Fig 2: Pre-treatment with ceftriaxone did not increase GLT1 protein expression in neonatal rat brain tissue. A, GLT1 protein expression in P7 rat brain tissue following pre-treatment with different dosages of ceftriaxone and saline; B, statistic analysis showed no difference in GLT1 protein expression among these groups. one-way ANOVA, p = 0.95 (n = 5 in each group). Saline group, animals received hypoxic-ischemic procedures and given saline injection.

Fig 3: Chronic stress exposure induced a blunted glutamate response after a cocaine i. p. injection and a downregulation of GLT-1 in the NAcore. (A) Schematic timeline of the experimental protocol. (B,C) Glutamate levels in the NAcore and NAshell respectively. Baseline: measurements were made during 120 min. Saline was administered at 120 min, while cocaine (15 mg/kg i.p.) was injected at 240 min. NAcore: Two-way ANOVA with repeated measures over time showed significant interaction (Non-CRS vs. CRS x time; F(12,156) = 5.04, p < 0.0001), significant effect of stress (Non-CRS vs. CRS; F(1,13) = 12.47, p < 0.01) and time (F(12,156) = 7.74, p < 0.0001). Bonferroni post-hoc test revealed significant differences between Non-CRS and CRS after a cocaine i. p. injection at 330 (**p < 0.01), 360 and 390 min (****p < 0.0001). No differences were found comparing Non-CRS with CRS after a saline i. p. injection (n.s.). The effect of cocaine on glutamate efflux was evaluated in each experimental group comparing results at 240 min (last data after a saline i. p. injection) and 270 min (first data in response to a cocaine challenge). Two-tailed unpaired t-test revealed an acute effect of cocaine on glutamate release (Non-CRS: t14 = 2.29; & p < 0.05–CRS: t12 = 2.26; # p < 0.05). NAshell: Two-way ANOVA with repeated measures over time revealed no interaction (Non-CRS vs. CRS x time; F(12,120) = 0.69, n. s.) and no effect of stress (Non-CRS vs. CRS; F(1,10) = 0.29, n.s.) but a significant main effect of time (F(12,120) = 4.41, p < 0.0001). No differences were found comparing Non-CRS with CRS after a saline i.p. injection (n.s.). The effect of cocaine on glutamate efflux was evaluated in each experimental group comparing data at 240 min and 270 min. Two-tailed unpaired t-test revealed an acute effect of cocaine on glutamate release (Non-CRS: t10 = 3.34; & p < 0.05–CRS: t10 = 2.45; # p < 0.05). Two-tailed unpaired t-test (Non-CRS vs. CRS) revealed that baseline glutamate levels were not significantly different between experimental groups either in the NAcore (t13 = 1.74; p = 0.11) or NAshell (t10 = 0.15; p = 0.89). However, in the NAcore, but not the NAshell, a marked trend was observed (Baseline glutamate levels–NAcore: Non-CRS = 24.63 ± 10.36 pmol/10 µl, CRS = 76.29 ± 29.61 pmol/10 µl; NAshell: Non-CRS = 34.86 ± 15.31 pmol/10 µl; CRS = 37.27 ± 5.95 pmol/10 µl). (D) Illustration of the location of the active membrane of the dialysis probe in the NAcore and NAshell according to Paxinos and Watson (2007). Dashed lines represent probe placements in the Non-CRS group, and solid lines those in the CRS group. (E) Schematic timeline of the experimental protocol. (F) GLT-1 protein content was reduced in the NAcore by chronic stress exposure. Two-way ANOVA revealed a significant main effect of stress (Non-CRS vs. CRS; F(1,28) = 5.80, *p < 0.05). There was no significant effect of drug (Sal vs. Coc; F(1,28) = 0.66, n.s.) or interaction (Non-CRS vs. CRS x Sal vs. Coc; F(1,28) = 0.82, n.s.). N is shown in brackets or bars. All data are shown as mean ± SEM. (CRS = chronic restraint stress, Sal, saline; Coc, cocaine).

Fig 4: Ceftriaxone restored the CRS-induced GLT-1 downregulation and increased basal concentrations of extracellular glutamate in the NAcore. (A) Schematic timeline of the experimental protocol. (B) Fitted regression lines of no-net-flux microdialysis in the NAcore (slopes–Non-CRS/Veh: 0.82 ± 0.07, CRS/Veh: 1.01 ± 0.10, Non-CRS/Cef: 0.99 ± 0.12 and CRS/Cef: 1.01 ± 0.08). Linear regression analysis showed non-significant differences among the slopes of the regression line of no-net-flux plot (F(3,108) = 0.83; p = 0.48). (C) Data obtained from extrapolation from the regression line of each subject belonging each experimental group with x axis showed the following basal extracellular glutamate concentrations: Non-CRS/Veh (2.46 ± 0.41) µM, CRS/Veh (5.27 ± 0.63) µM, Non-CRS/Cef (2.62 ± 0.42) µM and CRS/Cef (2.87 ± 0.34) µM. Two-way ANOVA revealed significant interaction (Non-CRS vs. CRS x Veh vs. Cef; F(1,25) = 6.28, p < 0.05), significant effect of stress (Non-CRS vs. CRS; F(1,25) = 8.90, p < 0.01) and ceftriaxone treatment (Veh vs. Cef; F(1,25) = 4.75, p < 0.05). Bonferroni post-hoc test showed significant differences comparing Non-CRS/Veh with CRS/Veh (**p < 0.01) and CRS/Veh with CRS/Cef (+ p < 0.05). Ceftriaxone treatment did not evidence a per se effect on basal extracellular glutamate concentrations as shown comparing Non-CRS/Cef with CRS/Veh (°°p < 0.01) and Non-CRS/Cef with Non-CRS/Veh (n.s.). (D) Illustration of the location of the active membrane of the dialysis probe in the NAcore and NAshell according to Paxinos and Watson (2007). Dashed lines represent probe placements in the Non-CRS group, solid lines those in the CRS group, lines shown in the left hemisphere correspond to the Veh group while lines in the right hemisphere correspond to the Cef group. (E) Schematic timeline of the experimental protocol. (F) Influence of ceftriaxone treatment on NAcore GLT-1 content. Two-way ANOVA revealed significant interaction (Non-CRS vs. CRS x Veh vs. Cef; F(1,28) = 9.56, p < 0.01). There were no significant effects of stress (Non-CRS vs. CRS; F(1,28) = 3.32, n.s.) or treatment (Veh vs. Cef; F(1,28) = 1.81, n.s.). Bonferroni post-hoc test showed significant differences comparing Non-CRS/Veh with CRS/Veh (*p < 0.05) and CRS/Veh with CRS/Cef (+p < 0.05). Ceftriaxone treatment did not evidence a per se effect on basal extracellular glutamate concentrations as shown comparing Non-CRS/Cef with Non-CRS/Veh (n.s.). Data are shown as mean ± SEM. N is shown in brackets. (CRS, chronic restraint stress; Veh, vehicle; Cef, ceftriaxone; Sal, saline; Coc, cocaine).

Fig 5: ß-Catenin partners with TCF-1 to positively regulate EAAT2 at the level of transcription in PDAs. PDAs were transfected with ß-catenin siRNA or Scrm-Si, and 48 h post transfection, cells were transfected with an EAAT2 promoter linked to luciferase, and 24 h later, luciferase activity was measured (a). b PDAs were transfected with pABC or pcDNA, and 24 h post transfection, cells were transfected with EAAT2 promoter linked to luciferase, and 24 h later, luciferase activity was measured. c Identification of putative TCF/LEF binding sites on the EAAT2 promoter. Shown in c is a ~2.5-kb promoter of EAAT2 illustrating TCF/LEF binding sites present in clusters and the regions selected for designing primers for real-time PCR. PDAs were transfected with siRNAs to knockdown TCF-1, TCF-3, TCF-4, LEF1, or scrambled siRNA (d, e), and 72 h post transfection, real-time qPCR was performed to measure the levels of EAAT2 mRNA (d) and protein by WB (e). PDAs (~2 × 106 cells) were cross-linked with 1.0 % formaldehyde and sonicated to generate chromatin preparations (f–h). ChIP was performed using antibodies against ß-catenin (f, g) or TCF-1 (h) (5 µg antibody/IP). PDAs were transfected with EAAT2-prom, and 24 h later, ChIP was run (h). Data shown are normalized to a nontargeting IgG control and reported as a fold change relative to IgG. Asterisks indicate a p < 0.05 relative to the respective control using Student’s t test. Error bars indicate SD