Role of the amygdala in processing auditory social signals

Rats emit 22 kHz and 50 kHz ultrasonic vocalizations (USVs) respectively, during negative and positive affective states. USVs are innately emitted in distinct situations and mediated by different neurochemical systems. It remains unclear whether behavioral responses to USVs are innate or acquired. Furthermore little is known about the neurobiology of responding to USVs. The first study (chapter 2) quantified behavioral and neuronal responses to USVs in experimentally-naive rats. Neuronal responses were analyzed in the amygdala (AM) because this structure is essential for emotional learning and memory. Playback of 22 kHz USVs and 50 kHz USVs respectively, elicited sustained increases and decreases in the firing rates of AM neurons. Perhaps surprisingly, 22 kHz USVs, which are sometimes called "alarm cries", did not elicit freezing behavior in experimentally-naive rats. The second study (chapter 3) investigated how rats learn to freeze to alarm cries. A prior aversive experience was necessary for freezing to 22 kHz USVs. Rats learned to freeze to alarm cries through a mechanism dubbed "auto-conditioning". This non-social learning mechanism was robust, reliable, and suitably stimulus-specific to be ethologically adaptive. The third study (chapter 4) examined neuronal responses in the lateral amygdala (LA) to 22 kHz and 50 kHz USVs as a function of auditory fear conditioning. A multidimensional classification scheme revealed several novel types of plasticity in LA neurons. First, conditioning increased the complexity of elicited firing patterns. Second, conditioning converted some unresponsive neurons to responsive ones and some responsive neurons to unresponsive ones. These opposing changes may reflect homeostatic mechanisms that stabilize network activity in the face of plasticity. Third, several neurons exhibited firing-rate increases during the precise interval in which the unconditional stimulus occurred during conditioning. These "US-timed" responses suggested that LA neurons can encode the inter-stimulus interval. Perhaps surprisingly, 22 kHz and 50 kHz USVs were equally effective in eliciting conditional freezing behavior. Overall this dissertation demonstrated that freezing to 22 kHz USVs is not innate, but acquired by virtue of autoconditioning. Moreover, auditory social signals elicit a surprisingly rich variety of firing patterns and plasticity in LA.