Quantitative Biology

Material Properties of the Postsynaptic Density Condensate Set AMPA-Receptor Retention Time: A Quantitative Phase-Separation and Escape-Kinetics Model of LTP Maintenance

Authors: Brent Hartshorn
Comments: 12 Pages.

A central puzzle in synaptic plasticity is how long-term potentiation (LTP) is maintained forhours to days despite continuous turnover of synaptic proteins. Recent experiments reframe theproblem in physical terms: the postsynaptic density (PSD) is a biomolecular condensate whose material state—its network connectivity and viscoelasticity—is causally coupled to AMPA-receptor (AMPAR) mobility and to plasticity. In particular, weakening scaffold percolation softens the PSD condensate, increases AMPAR mobility, and impairs plasticity, while CaMKII maintenance of LTP appears to be a structural (binding) rather than a downstream-enzymatic effect. These results are qualitative. Here we supply the missing quantitative framework. We model the PSD as a multivalent condensate using a Flory—Huggins free energy augmented by a percolation (gelation) order parameter, and we treat AMPAR retention as Kramers escape from the free-energy well created by the condensate. The model yields three results. (i) The receptor residence time depends exponentially on scaffold connectivity, so that retention collapses by orders of magnitude when the network is driven below its percolation threshold—a sharp "depotentiation cliff" rather than a graded loss. (ii) Treating an AMPAR nanodomain as a capillary cluster gives a critical size of tens of nanometres from independently measured condensate surface tension and binding-site concentration, an order-of-magnitude match to the ∼70—80 nm nanodomains seen by super-resolution microscopy. (iii) The framework reproduces, without invoking any downstream phosphorylation, why binding-competent but enzymatically dead CaMKII can sustain potentiation: the memory variable is the material state of the condensate, not the phosphorylation state of an effector. We give a parameter table with sources and uncertainties, and three quantitative, falsifiable predictions that distinguish this model from a purely biochemical account. We make no claims beyond synaptic biophysics; nothing here concerns quantum coherence, consciousness, or non-local correlations.
Category: Quantitative Biology