Development of an on-line automated sample clean-up method and liquid chromatography–tandem mass spectrometry analysis: application in an in vitro proteolytic assay

Abstract

Fluorescence detection has been a method of choice in industry for screening assays, including identification of enzyme inhibitors, owing to its high-throughput capabilities, excellent reproducibility, and sensitivity. Occasionally, inhibitors are identified that challenge the fluorescence assay limit, necessitating the development of more sensitive detection methods to assess these compounds. For data mining purposes, however, original assay conditions may be required. A direct method transfer to highly sensitive and specific LC-MS-based methods has not always been possible due to the presence of MS-incompatible neutral detergents and non-volatile salts in the assay matrix. Utilizing an in vitro proteolytic screening assay for the serine protease hepatitis C virus (HCV) nonstructural (NS) 3 protease as a test case, we report the development of an automated sample clean-up procedure implemented on-line with liquid chromatography–tandem mass spectrometry (LC-MS/MS) analysis to complement fluorescence detection. Ion exchange and peptide microtraps were employed to remove MS-incompatible assay matrix components. Three protease inhibitors were used to validate the MS/MS method. Comparable potencies were achieved for these compounds when assessed by fluorescence and MS/MS detection. Furthermore, four-fold less enzyme could be utilized when employing the MS/MS method compared to fluorescence detection. The longer analysis time, however, resulted in reduced sample capacity. The potency of our designed HCV NS3 protease inhibitors are thus routinely evaluated using a continuous fluorescence-based assay. Only pertinent inhibitors approaching the fluorescence assay sensitivity limit are subsequently analyzed further by LC-MS/MS. This methodology allows us to maintain a database and to compare results independent of the detection method. Despite the relatively slow sample turnaround time of this LC-MS approach, the versatility of the automated on-line clean-up procedure and sample analysis can be applied to assays containing reagents which were historically considered to be MS incompatible.