Title : MAPSNY.com: A platform for 3D morphological profiling and morphotyping of microglia in neuroinflammation
Abstract:
Background: Neuroinflammation underlies diverse neurological and psychiatric disorders, yet current tools lack the spatial resolution to characterize region-specific microglial activation states in three dimensions. We present MAPSNY.com, a computational platform for automated 3D morphological profiling of microglia from confocal z-stack images, extracting up to 28 threedimensional metrics per cell and classifying four morphotypes (homeostatic, hypertrophic, amoeboid, rod).
The platform was applied to three independent cohorts using an identical pipeline: (1) Primary LPS cohort - systemic LPS model (1 mg/kg i.p. × 7 days) in female Sprague-Dawley rats (n=4/group); 280 individually characterized microglia across prefrontal cortex (PFC) and basolateral amygdala (BLA); (2) PCOS external validation - prenatal androgenization model (PNA-PBA), BLA, n=6/group, 336 cells; (3) Prenatal methadone external validation - hippocampus, n=9–12/group, 230 cells.
Results: BLA-selective morphological remodeling under systemic LPS. Systemic LPS induced significant morphological remodeling exclusively in BLA. Eight independent 3D metrics were significantly reduced (Mann-Whitney, n=4/group)
Dissociated regional activation signature. PFC showed no significant change in any structural 3D metric; only IBA-1 fluorescence intensity was reduced (1532 ± 127 vs. 1089 ± 320 AU; p=0.029). BLA–PFC amoeboid proportions were uncorrelated in both conditions (Control: r=−0.21, p=0.787; LPS: r=−0.58, p=0.420), demonstrating region-autonomous microglial programming.
Conclusions: miMAPA detects disease-specific microglial remodeling with high sensitivity across three independent cohorts, distinct brain regions (BLA, PFC, hippocampus), and different pathological contexts (LPS, PCOS, prenatal opioid exposure). The platform outperforms existing tools in metric coverage and morphological resolution, and reveals region-specific activation signatures partially obscured by conventional 2D analysis.
