, 2010). To verify the role of Nrx in axonal changes, we applied a soluble form of Nrx fused to the Fc domain of human immunoglobulin G (Nrx1β-Fc) selleck products to the medium to block Cbln1-Nrx association (Matsuda

and Yuzaki, 2011; Uemura et al., 2010). Addition of Nrx1β-Fc (+S4) reduced the number of protrusions induced by GluD2-expressing HEK cells, whereas no effect was observed with Nrx1β-Fc (−S4), which does not inhibit Cbln1-Nrx interaction (Figures 4H and 4I). Taken together, these findings suggest that Cbln1-Nrx interaction is necessary for the PF structural change, which is directly induced by postsynaptic GluD2. To address whether PF structural changes occur during normal development in vivo, we examined PF morphology in immature cerebellum. GFP-encoding plasmids were introduced into the external granular layer (EGL) at P7 by electroporation in vivo (Konishi et al., 2004), and cerebella were fixed at various postnatal days. Most granule cells migrated into the internal granular layer by P10 when PFs were associated with few presynaptic structures (Figure 5A). Typical boutons started to appear by P14 (Figures 5A and 5E). At this time point, we confirmed the existence of PF protrusions, which were similar Y-27632 in vivo to

SPs and CPs observed in the slice culture (Figures 5A–C). These protrusions were closely apposed to the PC dendrites and GluD2 clusters (Figures 5A and 5C). We categorized protrusions and boutons according to the morphological criteria (Figure 1C). The density of PF protrusions peaked at P18 when the density of PF boutons was in the middle of

the growing phase (Figures 5D and 5E). PF protrusions decreased after P25 when the density of boutons plateaued. Thus, PF protrusions were found specifically before and during the peak of presynaptic bouton formation in vivo. To confirm that PF protrusions were associated with PC spines, we observed PFs by electron microscopy. PFs were labeled by injecting the neuron tracer biotin dextran amine (BDA), which can be visualized both by light and electron microscopy. To characterize individual protrusions, we utilized mice at P7 when there were very few protrusions. Matching Chlormezanone light and electron microscopic images was very difficult at later postnatal days because PF density had significantly increased. Despite with low incidence, some PF protrusions were found through careful observation by light microscopy (Figure 5F1). Successive observation by electron microscopy revealed that PF protrusions directly contacted to PC spines and occasionally encapsulated them (asterisk in Figure 5F2). In contrast, boutons were frequently observed in the BDA-labeled PFs of adult mice, but PF protrusions were not observed (Figure 5G1). Electron micrographs of the boutons revealed that they formed asymmetrical synapses with PC spines (Figure 5G2).