Limits...
NREM2 and Sleep Spindles Are Instrumental to the Consolidation of Motor Sequence Memories.

Laventure S, Fogel S, Lungu O, Albouy G, Sévigny-Dupont P, Vien C, Sayour C, Carrier J, Benali H, Doyon J - PLoS Biol. (2016)

Bottom Line: Results show that the Cond-NREM2 group had significantly higher gains in performance at retest than both the Cond-REM and NoCond groups.Also, only the Cond-NREM2 group yielded significant changes in sleep spindle characteristics during cueing.Finally, we found that a change in frequency of sleep spindles during cued-memory reactivation mediated the relationship between the experimental groups and gains in performance the next day.

View Article: PubMed Central - PubMed

Affiliation: Department of Psychology, University of Montreal, Montreal, Quebec, Canada.

ABSTRACT
Although numerous studies have convincingly demonstrated that sleep plays a critical role in motor sequence learning (MSL) consolidation, the specific contribution of the different sleep stages in this type of memory consolidation is still contentious. To probe the role of stage 2 non-REM sleep (NREM2) in this process, we used a conditioning protocol in three different groups of participants who either received an odor during initial training on a motor sequence learning task and were re-exposed to this odor during different sleep stages of the post-training night (i.e., NREM2 sleep [Cond-NREM2], REM sleep [Cond-REM], or were not conditioned during learning but exposed to the odor during NREM2 [NoCond]). Results show that the Cond-NREM2 group had significantly higher gains in performance at retest than both the Cond-REM and NoCond groups. Also, only the Cond-NREM2 group yielded significant changes in sleep spindle characteristics during cueing. Finally, we found that a change in frequency of sleep spindles during cued-memory reactivation mediated the relationship between the experimental groups and gains in performance the next day. These findings strongly suggest that cued-memory reactivation during NREM2 sleep triggers an increase in sleep spindle activity that is then related to the consolidation of motor sequence memories.

Show MeSH
Sleep spindle results.(A) Changes in parietal sleep spindle characteristics. Standardized (Z-score) differences in sleep spindle characteristics between the pre-matched and during-stimulation periods at Pz. Only the Cond-NREM2 group showed significant increase in spindle amplitude, frequency, and duration. Amplitude and frequency were significantly different between the Cond-NREM2 and NoCond groups. (B) Changes in the number of spindles at Pz in specific frequency ranges. Differences at Pz in the number of spindles categorized by frequency range between the pre-matched and during-stimulation sleep periods. Significance was determined using a Chi2 analysis with Bonferroni correction on the number of bins. The total number of spindles detected for each group and sleep period is shown in the “Spindle count” table. The results revealed a significant decrease of spindle in the 11–11.49 Hz range, but an increase in the 13.5–13.99 Hz range. (C) Changes in frequency at Pz mediates the relationship between the TMR protocol and MSL offline gains. The significant relation between the experimental protocol and the gains in performance on the MSL task (relation c) disappeared when the change in frequency in sleep spindles over the parietal cortex between the pre-matched and during-stimulation periods were included in the mediation model (direct effect: relation c’). The indirect effect composed of (1) the experimental protocol and the change in spindle frequency (relation a) and (2) the change in spindle frequency and the MSL offline gains (relation b) was significant, as demonstrated by the bootstrap analysis (Cl .004, .077). This pattern of results strongly suggests that sleep spindles occurring in the parietal regions are crucial to motor memory consolidation through an increase of spindles of higher frequency. Data deposited in the Dryad repository: http://dx.doi.org/10.5061/dryad.b4t60 [40]. * p < 0.05; ** p < 0.01; *** p < 0.001
© Copyright Policy
Related In: Results  -  Collection

License
getmorefigures.php?uid=PMC4816304&req=5

pbio.1002429.g003: Sleep spindle results.(A) Changes in parietal sleep spindle characteristics. Standardized (Z-score) differences in sleep spindle characteristics between the pre-matched and during-stimulation periods at Pz. Only the Cond-NREM2 group showed significant increase in spindle amplitude, frequency, and duration. Amplitude and frequency were significantly different between the Cond-NREM2 and NoCond groups. (B) Changes in the number of spindles at Pz in specific frequency ranges. Differences at Pz in the number of spindles categorized by frequency range between the pre-matched and during-stimulation sleep periods. Significance was determined using a Chi2 analysis with Bonferroni correction on the number of bins. The total number of spindles detected for each group and sleep period is shown in the “Spindle count” table. The results revealed a significant decrease of spindle in the 11–11.49 Hz range, but an increase in the 13.5–13.99 Hz range. (C) Changes in frequency at Pz mediates the relationship between the TMR protocol and MSL offline gains. The significant relation between the experimental protocol and the gains in performance on the MSL task (relation c) disappeared when the change in frequency in sleep spindles over the parietal cortex between the pre-matched and during-stimulation periods were included in the mediation model (direct effect: relation c’). The indirect effect composed of (1) the experimental protocol and the change in spindle frequency (relation a) and (2) the change in spindle frequency and the MSL offline gains (relation b) was significant, as demonstrated by the bootstrap analysis (Cl .004, .077). This pattern of results strongly suggests that sleep spindles occurring in the parietal regions are crucial to motor memory consolidation through an increase of spindles of higher frequency. Data deposited in the Dryad repository: http://dx.doi.org/10.5061/dryad.b4t60 [40]. * p < 0.05; ** p < 0.01; *** p < 0.001

Mentions: By contrast, when we compared changes in spindle characteristics between the pre-matched and during-stimulation sleep periods, the one-way ANOVA comparing percent change (Δ%) revealed a significant difference in peak amplitude (F1, 41 = 5.257, p = .03) and peak frequency (F1, 41 = 4.842, p = .03) between the Cond-NREM2 and NoCond groups (see Table 2 for details). A similar pattern of results was observed for spindles duration, although the effect did not reach significance (F1, 41 = 3.523, p = .07). Follow-up, one-sample t tests revealed that only the Cond-NREM2 group had a significant increase in Δ% in peak frequency (Cond-NREM2: t20 = 2.443, p = .02; NoCond: t21 = -.511, p = .62), Δ% in peak amplitude (Cond-NREM2: t20 = 2.394, p = .03; NoCond: t21 = -.481, p = .64), and Δ% in duration (Cond-NREM2: t20 = 3.013, p = .007; NoCond: t21 = .881, p = .39) (Fig 3A).


NREM2 and Sleep Spindles Are Instrumental to the Consolidation of Motor Sequence Memories.

Laventure S, Fogel S, Lungu O, Albouy G, Sévigny-Dupont P, Vien C, Sayour C, Carrier J, Benali H, Doyon J - PLoS Biol. (2016)

Sleep spindle results.(A) Changes in parietal sleep spindle characteristics. Standardized (Z-score) differences in sleep spindle characteristics between the pre-matched and during-stimulation periods at Pz. Only the Cond-NREM2 group showed significant increase in spindle amplitude, frequency, and duration. Amplitude and frequency were significantly different between the Cond-NREM2 and NoCond groups. (B) Changes in the number of spindles at Pz in specific frequency ranges. Differences at Pz in the number of spindles categorized by frequency range between the pre-matched and during-stimulation sleep periods. Significance was determined using a Chi2 analysis with Bonferroni correction on the number of bins. The total number of spindles detected for each group and sleep period is shown in the “Spindle count” table. The results revealed a significant decrease of spindle in the 11–11.49 Hz range, but an increase in the 13.5–13.99 Hz range. (C) Changes in frequency at Pz mediates the relationship between the TMR protocol and MSL offline gains. The significant relation between the experimental protocol and the gains in performance on the MSL task (relation c) disappeared when the change in frequency in sleep spindles over the parietal cortex between the pre-matched and during-stimulation periods were included in the mediation model (direct effect: relation c’). The indirect effect composed of (1) the experimental protocol and the change in spindle frequency (relation a) and (2) the change in spindle frequency and the MSL offline gains (relation b) was significant, as demonstrated by the bootstrap analysis (Cl .004, .077). This pattern of results strongly suggests that sleep spindles occurring in the parietal regions are crucial to motor memory consolidation through an increase of spindles of higher frequency. Data deposited in the Dryad repository: http://dx.doi.org/10.5061/dryad.b4t60 [40]. * p < 0.05; ** p < 0.01; *** p < 0.001
© Copyright Policy
Related In: Results  -  Collection

License
Show All Figures
getmorefigures.php?uid=PMC4816304&req=5

pbio.1002429.g003: Sleep spindle results.(A) Changes in parietal sleep spindle characteristics. Standardized (Z-score) differences in sleep spindle characteristics between the pre-matched and during-stimulation periods at Pz. Only the Cond-NREM2 group showed significant increase in spindle amplitude, frequency, and duration. Amplitude and frequency were significantly different between the Cond-NREM2 and NoCond groups. (B) Changes in the number of spindles at Pz in specific frequency ranges. Differences at Pz in the number of spindles categorized by frequency range between the pre-matched and during-stimulation sleep periods. Significance was determined using a Chi2 analysis with Bonferroni correction on the number of bins. The total number of spindles detected for each group and sleep period is shown in the “Spindle count” table. The results revealed a significant decrease of spindle in the 11–11.49 Hz range, but an increase in the 13.5–13.99 Hz range. (C) Changes in frequency at Pz mediates the relationship between the TMR protocol and MSL offline gains. The significant relation between the experimental protocol and the gains in performance on the MSL task (relation c) disappeared when the change in frequency in sleep spindles over the parietal cortex between the pre-matched and during-stimulation periods were included in the mediation model (direct effect: relation c’). The indirect effect composed of (1) the experimental protocol and the change in spindle frequency (relation a) and (2) the change in spindle frequency and the MSL offline gains (relation b) was significant, as demonstrated by the bootstrap analysis (Cl .004, .077). This pattern of results strongly suggests that sleep spindles occurring in the parietal regions are crucial to motor memory consolidation through an increase of spindles of higher frequency. Data deposited in the Dryad repository: http://dx.doi.org/10.5061/dryad.b4t60 [40]. * p < 0.05; ** p < 0.01; *** p < 0.001
Mentions: By contrast, when we compared changes in spindle characteristics between the pre-matched and during-stimulation sleep periods, the one-way ANOVA comparing percent change (Δ%) revealed a significant difference in peak amplitude (F1, 41 = 5.257, p = .03) and peak frequency (F1, 41 = 4.842, p = .03) between the Cond-NREM2 and NoCond groups (see Table 2 for details). A similar pattern of results was observed for spindles duration, although the effect did not reach significance (F1, 41 = 3.523, p = .07). Follow-up, one-sample t tests revealed that only the Cond-NREM2 group had a significant increase in Δ% in peak frequency (Cond-NREM2: t20 = 2.443, p = .02; NoCond: t21 = -.511, p = .62), Δ% in peak amplitude (Cond-NREM2: t20 = 2.394, p = .03; NoCond: t21 = -.481, p = .64), and Δ% in duration (Cond-NREM2: t20 = 3.013, p = .007; NoCond: t21 = .881, p = .39) (Fig 3A).

Bottom Line: Results show that the Cond-NREM2 group had significantly higher gains in performance at retest than both the Cond-REM and NoCond groups.Also, only the Cond-NREM2 group yielded significant changes in sleep spindle characteristics during cueing.Finally, we found that a change in frequency of sleep spindles during cued-memory reactivation mediated the relationship between the experimental groups and gains in performance the next day.

View Article: PubMed Central - PubMed

Affiliation: Department of Psychology, University of Montreal, Montreal, Quebec, Canada.

ABSTRACT
Although numerous studies have convincingly demonstrated that sleep plays a critical role in motor sequence learning (MSL) consolidation, the specific contribution of the different sleep stages in this type of memory consolidation is still contentious. To probe the role of stage 2 non-REM sleep (NREM2) in this process, we used a conditioning protocol in three different groups of participants who either received an odor during initial training on a motor sequence learning task and were re-exposed to this odor during different sleep stages of the post-training night (i.e., NREM2 sleep [Cond-NREM2], REM sleep [Cond-REM], or were not conditioned during learning but exposed to the odor during NREM2 [NoCond]). Results show that the Cond-NREM2 group had significantly higher gains in performance at retest than both the Cond-REM and NoCond groups. Also, only the Cond-NREM2 group yielded significant changes in sleep spindle characteristics during cueing. Finally, we found that a change in frequency of sleep spindles during cued-memory reactivation mediated the relationship between the experimental groups and gains in performance the next day. These findings strongly suggest that cued-memory reactivation during NREM2 sleep triggers an increase in sleep spindle activity that is then related to the consolidation of motor sequence memories.

Show MeSH