IBENS, ENS, PSL
Biophysical determinants of synaptic stability and size fluctuations
Developmental and activity-dependent modifications of synapses are thought to underlie learning and memory. However, synapses are microscopic structures often containing only tens to hundreds of neurotransmitter receptors responsible for synaptic transmission. As such, they are exposed to strong fluctuations, and an important problem of neuroscience remains to reconcile the molecular fluctuations of synapses with the long-term nature of the memories they encode, requiring a precise understanding of the biophysical mechanisms that govern the accumulation of neurotransmitter receptors at synaptic domains. In this talk, I will present our efforts to tackle this problem using biophysical modeling, in close collaboration with experimentalists. Focussing on inhibitory postsynaptic domains, which are less complex in molecular composition and morphology than excitatory synapses yet share many of the underlying organizing principles, we inferred receptor and scaffold protein kinetics and predicted synaptic receptor number fluctuations. Surprisingly, synapses are predicted to be rather stable, with small size fluctuations that decay on relatively fast (hour) timescales. We then set out to study the variability of biophysical parameters underlying these dynamics. By reanalyzing protein turnover data for different synapse size classes, we identified a range of parameter (co-)variation. Sampling parameters from Gaussian parameter distributions with accordingly constrained variances gave rise to skewed size distributions, matching experimental data surprisingly well. The inferred variability of synaptic parameters allows one to reconcile the predicted short decorrelation timescale with slow (days) synaptic size fluctuations reported in long-time imaging experiments of both excitatory and inhibitory synapses, which can be interpreted as slow parameter fluctuations. Beyond providing a mechanistic understanding of receptor and scaffold protein kinetics at inhibitory synapses, our model provides a framework to study size dynamics and variability of synapses, and more broadly to address the important question of synaptic weight volatility.