CROSS-REFERENCE TO RELATED APPLICATIONS

[01] This application claims benefit of priority to U.S. provisional patent application no. 62/586,800 filed on November 15, 2017, the content of which is herein incorporated by reference in its entirety.

REFERENCE TO SEQUENCE LISTING

[02] The official copy of the sequence listing is submitted electronically via EFS-Web as an ASCII formatted sequence listing with the file“PCT_sequences_WorkFile.txf’ created on November 14, 2018, filed on November 14, 2018 and having a size of 43 KB. The sequence listing contained in this ASCII formatted document forms part of the specification and is herein incorporated by reference in its entirety.

FIELD OF THE INVENTION

[03] The present invention relates to methods for preparing antibody libraries of improved functionality and relates to the design of antibody libraries of improved functionality and uses thereof.

BACKGROUND OF THE INVENTION

[04] Antibodies are immunoglobulins, or specialized immune proteins, having a heterodimeric structure. The antibody structure consists of two heavy chains and two light chains, folded into constant and variable domains. The variable domains of the heavy chains and light chains form the antigen binding site. Each variable region contains three hypervariable loops known as complementarity determining regions (CDRs) which alternate with less variable regions, called framework regions (FR). Monoclonal antibodies are antibodies that are made from identical immune cells that are all clones of one parent cell. Monoclonal antibodies possess unique characteristics such as specificity, affinity, potency, stability, solubility, and clinical tolerability.

[05] Hybridoma technology, also known as monoclonal antibody technology, is an efficient means to isolate single specificity antibodies and produce them in unlimited amounts (Kohler and Milstein, 1975).

This technology paved the way to generate antibodies for a diverse array of diagnostic and therapeutic applications (Reichert, 2017). In fact, antibody-based drugs represent the fastest-growing segment of all the therapeutic proteins in the biotechnology industry (Strohl, 2017). [06] Murom onab-CD3 (Orthoclone OKT3 ^® ) was the first United States Food and Drug

Administration (FDA) approved monoclonal antibody for use in therapeutic settings (Smith, 1996). The mouse monoclonal IgG2a antibody, which was developed by using hybridoma technology, blocks CD3- mediated activation of T cells and was instrumental in the prevention of organ rejection after transplantation (Emmons and Hunsicker, 1987). However, patients who were treated with Orthoclone OKT3 ^® developed a significant percentage of anti-drug antibodies (ADA). The development of ADA is also known as a“human anti-mouse antibody” (HAMA) response (Kimball et ah, 1993). The HAMA response leads to the inactivation and elimination of murine antibodies (Shawler et ah, 1985). Further, the HAMA response prevents the use of multiple administrations of an antibody, for example required for cancer therapy. Further complicating the use of murine monoclonal antibodies in human therapy is their association with the generation of severe allergic reactions. Together, such issues hamper the use of murine antibodies in human therapy (Shawler et al, 1985).

[07] To engineer more human-like antibodies and, thus, increase efficacy while decreasing immunogenicity, nonhuman variable (V) domains are combined with human constant (C) domains to generate molecules with 70% or more human content. This method is called chimerization and led to the approval of the first cancer-treating therapeutic antibody, Rituximab (Rituxan ^® ) (Morrison et al., 1984). Rituximab has been a tremendous medical and commercial success, currently being the fourth best-selling innovative drug of any kind (Strohl, 2017).

[08] Additional technology platforms emerged in parallel with Rituximab, which aimed to generate more human-like antibodies. Such technology platforms were perfected during the last three decades and include humanization (Jones et al., 1986), selection of fully human antibodies from Fv and Fab phage - displayed libraries (McCafferty et al., 1990) and the development of transgenic animals capable of generating fully human antibodies (Green et al., 1994; Lonberg et al., 1994).

[09] Humanized antibodies have more human content than chimeric antibodies but still do not eliminate the possibility of human anti-human antibody (HAHA) responses (Nechansky, 2010). Transgenic mice are capable of producing fully human antibodies. However, immunization does not always result in a successful in vivo antibody response with the desired level of affinity (Green and Jakobovits, 1998; Pruzina et al., 2011). This is particularly true for conserved epitopes between human and mouse orthologs. Transgenic rats and more recently, transgenic Omni Chicken, may partially mitigate this limitation (Ma et al, 2013 and Ching et al, 2017). However, toxic and unstable antigens and proteins with allosteric conformational changes are not well suited for an immunization approach and require an alternative solution for antibody discovery.

[ 10] Phage display technology connects proteins displayed at bacteriophage surface to their genes in the bacteriophage’s genome. Thus, phage display opened the possibility of designing and manipulating the repertoire of antibody genes to be used as source of antibodies in phage antibody libraries, hence enabling the selection of fully human antibodies (Hanes and Pluckthun, 1997; Francisco et al., 1993; Beerli et al., 2008; Cherf and Cochran, 2015; Finlay and Almagro, 2012). Moreover, these technology platforms allow selection for desired pre-defmed epitopes, thus avoiding immunodominant epitopes. Further, these technology platforms can also focus on specific conformations, rare cross-reactive epitopes, or conditions to select antibody variants with enhanced biophysical and biochemical properties.

[1 1] The concept that more diverse, functional, and larger libraries produce a larger number of specific antibodies with higher affinity antibodies is largely expected. One mathematic model has been proposed to formalize such a concept (Perelson and Oster, 1979). In this model, the probability (P) that an epitope is recognized by at least one antibody in a repertoire depends on the probability (p) that an antibody recognizes a random epitope with an affinity above a threshold value and on the size of the library (N) according to the equation P = l-e ^A (NP). This model predicts that the larger the size of the functional repertoire, the higher the chances of finding a specific and higher affinity antibody.

[12] In practice, it has been shown that a human antibody fragment phage display library of Ϊ0 members yields ~90 nM antibody fragments to four different proteins (Marks et al., 1991). Larger libraries, e.g., > 10 ¹⁰ (Vaughan et al., 1996; de Haard et al., 1999), have produced antibody fragments with single-digit nM or sub-nM affinities. Similarly, a library called Griffiths’ library (Griffiths et al.,

1994) of > 10° members produced sub-nM binders, whereas, antibody fragments of > 100 nM were obtained when a small portion of the library containing lCfclones was used in the selection process. Together, these results support the concept that larger libraries generate higher affinity antibodies.

[13] The theoretical number of unique antibody variants is virtually infinite. For instance, if the six CDRs that form the antigen-binding site were diversified with the 20 natural amino acids and it is considered that each CDR has on average seven positions to diversify, the corresponding theoretical number of unique variants is 20 ^{(6 x 7)} = 10 ⁵⁴ . However, only a very small fraction, which does not exceed 10 ¹¹ unique variants, can be cloned and displayed on the phage surface (Hoogenboom, 2005). More importantly, these 10 ¹¹ unique antibody variants should be functional to maximize the probability of producing antibodies with reasonable specificity and affinity, roughly in the low nM range.

[14] Several strategies have been described in the art to maximize the number of functional antibody variants in a phage display library. One typical strategy consists in cloning human antibody genes from natural sources, such as peripheral blood mononuclear cells (PBMCs) (de Haard et al., 1999). Other strategies are the construction of fully synthetic libraries, which are computationally designed and generated by chemical synthesis (Griffiths et al., 1994; Knappik et al., 2000; Shi et al., 2010), and semisynthetic libraries, which combine natural diversity and synthetic diversity (Hoet et al., 2005). [15] Antibody libraries composed of natural diversity, also known in the art as naive libraries, were the first generation of phage displayed antibody libraries (Marks et ah, 1991). Although successful, naive libraries included antibody genes toxic to E. coli and thus with low expression levels or no expression at all on the phage surface. These toxic genes severely compromised the number of functional antibodies displayed in the library. Synthetic antibody libraries followed and partially mitigated this limitation (Knappik et ah, 2000) (US patent 6,828,422). However, synthetic libraries must carefully be designed by making assumptions regarding the number of positions to diversify and type of amino acids to include in the design and proportion of each amino acid per position. These assumptions do not always hold true, particularly for CDR-H3, which is by far the most diverse region of the antigen-binding site and key in defining the specificity and affinity of antibodies. The structure of CDR-H1 and CDR-H2 can be predicted with an accuracy of < 1.0 A. This is a critical step in the successful design of the diversity of a synthetic library. However, no current method is available in the art to reliably predict the CDR-H3 structure (Almagro et al. 2014).

[16] The quality of the synthesis process also impacts the functionality of the library. Nucleotide sequences with stop codons leading to truncated sequences do not produce functional antibody fragments fused to a virion particle. Insertions or deletions of one or two nucleotides changing the reading frame of the antibody gene can generate stretches of amino acids that may impair folding or produce clones with hydrophobic amino acids resulting in aggregation. These non-functional antibody variants compromise the functionality of the library leading to poorly performing libraries and, overall, lower and often no production of antibodies with the desired specificity and affinity originally sought.

[17] The diversification methods used to generate synthetic libraries greatly influence the quality of the output antibodies. For instance, three libraries were designed for affinity maturation of anti- Oncostatin (anti-OSM) antibodies (US patent 8,580,714). The diversity of libraries was designed at positions frequently observed in contact with protein and peptide antigens (Raghunathan et al., 2012). The diversification regime, or amino acids and their frequencies per diversified position, were different in the three libraries. Two libraries were designed with few amino acids found in known antibody sequences. The third library was generated with random (NNK) codons, which produce the 20 amino acids plus one stop codon. However, the libraries designed with amino acids found in known antibody sequences yielded more diverse and higher affinity antibodies.

[18] Further, it has been realized that antibodies selected from naive, synthetic, and semisynthetic libraries often tend to fail during the development stages of formulation and manufacturing due to sub- optimal biophysical properties. This occurs despite possessing the desired specificity and affinity. Antibodies undergo posttranslational modifications (PTM) of some amino acids such as deamidation of asparagine (N), oxidation of methionine (M), and isomerization of aspartic acid (D) (Gilliland et al., 2012). These chemical modifications of the amino acids can result in heterogeneities in the antibody preparation and/or lack of potency if the amino acids are involved in the interaction with the antigen. In other instances, exposure of tryptophan (W) to the solvent can induce aggregation, thus leading to immunogenic reactions or lack of solubility at concentrations required for the therapeutic indication (Bethea et ah, 2012).

[19] Therefore, there has been a continued need in the art for making functional antibody libraries that produce antibodies amenable to clinical development and hence increase the probability of success along the preclinical and manufacturing processes (Urlinger et ah, US Patent 9,541,559). Computational methods and design principles, collectively called developability predictive methods (Kumar and Sigh, Developability of Biotherapeutics: Computational Approaches. CRC Press. 2015), have been perfected in the art to identify and remove residues that pose developability liabilities. It is critical to experimentally assess antibody variants as early as possible during the antibody discovery process. Developability predictive methods significantly improve the chances of successful antibody development, manufacturing, formulation, and stabilization to achieve the desire therapeutic effect.

[20] This invention discloses highly functional antibody libraries in addition to methods to make such highly functional antibody libraries. The exemplary antibody libraries of this invention combine highly stable and developable antibody variants with natural diversity in CDR-H3, as well as methods to remove non-productive antibody sequences such as out-of-frame variants and/or poorly folded antibodies due to imprecisions in the design, without compromising the functional size of the library.

SUMMARY OF THE INVENTION

[21] The present invention includes methods for producing highly functional antibody libraries, comprising the steps of: (1) designing and preparing primary antibody libraries (PLs); (2) applying a selection process (filtration) to enrich the PLs of step 1 with variants of improved functionality and/or developability to prepare intermediate filtered libraries (FLs); and (3) preparing highly functional secondary antibody libraries (SLs) by combining FLs with diverse CDR-H3 fragments.

[22] Two PLs were designed so that each of the PLs had one V _H scaffold binding to a Staphylococcus aureus Protein A. Two distinct V _L scaffolds were designed as counterpart of the V _H scaffold. One of the V _L scaffolds contained a short CDR-L1 loop (PL1). The other V _L scaffold contained a long CDR-L1 loop (PL2) (Figure 1). By changing the length of CDR-L1 from a short to a long loop, antibodies alter the preference to bind protein or peptide targets, respectively (Vargas-Madrazo et ah, 1995; Raghunathan et al., 2012). Therefore, by using the proper _L Vscaffold, the libraries of this invention can be used for selection of antibodies against protein or peptide targets. Selection of the two libraries, PL1 and PL2, would potentially generate antibodies that bind diverse types of epitopes on a given target.

[23] The scaffolds of the two PLs were diversified in positions that bind both protein and peptide targets and amino acids observed in human germline genes and natural antibodies. Amino acids associated with developability liabilities were avoided. Such amino acids included: (i) asparagine (N) followed by any amino acid but proline (XnoP) followed by serine (S) or threonine (S/T) [NXnoP(S/T)], which generates N-glycosylation sites; (ii) aspartic acid (D) followed by glycine (G) [DG], which tends to isomerize; (iii) asparagine (N) followed by glycine (G) or serine [NG/S], which tends to deamidate, (iv) exposed methionine (M), which tends to oxidize and (v) exposed tryptophan (W), which leads to aggregation spots.

[24] In one aspect of the invention, it was reasoned that by using only one CDR-H3 sequence to generate the PLs, the diversity of amino acids in contact with, or nearby, said CDR-H3 may be constrained to a few and specific residues to accommodate said CDR-H3 under specific selection conditions. Therefore, a set of CDR-H3 sequences called“neutral H3Js” was designed by starting from the repertoire of human D genes (IGDH) and combining them with the human heavy chain J genes (IGJH) in the germline configuration. Since antibodies from natural primary repertoires are mostly sequences in germline configuration, it was hypothesized that the set of neutral H3Js had enough diversity to avoid biases in amino acids at V _H and V _L in contact with, or nearby, the neutral H3J sequences, thus providing a diverse and favorable environment to select for developable diversified scaffolds and support cloning of natural CDR-H3 fragments in the third step described below.

[25] The designed V _H and V _L scaffolds were assembled as single chain Fv fragments (scFv) in a V _L - linker-V _H configuration and synthesized using trinucleotide phosphoramidites. Trinucleotide phosphoramidites synthesis, also known as trimer technology, is a type of synthesis that is based on synthetic codons instead of single nucleotides. Trinucleotide phosphoramidites synthesis generates precise combinations of amino acids at specifically targeted positions for diversifications while avoiding stop codons and unwanted amino acids which may disrupt the folding of the scaffolds used to generate the libraries. The quality control of the synthetic fragments was assessed via sequencing of 96 fragments in each library. The results indicated that 50% to 60% sequences were in-frame and matched the design.

[26] The synthetic fragments were then cloned in a phage display vector and displayed on the surface of M13 phage as fusion proteins to its minor coat protein pill to generate PL1 and PL2 following standard molecular biology protocols well known to those skilled in the art and described herein. Display on other platforms including yeast or related display technologies is a clear extension of the invention. A sample of individual clones chosen at random from PL1 and PL2 were submitted to Sanger sequencing. The percentage of in-frame sequences matching the design was 61.0% and 87.3%, for PL1 and PL2, respectively, which was in close agreement with the quality of the synthetic fragments.

[27] In a second aspect of the invention, there is provided the frequency of Protein A binders among the clones from the PLs displayed as scFvs after diverse incubation times and temperatures. It has been established that the least stable domain of the human IgGl is the _H E (Gilliland et al., 2012), which unfolds at 68°C. The question of whether a significant number of clones from the PLs were still stable at 68°C or above and hence, yielded developable antibodies when expressed as IgGl in a therapeutic antibody was addressed. The percentage of clones binding Protein A after incubation at 70°C for 10 min was 48.8% and 55.8% for PL1 and PL2, respectively, down from 58.5% and 69.8%, respectively in the PL1 and PL2. The difference of 9.7% and 14.0% less clones binding Protein A in PL1 and PL2, respectively, after heat shock indicated that some variants were unstable at 70°C.

[28] In a third aspect of the invention, PL1 and PL2 were incubated for 10 min at 70°C and well- folded and developable antibody fragments were rescued with Protein A, while unstable and non- developable antibody variants, which were either denatured or aggregated, were removed by a simple washing. Other harsh conditions include, but by no means are limited to, other temperatures and incubation times, high or low pH, high salt concentrations, and protease digestion. In one preferred embodiment, there exist counter-selections to remove antibodies with tendencies to aggregate, for example from the interactions of exposed hydrophobic residues.

[29] After incubation at 70°C for 10 min of PL1 and PL2 and rescue of well-folded antibody variants with Protein A, clones chosen at random from the filtered libraries (FLs) were sequenced and assayed for Protein A binding. The percentage of unique scFvs binding Protein A was around 90% in both libraries, 89.5% and 89.2%, respectively in FL1 and FL2, with virtually all the sequences being in-frame, 94.7% and 97.3%, respectively. Therefore, the functionality measured as the ability to bind Protein A was significantly improved by 31.0% and 19.4%, respectively in FL1 and FL2, after the heat-shock and filtering with Protein A.

[30] It should be noted that other natural ligands binding well-folded antibodies outside the antigen binding site have been described in the art. For example, Peptostreptococcus magnus Protein-L binds the V _L scaffold with the long CDR-L1 loop. The use of Protein-L as a ligand to select well-folded antibodies, alone and/or in conjunction with Protein A is a clear extension of this invention.

[31] Natural ligands that bind variable regions of antibodies, such as Protein A and Protein L, have extensively been used in the prior art to select for stable antibody domains after incubation under denaturing or destabilizing conditions. However, their use has been limited to enrich the final antibody library and/or select for variants after mutagenesis to improve stability. As shown herein, the selection with Protein A after submitting the libraries for harsh incubation conditions, removes a significant number of variants from the library. The decrease in the number of variants, and thus of the library size, depends on the library quality, scaffolds used to build the library, design of the library diversity, and selection conditions for stable and functional antibody variants. In the exemplary proofs of this invention, incubation at 70°C for 10 min led to a reduction of well-folded variants in PL1 and PL2 of 9.7% and

14.0% respectively. PL1 and PL2 are built with two human V _L scaffolds that belong to two different germline gene families and only share 68.3% identities. The kinetic of unfolding, as showed in the examples of this invention, is significantly different. However, in both cases, a consistent average decrease of -12% in Protein A binders represents a loss of 1.2 x 10 ⁹ unique and potentially functional antibody variants in a library containing 1 xlO ¹⁰ unique antibody variants. Therefore, improving stability after filtering came with the price of reducing the diversity of the libraries.

[32] In a forth aspect of the present invention, there is provided a method designed to restore the diversity of the library. In this step, the nucleotide sequences encoding the FLs were amplified by molecular biology techniques known to those skilled in the art and combined with diverse natural CDR-

H3 fragments, called“natural H3Js” to produce the SLs. It was reasoned that by replacing the neutral H3Js fragments in the FLs with natural H3Js, highly stable and functional libraries can be obtained. This rationale is supported by a substantial body of work indicating that CDR-H3 is key to determine the specificity and affinity of the antibodies and hence antibody libraries with highly diverse CDR-H3 fragments should increase the probability of obtaining diverse and high affinity antibodies. In the present invention, the natural H3 Js were isolated from a pool of 200 donors by molecular biology methods known to those skilled in the art. Amplification of diverse CDR-H3 fragments from other sources including CDR-H3 fragments obtained by synthetic means are clear extensions of this invention.

[33] Analysis of a sample of antibody variants chosen at random indicated that 68.3% and 75.6% in SL1 and SL2, respectively, bound Protein A when expressed as scFvs on the phage surface after incubation at 70°C for 10 min. Moreover, the sequence of all these clones had natural and unique CDR- H3 sequences. Comparatively to the PLs, this represents an increase of 19.5% and 19.8% in ability to survive a heat shock in SL1 and SL2, respectively. In relative numbers, these values translate in an average 36% to 40% increase of the stability of the SLs versus the PLs after incubation at 70°C for 10 min.

[34] It should be emphasized that it was not obvious from prior art that by combining the highly stable variants comprising the FLs with highly diverse natural CDR-H3 fragments, obtained from natural sources, which have not been under selection on harsh conditions such as incubation at 70°C for 10 min, the resulting SLs would retain the property of being highly stable. The selection of libraries of antibody fragments and antibody domains by Protein A had a profound effect on the diversity of the CDRs. Therefore, the gain in stability was at the expense of the library diversity. In the present invention, the combination of highly stable scaffolds used to build the libraries, a design based on positions found in contact with antigens and solvent exposed, germline gene diversity to diversify those positions, removal of liability developabilities, and the selection of PLs in conjunction with the set of neutral H3Js, led to a collection of highly stable variants suited to accommodate a collection of highly diverse natural H3J fragments, and hence, a highly functional antibody library.

[35] For each step of the construction, primary, filtrated and secondary libraries were analyzed by next generation sequencing (NGS) and established the conformity of the primary libraries to the intended design, a bias toward more hydrophilic sequences around CDR-H3 during the filtration process compensated by a diverse environment offered by the set of neutral HJ3 fragment, and the very high diversity of the secondary libraries. Analysis of Protein A binding from single clones provided the proof- of-concept that the overall filtration process resulted in stability improvement at the library level, while NGS analysis confirmed the retention of the in-frame character of the vast majority of the clones during the construction of the secondary libraries together with a boost in diversity.

[36] Finally, to demonstrate the potential of the SLs to produce specific and high affinity antibodies, pannings with two known antigen models have been performed. In one exemplary proof of this invention, the antigen model is Tumor Necrosis Factor (TNF). In another example, the antigen model is human serum albumin (HSA). After four rounds of selection against TNF and three for HSA, diverse and specific antibodies were obtained.

BRIEF DESCRIPTION OF THE DRAWINGS

[37] Those skilled in the art will recognize that the drawings described below are for illustrative purposes only. The drawings are not intended to limit the scope of the invention but to provide exemplary embodiments.

[38] Figure 1 provides a ribbon representation of the Y and V _L scaffolds. Figure 1A shows a drawing showing the ribbon representation of an Fv with a short Ll loop (PDB ID: 1ILC) [3-20/3-23]

Figure IB shows a drawing showing the ribbon representation of an Fv with a long CDR-L1 loop (PBD

ID: 1ILD) [4-01/3-23] These drawings show the ribbon representations of the Fvs used to build the semi-synthetic libraries of this invention. These Fvs have been solved by x-ray crystallography in a Fab configuration in association the CDR-H3 loop of a known therapeutic antibody (CNTO 888; (Obmolova et al, 2012). [39] Figure 2 provides a list of PDB ID complex structures used to design the primary repertoires.

This dataset of unique antibody: antigen complexes was obtained by starting from all the antibody structures compiled at the PDB and curated by the IMGT as of March ^l .l 2017. The initial dataset consisted of 2,645 antibody structures from diverse species and specificities. This initial data set was mined to extract the unique and well-solved antigen: antibody listed in the table.

[40] Figure 3 provides a table showing a diversification regime at CDR-H1 and CDR-H2 of the V _H scaffold. The residues in contact with antigens were determined in the set of structures listed in Figure 2 and mapped onto the structures of the V _H scaffold paired with the two V _L scaffolds PDB IDs: 1ILC and 1ILD. A total of 10 positions were targeted for diversification, four in the CDR-H1 and six in the CDR- H2. The diversification regime was designed using three sources of information: the dataset of curated antibody structures listed in Figure 2, the V _H sequences available at NCBI, and the germline genes of the human IGHV3 family compiled at the IMGT. The estimated number of amino acids per position is listed in the last column of the figure and yields a diversity of 5.4 x lO ⁵ unique amino acid V _H sequences.

[41] Figure 4 shows the diversification regime of the 3-20 ^scaffold. To identify positions for diversification and the diversification regime the same procedure than that for the scaffold was followed. The estimated number of amino acids per position is listed in the last column of the figure and yields a diversity of 1.1 x 10 ⁶ unique amino acid V _L sequences.

[42] Figure 5 shows the diversification regime of the 4-01 ^scaffold. To identify positions for diversification and the diversification regime the same procedure than that for the V scaffold was followed. The estimated number of amino acids per position is listed in the last column of the figure and yields a diversity of 1.3 x 10 ⁶ unique amino acid V _L sequences.

[43] Figure 6 provides the configuration of the two primary scFv repertoires, diversified regions and cloning sites. V _L and V _H are linked to form a scFv by the repetitive stretch of amino acids GS19. The V _L -linker-V _H configuration places the H3J fragment on the C-terminal side. Two Bgll/Sfil sites located on each side of the construct allow for cloning into the acceptor vector. Ncol and Kpnl sites are common to both repertoires, same for the JK1 anchor in the light chain sequence and the human VH Conserved Motif (CM) in the heavy chain sequence.

[44] Figure 7 provides a graph showing phage binding to Protein A by ELISA. The y-axis shows the optical density measurements at 490 nm and the x-axis shows dilutions in virions/ml for the 3-20/3-23 control scFv, the 4-1/3-23 control scFv, the 3-20/3-23 library, and the 4-01/3-23 library. COSTAR plates 3369 (Coming) were coated with Protein A (Sigma Aldrich, cat# P6031) at 4 pg/ml in TBS overnight at 4°C. After blocking with TBS with 0.1% Tween ^® 20 v/v (TBST) and 5% w/v nonfat dry milk for one hour, virions (~2.6 x 10 ¹² virions/ml) in TBST with 5% w/v nonfat dry milk were added to the wells, 2- fold serially diluted and incubated for 2 h at 37°C. As a reference, virions derived from the parent scaffolds with the CDR-H3 of CNTO 888 cloned in the same vector were added and similarly diluted on the plate. Bound phage was detected with A4G1.6 monoclonal antibody (Antibody Design Labs, San Diego, CA) conjugated to HRP. Binding of the secondary antibody, a murine IgGl, to Protein A was blocked by polyclonal human IgG at 100 pg/ml added to the incubation buffer.

[45] Figure 8 provides a graph showing, on the y-axis, optical density measurements at 490 nm taken at different temperature points during heat shock showing the thermal unfolding of the 4-01/3-23 and 3- 20/3-23 control scFvs. The 3-20/3-23 and 4-01/3-23 control scFvs displayed as fusion proteins to pill on the phage surface were incubated for 10 min over a range of temperatures, starting at 40°C and increasing the temperature up to 80°C in steps of 5°C. The unfolding process is monitored with a direct Protein A ELISA shown in Figure 7.

[46] Figure 9 provides a graph showing, on the y-axis, optical density measurements at 490 nm taken at different time points during heat shock showing the thermal unfolding of the 4-01/3-23 and 3-20/3-23 control scFvs at 60°C and 72°C. The 3-20/3-23 scFv and 4-01/3-23 scFv were displayed as fusion proteins to pill on the phage surface and the unfolding process is monitored with a direct Protein A ELISA shown in Figure 7.

[47] Figure 10 provides a bar graph showing different Protein A binding relative to control scFvs at 37°C. The binding shown is measuring PL1 single clone scFv-phages binding to Protein A relative to control scFv. A sample of 44 ampicillin-resistant colonies was chosen at random from PL1. The phage displaying scFvs were prepared in 96 deep-well plates and, after pelleting the bacteria by centrifugation,

100 pi aliquots were incubated for 10 min at 37°C in a 96-well PCR plate. After cooling down the samples for 30 min on ice, binding to Protein A shown in Figure 7 was performed. In-frame clones are indicated by the sign‘=’ and double transformants by the sign

[48] Figure 11 provides a bar graph showing different Protein A binding relative to control scFvs at 37°C. The binding shown is measuring PL2 single clone scFv-phages binding to Protein A relative to control scFv. A sample of 47 ampicillin-resistant colonies was selected from the PL2. Three clones, either with the parent vector or a partial insert were not further studied. The resulting 44 clones were analyzed as described in Figure 10. In-frame clones are indicated by the sign ‘=’ and double transformants by the sign

[49] Figure 12 provides a bar graph showing residual Protein A binding of PL1 scFv-phage single clones following 10 min at 70°C. The binding shown is measuring PL1 single clone scFv-phages binding to Protein A relative to control scFv. The 44 clones selected from PL1 were incubated for 10 min at 70°C as previously described in Figure 10. In-frame clones are indicated by the sign‘=’ and double transformants by the sign [50] Figure 13 provides a bar graph showing residual Protein A binding of PL2 scFv-phage single clones following 10 min at 70°C. The binding shown is measuring PL2 single clone scFv-phages binding to Protein A relative to control scFv. The 44 clones selected from PL2 were incubated for 10 min at

70°C as previously described in Figure 10. In-frame clones are indicated by the sign‘=’ and double transformants by the sign‘#\

[51] Figure 14 provides a bar graph showing Protein A binding relative to control scFvs after incubation at 70°C for 10 min and Protein A filtering. The binding shown is measuring FL1 single clone scFv-phages binding to Protein A relative to control scFv. A sample of 44 ampicillin-resistant colonies was chosen at random from FL1. Phage were prepared and assayed on Protein A as described in Figure 10. In-frame clones are indicated by the sign‘=’ and double transformants by the sign‘#\

[52] Figure 15 provides a bar graph showing Protein A binding relative to control scFvs after incubation at 70°C for 10 min and Protein A filtering. The binding shown is measuring FL2 single clone scFv-phage binding to Protein A relative to control scFv. A sample of 44 ampicillin-resistant colonies was selected at random from FL2. The phages were prepared and assayed on Protein A as described in

Figure 10. In-frame clones are indicated by the sign‘=’ and double transformants by the sign‘#\

[53] Figure 16 provides a drawing depicting a strategy for assembling seamlessly semisynthetic secondary repertoires. The top line shows a strategy for the filtrated primary library scaffold amplification, the middle line shows a strategy for an assembly with natural H3J fragments, and the bottom line of Figure 16 shows the final product. The diversified scaffold fragments from the filtrated primary libraries were amplified by a pair of primer located in the pelB leader and 5’ of the human VH Consensus Motif (CM). Two Bsal sites added at the end of the C-terminal of the Fv and at the beginning of the natural H3J fragments are used to assemble the complete scFv in a single digestion step plus ligation reaction. Finally, the assembled product is amplified as a whole scFv prior to ligation and cloning.

[54] Figure 17 shows a series of photographs of DNA gel electrophoresis, comparing between SL1 (on the left) and SL2 (on the right) library assembly by seamless amplification. Figure 17A shows an electrophoresis gel of the amplified filtrated primary library scaffolds. Figure 17B shows an electrophoresis gel before (left lane) and after (right lane) the fragment seamless assembly with natural

H3J fragments by simultaneous digestion by Bsal and ligation. Figure 17C shows an electrophoresis gel of the amplification of the full scFv fragments.

[55] Figure 18 provides a bar graph showing Protein A binding of SL1 single clones relative to control scFvs after incubation at 37°C for 10 min. A sample of 44 ampicillin-resistant colonies was selected at random from the SL1. The phage were prepared and assayed on Protein A as described in

Figure 7. In-frame clones are indicated by the sign‘=’ and double transformants by the sign‘#\ [56] Figure 19 provides a bar graph showing Protein A binding of SL2 single clones relative to control scFvs after incubation at 37°C for 10 min. A sample of 44 ampicillin-resistant colonies was selected at random from the SL2. The phages were prepared and assayed on Protein A as described in

Figure 7. In-frame clones are indicated by the sign‘=’ and double transformants by the sign‘#\

[57] Figure 20 provides a bar graph showing Protein A binding of SL1 single clones relative to control scFvs after incubation at 70°C for 10 min. The phages were prepared and assayed on Protein A as described in Figure 7. In-frame clones are indicated by the sign‘=’ and double transformants by the sign ^e #\

[58] Figure 21 provides a bar graph showing Protein A binding of SL2 single clones relative to control scFvs after incubation at 70°C for 10 min. The phages were prepared and assayed on Protein A as described in Figure 7. In-frame clones are indicated by the sign‘=’ and double transformants by the sign ^e #\

[59] Figure 22 provides a bar graph showing a comparison of Protein A survival between PLs and SLs after incubation at 70°C for 10 min. Survival is defined >10% with respect to the control scaffold in the Protein A binding ELISA after incubation for 10 min at 70°C for phage clones having 25% or more binding to Protein A relative to the control scaffolds.

[60] Figure 23 provides a graph showing, on the y-axis, optical density measurements at 450 nm versus 570 nm showing the binding of the purified anti-TNFalpha scFv TNF-E12 taken at different concentrations on the x-axis to either TNFalpha or BSA as a negative control.

DEFINITIONS

[61] Detailed descriptions of preferred embodiments are provided herein. It is to be understood, however, that the present invention may be embodied in various forms. Therefore, specific references to various forms are provided as a basis for the claims and for teaching one skilled in the present art to employ the present invention in appropriate system, structure, or manner.

[62] Unless defined otherwise, all technical and scientific terms used herein have the same meaning as is commonly understood by one of ordinary skill in the art to which this invention belongs. All patents, applications, published applications, and other publications referred to herein are incorporated by reference in their entirety. If a definition set forth in this section is contrary to or otherwise inconsistent with a definition set forth in the patents, applications, published applications, and other publications that are herein incorporated by reference, the definition set forth in this section prevails over the definition that is incorporated herein by reference.

[63] For the purposes of this specification, the following definitions will apply and whenever appropriate, terms used in the singular will also include the plural and vice versa. In the event that any definition set forth below conflicts with the usage of that word in any other document, including any document incorporated herein by reference, the definition set forth below shall always control for purposes of interpreting this specification and its associated claims unless a contrary meaning is clearly intended (for example in the document where the term is originally used). It is noted that, as used in this specification and the appended claims, the singular forms“a,”“an,” and“the,” include plural referents unless expressly and unequivocally limited to one referent. The use of“or” means“and/or” unless stated otherwise. For illustration purposes, but not as a limitation,“X and/or Y” can mean“X” or‘Ύ” or“X and Y”. The use of“comprise,”“comprises,”“comprising,”“include ,”“includes,” and“including” are interchangeable and not intended to be limiting. Furthermore, where the description of one or more embodiments uses the term“comprising,” those skilled in the art would understand that, in some specific instances, the embodiment or embodiments can be alternatively described using the language“consisting essentially of’ and/or“consisting of’. The term“and/or” means one or all of the listed elements or a combination of any two or more of the listed elements.

[64] The section headings used herein are for organizational purposes only and are not to be construed as limiting the described subject matter in any way. All literature cited in this specification, including but not limited to, patents, patent applications, articles, books, and treatises are expressly incorporated by reference in their entirety for any purpose. In the event that any of the incorporated literature contradicts any term defined herein, this specification controls. While the present teachings are described in conjunction with various embodiments, it is not intended that the present teachings be limited to such embodiments. On the contrary, the present teachings encompass various alternatives, modifications, and equivalents, as will be appreciated by those of skill in the art.

[65] The practice of the present invention may employ conventional techniques and descriptions of bacteriology, molecular biology (including recombinant techniques), cell biology, and biochemistry, which are within the skill of the art. Such conventional techniques include PCR, extension reaction, oligonucleotide synthesis and oligonucleotide annealing, ELISA. Specific illustrations of suitable techniques can be added by reference to the example herein below. However, other equivalent conventional procedures can, of course, also be used. Such conventional techniques and descriptions can be found in standard laboratory manuals such as Genome Analysis: A Laboratory Manual Series (Vols. I- IV), PCR Primer: A Laboratory Manual, and Molecular Cloning: A Laboratory Manual (all from Cold Spring Harbor Laboratory Press, 1989), Gait,“Oligonucleotide Synthesis: A Practical Approach” 1984, IRL Press, London, Nelson and Cox (2000), Lehninger, Principles of Biochemistry 3rd Ed., W. H. Freeman Pub., New York, N.Y. and Berg et al. (2002) Biochemistry, 5th Ed., W. H. Freeman Pub.,

New York, N.Y. Barbas CF et al, (2004) Phage Display: A Faboratory Manual. CSHF Press, all of which are herein incorporated in their entirety by reference for all purposes.

[66] The term“antibody”, as used herein, is used in the broadest sense and refers to monoclonal antibodies and one or more fragments of an antibody that retain the ability to specifically bind to an antigen (e.g., lysozyme). It has been shown that the antigen-binding function of an antibody can be performed by fragments of a full-length antibody. Examples of antigen-binding fragments encompassed a Fv fragment consisting of the Y and V _H domains of a single arm of an antibody, a Fab fragment, a F(ab)2, which is a bivalent fragment comprising two Fab fragments linked by a disulfide bridge at the hinge region fragment a monovalent fragment consisting of the Y V _H , CF and Cj 11 domains, a Fd fragment consisting of the V _H and C _H l domains.

[67] Furthermore, although the two domains of the Fv fragment, V _L and V _H domains are coded for by separate genes, they can be joined, using recombinant methods, by a synthetic linker that enables them to be made as a single protein chain in which the V _L and V _H regions pair to form monovalent molecules, known as single chain Fv (scFv). Such scFvs are also intended to be encompassed within the term "antibody". These antibody fragments are obtained using conventional techniques known to those of skill in the art and the fragments are screened using, but not limited to, phage, yeast and mammalian display for utility in the same manner as are intact antibodies.

[68] As used herein,“antibody variable domain” refers to the portions of the light and heavy chains of antibody molecules that interact specifically with an antigen (e.g., lysozyme). ‘(jV or“V _H domain” refers to a variable domain of an antibody heavy chain.“V _L ” or“V _L domain” refers to a variable domain of an antibody light chain. The Y domain is produced by the recombination of the IGVF and IGVJ germline genes, whereas the V _H domain is encoded by repertoires of IGVH, IGVD and IGJH germline genes. V _H and V _L contain the antigen-binding site and hence define the capacity of antibodies to bind virtually any antigen with exquisite specificity and high affinity.

[69] The term“antigen-binding site”, as used herein, contains the portion of the variable domains that interact with the antigens. Definitions of the antigen-binding site include the complementarity determining regions (CDRs) as defined by Kabat (Kabat et al., Sequences of Proteins of Immunological Interest, 5th ed. Bethesda, Md.: National Center for Biotechnology Information, National Fibrary of Medicine, 1991). There are three CDRs in V _L : CDR-F1, CDR-F2, and CDR-F3, and three in V _H CDR- Hl, CDR-H2, and CDR-H3. The CDRs alternate with conserved regions called Framework Regions (FRs), four in V _L : FR-F1, FR-F2, FR-F3 and FR-F4, and four in V _H : FR-H1, FR-H2, FR-H3 and FR-H4.

The antigen-binding site can also be defined as specificity-determining regions as defined and the SDR usage regions (SDRUs). There are six SDRs or SDRUs, which approximately correspond with three CDRs - a comparison of the different definitions of antigen-binding site can be found at Gilliland et al, Methods Mol Biol. 841:321, 2012.

[70] As used herein,“amino acid position” refers to a position of an amino acid located in V _H or V _L amino acid sequence. Several numbering systems have been proposed to identify an amino acid position.

In this invention, the Chothia’s numbering convention (Chothia and Lesk, 1987) is used.

[71] A“diversified position” refers herein to an amino acid position with different amino acids represented at said position. In one aspect of the invention, positions to diversify are determined by identifying amino acids in the antigen-binding site contact with antigens in the structure of antibody:antigen complexes determined by x-ray crystallography. Antibody amino acids in contact with the antigen are defined by the distance between an atom of said amino acid in the antibody and an atom of an amino acid in the antigen. Typically, two atoms are in contact when distance between said atoms is <

3.5 A. A compilation of amino acids in antibodies in contact with antigens in known antigen: antibody structures is compiled at the IMGT. In other aspects of the invention, the positions to diversify are defined by the exposure of the amino acids to the solvent, defined as accessible surface area (ASA).

[72] The“diversification regime” as used in this invention refers to amino acids used to diversify the amino acid position. The diversification regime is derived from amino acid sequences of known and/or naturally occurring antibodies or antigen binding fragments. Diversified positions are typically found in the CDRs in known and/or naturally occurring antibodies and their discovery are facilitated by the antibody sequences and structures compiled at Internet-based databases. Databases compiling amino acid structures include the Protein Data Bank (PDB; http://www.rcsb.org/pdb ). Antibody sequence databases include V-base (http://www2.mrc-lmb.cam.ac.uk/vbase/), The National Center for Biotechnology Information (NCBI; https://www.ncbi.nlm.nih.gov/), the IMGT, and Abysis (http://www.bioinf.org.uk/abs/). These electronic resources provide extensive collections and alignments of human light and heavy chain sequences and facilitate the determination of highly diverse positions in these sequences.

[73] As used herein,“repertoire” or“library” refers to a plurality of antibodies, antibody fragment sequences, antibody variable domains, diversified scaffolds, or the nucleic acids that encode these sequences, the sequences being different in the combination of variant amino acids that are introduced into these sequences according to the methods of the invention.

[74] A“scaffold”, as used herein, refers to a polypeptide or portion thereof that maintains a stable structure or structural element when a heterologous polypeptide or amino acid is inserted into the polypeptide. The scaffold provides for maintenance of a structural and/or functional feature of the polypeptide after the heterologous polypeptide has been inserted. In one embodiment, a scaffold comprises an antibody variable domain, and maintains a stable structure when a heterologous CDR or amino acids are inserted into the scaffold.

[75] The term“library size” as used herein refers to the number of phages that comprise the library.

Several methods can be used to estimate the library size, being the most common counting the number of antibiotic-resistant colonies after plating a dilution of the library. Following the electroporation, transformed bacteria are incubated in rich medium for no more than 55 min before plating dilutions on agar supplemented with the appropriate antibiotic. This procedure ensures that the original transformants are counted before the bacteria start to divide actively.

[76] The term“functional space” or“effective size” of a library as used herein refers to the number of unique antibody sequences in a library that produce functional antibody fragments fused to a phage particle. For instance, nucleotide sequences with stop codons like UAA ("ochre") or UGA ("opal") lead to truncated sequences and do not produce functional antibody fragments fused to a virion particle. Insertions or deletions of one or two nucleotides change the reading frame of the gene sequence leading to stretches of amino acids that may impair folding or lead to non-functional clones.

[77] The functional space of a library can correspond with the library size if all the sequences are functional or is a fraction of the library size. The functional space can also be called 'shape space' or ‘sequence space’.

[78] The term“stability” as used herein refers to the ability of a molecule to maintain a folded state such that it retains at least one of its normal functional activities, for example, binding to an antigen or to a molecule like Protein A. The stability of the molecule can be determined using standard methods. For example, the stability of a molecule can be determined by measuring the thermal melt (“Tm”) temperature. The Tm is the temperature in degrees Celsius at which ½ of the molecules become unfolded. Typically, the higher the Tm, the more stable the molecule.

[79] As used herein,“natural” or“naturally occurring” polypeptides or polynucleotides refers to a polypeptide or a polynucleotide having a sequence of a polypeptide or a polynucleotide identified from a no synthetic source. For example, when the polypeptide is an antibody or an antibody fragment, the no synthetic source can be a differentiated antigen-specific B cell obtained ex vivo, or its corresponding hybridoma cell line, or from the serum of an animal. Such antibodies can include antibodies generated in any type of immune response, either natural or otherwise induced. Natural antibodies include the amino acid sequences and the nucleotide sequences that constitute or encode these antibodies, for example, as identified in the Kabat database. As used herein, natural antibodies are different than“synthetic antibodies”, synthetic antibodies referring to antibody sequences that have been changed, for example, by the replacement, deletion, or addition, of an amino acid, or more than one amino acid, at a certain position with a different amino acid, the different amino acid providing an antibody sequence different from the source antibody sequence.

[80] A“plurality” or“population” of a substance, such as a polypeptide or polynucleotide of the invention, as used herein, generally refers to a collection of two or more types or kinds of the substance.

There are two or more types or kinds of a substance if two or more of the substances differ from each other with respect to a particular characteristic, such as the variant amino acid found at a particular amino acid position. In a non-limiting example, there is a plurality or population of polynucleotides of the invention if there are two or more polynucleotides of the invention that are substantially the same, preferably identical, in sequence except for one or more variant amino acids at particular CDR amino acid positions.

[81] The term“developability” as herein used, refers to a set of desirable antibody properties which, as a whole, facilitates clinical development and manufacturing of a therapeutic antibody. Properties that have been associated with developability include, but are not limited to, good expression (>30 mg/L in transient CHO cell expression), thermal stability (>70°C), low or no aggregation (< 1% high molecular weight aggregates; HMWA), and solubility of > 40 mg/ml for IV indications and > 100 mg/ml for subcutaneous indications.

[82] These properties are interrelated. For instance, chemical instabilities such as oxidation or clipping sites result in sample heterogeneity and eventually can impact the physical stability or lead to low solubility or aggregation. Poor physical stability can expose side-chains prone to oxidation or degradation, eventually leading to aggregation when these residues are degraded. Nevertheless, each of these parameters can be measured independently. For instance, expression can be assessed by measuring the amount of antibody in the culture media by ELISA, Octet™ or BIAcore™. Thermal stability can be measured by Thermal shift analysis or DSC. Aggregation can be measured by SEC-HPLC. Solubility can be measured by concentrating the antibody at diverse concentrations.

[83] Molecules that do not meet the success criteria above described are commonly deprioritized in the therapeutic antibody development process. Alternatively, if antibody leads were identified with promising biological activity but low performance during the developability assessment, they are targeted for developability enhancement by the methods described in the art during the optimization phase.

[84] “Phage display” is a technique by which variant polypeptides are displayed as fusion proteins to at least a portion of a coat protein on the surface of phage, e.g., filamentous phage, particles. A utility of phage display lies in the fact that large libraries of randomized protein variants can be rapidly and efficiently sorted for those sequences that bind to a target molecule with high affinity. Display of peptide and protein libraries on phage has been used for screening millions of polypeptides for ones with specific binding properties. Polyvalent phage display methods have been used for displaying small random peptides and small proteins through fusions to either gene III, VIII or IX of filamentous bacteriophage, see references cited therein.

[85] In monovalent phage display, a protein or peptide library is fused to a gene III or a portion thereof, and expressed at low levels in the presence of wild type gene III protein so that phage particles display one copy or none of the fusion protein. Avidity effects are reduced relative to polyvalent phage so that sorting is on the basis of intrinsic ligand affinity, and phagemid vectors are used, which simplify DNA manipulations.

[86] A“phagemid” is a plasmid vector having a bacterial origin of replication, e.g., ColEl, and a copy of an intergenic region of a bacteriophage. The phagemid may be used on any known bacteriophage, including filamentous bacteriophage and lambdoid bacteriophage. The plasmid will also generally contain a selectable marker for antibiotic resistance. Segments of DNA cloned into these vectors can be propagated as plasmids. When cells harboring these vectors are provided with all genes necessary for the production of phage particles, the mode of replication of the plasmid changes to rolling circle replication to generate copies of one strand of the plasmid DNA and package phage particles. The phagemid may form infectious or non-infectious phage particles. This term includes phagemids, which contain a phage coat protein gene or fragment thereof linked to a heterologous polypeptide gene as a gene fusion such that the heterologous polypeptide is displayed on the surface of the phage particle.

[87] As used herein,“amplify” refers to the process of enzymatically increasing the amount of a specific nucleotide sequence. This amplification is not limited to but is generally accomplished by PCR.

As used herein,“denaturation” refers to the separation of two complementary nucleotide strands from an annealed state. Denaturation can be induced factors such as, for example, ionic strength of the buffer, temperature, or chemicals that disrupt base pairing interactions.

[88] The terms“amplification cycle” and“PCR cycle” are used interchangeably herein and as used herein refers to the denaturing of a double -stranded polynucleotide sequence followed by annealing of a primer sequence to its complementary sequence and extension of the primer sequence.

[89] The terms“polymerase” and“nucleic acid polymerase” are used interchangeably and as used herein refer to any polypeptide that catalyzes the synthesis or sequencing of a polynucleotide using an existing polynucleotide as a template.

[90] As used herein,“DNA polymerase” refers to a nucleic acid polymerase that catalyzes the synthesis or sequencing of DNA using an existing polynucleotide as a template.

[91] All publications and patents mentioned herein are incorporated herein by reference for all purposes, including the purpose of describing and disclosing, for example, the constructs and methodologies that are described in the publications, which might be used in connection with the presently described invention. The publications discussed herein are provided solely for their disclosure prior to the filing date of the present application. Nothing herein is to be construed as an admission that the inventors are not entitled to antedate such disclosure by virtue of prior invention or for any other reason.

[92] Unless defined otherwise, all technical and scientific terms used herein have the same meaning as commonly understood to one of ordinary skill in the art to which this invention belongs. Although any methods, devices, and materials similar or equivalent to those described herein can be used in the practice or testing of the invention, the preferred methods, devices and materials are now described.

DETAILED DESCRIPTION OF THE INVENTION

[93] This invention includes methods to generate highly functional antibody libraries. In one embodiment, the methods comprise the steps of: (i) generation of a primary library (PL) of antibody fragments; (ii) a process to select for a well-folded antibody fragment library (FL) from the PL; and (iii) combining the antibody fragments selected in step 2 with diverse CDR-H3 fragments to generate a highly functional secondary antibody library (SL). A detailed description of modalities these steps and exemplary proofs of each step follow.

Primary Library (PL) Generation

[94] To demonstrate the methods in the present invention, two primary libraries were built with human antibody germline genes. Both libraries had a universal V scaffold and two \ scaffolds. The \ scaffolds provided two alternative antigen-binding site topographies. One of the V _L scaffolds had a short Ll loop (PL1), whereas, the other had a long Ll loop (PL2) (Figure 1). By changing the length of Ll from a short to a long loop, antibodies alter the preference to bind protein or peptide targets (Vargas- Madrazo et al., 1995). Therefore, by using the proper V _L scaffold, the repertoires exemplified herein can be used for selection of antibodies against protein or peptide targets. In one embodiment, libraries of the present invention built on these scaffolds were used in combination to generate antibodies that bind diverse types of epitopes on a given target.

[95] As shown in Example 1, the V _L scaffold with a short Ll loop was built by assembling the human IGKV3-20*0l germline gene with the human IGJV4*0l germline gene (SEQ ID NO: 1). The Y scaffold with the long Ll was built by assembling the human IGKV4-0l*0l germline gene combined with the same human IGJV4*0l germline gene (SEQ ID NO: 2). IGKV3-20*0l and IGJV4*0l belong to different IGKV families and share only 68% identities. Thus, the scaffolds built with these two genes represent two distantly related human germline genes and thus offer distinct exemplary proofs of this invention.

[96] As counterpart of the V _L scaffolds, a single V _H scaffold, built with the human IGHV3-23*0l and the human IGJH3*0l germline genes (SEQ ID NO: 3) was used. In one aspect of this invention, Protein A binds the framework region 3 (FR-3) of the V _H domain encoded by germline gene IGHV3-23*0l. The FR-3 is formed by discontinuous amino acid stretches distant in the primary sequence but brought together by folding. Therefore, Protein A has been used in prior art for its ability to bind only to well- folded V _H domains (Jespers et al., 2004).

[97] Other natural ligands that bind folded variable regions of antibodies outside the antigen-binding site have been described in the art. For example, Protein-F binds the V _L domain of antibodies encoded by the human IGKV-l, IGKV-2 and IGKV-4 gene families (Nilson et al., 1992). More specifically, the IGKV4-0l*0l germline gene, which belongs to the human IGKV4 family, binds Protein-F. The use of Protein-F as a ligand to select well-folded antibodies alone and/or in conjunction with Protein A is a clear extension of this invention. For instance, human V _H domains of antibodies encoded by the gene families IGHV-l, IGHV-2, IGHV-4, IGHV-5, IGHV-6 and IGHV-7, which do not bind Protein A, can be paired with libraries of V _L domains encoded by scaffolds built with members of the human IGKV-l, IGKV-2 and IGKV-4 gene families. These libraries and antibody pairs can then be submitted to diverse destabilizing conditions to select for well-folded antibodies with Protein-F.

[98] Yet in another aspect of the invention, the germline genes used to build the V _H and V _L scaffolds of this invention have frequently been observed in human antibodies elicited against a vast array of diverse antigens (Nilson et al., 1992). These genes have also been used as foundation to build numerous scFv and Fab libraries for antibody discovery in the art (US patent 9,062,305) (Shi et al., 2010), as well as in humanization of therapeutic antibodies (US patent 8,777,044). Therefore, it is expected that antibodies discovered from libraries built with these scaffolds will perform well both in vitro and in vivo settings and will be amenable to further therapeutic development.

[99] In other aspects of the invention, the Tm of the _H Vand V _L scaffolds in Fab format has been measured by Differential Scanning Calorimetry (DSC) yielding similar values of 75°C (Teplyakov et al., 2016). This Tm is almost l0°C above the Tm of the C _H 2 domain, which is estimated at 68°C (Gilliland et al., 2012). The C _H 2 is the least stable domain of the human IgGl molecule and hence the first domain that unfolds. Therefore, the antibodies isolated from the libraries herein described are expected to be highly stable.

[100] In yet in another related aspect of the invention, antibodies encoded by the exemplary V _H and V _L scaffolds herein disclosed have been solved by x-ray crystallography (Teplyakov et al, 2016) in association the CDR-H3 loop of a known therapeutic antibody, CNTO 888 (Obmolova et al., 2012). This knowledge facilitated the design of diversity in all the CDRs that form the antigen-biding site except in the CDR-H3.

[101] The diversity of the V _H scaffold was focused on the CDR-H1 and CDR-H2 and was designed at positions and amino acids commonly found in contact with protein and peptide antigens, positions accessible to antigens in the antigen-binding site and/or in contact with the CDR-H3. As exemplified in Example 2, to determine the positions in contact with protein and peptide targets, all the antibody structures compiled at the PDB and curated by the IMGT as of March 17 ^th , 2017 were analyzed. The initial dataset consisted of 2,645 antibody structures from diverse species and specificities. From this initial dataset, only anti-protein antibodies with different names and those solved at < 3Ά resolution were considered for identifying positions to diversify. In addition, only the antibody structures with the same length at CDR-H1 and CDR-H2 were studied. Although this filter is certainly stringent, it removes for the most part antigen-binding site structures with canonical structures classes (Chothia and Lesk, 1987) other than that of the V _H scaffold. Having the same canonical structures guarantees the same relative position of the amino acids at the antigen-binding site to contact antigens. As additional precaution to design the diversification of the V _H scaffold, the V _H sequences with the same length were compared by a clustering algorithm to remove those showing 100% of identity and thus avoided counting the same sequence twice. The final dataset contained 117 antigen: antibody complexes which are listed in Figure 2

[102] To identify the positions in contact with antigens, the contact tables of the 117 structures listed in Figure 2 were downloaded from the IMGT and the contact residues aligned. The residues in contact with antigens were then mapped onto the structure of the V _H scaffold (PDB ID: 5DIL) to determine their ASAs and structural environment. A total of 10 positions were targeted for diversification, four in the CDR-H1 and six in the CDR-H2. These positions have >70% ASA and thus should tolerate diverse amino acid side chains without disruption of the A scaffold. On other hand, all targeted positions but 30 at the CDR-H1 and 55 at the CDR-H2 were found in > 30% contacts, thus maximizing the probability of interacting with diverse antigens. Positions 30 and 55 are in the periphery of the antigen-binding site and are highly exposed to the solvent (> 65% ASA). Hence, although these positions are in contact antigens in less than 50% of the complexes, they were considered for diversification to supplement diversity for binding targets of bigger size than average proteins.

[103] The diversification regimes at CDR-H1 and CDR-H2 were designed using three sources of information: the dataset of curated antibody structures listed in Figure 2, the V _H sequences available at NCBI and curated by Abysis, and the germline genes of the human IGHV3 family compiled at the IMGT.

The number of V _H sequences available at Abysis amounts 76,000 V _H sequences, thus complementing the information obtained from the two other sources. The comparison with the frequency of amino acids encoded in an alignment of the human IGHV3 germline family ensured that the amino acids used for diversifying the scaffolds mimicked the human germline diversity.

[ 104] The amino acids targeted for diversification were examined in the structure of the primary library scaffolds (Figure 1) to avoid conflicts and/or inconsistencies between the nature of amino acids used for diversification and the environment of such amino acids in the context of the scaffolds. For instance, residues buried in the protein had to be hydrophobic. Residues in b-tums or maintaining the canonical structures such as position 54 at the CDR-H2 had to be consistent with the propensity of a residue in such conformation. Aromatic residues had to be close to other aromatic residues or residues with relatively long aliphatic side-chains. The final diversification regime at CDR-H1 and CDR-H2 of the V _H scaffold is summarized in Figure 3. The estimated number of amino acids per position is listed in the last column of the figure and yielded a diversity of 5.4 x lOmique amino acid \( sequences. Some positions are relatively conserved such as position 30 at the CDR-H1 with 2 amino acids allowed, whereas other positions are heavily diversified with up to 9 residues, e.g., position 50 of CDR-H2.

[105] To identify positions for diversification in the iV scaffolds the same procedure as in the A scaffold was followed. First, positions in contact with the antigen were identified in the dataset of 117 structures listed in Figure 2. Next, positions in contact with the antigen were mapped onto the structure the V _L scaffolds (Figure 1) to evaluate their solvent exposure, structural context and interaction with the V _H Scaffold.

[ 106] As in the V _H scaffold, the diversification regime of the V _L scaffolds was first determined by the frequency of amino acids occurring in the alignment of the 117 structures (Figure 2). This information was supplemented with the analysis of all the V _L sequences compiled by Abysis, total 20,764 sequences.

This information was then compared with the amino acids and frequencies of the human IGKV germline gens compiled at the IMGT. Fewer V _L than V _H sequences at Abysis reflects the fact that ~ 60% of the human antibody repertoire is k-type, and only k-type sequences are considered in design. Finally, the diversification regime of the V _L scaffolds was evaluated in the structures of V _H and V _L scaffolds (Figure 1). The final diversification regime of the V _L scaffolds is summarized in Figure 4 and Figure 5. The estimated number of amino acids per position is listed in the last column of each of the two figures and yielded a total of 1.1 x 10 ⁶ and 1.3 x 10 ⁶ unique amino acid V _L sequences for PL1 and PL2, respectively.

[107] In other aspect of the invention, it was reasoned that by using only one CDR-H3 sequence to generate the PLs, for instance the CDR-H3 of CNTO 888, the diversity of amino acids in contact with, or nearby, said CDR-H3 may be constrained to a few and specific residues to accommodate said CDR-H3 under specific selection conditions. Therefore, a set of CDR-H3 sequences called“neutral H3Js” was designed by starting from the repertoire of human IGDH genes and combining them with the human IGJH segments in germline configuration. Since antibodies from natural primary repertoires are mostly sequences in germline configuration, it was hypothesized that the neutral H3Js provide enough diversity to avoid biases in amino acids in contact with, or nearby, the neutral H3J sequences while contributing to select for developable antibodies. As exemplified in Example 3, the set of neutral H3Js fragments comprises ninety (90) fragments (SEQ ID Nos: 4-93).

[108] The designs of the diversified V _H and V _L scaffolds of this invention with the neutral H3Js were assembled in the configuration depicted in Figure 6. The V _L -linker-V _H scFv configuration places the H3J fragments on the C-terminal. V _L is physically linked to V _H by a linker of the 19 amino acids GGGGSGGGGSGGGSGGGGS (GS19). Two Bgll/Sfil sites on each side allows for in-frame cloning in the acceptor vector between the pelB leader peptide for periplasmic expression and the tags for detection.

Other configurations including Y _r linker-V _| and other linker sequences are obvious extensions of this invention.

[109] In one embodiment of the invention, the Ncol restriction site and the Kpnl restriction site are common to both repertoires, allowing for easy replacement of V _L - In other embodiments, the JK anchor sequence in V _L and the human VH Consensus Motif (CM) in V _H (Example 8) are used for exchange and cloning of natural H3J fragments in the third step of the invention or for exchanging the light chains. These two short sequence stretches are common to PL1 and PL2.

[110] The resultant scFv designs were synthesized using trimer phosphoramidite technology. Nevertheless, other various methods described in the art can be used to realize the designs, including mix of oligonucleotides and restriction enzymes. For instance, seamless DNA assembly using type IIs enzymes, e.g. Golden Gate assembly (New England Biolabs) can be used to concatenate^ ¥nd V _L diversified scaffolds and H3Js fragments in a single piece of DNA.

[111] The synthetic fragments of this invention were cloned as fusions to pill by digestion with Bgll restriction enzyme and ligated into pADL™-23c phagemid vector (Antibody Design Labs, San Diego, cat# PD0T11 ) as described in Example 4. The pADL™-23c phagemid vector is a classical type 3+3 phage display vector with a cloning site for display on the N-terminal side of the full-length gene III protein. Secretion in the periplasm of the fusion protein is driven by the PelB leader peptide. Display of the scFvs on the phage is obtained with the help of an amber-suppressive bacterial strain and M13K07 helper phage. Growth on non-suppressive strains results in the expression of free scFvs in the periplasm space, which can be subsequently purified by immobilized metal affinity chromatography (IMAC) with the help of the His tag and detected with the Myc tag as described in prior art.

[112] The initial diversity of PL1 and PL2 was 1.7 x 10 ⁹ and 2.3 x 10 ⁹ cfu, respectively. To prove that PL1 and PL2 displayed as fusion proteins on the phage surface bound Protein A and hence can be submitted to the step 2 of the invention, the libraries were assayed on Protein A by a direct phage ELISA (Example 5). Figure 7 shows that both primary libraries bound to Protein A, with PL2 showing a higher EC ₅₀ than PL 1.

[113] Although it was considered that these examples were the best modality in the field of this invention, there is no limitation to the origin of the PL, besides the understanding noted above that some constructions, use of specific scaffolds or natural sources may be more suitable for developable antibody generation. For instance, a PL can be obtained from natural sources such as PBMCs and lymphoid organs (lymph nodes, spleen), either from human or animals that have not been immunized against any specific antigen, hyper-immunized animals or individuals suffering or not from a debilitating condition or circumstance of immunological interest such as a recent vaccination or an infectious episode.

Selecting for Highly Stable Fragments and FLs

[114] In this step, the PLs are submitted to diverse conditions to select for highly stable and developable antibody variants. In the best modality of this invention, heat was used to eliminate unstable variants from the pool of primary antibodies. It has been shown that transient heat treatment of antibodies displayed on phage can lead to denaturation and aggregation of the less thermostable antibody variants in a library (Jespers et al., 2004). Therefore, following heat-induced denaturation and aggregation of variants with lower stability profiles, rescue by Protein A through direct binding yielded filtered libraries of improved functionality such as in-frame and thermostable clones. A well-known indirect effect to those skilled in the art of the process of selection and re-amplification of phage libraries as operated in the above filtration is the drastic decrease of diversity that can lead to the rapid collapse of libraries (Matochko et al., 2014). The main purpose of this invention is to provide a method to improve the diversity, hence functionality, of antibody library filtrated for improved developability without impacting the quality of the antibody variants collected along the filtration process.

[115] Conditions for selecting developable and stable antibody variants from PLs include, but are not limited to heat. Conditions capable to induce unfolding, or combination of thereof, are obvious extensions of this invention. For example, although without clear literature examples, but obvious from the skilled in the art, high or low pH, high salt concentrations or any chaotropic conditions such as urea may be used to interrogate antibody stability. In another application, partial unfolding of the antibody structure may lead to exposure of a stretch of the polypeptide backbone and protease sensitivity. This approach has been used to engineer antibody with high-protease resistance capable to survive the stress conditions found in the gastrointestinal track and is a clear extension of this invention (Hussack et al.,

2011). [116] Another fundamental aspect of antibody reactivity is the non-specificity or multi-specificity. Non-specific, multi-reactive B-cell clones are eliminated during the maturation of the B-cell repertoire.

Use of non-specific absorption and hydrophobic interaction may replicate such a process in vitro and are an obvious extension of this invention. In this case, rather than applying a positive selection of well- folded antibodies, e.g. with Protein-A, unwanted clones are passively eliminated from the pool. A cumbersome alternative to assay for non-specificity is to test each antibody lead along the pipeline against panels of antigens and proteins. Tools to predict antibody non-specificity applicable to entire libraries such as binding to chaperones, e.g., HSP70 or HPS90 (Kelly et al., 2017) are a clear extension of this invention.

[117] In an embodiment of this invention, well-folded antibody variants are rescued with a ligand that selectively binds said well-folded antibodies. As shown in Example 5, one such a binder is Protein A.

Herein is exemplified this critical step with incubation of PL1 and PL2 in a range of temperatures followed by the rescue of the folded variants by Protein A. To set up the optimal conditions for selection of developable antibody fragments, Example 6.1 shows the incubation for 10 min of the 3-20/3-23 and 4-

01/3-23 scFvs prior to diversification. The scFvs displayed as fusions to the minor phage coat protein pill on the phage surface were submitted for a range of temperatures, starting at 40°C and increasing the temperature up to 80°C in steps of 5°C. The unfolding process was monitored with a direct Protein A

ELISA (Figure 8). Since the M13 phage is stable at 80°C, a change in the ELISA signal in this range of temperatures is a direct consequence of the scFv unfolding.

[118] The unfolding process, which leads to aggregation of the scFv-phage and hence a drop in the ELISA signal depends on two parameters (Jespers et al., 2004): (i) thermal stability of the scFvs, and (ii) number of copies of the scFvs displayed on the phage surface, which varies from one to five copies. Once the scFv starts to unfold, a cooperative aggregation process takes place. ScFv-phages with a higher number of copies aggregate first, serving of aggregation seed for phages with a fewer number of scFv copies.

[119] Example 6.1 indicated that the 3-20/3-23 scFv started to unfold at 65°C, whereas, the 4-01/3-23 scFv started the unfolding process at 55°C. The Tm of the 3-20/3-23 scFv is 75°C, at which 50% was unfolded. The Tm of 4-01/3-23 scFv is 65°C. The unfolding kinetic at 60°C and 72°C of the 3-20/3-23 scFv and 4-01/3-23 scFv demonstrated that the 3-20/3-23 scFv remained folded at 60°C for up to one hour (Figure 9). At 72°C it unfolded slowly with approximately a 20% drop in the ELISA signal during the first 30 min. Afterwards, a quick drop of the ELISA signal was seen. The 4-01/3-23 scFv unfolded very slowly at 60°C. At 72°C it showed a quicker unfolding process than 3-20/3-23 scFv, with a drop in the signal of approximately 30% in the first 10 min. This process was accelerated afterwards, probably due to a massive aggregation of phage with a fewer number of 4-01/3-23 scFv copies. [120] Therefore, this example demonstrated that: (i) the 3-20/3-23 and 4-01/3-23 scFvs used as scaffolds to generate the exemplary PLs of this invention are stable at 60°C for at least one hour; (ii) said 3-20/3-23 and 4-01/3-23 scFvs can be incubated at 72°C for 10 min without significant unfolding and aggregation, i.e., < 30%; (iii) the 3-20/3-23 and 4-01/3-23 scFvs have specific Tm’s and dynamic of unfolding, which are intrinsic properties of the VL scaffold sequence and structure and can be used to tailor the selection conditions to generate a variety of developable antibodies by changing the harsh denaturing conditions.

[121] Accordingly, and as shown in Example 6.2, the exemplary PLs of this invention produced unique scFvs with a significant number, on average -52.3% (Table 3), of well-folded antibodies after incubation at 72°C for 10 min. This number was down from on average -64% in both PLs prior to the heat treatment, representing a combined number of more than 2 billion folded but unstable antibodies for PLs of 1 x 10 ¹⁰ clones or more each.

[122] Considering that the Tm of the 2 domain of the human IgGl is 68°C, as an example of selection for highly stable and developable antibodies in this invention, it was decided to submit PL1 and PL2 for incubation at 70°C for 10 min and rescue well-folded antibody variants with Protein A as shown in Example 7. The resulting libraries, FL1 and FL2, had both the hallmark of a selection by Protein A.

All single clones, but one in the respective sampling, were in-frame, giving a percentage of in-frame clones of 94.7% and 97.3% in FL1 and FL2, respectively. Longer or shorter incubation times at higher or lower temperatures are obvious extensions of the method herein disclosed.

[123] The use of Protein A to improve the functionality of antibody libraries is a well-described procedure, but is known to result in a significant loss of diversity (Jespers et ah, 2004; Famm et ah, 2008; Rouet et ah, 2014). For instace, antibody domains libraries displayed on filamentous phage have been submitted to 80°C in pH 7.4 by Jesper et ah, and folded domains have been selected by binding to Protein A. Using the same phage library repertoire, Famm et al. have extended the method to the selection of domains resistant to acid aggregation. In both cases, as expected, the number of stable clones have been reduced dramatically.

Construction of Secondary Libraries (SLs)

[124] The present invention also includes methods to restore diversity to an antibody library that has lost some of its diversity due to enrichment and/or selection. By combining the highly stable variants selected in the previous step with a collection of natural CDR-H3 fragments the diversity of the libraries can be restored. Two SL libraries, SL1 and SL2, were generated starting from plasmid DNA of FL1 and FL2 as substrate for amplification of well-folded scaffolds by PCR as exemplified in Example 8. The PCR-generated fragments included the pelB leader peptide-encoding nucleotide sequence and a stretch of nucleotides immediately before the CM (Figure 16). The sequence of the CM allowed the amplification of more than 95% of all antibody sequences found in circulating PBMCs as described in Example 8. During the amplification, a Bsal restriction site was added immediately after the end of the V _H fragment. Symmetrically, a repertoire of natural H3Js fragments was obtained from 200 healthy donors as described in Example 8.1. It was generated with a pair of primers matching the sequence of the CM and the pADL™-23c phagemid vector sequence downstream to the second Sfil site. In doing so, a Bsal site was added 5’ to the CM. Simultaneous digestion by Bsal of the two fragments and ligation led to the joining of the FLs with the natural H3J fragments (Figure 17). Subsequent amplification of complete scFvs by nested primers led to the successful cloning of very large SL1 and SL2 libraries (Example 8.2). The initial diversity was 1.4 x and 1.1 x 10 ¹⁰ cfii for SL1 and SL2, respectively. The use of the digestion/ligation joining reaction enabled the subsequent amplification starting from large amounts of well-assembled full-length scFvs. Other methods of molecular biology well-known to those skilled in the art, such as PCR by overlapping extension, are obvious extensions of the method to build the SL libraries.

[125] The quality of the SLs was assessed by Sanger sequencing as described in Example 8.3. It was found that 35 out of 41 clones (85.4%) were in-frame for both SL1 and SL2. These values were only on average 10% lower than those of the FLs, indicating that our method of the SLs construction retained the vast majority of the in-frame clones from the FLs. In addition, all the sequence studied happened to have unique natural CDR-H3 sequences. Therefore, the successful addition of natural H3J diversity into the FLs was performed to produce the SLs while retaining the in-frame character resulting from the filtration.

[126] When single clones from SL1 and SL2 were expressed as scFv on the phage surface, and after incubation at 72°C for 10 minutes, over 80% of them bound Protein A (Figure 20 and Figure 21). The comparison of the percentage of stable clones between PLs and SLs showed an increase in absolute value of 15% to 20%, equivalent - 1.5 xlO ⁹ to 2 xlO ⁹ additional variants in a library of 10 ¹⁰ variants (Figure 22). In relative values, the percentage of stable clones grew by an average of -33% in both SLs comparatively to the PLs using the library construction method of this invention. Since PLs are representative examples of typical synthetic libraries in the art, the newer created libraries of this invention, or SLs, expanded the functional space of the previous synthetic libraries. This improvement increased significantly the probability of isolating specific and higher affinity antibodies from the SLs.

[127] Most of the natural CDR-H3 loops are in-frame. Therefore, it was expected that most of the clones from the SLs resulted in in-frame scFv sequences. It was unexpected, that -70% of the SL clones survived the incubation for 10 min at 70 °C. Natural CDR-H3 fragments have not been exposed to such harsh conditions in physiological conditions, and hence have not evolved to be stable at such a high temperature. Moreover, it was not predictable from prior art that the natural CDR-H3 would interact well with the stable scaffolds from the FLs. The natural CDR-H3 fragments of this invention were isolated from PBMCs of a pool of 200 donors and amplified with primers to generate around 95% of all the CDR- H3 in those individuals. The CDR-H3 is highly variable, with length variation between 3 and more than 20 amino acids, recombined with over 40 IGHV and 6 IGHJ and paired with ~ 40 IGKV genes. This enormous diversity when cloned in the highly stable exemplified scaffolds of the invention, only impacted -10% of the overall library diversity while improving stability. Therefore, it is likely that an interplay between the selection among the human germline genes of highly stable scaffolds to build the libraries, with the designed diversification regime based on germline genes, compounded with removal of liability developabilities, expressed in conjunction with the set of neutral H3Js designed to be flexible and finally, the selection of PLs, produced to a collection of highly stable antibody variants suited to accommodate such a diverse natural collection of CDR-H3 fragments.

[128] As shown in Example 9, analysis of the libraries generated during the construction by next generation sequencing (NGS) confirmed the quality of PLs in accord with the intended design. Analysis of the filtration process showed as expected an increase in productive sequences and a decrease of the diversity after the filtration process. Distribution of the neutral D elements and JH fragments showed limited bias, with the exception of a trend toward the elimination of the most hydrophobic fragments, underlining the need for a varied local environment around CDR-H3 during the filtration process to minimize biases in the diversity as provided by the set of neutral H3J fragments. NGS analysis of the SL confirmed the presence of very high levels of diversity at both CDR-H3 level, with around 60% unique sequences, and full antibody sequence level with above 99% diversity at the depth of around one million sequences in each SL (Table 12). Analysis of the most prevalent CDR-H3 indicated a limited copy number for each clone, to the contrary of what one would expect from re-amplified libraries.

[129] As shown in Example 10, standard selection techniques applied with two known antigen models,

TNF and HSA , were performed. In both examples specific antibodies were isolated, demonstrating the potential of the SLs to produce specific antibodies

EXPERIMENTAL EXAMPLES

[130] The following examples are offered to illustrate, but not to limit the claimed invention. It is understood that the examples and embodiments described herein are for illustrative purposes only and that various modifications or changes in light thereof will be suggested to persons skilled in the art and are to be included within the spirit and purview of this application and scope of the appended claims. All publications, patents, and patent applications cited herein are hereby incorporated by reference herein in their entirety for all purposes.

EXAMPLE 1 : DESIGN OF TWO HUMAN SCAFFOLDS FOR RECOGNITION OF DIVERSE EPITOPES

[131] To maximize the functionality of the PLs, antibody scaffolds highly utilized in human with high expression profile and unique recognition properties were selected from the human germline repertoire. Two PLs were built with a V _L scaffold that has a short Ll loop (SEQ ID NO: 1) and a V _L scaffold that has a long Ll loop (SEQ ID NO: 2). As counterpart of the V _L scaffolds, a single V _H scaffold (SEQ ID NO: 3) was used. By changing the length of Ll from a short to a long loop, antibodies alter the preference to bind protein or peptide targets, respectively.

EXAMPLE 2: DESIGN OF DIVERSIFICATION REGIMES FOR TWO HUMAN

SCAFFOLDS

[132] The diversity of the V _H scaffold was focused on the CDR-H1 and CDR-H2 and was designed at positions and amino acids commonly found in contact with protein and peptide antigens, positions accessible to antigens in the binding site and/or in contact with CDR-H3. To determine the positions in contact with protein and peptide targets, 117 antigen: antibody complexes listed in Figure 2 were used.

The positions in contact with antigens were identified by downloading the contact tables of the 117 structures from the IMGT and aligning the contact residues. Then, the residues in contact with antigens were then mapped onto the structures of the V _H scaffold paired with the two V _L scaffolds (PDB IDs: 1ILC and 1ILD) to determine their AS As and structural environment.

[133] The diversification regime at CDR-H1 and CDR-H2 was designed using three sources of information: The dataset of curated antibody structures listed in Figure 2, the V _H sequences available at NCBI and curated by Abysis, and the germline genes of the human IGHV3 family compiled at the IMGT. The number of V _H sequences available at Abysis amounted for 76,000 V _H sequences, thus complementing the information obtained from the two other sources. The final diversification regime at CDR-H1 and CDR-H2 of the V _H scaffold is summarized in Figure 3. The estimated number of amino acids per position is listed in the last column of Figure 3 and yields a diversity of 5.4 x 10 ⁶ unique amino acid V _H sequences. [134] To identify positions for diversification in the V _L scaffolds, we followed the same procedure as for the \h scaffold. First, positions in contact with the antigen were identified in the dataset of 117 structures listed in Figure 2. Next, positions in contact with the antigen were mapped onto the structure of the V _L scaffolds (Figure 1) to evaluate their solvent exposure, structural context and interaction with the V _H scaffold. The diversification regime was first determined by the frequency of amino acids occurring in the alignment of the 117 structures (Figure 2). This information was supplemented with the analysis of all the V _L sequences compiled by Abysis, totaling 20,764 sequences. This information was then compared with the amino acids and frequencies of the human IGKV germline genes compiled at the IMGT. The final diversification regime of the V _L scaffolds is summarized in Figure 4 and Figure 5. The estimated number of amino acids per position is listed in the last column of Figure 4 and Figure 5 and yields a total of 1.1 x 10 ⁶ and 1.3 x 10 ⁶ unique amino acid V _L sequences for the 3-20 and 4-01 repertoires, respectively.

EXAMPLE 3: NEUTRAL CDR-H3 DESIGN

[135] To design the neutral H3Js, the IGDH germline genes compiled at the IMGT were translated in the three reading frames into amino acid sequences. Only productive sequences, i.e. without stop codons, were considered. Those IGDH genes with developability liabilities including methionine (M) and tryptophan (W) residues, as well as those encoding three or more hydrophobic residues were removed from the set of neutral H3Js. After this in silico selection process, 18 sequences from the IGDH germline genes were combined with 5 IGJH regions to produce ninety (90) fragments (Table I) (SEQ ID Nos: 4- 93). The length of the neutral H3Js varies between 18 and 27 amino acids.

TABLE I

“Neutral” D Elements J Regions

GYSGYDY AEYFQHWGQGTLVTVSS

GYSYGY DAFDVWGQGTMVTVS S

TTVT YFDYWGQGTLVTVSS

YSGSYY NWFDSWGQGTLVTVSS

DYGDY YGMDVWGQGTTVTVS S

DYSNY

SIAAR VQLER

YYDILTGYYN

YYYDSSGYYY

YYYGSGSYYN

EYSSSS

GITGT

GIVGAT

GTTGT

GYSSGY

LTG

VDIVATI

EXAMPLE 4: SYNTHESIS AND CLONING OF PRIMARY LIBRARIES

[136] In this example, the two scFv libraries with the diversity reported in Figures 3-5 and the configuration depicted in Figure 6 were submitted for chemical synthesis using trimer phosphoramidite mixtures at the position of degeneracy. The set of neutral H3J fragments (SEQ IDs: 0004-0093) was incorporated on the 3’ side of the synthetic fragments. The quality of the synthetic fragments was assessed by Sanger sequencing of 96 clones taken randomly from the final synthetic fragments. The results are summarized below (Table 2):

TABLE 2

Repertoire 3-20/3-23 4-01/3-23

Sample Size 96 clones 96 clones

In-frame Clones 62% 73%

Design Compliant Clones 50% 60%

[137] The 3-20/3-23 synthetic fragments exhibited 50% of sequences in-frame matching the design, whereas the 4-01/3-23 fragments had 60%. Around 10% of the fragments had either an in-frame insertion or a deletion. Although these sequences did not comply with the intended design, they could positively contribute to the diversity of the libraries.

[138] The expected and observed frequency of the IGHJ usage was evaluated in the in-frame sequences in both libraries and showed good correspondence between the designed and observed frequency. Comparison of the expected vs observed frequency of the amino acids per diversified position was also conducted. Good agreement was found, with differences falling within the expected variations of trimer- based oligonucleotide synthesis.

[139] The two libraries were cloned starting from one microgram of each synthetic fragment 3-20/3-23 or 4-01/3-23, to generate PL1 or PL2, respectively. Digested fragments with Bgll restriction enzyme overnight at 37°C were ligated into pADL™-23c phagemid vector (Antibody Design Lab, San Diego).

The ligation reactions were electroporated into electro-competent TG1 cells (Lucigen) and transformants were rescued on 2xYT medium supplemented with ampicillin at 37°C in the presence of glucose 1% w/v.

The initial diversity was 1.7 x 10 ⁹ and 2.3 x 10 ⁹ primary transformants for PL1 and PL2, respectively.

[140] Cells were harvested after overnight incubation, resuspended in fresh 2xYT medium supplemented with ampicillin and subsequently superinfected with M13K07 helper phage. No more than 55 min after transduction, kanamycin 50 pg/ml was added, the temperature was lowered to 30°C and the incubation was prolonged overnight. The morning after, virions were purified by PEG precipitation following standard protocols to those skilled in the art and references herein described.

[141] Sanger sequencing of PL1 and PL2 clones selected at random from the primary transformants showed no background vectors in 44 clones for PL1 and PL2. The frequency of scFv insert in-frame was 61% for PL1 and 84.1% for PL2. These percentages were in good correlation with the frequencies observed in the synthetic fragments. Three in 44 clones (6%) of PL1 and one in 44 (2.3%) of PL2 were chimeric clones with two plasmids detected by sequencing.

EXAMPLE 5: BINDING OF PRIMARY LIBRARIES TO PROTEIN- A

[142] PL1 and PL2 binding to Protein A was assessed by a direct ELISA (Figure 7). Both PLs bound Protein A, with PL2 showing a higher EC ₅₀ than PL1. A higher EC ₅₀ in PL2 can be explained by a higher quality of the synthetic fragments since 4-01/3-23 synthetic fragments exhibited 60% of in-frame sequences, whereas, the 3-20/3-23 fragments had only 50%. Comparison with the control scaffolds having the CDR-H3 of CNTO 888 and displayed as sFv fusion to pill showed similar signal in intensity at saturation level and inflexion at dilutions 20 to 6 times lower for PL1 and PL2, respectively. This observation was consistent with the previous observation that only a fraction of the virions displays fusions on their surface.

EXAMPLE 6: PROTEIN-A FILTERING AFTER HEAT SHOCK

[143] To set up the optimal conditions for selection of developable antibody fragments, two sets of experiments were performed: (1) Parent scaffolds unfolding at diverse temperatures followed by capture with Protein A and (2) unfolding of diversified scaffold variants followed by capture with Protein A. In the first set of experiments the unfolding process of the parent scaffolds was studied. In the second set of experiments it was assessed how diversification of the scaffolds affected their folding and stability.

EXAMPLE 6.1 : SCAFFOLDS UNFOLDING

[144] The 3-20/3-23 and 4-01/3-23 control scFvs with the CDR-H3 of CNTO 888 displayed as fusion proteins on the phage surface were incubated for 10 min in a range of temperatures, starting at 40°C and increasing the temperature up to 80°C in steps of 5°C (Figure 8). The unfolding process was monitored with a direct Protein A ELISA as described in Example 5. The 3-20/3-23 control scFv started to unfold at 65°C. The 4-01/3-23 scFv started the unfolding process at 55°C. The Tm (temperature at which 50% of the scFvs are unfolded) of 3-20/3-23 scFv had a value of 75°C. The Tm of 4-01/3-23 scFv was 65°C.

[145] The unfolding kinetic at 60°C and 72°C of the 3-20/3-23 control scFv and 4-01/3-23 control scFv (Figure 9) demonstrated that the 3-20/3-23 control scFv remained folded at 60°C for up to one hour. At 72°C it unfolded slowly with approximately a 20% drop in the ELISA signal during the first 30 min. The unfolding kinetic of 4-01/3-23 scFv was significantly different. It unfolded slowly at 60 °C. At 72°C the unfolding process accelerated, with a drop in the ELISA signal of approximately 30% in the first 10 min.

[146] Therefore, this example demonstrates that: (i) the 3-20/3-23 and 4-01/3-23 control scFvs are stable at 60°C for at least one hour; and (ii) 3-20/3-23 and 4-01/3-23 scFvs can be incubated at 72°C for

10 min without significant unfolding and aggregation.

EXAMPLE 6.2: FOLDING & STABILITY OF DIVERSIFIED SCAFFOLDS

[147] The folding and stability of the diversified scaffolds were assessed by analyzing both Protein A binding and unfolding at high temperature on single clones taken randomly from the primary libraries. A subset of 44 clones from PL1 and 44 from PL2 were studied. Protein A was restricted to in-frame clones in both libraries (Figure 10 and Figure 11). Only 4 chimeric clones, seen by the superimposition of two sequencing trace and most probably deriving from a double transformation were found, 3 in PL 1 and 1 in PL2. These clones were part of the experimental process but were not included in the statistics.

[148] Because the Protein A binding signal in the ELISA exhibited an obvious saturation effect, most likely due to saturation of the Protein A binding sites at the bottom of the plate, binding were expressed as percentage of the respective control scaffold signal, ranging from no binding (e.g. clone A6 in PL1 or Dl in PL2) to 100% (Figure 10 to Figure 13). Protein A binding of single clones was defined by at least 10% of the signal of the control scaffold. The characteristics of the selected clones in each library are summarized below (Table 3).

TABLE 3

Library ¹ PL1 PL2

Percentage Clones Percentage Clones

Single Clones 93.2% 41/44 97.7% 43/44

Double Transformants 6.8% 3/44 2.1% 1/44 Vector Background 0.0% 0/44 0.0% 0/44 In-Frame Clones 61.0% 25/41 83.7% 36/43

Protein A Binding ² 58.5% 24/41 69.8% 30/43 Heat Shock Survival ³ 48.8% 20/41 55.8% 24/43

[149] Protein A binding of single phage results from multiple factors such as: (1) as proper folding of the Fv fragment, (2) concentration of displayed scFv and (3) scFv valency, up to 5 per M13 phage. It should be noted that the last two properties are linked to better expression levels, e.g. the higher the expression, the higher the valency. Therefore, Protein A binding is linked to more developable antibodies (better folded or better expressed or both). On average 24 out of 41 single clones in PL1 bound Protein A (58.5%) and 30 out of 43 single clones in PL2 (69.8%) (Table 3). Therefore, the vast majority of the assayed clones had a signal between 20% and 90% with respect to the control scaffolds, showing a good dynamic range to assess changes in Protein A binding after heat treatment.

[150] The same clones were treated for 10 min at 70°C and analyzed the residual Protein A binding signal (Figure 12 and Figure 13). As shown in Example 6.1, the control scaffold scaffolds started to unfold significantly in this temperature range. A preliminary analysis on a few clones at different temperatures showed a good compromise between clones that lost most of their binding and clones resistant to the heat shock. To prevent overestimation of thermal resistance, it was considered only clones with at least 25% binding to the control scaffold at 37°C. Again, survival was defined as having a residual binding to Protein A of at least 10% of the control scaffold. On average 20 out of 41 single clones in PL1 survived the heat treatment (48.8%) and 24 out of 43 single clones survived in PL2 (55.8%) (Table 3). This observation indicated that around -50% of the PLs were not functional after incubation at 70°C for 10 min.

EXAMPLE 7: PROTEIN A FILTRATION OF THE PRIMARY LIBRARIES

[151] Twenty PCR tubes containing each one 100 mΐ of virions from PL 1 at a concentration of 1.3 x 10 ¹³ virions/ml (1.3 x l(? virions per tube), were incubated at said temperature of TiC for a 10 min period in a PCR machine and cooled down for 30 min on ice. In the meantime, 2 ml of magnetic beads (BioMag® Streptavidin #84660-5 from Polysciences, Inc., Warrington, PA) were washed two times with TBS with Tween 20 0.1% (TBST) and incubated with 500 pg biotinylated Protein A (Life Technologies,

Cat #29989) for 30 min with agitation. After 3 washes with TBST, the beads were aliquoted in 10 microfuge tubes, resuspended in 400 pl TBST with non-fat dry milk 5% w/v and blocked for 30 min at room temperature. One-hundred mΐ of heated library were added to each tube and incubated for 2 h at room temperature on a rocker.

[152] After 5 washes with TBST and 5 washes with TBS, the bound phage were successively eluted with Trypsin-EDTA (Life Technologies, Cat# 25200056, 500 mΐ per tube) and glycine 0.1 mM, pH 2.7 containing BSA 1 mg/ml (500 mΐ per tube) for 10 min at room temperature. After neutralization of the acid eluate with Tris 1 M, pH 8.0, both eluates were combined and used to transfect XL10 Gold cells (Agilent Technologies, San Diego, CA). Bacteria were grown overnight at 37°C in 2xYT medium supplemented with ampicillin and in the presence of glucose 1% w/v. The day after, plasmid DNA were prepared using a DNA MIDI kit preparation (Macherey Nagel, Germany).

[153] For PL2, only 10 tubes were processed in a similar way for a total of half the number of initial virions. Bacterial transfectants incubated no later than 55 min post transfection were counted by dilutions on agar plates supplemented with ampicillin and glucose. The filtrated library FL1 had a total of 7.2 x 10 ⁹ primary transfectants and FL2 had 5.1 x 10 ⁹ primary transfectants.

[154] To assess the quality of Protein A filtering, a sample of 44 ampicillin-resistant colonies was selected at random in each library. Analysis by colony PCR for the presence of scFv showed that all 88 clones had an insert at the expected size. Further analysis by Sanger sequencing found seven double transformants in each library with the superimposition of two scFv sequence traces and only one clone out of frame in each library, giving an observed frequency of in-frame clones of 94.7% and 97.3% for FL1 and FL2, respectively, after excluding the double transformants. [155] Phage displaying scFv for the single clones were prepared as described in Example 6.2 and binding to Protein A was demonstrated by EFISA. COSTAR plates 3369 (Coming) were coated with Protein A (Sigma Aldrich, cat# P6031) at 4 pg/ml in TBS overnight at 4°C. After blocking with TBST and 5% w/v nonfat dry milk for one hour, virions (~2.6 x 10 ¹² virions/ml) in TBST with 5% w/v nonfat dry milk were added to the wells and incubated for 2 h at 37°C. As a reference, virions derived from the parent scaffolds with the CDR-H3 of CNTO 888 cloned in the same vector were added and similarly incubated on the plate. Bound phage was detected with A4G1.6 monoclonal antibody (Antibody Design Labs, San Diego, CA) conjugated to HRP. Binding of the secondary antibody, a murine IgGl, to Protein A was blocked by polyclonal human IgG at 100 pg/ml added into the incubation buffer as described in

Figure 7. The characteristic of the selected clones are summarized below (Table 4).

TABLE 4

Library FL1 FL2

Percentage Clones Percentage Clones

Single Clones 84.1% 37/44 84.1% 37/44

Double Transformants 15.9% 6/44 15.9% 7/44 Vector Background 0.0% 0/44 4.3% 0/44

In-Frame Clones 94.7% 36/38 97.3% 36/37

Frequency of Protein A Binding 89.5% 34/38 89.2% 33/37

[156] The percentage of clones binding Protein A with > 10% Protein A binding with respect to the control scaffold was of 91.9% for FL1 and 89.2% for FL2 (Figure 14 and Figure 15, and Table 4). The very high frequency of insert clones in both libraries together with a frequency of Protein A binders around 90%, almost 50% more than the frequencies observed in PLs (58.5% and 65.2% for PL1 and PL2, respectively), is a clear signature of the effectiveness of the Protein A filtration. Based on the infectivity of the phage displaying scFv, around 5% in the pADL-23/scFv/Ml3K07 system, it was found retrospectively that the MOI was above 1 during the infection of the XL 10 Gold cells, explaining a posteriori the high frequency of double transformants in the FLs. Since virions were not prepared from these libraries but just the plasmid DNA, this was not consequential on the quality of the SL libraries.

EXAMPLE 8: SECONDARY REPERTOIRES [157] In this example, we will create secondary libraries by replacing the H3J fragments of the filtrated libraries exemplified in Example 7 by a pool of natural H3JK fragment derived from a large pool of human donors.

EXAMPLE 8.1 : NATURAL H3 J FRAGMENTS FROM HUMAN REPERTOIRES

[158] The natural H3J fragments were obtained from the PBMCs of 200 healthy donors, 100 females and 100 males under the age of 40 years. Each donor provided 5 x 10 ⁶ cells and thus potentially yielded 1 x 10 ⁶ unique H3J sequences. Therefore, the pool of 200 donors contained 2 x 10 ⁸ potentially unique H3J sequences. Starting from the PBMCs, total RNA (tRNA) was individually isolated using Trizol (Invitrogen; Cat# 15596026 and 15596018). Pools of tRNAs from 10 donors were generated after determining the concentration by UV spectrophotometry and mixing the donor tRNAs in equal amounts to generate 20 tRNA pools.

[159] Each of the 20 pools were processed to isolate Messenger RNA (mRNA) using polyA Spin™ mRNA Isolation Kit (NEB, Cat #: S1560S) following the manufacturer instructions. The mRNA was used as template to generate cDNA by reverse transcription using OneTaq® RT-PCR Kit (NEB, Cat #:

E5310S) and a poly-T oligonucleotide.

[ 160] To amplify the natural CDR-H all human VH genes in the FR3 region were aligned immediately before the CDR-H3. This region of the heavy chain exhibits highly conserved amino acids with two tyrosines at position 90 and 91 and a cysteine at position 92 (Chothia and Lesk, 1987). This conservation is reflected as well at the nucleotide level. Second, the alignment was filtered by considering only VH genes commonly found in circulating PBMCs, more precisely VH genes that had been found in at least 0.6% of the human circulating antibody repertoire (Glanville J et al., 2009). This particular set of antibodies has a highly conserved nucleotide sequence GACACGGCYGTGTATTACTGTGC (SEQ ID NO: 94) located at the FR-3 - CDR-H3 junction. There is polymorphism with a C or a T for the third nucleotide of the alanine codon at position 88, hence the Y (UIPAC code for C or T) symbol in this sequence. This motif is present in more than 95% of the VH sequences found in circulating PBMCs.

This human VH conserved motif (CM) was used to amplify the H3J fragments from the cDNA of the donor pool.

[161] Double-stranded DNA containing the repertoire of natural H3J fragments was obtained by PCR using a universal forward primer annealing to the CM and three reverse primers designed to amplify 95% of the human CDR-H3 fragments in circulating PBMCs (SEQ ID Nos: 95-97).

[162] The quality of H3J fragments was assessed by cloning an aliquot of final pool into a TOPO vector (Life Technologies). Sanger sequencing of 30 clones indicated that all H3J fragments were different, with length variation resembling the human CDR-H3 repertoire. The region introduced by the amplification primers for assembling the full scFvs and cloning into the vector matched 100% the expected nucleotide sequence.

EXAMPLE 8.2: CLONING OF A SECONDARY REPERTOIRE BY SEAMLESS ASSEMBLY

[163] The strategy for seamless assembly of the SLs is highlighted on Figure 16. The region corresponding to V _L , GS19 linker peptide and V _H just before the CM motif (Figure 16) was amplified by PCR from the DNA purified from the filtered primary libraries with the sfiFOR primer taken on the pelB leader-encoding sequence (SEQ ID NO: 98) and the ALT_hu3_23FR3_r primer annealing just before the CM motif (SEQ ID NO: 99). ALT_hu3_23FR3_r is extended on its 5’end by the sequence 5’- CACAGGTCTCG. This sequence contains a Bsa-I site, which, after digestion, creates a 4-base pair overhangs TGTC complementary to the first four nucleotides GACA of the CM motif.

[164] The second step of the scFv assembly involved a simultaneous digestion by Bsal and ligation by T4 DNA ligase at 37°C which joined the primary filtered fragments and the natural H3J fragments. The resulting full length scFv DNA was further amplified by PCR before cloning (Figure 16).

[165] Amplification of the primary filtered library DNA from library FL1 and FL2 with 17 cycles and an annealing temperature of 60°C using Phusion (New England Biolabs, MA) yielded single bands at the expected size (Figure 17, Panel A). For each library, natural H3J fragments (150 ng) were assembled with the primary filtered DNA fragments (600 ng) by simultaneous digestion with Bsal and ligation with T4 ligase for 4 h at 37°C. The natural H3J were in a slight excess and conversion of the filtered fragments was near 100% (Figure 17, Panel B). 60 ng of the joined products were further amplified by the nested primers padllib s (SEQ ID NO: 101) and ALT_huH3J_r (SEQ ID NO: 102) in a final volume of 300 pi using Phusion, 20 cycles and an annealing temperature of 60°C (Figure 17, Panel C).

[166] The two semisynthetic fragments so generated were ligated into the pADL-23c phagemid vector.

A total of 5 pg of ligated products was electroporated into electro-competent TG1 cells (800 ng/50 pi cells) as described in Example 1 for each secondary library. The number of primary transformants for the secondary libraries was 1.4 x 10 ¹⁰ cfii for the SL1 library and 1.1 x 10 ¹⁰ cfii for the SL2 library.

[167] The quality of each secondary library was assessed by Sanger sequencing of 44 clones picked randomly from a separate electroporation made with 10 times less DNA (80 ng/50 mΐ cells). The results are summarized in Table 5 below. As in Table 3, binding to Protein A is again defined as more than 10% binding relative to the respective control scaffold. We found 3 double transformants in each SL1 and SL2 libraries. These clones were excluded from the calculations. Finally, it was found that 35/41 (85.4%) were in-frame for the SL1 secondary library and 35/41 (85.4%) for the SL2 secondary library. These values were -10% lower than those of the filtered libraries indicating that the method of secondary library construction was able to retain most of the in-frame character of the filtrated libraries while adding natural H3J diversity in the heavy chain, thus compensating the loss of diversity during the step of filtration.

TABLE 5

Library SL1 SL2

Percentage Clones Percentage Clones

Single Clones 95.5% 42/44 95.5% 40/44 Double Transformants 6.8% 3/44 6.8% 3/44

Vector Background 0.0% 0/41 0% 0/40

In-Frame Clones 85.4% 35/41 85.4% 35/41 Frequency of Protein A Binding 82.9% 34/41 85.4% 35/41

Heat Shock Survival 68.3% 28/41 70.7% 29/41

EXAMPLES 8.3 : FOLDING & STABILITY OF THE SECONDARY LIBRARIES

[168] The 44 clones of each secondary library were tested for Protein A binding and survival after a heat shock at 70°C for 10 min (Figure 19 and Figure 20). All in-frame clones but one in the SL1 library gave strong binding to Protein A, giving a frequency of Protein A binding of 85.4% in that library. All in-frame clones in SL2 library were binding Protein A, giving a frequency of Protein A binding of 85.4% in that library. Therefore, not only in-frame clones that were selected by Protein A during the filtration remained in-frame but were for most of them were still binding Protein A after replacing the H3J sequences. Analysis of survival to the heat shock of the clones for clones having 25% or more binding to Protein A relative to the respective control scaffold prior to the heat shock (Figure 20 and Figure 21) found 28/41 (68.3%) for SL1 library, 19.5% better than PL1, and 29/41 (70.7%) for SL2, 14.9% better than the corresponding PL1. The same analysis for all clones found that 30/41 or 73.1% of the SL1 and 27/40 or 67.5% of the SL2 were surviving the heat shock. This was a significant increase comparatively to the PLs with 39.9% for PL1 and 33.3% for PL2 (Figure 22).

EXAMPLE 9: ASSESSING DIVERSITY OF THE LIBRARIES BY NEXT GENERATION SEQUENCING [169] To study the diversity along the filtration process and the construction of the secondary libraries, two amplicons were prepared from the phagemid DNA. Plasmid DNA were isolated using QIAGEN Plasmid Midi Kit (Cat No.: 12143) from the PL and SL bacterial cultures prior to helper phage superinfection and induction, and after overnight bacterial culture for the FL libraries. The DNA was used as a template to generate amplicons of approximately 300 bps. Amplicon 1 covered the V _L scaffolds (3-20 and 4-01) and was amplified with one forward primer for PL1 (SEQ ID 126) and PL2 (SEQ ID

127) and a reverse primer located in the J region (SEQ ID 128). This amplicon included the three V _L CDRs. Amplicon 2 covered the CDR-H2 and the H3J fragments and was amplified with one forward primer in FR1 region (SEQ ID 129) and one reverse primer located in the vector sequence immediately after the Sfil site (SEQ ID 130). The PCR reactions were performed as follows: 5 min start a€ 95 followed by 10 cycles of 1 min at 95°C, 1 min at 67°C, 1 min at 72°C and terminated by a lO-min extension at 72°C. The PCR fragments were gel-purified using QIAquick PCR Purification Kit (Cat No.: 28104) and used as template to prepare the samples for NGS following the manufacturer instructions. The sequencing was performed on a Miseq platform from Illumina. FASTQ files were processed with the software AptaAnalyzer™ (AptalT; Germany) using the BCR (B-cell receptors) functionality.

[170] The accepted output sequences were further curated with in-house Java scripts. In brief, for amplicon 1, sequences having insertions or deletions of two amino acids or more were discarded. For amplicon 2 sequences before FR-3 region having insertions or deletions of more than two amino acids were discarded and for CDR-H3, sequences not having the conserved cysteine H88 and tryptophan H102 were removed from the analysis as well.

EXAMPLE 9.1 : NGS ANALYSIS OF THE PRIMARY LIBRARIES AND THE FILTRATION PROCESS

[171] Over half million sequences were obtained from PL1 amplicon 1 and close to a million and a half for FL1 amplicon 1 (Table 6). A clear difference toward curated sequences from 76% in PL1 to up 95% in FL1 was observed, in agreement with an increase of productive sequences as a result of the filtration process. For PL2 and FL2, the total number of sequences was higher than for PL1, with over a million and a half for PL2 and over two million sequences for FL2 amplicon 1 (Table 7). The number of curated sequences in both PL2 and FL2 was higher than PL1 (88%) and similar for FL1 (90%). This observation agreed with the observation that the quality of synthetic PL2 fragments was superior to that of PL1. TABLE 6

Amplicon 1 PL1 FL1

Count % Count %

Total sequences 533,828 1,447,032

Accepted sequences 497,359 100.00 1,349,887 100.00 Curated Sequences 379,834 76.37 1,284,091 95.13 Unique sequences 222,465 44.73 483,262 35.80

TABLE 7

Amplicon 1 PL2 FL2

Count % Count %

Total sequences 1,466,411 2,249,262

Accepted sequences 1,341,533 100.00 2,172,872 100.00

Curated Sequences 1,285,500 95.82 1,984,567 91.33

Unique sequences 766,255 57.12 1,014,666 46.70

[172] The number of total sequences in amplicon 2 for PLs and FLs was similar with close to one million and a half sequences for each library (Table 8 and 9). The curated sequences were also similar, close to ~ 90%.

TABLE 8

Amplicon 2 PL1 FL1

Count % Count %

Total sequences 1,757,878 1,509,376

Accepted sequences 1,490,192 100.00 1,308,501 100.00

Curated unique sequences 695,605 46.68 584,346 44.66

TABLE 9

Amplicon 2 PL2 FL2

Count % Count % Total sequences 1,637,123 1,389,471

Accepted sequences 1,330,040 100.00 1,242,319 100.00 Curated unique sequences 656,092 49.33 586,474 47.21

[173] In both amplicons, the percentage of unique sequences was on average 50% in the PLs, providing a large coverage of the primary diversity. This percentage decreased on average of 5.8% during the filtration process, pointing to a net loss of diversity after treatment at 70°C for 10 min and re- amplification of the libraries.

[174] Analysis of the frequency of the neutral IGDH fragments contained in amplicon 2 before and after the filtration process (Table 10) indicated that all the designed fragments were represented in the

PLs. The frequency was within 1.3% of expected value of 6% of an even distribution (18 neutral fragments) and was similar in both PLs, suggesting no major bias due to the PLs preparation.

[175] After the filtration process the difference in frequency between PLs and FLs of some neutral IGDH fragments increased by more than 25% in both FLs with respect to the PLs, whereas others decreased by a similar frequency or more (Table 11). Both FL1 and FL2 showed a similar trend, although FL1 showed overall variations. Of particular interest was the IGDH fragment“VDIVATI”, which decreased by close to 75% in PL1 and slightly over 50% in PL2. This fragment contains five hydrophobic amino acids out of 7, suggesting an unfavorable selection for hydrophobic residues during the filtration process. Of the same trend was the consistent decrease after the filtration process of the longest IGDH fragments, all rich in tyrosine residues, and increase of the most hydrophilic IGDH “GITGT” and“GTTGT”

TABLE 10

Neutral D sequence PL1 PL2 FL1 FL2

% % % %

GYSGYDY 6.84% 6.83% 8.29% 7.90%

GYSYGY 6.88% 6.86% 7.12% 6.66%

TTVT 5.45% 5.53% 4.87% 5.00%

YSGSYY 4.63% 5.28% 4.49% 5.08%

DYGDY 5.51% 5.59% 6.07% 6.11%

DYSNY 5.53% 5.62% 5.76% 5.70%

SIAAR 5.40% 5.32% 6.08% 6.17%

VQLER 4.38% 4.41% 3.22% 4.03%

YYDILTGYYN 4.73% 4.43% 2.97% 3.28%

YYYDSSGYYY 4.53% 4.20% 1.94% 2.05% YYYGSGSYYN 4.82% 4.46% 3.66% 3.48%

[176] The five IGHJ germline genes were also within 4% of their even distribution (20%) with a slight 5% increase of IGHJ3 and IGHJ6 at the expense of IGHJ1 and IGHJ4 after filtration (Table 11). Again, the IGHJ frequencies were similar in both libraries, consistent with the suggestion that no major bias resulted from the PLs preparation.

TABLE 11

JH PL1 PL2 FL1 FL2

% % % %

JH1 20.51% 19.90% 17.10% 17.40%

JH3 21.02% 22.03% 27.02% 25.96%

JH4 20.51% 19.87% 14.84% 15.16%

JH5 21.63% 21.31% 22.00% 21.93%

JH6 16.33% 16.90% 19.04% 19.55%

JH1 20.51% 19.90% 17.10% 17.40%

EXAMPLE 9.2: NGS ANALYSIS OF THE SECONDARY LIBRARIES

[177] Over one million primary sequences were obtained for both SL1 and SL2. The number of accepted sequences was around 80% in both SLs, thus providing a large sampling of the secondary libraries at -850,000 and -990,000 sequence depth for SL1 and SL2, respectively (Table 12). Around 60% of these sequences had unique natural H3J fragments in each library. The top 30 most frequent sequences in each library barely added up to 1% of the total accepted sequences, indicating that the remaining 99% of all CDR-H3 sequences had only a few copies. Only half of the most prevalent 30 sequences were shared by the two libraries, pointing to the large diversity of CDR-H3 in the SLs.

TABLE 12

Amplicon 2 SL1 SL2

Count % Count %

Total sequences 1,077,746 100.00 1,227,176 100.00

Accepted sequences 853,327 79.18 990,864 80.74

Unique sequences (Synthetic

99.81 989,005 99.81

diversity and natural H3J fragments)

Unique CDR-H3 sequences 522,939 61.28 592,317 59.78

[178] The natural CDR-H3 length distribution followed a Gaussian curve typical of the human CDR- H3 sequences with an expected peak at 12 amino acids in length (Rabat’s definition, Table 13). No difference was observed between SL1 and SL2, indicating that the pool of natural H3J fragments from the 200 donors was not biased towards a particular CDR-H3 length during the SL preparation.

TABUE 13

CDR-H3 length SL1 SL2

Count % Count %

<4 3660 0.4 4588 0.5

4 3260 0.4 4179 0.4

5 4193 0.5 5333 0.5

6 9366 1.1 11640 1.2

7 19401 2.3 23589 2.4

8 32294 3.8 39128 3.9

9 52001 6.1 61615 6.2

10 69709 8.2 82621 8.3

11 84753 9.9 99212 10.0

12 99841 11.7 116562 11.8

13 95688 11.2 109993 11.1

14 85577 10.0 99002 10.0

15 76779 9.0 87474 8.8

16 66070 7.7 75163 7.6

17 54963 6.4 62589 6.3

18 43447 5.1 49095 5.0

19 31482 3.7 35744 3.6 20 19775 2.3 21975 2.2

>20 1093 0.1 1322 0.1

[179] The distribution of the IGHJ1-5 segments (Table 14) followed the expected usage seen in human antibody sequences except for IGHJ6. The reported IGJH6 frequency is ~ 40%, followed by IGHJ4 with ~ 30% (Amaout et al., 2011). In the SLs, IGHJ4 was the most prevalent joining region with -40%, followed by the IGHJ6 with 20%. However, it was so designed during the RT-PCR amplification of the natural H3J fragments since the IGHJ6 encodes a stretch of five Y residues in the N-terminal region, which could lead to a potential destabilizing effect. Thus, by lowering the proportion of natural H3J fragments encoded by IGHJ6 it was expected to select for a higher number of developable antibodies.

TABLE 14

JH SL1 SL2

Count % Count %

IGHJ4*0l 337668 40 392998 40

IGHJ6*0l 174385 20 201522 20

IGHJ3*02 94067 11 108560 11

IGHJ5*02 78043 9 89640 9

IGHJl*0l 55861 7 65708 7

IGHJ2*0l 42988 5 49448 5

IGHJ3*0l 24237 3 28729 3

IGHJ6*03 23977 3 27576 3

IGHJ5*0l 20613 2 24887 3

IGHJ6*04 1488 0 1796 0

[180] When the CDR-H2 diversity was analyzed together with the CDR-H3 diversity, more than 99% of all accepted sequences were unique (Table 10), in clear increase from the filtrated libraries and pointing to the exquisite diversity of the SLs, thus confirming at the NGS level the successful increase in sequence diversity achieved during the construction of the secondary libraries.

EXAMPLE 10: ASSES SING FUNCTIONALITY OF THE SECONDARY LIBRARIES USING TWO TARGET MODELS [181] To assess the potential of the secondary libraries to produce specific and developable antibodies, selections were performed with two non-related target models: TNF and HSA.

EXAMPLE 10.1 : SELECTION WITH TNF

[ 182] The selections with TNF were performed in solid phase following protocols well known to those skilled in art. In brief, one Immunotube (Thermofisher cat #: 341866) was coated with 4 ml of TNF at a concentration of 50 pg/ml in Carbonate Buffer (NaHC0 ₃ 50 mM, pH 9.6) for one hour at room temperature (RT). To avoid non-specific interactions, the Immunotube was blocked with MPBS (PBS +

3% w/v skim powder milk) for one additional hour at RT and washed three times with PBS. For the first round of selection, the SL1 and SL2 libraries (3 x ll phage) were mixed and diluted to 1 x ltf virion/ml in MPBS and 4 ml were added to the TNF-coated Immunotube. The Immunotube was incubated one hour at RT with slow shaking and one additional hour standing at RT. The unbound phages were washed away 10 times with TPBS (PBS + 0.1% Tween ® 20) washes and 10 washes with PBS. Bound phages were eluted with 0.2 M glycine/HC!, pH 2.2 and used to infect exponentially growing TG1 cells (OD _{600 nm} = 0.4). Infected cells were grown overnight at 37°C in 2xYT-agar plates containing 100 pg/ml carbenicillin and 1% w/v glucose. The cells were scrapped from the plate with 2xYT medium, expanded in 2xYT medium supplemented with carbenicillin and 1% w/v glucose, and superinfected with M13K07 helper phage. After exchanging the medium with fresh 2xYT without glucose, cells were incubated overnight in the presence of carbenicillin 100 pg/ml and kanamycin 50 pg/ml at 30°C and 250 rpm. Virions were then purified by PEG-precipitation and used as input at a concentration of 1 x 10 ¹² virion/ml for the following round of selection as described above.

[183] After the fourth round of panning, soluble scFvs of 45 clones chosen at random were assayed in ELISA for binding to TNF and BSA. As reporter reagent, Protein A/HRP was used. 12 positive clones for TNF but not for BSA were sent for Sanger sequencing to determine unique clones. Two unique clones were expressed as soluble scFvs and re-tested for binding to the target and BSA. The confirmed clones were expressed in lOO-mL culture and purified using HisTrap ^™ (Sigma; GE Catalogue # GE 17-5255-01 ).

The purified scFvs exhibited specific binding to the target in a direct ELISA as shown on Figure 23 or clone E12.

EXAMPLE 10.2: SELECTION WITH HS A [ 184] The selections with HSA were conducted following a similar procedure as for TNF with some modifications. That is, only SL2 was used for the selections and a COSTAR plate 3369 (Coming) instead of Immunotubes was used as solid phase. Ten wells were coated with 100 pl per well of HSA 4 pg/ml in PBS overnight at 4°C. After washing 3 times with PBS and blocking with 200 pl TPBS with BSA 3% w/v for one hour at 37°C, 100 mΐ of SL2 phage, 6.1 x 10 ¹¹ cfu/ml, per well in TPBS with BSA 3% w/v were incubated at 37°C for 2 h. After washing 5 times with TPBS and 5 times with TBS, bound phage were eluted by 100 mΐ glycine/HCl pH 2.2 BSA 1 mg/ml for 10 min at room temperature. Eluted fractions were collected and neutralized with Tris/HCl pH 8.0 1 M and used to infect exponentially growing TG1 cells as for the TNF screening.

[185] After 3 consecutive rounds of selection, 48 clones were analyzed by direct phage ELISA in 96- well format for binding to HSA with BSA as a negative control. 40 clones out of 48 showed a strong and specific binding to HSA (hit rate of 83%). Sanger sequencing of the 40 positive clones yielded 3 unique scFv sequences showing specific and intense binding to HSA in a phage ELISA and no binding to BSA as a negative control.

EXAMPLE 11 : SEQUENCES

[186] As mentioned throughout the aforementioned examples, several sequences were used in the construction of libraries as discussed in this application. Exemplary sequences to be used in the foregoing examples include those shown below, in Table 15.

TABLE 15

[187] While the foregoing invention has been described in some detail for purposes of clarity and understanding, it will be clear to one skilled in the art from a reading of this disclosure that various changes in form and detail can be made without departing from the true scope of the invention. For example, all the techniques and apparatus described above can be used in various combinations.

REFERENCES

[188] Almagro, J.C., Teplyakov, A., Luo, E, Sweet, R.W., Kodangattil, S., Hemandez-Guzman, F., et

al. (2014). Second antibody modeling assessment (AMA -II). Proteins 82(8), 1553-1562. doi:

10.1002/p rot.24567.

[189] Amaout R, Lee W, Cahill P, Honan T, Sparrow T, Weiand M, et al. High-resolution description

of antibody heavy-chain repertoires in humans. PLoS One 2011; 6:e22365.

[190] Beerli, R.R., Bauer, M., Buser, R.B., Gwerder, M., Muntwiler, S., Maurer, P., et al. (2008).

Isolation of human monoclonal antibodies by mammalian cell display. Proc Natl Acad Sci USA 105(38), 14336-14341. doi: l0.l073/pnas.0805942l05. [191] Bethea, D., Wu, S.J., Luo, L, Hyun, L., Lacy, E.R., Teplyakov, A., et al. (2012). Mechanisms of self-association of a human monoclonal antibody CNTO607. Protein Eng Des Sel 25(10), 531 - 537. doi: l0.l093/protein/gzs047.

[192] Bird, R.E., Hardman, K.D., Jacobson, J.W., Johnson, S., Kaufman, B.M., Lee, S.M., et al.

(1988). Single-chain antigen-binding proteins. Science 242(4877), 423-426.

[193] Cherf, G.M., and Cochran, J.R. (2015). Applications of Yeast Surface Display for Protein Engineering. Methods Mol Biol 1319, 155-175. doi: l0. l007/978-l-4939-2748-7_8.

[194] Ching, K.H., Collarini, E.J., Abdiche, Y.N., Bedinger, D., Pedersen, D., Izquierdo, S., et al.

(2017). Chickens with humanized immunoglobulin genes generate antibodies with high affinity and broad epitope coverage to conserved targets. MAbs , 1-10. doi:

10.1080/19420862.2017.1386825.

[195] Chothia, C., and Lesk, A.M. (1987). Canonical structures for the hypervariable regions of immunoglobulins. J Mol Biol 196(4), 901-917.

[196] de Haard, H.J., van Neer, N., Reurs, A., Hufton, S.E., Roovers, R.C., Henderikx, P., et al. (1999).

A large non-immunized human Fab fragment phage library that permits rapid isolation and kinetic analysis ofhigh affinity antibodies. J Biol Chem 274(26), 18218-18230.

[197] Emmons, C., and Hunsicker, L.G. (1987). Muromonab-CD3 (Orthoclone OKT3): the first monoclonal antibody approved for therapeutic use. Iowa Med 77(2), 78-82.

[198] Famm, K., Hansen, L., Christ, D., and Winter, G. (2008). Thermodynamically stable aggregation-resistant antibody domains through directed evolution. JMol Biol 376(4), 926-931. doi: 10.10 l6/j.jmb.2007.10.075.

[199] Finlay, W.J., and Almagro, J.C. (2012). Natural and man-made V-gene repertoires for antibody discovery. Front Immunol 3, 342. doi: 10.3389/fimmu.20l2.00342.

[200] Francisco, J.A., Campbell, R., Iverson, B.L., and Georgiou, G. (1993). Production and fluorescence-activated cell sorting of Escherichia coli expressing a functional antibody fragment on the external surface. Proc Natl Acad Sci U SA 90(22), 10444-10448.

[201] Gilliland, G.L., Luo, J., Vafa, O., and Almagro, J.C. (2012). Leveraging SBDD in protein therapeutic development: antibody engineering. Methods Mol Biol 841, 321-349. doi:

10.1007/978-1-61779-520-6 _ 14.

[202] Glanville J., Zhai W., Berka J., Telman D., Huerta G., Mehta G.R., Ni L, Mei L., Sundar P.D.,

Day G.M., Cox D., Rajpal A., and Pons J. (2009). Precise determination of the diversity of a combinatorial antibody library gives insight into the human immunoglobulin repertoire. Proc Natl Acad Sci USA. 106(48):20216-2 . [203] Green, L.L., Hardy, M.C., Maynard-Currie, C.E., Tsuda, H., Louie, D.M., Mendez, M.J., et al.

(1994). Antigen-specific human monoclonal antibodies from mice engineered with human Ig heavy and light chain YACs. Nat Genet 7(1), 13-21. doi: l0. l038/ng0594-l3.

[204] Green, L.L., and Jakobovits, A. (1998). Regulation of B cell development by variable gene complexity in mice reconstituted with human immunoglobulin yeast artificial chromosomes. J Exp Med 188(3), 483-495.

[205] Griffiths, A.D., Williams, S.C., Hartley, O., Tomlinson, I.M., Waterhouse, P., Crosby, W.L., et al.

(1994). Isolation of high affinity human antibodies directly from large synthetic repertoires. EMBO J 13(14), 3245-3260.

[206] Hanes, J., and Pluckthun, A. (1997). In vitro selection and evolution of functional proteins by

using ribosome display. Proc Natl Acad Sci USA 94(10), 4937-4942.

[207] Hoet, R.M., Cohen, E.H., Kent, R.B., Rookey, K., Schoonbroodt, S., Hogan, S., et al. (2005).

Generation of high-affinity human antibodies by combining donor-derived and synthetic complementarity-determining-region diversity. Nat Biotechnol 23(3), 344-348. doi: l0.l038/nbtl067.

[208] Hoogenboom, H.R. (2005). Selecting and screening recombinant antibody libraries. Nat Biotechnol 23(9), 1105-1116. doi: l0. l038/nbtl l26.

[209] Hussack, G., Hirama, T., Ding, W., Mackenzie, R., and Tanha, J. (2011). Engineered single

domain antibodies with high protease resistance and thermal stability. PLoS One 6(11), e282l8. doi: 10. l37l/joumal.pone.0028218.

[210] Jespers, L., Schon, O., Famm, K., and Winter, G. (2004). Aggregation-resistant domain antibodies selected on phage by heat denaturation. Nat Biotechnol 22(9), 1161-1165. doi: l0.l038/nbtl000.

[211] Jones, P.T., Dear, P.H., Foote, J., Neuberger, M.S., and Winter, G. (1986). Replacing the complementarity-determining regions in a human antibody with those from a mouse. Nature 321(6069), 522-525.

[212] Kelly, R.L., Geoghegan, J.C., Feldman, J., Jain, T., Kauke, M., Le, D., et al. (2017). Chaperone proteins as single component reagents to assess antibody nonspecificity. MAbs 9(7), 1036-1040. doi: 10.1080/19420862.2017.1356529.

[213] Kimball, J.A., Norman, D.J., Shield, C.F., Schroeder, T.J., Lisi, P., Garovoy, M., et al. (1993).

OKT3 antibody response study (OARS): a multicenter comparative study. Transplant Proc 25(1 Pt 1), 558-560.

[214] Knappik, A., Ge, L., Honegger, A., Pack, P., Fischer, M., Wellnhofer, G., et al. (2000). Fully synthetic human combinatorial antibody libraries (HuCAL) based on modular consensus frameworks and CDRs randomized with trinucleotides. J Mol Biol 296(1), 57-86. doi:

10. l006/jmbi.1999.3444.

[215] Kohler, G., and Milstein, C. (1975). Continuous cultures of fused cells secreting antibody of

predefined specificity. Nature 256, 495-497.

[216] Langone, J.J. (1982). Protein A of Staphylococcus aureus and related immunoglobulin receptors produced by streptococci and pneumonococci. Adv Immunol 32, 157-252.

[217] Lonberg, N., Taylor, L.D., Harding, F.A., Trounstine, M., Higgins, K.M., Schramm, S.R., et al.

(1994). Antigen-specific human antibodies from mice comprising four distinct genetic modifications. Nature 368(6474), 856-859. doi: l0.l038/368856a0.

[218] Ma, B., Osborn, M.J., Avis, S., Ouisse, L.H., Menoret, S., Anegon, I., et al. (2013). Human antibody expression in transgenic rats: comparison of chimeric IgH loci with human VH, D and

JH but bearing different rat C-gene regions. J Immunol Methods 400-401, 78-86. doi: l0.l0l6/j .jim.20l3.10.007.

[219] Marks, J.D., Hoogenboom, H.R., Bonnert, T.P., McCafferty, J., Griffiths, A.D., and Winter, G.

(1991). By-passing immunization. Human antibodies from V-gene libraries displayed on phage. JMol Biol 222(3), 581-597.

[220] Matochko, W.L., Cory Li, S., Tang, S.K., and Derda, R. (2014). Prospective identification of

parasitic sequences in phage display screens. Nucleic Acids Res 42(3), 1784-1798. doi:

10.1093/nar/gkt 1104.

[221] McCafferty, J., Griffiths, A.D., Winter, G., and Chiswell, D.J. (1990). Phage antibodies: filamentous phage displaying antibody variable domains. Nature 348(6301), 552-554.

[222] Morrison, S.L., Johnson, M.J., Herzenberg, L.A., and Oi, V.T. (1984). Chimeric human antibody molecules: mouse antigen-binding domains with human constant region domains. Proc Natl Acad Sci USA 81(21), 6851-6855.

[223] Nechansky, A. (2010). HAHA— nothing to laugh about. Measuring the immunogenicity (human anti-human antibody response) induced by humanized monoclonal antibodies applying ELISA and SPR technology. J Pharm Biomed Anal 51(1), 252-254. doi: l0T0l6/j.jpba.2009.07.0l3.

[224] Nilson, B.H., Solomon, A., Bjorck, L., and Akerstrom, B. (1992). Protein L from Peptostreptococcus magnus binds to the kappa light chain variable domain. JBiol Chem 267(4), 2234-2239.

[225] Obmolova, G., Teplyakov, A., Malia, T.J., Grygiel, T.L., Sweet, R., Snyder, L.A., et al. (2012).

Structural basis for high selectivity of anti-CCL2 neutralizing antibody CNTO 888. Mol Immunol 51(2), 227-233. doi: l0. l0l6/j.molimm.20l2.03.022.

[226] Perelson, A.S., and Oster, G.F. (1979). Theoretical studies of clonal selection: minimal antibody repertoire size and reliability of self-non-self discrimination. J Theor Biol 81(4), 645-670.

[227] Pruzina, S., Williams, G.T., Kaneva, G., Davies, S.L., Martin-Lopez, A., Bruggemann, M., et al.

(2011). Human monoclonal antibodies to HIV-l gpl40 from mice bearing YAC-based human immunoglobulin transloci. Protein Eng Des Sel 24(10), 791-799. doi: 10.1093/p rotein/gzr038.

[228] Raghunathan, G., Smart, J., Williams, J., and Almagro, J.C. (2012). Antigen-binding site anatomy and somatic mutations in antibodies that recognize different types of antigens. J Mol Recognit 25(3), 103-113. doi: l0.l002/jmr.2l58.

[229] Reichert, J.M. (2017). Antibodies to watch in 2017. MAbs 9(2), 167-181. doi:

10.1080/19420862.2016.1269580.

[230] Rouet, R., Lowe, D., and Christ, D. (2014). Stability engineering of the human antibody repertoire. FEBSLett 588(2), 269-277. doi: 10.1016/j febslet.2013.11.029.

[231] Shawler, D.L., Bartholomew, R.M., Smith, L.M., and Dillman, R.O. (1985). Human immune

response to multiple injections of murine monoclonal IgG. J Immunol 135(2), 1530-1535.

[232] Shi, L., Wheeler, J.C., Sweet, R.W., Lu, J., Luo, J., Tometta, M., et al. (2010). De novo selection of high -affinity antibodies from synthetic fab libraries displayed on phage as pIX fusion proteins. JMol Biol 397(2), 385-396. doi: l0.l0l6/j.jmb.20l0.0L034.

[233] Smith, S.L. (1996). Ten years of Orthoclone OKT3 (muromonab-CD3): a review. JTranspl

Coord 6(3), 109-119; quiz 120-101.

[234] Sondek, J., and Shortle, D. (1992). A general strategy for random insertion and substitution mutagenesis: substoichiometric coupling of trinucleotide phosphoramidites. Proc Natl Acad Sci USA 89(8), 3581-3585.

[235] Strohl, W.R. (2017). Current progress in innovative engineered antibodies. Protein Cell doi:

10.1007/s 13238-017-0457-8.

[236] Teplyakov, A., Obmolova, G., Malia, T.J., Luo, J., Muzammil, S., Sweet, R., et al. (2016).

Structural diversity in a human antibody germline library. MAbs 8(6), 1045-1063. doi: 10.1080/19420862.2016.1190060.

[237] Vargas-Madrazo, E., Lara-Ochoa, F., and Almagro, J.C. (1995). Canonical structure repertoire of the antigen-binding site of immunoglobulins suggests strong geometrical restrictions associated to the mechanism of immune recognition. JMol Biol 254(3), 497-504.

[238] Vaughan, T.J., Williams, A.J., Pritchard, K., Osbourn, J.K., Pope, A.R., Eamshaw, J.C., et al.

(1996). Human antibodies with sub-nanomolar affinities isolated from a large non-immunized phage display library. Nat Biotechnol 14(3), 309-314. doi: l0. l038/nbt0396-309.

All publications, patents, patent applications, and/or other documents cited in this application are incorporated by reference in their entirety for all purposes to the same extent as if each individual publication, patent, patent application, and/or other document were individually indicated to be incorporated by reference for all purposes.

Individual Applicant

Street : 4650 Lisann Street

City : San Diego

State : CA

Country : US

PostalCode : 92117

PhoneNumber : 858 945-4740

FaxNumber :

EmailAddress : pvaladon@abdesignlabs.com

<110> LastName : Valadon

<110> FirstName : Philippe

<110> Middlelnitial :

<110> Suffix :

Individual Applicant

Street : 320 Concord Avenue

City : Cambridge

State : MA

Country : US

PostalCode : 02138

PhoneNumber : 617 710-4487

FaxNumber :

EmailAddress : juan.c.almagro@gmail.com

<110> LastName : ALMAGRO

<110> FirstName : Juan Carlos

<110> Middlelnitial :

<110> Suffix :

Application Project

<120> Title : Highly Functional Antibody Libraries

<130> AppFileReference : PCT

<140> CurrentAppNumber :

<141> CurrentFilingDate :

Sequence

<213> OrganismName : artificial sequence

<400> PreSequenceString :

gaaattgtgt tgacgcagtc tccaggcacc ctgtctttgt ctccagggga acgtgccacc 60 ctctcctgcc gtgccagtca gagtgttagc agcagctact tagcctggta tcagcagaaa 120 cctggccagg ctccccgact cctcatctat ggtgcatcta gccgtgccac tggtatccca 180 gaccgtttca gtggcagtgg gtctgggaca gacttcactc tcaccatcag cagactggag 240 cctgaagatt ttgcagtgta ttactgtcag cagtatggta gctcacctct gacgttcggc 300 caaggtacca aggtggaaat caaa 324

<212> Type : DNA

<211> Length : 324

SequenceName : 1

SequenceDescription :

Sequence

<213> OrganismName : artificial sequence

<400> PreSequenceString :

gacatcgtga tgacccagtc tccagactcc ctggctgtgt ctctgggcga gcgtgccacc 60 atcaactgca agtccagcca gagtgtttta tacagctcca acaataagaa ctacttagct 120 tggtatcagc agaaaccagg acagcctcct aagctgctca tttactgggc atctacccgg 180 gaatccgggg tccctgaccg attcagtggc agcgggtctg ggacagattt cactctcacc 240 atcagcagcc tgcaggctga agatgtggca gtttattact gtcagcaata ttatagtact 300 cctctgacgt tcggccaagg taccaaggtg gaaatcaaa 339

<212> Type : DNA

<211> Length : 339

SequenceName : 2

SequenceDescription :

Sequence

<213> OrganisrnName : artificial sequence

<400> PreSequenceString :

gaggtgcagc tgttggagtc tgggggaggc ttggtacagc ctggggggtc cctgcgactc 60 tcctgtgcag cctctggatt cacctttagc agctatgcca tgagctgggt ccgccaggct 120 ccagggaagg ggctggagtg ggtgtcagct attagtggta gtggtggtag cacatactac 180 gcagactccg tgaagggccg gttcaccatc tcccgtgaca attccaagaa cacgctgtat 240 ctgcaaatga acagcctgcg tgccgaggac acggccgtgt attactgtgc gaaatacgac 300 ggtatctacg gtgaactgga cttctggggc caaggaaccc tggtcaccgt ttcctca 357

<212> Type : DNA

<211> Length : 357

SequenceName : 3

SequenceDescription :

Sequence

<213> OrganisrnName : artificial sequence

<400> PreSequenceString :

GYSGYDYAEY FQHWGQGTLV TVSS 24

<212> Type : PRT

<211> Length : 24

SequenceName : 4

SequenceDescription :

Sequence

<213> OrganisrnName : artificial sequence

<400> PreSequenceString :

GYSYGYAEYF QHWGQGTLVT VSS 23

<212> Type : PRT

<211> Length : 23

SequenceName : 5

SequenceDescription :

Sequence

<213> OrganisrnName : artificial sequence

<400> PreSequenceString :

TTVTAEYFQH WGQGTLVTVS S 21 <212> Type : PRT

<211> Length : 21 SequenceName : 6

SequenceDescription :

Sequence

<213> OrganismName : artificial sequence

<400> PreSequenceString :

YSGSYYAEYF QHWGQGTLVT VSS 23

<212> Type : PRT

<211> Length : 23

SequenceName : 7

SequenceDescription :

Sequence

<213> OrganismName : artificial sequence

<400> PreSequenceString :

DYGDYAEYFQ HWGQGTLVTV SS 22

<212> Type : PRT

<211> Length : 22

SequenceName : 8

SequenceDescription :

Sequence

<213> OrganismName : artificial sequence

<400> PreSequenceString :

DYSNYAEYFQ HWGQGTLVTV SS 22

<212> Type : PRT

<211> Length : 22

SequenceName : 9

SequenceDescription :

Sequence

<213> OrganismName : artificial sequence

<400> PreSequenceString :

SIAARAEYFQ HWGQGTLVTV SS 22

<212> Type : PRT

<211> Length : 22

SequenceName : 10

SequenceDescription :

Sequence

<213> OrganismName : artificial sequence

<400> PreSequenceString :

VQLERAEYFQ HWGQGTLVTV SS 22

<212> Type : PRT

<211> Length : 22

SequenceName : 11

SequenceDescription :

Sequence

<213> OrganismName : artificial sequence

<400> PreSequenceString :

YYDILTGYYN AEYFQHWGQG TLVTVSS 27

<212> Type : PRT

<211> Length : 27

SequenceName : 12

SequenceDescription : Sequence

<213> OrganismName : artificial sequence

<400> PreSequenceString :

YYYDSSGYYY AEYFQHWGQG TLVTVSS 27

<212> Type : PRT

<211> Length : 27

SequenceName : 13

SequenceDescription :

Sequence

<213> OrganismName : artificial sequence

<400> PreSequenceString :

YYYGSGSYYN AEYFQHWGQG TLVTVSS 27

<212> Type : PRT

<211> Length : 27

SequenceName : 14

SequenceDescription :

Sequence

<213> OrganismName : artificial sequence

<400> PreSequenceString :

EYSSSSAEYF QHWGQGTLVT VSS 23

<212> Type : PRT

<211> Length : 23

SequenceName : 15

SequenceDescription :

Sequence

<213> OrganismName : artificial sequence

<400> PreSequenceString :

GITGTAEYFQ HWGQGTLVTV SS 22

<212> Type : PRT

<211> Length : 22

SequenceName : 16

SequenceDescription :

Sequence

<213> OrganismName : artificial sequence

<400> PreSequenceString :

GIVGATAEYF QHWGQGTLVT VSS 23

<212> Type : PRT

<211> Length : 23

SequenceName : 17

SequenceDescription :

Sequence

<213> OrganismName : artificial sequence

<400> PreSequenceString :

GTTGTAEYFQ HWGQGTLVTV SS 22

<212> Type : PRT

<211> Length : 22

SequenceName : 18

SequenceDescription :

Sequence

<213> OrganismName : artificial sequence <400> PreSequenceString :

GYSSGYAEYF QHWGQGTLVT VSS 23

<212> Type : PRT

<211> Length : 23

SequenceName : 19

SequenceDescription :

Sequence

<213> OrganisrnName : artificial sequence

<400> PreSequenceString :

LTGAEYFQHW GQGTLVTVSS 20

<212> Type : PRT

<211> Length : 20

SequenceName : 20

SequenceDescription :

Sequence

<213> OrganisrnName : artificial sequence

<400> PreSequenceString :

VDIVATIAEY FQHWGQGTLV TVSS 24

<212> Type : PRT

<211> Length : 24

SequenceName : 21

SequenceDescription :

Sequence

<213> OrganisrnName : artificial sequence

<400> PreSequenceString :

GYSGYDYDAF DVWGQGTMVT VSS 23

<212> Type : PRT

<211> Length : 23

SequenceName : 22

SequenceDescription :

Sequence

<213> OrganisrnName : artificial sequence

<400> PreSequenceString :

GYSYGYDAFD VWGQGTMVTV SS 22

<212> Type : PRT

<211> Length : 22

SequenceName : 23

SequenceDescription :

Sequence

<213> OrganisrnName : artificial sequence

<400> PreSequenceString :

TTVTDAFDVW GQGTMVTVSS 20

<212> Type : PRT

<211> Length : 20

SequenceName : 24

SequenceDescription :

Sequence

<213> OrganisrnName : artificial sequence

<400> PreSequenceString :

YSGSYYDAFD VWGQGTMVTV SS 22 <212> Type : PRT <211> Length : 22

SequenceName : 25

SequenceDescription :

Sequence

<213> OrganisrnName : artificial sequence

<400> PreSequenceString :

DYGDYDAFDV WGQGTMVTVS S 21

<212> Type : PRT

<211> Length : 21

SequenceName : 26

SequenceDescription :

Sequence

<213> OrganisrnName : artificial sequence

<400> PreSequenceString :

DYSNYDAFDV WGQGTMVTVS S 21

<212> Type : PRT

<211> Length : 21

SequenceName : 27

SequenceDescription :

Sequence

<213> OrganisrnName : artificial sequence

<400> PreSequenceString :

SIAARDAFDV WGQGTMVTVS S 21

<212> Type : PRT

<211> Length : 21

SequenceName : 28

SequenceDescription :

Sequence

<213> OrganisrnName : artificial sequence

<400> PreSequenceString :

VQLERDAFDV WGQGTMVTVS S 21

<212> Type : PRT

<211> Length : 21

SequenceName : 29

SequenceDescription :

Sequence

<213> OrganisrnName : artificial sequence

<400> PreSequenceString :

YYDILTGYYN DAFDVWGQGT MVTVSS 26

<212> Type : PRT

<211> Length : 26

SequenceName : 30

SequenceDescription :

Sequence

<213> OrganisrnName : artificial sequence

<400> PreSequenceString :

YYYDSSGYYY DAFDVWGQGT MVTVSS 26

<212> Type : PRT

<211> Length : 26

SequenceName : 31

SequenceDescription : Sequence

<213> OrganismName : artificial sequence

<400> PreSequenceString :

YYYGSGSYYN DAFDVWGQGT MVTVSS 26

<212> Type : PRT

<211> Length : 26

SequenceName : 32

SequenceDescription :

Sequence

<213> OrganismName : artificial sequence

<400> PreSequenceString :

EYSSSSDAFD VWGQGTMVTV SS 22

<212> Type : PRT

<211> Length : 22

SequenceName : 33

SequenceDescription :

Sequence

<213> OrganismName : artificial sequence

<400> PreSequenceString :

GITGTDAFDV WGQGTMVTVS S 21

<212> Type : PRT

<211> Length : 21

SequenceName : 34

SequenceDescription :

Sequence

<213> OrganismName : artificial sequence

<400> PreSequenceString :

GIVGATDAFD VWGQGTMVTV SS 22

<212> Type : PRT

<211> Length : 22

SequenceName : 35

SequenceDescription :

Sequence

<213> OrganismName : artificial sequence

<400> PreSequenceString :

GTTGTDAFDV WGQGTMVTVS S 21

<212> Type : PRT

<211> Length : 21

SequenceName : 36

SequenceDescription :

Sequence

<213> OrganismName : artificial sequence

<400> PreSequenceString :

GYSSGYDAFD VWGQGTMVTV SS 22

<212> Type : PRT

<211> Length : 22

SequenceName : 37

SequenceDescription :

Sequence <213> OrganismName : artificial sequence

<400> PreSequenceString :

LTGDAFDVWG QGTMVTVSS 19

<212> Type : PRT

<211> Length : 19

SequenceName : 38

SequenceDescription :

Sequence

<213> OrganismName : artificial sequence

<400> PreSequenceString :

VDIVATIDAF DVWGQGTMVT VSS 23

<212> Type : PRT

<211> Length : 23

SequenceName : 39

SequenceDescription :

Sequence

<213> OrganismName : artificial sequence

<400> PreSequenceString :

GYSGYDYYFD YWGQGTLVTV SS 22

<212> Type : PRT

<211> Length : 22

SequenceName : 40

SequenceDescription :

Sequence

<213> OrganismName : artificial sequence

<400> PreSequenceString :

GYSYGYYFDY WGQGTLVTVS S 21

<212> Type : PRT

<211> Length : 21

SequenceName : 41

SequenceDescription :

Sequence

<213> OrganismName : artificial sequence

<400> PreSequenceString :

TTVTYFDYWG QGTLVTVSS 19

<212> Type : PRT

<211> Length : 19

SequenceName : 42

SequenceDescription :

Sequence

<213> OrganismName : artificial sequence

<400> PreSequenceString :

YSGSYYYFDY WGQGTLVTVS S 21

<212> Type : PRT

<211> Length : 21

SequenceName : 43

SequenceDescription :

Sequence

<213> OrganismName : artificial sequence

<400> PreSequenceString :

DYGDYYFDYW GQGTLVTVSS 20 <212> Type : PRT

<211> Length : 20

SequenceName : 44

SequenceDescription :

Sequence

<213> OrganisrnName : artificial sequence

<400> PreSequenceString :

DYSNYYFDYW GQGTLVTVSS 20

<212> Type : PRT

<211> Length : 20

SequenceName : 45

SequenceDescription :

Sequence

<213> OrganisrnName : artificial sequence

<400> PreSequenceString :

SIAARYFDYW GQGTLVTVSS 20

<212> Type : PRT

<211> Length : 20

SequenceName : 46

SequenceDescription :

Sequence

<213> OrganisrnName : artificial sequence

<400> PreSequenceString :

VQLERYFDYW GQGTLVTVSS 20

<212> Type : PRT

<211> Length : 20

SequenceName : 47

SequenceDescription :

Sequence

<213> OrganisrnName : artificial sequence

<400> PreSequenceString :

YYDILTGYYN YFDYWGQGTL VTVSS 25

<212> Type : PRT

<211> Length : 25

SequenceName : 48

SequenceDescription :

Sequence

<213> OrganisrnName : artificial sequence

<400> PreSequenceString :

YYYDSSGYYY YFDYWGQGTL VTVSS 25

<212> Type : PRT

<211> Length : 25

SequenceName : 49

SequenceDescription :

Sequence

<213> OrganisrnName : artificial sequence

<400> PreSequenceString :

YYYGSGSYYN YFDYWGQGTL VTVSS 25

<212> Type : PRT

<211> Length : 25

SequenceName : 50 SequenceDescription :

Sequence

<213> OrganismName : artificial sequence

<400> PreSequenceString :

EYSSSSYFDY WGQGTLVTVS S 21

<212> Type : PRT

<211> Length : 21

SequenceName : 51

SequenceDescription :

Sequence

<213> OrganismName : artificial sequence

<400> PreSequenceString :

GITGTYFDYW GQGTLVTVSS 20

<212> Type : PRT

<211> Length : 20

SequenceName : 52

SequenceDescription :

Sequence

<213> OrganismName : artificial sequence

<400> PreSequenceString :

GIVGATYFDY WGQGTLVTVS S 21

<212> Type : PRT

<211> Length : 21

SequenceName : 53

SequenceDescription :

Sequence

<213> OrganismName : artificial sequence

<400> PreSequenceString :

GTTGTYFDYW GQGTLVTVSS 20

<212> Type : PRT

<211> Length : 20

SequenceName : 54

SequenceDescription :

Sequence

<213> OrganismName : artificial sequence

<400> PreSequenceString :

GYSSGYYFDY WGQGTLVTVS S 21

<212> Type : PRT

<211> Length : 21

SequenceName : 55

SequenceDescription :

Sequence

<213> OrganismName : artificial sequence

<400> PreSequenceString :

LTGYFDYWGQ GTLVTVSS 18

<212> Type : PRT

<211> Length : 18

SequenceName : 56

SequenceDescription :

Sequence <213> OrganismName : artificial sequence

<400> PreSequenceString :

VDIVATIYFD YWGQGTLVTV SS 22

<212> Type : PRT

<211> Length : 22

SequenceName : 57

SequenceDescription :

Sequence

<213> OrganismName : artificial sequence

<400> PreSequenceString :

GYSGYDYNWF DSWGQGTLVT VSS 23

<212> Type : PRT

<211> Length : 23

SequenceName : 58

SequenceDescription :

Sequence

<213> OrganismName : artificial sequence

<400> PreSequenceString :

GYSYGYNWFD SWGQGTLVTV SS 22

<212> Type : PRT

<211> Length : 22

SequenceName : 59

SequenceDescription :

Sequence

<213> OrganismName : artificial sequence

<400> PreSequenceString :

TTVTNWFDSW GQGTLVTVSS 20

<212> Type : PRT

<211> Length : 20

SequenceName : 60

SequenceDescription :

Sequence

<213> OrganismName : artificial sequence

<400> PreSequenceString :

YSGSYYNWFD SWGQGTLVTV SS 22

<212> Type : PRT

<211> Length : 22

SequenceName : 61

SequenceDescription :

Sequence

<213> OrganismName : artificial sequence

<400> PreSequenceString :

DYGDYNWFDS WGQGTLVTVS S 21

<212> Type : PRT

<211> Length : 21

SequenceName : 62

SequenceDescription :

Sequence

<213> OrganismName : artificial sequence

<400> PreSequenceString : DYSNYNWFDS WGQGTLVTVS S 21

<212> Type : PRT

<211> Length : 21

SequenceName : 63

SequenceDescription :

Sequence

<213> OrganisrnName : artificial sequence

<400> PreSequenceString :

SIAARNWFDS WGQGTLVTVS S 21

<212> Type : PRT

<211> Length : 21

SequenceName : 64

SequenceDescription :

Sequence

<213> OrganisrnName : artificial sequence

<400> PreSequenceString :

VQLERNWFDS WGQGTLVTVS S 21

<212> Type : PRT

<211> Length : 21

SequenceName : 65

SequenceDescription :

Sequence

<213> OrganisrnName : artificial sequence

<400> PreSequenceString :

YYDILTGYYN NWFDSWGQGT LVTVSS 26

<212> Type : PRT

<211> Length : 26

SequenceName : 66

SequenceDescription :

Sequence

<213> OrganisrnName : artificial sequence

<400> PreSequenceString :

YYYDSSGYYY NWFDSWGQGT LVTVSS 26

<212> Type : PRT

<211> Length : 26

SequenceName : 67

SequenceDescription :

Sequence

<213> OrganisrnName : artificial sequence

<400> PreSequenceString :

YYYGSGSYYN NWFDSWGQGT LVTVSS 26

<212> Type : PRT

<211> Length : 26

SequenceName : 68

SequenceDescription :

Sequence

<213> OrganisrnName : artificial sequence

<400> PreSequenceString :

EYSSSSNWFD SWGQGTLVTV SS 22 <212> Type : PRT

<211> Length : 22 SequenceName : 69

SequenceDescription :

Sequence

<213> OrganismName : artificial sequence

<400> PreSequenceString :

GITGTNWFDS WGQGTLVTVS S 21

<212> Type : PRT

<211> Length : 21

SequenceName : 70

SequenceDescription :

Sequence

<213> OrganismName : artificial sequence

<400> PreSequenceString :

GIVGATNWFD SWGQGTLVTV SS 22

<212> Type : PRT

<211> Length : 22

SequenceName : 71

SequenceDescription :

Sequence

<213> OrganismName : artificial sequence

<400> PreSequenceString :

GTTGTNWFDS WGQGTLVTVS S 21

<212> Type : PRT

<211> Length : 21

SequenceName : 72

SequenceDescription :

Sequence

<213> OrganismName : artificial sequence

<400> PreSequenceString :

GYSSGYNWFD SWGQGTLVTV SS 22

<212> Type : PRT

<211> Length : 22

SequenceName : 73

SequenceDescription :

Sequence

<213> OrganismName : artificial sequence

<400> PreSequenceString :

LTGNWFDSWG QGTLVTVSS 19

<212> Type : PRT

<211> Length : 19

SequenceName : 74

SequenceDescription :

Sequence

<213> OrganismName : artificial sequence

<400> PreSequenceString :

VDIVATINWF DSWGQGTLVT VSS 23

<212> Type : PRT

<211> Length : 23

SequenceName : 75

SequenceDescription : Sequence

<213> OrganismName : artificial sequence

<400> PreSequenceString :

GYSGYDYYGM DVWGQGTTVT VSS 23

<212> Type : PRT

<211> Length : 23

SequenceName : 76

SequenceDescription :

Sequence

<213> OrganismName : artificial sequence

<400> PreSequenceString :

GYSYGYYGMD VWGQGTTVTV SS 22

<212> Type : PRT

<211> Length : 22

SequenceName : 77

SequenceDescription :

Sequence

<213> OrganismName : artificial sequence

<400> PreSequenceString :

TTVTYGMDVW GQGTTVTVSS 20

<212> Type : PRT

<211> Length : 20

SequenceName : 78

SequenceDescription :

Sequence

<213> OrganismName : artificial sequence

<400> PreSequenceString :

YSGSYYYGMD VWGQGTTVTV SS 22

<212> Type : PRT

<211> Length : 22

SequenceName : 79

SequenceDescription :

Sequence

<213> OrganismName : artificial sequence

<400> PreSequenceString :

DYGDYYGMDV WGQGTTVTVS S 21

<212> Type : PRT

<211> Length : 21

SequenceName : 80

SequenceDescription :

Sequence

<213> OrganismName : artificial sequence

<400> PreSequenceString :

DYSNYYGMDV WGQGTTVTVS S 21

<212> Type : PRT

<211> Length : 21

SequenceName : 81

SequenceDescription :

Sequence

<213> OrganismName : artificial sequence <400> PreSequenceString :

SIAARYGMDV WGQGTTVTVS S 21

<212> Type : PRT

<211> Length : 21

SequenceName : 82

SequenceDescription :

Sequence

<213> OrganisrnName : artificial sequence

<400> PreSequenceString :

VQLERYGMDV WGQGTTVTVS S 21

<212> Type : PRT

<211> Length : 21

SequenceName : 83

SequenceDescription :

Sequence

<213> OrganisrnName : artificial sequence

<400> PreSequenceString :

YYDILTGYYN YGMDVWGQGT TVTVSS 26

<212> Type : PRT

<211> Length : 26

SequenceName : 84

SequenceDescription :

Sequence

<213> OrganisrnName : artificial sequence

<400> PreSequenceString :

YYYDSSGYYY YGMDVWGQGT TVTVSS 26

<212> Type : PRT

<211> Length : 26

SequenceName : 85

SequenceDescription :

Sequence

<213> OrganisrnName : artificial sequence

<400> PreSequenceString :

YYYGSGSYYN YGMDVWGQGT TVTVSS 26

<212> Type : PRT

<211> Length : 26

SequenceName : 86

SequenceDescription :

Sequence

<213> OrganisrnName : artificial sequence

<400> PreSequenceString :

EYSSSSYGMD VWGQGTTVTV SS 22

<212> Type : PRT

<211> Length : 22

SequenceName : 87

SequenceDescription :

Sequence

<213> OrganisrnName : artificial sequence

<400> PreSequenceString :

GITGTYGMDV WGQGTTVTVS S 21 <212> Type : PRT <211> Length : 21

SequenceName : 88

SequenceDescription :

Sequence

<213> OrganisrnName : artificial sequence

<400> PreSequenceString :

GIVGATYGMD VWGQGTTVTV SS 22

<212> Type : PRT

<211> Length : 22

SequenceName : 89

SequenceDescription :

Sequence

<213> OrganisrnName : artificial sequence

<400> PreSequenceString :

GTTGTYGMDV WGQGTTVTVS S 21

<212> Type : PRT

<211> Length : 21

SequenceName : 90

SequenceDescription :

Sequence

<213> OrganisrnName : artificial sequence

<400> PreSequenceString :

GYSSGYYGMD VWGQGTTVTV SS 22

<212> Type : PRT

<211> Length : 22

SequenceName : 91

SequenceDescription :

Sequence

<213> OrganisrnName : artificial sequence

<400> PreSequenceString :

LTGYGMDVWG QGTTVTVSS 19

<212> Type : PRT

<211> Length : 19

SequenceName : 92

SequenceDescription :

Sequence

<213> OrganisrnName : artificial sequence

<400> PreSequenceString :

VDIVATIYGM DVWGQGTTVT VSS 23

<212> Type : PRT

<211> Length : 23

SequenceName : 93

SequenceDescription :

Sequence

<213> OrganisrnName : artificial sequence

<400> PreSequenceString :

gacacggcyg tgtattactg tgc 23

<212> Type : DNA

<211> Length : 23

SequenceName : 94

SequenceDescription : Sequence

<213> OrganismName : artificial sequence

<400> PreSequenceString :

tgttggcctc ccgggcctga agagacggtg accattgtcc 40

<212> Type : DNA

<211> Length : 40

SequenceName : 95

SequenceDescription :

Sequence

<213> OrganismName : artificial sequence

<400> PreSequenceString :

tgttggcctc ccgggcctga ggagacggtg accgtggtc 39

<212> Type : DNA

<211> Length : 39

SequenceName : 96

SequenceDescription :

Sequence

<213> OrganismName : artificial sequence

<400> PreSequenceString :

tgttggcctc ccgggcctga ggagacrgtg accagggt 38

<212> Type : DNA

<211> Length : 38

SequenceName : 97

SequenceDescription :

Sequence

<213> OrganismName : artificial sequence

<400> PreSequenceString :

cgctggattg ttattactcg eg 22

<212> Type : DNA

<211> Length : 22

SequenceName : 98

SequenceDescription :

Sequence

<213> OrganismName : artificial sequence

<400> PreSequenceString :

cacaggtctc gtgtcctcgg cacgcaggct gttcatttg 39

<212> Type : DNA

<211> Length : 39

SequenceName : 99

SequenceDescription :

Sequence

<213> OrganismName : artificial sequence

<400> PreSequenceString :

gtgtggtctc ggacacggcy gtgtattact gtgc 34

<212> Type : DNA

<211> Length : 34

SequenceName : 100

SequenceDescription :

Sequence <213> OrganismName : artificial sequence

<400> PreSequenceString :

tcgcggccca gccggccatg 20 <212> Type : DNA

<211> Length : 20

SequenceName : 101

SequenceDescription :

Sequence

<213> OrganismName : artificial sequence

<400> PreSequenceString :

gtgttggcct cccgggcctg 20

<212> Type : DNA

<211> Length : 20

SequenceName : 102

SequenceDescription :

Sequence

<213> OrganismName : Homo sapiens

<400> PreSequenceString :

caggtgcagc tggtgcagtc tggggctgag gtgaagaagc ctggggcctc agtgaaggtc 60 tcctgcaagg cttctggata caccttcacc ggctactata tgcactgggt gcgacaggcc 120 cctggacaag ggcttgagtg gatgggacgg atcaacccta acagtggtgg cacaaactat 180 gcacagaagt ttcagggcag ggtcaccagt accagggaca cgtccatcag cacagcctac 240 atggagctga gcaggctgag atctgacgac acggtcgtgt attactgtgc gagaga 296

<212> Type : DNA

<211> Length : 296

SequenceName : 103

SequenceDescription :

Sequence

<213> OrganismName : Homo sapiens

<400> PreSequenceString :

caggttcagc tggtgcagtc tggagctgag gtgaagaagc ctggggcctc agtgaaggtc 60 tcctgcaagg cttctggtta cacctttacc agctatggta tcagctgggt gcgacaggcc 120 cctggacaag ggcttgagtg gatgggatgg atcagcgctt acaatggtaa cacaaactat 180 gcacagaagc tccagggcag agtcaccatg accacagaca catccacgag cacagcctac 240 atggagctga ggagcctgag atctgacgac acggccgtgt attactgtgc gagaga 296

<212> Type : DNA

<211> Length : 296

SequenceName : 104

SequenceDescription :

Sequence

<213> OrganismName : Homo sapiens

<400> PreSequenceString : caggtgcagc tggtgcagtc tggggctgag gtgaagaagc ctggggcctc agtgaaggtt 60 tcctgcaagg catctggata caccttcacc agctactata tgcactgggt gcgacaggcc 120 cctggacaag ggcttgagtg gatgggaata atcaacccta gtggtggtag cacaagctac 180 gcacagaagt tccagggcag agtcaccatg accagggaca cgtccacgag cacagtctac 240 atggagctga gcagcctgag atctgaggac acggccgtgt attactgtgc gagaga 296

<212> Type : DNA

<211> Length : 296

SequenceName : 105

SequenceDescription :

Sequence

<213> OrganisrnName : Homo sapiens

<400> PreSequenceString :

caggtgcagc tggtgcagtc tggggctgag gtgaagaagc ctgggtcctc ggtgaaggtc 60 tcctgcaagg cttctggagg caccttcagc agctatgcta tcagctgggt gcgacaggcc 120 cctggacaag ggcttgagtg gatgggaggg atcatcccta tctttggtac agcaaactac 180 gcacagaagt tccagggcag agtcacgatt accgcggacg aatccacgag cacagcctac 240 atggagctga gcagcctgag atctgaggac acggccgtgt attactgtgc gagaga 296

<212> Type : DNA

<211> Length : 296

SequenceName : 106

SequenceDescription :

Sequence

<213> OrganisrnName : Homo sapiens

<400> PreSequenceString :

gaggtgcagc tgttggagtc tgggggaggc ttggtacagc ctggggggtc cctgagactc 60 tcctgtgcag cctctggatt cacctttagc agctatgcca tgagctgggt ccgccaggct 120 ccagggaagg ggctggagtg ggtctcagct attagtggta gtggtggtag cacatactac 180 gcagactccg tgaagggccg gttcaccatc tccagagaca attccaagaa cacgctgtat 240 ctgcaaatga acagcctgag agccgaggac acggccgtat attactgtgc gaaaga 296

<212> Type : DNA

<211> Length : 296

SequenceName : 107

SequenceDescription :

Sequence

<213> OrganisrnName : Homo sapiens

<400> PreSequenceString :

caggtgcagc tggtggagtc tgggggaggc gtggtccagc ctgggaggtc cctgagactc 60 tcctgtgcag cctctggatt caccttcagt agctatgcta tgcactgggt ccgccaggct 120 ccaggcaagg ggctagagtg ggtggcagtt atatcatatg atggaagtaa taaatactac 180 gcagactccg tgaagggccg attcaccatc tccagagaca attccaagaa cacgctgtat 240 ctgcaaatga acagcctgag agctgaggac acggctgtgt attactgtgc gagaga 296

<212> Type : DNA

<211> Length : 296

SequenceName : 108

SequenceDescription :

Sequence

<213> OrganisrnName : Homo sapiens

<400> PreSequenceString :

gaggtgcagc tggtggagtc tgggggaggc ttggtacagc ctggggggtc cctgagactc 60 tcctgtgcag cctctggatt caccttcagt agctatagca tgaactgggt ccgccaggct 120 ccagggaagg ggctggagtg ggtttcatac attagtagta gtagtagtac catatactac 180 gcagactctg tgaagggccg attcaccatc tccagagaca atgccaagaa ctcactgtat 240 ctgcaaatga acagcctgag agccgaggac acggctgtgt attactgtgc gagaga 296

<212> Type : DNA

<211> Length : 296

SequenceName : 109

SequenceDescription :

Sequence

<213> OrganisrnName : Homo sapiens

<400> PreSequenceString :

caggtgcagc tacagcagtg gggcgcagga ctgttgaagc cttcggagac cctgtccctc 60 acctgcgctg tctatggtgg gtccttcagt ggttactact ggagctggat ccgccagccc 120 ccagggaagg ggctggagtg gattggggaa atcaatcata gtggaagcac caactacaac 180 ccgtccctca agagtcgagt caccatatca gtagacacgt ccaagaacca gttctccctg 240 aagctgagct ctgtgaccgc cgcggacacg gctgtgtatt actgtgcgag agg 293

<212> Type : DNA

<211> Length : 293

SequenceName : 110

SequenceDescription :

Sequence

<213> OrganisrnName : Homo sapiens

<400> PreSequenceString :

caggtgcagc tgcaggagtc gggcccagga ctggtgaagc cttcggagac cctgtccctc 60 acctgcactg tctctggtgg ctccatcagt agttactact ggagctggat ccggcagccc 120 ccagggaagg gactggagtg gattgggtat atctattaca gtgggagcac caactacaac 180 ccctccctca agagtcgagt caccatatca gtagacacgt ccaagaacca gttctccctg 240 aagctgagct ctgtgaccgc tgcggacacg gccgtgtatt actgtgcgag aga 293

<212> Type : DNA

<211> Length : 293

SequenceName : 111

SequenceDescription :

Sequence

<213> OrganisrnName : Homo sapiens

<400> PreSequenceString :

gaggtgcagc tggtgcagtc tggagcagag gtgaaaaagc ccggggagtc tctgaagatc 60 tcctgtaagg gttctggata cagctttacc agctactgga tcggctgggt gcgccagatg 120 cccgggaaag gcctggagtg gatggggatc atctatcctg gtgactctga taccagatac 180 agcccgtcct tccaaggcca ggtcaccatc tcagccgaca agtccatcag caccgcctac 240 ctgcagtgga gcagcctgaa ggcctcggac accgccatgt attactgtgc gagaca 296

<212> Type : DNA

<211> Length : 296

SequenceName : 112

SequenceDescription :

Sequence

<213> OrganisrnName : Homo sapiens

<400> PreSequenceString :

caggtacagc tgcagcagtc aggtccagga ctggtgaagc cctcgcagac cctctcactc 60 acctgtgcca tctccgggga cagtgtctct agcaacagtg ctgcttggaa ctggatcagg 120 cagtccccat cgagaggcct tgagtggctg ggaaggacat actacaggtc caagtggtat 180 aatgattatg cagtatctgt gaaaagtcga ataaccatca acccagacac atccaagaac 240 cagttctccc tgcagctgaa ctctgtgact cccgaggaca cggctgtgta ttactgtgca 300 agaga 305

<212> Type : DNA

<211> Length : 305

SequenceName : 113

SequenceDescription :

Sequence

<213> OrganisrnName : Homo sapiens

<400> PreSequenceString :

gacatccaga tgacccagtc tccttccacc ctgtctgcat ctgtaggaga cagagtcacc 60 atcacttgcc gggccagtca gagtattagt agctggttgg cctggtatca gcagaaacca 120 gggaaagccc ctaagctcct gatctatgat gcctccagtt tggaaagtgg ggtcccatca 180 aggttcagcg gcagtggatc tgggacagaa ttcactctca ccatcagcag cctgcagcct 240 gatgattttg caacttatta ctgccaacag tataatagtt attctcc 287

<212> Type : DNA

<211> Length : 287

SequenceName : 114

SequenceDescription :

Sequence

<213> OrganisrnName : Homo sapiens

<400> PreSequenceString :

gacatccaga tgacccagtc tccatcttcc gtgtctgcat ctgtaggaga cagagtcacc 60 atcacttgtc gggcgagtca gggtattagc agctggttag cctggtatca gcagaaacca 120 gggaaagccc ctaagctcct gatctatgct gcatccagtt tgcaaagtgg ggtcccatca 180 aggttcagcg gcagtggatc tgggacagat ttcactctca ccatcagcag cctgcagcct 240 gaagattttg caacttacta ttgtcaacag gctaacagtt tccctcc 287

<212> Type : DNA

<211> Length : 287

SequenceName : 115

SequenceDescription :

Sequence

<213> OrganisrnName : Homo sapiens

<400> PreSequenceString :

gacatccaga tgacccagtc tccatcctcc ctgtctgcat ctgtaggaga cagagtcacc 60 atcacttgcc aggcgagtca ggacattagc aactatttaa attggtatca gcagaaacca 120 gggaaagccc ctaagctcct gatctacgat gcatccaatt tggaaacagg ggtcccatca 180 aggttcagtg gaagtggatc tgggacagat tttactttca ccatcagcag cctgcagcct 240 gaagatattg caacatatta ctgtcaacag tatgataatc tccctcc 287

<212> Type : DNA

<211> Length : 287

SequenceName : 116

SequenceDescription :

Sequence

<213> OrganisrnName : Homo sapiens

<400> PreSequenceString :

gatattgtga tgactcagtc tccactctcc ctgcccgtca cccctggaga gccggcctcc 60 atctcctgca ggtctagtca gagcctcctg catagtaatg gatacaacta tttggattgg 120 tacctgcaga agccagggca gtctccacag ctcctgatct atttgggttc taatcgggcc 180 tccggggtcc ctgacaggtt cagtggcagt ggatcaggca cagattttac actgaaaatc 240 agcagagtgg aggctgagga tgttggggtt tattactgca tgcaagctct acaaactcct 300 cc 302

<212> Type : DNA

<211> Length : 302

SequenceName : 117

SequenceDescription : Sequence

<213> OrganismName : Homo sapiens

<400> PreSequenceString :

gacatccaga tgacccagtc tccatcctcc ctgtctgcat ctgtaggaga cagagtcacc 60 atcacttgcc gggcaagtca gagcattagc agctatttaa attggtatca gcagaaacca 120 gggaaagccc ctaagctcct gatctatgct gcatccagtt tgcaaagtgg ggtcccatca 180 aggttcagtg gcagtggatc tgggacagat ttcactctca ccatcagcag tctgcaacct 240 gaagattttg caacttacta ctgtcaacag agttacagta cccctcc 287

<212> Type : DNA

<211> Length : 287

SequenceName : 118

SequenceDescription :

Sequence

<213> OrganismName : Homo sapiens

<400> PreSequenceString :

gaaattgtgt tgacacagtc tccagccacc ctgtctttgt ctccagggga aagagccacc 60 ctctcctgca gggccagtca gagtgttagc agctacttag cctggtacca acagaaacct 120 ggccaggctc ccaggctcct catctatgat gcatccaaca gggccactgg catcccagcc 180 aggttcagtg gcagtgggtc tgggacagac ttcactctca ccatcagcag cctagagcct 240 gaagattttg cagtttatta ctgtcagcag cgtagcaact ggcctcc 287

<212> Type : DNA

<211> Length : 287

SequenceName : 119

SequenceDescription :

Sequence

<213> OrganismName : Homo sapiens

<400> PreSequenceString :

gaaatagtga tgacgcagtc tccagccacc ctgtctgtgt ctccagggga aagagccacc 60 ctctcctgca gggccagtca gagtgttagc agcaacttag cctggtacca gcagaaacct 120 ggccaggctc ccaggctcct catctatggt gcatccacca gggccactgg tatcccagcc 180 aggttcagtg gcagtgggtc tgggacagag ttcactctca ccatcagcag cctgcagtct 240 gaagattttg cagtttatta ctgtcagcag tataataact ggcctcc 287

<212> Type : DNA

<211> Length : 287

SequenceName : 120

SequenceDescription :

Sequence

<213> OrganismName : Homo sapiens

<400> PreSequenceString :

gaaattgtgt tgacgcagtc tccaggcacc ctgtctttgt ctccagggga aagagccacc 60 ctctcctgca gggccagtca gagtgttagc agcagctact tagcctggta ccagcagaaa 120 cctggccagg ctcccaggct cctcatctat ggtgcatcca gcagggccac tggcatccca 180 gacaggttca gtggcagtgg gtctgggaca gacttcactc tcaccatcag cagactggag 240 cctgaagatt ttgcagtgta ttactgtcag cagtatggta gctcacctcc 290

<212> Type : DNA

<211> Length : 290

SequenceName : 121

SequenceDescription :

Sequence

<213> OrganisrnName : Homo sapiens

<400> PreSequenceString :

gacatcgtga tgacccagtc tccagactcc ctggctgtgt ctctgggcga gagggccacc 60 atcaactgca agtccagcca gagtgtttta tacagctcca acaataagaa ctacttagct 120 tggtaccagc agaaaccagg acagcctcct aagctgctca tttactgggc atctacccgg 180 gaatccgggg tccctgaccg attcagtggc agcgggtctg ggacagattt cactctcacc 240 atcagcagcc tgcaggctga agatgtggca gtttattact gtcagcaata ttatagtact 300 cctcc 305

<212> Type : DNA

<211> Length : 305

SequenceName : 122

SequenceDescription :

Sequence

<213> OrganisrnName : Homo sapiens

<400> PreSequenceString :

tcttctgagc tgactcagga ccctgctgtg tctgtggcct tgggacagac agtcaggatc 60 acatgccaag gagacagcct cagaagctat tatgcaagct ggtaccagca gaagccagga 120 caggcccctg tacttgtcat ctatggtaaa aacaaccggc cctcagggat cccagaccga 180 ttctctggct ccagctcagg aaacacagct tccttgacca tcactggggc tcaggcggaa 240 gatgaggctg actattactg taactcccgg gacagcagtg gtaaccatct 290

<212> Type : DNA

<211> Length : 290

SequenceName : 123

SequenceDescription :

Sequence

<213> OrganisrnName : Homo sapiens

<400> PreSequenceString :

aattttatgc tgactcagcc ccactctgtg tcggagtctc cggggaagac ggtaaccatc 60 tcctgcaccc gcagcagtgg cagcattgcc agcaactatg tgcagtggta ccagcagcgc 120 ccgggcagtt cccccaccac tgtgatctat gaggataacc aaagaccctc tggggtccct 180 gatcggttct ctggctccat cgacagctcc tccaactctg cctccctcac catctctgga 240 ctgaagactg aggacgaggc tgactactac tgtcagtctt atgatagcag caatca 296

<212> Type : DNA

<211> Length : 296

SequenceName : 124

SequenceDescription : Sequence

<213> OrganismName : Homo sapiens

<400> PreSequenceString :

tcctatgtgc tgactcagcc accctcagtg tcagtggccc caggaaagac ggccaggatt 60 acctgtgggg gaaacaacat tggaagtaaa agtgtgcact ggtaccagca gaagccaggc 120 caggcccctg tgctggtcat ctattatgat agcgaccggc cctcagggat ccctgagcga 180 ttctctggct ccaactctgg gaacacggcc accctgacca tcagcagggt cgaagccggg 240 gatgaggccg actattactg tcaggtgtgg gacagtagta gtgatcatcc 290

<212> Type : DNA

<211> Length : 290

SequenceName : 125

SequenceDescription :

Sequence

<213> OrganismName : artificial sequence

<400> PreSequenceString :

tctttcccta cacgacgctc ttccgatctt ggcagaaatt gtgttgacgc ag 52

<212> Type : DNA

<211> Length : 52

SequenceName : 126

SequenceDescription :

Sequence

<213> OrganismName : artificial sequence

<400> PreSequenceString :

tctttcccta cacgacgctc ttccgatctg actccctggc tgtgtctct 49

<212> Type : DNA

<211> Length : 49

SequenceName : 127

SequenceDescription :

Sequence

<213> OrganismName : artificial sequence

<400> PreSequenceString :

gtgactggag ttcagacgtg tgctcttccg atcccttggt accttggccg aac 53

<212> Type : DNA

<211> Length : 53

SequenceName : 128

SequenceDescription :

Sequence

<213> OrganismName : artificial sequence

<400> PreSequenceString :

tctttcccta cacgacgctc ttccgatctg tgcagcctct ggattcacc 49

<212> Type : DNA

<211> Length : 49

SequenceName : 129

SequenceDescription : Sequence

<213> OrganismName : artificial sequence

<400> PreSequenceString :

gtgactggag ttcagacgtg tgctcttccg atcgtggtga tggtgttggc etc 53

<212> Type : DNA

<211> Length : 53

SequenceName : 130

SequenceDescription :