Microglial Activation in an α-Synuclein Mouse Model of Parkinson's Disease

Last Updated Date: August 06, 2024

Authors: Laurent Potvin-Trottier, Ph.D., Robin Guay-Lord, Lionel Breuiland, Ph.D., Simone P. Zehntner, Ph.D., Elodie Brison, Ph.D., Jim Paskavitz, M.D., Barry J. Bedell, M.D., Ph.D.

Key Takeaways

• Microglia can be automatically classified as non-activated or activated based on their morphological characteristics.
• An "activation score" for each cell can be defined on a continuous scale using the sum of its morphological measures.
• In an α-synuclein preformed fibril (PFF) seeding mouse model, we have shown highly significant increases in activated microglia in regions with spread of phosphorylated α-synuclein aggregates.
• The phenotypic state of the microglia can be more finely quantified using (a) the distribution of activation scores, which can disentangle effects from the increase in cell density and the changes in their morphology, and (b) the distribution of activated microglia across different subclasses with distinct morphologies.
• Our novel approach may provide sensitive measures for the preclinical testing of the efficacy of disease-modifying therapeutics in animal models of Parkinson's disease.

Microglia are thought to play a key role in neurological diseases, such as Parkinson's disease (PD), Alzheimer's disease (AD), amyotrophic lateral sclerosis (ALS), frontotemporal dementia (FTD), and multiple sclerosis (MS). In these diseases, microglial activation is often linked with pathologic progression and disease severity.

Resting microglia, with their highly ramified processes, have the ability to shift into different functional states, with modification of their morphology (e.g. shorter processes), proliferation, phagocytic activity, antigen presentation, and release of cytokines and chemokines. While the number or density of microglia is typically measured on immunohistochemistry (IHC) or immunofluorescence (IF) sections, assessment of the morphological characteristics can provide additional information about the microglial phenotype.

We have developed a novel, fully-automated method for analysis of microglial morphology in neuroanatomical regions-of-interest on IHC sections. Our method leverages advanced computer vision and machine-learning algorithms. We have utilized this new technique for assessment of microglial activation in an α-synuclein preformed fibril (PFF) seeding & spreading mouse model of Parkinson's disease that is routinely used for preclinical assessment of putative disease-modifying therapeutic agents. 

In the PBS-injected control mice, the density of microglia was found to be consistent across regions-of-interest (ROIs) and have low variability, thereby demonstrating the precision of our approach. In the PFF-injected mouse brains, the density of microglia increased in most ROIs analyzed. This increase with higher in regions affected early in this α-synuclein PFF spreading model, such as the ipsilateral piriform and entorhinal cortices. We found that the "activated microglia" measure showed a clearer distinction between PFF and PBS groups of mice compared to the simple Iba-1 stain density measure and allowed for greater detection of statistically significant differences between groups.

Microglial activation is a complex process that is an active area of research. Reports have suggested both beneficial and deleterious microglial activation in neurological diseases. Attempts have been made to separate activated microglia into subclasses with different characteristics. To further provide insights into the microglial activation process, we subdivided the activated cells into four categories with distinct properties using unsupervised clustering based on morphological features. Identifying changes in the distribution across the subclasses, in conjunction with shifts in the activation scores, could provide unique insights into disease progress or therapeutics effects that may not be readily apparent by simple cell density measures. 

Presentation Highlights