New methyl compounds using the predicted data mining approach (PDMA), coupled with the biotransformation of Streptomyces peucetius O-methyltransferase

Abstract

The structural modification of natural products using biotransformation is an effective way to produce organic compounds in a regioselective and/or stereoselective manner. O-methylation is a major modification in nature. The predicted data mining approach (PDMA) can efficiently screen out biotransformable precursor candidates to produce new (functional) compounds from thousands of derivatives. Herein, an O-methyltransferase (SpOMT2884) from Streptomyces peucetius (ATCC 27952 strain) was selected to biotransform eight precursors, screened from 4364 commercially available natural compounds via PDMA. Seven of the eight precursors (calceolarioside B, isomangiferin, mangiferin, methyl chlorogenate, plantagoside, protosappanin B, and wedelolactone) could be biotransformed from SpOMT2884. Fourteen biotransformed compounds were confirmed to be methyl products with a high-resolution mass analysis. Three methyl products from two precursors (plantagoside and protosappanin B) were selected for further production and identified using nucleic magnetic resonance spectral methods. One methyl product of protosappanin B was identified as 10-O-methyl protosappanin B (1), which gained potent anti-inflammatory activity (IC50 = 76.1 ± 4.7 μM) compared with its precursor, protosappanin B (IC50 = 157.7 ± 5.0 μM). In a case study, two methyl products of plantagoside were indeed identified as novel 4′-O-methyl plantagoside (3) and 5′-O-methyl plantagoside (4). This study demonstrates that PDMA is a good in silico approach for screening biotransformable precursor candidates for the in vitro production of new or bioactive compounds from numerous natural products.