4D mitochondrial network assumes distinct and predictive phenotypes through human lung and intestinal epithelial development

Abstract

Mitochondria form a dynamic three-dimensional network within the cell that exhibits a wide range of morphologies and behaviors. Depending on cell state, cell type, and cell fate, a cells mitochondrial phenotype might range from relatively isolated mitochondrial segments to complex branching networks, and from stationary mitochondria to highly motile structures. While isolated mitochondrial phenotypes have been described for a subset of cell states, types, and fates, an integrated map of how mitochondrial phenotypes change over the full course of tissue development has so far been lacking. Here, we identify the mitochondrial phenotypes that appear throughout the course of lung and intestinal epithelial development from stem cells to differentiated tissue. Using human stem cell-derived intestinal and branching lung organoids that mimic developing human organs as model systems, we extract and analyze key mitochondrial biophysical phenotypes in human development. To achieve this, we employ lattice light-sheet microscopy (LLSM), which enables high-resolution, 4D (x, y, z, time) imaging of mitochondria in organoid tissues with minimal damage to the sample. We image at key developmental time points from stem cell differentiation into mature organoid tissue. For data processing, we utilize the MitoGraph and MitoTNT software packages along with our developed custom computational tools. These tools allow for automated 4D organoid to single cell image processing and quantitative 4D single cell mitochondrial temporal network tracking. This work represents the first 4D high spatiotemporal-resolution quantification of live human organoid tissues at the single-cell level through development. We identified distinct mitochondrial phenotypes unique to each organoid type and found correlations between mitochondrial phenotypes, cellular age, and cell type. Furthermore, we demonstrate that mitochondrial network characteristics can predict both organoid type and cell age. Our findings reveal fundamental aspects of mitochondrial biology that were previously unobservable, offering new insights into cell-type-specific mitochondrial dynamics and enabling new findings in relevant human model systems. We believe that our findings and methods will be essential for advancing 4D cell biology, providing a powerful framework for characterizing organelles in organoid tissues.