Genome-wide identification and characterization of functional neuronal activity-dependent enhancers

Abstract

Experience-dependent gene transcription is required for nervous system development and function. However, the DNA regulatory elements that control this program of gene expression are not well defined. Here we characterize the enhancers that function across the genome to mediate activity-dependent transcription in mouse cortical neurons. We find that the subset of enhancers enriched for monomethylation of histone H3 Lys4 (H3K4me1) and binding of the transcriptional coactivator CREBBP (also called CBP) that shows increased acetylation of histone H3 Lys27 (H3K27ac) after membrane depolarization of cortical neurons functions to regulate activity-dependent transcription. A subset of these enhancers appears to require binding of FOS, which was previously thought to bind primarily to promoters. These findings suggest that FOS functions at enhancers to control activity-dependent gene programs that are critical for nervous system function and provide a resource of functional cis-regulatory elements that may give insight into the genetic variants that contribute to brain development and disease.

Figure 1. Genome-wide analysis of H3K27ac ChIP-seq peaks

a) UCSC genome browser tracks of the Fos locus with data for indicated histone modifications and DNaseI hypersensitivity mapping from p56 mouse cortex (generated by the Mouse ENCODE project). H3K4me1, H3K4me3, H3K27me3 RNA Pol II, CBP and RNA-seq tracks were generated previously. b) Genomic distribution of H3K27ac peaks before and after membrane depolarization. Peaks +/− 1 kb of a TSS are listed as proximal. c) Overlap of gene distal (> 1 kb from a TSS) H3K27ac peaks before and after membrane depolarization d) Aggregate plot of H3K4me1, H3K4me3, H3K27ac and H3K27me3 signal before and after membrane depolarization centered at gene-distal DHS sites enriched for H3K27ac.

Figure 2. H3K27ac dynamics at activity-regulated enhancers

a) H3K4me1 ChIP-seq signal at CBP/H3K4me1-enriched sites before and after 2 h membrane depolarization by KCl (ρ= 0.96, Spearman’s rank correlation coefficient). b) H3K27ac ChIP-seq signal at CBP/H3K4me1-enriched sites before and after 2 h membrane depolarization by KCl (ρ = 0.79, Spearman’s rank correlation coefficient). c) Classification of putative activity-regulated enhancers by distinct H3K27ac dynamics in response to neuronal activity (see methods). d) Representative loci demonstrating distinct H3K27ac dynamics at enhancers in response to neuronal activity.

Figure 3. Functional analysis of enhancers with distinct H3K27ac dynamics

a) Functional testing of enhancers from distinct H3K27ac classes in cultured DIV7 cortical neurons. All results are displayed as a fold-increase over the Nptx2 reporter backbone without the Nptx2 enhancer element. Error bars represent standard error of the mean (S.E.M.) from 3 biological replicates with 3 technical replicates for each experiment. b) Average expression of the nearest gene to elements from each class (measured by RNA-seq; RPKM = reads/kb mapped). Gene expression analysis was performed following 0,1,or 6 h of membrane depolarization by KCl (generated by Kim et al.). *p = 7.83 × 10⁻¹¹; paired Wilcoxon signed rank test. The number of genes in each group is indicated below the x-axis. Error bars represent standard S.E.M. c) Average expression of the nearest gene to enhancers with increasing H3K27ac with and without CBP/H3K4me1 enrichment (*p = 3.21e⁻⁰⁸ ,**p < 2.2e⁻16). Error bars represent S.E.M.

Figure 4. FOS binding is highly enriched at neuronal activity-regulated enhancers

a) Position weight matrix of the AP-1 site identified by MEME de novo motif search, performed with a search window of 150 bp from the center of the CBP peak within each activity-regulated enhancer. b) Aggregate plots of ChIP-seq signal for FOS at activity-regulated enhancers. Signal from a ChIP-seq experiment performed with an additional FOS antibody (sc-7202) also indicated. c) Distribution of FOS peaks with respect to gene TSS. Distal binding sites were defined as > 1 kb from an NCBI annotated RefSeq TSS. d) Aggregate plots of FOSB and JUNB binding at activity-regulated enhancers e) Aggregate plot of ChIP-seq signal for H3K4me1 and H3K27ac before and after membrane depolarization at FOS binding sites genome-wide. f) Aggregate plot of CBP binding and RNA PolII binding before and after membrane depolarization at FOS binding sites genome-wide. g) Genome-wide overlap between H3K27ac peaks, CBP/H3K4me1-enriched sites, and FOS peaks throughout the genome. h) Fraction of increasing H3K27ac peaks bound by FOS

Figure 5. AP-1 transcription factors are required for proper function of activity-regulated enhancers

(a) Effect of AP-1 site mutation on FOS-bound, activity-dependent enhancer function in the Nptx2 reporter. AP-1 mutant enhancers have a single point mutation introduced into every AP-1 site within the enhancer that abrogates AP-1 complex binding. (b) Effect of Fos shRNA co-transfection on activity-regulated enhancer activity compared to a control shRNA plasmid (see methods). Error bars for each panel represent S.E.M for 3 biological replicates with 3 technical replicates for each experiment.

Figure 6. FOS activates an extensive gene program in neurons that regulates synaptic development and function

(a) Western blot for FOS, FOSB and proteins involved in activity-regulated signaling from cells expressing either control or Fos shRNA after 0,1,3, or 6 h of KCl stimulation. (b) Quantification of gene expression changes in cells expressing a lentiviral shRNA construct targeting FOS compared to a control shRNA targeting firefly luciferase. Neurons were treated with KCl for the indicated times (0,1,3,6 h) and gene expression was measured by microarray from biological replicate samples (n = 2 for each condition). The line indicates genes whose expression is induced by neuronal activity at least 3-fold at each of the 3 timepoints in the control condition. Red dots indicate activity-regulated microarray probes whose expression was reduced by at least 1.5-fold in the FOS shRNA condition compared to control shRNA. The number of total probesets at each timepoint and the activity-induced probes with 1.5-fold decreased expression in the Fos shRNA condition are listed on each scatterplot.

Figure 7. Integrated genomic analysis identifies direct targets of FOS

(a) Expression values measured by microarray of selected direct FOS target genes at each timepoint of KCl stimulation (error bars indicate S.E.M., n = 2) (b) Genome browser tracks for each locus shown in (a) indicating enhancer associated features and AP-1 transcription factor binding data. Putative enhancer elements are indicated by the boxed regions. Npas4 ChIP-seq data was generated previously.

Figure 8. Expression of FOS direct target genes in mouse visual cortex

The expression of each indicated direct FOS target gene in the primary visual cortex was measured by qRT-PCR. Seven week-old mice were dark housed for one week and exposed to light for the indicated time periods. Measured expression levels for each gene were normalized to the timepoint with the highest level of expression. Error bars indicate S.E.M. from 4 animals.