Trackosome: a computational toolbox to study the spatiotemporal dynamics of centrosomes, nuclear envelope and cellular membrane

Abstract

During the initial stages of mitosis, multiple mechanisms drive centrosome separation and positioning. How they are functionally coordinated to promote centrosome migration to opposite sides of the nucleus remains unclear. Imaging analysis software has been used to quantitatively study centrosome dynamics at this stage. However, available tracking tools are generic and not fine-tuned for the constrains and motion dynamics of centrosome pairs. Such generality limits the tracking performance and may require exhaustive optimization of parameters. Here, we present Trackosome, a freely available open-source computational tool to track the centrosomes and reconstruct the nuclear and cellular membranes, based on volumetric live-imaging data. The toolbox runs in MATLAB and provides a graphical user interface for easy and efficient access to the tracking and analysis algorithms. It outputs key metrics describing the spatiotemporal relations between centrosomes, nucleus and cellular membrane. Trackosome can also be used to measure the dynamic fluctuations of the nuclear envelope. A fine description of these fluctuations is important because they are correlated with the mechanical forces exerted on the nucleus by its adjacent cytoskeletal structures. Unlike previous algorithms based on circular/elliptical approximations of the nucleus, Trackosome measures membrane movement in a model-free condition, making it viable for irregularly shaped nuclei. Using Trackosome, we demonstrate significant correlations between the movements of the two centrosomes, and identify specific modes of oscillation of the nuclear envelope. Overall, Trackosome is a powerful tool to help unravel new elements in the spatiotemporal dynamics of subcellular structures. Recent advances in live-cell imaging and image analysis techniques made it possible to access the subcellular environment and quantitatively examine its underlying mechanisms. Commercial imaging software, like Imaris and Metamorph, are equipped with automatic particle tracking functions which have been used to track centrosome pairs in 3-dimensions (Collins et al. 2014; Yamashita et al. 2015; De Simone et al. 2016). In alternative, the open-source freeware ImageJ (Eliceiri et al. 2012) with its particle tracking plugin Trackmate (Tinevez et al. 2017), has also been used for centrosome tracking in several studies (Boudreau et al. 2018; Mahen 2018; Boudreau et al. 2019). However, these generic tracking tools are not fine-tuned for the appearance and motion dynamics/constraints of centrosome pairs, which limits their tracking performance often requiring exhaustive parameter optimization, particularly when high-quality videos are not available. Moreover, when studying the dynamics of spindle formation, it is often necessary to analyze centrosomes movement in reference to the cellular and nuclear membrane (and therefore, in a non-canonical coordinates system). To the best of our knowledge, the available computational tools do not directly allow the analysis of the coordinated changes between di...

The LINC complex ensures accurate centrosome positioning during prophase

“…This allowed us to accurately standardise cell shape and intracellular organization(Théry et al 2005; Nunes et al 2020). Then, we reconstructed the dynamics of centrosomes, nucleus, and cell membrane during the G2-M transition, using our previously developed tool, Trackosome (Castro et al 2020) (Fig. S1A-E).…”

“…Analysis of centrosome positioning and behaviour during mitotic entry was performed using a custom-designed MATLAB (The MathWorks Inc, USA; v2018b) script(Castro et al 2020) that contains a specialised workflow previously optimised for centrosome tracking, together with the reconstruction of both cellular and nuclear membranes in a 3D space. A pixel size of 0.176μm and a z-step of 0.5μm, were taken into consideration.…”

“…Coverslips were then blocked using 10% FBS in 0.1% Triton-X100 for 30 minutes, at RT. These coverslips were afterwards incubated with the following primary antibodies: rabbit MATLAB custom algorithm for centrosome tracking Analysis of centrosome positioning and behaviour during mitotic entry was performed using a custom-designed MATLAB (The MathWorks Inc, USA; v2018b) script 36 that contains a specialised workflow previously optimised for centrosome tracking, together with the reconstruction of both cellular and nuclear membranes in a 3D space. A pixel size of 0.176m and a z-step of 0.5m, were taken into consideration.…”