| dc.description.abstract | PSD95 is a core scaffold protein of the postsynaptic density (PSD) at excitatory synapses and a member of the MAGUK family, where it plays a central role in organizing glutamate receptors, ion channels, and signalling complexes into stable nanoscale assemblies. Its regulated membrane anchoring via palmitoylation and capacity for multivalent binding suggest that PSD95 is not randomly distributed within synapses but arranged into ordered domains that could regulate the fidelity of synaptic transmission. While synaptic transmission is inherently stochastic, with variability at every step from neurotransmitter release to receptor activation, neurons still maintain highly reliable information flow. The mechanisms by which scaffold proteins like PSD95 filter molecular noise, regulate order, and enhance information fidelity remain poorly understood.
To address this, we investigated the nanoscale organization, mobility, and phase behaviour of PSD95 in rat primary hippocampal neurons. Using a combination of super-resolution imaging modalities: direct stochastic optical reconstruction microscopy (dSTORM), single-particle tracking photoactivated localization microscopy (sptPALM), and DNA points accumulation for imaging in nanoscale topography (DNA-PAINT), we mapped PSD95 distributions at nanometre precision and tracked its real-time movements within living synapses. PSD95 was visualized through antibody labeling, intrabody cDNA transfection, and genetic fusion to fluorescent proteins.
We found that PSD95 exhibits a heterogeneous nanoscale distribution, forming high-density clusters or “nanodomains” within the PSD. Membrane-bound PSD95 localized almost exclusively to these nanodomains, and 2-Bromopalmitate-mediated palmitoylation blockade destabilized them, redistributing PSD95 across the broader PSD and synaptic compartment. Single-particle tracking revealed sequential confinement of PSD95 molecules from dendritic shafts to spine heads, then to the PSD, and finally into nanodomains. These trajectories, modelled as random harmonic oscillations, indicated that steep potential energy wells act as stabilizing features, confining PSD95 into discrete high-density states.
To understand the physical basis of nanodomain stability, we investigated the phase separation behaviour of different conformations of PSD95. Membrane-bound PSD95 showed the highest propensity for condensation within both PSD and nanodomain compartments, whereas cytosolic PSD95 had the lowest condensation propensity due to unfavourable energetics and a high critical cluster size. In-situ investigation of the phase behaviour of PSD95 indicated that it has a propensity for condensation and therefore could modulate the strength of synapse in response to plasticity, by regulating the recruitment of receptors to the condensates. To investigate the effect of activity-dependent plasticity paradigms on the phase behaviour of information reception machinery, we probed for the GluA2-subunit of AMPAR along with PSD95. Induction of various plasticity paradigms in hippocampal neurons revealed bidirectional control of phase behaviour: chemical paradigms that strengthen the synapse enhanced condensation propensity and reinforced information encoding capacity, while chemical paradigms that weaken the synapse reduced both. Homeostatic scaling up increased the number of PSD95 and GluA2 molecules without increasing condensation, creating a primed but non-encoding state.
Integrating these findings, we propose that PSD95 nanodomains function as ordered, low-entropy molecular assemblies that buffer synaptic noise and enable reliable signal transmission. High-density clustering, steep potential wells, and strong condensation propensity work together to reduce molecular disorder, generating local physical states optimal for stable and precise synaptic communication. By coupling quantitative nanoscale mapping with physical modelling, this work provides a mechanistic framework for how the modular assembly of synaptic proteins mediates molecular order in the face of synaptic stochasticity, offering a foundation for future studies on the physical principles of noise filtering in neural circuits. | en_US |
