No abstract

In contrast to easily formed fear memories, fear extinction requires prolonged training. The prelimbic cortex (PL), which integrates signals from brain structures involved in fear conditioning and extinction such as the ventral hippocampus (vHIP) and the basolateral amygdala (BL), is necessary for fear memory retrieval. Little is known, however, about how the vHIP and BL inputs to the PL regulate the display of fear after fear extinction. Using functional anatomy tracing in male rats, we found two distinct subpopulations of neurons in the PL activated by either the successful extinction or the relapse of fear. During the retrieval of fear extinction memory, the dominant input to active neurons in the PL was from the vHIP, whereas the retrieval of fear memory, regardless of the age of a memory and testing context, was associated with greater BL input. Optogenetic stimulation of the vHIP-PL pathway after one session of fear extinction increased conditioned fear, whereas stimulation of the vHIP inputs after several sessions of extinction decreased the conditioned fear response. This latter effect was, however, transient, as stimulation of this pathway 28 d after extinction increased conditioned fear response again. The results show that repeated fear extinction training gradually changes vHIP-PL connectivity, making fear suppression possible, whereas in the absence of fear suppression from the vHIP, signals from the BL can play a dominant role, resulting in high levels of fear.

Social support during exposure-based psychotherapy is believed to diminish fear and improve therapy outcomes. However, some clinical trials challenge that notion. Underlying mechanisms remain unknown, hindering the understanding of benefits and pitfalls of such approach. To study social buffering during fear extinction, we developed a behavioral model in which partner’s presence decreases response to fear-associated stimuli. To identify the neuronal background of this phenomenon, we combined behavioral testing with c-Fos mapping, optogenetics, and chemogenetics. We found that the presence of a partner during fear extinction training causes robust inhibition of freezing; the effect, however, disappears in subjects tested individually on the following day. It is accompanied by lowered activation of the prelimbic (PL) and anterior cingulate (ACC) but not infralimbic (IL) cortex. Accordingly, blocking of IL activity left social buffering intact. Similarly, inhibition of the ventral hippocampus–PL pathway, suppressing fear response after prolonged extinction training, did not diminish the effect. In contrast, inhibition of the ACC–central amygdala pathway, modulating social behavior, blocked social buffering. By reporting that social modulation of fear inhibition is transient and insensitive to manipulation of the fear extinction-related circuits, we show that the mechanisms underlying social buffering during extinction are different from those of individual extinction.