The Time-Limited Role of the Hippocampus in Memory Retrieval

Post by Elisa Guma

What's the science?

Episodic memory involves the retrieval of spatially and temporally specific events or situations. The hippocampus and its connections to the neocortex are thought to support memory recall, however, there is much debate over whether its role is time-limited. This week in PNAS, Gilmore and colleagues sought to test the two main hypotheses associated with autobiographical recall using in vivo human functional magnetic resonance imaging (fMRI) experiments: (1) the standard model of consolidation which asserts that a process of consolidation migrates from the hippocampus to the neocortex, which suggests that retrieval of more distant memories no longer requires the hippocampus, or (2) the multiple trace and trace transformation hypotheses, which state that each retrieval of an event is accompanied by another memory trace being formed and stored more broadly across the hippocampus, and that it is always required for vividly recalling memories.

How did they do it?

To assess the neural activity underlying memory retrieval, the authors acquired fMRI data — which measures brain activity — while participants described memories aloud, cued by photographs, for ~2 minutes from three different recall periods. These included earlier on the day of scanning, a period of 6-18 months prior, and a period of 5-10 years prior. 

The authors synced the spoken audio to functional time series data to analyze the relationship between recalled details and neural activity. Further, by using an overt recall design, the authors were able to pinpoint specific neural activity associated with event details, providing better estimates of retrieval-related activity, which would be lost in covert recall procedures (not spoken aloud). The authors analyzed whether hippocampal activity, across the anterior-posterior axis, was differentially associated with the three recall periods, and compared those to non-autobiographical control recordings. Finally, connectivity to other brain regions was investigated during the different recall events. The authors also used advanced data processing techniques (denoising) and evaluated the quality of their data after collection to ensure the fMRI data was not affected by motion due to the participants’ active speech.

What did they find?

First, the authors characterized the contents of the memories to ensure there were no biases based on differing amounts of detail between the three recall periods. They also investigated the effects of head motion in the scanner but found no difference between speech relative to non-speech periods.   

Next, they found that the posterior hippocampus showed greater activity for more recent (today and 6-18 months) than remote memories (5-10 years) and that only activity for the most recent conditions differed from the non-autobiographical control trials. This confirmed the temporally graded retrieval-related activity, consistent with the first standard model of consolidation hypothesis. Further, the authors observed differences in connectivity between the hippocampus and various neocortical regions (thought to support mental construction of scenes), depending on the recency of recall events. They observed greater connectivity strength for more recent recall than more remote recall, in the right hippocampus, with the same trend in the left, suggesting that both hippocampal activity and hippocampal-cortical interaction are temporally graded.

What's the impact?

The findings presented here show that both hippocampal activity and hippocampal-cortical interaction during autobiographical recall decrease as the events become more remote. This lends support for the standard model of consolidation hypothesis and provides further insight into the neural mechanisms underlying memory recall.  

Gilmore, AW et al., Evidence supporting a time-limited hippocampal role in retrieving autobiographical memories. PNAS (2021). The original scientific publication here.