In this project, I installed and adapted a new prototype of a SPICI postionisation module and combined it with surface-assisted LDI-MSI. The SPICI module was previously developed in cooperation with the University of Münster, Germany for the new version of the ion source installed in our MS system - the Spectroglyph Dual MALDI/ESI Injector. In the work group, we designed a new class E amplifier for the PI source and then, I optimized the system for LDI-PI-MSI – first, with classic organic matrices, and thereafter, with gold nanolayers, which were deposited on thin tissue sections by gas phase deposition. As a sample system for the optimisation process, I prepared a series of homogeneous tissue samples from cow brain that served as a standard and quality control throughout the whole project.
LDI-SPICI-MSI proved to be successful, after identification of critical parameters, e.g. in-source pressure, dopant concentration, and delay between the ablation laser and VUV pulse initiating the postionisation reaction. I could achieve major postionisation effects on glycerophopholipids and other lipids by adding acetone as a dopant, similar to the use of organic matrix.
For the next objective, I investigated the use of different dopants and different nanostructured layers in order to enhance less polar groups of analytes with the PI step. However, the ion funnel could not operate stably with dopants initiating electron transfer reactions, and even though postionisation effects were visible, I had to conclude that SPICI is limited to the use of proton transfer agents such as acetone and isopropanol in this ion funnel ion source. I next investigated different sputtering times, i.e. different thicknesses of deposited Au nanostructures and identified that the sputtering times used for LDI-MSI without PI yielded the strongest PI signals as well, confirming this sample preparation step as universal between both measurement modes. The use of other nano-layers did not yield better PI signals than Au.
One major observation in this work package was that – unlike the common glycerophospholipids – smaller metabolites <500 Da could successfully be postionized without the use of a chemical dopant in the gas phase. This also majorly reduced the intensity of background ion signals, allowing me to measure spatial metabolomics in a mass range previously not accessible. With this augmentation of analytical depth, I could successfully finish the second objective of MASS2, demonstrated on the analysis of animal tissue sections with high morphological distribution of numerous chemical compounds.
The acquired data was analyzed with the dedicated in-house software rMSI that allowed for identification of key parameters to detected the greatest analytical depth based on statistical means. Also, the reproducibility and reliability of the measurement strategy was monitored in this fashion. Therefore, I wrote a universal templated for a markdown in R language that included several analytical steps and facilitated the data analysis and the overview of technical advance. The detected signals were tentatively annotated based on data bases and with acquired MS/MS spectra, and the detected analyte classes listed in tables, indicating the measurement parameters for a successful detection, e.g. the used matrix, polarity, and dopant.
As the last step, a large set of UBC tissue was measured with Au-nanostructured LDI-MSI and yielded a high level of data quality for the entire set. The data analysis assessed the intra-tumor heterogeneity with unsupervised image segmentation strategies, and identified new prognostic and diagnostic biomarkers. For the second sample set of ZFE, we identified an optimal tissue embedding media allowing for an assay of up to 100 sections on a single MSI slide and compared the analytical depths of different MS systems (Orbitrap Exploris and timsTOF flex) using different matrices (DAN,NOR,DHB) and Au-nano structures in both ion modes. Furthermore, we conducted postionisation with LDI-SPICI-MSI combined with Au-nanostructures for spatial metabolomics in the smallest mass range <300Da complementary to the commercial MALDI-2 technique.
