Development of image analysis tools to characterize and classify cultured microglial cells in healthy and pathological conditions

Abstract

Microglia are the immune cells of the brain and they constantly survey the environment in order to prevent brain infection and phagocytose death cells and cell debris. In physiological conditions, microglia get rid of dysfunctional synapses (pruning) and neuronal molecules bind to microglia receptors to maintain the latter in a harmless state. However, after a brain damage, the balance can be disrupted. Under pathological situations such as neurodegeneration or stroke, damage-associated molecular pattern molecules (DAMPs) are released by injured neurons or astrocytes and activate microglia. The microglia response can cause inflammation and exacerbate neuronal death, but it is also involved in tissue regeneration . In the same brain area, surveillant and activated microglia may coexist. Microglia are highly plastic cells and adopt changes in their morphology in order to quickly reach the damaged area, and to phagocytose apoptotic cells or debris. Recent studies identified a sub-population of ¿neuroprotective¿ microglia among activated microglia in animal models of neurodegenerative diseases based on gene expression analysis suggesting heterogeneity within the microglia population. The goal of the present work is to characterize different subsets of microglia that presumably assume different functions and may coexist after neuronal injury . This objective is quite challenging, particularly in living cells, due to the morphological heterogeneity of microglia, shape transition and the presence of overcrossing cells. Methods We performed live cell imaging in cortical neuron/glia cultures prepared from embryos, resulting from the crossing between ROSA-CAG-tdTomato flox/flox mice and CX3CR1-CRE-ERT2 mice, in control conditions and after an excitotoxic challenge. In these cultures, microglia express tdTomato after exposure to 4-OH-tamoxifen. Time-lapse images were acquired in a Laser Scanning Confocal microscope. We developed plugins for ImageJ to provide an automated measurement of morphological features. Results We developed plugins for ImageJ to segment microglia and track individual cells in a semi-automated fashion. The plugin generates a mask which can be opened in ImageJ and used for morphometric measurements. The data generated were used for the classification of microglia in sub-groups. The novelty of our software is that it allows following-up every single microglia and its phenotype transition along time, and it assigns each cell into a cellular subtype according to the measured parameters. Conclusions Segmentation and dynamic morphometric measurements of cultured microglia in an in vivo time-lapse setting allow clustering the microglial cells into different subclasses.