LC-HRMS-Database Screening Metrics for Rapid Prioritization of Samples to Accelerate the Discovery of Structurally New Natural Products.

In order to accelerate the isolation and characterization of structurally new or novel secondary metabolites, it is crucial to develop efficient strategies that prioritize samples with greatest promise early in the workflow so that resources can be utilized in a more efficient and cost-effective manner. We have developed a metrics-based prioritization approach using exact LC-HRMS, which uses data for 24 618 marine natural products held in the PharmaSea database. Each sample was evaluated and allocated a metric score by a software algorithm based on the ratio of new masses over the total (sample novelty), ratio of known masses over the total (chemical novelty), number of peaks above a defined peak area threshold (sample complexity), and peak area (sample diversity). Samples were then ranked and prioritized based on these metric scores. To validate the approach, eight marine sponges and six tunicate samples collected from the Fiji Islands were analyzed, metric scores calculated, and samples targeted for isolation and characterization of new compounds. Structures of new compounds were elucidated by spectroscopic techniques, including 1D and 2D NMR, MS, and MS/MS. Structures were confirmed by computer-assisted structure elucidation methods (CASE) using the ACD/Structure Elucidator Suite.

Targeted Dereplication of Microbial Natural Products by High-Resolution MS and Predicted LC Retention Time.

A new strategy for the identification of known compounds in Streptomyces extracts that can be applied in the discovery of natural products is presented. The strategy incorporates screening a database of 5555 natural products including 5098 structures from Streptomyces sp., using a high-throughput LCMS data processing algorithm that utilizes HRMS data and predicted LC retention times (tR) as filters for rapid identification of known compounds in the natural product extract. The database, named StrepDB, contains for each compound the structure, molecular formula, molecular mass, and predicted LC retention time. All identified compounds are annotated and color coded for easier visualization. It is an indirect approach to quickly assess masses (which are not annotated) that may potentially lead to the discovery of new or novel structures. In addition, a spectral database named MbcDB was generated using the ACD/Spectrus DB Platform. MbcDB contains 665 natural products, each with structure, experimental HRESIMS, MS/MS, UV, and NMR spectra. StrepDB was used to screen a mutant Streptomyces albus extract, which led to the identification and isolation of two new compounds, legonmaleimides A and B, the structures of which were elucidated with the aid of MbcDB and spectroscopic techniques. The structures were confirmed by computer-assisted structure elucidation (CASE) methods using ACD/Structure Elucidator Suite. The developed methodology suggests a pipeline approach to the dereplication of extracts and discovery of novel natural products.