MSACL 2017 US Abstract

Exploring the Limits of DESI and MALDI MSI for Metabolite Identification

Andreas Dannhorn (Presenter)
Imperial College London

Bio: I obtained the license as pharmacist in January 2016. To obtain the license I studied for 4 years at Freie Universität Berlin, Germany followed by 1 year of mandatory internship placements. I spent the first 6 month of the practical year with the pharmaceutical company AstraZeneca in Mölndal, Sweden. During this placement I investigated the mechanism of inversion of the stereo-configuration of carboxylic acid containing compounds. The main methods used were human microsomal incubations followed by LC-MS based metabolite identification and LC-MS/MS based quantification. The second half of the practical year was a mandatory placement in a pharmacy. While studying I performed a 1 year internship at the German federal institute for risk assessment in Berlin. During this placement I worked on a LC-MS based metabolomics approach to identify biomarkers for exposure to polycyclic aromatic hydro

Authorship: Andreas Dannhorn (1), Maria Luisa Doria (1), Nicole Strittmatter (2), Anna Mroz (1), Renata Soares (1), John G. Swales (2), Richard J.A. Goodwin (2), Zoltan Takáts (1)
1) Department of Surgery and Cancer, Sir Alexander Fleming Building, Imperial College, London SW7 2AZ, UK 2) Drug Safety & Metabolism, AstraZeneca, Cambridge, CB4 0WG, UK

Short Abstract Mass spectrometry imaging (MSI) became a powerful tool in drug discovery applications. In the current study we explored the suitability of desorption electrospray ionization (DESI) and matrix-assisted laser desorption ionization (MALDI) MSI to map the spatial distribution of 4 cassette-dosed drugs and their metabolites in a rat brain, liver and kidney sections. Using these two different MSI techniques, we were able to map the spatial distribution of the 4 drugs and 7 out of 10 metabolites, identified by LC-MS analysis, together with endogenous compounds. Combined DESI full-scan ToF MS and MS/MS imaging experiments were performed to acquire full scan data of the tissue while confirming the identity and localization of the observed drugs and their metabolites.

Long Abstract

Mass spectrometry imaging (MSI) became a powerful tool in drug discovery applications. The possibility to map the spatial distribution of compounds while acquiring data for the tissue is crucial to fully understand effects observed in DMPK and toxicity studies. The observed effects can correlate with the compound tissue concentration obtained by LC-MS/MS analysis of tissue homogenates. When the compound accumulates is specific compartments of the tissue the overall tissue concentration might not correlate with the observed effects while the spatial distribution of the compound does.

In the current study we used matrix-assisted laser desorption ionization (MALDI) and desorption electrospray ionization (DESI) mass spectrometry imaging (MSI) to map the spatial distribution of 4 orally dosed drugs and their resulting metabolites in rat tissue sections and explore the boundaries of these MSI techniques.

Methods:

Rats were cassette-dosed with a combination of 4 low priority drugs (erlotinib, moxifloxacin, olanzapine and terfenadine). Brain, kidney and liver samples were collected from control animals, 2h and 6 h post-dose respectively. The snap-frozen tissues were cryosectioned to a thickness of 10 µm and thaw-mounted onto SuperFrost® Plus Glass slides (Thermo Fisher Scientific Inc, Waltham, MA, USA). To minimize the instrument dependent variability between different experiments for each organ tissue sections of the 3 time-points were mounted onto one slide and analyzed in the same imaging experiment.

MALDI MSI was performed on a MALDI Synapt G2 HDMS (Waters Corporation, Milford, MA) in positive ion mode using 2,5-Dihydroxybenzoic acid (DHB) as the MALDI matrix. Matrix application was performed using a TM sprayer (HTX Technologies, Chapel Hill, NC). All MALDI MSI experiments were performed with a pixel size of 100 µm.

DESI experiments were performed using a Xevo G2-XS Q-ToF (Waters Corporation, Milford, USA) operated in ToF-MS and ToF-MS/MS modes and a triple quadrupole Xevo TQ-S (Waters Corporation, Milford, USA) operated in multiple reaction monitoring (MRM) mode, both equipped with a 2D stage (Prosolia Inc., Indianapolis, USA) and the commercially available DESI sprayer from Waters. A mixture of 95:5 (v/v) methanol:water was used as solvent, delivered with a flow rate of 0.75 µL/min, nitrogen gas was used for nebulization at a pressure of 4 bar. Methods were initially developed on the Q-ToF and were then transferred and optimized to perform a fully targeted imaging approach using the TQ. All DESI MSI experiments were performed with a nominal pixel size of 50 µm.

LC-MS/MS experiments were performed on a Xevo TQ-S micro (Waters Corporation, Milford, MA) while a Xevo G2-XS Q-ToF was used for the identification of the metabolites. Both mass spectrometers were coupled to an Acquity UPLC micro (Waters Corporation, Milford, MA) equipped with an Acquity BEH C18 column. Water modified with 0.1% acetic acid was used as aqueous mobile phase and methanol modified with 0.1 % acetic acid as organic mobile phase. Methanolic extracts of a single tissue section were used to identify the metabolites present in the different organs and to quantify the drugs and semi-quantify their metabolites.

Results:

The MALDI imaging was performed in MS mode to fine map the distribution and their metabolites with a spatial resolution of 100 µm. To verify the identity of the observed drugs and metabolites while collecting information about the tissue itself the mass spectrometer was operated in a combined acquisition mode. DESI full scan MS mode was used to get data for the tissue while simultaneously several ToF MS/MS were acquired. Operating in combined acquisition mode allowed us to overlay the ion images for tissue marker (e.g. lipids) with the ion image for known MS/MS fragments of the observed compounds and to verify the compound/metabolite identity by comparing the on-tissue MS/MS spectrum to the MS/MS spectrum obtained from the drug standard. The resulting method was refined and developed into a fully targeted DESI MSI approach using a triple quadrupole (TQ) mass spectrometer operated in multiple reaction monitoring (MRM) mode.

Performing LC-MS analysis of the single tissue sections we could identify 10 metabolites of the given drugs within the different tissues. The identified metabolites were mainly phase 1 metabolites such as demetylated, hydroxylated or further oxidized varieties of the drugs and the acyl-glucoronide of moxifloxacin as a phase 2 metabolite. Applying the different MSI techniques we were able to determine the spatial distribution of the 4 drugs with a tissue concentration in the low nmol/g tissue range (>0.1 to 16.4 nmol/g). The tissue concentrations for the metabolites range from the low nmol/g (e.g. the carboxymetabolite of terfenadine) to the low pmol/g range (e.g. carboxyolanzapine and didesmethylerlotinib). We were able to map the distribution of 7 out of 10 metabolites missing the acyl-glucoronide and 2 low abundant metabolites.

The ion images for the drugs and their metabolites were found to correlate with the concentration of the drug present in tissue extracts, with the advantage that the spatial distribution information of the drug in the tissue was preserved. Olanzapine reaches the highest concentration in liver 6 h post dose giving the highest intensity in the imaging data. Fexofenadine, the carboxy-metabolite of terfenadine, has the highest concentration in kidney tissue 6 h post dose. The imaging experiments performed reveal that after 6 h the metabolite mainly accumulates in ducts while being eliminated, whereas at 2 h post dose it accumulates mainly in the medulla of the kidney.

Discussion:

The used MSI techniques showed a good coverage for the detection of the drugs and their metabolites. The images obtained from DESI performed on a Q-ToF and MALDI were comparable in terms of image quality and data output. Both techniques offer the possibility to collect information about the spatial distribution of drugs and their metabolites while collecting data for the tissue itself. MALDI lacks the possibility to simultaneously verify the identity of compounds performing MS/MS on the compounds of interest while acquiring full scan data. Usually a separate MS/MS experiment on an adjacent tissue section sample is necessary to further identify observed features.

While limiting the gained data to the raised question, in this case the distribution of the drugs and their metabolites within the tissues, the DESI TQ experiments showed a better sensitivity than the MS/MS experiments performed on the Q-ToF. Further improvement is necessary to raise both spatial distribution and image quality to the level gained with the Q-ToF. Using the Q-ToF setup, both the data acquisition of the mass spectrometer and the stage movement were synchronized through the MassLynx® (Waters Corporation, Milford, USA) software, resulting in good special resolution and high image quality. The stage movement of the TQ setup was operated through the Omni Spray 2D software (Prosolia Inc., Indianapolis, USA) and we assume that this desynchronization causes the lower image quality but the higher sensitivity of the TQ enables the possibility for faster data acquisition and the enhanced spectral reproducibility makes it a promising tool for the development of quantitative imaging approaches.

References & Acknowledgements:

Financial Disclosure

IP Royalty: no

Planning to mention or discuss specific products or technology of the company(ies) listed above:

no