Abstract: Feelings of fear and anxiety are evolutionarily conserved traits which have been selected to be protective against dangerous situations, but their effects can be debilitating when uncontrolled. Examples of uncontrolled fear responses include post-traumatic stress disorder (PTSD) and a variety of fear and anxiety disorders that afflict approximately 18% of the adult US population. Many of... read morethese disorders are characterized by having an impairment in the learning or extinction of fear, thus inhibiting the gradual decrease of the fear response that normally occurs over time. Understanding the functional synaptic connections between brain regions that participate in the fear and extinction circuits will offer insight into the neural mechanisms of fear and provide a basis for therapies of a dysfunctional circuit. The basal amygdala (BA) has been established as the primary control center in the brain that underlies contextual fear learning and memory. The bed nucleus of the stria terminalis (BNST), a limbic structure in the forebrain, has been implicated to preferentially support contextual fear, although its precise functional connectivity with the amygdala in contextual fear remains poorly understood. Here, we used a c-fos based reporter mouse (TetTag mouse) to investigate the activity of the BNST in contextual fear learning and memory. We found that the oval BNST subdivision (ovBNST) was activated during contextual fear conditioning while the anterodorsal BNST subdivision (adBNST) was not, underscoring the divergent functionality of these two dBNST subdivisions. To further characterize amygdalar input to the ovBNST, we used anterograde and retrograde tracers to identify a direct projection from the BA to the ovBNST. We then combined retrograde tracer injections with the TetTag mouse to investigate the activity of a BA-BNST pathway during the acquisition and retrieval of contextual fear. In support of a direct functional interaction between the BA and the BNST during the learning of fear, we identified BNST-projecting neurons in the posterior subdivision of the BA that were activated during contextual fear conditioning without reactivation during retrieval. We also identified a separate, non-BNST-projecting population of neurons in the anterior subdivision of the BA that were preferentially reactivated during fear retrieval. Our results suggest that a direct BA-ovBNST pathway might support the acquisition and/or consolidation of contextual fear, while BA neurons that do not project to the BNST support the storage of the contextual fear memory and are incorporated into the memory engram. We additionally used the TetTag system to investigate the extinction-induced synaptic changes that occur around neurons in the BA after contextual fear extinction. We found that extinction training leads to a decrease in fear response concomitant with a decreased reactivation of fear neurons in the mouse amygdala. This decrease in reactivated neurons was not seen in the hippocampus or medial prefrontal cortex, suggesting that the silenced BA neurons were a selective site of extinction-induced suppression. We found that extinction led to an increase in inhibitory perisomatic PV selectively around these silenced fear neurons in the BA, as well as an increase in CB1R localized to CCK perisomatic synapses around active fear neurons. These modulatory synaptic changes matched the silent or active state of the postsynaptic cell, inferring a target-specific mechanism through which the extinction circuit can directly interact with and suppress the fear circuit in the BA. Together, our findings add to the understanding of the complex fear circuit and the synaptic mechanisms underlying fear extinction.
Thesis (Ph.D.)--Tufts University, 2015.
Submitted to the Dept. of Neuroscience.
Advisor: Leon Reijmers.
Committee: Peter Juo, Jamie Maguire, and Maribel Rios.