Combining supervised and unsupervised analyses to quantify behavioral phenotypes and validate therapeutic efficacy in a triple transgenic mouse model of Alzheimer's disease

Abstract

Behavioral testing is an essential tool for evaluating cognitive function and dysfunction in preclinical research models. This is of special importance in the study of neurological disorders such as Alzheimer's disease. However, the reproducibility of classic behavioral assays is frequently compromised by interstudy variation, leading to ambiguous conclusions about the behavioral markers characterizing the disease. Here, we identify age- and genotype-driven differences between 3xTg-AD and non-transgenic control mice using a low-cost, highly customizable behavioral assay that requires little human intervention. Through behavioral phenotyping combining both supervised and unsupervised behavioral classification methods, we are able to validate the preventative effects of the immunosuppressant cyclosporine A in a rodent model of Alzheimer's disease, as well as the partially ameliorating effects of candidate drugs nebivolol and cabozantinib.

Figure 1.

Gait analysis of 3xTg-AD and WT mice. A) Example of WT (top) and 3xTg-AD (bottom) paw prints automatically recorded and categorized with the CatWalk XT system. B) Number of classified steps. C) Percentage of time spent in a normal step sequence pattern (NSSP). D) Manual print length, E) intermediate toe spread, F) mean intensity, and G) paw angle body axis for all four paws. The results are expressed as mean ± SEM. Statistics: two-way ANOVA followed by post hoc Tukey HSD test, * p < 0.05, ** p < 0.01, *** p < 0.001. N = 8 WT mice, N = 9 3xTg-AD mice.

Figure 2.

Overview of 8-cage imaging system and analysis pipeline. A) Mice are individually placed in a cage with access to food, water, and shelter in the form of a red hut. B) They are recorded for 22 hours, during which they are shown stimuli three times (hours 13, 15, and 17). C) The videos are analyzed using Fiji, DeepLabCut, and Keypoint MoSeq and D) A total of 125 behaviors are extracted. E) Significantly different behaviors in 3xTg-AD mice compared to age-matched WT mice. Green indicates an increase in 3xTg-AD mice compared to WT mice, while purple indicates a decrease. Statistics: Two-sample Wilcoxon test, * p < 0.05, ** p < 0.01. N = 8 WT mice, N = 16–18 3xTg-AD mice.

Figure 3.

Evaluating the effects of chronic treatment with cyclosporine, nebivolol, or cabozantinib in 3xTg-AD mice. A) Experimental timeline of treatment and behavioral phenotyping in drug-treated 3xTg-AD mice. B) Percent survival of cohorts across experimental time points. The black arrow indicates a change in the administered dose of nebivolol due to decreased survivability. C) Weight measurements across experimental timepoints. Error bars represent SEM. Statistics: two-way ANOVA followed by post hoc Tukey HSD test, * p < 0.05, ** p < 0.01, *** p < 0.001. N = 8 WT mice, N = 16–18 3xTg-AD mice, N = 8–10 3xTg-AD mice treated with cyclosporine, N = 8–10 3xTg-AD mice treated with nebivolol, N = 8–10 3xTg-AD mice treated with cabozantinib.

Figure 4.

Cluster analysis. A) Principal component analysis (PCA) and B) hierarchical cluster analysis of the 125-parameter behavioral profiles. WT_Y: young wild-type mice; WT_O: old wild-type mice; AD_Y: young 3xTg-AD mice; AD_O: old 3xTg-AD mice; AD_CSA_Y: young cyclosporine-treated 3xTg-AD mice; AD_CSA_O: old cyclosporine-treated 3xTg-AD mice; AD_NEB_Y: young nebivolol-treated 3xTg-AD mice; AD_NEB_O: old nebivolol-treated 3xTg-AD mice; AD_XL_Y: young cabozantinib-treated 3xTg-AD mice; AD_XL_O: old cabozantinib-treated 3xTg-AD mice.