Natural Products Drug Discovery: On Silica or In-Silico?

Abstract

Natural products have been the most important source for drug development throughout the human history. Over time, the formulation of drugs has evolved from crude drugs to refined chemicals. In modern drug discovery, conventional natural products lead-finding usually uses a top-down approach, namely bio-guided fractionation. In this approach, the crude extracts are separated by chromatography and resulting fractions are tested for activity. Subsequently, active fractions are further refined until a single active compound is obtained. However, this is a painstakingly slow and expensive process. Among the alternatives that have been developed to improve this situation, metabolomics has proved to yield interesting results having been applied successfully to drug discovery in the last two decades. The metabolomics-based approach in lead-finding comprises two steps: (1) in-depth chemical profiling of target samples, e.g. plant extracts, and bioactivity assessment, (2) correlation of the chemical and biological data by chemometrics. In the first step of this approach, the target samples are chemically profiled in an untargeted manner to detect as many compounds as possible. So far, NMR spectroscopy, LC-MS, GC-MS, and MS/MS spectrometry are the most common profiling tools. The profile data are correlated with the biological activity with the help of various chemometric methods such as multivariate data analysis. This in-silico analysis has a high potential to replace or complement conventional on-silica bioassay-guided fractionation as it will greatly reduce the number of bioassays, and thus time and costs. Moreover, it may reveal synergistic mechanisms, when present, something for which the classical top-down approach is clearly not suited. This chapter aims to give an overview of successful approaches based on the application of chemical profiling with chemometrics in natural products drug discovery.

Keywords: Antibiotics; Chemometrics; Correlation analysis; Discriminant analysis; In-silico; Metabolomics; anti-inflammatory; anticancer.