Drug discovery workflows are lengthy and resource intensive. Yet, innovative approaches to screen for novel drug leads can reduce labour and material cost significantly. We envisaged a reductionistic approach to screen for metalloenzyme binding small molecule natural products as potential leads for drug discovery. In lieu of the diverse libraries available to large pharma, we employed native and chemically perturbed bacterial cultures to generate ligand libraries and used immobilised metal affinity chromatography (IMAC) as a surrogate matrix in place of metalloenzymes.1 IMAC is a separation technique relying on coordination chemistry to retain compounds with metal binding properties. The immobilised metal (Cu(II), Ni(II), and Zn(II)) acted as a surrogate for the active site of metalloenzymes and thus could be used to selectively retain small molecule natural products with metal-binding affinity. The initial number of features (>23,000) in native and chemically perturbed Salinispora tropica cultures was sieved to 783 metal-selective metabolites using IMAC. Mass spectrometry-based metabolomics and genome mining analysis subsequently revealed 23 natural products previously undetected in S. tropica, 10 of which were unique to the perturbed cultures. A final list of 17 natural products was curated based on confidence of structural identification,2 similarity of associated biosynthetic gene clusters with the S. tropica genome, and metal selectivity. Ongoing studies aim to further verify compound identity by acquisition of available analytical standards, chemical synthesis, or isolation from upscaled S. tropica cultures. Annotated natural product hits will be validated against a panel of PDB-derived metalloenzymes by inverse virtual screening.3-4 This work is important, as it provides a transferrable and rapid workflow for screening of diverse natural product libraries (including unknowns) for metal-selective metabolites while eliminating non-binders and thus reducing complex mixtures based on their relevance in biological function.
References
1L.M.Roth, M.P.Gotsbacher, R.Codd. JMedChem., 2020, 63, 12116
2I.Blaženović, T.Kind, et al. JCheminform. 2017, 9, 32
3S.De Vita, G.Lauro, et al., JChemInfModel 2019, 59, 4678
4A.Y.Chen, R.N.Adamek, et al., ChemRev. 2019, 119, 1323