Cell Type-Specific Whole-Genome Landscape of ΔFOSB Binding in the Nucleus Accumbens After Chronic Cocaine Exposure

Abstract

Background: The ability of neurons to respond to external stimuli involves adaptations of gene expression. Induction of the transcription factor ΔFOSB in the nucleus accumbens, a key brain reward region, is important for the development of drug addiction. However, a comprehensive map of ΔFOSB's gene targets has not yet been generated.

Methods: We used CUT&RUN (cleavage under targets and release using nuclease) to map the genome-wide changes in ΔFOSB binding in the 2 main types of nucleus accumbens neurons-D1 or D2 medium spiny neurons-after chronic cocaine exposure. To annotate genomic regions of ΔFOSB binding sites, we also examined the distributions of several histone modifications. Resulting datasets were leveraged for multiple bioinformatic analyses.

Results: The majority of ΔFOSB peaks occur outside promoter regions, including intergenic regions, and are surrounded by epigenetic marks indicative of active enhancers. BRG1, the core subunit of the SWI/SNF chromatin remodeling complex, overlaps with ΔFOSB peaks, a finding consistent with earlier studies of ΔFOSB's interacting proteins. Chronic cocaine use induces broad changes in ΔFOSB binding in both D1 and D2 nucleus accumbens medium spiny neurons of male and female mice. In addition, in silico analyses predict that ΔFOSB cooperatively regulates gene expression with homeobox and T-box transcription factors.

Conclusions: These novel findings uncover key elements of ΔFOSB's molecular mechanisms in transcriptional regulation at baseline and in response to chronic cocaine exposure. Further characterization of ΔFOSB's collaborative transcriptional and chromatin partners specifically in D1 and D2 medium spiny neurons will reveal a broader picture of the function of ΔFOSB and the molecular basis of drug addiction.

Keywords: Addiction; CUT&RUN; ChIP-sequencing; Chromatin remodeling complex; Histone modifications; Nucleus accumbens; Transcription factor.

Figure 1.. Establishment of CUT&RUN for ΔFOSB in NAc.

(A) Motif analysis of called ΔFOSB peaks in NAc of male mice. (B) Venn diagram analysis of called ΔFOSB peaks after repeated (7 d) cocaine IP injections or chronic (10 d) cocaine self-administration and their saline controls. (C) Upper heatmap shows genomic regions where ΔFOSB binding is induced by repeated IP injections of cocaine vs. saline, and lower heatmap compares the effect of cocaine self-administration (SA) vs. IP injections. (D) Genomic distribution of ΔFOSB peaks in the NAc. (E) RNA and protein analysis of dorsal striatum from Fosb-null male mice. (F) Heatmap of ΔFOSB coverage at called ΔFOSB binding sites between control and Fosb-null groups. CUT&RUN-seq: 2 animals per replicate and n=3 per group

Figure 2.. CUT&RUN for ΔFOSB in NAc D1 MSNs and D2 MSNs after chronic cocaine exposure.

(A) Motif analysis of called ΔFOSB peaks. (B) Venn diagram analysis of called ΔFOSB peaks among saline- and cocaine-exposed male and female mice. (C) Genomic distribution of ΔFOSB peaks in NAc D1 and D2 MSNs of male and female mice under saline and cocaine conditions. (D) Heatmaps comparing the effect of chronic cocaine on ΔFOSB peaks in D1 vs. D2 MSNs in male (upper) and female (lower) mice. (E) Venn diagrams comparing the effect of chronic cocaine on ΔFosB peaks in D1 and in D2 MSNs in male vs. female mice. (F) Example of ΔFOSB binding at a promoter region of Setd1a in the male D1 cocaine dataset. (G) Example of ΔFOSB binding at a poised enhancer region of Efr3a in the male D1 cocaine dataset. (H) Example of ΔFOSB binding at an active enhancer region in the Drd2-Ankk1-Ttc12 locus in the male D2 cocaine dataset. CUT&RUN-seq: 5 animals per replicate and n=3 per group

Figure 3.. RRHO reveals convergence of ΔFOSB binding and gene expression.

(A) Promoter regions. (B) Gene body regions. Changes in ΔFOSB binding in response to chronic IP cocaine (Coc) vs. saline (Sal) injections were compared to cocaine-induced changes in gene expression in D1 and D2 MSNs using published datasets. Positive correlations were seen between ΔFOSB binding and D1 MSN gene expression in mice that received a priming dose of cocaine after a history of chronic drug exposure (Coc-Coc), with negative correlations seen prior to that priming dose (Coc-Sal). Opposite although weaker correlations were apparent in D2 MSNs. Patterns seen in promoter and gene body regions were similar, with larger magnitude effects in the latter.

Figure 4.. A majority (~70%) of ΔFOSB binding sites overlap with BRG1 and with histone marks indicative of active enhancers.

(A) Venn diagram analysis of called H3K4me1, H3K27ac, and ΔFOSB peaks in male D1 MSNs of NAc from cocaine mice. (B) Coverage of BRG1 and histone marks at called ΔFOSB+ / H3K4me1+ / H3K27ac+ peaks in the male D1 MSN cocaine group. (C) Venn diagram analysis of called H3K4me3 and ΔFOSB peaks in the male D1 MSN cocaine group. (D) Coverage of histone marks at called ΔFOSB+ / H3K4me3+ peaks in the male D1 MSN cocaine group. CUT&RUN-seq: 4 animals per replicate and n=3 per group

Figure 5.. Loci with ΔFOSB binding sites contain motifs recognized by homeobox and T-box transcription factors.

(A) In silico motif analyses of loci with ΔFOSB-bound AP1 sites (± 500 bp from AP1 sites) in D1 and D2 MSNs of male and female cocaine groups. (B) Venn diagram analysis of alternative splicing junctions and ΔFOSB binding sites. (C) Motif analysis of alternative splicing junctions with ΔFOSB peaks nearby.