A novel algorithm identifies stress-induced alterations in mitochondrial connectivity and inner membrane structure from confocal images

Mitochondria exist as a highly interconnected network that is exquisitely sensitive to variations in nutrient availability, as well as a large array of cellular stresses. Changes in length and connectivity of this network, as well as alterations in the mitochondrial inner membrane (cristae), regulate cell fate by controlling metabolism, proliferation, differentiation, and cell death. Given the key roles of mitochondrial dynamics, the process by which mitochondria constantly fuse and fragment, the measure of mitochondrial length and connectivity provides crucial information on the health and activity of various cell populations. However, despite the importance of accurately measuring mitochondrial networks, the tools required to rapidly and accurately provide this information are lacking. Here, we developed a novel probabilistic approach to automatically measure mitochondrial length distribution and connectivity from confocal images. This method accurately identified mitochondrial changes caused by starvation or the inhibition of mitochondrial function. In addition, we successfully used the algorithm to measure changes in mitochondrial inner membrane/matrix occurring in response to Complex III inhibitors. As cristae rearrangements play a critical role in metabolic regulation and cell survival, this provides a rapid method to screen for proteins or compounds affecting this process. The algorithm will thus provide a robust tool to dissect the molecular mechanisms underlying the key roles of mitochondria in the regulation of cell fate.Author summary: Mitochondria are the main providers of cellular energy and as such, play a crucial role in many cellular processes whose deregulation are linked to both neurodegenerative diseases and cancer. In order to perform their functions, mitochondria dynamically regulate their length, assemble into interconnected networks and modulate their internal structure. Measuring these mitochondrial dynamics can thus provide important information about cellular state. Although some automated methods can provide a partial assessment of mitochondrial networks, the gold standard remains manual quantification, a time-consuming process. Here, we developed a new algorithm that accurately identifies both mitochondrial elongation and fragmentation occurring in response to different cellular stresses, as well as concomitant changes in mitochondrial connectivity and their internal structure. Given its sensitivity and ease of use, the algorithm will provide a robust tool to dissect the mechanisms by which mitochondria regulate cell fate.