Picomolar affinity fibronectin domains engineered utilizing loop length diversity, recursive mutagenesis, and loop shuffling

The 10th type III domain of human fibronectin (Fn3) has been validated as an effective scaffold for molecular recognition. In the current work, it was desired to improve the robustness of selection of stable, high-affinity Fn3 domains. A yeast surface display library of Fn3 was created in which three solvent-exposed loops were diversified in terms of amino acid composition and loop length. The library was screened by fluorescence-activated cell sorting to isolate binders to lysozyme. An affinity maturation scheme was developed to rapidly and broadly diversify populations of clones by random mutagenesis as well as homologous recombination-driven shuffling of mutagenized loops. The novel library and affinity maturation scheme combined to yield stable, monomeric Fn3 domains with 3 pM affinity for lysozyme. A secondary affinity maturation identified a stable 1.1 pM binder, the highest affinity yet reported for an Fn3 domain. In addition to extension of the affinity limit for this scaffold, the results demonstrate the ability to achieve high-affinity binding while preserving stability and the monomeric state. This library design and affinity maturation scheme is highly efficient, utilizing an initial diversity of 2 × 10⁷ clones and screening only 1 × 10⁸ mutants (totaled over all affinity maturation libraries). Analysis of intermediate populations revealed that loop length diversity, loop shuffling, and recursive mutagenesis of diverse populations are all critical components.     

Simple method for production of randomized human tenth fibronectin domain III libraries for use in combinatorial screening procedures

Challenges such as the rapid development of detection reagents for emerging or engineered pathogens, the goal of identifying probes for every protein in the human proteome, and the development of therapeutic molecules require systems for development of epitope binding molecules that are faster and cheaper than conventional antibody development. To be practical and effective, antibody mimics must be small, stable molecules that contain exposed loops or surfaces that can be randomized and screened using selective combinatorial assays. The tenth human fibronectin type III domain (10Fn3) fits these requirements and has recently been developed as an antibody mimic for use in detection and therapeutic platforms. Previously described systems for working with 10Fn3 used PCR-based approaches to anneal multiple oligonucleotides to generate randomized 10Fn3 libraries. Here we describe a simplified approach for creating randomized 10Fn3 libraries and report the first use of a T7-based phage display system for screening these libraries.     

Efficient selection of DARPins with sub-nanomolar affinities using SRP phage display

There is an ever-increasing demand to select specific, high-affinity binding molecules against targets of biomedical interest. The success of such selections depends strongly on the design and functional diversity of the library of binding molecules employed, and on the performance of the selection strategy. We recently developed SRP phage display that employs the cotranslational signal recognition particle (SRP) pathway for the translocation of proteins to the periplasm. This system allows efficient filamentous phage display of highly stable and fast-folding proteins, such as designed ankyrin repeat proteins (DARPins) that are virtually refractory to conventional phage display employing the post-translational Sec pathway. DARPins comprise a novel class of binding molecules suitable to complement or even replace antibodies in many biotechnological or biomedical applications. So far, all DARPins have been selected by ribosome display. Here, we harnessed SRP phage display to generate a phage DARPin library containing more than 10¹⁰ individual members. We were able to select well behaved and highly specific DARPins against a broad range of target proteins having affinities as low as 100 pM directly from this library, without affinity maturation. We describe efficient selection on the Fc domain of human IgG, TNFα, ErbB1 (EGFR), ErbB2 (HER2) and ErbB4 (HER4) as examples. Thus, SRP phage display makes filamentous phage display accessible for DARPins, allowing, for example, selection under harsh conditions or on whole cells. We envision that the use of SRP phage display will be beneficial for other libraries of stable and fast-folding proteins.     

Exploring the capacity of minimalist protein interfaces: interface energetics and affinity maturation to picomolar KD of a single-domain antibody with a flat paratope

A major architectural class in engineered binding proteins ("antibody mimics") involves the presentation of recognition loops off a single-domain scaffold. This class of binding proteins, both natural and synthetic, has a strong tendency to bind a preformed cleft using a convex binding interface (paratope). To explore their capacity to produce high-affinity interfaces with diverse shape and topography, we examined the interface energetics and explored the affinity limit achievable with a flat paratope. We chose a minimalist paratope limited to two loops found in a natural camelid heavy-chain antibody (VHH) that binds to ribonuclease A. Ala scanning of the VHH revealed only three "hot spot" side chains and additional four residues important for supporting backbone-mediated interactions. The small number of critical residues suggested that this is not an optimized paratope. Using selection from synthetic combinatorial libraries, we enhanced its affinity by >100-fold, resulting in variants with Kd as low as 180 pM with no detectable loss of binding specificity. High-resolution crystal structures revealed that the mutations induced only subtle structural changes but extended the network of interactions. This resulted in an expanded hot spot region including four additional residues located at the periphery of the paratope with a concomitant loss of the so-called "O-ring" arrangement of energetically inert residues. These results suggest that this class of simple, single-domain scaffolds is capable of generating high-performance binding interfaces with diverse shape. More generally, they suggest that highly functional interfaces can be designed without closely mimicking natural interfaces.     

Antibody binding loop insertions as diversity elements

In the use of non-antibody proteins as affinity reagents, diversity has generally been derived from oligonucleotide-encoded random amino acids. Although specific binders of high-affinity have been selected from such libraries, random oligonucleotides often encode stop codons and amino acid combinations that affect protein folding. Recently it has been shown that specific antibody binding loops grafted into heterologous proteins can confer the specific antibody binding activity to the created chimeric protein. In this paper, we examine the use of such antibody binding loops as diversity elements. We first show that we are able to graft a lysozyme-binding antibody loop into green fluorescent protein (GFP), creating a fluorescent protein with lysozyme-binding activity. Subsequently we have developed a PCR method to harvest random binding loops from antibodies and insert them at predefined sites in any protein, using GFP as an example. The majority of such GFP chimeras remain fluorescent, indicating that binding loops do not disrupt folding. This method can be adapted to the creation of other nucleic acid libraries where diversity is flanked by regions of relative sequence conservation, and its availability sets the stage for the use of antibody loop libraries as diversity elements for selection experiments.     

The human combinatorial antibody library HuCAL GOLD combines diversification of all six CDRs according to the natural immune system with a novel display method for efficient selection of high-affinity antibodies

This article describes the generation of the Human Combinatorial Antibody Library HuCAL GOLD. HuCAL GOLD is a synthetic human Fab library based on the HuCAL concept with all six complementarity-determining regions (CDRs) diversified according to the sequence and length variability of naturally rearranged human antibodies. The human antibody repertoire was analyzed in-depth, and individual CDR libraries were designed and generated for each CDR and each antibody family. Trinucleotide mixtures were used to synthesize the CDR libraries in order to ensure a high quality within HuCAL GOLD, and a β-lactamase selection system was employed to eliminate frame-shifted clones after successive cloning of the CDR libraries. With these methods, a large, high-quality library with more than 10 billion functional Fab fragments was achieved. By using CysDisplay, the antibody fragments are displayed on the tip of the phage via a disulfide bridge between the phage coat protein pIII and the heavy chain of the antibody fragment. Efficient elution of specific phages is possible by adding reducing agents. HuCAL GOLD was challenged with a variety of different antigens and proved to be a reliable source of high-affinity human antibodies with best affinities in the picomolar range, thus functioning as an excellent source of antibodies for research, diagnostic, and therapeutic applications. Furthermore, the data presented in this article demonstrate that CysDisplay is a robust and broadly applicable display technology even for high-throughput applications.     

Direct selection of antibodies from complex libraries with the protein fragment complementation assay

The aim of the present study was to develop the protein fragment complementation assay (PCA) for the intracellular selection of specific binding molecules from the fully synthetic HuCAL antibody library. Here, we describe the first successful selections of specific antibodies by PCA, and we discuss the opportunities and limitations of this approach. First, we enriched an antibody specific for the capsid protein D of bacteriophage lambda (gpD) by ten successive rounds of competitive liquid culture selection. In an independent approach, we selected a specific antibody for the c-Jun N-terminal kinase 2 (JNK2) in a single-step selection setup. In order to obtain specific antibodies in only a single PCA selection round, the selection system was thoroughly investigated and several strategies to reduce the amount of false positives were evaluated. When expressed in the cytoplasm of Escherichia coli, the PCA-selected scFv antibody fragments could be purified as soluble and monomeric proteins. Denaturant-induced unfolding experiments showed that both antibody fragments are stable molecules, even when the disulfide bonds are reduced. Furthermore, antigen-specificity of the PCA-selected antibody fragments is demonstrated by in vivo and in vitro experiments. As antigen binding is retained regardless of the antibody redox state, both PCA-selected antibody fragments can tolerate the loss of disulfide bridge formation. Our results illustrate that it is possible to select well-expressed, stable, antigen-specific, and intracellular functional antibodies by PCA directly.     

Engineered Cystine-Knot Peptides that Bind αvβ3 Integrin with Antibody-Like Affinities

The α_vβ₃ integrin receptor is an important cancer target due to its overexpression on many solid tumors and the tumor neovasculature and its role in metastasis and angiogenesis. We used a truncated form of the Agouti-related protein (AgRP), a 4-kDa cystine-knot peptide with four disulfide bonds and four solvent-exposed loops, as a scaffold for engineering peptides that bound to α_vβ₃ integrins with high affinity and specificity. A yeast-displayed cystine-knot peptide library was generated by substituting a six amino acid loop of AgRP with a nine amino acid loop containing the Arg-Gly-Asp integrin recognition motif and randomized flanking residues. Mutant cystine-knot peptides were screened in a high-throughput manner by fluorescence-activated cell sorting to identify clones with high affinity to detergent-solubilized α_vβ₃ integrin receptor. Select integrin-binding peptides were expressed recombinantly in Pichia pastoris and were tested for their ability to bind to human cancer cells expressing various integrin receptors. These studies showed that the engineered AgRP peptides bound to cells expressing α_vβ₃ integrins with affinities ranging from 15 nM to 780 pM. Furthermore, the engineered peptides were shown to bind specifically to α_vβ₃ integrins and had only minimal or no binding to α_vβ₅, α₅β₁, and α_iibβ₃ integrins. The engineered AgRP peptides were also shown to inhibit cell adhesion to the extracellular matrix protein vitronectin, which is a naturally occurring ligand for α_vβ₃ and other integrins. Next, to evaluate whether the other three loops of AgRP could modulate integrin specificity, we made second-generation libraries by individually randomizing these loops in one of the high-affinity integrin-binding variants. Screening of these loop-randomized libraries against α_vβ₃ integrins resulted in peptides that retained high affinities for α_vβ₃ and had increased specificities for α_vβ₃ over α_iibβ₃ integrins. Collectively, these data validate AgRP as a scaffold for protein engineering and demonstrate that modification of a single loop can lead to AgRP-based peptides with antibody-like affinities for their target.     

Synthetic antibody libraries focused towards peptide ligands

Synthetic antibody libraries have proven immensely useful for the de novo isolation of antibodies without the need for animal immunization. Recently, focused libraries designed to recognize particular classes of ligands, such as haptens or proteins, have been employed to facilitate the selection of high-affinity antibodies. Focused libraries are built using V regions encoding combinations of canonical structures that resemble the structural features of antibodies that bind the desired class of ligands and sequence diversity is introduced at residues typically involved in recognition. Here we describe the generation and experimental validation of two different single-chain antibody variable fragment libraries that efficiently generate binders to peptides, a class of molecules that has proven to be a difficult target for antibody generation. First, a human anti-peptide library was constructed by diversifying a scaffold: the human variable heavy chain (V_H) germ line gene 3-23, which was fused to a variant of the human variable light chain (V_L) germ line gene A27, in which L1 was modified to encode the canonical structure found in anti-peptide antibodies. The sequence diversity was introduced into 3-23 (V_H) only, targeting for diversification residues commonly found in contact with protein and peptide antigens. Second, a murine library was generated using the antibody 26-10, which was initially isolated based on its affinity to the hapten digoxin, but also binds peptides and exhibits a canonical structure pattern typical of anti-peptide antibodies. Diversity was introduced in the V_H only using the profile of amino acids found at positions that frequently contact peptide antigens. Both libraries yielded binders to two model peptides, angiotensin and neuropeptide Y, following screening by solution phage panning. The mouse library yielded antibodies with affinities below 20 nM to both targets, although only the V_H had been subjected to diversification.     

Efficient Recovery of High-Affinity Antibodies from a Single-Chain Fab Yeast Display Library

Yeast display is a powerful technology for the isolation of monoclonal antibodies (mAbs) against a target antigen. Antibody libraries have been displayed on the surface of yeast as both single-chain variable fragment (scFv) and antigen binding fragment (Fab). Here, we combine these two formats to display well-characterized mAbs as single-chain Fabs (scFabs) on the surface of yeast and construct the first scFab yeast display antibody library. When expressed on the surface of yeast, two out of three anti-human immunodeficiency virus (HIV)-1 mAbs bound with higher affinity as scFabs than scFvs. Also, the soluble scFab preparations exhibited binding and neutralization profiles comparable to that of the corresponding Fab fragments. Display of an immune HIV-1 scFab library on the surface of yeast, followed by rounds of sorting against HIV-1 gp120, allowed for the selection of 13 antigen-specific clones. When the same cDNA was used to construct the library in an scFv format, a similar number but a lower affinity set of clones were selected. Based on these results, yeast-displayed scFab libraries can be constructed and selected with high efficiency, characterized without the need for a reformatting step, and used to isolate higher-affinity antibodies than scFv libraries.     

mRNA Display Design of Fibronectin-based Intrabodies That Detect and Inhibit Severe Acute Respiratory Syndrome Coronavirus Nucleocapsid Protein

The nucleocapsid (N) protein of severe acute respiratory syndrome (SARS) coronavirus plays important roles in both viral replication and modulation of host cell processes. New ligands that target the N protein may thus provide tools to track the protein inside cells, detect interaction hot spots on the protein surface, and discover sites that could be used to develop new anti-SARS therapies. Using mRNA display selection and directed evolution, we designed novel antibody-like protein affinity reagents that target SARS N protein with high affinity and selectivity. Our libraries were based on an 88-residue variant of the 10th fibronectin type III domain from human fibronectin (10Fn3). This selection resulted in eight independent 10Fn3 intrabodies, two that require the N-terminal domain for binding and six that recognize the C terminus, one with K_d = 1.7 nm. 10Fn3 intrabodies are well expressed in mammalian cells and are relocalized by N in SARS-infected cells. Seven of the selected intrabodies tested do not perturb cellular function when expressed singly in vivo and inhibit virus replication from 11- to 5900-fold when expressed in cells prior to infection. Targeting two sites on SARS-N simultaneously using two distinct 10Fn3s results in synergistic inhibition of virus replication.     

A simple hybridoma screening method for high-affinity monoclonal antibodies using the signal ratio obtained from time-resolved fluorescence assay

In hybridoma screening, quantitative kinetic evaluation is difficult since the concentration of each antibody in the hybridoma supernatant is unknown. From modeling calculations, we hypothesized that the ratio of two different antigen-antibody concentrations might allow discrimination of high-affinity monoclonal antibodies irrespective of the antibody concentration. Using anti-alpha-fetoprotein monoclonal antibodies of known affinity, we set the signal ratio of a time-resolved assay at >0.1, in which the antigen concentrations were 10 and 100 ng/mL. From anti-alpha-fetoprotein hybridoma screening with this assay, it was possible to effectively select high-affinity monoclonal antibodies with KD values below 1x10(-8) M. High-sensitivity sandwich enzyme-linked immunosorbent assay which detects domain III of alpha-fetoprotein has been established using selected high-affinity monoclonal antibodies. This screening method is useful for selection of high-affinity monoclonal antibodies of potential diagnostic value.     

Investigation and molecular mimicry of the antigen involved in the interaction between the monoclonal antibody 5D10 and the human breast cancer cell line MCF-7

Monoclonal antibody (mAb) 5D10 is directed against the human breast cancer cell line MCF-7. Biochemical characterization of the antibody epitope was attempted and revealed a complex, most likely carbohydrate-linked nature, which prevented isolation and further studies of the interaction. A major goal of this work was to generate structural mimics of the 5D10 epitope to serve as putative substitutes in such studies. A peptide library displayed on filamentous phage was used to select for mimotope peptide sequences. All positive phage clones selected from the library displayed the amino acid sequence H(2)N-QMNPMYYR-CO(2)H. This peptide sequence, as well as a branched form of the peptide, was found to bind mAb 5D10. Moreover, both peptide sequences were able to inhibit the binding of 5D10 to the MCF-7 cells in a concentration-dependent manner, with an EC(50) value in the range of 65 microM. According to these results, random phage peptide libraries can serve to identify mimotopic peptides for unknown complex cell surface epitopes.     

Simplified methods for construction, assessment and rapid screening of peptide libraries in bacteriophage.

An efficient strategy has been devised for the construction of diverse peptide libraries in bacteriophage vectors. This strategy was used to generate a library of 4 x 10(8) random decapeptide inserts in the pIII protein of bacteriophage fd. A novel method for evaluating the genetic diversity of bacteriophage libraries based on colony hybridization with partially degenerate oligonucleotides has been developed. The decapeptide library was affinity-selected with a previously characterized monoclonal antibody specific for the V3 loop of the gp120 protein of HIV-1. Immunological screening, an efficient technique for the rapid identification of putative binding bacteriophage, is described. Hexapeptide sequences similar to those obtained from affinity selection of a hexapeptide bacteriophage library were obtained from the decapeptide library in all five frames. Immunological screening of 20,000 clones from the two libraries after two rounds of affinity selection rapidly identified antibody-binding sequences; 93% and 86% of the sequences obtained from the hexapeptide and decapeptide libraries, respectively, had IC50 values < or = 10 mM as free peptides.     

High-affinity human antibodies from phage-displayed synthetic Fab libraries with a single framework scaffold

Phage-displayed synthetic antibody libraries were built on a single human framework by introducing synthetic diversity at solvent-exposed positions within the heavy chain complementarity-determining regions (CDRs). The design strategy of mimicking natural diversity using tailored codons had been validated previously with scFv libraries, which produced antibodies that bound to antigen, murine vascular endothelial growth factor (mVEGF), with affinities in the 100nM range. To improve library performance, we constructed monovalent and bivalent antigen-binding fragment (Fab) libraries, and explored different CDR-H3 diversities by varying the amino acid composition and CDR length. A Fab with sub-nanomolar affinity for mVEGF was obtained from a library with CDR-H3 diversity designed to contain all 20 naturally occurring amino acids. We then expanded the library by increasing the variability of CDR-H3 length and using tailored codons that mimicked the amino acid composition of natural CDR-H3 sequences. The library was tested against a panel of 13 protein antigens and high-affinity Fabs were obtained for most antigens. Furthermore, the heavy chain of an anti-mVEGF clone was recombined with a library of light chain CDRs, and the affinity was improved from low nanomolar to low picomolar. The results demonstrated that high-affinity human antibodies can be generated from libraries with completely synthetic CDRs displayed on a single scaffold.     

Construction of a large phage-displayed human antibody domain library with a scaffold based on a newly identified highly soluble, stable heavy chain variable domain

Currently, almost all U.S. Food and Drug Administration-approved therapeutic antibodies and the vast majority of those in clinical trials are full-size antibodies mostly in an immunoglobulin G1 format of about 150 kDa in size. Two fundamental problems for such large molecules are their poor penetration into tissues (e.g., solid tumors) and poor or absent binding to regions on the surface of some molecules [e.g., on the human immunodeficiency virus envelope glycoprotein (Env)] that are accessible by molecules of smaller size. We have identified a phage-displayed heavy chain-only antibody by panning of a large (size, ∼ 1.5 × 10¹⁰) human naive Fab (antigen-binding fragment) library against an Env and found that the heavy chain variable domain (V_H) of this antibody, designated as m0, was independently folded, stable, highly soluble, monomeric, and expressed at high levels in bacteria. m0 was used as a scaffold to construct a large (size, ∼ 2.5 × 10¹⁰), highly diversified phage-displayed human V_H library by grafting naturally occurring complementarity-determining regions (CDRs) 2 and 3 of heavy chains from five human antibody Fab libraries and by randomly mutating four putative solvent-accessible residues in CDR1 to A, D, S, or Y. The sequence diversity of all CDRs was determined from 143 randomly selected clones. Most of these V_Hs were with different CDR2 origins (six of seven groups of V_H germlines) or CDR3 lengths (ranging from 7 to 24 residues) and could be purified directly from the soluble fraction of the Escherichia coli periplasm. The quality of the library was also validated by successful selection of high-affinity V_Hs against viral and cancer-related antigens; all selected V_Hs were monomeric, easily expressed, and purified with high solubility and yield. This library could be a valuable source of antibodies targeting size-restricted epitopes and antigens in obstructed locations where efficient penetration could be critical for successful treatment.     

Selection of functional human antibodies from retroviral display libraries

Antibody library technology represents a powerful tool for the discovery and design of antibodies with high affinity and specificity for their targets. To extend the technique to the expression and selection of antibody libraries in an eukaryotic environment, we provide here a proof of concept that retroviruses can be engineered for the display and selection of variable single-chain fragment (scFv) libraries. A retroviral library displaying the repertoire obtained after a single round of selection of a human synthetic scFv phage display library on laminin was generated. For selection, antigen-bound virus was efficiently recovered by an overlay with cells permissive for infection. This approach allowed more than 10³-fold enrichment of antigen binders in a single selection cycle. After three selection cycles, several scFvs were recovered showing similar laminin-binding activities but improved expression levels in mammalian cells as compared with a laminin-specific scFv selected by the conventional phage display approach. Thus, translational problems that occur when phage-selected antibodies have to be transferred onto mammalian expression systems to exert their therapeutic potential can be avoided by the use of retroviral display libraries.     

Construction of Artificial TNF-Binding Proteins Based on the 10th Human Fibronectin Type III Domain Using Bacterial Display

Construction of antibody mimetics on the base of alternative scaffold proteins is a promising strategy for obtaining new products for medicine and biotechnology. The aim of our work was to optimize the cell display system for the 10th human fibronectin type III domain (¹⁰Fn3) scaffold protein based on the AT877 autotransporter from Psychrobacter cryohalolentis K5^T and to construct new artificial TNF-binding proteins. We obtained a ¹⁰Fn3 gene combinatorial library and screened it using the bacterial display method. After expression of the selected ¹⁰Fn3 variants in Escherichia coli cells and analysis of their TNF-binding activity, we identified proteins that display high affinity for TNF and characterized their properties.