During learning and-development, neural circuitry is refined in part through changes in synapse number and strength. Most studies of long-term synaptic changes have concentrated on Hebbian, synapse-specific forms of plasticity such as long-term potentiation (LTP). While Hebbian plasticity is important for the refinement of neuronal circuitry, it is probably not sufficient, because it tends to destabilize the activity of neuronal networks. We have used a cortical culture system to identify several forms of homeostatic plasticity that counteract the destabilizing effects of Hebbian plasticity by keeping neuronal firing rates within functional boundaries. One important stabilizing mechanism is a form of synaptic plasticity termed synaptic scaling, where long-lasting changes in activity bidirectionally regulate the amplitude of excitatory synapses between cultured cortical pyramidal neurons, so that increased activity reduces all of a neuron's synaptic strengths, and vice versa. This process is slow, independent of NMDA receptor activation, and occurs through a postsynaptic change in AMPA receptor number. Synaptic scaling is mediated by activity- dependent changes in the release of the neurotrophin brain- derived neurotrophic factor (BDNF). Here we will identify the intracellular transduction cascades through which BDNF regulates synaptic strengths, and the sites of action of BDNF within the cortical network. During synaptic scaling, activity modifies NMDA and AMPA currents proportionally. Here we will determine whether NMDA and AMPA currents are scaled by the same or independent mechanisms, and how this scaling interacts with Hebbian plasticity. Finally, we will ask whether synaptic scaling occurs in vivo. By examining the mechanism and function of synaptic scaling, we hope to illuminate the processes that allow cortical circuits to maintain both flexibility and stability during learning and memory. Understanding the mechanisms that promote stability in cortical networks will allow us to understand how these processes may go awry in disease states such as epilepsy.
Showing the most recent 10 out of 34 publications
