Assessment of gene expression is a good place to start a quest for mechanisms of neural plasticity (McClung and Nestler, 2008). Of course, these mRNA and subsequent protein-level changes must eventually lead to synaptic alterations in order to have a role in neural adaptation. Nonetheless, the efficiency of large-scale transcription studies makes it a powerful initial approach. Although this strategy has been used extensively within the field of drug addiction (Robison and Nestler, 2011), few studies have systematically evaluated brain transcriptional adaptation in response to metabolic challenge.

We recently completed a microarray analysis to identify changes in gene expression in response to a brief (5 days) and moderate (<10% weight loss) reduction in food intake (Guarnieri et al, 2012). Analysis was conducted on RNA from four brain regions: ventral tegmental area, prefrontal cortex, hypothalamus and nucleus accumbens. Stress responsive genes were prominent in the list of genes that change after restriction, and corticosterone was shown to be necessary and sufficient for many of these changes. There were some surprises in these results that deserve emphasis: (1) the gene changes occurred rapidly (within the first 24 h of restriction) before any weight loss was apparent and (2) the expression changes detected were not restricted to one region, and usually were seen in 2–4 of the regions surveyed. This fast and universal response was not something that was anticipated, but is reasonable if a systemic factor such as corticosterone is the key signal. Finally, the changes seen were implicated in increased food motivation that occurred after restriction.

Previously published work supports the role of stress response in mediating effects of caloric restriction. Work from the Bale lab has shown that a longer (21 days) caloric restriction leads to stress pathway activation, including corticosterone as well as corticotropin-releasing factor (CRF) in the bed nucleus of the stria terminalis (Pankevich et al, 2010). Restriction increased stress responses as well as binge eating. Interestingly, the changes in CRF did not reverse upon re-feeding, and subsequent analysis showed that the CRF regulatory sequences also maintained altered methylation patterns. Moreover, withdrawal from high-fat food produced similar adaptions in gene expression and chromatin state.

Other work with animals exposed to high-fat diet also supports a role for chromatin changes and transcriptional adaptation within dopamine circuits. Reyes and colleagues (Vucetic et al, 2011) have shown specific gene changes in the nucleus accumbens following long-term high-fat-diet exposure. Specifically, decreases in mu-opioid receptor expression correlated with increased methylation and MECP2 binding in the proximal regulatory regions. Moreover, H3K9 methylation was increased and acetylation was reduced, which is consistent with an inactive chromatin state.

In sum, changes in metabolic state can have specific and significant effects on the transcription profile of the brain. These changes are mediated by alterations in chromatin and, most importantly, are likely to have a role in behavioral adaptation to different food environments.

Defining these molecular responses will help us to understand how intake and weight control is influenced by previous metabolic experience. These pathways are targets for potential therapeutics that can reverse or otherwise modulate this plasticity and help with weight loss. Moreover, as with drugs of abuse, stress is a common cause for loss of control over eating. The above work suggests that the brain might interpret stress hormones as a state of hunger, and should inspire stress response attenuation as a potential therapeutic strategy for obesity.