g., at least about 10 3 , 10 4 , 10 5 , 10 6 , 10 7 , 10 8 , 10 9 , 10 10 , 10 11 , 10 12 , 10 13 , 10 14 , 10 15 , 10 16 , 10 17 , 10 18 , 10 19 , or 10 20 ). In certain embodiments, these sequences may be enriched during one or more enrichment steps, to provide libraries comprising at least about 10 2 , 10 3 , 10 4 , 10 5 , 10 6 , 10 7 , 10 8 , 10 9 , 10 10 , 10 11 , 10 12 , 10 13 , 10 14 , 10 15 , 10 16 , 10 17 , 10 18 , or 10 19 of the respective sequence(s).

Mutagenesis Approaches for Affinity Maturation

As described above, antibody leads can be identified through a selection process that involves screening the antibodies of a library of the invention for binding to one or more antigens, or for a biological activity. Coding sequences of these antibody leads may be further mutagenized in vitro or in vivo to generate secondary libraries with diversity introduced in the context of the initial antibody leads. Such mutagenized antibody leads can then be further screened for binding to target antigens or biological activity, in vitro or in vivo, following procedures similar to those used for the selection of the initial antibody lead from the primary library. Such mutagenesis and selection of primary antibody leads effectively mimics the affinity maturation process naturally occurring in a mammal that produces antibodies with progressive increases in the affinity to an antigen.

In some embodiments of the invention, only the CDRH3 region is mutagenized. In some embodiments of the invention, the whole variable region is mutagenized. In some embodiments of the invention one or more of CDRH1, CDRH2, CDRH3, CDRL1, CDRL2, and/ CDRL3 may be mutagenized. In some embodiments of the invention, “light chain shuffling” may be used as part of the affinity maturation protocol. In certain embodiments, this may involve pairing one or more heavy chains with a number of light chains, to select light chains that enhance the affinity and/or biological activity of an antibody. In certain embodiments of the invention, the number of light chains to which the one or more heavy chains can be paired is at least about 2, 5, 10, 100, 10 3 , 10 4 , 10 5 , 10 6 , 10 7 , 10 8 , 10 9 , or 10 10 . In certain embodiments of the invention, these light chains are encoded by plasmids. In some embodiments of the invention, the light chains may be integrated into the genome of the host cell.

Coding sequences of antibody leads may be mutagenized using any of wide variety of methods. Examples of methods of mutagenesis include, but are not limited to site-directed mutagenesis, error-prone PCR mutagenesis, cassette mutagenesis, and random PCR mutagenesis. Alternatively or additionally, oligonucleotides encoding regions with the desired mutations can be synthesized and introduced into the sequence to be mutagenized, for example, via recombination or ligation.

Site-directed mutagenesis or point mutagenesis may be used to gradually change the CDR sequences in specific regions. For example, this may be accomplished by using oligonucleotide-directed mutagenesis or PCR. For example, a short sequence of an antibody lead may be replaced with a synthetically mutagenized oligonucleotide in either the heavy chain or light chain region, or both. Such a method may not be efficient for mutagenizing large numbers of CDR sequences, but may be used for fine tuning of a particular lead to achieve higher affinity toward a specific target protein.

Cassette mutagenesis may alternatively or additionally be used to mutagenize the CDR sequences in specific regions. In a typical cassette mutagenesis, a sequence block, or a region, of a single template is replaced by a completely or partially randomized sequence. However, the maximum information content that can be obtained may be statistically limited by the number of random sequences of the oligonucleotides. Similar to point mutagenesis, this method may also be used for fine tuning of a particular lead to achieve higher affinity towards a specific target protein.

Error-prone PCR, or “poison” PCR, may be used to mutagenize the CDR sequences, for example, by following protocols described in U.S. Pat. No. 6,153,745; Caldwell and Joyce, PCR Methods and Applications, 1992, 2: 28; Leung et al., Technique, 1989, 1: 11; Shafikhani et al., Biotechniques, 1997, 23: 304; and Stemmer et al., PNAS, 1994, 91: 10747 (each of which is incorporated by reference in its entirety).

Conditions for error prone PCR may include, for example, (a) high concentrations of Mn 2+ (e.g., about 0.4 to about 0.6 mM) that efficiently induces malfunction of Taq DNA polymerase; and/or (b) a disproportionally high concentration of one nucleotide substrate (e.g., dGTP) in the PCR reaction that causes incorrect incorporation of this high concentration substrate into the template and produces mutations. Alternatively or additionally, other factors such as, the number of PCR cycles, the species of DNA polymerase used, and the length of the template, may affect the rate of misincorporation of “wrong” nucleotides into the PCR product. Commercially available kits may be utilized for the mutagenesis of the selected antibody library, such as the “Diversity PCR random mutagenesis kit” (CLONTECH™).

Primer pairs used in PCR-based mutagenesis may, in certain embodiments, include regions matched with the homologous recombination sites in the expression vectors. Such a design allows facile re-introduction of the PCR products back into the heavy or light chain chassis vectors, after mutagenesis, via homologous recombination.

Other PCR-based mutagenesis methods can also be used, alone or in conjunction with the error prone PCR described above. For example, the PCR amplified CDR segments may be digested with DNase to create nicks in the double stranded DNA. These nicks can be expanded into gaps by other exonucleases such as Bal 31. Gaps may then be filled by random sequences by using DNA Klenow polymerase at a low concentration of regular substrates dGTP, dATP, dTTP, and dCTP with one substrate (e.g., dGTP) at a disproportionately high concentration. This fill-in reaction should produce high frequency mutations in the filled gap regions. Such methods of DNase digestion may be used in conjunction with error prone PCR to create a high frequency of mutations in the desired CDR segments.

CDR or antibody segments amplified from the primary antibody leads may also be mutagenized in vivo by exploiting the inherent ability of mutation in pre-B cells. The Ig genes in pre-B cells are specifically susceptible to a high-rate of mutation. The Ig promoter and enhancer facilitate such high rate mutations in a pre-B cell environment while the pre-B cells proliferate. Accordingly, CDR gene segments may be cloned into a mammalian expression vector that contains a human Ig enhancer and promoter. Such a construct may be introduced into a pre-B cell line, such as 38B9, which allows the mutation of the VH and VL gene segments naturally in the pre-B cells (Liu and Van Ness, Mol. Immunol., 1999, 36: 461, incorporated by reference in its entirety). The mutagenized CDR segments can be amplified from the cultured pre-B cell line and re-introduced back into the chassis-containing vector(s) via, for example, homologous recombination.

In some embodiments, a CDR “hit” isolated from screening the library can be re-synthesized, for example using degenerate codons or trinucleotides, and re-cloned into the heavy or light chain vector using gap repair.

Other Variants of Polynucleotide Sequences of the Invention

In certain embodiments, the invention provides a polynucleotide that hybridizes with a polynucleotide taught herein, or that hybridizes with the complement of a polynucleotide taught herein. For example, an isolated polynucleotide that remains hybridized after hybridization and washing under low, medium, or high stringency conditions to a polynucleotide taught herein or the complement of a polynucleotide taught herein is encompassed by the present invention.

Exemplary low stringency conditions include hybridization with a buffer solution of about 30% to about 35% formamide, about 1 M NaCl, about 1% SDS (sodium dodecyl sulphate) at about 37° C., and a wash in about 1× to about 2×SSC (20×SSC=3.0 M NaCl/0.3 M trisodium citrate) at about 50° C. to about 55° C.

Exemplary moderate stringency conditions include hybridization in about 40% to about 45% formamide, about 1 M NaCl, about 1% SDS at about 37° C., and a wash in about 0.5× to about 1×SSC at abut 55° C. to about 60° C.

Exemplary high stringency conditions include hybridization in about 50% formamide, about 1 M NaCl, about 1% SDS at about 37° C., and a wash in about 0.1×SSC at about 60° C. to about 65° C.

Optionally, wash buffers may comprise about 0.1% to about 1% SDS.

The duration of hybridization is generally less than about 24 hours, usually about 4 to about 12 hours.

Sublibraries and Larger Libraries Comprising Libraries or Sub-Libraries of the Invention

Libraries comprising combinations of the libraries described herein (e.g., CDRH3 and CDRL3 libraries) are encompassed by the invention. Sublibraries comprising portions of the libraries described herein are also encompassed by the invention (e.g., a CDRH3 library in a particular heavy chain chassis or a sub-set of the CDRH3 libraries, for example based on length).

Moreover, libraries containing one of the libraries or sublibraries of the invention also fall within the scope of the invention. For example, in certain embodiments of the invention, one or more libraries or sublibraries of the invention may be contained within a larger library (theoretical or physical), which may include sequences derived by other means, for example, non-human or human sequence derived by stochastic or sitewise-stochastic synthesis. In certain embodiments of the invention, at least about 1% of the sequences in a polynucleotide library may be those of the invention (e.g., CDRH3 sequences, CDRL3 sequences, VH sequences, VL sequences), regardless of the composition of the other 99% of sequences. For the purposes of illustration only, one of ordinary skill in the art would readily recognize that a library containing 10 9 total members, where 10 7 members are members of the libraries of the invention (i.e., 1%) would have utility, and that members of the libraries of the invention could be isolated from such a library. In some embodiments of the invention, at least about 0.001%, 0.01%, 0.1%, 2%, 5%, 10%, 20%, 30%, 40%, 50%, 60%, 70%, 80%, 81%, 82%, 83%, 84%, 85%, 86%, 87%, 88%, 89%, 90%, 91,%, 92%, 93%, 94%, 95%, 96%, 97%, 98% or 99% of the sequences in any polynucleotide library may be those of the invention, regardless of the composition of the other sequences. In some embodiments, the sequences of the invention may comprise about 0.001% to about 1%, about 1% to about 2%, about 2% to about 5%, about 5% to about 10%, about 10% to about 15%, about 15% to about 20%, about 20% to about 25%, about 25% to about 30%, about 30% to about 35%, about 35% to about 40%, about 40% to about 45%, about 45% to about 50%, about 50% to about 55%, about 55% to about 60%, about 60% to about 65%, about 65% to about 70%, about 70% to about 75%, about 75% to about 80%, about 80% to about 85%, about 85% to about 90%, about 90% to about 95%, or about 95% to about 99% of the sequences in any polynucleotide library, regardless of the composition of the other sequences. Thus, libraries more diverse than one or more libraries or sublibraries of the invention, but yet still comprising one or more libraries or sublibraries of the invention, in an amount in which the one or more libraries or sublibraries of the invention can be effectively screened and from which sequences encoded by the one or more libraries or sublibraries of the invention can be isolated, also fall within the scope of the invention.

Alternative Scaffolds

As would be evident to one of ordinary skill in the art, the CDRH3 and/or CDRL3 polypeptides provided by the invention may also be displayed on alternative scaffolds. Several such scaffolds have been shown to yield molecules with specificities and affinities that rival those of antibodies. Exemplary alternative scaffolds include those derived from fibronectin (e.g., AdNectin), the β-sandwich (e.g., iMab), lipocalin (e.g., Anticalin), EETI-II/AGRP, BPTI/LACI-D1/ITI-D2 (e.g., Kunitz domain), thioredoxin (e.g., peptide aptamer), protein A (e.g., Affibody), ankyrin repeats (e.g., DARPin), γB-crystallin/ubiquitin (e.g., Affilin), CTLD 3 (e.g., Tetranectin), and (LDLR-A module) 3 (e.g., Avimers). Additional information on alternative scaffolds is provided, for example, in Binz et al., Nat. Biotechnol., 2005 23: 1257 and Skerra, Current Opin. in Biotech., 2007 18: 295-304, each of which is incorporated by reference in its entirety.

Additional Embodiments of the Invention

Library Sizes

In some embodiments of the invention, a library comprises about 10 1 to about 10 20 different polynucleotide or polypeptide sequences (encoding or comprising e.g., antibodies, heavy chains, CDRH3s, light chains, and/or CDRL3s). In some embodiments, the libraries of the invention are designed to include at least about 10 1 , 10 2 , 10 3 , 10 4 , 10 5 , 10 6 , 10 7 , 10 8 , 10 9 , 10 10 , 10 11 , 10 12 , 10 13 , 10 14 , 10 15 , 10 16 , 10 17 , 10 18 , 10 19 , or 10 20 , or more different antibody, heavy chain, CDRH3, light chain, and/or CDRL3 polynucleotide or polypeptide sequences. In some embodiments, a library of the invention may contain less than a particular number of polynucleotide or polypeptide sequences, where the number of sequences is defined using any one of the integers provided above. In certain embodiments of the invention, a particular numerical range is defined, using any two of the integers provided above as lower and upper boundaries of the range, inclusive or exclusive. All combinations of the integers provided, which define an upper and lower boundary, are contemplated.

In some embodiments, the invention provides libraries wherein a fraction of the members of the library are members produced according to the methods, systems, and compositions provided herein. One important property of the libraries of the invention is that they favorably mimic certain aspects of the human preimmune repertoire, including length diversity and sequence diversity. One or ordinary skill in the art will readily recognize that libraries provided by the invention include libraries where a subset of the members of the library are members produced according to the methods, systems, and compositions provided herein. For example, a library containing 10 8 members wherein 10 6 members are produced according to the methods, systems, and compositions provided herein, would contain 1% sequences produced according to the methods, systems, and compositions provided herein. One of ordinary skill in the art would recognize that one or more of the 10 6 members could readily be isolated using screening techniques known in the art. Therefore, said libraries fall within the scope of the invention. More specifically, libraries comprising at least about 1%, 2.5%, 5%, 10%, 15%, 20%, 25%, 30%, 40%, 45%, 50%, 55%, 60%, 65%, 70%, 75%, 80%, 85%, 90%, 95%, or 99% CDRH3, CDRL3, light chain, or heavy chain, and/or full-length antibody sequences provided herein fall within the scope of the invention. Libraries comprising at least about 10 3 , 10 4 , 10 5 , 10 6 , 10 7 , 10 8 , 10 9 , 10 10 , 10 11 , 10 12 , 10 13 , 10 14 , 10 15 CDRH3, CDRL3, light chain, heavy chain, and/or full-length antibody sequences provided herein also fall within the scope of the invention.

Human Preimmune Set

In some embodiments, the invention comprises the set of 3,571 curated human preimmune antibody sequences contained within the HPS, their corresponding CDRH3 sequences (Appendix A), and/or a representation of these CDRH3 sequences (and/or TN1, DH, N2, and/or H3-JH segments thereof) in a computer readable format. In certain embodiments, the invention comprises a method of producing a CDRH3 library, the method comprising matching candidate segments (i.e., TN1, DH, N2, and H3-JH) from a theoretical segment pool with CDRH3 sequences in the HPS and/or any other repertoire of CDRH3 sequences. In some embodiments, the invention comprises the candidate segments from the theoretical segment pools disclosed herein and/or the segments selected for inclusion in a physical library.

Embodiments

While the methods described herein demonstrate the production of theoretical segment pools of H3-JH and DH segments using a limited number of allelic variants, one of ordinary skill in the art will recognize that methods taught herein may be applied to any IGHJ and IGHD genes, including any other allelic variants and all non-human IGHJ and IGHD genes. Alternatively or additionally, methods described herein may be applied to any reference set of CDRH3 sequences, for example to extract additional TN1 and/or N2 segments. Alternatively or additionally, one of ordinary skill in the art will recognize that each of the described embodiments of the invention may be in polynucleotide or polypeptide form, within a vector, virus, or microorganism (e.g., a yeast or bacteria). Furthermore, since the invention involves synthetic libraries that are fully enumerated, certain embodiments of the invention relate to any of the embodiments described above in a computer readable format, and uses thereof.

Non-human antibody libraries also fall within the scope of the invention.

The present disclosure describes the removal of sequences containing Cys residues, N-linked glycosylation motifs, deamidation motifs, and highly hydrophobic sequences from the libraries of the invention. One of ordinary skill in the art will recognize that one or more of these criteria (i.e., not necessarily all) can be applied to remove undesirable sequences from any library of the invention. However, libraries containing one or more of these types of sequences also fall within the scope of the invention. Other criteria can also be used; those described herein are not limiting.

In certain embodiments, the invention provides libraries in which the number of times a particular sequence is repeated within the library (either theoretical, synthetic, or physical realization) is limited. For example, in some embodiments, the invention provides libraries wherein the frequency of occurrence of any of the sequences in the library (e.g., CDRH3, CDRL3, heavy chain, light chain, full-length antibody) is less than about 2, 5, 10, 15, 20, 25, 30, 35, 40, 45, 50, 55, 60, 65, 70, 75, 80, 85, 90, 95, 100, 200, 300, 400, 500, 600, 700, 800, 900, or 1000. In some embodiments, the frequency of occurrence of any of the sequences in the library is less than a multiple of the frequency of occurrence of any other sequence in the library, for examples less than about 2, 5, 10, 15, 20, 25, 30, 35, 40, 45, 50, 55, 60, 65, 70, 75, 80, 85, 90, 95, 100, 200, 300, 400, 500, 600, 700, 800, 900, or 1000 times the frequency of occurrence of any other sequence in the library.

In some embodiments, libraries are defined by the combinatorial diversity of the segments used to produce CDRH3 sequences, in particular the number of non-degenerate segment combinations that can be used to produce a particular CDRH3 sequence. In some embodiments, this metric may be calculated using, for example, a sample of about 2000, 5000, 10000, 20000, 50000, 100000, or more sequences from the CDRH3 library and “self-matching” using the segments used to generate the CDRH3 sequences of that library. In certain embodiments, the invention provides libraries wherein at least about 95%, 90%, 85%, 80%, 75%, 70%, 65%, 60%, 55%, 50%, 45%, 40%, 35%, 30%, 25%, 20%, 15%, 10%, or 5% of the CDRH3 sequences in the library may be formed by a single combination of segments.

In certain embodiments of the invention, a statistical bootstrap analysis was used to generate CDRH3 reference sets. While it may be advantageous to use this method, it is not required for every embodiment of the invention.

In some embodiments, the invention provides methods and systems of selecting polynucleotides to encode polypeptides of the invention, comprising selecting polynucleotide segments lacking (or containing) certain restriction sites individually and/or after combinatorial concatenation with other segments (e.g., see Example 9.3.7).

The exemplary libraries provided herein are not limiting and provided for exemplification only.

This invention is further illustrated by the following examples which should not be construed as limiting. The contents of all references, patents and published patent applications cited throughout this application are hereby incorporated by reference.

In general, the practice of the present invention employs, unless otherwise indicated, conventional techniques of chemistry, molecular biology, recombinant DNA technology, PCR technology, immunology (especially, e.g., antibody technology), expression systems (e.g., yeast expression, cell-free expression, phage display, ribosome display, and PROFUSION™), and any necessary cell culture that are within the skill of the art and are explained in the literature. See, e.g., Sambrook, Fritsch and Maniatis, Molecular Cloning: Cold Spring Harbor Laboratory Press (1989); DNA Cloning , Vols. 1 and 2, (D. N. Glover, Ed. 1985); Oligonucleotide Synthesis (M. J. Gait, Ed. 1984); PCR Handbook Current Protocols in Nucleic Acid Chemistry , Beaucage, Ed. John Wiley & Sons (1999) (Editor); Oxford Handbook of Nucleic Acid Structure , Neidle, Ed., Oxford Univ Press (1999); PCR Protocols: A Guide to Methods and Applications , Innis et al., Academic Press (1990); PCR Essential Techniques: Essential Techniques , Burke, Ed., John Wiley & Son Ltd (1996); The PCR Technique: RT - PCR , Siebert, Ed., Eaton Pub. Co. (1998); Antibody Engineering Protocols ( Methods in Molecular Biology ), 510, Paul, S., Humana Pr (1996); Antibody Engineering: A Practical Approach ( Practical Approach Series, 169), McCafferty, Ed., Irl Pr (1996); Antibodies: A Laboratory Manual , Harlow et al., C.S.H.L. Press, Pub. (1999); Current Protocols in Molecular Biology , eds. Ausubel et al., John Wiley & Sons (1992); Large - Scale Mammalian Cell Culture Technology , Lubiniecki, A., Ed., Marcel Dekker, Pub., (1990); Phage Display: A Laboratory Manual , C. Barbas (Ed.), CSHL Press, (2001); Antibody Phage Display , P O'Brien (Ed.), Humana Press (2001); Border et al., Nature Biotechnology, 1997, 15: 553; Border et al., Methods Enzymol., 2000, 328: 430; ribosome display as described by Pluckthun et al. in U.S. Pat. No. 6,348,315, and Profusion™ as described by Szostak et al. in U.S. Pat. Nos. 6,258,558; 6,261,804; and 6,214,553; and bacterial periplasmic expression as described in US20040058403A1. Each of the references cited in this paragraph is incorporated by reference in its entirety.

Further details regarding antibody sequence analysis using Kabat conventions and programs to analyze aligned nucleotide and amino acid sequences may be found, e.g., in Johnson et al., Methods Mol. Biol., 2004, 248: 11; Johnson et al., Int. Immunol., 1998, 10: 1801; Johnson et al., Methods Mol. Biol., 1995, 51: 1; Wu et al., Proteins, 1993, 16: 1; and Martin, Proteins, 1996, 25: 130. Each of the references cited in this paragraph is incorporated by reference in its entirety.

Further details regarding antibody sequence analysis using Chothia conventions may be found, e.g., in Chothia et al., J. Mol. Biol., 1998, 278: 457; Morea et al., Biophys. Chem., 1997, 68: 9; Morea et al., J. Mol. Biol., 1998, 275: 269; Al-Lazikani et al., J. Mol. Biol., 1997, 273: 927. Bane et al., Nat. Struct. Biol., 1994, 1: 915; Chothia et al., J. Mol. Biol., 1992, 227: 799; Chothia et al., Nature, 1989, 342: 877; and Chothia et al., J. Mol. Biol., 1987, 196: 901. Further analysis of CDRH3 conformation may be found in Shirai et al., FEBS Lett., 1999, 455: 188 and Shirai et al., FEBS Lett., 1996, 399: 1. Further details regarding Chothia analysis are described, for example, in Chothia et al., Cold Spring Harb. Symp. Quant Biol., 1987, 52: 399. Each of the references cited in this paragraph is incorporated by reference in its entirety.

Further details regarding CDR contact considerations are described, for example, in MacCallum et al., J. Mol. Biol., 1996, 262: 732, incorporated by reference in its entirety.

Further details regarding the antibody sequences and databases referred to herein are found, e.g., in Tomlinson et al., J. Mol. Biol., 1992, 227: 776, VBASE2 (Retter et al., Nucleic Acids Res., 2005, 33: D671); BLAST (world wide web at ncbi.nlm.nih.gov/BLAST/); CDHIT (bioinformaticslj crfedu/cd-hi/); EMBOSS (world wide web at hgmp.mrc.ac.uk/Software/EMBOSS/); PHYLIP (evolution.genetics.washington.edu/phylip.html); and FASTA (fasta.bioch.virginia.edu). Each of the references cited in this paragraph is incorporated by reference in its entirety.

Light Chain Libraries

Those skilled in the art will recognize, or be able to ascertain using no more than routine experimentation, many equivalents to the specific embodiments and methods described herein. Such equivalents are intended to be encompassed by the scope of the following claims.