SpinX: Time-resolved 3D Analysis of Mitotic Spindle Dynamics using Deep Learning Techniques and Mathematical Modelling

Time-lapse microscopy movies have transformed the study of subcellular dynamics. However, manual analysis of movies can introduce bias and variability, obscuring important insights. While automation can overcome such limitations, spatial and temporal discontinuities in time-lapse movies render methods such as object segmentation and tracking difficult. Here we present SpinX, a framework for reconstructing gaps between successive frames by combining Deep Learning and mathematical object modelling. By incorporating expert feedback through selective annotations, SpinX identifies subcellular structures, despite confounding neighbour-cell information, non-uniform illumination and variable marker intensities. The automation and continuity introduced allows precise 3-Dimensional tracking and analysis of spindle movements with respect to the cell cortex for the first time. We demonstrate the utility of SpinX using distinct spindle markers and drug treatments. In summary, SpinX provides an exciting opportunity to study spindle dynamics in a sophisticated way, creating a framework for step changes in studies using time-lapse microscopy.

Profiling Myxococcus xanthus swarming phenotypes through mutation and environmental variation

Myxococcus xanthus is a bacterium that lives on surfaces as a predatory biofilm called a swarm. As a growing swarm feeds on prey and expands, it displays dynamic multicellular patterns such as traveling waves called ripples and branching protrusions called flares. The rate at which a swarm expands across a surface, and the emergence of the coexisting patterns, are all controlled through coordinated cell movement. M. xanthus cells move using two motility systems known as Adventurous (A) and Social (S). Both are involved in swarm expansion and pattern formation. In this study, we describe a set of M. xanthus swarming genotype-to-phenotype associations that include both genetic and environmental perturbations. We identified new features of the swarming phenotype; recorded and measured swarm expansion using time-lapse microscopy; and compared the impact of mutations on different surfaces. These observations and analyses have increased our ability to discriminate between swarming phenotypes and provided context that allow us to identify some phenotypes as improbable ‘outliers’ within the M. xanthus swarming phenome. IMPORTANCE Myxococcus xanthus grows on surfaces as a predatory biofilm called a swarm. In nature, a feeding swarm expands by moving over and consuming prey bacteria. In the laboratory, a swarm is created by spotting cell suspension onto nutrient agar in lieu of prey. The suspended cells quickly settle on the surface as the liquid is absorbed into the agar, and the new swarm then expands radially. An assay that measures the expansion rate of a swarm of mutant cells is the first, and sometimes only, measurement used to decide whether a particular mutation impacts swarm motility. We have broadened the scope of this assay by increasing the accuracy of measurements and introducing prey, resulting in new identifiable and quantifiable features that can be used to improve genotype-to-phenotype associations.

Longitudinal tracking of neuronal mitochondria delineates PINK1/Parkin-dependent mechanisms of mitochondrial recycling and degradation

Altered mitochondrial quality control and dynamics may contribute to neurodegenerative diseases, including Parkinson’s disease, but we understand little about these processes in neurons. We combined time-lapse microscopy and correlative light and electron microscopy to track individual mitochondria in neurons lacking the fission-promoting protein dynamin-related protein 1 (Drp1) and delineate the kinetics of PINK1-dependent pathways of mitochondrial quality control. Depolarized mitochondria recruit Parkin to the outer mitochondrial membrane, triggering autophagosome formation, rapid lysosomal fusion, and Parkin redistribution. Unexpectedly, these mitolysosomes are dynamic and persist for hours. Some are engulfed by healthy mitochondria, and others are deacidified before bursting. In other cases, Parkin is directly recruited to the matrix of polarized mitochondria. Loss of PINK1 blocks Parkin recruitment, causes LC3 accumulation within mitochondria, and exacerbates Drp1KO toxicity to dopamine neurons. These results define a distinct neuronal mitochondrial life cycle, revealing potential mechanisms of mitochondrial recycling and signaling relevant to neurodegeneration.

Single Cell Variability of CRISPR-Cas Interference and Adaptation

CRISPR-Cas defence is a combination of adaptation to new invaders by spacer acquisition, and interference by targeted nuclease activity. While these processes have been studied on a population level, the individual cellular variability has remained unknown. Here, using a microfluidic device combined with time-lapse microscopy, we monitor invader clearance in a population of Escherichia coli across multiple generations. We observed that CRISPR interference is fast with a narrow distribution of clearance times. In contrast, for invaders with escaping PAM mutations we show large cell-to-cell variability of clearance times, which originates from primed CRISPR adaptation. Faster growth and cell division, as well as higher levels of Cascade, increase the chance of clearance by interference. In contrast, faster growth is associated with decreased chances of clearance by priming. A mathematical model explains the experimental findings, and identifies Cascade binding to the mutated invader DNA, rather than spacer integration, as the main source of priming heterogeneity. The highly stochastic nature of primed CRISPR adaptation implies that only subpopulations of bacteria are able to respond to invading threats in a timely manner. We conjecture that CRISPR-Cas dynamics and heterogeneity at the cellular level are crucial to understanding the strategy of bacteria in their competition with other species and phages.

Control of division and microtubule dynamics in Chlamydomonas by cyclin B/CDKB1 and the anaphase-promoting complex

In yeast and animals, cyclin B binds and activates the cyclin-dependent kinase ('CDK') CDK1 to drive entry into mitosis. We show that CYCB1, the sole cyclin B in Chlamydomonas, activates the plant-specific CDKB1 rather than the CDK1 ortholog CDKA1. Time-lapse microscopy shows that CYCB1 is synthesized before each division in the multiple fission cycle, then is rapidly degraded 3-5 minutes before division occurs. CYCB1 degradation is dependent on the anaphase-promoting complex (APC). Like CYCB1, CDKB1 is not synthesized until late G1; however, CDKB1 is not degraded with each division within the multiple fission cycle. The microtubule plus-end-binding protein EB1 labeled with mNeonGreen (EB1-NG) allowed detection of mitotic events in live cells. The earliest detectable step in mitosis, splitting of polar EB1-NG signal into two foci, likely associated with future spindle poles, was dependent on CYCB1. CYCB1-GFP localized close to these foci immediately before spindle formation. Spindle breakdown, cleavage furrow formation and accumulation of EB1 in the furrow were dependent on the APC. In interphase, rapidly growing microtubules are marked by 'comets' of EB1; comets are absent in the absence of APC function. Thus CYCB1/CDKB1 and the APC mitosis modulate microtubule dynamics while regulating mitotic progression.