Fig. 2: Exogenous NMDA-induced presynaptic depression of C-eEPSCs in the basal state, but presynaptic potentiation in the inflamed state.

a, b Typical examples (a) and quantitative summary (b) of synaptic depression of C-eEPSCs induced by acute bath application of NMDA (50 μM, 30 s) in NR1^fl/fl mice, but not in SNS-NR1^−/− mice in the basal state (n = 6–7). ***P < 0.001 by Kruskal–Wallis H test. c, d Time course (c) and quantitative summary (d) showing pre-LTD of C-eEPSCs evoked by long-term exposure to NMDA (50 μM, 10 min) in NR1^fl/fl mice, but not in SNS-NR1^−/− mice in the basal state (n = 9–10). ***P < 0.001 by unpaired t test. e, f Following CFA-evoked paw inflammation, acute bath application of NMDA led to presynaptic potentiation of C-eEPSCs in NR1^fl/fl mice, but not in SNS-NR1^−/− mice (representative examples are shown in e, quantitative summary shown in (f), n = 5–7). *P < 0.05 by Kruskal–Wallis H test. g, h Typical traces of recordings (g) and quantitative summary (h) showing long-term exposure of NMDA (50 μM, 10 min) evoked pre-LTP of C-eEPSCs in CFA-inflamed NR1^fl/fl mice, but not in inflamed SNS-NR1^−/− mice (n = 5–10). *P < 0.05 by unpaired t test. i, j Time course (i) and representative recording traces (j) as well as quantitative summary (k) showing the transition of NMDA-induced synaptic depression to potentiation following LFS-induced pre-LTP in the same spinal-PAG projection neuron (n = 12). MK-801 was present in the patch pipette to block postsynaptic NMDARs in the above recordings. *P < 0.05, **P < 0.01 by Kruskal–Wallis H test. Data are represented as mean ± S.E.M. See Supplemental Table 2 for detailed statistical information.