Continuous OFT-Based Behavioral Phenotyping Reveals Neural Stem Cell-Associated Shifts in Depression-Like Mice

Background: Depression-like phenotypes in animal models often show graded behavioral variation rather than uniform group-wise changes, yet treatment effects are still commonly evaluated using categorical comparisons. In particular, it remains difficult to determine whether treated animals retain model-like behavioral features or shift toward a more control-like state. Here, we developed a continuous behavioral phenotyping framework based on multidimensional open-field-test (OFT) features to quantify model-like behavioral states and treatment-associated behavioral shifts. Methods: In the present dataset, untreated CUMS- and LPS-induced mice jointly defined a shared model-like behavioral state characterized by reduced locomotor activity and altered spatial exploration. By integrating principal component analysis with k-nearest-neighbor-based behavioral inference, we derived two complementary indices: Δd, representing the relative distance of an animal from the model-like versus control-like behavioral centroids, and p_depression, a kNN-derived score representing similarity to the untreated model-like reference state. These indices were used to position individual animals along a spectrum between control-like and model-like behavioral states. Results: Higher p_depression and lower Δd indicated OFT phenotypes more similar to untreated model mice. Using this framework, NSC-treated animals showed treatment-associated shifts away from model-like behavioral patterns and toward more control-like states, while preserving substantial inter-individual variability. Behavioral scores also showed group-level correspondence with neuroinflammatory markers, suggesting preliminary biological relevance. Because the primary model-specific biological comparisons were based on a limited subgroup size (n = 4 per group), the observed NSC-associated treatment effects should be regarded as exploratory and require confirmation in larger, prospectively powered cohorts. Conclusions: Together, these findings support continuous OFT-based behavioral phenotyping as a promising exploratory approach for evaluating treatment-associated behavioral change and phenotypic heterogeneity in preclinical depression research.