Effects of prediction error and acute stress on retrieval-extinction of fear memories of different strengths

Abstract

<p id="p00005">Under the framework of Reconsolidation Interference of conditioned fear memory, Prediction Error (PE) has been demonstrated as a necessary condition of memory destabilization. However, the role of PE in destabilizing fear memories of different strengths is unclear. The degree of PE that is needed to effectively reactivate fear memories may be dependent on the strength of memory. It is unknown whether the PE used to reactivate weak memories is also effective in destabilizing stronger memories. Memory strength was proved to be an important boundary condition of memory reconsolidation; however, explorations of solutions to overcome the boundary are rare. Among factors that are possible to help to overcome the boundary condition, the effects of stress hormones are worth exploring. However, the manipulation of memory strength in human laboratory studies has not been well developed. Thus, the present study has three main aims: (1) to test the effect of fear memory strength in a human laboratory setting based on previous results in animal studies; (2) to examine the effect of PE during reactivation on destabilizing different strength memories and (3) to test the possible influence of post-reactivation exogenous stress to the retrieval-extinction of fear memories. <p id="p00010">These results indicate that PE used to destabilize weak memories is insufficient to destabilize strong memories; and that post-reactivation acute stress cannot nullify this deficit which is due to boundary conditions (e.g., strength). We discuss possible interpretations of these results and the implications for the translation of retrieval-extinction to clinical practice. <p id="p00015">The three days retrieval-extinction paradigm was adopted in the present study. We manipulated memory strength through two kinds of acquisition procedures on the first day, which varied the predictability of the unconditioned stimulus (US) occurrence after the conditioned stimulus (CS). Twenty-four hours later, a reminder containing a single PE was used to reactivate memories, which was followed by a stress task (Social Evaluate Cold Pressor test, SECPT) or not before extinction. After 24 hours, a test of spontaneous recovery and reinstatement was utilized to measure the return of fear in each condition. All participants were divided into three conditions: CS-Predictable US_no Stress Group, CS-Unpredictable US_no Stress Group and the CS-Unpredictable US_Stress Group. Skin conductance response (SCR) and fear-potentiated startle response (FPS) were used as measurements of conditioned fear. <p id="p00020">The results showed that there was a relatively stronger increase in fear response (SCR) from Day 1 to Day 2 in the CS-Unpredictable US condition than the CS-Predictable US condition, which may suggest a difference in memory strength between conditions. And for the weak fear memory (CS-predictable US), the reactivation that involved a single PE and was followed by extinction training prevented the spontaneous recovery, especially on the SCR measurement. On the other hand, in the enhanced memory condition (CS-unpredictable US), the extinguished memory returned in the memory test on the third day, which suggests a failure of memory destabilization. Furthermore, when the post-reactivation acute stress task was adopted in the enhanced memory condition, the return of fear further increased, compared with the no stress manipulation conditions.