Advances in target deconvolution have offered an increasing number of disease-relevant interactions that are difficult to address with traditional small- or large-molecule drugs. Peptides constitute a middle ground that provide large, yet synthetically accessible scaffolds, with the potential for oral delivery. This talk will cover the application of mRNA display to high-throughput peptide screening and the integration of this technology at J&J to enable lead discovery for challenging targets.

Macrocyclic peptides are a modality of high interest to the pharmaceutical industry as a way to inhibit protein-protein interactions. In recent years, DNA-encoded libraries (DEL) have been utilized to generate macrocyclic libraries utilizing non-canonical amino acids with a wide variety of ring-closing chemistries. This talk will describe our efforts to design and produce a novel macrocyclic DEL and highlight a case study using this DEL.

A major obstacle to cyclic peptide development is that little structural information is available, as most cyclic peptides adopt multiple conformations in solution. By combining molecular dynamics simulation and machine learning, we can now provide simulation-quality cyclic peptide structure predictions in seconds to enable structure-based design of cyclic peptides and an understanding of their sequence-activity relationships.

In this talk, we will delve into the integration of yeast surface display and explainable AI in crafting next-generation peptide scaffold libraries. These libraries are optimized for efficient, proper folding. Panning trials with these libraries have yielded peptides with enhanced stability and folding properties, marking a significant advancement in peptide therapeutics.

Cyclic peptides are attracting increasing attention as therapeutic modalities. However, their low membrane permeability significantly limits their applications for drug discovery. Recently, our group has shown that amide-to-ester substitution is an effective strategy to improve the membrane permeability of cyclic peptides. In this presentation, I will discuss the effect of amide-to-ester substitution on membrane permeability and conformational dynamics of cyclic peptides.

Non-degrading molecular glues are of high interest to drug discovery programs.  The natural products rapamycin, FK506, cyclosporin and related analogs have delivered several marketed drugs.  Despite these successes, there remain few approaches towards finding novel chemical matter in this area.  Rapafusyns are macrocyclic peptidic molecular glues that concomitantly bind FKBP12 and a disease target to form a ternary complex. These compounds can be made via modular synthesis in both DEL and array library format to screen against a broad range of difficult to drug intracellular and transmembrane targets.

I present cyclic glycosides related to endomorphin 1 and oxytocin that upon peripheral administration, effectively target their receptors in the CNS. Cyclization of linear peptides enhances serum stability in vivo and the cyclic nature of the peptides provides useful pharmacophores for advancement to the clinic. I will demonstrate how the degree of glycosylation affects biodistribution and BBB penetration, enhancing the promise of endogenous peptide hormones and neurotransmitters as lead compounds.

Loops and helices frequently occur at protein-protein interfaces. I present a new approach to design helical mimics of PPIs. Because loops are more diverse than helices, my method mimics interface loops containing several hot spots. I'll present validation of this dual capping strategy on PPis of medicinal value. The organic fragment of these macrocycles is smaller and more diverse than those derived from RNA-encoded libraries.

Disulfide constrained peptides (DCPs) are characterized by conserved cysteine residues that form intramolecular disulfide bonds. We designed and generated DCP phage libraries with enriched molecular diversity to enable the discovery of ligands against proteins of interest. Using these libraries, we identified highly specific antagonists with super affinity to vascular endothelial growth factor (VEGF), the primary driver for wet AMD. This new modality enables the discovery of next-generation ocular therapeutics.

A macrocyclic peptide was identified as an inhibitor of PD-L1 through an in vitro selection process.  A co-crystal structure of this macrocycle with PD-L1 enabled rapid optimization of this series with respect to PD-L1 inhibitory activity, while also providing insight as to strategies to mitigate off-target liabilities,  ultimately yielding BMS-986189.  This lead macrocycle progressed to the clinic, where PK/PD was evaluated in normal healthy volunteers. Details of these discoveries will be discussed.

Here we will describe a general computational method for identifying closed macrocycles composed of combinations of alpha, beta, gamma, and 17 other amino acids classes with distinct backbone chemistries. The method enables atomically accurate de novo design of permeable macrocycles composed of combinations of canonical and non-canonical backbones. We show using this methodology to develop selective and potent inhibitors of three protein targets.

This study aimed to develop potent, broad-spectrum small cationic peptides with enhanced selectivity for bacterial membranes. The synthesized cyclic peptides showed potent activity against drug-resistant Gram-positive (MIC=1.5-6.2 µg/mL) and Gram-negative (MIC=12.5-25 µg/mL) bacteria. Combined with antibiotics, they demonstrated significant synergistic effects against resistant pathogens. Additional studies confirmed their membranolytic action, supported by calcein dye leakage and electron microscopy. Cytotoxicity assays revealed higher specificity for bacteria over mammalian cells. In vivo experiments utilizing a mouse model with methicillin-resistant Staphylococcus aureus (MRSA) septicemia demonstrated promising pharmacokinetics and efficacy for the lead peptide.