Archives
s3i Neural activity resulting from sensory
Neural activity resulting from sensory experience is required for the refinement and maturation of neural circuits during cortical critical periods. This refinement is thought to be important for functional optimization of primary sensory cortices, enabling critical functions such as feature detection (Kang et al., 2013; Wang et al., 2013). Here we show that AMPAR-positive synapses are maintained at an equilibrium in V1, accompanied by a net loss of AMPAR-silent synapses during the critical period of cortical plasticity, resulting in a decrease in total glutamatergic synapse numbers. We demonstrate that Ng coordinates balanced, experience-driven AMPAR-silent synapse conversion and glutamatergic synapse elimination in L2/3 pyramidal s3i in V1 to achieve functional maturation and optimization of the cortical excitatory circuitry. Our results support the notion that AMPAR-positive synapses are reorganized during the critical period leading to excitatory circuit maturation. This circuit maturation is likely mediated by the combined effect of AMPAR-silent synapse maturation, defined by decreasing the proportion of silent synapses and increasing the AMPAR/NMDAR EPSC ratio (Huang et al., 2015) and the elimination of existing glutamatergic synapses (Bian et al., 2015; Holtmaat et al., 2005; Zuo et al., 2005a). These two processes are coordinated to maintain the equilibrium of AMPAR-positive synapses. It is noteworthy that conversion of AMPAR-silent synapses to AMPAR-positive synapses likely contributes to this progressive AMPAR-silent synapse maturation (Ashby and Isaac, 2011; Huang et al., 2015; Wu et al., 1996), which should lead to an overall enhancement of AMPAR-mediated synaptic transmission. However, the equilibrium in AMPAR-mediated synaptic transmission implies that some existing AMPAR-positive synapses must go through experience-dependent downregulation and counteract the AMPAR-silent synapse conversion for excitatory circuit maturation. The finding of a net loss of total synapses and AMPAR-silent synapses is in agreement with previous morphological studies showing a net decrease in spine density during development, preferentially in thin spines, which are thought to represent immature, weaker, or AMPAR-silent synapses (Bian et al., 2015; Holtmaat et al., 2005; Zuo et al., 2005a). Although our data do not provide direct evidence for the life cycle of glutamatergic synapses, based upon our functional analyses it is likely that both AMPAR-positive and AMPAR-silent synapses are eliminated during the critical period. Interestingly, experience-dependent re-silencing of AMPAR-positive synapses has been shown in the nucleus accumbens under special circumstances (Graziane et al., 2016), providing an intriguing possibility for AMPAR-positive synapse elimination. Further experiments are required to differentiate these possibilities and measure the dynamics of AMPAR-silent synapse maturation and synapse elimination. Importantly, our results imply that AMPAR-silent synapse maturation is part of the excitatory circuit maturation process via coordinated spine elimination and AMPAR-silent synapse conversion during heightened experience-dependent cortical plasticity. Consistent with previous reports, our results indicate that sensory deprivation prevents AMPAR-silent synapse maturation (Funahashi et al., 2013). Our results further show that sensory deprivation arrests synapses at the early developmental stage, with a higher proportion of AMPAR-silent synapses and a sustained number of AMPAR-positive synapses. It is likely that both experience-dependent synapse elimination and AMPAR-silent synapse conversion are prevented by sensory deprivation. Remarkably, the function of the CaM-complexing protein Ng is essential for this experience-dependent coordination. Loss of Ng broke the experience-dependent equilibrium of AMPAR-positive synapses, induced the loss of AMPAR-positive synapses, and halted AMPAR-silent synapse maturation. Deprivation of sensory input to V1 prevented the deterioration in synaptic transmission caused by loss of Ng and likely arrested the circuit at an earlier developmental stage prior to sensory deprivation (Figures 1 and 3). These results highlight the critical involvement of Ng in coordinating experience-dependent refinement of cortical excitatory circuitry.