Behavioral Neuroscience 19%, Vol. 110, No. 4,661-672 Copyright 1996 by the American Psychological Association, Inc. 0735-7044/96/13.00 Psychological Stress Impairs Spatial Working Memory: Relevance to Electrophysiological Studies of Hippocampal Function David M. Diamond University of Colorado Health Sciences Center and Veterans Administration Medical Center, Denver, CO Nan Ingersoll University of Colorado Health Sciences Center Monika Fleshner University of Colorado Gregory M. Rose University of Colorado Health Sciences Center and Veterans Administration Medical Center, Denver, CO Stress blocks hippocampal primed-burst potentiation, a low threshold form of long-term potentia- tion, thereby suggesting that stress should also impair hippocampal-dependent memory. There- fore, the effects of stress on working (hippocampal-dependent) and reference (hippocampal- independent) memory were evaluated. Rats foraged for food in seven arms of a 14-arm radial maze. After they ate the food in four of the seven baited arms, they were placed in an unfamiliar environment (stress) for a 4-hr delay. At the end of the delay they were returned to the maze to locate the food in the 3 remaining baited arms. Stress impaired only working memory. Stress interfered with the retrieval of previously stored information (retrograde amnesia), but did not produce anterograde amnesia. Stress appears to induce a transient disruption of hippocampal function, which is revealed behaviorally as retrograde amnesia and physiologically as a blockade of synaptic plasticity. Two largely independent fields in behavioral neuroscience have provided different perspectives on the involvement of the hippocampus in behavioral adaptation. Four decades of stud- ies in cognitive neuroscience have provided strong evidence that damage to the hippocampus results in memory impair- ments in animals and humans (Barnes, 1988; Scoville & Milner, 1957; Rasmussen, Barnes, & McNaughton, 1989; Squire, 1992). The neuroendocrine field also has a long history in studying the hippocampus, hut has emphasized that this structure is involved in regulating the hypothalamic-pituitary- adrenal (HPA) axis (de Kloet. 1991; Endroczi, Lissak, Bonus, & Kovacs, 1959; Fcndler, Karmos, & Telegdy, 1961; Kawakami cl al., 1968; Knigge & Hays, 1963; McEwen, 1991). Investiga- tors in this field have found that the hippocampus has the highest density of corticosterone receptors in the nervous system (Gerlach & McEwen, 1972; Reul & de Kloet, 1985; Spencer, Young, Choo, & McEwen, 1990) and is involved in David M. Diamond and Gregory M. Rose, Department nf Pharma- cology, University of Colorado Health Sciences Center, and Medical Research Service, Veterans Administration Medical Center; Monika Fleshner, Department of Psychology, University of Colorado; Nan Ingersoll, Department of Pharmacology, University of Colorado Health Sciences Center. This research was supported by grants from the Veterans Affairs Medical Research Service, National Institute on Aging (AG10755), and the Office of Naval Research (ONR N00014-91-J-1753). We thank Howard Eichenbaum and the two reviewers for their constructive criticisms of an earlier version of the manuscript, and Gary Zerbe for his assistance in the statistical analyses of chance level of performance in the radial-arm maze. Correspondence concerning this article should be addressed to David M. Diamond, Department of Pharmacology, University of Colorado Health Sciences Center, Box C23fi, Denver, Colorado 802fi2. Electronic mail may be sent via Internet to d.diamond@uchsc.edu. HPA negative feedback, circadian rhythms, and the behavioral response to stress (Dunn & Orr, 1984; luvone & Van Hartes- veldt, 1977; Magarinos, Somoza, & De Nicola, 1987; Redgate, 1976; Sapolsky, Krey, & McEwen, 1984; Sapolsky, Zola- Morgan, & Squire, 1991; Wilson, Greer, Greer, & Roberts, 1980). Only rarely have investigators examined how the cognitive and neuroendocrine components of hippocampal function interact (e.g., Bohus, de Kloet, & Veldhuis, 1982; Landfield, Baskin, & Pitler, 1981; Micco & McEwen, 1980; Nyakas, de Kloet, Veldhuis, & Bohus, 1983; Pavlides, Westlind-Daniels- son, Nyborg, & McEwen, 1991; Rigter, Veldhuis, & dc Klocl, 1984; Ryan, Springer, Hannigan, & Isaacson, 1985). In one avenue of investigation, the stress-sensitive and cognitive components of hippocampal function have been studied at great length. Several laboratories have shown that chronic stress, or prolonged hypersecretion of corticosterone, contrib- utes to hippocampal neuron loss and impaired learning (Arbel, Kadar, Silbermann, & Levy, 1994; Bodnoff et al., 1995; Dachir, Kadar, Robinzon, & Levy, 1993; Issa, Rowe, Gauthier, & Meaney, 1990; Kerr, Campbell, Applegate, Brodish, & Land- field, 1991; Luinc, Villegas, Martinez, & McEwen, 1994; Sapolsky, Uno, Rebert, & Finch, 1990; Stein-Behrens, Matt- son, Chang, Ych, & Sapolsky, 1994). Conversely, elimination of corticosterone blocks changes in hippocampal electrophysi- ology and the impairment of cognitive ability that can occur with aging (de Kloet, 1992; Landfield et al., 1981; Kerr, Campbell, Hao, & Landfield, 1989; Kerr, Campbell, Thibault, & Landfield, 1992). A second avenue of investigation has developed within the framework of the effects of stress on hippocampal electrophysi- ological plasticity. In 1987, Foy, Stanton, Levine, and Thomp- son (1987) first demonstrated that stress, produced by restraint and electric shock, blocked hippocampal long-term potentia- 661