Abstract
Tracing the motion of macromolecules, viruses, and nanoparticles adsorbed onto cell membranes is currently the most direct way of probing the complex dynamic interactions behind vital biological processes, including cell signalling, trafficking, and viral infection. The resulting trajectories are usually consistent with some type of anomalous diffusion, but the molecular origins behind the observed anomalous behaviour are usually not obvious. Here we use coarse-grained molecular dynamics simulations to help identify the physical mechanisms that can give rise to experimentally observed trajectories of nanoscopic objects moving on biological membranes. We find that diffusion on membranes of high fluidities typically results in normal diffusion of the adsorbed nanoparticle, irrespective of the concentration of receptors, receptor clustering, or multivalent interactions between the particle and membrane receptors. Gel-like membranes on the other hand result in anomalous diffusion of the particle, which becomes more pronounced at higher receptor concentrations. This anomalous diffusion is characterised by local particle trapping in the regions of high receptor concentrations and fast hopping between such regions. The normal diffusion is recovered in the limit where the gel membrane is saturated with receptors. We conclude that hindered receptor diffusivity can be a common reason behind the observed anomalous diffusion of viruses, vesicles, and nanoparticles adsorbed on cell and model membranes. Our results enable direct comparison with experiments and offer a new route for interpreting motility experiments on cell membranes.
Highlights
We find that the anomalous diffusion is not caused by mutivalent binding, as previously proposed, but by the hindered receptor diffusivity, which results in the particle trapping in regions rich in receptors
To take into account the variety of cell membranes observed in biological systems we employ both a fluid-like membrane model introduced in ref. 18 and 19 and a cross-linked model based on ref
An alternative approach, inspired by the accurate bi-Gaussian fit of the cumulative density functions, is to segment trajectories in distinct diffusive modes and assign separate models to each of them,[33] or to combine different anomalous diffusion models.[45,46,47]. We focus on the latter by invoking a so-called noisy continuous time random walk.[40,45]
Summary
Deducing the underlying molecular mechanisms from the trajectories is highly non-trivial as a large number of possible molecular mechanisms can result in similar anomalous motility.[10,11] For instance, high-speed single-particle tracking studies have reported anomalous diffusion of functionalised nanoparticles, vesicles, and virus-like particles bound to receptors on membranes in living cells[1] and supported bilayers.[2,3,4,5,6,7,8,9] Various physical and chemical effects have been proposed to underlie the observed anomalous diffusion,[12] including multivalent interactions between the nanoparticle and receptors, coupling between membrane leaflets,[2] molecular pinning,[2] receptor clustering,[13] formation of transient membrane domains,[14,15,16] and membranecytoskeleton interactions.[17]. We take a reverse approach: we simulate physical interactions between a nanoobject and deformable fluctuating membranes and measure the resulting trajectories for membranes of various properties. We characterise the resulting diffusion profiles of the nanoobject and match them to the underlying molecular mechanisms that are evident in molecular simulations. The particle binds to the membrane via multivalent interactions with the membrane receptors and locally deforms the membrane underneath it. We measure the nanoparticle’s diffusion profile within molecular dynamics (MD) simulations on fluid, gel-like, and fully cross-linked membranes, at varying receptor concentrations. We find a range of behaviours, from standard random walk to anomalous diffusion characterised by the particle’s hopping between regions of local trapping. We find that the anomalous diffusion is not caused by mutivalent binding, as previously proposed, but by the hindered receptor diffusivity, which results in the particle trapping in regions rich in receptors. We provide an in-depth numerical analysis of the data and a theoretical framework that characterises the observed anomalous diffusion
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
