Development of the first metabolite-based LC-MS(n) urine drug screening procedure-exemplified for antidepressants.

ABSTRACT

In contrast to GC-MS libraries, currently available LC-MS libraries for toxicological detection contain besides parent drugs only some main metabolites limiting their applicability for urine screening. Therefore, a metabolite-based LC-MS(n) screening procedure was developed and exemplified for antidepressants. The library was built up with MS(2) and MS(3) wideband spectra using an LXQ linear ion trap with electrospray ionization in the positive mode and full-scan information-dependent acquisition. Pure substance spectra were recorded in methanolic solution and metabolite spectra in urine from rats after administration of the corresponding drugs. After identification, the metabolite spectra were added to the library. Various drugs and metabolites could be sufficiently separated. Recovery, process efficiency, matrix effects, and limits of detection for selected drugs were determined using protein precipitation. Automatic data evaluation was performed using ToxID and SmileMS software. The library consists of over 700 parent compounds including 45 antidepressants, over 1,600 metabolites, and artifacts. Protein precipitation led to sufficient results for sample preparation. ToxID and SmileMS were both suitable for target screening with some pros and cons. In our study, only SmileMS was suitable for untargeted screening being not limited to precursor selection. The LC-MS(n) method was suitable for urine screening as exemplified for antidepressants. It also allowed detecting unknown compounds based on known fragment structures. As ion suppression can never be excluded, it is advantageous to have several targets per drug. Furthermore, the detection of metabolites confirms the body passage. The presented LC-MS(n) method complements established GC-MS or LC-MS procedures in the authors' lab.