Biochemical combinatorial techniques such as phage display, RNA display and oligonucleotide aptamers have proven to be reliable methods for generation of ligands to protein targets. Adapting these techniques to small synthetic molecules has been a long-sought goal. We report the synthesis and interrogation of an 800-million-member DNA-encoded library in which small molecules are covalently attached to an encoding oligonucleotide. The library was assembled by a combination of chemical and enzymatic synthesis, and interrogated by affinity selection. We describe methods for the selection and deconvolution of the chemical display library, and the discovery of inhibitors for two enzymes: Aurora A kinase and p38 MAP kinase.

Cellular toxicity resulting from nucleation-dependent polymerization of amyloid beta-peptide (Abeta) is considered to be a major and possibly the primary component of Alzheimer's disease (AD). Inhibition of Abeta polymerization has thus been identified as a target for the development of therapeutic agents for the treatment of AD. The intrinsic affinity of Abeta for itself suggested that Abeta-specific interactions could be adapted to the development of compounds that would bind to Abeta and prevent it from polymerizing. Abeta-derived peptides of fifteen residues were found to be inhibitory of Abeta polymerization. The activity of these peptides was subsequently enhanced through modification of their amino termini with specific organic reagents. Additional series of compounds prepared to probe structural requirements for activity allowed reduction of the size of the inhibitors and optimization of the Abeta-derived peptide portion to afford a lead compound, cholyl-Leu-Val-Phe-Phe-Ala-OH (PPI-368), with potent polymerization inhibitory activity but limited biochemical stability. The corresponding all-D-amino acyl analogue peptide acid (PPI-433) and amide (PPI-457) retained inhibitory activity and were both stable in monkey cerebrospinal fluid for 24 h.

Hydrogen exchange measurements on equine lysozyme show that amides in three of the four major helices of the native protein are significantly protected in a molten globule state formed at pH 2. The pattern of protection within the different helices, however, varies significantly. Examination of the pattern in the light of the native structure indicates that the side chains of the protected residues form a compact cluster within the core of the protein. We suggest that such a core is present in the molten globule state, indicating the existence of substantial native-like interactions between hydrophobic residues. The formation of clusters of this type during the early stages of folding could be crucial to directing polypeptide chains to their native structures.

The metalloprotease ADAMTS-5 is considered a potential target for the treatment of osteoarthritis. To identify selective inhibitors of ADAMTS-5, we employed encoded library technology (ELT), which enables affinity selection of small molecule binders from complex mixtures by DNA tagging. Selection of ADAMTS-5 against a four-billion member ELT library led to a novel inhibitor scaffold not containing a classical zinc-binding functionality. One exemplar, (R)-N-((1-(4-(but-3-en-1-ylamino)-6-(((2-(thiophen-2-yl)thiazol-4-yl)methyl)amino)-1,3,5-triazin-2-yl)pyrrolidin-2-yl)methyl)-4-propylbenzenesulfonamide (8), inhibited ADAMTS-5 with IC(50) = 30 nM, showing >50-fold selectivity against ADAMTS-4 and >1000-fold selectivity against ADAMTS-1, ADAMTS-13, MMP-13, and TACE. Extensive SAR studies showed that potency and physicochemical properties of the scaffold could be further improved. Furthermore, in a human osteoarthritis cartilage explant study, compounds 8 and 15f inhibited aggrecanase-mediated (374)ARGS neoepitope release from aggrecan and glycosaminoglycan in response to IL-1β/OSM stimulation. This study provides the first small molecule evidence for the critical role of ADAMTS-5 in human cartilage degradation.

Tuberculosis (TB) is one of the world's oldest and deadliest diseases, killing a person every 20 s. InhA, the enoyl-ACP reductase from Mycobacterium tuberculosis, is the target of the frontline antitubercular drug isoniazid (INH). Compounds that directly target InhA and do not require activation by mycobacterial catalase peroxidase KatG are promising candidates for treating infections caused by INH resistant strains. The application of the encoded library technology (ELT) to the discovery of direct InhA inhibitors yielded compound 7 endowed with good enzymatic potency but with low antitubercular potency. This work reports the hit identification, the selected strategy for potency optimization, the structure−activity relationships of a hundred analogues synthesized, and the results of the in vivo efficacy studies performed with the lead compound 65.

Despite their homologous structure, c-type lysozymes and alpha-lactalbumins have been found to differ profoundly in their unfolding behavior, in that the alpha-lactalbumins readily enter a partially unfolded collapsed state (the "molten globule"), whereas lysozymes unfold cooperatively to a highly unfolded state. The calcium-binding property of lysozyme from equine milk provides an evolutionary link between the two families of proteins. We demonstrate here that equine lysozyme undergoes a two-stage unfolding transition upon heating or in the presence of guanidine hydrochloride that is highly dependent on the state of calcium binding. Differential scanning calorimetry shows the two transitions to be particularly well resolved in the calcium-free protein, where the first transition occurs with a midpoint at 44 degrees C at pH 4.5 or in 0.8 M GdnHCl at pH 7.5, 25 degrees C, and the second occurs near 70 degrees C at pH 4.5 or in 3.7 M GdnHCl at pH 7.5, 25 degrees C. In the presence of calcium, the first transition takes place with a midpoint of 55 degrees C or in excess of 2.5 M GdnHCl, but the parameters for the second transition remain unchanged. Fluorescence emission and UV difference absorption spectroscopy suggest that the first transition generates an intermediate state in which sequestration of some aromatic side chains from solvent has occurred whereas the second represents denaturation to a highly unfolded state. CD and 1H NMR results indicate that the intermediate state possesses extensive secondary and tertiary structure, although the latter is substantially disordered.(ABSTRACT TRUNCATED AT 250 WORDS)

As a potential target for obesity, human BCATm was screened against more than 14 billion DNA encoded compounds of distinct scaffolds followed by off-DNA synthesis and activity confirmation. As a consequence, several series of BCATm inhibitors were discovered. One representative compound (R)-3-((1-(5-bromothiophene-2-carbonyl)-pyrrolidin-3-yl)oxy)-N-methyl-2′-(methylsulfonamido)-[1,1′-biphenyl]-4-carboxamide (15e) from a novel compound library synthesized via on-DNA Suzuki−Miyaura cross-coupling showed BCATm inhibitory activity with IC 50 = 2.0 μM. A protein crystal structure of 15e revealed that it binds to BCATm within the catalytic site adjacent to the PLP cofactor. The identification of this novel inhibitor series plus the establishment of a BCATm protein structure provided a good starting point for future structure-based discovery of BCATm inhibitors.

DNA Encoded Libraries (DELs) use unique DNA sequences to tag each chemical warhead within a library mixture to enable deconvolution following affinity selection against a target protein. With next-generation sequencing, millions to billions of sequences can be read and counted to report binding events. This unprecedented capability has enabled researchers to synthesize and analyze numerically large chemical libraries. Despite the common perception that each library member undergoes a miniaturized affinity assay, selections with higher complexity libraries often produce results that are difficult to rank order. In this study, we aimed to understand the robustness of DEL selection by examining the sequencing readouts of warheads and chemotype families among a large number of experimentally repeated selections. The results revealed that (1) the output of DEL selection is intrinsically noisy but can be reliably modeled by the Poisson distribution, and (2) Poisson noise is the dominating noise at low copy counts and can be estimated even from a single experiment. We also discuss the shortcomings of data analyses based on directly using copy counts and their linear transformations, and propose a framework that incorporates proper normalization and confidence interval calculation to help researchers better understand DEL data.