The method of cumulative EPSC amplitudes (Figure 5E) revealed a significantly smaller pool size in Robo3 cKO mice (8.6 ± 3.5 nA; n = 5) as compared to control mice (25.7 ± 4.8 nA; n = 7; p = 0.012). Furthermore, there was a significant reduction of release probability during the first EPSC,

as estimated by dividing the first EPSC amplitude by the pool size estimate (Figures 5E1 and 5E2). In summary, direct pre- and postsynaptic recordings at ipsilateral calyx of Held synapses indicate a more variable Ca2+ current density, a smaller size of the fast-releasable vesicle pool (FRP), and CH5424802 mouse a significant reduction of the initial release probability in Robo3 cKO mice (Figure 5). These data show that processes of synapse maturation, RG7204 order including the acquisition of fast transmitter release properties characteristic for the calyx of Held, fail to take place in Robo3 cKO mice. The finding of functional deficits at the calyx of Held synapses in Robo3 cKO mice suggests that axon midline crossing conditions the functional maturation of commissural output synapses. Alternatively, Robo3 could have a so far unknown direct role in synapse formation and synapse maturation. In a first series of experiments, we addressed this possibility by studying the developmental expression of Robo3, to verify whether Robo3 is expressed at the time of synaptogenesis

(Figures 6A–6C). In situ hybridization showed Robo3 expression at E14 in the developing VCN, but transcript levels in the VCN were essentially absent at E18 and undetectable

at P10 (Figure 6A). Using an anti-human Robo3 antibody which stained crossing hindbrain axons at E12.5 (Figure 6B), we next attempted to localize Robo3 protein in developing calyces of Held early postnatally, at P1 and P3 (Figure 6C), and at P5 and P8 (see Figure S1 available online). Robo3 was undetectable in developing calyx of Held axons, despite occasional non-specific signals (Figure 6C, arrow); the latter persisted in Robo3 cKO mice (Figure S1). This data indicates that at the calyx of SB-3CT Held projection, Robo3 expression is developmentally downregulated before E18, similar as at other commissural projections (Marillat et al., 2004; Sabatier et al., 2004; Tamada et al., 2008). The absence of Robo3 argues against a direct role of this protein in synapse development. To address a possible direct role of Robo3 in synapse development with an independent approach, we used a conditional KO approach with an inducible Cre mouse line, the CAGGS::CreERTM mouse line ( Guo et al., 2002; Livet et al., 2007). CAGGSCreERTM/+, Robo3lox/lox mice were injected at P0 with tamoxifen (referred to as Robo3 cKOTMX-P0 mice), in order to inactivate the floxed Robo3 allele following axon midline crossing, but before calyx of Held formation and – maturation.