Enabling cell-type-specific behavioral epigenetics in Drosophila: a modified high-yield INTACT method reveals the impact of social environment on the epigenetic landscape in dopaminergic neurons

Abstract

Background: Epigenetic mechanisms play fundamental roles in brain function and behavior and stressors such as social isolation can alter animal behavior via epigenetic mechanisms. However, due to cellular heterogeneity, identifying cell-type-specific epigenetic changes in the brain is challenging. Here, we report the first use of a modified isolation of nuclei tagged in specific cell type (INTACT) method in behavioral epigenetics of Drosophila melanogaster, a method we call mini-INTACT.

Results: Using ChIP-seq on mini-INTACT purified dopaminergic nuclei, we identified epigenetic signatures in socially isolated and socially enriched Drosophila males. Social experience altered the epigenetic landscape in clusters of genes involved in transcription and neural function. Some of these alterations could be predicted by expression changes of four transcription factors and the prevalence of their binding sites in several clusters. These transcription factors were previously identified as activity-regulated genes, and their knockdown in dopaminergic neurons reduced the effects of social experience on sleep.

Conclusions: Our work enables the use of Drosophila as a model for cell-type-specific behavioral epigenetics and establishes that social environment shifts the epigenetic landscape in dopaminergic neurons. Four activity-related transcription factors are required in dopaminergic neurons for the effects of social environment on sleep.

Fig. 1

mini-INTACT method affinity purifies cell-type-specific nuclei for epigenetic analysis in Drosophila. a Schematic of mini-INTACT method to affinity purify dopaminergic neurons from heads of Drosophila male flies after 4 days of social isolation or social enrichment. b Expression of SUN-tagged GFP (3XUAS-unc84-2XGFP) in dopaminergic neurons driven by TH-GAL4 in an adult male brain. c To assess the purity of dopaminergic nuclei, nuclei were obtained from a mixture of heads derived from flies expressing 10XUAS-unc84-tdTomfl under the control of the pan-neuronal elav-GAL4 driver (red) and flies expressing 3XUAS-unc84-2XGFP under the control of the TH-GAL4 driver (green). d After capturing green nuclei using bead-bound anti-GFP antibodies, red nuclei were washed away. e Bead-bound green nuclei were almost devoid of contaminating red nuclei

Fig. 2

Epigenome of mini-INTACT purified dopaminergic neurons measured by ChIP-seq and RNA-seq. a–f Genome-wide profiles of the levels for the six epigenetic marks shown as ngs.plot displays. Vertical axes indicate genome-wide average of read counts per million reads. a–c Activating marks were concentrated in the promoter and immediately downstream of the TSS. d H3K36me3, a mark associated with transcriptional elongation, was enriched in the gene body and skewed towards the TES. e, f Two repressive marks were depleted from the TSS and TES, concentrated in the gene body, and enriched upstream of the promoter region. g Epigenetic and transcriptional enrichment profiles surrounding the Ddc gene. Representative RNA-seq panels show that Ddc is more strongly expressed in GH (blue) than in SH (red) males. The distribution of epigenetic marks shown is representative of the SH dataset. The four activating marks (green panels) were high in this strongly expressed gene, while the two repressive marks (red) showed low levels. An example of single “input” DNA track, which is used as a control for mark levels, is shown in the final panel

Fig. 3

Epigenetic landscape of genes expressed in dopaminergic neurons is modulated by social experience. Heat map of eight groups identified by k-means clustering of the change in whole gene average mark levels and mRNA levels between GH and SH males. Red lines show genes whose marks or mRNA was higher in GH than SH males, blue lines show those that were higher in SH than GH males. Some clusters are enriched for genes with neural and regulatory functions, especially clusters 6–8. Enriched GO categories from each cluster are shown on the right. N (genes per cluster): 1:979, 2:612, 3:602, 4:413, 5:900, 6:637, 7:863, 8:544

Fig. 4

Activity-regulated genes (ARGs) are upregulated in dopaminergic neurons of GH males and correlate with transcriptional repression. a A zoomed in scatter plot of GH versus SH mRNA values. Ddc (Fig. 2) and four ARG-TFs defined as such in [61] and showing high fold change in our data are highlighted. b Box plots of mRNA log fold change z-scores (GH is positive; SH is negative) for groups of ARGs from two different studies. Genes with log fold change lower than 1.5 in Chen et al.’s study [61] are not over-represented in GH flies (Chen et al. low). However, the last two groups are significantly over-represented in GH flies. c Genes repressed (red) or activated (blue) by the ARG-TF Cbt (from Bartok et al., 2015) are shown on the same z-score scale as in (b). Genes repressed by Cbt have significantly lower mRNA and activating mark H3K27ac and significantly higher repressive mark H3K27me3. Genes activated by Cbt show the reverse pattern

Fig. 5

Knockdown of ARGs by RNAi affected social effects on daytime sleep. Knock-down of ARGs in dopaminergic neurons was achieved by driving RNAi transgenes with TH-GAL4; controls carried empty vectors without RNAi hairpin and TH-GAL4. a Example graph of sleep per 30 min over 24 h. Control single housed (SH) flies sleep less than group housed (GH) flies during the day (shaded gray area). Expressing RNAi for ARG-TF cabut in dopaminergic neurons significantly reduced this difference. b Daytime sleep was measured, and ΔSleep was compared between experimental males carrying the RNAi transgene and controls. ΔSleep is defined as minutes of daytime sleep for GH flies minus the same measure for SH flies (as described by Ganguly et al. [11]). c ΔSleep for controls and RNAi knockdowns. Error bars are mean ± SEM. In every case, RNAi knockdown significantly reduced the social effect on ΔSleep. Two different RNAi lines were tested for CrebA, each showing significant reductions