Sleep enhances reconsolidation-based strengthening of visuospatial memories

Abstract

Consolidated memories can be returned to a labile state upon reactivation. The re-stabilization of reactivated memories, or reconsolidation, can allow for change in previously established memories. Given the role of sleep in the initial consolidation of memories, sleep may be important for reconsolidation as well. However, effects of sleep on reconsolidation and specific aspects of sleep that may contribute are unclear. Here, participants learned 30 picture-location pairs. After overnight sleep, initial consolidation was tested. Following either one day (Experiment 1) or one week (Experiment 2), participants were tested again to reactivate their memory and then learned 30 novel picture-location pairs. Control groups (Experiment 1) received no reactivation prior to new learning. Twelve hours later, after daytime wakefulness or overnight sleep, participants completed a final memory test. Sleep participants underwent polysomnography between reactivation and final tests. In Experiment 1, reactivation led to preservation of memory compared to no reactivation. Sleep was associated with less post-reactivation memory decline than waking, with memory preservation positively related to time spent in non-rapid-eye movement sleep. In Experiment 2, sleep was associated with greater post-reactivation memory improvement than waking, with improvement positively related to sigma activity. These results suggest sleep enhances reconsolidation-based strengthening of episodic memories.

Figure 1

Experimental timeline and procedure. (a) In Experiment 1, participants encoded Set 1 in the evening of day 1 and were tested on Set 1 the morning of day 2 (12-h recall). On day 3 in the morning (wake group) or the evening (sleep group), participants in the reactivation groups were tested on Set 1 and then encoded Set 2, whereas participants in the control groups only encoded Set 2. Twelve hours later, participants were given a final test first on Set 1, then on Set 2 (final recall). (b) In Experiment 2, participants encoded Set 1 in the evening of day 1 and were tested on Set 1 the morning of day 2 (12-h recall). On day 8 in the morning (wake group) or the evening (sleep group), participants were tested on Set 1 prior to encoding Set 2. Twelve hours later, participants were given a final test first on Set 1, then on Set 2 (final recall).

Figure 2

Visuospatial task. During passive encoding, participants were shown the locations of 30 pictures. During active encoding, participants selected a location for each picture and were shown the correct location as feedback. During recall, participants selected a location for each picture and were not provided feedback.

Figure 3

Effect of sleep on reconsolidation-based memory strengthening. (a) Average Set 1 memory change between reactivation recall and final recall in Experiment 1. (b) Relationship between time in NREM sleep (NREM2 and SWS) and Set 1 memory change between reactivation recall and final recall in the sleep group in Experiment 1. (c) Average Set 1 memory change between reactivation recall and final recall in Experiment 2. (d) Relationship between NREM sigma density and Set 1 memory change between reactivation recall and final recall in the sleep group in Experiment 2. Error bars represent standard errors of means.

Figure 4

Consolidation of new learning. Average Set 2 memory change between encoding and final recall in Experiment 1 (a) and Experiment 2 (b). Error bars represent standard errors of means.