19.1 HIPPOCAMPAL AND PREFRONTAL MECHANISMS UNDERLYING RELATIONAL MEMORY FORMATION AND REASONING. (9th April 2019)

• Record Type: Journal Article
• Authors: Preston, Alison
• Is Part Of: Schizophrenia bulletin. Volume 45(2019)Supplement 2
• Page Start: S119
• Page End: S120
• Publication Date: 2019-04-09
• Subjects: Schizophrenia -- Periodicals; Schizophrenia -- Research -- Periodicals; 616.898005
• DOI: 10.1093/schbul/sbz022.075
• Languages: English
• ISSNs: 0586-7614

Abstract

Abstract: Background: Everyday behaviors require a high degree of flexibility, in which prior experience is applied to inform behavior in new situations. Such flexibility is thought to be supported by memory integration, a process whereby related experiences become interconnected in the brain through recruitment of overlapping neural populations. Integrating information acquired at different times allows memory to extend beyond direct experience by representing unobserved relationships between elements of different events. Integrated knowledge may then be flexibly deployed to promote higher-level cognitive functions such as reasoning. Across a series of functional magnetic resonance imaging (fMRI) studies in healthy young adults, we show that integrating related information in memory is mediated by interactions between the anterior hippocampus and medial prefrontal cortex. Furthermore, we will discuss evidence about how aberrant function of the hippocampal circuit in schizophrenia may impact memory integration mechanisms, leading to impairments in both memory and reasoning. Methods: Healthy young adults participants performed an associative inference task in which they learned about overlapping image pairs (AB, BC) followed by an inference test tapping their knowledge for the indirect relationships among pairs. Using a combination of high-resolution fMRI and multivariate statistical analyses, we examined how three component processes underlying memory integration supported learning of the overlapping experiences and inferential memory judgments. First, we used multivariate brain decoding techniques to test whether reinstatement of overlapping information (A items) facilitated or interfered with new learning (BC pair memory) and inference (AC decisions). Second, we tested the hypothesis that hippocampal novelty signals detecting deviations between present experience (BC pairs) and reinstated memories (AB pairs) would promote integration. Third, we used representational similarity analyses to test whether indirectly related memory elements (A and C items) are assimilated within hippocampus and medial prefrontal cortex. In a separate high-resolution fMRI study, we further tested how schizophrenia impacts hippocampal novelty signals essential for memory integration. Results: We found that memory reinstatement during new learning had a facilitative effect on memory and reasoning performance in healthy young adults. Co-activation of the anterior hippocampus and medial prefrontal cortex was observed during encoding of overlapping experiences, with novelty responses within the network predicting subsequent reasoning performance. Novelty responding in individuals with schizophrenia in a separate study was aberrant, suggesting one potential mechanism for impaired associative inference behavior in the disorder. In healthy young adults, representational similarity further evinced integration of indirectly related memory elements within anterior hippocampus and medial prefrontal cortex. Conclusions: Collectively, these data indicate that the anterior hippocampus and medial prefrontal cortex work in concert to assimilate new experiences into reinstated memory content, resulting in superior learning and reasoning performance. Our data further indicate that hippocampal novelty signals may be critical for initiating memory integration when new events deviate from reinstated memory-based predictions. Aberrant novelty processing in hippocampus may therefore underlie deficits in memory-based reasoning tasks observed in schizophrenia.