Is memory spared in monkeys with selective lesions of the hippocampus?

Abstract

Normal explicit/declarative memory requires a system of anatomically related structures in the medial temporal lobe (MTL) that includes the hippocampus and the entorhinal, perirhinal and parahippocampal cortex. Although extensive MTL damage in primates causes robust anterograde amnesia, defining the contribution of the hippocampus proper has proved challenging. We revisited that enduring puzzle here, directly addressing a number of factors suspected to contribute to conflicting results across earlier studies. Among them, we explored the effects of selective hippocampus lesions produced by two common excitotoxin methods, ibotenic acid (IBO) and N-methyl-D-aspartic acid (NMDA). Sample sizes were substantial and all behavioral testing was conducted postoperatively. Performance was assessed on several standard procedures designed for monkeys, including multiple variants of the delayed nonmatching-to-sample (DNMS) test of visual object recognition, a series of rapidly acquired two-choice object discriminations (OD), and a delayed response (DR) test of spatiotemporal memory. We used task-specific parametric manipulations (e.g., increasing retention intervals, and reduced stimulus set size) to systematically vary demands on memory, and test/retest analyses to enhance sensitivity for detecting impairment. Although both lesion groups sustained substantial hippocampal damage, their performance failed to differ from that of intact controls on any task, under any key test condition, regardless of data analytic strategy. The findings constrain plausible accounts of extant discrepancies in the literature, and moreover, highlight the need for fresh perspectives on the core operating characteristics of memory mediated by the primate hippocampus.