Beyond the Medial Regions of Prefrontal Cortex in the Regulation of Fear and Anxiety

Abstract

Fear and anxiety are adaptive responses but if left unregulated, or inappropriately regulated, they become biologically and socially maladaptive. Dysregulated emotions are manifest in a wide variety of psychiatric and neurological conditions but the external expression gives little indication of the underlying causes, which are inevitably multi-determined. To go beyond the overt phenotype and begin to understand the causal mechanisms leading to conditions characterized by anxiety and disorders of mood, it is necessary to identify the base psychological processes that have become dysregulated, and map them on to their associated neural substrates. So far, attention has been focused primarily on the medial regions of prefrontal cortex (PFC) and in particular their contribution to the expression and extinction of conditioned fear. However, functional neuroimaging studies have shown that the sphere of influence within the PFC is not restricted to its medial regions, but extends into dorsal, ventrolateral (vlPFC) and orbitofrontal (OFC) regions too; although the causal role of these other areas in the regulation of fear and anxiety remains to be determined and in the case of the OFC, existing findings are conflicting. Here, we review the evidence for the contribution of these other regions in negative emotion regulation in rodents and old world and new world monkeys. We consider a variety of different contexts, including conditioned and innate fear, learned and unlearned anxiety and cost-benefit decision-making, and a range of physiological and behavioral measures of emotion. It is proposed that both the OFC and vlPFC contribute to emotion regulation via their involvement, respectively, in the prediction of future outcomes and higher-order attentional control. The fractionation of these neurocognitive and neurobehavioral systems that regulate fear and anxiety opens up new opportunities for diagnostic stratification and personalized treatment strategies.

Keywords: anxiety; emotion regulation; fear; marmoset; orbitofrontal cortex; prefrontal cortex; primate; ventrolateral prefrontal cortex.

Figure 1

A schematic diagram depicting the location of lesions across studies in primates (upper panel) and rats (lower panel) in which the effects of ventral prefrontal cortex (PFC) damage on a variety of anxiety and conditioned/unconditioned fear tests have been investigated. AI, anterior insula; dAC, dorsal anterior cingulate; DLO, dorsolateral orbital area; IL, infralimbic cortex; LO, lateral orbital area; MO, medial orbital area; PL, prelimbic cortex; VO, ventral orbital area. Cytoarchitectonics of the ventral PFC are based on Carmichael and Price (; rhesus macaques), Burman and Rosa (; marmosets), Paxinos and Watson (; rats) and cortical granularity taken from Wise (2008).

Figure 2

The effects of permanent and temporary manipulations of the antOFC and vlPFC on the responsivity of marmosets to a variety of fear and anxiety-inducing stimuli. In (A) both lesioned groups, post-surgery, took the same number of sessions as controls to regain discriminative conditioned responding to a CS associated with aversive loud noise (i “Retention”). Following exposure to one session of partial extinction, however, whereas controls adapted their responding and took many sessions to then regain their discriminative responding (i “Recovery”), the lesioned groups maintained strong discriminative conditioning throughout (ii). *p < 0.05, **p < 0.01. Figures redrawn from Agustín-Pavón et al. (2012). In (B) pale gray dots represent the emotionality and coping strategy component scores of individual marmosets in the colony in response to a human intruder (HIT) and a model snake. Emotionality scores show a significant positive relationship such that animals scoring high on the HIT, score high on the Snake too and vice versa. Superimposed on these scores are the average scores of each of the lesioned and control groups. Both lesioned groups showed greater emotionality scores on the HIT and snake test whilst their coping strategies differed across the two tests. The vlPFC lesioned group displayed a higher, more active coping strategy on the HIT but both antOFC and vlPFC lesioned groups displayed a lower, more passive strategy score on the Snake, compared to controls. Data taken and redrawn from Agustín-Pavón et al. (2012) and Shiba et al. (2015). In (C) the effect of temporary inactivation of the antOFC and vlPFC on the effects of cost-benefit decision making are displayed. Inactivation of the vlPFC had no effect on response bias during reward only sessions (i, left side) but in the presence of punishment (Test Day 1) significantly increased responding away from punishment (i, right side). antOFC inactivations also had no effect on response bias on reward only sessions but enhanced responding away from the punished side the day after having received punishment (Test Day 2). The biases in both cases were ameliorated with concomitant treatment with the anxiolytic, diazepam (ii). The antOFC-induced punishment bias on the day after punishment was blocked by inactivation on Test Day 2 of the amygdala (Amyg) bilaterally, anterior hippocampus (Hipp) bilaterally, or amygdala and hippocampus unilaterally on opposite sides of the hemisphere (A-H disconnection; iii). The key below each graph in (C) indicates when and where infusions were made and whether punishment was present or not. *p < 0.05 on square-root transformed data. Figures redrawn from Clarke et al. (2015).