The study of active avoidance: A platform for discussion

Abstract

Traditional active avoidance tasks have advanced the field of aversive learning and memory for decades and are useful for studying simple avoidance responses in isolation; however, these tasks have limited clinical relevance because they do not model several key features of clinical avoidance. In contrast, platform-mediated avoidance (PMA) more closely resembles clinical avoidance because the response i) is associated with an unambiguous safe location, ii) is not associated with an artificial termination of the warning signal, and iii) is associated with a decision-based appetitive cost. Recent findings on the neuronal circuits of PMA have confirmed that amygdala-striatal circuits are essential for avoidance. In PMA, however, the prelimbic cortex facilitates the avoidance response early during the warning signal, perhaps through disinhibition of the striatum. Future studies on avoidance should account for additional factors such as sex differences and social interactions that will advance our understanding of maladaptive avoidance contributing to neuropsychiatric disorders.

Keywords: Accumbens; Amygdala; Fear; Prelimbic cortex; Striatum.

Figure 1.. Comparison of shuttle-box avoidance and platform-mediated avoidance tasks.

A. Typical behavior of rats during the shuttle-box avoidance task. Following the onset of the tone warning signal, rats learn to avoid by moving to the opposite side of the box, which both terminates the tone and prevents the generation of a footshock. Because rats must enter a place where they were previously shocked, no place is associated with safety. B. Typical behavior of rats during the platform-mediated avoidance task (PMA). Following the onset of the tone warning signal, rats learn to stop pressing the bar for sucrose pellets and move toward the platform, so that they are no longer in contact with the grid floor. Performing this avoidance response does not terminate the tone or prevent the generation of a footshock; rats must remain on the platform throughout the tone. In this way, rats are able to control their exposure to a shock, but not the presence of the shock or its predictive stimulus. While protecting them from shock, the platform also prevents rats’ access to sucrose reward. C. Acquisition of PMA across 10 days of training. As training proceeds, avoidance (time spent on platform) increases (orange lines), while freezing to the tone decreases (red lines). The rate of bar pressing during the inter-trial interval (ITI) increases back to its pre-conditioning rate (green lines). First data point shows the pre-conditioning baseline value (BL) to the first tone. All other data points are blocks of three tone trials. n=82 rats. All data shown as mean ± SEM. Modified from Bravo-Rivera et al. (2014).

Figure 2.. Proposed circuitry of platform-mediated avoidance at different timepoints during the tone.

A. (Left) Coronal schematics of BA (top) and PL (bottom) inactivations by infusing muscimol (MUS) or saline (SAL), showing location of cannula tips. (Right) Time spent on platform (in 3 sec bins) during the tone revealed that BLA inactivation blocked avoidance in MUS rats (top) across the entire tone compared to SAL controls (rptd meas ANOVA, post hoc Tukey). Data from Bravo-Rivera et al. (2014). PL inactivation significantly impaired in avoidance early, but not late, in the tone (bottom), in MUS rats compared to SAL controls (rptd meas ANOVA, post hoc Tukey). Modified from Diehl et al. (2018). B. (Left) Early during the tone, inhibitory responses are observed in PL neurons (dashed axon), which likely project to inhibitory interneurons in the VS. Inhibitory responses in PL would promote avoidance by disinhibiting the response of VS output neurons to BA inputs. PL neurons also show excitatory responses at platform mounting, which could project to BA neurons and facilitate avoidance. (Right) Late in the tone, as avoidance becomes more urgent, PL disinhibition of VS is maintained and BA excitation of VS increases in response to other inputs (thick axon), thereby over-riding inhibition from PL. Thus, expression of urgent avoidance (when shock is imminent) occurs independently of PL. Grey trace is hypothetical neural activity with black box indicating activity signal during the tone. All data shown as mean ±SEM. *p<0.05. Abbreviations: BA – basal amygdala; BLA – basolateral amygdala; CeA – central nucleus of the amygdala; IL – infralimbic cortex; LA – lateral amygdala; PL – prelimbic cortex; VS – ventral striatum.