One of these lines, B6;FVB-Tg(CAG-boNT/B,EGFP)U75-56Fwp/J (JAX Stock No. 018056), subsequently referred to as the iBot line, was retained for further analysis and validated in two ways. First, iBot mice were crossed to a transgenic line, where Cre is expressed in all cells (Tg(CMV-Cre); Dupé et al., 1997). Ubiquitous induction of the toxin should phenocopy the perinatal lethality of VAMP2 knockout mice (Schoch et al., 2001). Indeed, bigenic mice were born at a significantly lower rate compared to monogenic and wild-type littermates and notably, all bigenic offspring died

within 1 day after birth, whereas perinatal death rates among control mice were significantly lower (Figure 1B). Second, we tested Doxorubicin whether neuronal expression of BoNT/B inhibits

SNARE-dependent synaptic VE-822 cost transmission (Poulain et al., 1988). To this end, iBot mice were crossed with a line, in which the prion promoter targets an inducible version of Cre (CreERT) to cerebellar neurons (line 28.6, Tg(Prnp-CreERT); Weber et al., 2001). Whole-cell patch-clamp recordings from Purkinje cells revealed a strong reduction of synaptic responses to parallel fiber stimulation in Tamoxifen (Tam)-injected bigenic mice compared to Tam-injected monogenic littermates (Figure 1C). Immunohistochemical staining revealed the presence of EGFP, which is coexpressed with the toxin, in the cerebellar granular and molecular layer of bigenic mice (Figure 1D). Together, Montelukast Sodium these results proved that the iBot

line allows for cell-specific block of SNARE-dependent exocytosis. Next, we induced BoNT/B expression in retinal glia by crossing iBot mice with the Tg(Glast-CreERT2)T45-72 line, which expresses Tam-dependent Cre recombinase in Müller cells (Slezak et al., 2007). Since the toxin cannot be visualized, we used immunohistochemical staining for EGFP to detect the transgene in Müller cells. As shown in Figure 2A, adult Tam-injected bigenic mice showed EGFP expression in cellular retinaldehyde binding protein (CRALBP)-positive Müller cells, whereas monogenic littermates showed no EGFP staining. The targeting efficacy was on average 55% ± 8% EGFP-positive cells among all CRALBP-positive Müller cells (n = 4 bigenic mice). The limited sensitivity of immunohistochemical detection may underestimate the targeting efficacy. To test whether the toxin is active, we determined retinal VAMP2 levels by surface plasmon resonance (Ferracci et al., 2005). In retinal lysates from Tam-injected bigenic mice, levels of VAMP2 normalized to synaptophysin (SYP) content were significantly decreased compared to vehicle-injected bigenic mice or Tam-injected monogenic or wild-type littermates (Figure 2B) indicating proteolyic activity of the toxin in retinae from Tam-injected bigenic mice. To address whether BoNT/B expression blocks calcium-dependent exocytosis from Müller cells, we established an assay to measure calcium-induced glutamate release from acutely isolated cells.