Histone modifications around individual BDNF gene promoters in prefrontal cortex are associated with extinction of conditioned fear
Extinction of conditioned fear is an important model both of inhibitory learning and of behavior therapy for human anxiety disorders. Like other forms of learning, extinction learning is long-lasting and depends on regulated gene expression. Epigenetic mechanisms make an important contribution to persistent changes in gene expression; therefore, in these studies, we have investigated whether epigenetic regulation of gene expression contributes to fear extinction. Since brain-derived neurotrophic factor (BDNF) is crucial for synaptic plasticity and for the maintenance of long-term memory, we examined histone modifications around two BDNF gene promoters after extinction of cued fear, as potential targets of learning-induced epigenetic regulation of gene expression. Valproic acid (VPA), used for some time as an anticonvulsant and a mood stabilizer, modulates the expression of BDNF, and is a histone deacetylase (HDAC) inhibitor. Here, we report that extinction of conditioned fear is accompanied by a significant increase in histone H4 acetylation around the BDNF P4 gene promoter and increases in BDNF exon I and IV mRNA expression in prefrontal cortex, that VPA enhances long-term memory for extinction because of its HDAC inhibitor effects, and that VPA potentiates the effect of weak extinction training on histone H4 acetylation around both the BDNF P1 and P4 gene promoters and on BDNF exon IV mRNA expression. These results suggest a relationship between histone H4 modification, epigenetic regulation of BDNF gene expression, and long-term memory for extinction of conditioned fear. In addition, they suggest that HDAC inhibitors may become a useful pharmacological adjunct to psychotherapy for human anxiety disorders.
Substantial evidence indicates that extinction of conditioned fear, the reduction in responding to a feared cue when the cue is repeatedly presented without any adverse consequence, is new learning that inhibits the expression of a conditioned association rather than erasing it. For example, conditioned fear shows “spontaneous recovery” after the passage of time (Baum 1988), “reinstatement” after presentations of the unconditioned stimulus (US) alone (Rescorla and Heth 1975), and “renewal” when the feared cue is presented in a context different from that of extinction training (Bouton and King 1983). Efforts to understand the mechanisms of this form of learning have increased recently, particularly since it is an important model of anxiety disorder treatment.
Many forms of learning, including extinction, are dependent on changes in gene expression (Berman and Dudai 2001; Cammarota et al. 2003; Lin et al. 2003; Sangha et al. 2003; Vianna et al. 2003; Herry and Mons 2004; Suzuki et al. 2004; Yang and Lu 2005; Chhatwal et al. 2006; Herry et al. 2006; Lattal et al. 2006). Dynamic changes in chromatin structure make an important contribution to the regulation of tissue-specific gene expression. In particular, histone acetylation/deacetylation and dimethylation of specific lysine residues on nucleosomal histone proteins (i.e., H3-K9) and DNA methylation of CpG dinucleotides within promoter regions are ways that chromatin remodeling can influence ongoing transcription and synaptic plasticity (Martinowich et al. 2003; Levenson et al. 2006). Histone acetylation contributes an early step to the process of chromatin modification by disassembling nucleosomes to make DNA promoter regions accessible for transcription factor binding and for methylation. Histone acetylation states are regulated by specific enzymes, including histone deacetylases (HDACs), which can be both tissue- and cell-type-specific. Thus, the omnipresence and specificities of these enzymes may make them potential therapeutic targets for the treatment of neuropsychiatric disorders and disorders of learning and memory.
In addition to its trophic function during development, brain-derived neurotrophic factor (BDNF) is critical for learning-related synaptic plasticity and the maintenance of long-term memory. The role of BDNF in fear conditioning is well defined, and, within the amygdala of the rat, both fear conditioning and its extinction lead to an increase in BDNF protein and gene transcripts (Rattiner et al. 2004; Chhatwal et al. 2006; Ou and Gean 2006). Recent data indicate that the medial prefrontal cortex also plays an important role in fear extinction learning (Milad and Quirk 2002; Milad et al. 2004; Santini et al. 2004), but the function of BDNF in the prefrontal cortex during extinction remains undefined. Thus, regulation of BDNF in the prefrontal cortex is a reasonable candidate mechanism to make a contribution to extinction learning.
BDNF has four distinct transcripts each regulated by a specific promoter that is sensitive to epigenetic modification (Martinowich et al. 2003; Tsankova et al. 2004). We chose to examine histone acetylation around two of those promoters in the prefrontal cortex after fear conditioning or after fear conditioning followed by fear extinction in order to provide support for the hypothesis that distinct patterns of histone acetylation are associated with specific behavioral changes: histone acetylation patterns and behavior that might be mimicked by combining partial behavioral training with a drug that promotes histone acetylation.
We find that fear conditioning and extinction result in distinct patterns of histone acetylation of histones H3 and H4 around the P1 and P4 promoters of the BDNF gene. We also find that promoting histone acetylation potentiates partial fear extinction, by decreasing remembered fear, but only when combined with partial extinction training (training insufficient to generate significant extinction on its own). Finally, we demonstrate that the same partial extinction training, by itself, also fails to generate the characteristic histone acetylation changes of extinction, but, when combined with an HDAC inhibitor, mimics the histone acetylation pattern characteristic of strong extinction around the BDNF P4 promoter. This synergistic combination also increases BDNF exon IV-containing transcripts, while the combination does not specifically affect exon I-containing transcripts.
This work was supported in part by grants from the NIMH and the Tennenbaum Family Foundation (M.B.) and by postdoctoral fellowships from FRSQ and NSERC (T.W.B.).
Article is online at http://www.learnmem.org/cgi/doi/10.1101/lm.500907