This loss of inhibition facilitates competition-driven spine turnover on layer 5 pyramidal cells, presumably

through the loss of inhibitory synapses on these cells (Figures 3C–3E). The reduced inhibition outside the LPZ and the resulting changes in activity levels may trigger layer 2/3 cells to extend their axons along the gradient of reduced inhibition leading into the LPZ. These axons would thus provide novel inputs that facilitate functional reorganization after a retinal lesion. Together these data suggest a critical role for inhibitory structural changes in the initiation of circuit reorganization. All AZD2281 experimental procedures were carried out in compliance with the institutional guidelines of the Max Planck Society and the local government

(Regierung von Oberbayern). The left retinae of ketamine/xylazine anaesthetized adult mice were focally photocoagulated with multiple confluent lesions (300 μm, 500–600 mW, 200 ms, corresponding to 10–15 degrees vertically and 20–40 degrees horizontally of visual angle) through a laser-adapted operating microscope, as described previously (Keck et al., 2008). In a separate group of mice, both retinae were photocoagulated in their entirety by multiple confluent laser lesions, 300 μm, BLU9931 clinical trial 700–950 mW, 200 ms, directly aimed to and surrounding the optic disc in concentric circles to destroy all retinal ganglion cell fibers. Intrinsic imaging was used to determine the location of the LPZ following focal retinal lesions. Details of the imaging procedures and visual stimulation are described elsewhere (Mrsic-Flogel et al., 2005 and Schuett et al., 2002). Briefly, the visual cortex was illuminated with 707 nm light and images (600 ms in duration) were captured with a cooled

slow-scan CCD camera (ORA 2001, Optical Imaging, Rehovot, Israel), focused 200–300 μm below the cortical surface. During each 9 s stimulation trial, four blank frames were acquired, followed by visual stimulation, during which 11 frames were acquired. Visual stimuli consisted of square-shaped gratings (10–25°) presented MYO10 at 24 positions on a screen located in the contralateral visual hemifield. Retinotopic maps were computed as previously described (Keck et al., 2008, Mrsic-Flogel et al., 2005 and Schuett et al., 2002). We implanted cranial windows (Holtmaat et al., 2009) in ketamine/xylazine-anesthetized adult GAD65-GFP transgenic mice (age at surgery, 80–100 days), which express enhanced GFP under the GAD 65 promoter (López-Bendito et al., 2004). The skull overlying the right visual cortex was removed and replaced with a cover-glass window, leaving the dura intact. Animals recovered from surgery for at least 30 days before imaging started. We carried out two-photon imaging (Denk et al.