Mass spectrometry imaging (MSI) uses spatially resolved proteomics and metabolomics mass spectrometry methods to simultaneously record the distributions of hundreds of endogenous molecules directly from tissue samples. An essential component of biomarker discovery experiments is the need for histological specification because of the variety of cell types present in diseases such as cancer, or the distinct histoarchitecture in tissues such as brain that reflect distinct organs. The ability to apply histological stains to tissues after the MSI experiment, without significant loss of histological information, has greatly facilitated the alignment of MSI and histological images and thus the integration of MSI with histology.
The combination of mass spectrometry imaging and histology has proven a powerful approach for obtaining molecular signatures from specific cells/tissues of interest, whether to identify biomolecular changes associated with specific histopathological entities or to determine the amount of a drug in specific organs/compartments. Lack spatial correspondences between high resolution histology and low resolution MSI-data hampers their co-registration process to be automatically performed. Currently there is no software that is able to explicitly register mass spectrometry imaging data spanning different ionization techniques or mass analyzers. Accordingly, the full capabilities of mass spectrometry imaging are at present underexploited. We present a fully automated generic approach for registering mass spectrometry imaging data to histology and demonstrate its capabilities for multiple mass analyzers, multiple ionization sources, and multiple tissue types (
Abdelmoula et al., 2014). The Figure on the left shows an example of integrating MALDI-IMS data and histology of a thyroid cancer tissue. The top panel shows a preprocessed histological image; the middle panel shows a low dimensionality representation of the high-dimensional MALDI MSI data using tSNE (which is used as the moving image in the registration process); the bottom panel shows the fusion resultــــoverlay of the processed histological image and registered tSNE results.
Imaging mass spectrometry (IMS) holds great potential for uncovering biomolecules implicated in neurological disorders such as migraine. Currently the manual nature of the data analysis limits its full exploitation (it is a time and labor intensive process, requiring extensive expertise to correctly annotate the complex anatomy of brain tissues). We proposed an automatic pipeline to align the histological images of mice brain tissue and its associated IMS data to the Allen brain atlas (ABA) (
Abdelmoula et al., 2014a). This combined workflow makes it possible to: i) automatically annotate the tissue’s anatomy, ii) correlate the results with the anatomical structures, and iii) facile comparison of animal cohorts. The figure to the left summarizes the proposed pipeline. The figure shows an extensive example of the advantages of the proposed pipeline: 1) ABA-based Registration in which both MS data and its associated histology were aligned to their best reference section from the ABA. 2) ABA-based Segmentation in which the histo-anatomical structures can be automatically annotated using the segmentation maps provided in the ABA. 3) Correlation between the distinct anatomical regions and MS-data can be established that allows to automatically localizing region specific mass spectrometry signals.
Our recent developed pipelines (
Abdelmoula et al., 2014; Abdelmoula et al., 2014a) represent promising developments for the MSI community; developments that save time, effort and labor and assure high accuracy away from the observer variability issues. The developed pipelines have been recently used in two pre-clinical applications: 1) Investigation of Cortical Spreading Depression in a Mouse Model of Migraine ( Carreira et al., In press), 2) Precise anatomical localization of accumulated lipids in Mfp2 deficient murine brains ( Škrášková et el., In press). In addition, the MSI data could also be analyzed in the context of the gene expression data contained in the ABA, opening the doors for promising perspectives for the future of biology research. The figure shows a schematic of the workflow developed to analyze the effect of CSD in WT and R192Q mouse brains. Ninety six coronal brain sections were obtained from a total of 32 mouse brains (3 consecutive sections per animal). Proteins, peptides and metabolites were independently analyzed by MSI using optimized sample treatment for each molecular class as described in the Materials and Methods section. Each section was stained with Nissl reagent after matrix removal and the MSI datasets and histological images were aligned to the Allen Brain Atlas of the mouse brain ( Abdelmoula et al., 2014a). Automatic anatomical annotation of regions of interest allowed the extraction of MSI data from specific brain regions of interest and statistical analysis.