, 2008). Finally, models Apoptosis inhibitor for synaptic clustering have been proposed as a means to increase the computational capacity of dendrites (Larkum and Nevian, 2008) and as a form of long-term memory storage (Govindarajan et al., 2006). These models

have generally been derived from evidence of coordinated plasticity between excitatory synapses (Govindarajan et al., 2011 and Harvey and Svoboda, 2007). We find that clustered plasticity at the level of synapse formation and elimination can also occur between excitatory and inhibitory synapses and that these changes occur mainly within 10 μm of each other. This is a distance at which calcium influx and calcium-dependent signaling molecules from individual excitatory inputs can directly influence the plasticity of neighboring excitatory synapses (Govindarajan MDV3100 clinical trial et al., 2011, Harvey and Svoboda, 2007 and Harvey et al., 2008). Activation of excitatory inputs can also induce

translocation of calcium-dependent signaling molecules to inhibitory synapses resulting in enhanced GABA(A) receptor surface expression (Marsden et al., 2010). Further experiments using GABA uncaging also demonstrate selective inhibition of calcium transients in dendritic regions less than 20 μm from the uncaging site (Kanemoto et al., 2011). These findings and ours suggest that spatial constraints may influence coordinated plasticity between inhibitory and excitatory synapses along dendritic segments. Whereas we and others (Hofer et al., 2009) tuclazepam observe no increase in spine gain or loss on L2/3 pyramidal neurons during adult OD plasticity, the increased clustering of inhibitory synapse-dendritic spine remodeling in response to MD suggests that experience produces coordinated rearrangements between dendritic spines and inhibitory synapses. In the case of dually innervated spines, gating of the excitatory inputs can also be modified by the addition/elimination of inhibitory spine synapses. Thus, MD may still influence

excitatory synaptic plasticity in this cell type without altering the overall rate of spine turnover. These findings provide evidence that experience-dependent plasticity in the adult cortex is a highly orchestrated process, integrating changes in excitatory connectivity with the active elimination and formation of inhibitory synapses. For construction of the Cre expression plasmid (pFsynCreW), a Cre insert with 5′ NheI and 3′ EcoRI restriction sites was generated by PCR amplification from a WGA-Cre AAV vector ( Gradinaru et al., 2010) and subcloned into a pLL3.7syn lentiviral expression plasmid ( Rubinson et al., 2003). The Cre-dependent eYFP expression plasmid (pFUdioeYFPW) was constructed by subcloning a “double” floxed inverse orientation (dio) eYFP expression cassette (a gift from K. Deisseroth) into the pFUGW lentiviral expression plasmid ( Lois et al., 2002), replacing the GFP coding region between the 5′ BamHI and 3′ EcoRI restriction sites.