Novel Real-Time Library Search driven data acquisition strategy for identification and characterization of metabolites.

Abstract

Identification and structural characterization of novel metabolites in drug discovery or metabolomics experiments is one of the most challenging tasks. Multi-level fragmentation (MSn) based approaches combined with various dissociation modes are frequently utilized for facilitating structure assignment of the unknown compounds. As each of the MS precursors undergoes MSn, the instrument cycle time can limit the total number of precursors analyzed in a single run for complex samples. This necessitates splitting data acquisition into several LC/MS analyses where the results obtained in one acquisition inform the experimental design for the successive experiment. Here we present a new LC/MS data acquisition strategy, termed Met-IQ, where the decision to perform an MSn acquisition is automatically made in real time based on the similarity between an acquired experimental MS2 spectrum and a spectrum in a reference spectral library. Each MS2 spectrum is searched in real time against the spectra for the known compounds of interest. If a similarity to a spectrum in the library is found, the instrument performs a decision-dependent event, such as an MS3 scan. Compared to an intensity-based, data-dependent MSn experiment, only a selective number of MS3 are triggered using Met-IQ, increasing the overall MS2 instrument sampling rate. We applied this strategy to an Amprenavir sample incubated with human liver microsomes. The number of MS2 scan events increased nearly 3.5-fold compared to the standard data dependent experiment where MS3 was triggered for each precursor ion, resulting in identification and structural characterization of 55% more unique metabolites. Furthermore, the MS3 precursor fragments were selected intelligently, focusing on higher mass fragments of sufficient intensity to maximize acquisition of MS3 data relevant for structure assignment. The described Met-IQ strategy is not limited to metabolism experiments, and can be applied to analytical samples where the detection of unknown compounds structurally related to a known compound(s) is sought.