An open-source pipeline for analysing changes in microglial morphology.

Devin Clarke,Catherine N Hall,Hans S Crombag

doi:10.1098/rsob.210045

Abstract

Changes in microglial morphology are powerful indicators of the inflammatory state of the brain. Here, we provide an open-source microglia morphology analysis pipeline that first cleans and registers images of microglia, before extracting 62 parameters describing microglial morphology. It then compares control and 'inflammation' training data and uses dimensionality reduction to generate a single metric of morphological change (an 'inflammation index'). This index can then be calculated for test data to assess inflammation, as we demonstrate by investigating the effect of short-term high-fat diet consumption in heterozygous Cx3CR1-GFP mice, finding no significant effects of diet. Our pipeline represents the first open-source microglia morphology pipeline combining semi-automated image processing and dimensionality reduction. It uses free software (ImageJ and R) and can be applied to a wide variety of experimental paradigms. We anticipate it will enable others to more easily take advantage of the powerful insights microglial morphology analysis provides.

Highlights

Microglia, the brain’s resident immune cells, are involved in phagocytosis and regulation of the adaptive immune response [1,2]
We used our R package to combine the features best at discriminating between LPS conditions into a single ‘inflammation index’. We demonstrate that this inflammation index provides a stable measurement of cellular morphology that persists over an imaging session, we use it to assess the impact of high-fat diet (HFD) feeding on microglial morphology in C-X3-C motif chemokine receptor 1-green fluorescent protein (CX3CR1-GFP+/−) mice, and we further validate our technique using in vivo images of two-pore domain K+ channel THIK-1 (TWIK-related halothane-inhibited K+ channel) knockout microglia [10]
We developed an open-source image analysis pipeline that facilitates the largely automated processing of large imaging datasets and the creation of a single metric of morphological change using dimensionality reduction to allow users to leverage in vivo microglial morphological analyses

Summary

Introduction

The brain’s resident immune cells, are involved in phagocytosis and regulation of the adaptive immune response [1,2] While resting they have a small soma and long, dynamic processes, allowing them to survey their local environment [3] for pathogens or damage-associated molecular patterns (DAMPs) [4]. Such stimuli activate a morphological shift towards an amoeboid-like shape which facilitates migration to sites of injury and phagocytosis [3,5]. Analysing morphological changes is often done manually This can lead to rater error, demands a large time investment, and limits the number of cells that can be analysed. With complex studies demanding large sample sizes to detect small effects, a more efficient approach is necessary

Methods

Results

Conclusion