MLIBRA

Motivation

‍Brain atlases are an invaluable resource for the interpretation of structure-function relationships, provide a reference to the study of molecular heterogeneity among various brain cells, and a framework to formulate hypotheses about brain functioning and development. Although there exists today a wealth of knowledge regarding the different neuroanatomical regions, their precise chemical composition remains relatively unclear. This is unfortunate as brain operation largely relies on the electrochemical activity of neural cells. Lipids are a major component of the mammalian brain as they constitute half of the brain’s dry weight and are key components of every cell. Lipids are a chemically diverse class of molecules that participate in central metabolism, constitute the fabric of biomembranes and act as first and second messengers in signal transduction events. Different neural cell types produce different lipids, which impact many biological processes, including budding and fusion of synaptic vesicles, propagation of action potential, and insulation of axons as components of the myelin sheets. However, a systematic and fine-grained characterization of lipid metabolism in the different brain regions is currently unavailable. In this project, we propose compiling the first whole mammalian organ’s 3D metabolic atlas: the mouse Lipid Brain Atlas.  We aim at identifying homogenous regions of lipid expression across the mouse brain. The further goal is to study these regions and get insights about differences in lipid distributions within different sexes, during  pregnancy, and differences defined via gene expression change.

Proposed Approach / Solution

‍We developed an automatic hierarchical segmentation of brain slice MALDI readings. This  approach uses iterative bipartitioning and a generative probabilistic model. We then aim at identifying differences in brains of different types and relate these regions to gene expression and chemical pathways.

Impact

‍All research results will be collected in an interactive web atlas, in which researchers can explore the data and compare with their readings in their own experiments. Given the importance of mice brain for biomedicine, this could open venues for new lines of research or make relations between lipid distribution and other biological processes.