Acetylcholine signaling in the medial prefrontal cortex mediates the ability to learn an active avoidance response following learned helplessness training

Abstract

Increased brain levels of acetylcholine (ACh) are observed in subsets of patients with depression and increasing ACh levels chronically can precipitate stress-related behaviors in humans and animals. Conversely, optimal ACh levels are required for cognition and memory. We hypothesize that ACh signaling is important for encoding both appetitive and stress-relevant memories, but that excessive increases in ACh result in a negative encoding bias in which memory formation of a stressful event is aberrantly strengthened, potentially contributing to the excessive focus on negative experience that could lead to depressive symptoms. The medial prefrontal cortex (mPFC) is critical to control the limbic system to filter exteroceptive cues and stress-related circuits. We therefore evaluated the role of ACh signaling in the mPFC in a learned helplessness task in which mice were exposed to repeated inescapable stressors followed by an active avoidance task. Using fiber photometry with a genetically-encoded ACh sensor, we found that ACh levels in the mPFC during exposure to inescapable stressors were positively correlated with later escape deficits in an active avoidance test in males, but not females. Consistent with these measurements, we found that both pharmacologically- and chemogenetically-induced increases in mPFC ACh levels resulted in escape deficits in both male and female mice, whereas chemogenetic inhibition of ACh neurons projecting to the mPFC improved escape performance in males, but impaired escape performance in females. These results highlight the adaptive role of ACh release in stress response, but also support the idea that sustained elevated ACh levels contribute to maladaptive behaviors. Furthermore, mPFC ACh signaling may contribute to depressive symptomology differentially in males and females.

Figure 1.. ACh levels in the mPFC are predictive of escape behavior in the learned helplessness paradigm.

A. Experimental design. Virus containing a construct for the fluorescent ACh sensor GACh3.0 was infused into the mPFC of male and female WT mice. Mice were allowed to recover for 3 weeks prior to beginning learned helplessness training, which consisted of two successive days of 1 h induction sessions of 120 inescapable shocks each. On the third day, mice underwent active avoidance training during which shocks and the possibility to escape were presented simultaneously. B. Image of mPFC and a higher magnification inset with fiber tract and GACh3.0 infection (green fluorescence surrounding fiber tract). C. Exposure to shocks during induction sessions resulted in increased escape latencies in male (left) and female (right) mice. Using a k-means clustering algorithm, mice were divided into helpless and resilient groups based on both their escape latency and number of trials with a failure to escape during active avoidance testing. D. A Trend LogRank for escapes (censored for no escapes) indicated a significant difference in escape performance between helpless and resilient mice, in which resilient mice achieved full escape efficacy sooner than helpless mice. E. Helpless males displayed increased ACh signaling in response to 4-s shock (grayed area) during induction trials, (F) whereas female helpless and resilient mice displayed similar ACh signaling. Data shown are the average z-scored ΔF/F per group, per day. G. There was a significant correlation between average AUC of ACh signal in response to shock during induction trials and escape latency during subsequent active avoidance testing for individual male mice on both days 1 and 2, but not for female mice (H).

Figure 2.. Pharmacologically prolonging ACh signaling during LH induction increases escape deficits in later active avoidance testing.

A. Experimental design. Mice received injections of physostigmine (PHYSO) or saline (SAL) 30 min prior to their second induction session. B. Physostigmine administration increased escape latencies. C. Relative to saline, physostigmine decreased escapes in the active avoidance test (insert). Furthermore, a Trend LogRank for escapes (censored for no escapes) indicated a significant difference in escape performance between mice that received physostigmine and those that received saline, such that saline mice achieved full escape efficacy sooner than those treated with physostigmine.

Figure 3.. Effects of DREADD-mediated manipulation of cholinergic terminals in the mPFC.

A. Experimental design. Retrograde, Cre-dependent viral constructs expressing Gq- or Gi-DREADDs or mCherry control were infused into the mPFC of ChAT-Cre mice. 4 weeks later mice underwent learned helplessness induction and active avoidance testing. Mice were administered CNO 30 minutes prior to each induction session. B. Brain section containing mPFC, the target for viral infusion. C, D, E. Higher magnification image of mPFC showing ChAT in cyan (C), DREADD expression in mCherry (D), and both signals merged (E). F. Brain section containing nucleus basilis cell bodies infected with DREADD-containing retrograde virus injected into mPFC. G, H, I. Higher magnification image of nucleus basilis showing cell bodies (arrows) infected with DREADD-containing virus, with ChAT in cyan (G), DREADD in mCherry (H), and both signals merged (I). J. Shock exposure during induction increased escape latency in all groups relative to controls. Gq-mediated excitation of ChAT-expressing terminals in the mPFC during induction increased escape latency in active avoidance testing. K. In male mice, Gq-mediated excitation increased escape latencies relative to Gi-mediated inhibition, but not relative to controls. Gi-mediated inhibition also resulted in more resilient mice than those that received Gq-mediated excitation. L. In contrast, relative to controls, female mice showed increased escape latencies in active avoidance testing following both Gq-mediated excitation and Gi-mediated inhibition during induction trials.