06 ± 0.17, p = 0.81; 5 min: 1.04 ± 0.28 versus 1.07 ± 0.18, p = 0.93; 10 min: 1.04 ± 0.09 versus 0.97 ± 0.09, p = 0.64). These findings indicate that, in TSPAN7 absence, AMPAR internalization is increased. Given the uniform effects of TSPAN7 knockdown on GluA2 internalization over the 10 min period, in successive experiments (Figure 8), a single incubation period of 5 min was used. In the first set of experiments (Figure 8A), we further characterized TSPAN7′s effect on GluA2 trafficking. We checked Selleck MG-132 the specificity of TSPAN7 knockdown on GluA2 internalization by expressing siRNA14 alone or together with rescue WT. Rescue WT fully restored GluA2 internalization to control

levels (EGFP: 1.00 ± 0.03, siRNA14: 1.21 ± 0.09, ∗p = 0.01, rescue WT: 1.01 ± 0.04, p = 0.86; values normalized to EGFP). However, when siRNA14 was expressed with rescue ΔC, Selleck Ribociclib GluA2 internalization

was not restored to control levels (rescue ΔC 1.19 ± 0.07 ∗∗p = 0.008) (Figure 8A). We next investigated TSPAN7 overexpression, finding it had opposite effects to TSPAN7 knockdown: reduced GluA2 internalization compared to control (EGFP: 1.00 ± 0.03, TSPAN7: 0.70 ± 0.05, ∗∗∗p < 0.001, values normalized to EGFP). By contrast, TSPAN7ΔC overexpression had no effect on GluA2 internalization (TSPAN7ΔC: 0.91 ± 0.06 relative to EGFP, p = 0.17), clearly showing that the TSPAN7 C terminus is involved in regulating AMPAR trafficking (Figure 8A). In the next set of experiments (Figures 8B–8D), we investigated the combined influence of TSPAN7 and PICK1 on GluA2 trafficking, by directly manipulating expression of the two proteins. We knocked down PICK1 using a previously

characterized siRNA (siPICK1) (Citri et al., 2010). As expected, PICK1 silencing decreased GluA2 internalization relative to EGFP. When siPICK1 was coexpressed with siRNA14, GluA2 internalization was reduced as effectively as with siPICK1 alone, fully preventing the increase expected with TSPAN7 knockdown (Figures 8B and 8D, EGFP: 1.00 ± 0.04, siPICK1: 0.85 ± 0.01, ∗p = 0.02, siPICK1+siRNA14: 0.77 ± 0.07, ∗∗p = 0.006, values normalized to EGFP). Next, we overexpressed PICK1 (myc tagged) either alone or with TSPAN7 (pIRES-EGFP-TSPAN7). Neurons overexpressing only PICK1 had until greater GluA2 internalization than EGFP controls, consistent with findings showing that PICK1 overexpression decreases GluA2 surface levels (Terashima et al., 2004). When PICK1 and TSPAN7 were overexpressed together, PICK1 prevented the decrease in GluA2 internalization expected with TSPAN7 overexpression (Figures 8C and 8D, EGFP: 1.00 ± 0.04, PICK1: 1.26 ± 0.04, ∗∗∗p < 0.001, PICK1+TSPAN7: 1.29 ± 0.05, ∗∗∗p < 0.001 Tukey after ANOVA). These findings lead us to suggest a model whereby expression levels of TSPAN7 regulate PICK1-mediated AMPAR trafficking, possibly because TSPAN7 competes with AMPARs for PICK1 binding (Figure 6E) at the PDZ domain (Figures 6A–6D) (Dev et al., 1999 and Xia et al., 1999).