Novel high-content and open-source image analysis tools for profiling mitochondrial morphology in neurological cell models

Abstract

Mitochondria undergo dynamic morphological changes depending on cellular cues, stress, genetic factors, or disease. The structural complexity and disease-relevance of mitochondria have stimulated efforts to generate image analysis tools for describing mitochondrial morphology for therapeutic development. Using high-content analysis, we measured multiple morphological parameters and employed unbiased feature clustering to identify the most robust pair of texture metrics that described mitochondrial state. Here, we introduce a novel image analysis pipeline to enable rapid and accurate profiling of mitochondrial morphology in various cell types and pharmacological perturbations. We applied a high-content adapted implementation of our tool, MitoProfilerHC, to quantify mitochondrial morphology changes in i) a mammalian cell dose response study and ii) compartment-specific drug effects in primary neurons. Next, we expanded the usability of our pipeline by using napari, a Python-powered image analysis tool, to build an open-source version of MitoProfiler and validated its performance and applicability. In conclusion, we introduce MitoProfiler as both a high-content-based and an open-source method to accurately quantify mitochondrial morphology in cells, which we anticipate to greatly facilitate mechanistic discoveries in mitochondrial biology and disease.