Regret describes recognizing alternative actions could have led to better outcomes. It remains unclear whether regret derives from generalized mistake appraisal or instead comprises dissociable, action-specific processes. Using a neuroeconomic task, we found that mice were sensitive to fundamentally distinct types of regret following exposure to chronic social defeat stress or manipulations of CREB, a transcription factor implicated in stress action. Bias to make compensatory decisions after rejecting high-value offers (regret type I) was unique to stress-susceptible mice. Bias following the converse operation, accepting low-value offers (regret type II), was enhanced in stress-resilient mice and absent in stress-susceptible mice. CREB function in either the prefrontal cortex or nucleus accumbens was required to suppress regret type I but bidirectionally regulated regret type II. We provide insight into how maladaptive stress response traits relate to distinct forms of counterfactual thinking, which could steer therapy for mood disorders, such as depression, toward circuit-specific computations through a careful description of decision narrative.

BACKGROUND Recent research in humans and rodents has explored the use of deep brain stimulation (DBS) of the ventral capsule/ventral striatum (VS) as a possible treatment for drug addiction. However, the optimum electrode placement and optimum DBS parameters have not been thoroughly studied. Here we varied stimulation sites and frequencies to determine whether DBS of the VS could facilitate the extinction of morphine-induced conditioned place preference in rats. Methods Rats were implanted with DBS electrodes in the dorsal or ventral subregions of the VS and trained to the morphine-CPP. Subsequently, rats received extinction sessions over 9 days, combined with 60 min of either high (130 Hz) or low (20 Hz) frequency DBS. To study circuit-wide activations after DBS of the VS, c-fos immunohistochemistry was performed in regions involved in the extinction of drug seeking behaviors. Results High frequency DBS of the dorsal-VS impaired both extinction training and extinction memory, whereas high frequency DBS of the ventral-VS had no effect. In contrast, low frequency DBS of the dorsal-VS strengthened extinction memory when tested 2 or 9 days after the cessation of stimulation. Both DBS frequencies increased c-fos expression in the infralimbic prefrontal cortex, but only low frequency DBS increased c-fos expression in the basal amygdala and the medial portion of the central amygdala. Conclusions Our results suggest that low frequency (rather than high frequency) DBS of the dorsal-VS strengthens extinction memory and may be a potential adjunct for extinction-based therapies for treatment-refractory opioid addicts.

MiR-124 is a highly expressed miRNA in the brain and regulates genes involved in neuronal function. We report that miR-124 post-transcriptionally regulates PARP-1. We have identified a highly conserved binding site of miR-124 in the 3′-untranslated region (3′UTR) of Parp-1 mRNA. We demonstrate that miR-124 directly binds to the Parp-1 3′UTR and mutations in the seed sequences abrogate binding between the two RNA molecules. Luciferase reporter assay revealed that miR-124 post-transcriptionally regulates Parp-1 3′UTR activity in a dopaminergic neuronal cell model. Interestingly, the binding region of miR-124 in Parp-1 3′UTR overlapped with the target sequence of miR-125b, another post-transcriptional regulator of Parp-1. Our results from titration and pull-down studies revealed that miR-124 binds to Parp-1 3′UTR with greater affinity and confers a dominant post-transcriptional inhibition compared to miR-125b. Interestingly, acute or chronic cocaine exposure downregulated miR-124 levels concomitant with upregulation of PARP-1 protein in dopaminergic-like neuronal cells in culture. Levels of miR-124 were also downregulated upon acute or chronic cocaine exposure in the mouse nucleus accumbens (NAc)-a key reward region of brain. Time-course studies revealed that cocaine treatment persistently downregulated miR-124 in NAc. Consistent with this finding, miR-124 expression was also significantly reduced in the NAc of animals conditioned for cocaine place preference. Collectively, these studies identify Parp-1 as a direct target of miR-124 in neuronal cells, establish miR-124 as a cocaine-regulated miRNA in the mouse NAc, and highlight a novel pathway underlying the molecular effects of cocaine.