A long noncoding eRNA forms R-loops to shape emotional experience-induced behavioral adaptation

Abstract

Emotional experiences often evoke neural plasticity that supports adaptive changes in behavior, including maladaptive plasticity associated with mood and substance use disorders. These adaptations are supported in part by experience-dependent activation of immediate-early response genes, such as Npas4 (neuronal PAS domain protein 4). Here we show that a conserved long noncoding enhancer RNA (lnc-eRNA), transcribed from an activity-sensitive enhancer, produces DNA:RNA hybrid R-loop structures that support three-dimensional chromatin looping between enhancer and proximal promoter and rapid Npas4 gene induction. Furthermore, in mouse models, Npas4 lnc-eRNA and its R-loop are required for the development of behavioral adaptations produced by chronic psychosocial stress or cocaine exposure, revealing a potential role for this regulatory mechanism in the transmission of emotional experiences.

Figure 1.. Npas4 conserved enhancer produces a long non-coding enhancer RNA

(A) Genomic landscape of the Npas4 enhancer and gene coding region combined with total RNA-seq data from NAc of mice subjected to cocaine conditioning (“Coc cond”) or homecaged controls, as well as from mPFC of mice subjected to social defeat stress or homecaged controls. Additionally, human and vertebrate conservation are included. On the lower part, the positive strand is visualized at 100x scale in comparison to the data on top, together with total RNA-seq data from the human cortex. The enhancer region is indicated with a gray rectangle, and Npas4^eRNA is indicated with a dashed black rectangle. Kbp, kilo-base pairs. (B) 3C assay shows the Npas4 enhancer-promoter interaction in homecaged control mice. Cocaine conditioning and social defeat stress increase this interaction in the NAc and mPFC, respectively (n = 10 to 20). (C) 3C assay shows the time course of Npas4 enhancer-promoter interaction in the NAc (n = 10). Data display the mean +/− SEM. Statistical analyses were performed using a one or two-way analysis of variance (ANOVA) followed by Tukey’s post-hoc multiple comparison test, *P <0.05, **P <0.001, Detailed statistical analyses are provided in table S1.

Figure 2.. Npas4^eRNA is required for Npas4^mRNA expression.

(A) Experimental design showing CRISPRa recruitment to the Npas4 enhancer region using sgRNAs (sgRNA-E1 and sgRNA-E1’). VPR, VP64-p65-Rta. (B) CRISPRa + sgRNA-E1+E1’ increases Npas4^mRNA and Npas4^eRNA, Npas4^eRNA-shRNA reduces Npas4^eRNA, and CRISPRa-induced Npas4^mRNA in N2A cells (n = 11 to 18). (C) Npas4^eRNA-shRNA reduces Npas4^eRNA and cocaine conditioning induced Npas4^mRNA in the NAc of mice compared with scrambled control; levels of cFos are similar across conditions (n = 7 to 10) (D) Npas4^eRNA overexpression increases Npas4^eRNA and Npas4^mRNA in the NAc of home-caged mice compared with green fluorescent protein (GFP) overexpression; levels of cFos are not changed across conditions (n = 10 to12). Data plots show the mean +/− SEM. Statistical analyses were performed using a one-way ANOVA followed by a Tukey’s post-hoc multiple comparison test and Mann-Whitney U test,*P <0.05, ***P <0.001, ****P <0.0001, non-significant (ns) P >0.05. Detailed statistical analyses are provided in table S1.

Figure 3.. Npas4^eRNA forms RNA:DNA hybrid R-loops at the enhancer

(A) Diagram showing genomic fragments of the Npas4 enhancer region after enzymatic digestion with a %GC (guanine-cytosine) plot (calculated and adapted from SnapGene) as well as their GC-Skew. (B and C) DRIP-qPCR assay showing the percent enrichment of genomic fragments over the input in the NAc of mice with or without cocaine conditioning (B), the mPFC of mice with or without social defeat stress (C), and RNaseH1 pretreatment around Npas4 enhancer. The DRIP-qPCR assay shows enrichment of fragments A, B, and C in comparison to RNaseH1 control, with fragments B and C showing increased enrichment after cocaine conditioning (B) and social defeat stress (C) compared with the home-caged controls. Fragment α does not show the increased enrichment compared with RNaseH1 negative control (n = 6 to 20). (D) Correlation analyses of enrichment fragments A, B, and C with Npas4^mRNA expression in the NAc after 15 mins of cocaine conditioning (n = 9 or 10). (E and F) Time course of Npas4 mRNA expression and R-loops in the NAc after cocaine conditioning. Npas4^mRNA is increased in the NAc of mice after 15 mins of cocaine conditioning compared with those in the home cage. At 1 and 24 hours after cocaine conditioning, Npas4^mRNA expression is similar to home cage controls (n = 9 or 10) (E). The DRIP-qPCR shows enrichments of fragment A compared with RNaseH1 control, and fragments B and C show increased enrichment 15 min after cocaine conditioning compared with mice in the home cage. At 1 and 24 hours after cocaine conditioning, the enrichments are similar to home-caged controls (n = 7 to 10) (F). Data show the mean +/− SEM. Statistical analyses were performed using a one-way ANOVA followed by Tukey’s post-hoc multiple comparison test, *P <0.05, **P <0.01, ***P <0.001, ****P <0.0001s. Detailed statistical analyses are provided in table S1.

Figure 4.. R-loops at the Npas4 enhancer regulate activity-induced Npas4^mRNA expression and promoter-enhancer 3D chromatin looping

(A) (Left) Experimental design using dCas9-RNaseH1 active (CRISPR-H1) or enzymatically dead (control) with sgRNAs targeting the constitutive R-loop at the Npas4 enhancer (sgRNA-E1) or the activity-inducible R-loop (sgRNA-E2) to degrade R-loops in a locus specific manner. (Right) Diagram showing experimental strategy to assess R-loop function in the NAc after cocaine conditioning. (B and C) CRISPR-H1 recruited with sgRNA-E1 (B) and sgRNA-E2 (C) in the NAc reduces cocaine and saline conditioning-induced Npas4^mRNA expression compared with the enzymatically dead control [n = 5 to 6 (B), n = 6 to 10 (C)]s (D) Schematic of the experimental strategy to detect 3D looping via a chromosome conformation capture (3C) method with primers. (E) 3C analysis shows that Npas4 promoter-enhancer interaction in the NAc of control mice (CRISPR-H1d + sgRNA-E1/E2) is increased after 15 mins of cocaine conditioning, whereas CRISPR-H1 + sgRNA-E1/E2 eliminates the promoter-enhancer interaction in basal and cocaine-conditioned animals (n = 8 to10). Data display the mean +/− SEM. Statistical analyses were performed using two-way ANOVA followed by Tukey’s, Bonferroni’s or Sidak post-hoc multiple comparison test, *P <0.05, **P <0.01, ***P <0.001, ****P <0.001, ns: non-significant P >0.05. Detailed statistical analyses are provided in table S1.

Figure 5.. Npas4^eRNA and its R-loop are required for behavioral adaptation produced by cocaine or chronic stress experiences

(A) Diagram showing the experimental timeline of cocaine CPP with adult male and female mice that received AAV2-Npas4^eRNA shRNA, control shRNA, or Lenti-dCas9-RNaseH1 (CRISPR-H1) or enzymatically dead (control; CRISPR-H1d) together with Lenti-sgRNA-E2 in the NAc. (B) Male and female mice with control-shRNA in the NAc showed an increased CPP score after cocaine conditioning, whereas mice with Npas4^eRNA shRNA had a similar CPP score before and after conditioning (n = 10 to 12). (C) Male and female mice with Control lentivirus in the NAc showed an increased CPP score after cocaine conditioning, but those with CRISPR-H1 + sgRNA-E2 failed to develop cocaine CPP (n =10 to12). (D). Diagram showing the experimental timeline of CSDS followed by behavioral test battery including sucrose preference assay with adult male mice that received AAV2-Npas4^eRNA shRNA or control shRNA in the mPFC. (E) Control mice show decreased sucrose preference after CSDS whereas mice with Npas4^eRNA shRNA in the mPFC fail to reduce their sucrose preference after CSDS (n = 12 to 27). (F) Working model illustrating how the Npas4^eRNA forms R-loops that prime for emotional experience-induced enhancer-promoter chromatin looping, rapid Npas4 mRNA expression, and behavioral plasticity. Statistical analyses were performed using two-way ANOVA followed by Sidak or Tukey’s post-hoc multiple comparison test, *P < 0.05, **P < 0.01, ***P < 0.001, ns: non-significant P > 0.05. Detailed statistical analyses are provided in table S1.