Rapid analysis of antibiotic-containing mixtures from fermentation broths by using liquid chromatography-electrospray ionization-mass spectrometry and matrix-assisted laser desorption ionization-time-of-flight-mass spectrometry.

ABSTRACT

A crucial step in the isolation of antibiotic substances is establishing whether or not the isolated material represents a new chemical entity. Because of the importance of molecular weight to this process-known as dereplication-mass spectrometry has traditionally played an active role. In this communication a strategy for utilizing liquid chromatography-mass spectrometry (LC/MS) for novelty assessment is described. Crude extracts (20-50 µg) are chromatographed by conventional bore high-performance liquid chromatography (1 mL/min) after which a postcolumn split to divert roughly one-tenth of the sample to the mass spectrometer for molecular weight determination by electrospray ionization (ESI) mass spectrometry. The majority of the effluent is sent to a UV detector and ultimately collected as 1-min fractions for biological testing. As a secondary confirmation of molecular weight, an aliquot of each fraction (< 5%) is taken for analysis by matrix-assisted laser desorption ionization (MALDI). The improved efficiency of this approach over more traditional schemes utilizing off-line fraction collection and conventional ionization methods can be explained by several factors. First, the superior sensitivity of ESI and MALDI means that less material is required for successful analysis. Second, on-line LC/MS optimizes the efficiency of sample transfer and saves both time and labor. Furthermore, the concentration dependence of ESI allows a majority of the material injected for LC/MS to be recovered for biological testing without compromising the signal available for molecular weight determination. As a validation of the above method, crude extracts containing two well-characterized antibiotics-teicoplanin and phenelfamycin-were examined. Results from these analyses are presented along with data from the analysis of a potent unknown antifungal sample.