Automated Image-Quantification for Investigating the Mechanics of Cytoskeletal Spindles

Abstract

Propagation of a cell requires successful segregation of its genetic material into resulting daughter cells, a process brought about by the coordinated dynamics of cytoskeletal spindles. High resolution time-lapse microscopy has opened new avenues for understanding the dynamics of spindles within single cells. Automated computational image analysis allows the extraction of useful quantitative information from microscopy data in an unbiased, objective and reproducible manner. In the present work, I have developed three automated image analysis tools in MATLAB to investigate the quantitative aspects of subcellular dynamics such as particle motion, intensity and interaction. The tools are called Automated Multi-peak Tracking Kymography (AMTraK), Fluorescence Tracker (FluoreT) and Differential Interference Contrast Object Tracker (DICOT). While AMTraK and FluoreT are used to analyse time-lapse fluorescence images, DICOT applies to images from time-lapse DIC microscopy. The tools are validated with simulated noisy data, tested on experimental image-series and benchmarked by comparing their results with manual measurements, published literature and outputs of other software. The wide utility of these tools is demonstrated on diverse data ranging from in vitro microtubule gliding assays and clathrin assembly kinetics to in vivo axonal vesicle transport, DNA segregation in E. coli and cytoplasmic granule mobility in C. elegans embryos. ParM filaments that form a plasmid-segregating spindle in E. coli cells display dynamic instability and are also known to slide in vitro. Our Brownian dynamics simulations of the ‘search and capture’– based assembly of the ParMRC spindle suggest that dynamic instability does not alter the time taken by ParM filaments for plasmid capture. Quantitation of ParM sliding using AMTraK shows that the filaments slide against each other in a contractile, sub-diffusive manner at a speed higher than the rate of polymerization. The mobility of the mitotic spindle during the process of asymmetric positioning is shown to vary between nematode species, suggesting an evolutionary difference in spindle mechanics. Viscosity of the cytoplasm is hypothesized to give rise to these differences. Our novel, non-invasive quantification of cytoplasmic viscosity using DICOT suggests that viscosity may contribute to the observed difference in spindle motion patterns during positioning in the embryos of six related nematode species including C. elegans. My work highlights the insights that computational tool development can provide when applied to in vitro and in vivo dynamics of subcellular processes.