After 20 min of NMDA washout and reapplication of D-AP5, the I-V relationship was measured again (Figure 1A). We found that the +40mV EPSC was reduced by 55.0% ± 4.8% without

a significant corresponding reduction in the current at −60mV, which was only reduced by an average of 5.4% ± 3.6% (Figures 1B, 1F, and 1G; RG7204 mouse for nonleak subtracted currents, see Figure S1 available online). These changes in current amplitude are reflected in the increased rectification of the I-V relationship and a decrease in the RI to 0.21 ± 0.04 (Figures 1C and 1D; n = 20; p < 0.0001). The observed increase in rectification suggests that NMDAR activation causes a significant decrease in the proportion of synaptic CI-AMPARs. Moreover, the minimal change in amplitude at −60mV implies that there is a compensatory replacement with CP-AMPARs. NMDARs are ligand gated and voltage dependent, requiring both glutamate binding and depolarization to open the channel due to a channel block by Mg2+ ions. We determined whether postsynaptic NMDARs on RGCs were being directly activated with two controls. We either bath applied NMDA to voltage-clamped

cells (−60mV) without depolarizing the RGC or depolarized the RGC without NMDA application. In both cases, there was no change in rectification (Figure 1E; n = 6; percentage change from control RI = −4.7% ± 3.8%; n = 6; 3.5% ± 4.7%, respectively). Taken together, these changes show that direct postsynaptic NMDAR activation alters the AMPAR ratio of ON RGCs by selectively

replacing www.selleckchem.com/products/Vorinostat-saha.html synaptic CI-AMPARs with CP-AMPARs. We used a second, independent pharmacological approach to demonstrate NMDAR-driven changes in AMPAR subunit composition in ON RGCs. Philanthotoxin, (PhTX, 4 μM) a potent extracellular blocker of CP-AMPARs at negative membrane potentials (Bowie and Mayer, 1995; Kamboj et al., 1995; Koh et al., 1995), reduced the EPSC at −60mV to 67.5% ± 12.1% of the control response (Figures 1H and 1I; n = 7; p = 0.010). This reduction represents Thiamine-diphosphate kinase both a block of CP-AMPARs on the RGC itself and on upstream AII amacrine cells, which receive glutamatergic rod bipolar cell signals through CP-AMPARs (Ghosh et al., 2001; Singer and Diamond, 2003; Veruki et al., 2003). The philanthotoxin-resistant component of the light response was most likely carried by a separate pathway where rod signals pass into cones via a gap junction (Smith et al., 1986) and are conveyed to RGCs by ON cone bipolar cells, a pathway that does not utilize CP-AMPARs. Activation of NMDARs by bath application of NMDA paired with postsynaptic depolarization further reduced the EPSC to 28.8% ± 9.2% of the control response (p = 0.002 versus +PhTX, p = 0.0003 versus control), indicating that the response of CI-AMPARs was strongly reduced with NMDAR activation.