Emotional learning retroactively promotes memory integration through rapid neural reactivation and reorganization
- State Key Laboratory of Cognitive Neuroscience and Learning & IDG/McGovern Institute for Brain Research, Beijing Normal University, China
- Beijing Key Laboratory of Brain Imaging and Connectomics, Beijing Normal University, China
- Donders Institute for Brain, Cognition and Behaviour, Radboud University Medical Center, Netherlands
- School of Education, Chongqing Normal University, China
- Chinese Institute for Brain Research, China
Figures
Figure 1 with 6 supplements
Experimental design and behavioral performance.
(A) The experiment consisted of three phases. During the initial learning phase, participants were instructed to vividly imagine each face and its paired object interacting with each other. During …
Figure 1—figure supplement 1
Emotional ratings on the four voices.
Bar graphs depict averaged arousal (left) and valence (right) ratings of each voice clip in a pilot experiment with independent subjects (n=19). Error bars represent standard error of mean. Notes: …
Figure 1—figure supplement 2
Face item memory performance.
(A) Item memory performances in Study 1 (n=30), 2 (n=27, data from one participant was missing due to an equipment failure with stimuli presentation software), and 3 (fMRI, n=28). About 30 min after …
Figure 1—figure supplement 3
Associative memory performances on the four confidence levels.
Associative memory performances on each confidence rating, with 1 corresponding to very unconfident and 4 corresponding to very confident, in the three studies. Bar graphs depict averaged …
Figure 1—figure supplement 4
Relationships between vividness ratings and confidence ratings.
Trial-level relationships between vividness ratings during initial learning and subsequent confidence ratings during memory test in the three studies. Blue lines indicate the best linear fit and …
Figure 1—figure supplement 5
Low-confidence memory performance as a function of vividness and emotion.
Bar graphs depict averaged correct proportion for face-object associations remembered with low confidence, which were rated with low (LV) and high (HV) vividness during initial learning phase in …
Figure 1—figure supplement 6
Skin conductance levels (SCLs) in Study 3 (fMRI).
Bar graphs depict averaged SCLs in aversive and neutral conditions during initial and emotional learning phases separately (n=20, data from 8 participants were incomplete due to an equipment failure …
Figure 2 with 6 supplements
Trial-level neural reactivation of initial learning activity during emotional learning.
(A) An illustration of trial-level reactivation analysis. Example data was from one subject. During initial learning (left), sagittal views of activation maps for four trials were shown. During …
Figure 2—figure supplement 1
Brain systems involved in memory encoding effect during initial and emotional learning phases.
Widespread brain activation associated with an encoding effect (i.e. all encoding trials with 2 s duration from onset of each face-object/ face-voice association vs. fixation) during (A) the initial …
Figure 2—figure supplement 2
Condition-level neural reactivation of initial learning activity during emotional learning.
(A) An illustration of condition-level reactivation analysis. Example data was from one subject. (B) Bar graphs depict the average pattern similarities in aversive and neutral conditions during the …
Figure 2—figure supplement 3
Trial-level pattern similarity analysis approach.
Several similarity measures were computed for each trial. (1) Pair-specific similarity (black) was computed by correlating each face-voice pair’s multivoxel activity pattern during emotional …
Figure 2—figure supplement 4
Additional trial-level pattern similarity results.
(A–C) The bilateral hippocampal, vLOC and FFA ROIs used in trial-level pattern similarity analyses. (D–F) Bar graphs depict the average pair-specific, across-pair within-condition and across-pair …
Figure 2—figure supplement 5
Prediction relationships between hippocampal between-phases pattern similarity and associative memory performance.
(A) Scatter plots depict correlations between observed memory performance (i.e. face-object associative memory with high confidence) and predicted outcomes from machine-learning prediction analysis …
Figure 2—figure supplement 6
Prediction relationships between hippocampal within-phase pattern similarity and associative memory performance.
(A) Scatter plots depict correlations between observed memory performance (i.e. face-object associative memory with high confidence) and predicted outcomes from machine-learning prediction analysis …
Figure 3 with 3 supplements
Hippocampal connectivity with the amygdala and neocortical regions during emotional learning phase accounting for emotion-charged retroactive memory enhancement.
(A) The bilateral hippocampal seed used in task-dependent gPPI functional connectivity analysis (i.e. the same hippocampal ROI used in pattern similarity analysis). (B–D) Significant clusters in the …
Figure 3—figure supplement 1
Hippocampal connectivity in aversive and neutral conditions during initial and emotional learning phases.
Bar graphs depict averaged hippocampal connectivity with the right amygdala, left mFFA and left sLOC in aversive and neutral conditions during initial and emotional learning phases separately. Error …
Figure 3—figure supplement 2
Prediction relationships between hippocampal connectivity and associative memory performance in the neutral condition.
Scatter plots depict correlations of observed associative memory performance (i.e. remembered with high confidence) with predicted outcomes from prediction analysis based on (A) hippocampal-amygdala …
Figure 3—figure supplement 3
Alternative models for the relationships among hippocampal connectivity and associative memory in the aversive condition.
(A) Model 1 with hippocampal-mFFA and -sLOC connectivity as separate mediators showed no any reliable mediating effect, including (i) indirect effect (c*d): β=0.153, p=0.138, 95% CI: [–0.049, …
Figure 4 with 1 supplement
Hippocampal- and parahippocampal-neocortical connectivity at post-learning rests in relation to emotion-charged retroactive memory enhancement.
(A) An illustration of hippocampal-seeded functional connectivity analyses, and sagittal views of hippocampal connectivity maps at the group level for three rest scans (n=27, data from one …
Figure 4—figure supplement 1
Post-encoding hippocampal connectivity changes in relation to memory performances in aversive and neutral conditions.
Scatter plots depict partial correlations of (A) hippocampal-iLOC and (B) hippocampal-IPL/aPFC/PCC/mPFC connectivity changes with memory in the aversive condition controlling for neutral memory …
Author response image 1
Post-encoding hippocampal connectivity changes in relation to memory difference (i.e., aversive vs. neutral).
(A) Significant cluster in the lateral occipital cortex (LOC), showing its greater connectivity with the hippocampus at Rest 2 (vs. Rest 1) in positive relation to the memory difference. (B) …
Author response image 2
Hippocampal connectivity changes from baseline rest (Rest 1) to post-initial-learning rest (Rest 2) in relation with general/non-arousal memory performance.
(A) Significant clusters show their greater connectivity with the hippocampus at Rest 2 relative to Rest 1, with average memory across aversive and neutral conditions as the covariate of interest. …
Author response image 3
Additional condition-level pattern similarity results.
Bar graphs depict the average pattern similarities in aversive and neutral conditions during the presentations of face-voice associations (i.e., Face + Voice) separately for the bilateral …
Tables
Author response table 1
Post-encoding PPA connectivity changes in negative relation to memory in the aversive but not neutral condition.
| Brain Regions | Hemisphere | T values | MNI Coordinates | ||
| X | Y | Z | |||
| Rest 2 vs. Rest 1 | |||||
| Angular | R | -4.74 | 44 | -58 | 40 |
| Precuneus | R | -7.12 | 10 | -52 | 28 |
| Rest 3 vs. Rest 2 | |||||
| Superior frontal gyrus | L | -4.93 | -20 | -4 | 74 |
| Inferior frontal gyrus | L | -5.38 | -36 | 30 | 14 |
| Inferior Parietal Lobule | L | -4.61 | -40 | -50 | 58 |
| Supramarginal gyrus | L | -4.83 | -52 | -30 | 28 |
| Precentral gyrus | L | -8.48 | -50 | 0 | 6 |
| L | -4.62 | -34 | -2 | 62 | |
| Lingual gyrus | L | -5.17 | -12 | -90 | -8 |
Author response table 2
Hippocampal connectivity changes at post-initial-learning rest (Rest 2 vs.1).
| Brain Regions | Hemisphere | T values | MNI Coordinates | ||
| X | Y | Z | |||
| Related to average memory | |||||
| Superior temporal gyrus | R | 3.69 | 54 | 0 | -6 |
| Superior temporal pole | R | 3.99 | 54 | 16 | -20 |
| Insula | L | -3.62 | -38 | 14 | 10 |
| Cuneus | R | -3.84 | 20 | -66 | 24 |
| Supramarginal gyrus | R | -3.46 | 58 | -32 | 36 |
| Postcentral gyrus | L | -3.65 | -54 | -8 | 22 |
| R | -3.76 | 64 | 0 | 20 | |
| Related to memory in the neutral condition | |||||
| Superior temporal gyrus | R | 3.87 | 54 | 0 | -6 |
| Thalamus | R | 3.56 | 20 | -16 | 6 |
| Postcentral gyrus | R | -3.57 | 64 | 0 | 20 |
| Cuneus | R | -3.57 | 20 | -66 | 24 |
| Supramarginal gyrus | R | -3.36 | 58 | -32 | 36 |
Author response table 3
A set of fMRI studies with moderate sample size (N = 16 to 35).
| Author | Journal | Year | Sample Size (effective) | Main conclusion | Correlations(including mediating effects) | |
|---|---|---|---|---|---|---|
| 1 | Günseli, Aly | eLife | 2020 | N=29 | Hippocampus and vmPFC support memory-guided attention. | Correlation between vmPFC activity and hippocampal activity by skipped_pearson_correlation.m function (Figure 3B). |
| 2 | Keogh, Bergmann, Pearson | eLife | 2020 | N=32 | Cortical excitability is linked to individual differences in the strength of mental imagery. | Spearman rank correlations between cortical excitability and imagery strength (Figure 3 and also see Study design in Materials and methods). |
| 3 | Liu et al. | Nature Communication | 2016 | N=18 | Consolidation reconfigures neural pathways underlying the suppression of emotional memories. | Correlations between hippocampal functional connectivity/ pattern dissimilarity and behavioral suppression score by prediction analyses (Figure 4 and 6). |
| 4 | Gruber et al. | Neuron | 2016 | N=19 | Post-learning hippocampal dynamics predict reward-related memory advantages. | Pearson’s correlations between hippocampal functional connectivity/ reactivation and memory benefits (Figure 2B and 3C). |
| 5 | Schlichting, Preston | PNAs | 2014 | N=35 | Memory reactivation and hippocampal–neocortical functional connectivity during rest support subsequent learning. | Partial correlations between FFA reactivation/FFA-HPC connectivity and memory performance (Figure 2 and 3). |
| 6 | Wimmer, Shohamy | Science | 2012 | N=28 | Hippocampal activation, reactivation, and coupling predict decision bias. | Mediating effects of visual reactivation on the relationship between hippocampal activity and striatum activity. |
| 7 | Tambini, Ketz, Davachi | Neuron | 2010 | N=16 | Offline hippocampal-cortical interactions relate to subsequent associative memory. | Pearson’s correlations between offline hippocampal−LO correlations and associative memory (Figure 3D). |
Author response table 4
Statistical power for each significant correlation in main results (two-tailed tests).
| Correlations | Sample Size(n) | Significance Criterion(α) | Effect size(r) | Power(1-β) | |
|---|---|---|---|---|---|
| Pattern similarity | Correlation between pair-specific pattern similarity and associative memory in aversive condition (Figure 3C) | 28 | 0.05 | 0.46 | 0.72 |
| Task-dependent functional connectivity | Correlation between hippocampal-FFA connectivity and associative memory in aversive condition (Figure 5E) | 0.57 | 0.91 | ||
| Correlation between hippocampal-LOC connectivity and associative memory in aversive condition (Figure 5F) | 0.65 | 0.98 |
Additional files
-
Supplementary file 1
Acoustic characteristics of the four voice clips.
- https://cdn.elifesciences.org/articles/60190/elife-60190-supp1-v3.docx
-
Supplementary file 2
Brain regions involved in emotional effect on condition-level reactivation.
- https://cdn.elifesciences.org/articles/60190/elife-60190-supp2-v3.docx
-
Supplementary file 3
Brain regions involved in emotional effect on hippocampal connectivity during emotional learning.
- https://cdn.elifesciences.org/articles/60190/elife-60190-supp3-v3.docx
-
Supplementary file 4
Post-encoding hippocampal connectivity changes in positive relation to memory in the aversive condition but negative in the neutral condition.
- https://cdn.elifesciences.org/articles/60190/elife-60190-supp4-v3.docx
-
MDAR checklist
- https://cdn.elifesciences.org/articles/60190/elife-60190-mdarchecklist1-v3.docx
