Compound Activity Mapping for Natural Products Discovery

Roger Linington

Abstract

This project incorporated image-based biological screening data and high-resolution mass spectrometry data for the discovery of next-generation therapeutics. Project Description: Despite the widespread use of natural products as inspirations for existing drugs, the rate of development of new natural product-based drugs has slowed in recent years. This is due in part to an increase in the rates of rediscovery of existing scaffolds, as well as limited early characterization of biological functions of hits from primary screens. However, recent whole genome sequencing efforts are revealing a wealth of untapped biosynthetic diversity in environmental microorganisms, suggesting that there exists a vast reservoir of potential new drug leads from natural sources. Our laboratory is interested in the development of new high-content approaches to natural products discovery that use recent advances in image-based screening and high resolution mass spectrometry to create tools for the 'function-first' annotation of natural products libraries. The goal of this approach is to integrate high-throughput chemical and biological annotation of natural product libraries to provide accurate predictions of compound biological function early in the discovery process. The ultimate objective of this approach is to invert the natural products discovery model by creating a platform capable of characterizing the structural and biological features of all bioactive compounds in any natural products library prior to isolation and structure elucidation. Although the best case scenario would be the production of large screening libraries of pure natural products, all with known structure and of known concentration, this is unrealistic from a practical perspective, given that any natural products extract can contain up to 100 constituents. Instead, this platform aims to develop new informatic tools to allow the integration of multiparametric datasets from both high content screening and untargeted metabolomics to extract the data about bioactive constituents directly from these complex mixtures. Ultimately, this new platform will inform the process of drug discovery and development for a range of treatments including cancer therapeutics and new antibiotic discovery.