3D neural network-based particle tracking reveals spatial heterogeneity of the cytosol

The spatial structure and physical properties of the cytosol are not well understood. Measurements of the material state of the cytosol are challenging due to its spatial and temporal heterogeneity, the lack of truly passive probes, and the need for probes of many sizes to accurately describe the state across scales. Recent development of genetically encoded multimeric nanoparticles (GEMs) has opened up study of the cytosol at the length scales of multiprotein complexes (20-60 nm). Using these probes to spatially resolve diffusivity of a cytoplasmic volume within a cell requires accurate and automated 3D tracking methods. We developed an image analysis pipeline for whole-cell imaging of GEMs in the context of large, multinucleate fungi where there is evidence of functional compartmentalization of the cytosol for both the nuclear division cycle and branching. We apply a neural network to track particles in 3D, generate surface meshes to project data on representations of the cell, and create dynamic visualizations of local diffusivities. Using this pipeline, we have found that there is remarkable variability in the properties of the cytosol both within a single cell and between cells. By analyzing the spatial diffusivity patterns, we saw an enrichment of low diffusivity zones at hyphal tips and near some nuclei. These results show that the physical state of the cytosol varies spatially within a single cell and exhibits significant cell-to-cell variability. Thus, molecular crowding contributes to heterogeneity within individual cells and across populations.