Track: Discovery and Basic Research - Medicinal Chemistry - Rational Drug Design
Category: Poster Abstract
Multi-cytokine Inhibition and NKG2A+CD8 T-Cells in Humans with Immune and Hematologic Disorders: Computer-Aided Drug Discovery
Purpose: Cytokines play a role in a myriad of pathophysiological pathways and biological processes. Anticytokine therapy has been adopted in clinical practice to treat hepatitis B, multiple chlorosis, rheumatoid arthritis, and various carcinomas, to name a few. However, biologics are typically used for anticytokine therapy and are thought to have poor efficacy due to the poor pharmacokinetic profiles. In addition, anticytokine agents are selective, thus they are ineffective against disorders caused by multiple cytokines. The issue of drug selectivity is particularly challenging due to the unique synergistic mechanism of action of cytokines whereby inhibition of one will induce the activity of other cytokines to stimulate the same target cells. Our collaborator has developed a novel peptide (BNZ-1) that binds and inhibits γc/CD132, a cytokine receptor that is shared by IL-2, -4, -7, -9, -15 and -2. BNZ-1 multi-cytokine inhibitor (MCI) peptide is being studied for the treatment of human T-cell malignancies (Open label phase II clinical trial is in progress). Due to the pharmacokinetics and immunogenicity concerns of peptides as drugs, our research team has initiated the search for small molecules and or macrocycles that can mimic the behavior and mechanism of action of BNZ-1. Methods: We exploited the availability of the crystal structures of γc (PDB ID 2ERJ) and employed computer-aided drug design (CADD) approaches to accelerate the exploration and discovery of small molecule MCIs. High-throughput virtual screening (HTVS) of our in-house CUWSOP chemical library and ChemBridge macrocycles library as we recognize the difficulty in finding small molecules that can cover large surface areas of proteins as peptides. Our rigorous workflow included filtering chemical libraries to remove promiscuous, non-drug-like, and potentially toxic molecules using PAINs and ADME / Tox Filter prior to docking. To ensure sampling of structurally diverse molecules, we clustered the top-ranked hits and selected structures from each cluster-subset for biological screening. Results: Five molecules were initially selected, from the results of high-throughput virtual screening (HTVS) of our in-house CUWSOP chemical library, for bioassay-based screening. These molecules were selected based predicted binding affinities and on interactions with key residues of γc (Fig. 1) reported to maintain IL-2/γC interface. The initial biological screen suggests compound 7671 is a general inhibitor of IL-2 and 3 at concentrations > 0.5 μM (Fig. 2). However, compound 2090 exhibited promising results as a potential multi-cytokine modulator as it dose-dependently inhibited IL-2 action but enhanced IL-3 action (Fig. 2). Conclusion: Due to the promising preliminary biological results of compound 2090, we are conducting computer-guided search for analogs of 2090 for screening to extract meaningful structure-activity relationship (SAR) data. The development of a SAR library of 2090 analogs will guide our efforts to systematically and rationally optimize compound 2090.