Assessment Of Solid Phase Microextraction As A Sample Preparation Tool For Metabolomics Analysis Of Brain Tissue By Liquid Chromatography-Mass Spectrometry

This work presents an evaluation of solid-phase microextraction (SPME) SPME in combination with liquid chromatography-high resolution mass spectrometry (LC-HRMS) as an analytical approach for untargeted brain analysis. The study included a characterization of the metabolite coverage provided by C18, mixed-mode (MM, with benzene sulfonic acid and C18 functionalities), and hydrophilic lipophilic balanced (HLB) particles as sorbents in SPME coatings after extraction from cow brain homogenate at static conditions. The effects of desorption solvent, extraction time, and chromatographic modes on the metabolite features detected were investigated. Method precision and absolute matrix effects were also assessed. Among the main findings of this work, it was observed that all three tested coating chemistries were able to provide comparable brain tissue information. HLB provided higher responses for polar metabolites; however, as these fibers were prepared in-house, higher inter-fiber relative standard deviations were also observed. C18 and HLB coatings offered similar responses with respect to lipid-related features, whereas MM and C18 provided the best results in terms of method precision. Our results also showed that the use of methanol is essential for effective desorption of non-polar metabolites. Using a reversed-phase chromatographic method, an average of 800 and 1200 brain metabolite features detected in positive and negative modes, respectively, met inter-fibre RSD values below 30% (n=4) after removal of fibre and solvent artefacts from the associated datasets. For features detected using a lipidomics method, a total of 900 and 1800 features detected using C18 fibers in positive and negative mode, respectively, met the same criteria. In terms of absolute matrix effects, the majority of the model metabolites tested showed values between 80 and 120%, which are within the acceptable range. Overall, the findings of this work lay the foundation for further optimization of parameters for SPME-LC-HRMS methods suitable for in vivo and ex vivo brain (and other tissue) untargeted studies, and support the applicability of this approach for non-destructive tissue metabolomics