DNA-Encoded Libraries (DELs) of dual-display setup have been used for discovering simultaneously binding fragments. Furthermore, dual-display can now be combined with conventional split-and-pool derived single-display libraries yielding very large high-quality DELs by combinatorial assembly. Such libraries feature large structural diversity and selection results with these libraries will be presented for a panel of target proteins.

The growing scope of chemistries for DNA-encoded chemical library (DECL) synthesis has intensified the need for quantitative methods for validating their compatibility with DNA. We have developed a comprehensive approach for the assessment of chemical fidelity (reaction yield and purity), encoding fidelity (ligation efficiency), and readability (DNA compatibility), revealing the fate of the DNA-tag during DECL chemistry from a single platform. I will describe its utility in screening on-DNA chemistries.

While large generic DNA-encoded libraries (DEL) platforms are typically effective in drug discovery, they are expensive to make and can be associated with technical problems such as library heterogeneity and under-sampling. We generated DELs that are focused to specific protein classes allowing high screening productivity at low cost and providing reliable structure-activity information. The concept was demonstrated with a library targeting proteins with NAD(+)-binding pockets.

DNA-encoded libraries (DEL) sample vastly larger and more diverse chemical spaces than standard HTS collections, but rely on affinity selection of DNA-displayed small molecules to identify hits. The DNA tags are a known interference for some targets, particularly nucleic acid-binding proteins. We have developed an off-DNA macromolecular binding analysis
using solid-phase DELs, microfluidic droplets, and fluorescence polarization detection. Application to several targets will be discussed.

It is typical to find more candidate binders from DEL selections than is reasonable to pursue. Common practice is to limit synthesis to compounds estimated to be high-affinity binders; of the factors contributing to this estimation, the strength of the sequence signals of a compound is always important. We describe here methods and equations to relate sequence signal to the prospective estimation of compound affinity and chemical yield from DEL selection data.

Protein construct assessment before ELT selection is crucial for the success of ELT selection. We utilized affinity selection-mass spectrometry (ASMS), dynamic light scattering, and Bioanalyzer to characterize the constructs and optimize ELT selection conditions with various buffers and temperatures. Many features were successfully selected in the IDO1 ELT selection. To achieve high-throughput binder confirmation (HTBC), ASMS was used in the follow-up assay and confirmed 91 ELT primary binder.

We will discuss the discovery of a selective inhibitor of discoidin domain receptor1 (DDR1) and the progression from a DNA-encoded library hit to an in vivo active lead series, challenges of on-DNA synthesis of highly-functionalized and rigid heterocycles, and present possible solutions using the Pictet-Spengler reaction as an example.

Indoleamine 2,3-dioxogenase-1 (IDO1) is induced and activated in response to viral and bacterial infection causing a dysfunctional immune response in clearing pathogens. IDO1 inhibitors (IDO1i) have the potential to restore immune function in indications such as cancer and infection. A structurally-unique IDO1i class was discovered through the affinity selection of a novel DNA-encoded library. After additional medicinal chemistry iterations, the compound series was elaborated into potential best in class preclinical molecule.

At SyntheX, we have developed a novel approach to drug discovery by using synthetic biology to create cell-based drug discovery engines for challenging targets; ToRPPIDO is focused on protein-protein interaction (PPI) disruption and ToRNeDO is focused on discovering novel molecular degraders of proteins. Both technologies rely on functional first-pass selections to identify rare molecules that can perform complex intracellular functions. We aim to expand the drug discovery toolkit to enable access to targets that have been previously deemed 'undruggable'. We couple our intracellular selections with genetically-encoded libraries of peptides and macrocycles to generate molecular probes from a first pass screen. These initial molecules can then be used to discover biological insights such as new binding pockets and allosteric sites, or be turned into drugs using medicinal chemistry approaches.  

Using ToRPPIDO, we developed STX100, a peptide originating from an encoded library, targeting an intracellular protein-protein interaction in the homologous recombination DNA repair pathway. STX100-mediated cell killing is independent of canonical cell death mechanisms; it relies on acute calcium release from its target to elicit cell death. The mechanism translates to in vivo models, where a local delivery of STX100 and a combination of immune checkpoint blockade (ICB) agents can cure established tumors resistant to ICB therapies.