Quantitative molecular imaging of metabolites and neurotransmitter in brain sections using DESI MSI with internal stable isotope labeling.
Quantitative spatial distributions of small molecules, such as neurotransmitters, are vital for obtaining clinical insight into neurological diseases. Mass spectrometry imaging (MSI) using Desorption Electrospray Ionization (DESI) is an analytical approach for mapping the spatial distribution of endogenous molecules. DESI ionization often provides complementary information to other established MSI techniques such as matrix-assisted laser desorption ionization (MALDI). Here we showcase the utility of DESI imaging to obtain a relative quantitative profile of hydrophilic neurometabolites using stable isotope labels as internal standards, allowing for precision-driven quantitative measurements for compounds typically inaccessible in MALDI assays. The presented results are for research use only and not for use in diagnostic procedures.
Brain tissues were flash-frozen in liquid nitrogen before cryosectioning. The coronal sections were mounted on glass slides and stored in -80 ˚C until analysis. MSI was performed on a DESI system integrated with a quadrupole time of flight mass spectrometer with ion mobility separation (SYNAPT HDMS G2-Si; Waters Corporation, Milford, MA). DESI-MSI data were collected, processed, and analyzed using High Definition Imaging (HDI) 1.4 software with MassLynx 4.1 data acquisition control (Waters ). For quantitative profiling, stable-isotope labeled compounds (Cambridge Isotope Laboratories, Tewksbury, MA) were added to DESI sprayer solvent as internal standards.
The DESI source and mass spectrometer were optimized to perform untargeted molecular profiling of metabolites. Tissue preparation, such as removing phospholipid species with a series of alcohol/chloroform washes, was avoided to reduce molecular delocalization. Small molecules including choline, amino acids (e.g. taurine, glutamine), free fatty acids (e.g. arachidonic acid) and neurotransmitters (e.g. GABA, serotonin) were directly mapped from tissue in addition to the abundant phospholipids and sphingolipids. Contextual anatomical information of small molecules was obtained by overlaying ion images with a histological image or ubiquitous phospholipid species. In-line DESI lock mass was used to correct any mass drift in order to obtain a high mass accuracy (low ppm) for molecular assignment. Using a similar approach, multiple stable isotope labeled molecules were added as internal standards. The concentration of each compound was between 50 to 200 pg/μL, and was dependent on the ionization efficiency of the molecule. Relative quantitative imaging of the ion of interest was obtained by normalization of the data within the HDI software. During the normalization process, the intensity value of all the detected ions within the data set was divided by the intensity value of the internal standard on a pixel-by-pixel basis. Ion intensity ratios were calculated in regions of interest (ROIs) by exporting mass spectrum to MassLynx software. Molecular identification was obtained primarily based on accurate mass search against publically curated databases (e.g., METLIN, HMDB). In some cases, high-fidelity isotopic distribution, collisional cross sections (CCS), and/or MS/MS fragmentation was used to improve the identification confidence of the detected molecules. The presented preliminary work aims to show a robust approach for relative quantitative imaging of small molecules neurometabolites using DESI-MSI.