A perceptual-mnemonic role for the perirhinal cortex in age-associated cogntive decline
Item Type text; Electronic Dissertation
Authors Burke, Sara Nicole
Publisher The University of Arizona.
Rights Copyright © is held by the author. Digital access to this material is made possible by the University Libraries, University of Arizona. Further transmission, reproduction or presentation (such as public display or performance) of protected items is prohibited except with permission of the author.
Download date 05/10/2021 10:52:58
Link to Item http://hdl.handle.net/10150/195350 1
A PERCEPTUAL-MNEMONIC ROLE FOR THE PERIRHINAL CORTEX IN AGE-ASSOCIATED COGNITIVE DECLINE
by
Sara Nicole Burke
______
A Dissertation Submitted to the Faculty of the
GRADUATE INTERDISCIPLINARY PROGRAM IN NEUROSCIENCE
In Partial Fulfillment of the Requirements For the Degree of
DOCTOR OF PHILOSOPHY
In the Graduate College
THE UNIVERSITY OF ARIZONA
2 0 0 9 2
THE UNIVERSITY OF ARIZONA GRADUATE COLLEGE
As members of the Dissertation Committee, we certify that we have read the dissertation prepared by Sara Nicole Burke entitled
A perceptual-mnemonic role for the perirhinal cortex in age-associated cognitive decline and recommend that it be accepted as fulfilling the dissertation requirement for the Degree of Doctor of Philosophy
______Date: 4/17/09 Carol A. Barnes
______Date: 4/17/09 Lynn Nadel
______Date: 4/17/09 Lee Ryan
______Date: 4/17/09 Rebecca D. Burwell
Final approval and acceptance of this dissertation is contingent upon the candidate’s submission of the final copies of the dissertation to the Graduate College.
I hereby certify that I have read this dissertation prepared under my direction and recommend that it be accepted as fulfilling the dissertation requirement.
______Date: 4/17/09 Dissertation Director: Carol A. Barnes
3
STATEMENT BY AUTHOR
This dissertation has been submitted in partial fulfillment of requirements for an advanced degree at the University of Arizona and is deposited in the University Library to be made available to borrowers under rules of the Library.
Brief quotations from this dissertation are allowable without special permission, provided that accurate acknowledgment of source is made. Requests for permission for extended quotation from or reproduction of this manuscript in whole or in part may be granted by the head of the major department or the Dean of the Graduate College when in his or her judgment the proposed use of the material is in the interests of scholarship. In all other instances, however, permission must be obtained from the author.
SIGNED: Sara Nicole Burke 4
ACKNOWLEDGEMENTS
“To get through the hardest journey we need to take only one step at a time, but we must keep on stepping” (Chinese proverb). I would like to thank all of you that encouraged and helped me to keep stepping, without you this process would have been much less enjoyable and perhaps impossible. To my graduate student cohort: Marsha Penner, Lisa Marriott-Roko, Stephen Cowen, Alex Thome, Zaneta Navratilova, and Lan Hoang, you helped to stimulate my creativity through provoking discussions, informal journal clubs, and company during happy hour. Many of you were also my shoulders to cry on and my partners in crime when I needed either one. To the other students and post docs at NSMA both past and present: Monica Chawla, David Eusten, Nathan Insel, Peter Lipa, James Lister, Kaori Nishiuchi, Ephron Rosenzweig, and Vicki Sutherland, I have learned something from, or been inspired by, each of you. This dissertation work would not have been completed without the help of the amazing undergraduate and high school students I had the pleasure of mentoring: Ajay Uprety, Saman Nematollahi, Jenelle Wallace, and Ellen Wann. You dedicated countless hours to helping me spin tetrodes, collect data and cut clusters when I was travelling, training monkeys, or getting married. You also shared with me an enthusiasm for science that helped to keep me energized. I look forward with a great keenness to what the future holds for each of you. Administrative and technical support is at times thankless, but more often it is priceless. Thank you to Kim Bohne for the hyperdrive construction, Joseph Bohanick for data collection and always being good in an emergency, Jie Wang for cutting a heroic amount of clusters, Michael Montgomery for saving the day each time I had a computer or data meltdown, and to Michelle Carroll, Luann Snyder, and Erin Wolfe for your superb organizational skills and always tolerating my fear of paperwork. I have reflected many times on how fortunate I was to conduct my dissertation research with Carol A. Barnes. Your contributions to the field of the neurobiology of aging are immeasurable. You have taught me by example to think big, push myself, and the power of diplomacy. You also always knew when to rein me in and make me focus. I look forward to more years of science under your wisdom. Last, but certainly not least, I owe a great thanks to my phenomenal husband Andrew P. Maurer. To me, you have been a teacher, a sounding board, a database of the literature, a collaborator, an inspiration, and most importantly my best friend. You pushed me to challenge myself but also reminded me to slow down every once in a while. Thank you for going home to feed the dogs when I could not, and always knowing where my keys and cell phone are located. The best part of graduate school was finding you and I look forward to watching ourselves grow old together. Hopefully, you will still be able to tell me where my keys are when we are 70! 5
DEDICATION
This dissertation is dedicated to my family. You have always supported me and believed that I could do anything even when I did not believe it myself. To my mom and dad, Diane and Michael Burke, without your love and support I would not have made it this far. Thank you both for reminding me to take it easy every once in a while and coming to Tucson all those times I could not get away to see you. To my brother, Adam, thank you for always making me laugh when I needed it most. To my dear friend Sarah Marble, thank you for “talking me off the ledge” during prelims and every time science or life threw a curve ball at me. Finally, I want to dedicate this work to my perfect dogs Scout and Riley, although you missed out on many trips to the dog park and evening runs, you still loved me unconditionally and rested at my feet while I wrote most this. 6
TABLE OF CONTENTS
LIST OF FIGURES...... 11
LIST OF TABLES ...... 15
ABSTRACT...... 16
CHAPTER 1: INTRODUCTION...... 17
CHAPTER 2: AGE-ASSOCIATED CHANGES IN HIPPOCAMPAL
PLASTICITY...... 29
2.1 Introduction ...... 29
2.2 Morphology of the aging brain ...... 30
2.3 Biophysical properties of aged hippocampal neurons ...... 35
2.4 Changes in hippocampal cell–cell interactions ...... 37
CHAPTER 3: AGING ENSEMBLES: CIRCUIT CONTRIBUTIONS TO MEMORY DEFICITS ...... 51
3.1 Introduction...... 51
3.2 Fundamental Properties of Place Cells in Young and old Rats...... 53
3.3 Spike timing in the aged hippocampus ...... 58
3.4 Dynamic properties of hippocampal place cells in young and old rats ...... 63
3.5 From the map to the memory in aging ensembles ...... 75
CHAPTER 4: THE FUNCTION AND PHYSIOLOGY OF THE PERIRHINAL CORTEX ...... 87
4.1 Introduction...... 87
4.2 Perirhinal cortical-dependent behavior in rats and monkeys ...... 91
7
TABLE OF CONTENTS- continued
4.3 Biophysical and plasticity properties of perirhinal cortical neurons...... 103
4.4 Single-unit activity of perirhinal cortical neurons in behaving animals ...... 109
4.5 Discussion...... 118
CHAPTER 5: THE HUMAN PERIRHINAL CORTEX AND PROCESS THEORIES OF RECOGNITION MEMORY...... 120
5.1 Introduction...... 120
5.2 Knowing and recollecting: Single versus Dual-process models for recognition ...... 121
5.3 The functional neuroanatomy of human recognition memory . 128
5.4 Recognition memory and human aging...... 135
5.5 Discussion...... 140
CHAPTER 6: THE EFFECTS OF AGE AND CONTEXT ON SPONTANEOUS OBJECT RECOGNITION...... 141
6.1 Introduction...... 141
6.2 Methods ...... 144
6.2.1 Subjects and spatial memory testing...... 144
6.2.2 Spontaneous object recognition testing procedures ...... 146 6.2.3 Experiment 1: standard spontaneous object recognition (SOR) testing...... 147
6.2.4 Experiment 2: novel/repeat SOR testing...... 150
6.2.5 Experiment 3: SOR testing with context change ...... 153
6.2.6 Data analysis...... 156
6.3 Results...... 157 8
TABLE OF CONTENTS- continued
6.3.1 Morris swim task performance ...... 157
6.3.2 Experiment 1: standard SOR task results...... 160
6.3.3 Experiment 2: novel/repeat SOR task results...... 165
6.3.4 Experiment 3: SOR task with context change...... 169
6.3.5 Relationship between object recognition and spatial memory...... 175
6.4 Discussion...... 176
6.4.1 Summary of salient results ...... 176
6.4.2 The perceptual-mnemonic feature conjunctive (PMFC) model of the perirhinal cortex and object recognition during aging...... 184
6.4.3 Mesolimbic dopamine and novelty detection...... 186
6.4.4 Conclusions ...... 187
CHAPTER 7: THE EFFECTS OF AGE AND NOVELTY ON THE SINGLE-UNIT FIRING CHARACTERISTICS OF PERIRHINAL CORTICAL NEURONS ...... 189
7.1 Introduction...... 189
7.2 Methods ...... 190
7.2.1 Subjects and behavioral training...... 190
7.2.2 Surgical procedures ...... 194
7.2.3 Neurophysiology...... 196
7.2.4 Analyses and statistics ...... 198
7.3 Results...... 199
7.3.1 Morris swim task performance ...... 199
7.3.2 Running velocity in young and aged rats...... 201 9
TABLE OF CONTENTS- continued
7.3.3 Behavioral correlates of perirhinal cortical activity...... 207
7.3.4 Aging, novelty and perirhinal cortical neuron firing rate ...... 222
7.3.5 PRC neurons without object or food dish fields in young and aged rats...... 229 7.3.6 Object-related activity across delays...... 232
7.4 Discussion...... 235
7.4.1 Summary of salient results ...... 235
7.4.2 The lack of novelty and/or familiarity modulation of perirhinal cortical neuron activity in young and aged rats ...... 237
7.4.3 Object fields in young and aged rats ...... 240
7.4.4 Food dish fields in young and aged rats ...... 242
7.4.5 The influence of spatial location on object and food dish field activity ...... 243
CHAPTER 8 MIDDLE HIPPOCAMPAL CA1 ACTIVITY CHARACTERISTICS ARE MODULATED BY 3-DIMENSIONAL OBJECTS: EFFECTS ON PLACE FIELD SIZE, PROBABILITY OF PLACE FIELD EXPRESSION, AND MAP STABILITY ...... 247
8.1 Introduction...... 247
8.2 Methods ...... 254
8.2.1 Subjects and behavioral training...... 254
8.2.2 Surgical procedures ...... 257
8.2.3 Neurophysiology...... 259
8.2.4 Analyses and statistics ...... 261
8.3 Results...... 263 10
TABLE OF CONTENTS- continued
8.3.1 Morris swim task performance ...... 263
8.3.2 Running velocity...... 264
8.3.3 The effects of objects on the probability of place field expression and firing rate...... 268
8.3.4 Objects and place map stability ...... 272
8.3.5 The effect of objects on middle CA1 place field size...... 275
8.4 Discussion...... 281
CHAPTER 9: CONCLUSIONS ...... 288
9.1 Summary of salient observations...... 288
9.2 Suggestions of future experiments...... 292
9.2.1 The effect of perceptual difficulty on the performance of aged rats...... 292
9.2.2 Pattern separation in the perirhinal cortex...... 293
9.2.3 Behavioral relevance and perirhinal cortical activity ...... 294
9.2.4 The specificity of objects fields and the role of spatial location...... 295
9.2.5 The impact of objects in the aged middle hippocampus . 297
9.2.6 The impact of objects on place map stability ...... 297
9.3 Conclusions ...... 299
REFERENCES...... 301
11
LIST OF FIGURES
Figure 1.1: The comparative anatomy of the hippocampus in monkeys and rats adapted from Amaral and Lavenex, 2007 ...... 19
Figure 1.2: The comparative anatomy of the parahippocampal region in monkeys and rats adapted from Amaral and Witter, 2006; Burwell et al., 1995 ...... 20
Figure 2.1: The myth of brain aging from Burke and Barnes, 2006 ...... 32
Figure 2.2: Summary of age-related alterations in LTP and LTD between young and aged animals updated from Burke and Barnes, 2006 ...... 46
Figure 3.1: Investigating Neuronal Ensembles with Multiple Single Unit Recording Methods from Burke and Barnes, 2006 ...... 52
Figure 3.2: CA1 place fields in old rats ...... 54
Figure 3.3: Arc expression in the dentate gyrus from Small et al., 2004...... 56
Figure 3.4: Theta phase precession in young and old rats ...... 59
Figure 3.5: Phase precession during the first pass through a place field...... 61
Figure 3.6: Experience-dependent place field expansion in young and old rats...... 66
Figure 3.7: Hypothetical synaptic modification curves for young and aged CA1 neurons ...... 67
Figure 3.8: CA1 map stability in young and old rats from Barnes et al., 1997 . 69
Figure 3.9: Hippocampal map realignment adapted from Rosenzweig et al., 2003 ...... 73
Figure 3.10: Procedure for calculating the explained variance (EV) measure of activity pattern reactivation...... 78
Figure 3.11: The temporal bias calculation from Gerrard et al., 2008 ...... 81
Figure 3.12: NMDA receptor blockade and reactivation ...... 83
Figure 4.1: The rat perirhinal cortex (PRC) adapted from Burwell, 2001 ...... 88 12
LIST OF FIGURES- continued
Figure 4.2: A schematic of the spontaneous object recognition (SOR) task.... 95
Figure 4.3: The proposed convergence of object-feature representations as information passes through sequential processing stages in the ventral visual stream from Murray and Bussey, 1999 ...... 99
Figure 4.4: The perceptual oddity task from Bartko et al., 2007...... 100
Figure 4.5: The PRC laminae from Furtak et al., 2007...... 104
Figure 5.1: Signal-detection model of recognition memory adapted from Yonelinas, 2001 ...... 122
Figure 5.2: The process dissociation procedure, ROC curve asymmetry and recognition memory...... 124
Figure 5.3: ROC curves, Z-slope, and dual-process models ...... 126
Figure 5.4: Hippocampal and rhinal activation and associated memory confidence from Daselaar et al., 2006...... 138
Figure 6.1: Schematic of the standard SOR task used in Experiment 1 ...... 148
Figure 6.2: Schematic of the presentation of objects used in the novel/repeat SOR test (Experiment 2) ...... 152
Figure 6.3: The procedure for the SOR task with context change (Experiment 3)...... 155
Figure 6.4: Morris swim task performance of young and aged rats ...... 158
Figure 6.5: SOR task performance (Experiment 1)...... 162
Figure 6.6: Novel/repeat SOR task performance (Experiment 2) ...... 168
Figure 6.7: SOR task with context change performance (Experiment 3)...... 172
Figure 7.1: Behavioral procedures used for electrophysiological recordings ...... 192
Figure 7.2: Location of perirhinal cortical recordings, flap map image from Burwell, 2001 ...... 196 13
LIST OF FIGURES- continued
Figure 7.3: Performance on the Morris swim task ...... 200
Figure 7.4: Mean running velocity of young and aged rats during different behavioral conditions ...... 202
Figure 7.5: Mean running velocity for laps 1-2 versus laps 19-20...... 204
Figure 7.6: Behavioral correlates of PRC neuron activity ...... 210
Figure 7.7: Velocity by position...... 212
Figure 7.8: Information scores of PRC neurons...... 214
Figure 7.9: The effect of the food dishes on spatial information score...... 216
Figure 7.10: A representative PRC neuron with objects fields...... 217
Figure 7.11: Proportion of PRC cells with object fields in young and aged rats...... 219
Figure 7.12: Food dish-associated activity of PRC neurons ...... 221
Figure 7.13: Firing rate as a function of running velocity ...... 224
Figure 7.14: The difference in normalized firing rate between lap 1 and lap 10 ...... 227
Figure 7.15: The proportion of recorded PRC neurons that show a response decrement ...... 228
Figure 7.16: The proportion of inactive and non-selective PRC neurons during track running in young and aged rats...... 231
Figure 7.17: Correlated PRC activity patterns between behavioral epochs..... 234
Figure 7.18: Population coding in the PRC ...... 239
Figure 7:19: PRC activity during the configuration change condition...... 246
Figure 8.1: Tetrode track for distal middle CA1 recordings...... 258
Figure 8.2: Morris swim task performance...... 263 14
LIST OF FIGURES- continued
Figure 8.3: Running velocity during the different behavioral conditions...... 265
Figure 8.4: Place field expression across behavioral conditions ...... 270
Figure 8.5: The effect of objects on place field activity ...... 272
Figure 8.6: The effect of objects of CA1 place map stability...... 274
Figure 8.7: A representative example of the activity of two CA1 neurons recorded during epoch 1 ...... 276
Figure 8.8: The effect of objects on place field size...... 278
Figure 8.9: The effect of objects on place field size within a population of middle CA1 neurons...... 280
Figure 9.1: CA1 activity correlations between epochs ...... 298
15
LIST OF TABLES
Table 6.1: Exploration times during the familiarization phase of Experiment 1...... 160
Table 6.2 Exploration times during the familiarization phase of Experiment 2...... 167
Table 6.3 Exploration times during the familiarization phase of Experiment 3...... 171
Table 7.1: Numbers of perirhinal cortical (PRC) neurons recorded from young and aged rats ...... 207
Table 7.2: Firing rates in different cortical laminae ...... 209
Table 7.3: Single-unit firing characteristics of PRC neurons in young and aged rats...... 222
Table 8.1: Behavioral conditions during epochs 1 and 2 of track running..... 256
Table 8.2: Numbers of recorded middle CA1 pyramidal cells and place fields...... 268
Table 8.3: Mean firing rats for middle CA1 neurons during epoch 1 and epoch 2 ...... 271
16
ABSTRACT
Perirhinal cortical-dependent behavior and single-unit neuron activity in this brain region were compared between normal aged and young rats. Three different variants of the spontaneous object recognition task were used in these experiments, and the results confirmed previous reports that aged animals are impaired at stimulus recognition. The novel contribution of the present experiments was the identification that the behavioral deficit in the aged rats was due to the old animals treating novel objects as familiar, rather than to forgetting the previously experienced stimuli. This pattern of results in the old animals mirror data obtained from rats with perirhinal cortical lesions and promotes a hypothesis that this area of the brain serves a perceptual-mnemonic function. Additionally, multiple single-unit recordings were obtained from perirhinal cortical cells while young and aged rats traversed a track that contained several objects. Perirhinal neurons exhibited selective increases in their firing rates at object locations. We have called these areas of higher perirhinal cortical cell activity ‘object fields’. While both young and old rats expressed object fields, a lower proportion of perirhinal neurons showed this type of activity in the aged compared to the young rats. Although familiar and novel objects were placed on the track as part of a systematic design, there was no effect of novelty on the overall firing rates of perirhinal cortical neurons or the proportion of cells expressing object fields under these experimental conditions. These data suggest that the physiological correlate of stimulus recognition is not decrements in perirhinal cortical activity when a stimulus goes from novel to familiar. A final important observation made during these studies in young rats was that place fields in the middle hippocampal CA1 subregion are affected by placing objects in the track. Because this same manipulation increases perirhinal cortical activity, it could indicate that age-related changes in the perirhinal cortex might alter the function of other closely associated structures. 17
“Time is contagious, everybody’s getting old.” – Damien Rice
CHAPTER 1: INTRODUCTION
We are an aging world. In the year 2000, the net balance of the world’s elderly population grew by more than 795,000 people each month reaching over
420 million by the beginning of 2001 (Kinsella and Velkoff, 2001). In the United
States alone, according to U.S. Census Bureau projections, the number of individuals age 65 or older will more than double between 2000 and 2030 from
35 million to over 70 million (UNC, 2006). Declining birthrates and an increased life expectancy have resulted in a population age distribution with a larger proportion of elderly individuals than has ever occurred in recorded history.
Therefore, increasing our understanding of how advanced age impacts physiology and neuropsychological function is critical for insights into this growing demographic sector. The big challenges that face scientists interested in the aging process include, differentiating normal aging from pathology, understanding what makes some people age more successfully than others, and facilitating the development of potential therapeutics.
Although aging affects the entire body, its impact on the brain and cognition is something that individuals outside of the field of gerontology can appreciate, because we all have personal anecdotes of ‘forgetfulness’. Moreover, most people acknowledge that the frequency of these episodes of memory lapse increase with advanced age and self-reports of the impact of aging on cognitive 18 function are common. In fact, the famous Canadian psychologist D.O. Hebb, best known for his neuropsychological postulate of cell assembly formation and learning, wrote a personal account of the intellectual changes one faces in advanced aged. Hebb’s first description of what he termed a “memory blackout” occurred when he was 47:
I was reading a journal article that was directly relevant to some work I
had underway or was proposing to try. I thought, “I must make a note of
this”, but when I turned the page, I found a penciled note in my
handwriting! It was a real shock. I was used to forgetting things that didn’t
interest me, but even then I always knew when I was reading something a
second time. (Hebb, 1978)
Later, when Hebb reached his 70s he noted that his forgetfulness had increased, writing “I have long believed that the best way to remember to take something with me is to put it on the doorstep, so I see it when I leave; now this works better if the object is big enough to trip me” (Hebb, 1978). Because personal accounts of memory changes during aging are pervasive it is not surprising that research on the neurobiology of aging has focused on brain structures known to be critical for memory formation. In humans and other mammals, the medial temporal lobes are required for the encoding and consolidation of new memories (e.g., Scoville and Milner, 1957; Zola-Morgan and Squire, 1990). Interestingly, these structures show remarkable conservation across species, which makes animal models of age-related memory decline particularly tenable. Figure 1.1 shows the 19
A C
Monkey Rat
B D
Figure 1.1: The comparative anatomy of the hippocampus in monkeys and rats. (A) Magnetic resonance imaging of the monkey brain showing the hippocampus (dentate gyrus, hippocampus proper, and subiculum) in red and the entorhinal cortex in green. The left panel shows an oblique view and the right panel shows a lateral view. The white vertical line on the lateral view shows approximately where along the rostral-caudal axis the coronal section in (B) was taken from. (B) Nissl-stained section (left panel) and line drawing (right panel) of the different subregions of the hippocampus, which include the hippocampus proper (CA1, CA2 and CA3), the dentate gyrus (DG), and the subiculum (Sub). (C) Magnetic resonance imaging of the rat brain showing the hippocampus in red and the entorhinal cortex in green. The left panel shows an oblique view and the right panel shows a lateral view. Notice that in the rat brain the hippocampus is rostral to the entorhinal cortex but in the monkey the entorhinal cortex is anterior to the hippocampus. This is because in the primate the hippocampus and surrounding medial temporal cortex is rotated ventrally relative to the rat, presumably because of the greater cortical surface area in primates. Therefore, the anterior monkey hippocampus is homologous to the ventral rat hippocampus and posterior monkey hippocampus is homologous to the rat dorsal hippocampus. The white vertical line in the lateral view shows approximately where along the rostral-caudal axis the coronal section in (D) was taken from. (D) Nissl-stained section (left panel) and line drawing (right panel) of the different subregions of the hippocampus. Note, the similarities in the subdivisions between monkey and rat. Bar = 1mm in all bottom panels. Figure adapted from Amaral and Lavenex, 2007. hippocampus and entorhinal cortex in rats and non-human primates (Amaral and
Lavenex, 2007). Sensory association areas project to the parahippocampal region (Suzuki and Amaral, 1994; Burwell and Amaral, 1998), which includes the medial and lateral entorhinal cortices, the perirhinal cortex, and the postrhinal cortex (homologous to the parahippocampal cortex in primates) (Burwell et al.,
1995; Burwell and Amaral, 1998; Suzuki and Amaral, 2003). Figure 1.2 shows 20
A Monkey brain, Rat brain, ventral view lateral view
B Monkey Rat unfolded unfolded flat map flat map
Figure 1.2: The comparative anatomy of the parahippocampal region in monkeys and rats. (A) The parahippocampal region in monkeys (left) and rats (right). The perirhinal cortex is indicated by the green (area 36) and maroon (area 35) regions. In the primate area 35 is not visible because it is located in the rhinal sulcus (rs). The entorhinal cortex (EC) is shown in orange, and the parahippocampal cortex in the monkey (TH and TF) is shown in yellow. This region is homologous to the rat postrhinal cortex (POR). Figure adapted from Brown and Aggleton, 2001. (B) The two-dimensional unfolded flat maps of the parahippocampal region in monkeys (left) and rats (right). Although these cortical areas are located on the ventral surface of the monkey brain but are on lateral surface in the rat brain, their relative orientation to each other and there connections with other brain regions are remarkably similar between primates and rats. Figure adapted from Burwell et al., 1995. the parahippocampal region in monkeys and rats and the two-dimensional unfolded flat maps of these brain regions. These cortical structures share dense reciprocal connections with the hippocampus (for review, see Amaral and Witter,
1995). While the different areas of the medial temporal lobe work in concert to support learning and memory they also have dissociable functions (e.g.,
Eichenbaum, 2000; Brown and Aggleton, 2001), and it is likely that aging impacts 21 the hippocampus, perirhinal cortex, entorhinal cortex, and postrhinal cortex in distinct ways.
Research on the neurobiology of age-related memory decline has focused primarily on the hippocampus and relatively little is known regarding the impact of normal aging on the perirhinal, entorhinal and postrhinal cortices. There are several salient reasons why the hippocampus was the early focus of investigations searching for associations between age-related cognitive decline and brain function. First, it was observed that bilateral removal of the hippocampus lead to an inability to form new memories. In 1953, patient H.M. underwent experimental surgery, as treatment for his intractable seizure disorder, and emerged with profound amnesia. Remarkably, H.M.’s intellect and personality were unchanged even though his amnesia left him perpetually in the
1940s (Scoville and Milner, 1957). Prior to the historic collaboration of Scoville and Milner (1957), Karl Lashley’s principles of mass action and equipotentiality prevailed. These theories contended that the cortex acts as whole in many types of learning, and if certain parts of the brain are damaged other areas can take on the role of the damaged portion. Under this view, learning and memory deficits are proportional to the size of the lesion (Lashley, 1931; Lashley, 1949; Lashley et al., 1951). The landmark case study of Patient H.M. refuted ideas that there was no locus for memory in the brain and initiated a research field searching for the engram in the hippocampus. 22
On December 2, 2008 at the age of 82 H.M. passed away and we now know him as Henry Molaison. Because of his extended years, Mr. Molaison’s process of aging is of great interest. Without an intact hippocampus the memory lapses that concern many of us as we grow older never plagued him. There were, however, other changes in Mr. Molaison’s brain and cognitive status that were documented between the ages of 70 and 80 that could be associated with advanced age (Salat et al., 2006) and the post-mortem analysis of Mr. Molaison’s brain will provide additional insights regarding his clinical outcome and aging process (Corkin, 2002). Without being aware of it, Mr. Molaison’s contributions to neuroscience are immeasurable.
The second reason that research on cognitive aging has focused on the hippocampus is because it was in this area of the brain that the phenomenon of long-term potentiation (LTP), a possible biological mechanism for associative memory, was first observed. Specifically, it was noted that strong synaptic modification could be induced by high frequency stimulation of the perforant path input to the granule cells in the dentate gyrus of the hippocampus (Bliss and
Lomo, 1973). This marked the first physiological support for a mechanism that could underlie Hebb’s learning rule, which states that connections between co- active neurons are strengthened (Hebb, 1949). Direct support for the cooperativity principle was also demonstrated for LTP in 1978 (McNaughton et al., 1978), which further indicated that this was a plausible model of associative memory. Barnes showed that the decay rate, over days, of LTP in the dentate 23 gyrus correlated with accuracy of performance in a spatial memory task.
Importantly, the aged rats who showed the most profound memory deficits also had significantly faster rates of LTP decay relative to old rats with better spatial memory and to the young animals (Barnes, 1979). This suggested one possible mechanism for age-related learning and memory impairments, namely plasticity deficits in the aging brain. Over the last three decades, this avenue of research has been fruitful, providing a comprehensive understanding of plasticity impairments in the aged hippocampus. In Chapter 2, I will summarize what is currently known regarding age-related changes in plasticity in the different subregions of the hippocampus. This discussion will not be extended to the perirhinal cortex because there are currently no reports of perirhinal plasticity in the aged brain. By understanding how aging impacts hippocampal plasticity, however, we can begin to construct hypotheses regarding how these dynamic processes might be altered in the perirhinal cortex.
Another impetus to study the hippocampus in relation to the neurobiology of aging is the remarkable behavioral correlation of principal cell activity in this region of the brain. Specifically, when the activity of principal cells in the hippocampus is recorded from awake behaving rats, the firing rates of these neurons show patterned neural activity that is highly correlated with a rat’s position in space (O'Keefe and Dostrovsky, 1971). The discrete region of the environment that corresponds to a neuron’s increased firing rate is termed the
‘place field’ of the cell (O'Keefe, 1976). In fact, when the firing patterns of many 24 hippocampal neurons are recorded simultaneously, it is possible to reconstruct the position of a rat within an environment from the place field firing patterns alone (Wilson and McNaughton, 1993). The striking behavioral correlate of hippocampal neuron activity and the critical involvement of this brain area in memory (e.g., Scoville and Milner, 1957) made it clear that our understanding of age-related memory decline could be advanced by recording the ensemble activity of many hippocampal neurons simultaneously and comparing the place field characteristics between young and aged rats. Chapter 3 will describe what is currently known regarding age-associated changes in the ensemble activity characteristics of hippocampal neurons and the extent to which these alterations can explain memory impairments. The age-related changes in the ensemble dynamics of hippocampal neurons that have been revealed through multiple- single unit recordings can be used a framework to develop hypotheses regarding the impact of aging on perirhinal neuron populations, and how this contributes to cognitive impairments.
Although there are functional changes within the aged hippocampus, and many of these have been linked to age-related learning and memory deficits, we still do not know how changes in cortical areas upstream of the hippocampus contribute to age-associated alterations in hippocampal function. Conversely, how age-associated functional alterations within the hippocampus affect its cortical efferent regions remains undetermined. This thesis represents one attempt to elucidate the possible impact of aging on the functional integrity of a 25 cortical structure reciprocally connected to the hippocampus, the perirhinal cortex.
In young animals, there are several accounts of the physiological characteristics of perirhinal cortical cells and how this structure could support the learning of visual associations as well as stimulus recognition. To date, however, there are no reports of the activity characteristics of perirhinal neurons recorded in the aged rat brain. Chapter 4 will summarize what is known about the physiology and plasticity of the young perirhinal cortex with predictions for how these properties might be altered in the old animal, and Chapter 5 will discuss what is currently known about the human perirhinal cortex and its role in recognition memory.
Although empirical data regarding the impact of the normal aging on the functional integrity of the perirhinal cortex are limited, there are several reasons to assume that the perirhinal cortex is affected during advanced age. At the foundation of this hypothesis is the observation that old rats (Bartolini et al.,
1996; Wiig and Burwell, 1998; de Lima et al., 2005; Pieta Dias et al., 2007;
Platano et al., 2008), and old monkeys (Moss et al., 1988; Rapp and Amaral,
1991; Herndon et al., 1997; Moss et al., 1997; Shamy et al., 2006) show reduced object recognition relative to young animals. While the involvement of the hippocampus in this cognitive function is heavily debated (for example, see
Squire et al., 2007; Brown, 2008), it is widely accepted and clear from all the available lesion data that the perirhinal cortex is necessary for object recognition 26
(e.g., Ennaceur and Aggleton, 1997; Baxter and Murray, 2001a; Baxter and
Murray, 2001b; Prusky et al., 2004; Winters et al., 2004; Norman and Eacott,
2005; Winters and Bussey, 2005). Chapter 6 will present a comprehensive characterization of object recognition impairments in aged rats using the standard spontaneous object recognition task, and two variants of this task. In an attempt to account for the object recognition differences between young and old rats, possible functional alterations within perirhinal cortex were measured using single-unit recording from awake-behaving young and old rats. Chapter 7 will describe this series of electrophysiological experiments, which involve recording multiple-single perirhinal cortical neurons simultaneously as young and aged rats traverse a circular track containing either familiar or novel objects. Finally, in order to understand how functional alterations within the perirhinal cortex might impact hippocampal physiology, we recorded from the ventral CA1 subregion of the hippocampus of young rats under conditions in which perirhinal activity was putatively modulated by enriching the environment with objects. Chapter 8 describes how information processed in the perirhinal cortex may influence CA1 activity characteristics, giving us a direction for the exploration of how the interactions between different regions of the medial temporal lobe are affected by aging.
The primary role of the perirhinal cortex in cognition is still debated.
Specifically, some researchers have argued that the perirhinal cortex participates in a medial temporal lobe memory system but does not significantly contribute to 27 perception (e.g., Squire and Zola-Morgan, 1991; Buffalo et al., 1998;
Eichenbaum, 2000; Squire et al., 2004; Cate and Kohler, 2006; Eichenbaum et al., 2007). This view, however, does not adequately account for behavioral data showing that perirhinal cortical lesions lead to visual discrimination impairments between complex stimuli that share common features (Baxter and Murray, 2001;
Bussey et al., 2003; e.g., Buckley, 2005; Bartko et al., 2007; Bartko et al., 2007).
Moreover, models that attribute both perceptual and memory functions to the perirhinal cortex either not have specified how this brain region participates in both (e.g., Murray and Bussey, 1999; Bussey et al., 2005), or have asserted that distinct networks within this area support these two cognitive processes (Bogacz and Brown, 2002; Bogacz and Brown, 2003). A more parsimonious description of perirhinal cortical function would be one that could provide a unified account of how a single perirhinal network could subserve both perception and memory.
This dissertation will attempt to provide such an account.
When cognitive neuroscientists describe perceptual processes they typically refer to the ability of an organism to be aware of and derive meaning from sensory input. On the other hand, memory is the ability to store, retain and recall information. Paramount to both perceptual and memory processes, however, is the association of different elements of a stimulus or an episode. In this sense it may not be productive to attempt to arbitrarily define when perception ends and memory begins. In fact, in Chapters 4, 5, 6 and 7, I will argue that the perirhinal cortex serves a “perceptual-mnemonic” function (Murray 28 and Bussey, 1999; Bussey et al., 2005; Murray et al., 2007). For this dissertation, this perceptual-mnemonic process refers to the idea that the primary role of the perirhinal cortex is to bind and store features that are represented in lower sensory cortices. Because, the binding of information represented across disparate areas of the neocortex is believed to be the cornerstone of memory, behavioral deficits resulting from PRC lesions can present either as memory or perception impairments depending on the specific task demands. 29
CHAPTER 2: AGE-ASSOCIATED CHANGES IN HIPPOCAMPAL PLASTICITY
2.1 Introduction
Aging is associated with a decline in cognitive function that can, in part, be explained by changes in neural plasticity or cellular alterations that directly affect mechanisms of plasticity (Rosenzweig and Barnes, 2003; for review, see Burke and Barnes, 2006). Although several age-associated neurological changes have been identified during normal aging, they tend to be subtle compared with the alterations that are observed in age-associated diseases, such as Alzheimer’s disease (e.g., Coleman and Flood, 1987). Moreover, understanding age-related changes in cognition sets a background against which it is possible to assess the effects of pathological disease states. This chapter will provide a broad overview of the age-associated changes within the hippocampus that are believed to underlie the decline in learning and memory that humans and animals experience in advanced age. In particular, it is now understood that gross anatomical changes in brain morphology and cell number are not a requisite of normal aging (Morrison and Hof, 1997). Rather, the brain’s ability to change dynamically with experience appears to be affected by the aging process and this contributes to alterations in the neural circuits that support cognition. In this chapter the subtle changes in neuronal morphology, and cell–cell interactions that may contribute to alterations in plasticity in aged animals, are reviewed in detail. Chapter 3 will summarize how these changes in the dynamic properties of synapses may lead to alterations in network dynamics of aged neuronal 30 ensembles, ultimately resulting in selective behavioral impairments in learning and memory. The focus of these next two chapters emphasizes the neurobiology of aging in the hippocampus, because of the dearth of information on age-related alterations in the perirhinal cortex (PRC). Additionally, understanding how aging impacts structures densely connected with the PRC, such as the hippocampus, can provide a framework for investigations on PRC function in elderly humans and other animals.
2.2 Morphology of the aging brain
Age-related changes in the morphology of neurons are selective and it seems that there is no universal pattern across the entire brain. One finding that does seem to be consistent, however, is that in most brain areas neuronal loss does not have a significant role in age-related cognitive decline (for review, see
Morrison and Hof, 1997). Rather, small, region-specific changes in dendritic branching and spine density are more characteristic of the effects of aging on neuronal morphology (Figure 2.1). This is contrary to early investigations of the aged brain in which profound neuron loss was reported to occur in advanced age.
In 1955, Brody was the first to suggest that age-related reductions in brain weight were due, in part, to a decline in neuron number in all cortical layers
(Brody, 1955). Subsequent investigations corroborated this work, reporting a 10–
60% decline in cortical neuron density between late childhood and old age 31
(Coleman and Flood, 1987). In addition, profound cell loss was found in the hippocampus of aging humans (Ball, 1977) and the hippocampus and prefrontal cortex of non-human primates (Brizzee et al., 1980). The data obtained from these early reports, however, were confounded by a variety of technical and methodological issues, such as tissue processing and sampling design, that later called into question their accuracy (Morrison and Hof, 1997).
In the 1980s, when new stereological principles were developed, it became possible to identify and eliminate many of the confounding factors of the previous studies that had indicated a profound decline in neuron number occurring in advanced age (West, 1993). The resulting conclusion was that in humans (West et al., 1994; Pakkenberg and Gundersen, 1997), non-human primates (Peters et al., 1994; Gazzaley et al., 1997; Merrill et al., 2000; Keuker et al., 2003) and rodents (Rapp and Gallagher, 1996; Rasmussen et al., 1996;
Merrill et al., 2001), significant cell death in the hippocampus and neocortex is not characteristic of normal aging. Additionally, when cell number is examined for different sub-divisions of the of the parahippocampal region, there is no decline in total cell numbers with age in the perirhinal, postrhinal, lateral entorhinal, or medial entorhinal cortices of rats, even when the animals are impaired in tests of spatial learning and memory (Rapp et al., 2002). A notable exception to this finding has recently been reported, however. In aged non-human primates there is an approximately 30% reduction in neuron number in area 8A of the dorsal lateral prefrontal cortex in all layers, which significantly correlates with impaired 32 performance on a working memory task. By contrast, area 46 of the prefrontal
Figure 2.1: The myth of brain aging. A common misconception about normal aging is that significant cell loss and dramatic changes in neuronal morphology occur. (A) This example shows progressive loss of the dendritic surface in aged human dentate gyrus granule cells. These data, however, do not accurately reflect the subtle and selective morphological alterations that actually occur in aged neurons. Age-associated loss of dendritic extent in the dentate gyrus and CA1 was exaggerated by including demented and normal aged subjects in the same experimental group, and not using stereological controls. (B) Two representative granule cells filled with 5,6 carboxyfluorescein from the dentate gyrus of a 24 month old rat. In the rat dentate gyrus there is no significant change in dendritic extent between young and old age but there is a significant increase in electrotonic coupling. (C) Reconstructions of representative hippocampal CA1 neurons from young rats (2 months; top) and old rats (24 months; bottom). There is no reduction in dendritic branching or length with age in CA1. (D) Three electrotonically coupled CA3 neurons filled with 5,6-carboxyfluorescein. Figure from Burke and Barnes, 2006.
33 cortex shows conservation of neuron number (Smith et al., 2004). Similar to early reports of a decline in neuronal density with aging, early investigations of dendritic branching suggested massive deterioration in the human entorhinal cortex and hippocampus (Fig 2.1A; Scheibel et al., 1976; Scheibel, 1979). These experiments, however, included both demented and non-demented subjects.
Subsequent investigations, which controlled better for the subjects’ mental status and applied stereological methods, found that normal aged individuals had extensive dendritic branching in layer II of the parahippocampal gyrus (the origin of the perforant pathway to the dentate gyrus (Buell and Coleman, 1979; Buell and Coleman, 1981). Moreover, dendritic branching and length appeared to be greater in aged individuals compared with younger adults or patients with senile dementia. Other investigations have reported increased dendritic extent in the dentate gyrus of old compared with middle-aged humans (Flood et al., 1985;
Flood et al., 1987). In other subregions of the human hippocampus, however, including CA1 (Hanks and Flood, 1991), CA3 (Flood et al., 1987) and the subiculum (Flood, 1991), there are no changes in dendritic branching with age.
Studies of dendritic extent in other animals have, in general, confirmed that there is no regression of dendrites with age. In rats, there is no significant change in dendritic length of hippocampal granule cells between young (3 months), middle- aged (12–20 months) and aged rats (27–30 months). Although there is a trend towards an increase between middle-age (20 months) and old age (27 months)
(Flood, 1993). There is also no decrease in dendritic extent between young (3 34 months) and old rats (26 months) in CA1 (Turner and Deupree, 1991; Pyapali and Turner, 1996), although there is some evidence that a small subset of CA1 neurons from 24-month-old rats have increased basilar dendritic length and branching compared with 2-month-old rats (Pyapali and Turner, 1996; but see,
Markham et al., 2005). Unfortunately, there are no reports of age-associated changes in dendritric morphology in the parahippocampal region, and only recently has a comprehensive account of the different cell types of the PRC and their morphology been published for the young rat (Faulkner and Brown, 1999;
McGann et al., 2001; Moyer et al., 2002; Furtak et al., 2007).
Similar to the investigations on dendritic branching during aging, the data on spine density suggest that age-associated alterations are also region-specific.
Even in the hippocampus, changes in spine density are not consistent across subregions. In the dentate gyrus, there is no significant reduction in spine density in aged humans (Williams and Matthysse, 1986) or rats (Curcio and Hinds,
1983). There is also no reduction in spine density in CA1 in aged compared with young rats (Markham et al., 2005). In the subiculum of non-human primates, however, significant reductions in spine density with age have been observed in monkeys between the ages of 7 and 28 years (Uemura, 1985). The impact of age on spine density of PRC neurons is not known.
35
2.3 Biophysical properties of aged neurons of aged hippocampal neurons
In all subregions of the hippocampus most electrical properties remain constant over the lifespan (Barnes, 1994). These include resting membrane potential (Barnes, 1979; Barnes et al., 1992; Segal, 1982; Landfield and Pitler,
1984; Niesen et al., 1988; Kerr et al., 1989; Potier et al., 1993; Potier et al., 1992;
Turner and Deupree, 1991); membrane time constant (Pitler and Landfield, 1990;
Turner and Deupree, 1991; Barnes and McNaughton, 1980; Krause et al., 2008); input-resistance (Barnes and McNaughton, 1980; Segal, 1982; Landfield and
Pitler, 1984; Niesen et al., 1988; Kerr et al., 1989; Reynolds and Carlen, 1989;
Pitler and Landfield, 1990; Barnes et al., 1992; Potier et al., 1992; Potier et al.,
1993; Krause et al., 2008; but see, Luebke and Rosene, 2003); threshold to reach an action potential (Barnes and McNaughton, 1980; Niesen et al., 1988); and the width and amplitude of Na+ action potentials (Barnes and McNaughton,
1980; Segal, 1982; Niesen et al., 1988; Kerr et al., 1989; Turner and Deupree,
1991; Potier et al., 1992; Moyer et al., 1992; Potier et al., 1993). Numerous studies have shown an increase in Ca2+ conductance in aged neurons, however.
CA1 pyramidal cells in the aged hippocampus have an increased density of L-
type Ca2+ channels (Thibault and Landfield, 1996) that may lead to disruptions in
Ca2+ homeostasis (Toescu et al., 2004), potentially contributing to the plasticity
deficits that occur during aging (Foster and Norris, 1997; Landfield, 1988).
Moreover, Ca2+ activates outward K+ currents that are responsible for the after-
hyperpolarizing potential (AHP) following a burst of action potentials (Landfield 36 and Pitler, 1984; Kerr et al., 1989). Aged neurons in CA1 have an increase in the amplitude of the AHP that results, at least in part, from age-related increases in
Ca2+ conductance (Landfield and Pitler, 1984; Moyer et al., 1992). Other factors
that may contribute to the larger AHP in aged animals include reduced basal
cAMP levels (Tombaugh et al., 2005). This increased AHP in aged rats, however,
can be reversed by environmental enrichment (Kumar and Foster, 2007). The larger AHP observed in CA1 pyramidal cells has not been observed in PRC neurons (Brown TH, personal communication, 2008), indicating that Ca2+ dyshomeostasis may not occur in these cortical neurons.
The larger AHP in aged hippocampal neurons suggests that aged CA1 pyramidal cells are less excitable, as they are further from action potential threshold than are young neurons during the AHP. The only evidence that supports this notion is the finding that, in an in vitro hippocampal slice preparation, aged CA1 neurons fire fewer action potentials than do young neurons in response to a prolonged depolarization (Moyer et al., 1992). This is not the case, however, when pyramidal neurons are recorded in vivo under normal physiological conditions. In awake, behaving animals there is no difference in the firing rates of CA1 pyramidal neurons between aged and young rats (Barnes et al., 1983; Markus et al., 1994; Mizumori et al., 1996; Barnes et al., 1997; Shen et al., 1997; Tanila et al., 1997; Smith et al., 2000; Oler and
Markus, 2000; Wilson et al., 2005), and, in fact, the firing rates of CA3 pyramidal 37 neurons are actually slightly higher in the aged rat (Wilson et al., 2005). There are no reports of the biophysical properties of aged PRC, or entorhinal neurons.
2.4 Changes in hippocampal cell–cell interactions
Aged animals have alterations in the mechanisms of plasticity that contribute to cognitive functions. One functional alteration that could directly affect plasticity is reduced synapse number, which could make it more difficult to attain the sufficient amount of cooperativity among active synapses necessary for the depolarization that can lead to network modification. When all subregions of the aged hippocampus are examined together, however, there is no age-related change in the level of the synaptic protein synaptophysin, which is used as a marker of presynaptic terminals (Nicolle et al., 1999; Smith et al., 2000). In fact, changes in total synapse count between young and old animals, or additional alterations in synaptic integrity can only be detected when the connections between specific subregions are analyzed separately.
The anatomy of the hippocampus is commonly described as a tri-synaptic circuit with cortical input being transmitted to the dentate gyrus from layer II of the entorhinal cortex via the perforant path (for review, see Amaral and Witter, 1995).
Within the dendritic molecular layer of the dentate gyrus, afferent projections are anatomically distinct. The inner molecular layer of the dentate gyrus receives input primarily from the GABAergic inhibitory interneurons and the mossy cells of the polymorphic layer, while the middle and outer molecular layers get afferent 38 projections from the medial and lateral entorhinal cortices, respectively. An early electron microscopic investigation at the perforant path–granule cell synapse showed that aged rats have a 27% decrease in axodendritic synapse number in the middle molecular layer of the dentate gyrus compared with young rats
(Bondareff and Geinisman, 1976; Geinisman et al., 1977). Moreover, spatial memory deficits have been shown to correlate with a reduction in perforated synapses at the medial perforant path–granule cell synapse (Geinisman et al.,
1986). When these results were replicated with stereologically controlled measures of synapse number, the total number of synaptic contacts per neuron was found to be diminished significantly in the dentate gyrus middle molecular layer and inner molecular layer of aged rats relative to young adults (the outer molecular layer was not examined in this study), and both perforated and nonperforated axospinous synapses showed age-dependent decreases in numbers (Geinisman et al., 1992). The primary difference between the new stereological synapse count data and the old synapse count data is the observation that age-related synaptic loss involves axospinous, but not axodendritic, contacts. Consistent with these data, when the presynaptic marker synaptophysin is labeled in young and aged rats, old spatially-impaired animals show reduced levels of this protein in both the middle and outer molecular layers of the dentate gyrus, which indicates reduced synaptic contacts from axons originating in layer II of the medial and lateral entorhinal cortices, respectively
(Smith et al., 2000). Interestingly, the perforant path—CA3 synapse also shows a 39 similar decrease in synaptophysin with age in the stratum lacunosum-moleculare
(Nicolle et al., 1999). Together, these data indicate that synaptic projections originating in layer II of the entorhinal cortex are particularly vulnerable to the process of normal aging.
Electrophysiological data support the anatomical observation that there is a reduction in synapse number in the dentate gyrus. In the aged rat, the field excitatory postsynaptic potential (EPSP) recorded in the dentate gyrus is reduced
(Barnes, 1979; Barnes and McNaughton, 1980). This reduction is accompanied by a decrease in the presynaptic fiber potential amplitude at the perforant path– dentate gyrus granule cell synapse (Barnes and McNaughton, 1980; Barnes et al., 2000). Because there is no loss of entorhinal cells (Merrill et al., 2001; Rapp et al., 2002), this decrease is probably due to a reduction in axon collaterals from layer II of the entorhinal cortex to the granule cells. Interestingly, whereas the field EPSP in the aged dentate gyrus decreases, for a given magnitude of afferent fiber response, old animals show a larger synaptic field potential (Barnes and McNaughton, 1980), indicating that fewer fibers are able to elicit larger postsynaptic currents in aged animals. Consistent with this, the unitary EPSP is increased in old granule cells in response to stimulation of single afferent perforant path fibers (Foster et al., 1991). The increase in unitary EPSP size in the dentate gyrus is probably mediated by an increase in AMPA receptor currents and is suggestive of a compensatory mechanism that increases postsynaptic sensitivity in response to the reduced medial perforant path input 40
(Barnes and McNaughton, 1980; Foster et al., 1991; Barnes et al., 1992). In fact, when the NMDA receptor-mediated and AMPA receptor-mediated currents are isolated pharmacologically aged rats show enhanced AMPA receptor-mediated currents whereas the NMDA receptor-mediated field EPSP is decreased in old rats relative to young animals for given fiber potential. This suggests that there is a shift in the ratio of AMPA to NMDA receptors in the dentate gyrus over the lifespan (Yang et al. 2008; but see, Krause et al. 2008). Remarkably, the AMPA receptor-to-NMDA receptor field EPSP response ratio is not altered by extensive environmental enrichment, which significantly improves spatial memory in aged animals and the change in the ratio of AMPA receptor-to-NDMA receptor mediated currents in advanced age may be a consequence of a fixed developmental process (Yang et al., 2008).
Another possible mediator of the larger field EPSP currents in dentate gyrus granule cells is the increase in electrotonic coupling observed between aged hippocampal neurons. In fact, there is a 15% increase in the number of gap junctions between granule cells in aged compared with young rats (Fig 2.1B).
Old rats also show an increase in electrotonic coupling between neurons in CA1
(15% more gap junctions; Fig 2.1D) and CA3 (18% more gap junctions) compared with young rats (Barnes et al., 1987).
A reduction in axospinous synapses in the dentate gyrus is correlated with spatial memory deficits in aged rats (Geinisman et al., 1986; deToledo-Morrell et al., 1988). This is not the case for Schaffer collateral–CA1 synapses, as the total 41 synapse number remains the same across different age groups (Geinisman et al., 2004). Additionally, when the layer III entorhinal cortex perforant path–CA1 synapse is examined in young and aged rats using levels of the post-synaptic protein synaptophysin there is no difference between age groups (Smith et al.,
2000). Together these data indicate that the number of synaptic contacts in CA1 from two primary sources of input do not change in advanced age. Although synapse number in CA1 does not change with age, when the postsynaptic density (PSD) area of axospinous synapses in CA1 is compared between aged learning-impaired and learning-unimpaired rats, the impaired animals show a profound reduction in the PSD area of perforated synapses (Nicholson et al.,
2004). These findings support the notion that many hippocampal perforated synapses become non-functional or silent in aged learning-impaired rats, and this loss of functional synapses may contribute to the cognitive decline during normal aging.
Electrophysiological data also support the hypothesis that there is a loss of functional synapses in CA1. The amplitude of the field EPSP recorded in CA1 of aged rats is reduced, compared with young rats (Landfield et al., 1986; Barnes et al., 1992; Deupree et al., 1993). The unitary EPSP size in the CA1 region of the hippocampus in aged rats is preserved, however (Barnes et al., 1992), and there is no change in the amplitude of the Schaffer collateral pre-synaptic fiber potential (Barnes et al., 1997; Rosenzweig et al., 1997; Barnes et al., 2000), indicating that there is no axonal pruning of aged Schaffer collateral fibers. 42
Interestingly, the reduced field EPSP in aged rats can be reversed by protein phosphatase inhibitors suggesting that there could be an imbalance in kinase/phosphotase activity in aged CA1 neurons that contributes to reduced synaptic efficacy (Norris et al., 1998).
Although the PSD area of perforated synapses declines most significantly in aged learning-impaired Long Evans rats (Nicholson et al., 2004), the electrophysiological data indicate that the field EPSP amplitude is reduced in both impaired and unimpaired aged F344 rats (Tombaugh et al., 2002).
Importantly, plasticity mechanisms were only defective in the learning-impaired aged group (Tombaugh et al., 2002). Combined, these data implicate perforated synapse PSD area in plasticity mechanisms independent of fast synaptic transmission processes.
The effects of altered morphology, biophysical properties and synaptic connections of aged neurons on plasticity can be assessed by measuring age- associated alterations in long-term potentiation (LTP) and long-term depression
(LTD). These two processes can be divided into an induction phase (early-phase
LTP/LTD) and a maintenance phase (late-phase LTP/LTD). The induction phase involves the temporal association of presynaptic glutamate release with postsynaptic depolarization (necessary to eject Mg2+ ions from the pores of
NMDA receptors), which results in an increase in intracellular Ca2+ (Bliss and
Collingridge, 1993). Additionally, the NMDA receptor cannot be activated without
D-serine or another endogenous agonist binding to the NMDA receptor glycine 43 modulatory site (Mothet et al., 2000; Yang et al., 2003). All functional NMDA receptors are heterotetramers that are composed of both glycine binding NR1 subunits and glutamate binding NR2 subunits. Aging affects the conformation, distribution, and function of NMDA receptors within the hippocampus. Overall,
CA1 and CA3 shows reduced NMDA receptor binding in aged animals compared to young rats but intact AMPA receptor binding while the aged dentate gyrus shows intact NMDA receptor binding but reduced AMPA receptor binding (Wenk and Barnes, 2000). It was unclear from this investigation, however, if these age- related differences were due to a decline in actual receptor number, or other conformational changes that could affect receptor binding. When the levels of
NMDA receptor subtypes are measured individually, in the dentate gyrus of both aged rats (Newton et al., 2008), and aged monkeys (Gazzaley et al., 1996) there is a decrease in the levels of the NR1 subunit. Normal levels of the NR2 subunit in the dentate gyrus could explain why glutamate-NMDA receptor binding does not appear to be reduced in the aged dentate gyrus (Wenk and Barnes, 2000). In contrast, NR1, NR2A, and NR2B subunits show a reduction in the CA3 subregion of aged rats relative to young animals (Adams et al., 2008), and NR1 levels are decreased in aged compared to young animals in CA1 (Liu et al., 2008).
Interestingly, the endogenous ligand of the NR1 glycine modulatory site D-serine is reduced in aged rats compared to young animals (Junjaud et al., 2006), and this may contribute to age-associated plasticity deficits in LTP induction since
NR1 binding must occur to activate the receptor. 44
LTP maintenance is the continued expression of increased synaptic efficacy that persists after induction. It probably involves changes in gene expression and insertion of AMPA receptors into the postsynaptic membrane
(Malinow and Malenka, 2002). Aged rats have deficits in both LTP induction and maintenance. These deficits, however, are complex, protocol-dependent and region-specific.
Although there is a reduction in the field EPSPs recorded both in the dentate gyrus (Barnes and McNaughton, 1980; Foster et al., 1991) and in CA1
(Landfield et al., 1986; Barnes et al., 1992; Deupree et al., 1993), aged animals can show intact LTP induction at the perforant path–granule cell synapse
(Barnes, 1979; Diana et al., 1994; Diana et al., 1994), the CA3–CA1 Schaffer collateral synapse (Landfield and Lynch, 1977; Landfield et al., 1978; Kumar et al., 2007) and the perforant path–CA3 pyramidal cell synapse (Dieguez and
Barea-Rodriguez, 2004) when robust high-frequency, high current amplitude stimulation protocols are used (Fig 2.2A). Even when supra-threshold stimulation parameters are used, however, aged rats have a deficit in the maintenance of
LTP in both the dentate gyrus (Barnes, 1979; Barnes and McNaughton, 1980) and CA3 compared with young rats (Figure 2.2A; Dieguez and Barea-Rodriguez,
2004).
When peri-threshold stimulation parameters are used, LTP induction deficits can be observed in both the dentate gyrus and CA1. In the dentate gyrus, when weak presynaptic stimulation is combined with direct depolarization of the 45 granule cell, larger amplitude current injection is required to elicit LTP at the perforant path–granule cell synapse of aged rats compared with young rats
(Barnes et al., 2000). This indicates that aged granule cells in the dentate gyrus have an increased threshold for LTP induction. Additionally, when a 50% maximum current intensity is delivered to the perforant path in 10 bursts of 400
Hz to awake young and aged rats, only the young animals show a significant increase in the dentate gyrus field EPSP (Figure 2.2B; Sierra-Mercado et al.,
2008). A similar induction deficit is observed in aged rats at the perforant path- dentate gyrus granule cell synapse when tetanic stimulation is used (Dhanrajan et al., 2004). Interestingly, this impairment in LTP induction in aged rats can be alleviated by briefly exposing the old animals to novelty and then applying the peri-threshold LTP induction procedure in the novel environment (Sierra-Mercado et al., 2008). Novelty exposure, however, has little effect on the LTP maintenance deficits of the aged animals (Sierra-Mercado et al., 2008).
The pattern of age-related LTP deficits in CA1 pyramidal cells is different from that observed in the dentate gyrus. Aged neurons in CA1 do not have an increased threshold for LTP (Barnes et al., 1996), but when peri-threshold stimulation parameters are used, the level of LTP induction in aged rats is less than in young rats (Figure 2.2B; Deupree et al., 1991; Moore et al., 1993;
Rosenzweig et al., 1997; Tombaugh et al., 2002). For example, when LTP induction is measured in young and old rat hippocampal slices using four-pulse stimulation at 100 Hz (Deupree et al., 1991) or the primed-burst protocol, in 46
Figure 2.2: Summary of age-related alterations in LTP and LTD between young and aged animals. The Y-axes show the change in EPSP slope following LTP or LTD induction and the X-axes show the retention intervals for maintenance of LTP or LTD (red lines - young rats; green lines - aged rats). A. When supra-threshold stimulation parameters are used, LTP induction is intact at old hippocampal synapses (Barnes, 1979; Landfield et al., 1978; Dieguez and Barea-Rodriguez, 2004) but decay over days in the DG (Barnes, 1979) and CA3 (Dieguez and Barea-Rodriguez, 2004) is faster. B When peri-threshold stimulation parameters are used, aged rats can show LTP induction deficits (Sierra-Mercado et al., 2008; Moore et al., 1993; Barnes et al., 1996; Deupree et al., 1991). C In CA1, aged rats are more susceptible to LTD induction (Norris et al., 1996). LTD induction with low-frequency stimulation (LFS) in old rats, however, can be attenuated by agents, such as cyclopiazonic acid, that prevent the release of Ca2+ from internal Ca2+ stores (Kumar and Foster, 2005). D Aged rats are also more susceptible than are young rats to the reversal of LTP. The increase in EPSP slope that results from LTP-inducing stimuli can be attenuated by the application of LFS to the potentiated pathway. In young rats, LTP is not completely reversed by LFS and there is still some residual potentiation. In old rats, however, LFS returns the EPSP slope to the baseline pre-LTP levels (Norris et al., 1996). Figure was updated from Burke and Barnes, 2006. which a single priming pulse is followed 170 ms later by four stimulus pulses at
200 HZ (Moore et al., 1993), the increase in field EPSP slope is less in the aged rats than in the young rats. These induction deficits occur even if the stimulus 47 intensity of the Schaffer collaterals is increased to match field EPSP amplitude between young and aged rats (Rosenzweig et al., 1997). Although the aged rats’
Schaffer collateral axons can follow high-frequency stimulation as well as those of young rats, aged CA1 neurons show weaker temporal summation of the multiple EPSPs induced by high-frequency stimulation. Therefore, during high- frequency bursts, CA1 pyramidal cells are less depolarized, which explains the age-related LTP induction impairment in CA1 (Rosenzweig et al., 1997). One source of the decline in temporal summation that is observed in aged CA1 pyramidal cells could be the age-associated decline in D-serine, the endogenous ligand of the NR1 glycine modulatory site of the NMDA receptor, which is required for the receptor to be activated (Junjaud et al., 2006). In support of this idea is the observation that administration of D-cycloserine, which increases the levels of D-serine in the brain, can increase the induction of theta-burst LTP in
CA1 of aged rats (Billard and Rouaud, 2007).
It is possible that age-related changes in Ca2+ regulation cause some
portion of the observed age-related plasticity deficits. In particular, it has been proposed that postsynaptic intracellular levels of Ca2+ are involved in setting the
synaptic modification threshold. This threshold may then affect the probability
that a synapse will be depressed or potentiated at a given time (Bear et al., 1987;
Bear and Malenka, 1994; Foster and Norris, 1997). Ca2+ dyshomeostatis in aged
animals (Landfield, 1988; Thibault and Landfield, 1996; Foster and Norris, 1997)
could therefore alter the probability that synaptic activity will induce either LTP or 48
LTD. This idea is supported by Ca2+ imaging studies, which have shown that resting Ca2+ does not differ substantially with age in CA1. Greater elevation of
somatic Ca2+ and greater depression of EPSP frequency facilitation, however,
develop in aged CA1 neurons in response to stimulation (Thibault et al., 2001).
In line with the Ca2+ hypothesis of age-related plasticity impairments is the finding
that aged rats are more susceptible than are young rats to LTD (Fig 2.2C; Norris
et al., 1996) and to the reversal of LTP (Fig 2.2D; Foster and Norris, 1997).
Moreover, inhibition of Ca2+ release from intracellular Ca2+ stores attenuates LTD induction in aged CA1 neurons (Fig 2.2C; Kumar and Foster, 2005), while the administration of a Ca2+ chelator to aged rats increases LTP induction at the
Schaffer collateral-CA1 synapse and improves spatial memory (Tonkikh et al.,
2006). Finally, blocking L-type Ca2+ channels with the antagonist nifedipine
eliminates the induction of LTD following 1 Hz stimulation in aged rats and increases LTP induction following 5 Hz stimulation (Norris et al., 1998). Together these data indicate that Ca2+ dysregulation in aged CA1 pyramidal cells is
involved in the observed plasticity deficits. It has recently been shown, however,
that while both learning impaired and learning unimpaired aged rats have a
reduction in NDMA receptor-dependent LTP at the Schaffer collateral-CA1
synapse relative to young animals, the unimpaired aged rats only show an
increase in NDMA receptor-independent LTP when compared to young and aged
spatially impaired rats (Boric et al., 2008). This form of LTP relies on voltage-
gated L-type Ca2+ channels making these data puzzling since increases in L-type 49
Ca2+ channels and altered Ca2+ homeostasis are associated with cognitive
dysfunction (Campbell et al., 1996; Thibault and Landfield, 1996; Thibault et al.,
2001). Additionally, blockade of L-type Ca2+ channels can improve synaptic
plasticity and (Norris et al., 1998) cognition in aged animals (Veng et al., 2003).
The relative roles of L-type Ca2+ channels and metabotropic glutamate receptors
in eliciting NMDA receptor-independent LTP in CA1, however, was not
determined by Boric and colleagues (2008), nor were the currents through
voltage-gated calcium channels quantified between the different age groups and
the spatially-impaired and spatially unimpaired animals. Thus, enhanced NMDA
receptor-independent LTP might have occurred only in the age-unimpaired rats
because these animals had intact metabotropic glutamate receptor function in
CA1 of the hippocampus that is blunted in the aged impaired rats (Nicolle et al.,
1999). The exact role of the L-type Ca2+ channel and the metabotropic glutamate
receptor with regards to normal aging and plasticity needs to be explored further.
In summary, during the normal aging process, animals experience age-
related cognitive decline. Historically, it was thought that a primary contribution to
the etiology of this decline was massive cell loss (Brody, 1955) and deterioration
of dendritic branching (Scheibel et al., 1976; Scheibel, 1979). Now we know,
however, that the changes occurring during normal aging are more subtle and
selective than once believed. In fact, the general pattern seems to be that most
age-associated behavioral impairments result from region-specific changes in
dendritic morphology, cellular connectivity, and Ca2+ dysregulation, or other 50 factors that affect plasticity and ultimately alter the network dynamics of neural ensembles that support cognition. 51
CHAPTER 3: AGING ENSEMBLES: CIRCUIT CONTRIBUTIONS
TO MEMORY DEFICITS
3.1 Introduction
The normal aging process is accompanied by a decline in spatial and other forms of hippocampal-dependent memory (for review, see Rosenzweig and
Barnes, 2003; Wilson et al., 2006). As summarized in the previous chapter (see
Chapter 2), it is known that the neurobiological underpinnings of this cognitive decline involve subtle, region-specific changes in synapses, that alter mechanisms of plasticity rather than gross changes in cell number or neuron morphology. These observations imply that it will be productive to examine the interactions among, and the temporal dynamics of the neural ensembles in brain structures, such as the hippocampus that ultimately support learning and memory.
Advances in multiple single unit recording methods (Figure 3.1) have allowed the dynamics of hippocampal cell populations to be investigated in freely behaving rats (Wilson and McNaughton, 1993), and studies using these methods have shown that certain properties of these networks are compromised during aging. Interestingly, many of the age-related changes that have been discovered can be linked to plasticity deficits, because blockade of NMDA receptors in young rats results in ensemble dynamics that resemble those of aged rats (Kentros et al., 1998; Ekstrom et al., 2001). This chapter will review what is currently known about aged neural ensembles in the hippocampus and how alterations in the 52 dynamics of these circuits are linked to memory decline. Unfortunately, the focus of this chapter is limited to the hippocampus because relatively little is known about the impact of aging on the other structures of the medial temporal lobe.
Figure 3.1: Investigating Neuronal Ensembles with Multiple Single Unit Recording Methods. Recordings of more than 100 cells can be obtained from a ‘hyperdrive’ device that is permanently mounted on a rat’s head (Wilson and McNaughton, 1993; Gothard et al., 1996), enabling the recording of extracellular action potentials in freely moving animals. The tetrode recording probe used consists of four twisted 13μm wires, each providing a different recording channel (Wilson and McNaughton, 1993; Gothard et al., 1996; McNaughton et al., 1983). Cells can be distinguished from each other offline on the basis of the relative amplitude differences of their spikes. (A) Shows analogue waveforms from five hippocampal cells recorded from the four tetrode channels (different cells are shown in different colors). (B) Shows the amplitude distributions of the neurons from A. The top panel is peak amplitude on channel 1 compared with channel 2, and the bottom panel is the peak amplitude on channel 3 compared with channel 4. Note that individual cell amplitudes cluster distinctively for the different cells. Statistical clustering methods can be applied to data to identify individual cells enabling the rat’s behavior to be correlated with the activity of single neurons. For example, principal cells of the hippocampus will fire selectively when a rat is in a specific region of the environment (O'Keefe and Dostrovsky, 1971). The area of the environment where a hippocampal principal cell is active is referred to as the cell’s ‘place field’. (C) Place fields of seven CA1 pyramidal neurons when the rat traversed a circular track. Small dots correspond to individual spikes and the spikes from different neurons are shown in different colors (Burke et al., unpublished data). Multiple single unit recordings have been used to reveal differences in place cell ensemble dynamics between young and aged rats (see text). Figure from Burke and Barnes, 2006.
53
3.2 Fundamental Properties of Place Cells in Young and Old Rats
Neuronal recordings from the hippocampus of adult rats reveal that when a rat explores an environment, pyramidal (O'Keefe and Dostrovsky, 1971) and granule cells (Jung and McNaughton, 1993; Leutgeb et al., 2007) show patterned neural activity that is highly correlated with a rat’s position in space (the ‘place field’ of the cell; O'Keefe, 1976). Approximately 30–50% of CA1 pyramidal cells show place-specific firing in a given environment of intermediate size (e.g. Wilson and McNaughton, 1993; Muller et al., 1987; Gothard et al., 1996), earning these neurons the name ‘place cells’. When the firing patterns of many hippocampal neurons are recorded simultaneously, it is possible to reconstruct the position of a rat within an environment from the place cell firing patterns alone (Wilson and
McNaughton, 1993). The composite cell activity is ‘map-like’ and, in different environments, hippocampal place maps change markedly. Although these maps can be controlled by external environmental features (e.g. Knierim et al., 1995), internal events are also important and a new map may be generated in the same environment if the demands of the task change (Markus et al., 1995).
Given the extensive age-associated impairments in behaviors that are hippocampal dependent (for review, see Rosenzweig and Barnes, 2003) it is surprising that the basic characteristics of CA1 place fields such as: spike amplitude, spike width, mean and maximum firing rates, and inter-spike interval
(ISI) distributions are remarkably preserved in advanced age (e.g. Barnes et al.,
1983; Markus et al., 1994; Mizumori et al., 1996; Barnes et al., 1997; Shen et al., 54
1997; Oler and Markus, 2000). Moreover, upon the initial pass through a place field, or during a task in which the rat’s trajectory is not restricted (i.e., random foraging), the place fields of aged rats are just as specific as those of young rats
(Markus et al., 1994; Mizumori et al., 1996; Barnes et al., 1997; Shen et al.,
1997; Tanila et al., 1997; Oler and Markus, 2000). Figure 3.2 shows an example of the place field activity of an aged rat. Qualitatively, this CA1 activity is remarkably similar to what has been observed in young animals.
Figure 3.2: CA1 place fields in old rats. Each panel shows a firing rate map for one of 16 representative CA1 pyramidal cells that were active while an aged rat (27-months old) randomly foraged for chocolate sprinkles in a familiar environment. Each box reflects the environment explored by the rat in this experiment. Red indicates portions of the environment where the firing rate was high while blue indicates areas where the neuron did not fire. The CA1 place field characteristics of old rats during a random foraging task are not qualitatively different from those of a young rat (from Burke, unpublished data). In contrast, the basic properties of CA3 place fields show at least two age- related changes. First, the firing rates of CA3 pyramidal cells are higher in aged rats compared to young animals (Wilson et al., 2005). Additionally, during a 55 random foraging task, CA3 place fields are larger in old rats compared to young animals (Wilson et al., 2005). The observation that CA3 place fields were larger needs to be explored further, however. One potential confound is that the authors did not measure place field coherence over time, therefore, the larger place fields could simply have been due to CA3 place field location drifting more in the aged compared to the young rats. It is also known that CA3 place fields take more time to become stable relative to CA1 place fields (Leutgeb et al., 2004). In this experiment the rats were only exposed to each environment for 7 minutes and this may not have been enough time for the CA3 ensemble activity to fully develop. Moreover, the time course for place field development could be different between young and aged animals. These questions need to be explored, and the behavioral implications of age-related changes in CA3 pyramidal cell firing are just beginning to be elucidated (Wilson et al., 2005; for review, see Wilson et al.,
2006).
While there are no reports of electrophysiological recordings from granule cells in awake-behaving old rats, it is possible to label the neurons that were activated during an epoch of behavior by monitoring the expression of immediate early genes. Immediate-early genes (IEGs) are a class of genes that do not require the preceding activation of any other responsive genes or de novo protein synthesis to be expressed. Consequently, the transcription of IEGs is thought to be dynamically regulated by specific forms of patterned synaptic activity believed to underlie information storage (Cole et al., 1989). Moreover, it has been shown 56 that IEGs are expressed by cells that are activated during spatial exploration
(Guzowski et al., 1999). One of the IEGs of interest is the gene for activity- regulated cytoskeletal (Arc) protein. Arc codes for an effector protein that is involved in AMPA receptor trafficking (Chowdhury et al., 2006; Rial Verde et al.,
2006; Shepherd et al., 2006), and has been shown to be necessary for the maintenance of long-term memory, long-term potentiation (Guzowski et al.,
2000), and the induction and maintenance of long-term depression (Plath et al.,
2006).
Changes in the proportion of cells that express a particular immediate- early gene during behavior can be assessed using fluorescence in situ hybridization. This allows exact determination of which individual cells are expressing which genes. Using this method, it was determined that granule cells of the dentate gyrus, but not the pyramidal cells of CA1 and CA3, in aged rats
Figure 3.3 Arc expression in the dentate gyrus. The proportion of cells expressing the immediate-early gene Arc, following behavioral induction can be measured in specific brain structures with fluorescence in situ hybridization (from Small et al., 2004). Confocal images of fluorescence in situ hybridization for Arc mRNA in the dentate gyrus of a young rat (left panel) and an old rat (right panel). Granule cells are shown in red and Arc mRNA in yellow. More granule cells are positively labeled for Arc in young, relative to the old rats after spatial exploration. Figure from Small et al., 2004. 57 have a significantly smaller proportion of neurons that express Arc following spatial exploration (Small et al., 2004). These data suggest that fewer granule cells would express place fields in a given environment (Figure 3.3).
Alternatively, it is also possible that the same proportion of dentate gyrus granule cells express place fields but that during advanced age the transcription of Arc becomes de-coupled from the granule cell activity, which has been shown to occur in young CA1 pyramidal cells under certain circumstances (Guzowski et al., 2006; Fletcher et al., 2006). One mechanism by which Arc transcription could become decoupled from cellular activity in the aged dentate gyrus is through functional alterations in the NMDA receptor. NDMA receptor activation is necessary for the transcription of immediate-early genes (Cole et al., 1989).
There are two potential age-related changes in NMDA receptor function in the dentate gyrus that could increase the threshold for Arc induction in aged granule cells thereby decoupling cellular activity during behavior from Arc transcription.
First, aged rats show a reduction in NDMA receptor-mediated currents at the perforant path-dentate gyrus synapse following perforant path stimulation even when the reduced perforant path input is controlled for (Yang et al., 2008).
Additionally, levels of the NR1 subunit of the NDMA receptor are reduced in aged relative to young animals (Gazzaley et al., 1996; Newton et al., 2008). Because the NR1 subunit is necessary for binding glycine in order to activate the NDMA receptor, this could also contribute to less NMDA receptor activity thereby reducing Arc transcription. Neural recordings from the dentate gyrus of awake, 58 behaving young and old rats are required, however, to discriminate between differences in Arc transcription versus overall reduced dentate gyrus place field activity in aged animals. Interestingly, using other imaging methods, in humans and monkeys, the granule cells also seem to be particularly vulnerable to the effects of normal aging (Small et al., 2002; Small et al., 2004).
3.3 Spike timing in the aged hippocampus
In all subregions of the hippocampus the timing of principal neuron spikes shows a dynamic relationship to the hippocampal theta rhythm, a 5-10 Hz oscillation that is prominent in the rat hippocampus during active exploration
(Vanderwolf, 1969) and REM sleep (Vanderwolf et al., 1977). As a rat passes through a hippocampal principal cell’s place field, the timing of the spikes shows a clear shift relative to the local theta rhythm such that after entry into the place field the pyramidal cell discharges at earlier phases of the theta cycle (O'Keefe and Recce, 1993). Specifically, the first spikes appear late in the theta cycle but as the rat continues to advance through the place field the spikes shift to progressively earlier phases of theta (Figure 3.4A; O'Keefe and Recce, 1993;
Skaggs et al., 1996). For the majority of place fields recorded from CA1 pyramidal cells, this “theta phase precession” (O'Keefe and Recce, 1993) covers almost 360 degrees of phase shift, but not more (O'Keefe and Recce, 1993;
Skaggs et al., 1996; Shen et al., 1997; Ekstrom et al., 2001; Maurer et al., 2006).
This observation indicates that the rate of theta phase precession must be 59
Figure 3.4. Theta phase precession in young and old rats. (A) A schematic diagram of theta phase precession. As a rat passes through a principal cell’s place field, the timing of the spikes show a monotonic shift in their phase of firing such that as the rat first enters the field, the spikes occur late in phase but as the rat exits the field, the spikes occur relatively early in phase. For most CA1 pyramidal cells, place fields exhibit 360 degrees of phase advance and not more. (B) Shows phase by position plots for a young (left panel; open circles) and an old (right panel; filled circles) rat. The slope of this phase by position plot is significantly greater in the old rat compared to the young rat, indicating that while both young and old rats exhibit theta phase precession, the rate of precession is greater in old rats compared to young rats because the size of the field is larger in the young rats after the first traversals of the route. When the size of the place field is smaller, as in old rats, the spikes of the pyramidal cells need to, on average, precess more degrees per centimeter traveled by the rat in order to achieve 360 degrees of precession during a single pass through the place field. The result is a greater rate of theta phase precession. dynamically coupled to the size of a neuron’s place field, such that smaller place fields show a faster rate of precession compared to larger place fields (Shen et al., 1997; Ekstrom et al., 2001; Terrazas et al., 2005; Maurer et al., 2006).
Therefore, a single 360 degree cycle of phase precession can be used as an objective measure of place field size, as this definition does not require an arbitrary experimenter-defined threshold like the traditional firing rate measures of place field size require (Maurer et al., 2006). Moreover, it is possible for a single CA1 neuron to have multiple place fields that overlap in space. These 60 overlapping place fields can be distinguished from one another based on their discrete and independent cycles of theta phase precession. This feature of the phase precession definition of a place field is particularly advantageous compared with other methods that use only firing rates to determine place field boundaries, because the latter methods cannot easily distinguish between overlapping place fields (Maurer et al., 2006).
In the CA1 subregion of the hippocampus, pyramidal cells in both old and young rats exhibit theta phase precession (Shen et al., 1997). Moreover, there is no difference between the total phase change of place fields in old rats compared to young rats such that the place field spikes begin and end at approximately the same phase of theta in both young and old rats (Shen et al., 1997). Additionally, a complete cycle theta phase precession continues to occur after administration of an NMDA receptor antagonist (Ekstrom et al., 2001). This observation that theta phase precession does not require NMDA receptor activity suggests that this is not a dynamic phenomenon that requires plasticity, but rather that an asymmetry in the hippocampal CA1 network is established early in development, and this asymmetry is a likely source of theta phase precession in both young and aged rats. In support of this idea, we have recently shown that a full 360 degree cycle of theta phase precession occurs during the first pass through a place field in young rats, aged rats, and young rats administered the NMDA receptor antagonist CPP (Figure 3.5; Burke et al., 2007). 61
Figure 3.5. Phase precession during the first pass through a place field. Phase by normalized position (z-score) plots for all place fields recorded within a given age group and treatment condition. In these plots the direction of movement in from right to left. The black dots indicate the phase and normalized position for spikes that occurred during the first lap of track running while red dots indicate spikes that occurred during lap 15. Even during the first pass through a place field (lap 1) place field spikes occur during all phases of theta in all three groups of animals and are the phase by positions profiles of CA1 spikes during the first lap are indistinguishable from lap 15 in: 1) young rats, 2) young rats given the NMDA receptor antagonist CPP, and 3) old rats. This indicates that a complete cycle of theta phase precession occurs during the first pass through a place field independent of NMDA receptor activity. Note that there are fewer points in plot 2 because less neurons where recorded in the young rats during the CPP condition (Burke et al., 2007).
Interestingly, the place fields of CA1 pyramidal cells are smaller in old rats compared to young rats when the animal repeatedly takes the same path. This difference in place field size is due to a lack of NMDA receptor-dependent place field expansion plasticity in old rats (Mehta et al., 1997; Shen et al., 1997; Burke et al., 2008), which will be discussed more extensively in a later section. Because there is no difference in the total magnitude of phase advance between young and old rats, this indicates that the rate of phase precession must be greater in the old rats in order for the same magnitude of phase advance to be achieved in a smaller portion of space. Empirically, this has been shown and the slope of a phase versus location plot is significantly steeper in old rats (Figure 3.4B), indicating a more rapid phase precession with distance (Shen et al., 1997). A similar observation has been made in young rats given an NMDA receptor antagonist (Ekstrom et al., 2001). Therefore, phase precession is a basic 62 property of CA1 place fields that does not require NMDA receptor-dependent plasticity.
In addition to theta phase precession, other basic properties of place fields are maintained in advanced age. It is well documented that as the running speed of an animal increases the firing rate of dorsal CA1 neurons also increases proportionally (McNaughton et al., 1983). Even though old rats run at significantly slower velocities compared to young animals, the relationship between firing rate and the velocity is similar between young and old rats (Shen et al., 1997; Burke et al., 2008). Moreover, theta frequency increases with running speed similarly in both young and aged rats (Shen et al., 1997).
Although CA3 place fields show an increase in size and firing rate in aged compared to young rats, and there is a decrease in the proportion of active granule cells during spatial exploration, it is still unclear how these changes result in behavioral impairments. Because theta phase precession is tightly coupled to place field size, it is possible that aged CA3 neurons have a slower rate of phase precession or that during a single pass through a place field more than 360° of a phase advance occurs. These two possibilities need to be tested empirically.
Conversely, reports of age-related changes in the ensemble dynamics of CA1 place fields have been correlated with observed behavioral impairments (Barnes et al., 1997; Rosenzweig et al., 2003) in aged rats, but these changes can only be measured and understood by monitoring the activity of many hippocampal neurons as animals perform behaviors that activate hippocampal neurons. 63
3.4 Dynamic properties of hippocampal place cells in young and old rats
It is widely agreed that modifiable neuronal ensembles support cognition.
Therefore, alterations in how these networks adapt in response to stimuli or behavioral experience could lead to cognitive impairments observed with aging.
The physiological mechanisms by which synaptic strength can be altered in response to environmental stimuli and/or experience are similar to the experimentally-induced forms of plasticity such as LTP and LTD that were reviewed in the previous chapter. These forms of plasticity are believed to represent mechanisms by which the synaptic weight matrix can adapt and ultimately lead to modifications that reflect learning in the system. Findings of age-associated impairments in experimentally-induced LTP and LTD suggest that, while the basic properties of CA1 place fields are similar between young and aged rats, the dynamic properties of place fields that require NMDA receptor-dependent plasticity could be altered during the aging process.
Moreover, these alterations could be related to hippocampal-dependent age- related memory decline.
The initial expression of CA1 place fields within an environment does not require NMDA receptor-dependent plasticity (Kentros et al., 1998). Similarly, in young and old rats, the same proportion of CA1 neurons is active (Small et al.,
2004) and these cells express the same number of place fields in a single environment (Shen et al., 1997). There are, however, plasticity-dependent properties of place fields that are similarly disrupted both when the NMDA 64 receptor is blocked in young rats and in advanced age. Among these dynamic place field characteristics are experience-dependent place field expansion and hippocampal map stability.
In young rats CA1 place fields expand asymmetrically during repeated route following (for example, traversing a circular track), which results in a shift in the center of mass of place fields in the direction opposite to the rat’s trajectory
(Figure 3.6A; Mehta et al., 1997). This observation is consistent with neural network models dating back to Hebb's (1949) concept of the 'phase sequence' of cell assemblies, which have suggested that an associative, temporally asymmetric synaptic plasticity mechanism could serve to encode sequences or episodes of experience (Hebb, 1949). The magnitude of this place field expansion, however, significantly decreases in aged rats (Figure 3.6B; Shen et al., 1997). It is likely that this age-associated reduction in experience-dependent plasticity is due to LTP-like deficits, as place field expansion does not occur when young rats are administered the NMDA receptor antagonist CPP (Ekstrom et al.,
2001). Specifically, plasticity deficits at the Schaffer collateral CA3-CA1 synapse in old animals (Rosenzweig et al., 1997; Norris et al., 1996; Dieguez and Barea-
Rodriguez, 2004; Tombaugh et al., 2002; Moore et al., 1993; for review, see
Burke and Barnes, 2006) could be responsible for the diminished place field expansion in aged rats.
In young animals a rapid behaviorally-induced potentiation of feed-forward synapses from CA3 to CA1 can explain the experience-dependent expansion of 65
CA1 place fields (Mehta et al., 2000). During the first pass through a place field, a dorsal CA1 neuron may inherit its place-specific firing from neurons in layer III of the medial entorhinal cortex (Brun et al., 2002; Brun et al., 2008), and input to the CA1 neuron from CA3 would be relatively symmetric; that is, the CA1 neuron receives equal input from CA3 neurons with place fields just before and just after the location of the CA1 neuron place field. After repeated traversals, in which both directionally-selective CA3 and CA1 neurons are activated, a spike-timing dependent plasticity mechanism would strengthen synapses from those CA3 neurons that had a place field just before that of a CA1 neuron (Mehta et al.,
2000; Lee et al., 2004). The result is a backward, asymmetric shift of CA1 place fields (Figure 3.6A). If behaviorally-induced plasticity of the CA3 to CA1 Schaffer collateral synapse is the primary mechanism for place field expansion, then the decrease in LTP induction and the increase in LTD induction at this synapse could explain why aged rats fail to show this phenomenon under normal conditions. Interestingly, it has recently been shown that memantine (approved for treatment of the cognitive disorders associated with Alzheimer’s disease) can, at least partially, reinstate experience-dependent place field expansion plasticity in aged rats (Burke et al., 2008). Specifically, acute administration of memantine
20 minutes prior to an episode of track running results in an increase in the size of place fields, a shift in the center of mass of place fields, and an increase in spike number over the first several laps of tracking running (Figure 3.6C). The efficacy of memantine at reinstating experience-dependent place field expansion 66 plasticity may be attributed to memantine’s pharmacokinetics and binding affinity for the NMDA receptor (Rogawski and Wenk, 2003; Parsons et al., 1995).
Figure 3.6: Experience-dependent place field expansion in young and old rats. (A) A schematic diagram of experience-dependent place field expansion plasticity that occurs in young rats as they repeatedly traverse the same route traveling through a CA1 pryamidal cell’s place field multiple times. During the first pass through the place field, the field is symmetrical (shaded distribution). As the rat passes through the field multiple times, however, the field expands in the direction opposite to the rat’s trajectory, leading to a shift in the center of mass (COM) of the place field’s spikes. (B) Shows a plot of place field size by lap. Young rats show significant place field expansion plasticity (open circles) while old rats fail to show this form of experience-dependent plasticity robustly (filled circles; data from Shen et al., 1997). (C) Old rats given memantine (10 mg/kg; open circles) show improved place field expansion plasticity compared to saline controls (filled circles; data from Burke et al., 2008).
The observed plasticity deficits at the CA3-CA1 Schaffer collateral synapse in aged rats can be attributed, in part, to age-associated alterations in the regulation of intracellular Ca2+ levels. Aged CA1 pyramidal cells have increased Ca2+ conductances due to a higher density of L-type Ca2+ channels
(Thibault and Landfield, 1996), which was discussed more extensively in the previous chapter. This may lead to disruptions in Ca2+ homeostasis (for review, see Toescu et al., 2004) that ultimately contribute to age-related plasticity deficits 67
(Landfield, 1988; Foster and Norris, 1997; Kumar and Foster, 2005). Moreover, it has been hypothesized that post-synaptic intracellular levels of Ca2+ are involved
in setting the synaptic modification curve, which determines the probability that a synapse will be depressed or potentiated given the pattern of input (Bear et al.,
1987; Bear and Malenka, 1994; Foster and Norris, 1997). Calcium dyshomeostatis in aged animals (Landfield, 1988; Thibault and Landfield, 1996;
Foster and Norris, 1997) could therefore shift this modification curve thereby altering the probability that synaptic activity will induce either LTP or LTD. Figure
3.7 shows hypothetical synaptic modification curves for a young rat (green) and an aged rat (purple). The higher levels of intracellular Ca2+ following synaptic
activity in aged compared to young rats shifts this curve to the right. As a result,
even though the threshold for LTP is unchanged in the aged rats, the old animals
Long-term potentiation
s t a r
g n u o threshold Y
ts a r d e g A Long-term depression Synaptic modification Synaptic
Input Figure 3.7: Hypothetical synaptic modification curves for young (green) and aged (purple) CA1 neurons. The X-axis represents presynaptic input and the Y-axis is synaptic modification. In young and old CA1 neurons there is no change in the threshold to induce LTP, however, Ca2+ dyshomeostasis in aged CA1 neurons causes this curve to shift to the right. This results in a greater probability of LTP accompanied by a lower probability of LTP in aged CA1 neurons. 68 have an increased probability for LTD accompanied by a decreased probability for LTP when peri-threshold input is administered. In line with the Ca2+
hypothesis of age-related plasticity impairments, is the finding that the inhibition
of Ca2+-induced Ca2+ release from intracellular Ca2+ stores attenuates LTD
induction in aged CA1 neurons (Kumar and Foster, 2005). If the disruption in experience-dependent CA1 place field expansion in aged rats is related to Ca2+ dyshomeostasis then it is possible that reducing the amount of Ca2+ entering a
neuron through the activated NMDA receptor may reinstate this behaviorally- induced plasticity phenomenon. Memantine has low to moderate affinity for the
NMDA receptor channel, strong voltage-dependent channel blocking characteristics (similar to Mg2+), and fast channel unblocking kinetics (Parsons et
al., 1995). Therefore, it is possible that memantine restores place field expansion
through the mechanism just described. Moreover, if this is the mechanism for
memantine’s efficacy at reinstating behaviorally-induced plasticity then
antagonists of L-type Ca2+ channels should also be able to induce experience- dependent place field expansion in aged rats.
In addition to age-related alterations in experience-dependent place field
expansion, the stability of hippocampal maps (that is, the distribution of place
fields in an environment) also differs between young and old animals. In normal
young rats, a map for a given environment can remain stable for months
(Thompson and Best, 1990). That is, when a rat is repeatedly returned to the
same environment, the same map is retrieved. A similar stability of CA1 maps in 69 aged rats is observed within and between episodes of behavior in the same environment. Occasionally, however, if the old rat is removed from the environment and returned later, the original map is not retrieved and an independent population of place cells may be activated even in a familiar room
(Figure 3.8; Barnes et al., 1997; Gerrard et al., 2001). This global 'remapping' predicts that old rats should show bimodal performance on tasks that require the functional integrity of the hippocampus. For spatial tasks, good performance
Young rats Frequency
Aged rats Frequency
Figure 3.8: CA1 map stability in young and old rats. Place fields recorded from one young rat (upper panel) and one old rat (lower panel) on two sessions of track running recorded about one hour apart. Only a subset of cells recorded are shown, each cell is a different color in each rat. The dots represent spikes of a given CA1 pyramidal cell. Note the changes in the distribution of individual place fields across sessions for the old rat, suggesting “remapping”, but stability of the fields from one session to the next for the young rat. This is spontaneous remapping is probabilistic and is associated with bimodal performance on tasks that require the hippocampus. The right panel shows the frequency distribution of the performance of young rats (top) and aged rats (bottom) on the spatial version of the Morris swim task. Young rats have a unimodal frequency distribution. In contrast, performance in the aged rats is bimodal with aged rats finding the hidden platform with a short path in some trials but not recalling the location of the platform in other trials. Figure adapted from Barnes et al., 1997. 70 should correspond to retrieval of the original map and poor performance should correspond to retrieval of an incorrect map. This prediction is at least consistent with empirical observation. When trained on the spatial version of the Morris swim task, in early trials for both young and aged rats, performance is bimodal.
This means that for some trials rats find the hidden escape platform with a short path but for other trials the rats do not recall the location of the platform and take a longer path. By the final training trials, however, the young rats’ performance is unimodal with most rats taking a direct path to the platform. By contrast, the aged rats’ performance remains bimodal. The trials on which the old rats fail to correctly remember the location of the hidden escape platform could correspond to map retrieval failures (Figure 3.8; Barnes et al., 1997).
A probable mechanism for map retrieval failures is defective LTP in aged rats. Although map stability within an episode does not require plasticity, the maintenance of maps between episodes depends on an LTP-like mechanism. It has been shown in young rats that blockade of NMDA receptors (Kentros et al.,
1998) or protein synthesis inhibition (Agnihotri et al., 2004) results in map retrieval errors when the rat is returned to the same environment. Thus, effective long-term plasticity mechanisms are required for the map to become stabilized, possibly through binding the external features of the environment to the map. If this binding does not occur or is weakened by defective plasticity mechanisms, then the likelihood that the correct map will be recalled in a familiar environment decreases. Further support that a decrease in CA1 long-term place map stability 71 with age results from an inability to bind the map to external features of the environment comes from the observation that lesions of the perirhinal cortex also lead to re-mapping between episodes of behavior but not within episodes (Muir and Bilkey, 2001). Because the hippocampus receives extensive visual input from the perirhinal cortex, it is possible that binding the map to external visual features is disrupted by the absence of perirhinal input. The fact that the place field data obtained from animals without a perirhinal cortex mimics what has been observed in old rats indicate that investigating the impact of advanced age on the perirhinal cortex may be a productive avenue to be pursued.
Finally, the phenomenon of hippocampal map re-alignment also supports the idea that external features of the environment become bound to the “map”, and that process is disrupted by old age. Map re-alignment can be demonstrated by training rats to shuttle back and forth on a linear track between a start box mounted on a sliding rail, and a reward site at the opposite end of the track. On the initial part of all journeys out from the start box, CA1 place cells fire at fixed distances from the start box. As young rats approach the reward site, however,
CA1 place cells fire at fixed distances from the destination. Thus, on outward journeys from the start point, the position representation is updated by path integration. Farther along the journey, however, the place map becomes aligned on the basis of external stimuli. If the distance traveled from the start box conflicts with the external features of the environment because the start box was moved closer to the reward site during the rat’s outbound journey then the CA1 72 map realigns during the rat’s journey back to the start box from an alignment with the origin to an alignment with the destination (Gothard et al., 1996).
Hippocampal map realignment can also be observed in young rats that are required to run from a movable start box and are trained to pause at a hidden goal location in order to receive a medial forebrain bundle stimulation reward. If the origin of each lap is varied by shifting the start box, but the goal location is fixed with respect to stable distal cues, the CA1 place field activity eventually realigns, as the box is moved further from the original position, from a representation that was box-aligned to one that was room-aligned (Figure 3.9).
This realignment is necessary in order for the rat to find the goal location, which requires using a room-aligned coordinate frame. Importantly, the accuracy of young rats in pausing at the correct goal location is correlated with the position on the track of the map re-alignment from a start box reference frame to a room- aligned reference frame (Rosenzweig et al., 2003).
Using this ‘map re-alignment’ paradigm, it has been shown that old rats are delayed in the point on that the track where the map realigns to the fixed environmental cues; that is, the old rats move further from the start box before the hippocampal map shifts to a room-aligned reference frame. Consistent with these physiological findings, old rats are also less accurate at finding the goal location and there is a significant correlation between the ability to find the location where medial forebrain bundle stimulation will be delivered, and the location on the track of the map re-alignment (Rosenzweig et al., 2003). While 73
Figure 3.9: Hippocampal map realignment. A schematic of the apparatus and procedure used for measuring CA1 hippocampal map realignment in young and old rats. (A) The rat exits a start box and traverses a linear track. (B) At a fixed position in the room (aligned to distal room cues) is a hidden reward location. If the rat pauses at the correct location he receives a medial forebrain bundle stimulation reward. (C) For some trials the track and start box are moved but the reward location remains fixed. In order for the rat to find the goal location he must use a room-aligned reference frame rather than one that is box-aligned. Old rats are delayed at shifting from a box-aligned to a room-aligned reference frame and are thus impaired at finding the goal location compared with young rats. Figure adapted from Rosenzweig et al., 2003.
CA1 map re-alignment has not been studied in young rats administered an
NMDA receptor antagonist prior to the first experience on the track, one would predict that cessation of NMDA receptor activity would result in a hippocampal map re-alignment deficit.
A different picture emerges from age comparisons of CA3 pyramidal cell characteristics, which has lead to an increased understanding of the effects of aging on place map stability. In fact, it is now known that the effects of aging on hippocampal map stability are subregion specific. In young rats the ensemble 74 activity in CA1 and CA3 is dissociable (Lee et al., 2004; Leutgeb et al., 2004;
Vazdarjanova and Guzowski, 2004). This dissociation could reflect two competing functions of the hippocampal network: pattern completion versus pattern separation (e.g., Marr, 1971; McNaughton and Morris, 1987). Moreover, there is evidence of a dissociation of the effects of aging on CA1 and CA3 ensembles. While maps in CA1 and CA3 of aged rats are both stable within a single session of environmental exploration, in CA1 spatial representations are less stable across sessions in aged compared with young rats (Barnes et al.,
1997). By contrast, spatial representations in CA3 seem to be more rigid in aged rats. When an aged rat explores a familiar environment for 7 min and is then placed into a novel environment, spatial representations in CA3 remain the same even though the environment has changed (Wilson et al., 2005). In young rats, however, CA3 place maps are independent between familiar and novel environments (Leutgeb et al., 2004; Wilson et al., 2005). A disruption in the ensemble characteristics of dentate gyrus granule cells, a structure known to be particularly vulnerable to the aging process (Small et al., 2002; Small et al.,
2004), could contribute to the failure of aged CA3 networks to form new spatial representations. Theoretical considerations predict that the dentate gyrus distinguishes between similar inputs so the activity patterns stored in hippocampal networks are more dissimilar (i.e. pattern separation). This pattern separation may function to increase storage capacity (Marr, 1971). Because the transfer of information between granule cells and CA3 pyramidal cells may be 75 degraded, this could contribute to the inability of the aged CA3 network to form new spatial representations when it should (for review, see Wilson et al., 2006).
3.5 From the map to memory in aging ensembles
Since the seminal observation that hippocampal pyramidal cells are active at discrete locations in an environment (O'Keefe and Dostrovsky, 1971), a central question has been ‘what, in fact, do place cells encode?’. Because the external sensory input to the rat is not necessarily the input to the hippocampus, it might not be possible to determine what place cell firing reflects without knowing how external input is transformed and delivered to the hippocampus, and what internal inputs are received by the hippocampus (e.g. motivational states of the animal). Moreover, it has been suggested that place cell firing does not represent any single stimulus or even the configuration of stimuli. Rather, it has been proposed that hippocampal output provides an index code for each memory
(e.g., Marr, 1971; Teyler and DiScenna, 1986; McClelland et al., 1995;
Sutherland and McNaughton, 2000; Teyler and Rudy, 2007). In this model, the hippocampal ‘code’ is the distribution of place field activity across the principal neurons of the hippocampus. These hippocampal activity patterns become linked to the patterns of activity in other areas of the cortex, thereby acting as an index.
Thus, reactivating the hippocampal ‘code’ will evoke the associated patterns of activity in other brain structures (for review, see Sutherland and McNaughton,
2000; Teyler and Rudy, 2007). Understanding the process by which hippocampal 76 activity becomes associated with patterns of cortical activity can provide insight into how memory is consolidated and retrieved, and why memory impairments are present in aged humans and other animals.
It is clear from both human (Scoville and Milner, 1957; Squire et al., 1993;
Teng and Squire, 1999) and animal lesion studies (Winocur, 1990; Zola-Morgan and Squire, 1990; Kim and Fanselow, 1992; Kim et al., 1995) that the hippocampus is necessary for the initial acquisition of certain types of memory.
Although it is disputed whether or not episodic memory is ever independent of the hippocampus (e.g., Nadel and Moscovitch, 1997; Ryan et al., 2001), some theories of memory consolidation concur that the neocortex is the site of storage for long-term semantic memory, and that the hippocampus is necessary for indirectly associating independent regions of the neocortex. The hypothesis is that the hippocampus serves as an indirect linking system for neocortical regions that can maintain the memory more efficiently and for a longer period of time.
This indirect association between independent cortical sites is possible because the hippocampus receives highly processed, poly-modal input from most cortical areas and returns projections, via the entorhinal, perirhinal and postrhinal cortices back to these areas (Felleman and Van Essen, 1991; Swanson and
Kohler, 1986; Amaral and Witter, 1995; McIntyre et al., 1996; Insausti et al.,
1997; Insausti and Munoz, 2001; Lavenex et al., 2002). It is these back- projections from the hippocampus to the neocortex that provide a rapid linking system for associations distributed across the brain, which lack the intrinsic 77 connectivity to form these associations without the hippocampus (e.g. McClelland et al., 1995).
The physiological mechanism for creating connections between neocortical regions encoding different aspects of a memory is likely the reactivation of memory traces during periods where the brain is not acquiring new information. Such “off-line” reactivation during quiet wakefulness or sleep is hypothesized to lead to a reorganization of cortico-cortical connections and the strengthening of the memory trace (e.g., Marr, 1971; for review, see Sutherland and McNaughton, 2000; Teyler and Rudy, 2007).
Pavlides and Winson (1989) were the first to report that the firing rate of
CA1 neurons during behavior influenced the firing rate during the subsequent sleep episode (Pavlides and Winson, 1989). Since this initial report, additional empirical evidence that memories are reactivated during sleep has been obtained from ensemble recordings of multiple neurons in the hippocampus
(Wilson and McNaughton, 1994; Kudrimoti et al., 1999; Shen et al., 1998; Hirase et al., 2001) and neocortex (Qin et al., 1997; Hoffman and McNaughton, 2002; Ji and Wilson, 2007). When a young rat explores an environment, CA1 pyramidal cells with overlapping place fields have correlated firing patterns. It has been shown that cell pairs with correlated firing patterns due to overlapping place fields maintain this correlation in the subsequent slow-wave sleep period following this episode (Wilson and McNaughton, 1994; Shen et al., 1998). Therefore, a significant amount of the variance in correlated firing patterns of cell pairs 78 observed during a slow-wave sleep episode can be explained by the correlated activity patterns measured during the preceding behavior (Wilson and
McNaughton, 1994). This phenomenon can be quantified using the measure of explained variance (Figure 3.10). Explained variance describes how much of the variability in the firing patterns of pairs of neurons during a rest period can be explained by the activity patterns of the neuron pairs during the preceding behavioral experience. Therefore, an explained variance value that is “significant”
(or greater than expected by chance) reflects the case were the cell pair activity patterns established during behavior, are recapitulated during the subsequent rest episode.
Figure 3.10: Procedure for calculating the explained variance (EV) measure of activity pattern reactivation. The matrix of correlations of firing rates between all possible cell pairs that were active during an episode of behavior is computed (omitting within-tetrode correlations, black squares) for the rest 1 (pre, left panel), behavior (middle panel), and rest 2 (post, right panel) epochs. These are referred to as the R matrices. Both the X and the Y axis represent the cell ID. Red indicates cell pairs that have high firing rate correlations (r = ~0.4), while dark blue indicates that there was no correlation (see color bar) between the cell pair (r = 0.0). The correlations between each pair of R matrices are then calculated. Next, the explained variance between the behavior and subsequent rest epoch is calculated by using the partial correlation coefficient to eliminate the contribution of the pre-behavior rest on the rest 2-behavior correlation. This value is squared to yield the explained variance between the rest 2 and maze running epochs. Notice that the patterns of correlations in the R matrices are more similar between rest 2 and behavior compared to the patterns between rest 1 and behavior (Burke, unpublished data).
79
Interestingly, it has been shown that activity pattern reactivation in the hippocampus occurs primarily during large irregular amplitude local field potentials (Vanderwolf, 1969) called sharp wave events (Buzsaki, 1986).
Superimposed on the sharp wave events are high frequency (~200 Hz) oscillations called “ripples” (O'Keefe and Nadel, 1978). Ripples are associated with hippocampal pyramidal cell discharge and can be detected near the CA1 stratum pyramidale (Ylinen et al., 1995). Sharp waves occur about once or twice a second during immobilization and slow wave sleep (Buzsaki, 1986) and result from a depolarization of CA1 pyramidal cells and interneurons by the CA3 afferents (Buzsaki, 1986; Buzsaki, 1989; Chrobak and Buzsaki, 1994; Chrobak and Buzsaki, 1996; Ylinen et al., 1995).
Simultaneous recordings of hippocampal EEG and multiple pyramidal cells have shown that the explained variance measure of activity pattern reactivation is significantly higher during the sharp wave/ripple event compared to the inter-sharp wave interval (Kudrimoti et al., 1999). This is consistent with the idea that the sharp waves, and the associated ripple event, correspond to a convergence of the network onto an “attractor state” that represents a stored memory. This strongly suggests that these events lead to the reactivation of the memory trace (Shen and McNaughton, 1996). Moreover, sharp wave events during slow-wave sleep correlate with the transition from periods of low cortical activity (“down-state”) to periods of synchronous high cortical activity (“up-state”)
(Battaglia et al., 2004; but see, Sirota et al., 2003). The observation that sharp 80 waves facilitate transitions from neo-cortical down-states to up-states may reflect joint memory trace reactivation in the hippocampus and in the neocortex, possibly contributing to the formation of cortico-cortical connections and the consolidation of long-term memory. Importantly, the incidence and amplitude of sharp waves and the frequency of their corresponding ripple events are the same in young and old rats (Smith et al., 2000; Gerrard et al., 2001) even though in aged rats the inter-spike intervals are on average shorter during ripple events presumably because of Ca2+ dyshomeostasis.
Activity pattern reactivation remains intact in the aged rat (Gerrard et al.,
2001). For a single epoch of running on a maze, aged rats show significantly
greater explained variance for the subsequent sleep episode and the behavior
preceding it, than the observed explained variance between the behavior and the
pre-running sleep episode. Moreover, the magnitude of the explained variance is
similar between the young and the old rats (Gerrard et al., 2001). Importantly,
akin to the young rats, aged rats also show greater activity pattern reactivation
during the sharp wave/ripple epochs compared to the inter-sharp wave intervals
(Gerrard et al., 2001). These data imply that the mechanisms necessary to
reactivate patterns of neural ensembles that were involved in an episode of
behavior remain relatively intact in advanced age.
In addition to activity pattern reactivation, CA1 neurons have a temporal
order in their firing patterns in relation to one another during behavior, and they
retain this temporal order during subsequent sleep (Skaggs and McNaughton, 81
1996; Nadasdy et al., 1999; Lee and Wilson, 2002). Quantification of the extent of preservation of the temporal pattern of cell firing during behavior can be extracted from cross-correlograms among cells. That is, the cross-correlations of neurons that were active during an episode of behavior are compared with the cross-correlations between cells during both the preceding and the subsequent sleep episodes (Figure 3.11). From this, one can calculate the “temporal bias” or tendency for cell pairs to fire in the same sequence during post behavior rest
Figure 3.11: The temporal bias calculation. An example of the temporal bias analysis for a single pair of CA1 pyramidal cells with overlapping place fields. (A) Each spike during track running on the T-shaped track is plotted for the two neurons, cell 1 in black and cell 2 in gray. (B) The cross-correlogram of this cell pair for REST1 shows little correlated firing within the 1 sec window (top panel). The y-axis shows the count for each 10 ms bin shown on x-axis. For RUN, strong theta frequency ( 7–8 Hz) modulation is observed in the cross-correlogram, and in this cell pair, cell 2 tends to fire just before cell 1 (middle panel). This tendency is maintained during REST2 as the peak of the cross-correlogram occurs before zero (bottom panel). The difference in the area under the curve between –200 to 0 ms (black) and 0 to 200 ms (gray) for each cross-correlogram is calculated to obtain the temporal bias measure. Figure from Gerrard et al., 2008. 82 episodes as they did during the actual behavior. For example, if cell1 fires before cell 2 during behavior, this is reflected in the shape of the cross-correlogram.
During subsequent sleep, the cross-correlogram for cells 1 and 2 will retain a shape that is similar to what was observed during behavior. If the similarity between the cross-correlograms from the behavior and subsequent rest episode is greater compared to the similarity of the cross-correlograms between the behavior and the preceding rest episode, then this is taken as evidence that the temporal order of the activity among CA1 neurons during behavior is, in fact, preserved during sleep (Skaggs and McNaughton, 1996). In old rats, there is a significant decrease in the amount of temporal bias reactivation compared to young rats. This suggests that while the cell pairs active together during behavior in old rats reactivate during subsequent rest periods, they do not necessarily reactivate in the same temporal order as that during behavior (Gerrard et al.,
2008). Additionally, in both young and old rats the amount of temporal bias reactivation was significantly correlated with performance on the spatial version of the Morris swim task (Gerrard et al., 2008). Thus, sequence reactivation could be an essential mechanism for the consolidation of hippocampal-dependent tasks.
Although it is theoretically enticing to assume that a deficit in sequence reactivation could be related to the age-associated plasticity impairments, and the lack of experience-dependent place field expansion plasticity in aged CA1 neurons (Shen et al., 1997), recent evidence disputes this idea. When young rats 83 are injected with the NMDA receptor antagonist CPP, prior to an episode of track running, experience-dependent place field expansion was blocked (Ekstrom et al., 2001; Burke, unpublished data), however, both pattern reactivation and sequence reactivation during the subsequent sleep episode were intact (Burke, unpublished data). Figure 3.12A shows the explained variance (EV) of the activity patterns during sleep 2 minus the correlated activity between sleep 1 and behavior for the entire sleep episode (grey), the ripple epochs (black), and the inter-ripple intervals (IRI) when rats were injected with the saline control or CPP.
Figure 3.12B shows the amount of sequence reactivation for both drug conditions as quantified by the Pearson correlation coefficient (r-value) between behavior and sleep 1 (grey), and behavior and sleep 2 (black) for the temporal bias measures of all cell-pairs with activity during behavior. NMDA receptor blockade
A Pattern Reactivation B Sequence reactivation
0.4 0.2
0.3 ALL 0.1 0.2 Ripple 0.1 B-Sleep1 0.0 0.1 IRI B-Sleep2 R value 0 0.0 saline CPP Saline CPP 0.0 Condition Condition Figure 3.12: NMDA receptor blockade and reactivation. (A) Pattern reactivation, as measured by the variance of sleep 2 activity correlations that can be explained by the activity correlations during behavior, minus the EV between sleep 1 and behavior. The EV was calculated for activity that occurred during the entire sleep epochs (grey), during the ripple/sharp wave events (black) and inter-ripple intervals (IRI; white). Overall, the EV was greater during the ripples compared to the IRIs for both the CPP and saline control conditions (F[3,15] = 13.39, p < 0.05, repeated contrasts), but the EV during the ripple/sharp wave events for the saline control compared to the CPP condition were not significantly different (F[1,5] = 4.81, p = 0.08). Bin size = 50 ms. (B) Sequence reactivation following NMDA receptor blockade was also unaltered, and there was no significant decrease in sequence reactivation using the Skagg’s area under the curve measure when CPP was administered. Bin size = 10ms, window size = 200ms. 84 did not significantly disrupt reactivation for either measure. These data indicate that reactivation may occur in the absence of plasticity and the dynamic process of sequence reactivation could arise from the same intrinsic network asymmetry that accounts for theta phase precession.
Another mechanism that could promote the reactivation of the sequential firing patterns of neurons in sleep episodes could be sharp wave/ripple events that occur during the behavior. Recently, it has been shown that cell assemblies active during the acquisition of a memory are repeated during sharp wave/ripple events that occur during behavior (O'Neill et al., 2006; Foster and Wilson, 2006;
Jackson et al., 2006). While an animal is performing a behavior, the theta rhythm dominant in the EEG is interrupted when the animal stops at a food dish to eat, or stops to groom. When this occurs sharp wave/ripple events are observed
(O'Neill et al., 2006; Foster and Wilson, 2006; Jackson et al., 2006). Additionally, it has also been shown that theta during active exploration can be interrupted by sharp wave/ripple events (O'Neill et al., 2006). These non-rest sharp wave/ripple events contain the firing patterns of recently visited spatial locations along with the cell assembly active in the current location (Foster and Wilson, 2006; O'Neill et al., 2006). Therefore, non-rest sharp wave/ripple events may also facilitate the formation of place-related cell assemblies through facilitating connections between cells with overlapping place fields. Consistent with this idea, the number of sharp wave/ripple events emitted has been shown to increase with experience during behaviors that require repetition. It has been suggested that these non- 85 sleep sharp wave/ripple events arise from experience-dependent plasticity that occurs during behavior (Jackson et al., 2006), however, this has not been empirically proven. Because old rats have deficits in this form of plasticity (Shen et al., 1997), these sharp wave/ripple events during behavior may not occur with the same incidence in old animals as they do in young and this prediction needs to be examined directly.
In summary, aged rats have differences in the dynamic properties of CA1 place fields that have been shown to correspond to observed age-associated behavioral deficits. Specifically, aged rats fail to show experience-dependent place field expansion plasticity to the same extent as do young rats (Shen et al.,
1997). Because it has been suggested that a mechanism similar to this phenomenon could serve to encode sequences of events (Hebb, 1949), it is not surprising that aged animals are impaired at learning sequences (Oler and
Markus, 1998; Winter, 1997). Between episodes of experience in a single environment, aged rats are also impaired at maintaining stable spatial representations in the CA1 subregion of the hippocampus. This observation is consistent with the finding that the performance of old rats on tasks that require an allocentric spatial reference frame to be solved is impaired (Barnes et al.,
1997; Rosenzweig et al., 2003). Additionally, once a memory trace has been formed in the hippocampus, aged rats are able to reactivate the activity patterns of the CA1 neurons associated with the memory (Gerrard et al., 2001), but cannot accurately reactivate the sequential patterns of the neural activity 86
(Gerrard et al., 2002). Again, this electrophysiological correlate of consolidation is significantly associated with a given animal’s behavior on a spatial task (Gerrard et al., 2002). Thus, there is mounting evidence that some of the cognitive capabilities of old and young animals are reflected in the integrity of information processing in the hippocampus. How functional alterations in aged hippocampal circuits relate to information processing the other regions of the brain, however, has not been extensively investigated and it is unclear how neural networks just one synapse away from the hippocampus are impacted by advanced age. One aim of this thesis was to elucidate the extent to which aging modified the functional integrity of circuits in a structure closely related to the hippocampus, the perirhinal cortex. The next chapter will summarize what is known about the perirhinal cortex of young animals.
87
CHAPTER 4 – THE FUNCTION AND PHYSIOLOGY OF
THE PERIRHINAL CORTEX
4.1 Introduction
The aim of this dissertation work is to investigate the contribution of the perirhinal cortex to stimulus recognition and how changes in perirhinal cortical physiology, which might occur during aging, contribute to age-related cognitive decline. This chapter will review what is known about the physiology and cellular morphology of the perirhinal cortex in young animals, because this thesis marks the first formal examination of the electrophysiological characteristics of aged perirhinal cortical neurons.
The perirhinal cortex (PRC) is part of the parahippocampal region, which also includes the postrhinal cortex (POR; homologous to the parahippocampal cortex of primates), and the medial and lateral entorhinal cortices. These areas of neocortex, along with the hippocampus, comprise the medial temporal lobes. In both primates and rats, the PRC consists of 2 small strips of cortex (Broadman’s areas 35 and 36), which account for 5% of the total cortical surface area, and are associated with the most deeply invaginated portion of the rhinal sulcus (Burwell,
2001). This suggests that the PRC may scale up relative to cortical surface area.
In contrast, the entorhinal cortex scales down exponentially as cortical surface area increases, which suggests that the entorhinal cortex might be relatively less important farther up the evolutionary scale (Burwell, 2000). While the PRC occupies 5% of cortical surface area in both the rat and monkey, it is thicker in 88 the monkey brain relative to the rat brain and the laminar characteristics are much more prominent (Burwell, 2000). Despite these differences, the spatial arrangement of the PRC and its boundaries with adjacent cortical regions are remarkably similar between the rat, monkey and human.
In rats, the rostral border of the PRC is adjacent to the agranular and granular insular cortex. This border occurs between –2.45 and –2.8 mm relative to
bregma (Figure 4.1A). Dorsally, area 36 is bordered by ventral area TE of the
inferotemporal cortex. Area 36 has three subfields based on cytoarchitectural
differences: areas 36d, 36v, and A 36p. Areas 36d and 36v are PRC adjacent to each other in the
dorsoventral plane and occupy the
rostral two-thirds of the PRC while
B area 36p occupies the caudal one-
third (Figure 4.1B). Area 35 is
bordered ventrally and caudally by
the entorhinal cortex and comprises Figure 4.1: The rat perirhinal cortex. (A) Lateral view of the rat brain illustrating the two subfields, areas 35v and 35d locations of the perirhinal cortex (PRC), postrhinal (POR), and entorhinal cortices (ent). The PRC is shown in grey, area 36 in dark grey (Figure 4.1B). In general, the PRC and area 35 in light grey. The POR is shown in the cross-hatched pattern. The entorhinal cortex can also be distinguished from the (ent) is shown in grey with diagonal stripes, dark stripes for the medial entorhinal cortex (entm) and light stripes for the lateral entorhinal cortex dorsally adjacent and ventrally (entl). (B) Unfolded map of the PRC and POR. Figure from Burwell 2001. adjacent cortical regions by the 89 absence of heavily myelinated fibers (Burwell, 2001). While the subregions of the PRC have been distinguished by anatomical differences, such as interconnectivity and cytoarchitecture, there are no reports of any functional distinctions between the different subregions of the PRC. Data indicate, however, that the magnitude of sensory input from different modalities varies along the longitudinal and dorsal/ventral axes of the PRC. In particular, visual input is stronger to caudal areas of the PRC, olfactory and somatosensory input is stronger to rostral PRC, and auditory input terminates more medially. Finally, input from the insular cortex terminates along the entire rostrocaudal axis but primarily in more ventral regions (Burwell and Amaral, 1998). This anatomical organization suggests that stimulus encoding in the PRC might differ across subregions depending on the sensory modalities involved.
The PRC, along with postrhinal cortex and the entorhinal cortex, is heavily interconnected with other areas of the medial temporal lobe, including the hippocampus. In addition, the PRC is reciprocally connected with most other areas of the neocortical mantle. It is via the PRC, POR and entorhinal cortex that the hippocampus receives sensory information from polymodal and unimodal association cortices (Burwell et al., 1995; Suzuki and Amaral, 1994) and then returns projections to the neocortex through these same regions (Lavenex et al.,
2002). Although the PRC receives input from all sensory association cortices there is an important distinction between the cortical afferents of the PRC and the
POR. Specifically, in both rats and primates, the PRC receives input from areas 90 of the brain not directly involved with spatial (or motion) information processing, such as inferotemporal cortex and somatosensory association cortex.
Conversely, the POR receives more visuospatial input relative to the PRC, for example, the posterior parietal cortex sends a strong projection to the POR but only weakly projects to the PRC (Suzuki and Amaral, 1994; Burwell and Amaral,
1998; Furtak et al., 2007). The dissociation between non-spatial/perceptual information projecting to the PRC and spatial input to the POR continues into the entorhinal cortices where the lateral entorhinal cortex is heavily interconnected with the PRC, and the medial entorhinal cortex is heavily interconnected with the
POR (Burwell and Amaral, 1998; Kerr et al., 2007; Lavenex et al., 2002; Lavenex et al., 2004). Functionally, the distinct parallel processing streams of spatial and non-spatial information lead to a dissociation between the activity patterns of neurons in the lateral and medial entorhinal cortices. Specifically, while neurons in medial entorhinal cortex show spatially selective firing rate increases (Fyhn et al., 2004; Hargreaves et al., 2005; Hafting et al., 2005), neurons in the lateral entorhinal cortex do not (Hargreaves et al., 2005). In fact, neurons in the lateral entorhinal cortex selectively increase their firing rate at the location of objects
(Deshmukh and Knierim, 2008). The implications of these two distinct streams of input on hippocampal ensemble activity will be described more in Chapter 7.
91
4.2 Perirhinal cortical-dependent behavior in rats and monkeys
The reciprocal connections of the PRC with both the neocortex, and the hippocampus suggest that this brain region is involved in memory consolidation
(Squire et al., 2004). Lesion data indicate, however, that damage to the PRC can lead to deficits in certain behaviors that are dissociable from other structures in the medial temporal lobe (e.g., Baxter and Murray, 2001; Buffalo et al., 1999;
Bussey et al., 2000; Norman and Eacott, 2005), and the distinct roles of the hippocampus and PRC in different aspects of memory remain controversial. In particular, the results of many experiments are equivocal on the issue of whether or not judgments of recognition (i.e. the ability to know whether or not a stimulus has been previously experienced) require an intact hippocampus (e.g., Baxter and Murray, 2001; Zola and Squire, 2001; Broadbent et al., 2004; Clark et al.,
2001). One explanation for these conflicting results is proposed by a dual- process theory, which contends that decisions regarding the previous occurrence of a stimulus can be made with two different recognition systems (Mandler,
1980). One system involves the sensation of stimulus familiarity, and is rapid and automatic, potentially occurring independent of the hippocampus. The other system is recollection, which is slower, effortful and dependent on the hippocampus (for reivew, see Brown and Aggleton, 2001). If two different neurobiological processes can support stimulus recognition, it is possible that the demands of an object recognition task could favor one system over another. If the demands of a task favor recollection, such as having to remember a 92 particular object and a particular location, then a hippocampal lesion would lead to a deficit (Gaffan and Parker, 1996; Bussey et al., 2000). Conversely, if the task can be solved with more automatic judgments of familiarity (e.g., spontaneous recognition) then a hippocampal lesion would not produce an impairment in recognition (Winters et al., 2004). Unfortunately, studies directed at testing this dual-process theory, have not been able to account entirely for these conflicting results (see Chapter 5) and this issue remains a source of controversy (Squire et al., 2007; Brown, 2008).
In contrast to the hippocampus, lesion experiments in rats (Mumby and
Pinel, 1994; Ennaceur et al., 1996; Ennaceur and Aggleton, 1997; Bussey et al.,
2000; Kesner et al., 2001; Prusky et al., 2004; Gaffan et al., 2004; Winters and
Bussey, 2005) and monkeys (Meunier et al., 1993; Buffalo et al., 1999; Buffalo et al., 2000; Malkova et al., 2001) have unequivocally demonstrated that the PRC is necessary for object recognition. This result is consistently observed across lesion techniques (excitotoxic, electrolytic, aspiration or reversible inactivation), and even when different tasks are used to measure stimulus recognition.
Multiple tasks have been designed to measure object recognition. The delayed-nonmatching-to-sample (DNMS) task, which was first designed for monkeys (Mishkin and Delacour, 1975), can also be used to test a rat’s ability to discriminate between a sample stimulus and a novel stimulus. In this task, the animal must displace a single sample object to receive a food reward. After a variable delay, the rat or monkey is presented with two objects, the original 93 sample object and a novel object. If the animal displaces the novel object, it receives a food reward and this is recorded as a correct trial. The trial is incorrect if the animal displaces the sample object and it does not receive a reward
(Mishkin and Delacour, 1975; Mumby et al., 1992; Mumby et al., 1990; Mumby and Pinel, 1994). Lesions of the PRC in monkeys produce impairments on the
DNMS task when trial-unique stimuli are used (Meunier et al., 1993; Buffalo et al., 1999; Buffalo et al., 2000; Malkova et al., 2001). When a small set of familiar stimuli are used, however, PRC lesions in monkeys do not produce a significant impairment on the DNMS task (Eacott et al., 1994). Additionally, rats with PRC lesions perform worse on the DNMS task relative to control animals, or rats with lesions of the amygdala (Mumby and Pinel, 1994).
Several studies have measured the effect of advanced age on DNMS task performance and found that aged monkeys are also impaired on the DNMS task
(e.g., Moss et al., 1988; Rapp and Amaral, 1991; Shamy et al., 2006), which could indicate functional alterations of PRC circuits that support the recognition of visual stimuli. Although the DNMS task has never been used to measure object recognition in aged rats, the spatial variant of this task (delayed nonmatch-to- position) has been used to demonstrate that old rats do not effectively discriminate between the sample and the novel position when compared to young animals (Dunnett et al., 1988). Although perirhinal lesions produce similar deficits on the delayed nonmatch-to-position (DNMP) task in rats (Wiig and
Burwell, 1998), this age-associated impairment is difficult to interrupt for several 94 reasons. First, only 2 different stimuli are generally used in the DNMP task (left lever versus right lever); therefore, by virtue of this design the DNMP task cannot use trial unique stimuli, which may limit the requirement of an intact PRC for normal performance. Additionally, this task is spatial, and the PRC is only required for spatial tasks when the testing environment is rich in visual cues (for review, see Aggleton et al., 2004). For these reasons the DNMP task is not optimal for assessing recognition deficits that might arise from possible age- associated alterations in the PRC.
A more effective way to assess object recognition, particularly in rodents, is the spontaneous object recognition (SOR) task, which unlike the DNMS task requires no training. The SOR task, first described in 1988, capitalizes on a rat’s natural tendency to spend more time exploring novel objects relative to stimuli that have been previously encountered (Ennaceur and Delacour, 1988). In this task a rat in placed in a familiar testing arena with two identical novel objects, and given several minutes to explore. After this familiarization phase, there is a variable time delay, and then during the test phase, the rat is placed back in the same testing arena but with two different objects in the arena. One object is a triplicate copy of the objects used during the familiarization phase, while the other object is novel. The rat is given several minutes to explore the objects, and normal young rats will show significant novelty discrimination, spending more time exploring the novel object relative to the familiar object. Figure 4.2 shows a schematic representation of the SOR task. 95
Familiarization phase Test phase A1 A2 A3 B1
Figure 4.2: A schematic of the spontaneous object recognition task. In the object familiarization phase, a rat is place in a testing arena to explore duplicate copies of an object (e.g., A1 and A2). After this object familiarization phase, the rat is removed from the apparatus for a variable delay. Following the delay, during the test phase, the animal is returned to the apparatus to explore two different objects than had been used during acquisition. One object (A3) is the third copy of the triplicate set of the objects used in the familiarization phase, and the other was a novel object (B3).
When object recognition in rats is tested with the SOR task, rats with neurotoxic (Ennaceur et al., 1996; Ennaceur and Aggleton, 1997), electrolytic
(Norman and Eacott, 2005), or transient inactivation of the PRC do not show a significant exploratory preference for the novel object relative to the familiar object at delays longer than 5 min. Moreover, when compared to control rats, the animals without a functioning PRC, show less object discrimination during the test phase after all delays (Winters and Bussey, 2005a). In addition to an intact
PRC, it has also been shown that both glutamatergic and cholinergic transmission within the PRC is necessary for normal SOR task performance.
Antagonism of perirhinal NDMA receptors prior to the familiarization phase, impairs novelty discrimination on the SOR task for long delays (3 hours), but not short delays (5 min), between the familiarization and test phase. This indicates that NMDA receptor activity during encoding is required for long-term object 96 recognition (Winters and Bussey, 2005b). Additionally, if NMDA receptors are antagonized immediately following the familiarization phase of the SOR task, performance is disrupted suggesting that NMDA receptor activity is required for the consolidation of visual information necessary for discriminating between novel and familiar objects. AMPA receptor activity in the PRC, on the other hand, is necessary before the familiarization phase (encoding), immediately following this phase (consolidation), and during the test phase (retrieval) (Winters and
Bussey, 2005b).
In contrast to glutamate, the role of acetylcholine in object recognition appears to be primarily during stimulus encoding. If cholingeric projections to the
PRC from the basal forebrain nucleus are selectively lesioned by injecting the neurotoxin 192 IgG-saporin directly into the PRC object recognition even at short delays is disrupted. Additionally, pre-acquisition administration of the acetylcholine muscarinic receptor antagonist scopolamine to the PRC also disrupts the discrimination of a novel and a familiar object (Warburton et al.,
2003). Paradoxically, when scopolamine is infused into the PRC directly following the familiarization phase, it actually facilitates the performance of rats on the
SOR task (Winters et al., 2006). These data indicate that cholinergic modulation of the PRC is necessary for encoding object information necessary for a novelty preference during the test phase of the SOR task but, once the objects have been encoded, antagonizing muscarinic receptor activity can block the incidental 97 encoding of extraneous visual features that might interfere with the discrimination of objects during the test phase (Winters et al., 2006).
Plasticity within the PRC has also been linked to SOR task performance and blocking CREB phosphorylation, which is necessary for the induction of plasticity modulating immediate-early genes, subsequently impairs object recognition (Warburton et al., 2005). The likely necessity of plasticity for normal performance on the SOR task indicates that this task may be vulnerable to the process of normal aging. In fact, multiple investigations have reported age- related impairments in object recognition at longer delays between the familiarization and test phases (Bartolini et al., 1996; de Lima et al., 2005; Pieta
Dias et al., 2007; Pitsikas et al., 2005; Platano et al., 2008). A comprehensive account and the probable neurobiological etiology of these impairments will be described in Chapter 5.
Although it cannot be disputed that the PRC is involved in object recognition, whether or not this structure supports the recognition of stimuli via contributions to perception or as part of a ‘memory system’ is at the center of a contentious debate (e.g., Buckley and Gaffan, 2006; Cate and Kohler, 2006). The long-standing notion that the PRC, along with the entorhinal cortex, parahippocampal cortex and hippocampus, is part of a “medial temporal lobe memory system” (Squire and Zola-Morgan, 1991; Squire et al., 2004; Squire et al., 2007) has promoted the idea that PRC damage leads to ‘memory’ impairments. This assumption, however, is problematic and it is becoming 98 increasingly evident that there may be no anatomical loci in the brain where perception ends and the ‘memory system’ begins. Specifically, the brain is organized hierarchically with sensory cortices having access to limited information based on anatomical constraints (Felleman and Van Essen, 1991), and damage to these areas generally results in discrete perceptual deficits that are sensory-modality specific. Higher up the neural hierarchy, in multi-modal association cortices (e.g., the PRC), neurons within a given brain region receive input from more disparate areas of the cortex, and perceptual deficits may not be evident unless the task requires the conjoining of multiple features that are represented in the lower sensory cortices. In this view, a primary function of the
PRC is to bind together features represented in lower brain regions (e.g. auditory, somatosensory, visual association cortices). Since the binding of information represented across disparate areas of the neocortex is believed to be the cornerstone of memory, behavioral deficits resulting from PRC lesions can often present as memory impairments.
Murray & Bussey (1999) were the first to formally outline a role for the
PRC in both perception and memory. As part of a perceptual network, the PRC participates in the ventral visual cortical processing stream (although it is important not to discount input to the PRC from other modalities), which is devoted to the perception and identification of environmental stimuli. In this view, the special contribution of the PRC to this type of processing is the representation of complex conjunctions of stimulus features (Murray and Bussey, 99
Figure 4.3: The proposed convergence of object-feature representations as information passes through sequential processing stages in the ventral visual stream. A, B, C and D represent simple visual features encoded in early regions of the ventral visual stream. These features combine to form representations of increasing complexity. The PRC acts to bind together these features. Dotted lines represent putative divisions between adjacent cortical fields in inferotemporal cortex (IT). Figure from Murray & Bussey, 1999. 1999). Figure 4.3 shows a schematic representation of how the PRC binds the features represented in inferotemporal cortex and other caudal visual areas.
In support of the perceptual-mnemonic feature conjunctive role of the
PRC, lesion studies indicate that the PRC is required for visual discrimination, even in the absence of a memory load (0 delay), if the animal must discriminate between objects that share common features. When monkeys are trained to discriminate between objects in which no single feature can be used to solve the problem (that is, visual stimuli share features that are predictive of both a correct and incorrect response), animals with PRC lesions are impaired relative to controls. The lesioned monkeys are able to learn visual discrimination problems, however, if the rewarded and the unrewarded stimuli did not share common elements (Bussey et al., 2002; Bussey et al., 2003).
The same pattern of ‘perceptual’ impairments following PRC lesions has been shown in rats. When rats are tested on the SOR task with a zero delay, and 100 the novel and the familiar objects presented to the rat in the test phase contain overlapping features, rats with PRC lesions do not preferentially explore the novel object more than the familiar object. This pattern is not observed when the test objects are highly dissimilar, and rats with a lesion centered on the PRC can show normal object recognition (Bartko et al., 2007). To further illustrate that this is a perceptual deficit rather a problem with memory, rats with PRC lesions have also been tested on oddity discriminations. If a rat is simultaneously presented with 2 identical objects and 1 ‘odd’ object, a normal rat will spontaneously spend more time exploring the odd object. Figure 4.4 shows an example trial of this perceptual oddity task. Rats exhibit this behavior when presented with either 3-
Figure 4.4: The perceptual oddity task. A rat performing the perceptual oddity task with stimuli constructed from legos. Notice that the lego objects to the left and to the right of the rat are identical but the center object is different, however, all objects share common features. A normal rat will spend more time exploring the center object while rats with PRC lesions will explore all objects equally. Figure from Bartko et al., 2007. 101 dimensional (Bartko et al., 2007), or 2-dimensional stimuli (Forwood et al., 2007).
Rats with PRC lesions do not show significant oddity discrimination when the
‘odd’ object shares features with the other two identical stimuli. Because all three stimuli are presented simultaneously, it cannot be argued that this impairment is due to a memory deficit. Rather the data indicate that the PRC is necessary for an animal to detect differences between stimuli that share common elements
(Bartko et al., 2007). Therefore, both rat and monkey data confirm that the PRC is critical for processing complex visual stimuli.
The data discussed above suggest that it may not be useful to attempt to classify the function of the PRC as either perceptual or mnemonic, and I am of the opinion that perception and memory will not map onto anatomically distinct regions of the brain. Rather it is possible that the PRC acts as an index for the features of non-spatial stimuli that are represented across sensory association cortices, which do not have the anatomical connections to link together these stimulus elements directly. In this view, a PRC neuron active during the presentation of a specific stimulus set is analogous to the hippocampal ‘place cell’ in that the conjunction of features that will activate a given assembly of PRC neurons are randomly allocated. This also predicts that if a familiar stimulus is encountered in a new environment the ensemble of active PRC neurons will
‘remap’. If the PRC index is disrupted, a complex stimulus can no longer be represented as a whole. Given the demands of the behavioral task, this leads to perceptual impairments, memory deficits, or both. 102
The idea that the PRC operates as an index for stimulus features encoded by sensory cortices is supported by lesion data showing that, in addition to complex visual perception and object recognition, the PRC is also involved in learning associations between stimuli. Associative learning can be measured with either a delay or a trace conditioning task. In these tasks, a conditioned stimulus (CS) precedes the presentation of an unconditioned stimulus (US). After several trials of CS-US presentation, the animal will demonstrate the behavior initially elicited by the US during the CS presentation. For a delay conditioning task the CS and the US presentation overlap. In the trace conditioning variant, a delay is interjected between the CS and the US. For an auditory delayed fear conditioning task, in which a tone (CS) precedes a foot shock (US), rats with pre- training lesions of the PRC show impairments at learning the association between the CS and the US when the CS is a ‘complex stimulus’. In this example rat ultrasonic social signals are not learned, but a simple tone of a single frequency is learned (Lindquist et al., 2004). Importantly, although the amygdala is involved, the hippocampus is not required for delay fear conditioning (Selden et al., 1991). Therefore, these data indicate that when the CS is a simple stimulus, connections between subcortical structures and the amygdala can support delay fear conditioning. When the CS is a complex stimulus, however, the PRC is necessary for a conditioned fear response, independent of the hippocampus. This supports the idea that the PRC is involved in processing and/or maintaining representations of complex stimulus configurations (Murray 103 and Bussey, 1999). Interestingly, when the memory load is increased by testing animals on a trace fear conditioning task, rats with pre-training lesions of the
PRC are impaired regardless of CS complexity (Kholodar-Smith et al., 2008a).
Finally, PRC lesions also impair fear conditioning to the context in which an aversive US was given (Bucci et al., 2000; Kholodar-Smith et al., 2008a;
Kholodar-Smith et al., 2008b), suggesting that the deficit caused by a PRC lesion on fear conditioning to both cues and contexts reflects the essential role of PRC in binding stimulus elements together into unitary representations.
4.3 Biophysical and plasticity properties of perirhinal cortical neurons
In order to understand the functional physiology of the PRC it is important to appreciate the biophysical properties of perirhinal cells, which are distinct from other cortical neurons. The laminae of the PRC can be distinguished by their unique cytoarchitecture (Burwell, 2001; Furtak et al., 2007). Specifically, layer I does not contain neurons, and layer II/III has the highest density of small pyramidal cells. In layer V, the pyramidal cells are larger, less densely packed, and oriented perpendicular to the pial surface. In contrast, layer VI neurons are oriented parallel to the pial surface and external capsule (Furtak et al., 2007).
The PRC is agranular cortex and does not contain a layer IV (Burwell et al.,
1995; Burwell, 2001; Suzuki and Amaral, 2003). Figure 4.5 shows a Nissl stained cross-section of the PRC laminae with the different layers indicated by the black dashed lines. 104
Neurons with different
physiological characteristics (fast-
spiking, late-spiking, regular-spiking,
or burst-spiking) are present in
different proportions across the
different layers of the PRC (Faulkner
and Brown, 1999). Layer II/III contains
primarily late-spiking neurons, which
show long delays (up to seconds) to
initiate spiking following depolarizing
input, and regular-spiking neurons that
fire with shorter latencies but exhibit
strong accommodation. These two
Figure 4.5: Perirhinal cortical laminae. classes of neurons are represented in Nissl-stained tissue showing the laminar boundaries of the PRC. Features that easily approximately equal proportions in distinguish layers include the obvious horizontal orientation of layer VI, the highest layer II/III (Faulkner and Brown, 1999; apparent cell density in layer II/III, the largest cells in layer V, and the infrequent number of cells in layer I. Scale bar is 100 µm. Figure Beggs et al., 2000). In contrast, layer from Furtak et al., 2007. V contains primarily regular-spiking neurons (76%). In other areas of the neocortex, layer V is predominantly composed of burst-spiking neurons, however, in PRC these cells only represent
1 out of every 10 neurons. Finally, layer V also contains neurons with late-spiking characteristics (14%) (Moyer et al., 2002). In contrast, layer VI is composed of 105 primarily late-spiking neurons (McGann et al., 2001). The physiological characteristics of PRC neurons are notable because they are distinct from all other areas of the neocortex, which do not show substantial numbers of late- spiking cells (McGann et al., 2001). Moreover, the observation that some of these late-spiking cells exhibit firing latencies up to 12 seconds suggests that the
PRC is not a region of rapid throughput (i.e., a relay between sensory cortices and the hippocampus), but may be better viewed as a higher-order association cortex with unique processing capabilities (Moyer et al., 2002).
Data obtained from both slice physiology experiments and unanaesthetized multi-unit recordings in cats also support this idea that the PRC does not passively relay sensory information between the rest of the neocortex and the hippocampus. In fact, the probability of excitatory synaptic transmission between neocortical and entorhinal cortices through the PRC is low (<4%;
Pelletier et al., 2004; for review, see de Curtis and Pare, 2004). Moreover, when perirhinal and lateral entorhinal neurons are recorded simultaneously from awake cats, very few perirhinal-entorhinal neuron cross-correlations show significant correlated activity within a 200 msec time window (Paz et al., 2006). Moreover, slice physiology in concert with tract-tracing has established that the PRC contains both local inhibitory interneurons and long-range GABAergic projections that leave the PRC and synapse in the superficial layers of the entorhinal cortex, which is the origin of the perforant path projection to the hippocampus (Pinto et al., 2006; Apergis-Schoute et al., 2007). Combined with the high proportion of 106 late-spiking cells in the PRC (Beggs et al., 2000; Faulkner and Brown, 1999;
McGann et al., 2001; Moyer et al., 2002), the cumulative effects of local and long-range GABAergic inhibition could account for the low probability of excitatory impulse transfer between the neocortex and the entorhinal cortex.
Interestingly, entorhinal IPSPs evoked with PRC stimulation are reduced by agonism of cholingeric muscarinic receptors (Apergis-Schoute et al., 2007).
These data suggest that during periods when the brain may be encoding behaviorally relevant information, and acetylcholine levels are elevated (for review, see Hasselmo, 2006), the probability of excitatory impulse propagation between the PRC and entorhinal cortex is increased. This is particularly interesting considering the involvement of cholingeric neurotransmission in the
PRC for encoding stimulus information necessary for stimulus recognition during the SOR task (Warburton et al., 2003; Winters and Bussey, 2005).
The low probability of excitatory impulse propagation between the PRC and entorhinal cortex is interesting considering that many theories of memory formation and consolidation rely on the assumption that sensory information is transferred from the neocortex to the hippocampus (encoding) and then from the hippocampus back to the neocortex (consolidation) with high fidelity (for review, see McClelland et al., 1995; Sutherland and McNaughton, 2000). Specifically, empirical data show that hippocampal sharp wave events (high amplitude negative deflections in the EEG associated with increased synchrony; Buzsaki,
1989; Chrobak and Buzsaki, 1996), which occur during slow-wave sleep and 107 quiet wakefulness, correspond with the replay of neuronal activity initiated by behavior and could coordinate memory consolidation in the neocortex (see
Chapter 3; Kudrimoti et al., 1999; Gerrard et al., 2001; Lee and Wilson, 2002).
Moreover, sharp waves are correlated with episodes of higher neuronal spike synchrony in dorsal areas of neocortex (Sirota et al., 2003; Battaglia et al., 2004).
Sharp wave events recorded during quiet wakefulness in the entorhinal cortex, however, do not propagate to the PRC or lead to increased spike synchrony in this area of the brain (Pelletier et al., 2004). This indicates that the PRC does not passively transmit information between the hippocampus and other areas of the neocortex. Interestingly, consistent with theories contending that acetylcholine plays a critical role in learning and memory (Hasselmo, 2006; Hasselmo and
Stern, 2006), elevated levels of acetylcholine have been shown to increase the transfer of information from the PRC to the entorhinal cortex (Apergis-Schoute et al., 2007). This suggests that certain behavioral conditions could also modulate the excitatory transmission from the entorhinal cortex to the PRC, and there is empirical evidence to support this idea. During the acquisition of an appetitive- trace conditioning task, synchronized activity in the basolateral amygdala (Paz et al., 2006) or medial prefrontal cortex (Paz et al., 2007) leads to an increase in correlated activity between the PRC and lateral entorhinal cortex.
The low probability of synaptic transmission under most conditions and long-range GABAergic projections between the PRC and entorhinal cortex are not optimal for inducing synaptic weight changes experimentally. In support of 108 this there are no convincing reports of long-term potentiation (LTP) or long-term depression (LTD) between the PRC and the entorhinal cortex (but see, Ivanco and Racine, 2000). LTP can be elicited in CA1, however, following high frequency (400 Hz) stimulation of the PRC (Liu and Bilkey, 1996a), and this synaptic enhancement is NMDA receptor-dependent (Liu and Bilkey, 1996b).
One function of the PRC to CA1 synapse could be to bind the active hippocampal ensemble or place map to the external features of the environment.
Two lines of evidence support this idea. First, as mentioned in Chapter 3, lesions of the PRC produce place map instability between two episodes of behavior in the same environment. Additionally, putative modulation of perirhinal activity alters the characteristics of CA1 pyramidal cells (Chapter 7).
Synaptic weights within layer II/III of the PRC can also be modified experimentally by patterned stimulation of either layer II/III (intermediate pathway; Bilkey, 1996; Cho et al., 2000), or layer I (superficial pathway;
Ziakopoulos et al., 1999). Moreover, plasticity within the PRC can be evoked by stimulation of either dorsal input (temporal side) or ventral input (entorhinal side).
Importantly, this synaptic modification is input specific and both LTD
(Ziakopoulos et al., 1999; Cho et al., 2000) and LTP (Bilkey, 1996; Ziakopoulos et al., 1999) in the PRC demonstrate associative specificity, which is a requirement for Hebbian plasticity.
Experimentally induced plasticity in the PRC is relevant because LTP and
LTD in this area of the brain has been linked to a rat’s ability to discriminate 109 between a novel stimulus and a stimulus that has been previously experienced.
Specifically, when cholingeric transmission in the PRC is disrupted rats are impaired on the SOR task (Warburton et al., 2003; Winters and Bussey, 2005), and LTD in the PRC, but not LTP, is blocked (Warburton et al., 2003). Moreover, blocking CREB phosphorylation leads to LTP deficits in the PRC and subsequently impairs object recognition (Warburton et al., 2005).
4.4 Single-unit activity of perirhinal cortical neurons in behaving animals
Consistent with anatomical and lesion data, physiological recordings from awake-behaving animals have focused on the involvement of the PRC in stimulus recognition, and associative learning. Data obtained from different experiments attempting to show how PRC neural networks support these cognitive processes, however, cannot always be reconciled.
A long-standing hypothesis has been that recognition memory is supported by the response decrement measured in the activity of perirhinal cortical neurons when a stimulus goes from novel to familiar. MW Brown and colleagues (1987) were the first to report that when a monkey performed a
Konorski conditional delayed matching task, which requires an animal to indicate whether two successively presented stimuli are the same or different, that ~27% of visually responsive neurons in the PRC had significantly lower firing rates to the second presentation of the stimulus relative to the first. This occurred even when 15 intervening stimuli were presented between the first and second 110 presentation, and when stimuli were both novel and familiar, meaning that PRC neurons could potentially be signaling recency and familiarity (Brown et al.,
1987). This response decrement was not observed in hippocampal or subicular neurons, although the authors did not report the extent to which neurons in these brain regions showed visually selective activity (Brown et al., 1987). A subsequent investigation, using the same behavioral paradigm, confirmed these findings (Riches et al., 1991). The results reported for the hippocampal recordings, however, were inconsistent with what is currently accepted regarding hippocampal physiology (e.g., over 50% of the dentate gyrus granule cells responded to visual stimuli, and there were no firing rate differences between pyramidal cells and interneurons). Because their methods could not discriminate between cell types, this suggests that there could potentially be fundamental problems with how these recordings were obtained and one needs to be cautious with the interpretation of these data. Additional recordings from the non-human primate also suggested that ~10% of visually-responsive neurons in the PRC demonstrate lower firing rates for familiar relative to novel stimuli, and to the second compared to the first presentation of a familiar stimulus. This was observed whether or not the visual stimuli were relevant to the animal’s behavior
(Fahy et al., 1993). A later study reported that this response decrement persisted for a delay of up to 24 hours between repeated presentations of the stimulus, and was robust enough that when the average firing rate of the population of recorded PRC neurons was measured the overall mean firing rate was less 111 during the presentation of familiar stimuli compared to novel stimuli (Xiang and
Brown, 1998).
The response decrement patterns reported for PRC neurons in the primate have also been described in rats. Recordings from neurons in the PRC while visual stimuli are presented to rats show that many cells in the PRC have higher firing rates during the initial presentation of an object compared to the second presentation, and that the firing rates of PRC neurons are higher during the presentation of novel objects compared to familiar ones (Zhu et al., 1995;
Zhu and Brown, 1995), and this has been described even when animals are under anesthesia (Zhu and Brown, 1995). Although it is theoretically enticing to conclude that a response decrement in the activity of perirhinal neurons supports judgments of familiarity and recency, there are multiple alternative possibilities.
First, the response decrement occurs in a similar proportion of cells and in comparable magnitudes when the stimuli direct the animal to perform the correct behavior and when the stimuli are irrelevant for task performance. Additionally, the response decrement has also been observed in PRC neurons recorded from anesthetized animals. These observations are inconsistent with a neurobiological phenomenon that is to be associated with an active mnemonic process. Second, in the experiments described above, few to none of the recorded PRC neurons increased their firing rates during the second presentation of a stimulus. This is incompatible with observations that the PRC can show both experimentally- induced increases (Bilkey, 1996; Ziakopoulos et al., 1999; Warburton et al., 112
2003; Warburton et al., 2005) and decreases (Ziakopoulos et al., 1999; Cho et al., 2000; Warburton et al., 2003) in synaptic strength. Finally, all these electrophysiological recordings, and particularly those obtained from rodents, have been obtained under behavioral conditions where the animal was restrained or restricted to a small area of space. It is possible that neural recordings of PRC neurons in situations where rats are actively moving around a larger environment could lead to a different pattern of PRC neuronal activity.
While the idea that a response decrement could support stimulus recognition has been pervasive, other studies have shown that a proportion of neurons in the PRC increase their firing rates under certain circumstances. When rats are presented with olfactory stimuli during a continuous odor-guided DNMS task, many cells in the PRC are active during odor sampling and demonstrate odor-selective activity. During the delay and recognition phase of the task, odor memory coding is reflected in two general ways. First, a substantial proportion of cells show odor-selective activity throughout or at the end of the memory delay period. Second, odor-responsive cells show either odor-selective enhancement or suppression of activity during stimulus repetition in the recognition phase of the task, and a global decrease in PRC activity does not result from the repeated presentation of an olfactory stimulus (Young et al., 1997). Additionally, when
PRC neurons are recorded from non-human primates while they perform a delayed matching-to-sample task, with highly familiar visual stimuli (over a hundred presentations) and novel stimuli, it was reported that the overall 113 population of recorded PRC neurons had higher firing rates during the presentation of familiar stimuli relative to the novel stimuli, and the activity produced by novel stimuli slowly increased over many repeated stimulus presentations (Holscher et al., 2003). Therefore, it does not appear probable that a majority of neurons in the PRC are preferentially activated by novel versus familiar stimuli. This idea will be explored more in Chapter 6.
Although the notion that a response decrement in the activity of PRC neurons supports stimulus recognition is problematic for the reasons described above, this theory has gained momentum from imaging studies that use gene expression as a marker of neural activity (Wan et al., 1999; Warburton et al.,
2005; Zhu et al., 1995; Zhu et al., 1997). The immediate-early gene (IEG) c-fos can be used to map the distribution of thousands of neurons activated during specific behaviors across many anatomically distinct brain regions (Morgan et al.,
1987). Specifically, it is inferred that neurons expressing the protein products of c-fos, following the presentation of novel or familiar stimuli, had activity related to these specific behaviors. In the experiments that examined c-fos protein levels to measure a response decrement, rats nose-poked as novel images were presented to one eye and familiar images were displayed to the other eye. Thus, one hemisphere received novel input while the other received familiar input. In the PRC there were more c-fos positive cells in the hemisphere that received input from the images of novel items compared to the hemisphere that only viewed images of familiar items (Wan et al., 1999). 114
Similar to the data obtained from electrophysiological recordings, the interpretation of these results is confounded by the experimental methodology.
Again, the animals were restricted to a small area while stimuli were presented to them rather than actively exploring (nose poking for juice reward while images are passively viewed on 2-dimensional computer monitor). Moreover, the use of c-fos protein as a label for neurons active during a given episode is problematic.
Glutamate binding to metabotropic receptors on astrocytes can also lead to c-fos expression in glia (Edling et al., 2007). Therefore, it is unlikely that only neurons were included in these analyses. Additionally, it is known that immediate-early gene expression can be decoupled from neuronal activity under certain circumstances, and a neuron can be active without initiating the signaling cascades that lead to the induction of these genes (Guzowski et al., 2006;
Fletcher et al., 2006). Specifically, this electro-transcriptional decoupling is observed in the hippocampus after a massed exposure to an environment. The training used by Wan and colleagues (1999) involved presenting rats with the same visual stimuli many times in a day over the course of several days, and is a
‘massed’ exposure. Therefore, in the absence of electrophysiological confirmation, it is possible that a proportion of the activated cells did not express c-fos and that the difference in protein levels between hemispheres does not reflect a response decrement. Finally, experiments that have shown differences in c-fos expression following familiar versus novel stimuli have only made comparisons of protein levels across hemispheres during one episode of 115 behavior. Therefore, it cannot be determined whether a single neuron expresses c-fos after the presentation of a novel, but not a familiar, stimulus.
Although experiments investigating the role of the PRC in stimulus recognition have not fully elucidated the neurobiology of this cognitive process, other electrophysiological recordings have shown how the PRC may support associative learning. If the role of the PRC is to bind together elements of complex stimuli represented in lower sensory cortices, then the association of different stimuli should lead to observable changes in the activity of PRC neurons. This has been observed using several different associative learning tasks.
Monkeys can be trained to associate pairs of stimuli with the visual paired associate task. In this task, the animal must learn the association between at least 12 pairs of stimuli that were arbitrarily matched. The monkey is presented with a cue stimulus. After a delay, the paired associate of the cued stimulus and one stimulus from a different pair are shown. The monkey obtains a reward for touching the correct paired associate. Before learning, neurons in inferotemporal cortex do not show correlated activity patterns to the different stimuli within a paired associate. After learning, however, paired associates elicit significantly correlated responses in inferotemporal neurons, which does not occur when the
PRC is lesioned (Higuchi and Miyashita, 1996). Although this experiment directly implicates the PRC in visual associative learning, it does not indicate how PRC 116 activity supports this process. Other research recording from the PRC when monkeys perform similar tasks has gone on to show how this might occur.
Visual association learning can also be measured with a GO/NO-GO task, which requires monkeys to learn an association between a predictor stimulus and a choice stimulus in order to determine the correct behavioral response (bar release for a GO pair versus release for NO-GO pair). An important component of this task is that the same predictor stimulus can be paired with a choice stimulus indicating GO, or a different choice stimulus indicating NO-GO. This type of task is reminiscent of visual discrimination tasks that show that the PRC is necessary to disambiguate stimuli with overlapping features. As observed in the inferotemporal cortex, when a monkey learns to associate two different stimuli, responses to frequently paired predictor-choice stimuli are more similar to one another than was the case with infrequently paired stimuli. Thus, with long- term training, perirhinal neurons tend to link the representations of temporally associated stimuli (Erickson and Desimone, 1999). Another notable result is that while monkeys performed this GO/NO-GO task there was no difference in the activity of PRC neurons during the presentation of novel versus familiar stimuli
(Erickson and Desimone, 1999). In fact, rather than overall differences in firing rates of PRC cells, a subsequent investigation showed that when a monkey views images of novel stimuli neuronal response preferences for pairs of nearby neurons and far apart neurons are uncorrelated. As the stimuli become familiar, however, the response preferences of nearby neurons become correlated 117
(Erickson et al., 2000). These data suggest that visual experience leads to the development of localized cell assemblies in the PRC. Specifically, different neurons, which were initially activated by distinct visual stimuli, will develop correlated activity if the stimuli each neuron is selective for become closely associated. After this association is learned, presenting only one stimulus can activate both neurons. Anatomical evidence also supports the idea that visual learning leads to the formation of selective cell assemblies in the PRC.
Electrophysiological recordings in concert with retrograde tracers injected into area 36 of the PRC have shown that the degree of divergent projections from area TE of the inferotemporal cortex to neurons in area 36 is smaller from the TE neurons that are selective to learned pictures than from the nonselective TE neurons. Therefore, the anatomical difference (the divergence) correlates with the selectivity of TE neurons to the learned pictures (Yoshida et al., 2003).
In rats, the activity of PRC cells following associative learning paradigms has been investigated using trace conditioning tasks, which require an intact
PRC (Kholodar-Smith et al., 2008). Approximately 40% of the neurons in the rat
PRC respond to a single auditory stimulus that is not paired with a salient outcome (Allen et al., 2007; Furtak et al., 2007). When a conditioned auditory stimulus (CS) is paired with a foot shock (US), however, 73% of PRC neurons respond to the auditory CS after the CS-US association has been learned.
Additionally, a large proportion of the cells that are active during the CS remain active during the delay between the CS and the US (Furtak et al., 2007). These 118 results are analogous to the observation in non-human primates that one function of the PRC is to link the representations of temporally associated stimuli
(Erickson and Desimone, 1999).
4.5 Discussion
In conclusion, data obtained from more recent electrophysiological recordings from the PRC and behavioral experiments have questioned early assumptions about the role of the PRC in cognition. Due to the anatomical connections of the PRC with the hippocampus and the neocortex (for review, see
Burwell et al., 1995) it was believed that the PRC could serve as a relay station between the neocortex and the hippocampus, and that its primary function was to subserve memory (Squire et al., 2004). Moreover, early recordings from PRC neurons suggested that its unique contribution to memory was to provide a familiarity and/or a recency signal. New data require a reassessment of these assumptions, however. Particularly, results from recent lesion and electrophysiological experiments indicate that the PRC could contribute to cognition in one or more of the following ways:
1. It may bind together features of complex stimuli that are represented in
sensory cortices (Lindquist et al., 2004; Bussey et al., 2005). This function
is necessary to disambiguate stimuli that share common elements and
indicates that perirhinal activity may serve as an index for representations 119
in disparate areas of the neocortex that cannot easily form direct
associations.
2. The PRC may provide a link between stimuli that are temporally
associated (Erickson and Desimone, 1999; Furtak et al., 2007), this is
similar to Hebb’s concept of a ‘phase sequence of cell assemblies’ (Hebb,
1949).
3. Finally, the PRC may facilitate the transfer of behaviorally-salient sensory
information between the neocortical and entorhinal brain structures (Paz
et al., 2007; Paz et al., 2006), which suggests that this area of the brain
does not passively participate in memory consolidation as a relay station
but rather ensures that only the memory traces for episodes of behavioral
relevance are strengthened.
Future research should examine the involvement of the PRC in these distinct but complementary roles more closely. 120
CHAPTER 5: THE HUMAN PERIRHINAL CORTEX AND PROCESS
THEORIES OF RECOGNITION MEMORY
5.1 Introduction
While dozens of memory tests have been designed to examine a subject’s ability to distinguish novel from previously experienced stimuli, the cognitive variables (for review, see Malmberg, 2008), and the neurobiological underpinnings of this recognition memory process are still heavily debated (c.f.,
Eichenbaum et al., 2007; Squire et al., 2007). Models of human recognition memory either contend that it is supported by a unitary signal-detection process
(for review, see Wixted and Stretch, 2004), or that two independent processes of familiarity and recollection contribute to a subject’s ability to identify a stimulus as
‘old’ or ‘new’. Experiments with human subjects (c.f., Yonelinas, 1994; Yonelinas,
1997; Dunn, 2004), and more recently with rats (Fortin et al., 2004; Robitsek et al., 2008; Sauvage et al., 2008), tend to support the dual-process idea, and have argued that the hippocampus and perirhinal support recollection and familiarity, respectively. As discussed in the previous chapter, findings from animal lesion studies have led to conflicting results regarding the role of the hippocampus in recognition memory (e.g., Baxter and Murray, 2001; Zola and Squire, 2001;
Broadbent et al., 2004; Clark et al., 2001), and functional magnetic resonance imaging (fMRI) experiments in humans have also been inconsistent in identifying distinct contributions of medial temporal lobe subregions to recognition memory
(Daselaar et al., 2006a; Daselaar et al., 2006b; Kirwan et al., 2008; Shrager et 121 al., 2008). This chapter will summarize what is currently known regarding the neurobiology of human recognition memory in young adults and across the lifespan, and the potential role of the perirhinal cortex in age-associated declines in stimulus recognition.
5.2 Knowing and recollecting: Single versus Dual-process models for recognition
Early models of recognition memory assumed that it reflected a single familiarity process that could be described by signal-detection theory (for review, see Yonelinas, 2001; Dunn, 2004; Malmberg, 2008). In this view, old and new stimuli are part of two different ‘familiarity’ distributions that overlap, and recognition memory accuracy can be characterized by the single parameter d’, which is the distance between the peaks of the new and the old familiarity distributions (for review, see Yonelinas, 2001). Figure 5.1 shows the hypothetical familiarity distributions for old (purple) and new (green) stimuli, and the corresponding d’ value. A subject determines whether a stimulus is part of the new or the old familiarity distribution based on a response criterion (c; light blue vertical line in Figure 5.1). Within this d’ window a subject can respond with a conservative c, which is shifted closer to the peak of the ‘old’ distribution (i.e., with high confidence that a stimulus is old). This will decrease the rate of false positives (i.e., responding that a new stimulus is old; light green shaded area in
Figure 5.1), while increasing the rate of misses (i.e., responding that an old 122 stimulus is new; light purple shaded area in Figure 5.1). A subject can also respond with a liberal c (i.e., low confidence that a stimulus is old). A liberal c is one shifted towards the peak of the ‘new’ distribution, and would minimize the rate of misses while increasing the rate of false positives. Over the past several decades, several variants of the signal-detection model have been proposed (for review, see Malmberg, 2008), and all of these models assume that recognition memory is supported by a single process.
Figure 5.1: Signal-detection model of recognition memory. Familiarity distributions for old (purple) and new (green) stimuli. The difference between the peaks of these two distributions is d’. Because these two distributions overlap, the subject must set a response criterion (c; light blue vertical line). Within the d’ window it is possible to respond conservatively (i.e., with high confidence that a stimulus is old), which shifts c towards the peak of the old distribution. A subject can also respond liberally (i.e., with low condfidence that a stimulus is old), and this shifts c towards the peak of the new distribution. An optimal c will minimize the proportion of ‘misses’ (light purple), and ‘false alarms’ light green’ (adapted from Yonelinas, 2001). 123
The notion that recognition memory could be supported by either a simple familiarity judgment (i.e., the subject merely ‘knows’ the stimulus was encountered but may not recall where or when), and the more effortful process of recollection, was formalized decades ago (Mandler et al., 1969; Mandler, 1980).
This dual-process model of recognition memory gained momentum in the 1990s when Yonelinas (1994; 1997; 1999) capitalized on analyses of receiver operating characteristics (ROC) (Yonelinas, 1994; Yonelinas, 1997; Yonelinas, 1999). The tasks that Yonelinas used to distinguish between the different processes that could contribute to recognition were designed to allow analysis of the proportion of correct recognitions (i.e., ‘hits’), and the proportion of false alarms (Figure 5.1, light green area) across different confidence levels (i.e., response criteria). In a typical recognition experiment, a subject will identify a stimulus as old with different levels of confidence spontaneously. A high confidence level is equivalent to a conservative response criterion, and corresponds to the light blue vertical line in Figure 5.1 shifting to the right. When a subject has high confidence, the false alarm rate (light green area) is low but the rate of misses
(light purple area) is high. In contrast, a low confidence level corresponds to a liberal response criterion (the light blue vertical line in Figure 5.1 would shift to the left), which both increases the hit rate, and the rate of false alarms. If a subject can identify their confidence level for individual trials of a recognition experiment, the hit and false alarm rates can be plotted across these different confidence levels in order to generate the ROC curve. The shape of the ROC 124 curve can then be used to infer the cognitive variables that are involved in stimulus recognition (e.g., Yonelinas, 2001; Robitsek et al., 2008; Sauvage et al.,
2008).
Figure 5.2A shows familiarity distributions for old and new stimuli that have unequal variance. The different colored vertical lines represent different levels of confidence used to generate an ROC curve. Proponents of dual-process models argue that the increased variance of the ‘old’ distribution relative to the
‘new’ distribution is due to both familiarity and recollection contributing to recognition memory (e.g., Yonelinas, 2001). Figure 5.2B shows two hypothetical
A B 1
0.8
0.6 new old stimuli stimuli
Hit rate Hit 0.4
0.2
0 0 0.2 0.4 0.6 0.8 1 Familiarity False alarm rate Figure 5.2: The process dissociation procedure, ROC curve asymmetry and recognition memory. (A) Two hypothetical ROC curves (red and grey). In both examples the points on the left represent the hit and false rates obtained with the most conservative (high confidence) response criteria, while the points on the right are the hit and false alarm rates derived from the most liberal (low confidence) response criteria. The grey ROC curve represents hypothetical data that would be obtained if recognition memory was supported by a single process. Because it is asymmetrical, the red ROC curve is consistent dual-process models of stimulus recognition. (B) Familiarity distributions for new and old stimuli that do not have equal variance, presumably due to both familiarity and recollection supporting recognition judgements. The different colored vertical lines represent different response criteria. The hit and false alarm rates for each response criterion are shown in A in the corresponding color.
125
ROC curves (red and grey). In both examples the points on the left (nearest the origin) represent the hit and false rates obtained with higher confidence levels
(conservative response criteria), while the points on the right indicate the hit and false rates derived with lower confidence levels (liberal response criteria). The grey ROC curve represents hypothetical data that would be obtained if recognition memory was supported by a single process. This curve is symmetrical because the hypothetical familiarity distributions of the ‘old’ and
‘new’ stimuli have equal variance. If the familiarity distributions for the old and new stimuli do not have equal variance (Figure 5.2A), however, the ROC curve is asymmetrical (red curve). The different colored dots of the red ROC curve in
Figure 5.2B represent the hit and false alarm rates obtained with the different confidence levels represented by the vertical line of the corresponding color in
Figure 5.2A. When empirical data yielded asymmetrical ROC curves, and the degree of asymmetry was independent of recognition memory performance, proponents of dual-process models argued that these data could not be reconciled with the view that recognition memory is sustained by a single process. In this view, when stimulus recognition is supported by a familiarity judgment alone, the data produce a curvilinear symmetrical ROC curve (grey curve in Figure 5.2B). In contrast, when recollection is used along with familiarity, an asymmetrical ROC curve is observed (red curve in Figure 5.2B). The magnitude of ROC asymmetry can be quantified by transforming the hit and false alarm rates into Z-scores and plotting these values to obtain a ‘Z-slope’. If the 126
ROC curve is symmetrical, the z-slope will be close to 1 (or unity). The more asymmetrical the ROC curve is, the more the z-slope deviates from 1 (for review, see Yonelinas and Parks, 2007). For example, Figure 5.3 shows the z-score transformed values (right panel) of the ROC data shown in the left panel. The red
ROC curve is less symmetrical than the grey curve, and the z-slope of the red curve is also further from a slope of 1 (unity) relative to the grey z-slope
(Yonelinas and Parks, 2007). Experimental paradigms that encourage subjects to use recollection produce data that more closely resemble the red curve shown in
Figure 5.3. Importantly, because the relative contributions of recollection and familiarity can vary, accuracy and ROC asymmetry can vary independently (for review, see Yonelinas, 2001).
ROC Curves Z-slope 1 2
0.8 1.0
0.6 0
Hit rate 0.4 z(Hit rate)
-1.0 0.2
0 -2 0 0.2 0.4 0.6 0.8 1 -2 -1.0 0 1.0 2 False alarm rate z(False alarm rate) Figure 5.3: ROC curves, z-slope and dual-process models. When the hit and false alarm rates used to derive the ROC curve (left panel) are plotted as z-scores (right panel), symmetrical ROC curves have z-slopes that are close to 1 (grey). In contrast, asymmetrical ROC curves produce z-slopes that deviate from 1 (red). Asymmetrical ROC curves and Z- slopes that deviate from 1 have been used to agrue the recognition memory is supported by both recollection and familiarity. Figure adapted from Yonelinas & Park (2007).
127
Single-process models of recognition memory argue that the higher confidence responses reflect the degree of memory strength, and are supported by the same underlying process within the medial temporal lobe as the lower confidence responses. In this view, the difference between recognizing a stimulus with high versus low confidence varies along a continuum. Proponents of this ‘unitary memory strength’ hypothesis argue that when recognition memory tests do not control for differences in memory strength (i.e., both weakly and strongly remembered items are analyzed together) asymmetrical ROC curves can be obtained (Wais et al., 2006; Squire et al., 2007; Wais et al., 2009). Thus, the unitary memory strength variant of single-process models allows for the familiarity distributions of old and new stimuli to differ. Because both the dual process models and the newer unitary process model can account for asymmetrical ROC curves, these models are essentially are indistinguishable.
The debate between single versus dual-process models of recognition memory, however, has relied heavily on which model best fits the empirically-derived ROC curve (c.f., Yonelinas and Parks, 2007; Malmberg, 2008), with both camps arguing that their model better fits the data. Therefore, it has not been possible with behavioral data and memory models alone to determine if familiarity and recognition are two points on the same continuum or qualitatively different processes. Human lesion data have also been problematic for resolving this issue since large medial temporal lobe lesions produce deficits in recollection and familiarity (Reed et al., 1997; Yonelinas et al., 2002), and again proponents 128 of single- and dual-process models both contend that the lesion data provide supporting evidence for their respective models.
5.3 The functional neuroanatomy of human recognition memory
The availability of imaging techniques that can measure blood oxygen level-dependent (BOLD) activation in different brain areas during recognition memory tasks have made it possible to examine the neuroanatomy of this cognitive process in humans. Unfortunately, these models have not resolved the debate as to whether or not familiarity and recollection are distinct processes supported by different medial temporal lobe structures. Rather, similar to the
ROC analysis and lesion data, the proponents of single and dual-process models both argue that the imaging data support their respective view (c.f., Eichenbaum et al., 2007; Squire et al., 2007).
When event-related fMRI is used to image BOLD activation in the medial temporal lobe, dissociations between the magnitudes of this ‘activation’ have been revealed among different subregions of this brain area (in this section
‘activation’ refers to the BOLD signal, and is not meant to imply neural activity directly, although it is presumably a contributing factor). Specifically, perirhinal activation during the encoding phase of a recognition memory task appears to correlate with subsequent item recognition, but not with source recollection. In contrast, hippocampal and parahippocampal activation at encoding do not correlate with later item recognition, but rather are associated with whether or not 129 successful recognition was accompanied by contextual recollection (e.g.,
Davachi et al., 2003; Ranganath et al., 2004). While it may appear that these data support a dual-process account of recognition, there is one caveat. Although the BOLD signal in the perirhinal cortex is similar between correctly recognized items with and without correct contextual recollection, the perirhinal cortex still shows a stronger BOLD signal for recollected items relative to incorrect responses. This indicates that there may not be complete independence of recollection and familiarity, since a full encoding double dissociation between structures has not been demonstrated (Montaldi et al., 2006). Rather than complete process independence, it is possible that recognition only requires the hippocampus when recollection is involved. The perirhinal cortex, on the other hand, is always necessary for recognition regardless of what cognitive processes are engaged. This is consistent with animal lesion data that have produced conflicting results regarding the role of the hippocampus in stimulus recognition
(c.f., Baxter and Murray, 2001; Clark et al., 2001; Zola and Squire, 2001; Winters et al., 2004; Forwood et al., 2005; Wais et al., 2006), which presumably occurs because some tasks rely more on recollection strategies than others. In contrast, perirhinal cortical lesions always produce deficits in item recognition (e.g.,
Ennaceur and Aggleton, 1997; Buffalo et al., 1999; Buffalo et al., 2000; Eacott and Gaffan, 2005; Winters and Bussey, 2005). Together, these data suggest that recollection is always accompanied by the perception that a stimulus is familiar, and thus the idea of pure process independence is not supported. 130
The observation that increased medial temporal lobe BOLD activation is predictive of subsequent recognition memory performance has also been used to support the unitary strength theory. This theory argues that recollection and familiarity are not dissociable processes that are supported by the hippocampus and perirhinal cortex, respectively. Rather, the unitary strength theory contends that both regions respond to memory strength. They argue that the studies that have provided support for a distinction between medial temporal regions in recollection versus familiarity were confounded because they did not control for differences in memory strength (for review, see Squire et al., 2007). When BOLD activation is compared in studies that have varied the strength of item and source memory independently, the results can be interpreted as supporting this unitary memory strength theory. Specifically, activity in the hippocampus during learning varies in relation to the strength of subsequent item memory, even when source memory is held constant at chance levels (Kirwan et al., 2008; Shrager et al.,
2008). Additionally, when item memory strength was held constant, it was reported that hippocampal activity was not related to variations in source memory strength (although there was still a difference in hippocampal activation between recollected and non-recollected stimuli). In the perirhinal cortex, BOLD activation also has been related to the strength of item memory when source memory is held constant at chance levels; however, perirhinal activation levels show no relationship to subsequent recollection strength (Kirwan et al., 2008; Shrager et al., 2008). These data could be interpreted as consistent with the idea that the 131 perirhinal cortex supports familiarity but not recollection (Aggleton and Brown,
1999; Brown and Aggleton, 2001). Alternatively, these data could indicate that a signal for recollection in perirhinal cortex is not correlated with recollection strength because the relationship between recollection strength and activity in the perirhinal cortex is nonlinear (Squire et al., 2007; Kirwan et al., 2008; Shrager et al., 2008).
The magnitude of the BOLD signal in the medial temporal lobe has also been measured during the retrieval phase of recognition memory tasks. In the posterior hippocampus (equivalent to the dorsal hippocampus in rats), the BOLD signal increases as test items are correctly recognized with greater confidence
(e.g., Daselaar et al., 2006), or when recognized items are associated with source information (e.g., Weis et al., 2004). In contrast, accurate identification that a stimulus is old, in the absence of source information, has been associated with decreased activation in the perirhinal cortex (e.g., Weis et al., 2004;
Daselaar et al., 2006; Montaldi et al., 2006). These data have been interpreted as supporting the independent dual-process account of recognition such that the hippocampus is primarily involved in recollection and the perirhinal cortex is primarily involved in familiarity (for review, see Eichenbaum et al., 2007). The possibility cannot be ruled out, however, that high-confidence memories are simply stronger memories, and that memory strength is the driving force for this apparent dissociation between the hippocampus and perirhinal cortex (Wais et al., 2009; for review, see Squire et al., 2007). 132
Theories of the neurobiology of recognition memory that have been based on fMRI data are problematic on a number of levels. First, some studies have reported a decrease in activation in the perirhinal cortex for familiar stimuli relative to novel or forgotten stimuli (e.g., Weis et al., 2004; Daselaar et al., 2006;
Montaldi et al., 2006; for review, see Eichenbaum et al., 2007). Initially, these data appeared to support electrophysiological recordings from rats (Zhu and
Brown, 1995; Zhu et al., 1995) and monkeys (Fahy et al., 1993; Xiang and
Brown, 1998) that showed higher firing rates in the perirhinal cortex for novel relative to familiar stimuli. Because the studies that have reported novelty modulation of perirhinal neuron activity have significant methodological limitations (as discussed in Chapter 4), the interpretation of the human fMRI results with regard to the animal data are not straight forward. Moreover, other studies have shown that neurons in the perirhinal cortex actually increase their firing rates in response to very familiar stimuli (Holscher and Rolls, 2002;
Holscher et al., 2003; Rolls et al., 2005). Presumably, these would be the stimuli that the subject would judge as ‘old’ with the strongest confidence (and these stimuli should produce the greatest memory strength). These conditions do not match the circumstances under which the least amount of BOLD activation in human fMRI studies occurs (e.g., Weis et al., 2004; Daselaar et al., 2006;
Montaldi et al., 2006; for review, see Eichenbaum et al., 2007). Thus, it does not appear that perirhinal cell firing rates directly correlate with the BOLD signal under these recognition memory conditions. 133
A second issue with fMRI investigations of recognition memory is that these data cannot distinguish between single unitary memory strength (Squire et al., 2007), and the independent dual process views (Yonelinas, 1999; Yonelinas,
2001) of recognition memory. In order for this issue to be resolved, other methods are going to be needed. Alternatively, it may never be possible to determine which theory better reflects the empirical data since the two models have become essentially indistinguishable, and both may be flawed. The notion proposed by Squire (2007) that the distinct medial temporal lobe activation patterns observed for forgotten, familiar, and recalled stimuli are due to differences in memory strength relies on the idea that the hippocampus is necessary for judgments of item recognition based on familiarity alone. In most experiments with monkeys and rats, when the lesion is limited to the hippocampus, the data indicate that this assumption is not accurate (e.g., Baxter and Murray, 2001; Forwood et al., 2005; Good et al., 2007; Mumby, 2001;
Mumby et al., 2002; Winters et al., 2004). On the other hand, the dual-process model described by Yonelinas (1999; 2001) assumes that recollection and familiarity are independent, which is also unlikely. In animals, lesions of the perirhinal cortex always produce deficits in item recognition (e.g., Ennaceur and
Aggleton, 1997; Bussey et al., 2000; Malkova et al., 2001; Gaffan et al., 2004).
One would predict that if familiarity and recollection were independent processes, then recollection could compensate and promote accurate recognition when the perirhinal cortex was damaged and thus familiarity detection was compromised. 134
Moreover, the observation that the perirhinal cortex is necessary for auditory fear conditioning (Kholodar-Smith et al., 2008; Kholodar-Smith et al.,
2008; Lindquist et al., 2004) and contextual fear conditioning (Bucci et al., 2000), both of which require associating a stimulus with source information, suggests that the perirhinal cortex does indeed contribute to recollection. Together, these data support a novel account of recognition memory that is a hybrid between the independent dual-process account (Yonelinas, 1999; Yonelinas, 2001), and the unitary strength theory (Squire et al., 2007). In this view, some level of familiarity must always accompany recollection. Therefore, the perirhinal cortex, hippocampus, parahippocampal cortex (Hayes et al., 2007), and the prefrontal cortex (Davidson and Glisky, 2002; Kirwan et al., 2008) are all necessary for normal recollection. In contrast, the perirhinal cortex supports stimulus recognition by binding stimulus features encoded in upstream cortical regions, which allows humans and other animals to dissociate old from new stimuli (see
Chapter 4). This process is independent of the hippocampus when these judgments can be based on item familiarity alone. This account of the data should be considered with regards to age-related changes in human recognition memory, which have primarily been investigated within the framework of a dual- process account.
135
5.4 Recognition memory and human aging
Recognition memory deficits are consistently observed in normal elderly subjects relative to young controls. Specifically, aged humans have an increased rate of misses and false alarms (Gordon and Clark, 1974; Rankin and Kausler,
1979; Norman and Schacter, 1997; Schacter et al., 1997; Tun et al., 1998). The age-related increase in false alarms is particularly salient when lures that share common features with the study items are used as the novel stimuli (Schacter et al., 1997). Because it is believed that the perirhinal cortex may function to disambiguate stimuli that share common features (the perceptual-mnemonic theory, see Chapter 4), and the ability to distinguish between items that share common elements is disrupted by normal aging (Schacter et al., 1997), this provides support for the idea that perirhinal cortical function is disrupted by old age (see Chapters 6 and 7). Additionally, the increased rate of false alarms in aged humans parallels the observation that old rats show stimulus recognition impairments, because they are more likely to behave as if novel stimuli are familiar (see Chapter 6).
Proponents of the independent dual-process model have argued that age- associated deficits in recognition memory arise because one’s ability to use recollection is comprised by advanced age, but that the familiarity process is not affected in normal elderly humans. This model assumes that there are age- related functional changes in the hippocampus and prefrontal cortex that decrease one’s ability to recollect the source information associated with the 136 encoding of a new stimulus, while perirhinal cortical function remains intact.
Therefore, older subjects may rely more on familiarity in order to recognize a stimulus, and an increase in a perirhinal ‘familiarity signal’ may even compensate for recollection impairments (Daselaar et al., 2006). Although, it is evident that normal aging interferes with the ability to recollect, the dual-process account of aging and recognition memory may not accurately capture the impact of aging on familiarity-based stimulus recognition.
When older and younger adults perform recognition memory tests, older adults self-report using less recollection compared to younger subjects (Mantyla,
1993; Parkin and Walter, 1992; Bastin and Van der Linden, 2003). Moreover, declines in recollection correlate with both medial temporal lobe and frontal lobe dysfunction (Parkin and Walter, 1992; Davidson and Glisky, 2002). Although there is a general consensus in the aging literature that elderly humans have a reduced ability to use recollection in order to identify a stimulus as ‘old’, the evidence concerning familiarity is equivocal. Many studies that have used the process dissociation procedure have reported no effect of aging on measures of familiarity (Hay and Jacoby, 1999; Jennings and Jacoby, 1993; Jennings and
Jacoby, 1997; Titov and Knight, 1997), but other studies using this same procedure have found reduced familiarity with age (Davidson and Glisky, 2002;
Prull et al., 2006; Toth and Parks, 2006). Importantly, these familiarity impairments are associated with medial temporal lobe dysfunction (Davidson and 137
Glisky, 2002). Together these data suggest that although recollection is more vulnerable to the process of normal aging, familiarity can also be affected.
Recollection deficits may be more detectable during normal aging, because this process is affected if either the prefrontal cortex or the medial temporal lobe is functionally altered with aging. In contrast, familiarity apparently only declines in subjects that have medial temporal lobe dysfunction (Davidson and Glisky, 2002). Animal work specifically indicates that it is the perirhinal cortex that is essential for identifying a stimulus as familiar (see Chapter 4). Additionally, if the assertion that recollection is always accompanied by some level of familiarity is correct (see previous section), then subtle age-related changes in familiarity would also affect recollection. Therefore, during aging a decline in hippocampal function, prefrontal function, or perirhinal cortical function would contribute to reduced recollection. On the other hand, familiarity may be left intact unless the perirhinal cortex is disrupted during the aging process.
A number of experiments have used fMRI to investigate how BOLD activation in the brain regions associated with recognition memory change with age. The consistent observation has been reduced hippocampal activation during encoding and retrieval in older adults compared to young controls (Figure 5.4A;
Daselaar et al., 2006; Dennis et al., 2007; Dennis et al., 2008; Dennis et al.,
2008). In contrast, it has been reported that during retrieval aged subjects show an increased BOLD signal in the rhinal cortex for novel stimuli relative to young subjects, but as the subject identifies the stimulus as old with greater confidence, 138 the BOLD activation between young and aged subjects becomes similar.
Therefore, the difference in rhinal BOLD activation with increasing confidence that a stimulus has been seen before, is greater in aged compared to young subjects. Figure 5.4 shows the differences in activation associated with memory
Figure 5.4: Hippocampal and rhinal activation and associated memory confidence. The reported effect of aging on BOLD activation in the medial temporal lobe was a dissociation between the hippocampus and rhinal cortex. Data from young subjects are shown in yellow and the blue lines indicate data from aged humans. According the Daselaar et al., (2006) recollection-related activity in the hippocampus was attenuated by aging, and familiarity-related activity in the rhinal cortex was enhanced by aging. Figure from Daselaar et al., 2006. 139 confidence for the case of the hippocampus and rhinal cortex (Daselaar et al.,
2006). These data have been interpreted as older adults showing an enhanced rhinal-dependent familiarity signal in order to compensate for reduced recollection (Daselaar et al., 2006). This explanation of the data, however, is not necessarily consistent when the entire body of the dataset is considered. That is, the compensation notion implies that within an aged subject the decrease in the hippocampal recollection-associated BOLD signal should be accompanied by a larger decrease in the BOLD signal in the rhinal cortex for the most confidently remembered stimuli. At high confidence levels (“definitely old”), however, the rhinal BOLD signal is similar between old and young subjects (Figure 5.4B).
Additionally, the decrease in BOLD activation with increasing confidence that a stimulus is old was found in many other brain regions outside of the rhinal cortex, including: fusiform gyrus, lateral temporal cortex, prefrontal cortex, caudate nucleus, thalamus, occipital cortex, cerebellum, premotor cortex, frontal pole, anterior cingulate, superior parietal cortex, and the insula. Many of these regions have not been implicated in contributing to normal recognition memory. This raises questions concerning the anatomical specificity of the BOLD activation observed in this study. Although the fMRI data clearly show that hippocampal activity is lower in older subjects making the “definitely old” judgment, the fact there is greater hippocampal activation for stimuli judged as “new”, and that other structures show age-related changes similar to that observed in the rhinal cortex complicates a simple conclusion of hippocampal involvement in age-related 140 recollection impairments. Therefore, additional methods should be used to examine the neurobiology of age-associated item recognition impairments.
5.5 Discussion
Over the past several decades researchers have argued over whether recognition memory is a unitary process, or if it is supported by two independent components: recollection and familiarity (c.f., Malmberg, 2008; Yonelinas, 2001).
Additionally, there is no consensus regarding the neurobiological contributions to human stimulus recognition. Some researchers argue for a unitary strength model, which hypothesizes that all structures in the medial temporal lobe support familiarity. By this account, the reported differences in brain activation between recollection and familiarity are confounded by changes in memory strength
(Squire et al., 2007). Others contend that recollection and familiarity are two independent components of recognition (Brown and Aggleton, 2001;
Eichenbaum et al., 2007; Yonelinas et al., 2005). To date, behavioral data, models, and fMRI experiments have not been able to resolve these debates.
Moreover, it is unlikely that a completely accurate account of an animal’s ability to distinguish old from new stimuli has been described. Because this cognitive function declines during normal aging, it is important to gain a better understanding of the neurobiology of stimulus recognition. 141
CHAPTER 6: THE EFFECTS OF AGE AND CONTEXT ON SPONTANEOUS
OBJECT RECOGNITION
6.1 Introduction
The ability to discriminate novel stimuli from ones that have been previously encountered is a basic prerequisite for accurate stimulus recognition.
Moreover, this process is required for the retrieval of events that are familiar as well as the correct discrimination of events that are novel. Electrophysiological
(e.g. Higuchi and Miyashita, 1996; Erickson et al., 2000; Naya et al., 2001;
Holscher et al., 2003), animal lesion (e.g. Mumby et al., 2002; Baxter and
Murray, 2001; Baxter and Murray, 2001; Prusky et al., 2004; Nemanic et al.,
2004; Bussey et al., 2000; Bussey et al., 1999; Fahy et al., 1993; Malkova et al.,
2001; Winters and Bussey, 2005; Winters et al., 2004), and neuroimaging (e.g.
Pihlajamaki et al., 2004; Ranganath et al., 2004; Brozinsky et al., 2005) studies all support the notion that the perirhinal cortex is essential for stimulus recognition.
In rodents, the tendency to explore a novel object more than an object that has been experienced previously can be exploited as a sensitive test of stimulus recognition (Ennaceur and Delacour, 1988). Such spontaneous object recognition (SOR) tasks do not require a learning rule, nor do they rely on reinforcers such as food reward or foot shock (Ennaceur and Delacour, 1988).
Additionally, performance on the SOR task is critically dependent on the perirhinal cortex (Winters and Bussey, 2005; e.g. Ennaceur and Aggleton, 1997; 142
Abe et al., 2004), which was discussed extensively in Chapter 4, but it does not require an intact hippocampus (e.g., Winters et al., 2004; Forwood et al., 2005;
O'Brien et al., 2006; Good et al., 2007). Therefore, the SOR task allows for the evaluation of age-related cognitive decline that may be occurring independent of changes in hippocampal function.
While it is well documented that the normal aging process affects behaviors that require the medial temporal lobe, the impact of aging on object recognition has not been well-characterized. Furthermore, the data obtained from young and old rats that performed SOR tasks have been contradictory.
Specifically, when 24 month old male Wistar rats were tested on a SOR task they did not show novelty discrimination deficits relative to younger rats (Cavoy and
Delacour, 1993). This study, however, only used delays up to 5 min between the object familiarization and the test phases. At longer delays, decreases in novelty discrimination have been reported to occur in aged rats (Pieta Dias et al., 2007; de Lima et al., 2005; Pitsikas et al., 2005; Bartolini et al., 1996; Vannucchi et al.,
1997). These experiments often report that aged animals show a reduced
“discrimination ratio”, which is a measure of the preference the rat has for exploring the novel object relative to the familiar object (Dix and Aggleton, 1999).
The discrimination ratio is derived by calculating the difference in the amount of time that a rat spent exploring a novel object versus a familiar object and then dividing that difference by the total exploration time. Although it is assumed that a decline in the discrimination ratio results from a decline in recognition memory, it 143 does not clearly separate different processes that may influence performance of the task. Specifically, a decline in the discrimination ratio can result from either an increase in exploration of the familiar object or a decrease in exploration of the novel object, and these two outcomes necessarily lead to very different conclusions. The former implies a memory deficit, while the latter could be a consequence of a decline in novelty detection. These two processes, although related, likely have a distinguishable neurobiological etiology (Daselaar et al.,
2006) and require further investigation with regards to potential age-associated declines in stimulus recognition.
The current investigation examined the extent to which age-related deficits in object recognition could be attributed to memory impairments versus a decline in novelty detection. A deficit in memory should manifest itself if the rat behaves as if a familiar object is novel, and would be associated with increased exploration of the familiar object during the test phase of the SOR task. In contrast, if an aged rat does not correctly identify a stimulus that has never been experienced before as novel, one would expect the rat to behave as if the object was familiar by exploring it less during the test phase of the SOR task. These two alternate possibilities were examined using the standard SOR task (Experiment
1) and two additional variants of this task. The first variant of the SOR used in the current investigation (Experiment 2) simultaneously presented two identical objects during both the familiarization phase and the test phase. This enabled a direct comparison of object exploration between the familiarization and test 144 phases so that the effect of novel versus repeated object presentations could be evaluated. For the second variant of the SOR task (Experiment 3), both the room and the testing arena were changed between the familiarization and test phases in order to examine the effect of incongruent contexts on object recognition in young and old rats. The spatial learning and memory abilities of all rats were also characterized using the Morris swim task (Morris, 1984) enabling a comparison between hippocampal-dependent spatial memory, and perirhinal-dependent stimulus recognition performance across age groups.
6.2 Methods
6.2.1 Subjects and spatial memory testing
A total of forty-four young (7-9 months old) and fifty-one aged (24-25 months old) male F344 rats (from the National Institute on Aging's colony at
Harlan-Sprague Dawley, Indianapolis, IN) participated in these behavioral experiments. The rats were housed individually in plexiglas guinea pig tubs and maintained on a reversed 12-hr light-dark cycle. All behavioral testing occurred during the dark phase of the rats’ light-dark cycle and each animal was given access to food and water ad libitum for the duration of these experiments.
After arriving, rats were handled by experimenters over several days for at least 5-10 minutes per day. Once animals were comfortable while being handled, they were tested on the spatial and the visually-cued versions of the Morris swim task (Morris, 1984). This procedure has been described in detail previously (e.g. 145
Barnes et al., 1996; Shen et al., 1996). Briefly, during the spatial version of the
Morris swim task, all animals were tested over 4 days with 6 trials on each day.
During these trials, an escape platform was hidden below the surface of water, which was made opaque with non-toxic Crayola paint. Rats were released from seven different start locations around the perimeter of the tank, and each animal performed two successive trials. The order of the release locations was pseudo- randomized for each rat such that no rat was released from the same location on two consecutive trials. Immediately following the 24 spatial trials, animals were screened for visual ability with 2 days of cued visual trials (6 trials per day) in which the escape platform was above the surface of the water but the position of the platform changed between each trial. Rats’ performance on the swim task was analyzed offline with either in-house software (WMAZE, M. Williams) or a commercial software application (ANY-maze, Wood Dale, IL). Because different release locations and differences in swimming velocity produce variability in the latency to reach the escape platform, a corrected integrated path length (CIPL) was calculated to ensure comparability of the rats’ performance across different release locations (Gallagher et al., 1993). After the visually-cued trials of the
Morris swim test, eight aged rats and three young rats were excluded from the experiment due to poor vision (data not shown) and these animals did not participate any of the object recognition experiments.
146
6.2.2 Spontaneous object recognition testing procedures
Within two weeks of completing the Morris swim task procedures, each rat participated in one of three different behavioral experiments designed to evaluate stimulus recognition. Experiment 1 (Figure 6.1) was the standard spontaneous object recognition (SOR) test first described by Ennaceur and Delacour (1988), which has been used extensively to investigate the neurobiological origins of stimulus recognition (discussed in Chapter 4). In Experiment 2, a variant of the
SOR task that simultaneously presents either two novel, or two familiar objects during the test phase was used (Figure 6.2). This simultaneous novel or repeat presentation paradigm is better able to distinguish whether or not recognition deficits are caused by ‘forgetting’ versus impairments in novelty detection
(McTighe et al., 2008). Finally, in Experiment 3 both the room and the testing arena were changed between the object familiarization phase and the testing phase in order to evaluate the effect of incongruent contexts on object recognition in young and aged rats (Figure 6.3).
The apparatus used for Experiments 1 and 2, and for the familiarization phase of Experiment 3 was a box constructed from wood, and it was 30 cm by 30 cm and contained walls that were 30 cm tall. All walls of the apparatus were painted black, and the floor was black with a grid that was used to ensure that the location of objects did not change between object familiarization and test phases. Figure 6.1 shows a schematic diagram of this testing apparatus. Due to the requirement for a change in context, an additional apparatus was used for the 147 test phase of Experiment 3 (Figure 6.3; right arena). This testing box was identical in size (30 cm x 30 cm x 30 cm) to the original apparatus but the walls were brown and the floor was grey. For both apparati, an overhead camera and a video recorder were used to monitor and record the animal’s behavior for subsequent analysis.
6.2.3 Experiment 1: Standard spontaneous object recognition (SOR) testing
Stimulus recognition was measured in 16 young (7-9 months) and 18 aged (24-25 months) male F344 rats using the standard version of the spontaneous object recognition task. Before testing began, all rats were exposed to the empty apparatus (Figure 6.1) for 10 minutes on two consecutive days.
Recognition testing began the day immediately following this habitation procedure.
SOR testing was composed of an acquisition or object familiarization phase followed by a test phase. All rats participated in four object familiarization and test phases with 4 different delays between the acquisition and the recognition testing. Delays of 2 min, 15 min, 2 hours and 24 hours were used for
Experiment 1. The stimuli presented were triplicate copies of objects made of glass, plastic or wood that varied in shape, color, and size. In the object familiarization phase, duplicate copies of an object (Figure 6.1; A1 and A2) were placed near the two corners at either end of one side of the arena (10 cm from each adjacent wall). The pink cylinders and the yellow pyramid in Figure 6.1 148
Familiarization Phase Test Phase
delay
A1 A3
A2 B1
Figure 6.1: Schematic of the standard SOR task used in Experiment 1. The testing arena was 30 cm by 30 cm with 30 cm high walls and it was painted black. The floor of this apparatus had a grid painted on it to ensure that the object placement was consistent between the familiarization phase and the test phase. The pink cylinders and the yellow pyramid indicate the location that the objects were placed in during both phases of recognition testing, and the schematic white rat shows the orientation that the rat was placed in when it is first put in the arena. During the object familiarization phase (left box), a rat is placed in the arena and allowed to explore duplicate copies of an object (A1 and A2). After this object familiarization phase, the rat is moved out of the arena for a variable delay (2 min, 15 min, 2 hr, or 24 hr). Following the delay, during the test phase (right arena), the animal is returned to the arena to explore a triplicate copy of the objects presented during the familiarization phase (A3) and a novel object (B1). indicate the positions of the objects for both phases of recognition testing. The animal was placed into the arena facing the center of the opposite wall (Figure
6.1; schematic white rat), and then allowed a total of 4 min of exploration in the open arena. After this object familiarization phase, for the 2 minute delay, the rat was placed in a covered pot next to the apparatus. This prevented the animal from being exposed to extraneous visual stimuli. For the 15 min, 2 hr, and 24 hr delays, the rat was returned to its cage in the colony room. Rats were placed in covered pots for all transferring between the colony room and the experimental apparatus. 149
During the test phase, the animal was returned to the apparatus and placed back in the same start location as in the object familiarization phase
(Figure 6.1; schematic white rat). Again, the rat was allowed 4 min of exploration but was presented with two different objects than had been used during acquisition. One object (Figure 6.1; A3) was the third copy of the triplicate set of the objects used in the object familiarization phase (familiar), and the other was a novel object (Figure 6.1; B1). The pink cylinder and yellow pyramid in the right box of Figure 6.1 represent the location of the objects during the testing phase.
All objects and the testing apparatus were washed with 70% ethanol between every trial and before procedures began with another rat. The positions of the objects in the test phases and the objects used as novel or familiar were counterbalanced between the young and the aged animals.
Different sets of objects were used for each delay condition. Additionally, each rat only participated in one object familiarization/testing procedure a day. It was observed that rats habituate quickly to the objects, even when novel, and performing more than one testing session per day leads to reductions in overall exploration.
As stated previously, all episodes of exploration were videotaped for offline analysis. Exploratory behavior was defined as the animal directing its nose toward the object at a distance of ~2 cm. Any other behavior, such as resting against the object, or rearing on the object was not considered to be exploration.
Exploration was scored by an observer blind to the rat’s age and the duration of 150 the delay between the object familiarization and the test phases. Additionally, the amount of time spent exploring objects during the test phase was scored prior to measuring the amount of exploration during the object familiarization phase. This reverse-order of analysis ensured that the scorer was blind as to which object was familiar and which object was novel. The difference in time spent exploring the novel object compared with the familiar object divided by the time spent exploring both objects was calculated as the discrimination ratio (Dix and
Aggleton, 1999). A higher discrimination ratio is indicative of more exploration of the novel object relative to the familiar object during the test phase.
6.2.4 Experiment 2: Novel/repeat SOR testing
Object recognition was also tested using a variant of the spontaneous object recognition task that simultaneously presents two novel objects during the familiarization phase, but during the test phase either two different novel objects
(“novel” condition) are presented, or the objects from the familiarization phase are presented again ("repeat" condition; McTighe et al., 2008). A total of 10 young (7-9 months old) and 10 aged (24-25 months) F344 rats participated in this behavioral experiment.
Rats were habituated to the testing apparatus for 10 minutes per day over two consecutive days, and recognition testing began the day immediately following the final habituation episode. The stimuli presented were duplicate copies of objects, and each rat only participated in 2 object familiarization and 2 151 test phases that were separated by a 2 hr delay for both trials. In the first familiarization phase, duplicate copies of an object (e.g., C1 and C2) were placed near the two corners at either end of one side of the arena (10 cm from each adjacent wall). The animal was placed into the arena facing the center of the opposite wall (Figure 6.2; schematic white rat), and then the rat was allowed a total of 4 min to explore in the open arena. After this object familiarization phase, the rat was returned its home cage in the colony room for the 2 hr delay. After the delay, the rat was returned to the same testing apparatus. For half of the young and old rats the same two objects from the familiarization phase were again placed in the testing arena at the same locations (C1’ and C2’; repeat condition).
For the other 5 young and 5 aged rats, two identical novel objects (D1 and D2; novel condition) were placed in the arena in the same location as the objects from the familiarization phase. In both cases, the rats were given 4 min to explore. After the test phase, the rat was returned to the colony room for 24 hours before participating in another familiarization phase. During this second object familiarization phase, the rat was given 4 min to explore two identical novel objects that were distinct from any of the objects used the previous day (E1 and E2). Additionally, during this second familiarization phase in Experiment 2, the position of the objects in the arena was changed (Figure 6.2; left panel) in order to help promote more exploration (Cavoy and Delacour, 1993). After this object familiarization phase, the rat was returned to its home cage in the colony room for the 2 hr delay. After this delay, the rats that had participated in the 152 repeat condition on previous day were exposed to the two novel objects that they had not been presented with (D1 and D2; novel condition). Conversely, the 5 young and the 5 aged rats that were given a novel presentation for the test phase on the previous day were exposed to the same objects from the familiarization
Test Phase 1 Test Phase 2
Familiarization Familiarization Phase 1 – All rats Phase 2 – All rats C2’ D2
C1’ D1
Repeat Condition Novel Condition E2 C2 5 young/5 aged rats 5 young/5 aged rats 2 hr E1 2 hr C1 delay delay
D2 E2’
D1 E1’ Novel Condition 5 young/5 aged rats Repeat Condition 5 young/5 aged rats Figure 6.2: A schematic of the presentation of objects used in novel/repeat SOR task (Experiment 2). Experiment 2 consisted of familiarization phase 1 followed by a 2 hr delay and then test phase 1. The next day, rats participated in a familiarization phase 2, which was again followed by a 2 hr delay and a test phase. During the first familiarization phase, duplicate copies of an object (e.g., C1 and C2) were presented and the rat was allowed a total of 4 min of exploration in the open arena. After a 2 hour delay, the rat was returned to the same testing apparatus. For half of the young and the aged rats the same two objects from the familiarization phase were again placed in the testing arena in the same location (C1’ and C2’; repeat condition). For the other 5 young and 5 aged rats two identical novel objects (D1 and D2; novel condition) were placed in the arena in the same location as the objects from the familiarization phase. In both cases, the rats were given 4 min to explore. After the test phase, the rat was returned to the colony room for 24 hours before performing another object familiarization phase. During this second object familiarization phase, the rat was given 4 min to explore two identical novel objects that were distinct from any of the objects used the previous day (E1 and E2) and in a different position within the arena. After a 2 hour delay, the rats that had participated in the repeat condition on previous day were exposed to the two novel objects that they had not been presented with (D1 and D2; novel condition). Conversely, the 5 adult and the 5 aged rats that were given a novel presentation on the previous day were exposed to the same objects from the object familiarization phase (E1’ and E2’; repeat condition).
153 phase for the second test session (E1’ and E2’; repeat condition). Therefore, all rats performed 1 repeat condition and one novel condition with a 2 hr delay, and the design was counterbalanced to control for order effects. Figure 6.2 shows a schematic representation of the novel/repeat SOR testing procedure used in
Experiment 2.
As in Experiment 1, exploratory behavior was defined as the animal directing its nose toward the object at a distance of ~2 cm. The amount of time spent exploring objects was scored by an observer blind to the rat’s age, and to the phase of the experiment (familiarization or test). For this experiment, it was not possible to calculate a discrimination ratio since the objects and their relative novelty were always identical. Therefore, all comparisons between age groups, and condition (repeated versus novel) were made using total exploratory time.
6.2.5 Experiment 3: SOR testing with context change
The effect of an incongruent context between the object familiarization and the test phase of the spontaneous object recognition task was measured in 15 young (9 months) and 15 aged (24 months) F344 rats. The experimental procedures used in Experiment 3 were similar to those in Experiment 1. Before testing began, all rats were exposed to both of the apparati used for this behavior
(Figures 6.3 arena A and B) for 10 minutes in each arena on two consecutive days. The rat was exposed to arena A (Figure 6.3; left box) first for 10 minutes and then returned to its home cage in the colony room for several hours. After 154 this rest period, the rat was then exposed to arena B (Figure 6.3; right box) for 10 minutes. Recognition testing began the day immediately following this habituation procedure. Arena A was used for both familiarization phases, while the two testing sessions always occurred in arena B.
Each rat participated in two object familiarization and test phases with delays of 2 minutes and 24 hours between the familiarization phase and the test phase. The stimuli used were triplicate copies of objects. In the familiarization phase, duplicate copies of an object (e.g., F1 and F2) were placed near the two corners at either end of one side of the arena (10 cm from each adjacent wall).
The two green cubes in Figure 6.3 indicate the positions of the objects for the familiarization phase of recognition testing. The animal was placed into the arena facing the center of the opposite wall (Figure 6.3; schematic white rat), and then allowed a total of 4 min of exploration in the open arena. After this object familiarization phase, for the 2 minute delay, the rat was relocated, in a covered pot, to a different room that contained arena B and then objects recognition testing occurred in arena B. For the 24 hr delay the rat was immediately returned to its home cage in the colony room after the familiarization phase. After the 24 hr delay, the rat was taken from the colony room and transferred to arena B for testing in a covered pot to minimize extraneous visual cues. 155
Familiarization phase Test phase Room 1, Arena A Room 2, Arena B
delay G1 F2
F1 F3
Figure 6.3: The procedure for the SOR task with context change (Experiment 3). In the object familiarization phase, duplicate copies of an object (e.g., F1 and F2) were presented to the rat, which was then allowed a total of 4 min of exploration in the open arena. After the familiarization phase, and a 2 min or a 24 hour delay, the rat was relocated to a different room that contained arena B. During the test phase in arena B, the two objects were placed in similar locations relative to the walls of the apparatus as in the familiarization phase and the rat was allowed 4 min of exploration. During the test phase, one object (F3) was the third copy of the triplicate set of the objects used in the familiarization phase, and the other was a novel object (G1).
During the test phase in arena B, the two objects were placed in similar locations relative to the walls of the apparatus as in the familiarization phase.
These two locations are marked by the green cube and the blue cross in Figure
6.3. Additionally, the rat was placed in arena B at a similar start location relative to the walls and objects as was used during the object familiarization phase in arena A (Figure 6.3; schematic white rat). The rat was allowed 4 min of exploration in arena B with two different objects than had been used during acquisition. One object (F3) was the third copy of the triplicate set of the objects used in the familiarization phase (familiar), and the other was a novel object (G1).
Figure 6.3 shows the object presentation procedures used for Experiment 3. All 156 objects and the testing apparatus were washed with 70% ethanol between every trial and before procedures began with another rat.
Exploratory behavior was scored offline, from videotape, by an observer blind to the age of the rat using the same criteria as in Experiments 1 and 2.
Moreover, as in Experiment 1, the amount of time spent exploring objects during the test phase was measured prior to the scoring of exploration during the familiarization phase. This ensured that the observer was blind as to the relative familiarity of the objects. The difference in time exploring the novel object compared with the familiar object divided by the time spent exploring both objects was calculated as the discrimination ratio (Dix and Aggleton, 1999).
6.2.6 Data analysis
Data from the three different experiments were analyzed separately. In
Experiments 1 and 3, group means of four measures (the total time spent exploring objects during the familiarization phase, the discrimination ratio, the time spent exploring the novel object during the test phase, and the time spent exploring the familiar object during the test) were examined using repeated- measures analysis of variance (ANOVA). The within subjects factor of delay duration and the between-subjects factor of age group were the independent variables. For Experiment 2 the within-subjects factor of experimental condition
(repeat versus novel), the between-subjects factor of age group on the total time spent exploring objects during the familiarization phase, and total time spent 157 exploring objects during the test phase were analyzed with repeated-measures
ANOVA. All statistical tests and p-values were calculated using SPSS 9.0
(Chicago, Illinois) and alpha was set at the 0.05 level.
6.3 Results
6.3.1 Morris swim task performance
The mean performance of the aged rats on the spatial version of the
Morris swim task was significantly worse than the performance of the young rats.
Every aged rat that participated in recognition testing was able to learn the visual version of the task, however. Figure 6.4A shows the mean CIPL (Gallagher et al.,
1993) for the forty-three aged rats (purple circles), and the forty-one young rats
(green circles) that participated in these behavioral experiments. All animals were tested over 4 days with 6 spatial trials on each day. As reported previously (e.g.,
Barnes et al., 1997), the old rats had a significantly longer mean CIPL scores, compared with the young rats (F[1,312] = 86.98, p < 0.001; ANOVA). Post hoc
analysis revealed that this difference was due to the aged rats having
significantly longer CIPL scores on days 2, 3 and 4 of spatial testing (p < 0.001;
for all comparisons; Tukey HSD) while there was no significant difference in the
path lengths between young and aged rats on Day 1 of spatial testing (p = 0.99;
Tukey HSD). The old rats, however, did show a significant improvement in
finding the hidden platform between Day 1 and Day 4 of spatial testing (F[1,40] =
13.50, p < 0.01; repeated-measures ANOVA). This improvement did not differ 158
A 40 Young (spatial) Aged (spatial) 30 Young (visual) Aged (visual) 20 CIPL (m) CIPL 10
0 Day 1 Day 2 Day 3 Day 4 B 50 )
40
30
20
10 Spatial Trials Day 4 CIPL (m 0 Young Aged Figure 6.4: Morris swim task performance of young and aged rats. (A) The X-axis is the day of testing and the Y-axis is the mean corrected integrated path length (CIPL) score. Higher CIPL scores indicate longer path lengths to reach the escape platform. All rats completed 4 days of spatial trials (circles) in which the platform was hidden below the surface of the water. These spatial trials were followed by 2 days of visually-cued trials in which the platform was visible (triangles). During the spatial trials, the aged (purple) rats had significantly longer CIPL scores compared with the young (green) rats (F[1,312] = 86.98, p < 0.001; ANOVA). The performance of the aged rats, however, benefited significantly more when the escape platform was visible compared to the young rats (F[1,82] = 32.42, p < 0.001; repeated-measures ANOVA), and it is unlikely that the aged animals used in the current series of experiments were blind. Error bars represent +/-1 standard error of the mean. (B) The mean CIPL scores of individual young (green) and aged (purple) rats on Day 4 of spatial testing. The horizontal black lines indicate the mean CIPL for each age group. In the aged rats, CIPL score values below the grey horizontal line represent rats that performed within 1 standard deviation of the young animals. Only 7 aged rats met this criterion while the other 36 rats had CIPL scores that were at least 1 standard deviation above the young rat mean. significantly between the aged rats that participated in the three different experiments (F[2,40] = 2.52, p = 0.1; repeated-measures ANOVA), indicating that the groups of aged rats that participated in the three different experiments had comparable spatial learning impairments. Figure 6.4B shows the CIPL scores for the spatial trials on Day 4 of testing for individual rats. The horizontal black lines 159 indicate the mean CIPL for each age group. The CIPL scores for 36 of the aged rats were more than 1 standard deviation above the mean score for the young animals. Only 7 aged rats had a CIPL score within 1 standard deviation of the young animals, and none of the young rats had a CIPL score that was within 1 standard deviation of the aged rat mean. For this reason, analyses of object recognition measures only compared the young rats to the aged rats because the small number of “spatially-unimpaired” aged rats did not provide adequate statistical power to further subdivide the aged group into impaired and unimpaired.
To test whether the spatial impairments of the aged rats were due to visual problems, and to ensure that all rats had adequate vision to participate in recognition testing, the rats also performed 12 trials (6 trials/day) on the visually- cued version of the Morris swim task in which the platform was raised above the surface of the water, but its location changed every trial. Figure 6.4A shows the
CIPL values for the visual trials of the Morris swim task for the trials on Day 1 and
Day 2 for the aged (purple triangles) and the young (green triangles) animals.
Both the aged and the young rats showed a significant decrease in the CIPL on the 2nd day compared with the first day (F[1,78] = 52.53, p < 0.001; repeated-
measures ANOVA). There was an effect of age group on the CIPL score for Day
2 of visual testing, with aged rats having significantly longer path lengths
compared to the adult animals (T[82] = 2.52, p < 0.05; student’s T-test). If the CIPL
score from Day 4 of spatial testing was compared to the visual scores, however, 160 it was observed that the performance of the aged rats benefited significantly more when the escape platform was visible compared to the young rats (F[1,82] =
32.42, p < 0.001; repeated-measures ANOVA). Moreover, there was no correlation between the visual trial CIPL scores on Day 2 of testing and the spatial trial CIPL scores on Day 4 of testing in old rats (r = -0.01, p = 0.9). This indicates that the rats that did worse on the spatial trials were not the rats with poorer vision and it is unlikely that the aged animals used for the current series of experiments were significantly visually impaired.
6.3.2 Experiment 1: standard SOR task results
Figure 6.1 shows a schematic of the standard SOR task that was used for
Experiment 1. The young and the aged rats spent a comparable amount of time exploring the two identical objects during the familiarization phases, and the total exploration times of the young and the aged rats during the familiarization phases that preceded recognition testing are shown in Table 6.1. There was no significant difference in the amount of exploration time during the object
Table 6.1: Exploration times during the familiarization phase of Experiment 1 2 min 15 min 2 hr 24 hr
Young 41.2 sec 31.3 sec 43.0 sec 37.9 sec ± 5.0 ± 5.5 ± 6.9 ± 9.4 Aged 36.8 sec 28.2 sec 36.5 sec 19.6 sec ± 6.2 ± 4.2 ± 4.8 ± 3.9 161
familiarization phases between the two age groups (F[7,128] = 1.73, p = 0.11;
ANOVA).
When the discrimination ratios measured during the test phase were
examined, however, the aged rats had worse object recognition relative to the
young rats as indicated by the significantly smaller discrimination ratios
measured in the aged compared to the young animals (F[7,128] = 7.26, p < 0.001;
ANOVA). Planned comparisons revealed that the discrimination ratio was
significantly smaller in the aged rats relative to the young rats for the 15 min, 2
hr, and 24 hr delays (p < 0.05; repeated contrasts). The discrimination ratio was
not significantly different between the young and the aged rats at the 2 min delay,
however (p = 0.28, repeated contrast). This indicates that the aged rats were
able to discriminate between the novel and the familiar object at short delays,
and thus it is unlikely that the recognition impairments at the longer delays are
due to the age-associated vision, olfactory or somatosensory impairments or
changes in the general motivation to explore objects. Figure 6.5A shows the
discrimination ratios of the adult (green) and aged rats (purple) during the test
sessions for all delays. Although lower discrimination ratios are often interpreted
as impairments in recognition memory, this can occur either by increased
exploration of the familiar object or reduced exploration of the novel object and
the discrimination ratio measure is limited in that it cannot distinguish between
the these two possibilities. Therefore, we also compared the total time spent
exploring the novel object versus the familiar object during the test phase. 162
Analysis of exploration times of individual objects, during the test phase, showed that the age-associated reduction in the discrimination ratio at the 15 min, 2 hr, and 24 hr delays was associated with a reduction in novel object exploration in the aged compared to the young rats. In contrast, the amount of time spent exploring the familiar objects was similar in both age groups. Figure
6.5B shows the mean time spent exploring the novel and familiar objects for the
A 0.6
0.4
0.2 young
Discrimination ratio aged
0 B 2 min 15 min 2 hr 24 hr Delay 50 40 30 20 10 0 Exploration time (sec) time Exploration familiar novel familiar novel familiar novel f amiliar novel
2 min 15 min 2 hr 24 hr Delay Figure 6.5: SOR task performance (Experiment 1). (A) The mean discrimination ratio of the adult (green) and the aged rats (purple) measured during the test phase for the four different delay conditions. A higher discrimination ratio indicates that the animal spent more time exploring the novel object relative to the familiar object. Overall, the aged rats had significantly smaller discrimination ratios when compared to the adult rats (F[7,128] = 7.26, p < 0.001; ANOVA). (B) The mean amount of time young (green) and aged (purple) rats spent exploring the familiar and the novel object during the test phase for the four different delay conditions. There was a significant main effect of novel versus familiar objects on exploration time (F[1,128] = 122.03, p < 0.001; repeated-measures ANOVA). Additionally, the interaction effect of age group (adult versus aged), and object (novel versus familiar) on exploration time during the test phase was significant (F[1,128] = 22.08, p < 0.001). Error bars represent +/-1 standard error of the mean.
163 young (green), and aged rats (purple) for all four delay conditions. There was a significant main effect of novel versus familiar objects on exploration time (F[1,128]
= 122.03, p < 0.001; repeated-measures ANOVA). Additionally, the interaction
effect of age group, and object familiarity (novel versus familiar) on exploration
time during the test phase was also statistically significant (F[1,128] = 22.08, p <
0.001). Planned contrasts comparing the exploration times of the novel object
between the young and the age rats revealed that this interaction effect was due
to a significant reduction in the old rats’ exploration of novel objects for both the 2
hr and the 24 hr delay conditions (p < 0.01; repeated contrasts). The time spent
exploring the novel object in the test phase of the 2 min delay was not significantly different between aged groups, however (p = 0.48; repeated contrast). In contrast to the novel objects, there were no significant differences in the time spent exploring familiar objects in the test phase between the young and the aged rats during the 2 min (p = 0.83; repeated contrast), 15 min (p = 0.91;
repeated contrast), 2 hr (p = 0.11; repeated contrast), or the 24 hr (p = 0.15;
repeated contrast) delay conditions. These data suggest that the age-associated
reduction in the discrimination ratio at delays greater than 2 min is not due to
aged rats forgetting the familiar object, which would be associated with more
exploration of the familiar object. Rather this deficit selectively results from the
aged rats exhibiting reduced exploration of the novel object.
The observation that aged rats do not show a significant reduction in novel object exploration during the test phase after a 2 min delay is interesting and 164 suggests that there is not an overall decrease in motivation to explore during the test phase. One potential reason for the different results between the 2 min delay condition and the longer delay conditions is that this is the only delay condition where the rats remained in the same room as the testing arena. If novel object exploration decreases in aged rats after long delays because extraneous visual stimuli intervene during the delays, this might explain the disrupted detection of novel stimuli by the aged rats. Certainly, rats encounter fewer stimuli that could disrupt subsequent object recognition in the 2 min delay condition where they remain in the room in a covered pot. This possibility is discussed further in the discussion section.
In order to evaluate the hypothesis that the age effect on SOR task performance is not due to the old animals forgetting that a familiar object has been previously experienced, one would want to directly compare the time spent exploring objects between the familiarization and the test phases. If the aged rats were “forgetting” the stimulus, then their behavior would be similar between two sequential presentations of an object. Specifically, the old animals would behave as if a familiar object was novel and spend similar amounts of time exploring it during the first and the second presentation. If, however, the object was not forgotten, the aged rats should explore it less during the second exposure relative to when they first experienced the object. Such a direct comparison is problematic for the standard version of the SOR task, because in the familiarization phase the two objects are identical and both are novel, while 165 during the test phase the objects are different and only one is novel. As a result, the object familiarization phase and the test phase are distinguishable by both the relative familiarity of the objects and by the difference between the objects presented during the test phase. Moreover, the test phase of the standard SOR task requires a forced choice in exploratory behavior because when the rat explores the novel object it cannot explore the familiar object, and vice versa.
Therefore, a variant of the spontaneous object recognition task was implemented that controls for these confounds, and this task was administered to an additional group of young and aged rats (McTighe et al., 2008).
6.3.3 Experiment 2: Novel/repeated SOR task results
Object recognition was tested in Experiment 2 by presenting identical objects during both the familiarization and the test phases (Figure 6.2). Ten adult and 10 aged rats participated in this variant of the SOR task with 2 hr delays between test and familiarization phases. This delay duration was selected because aged rats reliably show a decrease in the discrimination ratio after a two hour delay and this enabled the object familiarization phase and the test phase to occur on the same day. For all 20 rats, two identical novel objects were presented during the first familiarization phase (Figure 6.2; left panel). For 5 of the adult and 5 of the aged rats, however, a different identical pair of objects was presented during the first test phase and this was the “novel” condition. For the other 5 adult and 5 aged rats, the same objects that were presented during the 166 familiarization phase were again presented to the animal in the test phase and this was the “repeat” condition. The following day, during familiarization phase 2
(Figure 6.2; left panel) a distinct pair of novel objects was presented to all of the rats during the familiarization phase but these objects were placed in a different location of the testing arena than the objects that had appeared the previous day.
During the test phase, the rats that had performed the repeat condition on the previous day were presented with novel objects during the test phase while the rats that had participated in the novel condition the day before, were given the same presentation of objects in both the familiarization and test phases (Figure
6.2; McTighe et al., 2008).
In Experiment 2, if the aged rats spent a similar amount of time exploring objects in both the familiarization and test phases of the “repeat” condition, the conclusion would be that the old animals forgot that the objects were previously experienced. This would be consistent with the idea that aged animals have recognition memory impairments. In contrast, if the aged rats show reduced object exploration between the familiarization and the test phase in the novel condition this could indicate that they incorrectly identify novel objects as familiar or that they habituate to object exploration and explore less during the second presentation of objects in the same location regardless of them being familiar or novel.
Statistical analysis of the data obtained in Experiment 2 revealed that both young and aged rats spent a similar amount of time exploring novel objects 167 during the two familiarization phases and there was not a significant effect of age group on exploration time (F[1,18] = 1.33, p = 0.26; ANOVA). Table 6.2 shows the
exploration times during the familiarization phases for the young and the aged
rats. Additionally, there was no significant difference in total exploration time
between the first object familiarization phase and the second familiarization
phase on the subsequent day in either age group (F[1,18] = 1.64, p = 0.22). This
could be due to the objects being in different locations between the first and the
second familiarization phases. The interaction effect between age group and
exploration time during the first and second object familiarization phases also did
not reach statistical significance (F[1,18] = 1.77, p = 0.20).
Table 6.2 Exploration times during the familiarization phase of Experiment 2 Novel Repeat condition condition
Young 23.5 sec 31.7 sec ± 2.6 ± 3.9 Aged 23.8 sec 22.2 sec ± 4.0 ± 2.5
Figure 6.6 shows the total mean time spent exploring objects by young
(green) and aged (purple) rats during both the familiarization and the test phases for the novel and the repeat conditions. Statistical analysis revealed that there was a significant main effect of experimental phase (object familiarization versus test) on total exploration time (F[1,36] = 69.45, p < 0.001), with more object
exploration occurring during the familiarization phase compared to the test 168
40
30
20
exploration time (sec) time exploration 10 adult aged 0 familiarizationnovel novel test familiarizationrepeat repeat test Figure 6.6: Novel/repeat SOR task performance (Experiment 2). The mean exploration time of the adult (green) and the aged rats (purple) measured during all four episodes of exploration. In the young rats (green), there was only a significant reduction in the total exploration time when the object presentation was the same between both the familiarization and the test phases (repeat condition). In contrast, the aged rats (purple) showed a reduction in total exploration time between the familiarization and test phases both when the object presentation was repeated and when it was novel. Error bars are +/-1 SEM. phases. There was also a significant interaction effect between experimental phase and condition (novel versus repeat) on exploration time (F[1,36] = 5.03, p <
0.05), and a significant three-way interaction effect of age group, experimental
phase, and condition (F[1,36] = 5.86, p < 0.05). Post hoc analysis revealed that
these interaction effects resulted from a significant decrease in exploration time
between the object familiarization and the test phase for the young rats during
the repeat condition but not during the novel condition (p < 0.02; Tukey HSD).
Specifically, the young group of animals spent similar amounts of time exploring
objects between the familiarization and test phases if the objects were novel in
both episodes (novel condition). If, however, the objects presented during the
familiarization phase were presented again during the test phase (repeat 169 condition), the young rats explored them less relative to their first presentation. In contrast, the aged rats always explored the objects less during the test phase relative to the familiarization phase regardless of whether the objects were novel or repeated during the test phase. These data support the idea that after long delays between familiarization and test phases, the aged rats may identify novel objects as familiar. These data do not support the idea that aged rats do not remember the previously experienced objects, which would be associated with an increase in the exploration of familiar objects. An alternative explanation for the results observed in Experiment 2 is that aged rats habituate to object exploration and will explore less during the test phase regardless of whether the objects are familiar or novel. To rule out this possibility, young and aged rats were tested on an object recognition task with incongruent contexts between the familiarization and test phases, which elicited increased exploration of both novel and familiar objects during the test phase after long delays.
6.3.4 Experiment 3: SOR task with context change
Figure 6.3 shows a schematic of the SOR task with context change.
Because it has been shown that objects become bound to the context that they were experienced in (e.g., Hayes et al., 2007; Piterkin et al., 2008), rats will express exploratory behavior as if a familiar object is novel when the object is presented in a different context from where it was first experienced (Eacott and
Gaffan, 2005; Eacott and Norman, 2004; Norman and Eacott, 2005). Therefore, 170 by measuring object recognition when the context is different between the familiarization and the test phase, we can assess whether aged rats show an overall habituation to object exploration or if they incorrectly indentify novel objects as being familiar. Specifically, if the context changes and the aged rats still show reduced exploration to both novel and familiar objects, then it is likely that old animals habituate to object exploration. In contrast, if the aged rats explore both novel and familiar objects more in a new context, then this would indicate that the reduced novel object exploration observed during Experiments 1 and 2 is because aged rats were impaired at identifying a new object as novel after long delays.
Unlike Experiments 1 and 2, there was a main effect of delay condition (2 min versus 24 hours) on total exploration time during the object familiarization phases (F[1,28] = 9.51, p < 0.01; repeated-measures ANOVA), with less
exploration of objects occurring during the familiarization phase of the 24 hr delay
relative to the 2 min delay. Additionally, there was a main effect of age on
exploration time during the familiarization phase (F[1,28] = 9.58, p < 0.01). Post hoc analysis revealed that the young rats spent significantly more time exploring objects during the familiarization phase of the 2 min delay relative to the familiarization phase of the 24 hr delay (p < 0.01; Tukey HSD). Additionally, the young rats spent more time exploring objects in the 2 min delay familiarization phase relative to the aged rats (p < 0.05; Tukey HSD). Importantly, the young and old rats explored the objects a similar amount of time during the object 171 familiarization phase that preceded the 24 hr delay (p = 0.85; Tukey HSD). The aged rats also spent a similar amount of time exploring the objects between the two familiarization phases (p = 0.67; Tukey HSD). Table 6.3 shows the total time in seconds that rats spent exploring the objects during the familiarization phase
(arena A) for the 2 min, and the 24 hr delays.
Table 6.3 Exploration times during the familiarization phase of Experiment 3 2 min 24 hr Young 30.1 sec 20.3 sec ± 2.2 ± 1.9 Aged 21.3 sec 18.1 sec ± 1.9 ± 1.8
Figure 6.7A shows the discrimination ratios measured during the test phase (arena B) for the young and the aged rats for both delay conditions. When the context was changed between the object familiarization and test phases, there was no effect of age on the magnitude of the discrimination ratio (F[1,28] =
0.06, p = 0.81). This is distinct from the standard SOR task used in Experiment 1,
which shows a reduction in the discrimination ratio of aged rats relative to young
rats at delays longer than 2 min. This suggests that the aged and the young rats
are affected similarly by incongruent contexts between the familiarization and test
phases of the SOR task. There was a main effect of delay duration (2 min versus
24 hr), such that the discrimination ratio during the test phase (in arena B) was
significantly larger for the 2 min compared to the 24 hr delay condition (F[1,28] =
11.08, p < 0.005; repeated-measures ANOVA). This occurred in both the young 172
A 0.6
0.4
0.2
0 adult discrimination ratio -0.2 aged 2 min 24 hr B delay 14 12 10 8 6 4
exploration time (sec) exploration time 2 0 familiar novel familiar novel
2 min 24 hr delay Figure 6.7: SOR task with context change task performance (Experiment 3). (A) The mean discrimination ratio of the young (green) and the aged rats (purple) measured during the test phase, in arena B, for the 2 min and 24 hour delay conditions. A higher discrimination ratio indicates that the animal spent more time exploring the novel object relative to the familiar object. Both the young and aged rats had a higher discrimination ratio after a 2 min delay compared to the 24 hour delay as indicated by the significant main effect of delay duration (F[1,28] = 11.08, p < 0.005; repeated-measures ANOVA). For the 24 hour delay, neither the young nor the aged rats distinguished between the novel and familiar object as indicated by the lack of a significant interaction effect between age group and delay condition (F[1,28] = 0.29, p = 0.60; repeated- measures ANOVA). (B) The mean amount of time young (green) and aged (purple) rats spent exploring the familiar and the novel object during the test phase for the 2 min and the 24 hour delay conditions. Unlike experiment 1, in which the context was congruent, when the context changed between the acquisition and the test phase, the adult and the aged rats spent significantly more time exploring the familiar objects for the 24 hr delay relative to the 2 min delay (p < 0.05; repeated contrasts). In contrast, the exploration time of the novel objects was not significantly different between delay conditions (p > 0.05 for all comparisons; Tukey HSD). Error bars represent +/-1 standard error of the mean. and the aged rats as the interaction effect between age group and delay condition was not significant (F[1,28] = 0.29, p = 0.60; repeated-measures
ANOVA). Moreover, the discrimination ratio following the 24 hr delay was not significantly different between the young and the aged rats (T[28] = 0.55, p = 0.59; 173 independent-samples T test). Importantly, when the discrimination ratio for the 24 hr delay is compared between the young rats that participated in the standard
SOR task and the young rats that participated the SOR task with context change, the discrimination ratio was significantly smaller during the context change experiment relative to when the context was the same between the object familiarization and the test phases (T[29] = 2.29, p < 0.03; independent-samples T test). These data suggest that objects become bound to the context they were experienced in but that this binding requires time (>2 min) to completely manifest
itself. The observation that an object becomes bound to the original context it was experienced in implies that when a familiar object is encountered in a different context it will be treated as if it were novel. This idea suggests that the
decrease in the discrimination ratio after the 24 hr delay in Experiment 3,
observed in both groups, should emerge from rats exploring the familiar object
more during the test phase. Specially, in the different context, both young and
aged rats could behave as if the familiar object is novel. To test this notion we
compared the exploration times of the novel versus the familiar objects during the
test phase of Experiment 3.
Figure 6.7B shows the mean time spent exploring the novel and the
familiar objects for the young (green), and aged rats (purple) for the 2 min and
the 24 hr delay conditions. The analysis of object exploration times during the
test phase in arena B showed that the reduction in the discrimination ratio at the
24 hr delay was associated with increased familiar object exploration in both the 174 young and the aged rats. Specifically, there was a significant main effect of novel versus familiar objects on exploration time (F[1,56] = 5.97, p < 0.02; repeated-
measures ANOVA). Additionally, the interaction effect of delay duration (2 min versus 24 hr), and object type (novel versus familiar) on exploration time during
the test phase was significant (F[1,56] = 4.07, p < 0.05), but there was no significant main effect of age on exploration time (F[1,56] = 3.54, p = 0.07). Planned contrasts
comparing the time spent exploring the familiar objects during the test phases for
2 min and the 24 hr delay conditions revealed that this interaction effect was due
to a significant increase in the exploration time of the familiar object during the 24
hr delay relative to the 2 min delay in both the young and the aged rats (p < 0.05;
repeated contrasts). The increased exploration of the familiar object after a 24 hr
delay in a different context in both young and old rats indicates that the reduced
novel object exploration observed in aged rats for Experiments 1 and 2 is not due
to the old animals showing an overall habituation to object exploration.
Finally, the amount of time spent exploring the novel objects during the
test phase was not significantly different between the 2 min and 24 hr delay
conditions (p > 0.05 for all comparisons; Tukey HSD). The aged rats, however,
explored the novel object significantly less compared to the young rats following
a 2 min delay (p < 0.05, Tukey HSD). This suggests that when the context
changes after a short delay aged rats show a reduction in novel object
exploration that is similar to what is observed at longer delays when the context
remains the same between familiarization and test phases. 175
6.3.5 Relationship between object recognition and spatial memory
Figure 6.4B shows the individual CIPL scores on Day 4 of spatial swim task testing for the young and the aged rats. Of the 18 aged rats that participated in Experiment 1, only 3 showed performance levels on the spatial version of the
Morris swim task within 1 standard deviation of the mean performance of the young rats. The performance of this group of spatially ‘unimpaired’ rats on the
SOR task, at all delay durations, was not significantly different from the rats that performed more than 1 standard deviation above the mean of the young animals on the spatial version of the Morris swim task (F[2,16] = 1.06, p = 0.37; ANOVA).
Additionally, when z-scores were calculated from the CIPL values on Day 4 of
spatial testing for aged and adult rats separately, no significant relationship was
found between spatial memory and object recognition in either age group for the
2 min (F[1,31] = 0.69, p = 0.41; linear regression), 15 min (F[1,31] = 0.55, p = 0.46;
linear regression), 2 hr (F[1,31] = 1.34, p = 0.26; linear regression), or the 24 hr
delays (F[1,31] = 0.25, p = 0.62; linear regression). Similarly, in Experiment 2,
three aged rats performed within 1 standard deviation of the young rats on the
Morris swim task. The performance of these rats on the simultaneous
novel/repeat SOR task was also not significantly different than the aged rats with
worse spatial memory performance (F[2,8] = 0.80, p = 0.49; ANOVA). This
supports the idea that the aging process affects different regions of the brain
independently. 176
In Experiment 3, only 1 aged rat performed within 1 standard deviation of the young animals on the spatial Morris swim task, which made a comparison between spatially impaired and spatially unimpaired animals untenable.
Although novelty discrimination across incongruent contexts requires the hippocampus under some circumstances (O'Brien et al., 2006), there was no significant relationship between object recognition in Experiment 3, in which the context between the familiarization and test phases was changed, and Morris swim task performance in either the young nor the aged rats at the 2 min delay
(F[2,27] = 1.29, p = 0.29; linear regression) or the 24 hr delay (F[2,27] = 0.15, p =
0.87; linear regression).
6.4 Discussion
6.4.1 Summary of salient results
The current investigation used the standard spontaneous object
recognition (SOR) task and two variants of this task in order to better
characterize object recognition in young and aged rats. When young and aged
rats participated in the standard SOR task, the data replicated previous findings
that aged rats show a delay-dependent deficit in object recognition (Figure 6.5A;
Vannucchi et al., 1997; de Lima et al., 2005; Pitsikas et al., 2005; Pieta Dias et
al., 2007; Pitsikas and Sakellaridis, 2007; Platano et al., 2008). Specifically, when
a 2 min delay between familiarization and test phases was used, aged rats
showed a novel object preference in their exploratory behavior that was similar to 177 the young animals. At longer delays, however, the aged rats showed a significant reduction in their exploratory preference for novel objects relative to the young animals. Traditionally this pattern of results has been interpreted as the aged rats having a deficit in recognition memory such that as the memory load increases, with longer delays, a behavioral impairment emerges. Because previous investigations of object recognition across the lifespan have only reported a measure that calculates the proportion of exploration time spent on the novel object relative to the total exploration time, such as the discrimination ratio (Dix and Aggleton, 1999), it has been unclear whether the decreased novelty preference results from old animals exploring the familiar objects more or from exploring the novel objects less. Importantly, either behavioral outcome would reduce the discrimination ratio, but they could result from different neurobiological etiologies. For example, increased exploration of the familiar object is consistent with a memory deficit suggesting that the animal does not recall experiencing the object during the familiarization phase. In contrast, decreased exploration of the novel object could indicate that the animal is not correctly discriminating the novel object from other stimuli that have been previously encountered. To examine these two distinct possibilities with the current data, the total exploration times of the novel and the familiar objects during the test phase of Experiment 1 were directly compared (Figure 6.5B). This analysis revealed that at longer delays the reduced discrimination ratio was due to the aged rats exploring the novel object less relative to the young rats, 178 indicating that the old animals behave as if novel objects are familiar at delays greater than 2 min. These results appear to be inconsistent with the idea that aged rats have recognition memory impairments. An alternative explanation that is consistent with the current results, however, is that during aging there is a reduced ability to discriminate novel objects from stimuli that have been previously encountered. This could result in aged rats behaving as if novel objects are familiar during the test phase. This leaves open the question of why the young and aged rats explore objects for similar amounts of time during the familiarization phase: It is possible that when objects are first encountered in a testing arena that had previously been empty, there are enough unique features in this experience that both young and old rats show similar amounts of object exploration, potentially because of overall increases in arousal. During the delay phase, however, the rat encounters stimuli that share common features with the novel object that will be presented during the test phase. The aged brain may be less able to disambiguate the common features that were observed during the delay from those of the novel object, and therefore, the novel object is identified as familiar. In young animals the stimuli that are encountered at longer delays do not appear to affect novelty discrimination during the test phase, presumably because they are better able to disambiguate different stimuli that share common features. In this view, aged rats show normal novelty discrimination during the 2 min delay condition because they remain in the testing room in a covered pot during the delay, which reduces the opportunity for intervening stimuli to affect 179 later object recognition. How age-related changes in the perirhinal cortex may contribute to this observed behavior will be discussed more below.
In order to evaluate the hypothesis that stimulus recognition impairments in aged rats are due to novel objects being identified as familiar rather than to memory deficits, a direct comparison between object exploration in the familiarization and the test phases is necessary. This direct comparison is difficult with the standard SOR task, because the repeated presentation of an object always occurs simultaneously with the presentation of a novel object. Moreover, during the test phase of the standard SOR task the rat must make a forced choice in exploratory behavior because when it is exploring the novel object it cannot explore the familiar object, and vice versa. Therefore, a variant of the spontaneous object recognition task was implemented that presents two identical objects during both phases of the SOR task. During one condition the same pair of identical objects is presented during both phases (“repeat condition”) separated by a 2 hr delay. In the other condition a different novel pair of objects is presented during familiarization and test phases (“novel condition”) with a 2 hr delay intervening between novel presentations (McTighe et al., 2008).
When this novel/repeat SOR task was used to test object recognition it was observed that total exploration time in the young rats was significantly reduced only when the object presentation was repeated during both the familiarization and test phases. When a novel pair of objects was used during both phases of the task, young rats explored the objects a similar amount of time 180 during the familiarization and test phases. In contrast, the aged rats always spent less time exploring objects during the test phase relative to the familiarization phase whether the objects were repeated or novel (Figure 6.6). Importantly, these data show a similar pattern to what is observed in young rats with perirhinal cortex lesions (McTighe et al., 2008).
An alternative explanation for the results obtained from Experiments 1 and
2 is that the aged rats can identify that the objects are novel, but they habituate to object exploration. Therefore, they do not express the same behavior as the young animals. There are several reasons to believe that is not the case. First, in the standard SOR task, both the young and aged rats show a similar amount of novel object exploration after a 2 min delay. Moreover, during the second familiarization phase of the novel/repeat SOR task, when the position of the novel objects in arena A was different than it had been during the previous familiarization and test phases, the aged and the young rats explored the novel objects during the second familiarization phase for similar amounts of time (Table
6.2). Finally, under conditions in which the context changed between the familiarization and the test phase (Experiment 3), after a 24 hr delay, both young and aged rats explored the familiar object more when it was in a different context from where it was first experienced (Figure 6.7B). This suggests that both young and aged rats behaved as if the familiar object was novel when it is experienced in a new context and indicates that the aged rats do not show an overall habituation to all object exploration during the test phase. 181
Taken together, the current data suggest that the primary age-related deficit in object recognition at longer delays arises from old rats behaving during the test phase as if the novel object is familiar rather than from forgetting.
Because reduced novel object exploration is not observed in aged rats after short delays and when objects are first encountered during the familiarization phase, this age-associated deficit could be due to the intervening stimuli encountered during long delay periods. Specifically, the 2 min delay of standard SOR task is the only condition in which the rats remained in the same room as the testing arena for the delay and were in a covered pot for the entire delay duration.
Therefore, in this condition the amount of extraneous stimuli that the rat encountered before the test phase was reduced relative to the other conditions.
This suggests that, for the aged rats only, the stimuli that are spontaneously encountered during longer delay periods could share common features with the objects that will be presented during the test phase and these common features disrupt the animal’s ability to subsequently detect a novel stimulus. This idea suggests that if an aged rat was exposed to more stimuli during the 2 min delay period, then novel objects would be identified as familiar even at this short delay.
In fact, this was observed. During Experiment 3, all of the rats were moved from the room containing arena A, after the familiarization phase, to a different room containing arena B for the test phase. In this experiment the aged rats showed less exploration of the novel object relative to the young rats. 182
It is also possible, however, that the observed age-related changes in object recognition arose because of vision impairments or differences in encoding during the object familiarization phase. The former is unlikely because all animals were screened for visual ability using the visually-cued Morris swim task, and rats that could not successfully find the visible platform were excluded from these experiments. Moreover, both age groups significantly benefitted from the platform being visible and showed shorter path lengths during visual trials relative to the hidden spatial trials. It is still possible, however, that the visually- cued Morris swim task was not sensitive enough to detect vision problems that could disrupt object recognition but leave swim task performance intact. This possibility is also unlikely since the aged rats showed novelty discrimination comparable to the adult rats at the 2 min delay of the SOR task (Experiment 1).
Additionally, it is untenable that the observed age effect on object recognition is due to encoding differences during the object familiarization phase.
During the standard SOR task (Experiment 1) adult and aged rats spent a similar amount of time exploring objects during the familiarization phase for the 2 min,
15 min, and 2 hr delays (Table 6.1). The aged rats did explore the objects less during the familiarization procedure for the 24 hr delay but since the age- associated reduction in novelty discrimination appeared at the 15 min and the 2 hr delays it is unlikely that changes in behavior during the familiarization phase could account fully for the current results. Moreover, there were no differences in exploratory behavior between age groups during the familiarization phases of 183
Experiment 2 (novel/repeat object recognition; Table 6.2), even though there was a significant effect of age on exploratory behavior during the test phase.
Importantly, the current data indicate that novelty detection in aged animals is similar to young animals under some circumstances. Specifically, the aged rats showed similar amounts of exploration relative to the younger animals when two novel objects were presented during the familiarization phases in all three experiments. One explanation for this observation in that during the object familiarization phase, novel objects appear in a familiar context (arena A), previously devoid of stimuli. Under these circumstances an overall increase in arousal coupled with an orienting response to new salient features added to a familiar environment could elicit exploratory behavior in the aged rats that is comparable to the younger animals. In this view, aged animals have an increased threshold for detecting novelty because novel items often share features with familiar stimuli, therefore, more novel features or elements need to occur before an aged animal overtly responds. Data from Experiment 3 also support this idea. Unlike in Experiments 1 and 2, where discrimination ratios decreased because the aged animals spent less time exploring the novel stimuli, in Experiment 3 when the context in which an object was explored is changed between the familiarization and test phases, the discrimination ratio decreased because the rats explored the familiar object more. In fact, both the young and the aged rats had exploration times of the novel and the familiar objects during 184 the test phase that were similar to the time spent exploring during the familiarization phase.
The current data question previous ideas that impairments in object recognition in aged animals result from a decline in “memory”. The finding that the aged rats demonstrate exploratory behavior during the test phase indicating that they identify the novel objects as familiar requires an alternate explanation for age-associated stimulus recognition deficits. Because rats that have perirhinal cortex lesions show the same behavior as the aged rats used in the current study
(McTighe et al., 2008), an age-related decrease in the ability of the perirhinal cortex to disambiguate stimuli that share common features is a possible explanation for this observation. Moreover, a decline in mesolimbic dopamine activity that occurs during aging could also help to account for the current observations.
6.4.2 The perceptual-mnemonic feature conjunctive (PMFC) model of the perirhinal cortex and object recognition during aging
Lesions studies in monkeys (Bussey et al., 2002) and rats (Bartko et al.,
2007; Bartko et al., 2007) show that the perirhinal cortex is necessary for visual discrimination when the items to be discriminated contain overlapping features.
These results lend empirical support to the view that the perirhinal cortex participates in perception (Bussey et al., 2002; Bussey et al., 2003; Murray and
Bussey, 1999) as well as in memory (for reviews, see Eichenbaum, 2000; Squire 185 et al., 2004; Eichenbaum et al., 2007). Mechanistically, this dual role of the perirhinal cortex can be explained by the perceptual-mnemonic feature conjunctive (PMFC) model, which proposes that the perirhinal cortex associates representations of stimulus features stored in lower sensory cortical areas that cannot store conjunctive visual features directly. Therefore, when a visual discrimination problem can be solved using a single feature (for example, red versus blue) the perirhinal cortex is not necessary. When the visual discrimination problem involves discriminating between stimuli with overlapping features (for example, red cube versus red cylinder), however, the perirhinal cortex is required for visual discrimination (Bussey and Saksida, 2002; Bussey et al., 2003; Bussey et al., 2005).
In the current experiment the stimuli used to test object recognition did not contain overlapping features, and the aged rats still did not discriminate between the novel and familiar objects. The PMFC model proposes that this behavioral deficit in the aged rats could be explained by the greater amount of interfering stimuli that occurs during longer delays. When the delay between the familiarization and the test phase is short, and the old animal is not moved from the testing environment, aged rats show normal stimulus recognition. According to the PMFC model, however, during a long delay between the familiarization and test phases, the rat encounters numerous visual stimuli containing simple features that are common to the objects presented during the SOR task. The specific conjunction of visual features that comprises a given complex object is 186 unique, however. Therefore, over the course of the delay, interference can disrupt the representation of a unique object if the perirhinal cortex is unavailable to associate the unique configuration of the stimulus elements. Thus, an animal with an intact perirhinal cortex has an advantage over an animal with damage to this region because it can associate the unique conjunctions of stimulus features making it easier to disambiguate between novel and familiar stimuli that share common elements. In contrast, an animal with a compromised perirhinal cortex must rely on the spared representations of individual stimulus features in lower visual cortical areas in order to discriminate between the novel and familiar stimuli. The object recognition impairment in animals with PRC damage is exacerbated by longer delay conditions because this allows for more interference of the common visual features represented in caudal visual cortex (Cowell et al.,
2006).
6.4.3 Mesolimbic dopamine and novelty detection
The perirhinal cortex receives a direct projection from the basal ganglia, including the ventral tegmental area (Furtak et al., 2007) and sends a reciprocal projection back to the ventral striatum (McIntyre et al., 1996). Moreover, in adult humans (Bunzeck and Duzel, 2006), and monkeys (Schultz, 1998) novel visual stimuli are associated with an increase in ventral tegmental activity. Therefore, a ventral tegmental to perirhinal cortex projection could act to modulate cell activity 187 in the perirhinal cortex in a way that facilitates the discrimination of novel objects from stimuli that have been previously encountered.
During aging, there are several changes in the limbic dopamine system
(for review, see Barili et al., 1998). Additionally, in aged humans reductions in mesolimbic hemodynamic responses to novelty correlate with volumetric reductions of the substantia nigra and ventral tegmental area (Bunzeck et al.,
2007). Finally, aged monkeys show a reduced novelty preference, relative to young monkeys, during the visual paired comparison task as measured by their eye movements to view novel pictures. This task simultaneously presents a novel and a familiar image to a monkey and then monitors their visual scan paths.
Because the age-associated decline in preferential viewing of novel images was correlated with altered saccade dynamics that are known to be related to dopamine dysfunction, it was suggested that the decreased novelty preference observed in old monkeys may be due to decreased dopamine levels (Insel et al.,
2008). Taken together, these data suggest that age-associated alterations in dopamine transmission could also contribute to the results observed in the current study.
6.4.4 Conclusions
Age-associated impairments in object recognition, as measured by the
SOR task, do not appear to result from the aged rats ‘forgetting’ the objects that were explored during the sample phase. Rather, the current data are consistent 188 with a view that functional alterations in the perirhinal cortex, which might occur during advanced age, result in aged rats behaving as if novel objects are familiar.
This could arise if aged rats were more vulnerable to the interfering effects of intervening stimuli encountered during the delay period. One explanation for this increased susceptibility to interference could be a diminished ability in advanced age to pattern separate between complex stimuli that share common features.
Therefore, similar to rats with PRC lesions (McTighe et al., 2008), after long delays aged rats may have difficulty disambiguating the novel object from other stimuli that were encountered during the delay. It is also possible that age-related reductions in mesolimbic dopamine could lead to a decreased ability for aged animals to detect when a stimulus is new. 189
CHAPTER 7: THE EFFECT OF AGE AND NOVELTY ON THE SINGLE-UNIT
FIRING CHARACTERISTICS OF PERIRHINAL CORTICAL NEURONS
7.1 Introduction
The work reviewed in the previous chapters and the behavioral data obtained from young and old rats, which were described in Chapter 6, indicate that the function of the perirhinal cortex (PRC) is likely to be affected by the process of normal aging. Additionally, it is unclear how neural networks within the
PRC support object recognition in young animals (see Chapter 4). The present chapter describes experiments aimed at investigating this question in young and aged rats while they perform behaviors that putatively modulate the activity of
PRC cells.
Although over twenty years have passed since the first report of electrophysiological recordings from the PRC of an awake behaving monkey
(Brown et al., 1987), it is still debated whether the primary function of this region of the temporal lobe is perceptual or mnemonic. Moreover, it remains unclear whether PRC networks support judgments of familiarity with an overall decrease
(Fahy et al., 1993) or increase (Holscher et al., 2003) in population activity. The series of experiments discussed in this chapter were designed to answer the following questions:
1. Do PRC neurons in rats show visually selective changes in their firing
rates? 190
2. Is there a response decrement or response increment in the activity of
PRC neurons that develops with increased experience with an object?
3. Is the activity of PRC neurons maintained across episodes of behavior
when short or long delays are interposed between episodes?
4. Do old rats show PRC population activity differences relative to young
animals that could possibly account for the object recognition deficits that
are observed behaviorally?
7.2 Methods
7.2.1 Subjects and behavioral training
Electrophysiological studies were conducted on six young (8-10 months old), and six aged (24-27 months old) Fisher-344 male rats. All rats participated in these experiments in pairs of one young rat and one aged rat. The rats were housed individually and maintained on a 12:12 light–dark cycle. Before rats were implanted with the hyperdrive recording device, they were screened for spatial memory impairments and normal vision using the Morris swim task (Morris,
1984). All animals were tested over 4 days with 6 spatial trials on each day.
Animals were then screened for visual ability with 2 days of cued visual trials (6 trials/day) in which the escape platform was above the surface of the water but the position of the platform changed between each trial. This procedure has been described in detail previously (e.g. Chapter 6; Barnes et al., 1996; Shen et al.,
1996). Rats’ performance on the swim task was analyzed offline with either in- 191 house software (WMAZE, M. Williams) or a commercial software application
(ANY-maze, Wood Dale, IL). Because different release locations and differences in swimming velocity produce variability in the latency to reach the escape platform, a corrected integrated path length (CIPL) was calculated to ensure comparability of the rats’ performance across different release locations
(Gallagher et al., 1993).
During electrophysiological recordings, the animals were food deprived to about 85% of their ad libitum weight and trained to run on a circular track (~335 cm in circumference) in both the counterclockwise and clockwise directions for food reinforcement. The food reward was a mixture of rat food pellets made soft by soaking them in water, applesauce, and the diet supplement Ensure. All electrophysiological recordings took place during the dark phase of the rats’ light–dark cycle. Food rewards were given in a small plastic food dish (4 cm x 4 cm) at two positions on the track. Both food dishes were located at the position on the track that marked the completion of one lap on opposite sides of a barrier, that is, where the rat was required to turn around and run in the opposite direction (Figure 7.1A). During all electrophysiological recording sessions, rats were required to run 20 laps (10 in the counterclockwise direction and 10 in the clockwise direction) during two distinct episodes of behavior. Each track running epoch was flanked by a rest period in which the rat was placed in a towel-lined pot located in a position that was central to the circumference of the track. Thus, the activity of PRC neurons was monitored during an initial rest session (before 192 behavior), during the first epoch of track running, during a second rest session after epoch 1 that was either 20 min or 2 hours long, during a second epoch of track running, and then finally during a third rest period. Data from the rest periods were used to assess firing stability across the entire recording session.
ABEpoch 1 Epoch 1
X barrier 20 min or 2 hr rest 20 minute rest Epoch 2 Epoch 2 food dishes 106.7 cm
Objects- both epochs Configuration condition change condition Figure 7.1: Behavioral procedures used for electrophysiological recordings. The track used for behavior during all electrophysiological recordings. Rats were required to run 20 laps bi-directionally (10 counterclockwise, 10 clockwise) for a food reward. (A) Rewards were given in two food dishes located on opposite sides of a barrier, at the position where the rat was required to turn around. The “X” indicates the location of the pot that the rat was placed in during rest episodes. (B) Examples of behavioral procedures where objects were placed on the track. In the “objects-both epochs” condition 8 novel objects were placed at discrete locations around the track for the first epoch of behavior (top panel), and the rat had to run past the objects to obtain the food reward. During the second epoch of behavior (bottom panel), 6 of the 8 objects used in epoch 1 were placed on the track at the same location as in epoch 1, while 2 of the 8 objects were removed and substituted with 2 novel objects. The grey squares indicate the two novel objects that replaced the objects from the first epoch of track running. The “configuration change” condition occurred on the day following the objects-both epochs condition. For this behavioral procedure, the same objects used in epoch 1 from the previous day were again placed on the track at the same location as in epoch 1, and the rat ran 20 laps (top panel). During the second epoch of track-running, the positions of the eight familiar objects were pseudo-randomly shuffled such that no object was at the same position between epoch 1 and 2 (bottom panel).
193
All twelve rats participated in the same behavioral procedures. During the first procedure (day 1) rats ran on an empty track (no objects; Figure 7.1A) for both epochs of behavior (no objects condition). A 20 minute rest period occurred between epoch 1 and epoch 2. For the second procedure (day 2), during epoch
1, eight novel objects that varied in size, color and texture were placed at eight different locations along the track. The side of the track that the objects were placed on alternated between the left and the right side and the rat had to run past these objects in order to obtain the food reward. Following either a 20 min or a 2 hr rest period, during epoch 2, six of the same objects used in epoch 1 remained in the same location on the track and two objects that were on the track during epoch 1 were replaced with two novel objects (objects-both epochs condition; Figure 7.1B). During the third procedure (day 3), the same eight objects that were placed around the track during epoch 2 from the previous day of recording were again placed on the track in the same positions as in day 2 for the first epoch of behavior. The rats ran 20 laps around these familiar objects, and then rested for 20 minutes. Following the 20 minute rest period, the rats ran another 20 laps. During this epoch of behavior, however, the positions of the eight familiar objects were pseudo-randomly shuffled such that no object was located in the same position between epochs 1 and 2 (configuration change condition; Figure 7.1B). Each rat completed all of these behavioral procedures on three consecutive days, and this process was repeated a minimum of 2 times 194 and a maximum of 6 times. For one aged rat only, no single-unit data were collected for the objects-both epochs condition with a 2 hr delay.
For all conditions in which objects were placed on the track, the objects were fixed in place using Velcro, thus, rats could actively explore, rear, and climb on the objects without displacing them. Additionally, during all rest periods the objects were removed from the track so that the rat could not see them during the intervening delay period.
7.2.2 Surgical Procedures
Surgery was conducted according to National Institutes of Health guidelines for rodents and protocols approved by the University of Arizona
Institutional Animal Care and Use Committee. Prior to surgery, the rats were administered penicillin G (30,000 units intramuscularly in each hind limb) to combat infection. The rats were implanted with a “hyperdrive” manipulator device that held an array of 14 separately moveable tetrode recording probes. During surgical implantation the rats were maintained under anesthesia with isoflurane administered at doses ranging from 0.5% to 2.5%. The hyperdrive recording device, implantation methods, and the parallel recording methods have been described in detail elsewhere (Gothard et al., 1996). Briefly, each hyperdrive consisted of 14 drive screws coupled by a nut to a guide cannula. Twelve of these cannulae contained tetrodes (McNaughton et al., 1983; Recce and
O'Keefe, 1989), four-channel electrodes constructed by twisting together four 195 strands of insulated 13 µm nichrome wire (H. P. Reid, Inc., Neptune, NJ). Two additional tetrodes with their individual wires shorted together served as an indifferent reference and an electroencephalogram (EEG) recording probe. A full turn of the screw advanced the tetrode 318 µm and all tetrodes were lowered between 4.0 and 6.0 mm ventral to the surface of brain. In all rats, recordings were made from the middle to caudal PRC region (between 4.0 and 6.5 posterior, 6.0 lateral to bregma, and angled 14° towards the midline). Following experimental procedures, 20 µA of DC current was administered to each tetrode and tetrode location was verified histologically. Only the units recorded from tetrodes in the PRC were used in the current analyses and neurons recorded from other brain regions (e.g. ventral CA1 or area TE of the inferotemporal cortex) were excluded.
The majority of tetrodes were located in area 36 of the perirhinal cortex, but in two young and two aged rats four tetrodes reached dorsal area 35.
Moreover, in the young rats neurons were recorded from both layer V and layers
II/III, but for the aged rats only one rat had tetrodes in layers II/III, and all of the other single-unit recordings were from neurons in layer V. Figure 7.2A shows
Nissl-stained coronal sections from three different rat brains with representative tetrode recording tracks and lesions, and Figure 7.2B shows the extent of perirhinal cortex that neurons were recoded from.
The implant was cemented in place with dental acrylic anchored by small screws. Immediately after surgery, all tetrodes were lowered approximately 1 mm 196 into the cortex, and rats were orally administered 26 mg of acetaminophen
(Children’s Tylenol Elixir, McNeil, PA) for analgesia. Oral administration of acetaminophen was continued for 3-5 days after surgery. Additionally, all rats were given either 25 mg of ampicillin (Bicillin, Wyeth Laboratories, Madison, NJ) or a combination of 20 mg of sulfamethoxale and 0.4 mg trimethoprin (Hi-Tech
Pharmacal Co., Inc, Amityville, NY) on a 10 days on/10 days off regimen for the duration of the experiment.
Caudal A young rat young rat aged rat B Rostral
Figure 7.2: Location of perirhinal cortical recordings. (A) Coronal Nissl stained sections of two young rat brains (left and middle panel), and one aged rat brain (right panel) showing representative tetrode tracks (black lines) and lesions (red circles) for perirhinal cortical recordings. In the young rats the tetrodes recorded neurons in layer V (left panel) and layers II/III (right panel). In the aged rats, most perirhinal cortical neurons were recorded from layer V. The red arrows indicate the location of the rhinal sulcus. (B) An unfolded flat map of the perirhinal and postrhinal cortices. The red oval delineates the location of perirhinal cortical recordings. In both young and aged rats most neurons were recorded from area 36 dorsal, ventral and posterior. Flat map image is from Burwell, 2001.
7.2.3 Neurophysiology
After surgery, tetrodes were lowered into the PRC over several weeks.
The neutral reference electrode was advanced with other tetrodes and when an area of cortex was reached that did not record any unit activity, it was not moved again. The four channels of each tetrode were attached to a 50-channel unity- 197 gain head stage (Neuralynx, Inc., Tucson, AZ). A multi-wire cable connected the head stage to digitally programmable amplifiers (Neuralynx, Inc.). The spike signals were amplified by a factor of 1,000 – 5,000, band pass-filtered between
600 Hz and 6 kHz, and transmitted to the Cheetah Data Acquisition system
(Neuralynx, Inc.). Signals were digitized at 32 kHz, and events that reached a predetermined threshold were recorded for a duration of 1 ms. Spikes were sorted offline on the basis of the amplitude and principal components from the four tetrode channels by means of a semiautomatic clustering algorithm
(KlustaKwik, author: K. D. Harris, Rutgers–Newark). The resulting classification was corrected and refined manually with custom-written software (MClust, author: A. D. Redish, University of Minnesota; updated by S. L. Cowen and D. R.
Euston, University of Arizona), resulting in a spike-train time series for each of the well-isolated cells. No attempt was made to match cells from one daily session to the next. Therefore, the numbers of recorded cells reported does not take into account possible recordings from the same cells on consecutive days.
Putative principal neurons in the deep and superficial layers of the PRC were identified by means of their waveform characteristics and autocorrelogram features (Bartho et al., 2004). Specifically, neocortical principal cells tend to have auto-correlograms with peaks at 3-6 ms followed by an exponential decay, which is indicative of “bursting” cells, or an auto-correlogram with an exponential rise from 1 to tens of milliseconds. These cells are considered regular-spiking 198 neurons. In contrast, the autocorrelograms of putative interneurons are not as fast decaying or slow rising as those of pyramidal neurons (Bartho et al., 2004).
Continuous recording was also taken from each tetrode. Several diodes were mounted on the head stage to allow position tracking. The position of the diode array was detected by a TV camera placed directly above the experimental apparatus and recorded with a sampling frequency of 60 Hz. The sampling resolution was such that a pixel was approximately 0.3 cm.
7.2.4 Analyses and Statistics
Spike activity diagrams were constructed by plotting the circular trajectories of the animals on a linearized, one-dimensional scale, using a linear interpolation (Maurer et al., 2005). For each cell, the maximum firing rate, mean firing rate, and spatial information content (spikes/bit) was calculated. Maximum and mean firing rate were obtained after normalizing spike activity by the occupancy of the animal, and information content was calculated from 161 bins of ~4.1 cm with the formula:
ΣPi(Ri/R)log2(Ri/R)
Where Pi was the probability of occupancy for a bin, Ri was the firing rate of the
bin, and R was the mean firing rate of cell (Skaggs et al., 1993). Information
content was used to examine the activity correlates of PRC neurons. Specifically,
it was observed that many PRC neurons increased there firing rates at the
location of objects (see Results). These patterns of activity were termed ‘object 199 fields’, and a PRC neuron was considered to have an object field if it had a spatial information score greater than 0.5 bits/spike, and the occupancy normalized mean firing rate within a bin exceeded the mean firing rate for at least
4 consecutive bins. Finally, if the 4 consecutive bins of higher firing overlapped with the area of the track that contained an object, the neuron was considered to have an ‘object field’.
In order to examine any possible age-related differences the mean for all cells was calculated for every rat. This ensured that rats in which more neurons were recorded from did not contribute more to the analysis, and that there would not be a bias towards finding an age difference due to the high statistical power obtained by the large number of cells. Finally, because data were always obtained from pairs of young/aged rats, repeated-measures or paired statistical analyses were used for significance testing and alpha was set to the 0.05 level.
7.3 Results
7.3.1 Morris swim task performance
The mean performance of the aged rats on the spatial version of the
Morris swim task was significantly worse than the performance of the young rats.
Every aged rat used for the current experiment was able to learn the visual version of the task, however. Figure 7.3 shows the mean CIPL (Gallagher et al.,
1993) for the six aged rats (purple circles), and the six young rats (green circles) that participated in these behavioral experiments. As reported previously (e.g., 200
40 Young (spatial) Aged (spatial) Young (visual) 30 Aged (visual)
20 CIPL (m)
10
0 Day 1Day 2Day 3Day 4 Figure 7.3: Performance on the Morris swim task. The results from the Morris Swim task for the young (green) and the aged (purple) rats. The X-axis is the day of testing and the Y-axis is the mean corrected integrated path length (CIPL) score. Higher CIPL scores indicate longer path lengths to reach the escape platform. All rats completed 4 days of spatial trials (circles) in which the platform was hidden below the surface of the water. During the spatial trials, the aged rats had significantly longer CIPL scores compared with the adult rats (F[1,40] = 6.12, p < 0.02; ANOVA). The spatial trials were followed by 2 days of visually-cued trials (triangles) in which the platform was visible but its located changed after every trial. During the visual trials the CIPL scores were not significantly different between the young and aged rats (F[1,10] = 0.07, p = 0.79; repeated-measures ANOVA). Error bars represent +/-1 standard error of the mean (SEM).
Barnes et al., 1997), the old rats had a significantly longer CIPL scores compared to the young rats (F[1,40] = 6.12, p < 0.02; ANOVA). Post hoc analysis revealed that this difference was due to the aged rats having significantly longer CIPL scores on days 3 and 4 of spatial testing (p < 0.05; for all comparisons; Tukey
HSD) while there was no significant difference in the path lengths between young and aged rats on day 1 or day 2 spatial testing (p > 0.2 for both comparisons;
Tukey HSD).
To test whether the spatial impairments of the aged rats were due to
visual problems, and to ensure that all rats had adequate vision to participate in 201 the electrophysiological recordings, the rats also performed 12 trials, over 2 consecutive days, of the visually-cued version of the Morris swim task. For these trials the platform was raised above the surface of the water but its location was changed between every trial. Figure 7.3 shows the CIPL for the visual trials of the
Morris swim task for the trials on day 1 and day 2 for the aged (purple triangles) and the young rats (green triangles). There was no significant effect of age on the CIPL values during either day 1 or day 2 of the visual swim task testing (F[1,10]
= 0.07, p = 0.79; repeated-measures ANOVA). Therefore, it is unlikely that any of
the aged animals used in the current series of experiments had significant visual
impairment.
7.3.2 Running velocity in young and aged rats
Rats have a natural tendency to explore objects, or other stimuli, that are
novel (Chapter 6; Ennaceur and Delacour, 1988). In the current task this
naturally occurring behavior could lead to slower overall velocities in behavioral
conditions with novel objects, as the rats’ running would be interrupted by periods
of object exploration. Additionally, young rats tend to have faster running
velocities relative to aged rats (e.g., Shen et al., 1997). Therefore, the running
velocity during epochs 1 and 2 for the different behavioral conditions was
compared across age groups. Figure 7.4 shows the mean running velocities for
both age groups during the different behavioral conditions of epoch 1 and epoch
2. As expected, the aged rats ran around the track significantly more slowly 202
compared to the young animals (F[1,73] = 69.5, p < 0.001, ANOVA). Overall, the rats had significantly faster running velocities during epoch 2 relative to epoch 1
(F[1,73] = 4.5, p < 0.04; ANOVA), and there was also a significant effect of behavioral condition on the running speed of the rats (F[3,73] = 4.6, p < 0.01;
ANOVA). Post hoc analysis indicated that this was due to both young and aged rats running faster when there were no objects on the track relative to conditions with objects (p < 0.05 for all comparisons; Tukey HSD). Finally, although there were no significant interaction effects between any of the factors of age, epoch or behavioral condition (p > 0.23 for all comparisons), the aged rats did not run
45
40
35
30
25
20 No objects
15 Objects-both epochs Velocity (cm/sec) Velocity (20 min delay) 10 Objects-both epochs (2 hr delay) Configuration 5 change
0 Young rats; epoch 1 Young rats; epoch 2 Aged rats; epoch 1 Aged rats; epoch 2 Figure 7.4: Mean running velocity of young and aged rats during the different behavioral conditions. The aged rats ran significantly more slowly than did the young animals (F[1,73] = 69.5, p < 0.001, ANOVA). Additionally, all rats ran faster during epoch 2 compared to epoch 1 (F[1,73] = 4.5, p < 0.04; ANOVA). Finally, behavioral condition also significantly affected the running speed (F[3,73] = 4.6, p < 0.01; ANOVA), and both young and old rats ran significantly faster during the no objects condition (dark grey bars) compared to the objects-both epochs with a 20 min delay (blue bars), the objects-both epochs with a 2 hr delay (yellow bars), and the configuration change (red bars) conditions (p < 0.05 for all comparisons, Tukey HSD). Error bars represent +/- 1 SEM.
203 faster during epoch 2 of the no objects condition relative to epoch 1 (26 cm/sec versus 25 cm/sec).
In order to examine further if novel objects on the track lead to slower
velocities, the running speed of the old and young rats during the first 2 laps and
the last 2 laps were compared for the different behavioral conditions. If novel
objects on the track lead to increased exploration, then the rats should have slower velocities during the first several laps of epoch 1, when the objects were novel, relative to the last laps in which the objects became more familiar.
Moreover, if novel objects on the track are a significant factor contributing to slower running velocities, then the difference in the rats’ running speeds between the early laps and the late laps should be less during epoch 2 and conditions that did not have novel objects on the track (configuration change, no objects). Figure
7.5 shows the mean running velocity during laps 1-2 and laps 19-20 for the no objects (dark grey), the objects-both epochs with a 20 min delay (blue), the objects-both epochs with a 2 hr delay (yellow), and the configuration change
(red) conditions for (A) young rats, epoch 1, (B) young rats, epoch 2, (C) aged rats, epoch 1, and (D) aged rats epoch 2. Because rats ran laps in both the counterclockwise and clockwise directions, lap 1 is the first lap in the counterclockwise direction and lap 2 is the first lap in the clockwise direction.
Similarly, lap 19 is the final lap that the rats ran in the counterclockwise direction
and lap 20 was the last lap run in the clockwise direction within an epoch of track
running. 204
50 A Young Rats - Epoch 1 50 B Young Rats - Epoch 2
40 40
30 30
no objects 20 20 Velocity (cm/sec) Velocity (cm/sec) Velocity (cm/sec) (cm/sec) Velocity Objects-both epochs (20 min delay) 10 10 Objects-both epochs (2 hr delay) configuration change 0 0 laps 1-2 laps 19-20 laps 1-2 laps 19-20 50 C Aged Rats - Epoch 1 50 D Aged Rats - Epoch 2
40 40
30 30
20 20 Velocity (cm/sec) (cm/sec) Velocity Velocity (cm/sec) Velocity
10 10
0 0 laps 1-2 laps 19-20 laps 1-2 laps 19-20 Figure 7.5: Mean running velocity for laps 1-2 versus laps 19-20. The running velocity for the first two laps and the last two laps during the no objects (blue), objects-both epochs with a 20 min delay (blue), objects-both epochs with a 2 hr delay (yellow), and the configuration change (red) conditions for A) young rats, epoch 1, (B) young rats epoch 2, (C) aged rats, epoch 1, and (D) aged rats, epoch 2. For laps 1 and 19 the rats were running counterclockwise and for laps 2 and 20 the rats were running clockwise. Errors bars represent +/- 1 SEM.
When the mean running speed of the rats for the first 2 laps was compared to the velocity for laps 19-20, statistical analysis revealed that overall, both the young and aged rats ran significantly more slowly during laps 1-2 relative to laps 19-20 (F[1,168] = 201.3, p < 0.001; repeated-measures ANOVA).
There was also a significant effect of epoch (F[1,168] = 17.0, p < 0.001; repeated- measures ANOVA) and behavioral condition (F[3,168] = 8.0, p < 0.001; repeated- measures ANOVA) on the difference in running velocity between laps 1-2 and 205 laps 19-20. Post hoc analysis indicated that the difference in running speed between the first laps and the last laps was greater for epoch 1 compared to epoch 2 (p < 0.001, Tukey HSD). This effect was due to the significantly slower running velocities measured in both young and aged rats for the first laps of epoch 1 relative to the velocities of the first laps of epoch 2 (F[1,38] = 69.28, p <
0.001; repeated-measures ANOVA). There was also a significant interaction effect between behavioral condition and the difference in velocity for the first laps of epoch 1 versus the first laps of epoch 2 (F[3,38] = 5.04, p < 0.01, repeated-
measures ANOVA). Post hoc analysis indicated that this interaction effect was
due to the lower velocity during the early laps of epoch 1 compared to epoch 2
for all behavioral conditions (p < 0.01; for all comparisons) except the
configuration change condition (p = 0.16). In contrast to laps 1and 2, during laps
19-20 there was no significant difference in the rats’ running velocities between epoch 1 and epoch 2 (T[182] = 0.5, p = 0.7; independent samples T test).
Moreover, in both the young and the aged rats, running velocities during laps 1-2
of epoch 1 were significantly greater for conditions that did not contain novel
objects (no objects and configuration change) compared to conditions where the
objects on the track were novel during epoch 1 (objects-both epoch with a 20 min
delays or a 2 hr delay; p < 0.05 for all comparisons; Tukey HSD). In the young
rats, behavioral condition did not significantly affect running velocities for laps 19-
20 during epoch 1 (p > 0.05 for all comparisons; Tukey HSD), which indicates that after the objects became familiar the young rats had similar running speeds 206 during all behavioral conditions. In contrast, the aged rats had faster running velocities during laps 19-20 for the no objects condition relative to the other conditions (p > 0.05 for all comparisons; Tukey HSD). Taken together these data suggest that running velocities are slower, in both young and aged rats, when novel objects are on the track relative to conditions when the objects are familiar or when the track is empty. In contrast, the running velocity of both young and aged rats did not decrease when familiar objects were moved into novel locations. These data are consistent with the idea that rats in both age groups spent more time exploring objects when they were the most novel during laps 1-2 of epoch 1 for the objects-both epoch conditions with a 20 min or 2 hr delay.
The observation that neither the young nor the aged rats showed a decrease in running velocity when familiar objects were moved to novel locations during the configuration change condition may appear inconsistent with previous studies. Specifically, it has been shown that when a rat is placed into an open arena with two identical objects (one object is in a familiar location and the other object has been moved to a novel location), the rat will spend more time exploring the object that is in a novel location (Ennaceur and Delacour, 1988;
Cavoy and Delacour, 1993). The conditions in the current experiment, however, were quite different from the freely-exploring, non-food restricted conditions of the experiments that have shown an effect of novel locations on exploration. That is, the rats were food restricted and running on a confined path to obtain a food reward. Under these conditions, the reconfiguration of familiar objects was not 207 enough of a salient change to provoke additional object exploration, would delay obtaining the food reward. Interestingly, these data suggest that when novel objects are on the track, both young and aged rats will postpone obtaining the food reward to explore a stimulus that has not previously been encountered. This behavior does not occur when familiar objects are placed in novel locations.
7.3.3 Behavioral correlates of perirhinal cortical activity
The activity of 2,405 PRC principal neurons was recorded in this experiment. A summary of the database from which the current results were derived is given in Table 7.1. The parentheses indicate the layer of the PRC that the cells were recorded from. The numbers of PRC neurons that were recorded during these experiments were not significantly different between the young and the aged rats (F[1,19] = 0.03, p = 0.87). Additionally, all single-unit neurons that
Table 7.1: Numbers of PRC neurons recorded from young and aged rats.
Pairs of Young/Aged rats
Young rat 20 min 2 hr delay Order No objects Aged rat 20 min 2 hr delay Order No objects delay change delay change 8412 8 11 17 7 8413 34 45 48 62 (V) (V) (V) (V) (V) (V) (V) (V) 8509 14 12 8 7 8507 17 13 20 (V) (V) (V) (V) (V) (V) (V) 8583 94 83 84 93 8615 56 62 101 99 (II/III) (II/III) (II/III) (II/III) (V) (V) (V) (V) 8661 33 & 12 25 & 10 52 & 19 34 & 14 8662 41 59 84 99 (II/III & V) (II/III & V) (II/III & V) (II/III & V) (V) (V) (V) (V) 8670 48 & 58 21 & 27 72 & 90 60 & 68 8696 44 27 77 39 (II/III & V) (II/III & V) (II/III & V) (II/III & V) (V) (V) (V) (V) 8883 30 12 35 35 8865 4 & 44 2 & 35 9 & 53 7 & 31 (V) (V) (V) (V) (II/III & V) (II/III & V) (II/III & V) (II/III & V) total 297 201 377 318 240 230 385 357 208 were included in the current analyses showed stability during an entire recording session, and there was no significant difference in the mean firing rate of neurons during rest 1 and rest 3 (F[1,10] = 1.01, p = 0.34; repeated-measures ANOVA).
Moreover, the stability of the mean firing rates between the beginning of a
recording session and the end of a recording session was not significantly
different between young and aged rats (F[1,10] = 0.02, p = 0.67; repeated- measures ANOVA). Finally, the mean firing rate during sleep episodes was 1.3
Hz for the young rats and 1.1 Hz for the aged rats, and these firing rates where not statistically different from each other (F[1,10] = 2.26, p = 0.16). The majority of
neurons were recorded from layer V of area 36 of the PRC. In three young rats,
699 neurons were recorded from layers II/III, while in only one aged rat 22
neurons where recorded from this region (Table 7.1). Table 7.2 shows the firing
rates by condition for all rats. The firing rates for layers II/III and layer V are
shown separately. In the three rats where neurons from the different cortical
laminae were recorded simultaneously, there was not a significant difference in
the firing rates between layers II/III and layer V (T[23] = 1.23, p = 0.23, paired-
samples T test). Therefore, for analyses involving firing rate comparisons
between age groups neurons from the different layers were analyzed together.
Additionally, for other age comparisons, firing patterns between the deep and superficial layers for the 3 rats (2 young, 1 old) in which neurons were recorded from both layers II/III and V simultaneously were first evaluated before additional 209
Table 7.2: Firing rates (Hz) in different cortical laminae
Pairs of Young/Aged rats Firing Rate (Hz)
Young rat 20 min 2 hr delay Order No objects Aged rat 20 min 2 hr delay Order No objects delay change delay change 8412 2.7 0.9 1.7 0.8 8413 1.2 0.9 1.4 1.0 (V) (V) (V) (V) (V) (V) (V) (V) 8509 1.7 1.6 1.6 0.8 8507 1.1 0.7 1.0 (V) (V) (V) (V) (V) (V) (V) 8583 1.21 1.5 1.2 1.7 8615 1.0 0.9 1.1 1.2 (II/III) (II/III) (II/III) (II/III) (V) (V) (V) (V) 8661 3.0 & 3.3 2.2 & 1.2 4.1 & 2.4 2.8 & 1.3 8662 1.3 1.6 1.2 1.2 (II/III & V) (II/III & V) (II/III & V) (II/III & V) (V) (V) (V) (V) 8670 1.6 & 1.6 1.5 & 2.1 1.3 & 1.9 2.0 & 2.4 8696 1.8 0.8 1.3 1.3 (II/III & V) (II/III & V) (II/III & V) (II/III & V) (V) (V) (V) (V) 8883 1.1 0.7 1.9 1.6 8865 3.4 & 0.8 0.7 & 0.8 0.8 & 2.1 3.3 & 1.1 (V) (V) (V) (V) (II/III & V) (II/III & V) (II/III & V) (II/III & V) statistical comparisons were conducted. This ensured the validity of collapsing the data across layers.
When young and old rats traversed the track with objects on it, many PRC neurons in both layers II/III and layer V showed a selective increase in their firing rates at the locations of objects. This occurred when objects were novel (during epoch 1 of the objects, both epochs conditions), as well as when objects were familiar (during epoch 2 of the objects, both epochs conditions, and the configuration change condition) for both age groups. This increase in spiking activity in the vicinity of objects will be referred to as “object fields”. Moreover, during the no objects condition many PRC neurons in both young and aged rats had high firing rates at the food dishes relative to other areas of the track. This
PRC neuron firing pattern will be referred to as “food dish fields”. The quantification of object and food dish field activity will be described in more detail below. Figure 7.6 shows a representative example of the activity of three PRC 210
A 6 7 8
1
5 2
3 B 4 9 Hz 6 Hz 2 Hz
0.5 Hz 1 Hz 1 Hz
Figure 7.6: Behavioral correlates of PRC neuron activity. (A) The activity of three representative PRC neurons under conditions with objects on the track (left panel) or without objects (middle and right panels). The black trace indicates the path of the rats and the red spots indicate the locations of spikes. In the left panel, the blue numbers represent the locations of objects. Notice that in this example the neuron fires at locations 6 and 8. In the no objects conditions many PRC cells increase their firing rates near the food dish locations (middle panel), or selectively show activity at the food dish (right panel). (B) The occupancy-normalized firing rate maps of the cells shown in A. neurons recorded from a young rat when there were objects on the track (left panel), and when the rat traversed the track in a behavioral condition without objects (middle and right panel). Because the object positions alternated between the left and the right side of the track, the rats had to modify their trajectory in order to run around the objects. This also ensured that the rats had to briefly view the objects in order to avoid colliding with them.
In order to quantify the number of PRC neurons that increased their firing rate at locations containing objects (i.e., the proportion of PRC neurons expressing ‘object fields’) the spatial information score (Skaggs et al., 1993) was calculated for each neuron, and then neurons with an information score above 211
0.5 bits/spikes were considered to have an object field if their occupancy normalized firing rate histogram exceeded the mean firing rate for at least 4 consecutive 4.1 cm bins, and this area of higher firing rate overlapped with an area of the track that contained an object. The criterion of 4 consecutive bins applied an arbitrary lower limit that an object field could not be smaller than 16.4 cm. When fewer than 4 bins was used to define an object field, however, several
PRC neurons reached this threshold that did not show consistent spiking at an object over multiple laps. Using the criteria described above, the mean size of object fields was 25.6 cm in the young rats and 26.9 cm in the aged rats, which was not significantly different (p = 0.32). Because the size of object fields was similar between age groups, it does not appear that the method of determining object fields was biased towards detecting them in one age group and not the other.
The area of the track within 7 bins of a food dish (28.7 cm) was excluded for the calculation of information score, because this activity could presumably be related to the food dish and/or reward rather than to the objects, and the activity near the food dish was analyzed separately. The regions within 28.7 cm of a food dish were selected for exclusion because when the rats were within this area of the track their running speed was either zero or it was changing rapidly as the rat was stopping to obtain reward or accelerating after eating. Figure 7.7 shows representative examples of the velocity by position plots for a young rat (Figure
7.7A), and an aged rat (Figure 7.7B) during the second epoch of the objects-both 212
A B 80 80
70 70
60 60
50 50
40 40
30 30 Velocity (cm/sec)Velocity Velocity (cm/sec) Velocity 20 20
10 10
0 0 0 50 100 150 200 250 300 0 50 100 150 200 250 300 Location (cm) Location (cm) Figure 7.7: Velocity by position. A representative example of a (A) young rat’s, and an (B) aged rat’s velocity by position. The black lines indicate the path of the rat for the area of the track that was included in analyses of object-related activity. The red lines indicate the path of the rat when it was within seven 4.1 cm bins (28.7 cm) of a food dish. This area of the track was excluded from analyses of object-related activity because in this region large change in velocity indicated that the rat was slowing down, or stopped in order to obtain the food reward. Moreover, for the behavioral conditions with objects this area of the track contained both an object and the food dish. Therefore, it was not possible to determine which variable was contributing to the increase in a PRC neuron’s firing rate. epochs condition with a 20 min delay. The black lines indicate the path of the rat on the area of the track that was included in the object-related activity analyses, and the red lines indicate the area that was excluded.
When the spatial information score was calculated for all cells that showed activity on the maze (a maximum firing rate greater than 2 spikes/bin occupancy), there was not a significant difference in mean information score between epochs
1 and 2 (T[45] = 0.10, p = 0.94; paired-samples T test). Moreover, for the three
rats (two young, and one aged) that had tetrodes in both the deep and the
superficial layers of the PRC, the information score for neurons in layer V (0.61
bits/spike) versus neurons in layers II/III (0.53 bits/spike) was not significantly
different (T[5] = 1.51, p = 0.19; paired-samples T test). Therefore, the data
obtained from the different epochs of track running, and from the different cortical 213 laminae were combined for additional analysis of the effects of behavioral condition and age.
A previous report showed that in an open arena without objects, the mean spatial information score of PRC neurons is low (0.19 bits/spike) relative to the hippocampus (1.05 bits/spike) and to the medial entorhinal cortex (0.61 bits/spike; Hargreaves et al., 2005). The increase in activity of PRC neurons that was observed in the present study at object locations, however, should increase the spatial information score. When the spatial information score was calculated for all cells that showed activity on the track during the behavioral conditions with objects (a maximum firing rate greater than 2 spikes/bin occupancy), the mean information score was 0.72 bits/spike for the young rats and 0.50 bits/spike for the aged rats. By contrast, for the no objects conditions the information score was 0.41 bits/spike in the young rats and 0.36 bits/spike in the aged rats. Thus scores were higher for track conditions with objects relative to conditions without objects. Specifically, when the data from both age groups were analyzed together, and the food dish regions were excluded, there was a significant effect of behavioral condition on the spatial information score (F[3,40] = 3.03, p < 0.05; repeated-measures ANOVA). This effect was due to the significantly lower information score for the no objects condition relative to the novel objects conditions (with a 20 min and a 2 hr delay; p < 0.02 for all comparisons; simple contrasts). Interestingly, the spatial information score between the no objects condition and the configuration change condition was not significantly different (p 214
= 0.15), which suggests that changing the locations of objects reduced the information score of PRC neurons. The potential reason for this reduction will be examined in the discussion.
As mentioned above, age affected the spatial information score. Overall, the mean information score of PRC neurons was significantly larger in young compared to aged rats (F[1,40] = 8.23, p < 0.01; repeated-measures ANOVA).
Additional post hoc comparisons revealed that the spatial information score was
significantly larger in the young relative to the aged rats during conditions with
objects (T[32] = 2.90, p < 0.01; paired-samples T Test), but was there was no
significant difference in the information score between age groups for the no
1 young aged 0.8
0.6
0.4
Spatial information (bits/spike) information Spatial 0.2
0 No objects Objects-both epochs Objects-both epochs Configuration (20 min) (2 hr) change Figure 7.8: Information scores of PRC neurons. The mean spatial information score of the PRC neurons that showed activity during track running for the young (green) and aged (purple) rats in the different behavioral conditions. Error bars represent +/-1 SEM.
215
objects condition (T[11] = 0.64, p = 0.54; paired-samples T Test). Figure 7.8
shows the mean information scores of PRC neurons recorded during the 4
behavioral conditions. When objects were placed on the track, the information
score of PRC neurons increased in both young and aged rats, but this increase
was larger in the young compared to the old animals. Finally, to determine
whether the object-related firing properties of PRC cells were present from the
very first exposure to objects on the track, the mean spatial information scores calculated during epoch 1 of the first behavioral experience with objects were
compared to epoch 2 of the last behavioral experience with objects. In both the
young and the aged rats there was no difference in the mean spatial information
scores between the first exposure to objects and the last (T[11] = 0.41, p = 0.69,
paired-samples T test). Because the information scores for the first and last
exposure were virtually identical (0.58 bits/spike versus 0.63 bits/spike), it does
not appear that there was a slow development of object-related activity, rather
encountering objects on the track increased information scores de novo.
Because it was difficult to isolate the contributions of the food dishes that
were distinct from the objects located adjacent to the food dishes, the no objects
condition was used to evaluate the effect of the food dish on PRC neuron spatial
information scores. Specifically, the information scores during the no objects
condition were calculated for the entire track and compared to the information
scores obtained when the food dish regions were excluded. The difference
between these information scores could then be attributed to the food dishes. 216
The results of this analysis indicated that the food dishes alone increased the mean information score of PRC neurons, and this increase was significantly greater than zero in both the young (T[11] = 3.67, p < 0.001; one-sample T test),
and the aged rats (T[11] = 3.86, p < 0.001; one-sample T test). The increased
information score that could be attributed to the food dishes did not differ
significantly between young and old rats (T[11] = 1.62, p = 0.12; paired-samples T
test). Figure 7.9 shows the spatial information score that could be attributed to
the food dishes alone in young and aged rats for the no objects condition.
0.6 young ) 0.5 aged
0.4
0.3
0.2 Spatial information difference (bits/spike)
Spatial information (bits/spike 0.1
0 Food dish regions - no objects conditions Figure 7.9: The effect of the food dishes on spatial information score. The difference in spatial information scores when they were calculated for the entire track area minus when the food dish regions were excluded. In both young (green) and aged rats (purple), the food dish regions significantly increased the spatial information score, but age did not affect this measure. Error bars represent +/- 1 SEM.
Because adding objects to the track significantly increased the mean spatial information scores of PRC neurons, the proportion of cells with ‘object fields’ was determined using the criteria that a PRC neuron was considered to 217 have an object field if its information score was above 0.5 bits/spike when the food dish regions were excluded, and there were at least four consecutive 4.1 cm bins where the neuron’s firing rate exceeded the mean firing rate. Additionally, the boundaries of an object field had to overlap with a region of the track that contained an object. A PRC neuron was considered to have a ‘food dish field’ if its information score was greater than 0.5 bits/spike when no objects were on the track, and the mean occupancy normalized firing rate at the food dish exceeded the mean firing rate for the entire track. Figure 7.10 shows the firing rate
15 Spatial information score = 0.51 bits/spike
10
5 Firing rate (spikes/occupancy) rate Firing
0 -300 -200 -100 0 100 200 300 Location (cm) Figure 7.10: A representative PRC neuron with object fields. The occupancy normalized firing rate histogram of a PRC neuron recorded during epoch 1 of a behavioral condition with objects. The spatial information score for this neuron was 0.51 bits/spike. The horizontal red line indicates the mean firing rate of the cell. The ‘object fields’ were considered to be those portions of the track where the occupancy normalized firing rate histogram exceeded the mean for at least 4 consecutive 4.1 cm bins. The red circles indicate the boundaries of the six object fields that were expressed by this neuron.
218 histogram of a representative PRC neuron with a spatial information score of
0.51 bits/spike and multiple object fields. The horizontal red line indicates the mean firing rate, and the red circles indicate the boundaries of the object fields.
When the cells expressing object fields were determined using the criteria described above, the proportion of PRC neurons with object fields was not significantly different between epoch 1 and epoch 2 for any of the behavioral conditions (F[1,42] = 0.04, p = 0.85; repeated-measures ANOVA). Moreover, in
both the young and the aged rats there was no difference in the proportion of
cells with object fields during epoch 1 of the first behavioral experience with
objects and epoch 2 of the last behavioral experience with objects (T[11] = 0.64, p
= 0.54, paired-samples T test). This indicates that PRC neurons were capable of
expressing object fields without any previous experience with objects on the track. Finally, for the three rats that had tetrodes in both layer V and layers II/III of the PRC there was not a significant difference in the proportion of cells with object fields between the different layers of cortex (T[5] = 0.24, p = 0.82; paired-
sample T test). Therefore, data obtained from the two different epochs of
behavior, and from the different cortical laminae were combined for additional
analyses of the effects of behavioral condition and age group.
The young rats had a significantly higher proportion of PRC cells that
expressed object fields compared to the aged rats (F[1,40] = 44.53, p < 0.001;
repeated-measured ANOVA). Additionally, behavioral condition had a significant
effect on the proportion of PRC neurons with object fields (F[3, 40] = 10.14, p < 219
0.001; repeated-measures ANOVA). Post hoc comparisons indicated that during
the no objects condition, significantly fewer neurons met the criteria for having
object fields relative to all three behavioral conditions with objects (p < 0.05 for all comparisons; Tukey HSD). When the proportion of PRC neurons that expressed object fields was compared between young and aged for the different behavioral conditions separately, it was observed that the proportion of PRC neurons with
object fields was not significantly different between the young and aged rats for
the no objects condition (T[11] = 1.68, p = 0.12; paired-samples T test). In
contrast, young rats had significantly more cells with objects fields relative to the
aged animals for the objects, both epochs with a 20 min delay (T[11] = 3.50, p <
0.4 young aged 0.3
0.2 Proportion
0.1
0 No objects Objects-both epochs Objects-both epochs Configuration (20 min) (2 hr) change Figure 7.11: Proportion of PRC cells with object fields in young and aged rats. During the no objects condition, in both young (green) and aged (purple) rats, less than 15% of PRC neurons had information scores above 0.5 bits/spike and activity that exceeded the mean occupancy normalized firing for at least 4 consecutive 4.1 cm bins. During the three behavioral conditions with objects, however, the young rats had a significantly higher proportion of PRC cells that expressed object fields compared to the aged rats. Error bars represent +/-1 SEM.
220
0.01; paired-samples T test), the objects, both epochs with a 2 hr delay (T[9] =
3.26, p < 0.05; paired-samples T test), and the configuration change (T[11] = 4.38, p < 0.01; paired-samples T test) conditions. Figure 7.11 shows the mean proportion of PRC neurons that had object fields for the different behavioral conditions when the food regions were excluded. These data indicate that PRC neurons in both young and aged rats show object-related spiking, but that the young rats have a higher proportion of neurons that show object fields.
The proportion of PRC neurons showing food dish fields was evaluated using the criteria that a cell has an information score above 0.5 bits/spike, and a mean occupancy normalized firing rate for the 7 bins located at and before the food dish that exceeded the total mean firing rate for the entire track. Additionally, cells that were active as the rat was leaving the food dish were excluded. Finally, cells that had a food dish field at only one food dish were counted separately from the cells that showed an increase in activity at both food dishes. These neurons were analyzed separately because a single food dish field implies that the neuron was responsive to the conjunction of the food dish and its particular location on the track, while a neuron that is active at both food dishes may be responding to the reward regardless of the spatial location.
This analysis revealed that the proportion of PRC neurons with food dish fields was significantly greater in the young compared to the aged rats (F[1,22] =
18.86, p < 0.001; repeated-measures ANOVA). There was no significant interaction effect between age group and the proportion of cells with 1 food dish 221
field versus fields at both food dishes (F[1,22] = 3.78, p = 0.07; repeated-measures
ANOVA). Figure 7.12 shows the proportion of PRC neurons that showed no food
dish activity (white, solid line), an activity at 1 food dish (dark blue) or at both food
dishes (grey). The dashed white bar is the probability for observing two food dish
fields by chance. In both the young and the aged rats the proportion of PRC
neurons showing increased activity at both food dishes was significantly greater
than what was expected by chance (p < 0.05 for both comparisons; one-sample
T test). This suggests that although some PRC neurons only responded to one
food dish, and may have been selective for the different food dish locations, a
proportion of the PRC neurons in both young (both food dishes – chance = ~6%)
1 No food dish fields
1 Food dish field 0.8 2 Food dish fields
0.6 Chance probability of 2 food dish fields
0.4 Proportion
0.2
0 young aged Figure 7.12: Food dish-associated activity of PRC neurons. The proportion of PRC neurons with no food dish fields (white, solid line), with 1 food dish field (dark blue), with fields at both of the food dishes (grey), and the chance probability for observing two food dish fields (white, dashed line) if having a field at one food dish was not correlated with having a field at the other food dish. In both the young and the aged rats the proportion of PRC neurons showing increased activity at both food dishes was significantly greater than what was expected by chance (p < 0.05 for both comparisons; one-sample T test). 222 and aged (both food dishes – chance = ~5%) rats showed increased activity near a food dish, regardless of its location.
7.3.4 Aging, novelty and perirhinal cortical neuron firing rate
Table 7.3 shows the mean and maximum firing rate of all PRC neurons recorded from the young and aged rats in the different behavioral conditions.
There was no difference in these spike rate parameters between epochs 1 and 2 for the different behavioral conditions (F[1,18] = 0.06, p = 0.81; repeated-measures
ANOVA). Additionally, when the firing rates of neurons recorded from the
different cortical layers were compared for the three rats that had tetrodes in both
layer V and layers II/III (2 young and 1 old), the firing rates between the different
layers were not significantly different (T[23] = 1.23, p = 0.23; paired-samples T
test). Therefore, the data obtained from the two different epochs and for the
Table 7.3: Single-unit firing characteristics of PRC neurons in young and aged rats
20 min 2 hr Configuration No objects p value delay delay change
Young 1.84 Hz 1.44 Hz 1.96 Hz 1.59 Hz p = 0.33 Mean firing rats ± 0.21 ± 0.20 ± 0.24 ± 0.19 rate Aged 1.26 Hz 1.00 Hz 1.28 Hz 1.21 Hz p = 0.16 rats ± 0.09 ± 0.11 ± 0.11 ± 0.06 p value p < 0.05 p < 0.05 p < 0.05 p < 0.05
Young 5.31 Hz 5.48 Hz 5.96 Hz 4.25 Hz p = 0.07 Maximum rats ± 0.31 ± 0.60 ± 0.36 ± 0.33 firing rate Aged 4.78 Hz 4.69 Hz 5.16 Hz 4.97 Hz p = 0.24 rats ± 0.10 ± 0.18 ± 0.18 ± 0.24 p value p = 0.44 p = 0.83 p = 0.07 p = 0.16
223 different cortical laminae were combined for comparisons between age groups.
There was a significant effect of age on the mean firing rate (F[1,40] =
22.52, p < 0.001; repeated-measures ANOVA), but age did not significantly affect the maximum firing rate (F[1,40] = 0.31, p = 0.58; repeated-measures ANOVA).
Table 7.3 shows the p values for age comparisons of the mean and maximum
firing rates for the different behavioral conditions. Behavioral condition did not
significantly affect the mean (F[3,40] = 0.31, p = 0.58; repeated-measures ANOVA)
or maximum firing rate (F[3,40] = 1.87, p = 0.15; repeated-measures ANOVA) in
either age group.
There are several possible explanations for the higher mean firing rate of the young rats relative to the old animals. Among these include the fact that aged rats have slower running velocities. In the hippocampus, the firing rate of CA1 pyramidal neurons is modulated by velocity (McNaughton et al., 1983; Maurer et al., 2005) in both young and old animals (e.g. Shen et al., 1997), therefore, slower overall running speed could lead to lower mean firing rates in regions of the brain where firing rate increases as the rat moves faster. In order to explore this possibility, firing rate was examined as a function of velocity in the young and the aged rats for velocities between 3 cm/sec to 51 cm/sec. Figure 7.13 shows the relationship between mean firing rate and velocity for the (A) no objects, (B)
objects, both epochs with a 20 min delay, (C) objects, both epochs with a 2 hr
delay, and (D) configuration change conditions. There were no significant
correlations between running velocity and PRC neuron mean firing rate in either 224
(A) No objects (B) Objects-both epochs (20 min delay) 3 3 ) ) 2 2 Firing rate(Hz Firing rate (Hz 1 1 Firing rate (Hz) rate Firing Firing rate (Hz) rate Firing
young young aged aged
0 0 3 9 15 21 27 33 39 45 51 3 9 15 21 27 33 39 45 51 Velocity (cm/sec) Velocity (cm/sec) (C) Objects-both epochs (2 hr delay) (D) Configuration change 3 3 ) ) 2 2 Firing rate (Hz) Firing rate rate(Hz(Hz Firing rate (Hz) Firing Firing rate (Hz 1 1
0 0 3 9 15 21 27 33 39 45 51 3 9 15 21 27 33 39 45 51 Velocity (cm/sec) Velocity (cm/sec) Figure 7.13: Firing rate as a function of running velocity. Mean firing rat was not significantly modulated by velocity in either the young (green), or the aged (purple) rats for the (A) no objects, (B) objects-both epochs with a 20 min delay, (C) objects-both epochs with a 2 hr delay, and (D) configuration change conditions (r[53] < 0.1, p > 0.5 for all comparisons; Pearson’s correlation coefficient. Error bars represent +/-1 SEM. the young or the aged rats for any behavioral condition (r[53] < 0.1, p > 0.5 for all comparisons; Pearson’s correlation coefficient). These data suggest that the firing rate difference between young and old rats cannot be accounted for by the aged rats’ slower running speeds. Moreover, in contrast to neurons in the hippocampus (McNaughton et al., 1983; Maurer et al., 2005), and medial 225 entorhinal cortex (Sargolini et al., 2006), PRC neurons do not show significant velocity modulation in their firing rates.
An alternate explanation for the decreased firing rates in aged compared to young rats could be a differential effect of relative novelty and familiarity on
PRC neuron firing rates. Previous studies have reported that in young rats PRC neurons have lower firing rates for familiar versus novel stimuli, and that this decrease in PRC neuron activity to previously experienced stimuli may provide a
‘familiarity signal’ that supports recognition memory (Zhu and Brown, 1995; Zhu et al., 1995). Moreover, it has been hypothesized that aged rats (Robitsek et al.,
2008), and humans (Daselaar et al., 2006) may have an enhanced familiarity signal in the PRC that could possibly compensate for a decline in hippocampal- dependent recollection (see Chapter 5). Thus, a major hypothesis examined in the current experiment was whether novelty and/or familiarity modulate the firing characteristics of PRC cells. Additionally, it was hypothesized that potential age- related differences in novelty/familiarity modulation may account for the impairments in stimulus recognition that are observed in old animals (see
Chapter 6).
In order to examine the predicted ability of novelty to modulate PRC discharge, the firing rate was measured across laps for epochs 1 and 2 of the different behavioral conditions, and then was normalized within a cell by calculating the Z-score firing rate for each lap. Neuron activity within 28.7 cm of a food dish was excluded from this analysis because of possible contamination 226 from the increased activity that occurred in this region. Finally, the counterclockwise (CC) and clockwise (clock) laps were analyzed separately.
There was no systematic change in the normalized firing rate over laps during any of the behavioral conditions for either the young or the aged rats
(F[15,120] = 0.41, p = 0.97; repeated-measures ANOVA). Moreover, the change in
normalized firing rate between lap 1 and lap 10 was not significantly different
between young and aged rats (F[1,10] = 0.20, p = 0.66; repeated-measures
ANOVA). Therefore, contrary to the prediction that novelty powerfully influences
PRC cell firing rates, there was no observed difference in the firing rates of PRC
neurons between novel and relatively familiar objects. Figure 7.14 shows the
mean difference in the normalized firing rate between lap 1 and lap 10 for the (A)
no objects, (B) objects, both epochs with a 20 min delay, (C) objects, both
epochs with a 2 hr delay, and (D) configuration change conditions. These data
indicate that the mean population firing rates within an epoch of behavior, in both
young and old PRC neurons, did not differ between novel and familiar objects.
Moreover, the difference in normalized firing rate between lap 1 and lap 10 for
the 3 conditions with objects was similar to the no objects condition (Figure
7.14A).
Although the population of recorded PRC neurons did not show a change
in firing rate between conditions with novel objects compared to the other
behavioral conditions, it is still possible that a subset of PRC neurons might show
a response decrement as objects go from being novel to relatively familiar that 227
1 (A) No objects 1 (B) 20 min delay Young Aged 0.5 0.5
0 0 (z-score) -0.5 -0.5
Lap 1-lap 10 firing rate rate firing 10 1-lap Lap -1 -1 CC clock CC clock CC clock CC clock Epoch 1 Epoch 2 Epoch 1 Epoch 2 (C) 2 hr delay (D) Configuration change 1 1
0.5 0.5
0 0 (z-score) -0.5 -0.5 Lap 1-lap 10 firing rate rate firing 10 1-lap Lap -1 -1 CC clock CC clock CC clock CC clock Epoch 1 Epoch 2 Epoch 1 Epoch 2 Figure 7.14: The difference in normalized firing rate between lap 1 and lap 10. In both the young (green) and the aged (purple) rats there was no systematic change in firing rate between lap 1 and lap 10 during either epoch 1 or epoch 2. Moreover, there was no significant difference in normalized firing rate between lap 1 and lap 10 between the (A) no objects, (B) objects-both epochs with a 20 min delay, (C) objects-both epochs was a 2 hr delay, and (D) configuration change conditions. Error bars represent +/-1 SEM. was not detected when the data were collapsed across all active neurons. To examine this possibility, PRC neurons that had firing rates at least 2 standard deviations above mean firing rate during lap 1 were identified. The proportion of recorded PRC neurons that met this criterion were then compared between the different behavioral conditions and between epochs 1 and 2. Approximately 5% of all the recorded PRC neurons fit this description; however, the proportion of neurons that were in this category was not significantly different across any of the 228
(A) No objects (B) 20 min delay 0.4 0.4 Young Young 0.3 0.3 Aged Aged 0.2 0.2 Proportion 0.1 0.1
0 0 CC clock CC clock CC clock CC clock Epoch 1 Epoch 2 Epoch 1 Epoch 2 (C) 2 hr delay (D) Configuration change 0.4 0.4
0.3 0.3
0.2 0.2
Proportion 0.1 0.1
0 0 CC clock CC clock CC clock CC clock Epoch 1 Epoch 2 Epoch 1 Epoch 2 Figure 7.15: The proportion of recorded PRC neurons that show a response decrement. In both young (green) and aged (purple) rats no significant difference was found in the proportion of cells that had a response decrement between the (A) no objects (B) objects, both epochs with a 20 min delay, (C) objects, both epochs with a 2 hr delay, and (D) configuration change conditions during both epochs of track running (F[15, 120] = 1.51, p = 0.21; repeated-measures ANOVA). Additionally, the young and the aged rats did not differ in the proportion of cells that showed a response decrement between the first lap and subsequent laps (F[1,10] = 3.4, p = 0.11, repeated measures ANOVA). Error bars represent +/-1 SEM.
behavioral conditions, including the no objects condition (F[15, 120] = 1.51, p =
0.21; repeated-measures ANOVA). This suggests that a response decrement was just as likely to occur in the objects conditions as it was in the novel objects conditions. Moreover, the proportion of cells that showed a response decrement between lap 1 and the subsequent laps was not significantly different between the young and the aged rats (F[1,10] = 3.4, p = 0.11, repeated measures ANOVA).
Figure 7.15 illustrates the mean proportion of the recorded PRC neurons in 229 young and aged rats during epochs 1 and 2 that had a firing rate at least 2 standard deviations above their average firing rate in the (A) no objects (B) objects, both epochs with a 20 min delay, (C) objects, both epochs with a 2 hr delay, and (D) configuration change conditions. Together, these data indicate no evidence of a response decrement in the activity of PRC neurons as objects went from novel to familiar in either the young or the aged rats.
7.3.5 PRC neurons without object or food dish fields in young and aged rats
The finding that differences in running speed between young and aged rats cannot account for age-associated firing rate changes in PRC neurons, and the observation that PRC neurons in both age groups did not show differential responses to novel versus familiar stimuli, indicates that there must be an alternate explanation for the effect of age on PRC neuron firing rate. Importantly, the firing rates of PRC neurons during sleep were not significantly different between age groups (1.3 Hz in aged rats, and 1.1 Hz in young rats; F[1,10] = 2.26, p = 0.16; repeated-measures ANOVA). This indicates that in advanced age PRC neurons do not become generally less excitable. Rather, it appears that during behavior, fewer PRC neurons in old animals show objects fields (Figure 7.11), or food dish fields (Figure 7.12). Consistent with this finding, is the observation that the proportions of neurons that did not show any activity during an episode of track running (maximum firing rate less than 2 spike/occupancy, including the food dish regions) was significantly higher in the aged compared to the young 230
rats (F[1,40] = 34.80, p < 0.001; repeated-measures ANOVA). This difference was
observed during all behavioral conditions (p < 0.05 for all comparisons, Tukey
HSD). In both young and old rats, however, there was not a significant effect of
condition on the proportion of inactive neurons (F[3,40] = 0.82, p = 0.49; repeated-
measures ANOVA). Figure 7.16A shows the mean proportion of PRC neurons
that did not show activity on the track in young (green) and aged (purple) rats for
the four different behavioral conditions. Because there was no difference in the
numbers of inactive cells between different behavioral conditions but the
proportion of object fields significantly increases between no objects condition and the conditions with objects, it does not appear that the ‘quiet neurons’ were the cells that acquired objects or food dish fields.
Figure 7.16B shows the proportions of neurons that showed non-selective activity on the track during the four different behavioral conditions. There was a significant effect of condition on the proportion of neurons with non-selective spiking on the track (F[3,40] = 15.14, p < 0.001; repeated-measures ANOVA).
Specifically, in both young and aged rats, significantly more PRC neurons
showed non-selective activity on the track during the no objects conditions
relative to the conditions with objects (p < 0.5 for all comparisons). These data
indicate that it is the neurons with non-selective activity on the track that will
potentially develop object or food dish fields when stimuli are added to the track.
Thus, it is possible that a level of baseline activity prepares a portion of the PRC 231
0.8 A young aged
0.6
0.4
0.2 Proportion of inactive cells cells inactive inactive of of Proportion
0 No objects Objects-both Objects-both Configuration epochs (20 min) epochs (2 hr) change B 0.8
0.6
0.4
0.2 Proportion of non-selective cells non-selective of Proportion
0 No objects Objects-both Objects-both Configuration epochs (20 min) epochs (2 hr) change Figure 7.16: The proportion of inactive and non-selective PRC neurons during track running in young and aged rats. (A) Significantly more PRC neurons were inactive during track running in the aged (purple) compared to the young (green) rats. This was observed during all behavioral conditions. (B) The proportions of neurons that were active on the maze, but did not show selective spiking at an object location or a food dish were significantly greater for the no objects condition relative to the conditions with objects. Aged rats showed significantly fewer neurons with non-selective activity on the track relative to the young rats during the no objects conditions, but not during conditions with objects. Error bars are +/- 1 SEM. neuron population to respond when a salient feature is encountered in an environment.
Age also significantly affected the proportion of neurons with non-selective spiking on the track (F[1,40] = 7.30, p < 0.01; repeated-measured ANOVA). 232
Specifically, the young rats had a significantly larger proportion of non-selective neurons, relative to the aged rats, during the no objects conditions (T[11] = 4.10, p
< 0.01; paired-samples T test), but not during conditions with objects (p > 0.4 for
all comparisons). Interestingly, there was an approximate 17% reduction in the
proportion of non-selective neurons between the young and aged rats. This is
roughly equivalent to the ~14% reduction in the proportion of neurons expressing
object fields between the young and old animals. This provides additional
evidence that it is the cells with some level of activity that will be recruited to
express object or food dish fields. Moreover, it is possible that the old rats have
fewer neurons with object or food dish fields because initially, under conditions
without objects, they have a smaller pool of neurons that are prepared to show
selective spiking when salient features are added to an environment.
7.3.6 Object-related activity across delays
In order to evaluate if PRC neuron activity patterns were similar between
epochs, the correlation of PRC neuron firing rates between epoch 1 and epoch 2
was calculated for the different behavioral conditions. To determine the
correlated activity patterns between epochs, the track was divided into 80 bins of
~4.1 cm, and the firing rate of an individual neuron was calculated for each bin.
Laps that the rats ran in the counterclockwise direction were separated from laps
where the running direction was clockwise, and bins within ~28.7 cm of either
food dish (14 bins) were excluded because these regions of the track show 233 elevated firing rates regardless of the behavioral condition and could inflate the measure of correlated activity. Therefore, the firing rate was determined for a total of 156 bins resulting in a 156 x 1 firing rate vector. The Pearson’s correlation coefficient between the epoch 1 and the epoch 2 firing rate vectors was then calculated for all PRC neurons across the four different behavioral conditions in the two age groups.
When correlation values of the firing rate vectors were measured between epochs 1 and 2 for both age groups together, behavioral condition did not have a significant effect on the correlated activity (F[3,27] = 1.21, p = 0.36; repeated- measures ANOVA). The aged rats, however, showed significantly higher levels of correlated activity between epochs 1 and 2 compared to the young rats (F[1,9] =
6.54, p < 0.05; repeated-measures ANOVA). Moreover, there was a significant
interaction effect of age group and behavioral condition on the activity pattern
correlations between the two epochs of track running (F[3,27] = 4.61, p < 0.02;
repeated-measures ANOVA). Post hoc analysis revealed that the correlated
activity patterns between epoch 1 and 2 for the objects, both epochs conditions
with a 20 min or a 2 hr delay was not significantly different between young and
old rats (p > 0.5 for both comparisons; Tukey HSD). On the other hand, the aged
rats showed significantly higher mean correlated activity patterns between
epochs 1 and 2 for the configuration change and no objects conditions relative to
the young rats (p < 0.01 for all comparisons; Tukey HSD). Within an age group,
activity correlations between epochs were significantly modulated by behavioral 234
condition in the young rats (F[3,15] = 4.14, p < 0.05; repeated-measures ANOVA),
but not in the aged rats (F[3,15] = 2.36, p = 0.12; repeated-measures ANOVA).
Specifically, for the young rats, activity correlations between epochs were significantly higher for the objects, both epochs conditions compared to the configuration change and the no objects conditions (p < 0.05 for all comparisons). Figure 7.17 shows the mean activity correlations between epochs
1 and 2 for the different behavioral conditions in young (green) and the aged
(purple) rats. These data suggest that when objects and their locations do not
0.8
) r (
) 0.6
Young 0.4 Aged
Mean correlation(r 0.2 Mean Pearson’s correlation
0 No objects Objects-both Objects-both Configuration epochs (20 min epochs (2 hr change delay) delay) Figure 7.17: Correlated activity patterns of PRC neurons between behavioral epochs. The mean Pearson’s correlation coefficient of binned firing rates between epoch 1 and epoch 2 for the different behavioral conditions in young (green) and aged (purple) rats. There was a significant interaction effect between age group and behavioral condition on the activity pattern correlations (F[3,27] = 4.61, p < 0.02; repeated-measures ANOVA). The young rats showed significantly lower mean activity correlations between epochs 1 and 2 for the configuration change and no objects conditions relative to the aged rats (p < 0.05 for all comparisons). Error bars represent +/-1 standard error of the mean.
235 change between epochs of track running, ensemble activity in the PRC between epochs is similarly correlated in young and aged rats. This was observed even when there was a 2 hr delay between epochs. In contrast, when the locations of objects was changed between epoch 1 and epoch 2 (configuration change condition), the activity of young PRC cells was less correlated between epochs relative to the objects-both epochs conditions. This was not observed in the aged rats, suggesting that the aged PRC may be more likely than the young PRC to retain an ‘object field’ even when the object is swapped with a different one.
7.4 Discussion
7.4.1 Summary of salient results
Three major novel findings have emerged from the current experiment: 1) the firing rates of perirhinal cortical (PRC) neurons in layer V of area 36 are not modulated by the relative novelty or familiarity of a stimulus. Although fewer neurons were recorded in layer II/III and area 35, these cells were also unaffected by the relative familiarity or novelty of objects that were encountered while traversing a circular track (Figures 7.14, 7.15); 2) the firing characteristics of PRC neurons are increased in areas of the track in which objects were located. We refer to the areas of higher PRC activity as ‘object fields’ because this firing property was not observed in conditions in which the track did not have objects (Figures 7.6, 7.10); 3) in addition, many PRC neurons exhibit firing rate increases at the locations of the track that contain a food dish (‘food dish fields’; 236
Figures 7.6, 7.12). Moreover, unlike hippocampal (e.g., McNaughton et al., 1983;
Shen et al., 1997) or medial entorhinal cortical neurons (Sargolini et al., 2006;
Maurer, personal communication), PRC neurons show no relationship between firing rate and running velocity (Figure 7.13). The implications of the lack of velocity modulation in the PRC will be discussed further in Chapter 8.
In addition to these novel observations, the current experiment provides insights into the changes in PRC physiology that might contribute to age- associated deficits in object recognition (see Chapter 6). We observed several differences between young and old rats in the activity characteristics of neurons recorded from this area. Specifically, a lower proportion of PRC neurons express object fields in the aged rats relative to the young animals (Figure 7.11), and fewer aged PRC neurons show increased activity at the food dish locations
(Figure 7.12). This is consistent with the observation that aged rats had more
PRC neurons that were inactive on the track, and overall aged PRC neurons had lower firing rates during behavior. Importantly, the firing rates of PRC neurons were not different between young and aged rats during sleep, indicating that there is not an overall reduction in the excitability of these cells. Finally, for the young rats only, the activity pattern correlations of PRC neurons between epochs
1 and 2 were greater for the objects, both epochs conditions (with 20 min delay or a 2 hr delay) relative to the configuration change and the no objects conditions
(Figure 7.17). This was not the case in the aged rats in which all behavioral conditions were similarly correlated between epochs. Moreover, the activity 237 pattern correlations were higher in the aged rats for the no objects and configuration change conditions compared to the young rats. Two possible explanations for these observations can be suggested. First, aged PRC neurons are more likely than young PRC neurons to retain an object field in a particular location after the object is replaced with an alternate object. Second, because there are fewer PRC neurons that show non-selective activity in the aged compared to the young PRC, this non-selective activity must have greater stability across epochs in the old rats.
7.4.2. The lack of novelty and/or familiarity modulation of perirhinal cortical neuron activity in young and aged rats
As mentioned previously, there was no significant effect of novel versus familiar objects on the overall firing rates of PRC neurons in the current experiment. This suggests that a response decrement of PRC neuron activity as a stimulus becomes familiar may not be the physiological correlate of PRC- dependent object recognition. These results have implications regarding the type of neural coding scheme employed by the PRC and how this could be affected during normal aging. A long standing view is that the PRC encodes information about the relative familiarity and recency of a stimulus with a simple rate code.
Specifically, neuron firing rates are higher if a stimulus is novel or if a stimulus has not been experienced recently (for review, see Brown and Aggleton, 2001).
The current results do not support the idea that response decrements support 238 stimulus recognition. This type of rate code is vulnerable to noise and spike failures. Given that the probability of synaptic transmission between the PRC and its cortical efferents is low (Pelletier et al., 2004), the signal-to-noise ratio for relaying a ‘response decrement’ might not be optimal. In light of the current findings, it is more plausible that the PRC utilizes a population code, which relies on the joint activities of a number of neurons and each neuron has a different distribution of responses over some set of inputs.
If network capacity is defined as the number of discrete states available to represent different stimuli, then in order for a population of neurons to have adequate capacity for storing distinct representations of different stimuli, three criteria should be satisfied (cf. McNaughton and Morris, 1987). First, there should be extensive connectivity between the neurons in the network. The PRC at least partially satisfies this criterion as the intrinsic connections of this region are dense (Burwell and Amaral, 1998). Second, the activity responses of different neurons in the network across different stimulus sets should not be more correlated than expected by chance (i.e., they should be orthogonal). The orthogonality requirement is empirically supported by the observation that before stimuli are learned, or associated with other stimuli, PRC neurons do not show correlated activity patterns (Erickson et al., 2000). Finally, an efficient population code is maximally distributed; that is, half of the available neurons are used to code each stimulus set (e.g., McNaughton and Morris, 1987). Figure 7.18 shows how the capacity of a theoretical network is affected by increasing or decreasing 239
Young rats
Aged rats network capacitynetwork
r neurons from a set of n Figure 7.18: Population coding in the PRC. The capacity of a hypothetical network to represent distinct stimuli based on the degree to which the population code is distributed. The X-axis is the number of neurons (r) in a set of n neurons. Network capacity is maximal when half of the neurons are activated by a given stimulus. In the PRC cortex fewer neurons are activated by a stimulus set in aged (purple) compared to young (green) rats. This could have the result of decreasing the capacity of the aged PRC to represent different stimuli distinctively. the number of neurons that are active when a stimulus set is presented. In young rats approximately 33% of PRC neurons expressed object fields, while less than
20% of PRC neurons showed object fields in aged rats. The impact that this decrease in neural activity could have on PRC network capacity in aged rats is shown in Figure 7.18.
A compromise in any of the network parameters described above will decrease the capacity of a population-based code to represent distinct stimulus sets. During normal aging, behavioral data indicate a reduced ability of the aged relative to the young PRC to differentiate novel stimuli from ones that have been previously encountered (Chapter 6). Although it is unknown how aging might 240 affect the intrinsic connectivity of the PRC, the current data suggest that neurobiological changes within this brain region in advanced age result in fewer neurons being activated by objects. The functional consequences of this are a reduced capacity of PRC neurons to represent different objects distinctly, which could contribute to aged rats regarding novel objects as familiar (Chapter 6).
7.4.3 Object fields in young and aged rats
Compared to the young rats, the aged rats had a lower proportion of neurons with object fields (Figure 7.11) and a higher proportion of neurons that were inactive during track running (Figure 7.16A). Interestingly, there was no evidence that the inactive neurons in either young or aged rats were the cells that would potentially express object fields. In fact, the proportions of inactive neurons were similar across conditions without objects and conditions with objects in both age groups. It appears that the neurons with non-selective activity (i.e., elevated firing that did not correspond to an object location) may have the ability to develop object fields. This would be consistent with the observation that there is a significant reduction in the proportion of neurons with non-selective activity between the no objects and objects conditions. This reduction was observed in both young and aged rats. Finally, the numbers of these non-selective neurons were similar between young and old rats during conditions with objects.
Interestingly, the aged rats also exhibited a significant reduction in neurons with non-selective activity during the no objects condition. Together these data 241 support two novel hypotheses regarding PRC function. First, it appears that when the track was relatively empty, a subset of PRC neurons fire in systematic locations around the track. This activity could reflect circuit dynamics that predispose specific neurons to develop an object field when a salient feature is encountered in the environment. Second, with respect to aged rats, the decreased numbers of neurons showing non-selective activity on the empty track could arise from the PRC neurons in old rats being under tighter inhibitory control during behavior (but not during rest/sleep). This idea is consistent with two observations in the present data: 1) the proportion of cells that exhibit non- selective activity and object fields are reduced in aged relative to young rats, 2) when these neurons do fire, their discharge patterns are more regular, resulting in higher spatial correlations between epochs 1 and 2.
A question remains regarding the function of the inactive PRC neurons and what distinguishes these cells from those that show non-selective activity.
One possibility is that these ‘quiet neurons’ may be under more inhibitory control than the cells with non-selective activity. Presumably, these cells will only become active when the animal encounters a stimulus that is emotionally salient or relevant for its survival. In line with this hypothesis is the observation that more
PRC neurons respond selectively to an auditory cue after it has been paired with a foot shock (Furtak et al., 2007). Additionally, excitatory transmission from the perirhinal to the entorhinal cortex is increased when the basolateral amygdala is active following an unexpected reward (Paz et al., 2006). In the framework of this 242 hypothesis, the observation that aged rats have more inactive cells suggests that during waking behaviors there may be a difference in the ratio of inhibition to excitation between old and young animals.
7.4.4 Food dish fields in young and aged rats
Many PRC neurons showed firing rate increases at one or both food dishes (35% in young rats and 21% in aged rats). Although these food dish fields were more prevalent in young compared to old rats, in both age groups more cells were active at a food dish than were active for any single object. These data suggest that either PRC neurons are more likely to respond to stimuli that are behaviorally relevant, or they may respond to the food reward itself. It is unclear from the current data, however, if a food reward alone (i.e., a reward not placed in an actual dish that is very familiar to the rat) would elicit PRC neuron activity, or if it associating a reward with a physical object was necessary for the expression of food dish fields. Regardless, these data suggest that rewarding stimuli may provoke more PRC activity than stimuli that are neutral (e.g., objects that are not associated with reward). This idea may be consistent with the possibility that dopamine modulates PRC activity patterns and promotes the expression of food dish fields.
The perirhinal cortex receives a direct projection from the basal ganglia, including the ventral tegmental area (Furtak et al., 2007) and sends a reciprocal projection back to the ventral striatum (McIntyre et al., 1996). Therefore, the 243 release of dopamine from ventral tegmental neurons in anticipation of, or during consummation of a food reward (e.g., Nishino et al., 1987; for review, see Wise,
2006), may facilitate PRC neuron activity that is selective for the reward or stimuli associated with the reward. In this view, aged rats potentially show fewer food dish fields because of a decline in dopamine function. As mentioned in the
Discussion of Chapter 6, there is evidence for reduced activity in the limbic dopamine system in advanced age (for review, see Barili et al., 1998).
Additionally, there is a lower proportion of ventral tegmental neurons that are active during rewarding behaviors in aged relative to young rats (Hoang et al.,
2008). This reduced dopamine activity in old animals may make it more difficult for PRC neurons to escape inhibitory control and show activity at the food dish.
7.4.5 The influence of spatial location on object and food dish field activity
The current experiment was designed specifically to examine the effect of relative novelty and familiarity on PRC neuron firing rates. Therefore, the experimental procedures were not optimized for evaluating the parameters that control object and food dish fields. Although the current data indicate that object fields are only present when objects are on the track, the specificity of this activity to a particular object remains to be determined. Future experiments will address this issue (see Chapter 9). Additionally, there was always a food dish when the rats were given the reward, and a reward was always placed in the food dish as the rats approached it. Therefore, it is unclear whether the food dish field activity 244 would persist if, contrary to the rat’s expectations, the reward was not given. It is also unknown whether a reward without any association with a physical object would elicit potential “reward fields”.
The role of spatial location or context on object and food dish fields requires further exploration. At present, it is unclear whether these fields represent the conjunctive encoding of a stimulus in a particular location, or if a stimulus in any location will elicit the same PRC neuron activity. There are several lines of evidence that support the idea that spatial location plays a role in object and food dish field activity. First, a proportion of food dish fields (24% in young rats and 12% in aged rats) were only selective for one food dish. Because the two food dishes were identical and both were an equal distance from the barrier, this indicates that the activity in at least some PRC neurons was modulated by the location of the food dish. It cannot be ruled out, however, that a small difference between the food dishes, not detected by the experimenters, allowed the rats to regard them as unique stimuli. Moreover, because the food dishes never moved, it remains to be determined if any of these food dish fields would continue to respond after the location is changed. Another observation consistent with a role for space in object fields is the fact that many PRC cells fired to an object in one place, but ceased firing in the configuration change in which the object was moved to another place.
The large number of objects used in the current study, the tendency for a single PRC neuron to express multiple object fields, and the population based 245 analyses employed, made it difficult to determine the extent to which individual object fields followed a particular object. The significantly reduced activity correlation in young rats between epochs for the configuration change condition, relative to the other conditions with objects, strongly suggest that at least some
PRC object fields altered their spiking patterns after the objects changed locations. Although the activity correlation was reduced in the configuration change condition, it was still high (r = ~0.5) relative to no correlation (r = 0). This suggests that a portion of active PRC neurons may have retained an object field in the same location even when the object was changed. It is also possible that the firing rate of some PRC neurons changed when the object was moved, but the field remained fixed to the same location. Within individual neurons there are examples of PRC cells that exhibited significantly reduced firing rates after the configuration change condition, as shown in Figure 7.19. The proportions of cells in which object fields followed an object, showed new/independent activity, or retained their firing field independent of objects is currently under examination.
Several additional experiments are being conducted to address some these questions. First, a condition has been designed where completely different objects are placed on the track between epochs to determine the relative specificity of an object field for a particular object. Second, a condition in which multiple identical objects are placed on the track at different locations (i.e., 3 sets of 2 identical objects) will help to determine the extent to which PRC neurons respond to the conjunction of a particular object in a specific location in space. 246
Epoch 1 – Familiar Objects Epoch 2 – Familiar Objects Familiar Configuration Novel Configuration A 30 30 20 20
10 10occ occ
0 Spikes/ 0 Spikes/ spikes/occ -300 -200 -100 0 100 200 300 spikes/occ -300 -200 -100 0 100 200 300 B12 12 10 10 8 8 6 6 # oflaps 4 # of laps laps 4 # oflaps 2 2 0 0 -300 -200 -100 0 100 200 300 -300 -200 -100 0 100 200 300 cm cm 1 1 Figure 7.19: PRC activity during the configuration change condition. The firing patterns for a representative PRC neuron rduring the two epochs of behavior. The positions of 8 familiar objects on the track were changed between the two epochs. (A) The firing rate histogram by ‘linearized’ position during the first epoch (left) and the second epoch (right). The X axis is position on the track (cm) with zero indicating the position of the barrier. Positive numbers are for laps when the rat was running in the clockwise direction while negative numbers indicate the position when the rat was running in the counterclockwise direction. The Y axis is the occupancy normalized firing rate of the neuron. (B) The raster plots by lap. Each horizontal line indicates a lap and blue lines are the laps in which the rat ran in the counterclockwise direction while red lines represent laps run in the clockwise direction. The black vertical lines indicate the position and extent of the objects. The activity of this neuron was decreased by the reconfiguration of objects.
These experiments will be discussed further in Chapter 9. Although there are several questions left to be answered regarding the behavioral correlates of PRC neuron activity, the experiment reported in this chapter provides strong evidence against the theory that decremental PRC neuron firing can support stimulus recognition. 247
CHAPTER 8: MIDDLE HIPPOCAMPAL CA1 ACTIVITY CHARACTERISTICS
ARE MODULATED BY 3-DIMENSIONAL OBJECTS: EFFECTS ON PLACE
FIELS SIZE, PROBABILITY OF PLACE FIELD EXPRESSION,
AND MAP STABILITY
8.1 Introduction
The previous chapter described how neurons in the perirhinal cortex
(PRC) showed a consistent increase in their firing rates near locations with objects and areas associated with a food reward. Additionally, there was a reduction in the proportion of PRC neurons in aged compared to young rats that increased their firing rate in response to encountering objects. One question that arises from these results is how age-related changes in PRC cell firing properties impact information processing in downstream structures. The PRC sends projections to the CA1 subfield of the hippocampus both directly (Canning and
Leung, 1997; Liu and Bilkey, 1997; Naber et al., 1999) and through the lateral entorhinal cortex (Naber et al., 1999; Burwell, 2000; Witter et al., 2000), and age- related changes in CA1 neuron firing characteristics are well documented (see
Chapter 2). Although the PRC and the lateral entorhinal cortex are a major source of input to the hippocampus, relatively little is known regarding the contributions of these structures to the activity patterns observed in hippocampal cells.
The hippocampus is hypothesized to play a role in episodic memory though the association of items within the context in which they were 248 experienced (e.g., Nadel et al., 1985; Nadel and Hardt, 2004; Eichenbaum et al.,
2007). Moreover, the ability to use the original context in which a stimulus was perceived to facilitate recognition (i.e., the capacity for stimulus recollection) also requires an intact hippocampus (Fortin et al., 2004). Importantly, both episodic memory (e.g. Scoville and Milner, 1957; Mishkin et al., 1998; Ekstrom and
Bookheimer, 2007) and recollection (Fortin et al., 2004) involve the association of spatial and non-spatial information. While there is a large body of empirical research describing the effect of spatial (e.g., O'Keefe and Dostrovsky, 1971;
O'Keefe, 1976; O'Keefe and Conway, 1978; McNaughton et al., 1983; Muller and
Kubie, 1987) and self-motion related input on hippocampal neuron firing characterizes (e.g., McNaughton et al., 1983; Maurer et al., 2005; Terrazas et al.,
2005), less is known about the influence that non-spatial information may have on hippocampal activity patterns.
Neocortical sensory and sensory association areas can, in large part, be divided into those regions that are primarily involved in processing spatial and movement-related information, and those regions that are more involved with the perception of stimulus features. For example, early in the visual system, information necessary for perceiving movement, and for action planning appears to engage anatomical regions associated with the “dorsal stream”, such as area
MT and the posterior parietal cortex (e.g., Goodale, 1993; Ungerleider and
Haxby, 1994; Milner and Goodale, 2008). In contrast, information related to sensory features necessary for determining stimulus identity appears to engage 249 anatomical regions associated with the “ventral stream”, such as the inferotemporal cortex (e.g., Goodale, 1993; Ungerleider and Haxby, 1994; Milner and Goodale, 2008). In primates and rats the anatomical separation of spatial/motion versus non-spatial/perceptual information continues into the medial temporal lobe (Amaral and Witter, 1995; Burwell, 2000). Specifically, the
PRC and lateral entorhinal cortex receive more input from the inferior temporal cortex and other cortical regions associated with the ventral visual stream than does the medial entorhinal cortex (Suzuki and Amaral, 1994a; Suzuki and
Amaral, 1994b; Burwell et al., 1995; Burwell and Amaral, 1998a; Burwell and
Amaral, 1998b; Burwell, 2000; Furtak et al., 2007; Kerr et al., 2007). Conversely, the medial entorhinal cortex and the postrhinal cortex (Area TH and TF of the parahippocampal cortex in the primate) receive more extensive afferent connections from the poster parietal cortex, retrosplenial cortex, and other regions considered to be part of the dorsal visual stream (Suzuki and Amaral,
1994a; Suzuki and Amaral, 1994b; Burwell et al., 1995; Burwell and Amaral,
1998a; Burwell and Amaral, 1998b; Burwell, 2000; Furtak et al., 2007; Kerr et al.,
2007).
When different subregions of the medial temporal lobe are lesioned, dissociable behavioral deficits are observed. This suggests that the divergent anatomical connectivity within the medial temporal lobe results in the postrhinal/entorhinal cortices, and the perirhinal/lateral entorhinal cortices making distinct contributions to episodic memory and recollection. For example, PRC 250 lesions lead to object recognition impairments (Norman and Eacott, 2005;
Winters and Bussey, 2005), while lesions of the postrhinal cortex result in deficits in recalling where an object has previously appeared (Norman and Eacott, 2005).
A similar dissociation has been observed between damage to the medial versus lateral entorhinal cortices. Lesions of the medial entorhinal cortex produce spatial learning and memory impairments (Steffenach et al., 2005), while inactivation of the lateral entorhinal cortex leads to memory deficits in a visual discrimination task (Myhrer and Andersen, 2001).
Given the differences among afferent inputs, and the functional dissociations between the medial and lateral entorhinal cortices, it is not surprising that the behavioral correlates of neurons recorded in the different subregions of the entorhinal cortex are also distinct. Neurons in layers II and III of the medial entorhinal cortex show strong spatial selectivity (Fyhn et al., 2004;
Hafting et al., 2005; Hargreaves et al., 2005; Fyhn et al., 2008). These cells are activated when the animal's position coincides with any vertex of a regular grid of equilateral triangles spanning the surface of the environment, earning these neurons the name ‘grid cells’ (Hafting et al., 2005). In contrast to the strong correlation of activity in the medial entorhinal cortex with an animal’s position in space, the lateral entorhinal cortex shows very little spatial selectivity
(Hargreaves et al., 2005). Rather these cortical cells show increased activity at the locations of objects (Deshmukh and Knierim, 2008). Similarly, we have 251 shown that PRC neurons exhibit ‘object fields’ and reward-associated firing characteristics (Chapter 7).
Consistent with the activity patterns of cells in the medial entorhinal cortex, the principal cells of the hippocampus that receive this afferent input also exhibit spatial selectivity. Principal cells in the hippocampus increase their firing rates at discrete locations of an environment, known as the neuron’s “place field”
(O'Keefe and Dostrovsky, 1971; O'Keefe and Nadel, 1978). When the firing patterns of many CA1 hippocampal neurons are recorded simultaneously, it is possible to reconstruct the position of a rat within an environment from the cell firing patterns alone (Wilson and McNaughton, 1993). The composite CA1 cell activity is ‘map-like’ and, in different environments, hippocampal place maps change markedly, exhibiting “global remapping” (e.g., Muller and Kubie, 1987).
Multiple lines of evidence indicate that the medial entorhinal cortex exerts a strong influence on CA1 pyramidal neuron activity. First, the scale at which space is represented (that is, the size of place fields and grid cell firing field size and spacing) increases systematically along the dorsoventral axis in both the CA1 subregion of the hippocampus (Jung et al., 1994; Maurer et al., 2005; Maurer et al., 2006) and the medial entorhinal cortex (Hafting et al., 2005). This implies that the size of medial entorhinal cortical grid fields contributes to the determination of the spatial extent over which a CA1 neuron will be active. Second, when global remapping occurs in populations of CA1 cells, there is an associated realignment and registration of the firing fields of the grid cells (Fyhn et al., 2007). Finally, a 252 specific lesion of cells in layer III of the medial entorhinal cortex decreases the spatial selectivity of CA1 place fields (Brun et al., 2008). These data have lead to the development of place field models that hypothesize that the spatial selectivity of hippocampal pyramidal neurons arises from the linear summation of inputs from medial entorhinal cortical grid cells (McNaughton et al., 2006; Solstad et al.,
2006).
In contrast to the medial entorhinal cortex, there are few empirical data on the activity pattern characteristics of neurons in the lateral entorhinal and perirhinal cortices. This has made it more difficult to propose a model for how these structures may affect CA1 place field firing characteristics. Moreover, lesions of the lateral entorhinal input have not been studied in relation to CA1 place field activity. Previous reports, however, did demonstrate that lesions of the
PRC (Muir and Bilkey, 2001), or combined lesions of both the medial and lateral entorhinal cortices (Van Cauter et al., 2008) in young rats can result in CA1 place map instability. Specifically, when intact young rats are returned to a familiar environment, the active CA1 neurons have place fields in the same region of the environment as observed during previous recording sessions (e.g., Muller et al.,
1987; Thompson and Best, 1990). When the PRC or entorhinal cortex is lesioned, however, the place field map is stable within an episode of environmental exploration, but when the young rat is removed, and later returned to the familiar environment, a new map is retrieved (Muir and Bilkey, 2001; Van
Cauter et al., 2008). Interestingly, these data obtained from young rats with PRC 253 or entorhinal lesions are similar to the place map multi-stability that is observed in aged rats (Barnes et al., 1997). One possible conclusion to be drawn from these observations is that input from the PRC and entorhinal cortex may be involved in the association of the hippocampal map with the external features of an environment, and that this input could be compromised during normal aging.
Therefore, it is not unreasonable to hypothesize that the prominent projections from the PRC and lateral entorhinal cortex, which convey much of the non-spatial information from the ventral visual processing stream to the hippocampus, influence CA1 unit activity characteristics.
To begin to understand how non-spatial information may modulate CA1 pyramidal cell firing properties, the current experiment investigated how adding
3-dimensional objects to a familiar environment altered the probability of place field expression, size, and map stability. Young rats were trained to traverse a track for food reward while the activity of multiple place fields was recorded. Data from the present dissertation (Chapter 7) suggest that PRC neurons are responsive under conditions in which objects are placed in an environment, and lateral entorhinal cortical neurons also increase their firing rates near object locations (Deshmukh and Knierim, 2008). Thus, it was hypothesized that if rats were required to navigate around 3-dimensional objects, then the object-related firing of PRC and lateral entorhinal cortical cells would be increased, and this could result in altered place field characteristics of the CA1 cells that receive this input. 254
8.2 Methods
8.2.1 Subjects and behavioral training
Electrophysiological studies were conducted on four Fisher-344 male rats between 8 and 12 months old. The rats were housed individually and maintained on a 12:12 light–dark cycle. Before rats were implanted with the hyperdrive recording device, all rats were screened for spatial memory impairments and normal vision using the Morris swim task (Morris, 1984). The behavioral procedures for testing on the swim task were identical to those described in
Chapter 6 and Chapter 7. Briefly, all animals were tested over 4 days with 6 spatial trials on each day. Animals were then screened for visual ability with 2 days of cued visual trials (6 trials/day) in which the escape platform was above the surface of the water but the position of the platform changed between each trial. Rats’ performance on the swim task was analyzed offline with either in- house software (WMAZE, M. Williams) or a commercial software application
(ANY-maze, Wood Dale, IL). Because different release locations and differences in swimming velocity produce variability in the latency to reach the escape platform, a corrected integrated path length (CIPL) was calculated to ensure comparability of the rats’ performance across different release locations
(Gallagher et al., 1993).
During electrophysiological recordings, the animals were food deprived to about 85% of their ad libitum weight and trained to run on a circular track (~335 cm in circumference) in both the counterclockwise and clockwise directions for 255 food reinforcement. All electrophysiological recordings took place during the dark phase of the rats’ light–dark cycle. Food rewards were given in a small plastic food dish (4 cm x 4 cm) at two positions on the track. Both food dishes were located at the position on the track that marked the completion of one lap on opposite sides of a barrier, that is, where the rat was required to turn around and run in the opposite direction (Figure 7.1A).
During all electrophysiological recording sessions the behavioral procedures were similar to those used in the experiments described in Chapter 7.
Rats were required to run 20 laps (10 in the counterclockwise direction and 10 in the clockwise direction) during two distinct episodes of behavior. Each track running epoch was flanked by a 10 or 20 minute period in which the rat rested in the same room in a pot located at the center of the track. Thus, the activity of
CA1 neurons was monitored during an initial rest session (before behavior), during the first epoch of track running, during a second rest session after epoch
1, during a second epoch of track running, and then finally during a third rest period. Data from the rest periods were used to assess firing stability across the entire recording session (data not shown).
All four rats participated in three different behavioral procedures. During the first procedure (day 1) rats ran on an empty track (Figure 7.1A) for both epochs of behavior (no objects, both epochs condition; Table 8.1). A 20 minute rest period occurred between epoch 1 and epoch 2. For the second procedure
(day 2), during epoch 1, eight novel objects that varied in size, color and texture 256 were placed at eight different locations along the track. The rat had to run past these objects in order to obtain the food reward. Following the 20 minute rest period, during epoch 2, six of the same objects used in epoch 1 remained in the same location on the track and two objects that were on the track during epoch 1 were replaced with two novel objects (objects, both epochs condition; Figure
7.1B; Table 8.1). During the third procedure (day 3), the same eight objects that were placed around the track during epoch 2 from the previous day of recording were again placed on the track in the same positions as in day 2 for the first epoch of behavior. The rats ran 20 laps around these familiar objects, and then rested for 20 minutes. Following the 20 minute rest period, the rats ran another
20 laps. During this epoch of behavior, however, the positions of the eight familiar objects were pseudo-randomly shuffled such that no object was located in the same position between epochs 1 and 2 (configuration change condition;
Figure 6.1B; Table 8.1). Each rat completed all of these behavioral procedures on three consecutive days, and this process was repeated four times for each rat for a total of 12 days of electrophysiological recording.
Table 8.1: Behavioral conditions during epochs 1 and 2 of track running
Condition Epoch 1 Delay Epoch 2 # of rats No objects, both epochs No objects 20 min No objects 4
Objects, both epochs 8 novel objects 20 min 6 familiar objects from 4 epoch 1, and 2 novel objects
Configuration change 8 familiar objects, in familiar 20 min 8 familiar objects with all 4 locations objects locations different from epoch 1 No objects-objects No objects 20 min 8 novel objects 2
Objects-no objects 8 novel objects 20 min No objects 2
257
In two of the four rats, electrophysiological recordings were conducted during two additional behavioral procedures. These included a “no object-objects condition” (Table 8.1), and an “objects-no objects condition” (Table 8.1). For the no objects-objects condition, the rats ran 20 laps during epoch 1 with nothing on the track except the two food dishes and the divider. During epoch 2, however, eight novel objects were placed on the track. Conversely, in the objects-no objects condition, eight novel objects were placed on the track for epoch 1, but these objects were then removed during epoch 2. Two rats ran the no objects- objects and the objects-no objects procedures four times each (eight additional days of recording).
For all conditions in which objects were placed on the track, the objects were fixed in place using Velcro, thus, rats could actively explore, rear, and climb onto the objects without displacing them. Additionally, during all rest periods the objects were removed from the track so that the rat could not see them during the intervening 20 minute delay.
8.2.2 Surgical Procedures
Surgery was conducted according to National Institutes of Health guidelines for rodents and protocols approved by the University of Arizona
Institutional Animal Care and Use Committee. Prior to surgery, the rats were administered penicillin G (30,000 units intramuscularly in each hind limb) to combat infection. The rats were implanted with a “hyperdrive” manipulator device 258 that held an array of 14 separately moveable tetrode recording probes. During surgical implantation the rats were maintained under anesthesia with isoflurane administered at doses ranging from 0.5% to 2.5%. The hyperdrive recording device, implantation methods, and the parallel recording methods have been described in detail elsewhere (Gothard et al., 1996). Briefly, each hyperdrive consisted of 14 drive screws coupled by a nut to a guide cannula. Twelve of these cannulas contained tetrodes (McNaughton et al., 1983; Recce and
O'Keefe, 1989), four-channel electrodes constructed by twisting together four strands of insulated 13 µm nichrome wire (H. P. Reid, Inc., Neptune, NJ). Two additional tetrodes with their individual wires shorted together served as an indifferent reference and an electroencephalogram (EEG) recording probe. A full turn of the screw advanced the tetrode 318 µm. For two rats, recordings were
made from the distal middle CA1 region,
which receives a direct projection from
neurons in the PRC and from neurons in
layer III of the LEA (6.0 posterior, 3.0
lateral to bregma, and angled 14°
towards the midline), and tetrodes for Figure 8.1: Tetrode tracks for distal middle CA1 recordings. A nissl stained this hippocampal location were verified coronal section of the right hemishere of a rat that participated in the current histologically. The other two rats were experiment. This section is approximately -6.0 mm posterior to bregma and shows 3 tracks, made by the tetrode recording implanted to obtain recordings from the probes, that reached the pyramidal cell layer (red circles) of the distal region of perirhinal cortex (Chapter 7). Therefore, middle CA1.
259 the implant coordinates were more lateral than the other two rats (6.0 posterior,
5.5 lateral to bregma, and angled 14° towards the midline). Nonetheless, in these two rats, 3-6 tetrodes were medial enough to reach the distal CA1 subregion of the middle hippocampus. Figure 8.1 shows a representative example of tetrode recording tracks in this region of CA1. Qualitatively similar properties were observed in the pyramidal neurons recorded from tetrodes adjacent to the subiculum and tetrodes in more lateral areas of distal CA1.
The implant was cemented in place with dental acrylic anchored by small screws. Immediately after surgery, all tetrodes were lowered approximately 1 mm into the cortex, and rats were orally administered 26 mg of acetaminophen
(Children’s Tylenol Elixir, McNeil, PA) for analgesia. Oral administration of acetaminophen was continued for 3-5 days after surgery. Additionally, all rats were given either 25 mg of ampicillin (Bicillin, Wyeth Laboratories, Madison, NJ) or a combination of 20 mg of sulfamethoxale and 0.4 mg trimethoprin (Hi-Tech
Pharmacal Co., Inc, Amityville, NY) on a 10 days on/10 days off regimen for the duration of the experiment.
8.2.3 Neurophysiology
The tetrodes were lowered after surgery into the hippocampus, allowed to stabilize for several days just above the middle CA1 hippocampal subregion, and then gradually advanced into the CA1 stratum pyramidale. The neutral reference electrode was placed in or near the corpus callosum. The four channels of each 260 tetrode were attached to a 50-channel unity-gain head stage (Neuralynx, Inc.,
Tucson, AZ). A multi-wire cable connected the head stage to digitally programmable amplifiers (Neuralynx, Inc.). The spike signals were amplified by a factor of 1,000 – 5,000, band pass-filtered between 600 Hz and 6 kHz, and transmitted to the Cheetah Data Acquisition system (Neuralynx, Inc.). Signals were digitized at 32 kHz, and events that reached a predetermined threshold were recorded for a duration of 1 ms. Spikes were sorted offline on the basis of the amplitude and principal components from the four tetrode channels by means of a semi-automated clustering algorithm (KlustaKwik, author: K. D.
Harris, Rutgers–Newark). The resulting classification was corrected and refined manually with custom-written software (MClust, author: A. D. Redish, University of Minnesota; updated by S. L. Cowen and D. R. Euston, University of Arizona), resulting in a spike-train time series for each of the well-isolated cells. No attempt was made to match cells from one daily session to the next. Therefore, the numbers of recorded cells reported does not take into account possible recordings from the same cells on consecutive days; however, because the electrode positions were adjusted from one day to the next, recordings from the same cell over days were probably relatively infrequent. Putative pyramidal neurons were identified by means of the standard parameters of firing rate, burstiness, and spike waveform (Ranck, 1973). Continuous recording was also taken from each tetrode and the EEG from the tetrode with the most pyramidal cell activity was analyzed and filtered for the hippocampal theta rhythm. This is a 261 prominent 6–10 Hz oscillation that occurs in the hippocampus during active exploration (Vanderwolf, 1969). The theta signal was band pass-filtered between
1 and 300 Hz and sampled at a frequency of 2.4 kHz. Several diodes were mounted on the head stage to allow position tracking. The position of the diode array was detected by a TV camera placed directly above the experimental apparatus and recorded with a sampling frequency of 60 Hz. The sampling resolution was such that a pixel was approximately 0.3 cm.
8.2.4 Analyses and Statistics
Place field occupancy normalized firing rate histograms and spike raster diagrams were constructed by plotting the circular trajectories of the animals on a linearized, one-dimensional scale, using a linear interpolation (Maurer et al.,
2005). Place field location and size were determined using two different criteria.
The first criterion capitalized on the observation that a unitary place field does not extend beyond a single cycle of theta phase precession (Maurer et al., 2006). It is well documented that as a rat passes through a CA1 neuron’s place field, the spike timing shows a systematic shift such that the spikes occur at successively earlier phases of the theta cycle until there is a complete cycle of phase advancement (O'Keefe and Recce, 1993; Skaggs et al., 1996). By using the phase precession definition, the start location of a field was defined as the location at which spikes fired late in theta phase (approximately 360°) and began a cycle of precession. The end location of the place field was determined to be 262 the location of spikes with the largest phase angle. This was accomplished by creating a two-dimensional plot of occupancy normalized position (x-axis) versus phase of spikes (y-axis), as well as occupancy normalized position versus phase density plots. To construct the theta phase of firing versus position plots, we assigned each spike a nominal theta phase. Precisely, the phase assigned to a spike at time t was 360° (t – t0)/(t1 – t0), where t0 and t1 are the times of the
preceding and following peaks of the filtered reference EEG signal (Skaggs et al.,
1996). The phase is always a number between 0 and 360. Place field boundaries
were manually drawn using the theta phase precession information by an
experimenter who was blind to the behavioral condition of the recording session.
Place fields that overlap with a food dish often show an abrupt halt in theta phase
precession and can show considerably less than 360° of precession. For this
reason, place fields that overlapped with a food dish were excluded.
Additionally, place field size was quantified using a firing rate definition.
For this criterion, the track was divided into 87 bins of approximately 3.84 cm. A
place field was considered to be the portion of the track that the occupancy
normalized firing rate of the neuron exceeded 10% of the maximum occupancy
normalized firing rate for at least 5 consecutive bins.
For all analyses, the mean of the dependent variable of interest was
determined for each rat, and all statistics were calculated using a rat mean rather
than the recording session or cell mean, which can inflate statistical power due to 263 the large number of observations. All statistical tests and p values were calculated using SPSS 9.0 (Chicago, Illinois), with alpha set at the 0.05 level.
8.3 Results
8.3.1 Morris Swim Task Performance
The mean performance of the rats used in the current experiment on the spatial version of the Morris swim task significantly improved over the four days of testing (Figure 8.2; black line). This is indicated by the significantly shorter corrected integrated path length (CIPL) between day 1 and day 4 (T[3] = 14.9, p <
0.005, paired-samples T-test). This pattern was observed in all rats. Additionally, rats performed significantly better when the platform was visible (Figure 8.1; grey
line), compared to the fourth day of spatial trials (T[6] = 4.52, p = 0.05). These
data indicate that no rat had inadequate vision since they all benefitted from
40.0
Spatial trials Visual trials 30.0
20.0 CIPL (m) CIPL
10.0
0.0 Day 1 Day 2 Day 3 Day 4 Figure 8.2: Morris swim task performance. The four rats used in the current experiment showed a significant improvement in locating the hidden escape platform between the first day of spatial testing (black line) and the fourth day (T[3] = 14.9, p < 0.005, paired-samples T-test). Additionally, rats performed significantly better on the cued visual trials, when the platform was visible (grey line), compared to day four of the spatial trials (T[6] = 4.52, p = 0.05), indicating that they had adequate vision. Error bars represent +/- one standard error of the mean.
264 platform visibility. Figure 8.2 shows the mean results for all rats on the Morris swim task for both the spatial (black line) and the visual trials (grey line).
8.3.2 Running Velocity
Rats have a natural tendency to explore objects, or other stimuli, that are novel (Ennaceur and Delacour, 1988). In the current experiment, this naturally occurring behavior could lead to slower overall velocities in behavioral conditions with novel objects, as the rats’ running would be interrupted by periods of exploration. To examine this possibility we compared the running velocity for the different behavioral conditions with objects versus conditions where the track was bare. Figure 8.3A shows the mean velocity for epoch 1 (white) and epoch 2
(grey) during the different behavioral conditions. The running speed across all laps was significantly different between epoch1 and epoch 2 (F[1,11] = 11.14, p <
0.01; repeated-measures ANOVA), but behavioral condition did not significantly
affect velocity (F[1,11] = 1.29, p = 0.33; repeated-measures ANOVA). There was a significant interaction effect of epoch and behavioral condition on velocity, however (F[4,11] = 15.85, p < 0.001; repeated-measures ANOVA). Post hoc
analysis revealed that this interaction effect was due to the significant difference
in velocity between epoch 1 and epoch 2 of the no objects-objects and the
objects-no objects conditions (p < 0.05; Tukey HSD). Specifically, within a
recording session, when rats ran on the track and objects were on it, the velocity
was significantly slower than when the track was bare. In contrast, when both 265 epochs of track running within a recording session had objects or did not have objects velocity was not significantly different between epochs (p > 0.05 for all comparisons; Tukey HSD). Finally, there was no significant effect of running direction (counterclockwise versus clockwise) on velocity during epoch 1 (F[1,6] =
0.08, p = 0.78, repeated-measures ANOVA), or epoch 2 (F[1,20] = 0.19, p = 0.66;
ANOVA).
60 A Epoch 1 Epoch 2 50
40
30
Velocity (cm/sec) 20
10
0 no objects, both objects, both configuration no objects- objects-no epochs epochs change objects objects
50 B Epoch 1 C 50 Epoch 2
40 40
30 30
20 20 Velocity (cm/sec) Velocity Velocity (cm/sec) Velocity No objects No objects 10 Novel objects 10 Novel objects Configuration change Familiar objects Configuration change 0 0 Laps 1-2 Laps 19-20 Laps 1-2 Laps 19-20 Figure 8.3: Running velocity during the different behavioral conditions. (A) The mean running velocity across all laps during epoch 1 (white) and epoch 2 (grey) for the different behavioral conditions. (B) The mean running velocity during epoch 1 for laps 1-2 and laps 19-20 for conditions with no objects (dark grey), conditions with novel objects (yellow), and the configuration change condition (red). (C) The mean running velocity during epoch 2 for laps 1-2 and laps 19-20 for conditions with no objects (dark grey), conditions with novel objects (yellow), conditions with familiar objects (blue), and the configuration change condition (red). 266
In order to examine further if novel objects on the track lead to slower velocities, the mean running speed of the rats during the first 2 laps and the last
2 laps were compared over the different behavioral conditions. If novel objects on the track lead to increased exploration, then the rats should have slower velocities during the first several laps of epoch 1 relative to the last laps in which the objects became more familiar. Moreover, if novel objects on the track are a significant factor contributing to slower running velocities, then the difference in the rats’ running speeds between the early laps and the late laps should be less during epoch 2 and during conditions that did not have novel objects on the track.
Figure 8.3B shows the mean running velocity during laps 1-2 and laps 19-20 during epoch 1 for conditions with no objects (no objects-both epochs, and no objects-objects; dark grey), conditions with novel objects (objects, both epochs, and objects-no objects; yellow), and the configuration change condition (red).
Figure 8.3C shows the mean running velocity during epoch 2 for laps1-2 and laps
19-20 during conditions with no objects in epoch 2 (no objects-both epochs, and objects-no objects; dark grey), conditions with novel objects (no objects-objects; yellow), conditions with familiar objects (objects, both epochs; blue), and the configuration change condition (red). Because rats ran laps in both the counterclockwise and clockwise directions, lap 1 is the first lap in the counterclockwise direction and lap 2 is the first lap in the clockwise direction.
Similarly, lap 19 is the final lap that the rats ran in the counterclockwise direction 267 and lap 20 was the last lap run in the clockwise direction within an epoch of track running.
When the mean running speed of the rats for the first 2 laps was compared to the velocity for laps 19-20, statistical analysis revealed that the rats ran significantly more slowly during laps 1-2 relative to laps 19-20 (F[1,22] = 32.80,
p < 0.001). Post hoc analysis revealed that the difference in running velocity
between the first two laps and last two laps was significantly different for conditions with novel objects (epoch 1 of the objects, both epochs condition; epoch 1 of the objects-no objects condition; epoch 2 of the no objects-objects condition; p < 0.05 for all comparisons). Additionally, the rats’ running velocity was significantly slower during laps 1-2 relative to laps 19-20 for epoch 2 of the configuration change where familiar objects were placed in a novel spatial arrangement (p < 0.05). During all behavioral conditions without objects, or with familiar objects in familiar positions on the track, there was no significant difference in running speed between laps 1-2 and laps 19-20 (p > 0.1 for all comparisons). Moreover, running velocities during laps 1-2 of epoch 1 were significantly greater for conditions that did not contain objects (no objects, both epochs; no objects-objects) compared to conditions where the objects on the track were novel during epoch 1 (objects-both epoch, objects-no objects; p <
0.05 for all comparisons; Tukey HSD). In contrast, behavioral condition did not significantly effect running velocities for laps 19-20 during epoch 1 (p > 0.05 for all comparisons; Tukey HSD), which indicates that after the objects became 268 familiar the young rats had similar running speeds during all behavioral conditions. Taken together these data suggest that running velocities are slower when novel objects are on the track relative to conditions when the objects are familiar or when the track is empty. These data are consistent with the idea that rats spent more time exploring objects when they were the most novel during laps 1-2 of epoch 1. Once the objects were familiar, however, their presence on the track did not significantly impede the rats’ ability to run for a food reward.
Therefore, any potential effect of behavioral condition of the CA1 pyramidal cells cannot be explained by overall differences in velocity.
8.3.3 The Effects of Objects on the Probability of Place Field Expression and
Firing Rate
The activity of 1036 pyramidal cells from the middle CA1 region was recorded in this experiment. A summary of the database from which the current results were derived in given in Table 8.2. Additionally, Table 8.2 shows the
Table 8.2: Numbers of recorded middle CA1 pyramidal cells and place fields
No objects, Objects, Configuration No objects Objects-no Total both epochs both epochs change objects objects
Total sessions 20 14 14 8 (2 rats) 8 (2 rats) 64 with usable CA1 neurons Total cells 321 170 190 194 161 1036 Epoch 21 1 2 1 2 1 2 1 2
Cells without 214 213 84 87 126 113 117 101 91 100 place fields Cells with at least 107 108 86 83 80 86 77 93 70 61 1 place field Proportion with 0.33 0.34 0.51 0.49 0.42 0.45 0.40 0.48 0.44 0.38 ≥1 place field
269 number of CA1 neurons without a place field versus neurons with one or more place fields. In both epochs 1 and 2, significantly more middle CA1 pyramidal cells had at least one place field during track-running conditions with objects compared to conditions without objects (F[3,2] = 9.65, p < 0.01). Consistent with
this observation, in the two rats that traversed the track with objects in one epoch
but no objects in the other epoch, when the proportion of cells without place
fields was compared between the first and the second epoch, a significantly
larger proportion of cells showed place field activity during the epochs with
objects compared to epochs with an empty track (T[15] = 2.95, p < 0.01). This
suggests that, within the same population of CA1 neurons, encountering objects
on the track increased the probability of place field expression.
During the objects condition, in which objects appeared in both epochs but
were novel during epoch 1 and familiar during epoch 2, there was no significant difference between the proportion of cells with at least one place field between epochs (T[3] = 0.53, p = 0.63; paired-samples T test). Moreover, the proportion of
CA1 neurons that had place field activity was not significantly different between
any of the different behavioral conditions with objects (F[3,16] = 0.60, p = 0.62;
repeated-measures ANOVA). This indicates that novelty did not significantly
influence the probability of place field expression and that the primary factor
increasing the proportion of distal middle CA1 neurons that demonstrated place
field activity was placing objects on the track. 270
Figure 8.4A shows the mean proportion of cells with zero through eight place fields for conditions without objects (dark grey) versus the behavioral conditions where the rats had to run past objects (light grey). When there were no objects on the track, these proportions were comparable to those observed in other multiple-single unit recording studies from the middle CA1 region of the hippocampus (Jung et al., 1994; Maurer et al., 2006). For behavioral conditions with objects on the track, however, there was an approximately 10-15% increase in the proportion of CA1 neurons expressing place fields, and this increase was statistically significant (F[3,2] = 9.65, p < 0.01). Additionally, the mean number of place fields expressed by each individual CA1 pyramidal neuron that was active during track running was significantly greater for conditions with objects
A 1 2.5 B No Objects No Objects Objects Objects 0.8 2
0.6 1.5
Proportion 0.4 1 # of place/cell # of # of place fields/cell # ofplace
proportion of neurons proportion 0.2 0.5
0 0 012345678 # of# of place place fields fields Figure 8.4: Place field expression across behavioral conditions. (A) The proportion of cells in middle CA1 with n = 0,1,2,3,4,5,6,7 or 8 place fields when there were no objects on the track (dark grey) versus conditions with objects on the track (light grey). A significantly larger proportion of middle CA1 cells expressed place fields during track running conditions with objects (F[3,2] = 9.65, p < 0.01). (B) The mean number of place fields per active neuron during track-running epochs with no objects (dark grey) versus the conditions with objects (light grey). The CA1 neurons that showed a significant firing rate increase during the track running epochs expressed signigicantly more place fields within an epoch of behavior when objects were place on the track relative to when the track was empty (F[3,18] = 4.56, p < 0.03; repeated-measures ANOVA). All place fields were determined using the phase precession definition and error bars represent +/- 1 standard error of the mean.
271
compared to conditions when the track was empty (F[3,18] = 4.56, p < 0.03;
repeated-measures ANOVA). Figure 8.4B shows the mean number of place
fields expressed by the individual CA1 neurons that were active during track
running. When the track was empty a single CA1 cell with place specific activity had an average of 1.6 place fields within an epoch of track running, but when the
track contained objects each active neuron had an average of 2.1 place fields.
Together these data indicate that when rats traverse a track with objects, the
probability that a middle CA1 pyramidal cell will have a place field increases.
Moreover, the neurons with place fields are more likely to have multiple fields
within a single environment compared to when the track is empty.
Table 8.3 shows the single-unit firing characteristics of the neurons recorded during each behavioral condition for epoch 1 and epoch 2. Although, a larger proportion of CA1 cells show place field activity during track-running conditions with objects, there was no significant difference in the overall behavioral firing rate for any of the different track-running conditions during epoch 1 or epoch 2 (F[3,9] = 1.18, p = 0.41). This suggests that the presence of objects did not affect the global excitability of the middle CA1 neuron population.
Table 8.3: Mean firing rates for middle CA1 neurons during epoch 1 and epoch 2 No objects, Objects, Configuration No objects- Objects- both epochs both epochs change objects no objects
Epoch 1 2 1 2 1 2 1 2 1 2
Mean all behavior 2.1 2.4 1.8 2.0 2.0 1.9 1.6 1.6 1.6 1.6 firing rate (Hz) ± 0.3 ± 0.4 ± 0.3 ± 0.3 ± 0.5 ± 0.4 ± 0.6 ± 0.6 ± 0.6 ± 0.6
Mean in-field 14.9 11.1 25.5 13.4 16.8 11.3 14.7 17.2 10.2 21.5 firing rate (Hz) ± 6.3 ± 2.4 ± 12.3 ± 2.9 ± 7.7 ± 1.9 ± 1.7 ± 8.5 ± 5.5 ± 4.3 272
Additionally, there was not a significant effect of track running condition on the mean in-field firing rate during epoch 1 or epoch 2 (F[3,9] = 0.38, p = 0.92).
8.3.4 Objects and CA1place map stability
The observation that 3-dimensional objects can increase the probability that a neuron will express a place field, and lead to an increase in the number of place fields expressed by a single active neuron suggests that objects may influence place map stability. Figure 8.5A shows a representative example of the
A B No objects, epoch 1 Objects, epoch 2 Objects, epoch 1 No objects, epoch 2
*** * * * ** ********
Figure 8.5: The effect of objects on place field activity. (A) A representative example of the activity of a single CA1 neuron when no objects were on the track during epoch 1 (left panel), and then objects were placed on the track for epoch 2 (right panel). This neuron only had place field activity near the food dish (far left area) in the absence objects but when objects were added to the track, the cell expressed three place fields. (B) A representative example of the activity of a CA1 neuron when objects were on the track during epoch 1 (left panel) but the objects were removed for epoch 2 (right panel). In this example, the neuron was active at three discrete locations and near the food dish when objects were present. When the objects were removed, however, only one of the three regions of high firing near objects remained. In both (A) and (B) the X-axes are position on the track in cm, and the rat was running in counterclockwise direction, which is represented as left to right. The black asterisks indicate the positions of the objects. The top panels show the firing rate histograms for both cells. The middle panels are the raster plots of spikes by lap, and the bottom panels show theta phase (y-axis) by position (x- axis) plots. Each dot represents a spike. Note that two cycles of theta phase precession are plotted for both panels. 273 activity of a single CA1 neuron when there are no objects on the track during epoch 1 (left panels), and then when objects are placed on the track for epoch 2
(right panels). Conversely, Figure 8.5B shows a representative example of the activity of a single CA1 neuron when there are objects on the track during epoch
1 (left panels), and then when the objects are removed for epoch 2 (right panels).
In Figure 8.5A and B the top panels show the occupancy normalized firing rate histograms. The middle panels show the raster plots of spikes by lap, and the bottom panel is the theta phase (y-axis) by position (x-axis). In both examples the spatial firing of the same CA1 neuron was different in the epoch with objects relative to the epoch with no objects.
In order to quantify whether or not CA1 place field activity was different between track-running epochs without objects and epochs with objects, within a single recording session, the correlated activity between epoch 1 and epoch 2 was calculated for each neuron. Specifically, the track was divided into 72 bins of
~4.6 cm and the firing rate of an individual neuron was calculated for each bin.
Laps that the rats ran in the counterclockwise direction were measured separately from laps where the running direction was clockwise; therefore, the firing rate was determined for a total of 144 bins resulting in a 144 x 1 firing rate vector. The Pearson’s correlation coefficient between the epoch 1 and the epoch
2 firing rate vectors was then calculated for all CA1 neurons across the five different behavioral conditions (no objects, both epochs; objects, both epochs; configuration change; no objects-objects; and objects-no objects conditions). 274
Figure 8.6 shows the proportion normalized frequency histograms of the correlation coefficients (r values) of the neurons recorded during each of the five track-running conditions. For the no objects, both epochs (Figure 8.6A) and the objects, both epochs conditions (Figure 8.6B) the mean correlations between epochs were 0.65 and 0.72, respectively. In contrast, the mean correlation for the no-objects, objects (Figure 8.6D) and objects-no objects (Figure 8.6E) conditions was 0.49 and 0.55, respectively. As anticipated, the CA1 activity between the epochs in which the track-running experience was identical across epochs was significantly more correlated than the activity between the conditions with the track-running conditions changed between epochs (Figure 8.6F (#); T[3] = 3.15, p
ABC
0.2 No objects, both epochs No objects, both epochs 0.2 NoObjects, Objects,objects, both both both epochs epochs epochs Configuration Change 0.2 Configuration change
0.15 0.15 0.15
0.1 0.1 0.1 proportion proportion proportion
0.05 proportion proportion 0.05 0.05 proportion
0 0 0 0.00 0.10 0.20 0.30 0.40 0.50 0.60 0.70 0.80 0.90 0.00 0.10 0.20 0.30 0.40 0.50 0.60 0.70 0.80 0.90 0.00 0.10 0.20 0.30 0.40 0.50 0.60 0.70 0.80 0.90 # DE F 0.2 Objects-no objects 0.2 No objects-objects Objects-no objects 0.8 *
0.15 ) 0.15 0.6
0.1 0.1 0.4
proportion 0.05 proportion proportion 0.05 0.2 mean correlation (r correlation mean
0 0 0
0.00 0.10 0.20 0.30 0.40 0.50 0.60 0.70 0.80 0.90 0.00 0.10 0.20 0.30 0.40 0.50 0.60 0.70 0.80 0.90 (r) correlation mean no objects, objects, configuration no objects- objects-no both epochs both epochs change objects objects Figure 8.6: The effect of objects on CA1 place map stability. The proportion normalized frequency distribution of the activity correlations (r) between epoch 1 and epoch 2 for CA1 neurons in the (A) no objects, both epochs, (B) objects, both epochs, (C) configuration change, (D) no objects-objects, and (E) objects-no objects conditions. (F) The mean r value for the five different behavioral conditions. The CA1 activity during the no objects, both epochs (grey) and the objects, both epochs (blue) conditions was significantly more correlated between epochs 1 and 2 relative to the no objects-objects (green), and objects-no objects conditions (#; T[3] = 3.15, p < 0.05; paired-samples T test). The configuration change conditions showed an intermediate r value between the “no change” and “change” conditions, and was significantly less correlated across epochs of track running than the objects, both epochs condition (*; p < 0.05; Tukey HSD), but was not significantly different from the other conditions (p > 0.05; Tukey HSD). 275
< 0.05; paired samples T-test). Additionally, the mean correlation between
epoch 1 and epoch 2 of the configuration change condition (Figure 8.6C) was
0.60, and intermediate in value between the “no change” and “change”
conditions. This measure of between epoch place stability for the configuration
change condition was only significantly different from the mean correlation
measured during the objects, both epochs condition (Figure 8.6F (*); p < 0.05;
Tukey HSD). Figure 8.6F shows the mean correlation values between epoch 1
and epoch 2 for the different behavioral conditions. These data suggest that changing the location of objects caused some neurons to remap but others continued to fire in the same location on the track regardless of the identity of the
nearest object.
8.3.5 The Effect of Objects on Middle CA1 Place Field Size
Figure 8.7 shows representative examples of the activity of a CA1 neuron
when no objects were on the track (Figure 8.7; left panels) and when the rats
were required to traverse the track with objects (Figure 8.7; right panels). The
neurons exhibited activity over larger portions of the track during the no object
conditions relative to the conditions with objects, which is particularly evident in
the occupancy normalized firing rate histograms (Figure 8.7A). Moreover, the
place field size was consistent throughout all laps, as can be observed in the lap
by lap spike raster plots (Figure 8.7B). Finally, when the track was empty, the rat
had to traverse a greater distance for the cell to precess 360° with respect to the 276
A No objects, epoch 1 Objects, epoch 1
B
***** * ** C
Figure 8.7: A representative example of the activity of two CA1 neurons recorded during epoch 1. One cell was recorded when there were no objects on the track (left panels) while the other neuron was recorded on a different day when objects were placed on the track during epoch1 (right panels). In both examples, the rat was running in the counterclockwise direction and the position is shown linearized such that the rat was running from left to right. The X-axes are position on the track in cm. (A) The firing rate histograms for both cells. (B) The raster plot of spikes by lap. (C) The theta phase (y-axis) by position (x-axis) plots. Each dot represents a spike. Note that two cycles of theta phase precession are plotted for both panels. The asterisks in the right panel indicate the position of objects. theta rhythm compared to when objects were on the track (Figure 8.7C). This example neuron recorded from a no-objects condition was active for more than an 80 cm area of the track compared to only approximately 40 cm for the example neuron recorded during a behavioral condition with objects on the track.
In order to quantify a potential effect of objects on place field size, the spatial extent of fields was determined using two different measures. First, the amount of distance traversed in order for the spikes to precess 360° relative to 277 the theta rhythm (Maurer et al., 2006) was measured for each place field.
Additionally, the results from this analysis were verified using a standard firing rate definition of place field size (e.g. Mehta et al., 1997; Ekstrom et al., 2001).
Specifically, the second method set the criterion of a place field to be the region of the track where the occupancy normalized firing rate exceeded 10% of the maximum occupancy normalized firing for at least 5 consecutive ~3.84 cm bins.
This measure, by its definition, applies an arbitrary lower limit such that no place field can be smaller than 19.2 cm.
When place field boundaries were determined by measuring the distance that a rat traversed in order for a CA1 neuron’s spikes to precess 360° relative to the theta rhythm, the mean size of place fields for behavioral conditions with no objects on the track was ~80 cm. The presence of objects on the track significantly reduced the mean size of place fields to ~55 cm (T[7] = 4.07, p <
0.01; paired-samples T-test). Figure 8.8A shows the mean size of place fields for
conditions with no objects versus conditions with objects.
When both epochs 1 and 2 contained objects or when both epochs did not
have objects on the track, the size of place fields was not significantly different
between the first epoch of track running and the second (T[7] = 0.66, p = 0.53;
paired-samples T-test). Moreover, place field size was not significantly different
between behavioral conditions with novel objects and conditions with familiar
objects (F[2,3] = 2.02, p = 0.25; repeated-measures ANOVA). Together these data
indicate that although novelty did not appear to affect place field size, enriching a 278
track with objects lead to a significant reduction in the size of CA1 place fields.
Figure 8.8B shows the proportion normalized frequency distribution of place field
size when rats traversed a bare track (dark grey) and when the track contained
objects (light grey). This figure shows that the range of place field sizes between
behavioral conditions with and without objects is similar, which suggests that
adding objects to the track does not change the spatial scale of place field size in
middle CA1. Rather, it appears from Figure 8.8B that when rats traverse a track
with objects, the proportion of small place fields increases while the proportion of
large place fields decreases.
AB100 0.06
No Objects 0.05 80 Objects No Objects Objects 0.04 60
0.03
40 Proportion 0.02 Place field size (cm) size field Place
20 0.01
0 0 0 50 100 150 200 250 Place field size (cm) Figure 8.8: The effect of objects on place field size. (A) The mean size of place fields for behavioral procedures with no objects on the track (dark grey) versus conditions with objects (light grey). The presence of objects on the track significantly reduced the mean size of place fields (T[7] = 4.07, p < 0.01; paired-samples T-test). Error bars represent +/- one standard error of the mean. (B) The proportion-normalized frequency distribution of place field size when rats traversed an empty track (dark grey) versus when the track had objects on it (light grey). Both frequency distributions were smoothed with a hanning window of 5. Although the range of place field size is similar between groups, when the track was enriched with objects a larger proportion of place fields were small and there was a reduction in the number of large place fields.
279
The above analyses, however, compared the size of place fields for neurons that were recorded on different days. Thus, with these data, it could not be determined if the same population of neurons can have large place fields when no objects are present but smaller place fields when objects are added to the environment. In order to investigate this possibility, the size of place fields were compared between epoch 1 and epoch 2 for the two rats that performed the objects-no objects and the no objects-objects behavioral conditions. Because most place fields showed global remapping between object versus no object epochs (see Figure 8.6D and E), it was not possible to determine whether the size of an individual place field, at a given location, was modulated by the presence of an object. In the population of recorded CA1 neurons that re- mapped, the size of the place fields in the condition with objects was significantly smaller than in the condition with no objects (F[1,3] = 25.69, p < 0.05, repeated-
measures ANOVA). These data are consistent with the interpretation that field
size is smaller when objects are on the track relative to when the track is empty.
Importantly, for the behavioral procedures where both epochs of track
running had objects on the track or both epochs did not have objects on the
track, there was no significant difference in place field size between epochs (F[1,3]
= 4.52, p = 0.12; repeated-measures ANOVA). Figure 8.9 shows the absolute
value of the mean difference in place field size between epoch 1 and epoch 2 for
the five different behavioral procedures. These data, together with the
observation that 3-dimensional objects can affect place map stability (Figure 8.6), 280 suggest that when objects are introduced or removed from the track CA1 place fields globally remap and the mean place field size for the hippocampal map when objects are on the track is smaller relative to when the track is bare. In contrast, when objects were on the track for both epochs, or if there were no objects present for both epochs, the size of place fields remained relatively stable between the two epochs of track running.
Epoch 1 versus Epoch 2 35
30 25
20 15
10
5 0 place field size difference (cm) no objects, objects, both configuration no objects- objects-no both epochs epochs change objects objects Figure 8.9: The effect of objects on place field size within a population of middle CA1 neurons. The absolute value of the mean difference in place field size between epoch 1 and epoch 2. Adding or removing objects from the track between the two epochs of track running significantly changed the size of place fields (F[1,2] = 25.69, p < 0.05, repeated-measures ANOVA). For the behavioral procedures where both epochs had objects on the track or both epochs did not have objects on the track, there was no significant difference in place field size between epochs (F[1,3] = 4.52, p = 0.12; repeated-measures ANOVA). Error bars represent +/- one standard error of the mean.
When a firing rate criterion was used to determine place field size, the mean size of place fields for conditions with no objects on the track was ~60 cm.
When objects were on the track, the mean size of place fields was ~48 cm, which was significantly smaller compared to when the track was empty (T[7] = 5.35, p <
0.05; paired-sample T-test). In both the no objects and objects conditions, the 281 firing rate measurement of place field size yielded smaller fields compared to when the theta phase precession definition was used. This difference was not significant, however (F[1,6] = 0.78, p = 0.41; repeated-measures ANOVA). These
analyses provide independent confirmation that when objects are located around
the track the place fields are smaller compared to when the track is empty.
8.5 Discussion
The main novel observation from the present experiment is that the
activity characteristics of ensembles of middle CA1 pyramidal cells is different
under conditions in which rats traverse an empty track, versus when the track
contained salient visual, tactile, and olfactory features. Specifically, placing 3-
dimensional objects on the track resulted in several significant changes in the
spiking pattern of these neurons: 1) an increase in the probability that a cell
would express a place field (Table 8.2; Figure 8.4A); 2) an increase in the mean
number of place fields per neuron of those cells that showed activity on the track
(Figure 8.4B); 3) global remapping (Figure 8.6); and 4) a decrease in the mean
size of place fields expressed by middle CA1 neurons (Figure 8.8). Although
there was a substantial decrease in the proportion of large place fields (>75 cm),
and a dramatic increase in small place fields (<40 cm) in the object conditions, the range of place field sizes was comparable between conditions (Figure 8.8B).
Since the observed change was a shift in distribution towards smaller place fields, this suggests that the overall spatial scale of this subregion of CA1 was 282 preserved in the object conditions. Finally, the mean and in-field firing rates during behavior were not different between object and no-object conditions
(Table 8.3), suggesting that the global excitability of middle CA1 is not affected by adding feature information to the track.
When rats traversed an empty 335 cm track, approximately 35% of the
CA1 neurons recorded from the middle portion of the dorsoventral axis of the hippocampus expressed place fields. This is consistent with the report that approximately 39% of neurons in this region were active on an empty 380 cm track (Maurer et al., 2006). When objects were added to the track in the current experiment, approximately 48% of middle CA1 neurons showed place field activity (see Figure 8.4A). The increase in the probability that a given middle CA1 neuron would express a place field occurred in both novel and familiar object conditions, suggesting that novelty was not required to produce increases in the number of neurons active within an environment.
The present results stand in apparent contrast to those reported by
Battaglia et al. (2004) who also compared CA1 activity in ‘cue-rich’ versus ‘cue- poor’ environments. In this study, although more neurons were active in the cue- rich conditions relative to the cue poor conditions, these differences did not reach statistical significance (Battaglia et al., 2004). At least one possible explanation for these apparent discrepancies can be offered. The CA1 neurons examined in the Battaglia et al. (2004) study were recorded in a more dorsal region of the hippocampus (3.8 mm posterior to bregma) relative to the current experiment 283
(6.0 mm posterior to bregma). Because higher proportions of neurons are active in dorsal compared to ventral hippocampal regions (Maurer et al., 2006; Jung et al., 1994), it is possible that the cue-poor condition was close to the ceiling for the numbers of active neurons. This would be consistent with the idea that, in order to optimize storage capacity, the proportion of active neurons in a single environment, even in the dorsal hippocampus, does not normally exceed 50%
(e.g. Wilson and McNaughton, 1993; Guzowski et al., 1999). Because more cells were active in cue-poor conditions (48%; Battaglia et al., 2004), there was less opportunity for the recruitment of additional neurons during the cue-rich condition. In more ventral regions of CA1, a smaller proportion of neurons are active within an environment (Maurer et al., 2006; Jung et al., 1994), as was observed in the current experiment. This may have allowed for the recruitment of additional active cells when objects were added, without resulting in a significant decrease in the storage capacity of the hippocampus.
The increase in the probability that a CA1 neuron would express a place field in the object conditions was also associated with a decrease in the mean size of place fields (Figure 8.8). This observation is consistent with the results of
Battaglia et al. (2004) who showed a faster decline in the population vector correlation on the cue-rich track relative to the cue-poor track. Because the rate of the population vector de-correlation reflects the distance that the rat must traverse for the activity of CA1 ensembles to orthogonalize, it is an indirect measure of place field size (Battaglia et al., 2004; Maurer et al., 2005; Terrazas 284 et al., 2005; Pastalkova et al., 2008). The observed decrease in place field size in the cue-rich condition (Battaglia et al., 2004), and under object conditions in the current experiment, could operate to maintain levels of sparsity within the hippocampus. Specifically, when more CA1 cells are active within an environment the size of place fields must be smaller in order for the proportion of neurons active at a given time to go unchanged (Skaggs and McNaughton,
1992).
Although the percent of active neurons increased, and the mean size of place fields decreased between the no-object and the object conditions, there was no overall increase in the global excitability. That is the mean and in-field firing rates of CA1 neurons during behavior in both conditions were similar (Table
8.3). This observation reflects the fact that although place fields were smaller when objects were on the track (Figure 8.8), each active CA1 neuron was more likely to express multiple place fields relative to when the track was bare (Figure
8.4B, ~2.3 versus ~1.6 place fields/cell). In order to calculate the total track area over which a given neuron was active, the mean size of a given place field was multiplied by the mean number of place fields per active neuron. The mean total area over which a given cell fired for all rats and all place fields was ~129 cm for the no-object conditions and ~131 cm for the object conditions, which was not statistically different (p = 0.94). This supports the idea that there is a conservation of the amount of space over which a given CA1 neuron fires. A possible consequence of this property is that it may serve to maintain a constant level of 285 excitation in the CA1 neuron population when behavioral conditions change, possibility resulting from homeostatic maintenance of synaptic gain within the hippocampus (Buzsaki et al., 2002).
A number of theories take into account the sensory input required to update the spatial firing of CA1 pyramidal cells (Burgess and O'Keefe, 1996;
Samsonovich and McNaughton, 1997). It is postulated that hippocampal activity is tightly coupled to voluntary motion (Samsonovich and McNaughton, 1997), and that variations in the gain of a movement-speed signal to the hippocampus lead to changes in spatial selectivity (Terrazas et al., 2005; for review, see
McNaughton et al., 2006). According to this ‘path integration’ model, the decrease in the responsiveness of ventral CA1 neurons to increasing velocity is the basis for the systematic variation in spatial scaling along the dorsoventral axis of the hippocampus (Maurer et al., 2005). While the path integration model accurately accounts for spatial scaling in the medial entorhinal cortex (Hafting et al., 2005; for review, see McNaughton et al., 2006), it may not completely account for the changes in place field size in middle CA1 observed in the present study. It is clear from the data presented in this chapter that place field size within the middle CA1 subregion can be decreased by providing additional non-spatial information. Moreover, when the object locations were shuffled between epochs of track running, place field activity was significantly less correlated than when the objects remained in the same spatial configuration between epochs (Figure
8.6F). Together, these data indicate that 3-dimensional objects can influence 286
CA1 spatial representations. Therefore, both path integration information and non-spatial input to the hippocampus can operate in concert to adjust place field activity patterns in the middle hippocampus. This, non-spatial information presumably arises from lateral entorhinal and perirhinal cortical input and should be considered in future models of place field selection and specificity.
The observation that adding 3-dimensional objects to an environment increased the number of active neurons in a CA1 ensemble, and decreased the mean size of place fields also has important implications for theories of memory consolidation (e.g., Squire and Zola-Morgan, 1991; Nadel and Moscovitch, 1997;
Moscovitch et al., 2005). It is generally agreed that the neocortex serves as the permanent store of long-term semantic memories (Squire and Zola-Morgan,
1991; McClelland et al., 1995), but the overall sparse interconnectivity of the neocortex (about 10−6 on average) makes it difficult to sustain the associations
necessary for the rapid formation of new memories directly. Marr was perhaps
the first to suggest that the hippocampus could serve as an indirect link to enable
rapid associations among different areas of the neocortex (Marr, 1971).
According to this general theory, during the initial encoding of a memory, output
from the hippocampus becomes associated with patterns of activity across
distinct neocortical areas. This hippocampal output could thus provide an index
code for each memory (Teyler and DiScenna, 1986; Teyler and Rudy, 2007). If
more feature information is added to an environment, however, the hippocampal
index would have to be altered in order to represent the additional sensory input. 287
The increased resolution of the spatial firing properties of neurons under object conditions could therefore reflect changes in the hippocampal index that are necessary for the objects to be incorporated into the memory. In this view, non- spatial input from the PRC and lateral entorhinal cortex may be as important in determining the firing characteristics of CA1 neurons as the self-motion/spatial input from the medial entorhinal cortex. Moreover, the influence that these two distinct types of information have on hippocampal activity patterns can account for the involvement of the hippocampus in recollection (Fortin et al., 2004) and episodic memory (e.g., Nadel et al., 1985; Nadel and Hardt, 2004; Eichenbaum et al., 2007), both of which require the association of ‘what’ stimulus was encountered with ‘where’ it was experienced.
288
CHAPTER 9: CONCLUSIONS
9.1 Summary of salient observations
The current series of experiments contribute to a growing understanding of the neurobiological underpinnings of age-associated cognitive decline.
Additionally, in an attempt to elucidate how the perirhinal cortex (PRC) might be altered during normal aging, the present results appear to require a significant re- evaluation of the prevailing theory of the PRC in stimulus recognition in young animals. Finally, these experiments resulted in the novel observation that 3- dimensional objects modulated both PRC neuron activity and middle hippocampal CA1 firing patterns. The primary contributions of this dissertation are as follows:
1. It is traditionally believed that aged rats do not express preference for
novelty in their object exploration behavior at long delays because they
‘forget’ the familiar stimulus (Bartolini et al., 1996; de Lima et al., 2005;
Platano et al., 2008). The data described in Chapter 6 indicate that this is
not the case. In fact, it was demonstrated that the aged rats behave as if
the novel objects are familiar. This observation is consistent with the
behavior expressed by young rats with PRC lesions (McTighe et al.,
2008). These data call into question the prevailing idea that the primary
contribution of the PRC is long-term object recognition memory (Brown
and Aggleton, 2001), and supports the perceptual-mnemonic feature 289
conjunction idea of Murray and Bussey (1999) that was formally modeled
by Cowell et al., (2006).
2. When young and aged rats traverse a track with 3-dimensional objects,
the primary behavioral correlate of PRC single-unit activity was increased
spiking at the location of objects, and at the food dishes. We termed these
patterns of activity ‘object fields’ and ‘food dish fields’, respectively
(Chapter 7). Moreover, the proportion of neurons with increased activity at
both food dishes was greater than expected by chance. These cells that
were active at both food dishes regardless of their locations may have
been responding to the reward itself. Although a portion of food dish fields
occurred at both food dishes, the observation that others only occurred at
one of two identical food dishes suggests a possible
location/object/reward interaction.
3. Contrary to previous electrophysiological recordings from rats (Zhu and
Brown, 1995; Zhu et al., 1995), we did not observe a response decrement
in PRC neurons as objects went from being novel to familiar (Chapter 7).
Possible reasons for this discrepancy include behavioral and
methodological differences between this current experiment and previous
studies. Specifically, in the current study rats were moving through space
and actively exploring objects, as opposed to nose poking while a stimulus 290
was presented. Additionally, the current experiment used chronic
recordings as opposed to acute, anesthetized conditions. The use of
chronic recording devices may minimize sampling bias confounds that can
result during acute conditions in which there is a tendency to sample
higher firing rate neurons. The observation that PRC neurons in both
young and aged rats did not show differential firing rates between novel
and familiar objects suggests that a response decrement is not the neural
correlate of a familiarity signal.
4. A salient difference in the PRC activity characteristics between young and
aged rats was that a lower proportion of old PRC neurons expressed
objects fields and food dish fields compared to young PRC neurons
(Chapter 7). Because fewer aged PRC cells had object or food dish fields,
this resulted in the aged rats having overall lower mean firing rate
compared to the young animals in all behavioral conditions. This
observation suggests that the aged PRC may be under tighter inhibitory
control compared to the young PRC (Chapter 6).
5. In both young and aged rats, the firing rate of PRC neurons was not
significantly modulated by the running velocity of the animal (Chapter 7).
This is a striking difference from neurons in the hippocampus
(McNaughton et al., 1983) and medial entorhinal cortex (Sargolini et al., 291
2006; Maurer, personal communication), supporting an idea that there are
two parallel streams of input that project to the hippocampus (Hargreaves
et al., 2005; Furtak et al., 2007). One stream is relayed through the
postrhinal and medial entorhinal cortices, and is more concerned with
information related to self-motion and space. The other stream involves
the PRC and lateral entorhinal cortex and is more involved with perception
and the identification of ‘what’ stimuli are in a space. This is somewhat
analogous to the dorsal and ventral visual streams that are associated
with ‘action’ and ‘perception’, respectively (Milner and Goodale, 1993).
Importantly, the postrhinal/medial entorhinal cortices receive more input
from the dorsal stream while the PRC/lateral entorhinal cortex is more
anatomically connected with the ventral visual stream.
6. In young animals putative modulation of PRC activity, by enriching the
track with objects, lead to a decrease in the size of place fields in neurons
recorded from the distal region of middle CA1 (Chapter 8). This indicates
that hippocampal spatial representations can be modulated significantly by
3-dimensional objects.
7. Adding objects to the track also leads to an increase in the proportion of
middle CA1 neurons that express place fields in young rats (Chapter 8).
These data lend further support to the notion that there are two distinct 292
streams of input to the hippocampus that can result in different forms of
modulation of CA1 activity patterns. The affect of the dorsal/action stream
has been well characterized, and the effects of the ventral/perceptual
stream affects are reported here.
9.2 Suggestions for future experiments
It is my hope that the completion of this dissertation work will mark the beginning of a previously under-explored avenue of research. The current data have lead me to the conclusion that an entirely novel approach needs to be taken to understand how aging of the PRC may affect perception versus memory.
Below is a brief list of experiments that I would like to see conducted if given access to endless time and resources. Hopefully, one or two of these may further advance our understanding of aging and PRC function.
9.2.1 The effect of perceptual difficulty on the performance of aged rats
In both monkeys (Murray and Bussey, 1999; Bussey et al., 2002; Bussey et al., 2003) and rats (Norman and Eacott, 2004; Bartko et al., 2007; Bartko et al., 2007; Forwood et al., 2007), the role of the PRC in disambiguating distinct stimuli with overlapping features is well established. The observation that aged rats are able to recognize stimuli that are repeated but do not respond to novel stimuli with increased exploratory behavior to the same extent as young animals, as discussed in Chapter 6, mirrors recent PRC lesion data (McTighe et al., 293
2008). These data are at lest consistent with the hypothesis that aged rats may have difficulty resolving feature ambiguity.
The perceptual oddity task and the standard object recognition task with a
0 second delay can reliably measure the impact of stimulus feature ambiguity on perceptual discrimination in rats with PRC lesions (Bartko et al., 2007). The obvious next step, with respect to the aging issue, is to test aged rats on these tasks and to measure the extent to which their performance differs from young animals. In fact, I regret lacking the foresight to use these tasks as a behavioral screen before conducting electrophysiological recordings. The potential to measure a possible correlation between the proportion of PRC neurons that express object fields and a rat’s ability disambiguate stimuli with overlapping features is extremely alluring.
9.2.2 Pattern separation in the perirhinal cortex
The belief that the PRC operates to disambiguate stimuli with overlapping features (Murray and Bussey, 1999; Bussey et al., 2002; Bussey et al., 2005;
Bussey et al., 2006; Cowell et al., 2006) suggests that this network has the capacity to pattern separate. That is, stimuli that are very similar, presumably activating overlapping input to the PRC, can be represented as more dissimilar if
PRC ensemble activity can assist in orthogonalizing these inputs. In the hippocampus, pattern separation has been examined using experiments in which the environment incrementally changes, or is “morphed” (often from a circle to a 294 square arena, or vice versa) (Leutgeb et al., 2005; Wills et al., 2005; Leutgeb et al., 2007). Using similar logic it could be fruitful to record from PRC neurons while two distinct visual stimuli are morphed into each other. Also, of interest would be to examine a potential correlation between when a PRC neuron detects a perceptual change versus when the animal can behaviorally indicate the change.
Admittedly, this experiment may be better suited for non-human primates, or least an animal with a more sophisticated visual system than a rat.
9.2.3 Behavioral relevance and perirhinal cortical activity
Unfortunately, the PRC electrophysiology experiment (Chapter 7) was designed specifically to address whether or not there were differences in the response decrement of PRC neurons between young and old rats. Although having rats traverse a track with many objects spaced relatively close together may have been an optimal design for detecting a response decrement (had one occurred), it was not ideal for answering other questions. Specifically, other than the food dishes, the objects were not relevant to the rats’ behavior except that they forced them to alter their path slightly. Of particular interest to me is how behavioral relevancy modulates object field expression in young and old rats.
There are several reasons to assume the behavioral salience of a stimulus affects PRC neural representations. First, it has been shown that more PRC neurons respond to an auditory stimulus after it has been paired with a foot shock (Allen et al., 2007). Moreover, in the current study a large proportion of 295
PRC neurons (~35% in young rats and ~21% in aged rats) were active at the food dishes. I would be interested to know if an object would provoke more PRC neuronal activity if it was paired with something salient to the rat’s behavior. An easy manipulation to address the above question would be to associate the food reward with a particular object that would not be fixed at the same spatial location. Additionally, experiments are currently underway in which, after several days of the food dish being paired with a food reward, the food reward is no longer given in the dish. It will be interesting to observe whether or not PRC neuron activity related to the food dish is altered after that object is no longer associated with a reward.
9.2.4 The specificity of object fields and the role of spatial location
As discussed in Chapter 7, the behavioral procedures used for the electrophysiological recordings from the PRC were not optimized to evaluate the parameters that control object and food dish fields. Particularly, the specificity of object fields and the contribution of spatial location remain to be determined. The specificity of object fields can easily be assessed by having rats run two behavioral epochs with different sets of objects. During the first epoch, 6 objects will be placed on track and the rats will run laps around these objects as they did in the previous experiment. During the second epoch, however, the objects will be replaced by 6 distinct novel objects. If the PRC activity patterns are not correlated between epochs, then it would appear that the object fields are 296 selective for particular objects. If the activity patterns in the PRC are not different after all of the objects are replaced then different conclusions would have to be made. This would indicate that any item on the track that compels the rat to alter its path (even slightly) can induce PRC activity selective for the region of space the object is at, but not for the object itself. If this result is observed, then discarding the name “object fields” would be sensible.
Two observations suggest that neurons in the PRC may be selective for the conjunction of a stimulus in a particular spatial location. First, many food dish fields only occurred at one food dish. Second, the reconfiguration of objects resulted in reduced firing rates in a portion of PRC neurons, and a decline in the activity correlations of PRC neurons between epochs. Although the PRC is often regarded as the highest level of association cortex in the ventral visual processing stream (“the what pathway”) (for review, see Murray and Bussey,
1999; Murray et al., 2000; Murray et al., 2007), it also receives a strong projection from the postrhinal cortex (parahippocampal cortex in primates;
Burwell and Amaral, 1998b). Because many neocortical regions associated with visuo-spatial information send afferent input to postrhinal cortex (Burwell and
Amaral, 1998a), it is likely the PRC receives spatial information.
To examine whether spatial location modulates the expression of object fields, if rats traverse at track with 3 pairs of identical objects, the extent to which object fields are observed at both identical objects can be used to determine the influence of space. Between epochs 1 and 2 the positions of identical objects 297 could be swapped to control for variables other than space that the rat may use to distinguish the matching objects from each other.
9.2.5 The impact of objects on CA1 activity patterns in the aged middle hippocampus
A lab maxim is that “science is long but life is short”. Over the years, I found this to become increasing true as there is often not enough time to complete all of the experiments that one would like to. This applies particularly to the experiment discussed in Chapter 8, which showed that the activity of middle
CA1 neurons was affected by adding objects to the track. The observation that old rats show a lower proportion of object fields compared to young animals implies that cells recorded in middle CA1 may show less of an effect of object- modulation on place field size and expression in old rats than was observed in the young rats. Therefore, I would really like to see this experiment extended using both young and aged rats.
9.2.6 The impact of objects on hippocampal map stability
In Chapter 8 the effect of objects on hippocampal map stability was described. In young rats place field activity was highly correlated between two epochs of track running for the no objects, both epochs condition and the objects, both epochs condition. There was a non-significant pattern, however, for the recorded CA1 neurons to be more correlated between epochs when there were 298 objects on the track versus when the track was empty. Figure 9.1 shows the normalized frequency distributions of the firing rate correlations between epoch 1 and epoch 2 for the no objects, both epochs condition (black), and the objects, both epochs conditions (blue). In Figure 9.1 it appears that the CA1 neurons may have an increased frequency of higher correlation values when objects are on the track. This was not significant, however, and it is likely that the near-ceiling level correlation values in the young rats during the no objects, both conditions made it difficult to detect an influence of objects on place map stability. This is
0.2 No objects, both epochs 0.18 Objects, both epochs 0.16 Objects, both epochs
0.14
0.12
0.1
0.08
0.06 Normalized proportion
0.04
0.02
0 0 0.2 0.4 0.6 0.8 1 Correlation (r) Figure 9.1: CA1 activity correlations between epochs. The normalized frequency distributions of correlation values obtained from CA1 neurons recorded the in distal middle region of the hippocampus. The black line is for neurons that were recorded during a no objects, both epochs condition, and the blue is for cells recorded during the objects both epochs conditions. 299 consistent with the observation made in young rats, where CA1 neurons do not change their spatial selectively between epochs (e.g., Thompson and Best,
1990) unless the environment (Muller and Kubie, 1987) or behavioral demands of a task (Markus et al., 1995) change substantially. In contrast, hippocampal maps in old rats can be unstable between episodes of behavior in a familiar environment (see Chapter 3). If objects also influence CA1 firing characteristics in old rats (admittedly they may not influence CA1 to the degree observed in young animals, see above), it is possible that objects may help to ‘stabilize’ the aged place map and this question should be addressed in concert with the experiment proposed above.
9.3 Conclusions
When the experiments described in Chapters 6 and 7 were first planned, it was my hypothesis that the observed impairments on the spontaneous object recognition task in aged rats were due to memory deficits. It was also my belief that a likely explanation for recognition memory deficits in aged rats would be an alteration in the well-described perirhinal neuron response decrement phenomenon. This dissertation work did not confirm either of these hypotheses.
Around the time that these experiments were beginning, Andrew Maurer’s thesis research was starting to confirm the observation that place fields of ventral hippocampal CA1 neurons are larger relative to place fields of neurons in dorsal
CA1 (Maurer et al., 2006; Maurer et al., 2005; Jung et al., 1994). Due to a 300 fortunate accident, several of the tetrodes in the first two rats I implanted with a hyperdrive aimed at the perirhinal cortex ended up in the ventral hippocampus.
Much to my surprise, I observed that ventral CA1 pyramidal cell place fields could be small under conditions with objects. After this initial observation, it was also revealed that PRC neurons exhibited activity at the location of objects.
Therefore, this thesis represents the evolution of my ideas regarding the function of the perirhinal cortex and how this structure may modulate activity in other medial temporal lobe areas.
My current theory is that aged rats show object recognition deficits because they are less able to distinguish new stimuli from those that have been previously encountered (Chapter 6). This deficit is exacerbated when stimuli share common features. The aged compared to the young PRC may be less able to disambiguate (i.e., pattern separate) distinct stimuli because of the decreased proportion PRC neurons that are activated by stimuli (Chapter 7). Finally, the data described in Chapter 8 suggest that input from the PRC may influence hippocampal activity patterns along with self-motion input from the medial entorhinal cortex (for review, see McNaughton et al., 2006). This can possibly account for the involvement of the hippocampus in recollection (Fortin et al.,
2004) and episodic memory (e.g., Nadel et al., 1985; Nadel and Hardt, 2004;
Eichenbaum et al., 2007), both of which require the association of ‘what’ stimulus was encountered with ‘where’ it was experienced. I look forward to future data that may possibly confirm these ideas, or more likely refute them. 301
References
Abe, H.,Ishida, Y. and Iwasaki, T. (2004). Perirhinal N-methyl-D-aspartate and muscarinic systems participate in object recognition in rats. Neuroscience Letters 356(3): 191-4. Adams, M.M.,Shi, L.,Linville, M.C.,Forbes, M.E.,Long, A.B.,Bennett, C.,Newton, I.G.,Carter, C.S.,Sonntag, W.E.,Riddle, D.R. and Brunso-Bechtold, J.K. (2008). Caloric restriction and age affect synaptic proteins in hippocampal CA3 and spatial learning ability. Experimental Neurology 211(1): 141-9. Aggleton, J.P. and Brown, M.W. (1999). Episodic memory, amnesia, and the hippocampal-anterior thalamic axis. Behav Brain Sci 22(3): 425-44; discussion 444-89. Aggleton, J.P.,Kyd, R.J. and Bilkey, D.K. (2004). When is the perirhinal cortex necessary for the performance of spatial memory tasks? Neuroscience and biobehavioral reviews 28(6): 611-24. Agnihotri, N.T.,Hawkins, R.D.,Kandel, E.R. and Kentros, C. (2004). The long- term stability of new hippocampal place fields requires new protein synthesis. Proceedings of the National Academy of Sciences, U S A 101(10): 3656-61. Allen, T.A.,Furtak, S.C. and Brown, T.H. (2007). Single-unit responses to 22 kHz ultrasonic vocalizations in rat perirhinal cortex. Behavioural Brain Research 182(2): 327-36. Amaral, D. and Witter, M. (1995). Hippocampal Formation. The Rat Nervous System. G, P. San Diego, Academic Press: 443-486. Amaral, D.G. and Lavenex, P. (2007). Hippocampal Neuroanatomy. The Hippocampus Book. Andersen, P., Morris, R.G., Amaral, D.G., Bliss, T.V. and O'Keefe, J. Oxford, Oxford University Press: 37-131. Apergis-Schoute, J.,Pinto, A. and Pare, D. (2007). Muscarinic control of long- range GABAergic inhibition within the rhinal cortices. Journal of Neuroscience 27(15): 4061-71. Ball, M.J. (1977). Neuronal loss, neurofibrillary tangles and granulovacuolar degeneration in the hippocampus with ageing and dementia. A quantitative study. Acta neuropathologica 37(2): 111-8. Barili, P.,De Carolis, G.,Zaccheo, D. and Amenta, F. (1998). Sensitivity to ageing of the limbic dopaminergic system: a review. Mechanisms of ageing and development 106(1-2): 57-92. Barnes, C.A. (1979). Memory deficits associated with senescence: a neurophysiological and behavioral study in the rat. Journal of Comparative of Physiological Psycholgy 93(1): 74-104. Barnes, C.A. (1994). Normal aging: regionally specific changes in hippocampal synaptic transmission. Trends in Neurosciences 17(1): 13-8. Barnes, C.A. and McNaughton, B.L. (1980). Physiological compensation for loss of afferent synapses in rat hippocampal granule cells during senescence. Journal of Physiology 309: 473-85. 302
Barnes, C.A.,McNaughton, B.L. and O'Keefe, J. (1983). Loss of place specificity in hippocampal complex spike cells of senescent rat. Neurobiology of Aging 4(2): 113-9. Barnes, C.A.,Rao, G.,Foster, T.C. and McNaughton, B.L. (1992). Region-specific age effects on AMPA sensitivity: electrophysiological evidence for loss of synaptic contacts in hippocampal field CA1. Hippocampus 2(4): 457-68. Barnes, C.A.,Rao, G. and Houston, F.P. (2000). LTP induction threshold change in old rats at the perforant path--granule cell synapse. Neurobiology of Aging 21(5): 613-20. Barnes, C.A.,Rao, G. and McNaughton, B.L. (1987). Increased electrotonic coupling in aged rat hippocampus: a possible mechanism for cellular excitability changes. Journal of Comparative Neurology 259(4): 549-58. Barnes, C.A.,Rao, G. and McNaughton, B.L. (1996). Functional integrity of NMDA-dependent LTP induction mechanisms across the lifespan of F-344 rats. Learning and Memory 3(2-3): 124-37. Barnes, C.A.,Rao, G. and Orr, G. (2000). Age-related decrease in the Schaffer collateral-evoked EPSP in awake, freely behaving rats. Neural Plasticity 7(3): 167-78. Barnes, C.A.,Rao, G. and Shen, J. (1997). Age-related decrease in the N-methyl- D-aspartateR-mediated excitatory postsynaptic potential in hippocampal region CA1. Neurobiology of Aging 18(4): 445-52. Barnes, C.A.,Suster, M.S.,Shen, J. and McNaughton, B.L. (1997). Multistability of cognitive maps in the hippocampus of old rats. Nature 388(6639): 272-5. Bartho, P.,Hirase, H.,Monconduit, L.,Zugaro, M.,Harris, K.D. and Buzsaki, G. (2004). Characterization of neocortical principal cells and interneurons by network interactions and extracellular features. Journal of Neurophysiology 92(1): 600-8. Bartko, S.J.,Winters, B.D.,Cowell, R.A.,Saksida, L.M. and Bussey, T.J. (2007). Perceptual functions of perirhinal cortex in rats: zero-delay object recognition and simultaneous oddity discriminations. Journal of Neuroscience 27(10): 2548-59. Bartko, S.J.,Winters, B.D.,Cowell, R.A.,Saksida, L.M. and Bussey, T.J. (2007). Perirhinal cortex resolves feature ambiguity in configural object recognition and perceptual oddity tasks. Learning and Memory 14(12): 821-32. Bartolini, L.,Casamenti, F. and Pepeu, G. (1996). Aniracetam restores object recognition impaired by age, scopolamine, and nucleus basalis lesions. Pharmacology, Biochemistry and Behavior 53(2): 277-83. Bastin, C. and Van der Linden, M. (2003). The contribution of recollection and familiarity to recognition memory: a study of the effects of test format and aging. Neuropsychology 17(1): 14-24. Battaglia, F.P.,Sutherland, G.R. and McNaughton, B.L. (2004). Hippocampal sharp wave bursts coincide with neocortical "up-state" transitions. Learning and Memory 11(6): 697-704. 303
Battaglia, F.P.,Sutherland, G.R. and McNaughton, B.L. (2004). Local sensory cues and place cell directionality: additional evidence of prospective coding in the hippocampus. Journal of Neuroscience 24(19): 4541-50. Baxter, M.G. and Murray, E.A. (2001). Effects of hippocampal lesions on delayed nonmatching-to-sample in monkeys: a reply to Zola and Squire (2001). Hippocampus 11(3): 201-3. Baxter, M.G. and Murray, E.A. (2001). Impairments in visual discrimination learning and recognition memory produced by neurotoxic lesions of rhinal cortex in rhesus monkeys. European Journal of Neuroscience 13(6): 1228-38. Baxter, M.G. and Murray, E.A. (2001). Opposite relationship of hippocampal and rhinal cortex damage to delayed nonmatching-to-sample deficits in monkeys. Hippocampus 11(1): 61-71. Bear, M.F.,Cooper, L.N. and Ebner, F.F. (1987). A physiological basis for a theory of synapse modification. Science 237(4810): 42-8. Bear, M.F. and Malenka, R.C. (1994). Synaptic plasticity: LTP and LTD. Current Opinion in Neurobiology 4(3): 389-99. Beggs, J.M.,Moyer, J.R., Jr.,McGann, J.P. and Brown, T.H. (2000). Prolonged synaptic integration in perirhinal cortical neurons. Journal of Neurophysiology 83(6): 3294-8. Bilkey, D.K. (1996). Long-term potentiation in the in vitro perirhinal cortex displays associative properties. Brain Research 733(2): 297-300. Billard, J.M. and Rouaud, E. (2007). Deficit of NMDA receptor activation in CA1 hippocampal area of aged rats is rescued by D-cycloserine. European Journal of Neuroscience 25(8): 2260-8. Bliss, T.V. and Collingridge, G.L. (1993). A synaptic model of memory: long-term potentiation in the hippocampus. Nature 361(6407): 31-9. Bliss, T.V. and Lomo, T. (1973). Long-lasting potentiation of synaptic transmission in the dentate area of the anaesthetized rabbit following stimulation of the perforant path. Journal of Physiology 232(2): 331-56. Bogacz, R. and Brown, M.W. (2002). The restricted influence of sparseness of coding on the capacity of familiarity discrimination networks. Network 13(4): 457-85. Bogacz, R. and Brown, M.W. (2003). Comparison of computational models of familiarity discrimination in the perirhinal cortex. Hippocampus 13(4): 494- 524. Bondareff, W. and Geinisman, Y. (1976). Loss of synapses in the dentate gyrus of the senescent rat. American journal of anatomy 145(1): 129-36. Boric, K.,Munoz, P.,Gallagher, M. and Kirkwood, A. (2008). Potential adaptive function for altered long-term potentiation mechanisms in aging hippocampus. Journal of Neuroscience 28(32): 8034-9. Brizzee, K.R.,Ordy, J.M. and Bartus, R.T. (1980). Localization of cellular changes within multimodal sensory regions in aged monkey brain: possible 304
implications for age-related cognitive loss. Neurobiology of Aging 1(1): 45- 52. Broadbent, N.J.,Squire, L.R. and Clark, R.E. (2004). Spatial memory, recognition memory, and the hippocampus. Proc Natl Acad Sci U S A 101(40): 14515- 20. Brody, H. (1955). Organization of the cerebral cortex. III. A study of aging in the human cerebral cortex. Journal of Comparative Neurology 102(2): 511-6. Brown, M.W. (2008). Hippocampal and perirhinal functions in recognition memory. Nature Reveiws Neuroscience 9(5): 405; author reply 405. Brown, M.W. and Aggleton, J.P. (2001). Recognition memory: what are the roles of the perirhinal cortex and hippocampus? Nature Reveiws Neuroscience 2(1): 51-61. Brown, M.W.,Wilson, F.A. and Riches, I.P. (1987). Neuronal evidence that inferomedial temporal cortex is more important than hippocampus in certain processes underlying recognition memory. Brain Research 409(1): 158-62. Brozinsky, C.J.,Yonelinas, A.P.,Kroll, N.E. and Ranganath, C. (2005). Lag- sensitive repetition suppression effects in the anterior parahippocampal gyrus. Hippocampus 15(5): 557-61. Brun, V.H.,Leutgeb, S.,Wu, H.Q.,Schwarcz, R.,Witter, M.P.,Moser, E.I. and Moser, M.B. (2008). Impaired spatial representation in CA1 after lesion of direct input from entorhinal cortex. Neuron 57(2): 290-302. Brun, V.H.,Otnass, M.K.,Molden, S.,Steffenach, H.A.,Witter, M.P.,Moser, M.B. and Moser, E.I. (2002). Place cells and place recognition maintained by direct entorhinal-hippocampal circuitry. Science 296(5576): 2243-6. Bucci, D.J.,Phillips, R.G. and Burwell, R.D. (2000). Contributions of postrhinal and perirhinal cortex to contextual information processing. Behavioral Neuroscience 114(5): 882-94. Buckley, M.J. (2005). The role of the perirhinal cortex and hippocampus in learning, memory, and perception. Quarterly journal of experimental psychology. B, Comparative and physiological psychology 58(3-4): 246- 68. Buckley, M.J. and Gaffan, D. (2006). Perirhinal cortical contributions to object perception. Trends in Cognitive Sciences 10(3): 100-7. Buell, S.J. and Coleman, P.D. (1979). Dendritic growth in the aged human brain and failure of growth in senile dementia. Science 206(4420): 854-6. Buell, S.J. and Coleman, P.D. (1981). Quantitative evidence for selective dendritic growth in normal human aging but not in senile dementia. Brain Research 214(1): 23-41. Buffalo, E.A.,Ramus, S.J.,Clark, R.E.,Teng, E.,Squire, L.R. and Zola, S.M. (1999). Dissociation between the effects of damage to perirhinal cortex and area TE. Learning and Memory 6(6): 572-99. 305
Buffalo, E.A.,Ramus, S.J.,Squire, L.R. and Zola, S.M. (2000). Perception and recognition memory in monkeys following lesions of area TE and perirhinal cortex. Learning and Memory 7(6): 375-82. Buffalo, E.A.,Reber, P.J. and Squire, L.R. (1998). The human perirhinal cortex and recognition memory. Hippocampus 8(4): 330-9. Bunzeck, N. and Duzel, E. (2006). Absolute coding of stimulus novelty in the human substantia nigra/VTA. Neuron 51(3): 369-79. Bunzeck, N.,Schutze, H.,Stallforth, S.,Kaufmann, J.,Duzel, S.,Heinze, H.J. and Duzel, E. (2007). Mesolimbic novelty processing in older adults. Cerebral Cortex 17(12): 2940-8. Burgess, N. and O'Keefe, J. (1996). Neuronal computations underlying the firing of place cells and their role in navigation. Hippocampus 6(6): 749-62. Burke, S.N. and Barnes, C.A. (2006). Neural plasticity in the ageing brain. Nature Reveiws Neuroscience 7(1): 30-40. Burke, S.N.,Maurer, A.P.,Navratilova, Z.,McNaughton, B.L. and Barnes, C.A. (2007). Young and Old Rats Show a Full Cycle of Theta Phase Precession During the First Pass Through a Place Field in a Familiar Environment. Society for Neuroscience Abstracts. Burke, S.N.,Maurer, A.P.,Zhiyoung, Y.,Navratilova, Z. and Barnes, C.A. (2008). Glutamate receptor-mediated restoration of experience-dependent place field expansion plasticity in aged rats. Behavioral Neuroscience in press. Burwell, R.D. (2000). The parahippocampal region: corticocortical connectivity. Annals of the New York Academy of Sciences 911: 25-42. Burwell, R.D. (2001). Borders and cytoarchitecture of the perirhinal and postrhinal cortices in the rat. Journal of Comparative Neurology 437(1): 17-41. Burwell, R.D. and Amaral, D.G. (1998). Cortical afferents of the perirhinal, postrhinal, and entorhinal cortices of the rat. Journal of Comparative Neurology 398(2): 179-205. Burwell, R.D. and Amaral, D.G. (1998). Perirhinal and postrhinal cortices of the rat: interconnectivity and connections with the entorhinal cortex. Journal of Comparative Neurology 391(3): 293-321. Burwell, R.D. and Amaral, D.G. (1998a). Cortical afferents of the perirhinal, postrhinal, and entorhinal cortices of the rat. Journal of Comparative Neurology 398(2): 179-205. Burwell, R.D.,Witter, M.P. and Amaral, D.G. (1995). Perirhinal and postrhinal cortices of the rat: a review of the neuroanatomical literature and comparison with findings from the monkey brain. Hippocampus 5(5): 390- 408. Bussey, T.J.,Duck, J.,Muir, J.L. and Aggleton, J.P. (2000). Distinct patterns of behavioural impairments resulting from fornix transection or neurotoxic 306
lesions of the perirhinal and postrhinal cortices in the rat. Behavioural Brain Research 111(1-2): 187-202. Bussey, T.J.,Muir, J.L. and Aggleton, J.P. (1999). Functionally dissociating aspects of event memory: the effects of combined perirhinal and postrhinal cortex lesions on object and place memory in the rat. Journal of Neuroscience 19(1): 495-502. Bussey, T.J. and Saksida, L.M. (2002). The organization of visual object representations: a connectionist model of effects of lesions in perirhinal cortex. European Journal of Neuroscience 15(2): 355-64. Bussey, T.J.,Saksida, L.M. and Murray, E.A. (2002). Perirhinal cortex resolves feature ambiguity in complex visual discriminations. European Journal of Neuroscience 15(2): 365-74. Bussey, T.J.,Saksida, L.M. and Murray, E.A. (2003). Impairments in visual discrimination after perirhinal cortex lesions: testing 'declarative' vs. 'perceptual-mnemonic' views of perirhinal cortex function. European Journal of Neuroscience 17(3): 649-60. Bussey, T.J.,Saksida, L.M. and Murray, E.A. (2005). The perceptual- mnemonic/feature conjunction model of perirhinal cortex function. Quarterly journal of experimental psychology. B, Comparative and physiological psychology 58(3-4): 269-82. Bussey, T.J.,Saksida, L.M. and Murray, E.A. (2006). Perirhinal cortex and feature-ambiguous discriminations. Learning and Memory 13(2): 103-5; author reply 106-7. Buzsaki, G. (1986). Hippocampal sharp waves: their origin and significance. Brain Research 398(2): 242-52. Buzsaki, G. (1989). Two-stage model of memory trace formation: a role for "noisy" brain states. Neuroscience 31(3): 551-70. Buzsaki, G.,Csicsvari, J.,Dragoi, G.,Harris, K.,Henze, D. and Hirase, H. (2002). Homeostatic maintenance of neuronal excitability by burst discharges in vivo. Cerebral Cortex 12(9): 893-9. Campbell, L.W.,Hao, S.Y.,Thibault, O.,Blalock, E.M. and Landfield, P.W. (1996). Aging changes in voltage-gated calcium currents in hippocampal CA1 neurons. Journal of Neuroscience 16(19): 6286-95. Canning, K.J. and Leung, L.S. (1997). Lateral entorhinal, perirhinal, and amygdala-entorhinal transition projections to hippocampal CA1 and dentate gyrus in the rat: a current source density study. Hippocampus 7(6): 643-55. Cate, A.D. and Kohler, S. (2006). The missing whole in perceptual models of perirhinal cortex. Trends in Cognitive Sciences 10(9): 396-7. Cavoy, A. and Delacour, J. (1993). Spatial but not object recognition is impaired by aging in rats. Physiology and Behavior 53(3): 527-30. Cho, K.,Kemp, N.,Noel, J.,Aggleton, J.P.,Brown, M.W. and Bashir, Z.I. (2000). A new form of long-term depression in the perirhinal cortex. Nature Neuroscience 3(2): 150-6. 307
Chowdhury, S.,Shepherd, J.D.,Okuno, H.,Lyford, G.,Petralia, R.S.,Plath, N.,Kuhl, D.,Huganir, R.L. and Worley, P.F. (2006). Arc/Arg3.1 interacts with the endocytic machinery to regulate AMPA receptor trafficking. Neuron 52(3): 445-59. Chrobak, J.J. and Buzsaki, G. (1994). Selective activation of deep layer (V-VI) retrohippocampal cortical neurons during hippocampal sharp waves in the behaving rat. Journal of Neuroscience 14(10): 6160-70. Chrobak, J.J. and Buzsaki, G. (1996). High-frequency oscillations in the output networks of the hippocampal-entorhinal axis of the freely behaving rat. Journal of Neuroscience 16(9): 3056-66. Clark, R.E.,West, A.N.,Zola, S.M. and Squire, L.R. (2001). Rats with lesions of the hippocampus are impaired on the delayed nonmatching-to-sample task. Hippocampus 11(2): 176-86. Cole, A.J.,Saffen, D.W.,Baraban, J.M. and Worley, P.F. (1989). Rapid increase of an immediate early gene messenger RNA in hippocampal neurons by synaptic NMDA receptor activation. Nature 340(6233): 474-6. Coleman, P.D. and Flood, D.G. (1987). Neuron numbers and dendritic extent in normal aging and Alzheimer's disease. Neurobiology of Aging 8(6): 521- 45. Corkin, S. (2002). What's new with the amnesic patient H.M.? Nature Reveiws Neuroscience 3(2): 153-60. Cowell, R.A.,Bussey, T.J. and Saksida, L.M. (2006). Why does brain damage impair memory? A connectionist model of object recognition memory in perirhinal cortex. Journal of Neuroscience 26(47): 12186-97. Curcio, C.A. and Hinds, J.W. (1983). Stability of synaptic density and spine volume in dentate gyrus of aged rats. Neurobiology of Aging 4(1): 77-87. Daselaar, S.M.,Fleck, M.S. and Cabeza, R. (2006). Triple dissociation in the medial temporal lobes: recollection, familiarity, and novelty. Journal of Neurophysiology 96(4): 1902-11. Daselaar, S.M.,Fleck, M.S.,Dobbins, I.G.,Madden, D.J. and Cabeza, R. (2006). Effects of healthy aging on hippocampal and rhinal memory functions: an event-related fMRI study. Cerebral Cortex 16(12): 1771-82. Davachi, L.,Mitchell, J.P. and Wagner, A.D. (2003). Multiple routes to memory: distinct medial temporal lobe processes build item and source memories. Proceedings of the National Academy of Sciences, U S A 100(4): 2157- 62. Davidson, P.S. and Glisky, E.L. (2002). Neuropsychological correlates of recollection and familiarity in normal aging. Cogn Affect Behav Neurosci 2(2): 174-86. de Curtis, M. and Pare, D. (2004). The rhinal cortices: a wall of inhibition between the neocortex and the hippocampus. Prog Neurobiol 74(2): 101-10. de Lima, M.N.,Laranja, D.C.,Caldana, F.,Bromberg, E.,Roesler, R. and Schroder, N. (2005). Reversal of age-related deficits in object recognition memory in rats with l-deprenyl. Experimental Gerontology 40(6): 506-11. 308
Dennis, N.A.,Hayes, S.M.,Prince, S.E.,Madden, D.J.,Huettel, S.A. and Cabeza, R. (2008). Effects of aging on the neural correlates of successful item and source memory encoding. J Exp Psychol Learn Mem Cogn 34(4): 791- 808. Dennis, N.A.,Kim, H. and Cabeza, R. (2007). Effects of aging on true and false memory formation: an fMRI study. Neuropsychologia 45(14): 3157-66. Dennis, N.A.,Kim, H. and Cabeza, R. (2008). Age-related differences in brain activity during true and false memory retrieval. J Cogn Neurosci 20(8): 1390-402. Deshmukh, S.S. and Knierim, J.J. (2008). Properties of lateral entorhinal cortex neurons in an environment with discrete objects Society for Neuroscience Abstracts. deToledo-Morrell, L.,Geinisman, Y. and Morrell, F. (1988). Age-dependent alterations in hippocampal synaptic plasticity: relation to memory disorders. Neurobiology of Aging 9(5-6): 581-90. Deupree, D.L.,Bradley, J. and Turner, D.A. (1993). Age-related alterations in potentiation in the CA1 region in F344 rats. Neurobiology of Aging 14(3): 249-58. Deupree, D.L.,Turner, D.A. and Watters, C.L. (1991). Spatial performance correlates with in vitro potentiation in young and aged Fischer 344 rats. Brain Research 554(1-2): 1-9. Dhanrajan, T.M.,Lynch, M.A.,Kelly, A.,Popov, V.I.,Rusakov, D.A. and Stewart, M.G. (2004). Expression of long-term potentiation in aged rats involves perforated synapses but dendritic spine branching results from high- frequency stimulation alone. Hippocampus 14(2): 255-64. Diana, G.,Domenici, M.R.,Loizzo, A.,Scotti de Carolis, A. and Sagratella, S. (1994). Age and strain differences in rat place learning and hippocampal dentate gyrus frequency-potentiation. Neuroscience Letters 171(1-2): 113- 6. Diana, G.,Scotti de Carolis, A.,Frank, C.,Domenici, M.R. and Sagratella, S. (1994). Selective reduction of hippocampal dentate frequency potentiation in aged rats with impaired place learning. Brain Res Bull 35(2): 107-11. Dieguez, D., Jr. and Barea-Rodriguez, E.J. (2004). Aging impairs the late phase of long-term potentiation at the medial perforant path-CA3 synapse in awake rats. Synapse 52(1): 53-61. Dix, S.L. and Aggleton, J.P. (1999). Extending the spontaneous preference test of recognition: evidence of object-location and object-context recognition. Behavioural Brain Research 99(2): 191-200. Dunn, J.C. (2004). Remember-know: a matter of confidence. Psychol Rev 111(2): 524-42. Dunnett, S.B.,Evenden, J.L. and Iversen, S.D. (1988). Delay-dependent short- term memory deficits in aged rats. Psychopharmacology (Berl) 96(2): 174- 80. 309
Eacott, M.J.,Gaffan, D. and Murray, E.A. (1994). Preserved recognition memory for small sets, and impaired stimulus identification for large sets, following rhinal cortex ablations in monkeys. European Journal of Neuroscience 6(9): 1466-78. Eacott, M.J. and Gaffan, E.A. (2005). The roles of perirhinal cortex, postrhinal cortex, and the fornix in memory for objects, contexts, and events in the rat. Quarterly journal of experimental psychology. B, Comparative and physiological psychology 58(3-4): 202-17. Eacott, M.J. and Norman, G. (2004). Integrated memory for object, place, and context in rats: a possible model of episodic-like memory? Journal of Neuroscience 24(8): 1948-53. Edling, Y.,Ingelman-Sundberg, M. and Simi, A. (2007). Glutamate activates c-fos in glial cells via a novel mechanism involving the glutamate receptor subtype mGlu5 and the transcriptional repressor DREAM. Glia 55(3): 328- 40. Eichenbaum, H. (2000). A cortical-hippocampal system for declarative memory. Nature Reveiws Neuroscience 1(1): 41-50. Eichenbaum, H.,Yonelinas, A.P. and Ranganath, C. (2007). The medial temporal lobe and recognition memory. Annual review of neuroscience 30: 123-52. Ekstrom, A.D. and Bookheimer, S.Y. (2007). Spatial and temporal episodic memory retrieval recruit dissociable functional networks in the human brain. Learning and Memory 14(10): 645-54. Ekstrom, A.D.,Meltzer, J.,McNaughton, B.L. and Barnes, C.A. (2001). NMDA receptor antagonism blocks experience-dependent expansion of hippocampal "place fields". Neuron 31(4): 631-8. Ennaceur, A. and Aggleton, J.P. (1997). The effects of neurotoxic lesions of the perirhinal cortex combined to fornix transection on object recognition memory in the rat. Behavioural Brain Research 88(2): 181-93. Ennaceur, A. and Delacour, J. (1988). A new one-trial test for neurobiological studies of memory in rats. 1: Behavioral data. Behavioural Brain Research 31(1): 47-59. Ennaceur, A.,Neave, N. and Aggleton, J.P. (1996). Neurotoxic lesions of the perirhinal cortex do not mimic the behavioural effects of fornix transection in the rat. Behavioural Brain Research 80(1-2): 9-25. Erickson, C.A. and Desimone, R. (1999). Responses of macaque perirhinal neurons during and after visual stimulus association learning. Journal of Neuroscience 19(23): 10404-16. Erickson, C.A.,Jagadeesh, B. and Desimone, R. (2000). Clustering of perirhinal neurons with similar properties following visual experience in adult monkeys. Nature Neuroscience 3(11): 1143-8. Fahy, F.L.,Riches, I.P. and Brown, M.W. (1993). Neuronal activity related to visual recognition memory: long-term memory and the encoding of recency and familiarity information in the primate anterior and medial 310
inferior temporal and rhinal cortex. Experimental Brain Research 96(3): 457-72. Faulkner, B. and Brown, T.H. (1999). Morphology and physiology of neurons in the rat perirhinal-lateral amygdala area. Journal of Comparative Neurology 411(4): 613-42. Felleman, D.J. and Van Essen, D.C. (1991). Distributed hierarchical processing in the primate cerebral cortex. Cerebral Cortex 1(1): 1-47. Fletcher, B.R.,Calhoun, M.E.,Rapp, P.R. and Shapiro, M.L. (2006). Fornix lesions decouple the induction of hippocampal arc transcription from behavior but not plasticity. Journal of Neuroscience 26(5): 1507-15. Flood, D.G. (1991). Region-specific stability of dendritic extent in normal human aging and regression in Alzheimer's disease. II. Subiculum. Brain Research 540(1-2): 83-95. Flood, D.G. (1993). Critical issues in the analysis of dendritic extent in aging humans, primates, and rodents. Neurobiology of Aging 14(6): 649-54. Flood, D.G.,Buell, S.J.,Defiore, C.H.,Horwitz, G.J. and Coleman, P.D. (1985). Age-related dendritic growth in dentate gyrus of human brain is followed by regression in the 'oldest old'. Brain Research 345(2): 366-8. Flood, D.G.,Buell, S.J.,Horwitz, G.J. and Coleman, P.D. (1987). Dendritic extent in human dentate gyrus granule cells in normal aging and senile dementia. Brain Research 402(2): 205-16. Flood, D.G.,Guarnaccia, M. and Coleman, P.D. (1987). Dendritic extent in human CA2-3 hippocampal pyramidal neurons in normal aging and senile dementia. Brain Research 409(1): 88-96. Fortin, N.J.,Wright, S.P. and Eichenbaum, H. (2004). Recollection-like memory retrieval in rats is dependent on the hippocampus. Nature 431(7005): 188- 91. Forwood, S.E.,Bartko, S.J.,Saksida, L.M. and Bussey, T.J. (2007). Rats spontaneously discriminate purely visual, two-dimensional stimuli in tests of recognition memory and perceptual oddity. Behavioral Neuroscience 121(5): 1032-42. Forwood, S.E.,Winters, B.D. and Bussey, T.J. (2005). Hippocampal lesions that abolish spatial maze performance spare object recognition memory at delays of up to 48 hours. Hippocampus 15(3): 347-55. Foster, D.J. and Wilson, M.A. (2006). Reverse replay of behavioural sequences in hippocampal place cells during the awake state. Nature 440(7084): 680- 3. Foster, T.C.,Barnes, C.A.,Rao, G. and McNaughton, B.L. (1991). Increase in perforant path quantal size in aged F-344 rats. Neurobiology of Aging 12(5): 441-8. Foster, T.C. and Norris, C.M. (1997). Age-associated changes in Ca(2+)- dependent processes: relation to hippocampal synaptic plasticity. Hippocampus 7(6): 602-12. 311
Furtak, S.C.,Allen, T.A. and Brown, T.H. (2007). Single-unit firing in rat perirhinal cortex caused by fear conditioning to arbitrary and ecological stimuli. Journal of Neuroscience 27(45): 12277-91. Furtak, S.C.,Moyer, J.R., Jr. and Brown, T.H. (2007). Morphology and ontogeny of rat perirhinal cortical neurons. Journal of Comparative Neurology 505(5): 493-510. Furtak, S.C.,Wei, S.M.,Agster, K.L. and Burwell, R.D. (2007). Functional neuroanatomy of the parahippocampal region in the rat: the perirhinal and postrhinal cortices. Hippocampus 17(9): 709-22. Fyhn, M.,Hafting, T.,Treves, A.,Moser, M.B. and Moser, E.I. (2007). Hippocampal remapping and grid realignment in entorhinal cortex. Nature 446(7132): 190-4. Fyhn, M.,Hafting, T.,Witter, M.P.,Moser, E.I. and Moser, M.B. (2008). Grid cells in mice. Hippocampus 18(12): 1230-8. Fyhn, M.,Molden, S.,Witter, M.P.,Moser, E.I. and Moser, M.B. (2004). Spatial representation in the entorhinal cortex. Science 305(5688): 1258-64. Gaffan, D. and Parker, A. (1996). Interaction of perirhinal cortex with the fornix- fimbria: memory for objects and "object-in-place" memory. J Neurosci 16(18): 5864-9. Gaffan, E.A.,Healey, A.N. and Eacott, M.J. (2004). Objects and positions in visual scenes: effects of perirhinal and postrhinal cortex lesions in the rat. Behavioral Neuroscience 118(5): 992-1010. Gallagher, M.,Burwell, R. and Burchinal, M. (1993). Severity of spatial learning impairment in aging: development of a learning index for performance in the Morris water maze. Behavioral Neuroscience 107(4): 618-26. Gazzaley, A.H.,Siegel, S.J.,Kordower, J.H.,Mufson, E.J. and Morrison, J.H. (1996). Circuit-specific alterations of N-methyl-D-aspartate receptor subunit 1 in the dentate gyrus of aged monkeys. Proceedings of the National Academy of Sciences, U S A 93(7): 3121-5. Gazzaley, A.H.,Thakker, M.M.,Hof, P.R. and Morrison, J.H. (1997). Preserved number of entorhinal cortex layer II neurons in aged macaque monkeys. Neurobiology of Aging 18(5): 549-53. Geinisman, Y.,Bondareff, W. and Dodge, J.T. (1977). Partial deafferentation of neurons in the dentate gyrus of the senescent rat. Brain Research 134(3): 541-5. Geinisman, Y.,de Toledo-Morrell, L. and Morrell, F. (1986). Loss of perforated synapses in the dentate gyrus: morphological substrate of memory deficit in aged rats. Proceedings of the National Academy of Sciences, U S A 83(9): 3027-31. Geinisman, Y.,de Toledo-Morrell, L.,Morrell, F.,Persina, I.S. and Rossi, M. (1992). Age-related loss of axospinous synapses formed by two afferent systems in the rat dentate gyrus as revealed by the unbiased stereological dissector technique. Hippocampus 2(4): 437-44. 312
Geinisman, Y.,Ganeshina, O.,Yoshida, R.,Berry, R.W.,Disterhoft, J.F. and Gallagher, M. (2004). Aging, spatial learning, and total synapse number in the rat CA1 stratum radiatum. Neurobiology of Aging 25(3): 407-16. Gerrard, J.L.,Burke, S.N.,McNaughton, B.L. and Barnes, C.A. (2008). Sequence reactivation in the hippocampus is impaired in aged rats. Journal of Neuroscience 28(31): 7883-90. Gerrard, J.L.,Kudrimoti, H.,McNaughton, B.L. and Barnes, C.A. (2001). Reactivation of hippocampal ensemble activity patterns in the aging rat. Behavioral Neuroscience 115(6): 1180-92. Gerrard, J.L.,Kudrimoti, H.K.,Cowen, S.L.,Redish, A.D.,Rosenzweig, E.S.,Barnes, C.A. and McNaughton, B.L. (2002). Dissociation of pattern and sequence reactivation efficiency in the aged rat hippocampus. Soc Neurosci Abstr 678.7. Good, M.A.,Barnes, P.,Staal, V.,McGregor, A. and Honey, R.C. (2007). Context- but not familiarity-dependent forms of object recognition are impaired following excitotoxic hippocampal lesions in rats. Behavioral Neuroscience 121(1): 218-23. Goodale, M.A. (1993). Visual pathways supporting perception and action in the primate cerebral cortex. Current Opinion in Neurobiology 3(4): 578-85. Gordon, S.K. and Clark, W.C. (1974). Application of signal detection theory to prose recall and recognition in elderly and young adults. Journal of Gerontology 29(1): 64-72. Gothard, K.M.,Skaggs, W.E. and McNaughton, B.L. (1996). Dynamics of mismatch correction in the hippocampal ensemble code for space: interaction between path integration and environmental cues. Journal of Neuroscience 16(24): 8027-40. Gothard, K.M.,Skaggs, W.E.,Moore, K.M. and McNaughton, B.L. (1996). Binding of hippocampal CA1 neural activity to multiple reference frames in a landmark-based navigation task. Journal of Neuroscience 16(2): 823-35. Guzowski, J.F.,Lyford, G.L.,Stevenson, G.D.,Houston, F.P.,McGaugh, J.L.,Worley, P.F. and Barnes, C.A. (2000). Inhibition of activity-dependent arc protein expression in the rat hippocampus impairs the maintenance of long-term potentiation and the consolidation of long-term memory. Journal of Neuroscience 20(11): 3993-4001. Guzowski, J.F.,McNaughton, B.L.,Barnes, C.A. and Worley, P.F. (1999). Environment-specific expression of the immediate-early gene Arc in hippocampal neuronal ensembles. Nature Neuroscience 2(12): 1120-4. Guzowski, J.F.,Miyashita, T.,Chawla, M.K.,Sanderson, J.,Maes, L.I.,Houston, F.P.,Lipa, P.,McNaughton, B.L.,Worley, P.F. and Barnes, C.A. (2006). Recent behavioral history modifies coupling between cell activity and Arc gene transcription in hippocampal CA1 neurons. Proceedings of the National Academy of Sciences, U S A 103(4): 1077-82. 313
Hafting, T.,Fyhn, M.,Molden, S.,Moser, M.B. and Moser, E.I. (2005). Microstructure of a spatial map in the entorhinal cortex. Nature 436(7052): 801-6. Hanks, S.D. and Flood, D.G. (1991). Region-specific stability of dendritic extent in normal human aging and regression in Alzheimer's disease. I. CA1 of hippocampus. Brain Research 540(1-2): 63-82. Hargreaves, E.L.,Rao, G.,Lee, I. and Knierim, J.J. (2005). Major dissociation between medial and lateral entorhinal input to dorsal hippocampus. Science 308(5729): 1792-4. Hasselmo, M.E. (2006). The role of acetylcholine in learning and memory. Current Opinion in Neurobiology 16(6): 710-5. Hasselmo, M.E. and Stern, C.E. (2006). Mechanisms underlying working memory for novel information. Trends in Cognitive Sciences 10(11): 487- 93. Hay, J.F. and Jacoby, L.L. (1999). Separating habit and recollection in young and older adults: effects of elaborative processing and distinctiveness. Psychol Aging 14(1): 122-34. Hayes, S.M.,Nadel, L. and Ryan, L. (2007). The effect of scene context on episodic object recognition: parahippocampal cortex mediates memory encoding and retrieval success. Hippocampus 17(9): 873-89. Hebb, D. (1949). The Organization of Behavior: A Neurophysiological Theory. Hebb, D. (1978). On watching myself get old. Psychology Today November. Herndon, J.G.,Moss, M.B.,Rosene, D.L. and Killiany, R.J. (1997). Patterns of cognitive decline in aged rhesus monkeys. Behavioural Brain Research 87(1): 25-34. Higuchi, S. and Miyashita, Y. (1996). Formation of mnemonic neuronal responses to visual paired associates in inferotemporal cortex is impaired by perirhinal and entorhinal lesions. Proceedings of the National Academy of Sciences, U S A 93(2): 739-43. Hirase, H.,Leinekugel, X.,Czurko, A.,Csicsvari, J. and Buzsaki, G. (2001). Firing rates of hippocampal neurons are preserved during subsequent sleep episodes and modified by novel awake experience. Proceedings of the National Academy of Sciences, U S A 98(16): 9386-90. Hoang, L.T.,Fellous, J.M. and Barnes, C.A. (2008). Expression of the immediate- early gene Arc in ventral tegmental neurons during aging. Society for Neuroscience, Washington D.C. Hoffman, K.L. and McNaughton, B.L. (2002). Coordinated reactivation of distributed memory traces in primate neocortex. Science 297(5589): 2070- 3. Holscher, C. and Rolls, E.T. (2002). Perirhinal cortex neuronal activity is actively related to working memory in the macaque. Neural Plasticity 9(1): 41-51. Holscher, C.,Rolls, E.T. and Xiang, J. (2003). Perirhinal cortex neuronal activity related to long-term familiarity memory in the macaque. European Journal of Neuroscience 18(7): 2037-46. 314
Insausti, R.,Herrero, M.T. and Witter, M.P. (1997). Entorhinal cortex of the rat: cytoarchitectonic subdivisions and the origin and distribution of cortical efferents. Hippocampus 7(2): 146-83. Insausti, R. and Munoz, M. (2001). Cortical projections of the non-entorhinal hippocampal formation in the cynomolgus monkey (Macaca fascicularis). European Journal of Neuroscience 14(3): 435-51. Insel, N.,Ruiz-Luna, M.L.,Permenter, M.,Vogt, J.,Erickson, C.A. and Barnes, C.A. (2008). Aging in rhesus macaques is associated with changes in novelty preference and altered saccade dynamics. Behavioral Neuroscience 122(6): 1328-42. Ivanco, T.L. and Racine, R.J. (2000). Long-term potentiation in the reciprocal corticohippocampal and corticocortical pathways in the chronically implanted, freely moving rat. Hippocampus 10(2): 143-52. Jackson, J.C.,Johnson, A. and Redish, A.D. (2006). Hippocampal sharp waves and reactivation during awake states depend on repeated sequential experience. Journal of Neuroscience 26(48): 12415-26. Jennings, J.M. and Jacoby, L.L. (1993). Automatic versus intentional uses of memory: aging, attention, and control. Psychol Aging 8(2): 283-93. Jennings, J.M. and Jacoby, L.L. (1997). An opposition procedure for detecting age-related deficits in recollection: telling effects of repetition. Psychol Aging 12(2): 352-61. Ji, D. and Wilson, M.A. (2007). Coordinated memory replay in the visual cortex and hippocampus during sleep. Nature Neuroscience 10(1): 100-107. Jung, M.W. and McNaughton, B.L. (1993). Spatial selectivity of unit activity in the hippocampal granular layer. Hippocampus 3(2): 165-82. Jung, M.W.,Wiener, S.I. and McNaughton, B.L. (1994). Comparison of spatial firing characteristics of units in dorsal and ventral hippocampus of the rat. Journal of Neuroscience 14(12): 7347-56. Junjaud, G.,Rouaud, E.,Turpin, F.,Mothet, J.P. and Billard, J.M. (2006). Age- related effects of the neuromodulator D-serine on neurotransmission and synaptic potentiation in the CA1 hippocampal area of the rat. Journal of Neurochemistry 98(4): 1159-66. Kentros, C.,Hargreaves, E.,Hawkins, R.D.,Kandel, E.R.,Shapiro, M. and Muller, R.V. (1998). Abolition of long-term stability of new hippocampal place cell maps by NMDA receptor blockade. Science 280(5372): 2121-6. Kerr, D.S.,Campbell, L.W.,Hao, S.Y. and Landfield, P.W. (1989). Corticosteroid modulation of hippocampal potentials: increased effect with aging. Science 245(4925): 1505-9. Kerr, K.M.,Agster, K.L.,Furtak, S.C. and Burwell, R.D. (2007). Functional neuroanatomy of the parahippocampal region: the lateral and medial entorhinal areas. Hippocampus 17(9): 697-708. Kesner, R.P.,Ravindranathan, A.,Jackson, P.,Giles, R. and Chiba, A.A. (2001). A neural circuit analysis of visual recognition memory: role of perirhinal, medial, and lateral entorhinal cortex. Learning and Memory 8(2): 87-95. 315
Keuker, J.I.,Luiten, P.G. and Fuchs, E. (2003). Preservation of hippocampal neuron numbers in aged rhesus monkeys. Neurobiology of Aging 24(1): 157-65. Kholodar-Smith, D.B.,Allen, T.A. and Brown, T.H. (2008). Fear conditioning to discontinuous auditory cues requires perirhinal cortical function. Behavioral Neuroscience 122(5): 1178-85. Kholodar-Smith, D.B.,Boguszewski, P. and Brown, T.H. (2008). Auditory trace fear conditioning requires perirhinal cortex. Neurobiol Learn Mem 90(3): 537-43. Kim, J.J.,Clark, R.E. and Thompson, R.F. (1995). Hippocampectomy impairs the memory of recently, but not remotely, acquired trace eyeblink conditioned responses. Behavioral Neuroscience 109(2): 195-203. Kim, J.J. and Fanselow, M.S. (1992). Modality-specific retrograde amnesia of fear. Science 256(5057): 675-7. Kinsella, K. and Velkoff, V.A. (2001). An aging world: 2001. U.S. Census Bureau Series P95/01-1. Kirwan, C.B.,Wixted, J.T. and Squire, L.R. (2008). Activity in the medial temporal lobe predicts memory strength, whereas activity in the prefrontal cortex predicts recollection. Journal of Neuroscience 28(42): 10541-8. Knierim, J.J.,Kudrimoti, H.S. and McNaughton, B.L. (1995). Place cells, head direction cells, and the learning of landmark stability. Journal of Neuroscience 15(3 Pt 1): 1648-59. Krause, M.,Yang, Z.,Rao, G.,Houston, F.P. and Barnes, C.A. (2008). Altered dendritic integration in hippocampal granule cells of spatial learning- impaired aged rats. Journal of Neurophysiology 99(6): 2769-78. Kudrimoti, H.S.,Barnes, C.A. and McNaughton, B.L. (1999). Reactivation of hippocampal cell assemblies: effects of behavioral state, experience, and EEG dynamics. Journal of Neuroscience 19(10): 4090-101. Kumar, A. and Foster, T. (2007). Environmental enrichment decreases the afterhyperpolarization in senescent rats. Brain Research 1130(1): 103-7. Kumar, A. and Foster, T.C. (2005). Intracellular calcium stores contribute to increased susceptibility to LTD induction during aging. Brain Research 1031(1): 125-8. Kumar, A.,Thinschmidt, J.S.,Foster, T.C. and King, M.A. (2007). Aging effects on the limits and stability of long-term synaptic potentiation and depression in rat hippocampal area CA1. Journal of Neurophysiology 98(2): 594-601. Landfield, P.W. (1988). Hippocampal neurobiological mechanisms of age-related memory dysfunction. Neurobiology of Aging 9(5-6): 571-9. Landfield, P.W. and Lynch, G. (1977). Impaired monosynaptic potentiation in in vitro hippocampal slices from aged, memory-deficient rats. Journal of Gerontology 32(5): 523-33. Landfield, P.W.,McGaugh, J.L. and Lynch, G. (1978). Impaired synaptic potentiation processes in the hippocampus of aged, memory-deficient rats. Brain Research 150(1): 85-101. 316
Landfield, P.W. and Pitler, T.A. (1984). Prolonged Ca2+-dependent afterhyperpolarizations in hippocampal neurons of aged rats. Science 226(4678): 1089-92. Landfield, P.W.,Pitler, T.A. and Applegate, M.D. (1986). The effects of high Mg2+-to-Ca2+ ratios on frequency potentiation in hippocampal slices of young and aged rats. Journal of Neurophysiology 56(3): 797-811. Lashley, K.S. (1931). Mass Action in Cerebral Function. Science 73(1888): 245- 254. Lashley, K.S. (1949). Persistent problems in the evolution of mind. Q Rev Biol 24(1): 28-42. Lashley, K.S.,Chow, K.L. and Semmes, J. (1951). An examination of the electrical field theory of cerebral integration. Psychol Rev 58(2): 123-36. Lavenex, P.,Suzuki, W.A. and Amaral, D.G. (2002). Perirhinal and parahippocampal cortices of the macaque monkey: projections to the neocortex. Journal of Comparative Neurology 447(4): 394-420. Lavenex, P.,Suzuki, W.A. and Amaral, D.G. (2004). Perirhinal and parahippocampal cortices of the macaque monkey: Intrinsic projections and interconnections. Journal of Comparative Neurology 472(3): 371-94. Lee, A.K. and Wilson, M.A. (2002). Memory of sequential experience in the hippocampus during slow wave sleep. Neuron 36(6): 1183-94. Lee, I.,Rao, G. and Knierim, J.J. (2004). A double dissociation between hippocampal subfields: differential time course of CA3 and CA1 place cells for processing changed environments. Neuron 42(5): 803-15. Lee, I.,Yoganarasimha, D.,Rao, G. and Knierim, J.J. (2004). Comparison of population coherence of place cells in hippocampal subfields CA1 and CA3. Nature 430(6998): 456-9. Leutgeb, J.K.,Leutgeb, S.,Moser, M.B. and Moser, E.I. (2007). Pattern separation in the dentate gyrus and CA3 of the hippocampus. Science 315(5814): 961-6. Leutgeb, J.K.,Leutgeb, S.,Treves, A.,Meyer, R.,Barnes, C.A.,McNaughton, B.L.,Moser, M.B. and Moser, E.I. (2005). Progressive transformation of hippocampal neuronal representations in "morphed" environments. Neuron 48(2): 345-58. Leutgeb, S.,Leutgeb, J.K.,Treves, A.,Moser, M.B. and Moser, E.I. (2004). Distinct ensemble codes in hippocampal areas CA3 and CA1. Science 305(5688): 1295-8. Lindquist, D.H.,Jarrard, L.E. and Brown, T.H. (2004). Perirhinal cortex supports delay fear conditioning to rat ultrasonic social signals. Journal of Neuroscience 24(14): 3610-7. Liu, P. and Bilkey, D.K. (1996). Direct connection between perirhinal cortex and hippocampus is a major constituent of the lateral perforant path. Hippocampus 6(2): 125-35. Liu, P. and Bilkey, D.K. (1996). Long-term potentiation in the perirhinal- hippocampal pathway is NMDA dependent. Neuroreport 7(7): 1241-4. 317
Liu, P. and Bilkey, D.K. (1997). Current source density analysis of the potential evoked in hippocampus by perirhinal cortex stimulation. Hippocampus 7(4): 389-96. Liu, P.,Smith, P.F. and Darlington, C.L. (2008). Glutamate receptor subunits expression in memory-associated brain structures: regional variations and effects of aging. Synapse 62(11): 834-41. Luebke, J.I. and Rosene, D.L. (2003). Aging alters dendritic morphology, input resistance, and inhibitory signaling in dentate granule cells of the rhesus monkey. Journal of Comparative Neurology 460(4): 573-84. Malinow, R. and Malenka, R.C. (2002). AMPA receptor trafficking and synaptic plasticity. Annual review of neuroscience 25: 103-26. Malkova, L.,Bachevalier, J.,Mishkin, M. and Saunders, R.C. (2001). Neurotoxic lesions of perirhinal cortex impair visual recognition memory in rhesus monkeys. Neuroreport 12(9): 1913-7. Malmberg, K.J. (2008). Recognition memory: a review of the critical findings and an integrated theory for relating them. Cogn Psychol 57(4): 335-84. Mandler, G. (1980). Recognizing: The Judgement of Prior Occurence. Psychological Review 87(3): 252-271. Mandler, G.,Pearlstone, Z. and Koopsmans, H.R. (1969). Effects of organization and semantic similarity on recall and recognition. Journal of Verbal Learning and Verbal Behavior 8: 410-423. Mantyla, T. (1993). Knowing but not remembering: adult age differences in recollective experience. Mem Cognit 21(3): 379-88. Markham, J.A.,McKian, K.P.,Stroup, T.S. and Juraska, J.M. (2005). Sexually dimorphic aging of dendritic morphology in CA1 of hippocampus. Hippocampus 15(1): 97-103. Markus, E.J.,Barnes, C.A.,McNaughton, B.L.,Gladden, V.L. and Skaggs, W.E. (1994). Spatial information content and reliability of hippocampal CA1 neurons: effects of visual input. Hippocampus 4(4): 410-21. Markus, E.J.,Qin, Y.L.,Leonard, B.,Skaggs, W.E.,McNaughton, B.L. and Barnes, C.A. (1995). Interactions between location and task affect the spatial and directional firing of hippocampal neurons. Journal of Neuroscience 15(11): 7079-94. Marr, D. (1971). Simple memory: a theory for archicortex. Philosophical Transactions of the Royal Society of London Series B, Biological Sciences 262(841): 23-81. Maurer, A.P.,Cowen, S.L.,Burke, S.N.,Barnes, C.A. and McNaughton, B.L. (2006). Organization of hippocampal cell assemblies based on theta phase precession. Hippocampus 16(9): 785-94. Maurer, A.P.,Cowen, S.L.,Burke, S.N.,Barnes, C.A. and McNaughton, B.L. (2006). Phase precession in hippocampal interneurons showing strong functional coupling to individual pyramidal cells. Journal of Neuroscience 26(52): 13485-92. 318
Maurer, A.P.,Vanrhoads, S.R.,Sutherland, G.R.,Lipa, P. and McNaughton, B.L. (2005). Self-motion and the origin of differential spatial scaling along the septo-temporal axis of the hippocampus. Hippocampus 15(7): 841-52. McClelland, J.L.,McNaughton, B.L. and O'Reilly, R.C. (1995). Why there are complementary learning systems in the hippocampus and neocortex: insights from the successes and failures of connectionist models of learning and memory. Psychol Rev 102(3): 419-57. McGann, J.P.,Moyer, J.R., Jr. and Brown, T.H. (2001). Predominance of late- spiking neurons in layer VI of rat perirhinal cortex. Journal of Neuroscience 21(14): 4969-76. McIntyre, D.C.,Kelly, M.E. and Staines, W.A. (1996). Efferent projections of the anterior perirhinal cortex in the rat. Journal of Comparative Neurology 369(2): 302-18. McNaughton, B.L.,Barnes, C.A. and O'Keefe, J. (1983). The contributions of position, direction, and velocity to single unit activity in the hippocampus of freely-moving rats. Experimental Brain Research 52(1): 41-9. McNaughton, B.L.,Battaglia, F.P.,Jensen, O.,Moser, E.I. and Moser, M.B. (2006). Path integration and the neural basis of the 'cognitive map'. Nature Reveiws Neuroscience 7(8): 663-78. McNaughton, B.L.,Douglas, R.M. and Goddard, G.V. (1978). Synaptic enhancement in fascia dentata: cooperativity among coactive afferents. Brain Research 157(2): 277-93. McNaughton, B.L. and Morris, R.G. (1987). Hippocampal synaptic enhancement and information storage within a distributed memory system. Trends in Neurosciences 10(10): 408-415. McNaughton, B.L.,O'Keefe, J. and Barnes, C.A. (1983). The stereotrode: a new technique for simultaneous isolation of several single units in the central nervous system from multiple unit records. Journal of Neuroscience Methods 8(4): 391-7. McTighe, S.M.,Cowell, R.A.,Winters, B.D.,Bussey, T.J. and Saksida, L.A. (2008). Novel objects appear familiar following perirhinal cortical damage Society for Neuroscience Abstracts. Mehta, M.R.,Barnes, C.A. and McNaughton, B.L. (1997). Experience-dependent, asymmetric expansion of hippocampal place fields. Proceedings of the National Academy of Sciences, U S A 94(16): 8918-21. Mehta, M.R.,Quirk, M.C. and Wilson, M.A. (2000). Experience-dependent asymmetric shape of hippocampal receptive fields. Neuron 25(3): 707-15. Merrill, D.A.,Chiba, A.A. and Tuszynski, M.H. (2001). Conservation of neuronal number and size in the entorhinal cortex of behaviorally characterized aged rats. Journal of Comparative Neurology 438(4): 445-56. Merrill, D.A.,Roberts, J.A. and Tuszynski, M.H. (2000). Conservation of neuron number and size in entorhinal cortex layers II, III, and V/VI of aged primates. Journal of Comparative Neurology 422(3): 396-401. 319
Meunier, M.,Bachevalier, J.,Mishkin, M. and 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(12): 5418-32. Milner, A.D. and Goodale, M.A. (1993). Visual pathways to perception and action. Progress in Brain Research 95: 317-37. Milner, A.D. and Goodale, M.A. (2008). Two visual systems re-viewed. Neuropsychologia 46(3): 774-85. Mishkin, M. and Delacour, J. (1975). An analysis of short-term visual memory in the monkey. J Exp Psychol Anim Behav Process 1(4): 326-34. Mishkin, M.,Vargha-Khadem, F. and Gadian, D.G. (1998). Amnesia and the organization of the hippocampal system. Hippocampus 8(3): 212-6. Mizumori, S.J.,Lavoie, A.M. and Kalyani, A. (1996). Redistribution of spatial representation in the hippocampus of aged rats performing a spatial memory task. Behavioral Neuroscience 110(5): 1006-16. Montaldi, D.,Spencer, T.J.,Roberts, N. and Mayes, A.R. (2006). The neural system that mediates familiarity memory. Hippocampus 16(5): 504-20. Moore, C.I.,Browning, M.D. and Rose, G.M. (1993). Hippocampal plasticity induced by primed burst, but not long-term potentiation, stimulation is impaired in area CA1 of aged Fischer 344 rats. Hippocampus 3(1): 57-66. Morgan, J.I.,Cohen, D.R.,Hempstead, J.L. and Curran, T. (1987). Mapping patterns of c-fos expression in the central nervous system after seizure. Science 237(4811): 192-7. Morris, R. (1984). Developments of a water-maze procedure for studying spatial learning in the rat. Journal of Neuroscience Methods 11(1): 47-60. Morrison, J.H. and Hof, P.R. (1997). Life and death of neurons in the aging brain. Science 278(5337): 412-9. Moscovitch, M.,Rosenbaum, R.S.,Gilboa, A.,Addis, D.R.,Westmacott, R.,Grady, C.,McAndrews, M.P.,Levine, B.,Black, S.,Winocur, G. and Nadel, L. (2005). Functional neuroanatomy of remote episodic, semantic and spatial memory: a unified account based on multiple trace theory. Journal of Anatomy 207(1): 35-66. Moss, M.B.,Killiany, R.J.,Lai, Z.C.,Rosene, D.L. and Herndon, J.G. (1997). Recognition memory span in rhesus monkeys of advanced age. Neurobiology of Aging 18(1): 13-9. Moss, M.B.,Rosene, D.L. and Peters, A. (1988). Effects of aging on visual recognition memory in the rhesus monkey. Neurobiology of Aging 9(5-6): 495-502. Mothet, J.P.,Parent, A.T.,Wolosker, H.,Brady, R.O., Jr.,Linden, D.J.,Ferris, C.D.,Rogawski, M.A. and Snyder, S.H. (2000). D-serine is an endogenous ligand for the glycine site of the N-methyl-D-aspartate receptor. Proceedings of the National Academy of Sciences, U S A 97(9): 4926-31. Moyer, J.R., Jr.,McNay, E.C. and Brown, T.H. (2002). Three classes of pyramidal neurons in layer V of rat perirhinal cortex. Hippocampus 12(2): 218-34. 320
Moyer, J.R., Jr.,Thompson, L.T.,Black, J.P. and Disterhoft, J.F. (1992). Nimodipine increases excitability of rabbit CA1 pyramidal neurons in an age- and concentration-dependent manner. Journal of Neurophysiology 68(6): 2100-9. Muir, G.M. and Bilkey, D.K. (2001). Instability in the place field location of hippocampal place cells after lesions centered on the perirhinal cortex. Journal of Neuroscience 21(11): 4016-25. Muller, R.U. and Kubie, J.L. (1987). The effects of changes in the environment on the spatial firing of hippocampal complex-spike cells. Journal of Neuroscience 7(7): 1951-68. Muller, R.U.,Kubie, J.L. and Ranck, J.B., Jr. (1987). Spatial firing patterns of hippocampal complex-spike cells in a fixed environment. Journal of Neuroscience 7(7): 1935-50. Mumby, D.G. (2001). Perspectives on object-recognition memory following hippocampal damage: lessons from studies in rats. Behavioural Brain Research 127(1-2): 159-81. Mumby, D.G.,E.R., W. and Pinel, J.P. (1992). Object recognition memory is only mildly impaired in rats with lesions of the hippocampus and amygdala Psychobiology 20: 18-27. Mumby, D.G.,Gaskin, S.,Glenn, M.J.,Schramek, T.E. and Lehmann, H. (2002). Hippocampal damage and exploratory preferences in rats: memory for objects, places, and contexts. Learning and Memory 9(2): 49-57. Mumby, D.G.,Glenn, M.J.,Nesbitt, C. and Kyriazis, D.A. (2002). Dissociation in retrograde memory for object discriminations and object recognition in rats with perirhinal cortex damage. Behavioural Brain Research 132(2): 215- 26. Mumby, D.G. and Pinel, J.P. (1994). Rhinal cortex lesions and object recognition in rats. Behavioral Neuroscience 108(1): 11-8. Mumby, D.G.,Pinel, J.P. and E.R., W. (1990). Nonrecurring items delayed nonmatching-to-sample in rats: A new paradigm for testing nonspatial working memory in rats. Psychobiology 18: 321-326. Murray, E.A. and Bussey, T.J. (1999). Perceptual-mnemonic functions of the perirhinal cortex. Trends in Cognitive Sciences 3(4): 142-151. Murray, E.A.,Bussey, T.J.,Hampton, R.R. and Saksida, L.M. (2000). The parahippocampal region and object identification. Annals of the New York Academy of Sciences 911: 166-74. Murray, E.A.,Bussey, T.J. and Saksida, L.M. (2007). Visual perception and memory: a new view of medial temporal lobe function in primates and rodents. Annual review of neuroscience 30: 99-122. Myhrer, T. and Andersen, J.M. (2001). Interference with visual memory in rats following infusion of the functional NMDA receptor antagonist, HA-966, into temporal regions. European Journal of Pharmacology 428(3): 323-30. Naber, P.A.,Witter, M.P. and Lopez da Silva, F.H. (1999). Perirhinal cortex input to the hippocampus in the rat: evidence for parallel pathways, both direct 321
and indirect. A combined physiological and anatomical study. European Journal of Neuroscience 11(11): 4119-33. Nadasdy, Z.,Hirase, H.,Czurko, A.,Csicsvari, J. and Buzsaki, G. (1999). Replay and time compression of recurring spike sequences in the hippocampus. Journal of Neuroscience 19(21): 9497-507. Nadel, L. and Hardt, O. (2004). The spatial brain. Neuropsychology 18(3): 473-6. Nadel, L. and Moscovitch, M. (1997). Memory consolidation, retrograde amnesia and the hippocampal complex. Current Opinion in Neurobiology 7(2): 217- 27. Nadel, L.,Wilner, J. and Kurz, E.M. (1985). Cognitive maps and environmental context. Context and Learning. Tomie, P.D.B.A. Hillsdale, N.J., Lawrence Earlbaum: 385-406. Naya, Y.,Yoshida, M. and Miyashita, Y. (2001). Backward spreading of memory- retrieval signal in the primate temporal cortex. Science 291(5504): 661-4. Nemanic, S.,Alvarado, M.C. and Bachevalier, J. (2004). The hippocampal/parahippocampal regions and recognition memory: insights from visual paired comparison versus object-delayed nonmatching in monkeys. Journal of Neuroscience 24(8): 2013-26. Newton, I.G.,Forbes, M.E.,Linville, M.C.,Pang, H.,Tucker, E.W.,Riddle, D.R. and Brunso-Bechtold, J.K. (2008). Effects of aging and caloric restriction on dentate gyrus synapses and glutamate receptor subunits. Neurobiology of Aging 29(9): 1308-18. Nicholson, D.A.,Yoshida, R.,Berry, R.W.,Gallagher, M. and Geinisman, Y. (2004). Reduction in size of perforated postsynaptic densities in hippocampal axospinous synapses and age-related spatial learning impairments. Journal of Neuroscience 24(35): 7648-53. Nicolle, M.M.,Colombo, P.J.,Gallagher, M. and McKinney, M. (1999). Metabotropic glutamate receptor-mediated hippocampal phosphoinositide turnover is blunted in spatial learning-impaired aged rats. Journal of Neuroscience 19(21): 9604-10. Nicolle, M.M.,Gallagher, M. and McKinney, M. (1999). No loss of synaptic proteins in the hippocampus of aged, behaviorally impaired rats. Neurobiology of Aging 20(3): 343-8. Niesen, C.E.,Baskys, A. and Carlen, P.L. (1988). Reversed ethanol effects on potassium conductances in aged hippocampal dentate granule neurons. Brain Research 445(1): 137-41. Nishino, H.,Ono, T.,Muramoto, K.,Fukuda, M. and Sasaki, K. (1987). Neuronal activity in the ventral tegmental area (VTA) during motivated bar press feeding in the monkey. Brain Research 413(2): 302-13. Norman, G. and Eacott, M.J. (2004). Impaired object recognition with increasing levels of feature ambiguity in rats with perirhinal cortex lesions. Behavioural Brain Research 148(1-2): 79-91. 322
Norman, G. and Eacott, M.J. (2005). Dissociable effects of lesions to the perirhinal cortex and the postrhinal cortex on memory for context and objects in rats. Behavioral Neuroscience 119(2): 557-66. Norman, K.A. and Schacter, D.L. (1997). False recognition in younger and older adults: exploring the characteristics of illusory memories. Mem Cognit 25(6): 838-48. Norris, C.M.,Halpain, S. and Foster, T.C. (1998). Alterations in the balance of protein kinase/phosphatase activities parallel reduced synaptic strength during aging. Journal of Neurophysiology 80(3): 1567-70. Norris, C.M.,Halpain, S. and Foster, T.C. (1998). Reversal of age-related alterations in synaptic plasticity by blockade of L-type Ca2+ channels. Journal of Neuroscience 18(9): 3171-9. Norris, C.M.,Korol, D.L. and Foster, T.C. (1996). Increased susceptibility to induction of long-term depression and long-term potentiation reversal during aging. Journal of Neuroscience 16(17): 5382-92. O'Brien, N.,Lehmann, H.,Lecluse, V. and Mumby, D.G. (2006). Enhanced context-dependency of object recognition in rats with hippocampal lesions. Behavioural Brain Research 170(1): 156-62. O'Keefe, J. (1976). Place units in the hippocampus of the freely moving rat. Experimental Neurology 51(1): 78-109. O'Keefe, J. and Conway, D.H. (1978). Hippocampal place units in the freely moving rat: why they fire where they fire. Experimental Brain Research 31(4): 573-90. O'Keefe, J. and Dostrovsky, J. (1971). The hippocampus as a spatial map. Preliminary evidence from unit activity in the freely-moving rat. Brain Research 34(1): 171-5. O'Keefe, J. and Nadel, L. (1978). The Hippocampus as a Cognitive Map. Oxford, Claredon Press. O'Keefe, J. and Recce, M.L. (1993). Phase relationship between hippocampal place units and the EEG theta rhythm. Hippocampus 3(3): 317-30. O'Neill, J.,Senior, T. and Csicsvari, J. (2006). Place-selective firing of CA1 pyramidal cells during sharp wave/ripple network patterns in exploratory behavior. Neuron 49(1): 143-55. Oler, J.A. and Markus, E.J. (1998). Age-related deficits on the radial maze and in fear conditioning: hippocampal processing and consolidation. Hippocampus 8(4): 402-15. Oler, J.A. and Markus, E.J. (2000). Age-related deficits in the ability to encode contextual change: a place cell analysis. Hippocampus 10(3): 338-50. Pakkenberg, B. and Gundersen, H.J. (1997). Neocortical neuron number in humans: effect of sex and age. Journal of Comparative Neurology 384(2): 312-20. Parkin, A.J. and Walter, B.M. (1992). Recollective experience, normal aging, and frontal dysfunction. Psychol Aging 7(2): 290-8. 323
Parsons, C.G.,Quack, G.,Bresink, I.,Baran, L.,Przegalinski, E.,Kostowski, W.,Krzascik, P.,Hartmann, S. and Danysz, W. (1995). Comparison of the potency, kinetics and voltage-dependency of a series of uncompetitive NMDA receptor antagonists in vitro with anticonvulsive and motor impairment activity in vivo. Neuropharmacology 34(10): 1239-58. Pastalkova, E.,Itskov, V.,Amarasingham, A. and Buzsaki, G. (2008). Internally generated cell assembly sequences in the rat hippocampus. Science 321(5894): 1322-7. Pavlides, C. and Winson, J. (1989). Influences of hippocampal place cell firing in the awake state on the activity of these cells during subsequent sleep episodes. Journal of Neuroscience 9(8): 2907-18. Paz, R.,Bauer, E.P. and Pare, D. (2007). Learning-related facilitation of rhinal interactions by medial prefrontal inputs. Journal of Neuroscience 27(24): 6542-51. Paz, R.,Pelletier, J.G.,Bauer, E.P. and Pare, D. (2006). Emotional enhancement of memory via amygdala-driven facilitation of rhinal interactions. Nature Neuroscience 9(10): 1321-9. Pelletier, J.G.,Apergis, J. and Pare, D. (2004). Low-probability transmission of neocortical and entorhinal impulses through the perirhinal cortex. Journal of Neurophysiology 91(5): 2079-89. Peters, A.,Leahu, D.,Moss, M.B. and McNally, K.J. (1994). The effects of aging on area 46 of the frontal cortex of the rhesus monkey. Cerebral Cortex 4(6): 621-35. Pieta Dias, C.,Martins de Lima, M.N.,Presti-Torres, J.,Dornelles, A.,Garcia, V.A.,Siciliani Scalco, F.,Rewsaat Guimaraes, M.,Constantino, L.,Budni, P.,Dal-Pizzol, F. and Schroder, N. (2007). Memantine reduces oxidative damage and enhances long-term recognition memory in aged rats. Neuroscience 146(4): 1719-25. Pihlajamaki, M.,Tanila, H.,Kononen, M.,Hanninen, T.,Hamalainen, A.,Soininen, H. and Aronen, H.J. (2004). Visual presentation of novel objects and new spatial arrangements of objects differentially activates the medial temporal lobe subareas in humans. European Journal of Neuroscience 19(7): 1939- 49. Pinto, A.,Fuentes, C. and Pare, D. (2006). Feedforward inhibition regulates perirhinal transmission of neocortical inputs to the entorhinal cortex: ultrastructural study in guinea pigs. Journal of Comparative Neurology 495(6): 722-34. Piterkin, P.,Cole, E.,Cossette, M.P.,Gaskin, S. and Mumby, D.G. (2008). A limited role for the hippocampus in the modulation of novel-object preference by contextual cues. Learning and Memory 15(10): 785-91. Pitler, T.A. and Landfield, P.W. (1990). Aging-related prolongation of calcium spike duration in rat hippocampal slice neurons. Brain Research 508(1): 1- 6. 324
Pitsikas, N.,Rigamonti, A.E.,Cella, S.G.,Sakellaridis, N. and Muller, E.E. (2005). The nitric oxide donor molsidomine antagonizes age-related memory deficits in the rat. Neurobiology of Aging 26(2): 259-64. Pitsikas, N. and Sakellaridis, N. (2007). Memantine and recognition memory: Possible facilitation of its behavioral effects by the nitric oxide (NO) donor molsidomine. European Journal of Pharmacology 571(2-3): 174-9. Platano, D.,Fattoretti, P.,Balietti, M.,Bertoni-Freddari, C. and Aicardi, G. (2008). Long-term visual object recognition memory in aged rats. Rejuvenation Res 11(2): 333-9. Plath, N.,Ohana, O.,Dammermann, B.,Errington, M.L.,Schmitz, D.,Gross, C.,Mao, X.,Engelsberg, A.,Mahlke, C.,Welzl, H.,Kobalz, U.,Stawrakakis, A.,Fernandez, E.,Waltereit, R.,Bick-Sander, A.,Therstappen, E.,Cooke, S.F.,Blanquet, V.,Wurst, W.,Salmen, B.,Bosl, M.R.,Lipp, H.P.,Grant, S.G.,Bliss, T.V.,Wolfer, D.P. and Kuhl, D. (2006). Arc/Arg3.1 is essential for the consolidation of synaptic plasticity and memories. Neuron 52(3): 437-44. Potier, B.,Lamour, Y. and Dutar, P. (1993). Age-related alterations in the properties of hippocampal pyramidal neurons among rat strains. Neurobiology of Aging 14(1): 17-25. Potier, B.,Rascol, O.,Jazat, F.,Lamour, Y. and Dutar, P. (1992). Alterations in the properties of hippocampal pyramidal neurons in the aged rat. Neuroscience 48(4): 793-806. Prull, M.W.,Dawes, L.L.,Martin, A.M., 3rd,Rosenberg, H.F. and Light, L.L. (2006). Recollection and familiarity in recognition memory: adult age differences and neuropsychological test correlates. Psychology and aging 21(1): 107- 18. Prusky, G.T.,Douglas, R.M.,Nelson, L.,Shabanpoor, A. and Sutherland, R.J. (2004). Visual memory task for rats reveals an essential role for hippocampus and perirhinal cortex. Proceedings of the National Academy of Sciences, U S A 101(14): 5064-8. Pyapali, G.K. and Turner, D.A. (1996). Increased dendritic extent in hippocampal CA1 neurons from aged F344 rats. Neurobiology of Aging 17(4): 601-11. Qin, Y.L.,McNaughton, B.L.,Skaggs, W.E. and Barnes, C.A. (1997). Memory reprocessing in corticocortical and hippocampocortical neuronal ensembles. Philosophical Transactions of the Royal Society of London Series B, Biological Sciences 352(1360): 1525-33. Ranck, J.B., Jr. (1973). Studies on single neurons in dorsal hippocampal formation and septum in unrestrained rats. I. Behavioral correlates and firing repertoires. Experimental Neurology 41(2): 461-531. Ranganath, C.,Yonelinas, A.P.,Cohen, M.X.,Dy, C.J.,Tom, S.M. and D'Esposito, M. (2004). Dissociable correlates of recollection and familiarity within the medial temporal lobes. Neuropsychologia 42(1): 2-13. Rankin, J.L. and Kausler, D.H. (1979). Adult age differences in false recognitions. Journal of Gerontology 34(1): 58-65. 325
Rapp, P.R. and Amaral, D.G. (1991). Recognition memory deficits in a subpopulation of aged monkeys resemble the effects of medial temporal lobe damage. Neurobiology of Aging 12(5): 481-6. Rapp, P.R.,Deroche, P.S.,Mao, Y. and Burwell, R.D. (2002). Neuron number in the parahippocampal region is preserved in aged rats with spatial learning deficits. Cerebral Cortex 12(11): 1171-9. Rapp, P.R. and Gallagher, M. (1996). Preserved neuron number in the hippocampus of aged rats with spatial learning deficits. Proc Natl Acad Sci U S A 93(18): 9926-30. Rasmussen, T.,Schliemann, T.,Sorensen, J.C.,Zimmer, J. and West, M.J. (1996). Memory impaired aged rats: no loss of principal hippocampal and subicular neurons. Neurobiol Aging 17(1): 143-7. Recce, M.L. and O'Keefe, J. (1989). The tetrode: an improved technique for multiunit extracellular recording. Soc Neurosci Abstr 15:1250. Reed, J.M.,Hamann, S.B.,Stefanacci, L. and Squire, L.R. (1997). When amnesic patients perform well on recognition memory tests. Behavioral Neuroscience 111(6): 1163-70. Reynolds, J.N. and Carlen, P.L. (1989). Diminished calcium currents in aged hippocampal dentate gyrus granule neurones. Brain Research 479(2): 384-90. Rial Verde, E.M.,Lee-Osbourne, J.,Worley, P.F.,Malinow, R. and Cline, H.T. (2006). Increased expression of the immediate-early gene arc/arg3.1 reduces AMPA receptor-mediated synaptic transmission. Neuron 52(3): 461-74. Riches, I.P.,Wilson, F.A. and Brown, M.W. (1991). The effects of visual stimulation and memory on neurons of the hippocampal formation and the neighboring parahippocampal gyrus and inferior temporal cortex of the primate. Journal of Neuroscience 11(6): 1763-79. Robitsek, R.J.,Fortin, N.J.,Koh, M.T.,Gallagher, M. and Eichenbaum, H. (2008). Cognitive aging: a common decline of episodic recollection and spatial memory in rats. Journal of Neuroscience 28(36): 8945-54. Rogawski, M.A. and Wenk, G.L. (2003). The neuropharmacological basis for the use of memantine in the treatment of Alzheimer's disease. CNS Drug Reviews 9(3): 275-308. Rolls, E.T.,Franco, L. and Stringer, S.M. (2005). The perirhinal cortex and long- term familiarity memory. Quarterly journal of experimental psychology. B, Comparative and physiological psychology 58(3-4): 234-45. Rosenzweig, E.S. and Barnes, C.A. (2003). Impact of aging on hippocampal function: plasticity, network dynamics, and cognition. Prog Neurobiol 69(3): 143-79. Rosenzweig, E.S.,Rao, G.,McNaughton, B.L. and Barnes, C.A. (1997). Role of temporal summation in age-related long-term potentiation-induction deficits. Hippocampus 7(5): 549-58. 326
Rosenzweig, E.S.,Redish, A.D.,McNaughton, B.L. and Barnes, C.A. (2003). Hippocampal map realignment and spatial learning. Nature Neuroscience 6(6): 609-15. Ryan, L.,Nadel, L.,Keil, K.,Putnam, K.,Schnyer, D.,Trouard, T. and Moscovitch, M. (2001). Hippocampal complex and retrieval of recent and very remote autobiographical memories: evidence from functional magnetic resonance imaging in neurologically intact people. Hippocampus 11(6): 707-14. Salat, D.H.,van der Kouwe, A.J.,Tuch, D.S.,Quinn, B.T.,Fischl, B.,Dale, A.M. and Corkin, S. (2006). Neuroimaging H.M.: a 10-year follow-up examination. Hippocampus 16(11): 936-45. Samsonovich, A. and McNaughton, B.L. (1997). Path integration and cognitive mapping in a continuous attractor neural network model. Journal of Neuroscience 17(15): 5900-20. Sargolini, F.,Fyhn, M.,Hafting, T.,McNaughton, B.L.,Witter, M.P.,Moser, M.B. and Moser, E.I. (2006). Conjunctive representation of position, direction, and velocity in entorhinal cortex. Science 312(5774): 758-62. Sauvage, M.M.,Fortin, N.J.,Owens, C.B.,Yonelinas, A.P. and Eichenbaum, H. (2008). Recognition memory: opposite effects of hippocampal damage on recollection and familiarity. Nature Neuroscience 11(1): 16-8. Schacter, D.L.,Koutstaal, W.,Johnson, M.K.,Gross, M.S. and Angell, K.E. (1997). False recollection induced by photographs: a comparison of older and younger adults. Psychol Aging 12(2): 203-15. Scheibel, A.B. (1979). The hippocampus: organizational patterns in health and senescence. Mechanisms of ageing and development 9(1-2): 89-102. Scheibel, M.E.,Lindsay, R.D.,Tomiyasu, U. and Scheibel, A.B. (1976). Progressive dendritic changes in the aging human limbic system. Experimental Neurology 53(2): 420-30. Schultz, W. (1998). Predictive reward signal of dopamine neurons. Journal of Neurophysiology 80(1): 1-27. Scoville, W.B. and Milner, B. (1957). Loss of recent memory after bilateral hippocampal lesions. Journal of Neurochemistry 20(1): 11-21. Segal, M. (1982). Changes in neurotransmitter actions in the aged rat hippocampus. Neurobiology of Aging 3(2): 121-4. Selden, N.R.,Everitt, B.J.,Jarrard, L.E. and Robbins, T.W. (1991). Complementary roles for the amygdala and hippocampus in aversive conditioning to explicit and contextual cues. Neuroscience 42(2): 335-50. Shamy, J.L.,Buonocore, M.H.,Makaron, L.M.,Amaral, D.G.,Barnes, C.A. and Rapp, P.R. (2006). Hippocampal volume is preserved and fails to predict recognition memory impairment in aged rhesus monkeys (Macaca mulatta). Neurobiology of Aging 27(10): 1405-15. Shen, B. and McNaughton, B.L. (1996). Modeling the spontaneous reactivation of experience-specific hippocampal cell assembles during sleep. Hippocampus 6(6): 685-92. 327
Shen, J.,Barnes, C.A.,McNaughton, B.L.,Skaggs, W.E. and Weaver, K.L. (1997). The effect of aging on experience-dependent plasticity of hippocampal place cells. Journal of Neuroscience 17(17): 6769-82. Shen, J.,Barnes, C.A.,Wenk, G.L. and McNaughton, B.L. (1996). Differential effects of selective immunotoxic lesions of medial septal cholinergic cells on spatial working and reference memory. Behavioral Neuroscience 110(5): 1181-6. Shen, J.,Kudrimoti, H.S.,McNaughton, B.L. and Barnes, C.A. (1998). Reactivation of neuronal ensembles in hippocampal dentate gyrus during sleep after spatial experience. Journal of Sleep Research 7 Suppl 1: 6-16. Shepherd, J.D.,Rumbaugh, G.,Wu, J.,Chowdhury, S.,Plath, N.,Kuhl, D.,Huganir, R.L. and Worley, P.F. (2006). Arc/Arg3.1 Mediates Homeostatic Synaptic Scaling of AMPA Receptors. Neuron 52(3): 475-484. Shrager, Y.,Kirwan, C.B. and Squire, L.R. (2008). Activity in both hippocampus and perirhinal cortex predicts the memory strength of subsequently remembered information. Neuron 59(4): 547-53. Sierra-Mercado, D.,Dieguez, D., Jr. and Barea-Rodriguez, E.J. (2008). Brief novelty exposure facilitates dentate gyrus LTP in aged rats. Hippocampus 18(8): 835-43. Sirota, A.,Csicsvari, J.,Buhl, D. and Buzsaki, G. (2003). Communication between neocortex and hippocampus during sleep in rodents. Proceedings of the National Academy of Sciences, U S A 100(4): 2065-9. Skaggs, W.E. and McNaughton, B.L. (1992). Computational approaches to hippocampal function. Current Opinion in Neurobiology 2(2): 209-11. Skaggs, W.E. and McNaughton, B.L. (1996). Replay of neuronal firing sequences in rat hippocampus during sleep following spatial experience. Science 271(5257): 1870-3. Skaggs, W.E.,McNaughton, B.L.,Gothard, K.M. and Markus, E.J. (1993). An information-theoretic to deciphering the hippocampal code. Advances in neural information processing Hanson, S.J., Cowan, J.D. and Giles, C.L. San Mateo, CA, Morgan Kaufmann Publishing. 5: 1030-1037. Skaggs, W.E.,McNaughton, B.L.,Wilson, M.A. and Barnes, C.A. (1996). Theta phase precession in hippocampal neuronal populations and the compression of temporal sequences. Hippocampus 6(2): 149-72. Small, S.A.,Chawla, M.K.,Buonocore, M.,Rapp, P.R. and Barnes, C.A. (2004). Imaging correlates of brain function in monkeys and rats isolates a hippocampal subregion differentially vulnerable to aging. Proceedings of the National Academy of Sciences, U S A 101(18): 7181-6. Small, S.A.,Tsai, W.Y.,DeLaPaz, R.,Mayeux, R. and Stern, Y. (2002). Imaging hippocampal function across the human life span: is memory decline normal or not? Annals of Neurology 51(3): 290-5. Smith, A.C.,Gerrard, J.L.,Barnes, C.A. and McNaughton, B.L. (2000). Effect of age on burst firing characteristics of rat hippocampal pyramidal cells. Neuroreport 11(17): 3865-71. 328
Smith, D.E.,Rapp, P.R.,McKay, H.M.,Roberts, J.A. and Tuszynski, M.H. (2004). Memory impairment in aged primates is associated with focal death of cortical neurons and atrophy of subcortical neurons. Journal of Neuroscience 24(18): 4373-81. Smith, T.D.,Adams, M.M.,Gallagher, M.,Morrison, J.H. and Rapp, P.R. (2000). Circuit-specific alterations in hippocampal synaptophysin immunoreactivity predict spatial learning impairment in aged rats. Journal of Neuroscience 20(17): 6587-93. Solstad, T.,Moser, E.I. and Einevoll, G.T. (2006). From grid cells to place cells: a mathematical model. Hippocampus 16(12): 1026-31. Squire, L.R.,Knowlton, B. and Musen, G. (1993). The structure and organization of memory. Annual Review of Psychology 44: 453-95. Squire, L.R.,Stark, C.E. and Clark, R.E. (2004). The medial temporal lobe. Annual review of neuroscience 27: 279-306. Squire, L.R.,Wixted, J.T. and Clark, R.E. (2007). Recognition memory and the medial temporal lobe: a new perspective. Nature Reveiws Neuroscience 8(11): 872-83. Squire, L.R. and Zola-Morgan, S. (1991). The medial temporal lobe memory system. Science 253(5026): 1380-6. Steffenach, H.A.,Witter, M.,Moser, M.B. and Moser, E.I. (2005). Spatial memory in the rat requires the dorsolateral band of the entorhinal cortex. Neuron 45(2): 301-13. Sutherland, G.R. and McNaughton, B. (2000). Memory trace reactivation in hippocampal and neocortical neuronal ensembles. Current Opinion in Neurobiology 10(2): 180-6. Suzuki, W.A. and Amaral, D.G. (1994). Perirhinal and parahippocampal cortices of the macaque monkey: cortical afferents. Journal of Comparative Neurology 350(4): 497-533. Suzuki, W.A. and Amaral, D.G. (1994). Topographic organization of the reciprocal connections between the monkey entorhinal cortex and the perirhinal and parahippocampal cortices. Journal of Neuroscience 14(3 Pt 2): 1856-77. Suzuki, W.A. and Amaral, D.G. (2003). Perirhinal and parahippocampal cortices of the macaque monkey: cytoarchitectonic and chemoarchitectonic organization. Journal of Comparative Neurology 463 (1): 67-91. Swanson, L.W. and Kohler, C. (1986). Anatomical evidence for direct projections from the entorhinal area to the entire cortical mantle in the rat. Journal of Neuroscience 6(10): 3010-23. Tanila, H.,Shapiro, M.,Gallagher, M. and Eichenbaum, H. (1997). Brain aging: changes in the nature of information coding by the hippocampus. Journal of Neuroscience 17(13): 5155-66. Teng, E. and Squire, L.R. (1999). Memory for places learned long ago is intact after hippocampal damage. Nature 400(6745): 675-7. 329
Terrazas, A.,Krause, M.,Lipa, P.,Gothard, K.M.,Barnes, C.A. and McNaughton, B.L. (2005). Self-motion and the hippocampal spatial metric. Journal of Neuroscience 25(35): 8085-96. Teyler, T.J. and DiScenna, P. (1986). The hippocampal memory indexing theory. Behavioral Neuroscience 100(2): 147-54. Teyler, T.J. and Rudy, J.W. (2007). The hippocampal indexing theory and episodic memory: updating the index. Hippocampus 17(12): 1158-69. Thibault, O.,Hadley, R. and Landfield, P.W. (2001). Elevated postsynaptic [Ca2+]i and L-type calcium channel activity in aged hippocampal neurons: relationship to impaired synaptic plasticity. Journal of Neuroscience 21(24): 9744-56. Thibault, O. and Landfield, P.W. (1996). Increase in single L-type calcium channels in hippocampal neurons during aging. Science 272(5264): 1017- 20. Thompson, L.T. and Best, P.J. (1990). Long-term stability of the place-field activity of single units recorded from the dorsal hippocampus of freely behaving rats. Brain Research 509(2): 299-308. Titov, N. and Knight, R.G. (1997). Adult age differences in controlled and automatic memory processing. Psychology and aging 12(4): 565-73. Toescu, E.C.,Verkhratsky, A. and Landfield, P.W. (2004). Ca2+ regulation and gene expression in normal brain aging. Trends in Neurosciences 27(10): 614-20. Tombaugh, G.C.,Rowe, W.B.,Chow, A.R.,Michael, T.H. and Rose, G.M. (2002). Theta-frequency synaptic potentiation in CA1 in vitro distinguishes cognitively impaired from unimpaired aged Fischer 344 rats. Journal of Neuroscience 22(22): 9932-40. Tombaugh, G.C.,Rowe, W.B. and Rose, G.M. (2005). The slow afterhyperpolarization in hippocampal CA1 neurons covaries with spatial learning ability in aged Fisher 344 rats. Journal of Neuroscience 25(10): 2609-16. Tonkikh, A.,Janus, C.,El-Beheiry, H.,Pennefather, P.S.,Samoilova, M.,McDonald, P.,Ouanounou, A. and Carlen, P.L. (2006). Calcium chelation improves spatial learning and synaptic plasticity in aged rats. Experimental Neurology 197(2): 291-300. Toth, J.P. and Parks, C.M. (2006). Effects of age on estimated familiarity in the process dissociation procedure: the role of noncriterial recollection. Mem Cognit 34(3): 527-37. Tun, P.A.,Wingfield, A.,Rosen, M.J. and Blanchard, L. (1998). Response latencies for false memories: gist-based processes in normal aging. Psychology and aging 13(2): 230-41. Turner, D.A. and Deupree, D.L. (1991). Functional elongation of CA1 hippocampal neurons with aging in Fischer 344 rats. Neurobiology of Aging 12(3): 201-10. 330
Uemura, E. (1985). Age-related changes in the subiculum of Macaca mulatta: synaptic density. Experimental Neurology 87(3): 403-11. UNC, I.o.A. (2006). United States Aging Demographics. Ungerleider, L.G. and Haxby, J.V. (1994). 'What' and 'where' in the human brain. Current Opinion in Neurobiology 4(2): 157-65. Van Cauter, T.,Poucet, B. and Save, E. (2008). Unstable CA1 place cell representation in rats with entorhinal cortex lesions. European Journal of Neuroscience 27(8): 1933-46. Vanderwolf, C.H. (1969). Hippocampal electrical activity and voluntary movement in the rat. Electroencephalogram Clinical Neurophysiolology 26(4): 407- 18. Vanderwolf, C.H.,Kramis, R. and Robinson, T.E. (1977). Hippocampal electrical activity during waking behaviour and sleep: analyses using centrally acting drugs. Ciba Found Symp(58): 199-226. Vannucchi, M.G.,Scali, C.,Kopf, S.R.,Pepeu, G. and Casamenti, F. (1997). Selective muscarinic antagonists differentially affect in vivo acetylcholine release and memory performances of young and aged rats. Neuroscience 79(3): 837-46. Vazdarjanova, A. and Guzowski, J.F. (2004). Differences in hippocampal neuronal population responses to modifications of an environmental context: evidence for distinct, yet complementary, functions of CA3 and CA1 ensembles. Journal of Neuroscience 24(29): 6489-96. Veng, L.M.,Mesches, M.H. and Browning, M.D. (2003). Age-related working memory impairment is correlated with increases in the L-type calcium channel protein alpha1D (Cav1.3) in area CA1 of the hippocampus and both are ameliorated by chronic nimodipine treatment. Brain Res Mol Brain Res 110(2): 193-202. Wais, P.E.,Squire, L.R. and Wixted, J.T. (2009). In Search of Recollection and Familiarity Signals in the Hippocampus. J Cogn Neurosci. Wais, P.E.,Wixted, J.T.,Hopkins, R.O. and Squire, L.R. (2006). The hippocampus supports both the recollection and the familiarity components of recognition memory. Neuron 49(3): 459-66. Wan, H.,Aggleton, J.P. and Brown, M.W. (1999). Different contributions of the hippocampus and perirhinal cortex to recognition memory. Journal of Neuroscience 19(3): 1142-8. Warburton, E.C.,Glover, C.P.,Massey, P.V.,Wan, H.,Johnson, B.,Bienemann, A.,Deuschle, U.,Kew, J.N.,Aggleton, J.P.,Bashir, Z.I.,Uney, J. and Brown, M.W. (2005). cAMP responsive element-binding protein phosphorylation is necessary for perirhinal long-term potentiation and recognition memory. Journal of Neuroscience 25(27): 6296-303. Warburton, E.C.,Koder, T.,Cho, K.,Massey, P.V.,Duguid, G.,Barker, G.R.,Aggleton, J.P.,Bashir, Z.I. and Brown, M.W. (2003). Cholinergic neurotransmission is essential for perirhinal cortical plasticity and recognition memory. Neuron 38(6): 987-96. 331
Weis, S.,Specht, K.,Klaver, P.,Tendolkar, I.,Willmes, K.,Ruhlmann, J.,Elger, C.E. and Fernandez, G. (2004). Process dissociation between contextual retrieval and item recognition. Neuroreport 15(18): 2729-33. Wenk, G.L. and Barnes, C.A. (2000). Regional changes in the hippocampal density of AMPA and NMDA receptors across the lifespan of the rat. Brain Research 885(1): 1-5. West, M.J. (1993). New stereological methods for counting neurons. Neurobiology of Aging 14(4): 275-85. West, M.J.,Coleman, P.D.,Flood, D.G. and Troncoso, J.C. (1994). Differences in the pattern of hippocampal neuronal loss in normal ageing and Alzheimer's disease. Lancet 344(8925): 769-72. Wiig, K.A. and Burwell, R.D. (1998). Memory impairment on a delayed non- matching-to-position task after lesions of the perirhinal cortex in the rat. Behavioral Neuroscience 112(4): 827-38. Williams, R.S. and Matthysse, S. (1986). Age-related changes in Down syndrome brain and the cellular pathology of Alzheimer disease. Progress in Brain Research 70: 49-67. Wills, T.J.,Lever, C.,Cacucci, F.,Burgess, N. and O'Keefe, J. (2005). Attractor dynamics in the hippocampal representation of the local environment. Science 308(5723): 873-6. Wilson, I.,S, I.,Gallagher, M.,Eichenbaum, H. and Tanila, H. (2005). Age- associated alterarions in place cells are subregion specific. Journal of Neuroscience 25: 6877-6886. Wilson, I.A.,Gallagher, M.,Eichenbaum, H. and Tanila, H. (2006). Neurocognitive aging: prior memories hinder new hippocampal encoding. Trends in Neurosciences 29(12): 662-70. Wilson, M.A. and McNaughton, B.L. (1993). Dynamics of the hippocampal ensemble code for space. Science 261(5124): 1055-8. Wilson, M.A. and McNaughton, B.L. (1994). Reactivation of hippocampal ensemble memories during sleep. Science 265(5172): 676-9. Winocur, G. (1990). Anterograde and retrograde amnesia in rats with dorsal hippocampal or dorsomedial thalamic lesions. Behavioural Brain Research 38(2): 145-54. Winter, J.C. (1997). The effects of age on continuous learning in the radial maze. Physiology and Behavior 61(4): 609-12. Winters, B.D. and Bussey, T.J. (2005). Glutamate receptors in perirhinal cortex mediate encoding, retrieval, and consolidation of object recognition memory. Journal of Neuroscience 25(17): 4243-51. Winters, B.D. and Bussey, T.J. (2005). Removal of cholinergic input to perirhinal cortex disrupts object recognition but not spatial working memory in the rat. European Journal of Neuroscience 21(8): 2263-70. Winters, B.D. and Bussey, T.J. (2005). Transient inactivation of perirhinal cortex disrupts encoding, retrieval, and consolidation of object recognition memory. Journal of Neuroscience 25(1): 52-61. 332
Winters, B.D.,Forwood, S.E.,Cowell, R.A.,Saksida, L.M. and Bussey, T.J. (2004). Double dissociation between the effects of peri-postrhinal cortex and hippocampal lesions on tests of object recognition and spatial memory: heterogeneity of function within the temporal lobe. Journal of Neuroscience 24(26): 5901-8. Winters, B.D.,Saksida, L.M. and Bussey, T.J. (2006). Paradoxical facilitation of object recognition memory after infusion of scopolamine into perirhinal cortex: implications for cholinergic system function. Journal of Neuroscience 26(37): 9520-9. Wise, R.A. (2006). Role of brain dopamine in food reward and reinforcement. Philosophical Transactions of the Royal Society of London Series B, Biological Sciences 361(1471): 1149-58. Witter, M.P.,Naber, P.A.,van Haeften, T.,Machielsen, W.C.,Rombouts, S.A.,Barkhof, F.,Scheltens, P. and Lopes da Silva, F.H. (2000). Cortico- hippocampal communication by way of parallel parahippocampal-subicular pathways. Hippocampus 10(4): 398-410. Wixted, J.T. and Stretch, V. (2004). In defense of the signal detection interpretation of remember/know judgments. Psychon Bull Rev 11(4): 616- 41. Xiang, J.Z. and Brown, M.W. (1998). Differential neuronal encoding of novelty, familiarity and recency in regions of the anterior temporal lobe. Neuropharmacology 37(4-5): 657-76. Yang, Y.,Ge, W.,Chen, Y.,Zhang, Z.,Shen, W.,Wu, C.,Poo, M. and Duan, S. (2003). Contribution of astrocytes to hippocampal long-term potentiation through release of D-serine. Proceedings of the National Academy of Sciences, U S A 100(25): 15194-9. Yang, Z.,Krause, M.,Rao, G.,McNaughton, B.L. and Barnes, C.A. (2008). Synaptic commitment: developmentally regulated reciprocal changes in hippocampal granule cell NMDA and AMPA receptors over the lifespan. Journal of Neurophysiology 99(6): 2760-8. Ylinen, A.,Bragin, A.,Nadasdy, Z.,Jando, G.,Szabo, I.,Sik, A. and Buzsaki, G. (1995). Sharp wave-associated high-frequency oscillation (200 Hz) in the intact hippocampus: network and intracellular mechanisms. Journal of Neuroscience 15(1 Pt 1): 30-46. Yonelinas, A.P. (1994). Receiver-operating characteristics in recognition memory: evidence for a dual-process model. J Exp Psychol Learn Mem Cogn 20(6): 1341-54. Yonelinas, A.P. (1997). Recognition memory ROCs for item and associative information: the contribution of recollection and familiarity. Mem Cognit 25(6): 747-63. Yonelinas, A.P. (1999). The contribution of recollection and familiarity to recognition and source-memory judgments: a formal dual-process model and an analysis of receiver operating characteristics. J Exp Psychol Learn Mem Cogn 25(6): 1415-34. 333
Yonelinas, A.P. (2001). Components of episodic memory: the contribution of recollection and familiarity. Philosophical Transactions of the Royal Society of London Series B, Biological Sciences 356(1413): 1363-74. Yonelinas, A.P.,Kroll, N.E.,Quamme, J.R.,Lazzara, M.M.,Sauve, M.J.,Widaman, K.F. and Knight, R.T. (2002). Effects of extensive temporal lobe damage or mild hypoxia on recollection and familiarity. Nature Neuroscience 5(11): 1236-41. Yonelinas, A.P.,Otten, L.J.,Shaw, K.N. and Rugg, M.D. (2005). Separating the brain regions involved in recollection and familiarity in recognition memory. Journal of Neuroscience 25(11): 3002-8. Yonelinas, A.P. and Parks, C.M. (2007). Receiver operating characteristics (ROCs) in recognition memory: a review. Psychol Bull 133(5): 800-32. Yoshida, M.,Naya, Y. and Miyashita, Y. (2003). Anatomical organization of forward fiber projections from area TE to perirhinal neurons representing visual long-term memory in monkeys. Proc Natl Acad Sci U S A 100(7): 4257-62. Young, B.J.,Otto, T.,Fox, G.D. and Eichenbaum, H. (1997). Memory representation within the parahippocampal region. Journal of Neuroscience 17(13): 5183-95. Zhu, X.O. and Brown, M.W. (1995). Changes in neuronal activity related to the repetition and relative familiarity of visual stimuli in rhinal and adjacent cortex of the anaesthetised rat. Brain Research 689(1): 101-10. Zhu, X.O.,Brown, M.W. and Aggleton, J.P. (1995). Neuronal signalling of information important to visual recognition memory in rat rhinal and neighbouring cortices. European Journal of Neuroscience 7(4): 753-65. Zhu, X.O.,Brown, M.W.,McCabe, B.J. and Aggleton, J.P. (1995). Effects of the novelty or familiarity of visual stimuli on the expression of the immediate early gene c-fos in rat brain. Neuroscience 69(3): 821-9. Zhu, X.O.,McCabe, B.J.,Aggleton, J.P. and Brown, M.W. (1997). Differential activation of the rat hippocampus and perirhinal cortex by novel visual stimuli and a novel environment. Neuroscience Letters 229(2): 141-3. Ziakopoulos, Z.,Tillett, C.W.,Brown, M.W. and Bashir, Z.I. (1999). Input-and layer-dependent synaptic plasticity in the rat perirhinal cortex in vitro. Neuroscience 92(2): 459-72. Zola-Morgan, S.M. and Squire, L.R. (1990). The primate hippocampal formation: evidence for a time-limited role in memory storage. Science 250(4978): 288-90. Zola, S.M. and Squire, L.R. (2001). Relationship between magnitude of damage to the hippocampus and impaired recognition memory in monkeys. Hippocampus 11(2): 92-8.