1 pA, 12.1 Hz ± 0.8 Hz, n = 7; RU486+CORT: 32.4 ± 4.9 pA, 11.3 ± 0.98 Hz, n = 9, p > 0.05) but not the MR antagonist RU28318 (10 μM, RU28318: 33.3 pA ± 4.7 pA,

11.8 Hz ± 1.3 Hz, n = 7; RU28318+CORT: 22.9 pA ± 1.4 pA, 7.4 Hz ± 1.4 Hz, n = 9, p < 0.05), suggesting that GR mediates the effect of chronic CORT treatment. To test whether the CORT-induced reduction of mEPSC was due to the decreased number of AMPARs at synapses, we performed immunocytochemical experiments to measure the cluster density (# clusters/50 μm dendrite) of total GluR1 and synaptic GluR1 (colocalized with the synaptic marker PSD-95) in PFC cultures. As shown in Figures 4E and 4F, CORT treatment (100 nM, 7 day) significantly reduced total GluR1 cluster density INCB024360 cost (control: 26.6 ± 3.1, n = 14; CORT: 15.6 ± Apoptosis Compound Library price 1.3, n = 12, p < 0.01) and synaptic GluR1 cluster density (control: 14.0 ± 1.0, n = 11; CORT: 7.8 ± 0.7, n = 12, p < 0.01). Taken together, these results suggest that, similar to in vivo repeated stress, prolonged in vitro CORT treatment also reduces AMPAR expression and function through GR activation. Since the total level of NR1 and GluR1 was reduced in repeatedly stressed animals, we examined whether it could be due to the decreased synthesis or increased degradation of glutamate receptors. As shown in Figure S4, repeated stress

did not significantly alter the mRNA level of AMPAR and NMDAR subunits, suggesting that protein synthesis is intact. PDK4 Thus, the reducing effect of repeated stress on NR1 and GluR1 expression may be due to the increased ubiquitin/proteasome-dependent

protein degradation. Consistent with this, the level of ubiquitinated GluR1 and NR1 was significantly increased in animals exposed to repeated restraint stress (Figures 5A and 5B, Ub-GluR1: 121.6% ± 28.3% increase, Ub-NR1: 135.9% ± 35.6% increase, n = 6 pairs, p < 0.01), which was abolished by RU486 injection (n = 3). The level of ubiquitinated GluR2, NR2A, or NR2B subunits remained unchanged (n = 4 pairs, Figure 5C). Repeated stress also failed to alter the ubiquitination of SAP97 (a GluR1 binding protein) and PSD-95 (an NR1 binding protein, n = 3 pairs, Figure 5C). These results provide direct evidence showing that prolonged GR activation selectively increases ubiquitin conjugation of GluR1 and NR1 subunits in PFC and thus enhances the susceptibility of these proteins to proteasome-mediated degradation. To further test the role of glutamate receptor degradation in chronic stress-induced reduction of synaptic transmission, we injected the proteasome inhibitor MG132 into PFC via an implanted cannula (0.5 μg each side; 21 pmol/g body weight, daily at 1 hr before stress). As shown in Figures 6A and 6B, the effects of repeated restraint stress on glutamatergic transmission were significantly different in saline- versus MG132-injected animals (AMPA: p < 0.