Atmospheric Pressure Matrix-Assisted Laser Desorption/Ionization mass spectrometry imaging (AP-MALDI MSI) is a variant of the MALDI technique. The capacity of AP-MALDI MSI to work in an atmospheric environment eliminates the need for vacuum chambers, allowing for the preservation of native hydrated samples as well as the analysis of vacuum-incompatible compounds. This facilitates integration with other analytical techniques and increases sample preparation versatility. It has found applications in many fields such as biology, including spatial dermatology, where the analysis of complex biomolecules is essential. Here we introduce Pyxis, novel vendor neutral software for comprehensive AP-MALDI MSI data analysis, to investigate the spatial lipidome distribution and alteration within sunfilter-protected and -unprotected reconstructed human epidermis (RHE) sections, submitted to UV radiations.
Methods
RHE sections subjected to three test conditions: no UV stress and no sun filter (n=12), UV stress and no sun filter (n=12), and UV stress and sun filter (n=12), were washed, coated with HCCA matrix using the SunCollect MALDI Sprayer (SunChrom GmbH, Germany), and analysed by AP-MALDI MS in both positive and negative ion mode. Here, the compact AP-MALDI (ng) UHR system (MassTech Inc., Columbia, MD), was coupled to a high resolution Orbitrap Exploris 480 mass spectrometer (ThermoFisher, San Jose, CA). Imaging experiments were performed at spatial resolution of 5 µm per pixel, over a mass range of 205–2000 Da and at a mass resolution of 240,000@m/z 200. All data analysis and identification was performed using Pyxis (Mass Analytica, Spain).
Preliminary data
Applications of this unique analytical and data analysis workflow to RHE sections enabled lipids species accumulation within sections to be visualised and identified, providing insight into the metabolomics hallmarks of different sunfilters.
Leveraging the segmentation, based on clustering algorithms, and co-localization capabilities of the Pyxis software, we delineated regions of interest (ROIs) on the RHE sections to perform supervised statistical analyses, i.e. partial least squares discriminant analysis (PLS-DA). The comparison between RHE sections exposed and unexposed to UV light revealed distinctive changes in m/z values across specific cell layers. It allowed us to pinpoint inflammation biomarkers associated with UV exposure. Using these biomarkers, we directly visualised the efficacy of a test sunfilter, annotating the biomarkers, directly within Pyxis, with the integrated LIPID MAPS and HMDB databases. Annotations were ranked using a scoring system, enabling the precise identification of signal and spatial alterations in particular lipid classes, notably sterols. For instance, the oxysterol , 25-hydroxy-cholesterol 3-sulfate, recognized for its involvement in anti-inflammatory response, emerged as a key indicator. While this oxysterol is significantly expressed in UV-exposed and unprotected RHE sections, the opposite was true in the RHE sections not exposed to UV or protected by the sunscreen. Hence, the effectiveness of the sunfilter in mitigating this pathway was discerned by the observed modulation of this lipid. Our findings highlight the utility of the Pyxis platform, providing a user-friendly interface and a full exploitation of comprehensive AP-MALDI MSI data processing workflow, offering insights into metabolomic signatures relevant to dermatological research.