Gamma oscillations during episodic memory processing reveal reversal of information flow between the hippocampus and prefrontal cortex
Summary (4 min read)
Introduction
- Prefrontal monitoring and control during episodic memory processing is thought to be critical for contextually mediated memory retrieval (Miller, 2013; Preston and Eichenbaum, 2013 ).
- Evidence supporting this model has come from rodent investigations employing lagged correlation between the hippocampus and PFC in theta band oscillatory power (e.g. Place et al., 2016) .
Encoding Retrieval
- Here the authors sought evidence of reversal of information flow between the hippocampus and prefrontal cortex during the encoding versus the retrieval of episodic memories.
- The authors did this by taking advantage of a unique dataset obtained from 77 human patients implanted with stereo EEG electrodes for seizure mapping purposes who performed a verbal free recall paradigm.
- During the study and recall phases of the task, the authors identified activation peaks in gamma oscillations from 40 to 120 Hz, using the onset of gamma activation as an estimate of the initial timing of activity in a given brain region.
- As their data set included subjects with electrodes implanted in both the hippocampus and PFC (in addition to other cortical locations), the authors were able to directly compare the timing of.
Behavioral Performance
- Across participants, the average probability of recall for all words was 24.4%.
- The average percentage of list intrusions (recall errors) per subject was 12.8%.
- The authors derived an estimate of temporal clustering (the tendency for items adjacent to each other in the study list to be recalled sequentially) to determine if temporal contextual factors were operating at retrieval (Watrous and Ekstrom, 2014) .
- The mean clustering factor across all participants was 0.642, robustly higher than the chance value of 0.500 ( (36) = 8.294, < 0.001), indicating that participants incorporated temporal contextual information into encoded representations of the study words (Sederberg et al., 2010) .
sEEG Data
- For their principal analysis, the authors identified the lag in onset of activation (Δ ) for five prefrontal locations relative to the hippocampus (positive Δ indicating activation following the hippocampus, negative Δ indicating activation preceding the hippocampus).
- The left aVLPFC exhibited a mean activation lag relative to the hippocampus during successful encoding of +13.4 msec (FDR corrected < 0. (continued) compared to hippocampus), and a reversal of this effect during retrieval, such that the region led the hippocampus by -10.4 msec (FDR corrected = 0.0116).
- Moreover, the Δ distributions for encoding and retrieval were significantly different (FDR corrected < 0.001) across electrodes when compared with a paired t-test.
- Further, the authors analyzed prior list intrusions (PLI) to test more directly whether Δ reversal is associated with the transmission of contextual information, as hypothesized by the reciprocal flow model.
Discussion
- The authors data reveal direct human electrophysiological evidence of the reversal of information flow between the hippocampus and prefrontal cortex (left aVLPFC) during episodic memory encoding versus retrieval.
- The authors acknowledge however that their data by itself does not allow us to make strong claims regarding the content of information characterized by reciprocal flow.
- CC-BY 4.0 International license a certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity.
Place et al., 2016).
- Taken together, their data support the relevance of the reciprocal flow hypothesis to human memory and establish lagged gamma activation as a method to identify functional interactions between memory-relevant regions in humans.
- The identification of electrode contacts in the aVLPFC that exhibit these functional properties may be a strategy for the identification of propitious targets of neuromodulation to treat memory disorders.
Methods and Materials
- Participants 77 patients with medically intractable epilepsy who underwent stereoelectroencephalography surgery for clinical purposes were recruited to participate in this study.
- Frontal contacts were divided into 5 regions: the anterior ventro-lateral prefrontal cortex (principally BA45), the posterior ventro-lateral prefrontal cortex (BA8, BA44), the dorsolateral prefrontal cortex (BA9, BA46), the medial orbitofrontal cortex (BA10, BA11, BA12), and the anterior cingulate cortex (ACC) (BA24).
- The anatomical features described above were used in expert neuroradiology review to localize all electrodes and in situations of conflict between the reviewed anatomical location and Talairach-based assignment to Brodmann areas the former was used for 7 of 12 .
- CC-BY 4.0 International license a certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity.
Experimental Paradigm
- Participants preformed a free recall task consisting of multiple study/test cycles.
- Participants were then instructed to verbally recall as many items as possible from the immediately prior list in no particular order.
- A full session consists of 12 full study/test cycles and 1 practice study/test cycle which was excluded from analysis.
- One complete session yielded electrophysiological recordings from 144 word encoding epochs (12 lists x 12 words) and a variable number of retrieval epochs.
- The authors used the temporal clustering factor, which is a measure of temporal contiguity for each recall transition relative to all possible recall transitions at a given time, to determine if contextual factors were operating at retrieval (Sederberg et al., 2008) .
Data Processing
- Stereo-EEG data were recorded using a Nihon Kohden EEG-1200 clinical system.
- Raw signals were subsequently re-referenced to a bipolar montage, with each contact referenced to the superficial adjacent contact.
- All analyses were conducted using MATLAB with both built-in and custom-made scripts.
- Retrieval trials were isolated such that each included trial was isolated from any other retrieval events by at least 1200 msec before the onset of vocalization and 200 msec after the onset of vocalization (this led to the exclusion of 3,258/12,791 [25.5% trials]).
Activation Onset Detection (Calculation of )
- The authors compared the temporal patterns of high gamma band power changes in the hippocampus and frontal cortex in the 1000 msec immediately following study item presentation and the 1000 msec immediately preceding word vocalization .
- CC-BY 4.0 International license a certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity.
- To obtain an estimate of gamma power, a threshold was defined as the mean spectral power within each frequency band across all trials for a given session and for each condition.
- Prior to obtaining the threshold for encoding trials, the trials were further divided into subsequently recalled and non-recalled items to account for possible power differences due to memory success.
- This was repeated for each frequency band and trial, and the was averaged across all bands and then across all trials for the recalled epochs, non-recalled epochs, and retrieval epochs to obtain a single time estimate of activation onset, , for each hippocampal and PFC electrode and each condition .
Cross-Correlation
- In a convergent analysis, the authors used cross-correlation of the gamma band amplitude envelope to investigate the temporal dynamics between the hippocampus and the PFC.
- CC-BY 4.0 International license a certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity.
- For each hippocampal-PFC electrode pair, the normalized cross-correlation was calculated on the meancentered amplitude envelope on a trial-by-trial basis using an 800 msec moving window with a 1 msec step size and a maximum lag of 150 msec.
- The initial 800 msec cross-correlation moving window for the encoding epochs was centered 100 msec prior to word presentation and stepped by 1 msec until 1700 msec after word presentation.
- Lastly, the correlation lag time with the highest z-score (i.e. the correlation lag where the correlation coefficient was maximally greater than zero across trials) was determined for each correlation moving window to produce a 1 by 1800 matrix for all recalled and non-recalled study trials and a 1 by 1200 matrix for all retrieval trials for each electrode pair.
Statistical Procedure
- To test for significant Δ across the electrodes within each hippocampal-PFC region pair during the encoding (subsequently recalled only) and retrieval conditions, the authors combined the Δ for all electrode pairs into a single matrix and used a t-test to compare the distribution of Δ against a null hypothesis of zero lag in onset activation (Δ = 0).
- In order to account for the type I error rate, p-values were false discovery rate (FDR) corrected.
- A one-sample t-test was used to compare the distribution of lag estimates across electrode pairs against a null hypothesis of zero lag (i.e. the correlation coefficient is maximized at zero lag) for each condition.
- The left aVLPFC shows a Δ reversal between conditions that is consistent with the result using the original threshold.
- Z-scores were calculated for each region using a paired t-test between gamma power for recalled words and non-recalled words.
Did you find this useful? Give us your feedback
Citations
252 citations
References
84 citations
"Gamma oscillations during episodic ..." refers background in this paper
...Furthermore, unlike rodents, human hippocampal recordings do not universally exhibit theta modulation as a function of memory processing (although this might be more prevalent in posterior hippocampal locations) (Lin et al., 2017; Watrous and Ekstrom, 2014)....
[...]
54 citations
"Gamma oscillations during episodic ..." refers background or methods in this paper
...Furthermore, unlike rodents, human hippocampal recordings do not universally exhibit theta modulation as a function of memory processing (although this might be more prevalent in posterior hippocampal locations) (Lin et al., 2017; Watrous and Ekstrom, 2014)....
[...]
...…factor when testing the reciprocal flow hypothesis in humans is that there appear to be multiple oscillations within traditional theta frequency bands, and the dominant theta frequency in the hippocampus may differ that in the neocortex (Lega et al., 2012; Miller, 2013; Watrous and Ekstrom, 2014)....
[...]
...We derived an estimate of temporal clustering (the tendency for items adjacent to each other in the study list to be recalled sequentially) to determine if temporal contextual factors were operating at retrieval (Watrous and Ekstrom, 2014)....
[...]
48 citations
42 citations
30 citations
"Gamma oscillations during episodic ..." refers background or result in this paper
...Of interest, fMRI, BOLD SMEs are consistently reported for the aVLPFC and surrounding regions in study tasks requiring elaborative encoding of verbal items (for a review, see Kim, 2019)....
[...]
...In humans, noninvasive data have stimulated the hypothesis that the VLPFC is necessary for generating retrieval cues during episodic memory search (Kim, 2019), consistent with rodent findings, and lesion studies suggest that patients with frontal lobe dysfunction have difficulty recalling items…...
[...]