(n=1-3) for Incorporation into a Searchable Database

Ernst Pittenauer¹; Martin Zehl¹; Robert Mistrik² and Günter Allmaier¹

¹Vienna University of Technology, Vienna, Austria
²HighChem, Ltd., Bratislava, Slovakia

Introduction

Searchable mass spectrometric databases (e.g. Wiley, NIST, Maurer-Pfleger) are well established tools for compound identification/verification with EI-MS. The advantage of these databases is that EI-spectra are usually instrument-independent and are rich in fragment ions. In contrast to EI, ESI/MALDI-full scan mass spectra do not exhibit abundant fragmentation. Therefore, CID MSⁿ experiments (n=2,3) are necessary to induce abundant diagnostic fragmentation from the various forms of precursor ions. As different types of analyzers favor different fragmentation pathways for the same compound due to different kinetic energies of the precursors (eV vs. keV), different collision energy regimes, few or multiple collisions and unimolecular or consecutive decays, data from different types of analyzers cannot be easily incorporated in one database.

Methods

All ESI-MSⁿ-data were measured on 3D-ion traps (Finnigan LCQ Deca XP and Bruker Esquire 3000plus). Roughly 600 compounds are incorporated into the database including various classes of antibiotics, steroidal hormones, peptide hormones, human/veterinary drugs, illegal drugs (e.g. THC, opioids) and pesticides. ESI-data for all compound are acquired in the positive- (precursors: [M+nH]n+, n=1-3; [M+NH4]+; [M+Na]+; [M+K]+) as well as in the negative-ion mode (precursors: [M-nH]n-, n=1-3) whenever possible under comparable CID conditions [standard (4 u ) and extended isolation width (6 u); precursor ion reduction by CID to 5-20 % of the 100 % product ion; standard (+/- 5 u) and broadband activation width (+/- 20 u)]. Selected MALDI-MSⁿ-data were acquired on a 3D-ion trap/reflectron TOF instrument (Kratos Axima-QIT).

Figure 1. Management of database entries using Mass Frontier™ software

Abstract

Especially with LC or CE/ESI-MS, ion traps offer the possibility of acquisition of full scan MSn spectra (n=2,3) on a chromatographic or capillary electrophoretic time scale at high sensitivity compared to most other types of analyzers. Therefore, an ESI-MSⁿ-3D-ion trap-based library will provide the possibility of searching CID-spectra of unknowns from real-life samples for a match with incorporated reference spectra. The above discussed “standardized” acquisition conditions for ESI-MSⁿ-spectra were found to enable extremely good long-term reproducibility with the same instrument as well as intercomparison with different ESI instruments from various manufactors, which are both of fundamental value when creating a searchable database. With respect to the chemistry of the compounds investigated, multiple precursors were selected (protonated, cationized, deprotonated as well as multiply charged ions) for measurement to provide compound searches with high flexibility. Special emphasis is also given to elucidate major fragmentation pathways based on multiple precursor selection of the most abundant first generation product ions. Recently, the combination of MALDI with ion trap hybrid technology became commercially available. In order to enlarge the scope of the ESI-MSⁿ-database, MALDI-MSⁿ-data of some selected pharmaceutical compounds (Vasopressin, Desmopressin, Oxytocin, Val-Gramicidine A, Moenomycin A, Tylosin A and Salinomycin) are compared and evaluated regarding the possibility of incorporation of MALDI-data into the database. First results indicate that MALDI-MSⁿ data are in good agreement with ESI-MSn-data allowing in principle incorporation into a searchable database. Major differences between the ESI ion traps to MALDI ion trap/reflectron TOF (Kratos Axima-QIT) are the limited number of possible precursors (only singly charged) for the latter instrument related to the MALDI sample preparation technique and different m/z-dependent trapping characteristics. New features of this database compared to existing ESI-libraries are MSn-data up to MS3 of all abundant first generation products, a wide variety of precursors as well as negative-ion MSⁿ-data.

Figure 2. Real-life human plasma sample of a victim of Colchicine intoxication, (A) TIC-trace of LC/ESI-MS2 of m/z 400.2 (amount: 70 ng/ml) , (B) MS2-spectrum of the peak corresponding to Colchicine eluting at Rt = 10.0 min [all data: Finnigan LCQ Classic], and (C) reference MS2-spectrum of Colchicine incorporated into the database [Finnigan LCQ XP].