An N-methyl-D-aspartate receptor agonist facilitates sleep-independent synaptic plasticity associated with working memory capacity enhancement

Abstract

Working memory (WM) capacity improvement is impacted by sleep, and possibly by N-methyl-D-aspartate (NMDA) agonists such as D-cycloserine (DCS), which also affects procedural skill performance. However, the mechanisms behind these relationships are not well understood. In order to investigate the neural basis underlying relationships between WM skill learning and sleep, DCS, and both sleep and DCS together, we evaluated training-retest performances in the n-back task among healthy subjects who were given either a placebo or DCS before the task training, and then followed task training sessions either with wakefulness or sleep. DCS facilitated WM capacity enhancement only occurring after a period of wakefulness, rather than sleep, indicating that WM capacity enhancement is affected by a cellular heterogeneity in synaptic plasticity between time spent awake and time spent asleep. These findings may contribute to development, anti-aging processes, and rehabilitation of higher cognition.

Figure 1. Experimental schedules for each of the 6 treatment groups.

Subjects were assigned to 3 pairs of randomized, blind, placebo-controlled, 2-arm experimental groups (A–F). Each group underwent a specific schedule consisting of an initial training session and a retest session. Subjects were administered encapsulated DCS (groups B, E, and F) or a placebo (groups A, C, and D) on an empty stomach 1.5 h prior to the training session. Subjects in groups A and B were placed on a diurnal experimental schedule, in which they were trained at 20:00 h and were then retested at 20:00 h the next day (e.g., 24 h later). Subjects in groups C–F were placed on a semidiurnal experimental schedule. Individuals in groups C and E were trained at 20:00 h and were retested at 08:00 h the following day (e.g., 12 h later) after a night of sleep. Individuals in groups D and F were trained at 08:00 h and were retested at 20:00 h after 12 h of wakefulness. All subjects performed spatial n-back WM tasks (n = 1–7) during each test session. All post-sleep retests were performed at least 1 h after awakening.

Figure 2. Changes in working memory (WM) capacity during the training period.

Subjects were administered either a placebo (open circles with dotted line; groups A, C, and D) or D-cycloserine (DCS; closed circles with continuous line; groups B, E, and F). Subjects in both groups showed significant improvements in n-back accuracy levels across the 3 trials, but the rates of improvement were not related to medication. By the end of training, rates of improvement in both groups had nearly plateaued, and accuracy levels in the 2 groups differed by only 0.06 ± 0.18 [mean ± standard error of the mean (SEM)]. The heights of the circles and error bars represent means and SEMs, respectively. Asterisks (**) indicate p < 0.0001.

Figure 3. Changes in working memory (WM) capacity between and within testing sessions in experimental groups on the diurnal schedule (24-h interval between the testing sessions).

(A) The n-back accuracy level significantly improved between the training and retest periods in both groups (p < 0.0001). However, the significant session × medication interaction (p = 0.041) suggests that DCS administration facilitated the greater improvements observed in group B. Light and dark green bars represent n-back accuracy levels at the training and retest sessions, respectively. (B) Inter-session improvements in n-back accuracy level in groups A and B. The n-back accuracy level improved significantly more (p < 0.005, indicated by the asterisk) in group B, which received the DCS treatment, than in group A, which received the placebo treatment. The heights of the brown bars and error bars represent the mean and SEM, respectively.

Figure 4. Changes in working memory (WM) capacity between and within testing sessions in experimental groups on the semidiurnal schedule (12-h interval between testing sessions).

(A) WM capacity significantly improved (p < 0.005) between the training and retest sessions for groups C–F. Light and dark green bars represent n-back accuracy levels at the training and retest sessions, respectively. (B) Inter-session differences in n-back accuracy levels in groups C–F. In groups receiving the placebo (PLC) treatment (C, D), sleep seemed to facilitate improvement in accuracy more than wakefulness. However, improvements in n-back accuracy levels were similar in groups receiving the D-cycloserine (DCS) treatment, regardless of whether they followed training with a period of wakefulness (F) or sleep (E). Blue and yellow bars represent changes in n-back accuracy level during sleep and wake periods, respectively. The heights of the bars and error bars represent the mean and SEM, respectively. The crosses (†) indicate a trend toward significance (p < 0.05).