| Abstract | A fundamental unsolved question in the post-genomic era is how the 3-dimensional geometry of cellular structures is generated by networks of molecular interactions. Mitochondria are an ideal system in which to approach this question. Mitochondrial morphology ranges from small individual organelles to giant reticular networks and affects mitochondrial functions. In budding yeast mitochondria compose a dynamic tubular network at the cell periphery, which undergoes subtle, consistent changes in morphology under different physiological conditions. We aim to understand the mechanisms that regulate the normal range of reticular structures. We have developed a method to quantify the 3D interconnected morphology of the mitochondrial network by considering the mitochondria as a network of edges (the tubules) and nodes (the branch points where tubules connect). Live mitochondria are rapidly imaged in 3D at high resolution using Spinning-disk Confocal microscopy. Specialized segmentation and 3D skeletonization methods are applied to the image stacks to extract the 3D mathematical graph of the mitochondrial network. For validation we create theoretical microscope images from pre-determined skeletonized structures and compare these to the 3D skeletons obtained by our “mito-graph” methodology. These tests are used to optimize the accuracy of the 3D skeletons and determine their limits, imposed by microscopy imaging. We then apply concepts and methods from complex networks research to characterize the connectivity and clustering of the tubules. This quantification technique provides us with measurements of the arc lengths, density, and connectivity of mitochondrial tubules in the cell and how the connections change as individual tubules undergo constant fusion and fission dynamics or individual cells progress through the cell cycle. In our first analysis we focus on how the properties of the network change with tubule density at the surface and which properties depend on fusion and fission dynamics. We also compare measurements of real mitochondrial networks with simulated networks created by joining points (nodes) on the surface of a sphere with edges following a set of rules to explore the minimum requirements necessary to generate quantitatively realistic mitochondrial networks. |
|---|
