[PDF][PDF] Hippocampal contribution to the novel use of relational information in declarative memory

AR Preston, Y Shrager, NM Dudukovic, JDE Gabrieli - Hippocampus, 2004 - Citeseer
AR Preston, Y Shrager, NM Dudukovic, JDE Gabrieli
Hippocampus, 2004Citeseer
Declarative memory, which encompasses the acquisition, storage, and retrieval of events
and facts (Cohen and Squire, 1980), can be expressed flexibly in novel situations different
from the original learning environment (Eichenbaum and Cohen, 2001). Declarative memory
relies on the integrity of multiple structures in the medial temporal lobe (MTL), but animal
research (Bunsey and Eichenbaum, 1996; Dusek and Eichenbaum, 1997; Honey et al.,
1998) indicates that the hippocampus specifically underlies the flexible expression of …
Declarative memory, which encompasses the acquisition, storage, and retrieval of events and facts (Cohen and Squire, 1980), can be expressed flexibly in novel situations different from the original learning environment (Eichenbaum and Cohen, 2001). Declarative memory relies on the integrity of multiple structures in the medial temporal lobe (MTL), but animal research (Bunsey and Eichenbaum, 1996; Dusek and Eichenbaum, 1997; Honey et al., 1998) indicates that the hippocampus specifically underlies the flexible expression of declarative memories. In the present report, we demonstrate selective hippocampal activation in humans during recognition of pairs of items whose relationship was not explicitly learned but could be mediated through an overlapping relation with an explicitly learned common item. This result suggests a specific role for the human hippocampus in the novel expression of declarative memories. The MTL comprises several structures, including the hippocampal formation (dentate gyrus, CA fields, and subiculum), and surrounding entorhinal, perirhinal, and parahippocampal cortices. These structures are likely to contribute differently to declarative memory, but the unique contribution of each structure remains a subject of debate. An emerging hypothesis, supported by evidence from studies of rats (Bunsey and Eichenbaum, 1996; Dusek and Eichenbaum, 1997; Honey et al., 1998) and monkeys (Gaffan and Parker, 1996; Brasted et al., 2003), proposes that the hippocampus plays an essential role in the flexible expression of declarative memories. According to this view, experiences yield memories that are learned directly from the environment (rote associations). Flexibility is needed when elements of individual experiences must be related or conjoined in new ways to deal with novel situations. This flexibility is thought to be a powerful characteristic of declarative memory because it allows for the generative use of the elements of experience to address new questions posed by the environment (Eichenbaum and Cohen, 2001).
In rats, damage to the hippocampal system, either by lesion to the hippocampus proper (Bunsey and Eichenbaum, 1996) or through disconnection of the hippocampus from its cortical and subcortical output pathways (Dusek and Eichenbaum, 1997), impairs the flexibility of declarative memories without disrupting the ability to learn explicitly trained memories (rote associations). However, it is unknown in any mammalian species, including in humans, whether the hippocampus plays a unique role in the flexible expression of declarative memories. Neuroimaging studies of humans have suggested hippocampal contributions to encoding (Davachi and Wagner, 2002; Davachi et al., 2003) and retrieval (Eldridge et al., 2000; Cansino et al., 2002) of declarative information, but none has specifically addressed questions regarding flexible expression of relational information as in previous rat studies. In the present investigation, we modeled a study with rats (Bunsey and Eichenbaum, 1996) and used functional magnetic resonance imaging (fMRI) to examine whether the human hippocampus plays a specific and unique role in the flexible expression of declarative memory. Prior to scanning, participants received explicit training on three sets of paired associates (Fig. 1A). Participants first learned to associate specific faces (stimuli A) with specific houses (stimuli B). Then, participants learned to associate another set of faces (stimuli C) with the same specific houses (stimuli B). In both training phases, each face-house pair was presented four times. Thus, each house was associated with two different faces in the two …
Citeseer