Synaptic learning rules in the cortex: From spike timing-dependent to network state-dependent synaptic plasticity

Spike timing-dependent plasticity (STDP) has emerged as a computationally attractive learning rule for cortical circuit refinement during development. In STDP, the order and precise timing of pre- and postsynaptic action potentials determine the polarity of synaptic change: If the presynaptic input is active before the postsynaptic spike, then long-term potentiation (LTP) occurs, whereas long-term depression (LTD) is induced if this order is reversed. Both LTP and LTD require NMDA receptors, but whereas LTP always requires postsynaptic NMDA receptors, activation of presynaptic NMDA receptors may induce LTD. This raises the possibility that LTD could be induced without involvement of the postsynaptic neuron. Indeed, specific spike patterns in the presynaptic neuron can induce LTD without any requirement of postsynaptic mechanisms. That result calls for investigations into the plasticity rules that operate in vivo. Using whole-cell recordings and optogenetic stimulation of presynaptic input in urethane-anesthetized mice, which exhibit slow-wave-sleep (SWS)-like activity, we show that synaptic plasticity rules are gated by cortical dynamics: Active network states are biased towards synaptic depression, with presynaptic stimulation alone leading to LTD. This latter plasticity rule provides an attractive mechanism for SWS-related synaptic downscaling and circuit refinement. In conclusion, synaptic plasticity rules are diverse and network state-dependent.