The endoplasmic reticulum (ER) is the largest organelle in eukaryotic cells and consists of a dynamic network of stacked sheets and lipid membrane tubules stretching from the nucleus to the cell periphery. The ER serves an essential role in the distribution of lipids, ions and proteins throughout the cell, but the impact of network morphology on transport remains poorly understood. To that end, analytic mean first passage times for particles diffusing on extracted ER network structures are used as a measure of accessibility for different regions of the ER, which exhibit a wide heterogeneity due to varying tubule density and connectivity patterns. Moving beyond average search times, we combine agent-based simulations of particles diffusing on tubular networks with in vivo data on the spreading of photoactivated membrane proteins in COS-7 cells, in order to quantify how local ER structure determines protein spreading. We demonstrate that static local connectivity serves to explain some but not all of the observed behavior of proteins in the ER. Network dynamics, such as tubule growth and rearrangement are also implicated. In order to quantify these effects, we develop a dynamic model of the peripheral ER that aims to capture its unique, constantly evolving structure. This minimal network model reproduces key features that are observed experimentally in the ER of COS-7 cells such as length and area distributions. Intriguingly, the effective node diffusivity predicted from the model is similar to the observed diffusivity of ER exit site structures. Pairing analytic calculations and theoretical modeling with live-cell imaging, we highlight the role the heterogeneous ER network morphology plays in the non-uniform transport of proteins and gain fresh insights into the dynamic nature of the network itself.
Read full abstract