Fig. 9.

Multiple superimposed layers of inhibition between the input and output stations of the amygdala. Scheme summarizes the major direct excitatory (solid) and indirect inhibitory (shaded) pathways linking the basolateral amygdala to the central nucleus. LA neurons target ITCd cells (arrow 1) that in turn inhibit CeL neurons (arrow 2), whereas BL neurons drive ITCv cells (arrow 3), resulting in the inhibition of CeM neurons (arrow 4). In addition, there are inhibitory projections from ITCd to ITCv cells (arrow 5), as well as from CeL to CeM (arrow 6). Note that the recruitment of ITCd cells by LA neurons during the conditioned stimulus should produce a parallel inhibition of ITCv cells and CeL neurons, resulting in the disinhibition of fear output neurons in CeM. We hypothesize that fear conditioning produces a potentiation of LA inputs to ITCd cells, whereas extinction training causes a potentiation of BL inputs to ITCv neurons. The latter view is in apparent contradiction with the fact that IL projects to both ITCd and ITCv clusters and that ITCd cells inhibit ITCv neurons. This raises the following question: can IL inputs to ITCv cells overcome the inter-ITC inhibition to inhibit CeM cells? According to a recent modeling study (Li et al. 2011), IL inputs are so strong that they can overcome the inter-ITC inhibition to cause a marked increase in the firing rate of ITCv cells, leading to a persistent decrease in CeM output. Last, it should be mentioned that to improve readability, the scheme does not include 2 well-established facts about this circuit, namely, the existence of BL projections to CeL (Krettek and Price 1978) and the presence of two CeL cell types with opposite responses to conditioned stimuli but unknown connectivity with ITC cells. The latter are discussed in the text.