Epigenetic Regulation of Hippocampal Fosb Expression Controls Behavioral Responses to Cocaine

Abstract

Drug addiction results in part from maladaptive learning, including the formation of strong associations between the drug and the circumstances of consumption. However, drug-induced changes in gene expression underlying the saliency of these associations remain understudied. Consolidation of explicit memories occurs within the hippocampus, and we have shown that spatial learning induces expression of the transcription factor ΔFosB in hippocampus and that this induction is critical for learning. Drugs of abuse also upregulate ΔFosB in hippocampus, but the mechanism of its induction by cocaine and its role in hippocampus-dependent cocaine responses is unknown. We investigated differences in mouse dorsal and ventral hippocampal ΔFosB expression in response to chronic cocaine, because these regions appear to regulate distinct cocaine-related behaviors. We found that cocaine-mediated induction of ΔFosB was subregion-specific, and that ΔFosB transcriptional activity in both the dorsal and ventral hippocampus is necessary for cocaine conditioned place preference. Further, we characterize changes in histone modifications at the FosB promoter in hippocampus in response to chronic cocaine and found that locus-specific epigenetic modification is essential for FosB induction and multiple hippocampus-dependent behaviors, including cocaine place preference. Collectively, these findings suggest that exposure to cocaine induces histone modification at the hippocampal FosB gene promoter to cause ΔFosB induction critical for cocaine-related learning.SIGNIFICANCE STATEMENT Although cocaine addiction is driven in part by the formation of indelible associations between the drug and the environment, paraphernalia, and circumstances of use, and although this type of associative learning is dependent upon changes in gene expression in a brain region called the hippocampus, the mechanisms by which cocaine alters hippocampal gene expression to drive formation of these associations is poorly understood. Here, we demonstrate that chronic cocaine engages locus-specific changes in the epigenetic profile of the FosB gene in the hippocampus, and that these alterations are required for cocaine-dependent gene expression and cocaine-environment associations. This work provides novel insight into addiction etiology and potential inroads for therapeutic intervention in cocaine addiction.

Keywords: cocaine; epigenetics; hippocampus; histone; transcription; ΔFosB.

Figure 1.

Cocaine effects on hippocampal expression of FosB isoforms. A, Mice were killed 1 h after the last exposure to saline or cocaine paired with a novel environment. qPCR reveals cocaine-dependent increases in the transcript levels for FosB and ΔFosB in whole dorsal hippocampus (dHPC). B, Ventral hippocampus was microdissected to isolate the vDG, which displayed no changes in full-length FosB or ΔFosB mRNA. C, Specific microdissection of vCA1 revealed a trend toward an increase in both FosB and ΔFosB mRNA. D, dHPC displayed no changes in either FosB or ΔFosB protein. E, F, vDG did not show any significant changes in FosB or ΔFosB (E), and vCA1 showed significant induction of both FosB and ΔFosB (F). At the bottom of each graph are representative images from Western blots showing FosB or ΔFosB, and GAPDH (n = 8/group; *p < 0.05, #p ≤ 0.06).

Figure 2.

Cocaine effects on FosB-isoform neuronal expression in dorsal and ventral hippocampal subregions. A, B, Representative 40× images of coronal sections of dorsal (A) and ventral (B) hippocampus stained for FosB isoforms (and enriched for ΔFosB) from mice given saline or cocaine in a novel environment. Arrowheads indicate examples of FosB-isoform+ cells. C, D, Cocaine increased FosB-isoform+ cells specifically in the CA1 subregion of dorsal (C) and ventral (D) hippocampus (n = 5 mice; *p < 0.05). E, Quantified FosB-isoform intensity per cell in dorsal hippocampus CA1 neurons after chronic cocaine exposure (n = 25–60 cells; *p < 0.05). Representative images show FosB-isoforms in red and DAPI in blue. Arrowhead indicates an example cell in which FosB isoforms are induced at high levels following chronic cocaine exposure.

Figure 3.

Transcriptional silencing of hippocampal ΔFosB impairs cocaine conditioned place preference. A, Example of dorsal hippocampal targeting of viral-mediated GFP expression (left) and injection sites of a representative cohort of dorsal hippocampal surgeries with AAV-GFP or AAV-ΔJunD (right). B, Example of ventral hippocampal targeting (left) and injection sites (right). C, Mice received ventral hippocampal infusions of AAV-GFP or AAV-ΔJunD. GFP mice spent significantly more time with the novel object than the familiar object, whereas the ΔJunD mice showed impaired novel object recognition. There was no difference between ventral hippocampus AAV-GFP or AAV-ΔJunD in CFC freezing behavior (D), center zone time in OF (E), or open arm time in EPM (F; n = 6–7; *p < 0.05). Mice were injected in dorsal (G) or ventral (H) hippocampus with HSV-GFP or HSV-ΔJunD and subjected to cocaine CPP (5 mg/kg, i.p.). Control mice formed a significant preference for the cocaine-paired chamber, which was blocked by HSV-ΔJunD in either hippocampal region (n = 8, *p < 0.05).

Figure 4.

Reduction of H3K9me2 at the FosB promoter is required for cocaine induction of ΔFosB in ventral hippocampus. A, Mice were injected with cocaine (20 mg/kg, i.p.) or saline and placed in a novel environment for 30 min/d for 10 d. Chromatin immunoprecipitation for H3K9me2 was conducted on microdissected hippocampus, which was taken 1 h after the last injection. Significant decreases in H3K9me2 enrichment were found at multiple sites along the FosB gene (n = 8; *p < 0.05). B, Diagram of histone posttranslational modifications at the FosB gene before and after cocaine exposure, and depicting locus-specific H3K9me2 deposition at the FosB promoter. Graphs (C) and representative images (D) of ventral hippocampus CA1 neurons in saline and cocaine-treated mice. ΔFosB (red) was significantly increased by cocaine in neurons of mice injected with HSV-GFP, but not HSV-FosB-ZFP-G9a (green; n = 16–30; *p < 0.05).

Figure 5.

Epigenetic remodeling by FosB-ZFP-G9a in dorsal and ventral hippocampus impairs general learning and memory. A, Dorsal hippocampus HSV-GFP mice spent significantly more time with the novel object than the familiar object, whereas FosB-ZFP-G9a mice spent a similar amount of time with both objects (n = 8–18; *p < 0.05). B, After CFC, FosB-ZFP-G9a mice spent the same amount of time freezing as GFP controls. C, There was no difference locomotor activity between groups. D, Short-term memory, as measured by time with the novel object only 4 h after familiarization, was not affected FosB-ZFP-G9a expression in dorsal hippocampus. E, Ventral hippocampus HSV-GFP mice spent significantly more time with the novel object than the familiar object, whereas FosB-ZFP-G9a mice spent a similar amount of time with both objects. There were no differences in CFC (E) or locomotor activity (F; n = 8–18; *p < 0.05). G, There was no difference locomotor activity between groups.

Figure 6.

Epigenetic remodeling by FosB-ZFP-G9a in dorsal and ventral hippocampus impairs cocaine CPP. Mice were injected in dorsal or ventral hippocampus with HSV-GFP or HSV-FosB-ZFP-G9a and were subjected to cocaine CPP (5 mg/kg, i.p.). Control mice formed a significant preference for the cocaine-paired chamber, which was blocked by injection of HSV-FosB-ZFP-G9a in dorsal (A) or ventral (B) hippocampus. Center time is also displayed (white bar), with no significant differences between the groups (n = 6–15; *p < 0.05).