Characterization of Mitochondrial Dynamics during Mitosis by Lattice Light Sheet Microscopy and Numerical 3D Morphological Analysis Chung-Chih Lin1*, Yeh-Hsio, Chu2, Yu-Ting Tseng1, An-Chi Chen1, Lung-Sen Kao1 and Bi-Chang Chen3, 1Department of Life Sciences, National Yang-Ming University; 2Brain Research Center, National Yang-Ming University; 3Research Center for Applied Sciences, Academia Sinica, Taipei, Taiwan *Corresponding Author
Abnormal mitochondrial partition is correlated with abnormal mitochondrial mitosis. The current model of mitochondrial partition is random and passive. But there is no numerical analysis of mitochondrial morphology to support this. According to simulation and correlation of correct mitosis and variance of mitochondrial partition, more than 10% mitochondrial partition variance results in incorrect mitosis and CHO cells needs more than 700 mitochondria to avoid such variance if mitochondrial partition is random. To avoid effects of photobleach mitochondrial morphology and missing data points due to low temporal resolution, we use lattice light sheet microscopy to acquire 3D+t images of nucleus and mitochondria with high spatial and temporal resolution (depth of optical section=350nm; 80-100 slices/stack; 1stack/s). Numerical 3D analysis of mitochondrial morphology shows that mitochondria are mostly in tubules or networks and mitochondrial number is ranging 80-700. Our results show that mitochondrial number is not enough for even mitochondrial partition via random partition. There are two possible ways to partition mitochondria. The first is self-volume exclusion (passive) and the second is spindle-dependent (active). Simulation shows that self-volume exclusion has more accurate than random partition. Mitochondrial sizes are heterogeneous, and largest mitochondrial sizes are ranging 3%-50% of total mitochondrial volume in cells of accurate mitosis. When sizes of largest mitochondria are higher, variance of mitochondria increases. Because sizes of some largest mitochondria are more than 50%, these cells need mitochondrial fission for even partition. Moreover, some mitotic events result in one daughter cell without nuclei but variance of mitochondrial partition is not significant. These results show that self-volume exclusion model is not sufficient to explain mitochondrial partition in cells with larger mitochondria. When we inspect mitochondrial dynamics from metaphase to telophase, mitochondrial cluster, stretching and thinning before fission at division furrow is correlated with correct mitosis. Disruption of microtubule assembly reduces phenomena described above and increases higher abnormal mitotic effects. Reduction of mitochondrial fission decreases accuracy in mitochondrial partition and abnormal mitosis is also increased. Our findings show that mitochondrial partition is active instead random and passive. Mitochondrial clustering, stretching and thinning before fission is important to accurate mitosis and regulated by microtubule assembly and mitochondrial fission.