Development of a MATLAB-based Toolkit to Quantify Neurodegenerative Histopathology Markers Applied to the Evaluation of a CRISPR-based Therapeutic for Alzheimer’s Disease
Nihal Satyadev1, Brent Aulston2, Kristen Branes2, Nidhi Checka2, Subhojit Roy2
1University of Medicine and Health Sciences, 2UCSD
Objective:
The objective of this study was to develop a comprehensive, easy-to-use set of MATLAB-based tools to quantify Alzheimer’s disease (AD) associated histopathological markers in immunolabeled brain tissue.  We then utilized this MATLAB-based toolkit to assess the efficacy of a novel CRISPR-based, amyloid precursor protein (APP)-editing approach in a mouse model of AD.
Background:

AD is characterized immunohistologically by the presence of extracellular amyloid plaque deposits and associated neuroinflammation throughout the cerebral cortex and hippocampus. Amyloid plaques are aggregates of amyloid beta (Aβ) peptides that are released by successive cleavages of full length APP. Our lab has developed a novel CRISPR-based, APP-editing approach that effectively deletes the YENPTY endocytic domain at the C-terminus of APP, thus attenuating pathologic beta-cleavage and up-regulating neuroprotective alpha-cleavage. In order to assess the efficacy of our approach in vivo, we used CRISPRs to generate germ-line and somatic edits in an APP knock in (APP NL/G/F) mouse model of AD. 

Design/Methods:

Immunohistochemistry (IHC): Tissues from appropriately aged control and CRISPR-edited APP NL/G/F mice were immunostained for Aβ plaques (6E10 antibodies), astrocytes (anti-GFAP antibodies) and microglia (anti-Iba1 antibodies).

 

MATLAB analyses: After images were optimally contrasted, a color-channel filter was applied to isolate cells and plaques from the background. Our MATLAB program was then able to segment individual cells/plaques and measure their area, distance to nearby cells/plaques, and percentage of overlap. Activated microglia were defined as those with an area of 25 μm2 or greater.

Results:
MATLAB quantification of IHC-stained sections revealed a robust decrease in Aβ deposition and reactive astrogliosis in CRISPR-treated APP NL/G/F tissues. These results matched supporting qualitative assessments and biochemical analyses.
Conclusions:

Our MATLAB toolkit can reliably quantify common neuropathological IHC markers and may be useful for evaluating the efficacy of neurotherapeutics.

10.1212/WNL.0000000000204039