Creating and screening natural product libraries

Abstract

Covering: up to 2020The National Cancer Institute of the United States (NCI) has initiated a Cancer Moonshot program entitled the NCI Program for Natural Product Discovery. As part of this effort, the NCI is producing a library of 1 000 000 partially purified natural product fractions which are being plated into 384-well plates and provided to the research community free of charge. As the first 326 000 of these fractions have now been made available, this review seeks to describe the general methods used to collect organisms, extract those organisms, and create a prefractionated library. Importantly, this review also details both cell-based and cell-free bioassay methods and the adaptations necessary to those methods to productively screen natural product libraries. Finally, this review briefly describes post-screen dereplication and compound purification and scale up procedures which can efficiently identify active compounds and produce sufficient quantities of natural products for further pre-clinical development.

Fig. 1

Natural products may be under-utilized in high throughput screening. A count of publications available on pubmed (https://www.ncbi.nlm.nih.gov/pubmed/) using either “natural products and high throughput screening” (green boxes) or “high throughput screening alone” (red boxes) reveals a profound disparity in publication counts. Green dashed line is at the 50 publication point.

Fig. 2

Overview of automated and high-throughput processes developed at the NCI National Program for Natural Products Discovery (NPNPD) to facilitate the production of a natural product-based screening library. (1) Since 1986, more than 80 000 samples have been acquired through collection agreements based on the NCI letter of collection with each participating source country or their representatives, which stipulates equitable benefit sharing from commercial products derived from discoveries made through these collections. (2) Extracts in the NCI NPR are prepared in a high-throughput manner using both an aqueous and organic solvent extraction process, resulting in two sequential extracts per collected specimen/sample. At present, the US National Cancer Institute’s (NCI) natural product repository contains over 230 000 unique extracts derived from plant, marine and microbial organisms. (3) Extracts (n = 88) are prefractionated on a customized Positive Pressure Solid Phase Extraction workstation (PPSPE) with two robotic arms working in parallel to produce seven fractions per extract (3.5 h; n = 616 fractions). (4) Fractions are dried using high-capacity centrifugal evaporation systems (18 h; n = 2304), and the final mass of each fraction is determined on an automated weighing station (5). (6) An automated sample management system with the capacity to store 1.1 million 2D-barcoded tubes (10 ml) is integrated with robotic systems designed to generate 384-well microtiter plates for HTS and 96-well plates for secondary HPLC-based fractionation of active primary fractions.

Fig. 3

Representative examples of types of active extract samples. NF1-mutant astrocytoma cells expressing luciferase reporters for proliferation and cell viability were treated for 48 h with crude or partially purified (“prefractionated”) natural product extracts (10 μg ml⁻¹). Open bars represent green signal (i.e., proliferative index), black bars represent red signal (i.e., cell health index). (A) Active fraction and inactive crude extract; (B) fraction and crude extract are active; (C) only crude extract is active; (D) toxic crude extract and active fraction *active sample (i.e., >80% reduction in cell proliferative index, <50% reduction in cell health index). Reprinted with permission from C. J. Henrich, L. K. Cartner, J. A. Wilson, R. W. Fuller, A. E. Rizzo, K. M. Reilly, J. B. Mcmahon and K. R. Gustafson, J. Nat. Prod., 2015, 78, 2776–2781. Copyright 2015, American Chemical Society.

Fig. 4

Fluorescence interference assay. Compound 1 was isolated from a fluorescence based protease inhibitor assay and was then rescreened in dose response against the fluorophore alone (7-aminomethylcoumarin) and found not to interfere with fluorescence detection throughout the IC₅₀ range of compound 1. Reprinted with permission from: T. D. Tran, B. A. P. Wilson, C. J. Henrich, L. M. Staudt, L. R. H. Krumpe, E. A. Smith, J. King, K. L. Wendt, A. M. Stchigel, A. N. Miller, R. H. Cichewicz, B. R. O’keefe and K. R. Gustafson, J. Nat. Prod., 2019, 82, 154–162. Copyright 2019 American Chemical Society.

Fig. 5

Pains aware assay development workflow. A schematic of a generalized assay development workflow in anticipation of and incorporating countermeasures against pan assay interference compounds (pains).

Fig. 6

Funding for High Throughput Screening (HTS) for Natural Products Discovery (NPD) has remained stable but 5-fold lower than other HTS campaigns. Compilation of NIH reporter data (https://projectreporter.nih.gov/reporter.cfm) for specific search terms within awarded NIH grant abstracts (green: “high throughput” and “natural products”; red: “high throughput” and “drug discovery”) reveals dramatically fewer grants awarded specifically for NPD.

Fig. 7

Structures and NCI-60 human tumor cell lines screen activity of natural products isolated in two steps from a prefractionated library.