“We have, each of us, a life-story, an inner narrative whose continuity, whose sense, is, our lives. It might be said that each of us constructs and lives, a ‘narrative’, and that this narrative is us, our identities … to be ourselves, we must have ourselves … we must recollect ourselves …”
Oliver Sacks, who authored this quote, is a neurologist well known for his compelling essays about his case studies (excerpt from “A matter of identity” in The Man who mistook his wife for a hat, p.105). Here, he is reflecting on his patient Mr. Thompson, who suffers from severe Korsakoff’s syndrome. Anterograde amnesia is a major symptom of Korsakoff’s that prevents Mr. Thompson from forming new memories. Interestingly, Mr. Thompson continually fabricates seemingly genuine memories when talking to others, in an apparent attempt to replace the narrative that he has lost since he obtained amnesia. While reading about Mr. Thompson desperately trying to make sense of his life, the importance of a functional memory in going about our lives is painfully appreciated. Fortunately, labs across the world are researching to understand the mechanics of memory and how it catastrophically fails. Recent insights have come from the lab of Lila Davachi, the upcoming seminar speaker for UCSD neuroscience.
Memory is often understood as having three principal underlying processes: encoding, consolidation, and retrieval. In terms of anterograde amnesia, the pathology may arise from deficits in any of these three processes. First, perceptions must be encoded into the brain as a construct that can be efficiently accessed and computed on. For long-term function, the memories must be consolidated by stabilizing its neural representation, so that the past information can later be recalled (retrieval). A prevalent theory is that memory consolidation occurs as the neocortex and hippocampus communicate in order to establish a hippocampal-independent representation of the thing to be remembered. This idea is well supported, including recent work showing that disruption of hippocampal reactivation during slow wave sleep impairs subsequent memory performance (Girardeau et al, 2009).
Earlier this decade, Lila Davachi’s lab enhanced our knowledge of memory consolidation in humans (Tambini, Ketz, & Davachi, 2010). Using functional magnetic resonance imaging (fMRI), they were able to quantify the interactions between different brain regions by correlating the recorded blood-oxygen-level dependent (BOLD) signals between defined regions of interest (see Figure 1 below). Interestingly, specific cortico-cortical and hippocampal-cortical interactions were predictive of future memory performance.
Figure 1. During and after an object-face association task, there was an increased correlation between the BOLD responses in the fusiform face area and the lateral occipital complex. (Figure 2A in Tambini, Ketz, & Davachi)
While in the scanner, subjects completed 2 different tasks, which both began and ended with a resting period for potential consolidation. In the first task, the subjects observed object-face pairs, while the second task contained several examples of scenes associated with faces. Importantly, subjects’ performance on subsequent memory tests were significantly different: object-face pairs were recalled more reliably than the scene-face pairs. This is because the scenes are less “distinct” from one another than pairs of objects, and so it is more difficult to make separable associations (to consolidate!).
This contrast in performance is related to the discrepancy in cortico-cortical interactions during the rest periods after each task. After the object-face (OF) association task, the cortical face region (fusiform face area, FFA) and the cortical object region (lateral occipital complex, LO) maintained higher functional connectivity (see Figure 2 below). However, this was not the case after the scene-face (SF) association task, as the correlation between the FFA and the cortical region for scenes (parahippocampal face area, PPA) was unchanged from baseline rest conditions.
Figure 2. Functional connectivity between relevant cortical regions was increased after the object-face association task, but not the scene-face association task. (Figure 2D in Tambini, Ketz, & Davachi)
Tambini, Ketz, and Davachi also quantified functional connectivity between the hippocampus and cortex. They found that the these interactions were also increased only for the rest period immediately following the object-face association task. Together with the behavioral results, these findings support the previously mentioned theory that hippocampal-cortical interactions underlie memory consolidation.
The cross-subject relationship between this functional hippocampal-cortical connectivity and subsequent memory performance is the most exciting finding of this paper. As seen in Figure 3 below, subjects with higher hippocampal-LO correlations following a task tended to retain those memories better.
Figure 3. Differences among individuals in their subsequent memory performance can be explained by the degree of hippocampal-cortical interactions during the post-task rest period. (Figure 3D in Tambini, Ketz, & Davachi)
It is possible that active rehearsal by the subjects could confound the observed effect. Rehearsal would suggest correlations between the prefrontal cortex and the lateral occipital cortex. However, this correlation did not carry information about subsequent memory performance, so rehearsal was likely not confounding.
This is the first time that hippocampal-cortical interactions have been directly implicated in enhancing long-term memory consolidation. Additionally, this study is groundbreaking in that it demonstrates memory consolidation improvements in an awake rest period, as prior studies had only proven this association during sleep periods. Hippocampal-cortical and cortico-cortical correlations had been seen in human resting-state fMRI, but their behavioral relevance was not identified until the current study.
Interested in learning about the more recent developments in our knowledge of memory consolidation? On Tuesday, April 28 at 4pm, Dr. Lila Davachi will give a talk entitled “Behavioral and neural investigations of human memory consolidation”, in the Center for Neural Circuits and Behavior Marilyn C. Farquhar Conference Room.
Scott Cole is a first-year Neurosciences student currently rotating with Dr. Eric Halgren. When his memory encoding and retrieval are operational, he likes to compare and critique carne asada burritos across San Diego, which make the optimal post-volleyball meal. You can follow him on Twitter @scottrcole.