Psychophysiology of Delayed Extinction and Reconsolidation in Humans

1. BACKGROUND AND SIGNIFICANCE 1.1. Reconsolidation and its modification. Recent animal research suggests that reactivation (retrieval) of a consolidated memory can return it to a labile state from which it must be restabilized in order to persist. This stabilization process has been termed "reconsolidation," and various behavioral and pharmacologic interventions have been found to modify or block it. Although extinction can reduce or even apparently eliminate conditioned fear, it typically does so by inhibiting the expression of the fear memory, not by erasing it. In contrast, blockade of reconsolidation is believed to erase, or at least diminish, the fear memory trace. It is not uncommon for reconsolidation articles to conclude with a statement that this offers novel therapeutic possibilities for PTSD, which may in part be characterized as a disorder of over-consolidated, persistently disturbing memories that do not go away.1
2. SPECIFIC AIM The aim of this project is to create an experimental assay in the form of an optimal Pavlovian differential fear conditioning paradigm, within which the relative strengths of various novel pharmacological and behavioral, reconsolidation-blocking interventions can be tested. As impressive as the Kindt8 and Schiller10 studies are for demonstrating pharmacological and behavioral fear reconsolidation-blocking interventions in normal humans, the results produced by these interventions make them nearly useless as potential assays for reconsolidation blockade. Specifically, both interventions produced an apparent total abolition of the fear memory, i.e., they both contain a floor effect. The aim in this work is to develop a normal human paradigm within which both propranolol and delayed extinction produce significant, but only partial reduction of the CR, which might then be improved upon by more powerful reconsolidation-blocking interventions. This study will not be used to support a label, advertising, or indication change of propanolol, and will not be prescribed in a patient population, route of administration, or dosage that significantly increases risk.

2.1. Elements of a new design to eliminate the floor effect 2.1.1. A stronger US. One way to produce a CR that is more resistant to reconsolidation blockade, i.e., to reduce the floor effect, would be to use a stronger US. However, because in our ongoing human conditioning studies, we are already using an electric shock selected in advance by subjects to be "highly annoying but not painful," we believe that to use a level of shock above this would be ethically impermissible, and hence we will not pursue this possibility.

2.1.2. A more highly "prepared" CS. It has long been shown that certain classes of CSs, when paired with a US, produce a stronger fear CR, i.e., they are more "prepared" to enter into an association with the US.13 The Schiller design10 did not use a prepared CS, but the Kindt study8 did, viz., still pictures of spiders. In our design, we will enhance the preparedness of the CS by using 12-sec.14 high-definition video clips of three different crawling tarantulas, each conspicuously different in appearance. Importantly, however, the present project will not be a study of spider fear or phobia; we will only use high-definition videoto enhance CS preparedness.

2.1.3. More sensitive subjects. We will limit recruitment to subject candidates who fall within the upper half of the distribution of normal humans on the Spider Phobia Questionnaire-1515 (SPQ-15) and for whom the CSs are likely to be especially salient. However in order to avoid potential adverse clinical consequences of participation, we will exclude subjects with features of diagnosable spider phobia (or the presence of any other Axis I mental disorder).

2.1.4. Selecting subjects with stronger CRs. In order for a subject who has completed the Day 1 Acquisition phase of the experiment to proceed to the remaining phases, we will require that they show strong evidence of conditioning, as manifest in CRs greater than 0.25 µSiemens to at least two CS+s. The Schiller study10 employed a differential conditioning cut-off of 0.10 µSiemens, but based upon our own experience, we favor the raw score cut-off. The Kindt study8 employed no such cut-off.