Recent Advances on the Selection Methods of DNA-Encoded Libraries

DNA-encoded libraries (DEL) have come of age and become a major technology platform for ligand discovery in both academia and the pharmaceutical industry. Technological maturation in the past two decades and the recent explosive developments of DEL-compatible chemistries have greatly improved the chemical diversity of DELs and fueled its applications in drug discovery. A relatively less-covered aspect of DELs is the selection method. Typically, DEL selection is considered as a binding assay and the selection is conducted with purified protein targets immobilized on a matrix, and the binders are separated from the non-binding background via physical washes. However, the recent innovations in DEL selection methods have not only expanded the target scope of DELs, but also revealed the potential of the DEL technology as a powerful tool in exploring fundamental biology. In this Review, we first cover the “classic” DEL selection methods with purified proteins on solid phase, and then we discuss the strategies to realize DEL selections in solution phase. Finally, we focus on the emerging approaches for DELs to interrogate complex biological targets.     

Protein Chemical Synthesis Combined with Mirror-Image Phage Display Yields d-Peptide EGF Ligands that Block the EGF–EGFR Interaction

The epidermal growth factor (EGF) pathway, being overactive in a number of cancers, is a good target for clinical therapy. Although several drugs targeting the EGF receptor (EGFR) are on the market, tumours acquire resistance very rapidly. As an alternative, small molecules and peptides targeting EGF have been developed, although with moderate success. Herein, we report the use of mirror-image phage display technology to discover protease-resistant peptides with the capacity to inhibit the EGF–EGFR interaction. After the chemical synthesis of the enantiomeric protein d -EGF, two phage-display peptide libraries were used to select binding sequences. The d versions of these peptides bound to natural EGF, as confirmed by surface acoustic waves (SAWs). High-field NMR spectroscopy showed that the best EGF binder, d -PI_4, interacts preferentially with an EGF region that partially overlaps with the receptor binding interface. Importantly, we also show that d -PI_4 efficiently disrupts the EGF–EGFR interaction. This methodology represents a straightforward approach to find new protease-resistant peptides with potential applications in cancer therapy.     

Opportunities for Expanding Encoded Chemical Diversification and Improving Hit Enrichment in mRNA-Displayed Peptide Libraries

DNA-encoded small-molecule libraries and mRNA displayed peptide libraries both use numerically large pools of oligonucleotide-tagged molecules to identify potential hits for protein targets. They differ dramatically, however, in the ‘drug-likeness’ of the molecules that each can be used to discover. We give here an overview of the two techniques, comparing some advantages and disadvantages of each, and suggest areas where particularly mRNA display can benefit from adopting advances developed with DNA-encoded small molecule libraries. We outline cases where chemical modification of the peptide library has already been used in mRNA display, and survey opportunities to expand this using examples from DNA-encoded small molecule libraries. We also propose potential opportunities for encoding such reactions within the mRNA/cDNA tag of an mRNA-displayed peptide library to allow a more diversity-oriented approach to library modification. Finally, we outline alternate approaches for enriching target-binding hits from a pooled and tagged library, and close by detailing several examples of how an adjusted mRNA-display based approach could be used to discover new ‘drug-like’ modified small peptides.     

Chemical Space Overlap with Critical Protein–Protein Interface Residues in Commercial and Specialized Small-Molecule Libraries

There is growing interest in the use of structure-based virtual screening to identify small molecules that inhibit challenging protein–protein interactions (PPIs). In this study, we investigated how effectively chemical library members docked at the PPI interface mimic the position of critical side-chain residues known as “hot spots”. Three compound collections were considered, a commercially available screening collection (ChemDiv), a collection of diversity-oriented synthesis (DOS) compounds that contains natural-product-like small molecules, and a library constructed using established reactions (the “screenable chemical universe based on intuitive data organization”, SCUBIDOO). Three different tight PPIs for which hot-spot residues have been identified were selected for analysis: uPAR⋅uPA, TEAD4⋅Yap1, and Ca_Vα⋅Ca_Vβ. Analysis of library physicochemical properties was followed by docking to the PPI receptors. A pharmacophore method was used to measure overlap between small-molecule substituents and hot-spot side chains. Fragment-like conformationally restricted small molecules showed better hot-spot overlap for interfaces with well-defined pockets such as uPAR⋅uPA, whereas better overlap was observed for more complex DOS compounds in interfaces lacking a well-defined binding site such as TEAD4⋅Yap1. Virtual screening of conformationally restricted compounds targeting uPAR⋅uPA and TEAD4⋅Yap1 followed by experimental validation reinforce these findings, as the best hits were fragment-like and had few rotatable bonds for the former, while no hits were identified for the latter. Overall, such studies provide a framework for understanding PPIs in the context of additional chemical matter and new PPI definitions.     

Highly Diverse Protein Library Based on the Ubiquitous (β/α)8 Enzyme Fold Yields Well‐Structured Proteins through in Vitro Folding Selection

Proper protein folding is a prerequisite for protein stability and enzymatic activity. Although directed evolution can be a powerful tool to investigate enzymatic function and to isolate novel activities, well‐designed libraries of folded proteins are essential. In vitro selection methods are particularly capable of searching for enzymatic activities in libraries of trillions of protein variants, yet high‐quality libraries of well‐folded enzymes with such high diversity are lacking. We describe the construction and detailed characterization of a folding‐enriched protein library based on the ubiquitous (β/α)₈ barrel fold, which is found in five of the six enzyme classes. We introduced seven randomized loops on the catalytic face of the monomeric, thermostable (β/α)₈ barrel of glycerophosphodiester phosphodiesterase (GDPD) from Thermotoga maritima. We employed in vitro folding selection based on protease digestion to enrich intermediate libraries containing three to four randomized loops for folded variants, and then combined them to assemble the final library (10¹⁴ DNA sequences). The resulting library was analyzed by using the in vitro protease assay and an in vivo GFP‐folding assay; it contains ～10¹² soluble monomeric protein variants. We isolated six library members and demonstrated that these proteins are soluble, monomeric and show (β/α)₈‐barrel fold‐like secondary and tertiary structure. The quality of the folding‐enriched library improved up to 50‐fold compared to a control library that was assembled without the folding selection. To the best of our knowledge, this work is the first example of combining the ultra‐high throughput mRNA display method with selection for folding. The resulting (β/α)₈ barrel libraries provide a valuable starting point to study the unique catalytic capabilities of the (β/α)₈ fold, and to isolate novel enzymes.     

Correctors of protein trafficking defects identified by a novel high-throughput screening assay

High-throughput small-molecule screens hold great promise for identifying compounds with potential therapeutic value in the treatment of protein-trafficking diseases such as cystic fibrosis (CF) and nephrogenic diabetes insipidus (NDI). The approach usually involves expressing the mutant form of the gene in cells and assaying function in a multiwell format when cells are exposed to libraries of compounds. Although such functional assays are useful, they do not directly test the ability of a compound to correct defective trafficking of the protein. To address this we have developed a novel corrector-screening assay for CF, in which the appearance of the mutant protein at the cell surface is measured. We used this assay to screen a library of 2000 compounds and have isolated several classes of trafficking correctors that had not previously been identified. This novel screening approach to protein-trafficking diseases is robust and general, and could enable the selection of molecules that could be translated rapidly to a clinical setting.     

Molecules that target beta-amyloid

The devastating effects of Alzheimer's and related amyloidogenic diseases have inspired the synthesis and evaluation of numerous ligands to understand the molecular mechanism of the aggregation of the beta-amyloid peptide. Our review focuses on the current knowledge in this field with respect to molecules that have been demonstrated to interact with either oligomeric or fibrillar forms of the beta-amyloid peptide. We describe natural proteins, peptides, peptidomimetics, and small molecules that have been found to interfere with beta-amyloid aggregation. We also detail recent efforts in selecting molecules that target beta-amyloid isolated from antibody, protein, and peptide libraries. These new molecules will likely aid in deciphering the details of the aggregation pathway for the beta-amyloid peptide and provide reagents that may stabilize relevant oligomeric intermediates which likely have bearing on the pathophysiology of Alzheimer's disease. Moreover, the described anti-amyloid molecular toolbox will also provide an avenue for designing new diagnostic and therapeutic reagents.     

Engineering of the Escherichia coli Im7 immunity protein as a loop display scaffold

Protein scaffolds derived from non-immunoglobulin sources are increasingly being adapted and engineered to provide unique binding molecules with a diverse range of targeting specificities. The ColE7 immunity protein (Im7) from Escherichia coli is potentially one such molecule, as it combines the advantages of (i) small size, (ii) stability conferred by a conserved four anti-parallel alpha-helical framework and (iii) availability of variable surface loops evolved to inactivate members of the DNase family of bacterial toxins, forming one of the tightest known protein-protein interactions. Here we describe initial cloning and protein expression of Im7 and its cognate partner the 15 kDa DNase domain of the colicin E7. Both proteins were produced efficiently in E.coli, and their in vitro binding interactions were validated using ELISA and biosensor. In order to assess the capacity of the Im7 protein to accommodate extensive loop region modifications, we performed extensive molecular modelling and constructed a series of loop graft variants, based on transfer of the extended CDR3 loop from the IgG1b12 antibody, which targets the gp120 antigen from HIV-1. Loop grafting in various configurations resulted in chimeric proteins exhibiting retention of the underlying framework conformation, as measured using far-UV circular dichroism spectroscopy. Importantly, there was low but measurable transfer of antigen-specific affinity. Finally, to validate Im7 as a selectable scaffold for the generation of molecular libraries, we displayed Im7 as a gene 3 fusion protein on the surface of fd bacteriophages, the most common library display format. The fusion was successfully detected using an anti-Im7 rabbit polyclonal antibody, and the recombinant phage specifically recognized the immobilized DNase. Thus, Im7 scaffold is an ideal protein display scaffold for the future generation and for the selection of libraries of novel binding proteins.     

Applications of one-bead one-compound combinatorial libraries and chemical microarrays in signal transduction research

The "one-bead-one-compound" (OBOC) combinatorial library method synthesizes millions of random compounds such that each bead displays only one compound. Bead libraries are screened, and positive beads are isolated for structure analysis. Peptide substrates and inhibitors of protein kinases, and peptide ligands for cell surface receptors have been identified using this method. A novel encoding strategy for OBOC libraries has been developed to identify peptidomimetic and small-molecule ligands that specifically interact with cellular proteins. These ligands will be tested for their effects on cell signaling and used to construct chemical microarrays for further characterization of ligand-protein interactions.     

Biofragments: An Approach towards Predicting Protein Function Using Biologically Related Fragments and its Application to Mycobacterium tuberculosis CYP126

We present a novel fragment‐based approach that tackles some of the challenges for chemical biology of predicting protein function. The general approach, which we have termed biofragments, comprises two key stages. First, a biologically relevant fragment library (biofragment library) can be designed and constructed from known sets of substrate‐like ligands for a protein class of interest. Second, the library can be screened for binding to a novel putative ligand‐binding protein from the same or similar class, and the characterization of hits provides insight into the basis of ligand recognition, selectivity, and function at the substrate level. As a proof‐of‐concept, we applied the biofragments approach to the functionally uncharacterized Mycobacterium tuberculosis (Mtb) cytochrome P450 isoform, CYP126. This led to the development of a tailored CYP biofragment library with notable 3D characteristics and a significantly higher screening hit rate (14 %) than standard drug‐like fragment libraries screened previously against Mtb CYP121 and 125 (4 % and 1 %, respectively). Biofragment hits were identified that make both substrate‐like type‐I and inhibitor‐like type‐II interactions with CYP126. A chemical‐fingerprint‐based substrate model was built from the hits and used to search a virtual TB metabolome, which led to the discovery that CYP126 has a strong preference for the recognition of aromatics and substrate‐like type‐I binding of chlorophenol moieties within the active site near the heme. Future catalytic analyses will be focused on assessing CYP126 for potential substrate oxidative dehalogenation.     

Synthesis of DNA-Encoded Disulfide- and Thioether-Cyclized Peptides

DNA-encoded chemical library technologies enable the screening of large combinatorial libraries of chemically and structurally diverse molecules, including short cyclic peptides. A challenge in the combinatorial synthesis of cyclic peptides is the final step, the cyclization of linear peptides that typically suffers from incomplete reactions and large variability between substrates. Several efficient peptide cyclization strategies rely on the modification of thiol groups, such as the formation of disulfide or thioether bonds between cysteines. In this work, we established a strategy and reaction conditions for the efficient chemical synthesis of cyclic peptide–DNA conjugates based on linking the side chains of cysteines. We tested two different thiol-protecting groups and found that tert-butylthio (S-tBu) works best for incorporating a pair of cysteines, and we show that the DNA-linked peptides can be efficiently cyclized through disulfide and thioether bond formation. In combination with established procedures for DNA encoding, the strategy for incorporation of cysteines may be readily applied for the generation and screening of disulfide- and thioether-cyclized peptide libraries.     

A novel affinity protein selection system based on staphylococcal cell surface display and flow cytometry

Here we describe the first reported use of a Gram-positive bacterial system for the selection of affinity proteins from large combinatorial libraries displayed on the surface of Staphylococcus carnosus. An affibody library of 3 x 10(9) variants, based on a 58 residue domain from staphylococcal protein A, was pre-enriched for binding to human tumor necrosis factor-alpha (TNF-alpha) using one cycle of phage display and thereafter transferred to the staphylococcal host ( approximately 10(6) variants). The staphylococcal-displayed library was subjected to three rounds of flow-cytometric sorting, and the selected clones were screened and ranked by on-cell analysis for binding to TNF-alpha and further characterized using biosensor analysis and circular dichroism spectroscopy. The successful sorting yielded three different high-affinity binders (ranging from 95 pM to 2.2 nM) and constitutes the first selection of a novel affinity protein using Gram-positive bacterial display. The method combines the simplicity of working with a bacterial host with the advantages of displaying recombinant proteins on robust Gram-positive bacteria as well as using powerful flow cytometry in the selection and characterization process.     

Genetic Code Evolution Investigated through the Synthesis and Characterisation of Proteins from Reduced-Alphabet Libraries

The universal genetic code of 20 amino acids is the product of evolution. It is believed that earlier versions of the code had fewer residues. Many theories for the order in which amino acids were integrated into the code have been proposed, considering factors ranging from prebiotic chemistry to codon capture. Several meta-analyses combined these theories to yield a feasible consensus chronology of the genetic code's evolution, but there is a dearth of experimental data to test the hypothesised order. We used combinatorial chemistry to synthesise libraries of random polypeptides that were based on different subsets of the 20 standard amino acids, thus representing different stages of a plausible history of the alphabet. Four libraries were comprised of the five, nine, and 16 most ancient amino acids, and all 20 extant residues for a direct side-by-side comparison. We characterised numerous variants from each library for their solubility and propensity to form secondary, tertiary or quaternary structures. Proteins from the two most ancient libraries were more likely to be soluble than those from the extant library. Several individual protein variants exhibited inducible protein folding and other traits typical of intrinsically disordered proteins. From these libraries, we can infer how primordial protein structure and function might have evolved with the genetic code.     

Automated design of degenerate codon libraries

Degenerate codon libraries are frequently used in protein engineering and evolution studies but are often limited to targeting a small number of positions to adequately limit the search space. To mitigate this, codon degeneracy can be limited using heuristics or previous knowledge of the targeted positions. To automate design of libraries given a set of amino acid sequences, an algorithm (LibDesign) was developed that generates a set of possible degenerate codon libraries, their resulting size, and their score relative to a user-defined scoring function. A gene library of a specified size can then be constructed that is representative of the given amino acid distribution or that includes specific sequences or combinations thereof. LibDesign provides a new tool for automated design of high-quality protein libraries that more effectively harness existing sequence-structure information derived from multiple sequence alignment or computational protein design data.     

A simple dual selection for functionally active mutants of Plasmodium falciparum dihydrofolate reductase with improved solubility

Sufficient solubility of the active protein in aqueous solution is a prerequisite for crystallization and other structural studies of proteins. In this study, we have developed a simple and effective in vivo screening system to select for functionally active proteins with increased solubility by using Plasmodium falciparum dihydrofolate reductase (pfDHFR), a well-known malarial drug target, as a model. Prior to the dual selection process, pfDHFR was fused to green fluorescent protein (GFP), which served as a reporter for solubility. The fusion gene was used as a template for construction of mutated DNA libraries of pfDHFR. Two amino acids with large hydrophobic side chains (Y35 and F37) located on the surface of pfDHFR were selected for site-specific mutagenesis. Additionally, the entire pfDHFR gene was randomly mutated using error-prone PCR. During the first step of the dual selection, mutants with functionally active pfDHFR were selected from two libraries by using bacterial complementation assay. Fluorescence signals of active mutants were subsequently measured and five mutants with increased GFP signal, namely Y35Q + F37R, Y35L + F37T, Y35G + F37L and Y35L + F37R from the site-specific mutant library and K27E from the random mutant library, were recovered. The mutants were expressed, purified and characterized as monofunctional pfDHFR following excision of GFP. Our studies indicated that all mutant pfDHFRs exhibited kinetic properties similar to that of the wild-type protein. For comparison of protein solubility, the maximum concentrations of mutant enzymes prior to aggregation were determined. All mutants selected in this study exhibited 3- to 6-fold increases in protein solubility compared with the wild-type protein, which readily aggregated at 2 mg/ml. The dual selection system we have developed should be useful for engineering functionally active protein mutants with sufficient solubility for functional/structural studies and other applications.     

Modulating Protein–Protein Interactions with Visible-Light-Responsive Peptide Backbone Photoswitches

Life relies on a myriad of carefully orchestrated processes, in which proteins and their direct interplay ultimately determine cellular function and disease. Modulation of this complex crosstalk has recently attracted attention, even as a novel therapeutic strategy. Herein, we describe the synthesis and characterization of two visible-light-responsive peptide backbone photoswitches based on azobenzene derivatives, to exert optical control over protein–protein interactions (PPI). The novel peptidomimetics undergo fast and reversible isomerization with low photochemical fatigue under alternatively blue-/green-light irradiation cycles. Both bind in the nanomolar range to the protein of interest. Importantly, the best peptidomimetic displays a clear difference between isomers in its protein-binding capacity and, in turn, in its potential to inhibit enzymatic activity through PPI disruption. In addition, crystal structure determination, docking and molecular dynamics calculations allow a molecular interpretation and open up new avenues in the design and synthesis of future photoswitchable PPI modulators.     

Site-directed protein recombination as a shortest-path problem

Protein function can be tuned using laboratory evolution, in which one rapidly searches through a library of proteins for the properties of interest. In site-directed recombination, n crossovers are chosen in an alignment of p parents to define a set of p(n + 1) peptide fragments. These fragments are then assembled combinatorially to create a library of p(n+1) proteins. We have developed a computational algorithm to enrich these libraries in folded proteins while maintaining an appropriate level of diversity for evolution. For a given set of parents, our algorithm selects crossovers that minimize the average energy of the library, subject to constraints on the length of each fragment. This problem is equivalent to finding the shortest path between nodes in a network, for which the global minimum can be found efficiently. Our algorithm has a running time of O(N(3)p(2) + N(2)n) for a protein of length N. Adjusting the constraints on fragment length generates a set of optimized libraries with varying degrees of diversity. By comparing these optima for different sets of parents, we rapidly determine which parents yield the lowest energy libraries.     

BonMOLière: Small-Sized Libraries of Readily Purchasable Compounds,Optimized to Produce Genuine Hits in Biological Screens across the Protein Space

Experimental screening of large sets of compounds against macromolecular targets is a key strategy to identify novel bioactivities. However, large-scale screening requires substantial experimental resources and is time-consuming and challenging. Therefore, small to medium-sized compound libraries with a high chance of producing genuine hits on an arbitrary protein of interest would be of great value to fields related to early drug discovery, in particular biochemical and cell research. Here, we present a computational approach that incorporates drug-likeness, predicted bioactivities, biological space coverage, and target novelty, to generate optimized compound libraries with maximized chances of producing genuine hits for a wide range of proteins. The computational approach evaluates drug-likeness with a set of established rules, predicts bioactivities with a validated, similarity-based approach, and optimizes the composition of small sets of compounds towards maximum target coverage and novelty. We found that, in comparison to the random selection of compounds for a library, our approach generates substantially improved compound sets. Quantified as the “fitness” of compound libraries, the calculated improvements ranged from +60% (for a library of 15,000 compounds) to +184% (for a library of 1000 compounds). The best of the optimized compound libraries prepared in this work are available for download as a dataset bundle (“BonMOLière”).     

Identification of Inhibitors of Biological Interactions Involving Intrinsically Disordered Proteins

Protein–protein interactions involving disordered partners have unique features and represent prominent targets in drug discovery processes. Intrinsically Disordered Proteins (IDPs) are involved in cellular regulation, signaling and control: they bind to multiple partners and these high-specificity/low-affinity interactions play crucial roles in many human diseases. Disordered regions, terminal tails and flexible linkers are particularly abundant in DNA-binding proteins and play crucial roles in the affinity and specificity of DNA recognizing processes. Protein complexes involving IDPs are short-lived and typically involve short amino acid stretches bearing few “hot spots”, thus the identification of molecules able to modulate them can produce important lead compounds: in this scenario peptides and/or peptidomimetics, deriving from structure-based, combinatorial or protein dissection approaches, can play a key role as hit compounds. Here, we propose a panoramic review of the structural features of IDPs and how they regulate molecular recognition mechanisms focusing attention on recently reported drug-design strategies in the field of IDPs.     

A Minimalist Approach to the Design of Complexity‐Enriched Bioactive Small Molecules: Discovery of Phenanthrenoid Mimics as Antiproliferative Agents

Over the last decades, much effort has been devoted to the design of the “ideal” library for screening, the most promising strategies being those which draw inspiration from biogenic compounds, as the aim is to add biological relevance to such libraries. On the other hand, there is a growing understanding of the role that molecular complexity plays in the discovery of new bioactive small molecules. Nevertheless, the introduction of molecular complexity must be balanced with synthetic accessibility. In this work, we show that both concepts can be efficiently merged—in a minimalist way—by using very simple guidelines during the design process along with the application of multicomponent reactions as key steps in the synthetic process. Natural phenanthrenoids, a class of plant aromatic metabolites, served as inspiration for the synthesis of a library in which complexity‐enhancing features were introduced in few steps using multicomponent reactions. These resulting chemical entities were not only more complex than the parent natural products, but also interrogated an alternative region of the chemical space, which led to an outstanding hit rate in an antiproliferative assay: four out of twenty‐six compounds showed in vitro activity, one of them being more potent than the clinically useful drug 5‐fluorouracil.