Detection of N-labeled metabolites in microbial extracts using AI-designed broadband pulses for H, N heteronuclear NMR spectroscopy.

Approximately 40% of bacterial and mammalian metabolites contain nitrogen-based chemical moieties such as amides, amines, and imines. The identification and quantification of these groups 2D H,N heteronuclear NMR spectroscopy have broadened the catalog of NMR-detected metabolites. However, these NMR experiments necessitate broadband radiofrequency (RF) pulses for inversion and refocusing operations to encompass the full range of N chemical shifts, a challenge that becomes increasingly apparent at high and ultra-high magnetic fields. Here, we show that a newly AI-designed broadband N universal 180° pulse for both inversion and refocusing incorporated in the 2D H, N heteronuclear single quantum coherence (2D H-N BB-HSQC) experiment significantly enhances spectral sensitivity. We demonstrate the advantage of the new technique by analyzing the crude extract of sp. WMMC264, a microbial strain that produces siderophores for iron absorption from the environment. The implementation of the AI-designed pulse in the 2D H-N BB-HSQC experiment will contribute to advancing the analysis of nitrogen-containing metabolites in biological fluids and cell extracts.