Homeostatic regulation across fast and slow timescales through aggregate synaptic dynamics
Neuronal firing rates must remain stable despite continuous synaptic modifications driven by Hebbian plasticity. Two compensatory mechanisms have been implicated in this stability. Heterosynaptic plasticity, which rapidly redistributes synaptic strength among competing inputs, and synaptic scaling, which adjusts synaptic efficacies globally on slow timescales. How these interact to maintain stable activity remains unclear. Here, we propose a unified framework in which local synaptic competition for a shared resource pool, combined with slow cell-wide homeostatic regulation of the total pool size, prevents runaway synaptic dynamics and ensures firing rate homeostasis. We show analytically that this architecture guarantees stability regardless of the timescale difference between Hebbian and homeostatic plasticity, and that the ratio of sensing and regulatory timescales determines whether convergence is monotonic or oscillatory. Fitting the model to activity blockade data reveals an inverse coupling between these timescales. When competition is spatially restricted to dendritic compartments, the framework reproduces experimentally observed synaptic weight distributions and local homeostatic upscaling following denervation, suggesting that heterosynaptic competition and synaptic scaling share a common mechanistic origin.