The suppressive effect of an intra-prefrontal cortical infusion of BDNF on cocaine-seeking is Trk receptor and extracellular signal-regulated protein kinase mitogen-activated protein kinase dependent

Abstract

Cocaine-mediated neuroadaptations in the prefrontal cortical-nucleus accumbens pathway underlie drug-seeking in animals with a cocaine self-administration (SA) history. Neuroplasticity in the cortico-accumbens pathway is regulated, in part, by the expression and availability of neurotrophic factors, such as BDNF. We have previously demonstrated that infusion of BDNF into the dorsomedial prefrontal cortex (dmPFC) immediately after the last of 10 cocaine SA sessions attenuates contextual, cue- and cocaine prime-induced reinstatement of cocaine-seeking (Berglind et al., 2007) and normalizes cocaine-induced disruption of glutamatergic transmission in the nucleus accumbens (Berglind et al., 2009). In the present study, the suppressive effect of intra-dmPFC BDNF on cocaine-seeking is shown to depend on Trk receptor-mediated activation of extracellular signal-regulated kinase (ERK) signaling in the dmPFC. The tyrosine kinase inhibitor, K252a, and the mitogen-activated protein/extracellular signal-regulated kinase kinase inhibitor, U0126 (1,4-diamino-2,3-dicyano-1,4-bis[2-aminophenylthio]butadiene), prevented BDNF's suppressive effects on cocaine-seeking. Vehicle-infused rats with a cocaine SA history showed significant decreases in ERK and cyclic AMP response element binding protein (CREB), but not Akt, phosphorylation after the final cocaine SA session that were reversed by intra-dmPFC BDNF. Additionally, BDNF's ability to normalize cocaine-mediated decreases in ERK and CREB phosphorylation was blocked by U0126, demonstrating that ERK/MAPK activation mediated the behavioral effects. This study elucidates a mechanism whereby BDNF/TrkB (tropomyosin receptor kinase B) activates ERK-regulated CREB phosphorylation in the dmPFC to counteract the neuroadaptations induced by cocaine SA and subsequent relapse to cocaine-seeking.

Figure 1.

Schematic representation of the experimental paradigms used in experiments 1–4.

Figure 2.

Graphical representation of histological placement from experiments 1–4. A, In experiment 1, bilateral cannula placements within the anterior cingulate and prelimbic regions (dmPFC) are represented by black dots. Rats with tract placements that were not within the dmPFC were excluded from the schematic maps and the subjects' data (n = 2) were excluded from the data analysis. B, Rats from experiments 2, 3, and 4 were decapitated and cannula placement within the dmPFC was verified during dissection.

Figure 3.

K252a and U0126 blocked BDNF's suppressive effect on cocaine-seeking. A, Post-abstinence test. Left, Before intra-dmPFC infusions, all rats show similar responding during the last 3 d of cocaine SA. On the final day of cocaine SA, rats received one of the following treatments: DMSO vehicle followed by PBS (V-P, black bar), DMSO vehicle followed by BDNF (V-B, black bar with grid pattern), K252a followed by PBS (K-P, white bar), K252a followed by BDNF (K-B, white bar with grid pattern), U0126 followed by PBS (U-P, gray bar), or U0126 followed by BDNF (U-B, gray bar with grid pattern). Right, Rats receiving V-B treatment showed significantly less cocaine seeking during the test compared with V-P controls (***p < 0.001). All other treatment groups demonstrated significantly greater cocaine-seeking in the post-abstinence test compared with cocaine SA responding (⁺p < 0.05, ⁺⁺p < 0.01, ⁺⁺⁺p < 0.001). B, Cue-induced reinstatement test. Left, All rats extinguished lever pressing and reached a common baseline of responding. Right, Rats receiving V-B treatment showed significantly less cocaine seeking during the cue test than V-P controls (**p < 0.01). All treatment groups demonstrated greater cocaine-seeking in the cue test compared with extinction responding (⁺p < 0.05, ⁺⁺p < 0.01, ⁺⁺⁺p < 0.001). C, Cocaine prime-induced reinstatement test. Left, All rats were brought back to a common baseline of responding during extinction. Right, Rats receiving V-B treatment showed significantly less cocaine seeking during the cocaine prime test than V-P controls (*p < 0.05). All treatment groups demonstrated greater cocaine-seeking in the cocaine prime test compared with extinction responding (⁺p < 0.05, ⁺⁺⁺p < 0.01). n = 9–16 per group.

Figure 4.

Intra-dmPFC BDNF activates phosphoprotein signaling in the dmPFC of drug-naive rats. A, p-ERK expression was significantly increased in the dmPFC 0.5 h (black bar with grid pattern) and 2 h (white bar with grid pattern), but not 6 h (gray bar with grid pattern), postinfusion in BDNF-infused rats (*p < 0.05) compared with PBS-infused control rats (solid black bar for 0.5 h, white bar for 2 h, and gray bar for 6 h). Above graph, Representative Western blot for p-ERK and total (t)-ERK expression. B, p-CREB expression was significantly increased in the dmPFC 0.5 h (black bar with grid pattern), 2 h (white bar with grid pattern), and 6 h (gray bar with grid pattern) post-BDNF infusion (*p < 0.05) compared with PBS-infused rats (solid black bar for 0.5 h, white bar for 2 h, and gray bar for 6 h). Above graph, Representative Western blot for p-CREB and t-CREB expression. C, p-Akt expression was significantly increased in the dmPFC 2 h (white bar with grid pattern), but not 0.5 h (black bar with grid pattern) or 6 h (gray bar with grid pattern) post-BDNF infusion (*p < 0.05) compared with PBS-infused rats (solid black bar for 0.5 h, white bar for 2 h, and gray bar for 6 h). Above graph, Representative Western blot for p-Akt and t-Akt expression.

Figure 5.

Intra-dmPFC BDNF counteracts cocaine-mediated decreases in p-ERK and p-CREB, but not p-Akt, expression 2 h postinfusion, and this normalization of cocaine-mediated decreases in p-ERK and p-CREB expression is blocked by U0126. A, p-ERK expression was suppressed in the cocaine SA group compared with the yoked-saline group (*p < 0.05 C-P vs S-P). Intra-dmPFC BDNF infusion significantly increased p-ERK expression in the yoked-saline group (**p < 0.01 S-B vs S-P) and normalized p-ERK expression in the cocaine SA (⁺⁺⁺p < 0.05 C-B vs C-P) group. Above graph, Representative Western blot for p-ERK and t-ERK expression. B, p-CREB expression was suppressed in the cocaine SA group compared with the yoked-saline group (***p < 0.001 C-P vs S-P). Intra-dmPFC BDNF infusion significantly increased p-CREB expression in the yoked-saline group (*p < 0.05 S-B vs S-P) and normalized p-CREB expression in the cocaine SA (⁺⁺⁺p < 0.01 C-B vs C-P) group. Above graph, Representative Western blot for p-CREB and t-CREB expression. C, Intra-dmPFC BDNF infusion significantly increased p-Akt expression in the yoked-saline group (***p < 0.001 S-B vs S-P) and in the cocaine SA (***p < 0.001 C-B vs C-P) group. Above graph, Representative Western blot for p-Akt and total Akt expression. D, p-ERK expression was suppressed in both cocaine SA groups compared with the yoked-saline group (***p < 0.001 S/U-P vs C/U-P and C/U-B). Intra-dmPFC BDNF infusion failed to increase p-ERK expression in the presence of U0126 for yoked-saline (S/U-B and C/U-B) or cocaine SA rats. Above graph: Representative Western blot for p-ERK and t-ERK expression. E, p-CREB expression was suppressed in both cocaine SA groups compared with the yoked-saline group (**p < 0.01 S/U-P vs C/U-P and C/U-B). Intra-dmPFC BDNF infusion failed to increase p-CREB expression in the presence of U0126 for yoked-saline or cocaine SA rats. Above graph, Representative Western blot for p-CREB and t-CREB expression. F, Intra-dmPFC BDNF infusion significantly increased p-Akt expression in the yoked-saline group (**p < 0.01 S/U-B vs S/U-P) and in the cocaine SA (**p < 0.01 C/U-B vs C/U-P) groups regardless of U0126 infusion. Above graph, Representative Western blot for p-Akt and total Akt expression.