Neither the frequency of bursts (control: 15.02 ± find more 2.06 min−1, TTX: 17.92 ± 1.23 min−1), the frequency of local calcium transients per synapse (control: 0.58 ± 0.09 min−1, TTX: 0.72 ± 0.11min−1), nor the density of functional synapses (control: 39.5 ± 14.8 mm−1, TTX: 57.6 ± 23.6 mm−1) was significantly different between control and TTX treated cells. And, as expected, the fine-scale organization of synaptic

inputs in control cells was indistinguishable from that in our first set of experiments (compare Figure 5C and Figure 6). In contrast, the relationship between distance and input correlation was entirely abolished in cells that developed in the absence of neuronal spiking (Figure 6A). Interestingly, we observed not only a significant reduction of coactivation at neighboring synapses, but also an increase in coactivation in synapse pairs of intermediate distances (50–100 μm). This suggested that spiking activity

led to the stabilization of neighboring coactive synapses and a depletion of synapses that are coactive at intermediate distances. The latter conclusion is further supported by the observations that very distant synapse pairs (>100 μm) exhibit higher correlations than those of intermediate distance (Figures 5D and 6A) and that the correlation of very distant synaptic pairs was identical in TTX treated and control cells (Figure 6A). Finally SB203580 we investigated whether NMDA receptors, which mediate calcium signaling at the synapse (Figure 1H), but are dispensable for bursting, are required for the activity-dependent

these development of synaptic clustering. Slices were incubated in medium containing APV for 3–4 days. Subsequently, APV was washed out and synapses were mapped functionally. Very similar to TTX, APV abolished the clustering of functional synaptic inputs (Figure 6B), indicating that sorting functional inputs along developing dendrites is mediated by network firing activity and NMDA mediated synaptic plasticity. The patterns of synaptic activation received by a developing neuron are crucial for the fine-tuning of its synapses. Here, we mapped the spatiotemporal activity patterns of large populations of synaptic inputs onto hippocampal pyramidal cells using calcium imaging combined with patch-clamp recordings. Our analysis gave several new insights into the fine-scale synaptic organization during development. First, we found that different sets of synapses are activated during successive bursts of synaptic inputs. Second, even though activation patterns vary from burst to burst, they are not completely random: synapses that are located close to each other are much more likely to be coactive than more distant ones. Third, the emergence of this fine-scale input organization requires spiking activity and NMDA receptor activation.