Fear conditioning- and extinction-induced neuronal plasticity in the mouse amygdala

Experience-dependent changes in behavior are mediated by long-term functional modifications in brain circuits. To study the underlying mechanisms, our lab is using classical auditory fear conditioning, a simple and robust form of associative learning. In classical fear conditioning, the subject is exposed to a noxious unconditioned stimulus (US), such as a foot-shock, in conjunction with a neutral conditioned stimulus (CS), such as a tone or a light. As a result of the training, the tone acquires aversive properties and when subsequently presented alone, will elicit a fear response. In rodents, such responses include freezing behavior, alterations in autonomic nervous system activity, release of stress hormones, analgesia, and facilitation of reflexes. Subsequently, conditioned fear can be suppressed when the conditioned stimulus is repeatedly presented alone, a phenomenon called fear extinction.
It emerges from a large number of studies in animals and humans that the amygdala is a key brain structure mediating fear conditioning. The amygdala consists of several distinct nuclei, including the lateral (LA) and basal (BA) nuclei, and the central nucleus (CEA). In the classical circuit model of fear conditioning, the LA is thought of as the primary site where CS-US associations are formed and stored. The formation of CS-US associations in the LA is mediated by N-methyl-D-aspartate (NMDA) receptor-dependent long-term potentiation (LTP) at glutamatergic sensory inputs originating from auditory thalamus and cortex. In contrast to the LA, the CEA has been considered to be primarily involved in the behavioral expression of conditioned fear responses.
While the mechanisms and the circuitry underlying fear conditioning in the LA have been extensively studied, much less is known about the neuronal substrates underlying fear extinction. The question of how conditioned fear can be inhibited by extinction is attracting increasing interest because of its clinical importance for the therapy of anxiety disorders. The amygdala is also a potential site of extinction-associated plasticity since intra-amygdala blockade of NMDA receptors or the MAPK signaling pathway prevents extinction. In the first part of this thesis, a combination of behavioral, pharmacological and in vivo electrophysiological approaches was used to study the role of distinct amygdala sub-nuclei in fear exinction. Single unit recordings in behaving mice revealed that the BA contains distinct types of neurons that are specifically activated upon fear conditioning or extinction, respectively. During acquisition of extinction, the activity of “fear neurons” gradually declines, while “extinction neurons” increase their activity. Conversely, when extinguished fear responses are recovered by placing the animal in an unsafe environment, “extinction neurons” switch off, while “fear neurons” switch on. Using local micro-iontophoretic injection of the GABAA receptor agonist muscimol, we found that inactivation of the BA completely prevents the acquisition of extinction or context-dependent fear recovery, depending on the injection time point. Finally, we could show that “fear neurons” and “extinction neurons” are differentially connected with the medial prefrontal cortex (mPFC) and the ventral hippocampus (vHC), two brain areas involved in context-dependent extinction. In contrast to previous models suggesting that amygdala neurons are active during states of high fear and inactive during states of low fear, our findings indicate that activity in specific neuronal circuits within the amygdala may cause opposite behavioral outcomes, thus providing a new framework for understanding context-dependent expression and extinction of fear behavior.
In the second part of the thesis, I examined how inhibitory circuits in the central nucleus of the amygdala (CEA) contribute to fear conditioning. While many studies have demonstrated that neuronal plasticity in the LA is necessary for fear conditioning, the role of the CEA, which is mainly composed of GABAergic inhibitory neurons, is poorly understood. In the classical circuit model, the CEA has been thought of as a passive relay station conveying LA output to downstream targets in the hypothalamus and in the brain stem. However, recent in vivo pharmacological experiments suggest a more active role for the CEA during fear conditioning. To address the role of CEA inhibitory circuits in fear conditioning, we obtained single unit recordings from neurons located in the lateral (CEl) and medial (CEm) subdivisions of the CEA in behaving mice. We found that CEm output neurons, that control fear behavior via projections to brainstem targets, are under tight inhibitory control from a subpopulation of neurons located in CEl. Fear conditioning induced opposite changes in phasic and tonic inhibition in the CEl to CEm pathway. Targeted pharmacological inactivation of CEl and CEm revealed that whereas plasticity of phasic inhibition is necessary for gating CEm output during fear learning and expression, changes in tonic inhibitory network activity control signal-to-noise ratio and stimulus discrimination. Our results identify CEA inhibitory circuits as a major site of plasticity in fear conditioning, and suggest that regulation of tonic activity of inhibitory circuits may be an important mechanism for controlling sensitivity and specificity in associative learning.
Taken together, these findings suggest that the amygdala is not a functionally homogeneous structure. Rather, our results reveal that the BA and the CEA contain specialized and discrete neuronal populations that contribute to distinct aspects of fear conditioning and extinction. Ultimately, elucidating these mechanisms is fundamental for an understanding of memory processes in the brain in general, and should also inform novel therapeutic strategies for psychiatric disorders involving excessive fear responses associated with amygdala hypersensitivity such as post-traumatic stress disorder and other anxiety disorders.