Waigh, T. A. | bioRxiv Science

bioRxiv2020

Network Organisation and the Dynamics of Tubules in the Endoplasmic Reticulum

The endoplasmic reticulum (ER) is a eukaryotic subcellular organelle composed of tubules and sheet-like areas of membrane connected at junctions. The tubule network is highly dynamic and undergoes rapid and continual rearrangement. There are currently few tools to evaluate network organisation and dynamics. We quantified ER network organisation in Vero and MRC5 cells, and developed a classification system for ER dynamics in live cells. The persistence length, tubule length, junction coordination number and angles of the network were quantified. Hallmarks of imbalances in ER tension, indications of interactions with microtubules and other subcellular organelles, and active reorganisation and dynamics were observed. Live cell ER tubule dynamics were classified using a Gaussian mixture model, defining tubule motion as active or thermal and conformational phase space analysis allowed this classification to be refined by tubule curvature states. STATEMENT OF SIGNIFICANCEThe endoplasmic reticulum (ER), a subcellular organelle, is an underexplored real-world example of active matter. Many processes essential to cell survival are performed by the ER, the efficacy of which may depend on its organisation and dynamics. Abnormal ER morphology is linked to diseases such as hereditary spastic paraplegias and it is possible that the dynamics are also implicated. Therefore, analysing the ER network in normal cells is important for the understanding of disease-related alterations. In this work, we outline the first thorough quantification methods for determining ER organisation and dynamics, deducing that tubule motion has a binary classification as active or thermal. Active reorganisation and dynamics along with indications of tension imbalances and membrane contact sites were observed.

biophysics↗

bioRxiv2019

Deciphering anomalous heterogeneous intracellular transport with neural networks

Biological intracellular transport is predominantly heterogeneous in both time and space, exhibiting varying non-Brownian behaviour. Characterisation of this movement through averaging methods over an ensemble of trajectories or over the course of a single trajectory often fails to capture this heterogeneity adequately. Here, we have developed a deep learning feedforward neural network trained on fractional Brownian motion, which provides a novel, accurate and efficient characterization method for resolving heterogeneous behaviour of intracellular transport both in space and time. Importantly, the neural network requires significantly fewer data points compared to established methods, such as mean square displacements, rescaled range analysis and sequential range analysis. This enables robust estimation of Hurst exponents for very short time series data, making possible direct, dynamic segmentation and analysis of experimental tracks of rapidly moving cellular structures such as endosomes and lysosomes. By using this analysis, we were able to interpret anomalous intracellular dynamics as fractional Brownian motion with a stochastic Hurst exponent.

cell biology↗