Open Collections

UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

A reexamination of the role of the hippocampus in object-recognition memory using neurotoxic lesions… Duva, Christopher Adam 1996

Your browser doesn't seem to have a PDF viewer, please download the PDF to view this item.

Item Metadata

Download

Media
831-ubc_1996-14741X.pdf [ 11.68MB ]
Metadata
JSON: 831-1.0087745.json
JSON-LD: 831-1.0087745-ld.json
RDF/XML (Pretty): 831-1.0087745-rdf.xml
RDF/JSON: 831-1.0087745-rdf.json
Turtle: 831-1.0087745-turtle.txt
N-Triples: 831-1.0087745-rdf-ntriples.txt
Original Record: 831-1.0087745-source.json
Full Text
831-1.0087745-fulltext.txt
Citation
831-1.0087745.ris

Full Text

A R E E X A M I N A T I O N O F T H E R O L E O F T H E H I P P O C A M P U S IN O B J E C T - R E C O G N I T I O N M E M O R Y U S I N G N E U R O T O X I C L E S I O N S A N D I S C H E M I A IN R A T S By C H R I S T O P H E R A D A M D U V A B.A., California State University, San Bernardino, 1987 M . A . , California State University, San Bernardino, 1990 A THESIS S U B M I T T E D IN P A R T I A L F U L F I L L M E N T O F T H E R E Q U I R E M E N T S F O R T H E D E G R E E O F D O C T O R O F P H I L O S O P H Y in T H E F A C U L T Y O F G R A D U A T E STUDIES (Department of Psychology) We accept this thesis as conforming to the required standards T H E U N I V E R S I T Y O F BRITISH C O L U M B I A J U L Y , 1996 © Christopher Adam Duva, 1996 In presenting this thesis in partial fulfilment of the requirements for an advanced degree at the University of British Columbia, I agree that the Library shall make it freely available for reference and study. I further agree that permission for extensive copying of this thesis for scholarly purposes may be granted by the head of my department or by his or her representatives. It is understood that copying or publication of this thesis for financial gain shall not be allowed without my written permission. Department The University of British Columbia Vancouver, Canada DE-6 (2/88) A B S T R A C T II Paradoxical results on object-recognition delayed nonmatching-to-sample (DNMS) tasks have been found in monkeys and rats that receive either partial, ischemia-induced hippocampal lesions or complete hippocampal ablation. Ischemia results in severe D N M S impairments, which have been attributed to circumscribed CA1 cell loss. However, ablation studies indicate that the hippocampus plays only a minimal role in the performance of the D N M S task. Two hypotheses have been proposed to account for these discrepant findings (Bachevalier & Mishkin, 1989). First, the "hippocampal interference" hypothesis posits that following ischemia, the partially damaged hippocampus may disrupt activity in extrahippocampal structures that are important for object-recognition memory. Second, previously undetected ischemia-induced extrahippocampal damage may be responsible for the D N M S impairments attributed to C A 1 cell loss. T o test the "hippocampal interference" hypothesis, the effect of partial N M D A -induced lesions of the dorsal hippocampus were investigated on D N M S performance in rats. These lesions damaged much of the same area, the CA1, as did ischemia; but did so without depriving the entire forebrain of oxygen, thereby reducing the possibility of extrahippocampal damage. In Experiment 1, rats were trained on the D N M S task prior to receiving an NMDA-lesion. Postoperatively, these rats reacquired the nonmatching rule at a rate equivalent to controls and were unimpaired in performance at delays up to 300 s. In Experiment 2, naive rats were given NMDA-lesions and then trained on D N M S . These rats acquired the D N M S rule at a rate equivalent to controls and performed normally at delays up to 300 s. These findings suggest that interference from a partially damaged hippocampus cannot account for the ischemia-induced D N M S impairments and that they are more likely produced by extrahippocampal neuropathology. In Experiment 3, rats from the previous study were tested on the Morris water-maze. Compared to sham-lesioned animals, rats with partial lesions of the dorsal hippocampus were impaired in the acquisition of the water-maze task. Thus, subtotal NMDA-lesions of the hippocampus impaired spatial memory while leaving nonspatial memory intact. Mumby et al. (1992b) suggested that the ischemia-induced extrahippocampal Ill damage underlying the D N M S deficits is mediated or produced by the postischemic hippocampus. T o test this idea, preoperatively trained rats in Experiment 4 were subject to cerebral ischemia followed within Ihr by hippocampal aspiration lesions. It was hypothesized that ablation soon after ischemia would block the damage putatively produced by the postischemic hippocampus and thereby prevent the development of postoperative D N M S deficits. Unlike "ischemia-only" rats, the rats with the combined lesion were able to reacquire the nonmatching rule at a normal rate and performed normally at delays up to 300 s. Thus, hippocampectomy soon after ischemia eliminated the pathogenic process that lead to ischemia-induced D N M S deficits. Experiment 5 investigated the role of ischemia-induced CA1 cell death as a factor in the production of extrahippocampal neuropathology. Naive rats were given NMDA-lesions of the dorsal hippocampus followed 3 weeks later by cerebral ischemia. If the ischemia-induced CA1 neurotoxicity is responsible for producing extrahippocampal damage then preischemic ablation should attenuate this process and prevent the development of D N M S impairments. This did not occur: Rats with the combined lesion were as impaired as the "ischemia-only" rats in the acquisition of the D N M S task. This suggests that the ischemia-induced pathogenic processes that result in extrahippocampal neuropathology comprise more than C A 1 neurotoxicity. The findings presented in this thesis are consistent with the idea that ischemia-induced D N M S deficits in rats are the result of extrahippocampal damage mediated or produced by the postischemic hippocampus. The discussion focuses on three main points: 1) How might the post-ischemic hippocampus be involved in the production of extrahippocampal neuropathology? 2) In what brain region(s) might this damage be occurring? 3) What anatomical, molecular, or functional neuropathology might ischemia produce in extrahippocampal brain regions? The results are also discussed in terms of a specialized role for the hippocampus in mnemonic functions and the recently emphasized importance of the rhinal cortex in object-recognition memory. iv T A B L E O F C O N T E N T S Abstract ii T a b l e of Contents iv L i s t of Figures xiii A c k n o w l e d g m e n t s x Chapter 1: General Introduction 1 The Case of H . M . and the Amnesic Syndrome 2 H i s t o r y 3 Neuropsychological Profile 4 Neuropathology 6 The Medial Temporal Lobe Amnesic Syndrome 7 Animal Models of Medial Temporal Lobe Amnesia: Specifying the Locus of Damage 11 Specifying the Locus of Damage: Anatomical Background 14 Hippocampus 17 Entorhinal Cortex 25 Specifying the Locus of Damage: Neurobehavioral Analyses 28 Monkey Models of Medial Temporal Lobe Amnesia 28 The Temporal Stem Hypothesis 28 The A m y g d a l a 30 The Hippocampus 31 Rat Models of Medial Temporal Lobe Amnesia 35 Amnesia Produced By Brain Damage Outside of the Medial Temporal Lobe 40 Diencephalon 40 Basal Forebrain 46 Prefrontal Cortex 48 Ischemia-Induced Amnesia 50 Ischemia-Induced Amnesia in Humans 50 V The Case of Patient R.B 51 Neuropathology of R.B 52 Animal Models of Cerebral Ischemia 53 M o n k e y s 53 Rats 54 Mechanisms of Ischemic Brain Damage 58 Excitotoxicity 58 E x p l a i n i n g the Paradox 61 Rationale for Experiments and Hypotheses 63 Chapter 2: General Methods 67 Subjects 67 Apparatus 67 Behavioral Procedure 67 Surgical Procedure 69 Cerebral Ischemia 69 Neurotoxic Lesions 70 Histological Procedure 73 Chapter 3 : A n Analysis of the Effect of Partial Hippocampal Damage on the Performance of Nonspatial and Spatial Memory Tasks 74 Experiment 1: The Effect of Partial Lesions of the Dorsal Hippocampus on D N M S Performance in Rats with Preoperative Training 74 Introduction 74 M e t h o d 74 Results 76 D i s c u s s i o n 83 Experiment 2: The Effect of Partial Lesions of the Dorsal Hippocampus on D N M S Performance in Naive Rats 87 Introduction 88 vi M e t h o d 88 Results 88 D i s c u s s i o n 95 Experiment 3: The Effect of Partial Lesions of the Dorsal Hippocampus on Spatially-Guided Behavior 96 Introduction 96 M e t h o d 97 Results 98 D i s c u s s i o n 103 Chapter 4: Ischemia-Induced Object-Recognition Memory Deficits in Rats: A n Analysis of the Locus of Damage Using Hippocampal Ablation and Cerebral Ischemia, Separately and in Combination 108 Experiment 4: Prevention of Object-Recognition Deficits by Hippocampal Ablation Immediately Following Cerebral Ischemia 108 Introduction 108 M e t h o d 109 Results 110 D i s c u s s i o n 120 Chapter 5: A n Analysis of the Ability of Partial Hippocampal Lesions to Prevent Ischemia-Induced D N M S Acquisition Deficits 124 Experiment 5: NMDA-Lesions of the Dorsal Hippocampus Fail to Prevent Ischemia-Induced D N M S Acquisition Deficits 124 Introduction 124 M e t h o d 124 Results 125 D i s c u s s i o n 133 Chapter 6: General Discussion 135 Ischemia-Induced Extrahippocampal Neuropathology: H o w , Where, A n d What? 139 vii How May the Hippocampus Mediate or Produce Ischemia-Induced Extrahippocampal Neuropathology? 140 Where is the Possible Anatomical Locus of the Extrahippocampal Neuropathology? 145 Rhinal Cortex 145 Diencephalon 146 Basal Forebrain 147 Prefrontal Cortex 147 What is the Possible Nature of the Extrahippocampal Neuropathology? 148 Anatomical Changes 148 Molecular Changes 150 Functional Changes 150 Synthesis 151 Implications for the Neurobiological Study of Memory 153 References 156 LIST O F F I G U R E S viii Figure 1: Ventral surface of the monkey brain 16 Figure 2: Coronal section through the dorsal hippocampus of the rat 19 Figure 3: Three-dimensional position of the hippocampal formation in the rat brain 22 Figure 4: A photomicrograph of the hippocampus of rats subjected to neurotoxic lesions or cerebral ischemia 72 Figure 5: Mean trials to criterion on object discrimination for the N M D A - L and N M D A - S groups 78 Figure 6: Mean trials to criterion on D N M S for the N M D A - L and N M D A - S groups 80 Figure 7: Mean percent correct across delays for the N M D A - L and N M D A - S groups 82 Figure 8: Extent of the dorsal hippocampal lesions in the N M D A - L group 85-86 Figure 9: Mean trials to criterion object discrimination for the N M D A - N A I V E and S - N A I V E groups 90 Figure 10: Mean trials to criterion on D N M S for the N M D A - N A I V E and S - N A I V E groups 92 Figure 11: Mean percent correct across delays for the N M D A - N A I V E and S - N A I V E groups 94 Figure 12: Latency to find the submerged platform in the Morris water-maze for the N M D A - N A I V E and S - N A I V E groups 100 Figure 13: Mean path length to find the submerged platform in the Morris water-maze for the N M D A - N A I V E and S - N A I V E groups 102 Figure 14: Extent of the dorsal hippocampal lesions in the N M D A - N A I V E group 105-106 Figure 15: Mean trials to criterion on D N M S for the I S C - p l u s - A B L and ISC-only groups postsurgery 112 Figure 16: Mean percent correct across delays for the I S C - p l u s - A B L and ISC-only groups 114 Figure 17: Schematic representation of the extent of the hippocampal lesions in the I S C - p l u s - A B L goup 117 ix Figure 18: Percent of C A 1 cell loss in the ISC rats 119 Figure 19: Data from experiment 4 compared with the results of Mumby et al. (1992b) 122 Figure 20: Mean trials to criterion on object discrimination for the I S C - N M D A and ISC-S groups postoperatively 127 Figure 21: Mean trials to criterion on D N M S for the I S C - N M D A and ISC-S groups postoperatively 129 Figure 22: Extent of the dorsal hippocampal lesions in the I S C - N M D A group 132-133 Figure 23: A schematic diagram of hippocampal afferents to structures believed to be involved in memory 143 A C K N O W L E D G E M E N T S X First and foremost, I would like to thank my parents Frank and Barbara, without whose emotional and financial support this thesis would not have been possible. I would also like to thank*the members of my thesis committee: M y supervisor Tony Phillips for his patience with me, John Pinel for allowing me to work in his lab and for his support and honesty, and Eric Eich for aggreeing to serve on my committee. Don Wilkie was also extremely helpful as the chair of my departmental committee and in reading earlier versions of this thesis. I would also like to acknowledge Stan Floresco for his assisitance in behavioral testing and the construction of some of the figures in this thesis. Lucille Hoover also deserves a special thanks for always being the person who provided me with equipment and technical support when I needed it the most. Dave Mumby has also been very influential in the formulation of this thesis through our many discussions on the neural basis of memory. Jason Carr also deserves thanks for proof-reading a final version of this thesis. I would also like to thank my friend Tom Kornecook for our many long discussions that provided me with catharsis over the past 5 years and my brother Mark and my Aunt Joy for similar support. Finally, I would like to thank the rats, who paid the ultimate price in the name of science. 1 CHAPTER 1: GENERAL INTRODUCTION Considerable effort has been directed toward understanding the amnesic syndrome produced by lesions in the medial temporal lobe (MTL). In their seminal study, Scoville and Milner (1957) first demonstrated that humans with large bilateral lesions of this region had profound memory dysfunction. Since that time, numerous studies with other neuropsychological patients, nonhuman primates, and rodents have further attested to the importance of the integrity of the M T L for normal memory functions. However, despite nearly 4 decades of research, many important questions remain to be answered. The central and most important of these being what is the relationship between the locus of the damage within the M T L and amnesia? That is, lesions to which structure, or combination of structures, are sufficient to produce the amnesic syndrome ? This question has been difficult to answer for two reasons. First, the M T L is a large region of the brain that includes the hippocampus, amygdala, parahippocampal, periamygdaloid, and entorhinal and perirhinal cortices (Squire, 1992; Squire & Zola-Morgan, 1992). Humans with circumscribed lesions to candidate structures are rare or nonexistent, making any correlation between locus of damage and memory impairment tenuous. Second, producing a reliable animal model of M T L amnesia has proven to be problematic. It is difficult to produce a lesion to a discrete part of the M T L without damaging adjacent areas. Conventional aspiration lesions of the deep structures (i.e., the hippocampus and amygdala) always damage the underlying cortex, again making interpretation difficult. Despite the interpretational problems, several studies concluded that damage to the hippocampus was sufficient to produce amnesia in humans (Scoville & Milner, 1957; Milner, 1966) and, more recently, monkeys (Mahut, Zola-Morgan & Moss, 1982; Zola-Morgan & Squire, 1986). However, with the subsequent refinement of the animal models of M T L amnesia, this view has been called into question, and an intriguing paradox has developed. Experiments with monkeys (Alvarez, Zola-Morgan, & Squire, 1995; O'Boyle, Murray, & Mishkin, 1993) and rodents (Aggleton, Hunt, Rawlins, 1986; Mumby, Pinel, & Wood, 1992a; Mumby, Wood, Buday, Bussey, Pinel, & Phillips, 1992b), in which discrete lesions of the entire hippocampus were made without damaging related cortical 2 areas, have found little, if any, memory impairment on some amnesia-sensitive memory tasks. Moreover, a paradox becomes evident when these results are compared to studies in which subtotal hippocampal lesions are produced. Animals that receive apparent damage to only the C A 1 subfield of the hippocampus, as a result of ischemia, exhibit profound amnesic symptoms (Bachevalier & Mishkin, 1989; Wood, Mumby, Pinel, & Phillips, 1993; Mumby et al., 1992). It appears then that a partial lesion to the hippocampus is more detrimental to memory than a total lesion. How can a lesion to one subfield of the hippocampus be more disruptive to memory than total ablation? The objective of this thesis is to attempt to resolve the paradox that exists between the effects of partial or total hippocampal lesions. This will be done in the context of the more general question of the relationship between locus of damage within the M T L and the amnesic syndrome. Accordingly, the general introduction is divided into four sections. The first section will present an overview of the case of patient H . M . and a discussion of the amnesic syndrome. The second section will be concerned with the development of nonhuman primate and rodent models of M T L amnesia and the relationship between findings in human and animal research. Included in this section will be a description of the neuroanatomical structure and connectivity of the hippocamapal formation. The third section will detail the mnemonic deficits that have been attributed to ischemia-induced hippocampal damage and how these findings are paradoxical when compared with lesion studies. Mechanisms of ischemia-induced brain damage will also be discussed. The fourth and final section of the introduction will provide a rationale for and a brief description of the research presented in this thesis. T H E C A S E O F H . M . A N D T H E A M N E S I C S Y N D R O M E The case of H . M . (Scoville & Milner, 1957) marks the beginning of the modern scientific investigation of the relationship between M T L damage and memory dysfunction, though this relationship had been noticed as early as 1890 (Bekhterev, 1890). H . M . ' s case is unique for a variety of reasons. First, his memory impairments have been exceptionally well documented over a period of more than 40 years. This extensive investigation has allowed some conclusions to be drawn about the nature and extent of amnesia resulting 3 from damage to the M T L . Second, unlike many other amnesics who display a sequellae of cognitive dysfunction, H.M.'s disorder is characterized by pure amnesia. Despite his profound memory impairments, his attention, perception, intelligence, and motor skills have remained intact. The following section provides a description of H . M . ' s early history, a neuropsychological and neuropathological profile, and a discussion of the amnesic syndrome. HISTORY H . M . began experiencing epileptic attacks at the age of 8, presumably as a result of a head injury sustained in a bicycle accident the previous year. Petit mal seizures began 3 years later and were occurring with a frequency of up to 10 times per day by the age of 13. Grand mal seizures began around 16 years of age and increased in frequency and duration over the next 7 years until they were occurring an average of once per week. Heavy and varied anticonvulsant medications were ineffective at controlling these attacks. At this point, H . M . was effectively prevented from working, engaging in social acivities, or otherwise leading a normal life (Milner, 1966; Scoville, 1968). Electrographic examinations of H . M . revealed a diffuse rather than a focal abnormality, with a bilateral centro-temporal dominance (Milner, 1966; Scoville, 1968). Nevertheless, because of the severe and intractable nature of his epilepsy, he underwent a radical bilateral M T L resection in the fall of 1953. This experimental surgical procedure (of varying degrees) had been previously performed on approximately 30 psychotic patients in the hope that it would alleviate some of their symptoms without producing the undesirable side-effects of a complete frontal lobotomy (Scoville, 1954; Scoville, Dunsmore, Liberson, Henry, & Pepe, 1953). In such cases, it was largely unsuccessful and, initially, its effects on memory were masked by the psychotic symptoms of the patients. In the days immediately following surgery, H . M . was confused and unresponsive but then, as he became more alert, a severe memory impairment was noticed. He could no longer remember people he had just met and had no memory for the events just prior to his surgery. He seemed to retain little of the day-to-day events occurring in the hospital and could not learn his way around the ward nor the names of any staff members. Extensive, 4 though patchy retrograde amnesia was also noticed at this time. Despite its profound effects on memory, the surgical procedure was successful in reducing H . M ' s epileptic attacks. Grand mal seizures have nearly disappeared and petit mal attacks have diminished in frequency to five or six per month. NEUROPSYCHOLOGICAL PROFILE Since his surgery in 1953, H.M.'s mnemonic capabilities have been extensively studied and continue to be so at present. The most notable feature of H.M. 's memory impairments is his profound anterograde amnesia. He retains virtually nothing of ongoing events. For, example he fails to recognize people he has repeatedly met since his surgery and will read the same magazine over and over without any apparent retention. When asked about the nature of his memory disorder H . M . responded by saying that "each moment is like waking from a dream" (Milner, 1966, p. 115). Along with his severe anterograde amnesia, H . M . also displays some persistent retrograde amnesia. The full extent of this retrograde deficit has been difficult to characterize. Initially, it was believed to be restricted to a period of 1-2 years prior to his surgery, but more recent work suggests that it extends back at least 10 years or more. H.M.'s very remote memories, however, remain vivid and intact (Milner, 1966; Milner, Corkin, & Tueber, 1968). H.M. 's amnesia encompasses both verbal and nonverbal material. He is deficient in tests of verbal learning and recall (Scoville & Milner, 1957) and is similarly impaired on nonverbal tasks such as the delayed recall of a complex figure and the recognition of recurrent nonsense patterns (Scoville & Milner, 1957; Milner et al., 1968). His amnesia persists in the absence of any general intellectual loss or perceptual impairment (Milner, et al„ 1968; Scoville & Milner, 1957). In fact, his scores on standardized intelligence tests were substantially higher in 1955 and 1962 than they were in the year preceding his surgery. H . M . has a normal immediate memory as measured by digit span (Drachman & Arbit, 1966) and can hold items in short term memory by verbal rehearsal, sometimes of a very elaborate nature (Milner, et al., 1968). Thus, his primary deficit appears to be transfering information from short-term to long-term memory. Interestingly, H.M.'s anterograde amnesia does not appear to be absolute. He seems able to form some lasting 5 memories about particularly salient events and constant features of his environment. Mind you, the recollections appear as flashes of memory rather than detailed descriptions. For example, H . M knows that his father is now dead, an event that occurred following the onset of H . M . 's amnesia. When questioned about the circumstance surrounding his father's death, however, H . M . could produce little accurate information. Similarly, H . M . can remember a few outstanding public events. He was familiar with the assassination of president Kennedy and had some knowledge about the Apollo space program. H . M . 's recall of particular salient events appears to depend on "the frequent repetition of the items and their embedding in a constant framework" (Milner et al., 1968 p. 232). H . M . also displays the ability to learn and retain certain types of motor and perceptual tasks, though his initial and final levels of performance are typically inferior to normal subjects. One such task is mirror-drawing. In this paradigm, subjects are asked to trace a figure while being guided by the reflection of their hand in a mirror. Since the mirror reverses the images, initial performance is poor. Gradually, as the subject adapts, the reversed image performance improves. Milner (1962) found that H.M. 's performance on a mirror-drawing task steadily improved over a 3 day period, as indicated by a reduction in the time to complete the task and in the number of errors commited. Subsequently, Corkin (1968) found similar preserved motor-skill learning in rotary pursuit, bimanual tracking, and tapping tasks. One of the most striking features of this residual learning was that H . M . retained little, or no, memory of the actual learning episodes. Thus, despite improved performance over subsequent daily practice sessions, H . M . reported virtually no knowledge of ever being exposed to the task. Further studies with H . M demonstrated that his preserved learning capacities extended beyond motor learning to the domain of perceptual learning as well. Milner et al. (1968) found that H . M . could identify incomplete drawings in the Gollin picture completion paradigm at a rate equivalent to matched controls. This effect is now known as priming, and is defined as a phenomenon "in which previous exposure to a word or object on a study list facilitates its subsequent identification when degraded perceptual cues are provided" (Schacter, Chiu, & Ochsner, 1993 p. 160). More recently, H . M . has been tested 6 in several additional priming paradigms and has shown intact priming for visual patterns (Gabrieli, Milberg, Keane, & Corkin, 1990) and word-stem completion with both words and nonwords (Gabrieli & Keane, 1988). H . M . is, however, impaired in stem-completion priming with novel words, that is words that entered into the English language after the onset of his amnesia (Postle & Corkin, 1994). H.M.'s learning capacity for priming resembles that for motor learning. While he does show improvement with practice, his initial and final levels of performance are inferior to normal subjects and he recalls almost nothing of the learning situation. NEUROPATHOLOGY Scoville and Milner (1957) described what they believed to be the extent of H.M. 's medial temporal lobectomy. On the medial surface, the excision extended approximately 8 cm posteriorly from the anterior tips of the temporal lobe. The excision was limited superiorly and laterally by the temporal horns and extended approximately 3 cm posterior to the medial portion of the petrous ridge. Accordingly, the prepyriform gyrus, uncus, amygdala, hippocampus, and hippocampal gyrus were presumed to be destroyed bilaterally while the temporal neocortex was believed to have been entirely spared. A more accurate description of the extent of the M T L lesions has been provided by a recent magnetic resonance imaging (MRI) scan of H.M.'s brain. Corkin, Amaral, Johnson, and Hyman (1994) found that the bilateral lesions included the medial temporal pole, amygdala, and entorhinal cortex. The anterior 2 cm of the dentate gyrus, hippocampus, and subiculum are absent; the posterior 2 to 2.5 cm of these fields appear to be present but atrophic. The rostral perirhinal cortex is also absent but some of the posterior perirhinal and parahippocampal cortex remains intact. The cerebellum was also noticeably atrophic, which the authors attributed to H.M.'s long history of antiepileptic medication. Although H.M.'s medial temporal lobectomy appears to be less complete than previously believed, the functional status of the remaining tissue has yet to be determined. Future positron emission tomography (PET) studies are planned to address this question (Corkin et al., 1994). It is important to note that even with the use of modern imaging 7 technology, the full extent of H . M ' s lesion can only be accurately ascertained by a post mortem histological examination. Although such an examination may be able to clarify the extent of the excision made in 1953, H.M's previous history of severe epilepsy and his now advanced age may make a firm correlation between neuropathology and mnemonic deficits difficult, if not impossible. THE MEDIAL TEMPORAL LOBE AMNESIC SYNDROME The amnesic syndrome resulting from bilateral damage to the medial temporal lobe is characterized by a profound deficit in the ability to acquire new information or to remember day-to-day events, despite an intact immediate memory and normal intellectual functioning. Retrograde memory deficits of varying degrees may also be present. Memory for overlearned information such as general world knowledge and social and language skills tends to remain intact. Additionally, the ability to acquire and retain a variety of motor, perceptuo-motor, and cognitive skills, in the absence of any memory for the learning episode, is also present (Squire & Cohen, 1981). Based on their work with H . M . and several other neuropsychological patients (including those described by Sawa, Ueki, Arita, & Harada, 1954; Terzian, 1958; and Terzian & Dalle Ore, 1955), Scoville and Milner (1957) and Milner (1966) posited that the amnesic syndrome associated with lesions to the M T L was the result of bilateral damage to the hippocampal region. The hippocampal region was originally thought to include the hippocampus (presumably the dentate gyrus and Ammon's horn) and the parahippocampal cortex, a then poorly defined cortical area surrounding and underlying the hippocampus. This hypothesis was proffered for two reasons. First, the degree of memory impairment was positively correlated with the extent of hippocampal removal. If the lesion did not extend far enough posteriorly to damage the hippocampus memory impairments were absent. Additionally, no residual memory deficits were observed in patients after excisions of only the uncus or amygdala. Second, unilateral temporal lobectomy (which involves the removal of the uncus, amygdala, hippocampus, and varying amounts of the lateral neocortex) typically results in only a mild, material-specific memory disorder: Left temporal lobectomies produce deficits in verbal memory (Hermann, Wyler, Somes, Dohan, 8 Berry, & Clement, 1994; Milner, 1972), whereas excisions of the right temporal lobe produce deficits in visiospatial memory (Petrides, 1985; Pi gott & Milner, 1993; Smith & Milner, 1989). However, in a small subset of patients who had received a unilateral temporal lobectomy, a memory impairment nearly as severe as the one seen in H . M . developed (Penfield & Milner, 1958). Electrophysiological and post mortem examinations of these patients revealed substantial abnormalities in the hippocampal region contralateral to the excised lobes, presumably of an organic nature. In some cases, this hippocampal damage was actually greater than that sustained as a result of the temporal lobectomy. Thus, in effect, when the surgical procedure was performed an inadvertent bilateral hippocampal lesion was produced. The fact that both hippocampi had to be damaged before a severe memory deficit became apparent led to the conclusion that it was the locus of damage responsible for M T L amnesia. The pattern of memory loss characteristic of the amnesic syndrome provided some of the first experimental evidence that memory is not a unitary process, an idea that can be traced back to Tolman (1932). Bilateral damage to the M T L can result in profound anterograde amnesia, however, patients with this memory deficit can, nevertheless, learn and retain certain types of information. Initially, these preserved learning capacities were thought to be limited to motor-skills, but there is now considerable evidence that a wide variety of other behaviors not related to motor skills can be learned and remembered. The inability to form new long-term memories combined with preserved motor and perceptual learning indicated the existence of at least two separate and dissociable forms of memory, most likely subserved by different anatomical substrates. Though there is now a general consensus regarding the existence of multiple forms of memory, classifying them into a theoretically meaningful framework has been difficult (Nadel, 1992). In general, most theories emphasize a dichotomy between the mnemonic abilities that are impaired in amnesia and those that are preserved. Characterizing the essential differences between these abilities has become a major goal of both clinical and animal research on amnesia (Mishkin, Malamut, & Bachevalier, 1984). Graf and Schacter (1985) emphasized a distinction between implicit and explicit 9 memory. Implicit memory refers to the way in which prior learning experiences can be expressed as a facilitation in future performance, unintentionally and without conscious recollection of the initial learning episode. Explicit memory refers to the conscious retrieval of previously stored information. Explicit memory entails the storing of representations of objects, events, and the relations among them. Another distinction that has been applied to the preserved and impaired learning capacities in amnesics is that of procedural versus declarative knowledge (Cohen & Squire, 1980; Squire, 1992). Procedural learning, such as the acquisition of motor skills, can occur in the absence of memory for the stimulus material and/or the learning situation itself. By contrast, declarative memory, is memory for facts or information that is acquired through learning and is consciously accessible. The declarative memory system is based on neural representations of the specific outcomes or results of operations performed by those processing structures or procedures activated by the learning task. Declarative memory includes memories for fact, events, faces, and spatial layouts. It is referred to as knowing "that" in contrast to procedural memory which is referred to as knowing "how." Other theorists have suggested a dichotomy between what have they termed as "memory formation" and "habit formation" (Hirsh,1974; 1980; Mishkin et al., 1984). This idea was initially suggested as a potential way to reconcile stimulus-stimulus (S-S) learning theory (Tolman, 1932) and stimulus-response (S-R) learning theory (Hull, 1943) into an integrated framework. Accordingly, memories are thought to represent cognitive S-S relationships and would entail the storage of facts, events, temporal relations, and other similar "explicit" information. Habits, on the other hand, are believed to represent noncognitive S-R bonds, such as the learning of a motor sequence or perceptual identification. Neuroanatomically, memories are believed to be subserved by a corticolimbic system, which is damaged in M T L amnesia, whereas habits are thought to be mediated by a cortico-striatal system, which is preserved in M T L amnesics. Working memory and reference memory are two distinctions drawn primarily from the animal learning literature (Honig, 1978; Olton, Becker, & Handelman, 1979), but they map loosely onto the previous theories based largely on human memory loss. As much of 10 the current work in M T L amnesia involves the use of animal models, a description of this dichotomy is pertinent. In experimental learning paradigms, working memory represents information that is useful for only a small part of the experiment, usually one trial. This "trial unique" information is subject to temporal constraints and is constantly changing. Working memory can be viewed as a component of declarative or explicit memory. In contrast, reference memory is information that is useful for many trials and usually throughout the entire experiment. It is often operationalized as an invariant rule that the animal must learn in order to perform a task correctly. In the majority of cases, animals with hippocampal lesions are impaired on a variety of tasks that involve working memory but not reference memory, though this distinction does not always hold true (Jarrard, 1993, Olton, & Feustle, 1981). It should be emphasized that the labels applied to various mnemonic processes are not important and should not be the focus of debate. In many cases they are just slightly different ways of describing the same phenomena. What is important, however, is to consider the properties that have been associated with each term. It is apparent that several different conceptual frameworks (i.e., implicit/explicit; procedural/declarative) describe similar biologically real components of memory and that these two different forms of memory appear to be subserved by different neural systems (Nadel, 1992; Squire, 1992). A s it is impossible to remain atheoretical when describing memory processes, the implicit/explicit dichotomy will be adopted for the remainder of this thesis, but not with the idea that it is substantially different from other nomenclatures. Humans with M T L damage perform poorly on tests of explicit memory but show relatively intact implicit memory. This was first observed in patient H . M . who, with a few notable exceptions, could form no new explicit memories for facts or events but retained the capacity to learn certain perceptuo-motor tasks. A similar pattern of results has been demonstrated in other amnesics with various etiologies (Brooks & Baddely, 1976; Cohen, 1984; Cohen & Squire, 1980). Conversely, humans with lesions to the basal ganglia show a pattern of memory impairments that are the opposite of those displayed by patients with M T L lesions. That is, intact explicit memory but impaired implicit memory. Various 11 types of skill, motor, cognitive, and perceptual learning can be viewed as manifestations of implicit memory, as they are acquired slowly and can be performed without recollection of the prior learning situation. Humans with basal ganglia damage as a result of Huntington's or Parkinson's disease have been shown to be impaired on the pursuit rotor, mirror reading, and tower of Hanoi tasks; while performing normally on free-recall, cued-recall and recognition tests (Butters, Wolfe, Martone, Granholm, & Cermak, 1985; Marton, Butters, Payne, Becker, & Sax, 1984; Saint-Cyr, Taylor & Lang, 1988). The basal ganglia may be the anatomical locus of skill-based implicit learning. Another form of implicit learning that is preserved in human amnesics is priming. Unlike in skill-based implicit learning tasks, humans with basal ganglia damage display normal priming abilities. Deficits in priming are typically found only in humans with Alzheimer's disease (Bondi & Kaszniak, 1991; Heindel, Salmon, & Butters, 1990; Heindel, Salmon, Shults, Walicke, & Butters, 1989; Shimamura, Salmon, Squire, & Butters, 1987), a condition that causes widespread cortical degeneration . These findings indicate that priming may be subserved by the neocortex (Tulving & Schacter, 1990), an idea that has been supported by a recent P E T study (Squire, Ojemann, Meizen, Petersen, Videen, & Raichle, 1992). A N I M A L M O D E L S O F M E D I A L T E M P O R A L L O B E A M N E S I A : S P E C I F Y I N G T H E L O C U S O F D A M A G E Early attempts at producing an animal model of the memory deficits associated with resections of the M T L were guided by the suggestion that damage to the hippocampal region was responsible for producing the amnesic syndrome (Scoville & Milner, 1957; Milner, 1966). These first attempts, however, met with little success (Douglas, 1967; Isaacson, Douglas & Moore, 1961; Kimble 1963; Orbach, Milner, & Rasmussen, 1960; Correll & Scoville, 1965). Hippocampal lesions were shown to impair some forms of spatial memory, but their effects on other forms of memory were only minor (see O'Keefe & Nadel, 1978), especially when compared to the profound deficits seen in human amnesics. Initially, several reasons were offered to explain the divergent findings obtained 12 between human and animal hippocampal lesions. Mishkin (1978) suggested that this discrepancy could reflect either an evolutionary shift in the functions of the hippocampus or the use of incommensurate measures across species. Additionally, the demonstration of a wide variety of preserved learning capacities in amnesics indicated that the mnemonic demands of the tasks used to assess memory in the animal model could be a critical factor in its success. For a variety of reasons, the mnemonic demands of the learning paradigms employed in the initial studies were not analogous to the types of explicit memory tasks that human amnesics perform poorly on (Gaffan, 1974; Iversen, 1976). Early attempts at producing an animal model of M T L amnesia may have failed because they were using tasks that did not tap into the "right" kind of memory. Mishkin (1978) developed the first successful animal model of M T L amnesia in the monkey {Macaca mullata). T o test the monkeys' mnemonic abilities, he adapted a visual object-recognition memory paradigm that had been developed by Gaffan (1974). The monkeys were first presented with a baited object over the center well of a three-well food tray. Ten seconds later, the object was re-presented along with a single novel object over one of the lateral wells. The monkeys were required to displace the novel object in order to receive the food reward that was beneath it. Several hundred trials using different objects on each trial were required before the monkeys were able to master this task. Delays of 30, 60, and 120 s were then interposed between the presentation of the sample and the novel objects in order to test retention capabilities. The fact that monkeys had to hold a representation of an object in memory over a period of several minutes suggested that it may tap into the same kind of memory that was impaired in human amnesics. This was later confirmed when human amnesics were shown to be impaired on an analogous version of this task (Squire, Zola-Morgan, & Chen, 1988). This paradigm is now referred to as the nonrecuring items delayed nonmatching-to-sample (DNMS) task and is a central component of the monkey model of human amnesia. Mishkin (1978) found that preoperatively trained monkeys that received large bilateral lesions of the M T L took significantly longer to relearn the D N M S task and were impaired across delays, relative to unoperated controls. This was the first demonstration 13 that a lesion that approximated the one that H . M . had received could produce a severe, nonspatial memory deficit in animals. However, only lesions that included the hippocampus, amygdala, and their underlying cortical areas resulted in a significant memory deficit (this lesion is now known as the the H+A+ lesion where H is the hippocampus - A the amygdala and + the adjacent underlying cortex associated with each area). Lesions restricted to the hippocampus and its underlying cortex (H+) or the amygdala and its underlying cortex (A+) produced only mild amnesic symptoms. Mishkin concluded that conjoint damage to the hippocampus and amygdala was responsible for the amnesic syndrome in monkeys and, by extension, humans. He argued further that earlier attempts to develop an animal model had failed not only because they were using tasks that did not assess the type of memory that is impaired in amnesics but because investigators tried to mimic the human condition with hippocampal lesions alone. There are several notable features of the monkey model of amnesia. First, the deficit appears to be solely of memory. Once monkeys have learned the nonmatching rule their performance at short delays remains quite good. As the retention intervals increase beyond the capacity of immediate memory a delay dependent deficit is observed. Monkeys perform progressively more poorly as the delay is increased form 8 s to 10 min. The fact that animals perform well at short delays indicates that the deficits at long delays are not the result of a more generalized attentional or perceptual deficit (for a discussion and criticism of this idea see Ringo, 1988). Another critical feature of the monkey model of amnesia is its high degree of construct validity in relation to the human condition. Subsequent to finding deficits on D N M S , monkeys with M T L lesions were shown to be impaired on several other tasks that human amnesics perform poorly (Squire et al., 1988). These include delayed retention of object discrimination, concurrent object discrimination, and delayed response with and without distraction (Zola-Morgan & Squire, 1985). Similar to human amnesia, the memory deficits in monkeys appear to be multimodal, as they can be observed with the use of tactile as well as visual stimuli (Murray & Mishkin, 1984). Despite such multimodal memory impairments, monkeys with H+A+ lesions have been shown to perform normally 14 on several tests of perceptuo-motor learning (Zola-Morgan & Squire, 1982). Finally, the D N M S task may be seen as comprising two components. First, there is the rule learning component. T o succeed at the task, the monkeys must learn the nonmatching rule. This task takes advantage of the monkeys' innate tendency to explore novel stimuli and the nonmatching rule is typically learned in less than 300 trials (Mishkin 1978; Zola-Morgan & Squire, 1986). T o be considered to have learned the task the monkeys must be performing at 90% correct. The number of trials it takes to reach the criterion of 90% is used as one measure of D N M S performance. The other component of the D N M S task is the performance at the various delays. Once the monkeys have reached criterion at a relatively short delay, retention intervals from 10 s to 40 min are imposed between the presentation of the sample and novel objects. During delays, it is assumed that the monkey holds a mental representation of the sample object in memory until it can be compared to the novel object. Percent correct responses at each delay is the second measure of D N M S performance. Recent evidence indicates that the two components of the D N M S task may be subserved by different anatomical substrates (Meunier, Bachevalier, Mishkin, & Murray, 1993; Otto & Eichenbaum, 1992). S P E C I F Y I N G T H E L O C U S O F D A M A G E IN A M N E S I A : A N A T O M I C A L B A C K G R O U N D In primates, the medial temporal lobe is a large region of the brain that consists of the entorhinal and perirhinal cortex (often referred to in combination as the rhinal cortex), the parahippocampal gyrus, the hippocampus, and the amygdala (Figure 1). The periamygdaloid cortex and parts of the piriform cortex border these areas and are sometimes also included. The amygdala occupies the anterior pole of the M T L , the hippocampus is positioned posterior to the amygdala. On the ventral surface, the rhinal cortex underlies the amygdala and anterior hippocampus. The entorhinal cortex occupies the medial bank of the rhinal sulcus, the perirhinal cortex the lateral bank. Posterior to this lies the remaining cortex of the parahippocampal gyrus (Murray, 1992). Even though Milner (1966) had indicated that damage to the cortex adjacent to the hippocampus may contribute to amnesia, many early animal studies failed to acknowledge 15 Figure 1. A line drawing of the ventral surface of the monkey brain showing the position of the hippocampus (H) and amydala (A) and the underlying perirhinal (PRh) and entorhinal (ERh) cortices. Ots-occipital-temporal sulcus; rs-rhinal sulcus. From Murray, 1989 (used with permission). 17 the importance of these areas or gave them only a cursory mention. A n additional problem at this time was that these cortical regions were anatomically ill-defined. The cytoarchilecture and connectivity of the M T L has only recently been characterized and is still the subject of much debate (Amaral & Witter, 1995). Moreover, Mishkin's (1978) conclusion that combined hippocampal-amygdala damage was necessary to produce amnesia did much to focus attention on the deep structures of the M T L . As a prelude to a discussion of the effects of M T L lesions in animal models of amnesia, the following section will describe the anatomy and physiology of the hippocampus and related structures. It should be noted that the majority of the anatomical delineation of the hippocampus has been carried out in rodents. Although much of the structure and connectivity is most likely preserved across species, there are no doubt interspecies differences in anatomy. HIPPOCAMPUS In the literature, three terms are used to describe a group of anatomically related structures in the M T L of primates and the incipient temporal lobe of rodents: hippocampus, hippocampus proper, and hippocampal formation. The term hippocampus typically denotes all of the subfields of Ammon's horn and the dentate gyrus, whereas hippocampus proper refers only to Ammon's horn. Despite extensive investigation, however, there is still considerable debate as to which brain structures should be included under the term of hippocampal formation. For the purpose of this thesis, the schema of Amaral and Witter (1995) will be used. Accordingly, the hippocampal formation is said to consist of six cytoarchitectonically distinct regions: the entorhinal cortex, the dentate gyrus, the subiculum, presubiculum, and parasubiculum (often referred to collectively as the subicular complex), and the hippocampus proper. Following the cornu ammonis nomenclature of Lorento de No (1934), the hippocampus proper can be further divided into the C A 1-CA4 subfields. Figure 2 shows a coronal section through the dorsal hippocampus and dentate gyrus of the rat. The C A 1 sector consists of a band of cells occupying the most dorsal aspect of the hippocampus. Laterally, there is a clearly defined transition between the CA1 and C A 2 regions marked by 18 Figure 2. A coronal section through the dorsal hippocampus of the rat showing the subfields of Ammon's horn (CA1-CA2 -CA3) and the dentate gyrus (DG). Alv-alveus; PoDG-polymorphic layer of dentate gyrus. From Amaral and Witter, 1995 (used with permission). 1 9 20 an increase in size of the pyramidal cells. The transition from the C A 2 to the C A 3 is not observable in conventionally stained material but lies somewhere in the lateral curvature. The transition of C A 3 to C A 4 occurs within the blades of the V-shaped dentate gyrus (Schmidt-Kastner & Freund, 1991). The C A 4 is also difficult to distinguish in normal histological preparations and is sometimes mistakingly referred to as the hilus of the dentate gyrus (Amaral & Witter, 1995). Each of the subfields of Ammon's horn has a similar laminar structure. The primary cell layer of each field is composed of large pyramidal cells and is called stratum pyramidale. Superior to this is a narrow, relatively cell-free layer, the stratum oriens, and the fiber tracts of the alveus. Inferior to the pyramidal cell band is the stratum radiatum and stratum moleculare. In the C A 3 field only, a narrow acellular region just above the pyramidal cell layer contains the mossy fiber axons originating from the dentate gyrus. This area is referred to as the stratum lucidum. The dentate gyrus itself also consists of three lamina. A relatively cell-free molecular layer, a densely packed granule cell layer, and a polymorphic cell layer, which contains the mossy cells and a variety of other morphologically distinct neurons (Amaral & Witter, 1995; Van Groen & Wyss, 1990). In humans, the laminar, intertwined subfields resemble the shape of a seahorse, for which the hippocampus received its Greek name (Knowles, 1992). The three-dimensional shape and position of the rodent hippocampal formation is relatively complex (Figure 3). It appears grossly as an "elongated structure with its long axis extending in a C-shaped fashion from the septal nuclei of the basal forebrain rostrodorsally, over and behind the diencephalon, to the incipient temporal lobe caudoventrally" (Amaral & Witter, 1995, p. 444). The long axis is referred to as the septotemporal axis and the orthogonal axis is referred to as the transverse axis. A n interesting feature of the structure of the hippocampal formation is that not all of the subfields are present at any given level. For example, in the extreme septal pole only the the dentate gyrus and the C A 1 - C A 3 subfields can be seen. The subiculum does not appear until approximately a third of the way back toward the temporal pole. The presubiculum and parasubiculm appear even farther back along the septotemporal axis. The entorhinal cortex is located in the most caudal and ventral portion of the hippocampal formation 21 Figure 3. A line drawing of the rat brain illustrating the three-dimensional position of the C-shaped hippocampus. From Amaral and Witter, 1995 (used with permission). 22 23 (Amaral & Witter, 1989; Amaral & Witter, 1995). The primary reason for the inclusion of the six aformentioned stuctures under the construct of hippocampal formation is that they are connected, one to the next, by distinct and largely unidirectional projections. The major source of input to the hippocampus arises from the entorhinal cortex, via the perforant path, to the dentate gyrus. This projection is not reciprocal because none of the dentate cells send afferents back to the entorhinal cortex. Unidirectional projections are also seen from the dentate gyrus to the C A 3 , via the Mossy fibers, and from the C A 3 to the C A 1 , via the Schaffer collaterals. This circuit forms the classic trisynaptic pathway of the hippocampus. Connections from the CA1 to the subiculum are also largely unidirectional. Such unidirectional cortico-cortico projections are extremely atypical, and provide the primary justification for the inclusion of structures under the rubric of hippocampal formation. In most other cortical areas reciprocal projections dominate. For example, the entorhinal cortex, receives its major source of cortical input from the adjacent perirhinal cortex. The perirhinal-entorhinal projections are, however, reciprocal. Thus, despite its close functional relationship with the hippocampus and the entorhinal cortex, the perirhinal cortex is not considered to be part of the hippocampal formation (Amaral & Witter, 1995). Based on the anatomical and physiological evidence available at the time, Andersen, Bliss and Skrede (1971) proposed that information flow within the hippocampus occurred primarily in distinct lamellae (slices) and that these slices were oriented perpendicular to the long (septotemporal) axis of the hippocampus. Thus, stimuli would activate the entorhinal cortex and successively activate each component in the trisynaptic circuit. This activation would occur only in a small strip of the hippocampus that presumably received topographically organized projections from the entorhinal cortex. Accordingly, each lamalla may be suspected to operate as a functionally independent unit, with minimal influence from neighboring slices. The hippocampus could then be viewed as comprising of a series of slices that are stacked against one another to form the septotemporal axis. The major flow of information would be within each slice rather than between adjacent slices. The lamellar hypothesis of hippocampal organization was critical to the development of the in vitro hippocampal slice preparation (Amaral & Witter, 1989). 24 The veracity of the lamellar hypothesis has recently been questioned by Amaral and Witter (1989) who believe that recent anatomical and electrophysiological data do not support a lamellar organization. A good example of this is the recently discovered organization of entorhinal-dentate gyrus projections. Localized injections of radiolabeled amino acids or the anterograde tracer Phaseolus vulgaris leucoagglutinin ( P H A - L ) into the entorhinal cortex did not result in the labeling of discrete hippocampal lamellae. Instead, widespread labeling across the entire septotemporal axis of the hippocampus was found (Steward, 1976; Witter, Jorritsma-Byham, & Wouterlood, 1992; Wyss, 1981). It appears that projections from a focal point in the entorhinal cortex diverge to an extensive portion of the long axis of the dentate gyrus. If these data were to be interpreted in terms of the lamellar hypothesis, each lamella would be 2.5 mm in thickness (Amaral & Witter, 1989). Electrophysiological studies in the cat have also revealed that focal stimulation of the entorhinal cortex results in activation of at least half of the septotemporal extent of the dentate gyrus (Van Groen & Lopes da Silva, 1985). A similar pattern of longitudinal information flow can be observed in other areas of the hippocampal formation. Associational projections arising from the molecular layer of the dentate gyrus appear to be organized to promote information flow along its long axis. Injections of P H A - L into discrete regions of the dentate gyrus labeled terminals as far away as 3 mm (Ishizuka, Weber, & Amaral, 1990). Additionally, Schaffer collaterals do not travel parallel to the pyramidal cell layer as previously believed but have now been shown to project diversely across the septotemporal axis. Lamellar organization is, however, present in one hippocampal pathway, this being the mossy fiber projections from the dentate gyrus to the CA3. The limited extent of these projections was first demonstrated by the early Golgi studies and more recently by anterograde P H A - L experiments (Amaral & Witter, 1989). The aformentioned findings, among others, led Amaral and Witter (1989) to conclude that, aside from the Mossy fiber projections, none of the intrinsic circuitry of the hippocampus is organized in a lammellar fashion. In fact, the opposite appears to be true. There are extensive and diverse projections along the septotemporal axis that are as well 25 organized as those along the transverse axis. The conceptualization of information flow within the hippocampus should be revised to integrate its recently discovered three-dimensional organization. ENTORHINAL CORTEX The entorhinal cortex constitutes a major part of the parahippocampal area in lower species and the parahippocampal gyrus in primates (Jones, 1993). In the rat, the entorhinal cortex makes up the ventroposterior convexity of the cerebral hemispheres. It extends dorsolaterally to approach the rhinal fissure and borders the parasubiculum medially and the piriform cortex and the amygdaloid complex rostrally (Amaral &Witter, 1995). The entorhinal cortex receives massive input from neocortical multimodal association areas as well as direct input from primary sensory areas. Additionally, it receives projections from numerous subcortical structures including the raphe' nucleus, the ventral tegmental area, the locus coeruleus, the septum, the thalamus, the hypothalamus, the basal forebrain, the amygdala, and the claustrum (Jones, 1993). The close functional relationship between the entorhinal cortex and the hippocampus merits a further consideration of its anatomy and morphology. Structurally, the entorhinal cortex appears to be transitional between the neocortex (isocortex) and the hippocampus proper (allocortex). This has led to some confusion as to its laminar structure. Ramon y Cajal (1911) identified six layers, four cellular layers (layers II, III, V , and VI) and two acellular or plexiform layers (layer I and layer IV, the latter is also referred to as the lamina dessicans). Lorento de No (1934) identified six layers (cellular layers II, III, IV, V , and VI). Similar to Ramon y Cajal's nomenclature, layer I was a cell free layer and there was a cell free layer between layers III and IV. For the present purposes, I will adopt the nomenclature of Amaral and Witter (1995), which is a slight modification of Ramon y Cajal's (1911). Their schema, including a brief description of the cell types found in each layer is as follows. Layer I is the most superficial plexiform or molecular layer. It consists of a few widely dispersed GABAergic neurons, stellate and horizontal cells that project to the dentate gyrus and hippocampus proper. The principal component of layer I appears to be a large number of transversely oriented dendritic fibers. Layer II 26 contains mostly medium-sized to large stellate cells that tend to be clustered together. These stellate cells are considered to be the primary source of fibers for the perforant path. However, pyramidal, multipolar, and horizontal cells also contribute to the perforant path. Layer III contains a heterogeneous mixture of cells of various shapes and sizes, of which the pyramidal cell is the most numerous. These pyramidal cells give rise to axons that project directly to the CA1 and subiculum. Layer III also contains multipolar, stellate, fusiform, horizontal, and bipolar cells, the fibers of which contribute significantly to the perforant path. Layer IV, or the lamina dessicans, is a cell free layer located between layers III and IV. Layer V contains mostly large pyramidal cells with a lesser number of small pyramidal cells and other polymorphic cell types. The density of cells within this layer can vary considerably depending on the topographic location. Layer VI contains a heterogeneous population of cell sizes and shapes. At caudal levels, this layer appears to have a columnar organization. The rat entorhinal cortex can be subdivided into two general areas, the medial entorhinal area ( M E A ) and the lateral entorhinal area (LEA). These subdivsions are based upon the cytoarchitectonic differences observed in the various lamina. For example, cells in layer II of the L E A are more densely packed than in the M E A . This makes the border between layer II and III more pronounced in the L E A than in the M E A . Although the idea of two subdivisions of the entorhinal cortex is generally accepted (Amaral & Witter 1995), there have been several attempts to divide it even further (Insuasti, Herrorot, & Witter, 1994; Wyss, 1981). Again, these divisions are based on slight cytoarchitectonic differences in discrete regions of the entorhinal cortex. The functional or behavioral significance of these subdivisions remains to be determined. A s previously mentioned, the entorhinal cortex provides the major source of afferent input to the hippocampus, via the perforant pathway. The majority of these fibers arise from layers II and III, with a small contribution of fibers from the deeper layers, and terminate predominately in the molecular layer of the dentate gyrus (Lopes da Silva, Witter, Boeijinga, & Lohman, 1990). So substantial is this innervation, that it has been estimated that entorhinal synapses comprise 85% of the total population of synapses within the 27 dentate gyrus (Matthews, Cotman & Lynch, 1976). These projections contain glutamate and aspartate and are most likely excitatory (Di Lauro, Schmid & Meek, 1981; Storm-Mathisen, 1972), though peptide neurotransmitters somatostatin, enkephalin, and cholecystokinin are also present in the perforant path (Chavkin, Shoemaker, Mcginty, Bayon, & Bloom, 1985; Fredens, Steengard-Pedersen, & Larson, 1984; Morrison, Benoit, Magistretti, Ling, & Bloom, 1982). Additional projections from the entorhinal cortex directly to all other subfields of Ammon's horn also exist but are less well characterized. The subicular complex receives afferents primarily from layer III (Amaral & Witter, 1994) though some fibers originating from layer II have recently been found (Lingenhohl & Finch, 1991; Tamamaki & Nojyo, 1993). Similar to the dentate gyrus, the C A 2 and C A 3 receives direct projections from layer II of the entorhinal cortex (Steward & Scoville, 1976; Tamamaki & Nojyo, 1993). Entorhinal projections to the C A 1 arise from layer III rather than from layer II (Steward & Scoville, 1976) and are topographically organized. Septal parts of the CA1 receive projections from L E A , whereas the temporal regions receive input from the M E A (Van Groen & Lopes da Silva, 1985). Taken in combination with the previously mentioned topographic organization of the entorhinal cortex, Van Groen and Wyss (1990) suggested that each one third of the C A 1 may have a distinct role in hippocampal functioning; "the septal parts of the C A 1 having primarily neocortical interactions and temporal parts of the CA1 relating to visceral functions" (Van Groen & Wyss, 1990, p. 527). The entorhinal cortex receives projections from a number of different hippocampal regions. The most extensive of these being from the subicular complex. These projections terminate predominantly in the deep layers of both the L E A and the M E A . A small number of fibers also synapse in layer III (Amaral & Witter, 1995). These projections are believed to be predominantly excitatory (Jones, 1993). Fibers directly from the C A 1 to the entorhinal cortex provide the second major source of input. These projections arise from the full septotemporal and transverse axis of the C A 1 and terminate predominantly in layer VI. They appear to be organized topographically. Fibers from the septal one third of the hippocampus terminate primarily in the caudal portions of the entorhinal cortex whereas 28 projections that arise from temporal regions terminate in the lateral entorhinal cortex (Van Groen & Wyss, 1990). The recently discovered diversity of entorhinal-hippocampal connections all but obviates the classical trisynaptic circuit. The widely held notion that an exclusively serial or sequential flow of information from the entorhinal cortex to the dentate gyrus and through the subfields of Ammon's horn is no longer tenable. The entorhinal cortex gives rise to fibers that project to all fields of the hippocampus, bypassing the classical trisynaptic circuit. Thus, activity in the entorhinal cortex can influence hippocampal neurons in a variety of ways. For example, the CA1 can both receive information directly from the entorhinal cortex and via the perforant path. How these two sources of input into the CA1 may effect its function is presently unclear. Jones (1993) has suggested that the multiple pathways into the hippocampus may function in combination to differentially regulate the flow of information en route from the neocortex. The direct route to the C A 1 and C A 3 may be primarily involved in the moment-to-moment transfer of information ,whereas the perforant path fibers may become active only when the frequency of afferent activity to the entorhinal cortex increases. Whatever the functional consequences are, it is clear that the idea of a trisynaptic circuit, much like the notion of lamellar organization, is in need of substantial revision (Amaral & Witter, 1995). S P E C I F Y I N G T H E L O C U S O F D A M A G E IN M E D I A L T E M P O R A L L O B E A M N E S I A : N E U R O B E H A V I O R A L A N A L Y S E S The availability of a monkey model and, more recently, a rodent model of M T L amnesia has now made it possible to determine experimentally which are the critical structures that, when damaged, produce amnesia. The following section will describe how ablation studies have changed some previously held notions about amnesia and how they have contributed to our understanding of the anatomical locus of memory deficits produced by lesions in the M T L . MONKEY MODELS OF MEDIAL TEMPORAL LOBE AMNESIA The Temporal Stem Hypothesis Horel (1978) challenged the assertion that M T L amnesia in humans was caused by 29 damage to the hippocampal region. Instead he proposed that the severing of the white matter tracts that lie medial to the temporal lobe, "the temporal stem," as a result of M T L resections was responsible for the memory impairments. The temporal stem contains afferent and efferent fiber connections of the temporal cortex and links these areas with various subcortical targets. Temporal lobe resections were hypothesized to produce amnesia not by destroying the hippocampus and amygdala but by destroying the connections between cortical and subcortical structures involved in memory. Horel's theory was later applied to explain the memory deficits observed in the monkey model of amnesia. Accordingly, the most severe deficits are found following the combined H+A+ lesion; not because of the synergistic effect of destroying both structures, but because of destroying the white matter paths that lie adjacent to them. When the hippocampus or amygdala are lesioned separately an insufficient number of temporal stem fibers are destroyed to produce a memory impairment. The ablation of both structures in combination produces a much more radical transection of the white matter tracts than lesioning either structure alone. This extensive transection was postulated to be responsible for the amnesic symptoms in monkeys with H+A+ lesions. Zola-Morgan, Squire, and Mishkin (1982) tested the role of white matter damage in amnesia by comparing a group of monkeys with lesions of the temporal stem and a group of monkeys with H+A+ lesions to a group of normal monkeys. Monkeys with temporal stem lesions learned the task at a rate equivalent to controls, whereas monkeys with H+A+ lesions were significantly more impaired than either group. A similar pattern was found across delays of up to 10 min. The performance of temporal stem-transected monkeys did not differ from controls, and H+A+ lesioned monkeys were significantly impaired compared to both the temporal stem lesioned group and the normal monkeys. These findings indicate that damage to the temporal stem cannot account for the amnesic symptoms in monkeys with H+A+ lesions. Though Horel's ideas about the temporal stem appear to be incorrect, he was one of the first researchers to question seriously the relationship between hippocampal damage and amnesia. His foresight in emphasizing the importance of the cortical regions of the temporal lobes has not been sufficiently 3 0 recognized. The Amygdala Although additional studies did support the hypothesis that conjoint hippocampal-amygdala damage was necessary to produce a deficit in object-recognition memory (Murray & Mishkin, 1984; Saunders, Murray, & Mishkin, 1984) more recent evidence indicates that the severe D N M S deficits resulting from H+A+ lesions are more likely the result of damage to the cortex underlying and adjacent to the amygdala (the rhinal cortex) rather than damage to the amygdala itself. Initially, amygdala damage was thought to contribute to the memory deficits for three reasons. First, other than the hippocampus, the amygdala is the largest structure, by volume, that is damaged in temporal lobe ablations. Second, knowledge of the full anatomical extent of the perirhinal cortex was lacking. This lack of knowledge manifested itself as an incorrect assessment of neuropathology in some of these studies. Third, the amygdaloid complex was known to receive strong connections from modality-specific neocortical processing areas and to project to other brain regions thought to be important in memory, such as the mediodorsal nucleus of the thalamus and the prefrontal cortex (Murray, 1992). The first experimental evidence against the involvement of amygdala damage in M T L amnesia was provided by Zola-Morgan and Squire (1986). Unlike previous studies (Mishkin, 1978; Murray & Mishkin, 1984; Saunders et al., 1984), they found that damage to the hippocampal region alone (the H+ lesion) could produce a significant memory impairment, though it was not as severe as the one produced by the H+A+ lesion. The greater impairment in the H+A+ group could be caused by lesions of the amygdala, but amygdala damage is not the only component of this lesion that distinguishes it from the H+ lesion. The anterior/ventral surgical approach used to ablate the amygdala necessarily damages the adjacent periamygdaloid, entorhinal, and perirhinal cortices. Lesions to the underlying cortex, rather than to the amygdala itself, could be the reason that the H+A+ lesion produces a greater memory impairment than the H+ lesion. T o test this idea, electrolytic lesions of the amygdala were made in two groups of monkeys (Zola-Morgan, Squire, & Amaral, 1989). In one group, a circumscribed 31 electrolytic lesion of the amygdala, that did not damage its underlying cortex, were made (the A lesion). In the other group, similar amygdala lesions were made in combination with H+ lesions (the H+A lesion). The performance of these monkeys was then compared to a group of monkeys with H+ lesions and a group of normal monkeys. Monkeys with A lesions performed equivalent to controls on the D N M S task. Those with H+A lesions did not perform significantly worse than the H+ group, suggesting that amygdala damage contributes little to deficits in D N M S performance. Additionally, monkeys with an H+A+ lesion were significantly more impaired than both the H+ and H+A groups, indicating the damage to the cortical areas underlying the amygdala was adding to the deficits produced by the H+ lesion. This idea was confirmed in a subsequent study (Zola-Morgan, Squire, Clower, & Rempel, 1993) in which monkeys received an H+ lesion that was extended forward to include the anterior entorhinal cortex and perirhinal cortex (H++ lesion). When this additional cortical damage was added, the monkeys were as severely impaired as those with the H+A+ lesion. Several other pieces of evidence indicate that the cortex underlying and adjacent to the amygdala is involved in object-recognition memory. Based on new anatomical distinctions in this region, Murray (1992) reanalyzed the extent of some earlier lesions (Murray & Mishkin, 1984) in which memory impairments were attributed to combined hippocampal-amygdala damage. It was found that, in addition to the amygdala damage, most ablations included large portions of the perirhinal cortex. Furthermore, large lesions of the rhinal cortex and more restricted lesions of the perirhinal cortex have both been shown to impair D N M S performance (Meunier et al., 1993; Zola-Morgan, Squire, Amaral, & Suzuki, 1989). Accordingly, it is now believed that amygdala damage does not play a significant role in object-recognition memory deficits or human amnesia (Murray, 1992; Squire, 1992; Squire & Zola-Morgan, 1991). Damage to the adjacent cortical areas appears to be responsible for the memory impairments once attributed to amygdala damage. The Hippocampus Contrary to initial reports (Mishkin, 1978; Murray & Mishkin, 1984; Saunders et al., 1984), subsequent investigations found that lesions of the hippocampal region alone 32 could significantly impair D N M S performance (Mahut, Zola-Morgan, & Moss, 1982; Zola-Morgan & Squire, 1986). Naive monkeys with H+ lesions took longer to learn the D N M S task and were impaired at delays of up to 10 min, relative to controls. One possible explanation for the observed differences in the various experiments was that the monkeys had received different degrees of preoperative training experience (Zola-Morgan & Squire, 1986). Memory impairments following hippocampal lesions were most severe when the animals had received no preoperative training (Mahut et al., 1982; Zola-Morgan & Squire, 1986), only mild when preoperative training was given (Mishkin, 1978), and absent when there was extensive preoperative training (Murray & Mishkin, 1984). Lack of preoperative training appears to exacerbate the severity of the memory deficits produced by lesions of the hippocampal region. As with the amygdala, hippocampal ablations in primates also damage surrounding cortical tissue. In the case of hippocampal lesions, this damage occurs primarily to the posterior rhinal cortex and the parahippocampal gyrus. The contribution that this damage makes to the behavioral deficits following an H+ lesion has only recently been appreciated. Unlike previous studies, which have sought to determine the relative contributions of amygdalar or hippocampal damage to amnesia, current research has focused on trying to dissociate the mnemonic deficits produced by lesioning only the hippocampus with those produced by lesioning only the adjacent cortical areas. This has proven difficult because it is technically challenging to produce a circumscribed hippocampal lesion in monkeys. Conventional aspiration lesions always damage the adjacent underlying cortex, and stereotaxic lesions are unreliable because the size and shape of the hippocampus varies considerably from monkey to monkey (Alvarez-Royo, Clower, Zola-Morgan, & Squire, 1991). T o circumvent this last problem, Alvarez, Zola-Morgan, and Squire (1995; note that this finding was originally published in abstract form - Clower, Alvarez-Royo, Zola-Morgan, & Squire, 1991) produced circumscribed electrolytic lesions of the monkey hippocampus (the H lesion) that were guided by MRI images. Naive monkeys with H lesions learned the D N M S task at a rate equivalent to monkeys with H+ lesions and normal 33 monkeys, 4-8 weeks postsurgically and, again, 6-9 months later. When tested at delays of 8 s, 15 s, 60 s, 10 min, and 40 min the monkeys with the H lesions were found to be impaired only at the longest delays, 10 and 40 min. Moreover, it is important to note that during these long delays the monkeys were removed from the testing apparatus and returned to their home cages. The change of context during the retention period may have important implications for D N M S performance and could possibly have contributed to the deficits that were observed at the longer delays. Histological analysis revealed that the monkeys in the H group sustained hippocampal damage ranging from 38% to 75% with a mean of 57%. Those who argue in favor of the critical role of hippocampal damage in M T L amnesia emphasize the importance of the impairments at the 10 and 40 min delays. (Alvarez, Zola-Morgan, & Squire, 1995; Squire 1992; Squire & Zola-Morgan, 1991). Others believe that this pattern of results indicates that hippocampal damage plays only a minor role in amnesia (Mishkin & Murray, 1994; Murray, 1996; Nadel, 1992). Two additional studies in monkeys have used microinjections of neurotoxins guided by MRI images in an attempt to produce a restricted hippocampal lesion. These two studies have produced somewhat discrepant results. However, as they exist only in abstract form, the limited histological evidence available makes drawing any firm conclusions about locus of damage and memory impairments tenuous. Beason-Held, Rosene, & Moss (1993) found that naive monkeys with ibotenic acid lesions of the hippocampus took significantly longer to learn the D N M S task and were impaired at delays of 2 and 10 min. Preliminary histology from two of these animals revealed only 25% damage to the hippocampus. In contrast, O'Boyle et al. (1993) found no deficits in relearning the D N M S task or at delays of up to 2 min in preoperatively trained monkeys that had approximately 8 0 % of the hippocampus destroyed by ibotenic acid. T w o possible explanations can be put forth to account for these discrepant results. First, the differences could have been due to the different preoperative training experiences of the animals. Monkeys in the Beason-Held et al.'s study received no preoperative training, whereas those of O'Boyle et al. did. Second, it may be that a partial lesion to the hippocampus is more detrimental to memory than a more extensive lesion. The limited histology available 34 for these two experiments makes a critical evaluation of this hypothesis currently impossible. However, this issue will be covered in greater detail later in this thesis as it pertains to ischemia-induced hippocampal damage and amnesia. Another way to dissociate the effects of cortical damage from hippocampal damage is to lesion the cortex alone and to evaluate its effect on object-recognition memory. Zola-Morgan, Squire, Amaral, and Suzuki (1989) lesioned the perirhinal and parahippocampal cortices (the P R P H lesion) and found an interesting pattern of deficits on D N M S . Compared to normal monkeys or monkeys with H+A-i- lesions, monkeys with the P R P H lesion took significantly longer to learn the D N M S task. Some monkeys were unable to reach criterion even after more than 1500 trials. When tested at delays the P R P H lesioned monkeys were more impaired than normal monkeys, but were not quite as impaired as monkeys with H+A-t- lesions. These results are open to two interpretations. The fact that lesions that spared the hippocampus produced almost as severe a deficit as lesions that destroyed it is totally inconsistent with previous assertions about the critical role of hippocampal damage in D N M S . However, the fact that monkeys with the H+A-i- lesion were more impaired than P R P H lesioned monkeys led Zola-Morgan et al. (1989) to conclude that the additional hippocampal damage was responsible for this greater impairment. Meunier et al. (1993) have also found that lesions of the cortex adjacent to the hippocampus impair D N M S performance. Compared to normal monkeys, pretrained monkeys with rhinal cortex lesions (which included the most of the perirhinal and entorhinal cortices) were impaired at relearning the D N M S task and at performance at delays up to 2 min. Furthermore, Meunier et al. went on to dissociate the behavioral effects of entorhinal versus perirhinal lesions. Monkeys with ablations of the perirhinal cortex were impaired at relearning the D N M S task and were impaired at delays, but not as impaired as animals with complete rhinal cortex lesions. Monkeys with entorhinal lesions showed no deficits in relearning the D N M S task but were as impaired at delays as the perirhinal group. These findings suggest that the perirhinal cortex is disproportionately involved in the learning or maintenance of the D N M S rule, whereas both the entorhinal and 3 5 perirhinal areas contribute to performance over delays. The latter idea is underscored by the fact that lesions to either the entorhinal cortex or the perirhinal cortex produce submaximal delay-dependent deficits and that the greatest deficit was observed when these two lesions were combined. This fact may also explain why Zola-Morgan et al. (1989) found a greater memory deficit following H+A+ lesions than with lesions that only damaged the perirhinal cortex and parahippocampal gyrus. They attributed the greater deficit in the H+A+ group to the hippocampal damage that was not sustained in the P R P H lesion group. However, the H+A+ lesion also included some of the entorhinal cortex. This additional damage to the entorhinal cortex could be producing the deficits that were attributed to the hippocampal component of the H+A+ lesion (Meunier et al., 1993). RAT MODELS OF MEDIAL TEMPORAL LOBE AMNESIA The major difficulty in drawing firm conclusions about the role of the hippocampus in the primate model of M T L amnesia has been the inability to reliably produce circumscribed hippocampal lesions in monkeys. Studies with rats provide an alternative approach to dissociating the behavioral effects of hippocampal and/or amygdala damage from adjacent cortical damage. Rodents are particularly suitable for this type of investigation because their hippocampus lies beneath the parietal cortex and can be removed without necessarily damaging the entorhinal and perirhinal cortices. Several rodent models of M T L amnesia have recently been developed, all of which use a modified version of the D N M S paradigm (Aggleton,1985; Rothblat & Hayes, 1987). The task devised by Mumby, Pinel & Wood (1990) is the most similar to the D N M S task used with monkeys. Rats are first presented with a sample "junk" object over a previously baited food well at one end of an elevated runway. The rat must then displace the object with its head and/or forelimbs in order to obtain the concealed food reinforcement. The sample object is then removed and placed at the opposite end of the runway, behind an opaque guillotine door, and adjacent to the previously positioned novel object. The door is then opened and the rat must locomote to the other end of the box and displace the novel object in order for food reinforcement to be delivered. Norma] rats typically learn this task to criterion (85% correct over two consecutive sesions of 20 trials) in 200-400 trials. Once 3 6 the nonmatching rule has been learned, delays of 15-300 s are then interposed between the presentation of the sample and novel objects in order to test memory retention. The final level of performance of rats on the D N M S task is comparable to the level of performance attained by monkeys, at delays of less than 5 min. The manner in which rats solve this D N M S task has recently become the subject of some debate. In the standard method of administering this task, the novel object is always positioned over the food well before the sample object during the choice phase. Herremans, Hijzen, and Slangen (1995) found that when the order of placement of the sample and novel objects was varied randomly during the choice phase the performance of the rats dropped from 87% to 28% correct, less than would be expected by chance. They argued that rats do not use memory to solve this task but may instead may be using an olfactory discrimination based on which object was last handled by the experimenter. Because the objects are manually positioned, they are inadvertently scent marked (e.g., human sweat) by the experimenter. Due to the fixed order of positioning, when the rat goes to make a choice the scent mark on the novel object has had more time to wear off, and may therefore be less intense than the odor mark on the sample object. Subsequently, the rats may learn to discriminate between the two objects based on odor intensity and select the one with the less intense odor in order to receive reinforcement. At least three pieces of evidence do not support to this interpretation. First, Mumby, Kornecook, Wood, and Pinel (1995) attempted to train rats on two tasks designed to assess their ability to use olfactory cues in order to solve an ostensibly visually guided object-recognition tasks. In the first experiment, rats had to learn to discriminate betwen two visually identical stimulus objects (7 cm in diameter plastic bottle lids) based on olfactory cues. T o begin, one object was placed over a food well in the D N M S apparatus. This object was designated the S+ because it was handled first by the experimenter. Its choice was rewarded. Four seconds later a second lid was placed over the adjacent food well. This lid was designated the S- by virtue of having been touched last by the experimenter and was never rewarded. Remarkably, all rats were eventually able to learn the discrimination, apparently based on which object had been touched more recently, even 37 though they were touched a mere 4 s apart. However, two important findings suggest that this is not the way in which rats normally solve the D N M S task. Rats in this study required 520-600 trials to reach criterion on the olfactory discrimination, whereas rats typically reach criterion on the D N M S task in 200-400 trials (Mumby et al.,1990; Mumby et al., 1992a; Wood et al., 1993). Additionally, when the delay between the presentation of the S+ and S- was extended from 4 to 15 s, the rats performance dropped to chance. In D N M S , increasing the delay from 4 to 15 s has little effect on choice accuracy (Mumby et al. 1995; Mumby et al., 1990; Wood et al., 1993) In the other experiment, rats were trained on D N M S under the standard protocol or the one used by Herremans et al. (1995; i.e., the random ordering of the placement of the sample and novel objects during the choice phase). No difference was found between the number of trials required by either group to reach criterion. Rats in the group in which the ordering of placement of the novel and sample objects was varied randomly learned the task at the same rate as the group in which the sample object was always touched last. Moreover, when contingencies where reversed between the two groups performance was unchanged. Second, if duplicate sample objects are used such that a sample object can be placed at one end of the runway for the initial phase and a duplicate sample object and the novel object can be placed at the other end of the runway for the choice phase, the rats performance is unchanged (unpublished observations). The use of the duplicate sample object makes it possible to position both objects simultaneously and thereby eliminate the intensity of residual odor cues as a possible confound. Third, an automated version of the D N M S task for rats has now been developed (Bussey, Muir, Robbins, 1994). This task is quite similar to the manual D N M S task except that the stimuli are computer generated images of objects presented on video touch screens. Rats can solve this task at a rate similar to the manual version. Obviously, rats cannot use olfactory cues to solve this task and therefore this paradigm demonstrates that rats are at least capable of solving a D N M S task based solely on visual cues. Damage to the hippocampus and/or amygdala has been shown to produce, at best, only mild deficits in D N M S performance in rats. Rothblat and Kromer (1991) found no 3 8 deficits in preoperatively-trained rats following hippocampal lesions. These rats were able to reacquire the the D N M S rule at a rate equivalent to controls and performed normally at delays of up to 30 s. Similarly, Aggleton, Hunt, and Rawlins (1986) found that naive rats that had received hippocampal lesions were able to learn a D N M S task at a normal rate and to perform accurately at delays as long as 60 s. Conjoint amygdala/hippocampal damage was examined by Mumby, Wood, and Pinel (1992). Rats were trained on D N M S at delays of 4, 15, 60, 120, and 600 s and then received bilateral lesions of the hippocampus, amygdala, or both. When retested postsurgically, the three experimental groups relearned the D N M S task at a rate equivalent to controls. Comparisons at the various delays revealed that hippocamapal, amygdala, or combined amygdalo-hippocampal lesioned rats were impaired only at the 600 s delay. Interestingly, several rats in this study received inadvertent damage to the rhinal cortex. Compared to unoperated controls, these rats were impaired at reacquiring the D N M S task and were impaired at all retention intervals. Rats that received no training prior to hippocampal lesions are slightly more impaired on D N M S than those receiving preoperative training. Such rats required more trials to learn the nonmatching rule than controls and were impaired at delays of 120 s (Mumby, Pinel, Kornecook, Shen, & Redila, 1995). Aggleton, Blindt, and Rawlins (1989) found that rats were profoundly impaired in postoperative D N M S performance following combined amygdala-hippocampal lesions. They emphasized, however, that inadvertent damage to the piriform cortex may have contributed to these impairments. Two factors that may influence a rats' performance on D N M S are the amount of preoperative training experience and the physical characteristics of the stimulus objects. Similar to what is found with monkeys, preoperative experience with the D N M S task may have an effect on the magnitude of the observed memory impairment following hippocampal lesions. Rats that receive preoperative training (Mumby et al., 1992a; Rothblat & Kromer, 1991) display milder deficits than those that receive no preoperative training (Mumby et al., 1995; Aggleton et al., 1989). The characterisitics of the stimulus objects used as a possible factor in mediating the severity of memory impairment produced by hippocampal lesions has only recently 39 begun to be addressed. Generally, the more similar stimulus objects are to one another in physical characterisitics (e. g., size, shape, color, texture) the more suceptible the animal's performance will be to hippocampal lesions. Similarly, if the pool of stimulus objects is small, such as it changes the demands of the task from object-recognition to a recency discrimination the more likely one is to find a deficit following hippocampal lesions (Rawlins, Lyford, Seferiades, Deacon, & Cassady, 1993; Shaw & Aggleton, 1993). As hippocampal and/or amygdala lesions have been shown to have little effect on object-recogniton memory in rats, several studies have begun to address the effects of cortical damage. Mumby and Pinel (1993) lesioned the rhinal cortex of pretrained rats and then retested their D N M S performance. Rhinal cortex lesioned animals reacquired the D N M S task in approximately the same number of trials as control animals. However, the lesioned rats were significantly impaired at longer retention intervals. Kornecook, L u i , Duva, Anzarut, and Pinel (1995) also found that rhinal cortex lesions produced deficits at retention intervals greater than 30 s in pretrained animals but did not effect their ability to reattain criterion or perform at shorter delays. A similar pattern of results was obtained by Otto and Eichenbaum (1992) using an odor-guided D N M S task. Rats with rhinal cortex lesions were able to learn the task at a normal rate but displayed significant impairments when delays were introduced. Consistent with the findings in monkeys, the above results suggest that the rhinal cortex may be responsible for maintaining representations of stimuli over long periods of time. While the hippocampus and amygdala may not be critical for object-recognition memory in rats, they do appear to have some mnemonic functions. The hippocampus, for example, may be particularly involved in spatial memory. Lesions of the hippocampus have been shown to produce deficits in a wide variety of spatial memory tasks including the Morris water-maze (Morris, Garrund, Rawlins, & Okeefe, 1982), delayed matching-to-position (Hunt, Kesner, & Evans, 1994), the split T-maze (Thomas & Gash, 1988), and several versions of the radial arm-maze (for reviews see Jarrard, 1993; O'Keefe & Nadel, 1978; Nadel, 1991; Worden, 1992). The amygdala, on the other hand, may function in the attribution of emotional salience to memories (LeDoux, 1993). 4 0 A M N E S I A P R O D U C E D B Y B R A I N D A M A G E O U T S I D E O F T H E T E M P O R A L L O B E Damage to midline diencephalic structures, the basal forebrain, or the prefrontal cortex can also result in memory impairments, some of which are quite similar to those that result from lesions to the M T L . These areas have strong reciprocal connections with various regions of the M T L , and this interconnectivity makes it difficult to determine if the amnesia produced by lesions to these areas is qualitatively different from that produced by lesions in the M T L . One possibility is that structures in the M T L may form a circuit with other more distal regions of the brain and that this entire circuit is involved in the formation of memories. Damage to any of the structures in this circuit may produce a qualitatively similar type of memory dysfunction. Another possibility is that each structure within this circuit makes a unique contribution to memory. The deficits that are observed following lesions to these various structures are only superficial and appear similar because our ability to experimentally distinguish between different types of amnesias is poor. Thus, discrete brain structures may play differential roles in encoding, storage, or retrieval. Damage any one of several areas and amnesia is produced. The nature of the amnesia, however, may be quite different in each case. The following sections will describe the similarities and the differences between the mnemonic deficits produced by lesions of the M T L and other brain regions. Within this context, attempts to experimentally dissociate the nature of the amnesia associated with a specific brain lesion will also be discussed. DIENCEPHALON Damage to the midline diencephalon can produce amnesic symptoms that are remarkably similar to those that result from lesions to the M T L . In humans, such damage can be caused by ischemic infarcts (Barbizet, Degos, Louarn, Nguyen, & Mas, 1981; Mills & Swanson, 1978), traumatic injury (Squire & Moore, 1979), infections (Degirolami, Haas, & Richardson, 1974), tumors (McEntee, Biber, Perl, & Benson, 1976; Sprofkin & Sciarra, 1952), or as a result of Korsakoff's syndrome, a disorder associated with chronic alcoholism (Butters & Cermak, 1980; Mair, Warrington, & Weiskrantz, 1979; Victor, Adams, & Collins, 1989). The thalamus and the hypothalamus compose the brain region 41 typically referred to as the diencephalon. These structures themselves comprise dozens of discrete nuclei that are interconnected with functionally dissimilar areas of the brain. Important for the present discussion is the fact that hippocampal afferents to the diencephalon arise almost exclusively from the subicular complex (Amaral & Witter, 1995) and project primarily to the anterior thalamic nuclei, the midline thalamic nuclei, and the mammillary nuclei. The relatively complex nuclear structure of the diencephalon has made it difficult to determine exactly which structure or combination of structures, when damaged, produce amnesia. Much of our knowledge about the localization of damage in humans is derived from the study of Korsakoff's syndrome, the most well characterized example of diencephalic amnesia. First diagnosed by Russian psychiatrist S.S. Korsakoff in 1887, the disorder produces a wide range of cognitive abnormalities, the primary being severe anterograde and less severe retrograde amnesia. Other symptoms may include a generalized confusional state, as well as deficits in reasoning, planning, spatial organization, and visual abstraction (Victor, 1994). Korsakoff's amnesics are also particularly poor at timing and temporal relations (Squire, 1982). Clinicopathological studies have shown that Korsakoff's amnesics consistently have damage to the mediodorsal nucleus (MDN) of the thalamus and the mammillary bodies (MB) of the hypothalamus (Charness & DeLapaz, 1987; Shimamura, Jernigan & Squire, 1988; Victor et al., 1989). Damage may also be found in other thalamic nuclei, the hippocampus, the orbitofrontal cortex, the pons, and the medulla (Jernigan, Schafer, Butters, & Cermak, 1991; Mayes, Meudell, Mann, & Pickering, 1988; Victor et al., 1989). Cortical atrophy and an enlargement of the sulci and ventricles have also been detected (Epstein, Pisani, & Fawcett, 1977; Fox, Ramsey, Huckman, & Proske, 1976), though their relationship to cognitive dysfunction is questionable (Victor, 1994). Interestingly, alcohol neurotoxicity per se is not responsible for these abnormalities. They appear to be the result of thiamine deficiency brought on by the malnutrition that often accompanies chronic alcoholism (Victor, 1994). Because Korsakoff's syndrome produces cognitive impairments other than 4 2 amnesia, the lesion(s) specifically responsible for the memory deficit have been difficult to identify reliably. Initially, damage to the M B was thought to play a key role, as they were one of the most consistently damaged structures in patients with Korsakoff's amnesia. More recent studies have tended to focus on the thalamus, particularly the M D N , which is also damaged in a large percentage of Korsakoff's amnesics. The possible role of thalamic damage in producing diencephalic amnesia was further underscored by the case of patient N.A. In 1960, N.A. received a stab wound to the base of the brain from a miniature fencing foil. Since this time, he has displayed severe anterograde amnesia, particularly for verbal information (Squire & Slater, 1978; Tueber, Milner & Vaughan, 1968). Computerized axial tomography (CAT) scans have revealed that the lesion is restricted primarily to the left M D N of the thalamus. The unilateral left hemisphere lesion may account for the strong verbal nature of N.A.'s amnesia (Squire & Moore, 1979). Animal studies, while they have been successful in modeling some of the mnemonic deficits associated with diencephalic amnesia, have done little to identify the critical nuclear structures. In pretrained monkeys, large electrolytic lesions of the thalamus impair the postoperative reacquisition and performance of the D N M S task (Aggleton and Mishkin, 1981). More restricted lesions have a similar, albeit attenuated, effect. Aggleton & Mishkin (1983) lesioned the medial thalamus, which included most of the midline thalamic nuclei as well as medial portions of the anterior and medial thalamic nuclei, in pretrained monkeys. Postsurgically, these monkeys required more trials to relearn the D N M S task and were impaired at delays of 30, 60, and 120 s, relative to sham-operated animals. Post mortem histological analysis revealed that all of the monkeys with thalamic lesions also had bilateral degeneration of the M B , making it impossible to attribute their deficits to thalamic damage alone. Discrete lesions of the M D N nucleus in both monkeys and rats has also been shown to result in object-recognition deficits. Zola-Morgan and Squire (1985) produced circumscribed lesions of the M D N of the thalamus in naive monkeys. These monkeys took significantly longer to learn the D N M S task than did controls and were impaired at delays up to 10 min. Naive rats with MDN-lesions took longer to learn the D N M S task than did 4 3 controls and were impaired at delays ranging from 30 to 300 s, though they performed normally on an object discrimination task. Morever, rats that received training prior to surgery took longer to reattain criterion postoperatively and were impaired at delays from 4 to 300 s, relative to sham-lesioned rats (Mumby, Pinel, & Dastur, 1993). Neither of these two studies reported significant damage to the M B , demonstrating that lesions confined to the M D N can produce memory impaiments on tasks sensitive to diencephalic amnesia in humans. Lesions to the M B have, however, also been shown to impair D N M S performance, but not as severely as lesions to the thalamus. Aggleton and Mishkin (1985) found that pretrained monkeys required more trials to reattain criterion post-surgically than did control animals. There was, however, no significant difference between the groups at delays of up to 120 s. Zola-Morgan, Squire and Amaral (1989) reported only a transient impairment in naive monkeys following M B lesions. Their interpretation of the data is, however, questionable. Initially, M B lesioned animals took significantly longer to learn the D N M S task than controls. In fact, they learned it at a rate equivalent to monkeys with H+ lesions. When tested at delays, M B lesioned monkeys were impaired at 15 s, 60 s, and 10 min, relative to controls. In contrast, the H+ lesioned monkeys were impaired only at delays of 10 min. These results appear to indicate that M B lesions are sufficent to produce a deficit on D N M S and that this deficit may be more severe than that produced by an H+ lesion. This was not the conclusion reached by Zola-Morgan et al. for the following reasons. When the monkeys were retested on D N M S 18-26 months after the initial testing the H+ group took more trials to reattain criterion than did the control or M B groups (H+ = 106, N = 73, M B = 30). However, whether these differences are statistically significant was not addressed. The fact that there was one monkey in the H+ group with an extreme score and that the rest performed at levels comparable to the normal and M B monkeys makes it unlikely. When the animals were tested at delays of 15 and 30 s no difference between control, H+ or M B groups was found. It should be noted that the 60 s and 10 min delays were not administered at the second test. It was in these delays that the greatest difference was observed in the H+ and M B groups during the initial testing. It is 44 impossible to know what the monkeys performance at these delays would have been at a second testing. These results would indicate that neither the M B group nor the H+ group had a long-lasting impairment. Complicating this interpretation, however, is the fact that the H+ group was tested at delays of 15 and 30 s again at 30 and 56 months following surgery. Their performance at these delays at this time was equal to their initial performance. This led Zola-Morgan et al. to conclude that the H+ monkeys had a long lasting deficit whereas the M B monkeys had only a transient deficit. However, the M B animals were not tested at 30 or 50 months after initial testing and it is likely they also could have retained an impairment. Accordingly, stating that M B lesions have no effect on D N M S performance under such conditions appears rather suspect. In humans, monkeys, and rats there are striking similarities between the mnemonic deficits produced by lesions of the diencephalon and the M T L . There are at least five points of convergence: 1) Humans, monkeys, and rats with either M T L or diencephalic lesions demonstrate profound anterograde and, to a lesser extent, retrograde amnesia (Aggleton & Mishkin, 1983; Mumby et al., 1993; Squire, 1992; Victor, 1994), 2) Humans and monkeys with either M T L or diencephalic lesions show preserved perceptuo-motor learning (Aggleton & Mishkin, 1983; Zola- Morgan & Squire, 1985; Squire, 1992), 3) Humans with M T L or diencephalic amnesia display intact priming abilities (Gabrieli & Keane, 1990; Schacter et al., 1988), 4) Humans, monkeys, and rats with either M T L or diencephalic lesions are impaired on the D N M S paradigm (Mumby et al., 1990; Squire et al., 1988; Zola-Morgan & Squire, 1985), 5) Humans with M T L or diencephalic lesions perform poorly on other tests that are sensitive to similar damage in monkeys and rats (Mumby et al., 1995; Squire, 1992; Squire et al., 1988). Current research has begun to focus on the subtle, albeit possibly revealing, differences in the mnemonic deficits that may distinguish diencepahlic from M T L amnesia. In humans, rate of forgetting, the severity and extent of retrograde amnesia, and tendency to confabulate in order to cover up a memory deficit have all been suggested as possible criteria on which to make a differentiation (Parkin, 1984). 4 5 Mnemonic differences between M T L and diencephalic amnesia are also being explored with animal models. Using a battery of object-recognition behavioral tests (object discrimination, discrimination reversal, eight pair concurrent object discrimination, D N M S , D N M S with object lists, and temporal order discrimination), impaired and preserved mnemonic abilities are being established for various brain lesions in rats. These "memory profiles" can then be compared and any differences observed would indicate qualitative differences in the amnesia associated with the lesion. Mumby et al. (1995) tested naive rats with either hippocampal or amygdala lesions in this battery of tests. Relative to controls, rats with hippocampal lesions were impaired on object discrimination and concurrent object discrimination tasks. They also took longer to learn the D N M S task than did controls and, once they had acquired the rule, they were impaired at delays of 120 s. Rats with lesions of the amygdala were impaired on object discrimination but performed normally on concurrent object discrimination. They required more trials to learn the D N M S task than did control rats, but fewer trials than rats with hippocampal lesions. Once they had mastered the D N M S task, however, they performed normally at all delays. Neither hippocampectomy nor amygdalaectomy impaired performance on discrimination reversal, D N M S with lists, or temporal order discrimination. In comparison, rats with thalamic lesions were unimpaired on object discrimination but took significantly more trials to learn the discrimination reversal and concurrent object discrimination. These rats learned the D N M S task at a rate equivalent to controls but were impaired at delays greater than 15 s. Additionally, they were impaired on D N M S learning with lists but performend normally on the temporal order discrimination task (Kornecook & Pinel, 1994). The differences in the pattern of memory impairments following lesions to the amygdala, hippocampus, or thalamus suggests that there may be a qualitative difference in the mnemonic deficits produce by lesions to the M T L or the diencephalon. However, a memory profile for rats with cortical lesions (i.e., entorhinal/ perirhinal) has not been determined. As it is now believed that this damage contributes substantially to M T L amnesia, the critical comparison to make would be between rats with rhinal cortex or diencephalic lesions. 4 6 BASAL FOREBRAIN The basal forebrain consists of the septal nuclei, the nuclei of the diagonal band of Broca (DBB), and the nucleus basalis of Meynert (NBM). It provides the major source of cholinergic input to the neocortex and the hippocampus (Fibiger, 1991). Basal forebrain fibers project to the hippocampus in four bundles, collectively referred to as the septohippocampal pathway. The two most prominent of these are the fimbria, which travels along the lateral edge of the hippocampus; and the fornix which travels along the medial aspect. These two fiber paths converge near the nuclei of the basal forebrain where they are termed the fimbria/fornix (Dutar, Bassant, Senut, & Lamour, 1995). The fimbria and the fornix carry both efferent fibers from hippocampal formation to the basal forebrain and afferent fibers from the basal forebrain to the hippocampus. Thus, at all points along these fiber bundles, there will be axons traveling in both directions (Amaral & Witter, 1995). Septal afferent fibers terminate in essentially all fields of the hippocampal region, the most prominent connection being with the dentate gyrus. These projections have a loose topographic organization. Fibers terminating in the septal hippocampus originate primarily from the D B B whereas fibers terminating in the temporal hippocampus originate from both the D B B and the septal nuclei. Temporal lobe afferents to the basal forebrain arise from the entorhinal cortex, the CA1, the C A 3 , and the subicular complex (Amaral & Witter, 1994; Dutar etal., 1995). In humans, large basal forebrain lesions as a result of penetrating war injuries (Salazar, Grafman, Schlesselman, Vance, Mohr, Carpenter, Pevsner, Ludlow, & Weingartner, 1986), ruptures of the anterior communicating artery (Damasio, Graff-Radford, Eslinger, Damasio, & Kassell, 1985a), or herpes encephalitis (Damasio, Eslinger, Damasio, Van Hoesen, & Cornell, 1985b) have been found to produce significant amnesic symptoms, most notably severe anterograde and retrograde memory impairments. Deficits in temporal relations, motivation, and a variety of other non-mnemonic cognitive impairments may also be present (Damasio et al., 1985a, 1985b; Salazar et al, 1985). Attempts to correlate the memory deficits with damage to specific nuclear structures have been mixed, though at least one study emphasized the importance of 4 7 lesions to the D B B (Morris, Bowers, Chatterjee, & Heilman, 1992). Amnesia following basal forebrain damage may be the result of the depletion of cholinergic innervation of the neocortex (Damasio et al., 1985). Cholinergic forebrain systems have long been implicated in memory through pharmacological manipulations (Drachman & Leavit, 1974; Aigner & Mishkin, 1986) and their degeneration may be related to the memory deficits associated with Alzheimer's disease (Whitehouse, Price, Struble, Clark, Coyle, & Delong, 1982). In monkeys, basal forebrain lesions have been shown to produce deficits in a variety of memory tasks (Irle & Markowitsch, 1987; Ridley, Baker, Drewett & Johnson, 1985), including D N M S . Irle and Markowitsch (1987) found that lesions of the N B M impaired both the acquisition of the D N M S task and performance at delays of up to 120 s. Histological analysis revealed that the actual lesions were larger than intended and included some of the D B B , amygdala, globus pallidus, and claustrum. Aigner, Mitchell, Aggleton, DeLong, Struble, Price, Wenk, Pettigrew, and Mishkin (1991), however, found only a transient impairment in D N M S following large basal forebrain lesions in macaques. Pretrained monkeys received either a N B M lesion, a D B B lesion or both lesions combined. Only monkeys with the combined lesions were impaired on the D N M S task (both reacquisition and performance at delays) whereas the N B M and D B B lesioned animals performed similar to controls. After extensive practice this group was, however, able to reach control levels of performance. Histological analysis revealed a depletion in cortical cholinergic markers in the monkeys that displayed an initial memory impairment. In another study, a complete lack of impairment on D N M S and several other tests of object-recognition memory was found following ibotenic acid lesions of the septal nuclei, N B M , and D B B in combination (Voytko, Olton, Richardson, Gorman, Tobin, & Price, 1994). Moreover, cholinergic function was shown to be markedly reduced as a result of the lesion: Findings such as this add to a growing body of evidence suggesting that damage to basal forebrain cholinergic system alone might not be sufficient to produce a memory impairment (Dunnet, Everitt, & Robbins, 1991). Although basal forebrain lesions in rats have been shown to impair visual discrimination, nonmatching-to-position, working memory, and spatial memory (Dunnett, 1985; Everitt, Robbins, Evenden, Marston, Jones, & Sirkia, 1987; Knowlton, Wenk, Olton, & Coyle, 1985; McAlonan, Dawson, Wilkinson, Robbins, Everitt, 1995), their effect on D N M S performance is just begining to be investigated. Kornecook, Kippin, Pinel and Wood (1993) found that lesions of the septal nucleus and the D B B in pretrained rats impaired the postsurgical reacquisition and performance of D N M S at delays of up to 120 s. The exact nature of these deficits is unclear. The fact that rats with fornix lesions perform normally on D N M S (Shaw & Aggleton, 1993) suggest that these deficits are not due to cholinergic depletion. The authors did suggest that motivational and/or perceptual impairments may have contributed to the impaired performance. PREFRONTAL CORTEX The prefrontal cortex (PFC) is a heterogeneous brain region that can be grossly subdivided into medial and lateral aspects. Based upon cytoarchitectonic differences, the medial P F C can be further subdivided into the medial precentral, the dorsal anterior cingulate, and prelimbic areas. The lateral P F C can subdivided into the dorsal and the ventral agranular insular cortices (Van Eden & Uylings, 1985). The P F C receives prominent projections from the hippocampus, both directly from the C A 1 subfield as well as via the subicular complex (Amaral & Witter, 1995). The CA1-prefrontal projections arise from the temporal pole of the hippocampus and terminate primarily in the caudal prelimbic area (Jay, Glowinski, & Thierry, 1989; Van Groen & Wyss, 1990) and the adjacent ventral infralimbic area (Swanson, 1981). Projections to other subdivisions of the P F C and projections from the more septal levels of the CA1 have not been found. Importantly, the P F C receives substantial projections from the thalamus. In fact, the P F C was originally defined as the cortical projection area of the M D nucleus of the thalamus (Rose & Woolsey, 1948). Using this criterion, a P F C can be identified in all mammal brains. The diversity of these thalamic connections, however, varies considerably from species to species (Kolb, 1990). The close anatomical relationship between the P F C and the thalamus may have important implications for the interpretation of behavioral deficits that result from lesions to these areas. The precise role of the P F C in mnemonic functions has yet to be determined. In 49 humans, restricted frontal lobe lesions do not generally produce a full-blown amnesic syndrome. However, amnesics with frontal lobe pathology show a pattern of memory performance that is qualitatively different than amnesics without frontal lobe pathology. For example, spatiotemporal memory appears to be particularly effected in amnesics with concommitent frontal lobe pathology (Schacter, 1987). Deficits in working memory, habituation, and some types of associative learning have also been reported (Corkin, 1965; Luria & Homskaya, 1964; Petrides, 1985). In monkeys, the P F C may be preferentially involved in working memory and/or rule learning (Goldman & Rosvold, 1970; Goldman, Rosvold, Vest, & Galkin, 1971; Goldman-Rakic, 1995). Although the rule learning component and the temporal demands of the D N M S task suggest that its performance would be susceptible to P F C lesions, experimental findings have been somewhat inconsistent. Bachevalier and Mishkin (1986) found that lesions to the ventromedial P F C in pretrained monkeys produced a severe reacquisition deficit and a performance deficit at delays of up to 120 s. Lesions of the dorsolateral P F C had no effect on either aspect of D N M S performance. It should be cautioned, however, that animals in the ventromedial P F C group had significant cell loss in the thalamus that could have contributed to the D N M S deficit. In fact, D N M S reacquisition deficits have been found in monkeys with thalamic lesions (Aggleton & Miskin, 1981; 1983). In rats, Shaw and Aggleton (1993) found that P F C lesions had no effect on the acquisition or performance of a D N M S task. The size of the P F C lesion did not appear to be a mediating factor, as both rats with small or large lesions of the P F C performed similar to controls. These lesions had no effect even when the pool of stimulus items was reduced to make the task more difficult. Kolb, Burhman, MacDonald, and Sutherland (1994) tested rats on both the Aggleton (1985) and Mumby et al. (1990) versions of the rat D N M S task. No deficits in D N M S acquisition were found on either version of the task and a deficit at the 20 s delay was found only on the Aggleton version. On the contrary, Otto and Eichenbaum (1992) found that rats with orbital prefrontal cortex lesions were impaired in an odor guided version of D N M S . A dissociation between the behavioral effects P F C and 50 rhinal cortex lesions was also found. Rats with PFC lesions were impaired in D N M S acquisition, but performed normally at delays; rats with rhinal cortex lesions displayed the opposite pattern. This suggests that the P F C may participate selectively in the acquisition of the D N M S rule, whereas the rhinal cortex may be primarily involved in the maintenance of stimulus representations. I S C H E M I A - I N D U C E D A M N E S I A A general pattern has emerged from animal studies as to which areas of the M T L , when damaged, are sufficient to produce amnesia. In the past 10 years, a gradual shift has been occurring, a shift away from the importance of hippocampal damage in producing amnesia and more toward the importance of damage to areas around the rhinal sulcus (Mishkin and Murray, 1994; Murray, 1996). Such a conclusion, however, must be tempered by the findings that cerebral ischemia in humans, monkeys and rats can cause severe amnesia. In some cases, the extent of the memory impairment is substantially greater than that produced by total bilateral hippocampal ablations. Moreover, it has been argued that ischemic insult produces damage restricted to the hippocampus, primarily the C A 1 subfield, with little or no pathology occurring elsewhere in the brain. This, then, is the central paradox that will be addressed in this thesis. How can partial damage to the hippocampus produce as profound or a more profound deficit than total hippocampal ablation; particularly when lesion studies have demonstrated that the hippocampus plays little, if any, role in object-recognition memory? This section will describe the human clinical studies and the animal research that has delineated this paradox. Neuropathologies mechanisms of ischemia-induced amnesia will also be covered. I S C H E M I A - I N D U C E D A M N E S I A IN H U M A N S "Ischemia" is defined as a situation of severely reduced or completely blocked blood flow (Brierly & Graham, 1984; Siesjo, 1978). In contrast, hypoxia denotes a reduction of oxygen content in arterial blood (Schmidt-Kastner & Freund, 1991). In humans, global cerebral ischemia can be produced by attempted hanging (Berlyne & Strachman, 1968), carbon monoxide poisoning (Muramato, Kuru, Sugishita, & Toyokura, 1979), and, most commonly, cardiopulmonary arrest (Mcneill, Tidmarsh, & Rostal, 51 1965). A s a consequence of an ischemic episode, survivors often display severe and long lasting memory deficits (Adams, Brierly, Connor, & Triep, 1966; Brierly & Graham, 1984; Volpe, Herscovitch, Raichle, Gazzaniga, & Hirst, 1983, Volpe & Hirst, 1983; Volpe, Holtzman, & Hirst, 1986). Attempts to correlate ischemia-induced brain damage with mnemonic deficits have met with mixed results. In the majority of the published cases, reports of impaired memory function were largely anecdotal and/or the underlying neuropathology received only a cursory examination. In one case, however, that of patient R.B., extensive neuropsychological testing and a thorough post mortem histological analysis is available (Zola-Morgan, Squire, & Amaral, 1986). This case has provided a unique opportunity to examine the relationship between cognitive function and ischemia-induced neuropathology in humans. THE CASE OF PATIENT R.B. Patient R.B. suffered an ischemic episode following coronary bypass surgery in 1978. Approximately 6 months after his surgery, he was given a battery of standard neuropsychological test as part of his follow up evaluation. At this time, his full scale WAIS score was 111 (verbal 108 performance 114) indicating above average intelligence. However, his Wechsler Memory Scale (WMS) score was 91, 20 points lower than his full scale WAIS, indicative of a severe memory impairment. Over the next 4 years R.B.'s performance on a variety of tests useful in characterizing amnesic patients was examined. These tests evaluated: 1) the ability to acquire new information (to assess anterograde amnesia); 2) the ability to recall events that had occurred prior to the ischemic episode (to assess retrograde amnesia); and 3) the ability to perform cognitive tasks that do not depend heavily on memory. R.B. was impaired on three tests of new learning ability (paired associates, story recall, and diagram recall) at 6 , 9, and 23 months following the onset of his amnesia. R.B. was also impaired on tests of free-recall and recognition memory in the 4 years following his ischemic episode. These findings indicate that R B. had severe anterograde amnesia for both verbal and nonverbal material and that this impairment was relatively persistent over time. 52 In contrast to his severe impairment in acquiring new information, R.B. demonstrated little, if any, impairment on tests designed to assess retrograde memory deficits. These tests included recall and recognition of 88 news events that had occurred in one of the 4 decades from 1940-1970, the Boston famous faces test, which requires the identification of photographs of people who had come into prominence at various times between 1930 and 1979, multiple choice recognition tests for television programs that were broadcast for a single season, and a test of autobiographical memory. On each of these tests, R.B. performed normally, with the exception of a mild deficit for news items and television programs in the 3 years preceding his ischemia. Although R.B. did display severe anterograde amnesia, he performed normally on other tests of cognitive ability. The Boston naming test, Aphasia screening test, the Parietal lobe test battery, and the Apraxia test were also administered to R.B. The only deficit found was in a subtest of the Parietal lobe test battery, the tactile-visual finger test. R.B. was unable to say which fingers on his left hand were being touched by the examiner when his hand was hidden from view. Neuropathology of R . B . Upon his death in 1983, R.B.'s brain underwent a detailed histological analysis in an attempt to correlate the locus of brain pathology with his amnesic symptoms. The most prominent insult was a bilateral lesion involving the entire mediolateral and rostrocaudal extent of the C A 1 subfield of the hippocampus. Throughout most of the hippocampus the lesion was confined to the CA1 subfield, except for some patchy cell loss in the C A 3 and subiculum in the caudal one third of the hippocampus. Other temporal lobe structures appeared relatively normal. There was no observable cell loss or anatomical changes in dentate gyrus, or entorhinal cortex. A l l nuclei of the amygdala were present and appeared to contain the appropriate density of neurons except for a small, bilateral area of cell loss at the junction of the amygdala and the substantia innominata. Outside of the temporal lobe, minor pathologies were found in the right postcentral gyrus, the left globus pallidus, and the left intermedullary lamina of the thalamus. Additionally, small patches of cell loss were observed in the left internal capsule, the neocortex, and the cerebellum. 53 The pattern of brain damage observed in R.B. is consistent with that observed in other patients following cerebral ischemia (Adams et al., 1966; Kartsounis, Rudge, & Stevens, 1995). The majority of observable cell loss occurs within the hippocampus, primarily within the CA1 subfield, with little detectable damage elsewhere. This is most likely due to the selective vulnerability of hippocampal CA1 pyramidal cells to ischemic insult. This selective vulnerability to was first recognized by Sommer (1880) and is now thought to be caused by a massive release of excitatory amino acids during ischemia. Selective vulnerability will be covered in greater detail in the section on mechanisms of ischemic brain damage. A N I M A L M O D E L S O F C E R E B R A L ISCHEMIA. With the apparent discovery that CA1 lesions could cause amnesia, several attempts were made to produce an animal model of the neuropathology and mnemonic deficits exhibited by R.B. and other ischemic patients. The following section describes the results obtained in monkeys and rats. Careful attention should be paid to the studies which used a version of the D N M S task to assess memory. MONKEYS Bachevalier and Mishkin (1989) permanently occluded a branch of the posterior cerebral artery (PCA) in six Rhesus monkeys that had previously received D N M S training. Although this procedure is not directly analogous to global ischemia, it does render ischemic the tissue that was found to be most susceptible to damage in R.B. (i.e., the hippocampal region). Three of the six monkeys had bilateral lesions restricted primarily to the C A 1 subfield of the hippocampus. Extrahippocampal damage was minimal and inconsistent. Small infarcts were found in the dentate gyrus, subiculum, thalamic nuclei, and entorhinal cortex. When tested postoperatively on D N M S , the performance of these three monkeys were significantly worse than that of monkeys that had received an H+ lesion. The neuropathology of the remaining three monkeys was quite variable and their scores on D N M S were not included in the final statistical analysis. One monkey had virtually no detectable brain damage and performed at control levels. Another animal had a unilateral lesion to the CA1 and was mildly impaired on D N M S . It is the remaining 54 monkey, however, that is most revealing. This animal had large bilateral lesions of the entire hippocampus combined with damage to some of the surrounding cortical areas. Interestingly, this animal was impaired on D N M S but this impairment was milder than the impairment displayed by the three monkeys with circumscribed CA1 lesions. These results suggest that lesions of the CA1 produced by ischemia may be more detrimental to object-recognition memory than large lesions of the hippocampus. Monkeys with brain damage as a result of reversible ischemia have also been shown to have significant memory deficits. Zola-Morgan, Squire, Rempel, Clower, and Amaral (1992) produced a 15 min period of ischemia in monkeys by coupling carotid artery occlusion with pharmacologically-induced hypotension. Following recovery, these animals learned the D N M S task at a rate equivalent to controls and significantly faster than animals with H+ or H+A+ lesions. However, when tested at delays of 15, 60, or 600 s the ischemic monkeys were significantly impaired. They performed worse than control monkeys, but at a rate similar to H+monkeys and better than H+A+ monkeys. Interestingly, when compared to monkeys that received an H lesion (Alvarez et al., 1995) paradoxical results similar to those reported by Bachevalier and Miskin are found. Although both groups of animals were unimpaired at learning the D N M S task, the ischemic lesioned animals performed worse than the H monkeys at the longer delays. Neuropathological analysis of these animals revealed damage restricted primarily to the C A l region of the hippocampus with some slight cell loss in the C A 2 region, the hilus of the dentate gyrus, and cerebellum. RATS Studies utilizing rat models of cerebral ischemia have also demonstrated memory impairments, some of which mirror the deficits seen in monkeys. There are two common methods for inducing cerebral ischemia in rats. One procedure, the four vessel occlusion (4VO) method (Pulsinelli & Brierly, 1979), is performed in two stages. In the initial phase, the vertebral arteries are permanently occluded by cauterization and two atraumatic arterial clamps are surgically implanted around the common carotid arteries. After allowing 24 hr for recovery, ischemia is then induced in unanaesthetized animals by tightening the 5 5 clamps for periods up to 30 min. In this model, the most severe brain damage occurs in the C A 1 subfield of the hippocampus and the dorsolateral striatum. Additional, but less extensive, damage occurs in neocortical layers 3, 5, 6, the cerebellum, and the thalamus as the duration of ischemia increases. The other rat model of cerebral ischemia is the two vessel occlusion (2VO) method (Mudrick, Leung, Baimbridge, & Miller, 1988; Smith, Bendek, Dahlgren, Rosen, Weiloch, & Siesjo, 1984). In this model, ischemia is induced by first lowering the rats mean arterial blood pressure by hemorrhage and then temporarily clamping off the common carotid arteries for periods of up to 20 min. Observable cell loss is typically confined to the C A 1 subfield of the hippocampus, with some patchy loss occurring in the hilus of the dentate gyrus. Only a few studies have examined the effect of cerebral ischemia on object-recognition memory in rats. Wood, Mumby, Pinel, and Phillips (1993) found that ischemic rats that received no preoperative training took significantly longer to learn the D N M S task and were impaired across delays of 4, 15, 30, 60, 120 and 300 s, relative to controls. Moreover, rats that had received presurgery D N M S training took significantly longer to reattain preoperative levels of performance and were impaired at all delays, relative to controls. These findings have since been replicated (Mumby et al., 1992b). Importantly, ischemic rats were more impaired than rats that had received bilateral aspiration lesions of the hippocampus (Mumby et al., 1992a; Mumby et al., 1992b), a paradox which is identical to that found when comparing the performance of ischemic versus hippocampectomized monkeys (Alvarez et al., 1995; Bachevalier & Mishkin, 1989; Zola-Morgan etal., 1992). The vast majority of studies employing rat models of cerebral ischemia have used performance on spatial tasks as their index of learning. Several factors should be considered when trying to glean any useful information from these studies. First, it is important to realize that these tasks are not directly comparable to nonspatial object-recognition tasks. The mnemonic strategies required to solve each task may be quite different and they may be subserved by different neural structures. A s previously 56 mentioned, the rat hippocampus may be disproportionately involved in spatial memory (O'Keefe & Nadel, 1978; Nadel, 1991). Therefore, any comparison between the findings with rodents on spatial tasks and primates on nonspatial tasks is tenuous at best. Second, despite the fact that large lesions of the hippocampus in rats severely disrupts their spatial memory (Morris, Garrud, Rawlins, & O'Keefe, 1982; Jarrard, 1993), the findings regarding the relationship between ischemia-induced hippocampal lesions and spatial memory deficits do not fall into a readily interpretable pattern. The type of spatial paradigm used (eight-arm radial-maze, Morris water-maze) in combination with the method used to induce ischemia (2VO or 4VO) yield quite discrepant results. With this in mind, the following is intended as an overview of the types of deficits seen when using rat models of ischemia in conjunction with tasks designed to assess spatial memory. Two measures of performance on the eight-arm radial-maze are typically used when assessing the behavioral effects of ischemia in rats; reference memory and working memory. The reference memory aspect of the task requires that rats learn that only five of the eight arms have a food reward and that these five arms remain constant over all trials. The working memory component of the task requires that the rats remember which arms have been visited during a given trial and that once an arm is visited and the reward obtained that the arm should not be revisited again on that trial. The mnemonic demands of this task may be similar to those of D N M S because both tasks require the animals to utilize trial unique information. Ischemia induced by 4 V O has been shown to produce deficits in the working memory component and, to a lesser extent, the reference memory component of this task (Davis, Baranowski, Pulsinelli, & Volpe, 1986; Davis, Tribuna, Pusinelli, & Volpe, 1986; Grotta, Pettigrew, Rosenbaum, Reid, Rhoades, & McCandless, 1988; Kiyota, Miyamato, & Nagoaka, 1991;Volpe, Pusinelli, Tribuna, & Davis, 1984). Whether these deficits can be attributed solely to hippocampal damage is open for debate. Additional damage was typically detected in the cortex and striatum of these animals and reference memory type tasks have been shown to be sensitive to striatal lesions (Packard, Hish, & White, 1989). Attempts to correlate performance on the radial arm maze with CA1 damage have produced puzzling results. For example, Kiyota et al. (1991) found an inverse 57 correlation between the number of CA1 pyramidal cells and and the number of working and reference memory errors following 4VO-ischemia. Moreover, 2VO-ischemia produces no deficits in reference or working memory performance in the radial arm-maze, despite significant CA1 cell loss. (Wood, personal communication, 1992). Two versions of the Morris water-maze (Morris et al., 1982) have also been used to characterize ischemia-induced memory deficits. In one version, a simple place task, rats are required to learn the position of a hidden platform whose position remains constant throughout training. Latency to find the platform and the distance swam (path length) are used as dependent measures. Following training, a probe trial is run during which time the platform is removed. The amount of time the rat spends searching for it in its previous location is used as a further index of memory. In the learning set task (Whishaw, 1985), rats must learn a sequence of six different platform locations. Time spent searching for the platform in its previous location and path length to the platform are used as measures of learning. Rats subjected to 5-20 minutes of 4 V O are impaired in the acquisition of the place task in some studies (Block, Jasper, Heim, & Sontag, 1990; Hagan & Beaughard, 1990; Jaspers, Block, Heim, &Sontag, 1990; Nunn, LePeillet, Netto, Hodges, Gray, & Meldrum, 1994; Olsen, Scheel-Kruger, Moller, & Jensen, 1994a; 1994b) but not in others (Kiyota, Miyamoto, & Nagaoka, 1991). Discrepant results have also been found when the 2 V O model is used. Only mild and transient deficits in the acquisition of the place task were observed in rats with up to 6 0 % C A 1 cell loss (Auer, Jensen, & Wishaw, 1989; V o l l , Wishaw, & Auer, 1989). However, Jaspers et al. (1990) found an acquisiton impairment in place learning following 2VO-ischemia in the absence of any detectable CA1 cell loss. This deficit was attributed to functional changes in neurons following cerebral ischemia. These discrepant findings lead one to question just what exactly is the neural basis of spatial deficits in the water-maze following ischemia. Several attempts have been made to correlate performance on the water-maze with CA1 cell loss. Nunn et al. (1994) exposed rats to varying durations of 4VO-ischemia, 5, 10, 15, or 30 min. The amount of C A 1 cell loss increased in a linear manner with increasing durations of ischemia, but no 58 correlation between cell loss and water-maze deficits was found. These findings suggest that deficits on the water-maze cannot be solely attributable to CA1 cell loss. Failures to correlate CA1 damage with water-maze performance have also been reported by Olsen, Scheel-Kruger, Moller, and Jensen (1994a; 1994b). In one study in which a correlation was reported (Rod, Whishaw, & Auer, 1990), sham-operated animals with no brain damage were included in the correlational analysis. The inclusion of a group with a complete absence of variation in one variable is questionable, and makes their conclusion that CA1 cell loss is significantly correlated with performance in the water-maze less than convincing. A true correlation between CA1 cell loss and performance should reveal itself without the inclusion of a group with no CA1 damage (Nunn et al., 1994). M E C H A N I S M S O F I S C H E M I C B R A I N D A M A G E The brain is very sensitive to ischemic insult because it has high metabolic demands and low oxygen stores (Siesjo, 1978). Ischemia-induced brain damage results from the interaction of several complex pathophysiological events. The hippocampus has been the classic site for the study of these events due to its exquisite sensitivity to ischemia. Pyramidal cells in the CA1 region are the most vulnerable to ischemic injury of any cells in the entire forebrain (Schmidt-Kastner & Freund, 1991). Using the hippocampus as a model, the following section will provide an outline of the putative mechanisms of ischemia-induced brain damage, focusing primarily on the excitotoxic hypothesis of ischemia induced cellular necrosis (Rothman & Olney, 1986). Similar neuropathological processes may produce damage in other brain regions or, conversely, mechanisms of ischemic damage may be unique to individual brain sites (Kuroiwa & Okeda, 1994). EXCITOTOXICITY The transport of ions across the cell membrane is necessary to maintain ionic homeostasis. Neurons require oxygen to run the adenosine triphospate (ATP)-dependent sodium-potassium pump that contributes to the resting membrane potential by exchanging intracellular sodium ions for extracellular potassium ions. Depletion of A T P as a result of ischemia causes the sodium-potassium pump to shut down, thereby increasing intracellular sodium concentrations. The subsequent membrane depolarization results in the opening of 59 voltage-sensitive sodium channels, allowing more sodium to enter the cell and begin a positive-feedback loop. Membrane depolarization gives rise to an action potential which causes the the opening of voltage-sensitive calcium channels (VSCC's). Membrane depolarization is prolonged under ischemic conditions, leading to a massive increase in intracellular calcium levels. High concentrations of calcium have been shown to be neurotoxic in vitro (Kuroiwa & Okeda, 1994) and it is generally believed that an increase in intracellular calcium is the major precipitating event leading to irreversible ischemic brain damage (Siesjo, 1981, Raichle, 1983). The exact mechanism by which elevated calcium levels cause cell death are not fully understood but may involve the breakdown of membrane phospholipids (Siesjo, 1990), the production of free radicals (Schmidley, 1990), and the degradation of cytoskeletal proteins (Siesjo, 1990; Siesjo & Wieloch, 1985). Disruptions of calcium homeostasis alone, however, cannot solely account for ischemia-induced neuropathology, as all cells that are rendered ischemic have disruptions in calcium homeostasis but only a few are selectively vulnerable (Kuroiwa & Okeda, 1994; Seisjo, Memezawa, & Smith, 1991). Moreover, CA1 pyramidal cells display "delayed neuronal death" following ischemia. Markers that indicate neuronal injury do not appear until approximately 24 hr following ischemia (Kirino, 1982) and the majority of cell death occurs between 1 and 5 days postischemia (Schmidt-Kastner & Freund, 1991). Additional factors or mechanisms must be present in selectively vulnerable brain regions to account for these discrepancies. Research into these areas has suggested that the distribution of excitatory amino acids (EEA's) may be the key to understanding this selective necrosis and has led to the formulation of the excitotoxic hypothesis of ischemia-induced cell death (Rothman & Olney, 1986). Elevated levels of intracellular calcium also play a central role in this hypothesis but it is through glutamate receptors, not VSSC's, that this deleterious increase in calcium is thought to occur. The most prominent mechanism of glutamate-mediated calcium accumulation is believed to be through the NMDA-receptor/channel complex (Choi, 1988). Membrane depolarization as a result of ischemia produces a massive release of glutamate, 60 which causes the NMDA-channel to open, leading to an influx of calcium in the post-synaptic cell. The excitotoxic theory of ischemic neuronal damage is suppported by a variety of pharmacological, anatomical, and neurochemical evidence. First, a number of glutamate receptor antagonists, particularly those effective at blocking the NMDA-receptor, have been shown to reduce ischemia-induced CA1 cell loss (Rod & Auer, 1989; Swan, Evans, & Meldrum, 1988; Swan & Meldrum, 1990). Glutamate receptor antagonists that work on metabotropic receptors have also been found to lessen CA1 cell loss (Sheardown, Neilsen, Hansen, Jacobsen, & Honore, 1990; Diemer, Johansen, & Jorgenson, 1990). In rats, some of the memory impairments that result from cerebral ischemia can also be attenuated with N M D A antagonists (Rod, Whishaw, & Auer, 1990). Additionally, application of high concentrations of glutamate agonists and other E A A ' s directly to the C A 1 have been shown to cause cell death (Olney, 1978). E A A ' s have recently become quite popular for making discrete brain lesions in animals, as they can be applied to a relatively restricted area and do not damage fibers of passage (Nunn & Hodges, 1994). A second, and critical, piece of evidence implicating excitotoxicity in C A 1 necrosis is provided by studies in which the entorhinal cortex was disconnected from the hippocampus. As previously mentioned, the entorhinal cortex provides the major source of excitatory input into the hippocampus, via the perforant path. This pathway is known to contain glutamate and aspartate (Schmidt-Kastner & Freund, 1991). Transections of the perforant path (Wieloch, Lindvall, Blomqvist, & Gage, 1985) or ablations of the entorhinal cortex (Kaplan, Lasner, Nadler, & Crain, 1989) prior to ischemia attenuates CA1 cell loss. Additionally, neurochemical studies have shown a substantial increase in glutamate and aspartate levels in the CA1 following ischemia (Benveniste, Drejer, Schousboe, & Diemer, 1984; Drejer, Benveniste, Diemer, & Schousboe, 1985). The postischemic glutamate concentrations in the hippocampus are, in and of themselves, high enough to be neurotoxic. Although the excitotoxic hypothesis cannot account for all aspects of ischemia-induced cell death, it has provided a substantial impetus for research into the underlying 61 mechanisms and possible neuroprotective agents. Other factors such as lactacidosis (Schurr & Rigor, 1992) and an increase in nitric oxide (Moncada, Lekieffre, Arvin, & Meldrum, 1992) have also been suggested as possible contributors to ischemic injury. While these and other additional factors remain to be fully explored, it is generally accepted that glutamate neurotoxicity plays at least some role in the selective vulnerability of CA1 neurons to ischemic insult (Nunn & Hodges, 1994). E X P L A I N I N G T H E P A R A D O X There is no question that cerebral ischemia in humans, monkeys, and rats can lead to severe and persistent memory deficits on a wide range of tasks. The central question is, are circumscribed C A 1 lesions the anatomical locus of these deficits? At least for D N M S performance (object-recognition memory), lesion studies with both rats and monkeys indicate that the hippocampus plays little, if any, role. Thus, a paradox exists: Why do lesions that appear limited to the CA1 subfield of the hippocampus produce a severe deficit in D N M S performance whereas hippocampal ablations produce only minor effects? Two theories have been posited as possible explanations for this conundrum. Bachevalier and Mishkin (1989) suggested that a partially damaged hippocampus could result in the functional disorganization of the remaining hippocampal circuitry. This in turn would have a disruptive effect on other structures that receive afferents from the hippocampus, some of which may participate in D N M S learning. Small incomplete lesions may be more detrimental than hippocampectomy because if the hippocampus was totally absent these other structures would not be receiving any aberrant, potentially disruptive, input. This "hippocampal interference" hypothesis has received some support from experiments with monkeys in which circumscribed hippocampal lesions were made using E A A ' s . In one study, 85% of the hippocampus was destroyed with ibotenic acid. These monkeys displayed only a mild impairment in D N M S learning (O'Boyle, et al., 1993). However, in similar study, in which only 25% of the hippocampus was destroyed, the monkeys demonstrated an impairment in learning the D N M S rule and in performnace at delays of 2 and 10 min (Beason-Held et al., 1993). These findings suggest that there is an inverse relationship between the amount of hippocampal damage and deficits on D N M S 62 performance, and that small lesions can produce more severe deficits than large lesions. As these studies are published only in abstract form, such a conclusion is tentative at best. The second explanation is that, in addition to the CA1 cell loss found following ischemia, neuronal changes not visible by conventional histology are occurring in other brain regions important for memory (Bachevalier & Mishkin, 1989). When investigating ischemic brain damage researchers have focused primarily on the hippocampus, most likely due to its unique cytoarchitecture and vulnerability to hypoxic injury. Cell loss within the C A 1 is the typical measure used to evaluate ischemic injury and therapeutic interventions in animal models. However, many less apparent changes in neuronal function and morphology may be taking place elsewhere in the brain, some of which may contribute to amnesia. T o observe these changes it would require the use of sophisticated histological, immunohistochemical, or molecular techniques, techniques that may be beyond the scope of laboratories, such as ours, whose main focus is behavioral studies. These changes would therefore go unnoticed in the majority of studies trying to link ischemia-induced brain damage with memory deficits. T o support this, Bachevalier and Mishkin (1989) cite research that has demonstrated that ischemia, in addition to producing C A 1 cell loss, can cause long lasting changes in cerebral blood flow (Volpe, Herscovitch, Raichle, Hirst, & Gazzaniga, 1983) and in the responsivness of neocortical neurons to peripheral stimulation (Dietrich, Ginsberg, & Bustro, 1986). Behavioral studies have directly and indirectly addressed each of the two proposed hypotheses. Volpe, Davis,Towle, and Dunlap (1992) compared the effects of ischemia-induced CA1 cell loss with CA1 cell loss produced by ibotenic acid on the ability of rats to perform in a split T-maze. They reasoned that if extrahippocampal damage was contributing to the memory deficits in ischemic rats then rats that had the C A 1 lesioned by ibotenic acid (and presumably sustained no damage elsewhere) should perform differently. It was found that both groups of animals were similarly impaired on the T-maze and that their memory deficits were positively correlated with amount of C A 1 cell loss. The authors conluded that the deficits of ischemic rats in T-maze performance can be reasonably attributed to damage of the CA1. 63 Using the rat version of the D N M S task, Mumby et al. (1992b) designed a critical experiment to test both hypotheses simultaneously. Rats were trained on the D N M S task and then received either 20 min of transient forebrain ischemia or bilateral aspiration lesions of the hippocampus. A s had been reported previously, the ischemic rats were severely impaired postoperatively, whereas hippocampectomized rats displayed only mild impairments at long delays. The rats then received a second surgical manipulation. The rats that had been hippocampectomized were subjected to 2VO-ischemia, and ischemic rats received bilateral hippocampal aspiration lesions. It was reasoned that if additional extrahippocampal damage was responsible for the memory impairments in ischemic rats then the hippocampectomized rats that underwent ischemia should then perform more poorly on the task when retested. Additionally, if the "hippocampal interference" hypothesis is correct, then ischemic rats who have their damaged hippocampi removed should improve their performance on the D N M S task. It was found that the performance of hippocampectomized rats did not deteriorate following ischemia nor did the performance of ischemic rats improve as a result of hippocampectomy. These findings led Mumby et al. (1992b) to conclude that ischemia-induced extrahippocampal damage is responsible for the object-recognition deficits, but that this damage is mediated or produced by the post-ischemic hippocampus. R A T I O N A L E F O R E X P E R I M E N T S A N D H Y P O T H E S E S Numerous studies with monkeys and rats in which circumscribed lesions of the hippocampus were produced have found, at best, only mild impairments on D N M S . Contrary to the mild deficits resulting from hippocampal ablations, lesions to the adjacent rhinal cortex have been shown to produce profound deficits on object-recognition memory. Several recent reviews (Jarrard, 1993; Mishkin & Murray, 1995; Murray, 1996; Nadel, 1992) are consistent with the idea that hippocampal lesions alone are not sufficient to produce the temporal lobe amnesic syndrome in monkeys or rats and that damage to the rhinal cortical areas is more likely the critical determinant. It is, however, difficult to reconcile the preceding conclusions with the evidence from ischemia-induced hippocampal lesions. Ischemic insult produces a brain lesion that, 64 by conventional histological methods, appears to be limited primarily to the CA1 subfield of the hippocampus. Ischemic insult results in profound deficits in object-recognition memory. These deficits are typically much greater than those produced by lesions of the hippocampus alone and in some cases are greater than those produced by an H+ lesion. Initially, Bachevalier and Mishkin (1989) proposed two possible theories to explain this paradox. One idea was that a partially damaged, malfunctioning hippocampus was more detrimental to memory than total ablation. The partially damaged hippocampus may be causing aberrant activity in afferent structures important for memory. The fact that ablating the hippocampus in rats that had been previously subjected to ischemia did not attenuate their memory deficit does not support this hypothesis (Mumby et al.,1992b). If the partially damaged hippocampus was interfering with afferent structures responsible for memory it seems logical that removing it would eliminate these processes and thereby improve the D N M S performance of the ischemic animals. This did not occur. The second hypothesis proposed by Bachevalier & Mishkin (1989) was that ischemia may be producing extrahippocampal damage that is not detectable with standard histological techniques and that it is this undetected damage that is responsible for ischemia-induced memory impairments. Mumby et al. (1992b) tested this hypothesis indirectly. Previously hippocampectomized rats that were performing quite well on D N M S were subjected to ischemia. It was reasoned that if extrahippocampal damage was responsible for the memory deficits in ischemic animals then, hippocampectomized rats should develop a D N M S deficit following ischemia. This did not occur, leading Mumby et al. to conclude that the ischemia-induced extrahippocampal damage is mediated or produced by the postischemic hippocampus. Based on the aformentioned results and the known neuroanatomical connections of the hippocampus, a strategy for reconciling the data from the ablation and ischemia experiments is proposed here. It is my contention that the D N M S impairments in ischemic rats are caused by extrahippocampal damage and that this damage is somehow produced by pathogenic processes occuring in the postischemic hippocampus. The fact that D N M S deficits can be produced by lesions of the rhinal cortex, the thalamus, the prefrontal cortex, 6 5 and the basal forebrain suggest that this damage may be occuring in one or more of these areas. Additionally, the strong anatomical connections between these areas and the hippocampus suggests that the pathophysiological processes occurring in the hippocampus responsible for the CA1 loss may be producing damage in distal afferent structures. This damage could be manifest as a wide variety of pathophysiological changes, such as changes in gene expresion, protein synthesis, or receptor densities. A l l of which would not be detectable with conventional histological techniques but may, nevertheless, influence neuronal function. This position leads to two main hypotheses that will be addresed in this thesis. First, deficits on D N M S should not be found in rats following partial lesions of the hippocampus if such lesions are induced by a method other than ischemia. The reason being that such a lesion would be made without inducing the full spectrum of pathophysiological events that accompany ischemia, which are hypothesized to produce extrahippocampal damage. If no deficit is found following such a lesion, then one can reliably conclude that the partial hippocampal damage that results from ischemia is not responsible for D N M S deficits. However, rats with such lesions should be impaired on other tasks that are known to be sensitive to hippocampal damage (i.e., spatial memory tasks). Second, if the pathophysiological processes occurring in the hippocampus following ischemia were eliminated, then the extrahippocampal damage responsible for the memory deficits would not be produced. Therefore, lesioning the hippocampus shortly after ischemia should prevent this damage from occurring and animals in which such a procedure was undertaken should perform similar to controls on D N M S . These hypotheses are addressed in the following series of experiments. In Chapter 3, the effect of neurotoxic lesions to the dorsal hippocampus on nonspatial and spatial memory tasks is reported. The results of these experiments will be important to determine if partial lesions to the dorsal hippocampus can produce deficits in object-recognition memory and/or spatially-guided behavior. In Chapter 4, the effect of hippocampal ablation immediately following ischemia is described in rats with preoperative D N M S experience. These data are discussed in conjunction with the results of Mumby et al.'s (1992) study in 66 order to develop a coherent picture of ischemia and D N M S deficits. Such a comparison is essential if an understanding of the relationship between ischemia, extrahippocampal damage, and object-recognition deficits is to be approached. In Chapter 5, the ability of partial lesions to the hippocampus to block ischemia-induced object-recognition deficits is decribed. This experiment was designed to assess whether or not the ischemia-induced cell death in the CA1 is directly responsible for producing extrahippocampal damage that results in D N M S deficits. CHAPTER 2: GENERAL METHODS 67 This chapter describes the methods most common to the experiments presented in this thesis. Any changes or additions to the general methodology are described in the method section of the appropriate experiment. S U B J E C T S The subjects were experimentally naive male Wistar rats (Charles River Laboratories, Quebec) that weighed 300-350 g at the begining of each experiment. The rats were housed individually and maintained on a 12/12 hr dark-light cycle (light 0800 to 2000 hrs). Prior to behavioral testing, their body weights were reduced to 85% of ad libidum levels. This weight was maintained throughout the experiments by providing each rat with 15-25 g of rat chow per day. Behavioral training began after the rats were on the restricted feeding regimen for 14 days. Rats were allowed free access to water when in their home cages. A P P A R A T U S The D N M S testing apparatus has been described in detail elsewhere (Mumby et al., 1990). Briefly, it consisted of an elevated runway which was separated from identical goal areas at each end by opaque guillotine doors. Each goal area contained two food wells in which food pellets (45 mg Bio-Serv Inc., Frenchtown, N.J.) could be delivered by hand through plastic tubes. Test stimuli consisted of over 400 "junk" objects of various sizes, shapes, colors, and textures. Each object was large enough to cover the food well but small enough to be displaced easily by the rats. No objects with obvious odors were used and the objects were washed every few days in a solution of water and chlorine bleach. B E H A V I O R A L P R O C E D U R E A l l training and testing took place during the light phase of the dark-light cycle between 14 and 21 hr after the rats last meal. Animals were tested no more than once per day and no fewer than 5 times per week. Training consisted of three phases; habituation, simple object discrimination, and D N M S . During the habituation phase, rats were initially allowed to explore the apparatus and eat from previously baited food wells. A s the rats became accustomed to the apparatus, they were shaped to alternate between opposite ends 68 of the box in order to receive reinforcement. Once they were retrieving food pellets consistently, the operation of the guillotine doors was introduced. Rats were shaped to approach the doors by baiting a food well on the far side of a closed door. When the rat approached the door, the door was raised, allowing access to the food well. The same procedure was then repeated at the other end of the box until the rats were alternating between both ends without hesitation. Following habituation, the rats were trained on a simple object discrimination task. Two dissimilar objects were chosen as stimuli and were randomly assigned as S+ (reward) and S- (no reward). At the start of each session one door was closed and the S+ and S-were positioned over the food wells behind the door. The rat was placed in the center of the apparatus and, when it approached the closed door, the door was opened thereby allowing access to the test objects. If the rat displaced the S+ from over its food well then a food pellet was delivered; if the rat displaced the S-, no reward was delivered. Once the rat had displaced one of the objects, the far door was lowered and the S+ and S- were removed and placed behind it in preparation for the next trial. The positioning of the S+ and S- (left vs. right) varied from trial to trial in an irregular but balanced manner. Rats were allowed to correct themselves during the first session only. If a rat made a mistake and chose the S-, the S- was removed and the rat was then allowed to displace the S+, for which it was reinforced. On all remaining sessions, however, no correction was allowed; both the S+ and the S- were removed as soon as one had been displaced. A l l rats were trained until a criterion of at least 21 correct choices out of 25 on two consecutive sessions was reached. Once an animal reached criterion on simple object discrimination, D N M S training began. For each trial a different pair of objects was used. A trial began with the rat in the center of the box and both doors closed. Two objects were chosen; one of which was designated the sample object the other designated the novel object. The objects were placed over randomly designated food wells, one at each end. One of the doors was opened and allowed the rat access to the sample object. The rat displaced the sample object and received a food reward that was concealed beneath it. The sample object was removed 69 immediately and placed at the other end of the apparatus over the vacant food well. This process took approximately 4 s, after which time the other door was opened and the rat was allowed to displace one of the two objects. If the rat chose the novel object, it was rewarded and both objects were removed. If it chose the sample object, both objects were removed and no reinforcement was delivered. Only when a rat actually displaced an object was it considered to have made a choice, merely touching an object was not considered a choice. A s with object discrimination, correction was allowed for the first session only. Rats were then allowed to return to the center of the box at which time both doors were closed and a new trial was begun. Rats were not handled during the D N M S session. Each rat received 20 trials per session until it reached a criterion of at least 17 correct out of 20 on 2 consecutive days (85% correct). After reaching criterion at the 4 s retention interval, each rat received training at retention intervals ranging from 15-300 s. Individual training at these intervals continued until a rat had achieved criterion (17 out of 20 correct on 2 consecutive days) or for a maximum of five sessions. Once the training at the various delays was complete, rats then received 8 days of mixed-delay testing. During these sessions, rats were tested in an ascending delay manner starting with the shortest delay and finishing with the longest. This pattern was then reversed. The exact number of trials for each mixed-delay session varied slightly depending on the number of delays used. S U R G I C A L P R O C E D U R E C E R E B R A L I S C H E M I A Acute transient forebrain ischemia was induced by 2 V O following the method of Wood et al. (1993). This procedure produces observable cell loss primarily within the C A 1 subfield of the hippocampus. The cell loss is most pronounced in the septal pole of the hippocampus and becomes less severe in the temporal pole. Cell loss outside of the CA1 is patchy and inconsistent but most reliably occurs in the hilus of the dentate gyrus. A representative ischemic lesion of the CA1 is shown in Figure 4 (bottom). Rats were anesthetized with sodium pentobarbital (65 mg/kg i. p.) and treated with atropine sulfate (1.0 mg/kg i. p.) to reduce respiratory tract secretions. The common 70 carotid arteries were isolated by blunt instrument dissection and encircled with 5-0 silk ligatures. The femoral artery was isolated in a similar manner and was cannulated with polyethylene tubing (PE 50). The P E 50 tubing was connected to a saline primed reservoir via a pressure transducer (Elecromedics, Inc. Englewood CO) and arterial blood pressure was recorded on a Y - T single channel chart recorder (Gould Inc., Glen Burnie, M D ) . Initially, the rats received a 0.1 ml bolus of heperanized saline via the cannulated femoral artery. Additional small volumes (0.1 ml) of more dilute heparinized saline (20 units/ml) were administered periodically to prevent the cannula from becoming blocked. T o induce ischemia, the mean arterial blood pressure was first reduced to 30mm H G by allowing the rat to hemorrhage through the femoral catheter and into the saline primed reservoir. Once the blood pressure stabilized at this level, the carotid arteries were occluded with atraumatic arterial clamps for 20 min. During this time, the mean arterial blood pressure was maintained at 30 mm Hg by withdrawing or reinfusing blood as needed. After the period of ischemia, the clamps were removed from the carotid arteries and the collected blood was reinfused by applying 200 mm Hg back pressure to the reservior (over a period of 10-15 min). Mean arterial blood pressure quickly returned to preischemic values, at which point the catheter was removed, the femoral artery ligated and the incisions sutured. During the entire period of surgery, the rats core body temperature was maintained at 35° C with a temperature controlled heating pad (American Hospital Supplies, McGraw Park, U L ) . N E U R O T O X I C L E S I O N S Procedures for producing the neurotoxic lesions were adapted from Volpe et al. (1992) and Sugaimachi et al. (1992) and were designed to mimic the distribution of cell loss observed in the hippocampus of rats subjected to 2VO-ischemia (i.e., greater cell loss in the septal pole of the CA1 with less pronounced loss in the temporal pole). Such lesions were intended to do minimal damage to other subfields of the hippocampus or adjacent structures. Rats were anesthetized with sodium pentobarbitol (65 mg/kg i.p.) and placed in a stereotaxic apparatus. The scalp was incised in order to reveal the skull and holes were drilled over the hippocampus with a dental drill. Lesions were made by giving three 71 Figure 4. Photomicrographs of an NMDA-lesion of the dorsal hippocampus, restricted primarily to the CA1 subfield (top), and ischemia-induced CA1 cell loss in the hippocampus following 20 minutes of 2VO-ischemia (bottom). Compare with the intact hippocampus in Figure 2. 7 2 73 injections bilaterally of N-methyl, D-aspartate ( N M D A , Sigma, St. Louis, M O ) at the following coordinates: bregma -3.3 mm, midline +/-2.0 mm, depth -2.6 mm; bregma -4.0 mm, midline +/- 2.5 mm, depth -2.7 mm; bregma -4.5 mm, midline +1-2.9 mm, depth -2.9mm. Exact concentrations of N M D A varied slightly between experiments and will be described in the method section of each experiment. Delivery of the neurotoxin was accomplished via a 10/d Hamilton syringe and a Harvard apparatus pump set at an infusion rate of 0.14 /d/min. Following infusion, the injection cannulae were left in place for 5 min to allow the neurotoxin to diffuse into the surounding tissue. The cannulae were then removed and the wound was closed with stainless steel auto clips. A n NMDA-lesion limited to the CA1 subfield is shown in Figure 4 (top). Sham-lesioned rats received exactly the same treatment. However, in order to reduce the possibility of inadvertent C A 1 damage, an infusion of vehicle was not made. H I S T O L O G I C A L P R O C E D U R E Following the completion of behavioral testing, rats were anesthetized with sodium pentobarbital (100 mg/kg i.p.) and transcardially perfused with 10% formalin in 0.05% phosphate buffered formalin (pH 7.4). Their brains were immediately removed and stored in phosphate buffered formalin for at least 24 hr. The brains were then dehydrated in graded ethanols and xylene and embedded in paraffin. Coronal sections, 10 pim thick were cut throughout areas of interest. Every tenth section was mounted on a gelatin coated slide and Nissl stained with 0.1% cresyl violet. Exact methods for the quantification of neuropathology are included with each experiment. 74 CHAPTER 3: AN ANALYSIS OF THE EFFECT OF PARTIAL HIPPOCAMPAL DAMAGE ON THE PERFORMANCE OF NONSPATIAL AND SPATIAL MEMORY TASKS The three experiments presented in this chapter were designed to test the role of partial hippocampal damage, including the CA1 region, on memory. The first two experiments examined the effect of NMDA-lesions of the dorsal hippocampus on D N M S learning in rats that received preoperative training and in rats that were naive to D N M S procedures at the time of surgical manipulation, respectively. The third experiment examined the role of partial hippocampal damage on spatial memory. E X P E R I M E N T 1: T H E E F F E C T O F P A R T I A L L E S I O N S O F T H E D O R S A L H I P P O C A M P U S O N D N M S P E R F O R M A N C E IN R A T S W I T H P R E O P E R A T I V E T R A I N I N G INTRODUCTION Rats and monkeys that received preoperative training on D N M S and were then subjected to ischemia were found to be impaired when retested postoperatively. In rats, this impairment was manifested both as a deficit in the reacquisition of the nonmatching rule and as a deficit across retention intervals up to 5 min (Wood et al., 1993; Mumby et al., 1992b). Monkeys were found to have a mild postoperative reacquisition deficit and impairments at delays up to 120 s (Bachevalier & Mishkin, 1989). If a partial hippocampal lesion is sufficient to produce this deficit by interfering with afferent structures involved in memory (Bachevalier & Mishkin, 1989), then partial hippocampal damage as a result of a neurotoxic lesion should produce a similar behavioral effect. If, however, ischemia-induced damage to extrahippocampal structures is responsible for producing object-recognition deficits then neurotoxic lesions restricted to the dorsal hippocampus should have little effect on D N M S performance. METHOD Subjects The subjects were 11 experimentally naive male Wistar rats (Charles River 75 Laboratories, Quebec). Apparatus The apparatus was the same as described in the general methods section. Behavioral Procedure Prior to surgical manipulation, the rats were habituated to the D N M S apparatus. The object discrimination task was then administered, followed by the acquisition portion of the D N M S task. Once the animals had reached criterion on D N M S (85% correct for 2 consecutive days at a delay of 4 s) they were given training at delays of 30, 60, 120, and 300 s. Rats were trained at delays until a criterion of 85% correct was reached or for five sessions. Rats then received eight mixed-delay sessions that included delays of 4, 30, 60, 120, and 300 s. The mean score for each retention interval obtained during mixed-delay testing served as the animal's baseline measure of performance to which postsurgery scores would be compared. Based on their performance on the mixed-delay retention intervals, rats were matched into two groups and then received the appropriate surgical manipulation. Postoperative testing began with the reacquisition of the D N M S task. Similar to presurgery training, rats were required to achieve a score of 85% correct for 2 consecutive days in order to have reached criterion. Once they had reached criterion, rats were then given eight mixed-delay sessions in an identical manner in which they had received them preoperatively. Surgical Procedure Six rats received bilateral neurotoxic lesions of the dorsal hippocampus made with N M D A (Sugaimachi et al., 1992) and five rats received a sham surgical procedure (the N M D A - L and N M D A - S groups, respectively). For the lesion group, bilateral injections of N M D A (0.7 fil, 40nmol//^l) were made at three sites along the septotemporal extent of the dorsal hippocampus following the procedures outlined in the general methods section. The sham-lesioned rats were subject to exactly the same treatment but no infusion of N M D A was given. Immediately after surgery rats were housed in individual plastic cages and their condition was closely monitored. Unlike animals subjected to forebrain ischemia, rats that 76 received NMDA-lesions displayed no, or only very mild, overt convulsions. Survival following NMDA-lesions was 100%. The rats were allowed 3 weeks to recover before postoperative behavioral testing was initiated. RESULTS Behavior The mean number of trials required to learn the object discrimination task before surgery is presented in Figure 5. The N M D A - L group required a mean of 66.6 trials to learn the discrimination and the N M D A - S group required a mean of 60 trials. This difference was not significant (t (9) = 0.83, p = 0.25) and demonstrates that both groups of rats performed the object discrimination equivalently at the begining of the experiment. The mean number of trials to reach criterion pre- and postsurgery on D N M S for the N M D A - L and N M D A - S groups are shown in Figure 6. A two-factor repeated measures analysis of variance ( A N O V A ) demonstrated that there was no significant difference between the mean number of trials for the N M D A - L ( M = 220) and the N M D A - S ( M = 264) groups to reach criterion prior to surgery (F [1, 20] = .273, p = 0.60). Similarly, there was no significant difference in the mean number trials ( N M D A - L - M = 85; N M D A -S - M = 65) taken by both groups to reattain criterion postsurgically (F <1). In fact, both groups of animals reacquired the D N M S task significantly faster postsurgically than when they learned it initially (F [1, 20] = 36.37, p = 0.0001). Figure 7 shows the pre- and postsurgery scores of the N M D A - L and N M D A - S rats averaged across each delay in the eight mixed-delay sessions. Both groups exhibited a similar pattern of results. Pre- and postsurgery scores were at or above 85% at the 4 s interval and declined as the delay was increased to 300 s. A three-factor A N O V A (group x time x delay ) revealed no significant main effect for group ( F [1, 9] = 1.60, p = 0.24) or time (pre- vs. postsurgery; F < 1). Thus, the N M D A - L and N M D A - S groups did not differ from each other either pre- or postsurgically. There was a significant main effect for delay (F [3, 27] = 31.00, p > 0.0001) indicating that the scores at the long delays were significantly lower than scores at the short delays. There was no significant delay x group 77 Figure 5. Mean number of trials to reach criterion on object discrimination for the NMDA-L and NMDA-S groups prior to surgery. GROUP 79 Figure 6. Mean number of trials to reach criterion on D N M S for the N M D A - L and the N M D A - S groups pre- and postsurgery. 80 300 o cc o 12 CO < cc 200 H 100 H X • NMDA-L/PRESURGERY 0 NMDA-L/POSTSURGERY • NMDA-S/PRESURGERY 0 NMDA-S/POSTSURGERY GROUP Figure 7. Mean percent correct responses across delays for the N M D A - L and N M D A - S groups pre- and postsurgery. 82 100 90 H 80 H 70 H 60 H 50 NMDA-L/PRESURGERY NMDA-S/PRESURGERY NMDA-L/POSTSURGERY NMDA-S/POSTSURGERY T " 4 - 1 — 60 120 300 D E L A Y 83 interaction (F [3, 27] = 2.32, p = 0.10) or delay x time interaction (F [3, 27] = 1.16, p = 0.34), indicating that neither the lesion nor time of testing contributed to the declining scores over delays. Neuropathology Following the completion of behavioral testing, the rats were sacrificed and their brains prepared as described in the general procedure section. The entire septotemporal length of the hippocampus was then sliced into 10 pirn sections. Every tenth section was mounted on a gelatin coated slide and stained with 0.1% cresyl violet. Figure 8 shows the extent of the neurotoxic lesions at five septotemporal levels of the hippocampus. The extent of the lesions was determined by projecting the slide onto a schematic of the corresponding hippocampal level from the atlas of Paxinos and Watson (1986). As can be seen, all six of the N M D A - L rats received partial lesions to the dorsal hippocampus. The lesion included most of the CA1 across the dorsal hippocampus. Additional, but inconsistent, damage to the C A 2 - C A 4 , and dentate gyrus also occurred, predominantly at the more septal levels. Some incidental cortical damage also occurred as a result of the injection cannula and residual N M D A . DISCUSSION The results of Experiment 1 demonstrate that partial damage to the dorsal hippocampus in rats as a result of neurotoxic lesions is not sufficient to produce an impairment in object-recognition memory. Preoperatively trained rats that received N M D A -lesions reattained criterion on D N M S at a rate equivalent to sham-operated rats. Postsurgery scores at delays of 4, 60, 120, and 300 s were also equivalent between groups. Additionally, there was no within-groups difference between pre- and postsurgery scores for either the N M D A - L or N M D A - S groups. These findings contrast sharply with those obtained from rats that have received partial damage to the hippocampus as a result of cerebral ischemia (Mumby et al., 1992a; Wood et al., 1993). Postsurgically, ischemic rats typically require significantly more trials than sham-ischemic rats to reattain criterion on D N M S , are impaired at all delays relative to 84 Figure 8. The extent of the dorsal hippocampal lesions in the N M D A - L rats. C A 1 lesions are in red, lesions to other hippocampal subfields are green and incidental cortical damage is black. The numbers refer to individual rats. 8 6 87 sham-operated animals, and perform significantly more poorly on the D N M S task than they did preoperatively. The comparison of these discrepant results is particularly relevant to the question of the role of ischemia-induced hippocampal lesions in producing D N M S deficits. Partial lesions produced by ischemia cannot be solely responsible for such deficits as partial damage produced by neurotoxic lesions had no effect on D N M S performance. Although the neurotoxic lesions were not entirely restricted to the C A 1 , they did manage to destroy a large portion of this region along most of the septotemporal and transverse axis of the hippocampus. Additionally, the neurotoxic lesions were most extensive in the septal pole of the hippocampus, the area that also receives the greatest amount of damage following 2VO-ischemia (Auer, Jensen, & Whishaw, 1989; Wood et al., 1993) Based upon the known septotemporal extent of intrahippocampal projections (Amaral & Witter, 1991), the extent of these lesions should have been sufficient to produce some degree of functional disorganization of the remaining hippocampus. Whereas functional disorganization has been suggested as a possible mechanism to account for the behavioral paradox between partial ischemia-induced hippocampal lesions and complete hippocampal ablation (Bachevalier & Mishkin, 1989), the present results do not support this hypothesis. The lack of a signifcant impairment on D N M S following neurotoxic lesions that are anatomically and functionally similar to an ischemic lesion indicates that ischemia-induced D N M S deficits may occur as a consequence of damage to extrahippocampal structures. E X P E R I M E N T 2: T H E E F F E C T O F P A R T I A L LESIONS O F T H E D O R S A L H I P P O C A M P U S O N D N M S P E R F O R M A N C E IN N A I V E R A T S INTRODUCTION Experiment 2 was designed to investigate the effect of partial hippocampal lesions on D N M S learning in rats that received no preoperative D N M S training. In monkeys, preoperative training has been shown to influence the severity of a performance deficit following lesions to the hippocampal region, with the impairment being attenuated following preoperative training (Zola-Morgan et al., 1986). Additionally, ischemic rats with partial hippocampal lesions are severely impaired in acquiring and performing the 88 D N M S task (Wood et al., 1993) whereas rats with hippocampal ablations show only a mild impairment (Mumby, et al., 1994). Rats in the present experiment received N M D A -lesions to the dorsal hippocampus prior to being trained on the D N M S task. As with Experiment 1, failure to observe a D N M S deficit in rats with partial NMDA-induced lesions of the dorsal hippocampus would suggest the D N M S impairments in ischemic rats are not due solely to the result of CA1 lesions. METHOD Subjects The subjects were 13 experimentally naive male Wistar rats. Housing and food deprivation procedures were identical to those in Experiment 1. Apparatus The apparatus was identical to the one used in Experiment 1. Surgical Procedure Seven of the rats received bilateral N M D A lesions to the dorsal hippocampus ( N M D A - N A I V E ) and six received a sham surgical procedure (S-NAIVE). The lesioning method was identical to that in Experiment 2. However, in an attempt to produce less damage to the dentate gyrus, C A 2 - C A 4 , and overlying cortex, a smaller volume (0.4/<l) of the same concentration of N M D A was used. Behavioral Procedure Behavioral training and testing were identical to experiment 1 except that all training took place postsurgically. RESULTS Behavior Figure 9 shows the mean number of trials required by the N M D A - N A I V E and the S - N A I V E groups to learn the object discrimination. S - N A I V E animals required an average of 67 trials to learn the discrimination whereas N M D A - N A I V E animals required an average of 75 trials. This difference was not statistically significant (t (11) = 0.63, p = 0.54), indicating that both groups of rats performed the object discrimination equivalently at the 89 Figure 9. Mean number of trials to reach criterion on object discrimination for the N M D A - N A I V E and S - N A I V E groups. GROUP 91 Figure 10. Mean number of trials to reach criterion on D N M S for the N M D A -N A I V E and S - N A I V E groups. GROUP 93 Figure 11. Mean percent correct responses across the mixed-delay sessions for the N M D A - N A I V E and S - N A I V E groups. D E L A Y 95 start of the experiment. Figure 10 shows the mean number of trials required by each group to reach criterion on the D N M S task. The S - N A I V E rats required a mean of 330 trials to reach criterion whereas N M D A - N A I V E group required a mean of 334 trials. The difference between these groups was not statistically significant ( t (11) = 0 .09, p = 0.93). The mean scores at each delay for the N M D A - N A I V E and S - N A I V E groups are presented in Figure 11. Both groups scored approximately 90% correct at the 4 s delay and their performance declined monotonically as the delay was increased to 300 s. A 2-factor repeated measures A N O V A revealed that there was no significant main effect for group (F [3, 11] = 0.54, p = 0.48) indicating that, overall, the lesioned and the sham- lesioned groups performed equivalently. There was a significant main effect for delay (F [3,11] = 39.98, p = 0.0001) indicating that performance declined as the delay length was increased. However, the group x delay interaction was nonsignificant (F [3, 11] = 0.89, p = 0.46) demonstrating that the decline in perfomance with increasing delays was comparable between the N M D A - N A I V E and S - N A I V E groups. Neuropathology Rats from this experiment were also used in Experiment 3. Detailed histological results for these animals can be found in the neuropathology section of Experiment 3. Briefly, however, the NMDA-lesions damaged most of the CA1 subfield of the hippocampus, with the damage being most consistent in the medial and septal aspects. Damage to other hippocampal fields and the cortex was less than in Experiment 1. DISCUSSION NMDA-lesions of the dorsal hippocampus did not produce an impairment in the learning or the performance of the D N M S task in rats that had no preoperative experience with the task. Lesioned rats were able to acquire the nonmatching rule in approximately the same number of trials as sham-lesioned rats. Furthermore, the acquisition scores for both groups (approximately 250 trials) fell within the range typically required by controls to reach criterion (Mumby et al., 1990;, Mumby et al., 1992a; Wood et al., 1993). The N M D A - N A I V E and S - N A I V E groups also performed equivalently at delays of 4, 60, 120, 96 300 s. These findings are in sharp contrast to those obtained in ischemic rats with no preoperative D N M S experience. Such rats were profoundly impaired in the acquisition of the D N M S task, some to a point were they did not learn the nonmatching rule after 1000 trials. Additionally, when ischemic rats did manage to learn the nonmatching rule they were subsequently found to be impaired at delays up to 300 s, relative to sham-ischemic animals (Wood et al., 1993). These results underscore the fact that there is a qualitative difference in the locus and/or type of brain damage produced by ischemia as compared with NMDA-lesions. Rats in the present study sustained damage to much of the same area that is damaged as a result of ischemia, namely the dorsal-septal C A 1. Again, the amount of this damage could be reasonably expected to produce some type of functional disorganiztion in the remaining hippocampus. As these animals were not even slighlty impaired on the D N M S task, it would seem unlikely that the functional disorganization of the hippocampus is responsible for ischemia-induced memory deficits. E X P E R I M E N T 3: T H E E F F E C T O F P A R T I A L LESIONS O F T H E D O R S A L H I P P O C A M P U S O N S P A T I A L L Y - G U I D E D B E H A V I O R INTRODUCTION Experiment 3 was designed to determine the effect of NMDA-lesions of the dorsal hippocampus on spatial memory in the Morris water-maze (Morris et al., 1982). Although there is general agreement that the rodent hippocampus participates in spatial memory (Jarrard, 1993; Nadel, 1991; 1992), many questions about specific function remain to be answered. For example, it is still unclear what the mnemonic effect of damage to the different subfields may be. Ischemia typically results in impaired performance on the Morris water-maze, however, the amount of CA1 cell loss does not correlate strongly with the severity of the memory impairment (Nunn et al, 1994; Olsen et al., 1994a; 1994b). Neurotoxic lesions of the C A 1 have also been shown to impair performance on the water maze but the pattern of impairment is different from that observed following ischemia. These discrepant findings have led to the suggestion that ischemia-induced C A 1 cell loss may not be the only determinant of impaired spatial memory in the water-maze (Nunn & 9 7 Hodges, 1994; Nunn etal., 1994) The present experiment utilized the rats from Experiment 2 in an attempt to determine if a partial hippocampal lesion, which failed to produce a memory impairment in an object-recognition task, can produce a deficit on a task known to be sensitive to hippocampal damage. It is hypothesized that N M D A lesioned rats from Experiment 2 will be impaired in spatial learning on the Morris water-maze. This finding would indicate that partial hippocampal lesions are sufficient to impair some, but not all, types of memory and would suggest a selectivity in the mnemonic process mediated by the hippocampus. METHOD Subjects The subjects were the same rats that were used in Experiment 2. The N M D A -N A I V E group received bilateral lesions to the dorsal hippocampus 14-16 weeks prior to being tested on the water-maze. The S - N A I V E group underwent a sham surgical procedure within the same time frame. Rats were given free access to food and water when in their home cages. Apparatus The testing room measured 3 x 7 x 3.5 m and was illuminated by a 120 W incandescent light bulb in the ceiling near the center of the room. The walls of the room were painted beige and the room contained numerous extra-maze cues that the rats could use as landmarks. The experimenter was separated from the water-maze by a large brown room divider. Between trials the rats were kept in a plastic bin that was lined with paper towels and bedding. The Morris water-maze (Morris et al., 1982) consisted of a large white plastic pool 180 cm in diameter and 54 cm in height. The pool was filled with 20 cm of water which was rendered opaque by the addition of water-soluble, non-toxic, white powdered paint. The water temperature was approximately 22° C. The submerged escape platform was 18cm tall and 9 cm in diamater with a piece of wire mesh 13cm x 13cm fixed to the top. Animal movement in the pool was recorded by an overhead tracking system consisting of a video camera linked to a computer equipped with tracking software ( H V S image, 98 Hampton, UK). Prior to each trial the rats were marked with a black felt marker along the head and back to provide contrast with the water required by the video tracking system. The computer system tracked the swim paths, recorded the latency to find the submerged platform, and measured the path length for each rat as it searched for the platform. Procedure Training on the Morris water-maze began approximately 2 weeks after the completion of D N M S testing. For a single training trial, rats were placed tail first into the water, facing toward the pool wall. They were then allowed 90 s to search for the submerged platform. If a rat failed to find the platform in the alloted time, the experimenter would then manually guide him toward it. Once on the platform, the rat was allowed 15 s to observe the extramaze cues. The rat was then removed from the platform and placed in the holding bin until the next trial. The intertrial interval was approximately 4 min. The escape platform was always located in the northwest quadrant of the pool. The release points (N, S, E , or W) were varied randomly from trial to trial. Animals were given eight trials per day for 4 days. The data from every four trials was averaged to form eight trial blocks which were used in the final analysis. RESULTS Behavior Figure 12 shows the mean latency to find the hidden platform for the N M D A -N A I V E and the S - N A I V E groups over blocks of eight trials. As can be seen, both groups improved in their ability to find the submerged platform with daily practice sessions. A two-factor repeated measures A N O V A revealed a significant main effect for group ( F [1,11] = 7.48, p = 0.02) indicating that the rats with lesions to the dorsal hippocampus took longer to find the submerged platform than the sham-lesioned rats. The effect of trial block and the trial block x group interaction were both nonsignificant (F < 1). Figure 13 shows the mean path length to reach the sumberged platform for both groups over blocks of eight trials. As can be seen the N M D A - N A I V E group had longer path lengths than the S - N A I V E group. A two-factor repeated measures A N O V A revealed a significant main 99 Figure 12. Mean latency to find the submerged platform for the N M D A - N A I V E and S - N A I V E groups over blocks of eight trials. 100 TRIAL BLOCKS 101 Figure 13. Mean path length to find the hidden platform for the N M D A - N A I V E and S - N A I V E groups over blocks of eight trials. 102 103 effect for group ( F [1,11] = 17.32, p = 0.002) indicating that the N M D A - N A I V E rats traveled farther in the pool to find the platform than did the S - N A I V E group. The main effect for trial block and the group x trial block interaction were both nonsignificant ( F < 1 ) . N e u r o p a t h o l o g y Following the completion of behavioral testing, the rats were sacrificed and their brains prepared as described in the general procedure section. The method for quantifying the extent of brain damage was identical to that in Experiment 1. Figure 14 shows the extent of the neurotoxic lesions at five septotemporal levels of the hippocampus. As can be seen, all seven of the N M D A - N A I V E rats sustained partial lesions to the dorsal hippocampus. The lesion included most of the C A 1 across the dorsal hippocampus and was most severe in the septal pole. Additional, but inconsistent, damage to the C A 2 , C A 3 , C A 4 , and dentate gyrus also occurred, predominantely at the more septal levels. Some incidental cortical damage also occurred as a result of the injection cannula and residual N M D A . DISCUSSION Partial NMDA-induced lesions of the dorsal hippocampus were sufficient to produce an acquisition deficit on the Morris water maze task. Lesioned rats took longer to find the submerged platform and exhibited more random searching than did sham-lesioned animals. These findings are similar to those obtained in rats with ischemia-induced CA1 lesions (Jasper et al., 1990; Nunn et al., 1994; Olsen et al., 1994a; 1994b) and ibotenic acid lesions of the CA1 (Nunn & Hodges, 1994). The N M D A lesions in the present experiment produced widespread cell loss in the dorsal and septal regions of the C A 1 , a pattern similar to that found following cerebral ischemia (Nunn & Hodges, 1994; Wood et al., 1993). Addtional, but inconsistent, damage was also present in the C A 2 , C A 3 , C A 4 , and dentate gyrus at the most septal levels. It is possible that some of this damage also contributed to the deficits observed in the water-maze, as damage to C A 2 , C A 3 , and dentate gyrus has also been shown to produce water-maze deficits (Barone, Tandon, McGinty, & Tilson, 1991). However, in this study the lesions to these structures 1 Figure 14. The extent of the hippocampal lesions in the N M D A - N A I V E group. The extent of the CA1 lesions is shown in red, lesions to other hippocampal subfields is indicated by green, and incidental cortical damage is black. Numbers refer to individual rats. The shaded area in rat #5 represents an area of complete tissue loss. 105 1 0 6 107 were far more extensive than in the present study. The dissociation between the effect of partial hippocampal lesions on the performance of D N M S or the water-maze is consistent with the idea that the hippocampus is primarily involved in spatial learning and memory (Jarrard, 1993, Nadel, 1991; O'Keefe & Nadel , 1978). This idea is supported by the fact that hippocampal lesions disrupt spatial abilities in rats (Jarrard, 1993 Nadel, 1991; 1992; O'Keefe & Nadel, 1978), that there are specific cells in the hippocampus that respond differentialy to spatial locations, the so called "hippocampal place cells" (O'Keefe & Nadel, 1978), and the fact that the hippocampus is proportionately larger and contains more neurons in the brains of animals that rely on spatial abilities for surival, such as food caching birds (Healy & Krebs, 1989; Krebs, Sherry, Healy, Perry & Vacccarino, 1989; Sherry, Jacobs, & Gaulin, 1978). Interestingly, these hippocampal volume differences are not present at birth and appear to be dependent on experience rather than on maturational factors. Marsh tits that were prevented from caching had lower hippocampal volumes than those allowed to cache and retrieve food (Clayton, 1995; Clayton & Krebs, 1994). Partial lesions of the hippocampus may disrupt spatial behavior by disrupting the functional integrity of the hippocampus and thereby simulating a complete lesion. On the other hand, each subfield may make a unique contribution to spatial memory such that lesions of the various subfields, separately or in combination, may produce very different behavioral results (Eichenbaum, Otto, & Cohen, 1994; Nunn & Hodges, 1994). 108 CHAPTER 4: ISCHEMIA-INDUCED OBJECT-RECOGNITION M E M O R Y DEFICITS IN RATS: AN ANALYSIS OF T H E LOCUS OF D A M A G E USING HIPPOCAMPAL ABLATION AND CEREBRAL ISCHEMIA, SEPARATELY AND IN COMBINATION This chapter seeks to clarify the relationship between ischemia-induced object-recognition deficits and pathophysiological events occurring in the postischemic hippocampus. A s described in the introduction, Mumby et al.'s (1992b) study suggested that ischemia-induced D N M S deficits are caused by extrahippocampal damage that is undetectable with standard histological techniques, not the observable C A 1 pyramidal cell loss. This damage, however, appears to be mediated by the postischemic hippocampus. Experiment 4 addresses this issue in that hippocampal ablations were performed immediately following ischemia. Such a procedure is expected to prevent the occurrence of the aberrant, hippocampally-mediated neuronal events that produce extrahippocampal damage and, thereby, prevent subsequent ischemia-induced D N M S deficits. These data are then discussed in combination with the results of Mumby et al. (1992b) in order to describe more precisely the relationship between hippocampal pathophysiology and object-recognition deficits. This integration is crucial because Experiment 4 provides a direct test of the idea that the postischemic hippocampus mediates extrahippocampal damage. By comparing the D N M S performance among rats that received either ischemia, hippocampal ablation, or some combination of the two, it will be possible to determine the role of the postischemic hippocampus in producing object-recognition deficits and establish a time frame for when the extrahippocampal neuropathology may be produced. E X P E R I M E N T 4: P R E V E N T I O N O F O B J E C T - R E C O G N I T I O N M E M O R Y DEFICITS B Y H I P P O C A M P A L A B L A T I O N I M M E D I A T E L Y F O L L O W I N G C E R E B R A L I S C H E M I A INTRODUCTION The present experiment provides a direct test of the hypothesis that the postischemic hippocampus mediates or produces the extrahippocampal damage that is responsible for 109 ischemia-induced object-recognition memory deficits. Rats were trained on D N M S and then received 2 V O cerebral ischemia followed by bilateral hippocampal aspiration lesions. Unlike Mumby et al.'s (1992b) initial work, the period between ischemia and hippocampectomy was within 1 hr, not 8-10 weeks. It was hypothesized that the hippocampal ablation would eliminate or attenuate any ischemia-induced D N M S deficits by removing the aberrant pathophysiological activity in the hippocampus that may be producing damage in structures important for recognition memory. METHOD Subjects The subjects were 12 experimentally naive male Wistar rats (Charles River Laboratories, Quebec). Apparatus The apparatus was identical to the one described previously. Behavioral Procedure The habituation, object discrimination, and D N M S acquisition procedures were identical to those of Experiments 1 and 2. Training and testing at the longer delays and at the mixed-delays varied slightly in that a 15 s delay was added. Thus, rats received training and testing at delays of 4, 15, 60, 120, and 300 s. Surgical Procedure Eight of the rats received ischemia followed by hippocampal ablation (ISC-plus-A B L ) and four of the rats received ischemia (ISC) only. Prior to surgery, rats were anesthetized with sodium pentobarbitol (65 mg/kg i.p.) and treated with atropine sulfate (1.0 mg/kg i.p.) to reduce respiratory tract secretions. Additional supplements of pentobarbitol were administered periodically during surgery as required to mantain anesthetization. Ischemia was induced according to the procedure outlined in the general methods section. Within Ihr following ischemia, hippocampal ablations were made using a combination of aspiration to remove the dorsal hippocampus and electrolysis to lesion the ventral hippocampus, dentate gyrus, and subiculum (Mumby et al., 1992a). Rats were placed in a stereotaxic apparatus, and the scalp was incised to reveal the skull. Holes were 110 cut in the skull over each hippocampus. The holes extended from approximately 2mm posterior to the coronal suture to 2mm anterior to the lamboid suture, and from 1.5mm lateral to the sagittal suture to within 1 mm of the temporal ridge. The electrolytic lesions of the ventral hippocampus were made first, followed immediately by aspiration lesions of the dorsal hippocampus. The electrolytic lesions were made bilaterally at two sites (2 m A for 15 s) with a bipolar stainless steel electrode that was insulated with teflon except for approximately 1mm at its tip. Relative to bregma, the coordinates for the electrode sites was as follows: A P -4.8, M L 4.6, D V -9.4; A P -5.5, M L 4.6, D V -9.4. For the aspiration lesions, a portion of the posterior parietal cortex and white matter was aspirated using a Pasteur pipette, exposing the dorsal hippocampus. The dorsal hippocampus was aspirated, the.cavity filled with Gelfoam (Upjohn Co., Don Mills, Ontario), and the skin sutured. Following surgical manipulation, rats were allowed to regain consciousness under a heat lamp and were then transferred to a cool, dark, quiet room were they were kept for the next several days. During this time, their recovery was monitored closely. Some rats began to display overt convulsions and small amounts of Diazepam (10-15 mg/kg i. p.) were administered in order to suppress them. RESULTS Behavior Three of the ISC-plus-ABL rats and 2 of the ISC rats died as a result of the surgical manipulation. This left 5 and 2 rats per group, respectively. For statistical purposes, the data from the 2 ischemic rats in the present study were combined with the data from a pool of 8 other ischemic rats from our laboratory. This pool consisted of ISC rats from related experiments that employed the identical behavioral protocol and in many cases were tested by the same experimenters. Figure 15 shows the mean number of trials required by the ISC-plus-ABL rats and ISC-only rats to reattain criterion postsurgically. The ISC-plus-A B L rats required a mean of 80 trials to reattain criterion on the D N M S task, whereas the ISC rats required a mean of 282. This difference was statistically significant (t (13)= 2.11, p = 0.05) indicating that ISC rats were impaired at relearning the D N M S task 111 Figure 15. Mean trials to criterion on D N M S for the ISC-plus-ABL and ISC only groups postsurgery. 112 400 l GROUP 113 Figure 16. Mean percent correct responses across delays for the ISC-plus-ABL and ISC only group pre- and postsurgery. D E L A Y ( S E C ) 115 postsurgically. Figure 16 shows the mean percent correct responses across the mixed-delays (4, 15, 60,120, 300 s) for ISC-plus-ABL rats and ISC rats. A repeated measures A N O V A revealed that the postsurgery scores of the ISC rats were significantly lower than their presurgery scores (F [1,9] = 101.73, p < 0.0001). In contrast, there were no significant differences between the pre- and postsurgery scores of the ISC-plus-ABL rats (F [1,4] <1). There was also no significant difference between the postsurgery scores of the ISC and ISC-plus ABL-groups (F [1, 13] = 2.11, p > 0.05). Neuropathology Following the completion of behavioral testing, the brains of the ISC-plus-ABL rats were processed as decribed in the general methods section. They were then sliced on a microtome, placed on gelatin coated slides and stained with 0.1% cresyl violet. The extent of the hippocampal lesions were estimated by projecting a slide from a given level onto a schematic of the rat brain from the atlas of Paxinos and Watson (1989) in a manner similar to which the NMDA-lesions were estimated in Experiments 1 and 2. Figure 17 illustrates the largest and smallest lesions in the ISC-plus-ABL group. Overall, all of the dorsal hippocampus and most of the lateral and ventral hippocampus were removed. Additionally, some of the posterior parietal cortex and corpus callosum were also removed. The caudal extent of the lesions included portions of the subiculum, presubiculum, and parasubiculum bilaterally. Each lesion also extended rostrally to include the fimbria/fornix. The brains of the ISC-alone rats were processed in an identical manner as the brains of the ISC-plus-ABL rats. Quantification of the neuronal damage was obtained by direct visual counting of viable neurons using a light microscope at 40x power. C A 1 pyramidal cells were counted bilaterally at six different hippocampal levels equally distributed along its septotemporal axis. These cell counts were then averaged across the right and left hippocampi and expressed as the mean number of neurons per unit length of the pyramidal cell field (cells/125 pirn) for each level of the hippocampus. Figure 18 shows the mean number of CA1 pyramidal neurons per 125 pim at each of the six hippocampal levels for ISC rats in comparison to control rats from Mumby et al. (1992b). A repeated measures A N O V A was performed on the CA1 cell counts using group (ISC vs Control) and 116 Figure 17. A representation of the largest (black) and smallest (striped) hippocampal ablations in the ISC-plus-ABL group. 117 -2.56 -4.52 -6.70 118 Figure 18. Percent of CA1 cell loss across six septotemporal levels of the hippocampus for the ISC rats compared to controls from Mumby et al. (1992b). Figure adapted from Auer, Jensen, and Whishaw, (1989). 119 • ISC 120 septotemporal level (1-6) as the independent variable. A significant main effect was found for group (F [1,9] = 33.35, p < 0.0005), and a significant interaction between group and level was also found (F [5,45] = 3.94, p < 0.005). Planned comparisons between the groups at each of the six hippocampal levels showed that the ischemic rats had significantly fewer CA1 neurons than did controls at all levels (all p's < 0.005). DISCUSSION The results of the present experiment demonstrate that ablating the hippocampus of rats soon after ischemia can prevent the development of subsequent memory impairments. Rats that received ischemia followed within 1 hr by bilateral hippocampal aspiration lesions did not display postoperative D N M S deficits. They were able to reacquire the nonmatching rule significantly faster than ischemic animals and within the range of values typically reported for sham-operated controls (Wood et al., 1993). Additionally, their postoperative performance at delays was not significantly different from their preoperative performance. These findings differ markedly from those of ISC-alone rats. Postoperatively ISC rats were found to have a severe deficit in reacquiring the nonmatching rule; they took longer to learn the nonmatching rule postischemia than they did preischemia. Moreover, the ischemic rats were impaired postoperatively at delays of 4-300 s (Wood et al., 1993). The strength of the within-group finding is compromised somewhat by the inability to find a significant difference between the postsurgical scores of the ISC-plus-ABL and ISC only groups. However, it should be noted that since not all the rats in the ISC group were run at the same time as the ISC-plus-ABL group it was impossible to match presurgery scores. Under these circumstances, the within-group comparison represents a more valid measure when comparing the effects of the surgical manipulation. For comparison purposes, the data from this study are presented in combination with the results obtained by Mumby et al (1992b) in Figure 19. The D N M S impairment in the ISC rats was not attenuated following hippocampal lesions 8-10 weeks later, as would be expected if ongoing hippocampal interference of afferent structures involved in memory is responsible for ischemia-induced D N M S deficits (Figure 19B). This strongly suggests that ischemia-induced neuropathology outside the hippocampus is responsible for the 121 Figure 19. Results from Mumby et al. (1992b; graphs A and B) and the present study (graph C). Graph A shows the performance across delays for preoperatively trained rats that received hippocampal ablation (white squares). Their performance after receiving a second surgical manipulation of ischemia is represented by black squares. Their performance was unchanged by the second manipulation. Graph B shows the performance of preoperatively trained rats that received ischemia (white squares) and their performance following subsequent hippocampal aspiration lesions. Their performance was unchanged by the second manipulation. This pattern of results suggests that the D N M S impairments are the result of extrahippocampal neuropathology mediated or produced by the postischemic hippocampus (see text). Graph C shows the performance of preoperatively trained rats that received ischemia followed within 1 hr by hippocampal aspiration lesions. When retested, their performance on D N M S was unchanged, indicating that pathogenic events in the postischemic hippocampus are responsible for producing the neuropathology that results in memory impairments. O ill CC cr O o H Z LU O CC U i a 100 -.90 -80 70" 60 -50-40 o LU CC CC O o H Z LU o cc LU a K U LU Qt at a: u (J L U a. o 100 90 80 70 60 50 40 100 • 90-80-70-60-50" 40 - • ABL-ISC 122 BEFORE SECOND SURGERY AFTER SECOND SURGERY 15 60 - 120 DELAY (SEC) ISC-ABL 300 BEFORE SECOND SURGERY AFTER SECOND SURGERY B 4 15 60 120 DELAY (SEC) ISC-plus-ABL 300 ISC-plus-ABL Presurgery ISC-plus-ABL Postsurgery 4 IS 60 — T 120 300 DELAY (SEC) 123 D N M S deficits. The mild D N M S deficits in rats with hippocampal ablations was not exacerbated by ischemia (Figure 19A). The failure of ischemia under these conditions to produce a D N M S deficit suggests that the hippocampus mediates the formation of extrahippocampal damage responsible for the D N M S deficits. This hypothesis was confirmed in the present study when it was found that removal of the hippocampus within 1 hr following ischemia prevented subsequent D N M S deficits (Figure 19C). Thus, while each individual experiment provides only limited information about the relationship between ischemia-induced hippocampal neuropathology and object-recognition deficits, taken in combination they paint a clear picure that is consistent with the hippocampus being a mediator, not the source, of D N M S deficits following ischemia. 124 CHAPTER 5: AN ANALYSIS OF THE ABILITY OF PARTIAL HIPPOCAMPAL LESIONS TO PREVENT ISCHEMIA-INDUCED DNMS ACQUISITION DEFICITS E X P E R I M E N T 5: N M D A - L E S I O N S O F T H E D O R S A L H I P P O C A M P U S F A I L T O P R E V E N T I S C H E M I A - I N D U C E D D N M S A C Q U I S I T I O N DEFICITS INTRODUCTION Experiments 1 and 2 demonstrated that partial, NMDA-induced damage to the dorsal hippocampus is not sufficient to produce a deficit in the acquisition of the D N M S task or its performance at delays of up to 300 sec. Furthermore, the results of Experiment 4 are consistent with the hypothesis that ischemia-induced impairments in object-recognition memory are the result of damage to extrahippocampal structures that is produced or mediated by the pathogenic events occurring in the postischemic hippocampus. Ischemia-induced neurotoxicity is most pronounced in the C A 1 subfield (Schmidt-Kastner & Freund, 1991), and this area is also strongly interconnected with diverse brain regions known to be important for memory (Amaral & Witter, 1995; Lopes da Silva, 1990). It is therefore possible that ischemia-induced events in this subfield play a disproportionate role in contributing to the putative extrahippocampal neuropathology. Hippocampectomy prior to or soon after ischemia may prevent D N M S impairments by abolishing the deleterious effects of ischemia-induced CA1 neurotoxicity, while the concomitant removal of the other hippocampal subfields may contribute little to this effect. T o test this hypothesis, rats in Experiment 5 were given partial NMDA-lesions of the dorsal hippocampus that were designed to destroy much of the dorsal CA1. Following recovery, they were then subjected to ischemia and trained on the acquisition portion of the D N M S task. It was hypothesized that the partial lesions would approximate the effect of aspiration lesions and thereby prevent the ischemia-induced D N M S acquisition deficits that would otherwise develop, by attenuating the CA1 neurotoxicity produced by ischemia. 125 METHOD Subjects The subjects were 18 experimentally naive male Wistar rats. Housing and food deprivation conditions were identical to previous experiments. Apparatus The apparatus was the same as described in previous sections. Surgical Procedure Nine of the rats received bilateral lesions of the dorsal hippocampus ( I S C - N M D A ) and the remaining rats received a sham surgical procedure (ISC-S) in a manner identical to that of Experiment 2. After allowing 3 weeks for recovery, both groups then received 20 minutes of 2 V O transient forebrain ischemia following the procedures outlined in the general methods section. Following ischemia, the rats were allowed 4 weeks to recover before the initiation of the behavioral protocol. Behavioral Procedure A l l behavioral training and testing occurred postoperatively and was conducted in a manner similar to that of the previous experiments. Rats were first habituated to the D N M S apparatus and were then trained on a simple object discrimination task. Once rats had mastered the object discrimination task, training on D N M S was begun. Rats were required only to learn the D N M S rule, training and testing at longer retention intervals was not done. RESULTS Behavior Three of the rats in the I S C - N M D A group and 6 of the rats in the ISC-S group died as a result of ischemia, leaving 6 and 3 subjects per group, respectively. Figure 20 shows the mean number of trials required for I S C - N M D A and ISC-S rats to reach criterion on object discrimination (ISC-S, M = 125.0; I S C - N M D A , M =145.83). A t-test revealed no significant difference between the groups (t (7) = 0.54 p = 0.60) indicating that both groups of rats performed the object discrimination task at the same level. Figure 21 shows the mean number of trials required for each group to reach criterion on D N M S (ISC-S, M=640; I S C - N M D A , M= (903.33). A t-test revealed that this difference was not 126 Figure 20. Mean trials to criterion on object discrimination postoperatively for the ISC-NMDA and ISC-S groups. 127 G R O U P 128 Figure 21. Mean trials to criterion on D N M S postoperatively for the I S C - N M D A and ISC-S groups. 129 G R O U P 130 statistically significant ( t (7) = 2.18, p = 0.07). Neuropathology The brains of the I S C - N M D A and ISC-S rats were processed and the extent of the lesions were determined in a manner similar to the previous experiments. The extent of the combined NMDA-ischemic lesions is shown in Figure 22. In general, these animals sustained a similar degree of CA1 cell loss as rats in Experiments 1 and 2. However, because of the combined lesion, it is impossible to determine how much of this cell loss occurred as a result of the NMDA-lesion or how much occurred as a result of ischemia. Unlike the ISC-S rats, the I S C - N M D A rats had damage to the C A 2 , C A 3 , C A 4 , and dentate gyrus. This damage is most likely attributable to the NMDA-lesions, but the possibility that a synergistic effect resulting from the combination of the two lesioning techniques is responsible for some of this damage cannot be ruled out. The neuronal loss in the CA1 of the ISC-S rats was not quantified but the brains were subjected to a general microscopic analysis at lOx magnification. A l l the ISC-S rats had extensive bilateral loss of CA1 pyramidal neurons in a pattern similar to that which has been consistently reported from our lab (Mumby et al., 1992b; Wood et al., 1992, Wood et al., 1993). Visual examination of structures outside of the hippocampus revealed no gross abnormalities. DISCUSSION These results indicate that NMDA-lesions to the dorsal hippocampus prior to ischemia are not sufficient to prevent the development of subsequent D N M S acquisition deficits. The ISC-S group required 640 trials to master the D N M S task whereas as the I S C - N M D A group required 903 trials. Control rats require an average of 250-400 trials to attain criterion (Experiments 1 and 2; also Mumby et al; 1991, Wood et al., 1993). Ischemic rats require approximately 850 trials to reach criterion (Wood et al, 1993). If the lesion had blocked or attenuated the effects of ischemia, it would be expected that the performance of the I S C - N M D A group would be more similar to that of intact animals. In actuality, the NMDA-lesions of the dorsal hippocampus did nothing to attenuate the effects of ischemia. I S C - N M D A rats learned the D N M S task at a rate not significantly different 131 Figure 22. The extent of the dorsal hippocampal lesions in the ISC-NMDA group. CA1 lesions are in shown in red, lesions to other hippocampal subfields are indicated in green, and incidental cortical damage in black. Numbers refer to individual rats. 133 134 from that of ISC-S animals. There are at least two explanations for the failure of NMDA-lesions of the dorsal hippocampus to block object-recognition deficits in a manner similar to hippocampal aspiration lesions. The first concerns the anatomical extent of the NMDA-lesions. These lesions were designed to produce cell loss in the dorsal hippocampus only. The fact that such lesions did not attenuate ischemia-induced D N M S deficits suggests that C A 1 neurotoxicity is not the only pathogenic process contributing to the putative extrahippocampal neuropathology. Alternatively, it is possible that this extrahippocampal neuropathology is mediated or produced by postischemic CA1 neurotoxicity in the ventral C A 1 , an area not lesioned in the present study. The second explanation as to why NMDA-lesions of the dorsal hippocampus failed to block D N M S acquisition deficits has to do with the training protocol in the present experiment. In Chapter 4, hippocampectomy prior to or soon after ischemia blocked object-recognition deficits in rats that had received extensive preoperative D N M S training. Rats in Experiment 5, however, had no preoperative D N M S training. D N M S acquisition is more sensitive to hippocampal damage when no preoperative training is given (Zola-Morgan & Squire, 1986; Mumby et al., 1992a; Mumby et al., 1995). Therefore, acquisition deficits observed in the present experiment could be due to a lack of preoperative training. However, the severity of the acquisition deficits in the present study, as compared to the acquisition impairments in naive rats with hippocampal aspiration lesions (Mumby et al., 1995), suggests that they are due to more than just the partial hippocampal lesions. CHAPTER 6: GENERAL DISCUSSION 135 The primary concern of this thesis has been to address the paradox that exists between the mnemonic deficits produced by ischemia or hippocampal ablation. Specifically, why does an ischemic insult, which results in observable neuropathology confined primarily to the CA1 region of the hippocampus, produce a severe D N M S impairment whereas studies in which complete hippocampal ablations were made indicate that this structure plays little, if any, role in object-recognition memory? This paradox was first demonstrated in monkeys following P C A occlusion (Bachevalier & Mishkin, 1989) and later in rats (Mumby et al., 1992a; 1992b; Wood et al., 1993) following forebrain ischemia. A s a way to explain this discrepancy, Bachevalier and Mishkin (1989) suggested that, although object-recognition memory is not dependent on the hippocampus, partial lesions may produce a functional disorganization of the remaining hippocampal circuitry. The compromised integrity of the remaining hippocampus is suspected to somehow disrupt or interfere with the functioning of the hippocampal afferent structures that actually subserve D N M S learning. The results of Experiments 1 and 2 do not support this "hippocampal interference" hypothesis. Rats in these experiments received partial NMDA-induced lesions to the dorsal hippocampus. These lesions produced widespread cell loss in much of the same area, the C A 1 , that sustains damage as a result of ischemia and were less likely to produce extrahippocampal damage that may result from depriving the entire forebrain of oxygen. Despite such lesions, however, the rats displayed no impairments on any aspect of the D N M S paradigm. Rats in Experiment 1 that received preoperative training and were then given NMDA-lesions were not impaired on the reacquisition of the nonmatching rule or at the performance of the D N M S task at delays of 4, 60, 120, and 300 s. Similarly, in Experiment 2, naive rats with NMDA-lesions learned the nonmatching rule at a normal rate and were unimpaired on D N M S performance at any retention interval. The extent of the NMDA-lesions in present experiments could reasonably be considered to produce some type of functional disorganization of the remaining hippocampus. The diverse and widespread connections along both its septotemporal and transverse axis (Amaral & Witter, 136 1989) suggest that even small lesions of the hippocampus could effect internal neurotransmission. These findings, then, suggest that the functional disorganization of the remaining hippocampal circuitry, as a result of a partial lesion, cannot account for the ischemia-induced D N M S impairments in rats. Mumby et al. (1992b) reached a similar conclusion when it was found that removal of the partially damaged and putatively functionally compromised hippocampus did not attenuate existing D N M S impairments in ischemic rats. If interference by the partially damaged hippocampus with brain regions involved in object-recogntion memory is the source of ischemia-induced D N M S deficits, it seems reasonable to assume that ablating the hippocampus and eliminating this interference would attenuate the impairment. This did not happen. The present findings are also consistent with results obtained in monkeys. Monkeys with partial electrolytic lesions that damaged 38-75% of the hippocampus were able to learn the D N M S task at a normal rate and performed normally at delays up to 120 s (Alvarez et al., 1995). Similarly, ibotenic acid lesions that destroyed 85% of the monkey hippocampus produced no impairments in postsurgical D N M S reacquisition or performance (O'Boyle et al., 1993). One study, however, found that ibotenic acid lesions that damaged only 25% of the hippocampus produced a D N M S acquisition impairment and deficits at delays of 2 and 10 min (Beason-Held et al., 1993). A s this study exists only in abstract form, it is impossible to make a critical analysis of the neuropathology that may account for these deficits. Taken in combination, there is now considerable evidence that partial hippocampal lesions in monkeys and rats, when produced by methods other than ischemia, have little effect on D N M S performance. These findings are consistent with the idea that ischemia-induced D N M S deficits in rats are caused by extrahippocampal damage. It is possible that the NMDA-lesions in Experiments 1 and 2 failed to produce a D N M S deficit because the pattern of brain damage produced by the neurotoxin was not completely analogous to that produced by ischemia. Two aspects of the present study address this concern. First, there was variance in the size of the NMDA-lesions. Some lesions were primarily restricted to the CA1 whereas others included portions of the C A 2 -C A 4 , dentate gyrus, and the overlying cortex (see Figures 8 and 14). This variance can be 137 viewed as a positive factor such that a range of different lesions, which may be thought to produce different patterns of functional disorganization, produced no memory impairments. Second, although the NMDA-lesions of the dorsal hippocampus had no effect on D N M S performance, they did have an effect on performance in the Morris water-maze, a paradigm widely believed to assess spatial memory (Morris et al., 1982). This demonstrates that the lesions were extensive enough to produce an impairment in a task known to be sensitive to hippocampal damage while leaving nonspatial object-recognition memory intact. CA1 cell loss alone may account for this deficit, as both ischemia-induced and neurotoxic lesions of the C A 1 have been shown to disrupt spatial memory (Nunn & Hodges, 1994; Nunn et al., 1994; Olsen et al., 1994a; 1994b; Volpe et al., 1984; Volpe et al., 1992). Damage to other hippocampal regions as possible contributors to the water-maze deficits, however, cannot be completely ruled out. The apparent dissociation between the spatial and non-spatial mnemonic functions of the hippocampus will be addressed later in the discussion. As the ischemia-induced D N M S deficits do not appear to be caused by interference from a partially damaged hippocampus, Bachevalier and Mishkin's (1989) second hypothesis, that these deficits are the result of ischemia-induced extrahippocampal damage that is difficult to detect with conventional histological techniques appears to be more tenable. The fact that hippocampal ablation several weeks prior to an ischemic insult in rats prevented the subsequent object-recognition memory impairments from developing suggested that this damage, wherever it may be occurring, is in some way mediated or produced by the postischemic hippocampus (Mumby et al., 1992b). This idea was confirmed by the results of Experiment 4 when it was found that hippocampal ablation within 1 hr following ischemia prevented D N M S impairments from developing. Preoperatively trained rats that received a combined lesion were able to reacquire the nonmatching rule at a normal rate and their postsurgery scores at delays did not differ from their presurgery scores. Taken together, these findings suggest that the pathophysiological events resulting in ischemia-induced object-recognition memory impairments are mediated by the post-ischemic hippocampus and, that by eliminating these processes, the brain damage responsible for ischemia-induced D N M S deficits can be prevented. Little is known 138 about the potential mechanisms in the hippocampus that may actually be the substrates for producing damage in other brain regions. However, Experiment 5 demonstrated that the production of this extrahippocampal neuropathology is dependent on factors other than C A 1 neurotoxicity. Destroying a large portion of the CA1 prior to ischemia did not reduce subsequent D N M S acquisition deficits. It appears, then, that other hippocampal subfields (i.e., C A 2 , dentate gyrus), that also project to brain regions believed to be involved in memory, may be involved in producing the extrahippocampal neuropathology that results from ischemia. In addition to suggesting that hippocampal interference from a partial, ischemia-induced lesion is not responsible for producing D N M S deficits, the results of the present study also add to a growing body of literature that indicates that C A 1 cell loss is not the only neuropathology that results from ischemia. Furthermore, C A 1 cell loss appears not to be the only factor contributing to ischemia-induced memory impairments. This may be true regardless of the species, but has received the most empirical support in the rat. For example, the discrepancy between the mnemonic deficits produced by hippocampal ablation and cerebral ischemia are most pronounced in rats. Preoperatively trained rats subjected to ischemia are impaired both in the postoperative reacquisition of the nonmatching rule and performance at delays 4, 15, 60 120, and 300 s (Wood et al., 1993), whereas similarly trained rats that receive complete hippocampal ablation show only mild impairments. Ischemia-induced D N M S deficits are among the most severe to result from any brain manipulation, as they entail both a deficit in learning (or relearning) the nonmatching rule and a performance deficit even at the shortest delays (i.e. 4, 15 s when administered during mixed-delay sessions). CA1 lesions alone appear unlikely to be the sole source of these D N M S deficits because, as already noted, partial NMDA-induced lesions of the dorsal hippocampus, that damage much of the CA1 subfield, fail to produce a D N M S deficit in rats. If CA1 lesions are responsible for ischemia-induced D N M S deficits then it would be expected that some type of impairment would be found in rats with NMDA-lesions of the dorsal hippocampus. This was not the case. A n additional and complementary piece of evidence indicating that ischemia 139 produces neuropathology in addition to CA1 cell loss, and that this damage produces memory impairments, comes from studies which have examined the effects of ischemia on spatially-guided behavior in rats. The relationship between CA1 lesions and spatial memory deficits has recently been questioned by Nunn and Hodges (1994). In an extensive review, they concluded that impairments in spatial memory tasks following cerebral ischemia cannot be attributed solely to CA1 pyramidal cell loss. Two pieces of evidence were central to this conclusion. First, the pattern of memory deficits in the Morris water-maze is different between ischemic rats and ibotenic acid-lesioned rats that have comparable amounts of CA1 cell loss (Nunn et al., 1994). Additionally, the severity of spatial memory deficits following ischemia have not been shown to correlate reliably with the extent of CA1 necrosis (Nunn et al., 1994; Olsen et al., 1994a; 1994b). Thus, behavioral evidence in rats with both spatial and nonspatial memory tasks suggests that cerebral ischemia is producing neuropathology in addition to CA1 cell loss and that this neuropathology is contributing to the observed memory deficits. In accordance with the preceding ideas, several studies have reported mnemonic deficits in rats following brain manipulations in the absence of any observable neuropathology. For example, Lyeth, Jenkins, Hamm, Dixon, Phillips, Clifton, Young, and Hayes (1990) found impairments in radial arm maze performance in rats following traumatic brain injury that resulted in no observable neuropathology. Jaspers et al. (1990) found deficits in the water-maze following 2VO-ischemia in rats that produced no apparent C A 1 cell loss. Thus, there are known cases were unobservable brain damage produces a memory impairment. As a final point, in light of the current trend in the animal literature, it seems reasonable to speculate that extrahippocampal damage may contribute to ischemia-induced memory deficits in humans, such as patient R.B. (Zola-Morgan et al., 1986). I S C H E M I A - I N D U C E D E X T R A H I P P O C A M P A L N E U R O P A T H O L O G Y : H O W , W H E R E , A N D W H A T ? As it appears that cerebral ischemia in rats is producing neuropathology that extends beyond CA1 cell loss, the next section of this thesis addresses the possible locus of this damage. T o this end, anatomical and behavioral evidence are summarized with the goal of 140 proposing how this neuropathology may be mediated or produced by the postischemic hippocampus, which brain regions are possibly affected, and what the underlying structural and/or functional neuropathology may be. H O W M A Y T H E H I P P O C A M P U S M E D I A T E O R P R O D U C E I S C H E M I A - I N D U C E D E X T R A H I P P O C A M P A L N E U R O P A T H O L O G Y ? A mechanism by which the hippocampus mediates or produces ischemia-induced extrahippocampal damage can be postulated based on its known anatomical connections and neurochemistry. Central to the formulation of this hypothesis is the established role that the glutamatergic projections from the entorhinal cortex to the hippocampus play in producing ischemia-induced CA1 cell loss. Lesions of the entorhinal cortex (Kaplan et al., 1989) or transections of the perforant path (Wieloch, et al., 1985) eliminate the major source of excitatory input into the hippocampus and attenuate ischemia-induced CA1 cell loss, strongly implicating glutamatergic afferentation as contributing to the sensitivity of C A 1 neurons to ischemia. It is proposed here that excitatory projections (mainly glutamatergic) from the hippocampus to afferent structures may produce neuropathology in these brain regions in a manner similar to that which by glutamatergic projections from the entorhinal cortex underlie CA1 cell loss. Thus, there may be a chain of ischemia-induced neurotoxicity that begins with the glutamatergic projections from the entorhinal cortex and runs through the hippocampus to afferent brain regions. This may explain how pre- or postischemic hippocampal ablations prevents D N M S deficits, by breaking the chain of excitotoxicity and preventing neuronal injury in afferent hippocampal structures that subserve object-recognition memory. Glutamatergic projections from the hippocampus to several brain regions believed to be involved in object-recognition memory in rats are known to exist (Figure 23). The hippocampus sends glutamatergic projections to the rhinal cortex (both the perirhinal and entorhinal subdivisions), both directly and via the subicular complex (Amaral & Witter, 1995; Jones, 1993). Another such pathway connects the hippocampus with the PFC. These projections arise primarily from the temporal CA1 and subicular complex and project to the infralimbic (Swanson, 1981), prelimbic (Jay et al.,1989), and orbitofrontal (Jay & 141 Witter, 1989) subdivisions of the PFC. These pathways have been shown to use glutamate and/or aspartate as their neurotransmitters (Laroche, Jay, & Thierry, 1990). Diencephalic structures, including the anterior thalamic nuclei and the mammillary bodies of the hypothalamus, receive afferents directly from the hippocampus and by way of the subicular complex (Amaral & Witter, 1995; Lopes da Silva, 1990). Glutamate has also been found in these pathways, though the full extent of this innervation remains to be characterized (Ottersen, Hjell, Osen, & Laake, 1995; Price, 1995). Hippocampal projections to the basal forebrain arise from the CA1, CA3, and subicular complex and project primarily to the lateral septal nuclei and the nuclei of the diagonal band of Bfoca (Amaral & Witter, 1995). These pathways are rich in glutamate, especially in the lateral septum (Ottersen et al., 1995; Jakab & Leranth, 1995). Interestingly, although there is little glutamate in the reciprocal septohippocampal pathway (Ottersen et al., 1995), transections of the fimbria/fornix prior to ischemia attenuate CA1 cell loss (Buchanan & Pulsinelli, 1989), similarly to transections of the perforant path (Wieloch et al., 1985). Thus, there is some relationship between the basal forebrain and ischemia-induced neuropathology that may involve projections between the two structures. McEwen (1993) has suggested that one of the factors that may contribute to the hippocampus' vulnerability to injury is its plasticity. The hippcampal-PFC pathway has been shown to be capable of supporting L T P (Doyere, Burette, Redini-Del Negro, & Laroche, 1993). Thus, as plasticity, and the mechanisms for plasticity, have been demonstrated in afferent hippocampal pathways, it is possible that they too are particularly vulnerable to ischemic insult. Neurochemically, plasticity may be a double-edged sword; allowing for experience-dependent changes while being a potential precursor to neuropathology. The precise nature of the events in the postischemic hippocampus that mediate or produce extrahippocampal neuropathology are currently unclear. Neurotoxicity of CA1 neurons, by itself, does not appear to be the primary cause. NlVlDA-lesions that damaged much of the CA1, and presumably attenuated ischemia-induced CA1 neurotoxicity, did not attenuate D N M S impairments following ischemia. In conjunction with this, the fact that total hippocampal ablations do prevent ischemia-induced D N M S deficits (Mumby et al., 142 Figure 23. A diagram of the excitatory (mainly glutamatergic) projections from the hippocampus to brain areas believed to be involved in object-recognition memory (see text for anatomical origin of these projections within the hippocampus). 143 144 1992b) strongly suggest that processes in addition to CA1 cell death contribute to the production of extrahippocampal damage. It could be argued that the preischemic N M D A -lesions in the present study did not ablate the entire CA1 and that extrahippocampal damage from the death of the remaining cells could have been produced. Two things make this unlikely. First, the neurons in the ventral/temporal regions of the hippocampus are more resistant to ischemia than those in the dorsal/septal aspect (Schmidt-Kastner & Freund, 1991). Thus, it does not appear that ischemia produces much excitotoxicity in this region to begin with. Second, the fact that even a mild attenuation of D N M S acquisition deficits was not observed following preischemic NMDA-lesions argues against ischemia-induced C A 1 neurotoxicity as being the sole event in producing extrahippocampal damage. One process occurring in the postischemic brain that may produce widespread neuropathology is seizure activity. The temporal lobe, particularly the amygdala and hippocampus, can become focal points for seizures following a variety of neurological insults (Johnston & Brown, 1984; McEwen, 1993). Prolonged seizure activity following ischemia may occur in the hippocampus and produce damage in afferent structures that receive excitatory projections. Consistent with this idea is the fact that both amygdala-kindled rats (Gale, 1992; Wasterlain & Shirasaka, 1994) and human temporal lobe epileptics (Babb, 1991; Sisodiya, Free, Stevens, Fish, & Shorvon, 1995; Swanson, 1995) have neuropathology that extends to brain regions beyond the temporal lobe. Moreover, human epileptics often show cognitive deficits that are related to the severity and duration of their seizures (Besag, 1988; Klein, 1991). Seizures in monkeys following global ischemia (Zola-Morgan et al., 1992) or P C A occlusion (Bachevalier & Mishkin, 1989) have not been described, but they are a frequent occurrence in rodent models of ischemia (Schmidt-Kastner & Freund, 1989; Wood et al., 1993; personal observations). The degree of postischemic seizure activity may offer an explanation as to why complete hippocampal ablations prevent ischemia-induced D N M S deficits and partial lesions fail to do so. Complete lesions of the hippocampus prior to or immediately following ischemia may significantly reduce postischemic seizures and thereby prevent extrahippocampal neuropathology; partial lesions may still allow postischemic seizure activity to occur in the 145 remaining hippocampus, producing brain damage in afferent regions and concomittant memory deficits. W H E R E IS T H E P O S S I B L E A N A T O M I C A L L O C U S O F T H E E X T R A H I P P O C A M P A L D A M A G E ? One possible way to determine the locus of the ischemia-induced extrahippocampal neuropathology underlying object-recognition memory deficits is to use the D N M S paradigm as a "behavioral assay." Used in this manner, it is possible to compare the pattern of ischemia-induced D N M S deficits in rats with those that result from lesions to known structures. Central to this approach is the fact that different brain lesions produce different patterns of impairment on the D N M S task. Three separate aspects of D N M S performance can be used for analysis: 1) Is there a deficit in acquiring the nonmatching rule (in pretrained animals is there a deficit in postsurgical reacquisition)? 2) Are the rats impaired at the various delays? 3) If the rats do show impairments across delays, are these impairments delay dependent? That is, once the rats have acquired the nonmatching rule is their performance relatively normal at short delays and does it decline at long delays, which would indicate a true memory impairment; or, in contrast, do subjects display a delay-independent deficit which would suggest a more generalized impairment that may be unrelated to memory? As mentioned previously, ischemic rats (both naive and preoperatively trained) are significantly impaired in acquiring (or reacquiring) the nonmatching rule and are impaired at delays of 4, 15, 60, 120 and 300 s. By comparing this pattern of deficits, in which the locus of damage responsible for them is unknown, with the deficits produced by histologically confirmed lesions, it may guide us toward potential brain areas in which the ischemia-induced extrahippocampal damage may be occurring. RHINAL CORTEX At first glance, the close anatomical and functional relationship between the hippocampus and the rhinal cortex and the fact that the rhinal cortex is believed to be critically involved in object-recognition memory in monkeys and rats (Kornecook et al., 1995; Meunier et al., 1993; Mishkin & Murray, 1994; Mumby & Pinel, 1994; Murray, 146 1996) suggest that ischemia-induced D N M S deficits may be caused by damage to this area that has remained undetected in previous studies. A close examination of the behavioral data, however, reveals that this idea may be a bit too simplistic. Importantly, the pattern of D N M S deficits following ischemia or rhinal cortex lesions in rats is somewhat different. Preoperatively trained rats that were given rhinal cortex lesions did not display a postoperative reacquisition deficit (Mumby & Pinel, 1994), whereas ischemic rats were severely impaired in the postsurgical reacquisition of the nonmatching rule (Mumby et al., 1992b; Wood et al., 1993). Both rhinal cortex lesioned and ischemic rats were impaired at delays, however, the rhinal cortex lesioned rats were impaired at delays of 15 s or greater whereas ischemic rats were impaired even at the 4 sec delay. These discrepancies in behavioral performance indicate that, although some ischemia-induced D N M S deficits may be produced by rhinal cortex damage, additional factors must also be present to account for the full spectrum of impairments. DIENCEPHALON The diencephalic structures most frequently implicated in object-recognition memory are the mediodorsal nucleus of the thalamus and the mammillary bodies. In monkeys, lesions of various thalamic nuclei, including the M D nucleus, result in D N M S impairments (Aggleton & Mishkin, 1981; 1983; Zola-Morgan & Squire, 1985). In rats, the somewhat discrepant findings on D N M S performance make a direct comparison with ischemia-induced deficits difficult. For example, naive rats with M D nucleus lesions may be both impaired (Mumby et al., 1993) or unimpaired (Kornecook & Pinel, 1994) on the acquisition of the nonmatching rule. Additionally, preoperatively trained rats with M D lesions were impaired even at 4 s delays, similar to ischemic rats; whereas naive rats with M D nucleus lesions were impaired only at delays of 15 s or greater. This pattern is opposite to that found following lesions in the M T L : Preoperative training typically lessens the severity of D N M S impairments (Zola-Morgan & Squire, 1986; Mumby et al., 1992a; Mumby et al., 1995). Thus, although the behavioral deficits following M D nucleus lesions and ischemia are identical in preoperatively trained rats, they are somewhat different in naive rats. 147 The effect of mammillary body lesions on D N M S performance in rats has not been investigated. In monkeys, M B lesions have been shown to produce mild impairments in D N M S learning (Aggleton & Mishkin, 1985; Zola-Morgan et al., 1989). Thus, there is insufficient evidence at the present time to determine if mammillary body damage plays a role in ischemia-induced D N M S deficits. BASAL FOREBRAIN Basal forebrain lesions in monkeys have been shown to produce either significant (Irle & Marksowitch, 1987), transient (Aigner et al., 1991) or no (Voytko et al., 1994) impairment on D N M S . In rats, the D N M S deficits resulting from electrolytic lesions of the septal nuclei and the diagonal band of Broca were identical to those found following ischemia. Both manipulations produced deficits in the postsurgical reacquisition of the nonmatching rule and severe deficits in performance across delays (Kornecook et al., 1993; Mumby et al., 1992b; Wood et al., 1993). Moreover, both lesions impaired performance at very short delays (i.e., 4 s), a characteristic not typically found after lesions to the rhinal cortex, P F C . Based on the above similarities, ischemia could be postulated to result in D N M S deficits by producing neuropathology in the basal forebrain. However, whether basal forebrain lesions in rats produce "true" memory impairments is still uncertain. Several authors suggest that impairments in basal forebrain lesioned rats (Dunnet, 1991; Kornecook et al., 1993) and monkeys (Voytko, 1996; Voytoko et al., 1994) on tasks designed to assess memory may be the result of attentional deficits. Furthermore, the contribution of damage to the forebrain cholinergic system to the memory impairments in basal forebrain lesioned animals is also in question (Dunnet et al., 1991; Fibiger, 1991). Where once there appeared to be strong evidence for a role for the basal forebrain in memory, there are now many unanswered questions. PREFRONTAL CORTEX The role of the P F C in D N M S performance is somewhat unclear. In preoperatively trained monkeys, lesions of the ventromedial P F C produce an acquisition and performance deficit whereas lesions of the dorsolateral P F C have no effect (Bachevalier & Mishkin, 148 1986). In rats, lesions to the medial P F C produce no acquisition impairment and only a very mild performance deficit on the Aggleton (1985) D N M S task and no defcicits on the Mumby et al. (1990) version (Kolb et al., 1994). Shaw and Aggleton (1993) also found no impairments on D N M S following both large and small lesions of the PFC. However, Otto and Eichenbaum (1992) have shown that orbitofrontal lesions impair the acquisition of an odor-guided D N M S task, suggesting that this region participates in some type of rule learning. Failure to find an object-recognition deficit following P F C lesions in rats makes it difficult to attribute ischemia-induced D N M S impairments to neuropathology in this region. However, the functional heterogeneity of the rat prefrontal cortex is well known and different subregions of the P F C have been shown to underlie different mnemonic processes (Kolb et al., 1994; Seamans, Floresco, & Phillips, 1995). Lesions in the previously mentioned studies may have failed to produced a D N M S deficit because of the anatomical locus of the lesion within the PFC. Additional studies are required to determine if the functional heterogeneity of the P F C extends to D N M S performance. W H A T IS T H E P O S S I B L E N A T U R E O F T H E E X T R A H I P P O C A M P A L N E U R O P A T H O L O G Y ? Of all of the brain structures examined to date, the CA1 region of the hippocampus has received the most attention with regard to assessing ischemia-induced neuropathology. This is most likely due to its selective vulnerability and the relative ease with which hippocampal neuropathology can be quantified. However, the recent use of sophisticated analytical techniques is beginning to show that ischemia-induced neuropathology is not confined to the hippocampus. The following section will describe extrahippocampal neuropathology that is not detectable with conventional histological techniques (i.e., Nissl stains), the type that are used typically in experiments that attempt to correlate ischemia-induced neuropathology with behavior. This neuropathology can be broadly categorized as anatomical, molecular, or functional changes. ANATOMICAL CHANGES The majority of studies investigating the relationship between ischemia-induced 149 neuropathology and memory impairments have used Nissl stains or hematoxylin-eosin in combination with the visual counting of neurons in order to determine the locus and extent brain damage (Auer et al., 1989; Hagan & Beaughard, 1990; Kiyota et al., 1991; Jaspers et al., 1990; Mumby et al., 1992b; Volpe et al., 1984; Volpe et al., 1992; Wood et al., 1993; Zola-Morgan et al., 1992). Such analyses have reported widespread CA1 cell loss, with some patchy and inconsistent cell loss in extrahippocampal structures - the most consistent being striatal damage in the 4 V O rat model. The most revealing anatomical investigation of ischemia-induced extrahippocampal neuropathology was recently carried out by Nunn and Jarrard (1994). The impetus for this investigation was, no doubt, the conclusion reached by Nunn et al. (1994) that ischemia-induced CA1 cell loss is not the sole determinant of spatial memory deficits on the water-maze. Neuropathological analysis was carried out using the Fink-Heimer silver impregnation technique (Fink & Heimer, 1967), a technique that is very sensitive in the identification degenerating neurons. Rats were subjected to 4VO-ischemia and sacrificed after allowing 5 days for recovery. The brains were then sectioned and alternate slices were stained with either cresyl violet or silver stained. Both histological techniques revealed widespread degeneration of CA1 pyramidal cells. However, the damage revealed by cresyl violet was almost exclusively confined to the CA1 region of the hippocampus whereas silver staining revealed cell loss in several extrahippocampal structures. The most consistent damage was found in the cingulate area of the prefrontal cortex. This damage ranged from slight to 3 0 % cell loss and is a novel finding following cerebral ischemia in rats. Considerably less (10%) cell loss was also found in the dorsal perirhinal cortex. Additional, but inconsistent, neuropathology was also found in layer III of the somatosensory cortex, the dorsolateral striatum, the lateral septal nuclei, and the entorhinal cortex. Adjacent histological sections stained with cresyl violet revealed none of this extrahippocampal cell loss. Other investigators have also reported ischemia-induced extrahippocampal neuropathology using silver impregnation. Crain, Westerkam, Harrison, and Nadler (1988) found neuronal degeneration in the lateral septal nuclei, the somatosensory cortex, 150 and the striatum in gerbils. Freund and Buzsaki (1990) have reported damage in the neocortex, thalamus, neostriatum and substantia nigra following 4 V O ischemia in rats. No studies, however, have yet attempted to use silver impregnation to assess possible extrahippocampal neuropathology produced by 2VO-ischemia in rats. As 2 V O produces a slightly different pattern of brain damage than 4VO-ischemia, as demonstrated by Nissl stains, it seems likely that silver impregnation may differentiate between extrahippocampal cell loss produced by each model of ischemia. MOLECULAR CHANGES Alterations at the molecular level following ischemia have been well documented in the hippocampus. Such changes may involve the expression of immediate early genes (c-fos,c-jun), the activation of heat shock proteins, the disorganization of the neuronal cytoskeleton, and a decrease in protein synthesis (Schmidt-Kastner & Freund, 1991). Evidence that neuropathology at the molecular level is occurring in extrahippocampal brain regions following ischemia is limited at the present time. Araki, Kato, Kania, and Kogure (1993) investigated the effects of forebrain ischemia on long-term changes in second messenger systems in the gerbil brain. As would be expected, the most significant changes occurred in the hippocampus. Eight months after ischemia, autoradiography revealed changes in the binding sites for protein kinase C, inositol triphosphate, forskolin, and cyclic A M P . Importantly, however, changes in inositol triphosphate binding were also observed in the thalamus, while forskolin binding was effected in the substantia nigra and cyclic A M P binding was altered in the striatum. Thus, preliminary evidence suggests that ischemia can lead to long-term molecular changes in brain regions other than the hippocampus. FUNCTIONAL CHANGES Similar to postischemic molecular changes, functional changes in the hippocampus resulting from cerebral ischemia have received the most attention. In studies on this topic, various metabolic and electrophysiological abnormalities have been observed both in vitro and in vivo (Scmidt-Kastner & Freund, 1991). In light of the present discussion on ischemia-induced extrahippocampal brain damage, it seems parsimonious to assume that 151 any functional changes would be accompanied by an underlying neuropathology. The few studies that have looked for functional changes in extrahippocampal regions have demonstrated that ischemia can result in long-lasting alterations throughout the brain. For example, Deitrich, Ginsberg, and Busto (1986) found changes in the activation of a somatosensory circuit. Metabolic responsiveness as measured by 2-deoxyglucose autoradiography was reduced in the whisker barrel circuit of rats that were subjected to 30 min of 4VO-ischemia. They speculated that this change in metabolic activity may be due to modifications in dendritic and/or synaptic elements. This type of neuropathology would not be detectable with light microscopy. In addition to metabolic changes, extrahippocampal neurophysiological abnormalities have also been reported. Peruche, Klaasens, and Kreiglestein (1995) found a decrease in the amplitude of electrocorticograms of freely moving rats that had been subjected to 10 min of 2VO-ischemia. The cause of this deficit is unknown and the relationship between whole brain corticograms and hippocampal neuropathology is unclear at the present time. One interesting study, which examined only the hippocampus, may be useful when discussing the possibility of functional changes in the absence of observable brain damage. In this study, gerbils were given pentobarbitol as a neuroprotective agent prior to ischemia (Ishimaru, Takahashi, Ikarashi, & Maruyama, 1995). Histological analysis later revealed that this treatment prevented hippocampal C A 1 cell loss completely. However, additional assays showed that hippocampal acetylcholine levels were reduced, as was the release of acetylcholine. Thus, there were ischemia-induced changes in neurotransmitter levels in the hippocampus with no observable neuropathology. S Y N T H E S I S The findings presented in this thesis are consistent with the idea that extrahippocampal neuropathology, not CA1 cell loss, is responsible for the D N M S deficits in ischemic rats. This neuropathology is difficult to detect with conventional histological methods and appears to be mediated or produced by the postischemic hippocampus. Combined with the fact that ischemia-induced deficits on the Morris water-maze appear to be the result of more than CA1 damage (Nunn & Hodges, 1994) and that more 152 sophisticated histological techniques (i.e., silver impregnation) have revealed extrahippocampal ischemia-induced cell loss (Nunn & Jarrard, 1994), it seems reasonable to conclude that previous studies have overlooked the contribution of extrahippocampal neuropathology to ischemia-induced memory impairments. This affords a potential resolution to the paradox between the mnemonic deficits produced by either hippocampal ablation or cerebral ischemia. Based on the available anatomical and behavioral data, a tentative model can be constructed that links ischemia-induced D N M S deficits to extrahippocampal neuropathology that is mediated by the postischemic hippocampus. Athough this model is based on evidence from the rat, related behavioral results and neuroanatomical connections are similar in primates. Central to this model are the excitatory projections from hippocampus to other brain regions that participate in object-recognition memory (Figure 23). Following ischemia, excitatory projections from the hippocampus to these areas may be responsible for producing neuropathology in a manner similar to that which by the glutamatergic projections from the entorhinal cortex to the hippocampus underlie CA1 necrosis. This neuropathology may be manifested as anatomical damage, given that silver impregnation has revealed significant cell loss in extrahippocampal brain areas following 4VO-ischemia (Nunn & Jarrard, 1994) or, on the contrary, other molecular or functional changes may be responsible for, or contribute to, the D N M S deficits. Seizure activity in the post-ischemic hippocampus may set up a focus that drives this excitotoxic process. As it is estimated that nearly 90% of the synapses in the brain use amino acids as their neurotransmitter (Ottersen et al., 1995) merely the presence of glutamate in a given brain region is likely not sufficient to produce ischemia-induced neurotoxicity. However, the "delayed neuronal death" in the hippocampus and the postischemic seizure activity indicates that pathogenic processes are occurring following ischemia. It is the continued seizure activity in the postischemic hippocampus that could result in enhanced neurotoxicity and this may make the hippocampus unique in producing postischemic damage in afferent structures that are important for memory. Behavioral evidence from histologically confirmed lesions suggests several 153 potential sites for ischemia-induced extrahippocampal neuropathology to be occurring. Damage to the rhinal cortex may underlie ischemia-induced D N M S deficits at long delays, as this region is believed to maintain stimulus representations over time (Mumby & Pinel, 1994; Murray, 1996; Mishkin & Murray, 1994; Otto & Eichenbaum, 1992; Eichenbaum, et al., 1994). Additionally, ischemia-induced neuropathology in the P F C may be responsible for the impairments in the rule learning component of the D N M S task, though this is somewhat tentative in rats. This model, however, cannot account for the deficits in ischemic rats at short delays (i.e., 4 s). Such a deficit is not observed after either P F C or rhinal cortex lesions. It is possible that the presence of both lesions in combination, as may result from ischemia, would produce such a deficit, but this idea has not been tested in rats or monkeys. In some studies (Mumby et al., 1994), thalamic damage has been shown to produce deficits at very short delays. Thus, it is also a potential site for ischemia-induced neuropathology. Another possibility is that ischemic rats are impaired at short retention intervals due to nonmnemonic factors. As the pattern of deficits in ischemic rats is identical to that found after basal forebrain lesions, which may be caused by attentional deficits, ischemia may also result in an attentional deficit. This deficit may, in fact, be caused by ischemia-induced damage to the basal forebrain. Glutamatergic afferention of the lateral septal nuclei is quite extensive (Jakab & Leranth, 1995) and ischemia has been shown to result in the degeneration of these neurons in some studies (Crain et al., 1989; Nunn & Jarrard, 1994; Smith, Bendek, & Dahlgren, 1984). Finally, ischemia-induced D N M S impairments may be caused by a combination of attentional and mnemonic deficits that remain uncharacterized at the present time. I M P L I C A T I O N S F O R T H E N E U R O B I O L O G I C A L S T U D Y O F M E M O R Y The results of the present study are consistent with the hypothesis that the hippocampus does not play a critical role in object-recognition memory. A s such, these finding are in accordance with the most recent evidence rats and monkeys (Aggleton et al., 1986; Jarrard, 1993; Mumby & Pinel, 1994; Mumby et al., 1992a; Nadel, 1992; O'Boyle et al., 1993; Rothblat & Kromer, 1991). Two recent reviews (Mishkin & Murray, 1994; 154 Murray, 1996) have concluded that the role of the hippocampus in object-recognition memory in monkeys was overemphasized in early studies, due to the imprecise nature of aspiration lesions and a lack of knowledge about the cytoarchitecture and connectivity of the rhinal cortex. Accordingly, the rhinal cortex appears to be the only critical M T L structure for object-recognition memory. The hippocampus occupies a central role in many theories of learning and memory. A s previous ideas appear to be incorrect, notions about hippocampal function will need to be revised. Although there are numerous detailed theories on the mnemonic functions of the hippocampus (Eichenbaum et al., 1992; Hirsch, 1974; Jarrard, 1993; O'Keefe, & Nadel, 1978; Olton et al., 1979; Rawlins, 1985; Sutherland & Rudy, 1989) the broader question of whether the hippocampus is a general memory structure or one with a more limited function will be addressed here. One theory (Cohen & Squire, 1980; Squire, 1992; Squire & M a - M o r g a n , 1991) posits that the hippocampus is the critical brain structure mediating declarative memory. This general category of memory comprises the storage of facts, events, and relations among stimuli. A n important additional characteristic is that declarative memory is consciously accessible. Declarative memory has been referred to knowing "that" as opposed to procedural memory which is deemed knowing "how" (Cohen & Squire, 1980). In animals, object-recognition memory tasks are believed to tap into mnemonic processes analogous to human declarative memory. This idea is supported by the fact that human amnesics perform poorly on D N M S (Squire et al., 1988). As the hippocampus is no longer believed to be involved critically in object-recognition memory, work with animals suggests that it can no longer be considered a general "declarative memory structure," such that damage to it would produce a universal impairment in all types of declarative memory. Recent evidence is consistent with a more specialized role for the hippocampus in memory. One aspect of hippocampal function that has received the greatest amount of support is its role in spatial memory. Hippocampal involvement in spatial memory is well documented in rats (O' Keefe & Nadel 1978; Jarrard, 1993) but similar results have also been obtained in monkeys (Parkinson, Murray, & Mishkin, 1988) and humans (Cave & 155 Squire, 1991). Nadel (1991; 1992) has suggested that the hippocampus may act as a "spatial module" in terms of memory processing. The findings in this thesis are in agreement with a specialized role for the hippocampus in memory. Partial hippocampal lesions produced a deficit in a spatial memory task (i.e., the Morris water-maze) while leaving other forms of declarative memory (i.e., performance on D N M S ) intact. Another possible role for the hippocampus has to do with it somehow being involved in long-term retention of information. Rawlins (1985) has suggested that the hippocampus acts as type of temporary memory store or buffer. This idea is consistent with the D N M S impairments seen at very long delays in monkeys (Alvarez et al., 1995) and rats (Mumby et al., 1992a) and with the retrograde amnesia that results from hippocampal damage (Squire, 1992). Moreover, Vnek and Rothblat (1996) have recently demonstrated long-term deficits in the retention of three concurrent object discriminations in rats with lesions to the dorsal hippocampus. Initial levels of performance between lesioned and control groups were equal; however, when tested 3 weeks later, the rats with hippocampal lesions were impaired on the retention of the object discrimination problems, relative to control rats. The normal initial performance in combination with the impairments after a delay of 3 weeks led the authors to conclude that the hippocampus may be involved in the retention of information for weeks to, possibly, years. Murray (1996) recently pointed out that the M T L can no longer be viewed as a single functional unit due to the fact that damage to discrete regions leads to qualitatively different types of mnemonic impairments. Thus, the hippocampus can be seen as being one component of a larger medial temporal lobe memory system. A component that, most likely, performs a limited subset of "declarative type" functions (Nadel, 1992). Structures in the M T L themselves, however, appear to be part of a larger memory system (Mishkin & Murray, 1994) comprising diverse brain regions such as the diencephalon, P F C , and basal forebrain. Identification of the critical brain structures responsible for ischemia-induced object-recognition deficits may provide important clues as to the true nature of this extensive memory system. 156 REFERENCES Adams, J. H . , Bierly, J. B., Connor, R. C. T., & Triep, C. (1966). The effects systemic hypotension upon the human brain. Brain, 89, 135-154. Aggleton, J. P. (1985). One-trial object recognition by rats. Quarterly Journal of Experimental Psychology. 37. 279-294. Aggleton, J. P., Blindt, H . S., & Rawlins, J. N. P. (1989). Effects of amygdaloid and amygdaloid-hippocampal lesions on object recognition memory and spatial working memory in rats. Behavioral Neuroscience, 103, 962-974. Aggleton, J. P., Hunt, P. R., & Rawlins, J. N . P. (1986). The effects of hippocampal lesions upon spatial and non-spatial tests of working memory. Behavioral Brain Research. 19. 133-146. Aggleton, J. P. & Mishkin, M . (1983). Visual recognition impairment following medial thalamic lesions in monkeys. Neuropsychologia, 21, 189-197. Aggleton, J. P. & Mishkin, M . (1985). Mammillary-body lesions and visual recognition in monkeys. Experimental Brain Research, 58. 190-197. Aigner, T. G. & Mishkin, M . (1986). The effect of physostigmine and scopolamine on recognition memory in monkeys. Behavioral and Neural Biology, 45, 81-87. Aigner, T. G., Mitchell. S. J . , Aggleton, J. P., Delong, M . R., Struble, R. G., Price, D. L., Wenk, G L., Pettigrew, K. D., & Mishkin, M . (1991). Transient impairment of recognition memory following ibotenic-acid lesions of the basal forebrain in macaques. Experimental Brain Research, 86, 18-26. Alvarez-Royo, P., Clower, R. P., Zola-Morgan, S., & Squire, L . R. (1991). Stereotaxic lesions of the hippocampus in the monkey: Determination of the surgical coordinates and analysis of lesions using magnetic resonance imaging. Journal of Neuroscience Methods, 38. 223-232. Alvarez, P., Zola-Morgan, S., & Squire, L. R. (1995). Damage limited to the hippocampal region produces long-lasting memory impairments in monkeys. Journal of Neuroscience. 15. 3796-3807. Amaral, D. G. & Witter, M . P. (1989). The three-dimensional organization of the hippocampal formation: A review of anatomical data. Neuroscience. 3. 571-591. Amaral, D. G. & Witter, M . P. (1995). Hippocampal Formation. In G. Paxinos (Ed.), The rat nervous system (pp. 443-493) San Diego: Academic Press. Andersen, P., Bliss, T. V. P., & Skrede, K. K. (1971). Lamellar organization of the hippocampal excitatory pathways. Experimental Brain Research, 13, 222-228. Araki, T . , Kato, H . , Kanai, Y. , & Kogure, K. (1993). Postischemic changes of intracellular second messengers in the gerbil brain after long-term survivial: A n 157 autoradiographic study. Neuroscience, 53, 829-836. Auer, R. N . , Jensen, M . L., & Whishaw, I. Q. (1989). Neurobehavioral deficit due to ischemic brain damage limited to half of the CA1 sector of the hippocampus. Journal of Neuroscience, 9, 1641-1647. Babb, T. (1991). Bilateral damage in temporal lobe epilepsy. Canadian Journal of Neurological Science, 18, 645-648. Bachevalier, J. & Mishkin, M . (1986). Visual recognition impairment follows ventromedial but not dorsolateral prefrontal lesions in monkeys. Behavioral Brain Research. 20. 249-261. Bachevalier, J. & Mishkin, M . (1989). Mnemonic and neuropathological effects of occluding the posterior cerebral artery in Macaca mullata. Neuropsychologia, 27, 83-105. Barbizet, J . , Degos, J. D., Louarn, R , Nguyen, J. P., & Mas, J. L. (1981). Amnesia from bilateral ischemic lesions of the thalamus. Review of Neurology, 137, 415-424. Beason-Held, L. , Rosen, D. L. , & Moss., M . B. (1993). Memory deficits associated with ibotenic acid lesions of the hippocampal formation in rhesus monkeys. Society for Neuroscience Abstracts, 19, 438. Bekhterev, V. M . (1900). Demonstration eines gehirns mit zerstorung der vorderen und inneren theile der hirnrinde beider schlafenlappen. Neurol. Zentbl. 19, 900-991. Benveniste, H . , Drejer, J. , Schousbe, A . , & Diemer, N. H . (1984). Elevation of the extracellular concentrations of glutumate and aspartate in rat hippocampus during transient cerebral ischemia monitored by intracerebral microdialysis. Journal of Neurochemistry, 43, 1369-1374. Berlyne, N, & Strachan, M . (1968). Neuropsychiatry sequelae of attempted hanging. British Journal of Psychiatry, 114, 411-422. Besag, F. M . C. (1988). Cognitive deterioration in children with epilepsy. In M . R. Trimble and E. H . Reynolds (Eds.), Epilepsy, Behavior, and Cognitive Fucntion (pp. 113-127). New York: Wiley. Block, F., Jaspers, R. M . A . , Heim, C, & Sontag, K. H . (1990). S-Emopamil ameliorates ischemic brain damage in rats: Histological and behavioral approaches. Life Sciences, 47, 1511-1518 Bondi, M . W. & Kaszniak, W. (1991). Implicit and explicit memory in Alzheimer's disease and Parkinson's disease. Journal of Clinical and Experimental Neuropsychology, 13, 339-358. Brierly, J. B. & Graham, D. L. (1984). Hypoxia and vascular disorders of the central nervous system. In J. H . Adams, J. Corsellis, & L. W. Duchen (Eds.), Greenfield's Neuropathology (pp. 125-208). New York: Wiley. 158 Brooks, D. N. & Baddeley, A . D. (1976). What can amnesic patients learn? Neuropsychologia, 14, 111-122. Buchanan, A . M . & Pulsinelli, W. A . (1989). Fimbria-fornix lesions: The temporal profile for the protection against ischemic injury. Journal of Cerebral Blood Flow and Metabolism. 9, 749. Bussey, T. J. , Muir, J. L. , & Robbins, T. W. (1994). A novel automated touch screen procedure for assessing learning in the rat using computer graphic stimuli. Neuroscience Research Communications. 15. 103-110. Butters, N . , Wolfe, J . , Granholm, E . , & Cermak, L. S. (1985). Memory disorders associated with Huntington's disease: Verbal recall, verbal recognition, and procedural memory. Neuropsychologia, 23, 729-743. Cave, C. B. & Squire, L. R. (1991). Equivalent impairment of spatial and nonspatial memory following damage to the human hippocampus. Hippocampus, 1, 329-340. Charness, M . E. & DeLapaz, R. L. (1987). Mammillary body atrophy in Wernicke's encephalopathy: Antemortem identification using magnetic resonance imaging. Annals of Neurology. 22. 595-600. Chavkin, C , Shoemaker, W. J. , McGinty, J. F., Bayon, A . , & Bloom, F. E . (1985). Characterization of the prodynorphin and proenkephalin neuropeptide system in rat hippocampus. Journal of Neuroscience, 5, 808-816. Choi, D. (1988). Calcium-mediated neurotoxicity: Relationship to specific channel types and role in ischemic damage. Trends in Neuroscience, 10, 465-469. Clayton, N. S. (1995). The neuroethological development of food-storing in birds: A case of use it, or lose it! Behavioral Brain Research, 70, 95-102. Clayton, N. S. & Krebs, J. R. (1994). Hippocampal growth and attrition in birds affected by experience. Proceeding of the National Academy of Science, 91, 7410-7414. Clower, R., Alvarez-Royo, P., Zola-Morgan, S., & Squire, L. R. (1991). Recognition memory impairments in monkeys with selective hippocampal lesions. Society for Neuroscience Abstracts. 17, 338. Cohen, N. J. (1984). Preserved learning capacity in amnesia: Evidence for multiple memory systems. In L. R. Squire and N. Butters (Eds.), The Neuropsychology of Memory (pp. 83-103). New York: Guilford Press. Cohen, N. J. & Squire, L. R. (1980). Preserved learning and retention of pattern analyzing skills in amnesia: Dissociation of knowing how and knowing that. Science. 210. 207-210. Corkin, S. (1965). Tactually guided maze-learning in man: Effects of unilateral cortical excisions and bilateral hippocampal lesions. Neuropsychologia, 3, 339-351. 159 Corkin, S. (1968). Acquisition of a motor skill after bilateral medial-temporal lobe excision. Neuropsychologia. 6, 255-265. Corkin, S., Amaral, D. G., Johnson, K. A . , & Hyman, B. T . (1994). H . M . ' s MRI scan shows sparing of the posterior half of the hippocampus and parahippocampal gyrus. Society for Neuroscience Abstracts, 20. 876. Correll, R. E . & Scoville, W. B. (1965). Performance on delayed match following lesions of medial temporal lobe structures. Journal of Comparative and Physiological Psychology, 60, 360-367. Crain, B.J., Westerkam, W. D., Harrison, A . H . , & Nadler, J. V . (1988). Selective neuronal death after transient forebrain ischemia in the mongolian gerbil: A silver impregnation study. Neuroscience, 27,387-402. Damasio, A . R., Eslinger, P. J . , Damasio, H . , Van Hoesen, G. W . & Correl, S. (1985a). Multimodal amnesic syndrome following bilateral temporal and basal forebrain damage. Archives of Neurology, 42, 252-259. Damasio, A . R., Graff-Radford, N. R., Eslinger, P. J . , Damasio, H . , & Kassel, N. (1985b). Amnesia following basal forebrain lesions. Archives of Neurology, 42, 263-270. Davis, H . P., & Volpe, B. T. (1990). Memory performance after ischemic or neurotoxin damage of the hippocampus. In L. R. Squire & E. Lindenlaub (Eds.), The biology of memory (pp. 477-507). New York: Schattauer. Degirolami, V . , Haas, M . L., & Richardson, E. P. (1974). Subacute diencephalic angioencephalography: A clinicopathological case study. Journal of Neurological Sciences, 22, 197-210. Diemer, N. H . , Johansen, F. F., & Jorgensen, M . B. (1990). N-methyl-D-aspartate and non N-methyl-D-aspartate antagonists in global cerebral ischemia. Stroke, 21, 39-42. Dietrich, W. D., Ginsberg, M . D., & Bustro, R. (1986). Effect of transient cerebral ischemia on metabolic activation of a somatosensory circuit. Journal of Cerebral BloodFlow and Metabolism, 6, 405-413. D i Lauro, A . R., Schmid, R. W., & Meek, J. L. (1981). Is aspartic acid the neurotransmitter of the perforant pathway? Brain Research, 207,476-480. Douglas, R. J. (1967). The hippocampus and behavior. Psychological Bulletin, 67. 416-442. Doyere, V. , Burette, E , Redini-Del-Negro, C , & Laroche, S. (1993). Long-term potentiation of hippocampal afferents efferents to prefrontal cortex: Implications for learning and memory. Neuropsychologia, 31, 1031-1053. Drachman, D. A . & Arbit, J. (1966). Memory and the hippocampal complex. Archives of Neurology, 15,52-61. 160 Drachman, D. A . & Leavitt, J. (1974). Human memory and the cholinergic system. Archives of Neurology, 30, 113-121. Drejer, J. , Benveniste, H . , Diemer, N. H . , & Schousboe, A . (1985). Cellular origin of ischemia-induced glutamate release from brain tissue in vivo and in vitro. Journal of Neurochemistry, 45, 145-151. Dunnett, S. B. (1985). Comparative effects of cholinergic drugs and lesions of the nucleus basalis or fimbria-fornix on delayed matching in rats. Psychopharmaclogia, 87, 357-363. Dunnett, S. B., Everitt, B. J. , & Robbins, T. W . (1991). The basal forebrain-cortical cholinergic system: interpreting the functional consequences of excitotoxic lesions. Trends in Neuroscience, 14, 494-501. Dutar, P., Bassant, M . , Senut, M . , & Lamour, Y . (1995). The septohippocampal pathway: Structure and function of the central cholinergic system. Neuroscience, 75, 393-427. Epstein, P. S., Pisani, V. D., & Fawcett, J. A . (1977). Alcoholism and Cerebral Atrophy. Alcohol: Clinical and Experimental Research, 1, 61-65. Everitt, B. J. , Robbins, T. W., Evenden, J. L. , Jones, G. H . , & Sirkia, T . E . (1987). Effects of excitotoxic lesions of the substantia innominata, ventral and dorsal globus pallidus on the acquisition and retention of a conditional visual discrimination; implications for a cholinergic hypothesis of learning and memory. Neuroscience, 22, 441-469. Fibiger, H . C. (1991). Cholinergic mechanisms in learning, memory, and dementia: a review of recent evidence. Trends in Neuroscience, 14,220-223. Fink, R. P. & Heimer, L. (1967). Two methods for selective silver impregnation of degenerating axons and their synaptic endings in the central nervous system. Brain Research, 4, 369-374. Fox, J. H . , Ramsey, R. G., Huckman, M . S., & Proske, A . E . (1976). Cerebral ventricular enlargement: chronic alcoholics examined by computerized tomography. Journal of the American Medical Association, 236,365-368. Freden, K., Steengaard-Pedersen, K., & Larsson, L. I. (1984). Localization of enkephalin and cholecystokinin immunoreactivities in the perforant path terminal fields of the hippocampal formation. Brain Research, 304, 255-263. Freund, T. F., Buzsaki, G., Leon, A . , & Somogyi, P. (1990). Hippocampal cell death following ischemia: Effects of brain temperature and anesthesia. Experimental Neurology, 108. 251-260. Gabrieli, J. D. E. & Keane, M . (1988). Priming in the patient H . M . : New findings and a theory of intact in impaired priming in patients with memory disorders. Society for Neuroscience Abstracts, 14, 1290. Gabrieli, J. D. E , Milberg, W., Keane, M . , and Corkin, S. (1990). Intact 161 priming of patterns despite impaired memory. Neuropyschologia, 28. 417-427. Gaffan, D. (1974). Recognition impaired and association intact in the memory of monkeys after transection of the formix. Journal of Comparative and Physiological Psychology, 86. 1100-1109. Gale, K. (1992). Subcortical structures and pathways involved in convulsive seizure generation. Journal of Clinical Neurophysiology, 9, 264-277. Grotta, J. C , Pettigrew, L. C , Rosenbaum, D., Reid, C , Rhoades, H . , & McCandles, D. (1988). Efficacy and mechanism of action of a calcium channel blocker after global cerebral ischemia in rats. Stroke, 19, 447-454. Goldman, P. S. & Rosvold, H . E. (1970). Localization of function within the dorsolateral prefrontal cortex of the rhesus monkey. Experimental Neurlogy, 27, 291-304. Goldman, P. S., Rosvold, H . E . , Vest, B., & Galkin, T. W. (1971). Analysis of the delayed-alternation deficit produced by dorsolateral prefrontal lesions in the rhesus monkey. Journal of Comparative and Physiological Psychology, 77, 212-220. Goldman-Rakic, P. S. (1995). Cellular basis of working memory. Neuron, 14, 447-485. Graf, P. and Schacter, D. L. (1985). Implicit and explicit memory for new associations in normal and amnesic subjects. Journal of Experimental Psychology: Learning, Memory, and Cognition, 11, 501-508. Hagan, J. J. & Beaughard, M . (1990). The effects of forebrain ischemia on spatial learning. Behavioral Brain Research, 41, 151-160. Healy, S. D. & Krebs, J. R. (1989). Food storing and the hippocampus in corvids: Amount and volume are correlated. Proceedings of the Royal Society Bulletin, 248, 241-245. Heindel, W . C , Salmon, D. P., & Butters, N. (1990). Pictorial priming and cued recall in Alzheimer's and Huntington's disease. Brain and Cognition, 13, 282-295. Heindel, W. C , Salmon, D. P., Shults, C. W., Walicke, P. A . & Butters, B. (1989). Neuropsychological evidence for multiple implicit memory systems: a comparison of Alzheimer's, Huntington's and Parkinson's disease patients. Journal of Neuroscience, 9, 582-587. Hermann, B. P., Wyler, A . R., Somes, G., Dohan, F. C , Jr., Berry A . D. & Clement, L. (1994). Declarative memory following anterior temporal lobectomy in humans. Behavioral Neuroscience, 108. 3-10. Herremans, A . H . J . , Hijzen, T. H . , Slangen, J. L. (1995). The object delayed non-matching to sample task in rats does not depend on working memory. Neuroreport, 6, 1963-1965. Hirsch, R. (1974). The hippocampus and contextual retrieval of information from memory: A theory. Behavioral Biology, 12, 421-444. 162 Honig, W . K. (1978). Studies of working memory in the pigeon. In S. H . Hulse, H . Fowler, & W. K. Honig (Eds.) Cognitive processes in animal behavior (pp. 211-248). Hillsdale, NJ: Erlbaum. Horel. J. A . (1978). A critique of the hippocampal memory hypothesis. Brain, 101. 403-445. Hull. C. L. (1943). Principles of behavior. New York: Appleton. Irle, E . & Markowitsch, H . J. (1987). Basal forebrain-lesioned monkeys are severely impaired in tasks of association and recognition memory. Annals of Neurology. 22. 735-743. Ishimaru, H . , Takahashi, A . , Ikarashi, Y . , & Maryuma, Y . (1995). Pentobarbitol protects against CA1 pyramidal cell death but not dysfunction of hippocampal cholinergic neurons following transient ischemia. Brain Research, 673, 112-118. Ishizuka, N . , Weber, J. , & Amaral, D. G. (1991). Organization of intrahippocampal projections originating from CA3 pyramidal cells in the rat. The Journal of Comparative Neurology, 295, 580-623. Isaacson, R. L. , Douglas, R. J. & Moore, R. Y . (1961). The effect of radical hippocampal ablation on acquisition of avoidance response. Journal of Comparative and Physiological Psychology, 54, 625-635. Iversen, S. (1976). Do hippocampal lesions produce amnesia in animals? In C. C. Pfeiffer and J. R. Smythies (Eds.), International Review of Neurobiology V o l . 19 (pp. 1-49). New York: Academic Press. Jarrard, L. E. (1993). On the role of the hippocampus in learning and memory in the rat. Behavioral and Neural Biology, 60,9-26. Jaspers, R. M . A . , Block, F., Heim, C , & Sontag, K. H . (1990). Spatial learning is affected by transient occlusion of common carotid arteries (2VO): comparison of behavioural and histopathological changes after 2 V O and four-vessel occlusion in rats. Neuroscience letters, 117, 149-153. Jakab, R.L. & Leranth, C. (1995). Septum. In G. Paxinos (Ed.), The rat nervous system (pp. 405-442) San Diego: Academic Press. Jay, T. M . , Glowinski, J., & Thierry, A . M . (1989). Selectivity of the hippocampal projection to the prelimbic area of the prefrontal cortex of the rat. Brain Research, 505, 337-340. Jay, T. M . & Witter, M . P. (1991). Distribution of hippocampal C A 1 and subicular efferents in the prefrontal cortex of the rat studied by means of anterograde transport by Phaseolus vulgaris-leucoagglutinin. Journal of Comparative Neurology, 313, 574-586. Jernigan, T. L. , Schafer, K., Butters, N. & Cermak, L. S. (1991). Magnetic resonance imaging of alcoholic Korsakoff patients. Neuropsychopharmacology, 4, 163 175-186. Johnston, D., & Brown, T. H . (1984). The synaptic nature of the paroxysmal depolarizing shift in hippocampal neurons. Annals of Neurology, 16, 65-71. Jones, R. S. G. (1993). Entorhinal-hippocampal connections: A speculative view of their function. Trends in Neuroscience, 16, 58-64. Kaplan, T. M . , Lasner, T. M . , Nadler, J. V . , & Crain, B. J. (1989). Lesions of excitatory pathways reduce hippocampal cell death after transient forebrain ischemia in the gerbil. Acta Neuropathology, 78, 283-290. Kartsounis, L. D., Rudge, P., & Stevens, J. M . (1995). Bilateral lesions of the C A 1 and C A 2 fields of the hippocampus are sufficient to cause a severe amnesic syndrome in humans. Journal of Neurology, Neurosurgery, & Psychiatry, 59, 95-98. Kimble, D. P. (1963). The effects of bilateral hippocampal lesions in rats. Journal of Comparative and Physiological Psychology, 56, 273-283. Kirino, T. (1982). Delayed neuronal death in the gerbil hippocampus following ischemia. Brain Research, 239, 57-69. Kiyota, Y . , Miyamato, M . , & Nagaoka, A . (1991). Relationship between brain damage and memory impairments in rats exposed to transient forebrain ischemia. Brain Research. 538, 295-302. Klein, S. K. (1991). Cognitive factors and learning disabilities in children with epilepsy. In O. Devinsky and W. H . Theodore (Eds.), Frontiers of Clinical Neuroscience, Vol 12: Epilepsy and Behavior (pp. 171-179). New York: Wiley Liss. Knowles, W. D. (1992). Normal anatomy and neurophysiology of the hippocampal formation. Journal of Clinical Neurophysiology, 9, 252-263. Knowlton, B. J . , Wenk, G. L. , Olton, D. S., & Coyle, J. T. (1985). Basal forebrain lesions produce a dissociation of trial-dependent and trial-independent memory performance. Brain Research, 345, 315-321. Kolb, B. (1990). Animal models of human PFC-related disorders. In H . B. M . Uylings, C. G. Van Eden, J. P. C. De Bruin, M . A . Corner, & M . G. P. Feenstra (Eds.), Progress in Brain Research, 85, (pp. 501-519). Amsterdam: Elsevier. Kolb, B., Buhrman, K., MacDonald, R., & Sutherland, R. J. (1994). Dissociation of the medial prefrontal, posterior parietal, and posterior temporal cortex for spatial navigation and recognition memory in the rat. Cerebral Cortex, 4, 664-680. Kornecook, T . J . , Kippin, T. E . , Pinel, J. P. J . , & Wood, E . R. (1993). Impairment of reacquisition and performance of delayed nonmatching-to-sample following lesions to the medial septum and diagonal band in rats. Society for Neuroscience Abstracts. 19. 1233. Kornecook, T. J . , L u i , M . , Duva, C. A . , Anzarut, A . , & Pinel, J. P. J. (1995). Effects of perirhinal or rhinal cortex lesions on object-memory tasks in the rat. Society for 164 Neuroscience Abstracts, 21, 1935. Kornecook, T. J. & Pinel, J. P. J. (1994). Assessing the performance of rats with lesions to the mediodorsal nucleus of the thalamus on a battery of nonspatial memory tasks. Society for Neuroscience Abstracts, 20, 1213. Krebs, J. R., Sherry, D. F., Healy, S. D., Perry, V . H . , & Vaccarino, A . L. (1989) . Hippocampal specialization of food storing birds. Proceedings of the National Academy of Science. 86. 1388-1392. Kuroiwa, T. & Okeda, R. (1994). Neuropathology of cerebral ischemia and hypoxia: Recent advances in experimental studies on its pathogenesis. Pathology International, 44. 171-181. LeDoux, J. E. (1993). Emotional memory systems in the brain. Behavioral Brain Research. 58. 69-79. Lingenhohl, K. & Finch, D. (1991). Morphological characterization of rat entorhinal neurons in vivo: Soma-dendritic structure and axonal domains. Experimental Brain Research, 84. 57-74. Lopes da Silva, F. H . , Witter, M . P., Boeijinga, P. H . , & Lohman, A . H . M . (1990) . Anatomical organization and physiology of the limbic cortex. Physiological Reviews, 70. 453-511. Lorento de No, R. (1934). Studies on the structure of the cerebral cortex. II. Continuation of the study of the ammonic system. Psychology & Neurology, 46, 113-177. L o u , H . C . , Edvinsson, L . , & MacKenzie, E.T. (1987). The concept of coupling blood flow to brain function: Revision required? Annals of Neurology, 22,289-297. Luria, A . R. & Homskaya, E. D. (1964). Disturbances in regulative role of speech with frontal lobe lesions In J. M . Warren & K. Akert (Eds.), The frontal granular cortex and behavior. New York: MacGraw-Hill. Lyeth, B. G., Jenkins, L. W., Hamm, R. J. , Dixon, C. E . , Phillips, L . L . , Clifton, G. L. , Young, H . F., & Hayes, R. L. (1990). Prolonged memory impairment in the absence of hippocampal cell death following traumatic brain injury in the rat. Brain Research. 526. 249-258. Mahut, H . , Zola-Morgan, S., & Moss, M . (1982). Hippocampal resections impair associative learning and recognition memory in the monkey. Journal of Neuroscience, 2, 1214-1229. Mair, W. G. P., Warrington, E. K., & Weiskrantz, L. (1979). Memory disorder in Korsakoff's psychosis. A neuropathological and neuropyschological investigation of two cases. Brain , 102, 749-783. Martone, M . , Butters, N . , Payne, M . , Becker, J. T. , & Sax, D. S. (1984). Dissociations between skill learning and verbal recognition in amnesia and dementia. Archives of Neurology, 41, 965-970. 165 Mathews, D. A . , Cotman, C , & Lynch, G. (1976). A n electron microscope study of lesion-induced synaptogenesis in the dentate gyrus of the adult rat. I. Magnitude and time course of degeneration. Brain Research, 115, 1-21. Mayes, A . R., Meudell, P. R., Mann. D., & Pickering, A . (1988). Location of lesions in Korsakoff's syndrome: Neuropsychological and neuropathological data on two patients. Cortex, 24, 367-388. McAlona, G. M . , Dawson, G. R., Wilkinson, L. O., Robbins, T . W . , & Everitt, B. J. (1995). The effects of AMPA-induced lesions of the medial septum and vertical limb of the diagonal band of Broca on spatial delayed non-matching to sample and spatial learning in the water maze, European Journal of Neuroscience, 7, 1034-1049. McEntee, W. J. , Biber, M . P., Perl, D. P., & Benson, D. F. (1976). Diencephalic amnesia: A reappraisal. Journal of Neurology, Neurosurgery, and Psychiatry, 39, 436-441. McEwen, B. S. (1994). The plasticity of the hippocampus is the reason for its vulnerability. Seminars in the Neurosciences, 6, 239-246. McNeill , D. L . , Tidmarsh, D., & Rostall, M . L. (1965). A case of amnesic syndrome following cardiac arrest. British Journal of Psychiatry, 111, 697-699. Meunier, M . , Bachevalier, J., Mishkin, M . , & Murray, E. A . (1993). Effects on visual recognition of combined and separate ablations of the entorhinal and perirhinal cortex in rhesus monkeys. Journal of Neuroscience, 13, 5418-5432. Mills, R. P. & Swanson, P. D. (1978). Vertical occulomotor apraxia and memory loss. Annals of Neurology, 4, 149-153. Milner, B. (1962). Les trouble de la memiore accompagnant des lesions hippocampiques bilaterales. In Physiologie de l'hippocampe, 257-252. Centre National de la Recherche Scientifique, Paris. Milner, B. (1966) Amnesia following operations on the temporal lobes. In C. W. M . Whitty & O. L. Zangwill (Eds.), Amnesia (pp. 109-103). London: Butterworth. Milner, B. (1972). Disorders of learning and memory after temporal lobe lesions in man. Clinical Neurosurgery, 19,421-426. Milner, B., Corkin, S. & Tueber, H . L. (1968). Further analysis of the hippocampal amnesic syndrome: 14-year follow-up study of H . M . Neuropsychologia, 6, 215-234. Mishkin, M . (1978) Memory in monkeys severely impaired by combined but not by separate removal of amygdala and hippocampus. Nature, 273, 297-298. Mishkin, M . , Malamut, B., and Bachevalier, J. (1984). Memories and habits: T w o neural systems. In G. Lynch, J. L. McGaugh, and N. M . Weinberger (Eds.), Neurobiology of Learning and Memory (pp. 65-77). New York: Guilford Press. 166 Mishkin, M & Murray, E. A . (1994). Stimulus recognition. Current Opinion in Neurobiology. 4. 200-206 Moncada, C , Lekieffre, D., Arvin, B., & Meldrum, B. S. (1992). Effect of N O synthase inhibition on N M D A - and ischemia-induced hippocampal lesions. Neuroreport. 3, 530-532. Morris, M . K., Bowers, D., Chatterjee, & Heilman, K. M . (1992). Amnesia following a discrete basal forebrain lesion. Brain, 115, 1827-1847. Morris, R. G. M . , Garrud, P., Rawlins, J. N. P., & O'Keefe, J. (1982). Place navigation impaired in rats with hippocampal lesions. Nature. 297,681-683. Morrison, 3. H . , Benoit, R. , Magistretti, P. J . , Ling, N . , & Bloom, F. (1982). Immunocytochemical distribution of prosomatostatin-related peptides in the hippocampus. Neuroscience Letters, 34, 137-142. Mudrick, L . A . , Leung, P. P. H . , Baimbridge, K. G., & Miller, J.J. (1988). Neuronal transplants used in the repair of acute ischemic injury in the CNS. Progress in Brain Research. 74. 87-93. Mumby, D. G. , Kornecook, T. J. , Wood, E . R., & Pinel, J. P. J. (1995). The role of experimenter-odor cues in the performance of object-memory tasks by rats. Animal Learning & Behavior. 23. 447-453. Mumby, D. G. & Pinel, J. P. J. (1993) Rhinal cortex lesions and object recognition in rats. Behavioral Neuroscience, 108, 11-18. Mumby, D. G. , Pinel, J. P. J . , & Dastur, F. N. (1993). Mediodorsal thalamic lesions and object recognition in rats. Psychobiology, 21, 27-36. Mumby, D. G. , Pinel, J. P. J. , Kornecook, T. J. , Shen, M . J. , & Redila, V . A . (1995). Memory deficits following lesions of the hippocampus or amygdala in rat: Assessment by an object-memory test battery. Psychobiology, 23, 26-36. Mumby, D. G., Pinel, J. P. J . , & Wood, E. R. (1990). Nonrecurring-items delayed nonmatching-to-sample in rats: A new paradigm for testing nonspatial working memory. Psychobiology, 18, 321-326. Mumby, D. G., Wood, E. R., & Pinel, J. P. J. (1992a). Object-recognition memory is only mildly impaired in rats with lesions of the hippocampus and amygdala. Psychobiology. 20. 18-27. Mumby, D. G., Wood, E. R., Buday, M . E . , Bussey, T. J . , Pinel, J. P. J . , & Phillips, A . G. (1992b). Paradoxical effects of hippocampal lesions in rats: Ablation combined hernia-induced damage. Poster presented at Conference on the Neurobiology of Learning and Memory, Irvine Ca. Muramoto, O., Kuru, Y . , Sugishita, M . , & Toyokura, Y . (1979). Pure memory loss with hippocampal lesions. Archives of Neurology, 36, 54-56. Murray, E . A . (1992). Medial temporal lobe structures contributing to recognition 167 memory: The amygdaloid complex versus the rhinal cortex. In J. P Aggleton (Ed.), The amygdala: neurobiological aspects of emotion, memory, and mental dysfunction (pp.453-470). New York: Wiley-Liss. Murray, E. A . (1996). What have ablation studies told us about the neural substrates of stimulus memory? Seminars in the Neurosciences, 8,13-22. Murray, E. A . & Mishkin, M . (1984). Severe tactual as well as visual memory deficits follow combined removal of the amygdala and hippocampus in monkeys. The Journal of Neuroscience, 4, 2565-2580. Nadel, L. (1991). The hippocampus and space revisited. Hippocampus, 1, 221-229. Nadel, L. (1992). Multiple memory systems: What and why. Journal of Cognitive Neuroscience, 4, 179-187. Nunn, J. & Hodges, H . (1994). Cognitive deficits induced by global cerebral ischaemia: Relationship to brain damage and reversal by transplants. Behavioral Brain Research, 65, 1-31. Nunn, J. A . , LePeillet, E . , Netto, C. A . , Hodges, H . , Gray, J. A . , & Meldrum, B. S. (1994). Global Ischaemia: Hippocampal pathology and spatial deficits in the water-maze. Behavioral Brain Research, 62,41-54. O'Boyle, J. R., Murray, E. A . , & Mishkin, M . (1993). Effects of excitotoxic amygdalo-hippocampal lesions on visual recognition in rhesus monkeys. Society for Neuroscience Abstracts, 19, 438. O'Keefe, J , & Nadel, L. (1978). The hippocampus as a cognitive map. London: Oxford University Press. Olney, J. W. (1978). Neurotoxicity of excitatory amino acids. In E . G . McGeer, J. W. Olney, & P. L. McGeer (Eds.), Kainic acid as a tool in neurobiology New York: Raven. Olsen, G . M . , Scheel-Kruger, J. , Moller, A . , & Jensen, L . H . (1994a). Relation of spatial learning in rats in the Morris water maze task to the number of viable C A 1 neurons following four-vessel occlusion. Behavioral Neuroscience, 108, 681-690. Olsen, G. M . , Scheel-Kruger, J., Moller, A . , & Jensen, L. H . (1994b). Does neuronal damage of C A 1 relate to spatial memory performance of rats subject to transient forebrain ischemia? Acta Neurologica Scandinavica, 89, 204-209. Olton, D. S. & Feustle, W. A . (1981). Hippocampal function required for nonspatial working memory. Experimental Brain Research, 41, 380-389. Olton, D. S., Becker, J. T. , & Handelmann, G. E. (1979). Hippocampus, space, and memory. The Behavioral and Brain Sciences, 2, 311 -365. Orbach, J. , Milner, B., & Rasmussen, T. (1960). Learning and retention on monkeys after amygdala-hippocampal resection. Archives of Neurology, 3, 230-251. 168 Ottersen, O. P., Hjell, O. P., Osen, K. K., & Laake, J. H . (1995). Amino acid transmitters. In G. Paxinos (Ed.), The rat nervous system (pp. 1017-1040) San Diego: Academic Press. Otto, T. & Eichenbaum, H . (1992). Complimentary roles of the orbital prefrontal cortex and the perirhinal-entorhinal cortices in an odor-guided delayed-nonmatching-to-sample task. Behavioral Neuroscience, 106, 762-775. Packard, M. G. , Hirsh, R., & White, N. M . (1989). Differential effects of fornix and caudate nucleus lesions on two radial maze tasks: Evidence for multiple memory systems. Journal of Neuroscience. 9, 1465-1472. Parkin, A . J. (1984). Amnesic syndrome: A lesion specific disorder? Cortex. 20. 479-508. Parkinson, J. K. , Murray, E. A . , & Mishkin, M . (1988). A selective mnemonic role for the hippocampus is monkeys: Memory for the location of objects. Journal of Neuroscience, 8, 4159-4167. Paxinos, G. & Watson, C. (1986). The rat brain in stereotaxic coordinates. Toronto: Academic Press. Penfield, W., & Milner, B. (1958). Memory deficit produced by bilateral lesions in the hippocampal zone. Archives of Neurology and Psychiatry, 79,475-497. Peruche, B., Klaassens, H . , & Krieglstein, J. (1995). Quantitative analysis of the electrocorticogram after forebrain ischemia in the rat. Pharmacology, 50, 229-237. Petrides, M . (1985). Deficits on conditional associative learning tasks after frontal- and temporal-lobe in lesions man. Neuropsychologia, 23, 601-614. Pigott, S. & Milner, B. (1993). Memory for different aspects of complex visual scenes after unilateral temporal- or frontal-lobe resection. Neurospsychologia, 31, 1-15. Postle B. R. & Corkin, S. (1994). H . M . shows impaired word-stem completion priming with novel words. Society for Neuroscience Abstracts, 20, 876. Price, J. L. (1995). Thalamus. In G. Paxinos (Ed.), The rat nervous system (pp. 629-648) San Diego: Academic Press. Pulsinelli, W. A . , & Brierly, J. B. (1979). A new model of bilateral hemispheric ischemia in the unanesthetized rat. Stroke, 10, 267-272. Ramon y Cajal, S. (1911). "Histologic du Systeme Nerveux de l 'Homme et des Vertebres." Paris: Maloine. Rawlins, J. N. P. (1985). Associations across time: The hippocampus as a temporary memory store. Behavioral and Brain Science, 1, 211-225. Rawlins, J. N. P., Lyford, G. L. , Seferiades, A . , Deacon, R. M . J . , & Cassady, 169 H . J. (1993). Critical determinants of nonspatial working memory deficits in rats with conventional lesions of the hippocampus or fornix. Behavioral Neuroscience. 107. 420-433. Ridley, R., Baker, H . , Drewett, B., & Johnson, J. (1985). Effects of ibotenic acid lesions of the basal forebrain on serial reversal learning in marmosets. Psychopharmacology, 86, 438-443. Rod, M . R. & Auer, R. N. (1989). Pre- and postischemic administration of dizocilpine (MK801) reduces cerebral necrosis in the rat. Canadian Journal of Neurological Science, 16, 340-344. Rod, M . R., Whishaw, I. Q., Auer, R. N. (1990). The relationship of structural ischemic brain damage to neurobehavioral deficit: the effect of postischemic MK-801. Canadian Journal of Psychology, 44, 196-209. Rose, J. E. & Woolsey, C. N. (1948). The orbitofrontal cortex and its connections with mediodorsal nucleus in rabbit, sheep, and cat. Research Publications of the Association of Nervous and Mental Disorders, 27, 210-232. Rothblat, L. A . & Hayes, L. L. (1987). Short-term object recognition in the rat: Nonmatching with trial-unique stimuli. Behavioral Neuroscience, 101, 587-590. Rothblat, L . A . & Kromer, L. F. (1991). Object-recognition memory in the rat: The role of the hippocampus. Behavioral Brain Research, 42, 25-32. Rothman, S. M . & Olney, J. W. (1986). Glutamate and the pathophysiology of hypoxic-ischemic brain damage. Annals of Neurology, 19, 105-111. Saint-Cyr, J. A . , Taylor, A . E. & Lang, A . E. (1988). Procedural learning and neostriatal dysfunction in man. Brain, 111, 941-959. Salazar, A . M . , Grafman, J., Schlesselman, S., Vance, S. C , Mohr, J. P., Carpenter, M . , Pevsner, P., Ludlow, C , & Weingartner, H . (1986). Penetrating war injuries of the basal forebrain. Neurology, 36,459-465. Saunders, R. C , Murray, E. A . , & Mishkin, M . (1984). Further evidence that amygdala and hippocampus contribute equally to recognition memory. Neuropsychologia, 22, 785-796. Sawa, M . , Ueki, Y . , Arita, M . , & Harada, T. (1954). Preliminary report on the amygdaloidectomy on the psychotic patients, with interpretation of oral-emotional manifestations in schizophrenics. Folia Psychiatry and Neurology of Japan, 7, 309-329. Schacter, D. L. (1987). Memory, amnesia, and frontal lobe dysfunction. Psychobiology, 15, 21-36. Schacter, D. L. , Chiu, C.-Y. P. & Ochsner, K. N. (1993). Implicit memory: a selective review. Annual Review of Neuroscience, 16, 159-182. Schmidt-Kastner, R. & Freund, T. F. (1991). Selective vulnerability of the hippocampus in brain ischemia. Neuroscience, 40, 599-636. 170 Scoville, W. B. (1954). The limbic lobe in man. Journal of Neurosurgery, 11, 64-66. Scovile, W. B. (1968). Amnesia after bilateral mesial temporal-lobe excision: Introduction to case H . M . Neuropsychologia, 6. 211-213. Scoville, W. B., Dunsmore, R. H . , Liberson, W. T. , Henry, C. E. & Pepe. A . (1953). Observations on medial temporal lobotomoy and uncotomy in the treatment of psychotic states. Journal of Nervous and Mental Disorders, 31, 347-369. Scoville, W. B. & Milner, B. (1957). Loss of recent memory after bilateral hippocampal lesions. Journal of Neurology, Neurosurgery, and Psychiatry, 20, 11-21. Schmidley, J. W. (1990). Free radicals in central nervous system ischemia. Stroke. 21. 1086-1090. Schurr, A . & Rigor, B. M . (1992). The mechanisms of cerebral hypoxic-ischemic damage. Hippocampus, 2, 221-228. Seamans, J. K. , Floresco, S. B., & Phillips, A . G. (1995). Functional differences between the prelimbic and anterior cingulate regions of the rat prefrontal cortex. Behavioral Neuroscience, 109, 1063-1073. Shaw, C. & Aggleton, J. P. (1993). The effects of fornix and medial prefrontal lesions on non-matching-to-sample by rats. Behavioral Brain Research, 54, 91-102. Sheardown, M . J. , Nielsen, E. O., Hansen, A . J. , Jacobsen, P., & Honore, T . (1990). 2,3-Dihydroxy-6-nitro-7-sulfamoyl-benzo(F)quinoxaline: A neuroprotectant for cerebral ischemia. Science, 247, 571-574. Shimamura, A . P., Jernigan, T. L. , & Squire, L. R. (1988). Korsakoff's syndrome: Radiologic (CT) findings and neuropsychological correlates. Journal of Neuroscience, 8, 4400-4410. Shimamura, A . P., Salmon, D. P., Squire, L. R., & Butters, N. (1987). Memory dysfunction and word priming in dementia and amnesia. Behavioral Neuroscience, 101, 347-351. Siesjo, B. K. (1978). Brain energy metabolism. New York: Wiley. Siesjo, B. K. (1981). Cell damage in the brain: a speculative synthesis. Journal of Cerebral Blood H o w and Metabolism, 1, 155-185. Siesjo, B. K. (1990). Calcium, excitotoxins, and brain damage. News in the Physiological Sciences, 5, 120-125. Siesjo, B. K . , Memezawa, H . , & Smith, M . L. , (1991). Neurocytotoxicity: pharmacological implications. Fundamental and Clinical Pharmacology, 5, 755-767. Siesjo, B. K. & Wieloch, T. (1985). Molecular mechanisms of ischemic brain damage: Ca-H- related events. In F. Plum and W. Pulsinelli (Eds.), Cerebrovascular 171 Diseases, (pp. 187-200). New York: Raven Press. Sisodiya, S. M . , Free, S. L. , Stevens, J. M . , Fish, D. R., & Shorvon, S. D. (1995). Widespread cerebral changes in patients with cortical dysgenesis and epilepsy. Brain, 118, 1039-1050. Smith, M . L. & Milner, B. (1989). Right hippocampal impairment in the recall of spatial location: Encoding deficit or rapid forgetting? Neuropsychologia, 27, 71-81. Smith, M . L. , Bendek, G., Dahlgren, N . , Rosen., I., Wieloch, T . , & Seisjo, B. K. (1984). Models for studying long term recovery following forebrain ischemia in the rat. 2. A 2-vessel occlusion model. Acta Neurologica Scandinavica, 69385-401. Sommer, W. (1880). Erkankung des ammonshorns als aetiologisches moment der epilepsie. Archives of Psychiatry Nervenkr, 10, 631-675. Sprofkin, B. E. & Sciarra, D. (1952). Korsakoff's psychosis associated with cerebral tumors. Neurology, 2, 427-434. Squire, L. R. (1982). Comparisons between different forms of amnesia: Some deficits are unique to Korsakoff's syndrome. Journal of Experimental Psychology (Learning, Memory, and Cognition), 8, 560-571. Squire, L . R. (1992). Memory and the hippocampus: A synthesis from findings with rats, monkeys, and humans. Psychological Reviews, 99. 195-231. Squire, L. R. & Cohen, N. J. (1984). Human memory and amnesia. In G . L . Lynch, J. L . McGaugh, and N. M . Weinberger (Eds.). Neurobiology of learning and memory (pp. 3-64). New York: Guilford. Squire, L . R.& Moore, R. Y . (1979). Dorsal thalamic lesion in a noted case of human memory dysfunction. Annals of Neurology, 6,503-506. Squire, L . R., Ojemann, J. G. , Miezen, F. M . , Petersen, S. E . , Videen, T . O., & RaichleM. E. (1992). Activation of the hippocampus in normal humans: Afunctional anatomical study of memory. Proceedings of the National Academy of Science, 89, 1837-1841. Squire, L . R. & Slater, P. C. (1978). Anterograde and retrograde memory impairment in chronic amnesia. Neuropsychologia, 16,311-322. Squire, L . R. & Zola-Morgan, S. (1985). The neuropsychology of memory: new links between humans and experimental animals. In D. Olton, S. Corkin, & E. Gamzu (Eds.), Conference on memory dysfunctions (pp. 137-149). New York: New York Academy of Sciences Press. Squire, L. R. & Zola-Morgan, S. (1991). The medial temporal lobe memory system. Science, 253, 1380-1386. Squire, L. R., Zola-Morgan, S., & Chen., K. (1988) Human amnesia and animal models of amnesia: Performance of amnesiac patients on tests designed for the monkey. Behavioral Neuroscience, 11, 210-221. 172 Steward, O. (1976). Topographic organization of the projections from the entorhinal area to the hippocampal formation of the rat. The Journal of Comparative Neurology. 167. 285-314. Steward, O. & Scoville, S. A . (1976). Cells of origin of entorhinal cortical afferents to the hippocampus and fascia dentata of the rat. The Journal of Comparative Neurology, 169, 347-370. Storm-Mathisen, J. (1972). Glutamate decarboxylase in the rat hippocampal region after lesions of the afferent fiber system. Evidence that the enzyme is localized in intrinsic neurons. Brain Research, 40, 215-235. Sugaimachi, K . , Izawa, K., Nakamura, K., Kimura-Iwasaki, K. , Yamaguchi, M , Nakagawa. H . , & Oshino, N. (1992). Impairment of working memory by neuronal degeneration with N M D A in rat hippocampal CA-1. Behavioral Pharmacology, 3, 379-385. Sutherland, R. W. & Rudy, J. W. (1989). Configural association theory: The role of the hippocampal formation in learning, memory, and amnesia. Psychobiology, 17, 129-144. Swan, J. H . , Evans, M . C , & Meldrum, B. S. (1988). Long-term development of selective neuronal loss and the mechanisms of protection by 2-amino-7-phosphonoheptanoate in a rat model of incomplete forebrain ischaemia. Journal of Cerebral Blood Flow and Metabolism, 8, 64-78. Swan, J. H . & Meldrum, B. S. (1990). Protection by N M D A antagonists against selective cell loss following transient ischemia. Journal of Cerebral Blood Flow and Metabolism. 10.343-351. Swanson, L. W. (1981). A direct projection from Ammon's horn to prefrontal cortex in rat. Brain Research, 217, 150-154. Swanson, T. H . (1995). The pathophysiology of medial temporal lobe epilepsy. The Journal of Clinical Neurophysiology, 12, 2-22. Tamamaki, N. & Nojyo, Y . (1993). Projection of the entorhinal layer II neurons in the rat as revealed by intracellular pressure injection on neurobiotin, Hippocampus, 3, 471-480. Terzian, H . (1958). Obsevations on the clinical symptomology of bilateral partial or total removal of the temporal lobes in man. In M . Baldwin and P Bailet (Eds.), Temporal Lobe Epilepsy (pp. 510-529). Springfield, IL: Thomas. Terzian, H . & Dalle Ore, G. (1955). Syndrome of Kluver and Bucy reproduced in man by bilateral removal of the temporal lobes. Neurology, 5, 373-380. Teuber, H . L. , Milner, B., & Vaughan, H . G. (1968). Persistent anterograde amnesia after stab wound to the basal brain. Neuropsychologia, 6, 267-282. Thomas, G. J. & Gash, D. M . (1988). Differential effect of hippocampal 173 ablations on dispositional representational memory in the rat. Behavioral Neuroscience, 102. 635-642. Tolman, E. C. (1932). Purposive behavior in animals and men. New York: Century. Tulving, E. & Schacter, D. L. (1990). Priming and human memory systems. Science, 247, 301-306. Uylings, H . B. M . & Van Eden, C. G. (1990). Qualitative and quantitative comparison of the prefrontal cortex in rat and in primates, including humans. In H . B. M . Uylings, C. G . Van Eden, J. P. C. De Bruin, M . A . Corner, & M . G. P. Feenstra (Eds.), Progress in Brain Research, 85, (pp. 31-61) New York: Wiley-Liss. Van Eden, C. G . & Uylings, H . B. M . (1985). Cytoarchitectonic development of the prefrontal cortex in the rat. The Journal of Comparative Neurology, 241, 253-267. Van Groen, T. & Lopes da Silva, F. H . (1985). Septotemporal distribution of entorhinal projections to the hippocampus in the cat: Electrophysiological evidence. The Journal of Comparative Neurology, 238, 1-10. Van Groen, T. & Wyss, M . J. (1990). Extrinsic projections from area C A 1 of the rat hippocampus: Olfactory, cortical, subcortical, and bilateral hippocampal formation projections. The Journal of Comparative Neurology, 302, 515-528. Victor, M . (1994). Alcoholic dementia. Canadian Journal of Neurological Sciences, 21. 88-99. Victor, M . , Adams; R. D., Collins, G. H . (1989). The Wernicke-Korsakoff's Syndrome. A Clinical and Pathological Study of 245 patients, 82 with Postmortem Examinations. Philadelphia: F. A . Davis. Vnek, N . & Rothblat, L. A . (1996). The hippocampus and long-term object memory in the rat. Journal of Neuroscience, 16, 2780-2787. V o l l , C. L. , Whishaw, I. Q., & Auer, R. N. (1989). Postischemic insulin reduces spatial learning deficit following trasnsient forebrain ischemia in rats. Stroke, 20, 646-651. Volpe, B. T. , Davis, H . P., Towle, A . , & Dunlap, W. P. (1992). Loss of hippocampal CA1 pyramidal neurons correlates with memory impairment in rats with ischemic or neurotoxin lesions. Behavioral Neuroscience, 106, 457-464. Volpe, B.T., Pulsinelli, W. A . , Tribuna, J., & Davis, H . P. (1984). Behavioral performance of rats following transient forebrain ischemia. Stroke, 15, 558-562. Volpe, B. T. , Herscovitch, P., Raichle, M . E . , Gazzaniga, M . S., & Hirst, W. (1983). Cerebral blood flow and metabolism in human amnesia. Journal of Cerebral Blood Flow and Metabolism, 3, 5-6. Volpe, B. T. , & Hirst, W. (1983). The characterization of the amnesic syndrome after hypoxic ischemic injury. Archives of Neurology, 40, 436-440. 174 Volpe, B. T. , Holtzman, J. D., & Hirst, W. (1986). Further characterization of patients with amnesia after global hypoxic ischemic injury: Preserved recognition memory. Neurology. 36. 408-411. Voytko, M . L . (1996). Cognitive functions and the basal forebrain cholinergic system in monkeys: Memory or attention? Behavioral Brain Research, 75, 13-25. Voytko, M . L. , Olton, D. S., Richardson, R. T. , Gorman, L . K. , Tobin, J. R., & Price, D. L . (1994). Basal forebrain lesions in monkeys disrupts attention but not learning and memory. The Journal of Neuroscience, 14, 167-186. Wasterlain, C. G. & Shirasaka, Y . (1994). Seizure, brain damage, and brain development. Brain & Development, 16, 279-295. Wieloch, T. , Lindavall, O., Blomqvist, P., & Gage, F. H . (1985). Evidence for amelioration of ischemic neuronal damage in the hippocampal formation by lesions of the perforant path. Neurological Research, 7, 24-26. Whishaw, I. Q. (1985). Formation of a place learning-set task in the rat: A new procedure for neurobehavioral studies. Physiology & Behavior, 35, 845-851. Whitehouse, P., Price, D., Struble, R., Clark, A . , Coyle, J . , & Delong, M . (1982). Alzheimer's disease and senile dementia: Loss of neurons in the basal forebrain. Science. 215. 1327-1239. Witter, M . P., Jorritsma-Byham, B., & Wouterlood, F. G. (1992). Neurons of layer V a of the entorhinal cortex form part of the hippocampal cotrical output system. European Journal of Neuroscience, 2, 718-725. Wood, E. R., Bussey, T. J . , & Phillips, A . G. (1992). Performance of ischemic rats on win-stay and win-shift tasks. Society for Neuroscience Abstracts, 18, 1424. Wood, E . R., Mumby, D. G. , Pinel, J. P. J . , & Phillips, A . G . (1993). Impaired object recognition memory in rats following ischemia-induced damage to the hippocampus. Behavioral Neuroscience, 107, 51-62. Worden, R. P. (1992). Navigation by fragment fitting: A theory of hippocampal function. Hippocampus, 2, 165-187. Wyss, J. M . (1981). A n autoradiographic study of the efferent connections of the entorhinal cortex. The Journal of Comparative Neurology. 199, 495-512. Zola-Morgan, S. & Squire, L. R. (1984). Preserved learning in monkeys with medial temporal lesions: Sparing of motor and cognitive skills. The Journal of Neuroscience, 4, 1072-1085. Zola-Morgan, S. & Squire, L. R. (1985). Medial temporal lesions in monkeys impair memory on a variety of tasks sensitive to human amnesia. Behavioral Neuroscience, 99, 22-34. Zola-Morgan, S. & Squire, L. R. (1985). Amnesia in monkeys following lesions 175 of the mediodorsal nucleus of the thalamus. Annals of Neurology. 17. 558-564. Zola-Morgan, S. & Squire, L. R. (1986). Memory impairments in monkeys following lesions limited to the hippocampus. Behavioral Neuroscience, 100, 155-160. Zola-Morgan, S., Squire, L . R. & Amaral, D. G . (1986). Human amnesia and the medial temporal region: Enduring memory impairments following bilateral lesions limited to the field CA1 of the hippocampus. Journal of Neuroscience, 6, 2950-2967. Zola-Morgan, S., Squire, L. R. & Amaral, D. G. (1989a). Lesions of the amygdala that spare the adjacent cortical regions do not impair memory or exacerbate the impairment following lesions of the hippocampal formation. Journal of Neuroscience, 9, 1922-1936. Zola-Morgan, S., Squire, L. R. & Amaral, D. G. (1989b). Lesions of the hippocampal formation but not lesions of the fornix or the mammillary nuclei produce long-lasting memory impairments in monkeys. Journal of Neuroscience, 9, 898-913. Zola-Morgan, S., Squire, L. R., Amaral, D. G., & Suzuki, W . A . (1989). Lesions of the perirhinal and parahippocampal cortex that spare ther amygdala and hippocampal formation produce severe memory impairment. Journal of Neuroscience, 9, 4355-4370. Zola-Morgan, S., Squire, L . R., Clower, R. P., & Rempel, N . L . (1993). Damage to the perirhinal cortex exacerbates memory impairment following lesions to the hippocampus. Journal of Neuroscience, 13, 251-265. Zola-Morgan, S., Squire, L. R., Rempel, N. L . , Clower, R. P. & Amaral, D. G . (1992). Enduring memory impairment in monkeys after ischemic damage to the hippocampus. Journal of Neuroscience, 12, 2582-2596. Zola-Morgan, S. & Squire, L. R., & Mishkin, M . (1982). The neuroanatomy of amnesia: Amygdala-hippocampus versus temporal stem. Science, 218, 1337-1339. 

Cite

Citation Scheme:

        

Citations by CSL (citeproc-js)

Usage Statistics

Share

Embed

Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                        
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            src="{[{embed.src}]}"
                            data-item="{[{embed.item}]}"
                            data-collection="{[{embed.collection}]}"
                            data-metadata="{[{embed.showMetadata}]}"
                            data-width="{[{embed.width}]}"
                            async >
                            </script>
                            </div>
                        
                    
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:
https://iiif.library.ubc.ca/presentation/dsp.831.1-0087745/manifest

Comment

Related Items