THE AGING HIPPOCAMPUS: A MULTILEVEL ANALYSIS IN THE RAT 
IRA DRISCOLL 
M.S., The University of New Mexico, 2002 
B.A., The University of New Mexico, 2000 
A Thesis 
Submitted to the School of Graduate Studies 
of the University of Lethbridge 
in Partial Fulfilment of the 
Requirements for the Degree 
PH.D., NEUROSCIENCE 
Canadian Centre for Behavioural Neuroscience, Department of Neuroscience 
University of Lethbridge 
LETHBRIDGE, ALBERTA, CANADA 
© Ira Driscoll, 2005 
Abstract 
The purpose of the current thesis was twofold: (1) to examine various factors 
that might be contributing to age-related learning and memory deficits 
specifically related to the hippocampus, and (2) to validate our rat model of 
aging, employing a multilevel analysis. We found age-related deficits on both 
spatial and non-spatial hippocampus-dependent tasks that were accompanied by 
structural alterations observed both in vivo (volume, but not neuronal metabolic 
function) and post mortem (neuronal density and neurogenesis, but not synaptic 
or mitochondrial density). Furthermore, our results suggest that the observed 
hippocampal structural changes, namely decreased volume and neurogenesis, 
predict learning and memory deficits, and both can be accounted for by 
neurogenic reduction. In addition, the above-mentioned pattern of age-related 
deficits closely resembles that seen in humans, suggesting the present rat 
version of aging to be a very useful model for investigating hippocampal aging in 
humans. 
Acknowledgements 
First and foremost, I would like to thank my advisor, Dr. Robert Sutherland, 
for this amazing opportunity. I feel very lucky to have been given a chance 
and am grateful for your invaluable guidance, mentorship, and support. I 
would also like to thank my committee members, Dr. William Brooks, Dr. 
Bryan Kolb, Dr. Ian Whishaw, and Dr. Olga Kovalchuk, for their valuable 
recommendations and support pertaining to this thesis. Special thanks to Ian 
Whishaw, for going over and above his call of duty and providing precious 
advice and learning opportunity. One should be so lucky to always have 
someone like you around, Ian. Dr. Kolb, for the opportunity to learn and 
conduct some research in your laboratory. Also, my sincere gratitude to Dr. 
Brooks for his continuous support and consultation, and for coming to 
Lethbridge multiple times on my behalf. 
I would like to acknowledge the Calgary group from the Experimental 
Imaging Centre - Dr. Boguslaw Tomanek, Tadeus Foniok, and Dave Kirk. It 
was a pleasure working with all of you. Special thanks to Dave for providing 
the entertainment during scanning and to Tad for putting up with it. My 
deepest gratitude to Helen Petropoulos and Laura Rowland for providing a 
spectroscopy hotline, and for the friendship and support that you've shown to 
me throughout the years. 
Many, many thanks to Marie-Helene Monfils. A mere word of thanks is not 
enough for all that you have done, nor is a page to list all that I'm thankful 
for. Marie, I will always be happy to look through peepholes for you. Sarah 
Howard and Jared Stone, I feel privileged to have worked with you. Your 
hard work and friendship are greatly appreciated. Kristin Johnson, for sharing 
her Ki67 procedure with me. Doug Bray, for teaching me how to use an 
Electron Microscope and all that is associated with it, but also for having a 
seemingly inexhaustible amount of patience. 
Very special thanks to Glen Prusky for the opportunity to collaborate and to 
toboggan in Mountain View, and to the members of his laboratory for 
accepting me as one of their own, especially Trevor McGill and Byron Silver. 
Rob McDonald and his laboratory for collaborations, especially Nhung Hong -
it's been a pleasure working with you. 
Special thanks to Janice Sutherland for many years of support and friendship. 
Naomi Cramer, for being so kind and helpful from the very first moment I 
met you. Afra Foroud, for being a dear friend and talking me into dancing 
again. Christine Reinhart, for making sure I wasn't the only quirky one 
around - that's what friends are for. Penny Vandenberg, for making my life 
so much easier. Daniel Poda and Dawn Danka, for their help and friendship. 
Many thanks to various people at the CCBN whose kindness is greatly 
appreciated. And to those not so kind - it just goes to show that what 
doesn't kill you, certainly does make you stronger, so I suppose thanks are in 
order. 
And last, but not least, to my family. To my husband and my best friend, 
Jason, thank you for the many years of support and for always believing in 
me, but most of all for the immense amount of patience, understanding, and 
sacrifice on your part. To my parents, who sacrificed everything they had so 
that I could have opportunities. I hope I made you proud. To my sister, 
Marijam, who provided immeasurable support and entertainment over the 
years. And finally to the rest of my family and all of my friends, your 
encouragement is greatly appreciated. 
Table of Contents 
1. Introduction 1 
2. The Hippocampal Formation 3 
3. The Function of the Hippocampus 5 
3.1 Rationale for Memory Tests of Hippocampal Function 8 
3.2 Hippocampus-dependent Learning and Memory in 
Rodents 15 
3.2.1 Spatial Navigation 16 
3.2.2 Transverse Patterning Discrimination 18 
3.3 Hippocampus-dependent Learning and Memory in 
Humans 20 
3.3.1 Virtual Navigation 2
3.3.2 Transverse Patterning Discriminations 26 
4. The Aging Hippocampus 28 
4.1 Physiological Changes 2
4.2 Structural Changes 31 
4.3 Biochemical Changes 34 
4.4 Neurogenesis 9 
5. Human Model of Rat Aging 46 
6. Experiment 1 - Visual Acuity 53 
6.1 Background
6.2 Materials and Methods 55 
6.3 Results 61 
6.4 Discussion 2 
7. Experiment 2 - Long -Term Potentiation 64 
7.1 Background 64 
7.2 Materials and Methods 7 
7.3 Results 70 
7.4 Discussion 3 
8. Experiment 3 - Spatial Learning and Memory 78 
8.1 Background 7
8.2 Materials and Methods 80 
8.3 Results 82 
8.4 Discussion 5 
9. Experiment 4 - The Transverse Patterning Problem 88 
9.1 Background 8
9.2 Materials and Methods 91 
9.3 Results 96 
9.4 Discussion 8 
10. Experiment 5 - Non-spatial Learning and Memory 104 
10.1 Background 10
10.2 Materials and Methods 105 
10.3 Results 108 
10.4 Discussion 110 
11. Experiment 6 - Magnetic Resonance Imaging and Spectroscopy 113 
11.1 Background 11
11.2 Materials and Methods 116 
11.3 Results 120 
11.4 Discussion 1 
12. Experiment 7 - Light and Electron Microscopy 124 
12.1 Background 12
12.2 Materials and Methods 126 
12.3 Results 132 
12.4 Conclusion 4 
13. Experiment 8 - Neurogenesis 137 
13.1 Background
13.2 Materials and Methods 139 
13.3 Results 143 
13.4 Discussion 5 
14. Correlations 149 
14.1 Partial Correlations 155 
15. General Discussion 158 
16. Closing Comments 168 
17. Conclusion 170 
18. References 173 
List of Tables 
Discrimination training procedures. 
Partial correlations. 
List of Figures 
1. The Visual Water Task. 195 
2. Visual acuity - Static gratings. 196 
3. (A and B) Visual acuity - OptoMotry. 197 
4. Electrophysiology - Electrode placements. 198 
5. Effects of HFS sessions on the evoked potentials. 199 
6. Representative field potentials. 200 
7. Morris water task performance - hidden platform trials. 201 
8. Morris water task performance - 24-hour retention. 202 
9. Morris water task performance - visible platform trials. 203 
10.The Visual Discriminanda. 204 
11. A representative NMDA damage. 205 
12. Elemental discrimination performance - Lesions. 206 
13. Learning rate - Elemental discriminations. 207 
14. Transverse patterning discriminations - Lesions. 208 
15. Elemental discriminations - Aging. 209 
16. Transverse patterning discriminations - Aging. 210 
17. A representative MR image of the rat hippocampus. 1 
18. (A and B) Magnetic Resonance Spectroscopy (MRS) acquisition. 212 
18. (C) A representative raw MRS spectrum. 213 
19. Age-related differences in intracranial volume. 214 
20. Age-related differences in absolute hippocampal volume. 215 
21. Age-related differences in normalized hippocampal volume. 216 
22. Semi-thin (1 ^m) section selection for light microscopy. 217 
23. Electron micrograph depicting a larger area containing synapses. 218 
24. Electron micrograph depicting a larger area containing mitochondria. 219 
25. Changes in neuronal density with age. 220 
26. A confocal image of BrdU/NeuN double-labeled cells. 221 
27. Representative Doublecortin labeling. 222 
28. Representative Ki67 labeling. 223 
29. The number of BrdU labeled cells. 224 
30. The number of BrdU/NeuN double-labeled cells. 225 
31. The number of DCX-labeled cells. 226 
32. The number of Ki67-labeled cells. 7 
33. Correlations between MWT and hippocampal volume. 228 
34. Correlations between TPDT and hippocampal volume. 229 
35. Correlations between MWT and DCX-labeled cells. 230 
36. Correlations between TPDT and DCX-labeled cells. 231 
37. Correlations between hippocampal volume and DCX-labeled cells. 232 
38. Correlations between BrdU/NeuN- and DCX-labeled cells. 233 
39. Correlations between Ki67-labeled cells and TPDT. 234 
40. Correlations between Ki67-labeled cells and hippocampal volume. 235 
41. Correlations between Ki67- and DCX-labeled cells. 236 
1. Introduction 
When I was younger, I could remember anything, whether it had happened or 
not; but my faculties are decaying now and soon I shall be so I cannot 
remember any but the things that never happened. It is sad to go to pieces like 
this but we all have to do it. ~ Mark Twain 
The study of aging attempts to understand the changing capacities of the elderly 
as a normal developmental process and the understanding of the neurobiology 
of aging is an integral part of that framework. Developmental boundaries and 
failure to manifest certain changes in the usual course of development are easily 
recognized early in life, such as failure to walk or talk by certain age. On the 
other end of the developmental spectrum the boundaries of normal aging as 
distinct from pathology, however, are not clearly defined. Age-related changes 
do not seem to differ from manifestations of age-related pathology, at least not 
in the early stages. Both normal and pathological aging are characterized by 
learning and memory impairments, and there is no consensus on the precise 
nature of underlying neurobiological changes. 
Alzheimer's disease, on the other hand, is a debilitating neurological disorder of 
old age that is characterized by a progressive loss of cognitive function, memory 
loss being a hallmark. AD has damaging effects on the quality and enjoyment of 
life of the elderly people, both the ones suffering from the disease and those 
1 
caring for them. Currently, efforts are under way to develop treatments for AD, 
even though we do not fully understand what causes the disease. As critical as it 
is to find the cure for this devastating condition, a good understanding of 
boundaries between normal and pathology is necessary. 
Various cognitive domains, subserved by many different brain regions, are 
affected by aging. Here we restrict our considerations to the tasks that are 
dependent on the integrity of the hippocampal formation. This focus is based on 
the fact that age-related cognitive changes, normal and pathological, are 
intimately linked to the hippocampal formation. The hippocampus has 
maintained a central role in the realm of learning and memory research over the 
past 40 years, particularly in memory for cue conjunctions, best exemplified in 
memory for contexts or spatial locations. 
One of the broad aims here is to contribute to the understanding of 
hippocampus-dependent learning and memory by studying age-related changes 
in both behavior and in underlying neurobiology. First, the current state of the 
literature in relation to hippocampal function will be briefly reviewed. Then, the 
functional deficits will be described and appropriate parallels drawn between the 
human and rat behavioral literature. Further, magnetic resonance imaging and 
magnetic resonance spectroscopy findings will be considered to draw plausible 
links with hippocampus-dependent cognitive impairments that have been 
observed in both aged humans and rodents. Finally, the current findings will be 
c onsidered and to conclude, a prospective view will be offered. 
2 
2. The Hippocampal Formation 
The hippocampus, a medial temporal lobe structure, is a part of the cerebral 
cortex and the limbic system. Even though the hippocampus is not a part of the 
superficial cortex, it is not a "subcortical" structure. The hippocampus is 
bilaterally symmetrical and it is comprised of the hippocampus proper 
(principally CA1 and CA3) and the dentate gyrus (fascia dentata). Each field 
contains a densely packed sheet of cells, both pyramidal and granular, which are 
the principal cell types found in this structure. In addition, each field contains a 
variety of interneurons that are intermixed with the principal cell types. The 
hippocampus proper contains CA1, CA2, CA3, and CA4 fields and is primarily 
comprised of pyramidal cells. CA2 and CA4 fields are very small and not well 
studied. In contrast, CA1 and CA3 fields have been more extensively explored. 
CA1 cells enter the alveus and further project to the subiculum, the lateral 
septal nuclei, and the prefrontal cortex. The CA3 pyramidal cells project 
primarily to the CA1 region via the Schaffer collaterals, to the lateral septal 
nuclei via fornix, and have recurrent excitatory connections. The dentate gyrus, 
the principal recipient of projections from the entorhinal cortex, is primarily 
comprised of granule cells. It is commonly divided into an upper and a lower 
blade, between which reside the large polymorph cells in the hilus of the dentate 
gyrus. The hilar cells play an interesting role in providing efferents that form the 
part of the hippocampal commissural and ipsilateral association system, the role 
of which is to provide interhemispheric and intrahemispheric associative 
connections between components of the hippocampus. Axons from the dentate 
3 
gyrus make connections with dendrites of CA3 pyramidal cells via the mossy 
fibers. Input and output from the neocortex happens via the entorhinal cortex, 
which projects to the dentate gyrus, and CA1 and CA3 fields via the perforant 
path. The CA1 field has projections back to the entorhinal cortex via subiculum. 
For an extensive review of hippocampal anatomy we refer the reader to articles 
by Amaral & Witter (1989, 1995). 
4 
3. The Function of the Hippocampus 
There is now agreement that the hippocampus, although it may have multiple 
functions, is critically involved in learning and memory. On the other hand, no 
general consensus has emerged regarding the characteristics of memory 
processes that are dependent on the hippocampus. According to Nadel and 
Moscovitch (2001), three fundamental questions emerge in relation to the 
involvement of this structure in learning and memory. One question is whether 
all types of learning and memory are affected by hippocampal damage. The 
second question is linked to hippocampus-related amnesia, specifically whether 
amnesia is best characterized as a deficit in memory storage or retrieval. And 
the third question asks whether the hippocampus plays an equivalent role in 
remote or recent memories. To date, only the first of these three questions 
seems to be adequately resolved. 
Research on memory has traditionally focused on trying to characterize what 
amnesic patients can or cannot do. Nearly all investigators now agree that most 
insults to the hippocampal formation alter its function, whether permanently or 
temporarily, and as a consequence certain information, whether learned before 
or after the insult, is disrupted. Most of the early studies on hippocampal 
function sought to identify tasks that are either profoundly impaired or else 
completely spared following hippocampal damage. In considering the function of 
the hippocampus and the changes in hippocampus-dependent cognition as we 
age, it is necessary to consider it from a historical perspective first. 
5 
The pioneering reports on amnesic patient H.M. (Scoville and Miller, 1957), and 
similar subsequent patients with medial temporal lobe damage that included the 
hippocampus, revealed that short-term, procedural, and implicit memory are 
apparently spared even though a significant portion of the hippocampus and 
sometimes surrounding structures were missing. Cumulatively, assessments of 
medial temporal lobe damaged patients revealed that not all types of learning 
and memory are affected by hippocampal damage. Initially it was reported that 
the memory deficits were limited to the information from experiences within a 
few years prior to the damage while older memories remained spared. 
In particular, for memories acquired before the hippocampus was damaged, 
more recent memories were lost (retrograde amnesia) but more remote 
memories appeared to be spared. This kind of dissociation between recent and 
remote memories (temporally-graded retrograde amnesia) is consistent with the 
idea that neither retrieval nor very long-term memory storage require intact 
hippocampal circuitry, and further, that only for recently acquired memories 
hippocampus plays an essential role in storage and/or retrieval (Corkin, 1984; 
Milner, 1966). The claims about the temporally-graded processes in retrograde 
amnesia came under increasingly critical scrutiny in recent years, partly due to 
new information gained from additional memory testing in medial temporal lobe 
damaged patients, and partly as a consequence of the emerging evidence from 
functional imaging. 
6 
Several studies employing fMRI have measured hippocampal activation while 
participants were remembering recently acquired vs. remotely acquired 
information. The simplest prediction, considering the results showing temporally 
graded retrograde amnesia, is that there should be relatively less hippocampal 
activation as more remote memories are retrieved. In contrast to this prediction, 
studies found that the hippocampus is activated to the same extent during 
retrieval of recent and remote memories (Maguire, 1999; Ryan et al., 2001). 
Also, the period of retrograde amnesia in patients with damage restricted to the 
hippocampal formation is now known to extend for decades (Cipolloti et al., 
2001). 
The non-human animal experimental literature is no less controversial with 
retrograde gradients reported to last a few days to months to no retention of 
remote information after large hippocampal lesions (Anagnostaras et al., 1999; 
Bohbot et al., 1998, Clark et al., in press, 2002; Sutherland et al., 2001; 
Winocur, 1990; Winocur et al., 2001). The debate about differential involvement 
of the hippocampus in recent vs. remote memories will continue for some time 
to come before a consensus is achieved. Future research will likely center on two 
issues: the extent of the lesion and the types of memory tests employed. Also, 
neuropsychological studies of human memory impairments, which are generally 
limited to opportunistic encounters with patients, usually involve limited, but 
rarely precise or complete damage to the hippocampus. The precise 
experimental controls, in order to understand memory function in detail, can 
only be imposed through research on non-human animals. A major impediment 
7 
to non-human animal research, however, has been the assessment of 
declarative or episodic memory. 
Episodic memory (Tulving, 1986) is thought to be a characteristic unique to 
humans. Still, recent studies of memory for cache sites in food storing jays 
(Griffiths and Clayton, 2001) indicate that several features of episodic memory 
are present, suggesting that the essential properties may not be as unique to 
humans as some may think. Such research highlights the need for development 
of appropriate animal models in order to bridge the gap between human and 
non-human animal memory research. Therefore, the hippocampus-dependent 
tasks and theories of hippocampal function that emphasize cross-species 
comparisons are essential. To that end, the current state of the literature in 
relation to hippocampal function is briefly reviewed. 
3.1 Rationale for Memory Tests of Hippocampal Function 
There are many theories of hippocampal function. In fact, for the person outside 
the area, the variety is bewildering. Despite the variety, most modern theories 
agree that the hippocampus is essential for learning of certain kinds of 
information and that other systems outside the hippocampus are able to acquire 
other types of information independently. One common theme found in these 
theories is that the hippocampus creates a complex representation of an event 
or an episode such that conjunctions of co-occuring or related cues are bound 
8 
together into a unitary representation. For certain memories the hippocampus 
plays an essential role in storage and retrieval for at least months and perhaps 
for decades after the occurrence of the learning episode. The hippocampus is 
thought to play a central role in memories for facts and for personal events, and 
is considered to support conscious recollections of experiences. Many 
fundamental aspects of hippocampal functioning remain a matter of debate and 
so it would be a mistake to suggest that there is a satisfying consensus. In 
order to get a feel for the current state of literature on hippocampal function, 
some of the more influential theories of hippocampal function are reviewed 
below. 
A commonly held view is that the hippocampus plays a fundamental role in a 
formation of unified representations of an event by integrating multiple cues 
related to that event (Eichenbaum, 2001; Gaffan, 1994; O'Reilly and Rudy, 
2001; Rudy and Sutherland, 1995; Squire, 1992; Sutherland and Rudy, 1989; 
Wickelgren, 1979). Possibly the most influential theory of hippocampal function 
to emerge in the past three decades is the cognitive mapping theory of O'Keefe 
& Nadel (1978). 
The principal driving force behind the development of cognitive mapping theory 
was the discovery of place cells (O'Keefe and Dostrovsky, 1971). The most 
salient event controlling the firing rate of these cells in a freely moving animal is 
the location of the animal within a particular environment (O'Keefe and Conway, 
1978; O'Keefe and Dostrovsky, 1971). According to the cognitive mapping 
9 
theory, learning and memory are supported by two systems - the taxon system 
and the locale system. The taxon system acquires information independently of 
the hippocampus and is involved in habit learning or stimulus-response learning. 
The locale system, on the other hand, depends on the integrity of the 
hippocampal formation and supports cognitive mapping. The locale system, 
centered on the hippocampus, creates and maintains a unified map-like 
representation of an environment, which is essentially a representation of 
places, directions, and distances between them. Whenever the discrepancy 
between the stored map and the current perception of the environment is 
detected, information is updated in an all-or-none manner by exploration, in a 
way that is independent of goal-oriented behavior. 
Also, briefly, at approximately the same time, Hirsh (1974) distinguished 
between two systems, one system involving performance-line storage and the 
second memory system involving conditional relationships between events. Only 
the latter he linked to the hippocampal formation. O'Keefe and Nadel (1978) 
along with Hirsh (1974) appear to be the first theorists to have made explicit the 
multiple memory system approach and applied it to explain learning and 
memory deficits. 
Theories of associative learning, in which learning is described in terms of the 
associative strength among stimuli and events, as well as the error-correcting 
rules by which modifications of associative strengths occur (Mackintosh, 1975; 
Pierce and Hall, 1980; Rescorla and Wagner, 1972), provide a principled account 
10 
of how learning, including learning about places, occurs. Another psychological 
theory of hippocampal function, based on associative learning principles, is 
Sutherland & Rudy's (1989) configural association theory. Configural association 
theory (see also Rudy and Sutherland, 1985) proposes that the hippocampal 
formation contributes to learning and memory by storing configural 
representations and enhancing the activation or salience of a more limited set of 
configural or conjunctive representations in the cortex. 
According to this theory, learning is also dependent on multiple systems. One 
system depends upon the hippocampus and supports the creation of complex 
representations. It does so by binding two or more features or cues together, 
which are subsequently used to solve many spatial and non-spatial memory 
problems. In principle, such a system could make a contribution to resolving any 
discrimination problem that contains a relationship between two or more cues or 
events. Conjoining two or more cues, which can be spatial or non-spatial in 
nature, forms the configural representation. 
The configural association theory predicts that animals with hippocampal 
damage are especially susceptible to deficits in discrimination problems that 
require configural (non-linear) solutions (i.e. cannot be solved by the 
combination of the individual associative strengths of the individual stimuli 
comprising the compound). An example of a task that requires a configural 
solution is the transverse patterning task (A+B-; B+C-; C+A-; which underlies 
the problem of "Rock-Paper-Scissors" on a more familiar level), a task that will 
11 
be discussed in more detail later. Some other tasks that engage a configural 
solution include negative patterning (A+; B+; AB-), certain conditional 
discriminations (AC+; B+; AB-; C-), and bi-conditional discriminations (AC+; 
BD + ; AD-; BC-; see Rudy and Sutherland, 1985 for a review). Even when 
problems can be solved using the summation of the associative strengths of cue 
representations in nonhippocampal networks, it should be clear that in the intact 
animal the hippocampus will construct a configural (conjunctive) representation 
from the same cues. Depending upon the relative rate of associative strength 
acquisition and the number of learning episodes, it is possible that the 
hippocampus could dominate performance even in problems with linear 
solutions. 
In this regard, O'Reilly & Rudy (2001) proposed a theory that is akin to 
configural theory, but with additional assumptions about the different properties 
of learning in hippocampal and nonhippocampal systems. The emergence of this 
theory was in part motivated by reports showing that animals with hippocampal 
damage are able to learn to solve at least some configural problems. O'Reilly & 
Rudy (2001) suggest that two separate systems contribute to the formation of 
configural or conjunctive representations and they provide a principled basis for 
deciding when memory tasks would favor one system over another. One system 
is hippocampus-dependent. It rapidly and automatically forms and stores 
conjunctive (configural) representations during incidental learning. The second 
system is neocortex-dependent (and not hippocampus dependent), and it 
supports learning of simple discriminations, and can form configural 
12 
representations. This system also supports stimulus conjunctions 
(configurations), but it does so gradually only if driven by task demands. 
In a similar vein, Eichenbaum (1992, 2001) proposed a relational theory of 
hippocampal function. In their view, the hippocampus supports relational 
representations involving multiple events and plays a central role in declarative 
memory. Declarative memories involve conscious recollections of events and 
may be called upon in the future to solve novel problems. The relational theory 
proposes that the hippocampal relational network allows for logical inferences to 
be made about events or items that may have never been experienced together 
in the past. This representational flexibility, they claim, allows for previously 
learned information to be applied in novel situations involving different actions. 
Evidence bearing on central predictions of this theory is not unequivocal. In 
particular, the sort of transitive inference process claimed to be dependent on 
hippocampal circuitry has been shown computationally not to require relational 
processing (Frank et al., 2003). Furthermore, transitive inference has been 
shown to be unaffected by large hippocampal lesions (Van Elzakker et al., 
2003). Similarly problematic is the finding that rats with large hippocampal 
lesions show transfer of perfect discrimination performance to novel 
combinations of previously trained reinforced and non-reinforced visual cues 
(Driscoll et al., 2004b) in the Visual water task (Prusky et al, 2000a). 
13 
As a final contemporary theory discussed here, Squire (1992) has presented a 
position on hippocampal function that has been very influential over the years. 
This theory proposes that there are multiple memory systems with distinct 
anatomical organizations and distinct functions. The hippocampus, along with 
anatomically related structures, is fundamental for declarative memory, that is, 
consciously recalled, reportable memories for events or facts. The hippocampus 
is involved in storing a representation that is involved in binding distributed sites 
in the neocortex in order to form complete, accessible memories. Squire (1992) 
proposes that there is a gradual process of reorganization that permits 
memories to be retrieved via strengthened cortical connections independently of 
the hippocampus. Declarative memory is contrasted with implicit (procedural) 
memory, which does not dependent on the hippocampus. 
Taken together, these theories of hippocampal function agree that the 
hippocampus contributes to the binding of memories of the components of 
events, including the spatio-temporal contexts within which they occur. The 
precise processes taking place in the hippocampus remain, however, a matter of 
ongoing debates and the above discussion of theories of hippocampal function is 
by no means exhaustive. The fact that the hippocampus is essential to spatial 
navigation, however, is largely agreed upon. 
The position that is favored here is that the hippocampal formation is essential 
for initial encoding, and for at least initial storage and retrieval of information. 
Furthermore, the loss in processing ability afforded by this system produces 
14 
deficits in learning and memory that accompany the aging of the hippocampal 
formation. Thus, we are in agreement with many theories of the hippocampal 
function (Hirsch, 1974; Mishkin et al., 1984; O'Keefe and Nadel, 1978; O'Reilly 
and Rudy, 2001; Rudy and Sutherland, 1995; Sutherland and Rudy, 1989), 
which posit that people or animals with damage to the hippocampus solve 
simple discrimination problems involving unambiguous single cues by gradually 
incrementing and decrementing the associative value of the elemental cues, but 
have difficulty solving problems that require configural/conjunction solutions. 
3.2 Hippocampus-dependent Learning and Memory in Rodents 
As already mentioned, many memory tasks are sensitive to hippocampal 
damage. Non-human animal research provides us with a wealth of knowledge 
about tasks that are dependent on hippocampal integrity, permitting further 
exploration of structural, functional, and molecular substrates. Two tasks that 
have been initially employed in testing non-human animals, and subsequently 
applied to humans, are of particular interest to us: the transverse patterning 
discrimination task (TPDT) and the Morris water task (MWT); non-spatial and 
spatial hippocampus-dependent tasks respectively. 
15 
3.2.1 Spatial Navigation 
A disruption in spatial learning and memory has been repeatedly observed in 
animals with hippocampal damage (Morris et al., 1982; Olton and Samuelson, 
1976; Sutherland et al., 1982a). The Morris water task (1981) has become the 
gold standard for measuring spatial learning and memory in rodents. In its most 
commonly used version, the Morris water task involves animals navigating to a 
platform hidden in a fixed location in a pool of opaque water based on learning a 
configuration of room cues outside the pool. Learning is assessed by measuring 
approach latencies or path length traveled from the release location to the 
platform. Both measures should improve across days and trials in a normal rat. 
Memory retention is often assessed during a fixed duration probe trial with the 
hidden platform removed from the pool. Intact animals swim over the location 
where the platform had been during training several times during the probe trial 
and have a selective, biased search pattern that reflects a preference for this 
region of the pool. For a rat to display a selective search bias, the rat should 
spend more time in the quadrant where the platform was located during training 
compared to any other quadrant (the same measures of performance are 
adopted in assessing success in human performance on virtual navigation tasks, 
which will be discussed below). Several types of hippocampal damage, including 
lesions (Morris et al, 1982; Sutherland et al, 1982a), pharmacological 
manipulations (Packard and White, 1991; Sutherland et al., 1982b), genetic and 
environmental factors (Gianoulakis, 1990; Zaharia et al., 1996), and aging 
(Barnes et al., 1997) are known to disrupt normal place learning and abolish 
16 
memory for previously acquired information. Specific to the Morris water task 
performance, rats with hippocampal damage do not selectively search the pool 
during the probe trial, nor do they directly approach the platform location from 
the initial starting point. It is important to note that the loss of spatial navigation 
is independent of the disruption in motor, perceptual, or motivational 
components of the task (Morris et al., 1982; Morris et al., 1986). 
Similarly, there are numerous reports of age-related deficits in rats on the 
Morris water task. Age-related deficits in performance are well documented in 
the standard fixed location, hidden platform version (Gage et al, 1984,1988; 
Gallagher and Pelleymounter, 1988; Gallagher et al., 1993; Hebda-Bauer et al., 
1999; Lindner and Gribkoff, 1991; Menaey et al., 1991; Pelleymounter et al., 
1987; Rapp et al., 1987), and in the working memory version (Markowska, 
1999; Van der Staay and de Jonge, 1993; Wyss et al., 2000). Spatial working 
memory in the rat can be assessed, for example, by testing the acquisition of 
the hidden platform location from trial to trial so that the location of the platform 
is varied on each subsequent trial in the session (Frick et al., 1995), or by 
allowing two 1-minute training sessions and then testing 30 seconds after the 
second training session on a no-platform probe trial (Lee et al., 2000). Overall, 
aged rats are less efficient at acquiring the task compared to younger adult rats, 
and spatial learning and memory deficits are reported in the absence of 
significant sensory, motor, or motivational problems (Gallagher and 
Pelleymounter, 1988; Rapp etal., 1987). 
17 
This age-related spatial deficit in the Morris water task is not unique, given that 
aged rats and rats with hippocampal damage also exhibit deficits in performance 
on other spatial tasks, such as the radial arm maze (Barnes, 1979; Barnes et 
al., 1989; Gallagher et al., 1985; Jarrard, 1983; Olton and Samuleson, 1976), 
the T-maze (Bannerman et al., 2001; Rawlins and Olton, 1982), and the Barnes 
circular maze (Barnes, 1979; Barnes, 1988; Barnes and McNaughton, 1985; 
Barnes et al., 1989; Barnes et al., 1990). In addition, it has been demonstrated 
that aged animals exhibiting behavioral hippocampus-dependent deficits also 
exhibit different hippocampal physiological and morphological profiles compared 
to both young and old unimpaired animals (for a review see Geinisman et al., 
1995), certain aspects of which will also be considered shortly. 
3.2.2 Transverse Patterning Discriminations 
In addition to deficits in spatial tasks, rats with hippocampal damage have 
deficits in certain forms of non-spatial learning. In particular, they have 
difficulty solving configural/conjunctive discriminations, such as the transverse 
patterning problem, but no trouble solving similar elemental (simple) 
discriminations (Alvarado and Rudy, 1995a; Dusek and Eichenbaum, 1998). 
Elemental problems involve associating a discrete cue (e.g., A) with a response 
(X; e.g., A-X). Unlike elemental discriminations, simply responding to or 
avoiding a single cue cannot solve the transverse patterning problem, because 
each choice stimulus is ambiguous in predicting a correct response. Successful 
18 
performance on the transverse patterning problem involves associating a 
compound cue with a correct response (e.g., AB-X). A compound cue is formed 
when multiple stimuli (e.g., A and B) are combined to form a single 
representation (e.g., AB), The only way to successfully solve this problem is to 
utilize a configural solution, which in turn requires one to respond on the basis 
of the relationship between the cues. 
It has been demonstrated that rats with hippocampal damage are unable to 
learn transverse patterning discriminations, while elemental discrimination 
learning is spared (Alvarado and Rudy, 1995a). Also, Rondi-Reig et al. (2001) 
tested mice lacking CA1 NMDA receptors on the transverse patterning problem 
that required simultaneous acquisition of three odor discriminations comprising 
the transverse patterning problem contrasted with concurrent learning of three 
simple odor discriminations. They found that CA1 NMDA receptor knock-out mice 
were impaired in solving the transverse patterning odor discrimination compared 
with normal performance on the non-overlapping (elemental) concurrent odor 
discriminations (but see Bussey et al., 1998). To the best of our knowledge, 
prior to the present work, the transverse patterning task has not been 
previously employed in the studies of aged animals. 
19 
3.3 Hippocampus-dependent Learning and Memory in Humans 
The important role that has emerged for the hippocampus in spatial learning and 
memory in non-human animals has prompted examination of the human 
hippocampus involving similar tasks. Thus far, the evidence supports the notion 
that the hippocampus plays an important role in spatial learning and memory 
across many species. For example, damage to the hippocampal formation in 
species as widely divergent as pigeons and humans results in disrupted spatial 
learning. Also, recent neuroimaging investigations have provided evidence 
showing that the human hippocampus becomes activated during spatial 
navigation. Such studies will be discussed in detail below. 
3.3.1 Virtual Navigation 
A variety of tasks have been used to assess human spatial cognition, some likely 
engaging similar processes as in non-human animal experiments and others 
engaging very different processes. For example, mental rotation, spatial 
orientation, route learning, map learning, pointing to places, throwing to 
targets, intercepting projectiles, and measuring water levels have all been 
employed to measure human spatial ability. A key point in the human studies of 
spatial ability is that they often reveal a superior performance by males. For 
many, the mental rotation task originally described by Metzler & Shepard (1974) 
is the gold standard for measuring spatial cognition in humans. In important 
20 
ways, data from studies that have employed mental rotations do not lend 
themselves to a direct comparison with data on performance of a rat in the 
Morris water task. Even though both tasks are thought to be spatial in nature, 
evidence suggests that they involve a distinctly different circuitry (posterior 
parietal vs. hippocampal) and to date mental rotation performance has not been 
tested in the rat. A human analog of the Morris water task for measuring 
hippocampal processes has been developed, however, using virtual navigation 
through computer-generated environments. Many virtual navigation tasks, such 
as a virtual arena (Jacobs et al., 1998), a virtual reality town (Maguire et al., 
1998), a virtual maze (Moffat et al., 1998), a virtual Morris water task (Astur et 
al., 1998a; Astur et al., 2002; Driscoll et al., 2003a, 2005; Hamilton and 
Sutherland, 1999; Hamilton et al., 2002; Hamilton et al., 2003; Moffat and 
Resnick, 2002), and a Hebb-Williams maze (Shore et al., 2001) have been 
developed to facilitate the testing of human navigational processes in a 
laboratory setting. 
Hippocampal activation has been observed using PET in participants learning to 
navigate in virtual environments (Maguire et al., 1996, 1997, 1998). For 
example, Maguire et al. (1998) designed a clever experiment in which London 
taxi drivers were examined on their knowledge of complex routes versus 
landmarks around London. Differences in activation were seen in the parietal 
lobe, the cingulate cortex, the parahippocampal gyrus, and the right 
hippocampal formation. Further, functional neuroimaging of human brain activity 
during navigation in a virtual town revealed that navigational accuracy varied 
21 
directly with the degree of activation of the right hippocampus, while the left 
hippocampal activation was related to non-spatial aspects of navigation (Maguire 
et al., 1998). Damage to the right temporal lobe (including the hippocampus) 
due to thermo-coagulation along the amygdalo-hyppocampal axis in an attempt 
to alleviate epilepsy has also been shown to affect performance on procedures 
that emulate rodent spatial learning (Bohbot et al., 1998). The above-mentioned 
experiments comprise a growing body of evidence suggesting that the human 
hippocampus is involved in spatial information processing. Taken together, 
human brain lesion and neuroimaging studies correspond well with rat lesion 
studies in demonstrating that hippocampal damage compromises spatial 
navigation in significant ways. Other involved regions (such as right caudate 
nucleus, right inferior and bilateral medial parietal regions, and left frontal lobe) 
have been implicated, outlining a distributed network supporting human 
navigation, and linking the functions of the network regions to physiological 
observations already reported in other mammals. 
The computerized version of the Morris water task developed in our laboratory 
(Astur et al., 1998a; Hamilton and Sutherland, 1999) is part of an effort to 
translate non-human work into a virtual domain, in order to develop human 
models of rat memory processes. We have evaluated whether many 
psychological functions engaged by this task in rats are similarly engaged in 
humans (Hamilton et al., 2002). For example, we investigated human place 
learning in the virtual Morris water task under comparable conditions to those 
employed by Sutherland et al. (1987) with rats. More specifically, Sutherland et 
22 
al. (1987) tested seven different groups of rats in the Morris water task by 
systematically varying the opportunity to view and/or navigate through the 
environment. All groups in the Sutherland et al. (1987) experiment were 
released in one half of the pool, while physical and/or visual access to the other 
half of the pool where the hidden platform was located were restricted. When 
subsequently tested without restrictions, only rats that were able to view and 
navigate through the entire environment without restrictions performed well. 
Years later, the same conditions were imposed in a virtual environment on 
seven different groups of undergraduate students (Hamilton et al., 2002). Like 
the rats in the Sutherland et al. (1987) experiment, the human participants 
were trained to navigate to a hidden platform under the same seven conditions, 
but in the computerized task. Our results show that human performance in the 
virtual Morris water task closely matches the data obtained by Sutherland et al. 
(1987) with rats in all of the seven different conditions that were imposed on the 
subjects. Furthermore, our data indicate that there is some generality in the 
principles involved in place learning and provide a reasonable basis for cross-
species comparisons (Hamilton et al., 2002). 
In addition, we propose that the virtual Morris water task not only allows greater 
control over space than would be possible in a real-world navigation task, but 
that it, to an extent, taps similar processes to those engaged by the task in the 
rat. Thus, it enables one to draw direct connections to an existing body of 
animal research in this area, a literature that includes levels of description of 
23 
spatial memory processes involving specific networks, single neuron physiology, 
and molecular signaling. One may imagine making intimate contact between 
human cognitive research and the vast amount of animal literature on 
hippocampal function that is usually overlooked when designing human studies. 
Even though some might argue that a meaningful comparison would be 
inappropriate due to the inherent limitations of virtual environments, such as the 
lack of vestibular and proprioceptive cues, the work conducted in our (Hamilton 
et al., 2002) and other laboratories (Shore et al., 2001) provides a reasonable 
basis for believing that valid cross-species comparisons can be made using such 
tasks. So far, the virtual Morris water task has proven useful in studying sex 
differences in place learning and memory (Astur et al., 1998a, 2004; Driscoll et 
al., 2005), rules governing place learning (Hamilton and Sutherland 1999; 
Hamilton et al., 2002), deficits after hippocampal damage (Astur et al., 2002), 
aging (Driscoll et al., 2003a, 2005; Laurance et al., 2002; Moffat and Resnick, 
2002), and fetal alcohol syndrome (FAS; Hamilton et al., 2003). Together, these 
studies demonstrate that successful spatial navigation in humans, as assessed 
by virtual navigation, is also highly dependent on the integrity of the 
hippocampal formation. 
In a relatively direct test of this idea, Astur et al. (2002) tested ten patients with 
unilateral resection of the hippocampus on a virtual version of the Morris water 
task. Five of the patients had the left and five had the right hippocampus 
removed. They found that patients with hippocampal resections display severe 
24 
deficits in spatial navigation compared to the age-matched controls and age-
matched patients with extra-hippocampal resections, suggesting that the 
hippocampus is critical for spatial learning and memory regardless of which 
hemisphere was damaged. In addition, a preliminary evaluation of performance 
in a VMWT with functional magnetic resonance imaging (fMRI) by Astur et al. 
(2002) revealed a bilateral activation centered in the hippocampus, consistent 
with other functional imaging reports during virtual navigation (Aguirre et al., 
1996; Maguire et al., 1998). 
Similarly to aged rats, aged humans are thought to have spatial information 
processing deficits (Kirasic and Allen, 1985; Perlmutter et al., 1981). Recent 
evidence indicates that elderly people are impaired in spatial navigation in 
virtual environments (Driscoll et al., 2003a, 2003b, 2005; Laurance et al., 2002; 
Moffat and Resnick, 2002; Moffat et al., 2001). Taken together, the human 
imaging studies are consistent with the idea that the hippocampal formation 
contributes to the encoding and retrieval of topographical information in 
humans. Furthermore, human imaging and lesion studies compare well with 
non-human animal studies that show robust and persisting deficits in spatial 
ability following hippocampal damage. 
25 
3.3.2 Transverse Patterning Discriminations 
The transverse patterning discrimination problem is an important foundation for 
evidence supporting the configural/conjunctive association view of hippocampal 
function. Although there are reports that adult humans are unable to solve the 
transverse patterning (Berch and Israel, 1971), contrasting evidence offered by 
Astur and Sutherland (1998b) indicates that college-aged individuals have little 
difficulty solving this problem, especially if a stepwise approach is used as 
previously described by Alvarado and Rudy (1995a) with rats. Further, Astur 
and Sutherland (1998b) suggest that participants seem to adopt a configural 
strategy when presented with ambiguous stimulus pairs, even though an 
elemental strategy would suffice. Also, Rudy et al. (1993) reported that children 
at least 4.5 years old were able to solve transverse patterning and conditional 
discriminations, two problems that require configural association solutions (i.e., 
are hippocampus-dependent). In addition, they reported that younger children 
did not solve these problems, but were able to solve problems that permitted an 
elemental solution even though the problems were constructed from the same 
stimulus materials. They suggest that children may not gain access to configural 
solutions until they are about 4.5 years old and that the configural association 
system depends on different, later developing neural structures than does the 
elemental association system. 
A more recent study (Astur and Constable, 2004) attempted to assess the role 
of the hippocampal formation on transverse patterning discriminations using 
26 
functional magnetic resonance imaging. Astur and Constable (2004) report that 
activation maps and time course data both suggest that the hippocampus is 
involved in this task, but interestingly, the hippocampus was less active during 
the hippocampus-dependent period of the testing phase compared to both the 
hippocampus-independent test phase and the fixation control phase. This should 
alert us that the traditional assumptions advocating that the increase in blood-
oxygenation level dependent response is necessary in order for a brain region to 
be considered involved in a task might be inaccurate and that alternative 
explanations, such as the hippocampal 'dampening' in this case, should be 
considered. Impairment in solving the transverse-patterning problem has also 
been reported in amnesic patients with medial temporal lobe damage (Reed and 
Squire, 1999; Rickard and Graffman, 1998). 
27 
4. The Aging Hippocampus 
The hippocampal formation is often implicated in the age-related cognitive 
decline. There is agreement that many behavioral deficits exhibited during 
normal aging resemble the ones seen following bilateral disruption of the 
hippocampus (Geinisman et al., 1995). Important features of the age-related 
cognitive impairment include that, once acquired, the information is lost rapidly, 
and also that newly learned information is much more vulnerable compared to 
the information learned in the remote past. While qualitatively similar, the age-
related deficits are, however, believed to be quantitatively less severe. 
An advantage of the aging rodent studies is in the ability to separate and strictly 
control the contribution of genetic and environmental factors. As one may 
imagine, this is much harder to do with human participants. Nonetheless, many 
factors have been proposed as a neural substrate for age-related cognitive 
decline. Here we consider only a subset of them, namely physiological, 
structural, and biochemical changes. 
4.1 Physiological Changes 
One possible physiological source of age-related impairments in hippocampus-
dependent learning and memory that has been extensively investigated is the 
mechanism of long-term potentiation (LTP), also sometimes referred to as long 
28 
term enhancement (LTE). LTP is a relatively enduring form of synaptic plasticity 
that is well represented in the hippocampus. It is presently the best candidate 
for the neural mechanism that underlies information storage in the mammalian 
brain. It is presumed that the "activity-dependent synaptic plasticity is induced 
at appropriate synapses during memory formation and is both necessary and 
sufficient for the information storage underlying the type of memory mediated 
by the brain area in which the plasticity is observed" (Martin et al., 2000). 
Therefore, the change in the strength of connections between the neurons is 
assumed to be the mechanism by which memory traces are encoded and stored 
in the brain. So far, there is evidence that older rats show impairments in LTP 
induction in the Schaffer-collateral-CAl synapse when sub-optimal pulse 
parameters are used. It has also been demonstrated that a larger amplitude 
current is necessary for LTP to be induced at the perforant path granule cell 
synapses (Barnes et al., 2000). Together, these observations point to a change 
in the induction threshold at the synapse, and that LTP should be harder to 
induce in older rats. Age-related changes in LTP are often attributed to the 
reduction in the number of perforant path synaptic contacts on granule cells. A 
number of factors have been suggested to account for differences in synaptic 
plasticity in aged rats, such as possible age-related differences in protein 
synthesis, axonal transport, calcium regulation, and hippocampal synaptic 
interconnectivity. 
29 
It is also possible that different regions within the hippocampus differentially 
undergo physiologically detectable changes with aging (Rosenzweig and Barnes, 
2003). Overall, impairments indicative of alterations in neural plasticity are 
observed in a number of aged animals, specifically those that exhibit poor 
memory function. It is easy to see how such changes in network connectivity 
could be responsible for changes in plasticity with advanced age. Many reports 
by Barnes and colleagues (Barnes, 2003; Rosenzweig and Barnes, 2003) 
suggest, however, that old rats can display intact LTP induction but that the 
enhancement decays significantly faster. Disparate results like these in the 
aging and LTP literature are often attributed to critical methodological 
differences, such as the intensity of stimulation or whether induction or decay is 
measured. Sometimes even the strain of rat has been considered as a possible 
source of differences. 
The present overview merely scratches the surface when it comes to the topic of 
hippocampal synaptic plasticity and aging, in part because this has been the 
subject of excellent recent reviews (for example see Barnes, 2003; Geinisman et 
al., 1995; Rosenzweig and Barnes, 2003). Rosenzweig and Barnes (2003) not 
only review the work on synaptic plasticity and network dynamics in relation to 
aging but also the available evidence for the role of LTP in learning and memory. 
30 
4.2 Structural Changes 
Age related structural changes that are likely to be relevant to hippocampus-
dependent cognition have been identified. Regional differences in the aging brain 
are thought to be rather subtle in comparison to age-related global brain 
deterioration. Some evidence suggests that the hippocampal formation may be 
particularly vulnerable to the effects of aging, especially age-related pathology. 
This evidence includes reports of decrease in neuronal count, intracellular 
pathology, affinity of neurofibrillary tangles for this region (Raz, 2000), and a 
decrease in volume (Golomb et al., 1993; Jernigan et al., 1991). Age-related 
reductions in hippocampal volume have been observed repeatedly (Golomb et 
al., 1993; Jernigan et al., 1991), but there are also important exceptions that 
should be noted (Raz, 2000; Sullivan et al., 1995). Despite the equivocal 
findings at times, it is our position that it is doubtful that the hippocampal 
formation would completely escape the deleterious effects of old age, especially 
when one considers its involvement in age-related pathology. 
In vivo neuroimaging reveals only mild age-related shrinkage of the broadly 
defined hippocampal regions. Traditionally, volume loss was considered to be a 
direct result of neuronal death. Evidence indicates that observed changes might 
be due to the shrinking of neurons rather than their death (Esiri, 1994). 
Cytoarchitectonically, the hippocampus is a complex structure and one might 
expect that different subdivisions of the formation may be differentially affected 
31 
by age. Evidence from some in vivo studies of the hippocampus suggests a 
possible "anterio-posterior gradient of age-related vulnerability" (Raz, 2000). 
Petersen et al. (1998) demonstrated that the anterior hippocampus, unlike the 
body and the posterior part, is less likely to atrophy in early AD patients and 
non-AD controls with age-related memory problems. Further, hippocampal 
volumetric studies in which anterior hippocampus was not measured were more 
likely to show an age-related volume decline than the studies that included less 
vulnerable anterior slices (Raz, 2000). In turn, the functional distinction along 
the anterio-posterior axis of the hippocampus is of particular interest. 
In comparing highly specialized spatial navigation in London taxi drivers with 
that of control subjects, Maguire et al. (2000) found a positive correlation in the 
volume of posterior hippocampus and negative correlation in the volume of 
anterior hippocampus with the amount of time spent as a taxi drive. The same 
associations were found with their performance on a virtual navigation task. The 
results are in agreement with the idea that the posterior hippocampus is 
responsible for processing related to efficient usage or storage of spatial 
information. Similarly, there is some evidence that the posterior hippocampus 
in monkeys (Colombo et al., 1998) and the dorsal hippocampus in rodents 
(homologous with the posterior hippocampus in monkeys and humans) (Moser 
and Moser, 1998; Moser et al., 1993) are more important than the anterior (or 
ventral) regions for encoding spatial information. 
32 
MRI-volumetric measurements are being used clinically with increased 
frequency. It has been reported that the hippocampal volume accurately reflects 
the structural-functional relationships between memory deficits and hippocampal 
damage from normal aging to dementia (Petersen et al., 2000). The appeal of 
this technique is that, clinically, a relevant brain structure can be measured in 
vivo, and that this information may aid in diagnosis of pathology early on. 
Structural MR imaging is seldom performed on non-human animals. In non-
human animals, morphological analyses are customarily performed on post 
mortem tissue. Changes in the number of hippocampal neurons and synapses 
have been extensively investigated over the years in hope of finding structural 
substrates of age-related memory decline. Many, but not all, have reported a 
loss in the hippocampus in the number of neurons and/or synapses that is 
associated with old age (for a review see Gallagher and Rapp, 1997). 
Additionally, this structural loss has been correlated with the severity of memory 
impairments in the same animals. Few of such studies have, however, employed 
unbiased counting techniques. When such techniques were employed, age-
related alterations in neuron and synapse number were only convincingly 
demonstrated in humans, and not in the non-human animals with age-related 
memory deficits. 
We propose that MRI provides a valuable technique in bridging the gap between 
humans and non-human animals in that structural integrity can be measured in 
non-human animals in vivo employing the same techniques that are used in 
33 
human research. In addition, the non-human MR results can be easily 
complemented with cellular and molecular post-mortem measurements of the 
same tissue. Even though structural research on non-human animals has been 
extensive, the progress in technique development makes this research area a 
fertile ground for further exploration. 
4.3 Biochemical Changes 
The precise nature of the affected processes that lead to the age-related decline 
in memory is still uncertain, even though definitive age-related physiological and 
anatomical changes have been reported. A complimentary technique to MRI is 
magnetic resonance spectroscopy (MRS). MRS uses tools similar to the more 
familiar magnetic resonance imaging (MRI) to measure brain function in vivo. In 
the past ten years MRS has carved out a niche in the evaluation of metabolic 
processes in the living organism. MRS allows the measurement of the 
concentrations and synthesis rates of individual chemical compounds within a 
precisely defined brain region. Most common measurements are based upon the 
1H nucleus, and most often studied measurable metabolites include N-
acetylaspartate (NAA), creatine (Cre), choline (Cho), y-amino butyric acid 
(GABA), glucose, glutamine, glutamate, and lactate. These metabolites play 
critical roles in neuroenergetics, neurotransmission, and neuromodulation 
(Nicholls, 1989; Sokoloff, 1999). Often metabolic alterations can be detected 
before the more obvious structural alterations become apparent, both in 
34 
response to treatment and as a part of natural progression of the disease. By 
employing MRS, it is possible to measure levels of different brain metabolites in 
vivo in a particular region of interest. The application of proton magnetic 
resonance spectroscopy (XH-MRS) to the investigation of the brain most often 
involves measuring N-acetylaspartate (NAA), choline (Cho), and creatine (Cre) 
containing compounds. 
A decrease in NAA concentrations was classically believed to reflect atrophy, 
neuronal loss, and/or decreased neuronal density. Recent studies, however, 
suggest that lower levels of NAA/Cho do not seem to be necessarily related to 
neuronal loss, but may also be associated with metabolic dysfunction (Seilhean 
et al., 1993). For example, this appears to be the case in dementia complex in 
HIV patients with documented neuronal injury (Lipton, 1992). Consequently, it 
has been suggested that decrease in NAA values may also reflect the magnitude 
of neuronal injury and metabolic dysfunction. Decrease in NAA in the 
hippocampus has been reported in several conditions, including PTSD (Schuff et 
al., 1997a), epilepsy (Ende et al., 1997), Alzheimer's disease (Schuff et al., 
1997b) and normal aging (Driscoll et al., 2003a, 2005; Schuff et al., 1999). 
NAA is a putative neuronal marker since it is almost exclusively found in neurons 
(Moffet et al., 1991) and is virtually undetectable in other cell types (Urenjak et 
al., 1993). Nonetheless, the role of NAA in neuronal function and hence the 
exact significance of MRS measurements of NAA are still uncertain despite the 
fact that NAA is one of the most abundant amino acids in the brain. NAA has 
35 
been suggested to be a source of acetyl groups for lipid synthesis, a regulator of 
protein synthesis, a storage form of acetyl-CoA or aspartate, a breakdown 
product of N-acetyl-aspartyl-glutamate, or an osmolite (Birken and Olendorf, 
1989). Animal models of chronic neuronal injury provide reliable correlations 
between NAA levels and in vitro measures of neuronal survival, tempting one to 
equate NAA with neuronal density or to label it as neuronal marker. On the other 
hand, there are reports of reversible NAA deficiency due to either spontaneous 
or treatment-related recovery, suggesting that the initial drop could not be due 
to the loss of axons or neurons. 
Reversible decrease in NAA has been reported in various conditions involving 
brain injury (Davie et al., 1994; De Stefano et al., 1995; Hugg et al., 1996; 
Pavlakis et al., 1998), making the role of NAA as a specific marker uncertain. 
Also, Demougeot et al. (2001) reported that cellular dysfunction can cause 
greater reductions in NAA levels than the neuronal loss itself. Furthermore, their 
findings suggest that dysfunction in mitochondrial energy metabolism may be 
sufficient for a severe drop in NAA. Such observations suggest NAA to also be a 
marker of cellular dysfunction in certain cases. Moreover, elevated NAA levels 
are found in Canavan's disease, and clearly do not reflect an increase in 
neuronal density here, but rather the deficiency in an enzyme (aspartoacylase) 
responsible for NAA degradation. Taken together, the data indicate that NAA 
levels probably underestimate neuronal density. Clearly, NAA quantification does 
not provide a faithful estimate of neuronal density and should not replace 
histological examinations. NAA remains useful in assessing early cellular 
36 
dysfunction before histological lesions appear. Still, NAA does not appear to be a 
sensitive marker of neuronal density and other complementary assessments 
should be used before coming to such conclusions, and reversible cellular 
dysfunction must be considered as a plausible explanation. 
Choline (Cho) resonance originates from glycerophosphocholine, 
phosphocholine, and choline itself, which are all integral components of 
membrane phospholipids. Increased Cho levels are taken as a sign of membrane 
breakdown, inflammation or demyelination as seen in stroke (Olson et al., 1992) 
and multiple sclerosis (Davie et al., 1994). Creatine (Cre), whose resonance 
represents the sum of creatine and phosphocreatine, provides a measure of 
cellular energy currency. Choline and creatine are found in all brain cells, 
making the changes somewhat difficult to interpret and understand. 
Many MRS studies in humans report general age-related alterations of the 
aforementioned metabolites (Chang et al., 1996; Fukuzako et al., 1997), 
whereas others are more specific to hippocampal metabolite changes associated 
with volume loss (Schuff et al., 1999). Studies in different patient cohorts 
(Brooks et al., 1997, 1999; Jung et al., 1998) and normal populations (Jung et 
al., 1999) using quantitative spectroscopy and an extensive neurocognitive 
battery reveal a significant relationship between neurochemical markers of brain 
integrity and brain function. For example, NAA was related to the overall 
performance on neuropsychological battery of tests, especially the timed ones, 
in normal young adults (Jung et al., 1999) and in patients with systemic lupus 
37 
erythematosus (Brooks et al., 1999). Despite the limitations, not just in the 
method itself but in the interpretation of results, we believe *H MRS/MRSI to be 
a useful tool in assessing hippocampal neuronal integrity with aging. 
Furthermore, proton MR spectroscopy may provide a clue in determining a 
relationship between age-associated brain atrophy and brain function. 
An increasing number of techniques are being developed to non-invasively 
analyze structural, physiological, and chemical changes of the brain in vivo. MRS 
has a unique value in the realm of different imaging techniques - it provides 
information on a quantitative spectrum of multiple compounds simultaneously in 
the region of interest. The application of MRS in investigations of various disease 
progressions is increasing, and so is the knowledge on the unique biochemical 
profiles of different neurodegenerative disorders. 
Overall, non-invasive MRS measurements of NAA appear to be one of the best 
surrogate markers currently available for neuronal integrity. A deeper 
understanding of NAA biochemistry should help elucidate its function in normal 
and pathological states. Until then, 1H MRS spectrum of NAA and other 
metabolites have proven at times to be effective indicators of neuronal/axonal 
viability and are finding an increasing application in research and in clinical 
neurology. 
38 
4.4 Neurogenesis 
In discussing the hippocampus we find it difficult to ignore the topic of adult 
neurogenesis. We will, however, only scratch the surface on this topic here. 
Adult neurogenesis seems to be restricted to two regions in the brain, one of 
which is the dentate gyrus of the hippocampus. Demonstration of neurogenesis 
in adult brain represents a major advance in our understanding of the cellular 
mechanisms underlying neuronal circuit plasticity and complex behavior. 
The traditional view of the mammalian brain organization is that the brain 
becomes structurally stable shortly after birth, and no new neurons are added 
throughout adulthood. More recently this view has been supplemented by an 
appreciation of the importance of regressive events in sculpting mature circuitry 
postnatally, including neuron and synapse loss. Versions of this view have for 
centuries guided research on neurodegenerative diseases, brain injury, and 
recovery of function. A challenge has arisen in the last decade with the 
development of new methods that yielded evidence of adult neurogenesis 
(Cameron and McKay, 2001; Kuhn et al., 1996). The studies corroborated the 
early reports (Altman and Das, 1965; Kaplan and Hinds, 1977), which were 
largely ignored and had little impact on the field. There are at least two brain 
regions in which neurogenesis in adult animals has been demonstrated 
unequivocally: a forebrain subventricular zone (SVZ) and the subgranular zone 
(SGZ) in the dentate gyrus of the hippocampus. 
39 
Currently, there is agreement that granule cells in the SGZ continue to 
proliferate, migrate, and fully differentiate throughout adulthood in the 
hippocampus of every species investigated so far, including birds (Barnea and 
Nottebohm, 1994), mice (Kempermann et al., 1997), rats (Bayer, 1982; 
Cameron et al., 1993, Derrick et al., 2000), tree shrews (Gould et al., 1997,), 
monkeys (Gould et al., 1998, 1999b; Kornack and Rakic, 1999), and humans 
(Eriksson et al., 1998). Although the exact function of adult neurogenesis is 
unknown, several studies suggest that it is likely to play a role in the normal 
information processing in the circuits to which neurons are added and as such 
would play an important part in hippocampal plasticity. Some have also 
suggested that newly generated neurons in the dentate gyrus may play a role in 
certain aspects of learning and memory. 
The inability to generate replacement neurons has long been considered to be 
an important cause of neurological disease, age-related symptoms, and 
behavioral impairments. This raises the possibility that naturally occurring 
neurogenesis may be useful for slowing the progression of functional loss in the 
aging brain. The production of new neurons is only one of several steps, 
however, believed to be necessary to restore damaged circuits. More 
importantly, once born and differentiated, these neurons need to survive and 
become a functionally integrated part of their neuronal environment. Studies of 
normal animals have identified several factors that can affect the production and 
survival of new neurons in the adult brain. 
40 
Regulation of the rate of neurogenesis is controlled by a variety of behaviorally 
derived factors and endogenous regulators. It has been suggested that growth 
factors through their ability to promote neurogenesis may have a role in the 
treatment of age-related neurodegenerative disorders by stimulating 
proliferation of neuronal precursors in the adult brain. Although growth factors 
are numerous, whether and how the aged brain remains responsive to these 
factors is not well known. For example, basic fibroblast growth factor (bFGF or 
FGF-2) is one of the primary mitogens used for neural stem cell propagation in 
vitro and is hypothesized to be important in controlling in vivo cell proliferation. 
Indeed, repeated application of FGF-2 promotes generation and maintenance of 
neurons from the adult human subventricular zone (Goldman, 1998). Recently it 
has been reported that FGF-2 delivery into the rodent brain increased 
proliferation of progenitor cells in the dentate gyrus and stem cells in the 
subventricular zone (Jin et al., 2003). 
Another growth factor, erythropoietin (an inducible cytokine), appears to act 
directly on neural stem cells by promoting the production of neuronal 
progenitors at the expense of multipotent progenitors, and is found to regulate 
neurogenesis in vitro and in vivo (Shingo et al., 2001). In fact, erythropoietin 
induced a two- to three-fold increase in the rate of neuron production. It is 
beyond our scope to exhaustively review all of the neurotransmitters, hormones, 
modulators, and neurotrophins that affect adult neurogenesis. Instead, we 
briefly discuss below only a couple that are directly relevant to aging. 
41 
There are also numerous reports showing that living in an enriched environment 
enhances the survival of newly generated cells in the hippocampus (Barnea and 
Nottebohm, 1994; Kempermann et al., 1997, 1998; Nilsson et al., 1999), 
providing a neural substrate for "use it or lose it" hypothesis. Increased rate of 
neurogenesis accompanied by better learning, increase in exploratory behavior, 
and in locomotor activity have all been reported in aged mice living in an 
enriched environment (Kempermann et al., 2002). Similarly, environmental 
enrichment enhanced the survival of newly born neurons in the adult mouse 
dentate gyrus (van Praag et al., 1999). Together these studies suggest that the 
aging hippocampus retains a potential for cellular plasticity, and that experience 
or activity can influence the functional status of circuitry in this region possibly 
via an effect on neurogenesis. 
The exact function of hippocampal neurogenesis in the normal brain is still not 
known. Nevertheless, it has been suggested that the adult neurogenesis in the 
hippocampus promotes learning and memory. Correlational evidence is 
accumulating that supports a role for neurogenesis in compensating for dying 
neurons, thereby protecting the functional integrity of hippocampal circuitry 
(Snyder et al., 2001). Aging affects hippocampus-dependent cognitive ability, 
and age-related spatial memory impairments have been related to alterations in 
hippocampal plasticity (Barnes, 2003). These cognitive deficits can be 
ameliorated by the infusion of either nerve growth factor or insulin-like growth 
factor. Given that both of these treatments have been shown to increase 
42 
neurogenesis in the adult dentate gyrus (Poe et al., 2001), this may be a basis 
for improved memory performance. 
Recent data suggest that the failure in regulation of hippocampal neurogenesis 
might result in diminished capabilities of the hippocampus to adjust to a novel 
environment and to form new memories. For example, Kempermann et al. 
(2002) tested ten strains of recombinant mice, some known to be good learners 
and have high baseline levels of neurogenesis (C57BL/6) and some which are 
known to be poor learners and have a low baseline levels of neurogenesis 
(DBA/2), in order to examine whether the baseline levels of neurogenesis 
predict performance on the Morris water task. There was a significant correlation 
between the number of newly generated neurons in the dentate gyrus and the 
acquisition measures on the Morris water task for all the strains, suggesting 
specific neurogenic involvement in the function of the hippocampus - the 
acquisition of new information in particular. Similarly, it has been demonstrated 
that spatial memory performance of aged rats predicts levels of hippocampal 
neurogenesis (Drapeau et al., 2003). 
In addition, it has been suggested that some age-related disorders, such 
Alzheimer's disease (Haughey et al., 2002; Jin et al., 2004; Tatebayashi et al., 
2003; Wang et al., 2004), and Huntington's disease (Curtis et al., 2003; 
Tattersfield et al., 2004), are related to alterations in neurogenesis. Even though 
hippocampal neurogenesis declines with age, it remains regulated in response to 
external challenges. Some growth factors can be administered systemically to 
43 
stimulate neurogenesis (Aberg et al., 2000; Wagner et al., 1999), suggesting a 
relatively non-invasive means for exploring this approach to treatment. This 
possibility that hippocampus can be stimulated to produce new cells with the 
ability to replace dysfunctional or lost neurons could have important therapeutic 
implications of great significance to the aging population. Before neurogenesis 
can be harnessed to ameliorate the age-related deficits, however, its role in the 
normal brain must be understood. 
Currently there is no direct evidence suggesting that new neurons can aid in 
retarding the age-related functional decline, nor have there been direct prior 
tests. Furthermore, the enhancement of generation of new neurons has not 
been causally linked to recovery of lost function. For neurogenesis to be 
advantageous, the new neurons have to be able to not only differentiate and 
survive, but also to appropriately integrate into the existing neural circuitry. 
Once integrated, the new neurons should display functional properties similar to 
those of the neurons they had replaced, and should be generated in numbers 
that are comparable to the number of neurons that were lost. 
So, are these newly generated cells functional? There is mounting evidence that 
suggests that the newly generated cells are indeed functional. For example, we 
know that adult neurogenesis produces cells that are incorporated into the 
correct location, express granule cell-specific antigens, have normal dendritic 
arbors, and extend mossy fiber type axons into CA3 field - all characteristics of 
mature granule cells (Kempermann et al., 2002; Stanfield and Trice, 1988). In 
addition, van Praag and colleagues (2002) demonstrated in hippocampal slices 
44 
that weeks after birth new granule cells expressed the electrophysiological 
properties of mature granule cells. There is also a report of entorhinal cortex 
neurons infected with transynaptically transported pseudorabies virus that later 
localized to new granule cells and CA3 cells (Carlen et al., 2002), supporting 
functional integration into the hippocampal network. Similarly, taking advantage 
of the activity-dependent expression of c-fos, over 80% of new granule cells 
show c-fos activation by training in the Morris water task (Jessberger and 
Kempermann, 2003). 
Collectively, there are good reasons to lead us to believe that new granule cells 
participate in information processing within the hippocampal network. While 
none of the functional properties of mature granule cells have been 
demonstrated in the new ones, we do have a solid foundation for pursuing 
functional experiments related to restoration of damaged hippocampal circuits. 
For a more in-depth discussion regarding hippocampal neurogenesis we refer 
the reader to a couple of recent reviews (Goldman, 1998; Lie et al., 2004). 
45 
5. Human Model of Rat Aging 
The results of a great deal of research have shown alterations in the 
hippocampus related to the process of aging in both humans and non-human 
animals. On the other hand, there are important shortcomings associated with 
this field of research. One such shortcoming is that the age-related alterations 
are often investigated in the absence of information from the participants about 
their level of behavioral or cognitive functioning and it is not clear how they 
relate to learning and memory dysfunction. Also, the tests used to assess 
learning and memory in humans typically are very different from those 
employed to test non-human animals, making it hard to generalize or make 
comparisons. For example, human tests are almost always verbal. Furthermore, 
methodologies employed between the species are often quite different and do 
not lend themselves to direct cross-species comparisons. For example, non-
human animal studies are commonly performed on post mortem tissue, while 
studies on humans are almost always performed in vivo and often involve 
imaging. All in all, we are left with numerous reports on human aging, and just 
as many if not more reports on non-human animal aging, but despite the 
abundance of information there are clearly missed opportunities for the two 
avenues of research to inform one another. 
In order to address some of these shortcomings and to further investigate age-
related changes in hippocampus-dependent cognition we set out to develop a 
human model of animal aging. In addition, we searched for converging evidence 
46 
of changes that may underlie age-related deficits in hippocampus-dependent 
learning and memory. To that end, we conducted a study (Driscoll et al., 2003a) 
involving young and elderly human participants that employed the virtual Morris 
water task and the transverse patterning discrimination task, a spatial and non-
spatial hippocampus-dependent task respectively. In addition, magnetic 
resonance imaging (MRI) and magnetic resonance spectroscopic imaging (MRSI) 
were used on the same participants to measure structural and neurochemical 
integrity of the hippocampus respectively. 
Specifically, 16 young (M (age) = 26.1) and 16 elderly (M (age) = 77.6) 
volunteers participated in the experiment, 8 males and 8 females in each group. 
Elderly participants were recruited from the New Mexico Aging Process Study 
(NMAPS; Garry et al., 1982, 1992). Individuals with overt clinical conditions, 
such as coronary heart disease, significant peripheral vascular disease, insulin-
dependent diabetes, hepatic disease, history of internal cancer requiring 
surgery, x-rays or chemotherapy in the past 10 years, a positive test for 
hepatitis, untreated hypertension, and those taking prescription medications 
except for thyroid and estrogen replacement, were excluded. As a part of the 
NMAPS, participants were also annually assessed on a battery of measurements 
of cognitive and physical function, health habits and attitudes, morbidity, 
number of falls, diet, physical activity, nutritional status, and body composition. 
In addition, participants were genotyped for Apoliporprotein E (APOE) e4 allele 
to assess genetic predisposition for developing Alzheimer's disease. Only 
47 
participants that were not genetically predisposed to Alzheimer's disease were 
chosen for this study, all were APOE e3 homozygotes. 
We found age-related deficits in performance of the elderly, relative to young 
participants on the VMWT consistent with well established reports of age-related 
decline in hippocampus-dependent cognition noted in numerous previous studies 
of aging. More specifically, elderly exhibited deficits in place learning in the 
virtual Morris water task as measured by latency and path length traveled to 
reach the hidden platform during training. Elderly participants also exhibited 
memory deficits compared to young participants, assessed by the time spent 
searching and path length traveled in the training quadrant of the pool on the 
no-platform probe trial of the VMWT. We found no differences in performance 
between young and elderly participants on conditions that were independent of 
the hippocampal integrity, such as the visible platform trials of the VMWT. This 
last finding also suggested that the elderly were able to successfully interact 
with the computer and navigate about the virtual environment despite their 
more than likely limited computer experience. 
We also suggested that the observed deficit in performance on the VMWT was 
not solely caused by the inherent constraints of the virtual environment, such as 
the lack of locomotor-based proprioceptive and vestibular cues and a narrower 
filed of view, since all of our participants were deprived of such information. The 
observed deficits in spatial navigation were not sensory-motor or motivational in 
48 
nature either given that there were no differences in the speed of swimming or 
in the performance on visible platform trials. 
Cross-sectional experiments with just the young and old groups, however, can 
possibly lead to erroneous conclusions drawn about the shape of underlying 
aging functions. Ideally longitudinal studies would be conducted to address such 
questions, but longitudinal studies are often precluded by inherent constraints, 
such as a time commitment, high costs, and high participant attrition. Care 
should be taken to at least include multiple age groups in order to obtain 
relevant information about the shape of the aging curve for a particular 
measure. We conducted a purely behavioral study to assess spatial navigation 
across adulthood (Driscoll et al., 2005). In this study we tested young (20-39 
years of age; N = 24), middle-aged (40-59 years of age; N = 23), and elderly 
(60-85 years of age; N = 23) adults on the VMWT. The young group performed 
quite well, which was expected. Surprisingly, the deficit already became 
apparent in the performance of the middle-aged group, and the performance of 
the elderly group was additionally worse. Results from our cross-sectional study 
suggest that, on the VMWT, performance declines linearly with age. 
The same elderly participants also exhibited a deficit on the non-spatial 
hippocampus-dependent task, the transverse patterning discriminations, while 
performance on discriminations (elemental) not dependent on the hippocampus 
was spared (Driscoll et al., 2003a). More specifically, only 19% of elderly were 
able to learn the relationship among transverse patterning cues compared to 
49 
100% of younger adults. On the other hand, 94% of elderly and 100% of 
younger adults were able to successfully complete the elemental discriminations, 
which do not depend on hippocampal integrity. Elderly people also made 
significantly more errors and required significantly more trials to complete the 
task. Moreover, the performance on the virtual Morris water task predicted 
performance on the transverse patterning discriminations, suggesting that the 
observed behavioral age-related deficits are hippocampus-dependent rather 
than unique to a specific memory process (such as spatial memory per se) or 
related to a specific test format. 
Again, in order to obtain relevant information about the shape of the aging 
curve for performance on the transverse patterning task, we tested young (ages 
20-39; N = 24), middle aged (ages 40-59; N = 23), and elderly (>60; N = 23) 
participants (Driscoll et al., 2004a). Our findings suggest that young and middle-
aged adults have little trouble solving the problem, while elderly participants are 
significantly impaired (Driscoll et al., 2004a). All three groups had no problems 
solving elemental (hippocampus-independent) discriminations. We find this 
observation especially interesting in light of our previous findings with the VMWT 
suggesting that performance on the spatial task seems to decline linearly with 
age (Driscoll et al., 2005); curiously a pattern that is different from our findings 
with the transverse patterning task. 
In this study, we have also attempted to answer whether normal aging is 
associated with decrease in hippocampal volume, and if so whether there is a 
50 
relationship between volume and hippocampus-dependent cognition (Driscoll et 
al., 2003a). Participants underwent MR imaging in order to obtain structural 
images, which were to further serve for quantitation of hippocampal volumes. 
Our structural findings indicate a reduction in hippocampal volume with age and 
are consistent with many earlier reports (for a review see Raz, 2000). In 
addition, the volume of the hippocampus correlated with performance on 
hippocampus-dependent tasks, both spatial and non-spatial. Specifically, larger 
hippocampi were associated with more time spent searching in the correct 
quadrant during the probe trial on the VMWT, and to the lower number of errors 
committed while learning the transverse patterning discriminations. 
In assessing age-related changes in hippocampus-dependent cognition, we also 
wanted to examine the underlying hippocampal biochemistry and relate it to 
hippocampus-dependent cognitive performance in these elderly participants. We 
found lower NAA/Cre values in the hippocampi of our elderly participants as 
measured by MRSI. Cho/Cre remained stable in both the frontal white matter 
and the hippocampus. Thus, the results seem to be primarily driven by the 
decrease in NAA and not in the variation in Cho and/or Cre containing 
compounds. Significantly lower hippocampal, compared to frontal, NAA/Cre 
indicated that the observed reduction in hippocampal NAA was not a 
consequence of global age-related metabolic brain deterioration. Also, we found 
no differences in metabolic profiles between the two groups in the frontal white 
matter. In addition, we found that better performance on both hippocampus-
dependent tasks was significantly correlated with higher hippocampal NAA/Cre. 
51 
Our combined behavioral and imaging study of a carefully selected aging sample 
sheds some light on hippocampal aging, as well as the neurobiological 
substrates of human hippocampus-dependent learning and memory. We have 
detected age-related variations linked specifically to the hippocampus, even 
though regional brain differences caused by aging are thought to be rather 
subtle compared to global brain deterioration. 
Despite the fact that we developed a human model of animal aging using tests 
that allow for cross-species comparisons, and described hippocampal aging on 
several different levels of investigation, several questions arose. Specifically, 
one question that is often linked to MRI research and volumetry is related to 
what a decline in volume of a structure really tells us? Does it denote atrophy? 
Or is it simply a reduction in size? In addition, in respect to MRS, more and 
more the functional role of NAA in the brain is questioned? Is the view of NAA as 
a putative neuronal marker useful in interpreting neurospectroscopic results, or 
is such a view an oversimplification? 
In order to address some of the questions arising from the previous studies, but 
also to validate and better develop our animal model of aging, and to better 
describe the changes that occur with age specifically related to the hippocampus 
we propose a multilevel analysis in the rat. 
52 
6. Experiment 1 - Visual Acuity 
Assessment of visual acuity in the Fisher 344 X Brown Norway hybrid 
strain of rat. Does visual acuity vary with age? 
6.1 Background 
Animal models are helpful for gaining insight into normal aging and for 
determining the contribution of various factors to age differences in behavior. A 
specific advantage of rat models of aging is that the contribution of genetic and 
environmental factors can be strictly controlled and thus systematically 
investigated. Recently, the degeneration of the visual system with age and the 
effect it may have on the performance of tasks with a visual component are 
receiving increased attention (van der Staay, 2002). A worry is that poor 
performance in mnemonic or cognitive tasks often exhibited by aged rats is 
possibly confounded by poor vision. 
Vision-mediated behavioral tasks, such as the Morris water task (Morris, 1981), 
the radial arm maze (Walker and Olton, 1979), and the Barnes circular maze 
(Barnes, 1979) are the most often used tests of spatial learning and memory in 
the rat. Oftentimes the results of these tests are interpreted without specific 
knowledge of the visual capacity of the rodent strain used. This omission is in 
part due to inability to test visual acuity in the rat. 
53 
Recently, however, Prusky and colleagues (2000a, 2004) developed two tasks 
that allow assessment of the visual acuity in rats and mice, the grating acuity 
test (Prusky et al., 2000a) and the virtual optomotor system (Prusky et al., 
2004). 
Previously, Prusky et al. (2002) measured grating acuity of three pigmented and 
three albino laboratory strains, and a wild strain. The acuity of the albino strains 
was the lowest, around 0.5 cycles per degree (cpd). The Dark Agouti, Long-
Evans and wild strains clustered around 1.0 cpd. Fisher 344 X Brown Norway 
(FBNF1) hybrid rats, however, had a significantly higher acuity threshold, 
around 1.5 cpd. In addition, it has been reported that rats with reduced visual 
acuity exhibit a significant deficit in place learning, although eventually they can 
reach the same latencies as the rats whose vision is not compromised (Prusky et 
al., 2000b). Visually impaired animals were not impaired in locating the visible 
platform, suggesting that the visible platform trials, which are often considered 
a control condition for sensory-motor or motivational deficits, may not be 
sensitive enough to detect small but significant changes in visual acuity. This 
particularly becomes a problem for the studies of aging that utilize the Morris 
water task in particular and other vision-mediated behavioral tasks alike. 
In examinig aging and memory in 25-27-month-old albino rats, performance on 
a Morris water task was found to be dependent on the structural integrity of the 
54 
retina (Osteen et al., 1995), suggesting that it is essential to consider vision 
during design of experimental procedures, both in aging and in general. Despite 
the considerations, the visual acuity in aged rats has not been assessed in a 
systematic manner yet. 
Given the highest acuity threshold, FBNF1 rats make a good choice for 
experiments employing vision-mediated behavioral tasks. FBNF1 rats are also 
bred and maintained by the National Institute on Aging (NIA, Bethesda, MD) 
specifically for gerontology research purposes. The question remains, however; 
is the visual acuity of this strain affected by age? In order to address this 
question, we employed two different tasks. In one experiment we measured the 
grating acuity in 14 month and 26 month old rats, and in the other we employed 
the virtual optomotor system to quantify visual acuity of 4, 14, and 26 months 
old rats. 
6.2 Materials and Methods 
Animals. Six middle-aged (14 months old) and six old (26 months old) male 
Fisher 344 X Brown Norway rats (NIA/Harlan colonies) were tested on the Visual 
Water Task. Six young (4 months old), six middle-aged (14 months old), and six 
old (26 months old) female Fisher 344 X Brown Norway rats (NIA/Harlan 
colonies) were tested on the virtual optomotor system. All rats were free of 
cataracts. Animals were cared for and tested at The University of Lethbridge 
55 
vivarium according the National Institutes of Health guidelines and under the 
approval of the institutional animal care committee which only approves studies 
that are in accordance with the Canadian Council on Animal Care guidelines. All 
animals were housed in pairs in hanging plastic cages, in a room with an 
ambient temperature of 21° C and 35% relative humidity, 12/12/ light dark 
cycle. Food and water were available ad libitum. 
The Visual Water Task. The Visual Water Task (VWT) has been previously 
described in detail by Prusky et al. (2000a). Briefly, a trapezoidal metal tank 
(140 cm L x 80 cm W x 25 cm W x 55 cm H) was placed on a solid table and 
filled with tap water to a depth of 15 cm. The wider end wall of the tank was 
transparent. A midline divider (40 cm H) was placed in the tank and extended 
46 cm into the pool from the transparent wall creating the two arms at the end 
of which the two computer monitors (17" VGA; Viewsonic E70F) were positioned 
on the outside of the tank (Figure 1). The bottoms of the screen monitors were 
level with the water surface. The monitors were used to control the presentation 
of stimuli. An Apple Macintosh computer (Power PC G4, 400 mHz) and a 
computer program (Vista; http://www.cerebralmechanics.com) were used to 
control the stimulus presentation, right/left randomization pattern, record 
behavioral responses, and provide parameter control. A transparent Plexiglas 
platform (37 cm L x 13 cm W x 14 cm H) was submerged at the end of one of 
the arms, and always paired with the correct ( + ) stimulus. 
56 
The VWT task capitalizes on rat's natural inclination to escape water by finding a 
solid substrate (a platform in this case). The rats were first pre-trained to 
associate the escape from water with the platform by learning to discriminate 
between the low spatial frequency grating from a gray screen, spatial frequency 
always being paired with the platform according to the procedure outlined by 
Prusky et al. (2000a). A low spatial frequency, vertical sine wave grating (0.12 
cpd; 4 sine wave cycles on the computer screen), was displayed on one monitor, 
and if chosen rewarded ( + ) by escaping the water. A uniform gray of the same 
mean luminance (- stimulus) was displayed on the other monitor. The escape 
platform was located directly below the rewarded stimulus. On each trial, rats 
were removed from their holding cage and released into the pool facing the 
computer screens. Once on the platform, they were allowed to remain there for 
a few seconds before being returned to their holding cages. On the next trial the 
location of the grating and the platform were switched to the opposite arm of 
the pool, and the side of presentation alternated from then on. 
During the training phase, the rats were trained to discriminate between the 
same two stimuli as in the pre-training phase, the sine wave grating versus gray 
screen. Once the rats performed at 80% correct on 30 consequent trials the 
testing of grating threshold was initiated employing a method-of-limits 
procedure where incremental changes in the spatial frequency of the sine wave 
stimulus were implemented. More concretely, if a rat made a correct choice one 
sine wave cycle was added to the grating (~ 0.3 - 0.49 cpd) on the subsequent 
trial. This continued until the performance accuracy fell below 70%. 
57 
If an error occurred, additional trials with the same spatial frequency were 
administered until three correct choices were made in sequence, or seven 
correct choices out of 10 consecutive trials after which the sine wave cycle of 
the gratings was increased on the next trial. As the spatial frequency increased 
(~ 0.5 - 0.8 cpd), the minimal number of trials at each spatial frequency to 
allow the addition of one sine wave cycle was increased to three. A similar error 
correction procedure as described above was used; either five consecutive 
correct trials or seven out of 10 trials were required before spatial frequency 
could be increased. In the next block of spatial frequencies (~ 0.8 - 1.0 cpd), 
five consecutive correct trials or seven out of 10 trials were needed to warrant 
an increase in spatial frequency. For the gratings with spatial frequencies higher 
than ~1.0 cpd, seven out of 10 correct trials were required to proceed with an 
increase in spatial frequency. The choice was deemed correct if the rat swam to 
the platform without entering the arm displaying the non-rewarded stimulus. 
The rat was judged as entering the incorrect arm if the hind legs crossed into 
this arm of the pool, even if the rat had turned around before reaching the 
platform. The pattern of the rewarded stimulus/platform location was a left (L), 
right (R), LLRLRR sequence. The rats swam in a group of six, one at a time, and 
the inter-trial interval was approximately four minutes. 
A preliminary grating threshold was established when animals failed to achieve 
70% accuracy at a specific spatial frequency. To confirm the validity of the 
threshold the spatial frequency of the grating was decreased by 3-4 cycles (0.9 
58 
-1.2 cpd) and the experimental procedure described above was followed until 
the performance again dropped below 70%. The point at which the performance 
repeatedly dropped below 70 % accuracy was recorded as the grating acuity. 
The Virtual Optomotor System. A virtual cylinder comprised of a vertical sine 
wave grating was projected in 3D coordinate space on computer monitors 
arranged in a square around the testing arena. The arena consisted of a 
Plexiglas box (39 cm L x39 cm W x 32.5 cm H) with rectangular openings (33.5 
cm W x 26.5 cm H) on each wall painted white on the inside. A mirror with a 
small hole in the center of it was positioned tangentially to the bottom of the 
openings. A vented lid enclosed the top of the apparatus. A FireWire camera 
(iSight; Apple Computer Corporation) was attached to the lid with a holding 
sleeve and centered over the apparatus. 
A computer program (OptoMotry©; Cerebral Mechanics), running on a dual-
processor G4 Power Macintosh (Apple Computer Corporation), was used to drive 
Open GL-compatible video cards (Radeon 7000 Mac Edition; ATI) and project on 
the monitors a virtual cylinder in 3D coordinate space. Visual stimuli were 
presented on the walls of the cylinder and the image was reflected by the floor 
and ceiling mirrors such that each monitor appeared as a window onto the 
surrounding 3D world. The program controlled the speed of rotation and 
geometry of the cylinder, the spatial frequency and contrast of stimuli, and 
enabled live video feedback of the testing area. 
59 
Rats were placed on the platform in the center of the apparatus one at a time 
and the lid was closed. As the rat moved freely around the platform, the 
experimenter followed the rat's head with a crosshair superimposed on the video 
image. The crosshair's X, Y coordinates were used to center the rotation of the 
cylinder at the rat's viewing position so that the virtual walls of the cylinder were 
maintained at a constant distance from the animal. When a grating detectable 
by the rat was projected on the cylinder wall and the cylinder was rotated (12 
deg/sec), the rat would stop moving and would begin to track the grating with 
reflexive head movements. 
The experimenter assessed tracking by monitoring the video image of the 
cylinder, the animal, and the crosshair simultaneously. If the rat's head tracked 
the cylinder rotation (moved against stationary arms of the crosshair), it was 
judged that the animal could see the grating. Again, the staircase or method-of-
limits procedure was used to systematically increase the spatial frequency of the 
grating until the rat no longer responded. The highest spatial frequency that the 
rat could track in either clockwise or counterclockwise direction was recorder as 
the threshold. The apparatus and the training procedures are described in more 
detail elsewhere (Prusky et al., 2004). 
60 
6.3 Results 
The Visual Water Task. The effect of age on grating acuity was statistically 
assessed by performing a one-way univariate analysis of variance (ANOVA) with 
the grating acuity as a dependent variable and Group (Middle-aged vs. Old) as 
an independent variable. The mean grating acuity for the Middle-aged FBNF1 
rats was 1.543 cpd and for the Old ones 1.36 cpd, and this difference in acuity 
between the two groups was significant (F (1, 11) = 77.207, p < .001; Figure 
2). 
The Virtual Optomotor System. The visual acuity, as measured by a virtual 
optomotor system, was statistically assessed in relation to age group by 
performing a one-way ANOVA. A score for tracking the gratings in clockwise and 
counterclockwise direction respectively served as a dependent variable and 
Group (Young, Middle-aged, and Old) served as an independent variable. 
Although subtle, the difference in tracking in both clockwise (F (2,17) = 5.529, p 
= .016; Figure 3A) and counterclockwise (F (2,17) = 8.107, p = .004; Figure 
3B) direction was significant between age groups. Follow-up tests revealed no 
difference in acuity between Young and Middle-aged groups (p's > .2). The 
acuity of the Old group, however, was significantly lower than the acuity of the 
Young (p's < .015) or the Middle-aged (p's < .018) groups. 
61 
6.4 Discussion 
The main goal of the present study was to assess whether age has an effect on 
visual acuity of the FBNF1 rats. We found that the acuity, as measured by the 
grating acuity test did drop from ~1.54 cpd in the middle-aged rats to ~ 1.36 
cpd in the old rats. Our results suggest that this difference in acuity was 
significant. Similar results were observed with the virtual optomotor system, 
where there were no differences in acuity between the Young and Middle-aged 
rats, but the grating threshold of the Old group was significantly lower from that 
of the other two age groups. 
Our results also replicate findings reported by Prusky et al. (2002), that the 
visual acuity of the FBNF1 strain of rats clusters at about ~ 1.5 cpd. The rats in 
the present study with visual acuity of ~ 1.5 cpd, however, were 14 months of 
age while the rats tested by Prusky et al., (2002) were under 6 months of age. 
By comparing the present findings pertaining to the visual acuity of the FBNF1 
strain of rats to those reported by Prusky et al. (2002) we can draw two 
additional conclusions: (1) that there are no changes in visual acuity up to the 
14 months of age in this strain of rat, and (2) that the drop in acuity occurs 
somewhere between 14 month and 2 years of age. 
Notwithstanding the significant difference in the grating acuity, it is important to 
note that the acuity of old FBNF1 rats (age > 24 months; 1.36 cpd) is still much 
better than the acuity of young adult pigmented rats (such as the Long-Evans 
62 
rat for example), which clustered around 1.0 cpd (Prusky et al., 2002). 
Similarly, the acuity threshold of the old FBNF1 rats, as measured by the virtual 
optomotor system, is not significantly different from the acuity of the young 
Long-Evans rats (~ .54 cpd; unpublished observations). Collectively, our results 
suggest that the Fisher 344 x Brown Norway hybrid rats are a suitable strain for 
aging research employing vision-mediated behavioral tasks. 
63 
7. Experiment 2 - Long -Term Potentiation 
Is hippocampal synaptic plasticity affected by age? - A study of 
long-term potentiation 
7.1 Background 
In the early 1970's, Bliss and colleagues (Bliss & Lomo, 1973; Bliss and 
Gardner-Medwin, 1973) reported a long-lasting enhancement of synaptic 
strength in response to high-frequency stimulation. This phenomenon came to 
be commonly known as long-term potentiation (LTP; Douglas and Goddard, 
1975) or long-term enhancement (LTE). Early experiments suggested that LTP 
could be a cause of synaptic changes underlying memory formation (see 
Rosenzweig and Barnes, 2003 for a review). Research also suggests that age-
related learning and memory impairments are, at least in part, caused by 
altered hippocampal synaptic plasticity (Barnes, 2003; Barnes & McNaughton, 
1985; Geinisman et al., 1995; Rosenzweig & Barnes, 2003). Hence, altered 
hippocampal synaptic plasticity represents the primary electrophysiological 
correlate of age-related memory impairments. 
There is agreement that age does have an effect on LTP in the dentate gyrus of 
the hippocampus, but only under certain conditions. For example, such 
conditions, which have been blamed for the variable results reported in the 
64 
literature, include the stimulation parameters and the strain of rat. Whether 
age-related differences are seen also depends on whether induction, 
maintenance, or decay of LTP is being measured. Since the pioneering report by 
Barnes (1979), it has been widely accepted that LTP decays more quickly in the 
aged rats. In addition, performance on hippocampus dependent tasks was 
correlated with the decay rate of LTP in both rats (Barnes, 1979; Barnes and 
McNaughton, 1980) and mice (Bach et al., 1999). Whether aged rats have a 
deficit in LTP induction, however, remains a subject of debate. 
Many early experiments of the effect of age on LTP, using long, high-frequency 
stimulation, report no age-related deficits in LTP induction in the dentate gyrus 
of the hippocampus in vivo (Barnes, 1979; deToledo-Morrell and Morrell, 1985) 
and in vitro (Diana et al., 1994a,b). There are just as many experiments, 
however, reporting age-related LTP induction deficits (Lynch and Voss, 1994; 
McGahon et al., 1997; Murray and Lynch, 1998a,b; Tielen et al., 1983). Given 
that most of the experiments reporting deficits come from a single laboratory, 
they have been largely discounted and their findings attributed to crucial 
methodological differences and possibly the strain of rat used (see a review by 
Geinisman et al., 1995). 
Instead, the view proposed by Barnes and colleagues, that there are no age-
related deficits in the induction of LTP, has dominated this field for the past 20 
years. This view is based on the finding that old rats are capable of reaching the 
same maximum level of enhancement as the young rats, and does not take as 
65 
important the fact that the old rats take twice as long to reach the maximum 
than do young rats (Barnes and McNaughton, 1985). Recently Barnes et al. 
(2000) demonstrated, however, an increased threshold for LTP induction at the 
perforant path/dentate gyrus synapse, suggesting that there is an age-related 
induction deficit in addition to a deficit in LTP maintenance. 
For our current set of experiments, we chose the Fisher 344 X Brown Norway 
(FBNF1) hybrid rats bred by the NIA particularly for aging research purposes. 
We now know (from Experiment 1) that these rats have a high visual acuity, 
especially in young but reasonably so in old age, making them especially 
suitable subjects for testing learning and memory in vision-mediated behavioral 
tasks. We decided to evaluate hippocampal synaptic plasticity in the FBNF1 
strain using high-frequency stimulation before proceeding with assessment of 
learning and memory abilities. 
Our motivation for assessing synaptic plasticity in FBNF1 rats in relation to age 
revolved around the fact that the hippocampal synaptic plasticity had not been 
investigated much in the FBNF1 rats so we did not know what to expect in terms 
of synaptic plasticity in these animals. In addition, the strain of rat along with 
stimulation parameters has been seriously considered as a possible cause of 
disparate findings in the literature on LTP in the aged hippocampus. Hence, the 
primary goal of the present experiment is to identify whether there are age-
related changes in hippocampal synaptic plasticity in FBNF1 rats, given that we 
plan to use this strain of rat for further studies on the hippocampus and aging. If 
hippocampal LTP is compromised in aged FBNF1 rats and it recruits similar 
66 
synaptic plasticity mechanisms as memory formation, then we predict that age-
related LTP deficits should be associated with age-related memory deficits. The 
memory deficits should be particularly notable on tasks that are dependent upon 
the integrity of the hippocampus. 
7.2 Materials and Methods 
Animals. Ten middle-aged (12 month old) and ten old (24 month old) male 
FBNF1 rats (350-425g at the start of the experiment) were housed in pairs in 
clear plastic cages in a colony room at the University of Lethbridge that was 
maintained on a 12hr on/12hr off light cycle. Rats had access to food and water 
ad libitum. All animals were cared for and tested at the University of Lethbridge 
vivarium according to the National Institutes of Health guidelines and under the 
approval of the institutional animal care committee, which only approves studies 
that are in accordance with the Canadian Council on Animal Care guidelines. 
Surgical and Recording Procedures. All rats were implanted unilaterally with 
a stimulating electrode (Figure 4A) in the perforant path and a recording 
electrode in the dentate gyrus (Figure 4B). The rats were anesthetized with 
isofluorane (induction at 2.5- 4%; maintenance at 1- 2.5%; in 1.5 L/min 02) 
and stereotaxically implanted with a single stimulating and a single recording 
electrode. The electrodes were constructed using Teflon-coated stainless steel 
wire, 114 \im in diameter crimped onto gold-plated male amphenol pins (A & M 
67 
Systems). 
The recording electrode was lowered into the hilar region of the dentate gyrus 
(Paxinos and Watson, 1986; coordinates: 3.2 mm posterior to bregma, 1.6 mm 
lateral to the mid-sagittal suture and approximately 3.5 mm below the dura 
matter) and the stimulating electrode was lowered into the ipsilateral perforant 
path (Paxinos and Watson, 1986; coordinates: 8.1 mm posterior to bregma, 4.3 
mm lateral to the saggital suture, and approximately 3.0 mm below the dura 
matter). Three stainless steel jewelers' screws tapped into the skull served as 
reference and ground components of the differential recording circuit. The 
position of both the stimulating and recording electrode was optimized under 
electrophysiological guidance using single pulse stimulation (pulse duration = 
100 \is, amplitude = 300 fxA, frequency = 1 pulse/30 s). The electrodes were 
secured in place using dental acrylic. For the chronic experiment, the amphenol 
pins were inserted into a 9-pin Mclntyre-Molino plug, which was subsequently 
cemented onto the skull. 
All rats were habituated to the recording room for two 15-minute sessions for 
two days before recordings began. All recordings sessions were performed in the 
animal's own home cage. Evoked potentials were recorded from the 
hippocampus using Data Wave System Experimenter's Workbench Package. A 
computer was interfaced with a digital stimulator (AMPI Master 8) connected to 
a constant current stimulator/stimulus isolation unit (AMPI iso-flex) for electrical 
stimulation. All stimulation pulses were rectangular monophasic pulses, 100 \is 
68 
in duration. The evoked potentials were differentially recorded, amplified 200 
times, and filtered at half amplitude between 1 Hz and 10 kHz. The signal was 
acquired by the computer at a sampling rate of 33 kHz, triggered 5 ms before 
and continuing for 30 ms after the delivery of each stimulation pulse. An audio 
analogue of the signal was monitored, and all the individual sweeps were 
displayed on the computer monitor and acquired on the hard drive. 
One week after implantation two baseline input-output (I-O) measures were 
obtained from awake freely-behaving rats. Stimulation pulses were delivered at 
100, 300, 500, 700 and 900 \xA. Chronic LTP was induced using high frequency 
stimulation (HFS; 20 ms bursts of 500 Hz stimuli, repeated 10 times) delivered 
once per day for 10 consecutive days followed by 1-0 measures twice daily 1 
and 24 hours following HFS. 
For each rat, the evoked potentials were analyzed for population spike amplitude 
and excitatory postsynaptic potential (EPSP) slope, and the resultant values 
were averaged at each intensity. The population spike served as an index of 
granule cell excitability, and was measured as the area under the tangent 
between the spike onset and offset. The EPSP slope served as an index of 
synaptic strength, and was measured as the slope of a line between the field 
EPSP onset and the onset of the population spike. These two measures were 
derived using the Data Wave software program. All values obtained for the 
follow-up recordings were analyzed as averaged fractional changes from 
baseline. 
69 
7.3 Results 
Effects of HFS on population spike amplitude in middle-aged and old 
rats. Repeated measures analyses of variance (ANOVA) was conducted to 
compare the population spike (PS) amplitude for the middle-aged and old group 
either 1 hour or 24 hours post high-frequency stimulation (HFS) respectively 
across 10 consecutive recording days. A repeated measures ANOVA revealed an 
Age X Day interaction, whereby the HFS led to a differential effect on PS 
amplitude for middle-aged and old rats 1 hour post-stimulation across the 10 
days of recording (F (1,120) = 8.86, p < 0.001). For each repeated measures 
ANOVA, follow-up tests compared the middle-aged and old groups on each day. 
Follow-up tests revealed that whereas the PS amplitude of middle-aged and old 
rats did not differ 1 hour after the first HFS session, it differed between groups 
after each subsequent session (all significant comparisons: p < 0.05; see Figure 
5A). A t-test was computed to compare the baseline PS amplitude for the 
middle-aged and old rats to the PS amplitude obtained 1 hour following the first 
HFS session and revealed a significant increase in PS amplitude for both middle-
aged (t (9) = - 4.783, p = .001) and old rats (t (9) = -2.967, p = .016). The 
increase in the PS was evident in the middle-aged rats but not in the old rats 
following the initial increase after the first session of HFS. 
A separate repeated measures ANOVA revealed an Age X Day interaction, 
whereby the HFS led to a differential effect on PS amplitude for middle-aged and 
old rats 24 hours post stimulation (F (1,120) = 8.69, p < 0.001). Follow-up 
70 
tests revealed that whereas the PS amplitude of middle-aged and old rats did 
not differ 24 hours after the first HFS session (p's > 0.05), it differed between 
groups after each subsequent session (all p's < 0.05; see Figure 5B). Again, the 
increase was evident in the middle-aged but not old rats. 
A t-test was computed to compare the baseline PS amplitude to the PS 
amplitude obtained 24 hours following the first HFS session and revealed a 
significant increase in PS amplitude for middle-aged (t (9) = 8.683, p < .001) 
but not old rats (t (9) = 1.351, p = .214). Representative responses evoked in 
the dentate gyrus of the hippocampus by stimulation of the ipsilateral perforant 
path prior to, 1 hour after the first, and 24 hours after the 10th HFS session for 
the middle-aged and old rats are shown in Figure 6 A-H. 
Effects of HFS on the EPSP slope of middle-aged and old rats. Repeated 
measures ANOVAs were conducted to compare the excitatory post-synaptic 
potential (EPSP) slope for the middle-aged and old group either 1 hour or 24 
hours post high-frequency stimulation (HFS) respectively across 10 consecutive 
recording days. For each repeated measure ANOVA, follow-up tests compared 
the middle-aged and old group on each day. We found an Age X Day interaction, 
whereby the HFS led to a differential effect on EPSP slope for middle-aged and 
old rats 1 hour post-stimulation across the 10 days of recording (F (1,120) = 
2.124, p = 0.023). Follow-up tests revealed that whereas the EPSP slope of 
middle-aged and old rats did not differ 1 hour after the first HFS session, it 
differed between groups after each subsequent session (p's < 0.05; Figure 5C). 
71 
Post hoc analyses revealed that a significant increase in EPSP slope occurred 1 
hour following the second HFS session in the middle-aged rats, and 1 hour 
following the 7th HFS session in the old rats (p < 0.05). 
A t-test was computed to compare the baseline EPSP slope for the middle-aged 
and old rats to the EPSP slope obtained 1 hour following the first HFS session 
and revealed that there was no significant increase in EPSP slope in either the 
middle-aged or the old rats. A t-test was computed to compare the baseline 
EPSP slope for the middle-aged and old rats to the EPSP slope obtained 1 hour 
following the first HFS session and revealed no significant increase in both 
groups (p 's > .25). A separate t-test was computed to compare the baseline 
EPSP slope for the middle-aged and old rats to the EPSP slope obtained 1 hour 
following the second HFS session and revealed a significant increase in EPSP 
slope for middle-aged (t (9) = 2.595, p = .029) but not old rats (t (9) = .583, p 
= .576). The increase in the EPSP was evident in the middle-aged rats but not in 
the old rats following the initial increase after the first session of HFS. 
A separate repeated measures ANOVA revealed an Age X Day interaction, 
whereby the HFS led to a differential effect on the EPSP slope for middle-aged 
and old rats 24 hours post stimulation (F (1,120) = 4.92, p < 0.001). Follow-up 
tests revealed that whereas the EPSP slope of middle-aged and old rats did not 
differ 24 hours after the first HFS session, it differed after each subsequent 
session (p' s < 0.05; see Figure 5D). A t-test was computed to compare the 
baseline EPSP slope to the EPSP slope obtained 24 hours following the second 
72 
HFS session and revealed a significant increase in EPSP slope for middle-aged (t 
(9) = 3.946, p < .001), and a non-significant trend for the old rats (t (9) = 
1.896, p = .09). 
Additionally, a difference score was computed for both the PS amplitude and 
EPSP slope respectively across 11 recording days (baseline and 10 HFS sessions) 
in order to obtain a measure of 24-hour retention between the recordings 
obtained 1 hour and 24 hours after baseline and each subsequent HFS. The 
differences in the percentage of change in both PS amplitude and EPSP slope 
retained for 24-hours after each HFS were largely insignificant in both middle-
aged and old rats (p's > .07). The only significant differences in 24-hour 
retention were observed for EPSP slope after the second HFS (p = .017), which 
is also the first day when group differences emerged in the amount of change in 
the slope of EPSP. Significant differences in 24-hour retention of PS amplitude 
were observed at baseline (p = .024) and after 4th HFS session (p = .044). 
7.4 Discussion 
Despite the repeated HFS over 10 consecutive days, we find a substantially 
reduced responsiveness in the hippocampus of the 24-month old FBNF1 rats 
compared to the 12-month old rats. Even though both groups of animals 
potentiated to the same degree on the first day of stimulation, the hippocampal 
73 
circuitry of 12-month old FBNF1 rats seems to remain more plastic than that of 
the 24-month old rats. 
There was a significant change in PS amplitude compared to baseline after the 
very first HFS in both age groups, but not in the EPSP slope. For each following 
day of HFS, fractional change from baseline in PS amplitude of old rats changed 
very little from the level achieved after the first HFS. The fractional change from 
baseline in PS amplitude of middle-aged rats, however, increased significantly 
after the second HFS session and reached asymptote on day 8 (Figure 5A & B). 
Similarly, the fractional change in the EPSP slope increased significantly after 
the second day of HFS in the middle-aged animals and continued to significantly 
differ from baseline across the 10 days of stimulation. For the old rats, a 
significant increase in the EPSP slope occurred following the 7th HFS session. 
Although the possibility remains that eventually the old rats could potentiate to 
the same degree as the middle-aged rats, our results suggest that there is a 
deficit in the induction of LTP in aged FBNF1 rats. 
Similarly, the pioneering work by Barnes (1979) suggested that there were no 
age-related difference in the magnitude of LTP induction after the first day of 
stimulation. The difference emerged in persistence of LTP, and only after the 
third day of stimulation suggesting that repeated HFS is needed for sustained 
LTP even in the young animals. The effect of repeated stimulation on LTP 
persistence was not observed in the aged animals. Another important finding 
was that the performance on the circular maze correlated with the amount of 
74 
LTP induced after the third HFS session, providing the first support for the 
hypothesis that LTP and spatial learning may share similar mechanistic 
processes (Barnes, 1979). The question remained, however, whether old rats 
could reach the same amount of potentiation, possibly with more repetition. 
Barnes and McNaughton (1980) examined this possibility over 12 consecutive 
days of repeated stimulation and found that the old rats can eventually reach 
the same magnitude of LTP as the young rats, a finding that has for years been 
interpreted as the lack of age-related deficits in LTP induction. The young rats, 
however, reached their maximum level of potentiation on day 5, whereas the old 
rats reached the maximum on day 10 - clearly not a negligible difference in the 
induction. 
The discovery that received more attention was related to the finding that LTP in 
the old animals, even though it reached the same magnitude of potentiation as 
it did in the young rats, decayed much faster. This suggested that the rate of 
forgetting may follow a similar pattern. Indeed, this turned out to be the case 
(Barnes and McNaughton, 1985). At the same time, reports emerged of a 
significant relationship between LTP and performance on tasks shown to be 
dependent on the hippocampal integrity, specifically spatial learning and 
memory (see Geinisman et al., 1985, Rosenzweig and Barnes, 2003 for a 
review). 
Some of the more convincing evidence for the role of LTP in information 
processing and storage came from studies that involved manipulations that 
75 
affect both memory and LTP in a similar manner. For example, Morris et al., 
(1986) demonstrated that pharmacological agents that impair spatial memory in 
rats also block the induction of LTP. On a similar note, it has been demonstrated 
that agents that improve spatial memory in aged rats also improve the retention 
of LTP (deToledo-Morrell et al., 1988). 
Our results are similar to those reported by Barnes (1979), in that there are no 
age-related differences in LTP induction after a single LTP session. Nonetheless, 
with repeated high frequency stimulation protocols, our findings diverge. Barnes 
and McNaughton (1980) extended their stimulation protocol to 12 days and 
found that old rats just take longer to reach the same maximum as the young 
rats. Notwithstanding the slight difference in stimulation parameters and the 
strain of rats used in the present study compared to those employed by Barnes 
(1979, 2003) and colleagues (1980, 1985), our results suggest that old FBNF1 
rats did not benefit from repeated stimulation. After the initial incrase in the 
populations spike amplitude for both groups after the first HFS session, we 
observed no additional change in the population spike amplitude of old rats 
across the remaining nine days of HFS. A significant change in EPSP slope was, 
however, observed after day 7 in old animals suggesting that maybe they could 
have benefited from additional HFS. Nevertheless, the fact that there was no 
substantial change in PS amplitude with repeated HFS, and some change in 
EPSP but only after day 7 leads us to conclude that hippocampal synaptic 
plasticity is importantly compromised in aged FBNF1 rats. 
76 
The stimulation parameters and the strain differences remain to be reconciled. 
For the purpose of present experiments, our findings suggest that age-related 
changes in hippocampal plasticity may be expected, and may further underlie 
some of the behavioral deficits on hippocampus-dependent tasks seen with age. 
Hence, hippocampus-dependent learning and memory deficits are anticipated in 
the aged FBNF1 rats. 
77 
8. Experiment 3 - Spatial Learning and Memory 
The Morris water task: Does age have an effect on learning and memory 
of spatial hippocampus-dependent information? 
8.1 Background 
Over the years, numerous tasks have been developed to assess spatial learning 
and memory in rodents (for examples see a review by Geinisman et al., 1995). 
A task that is perhaps most commonly used to assess spatial learning and 
memory, and is considered a litmus test of hippocampus-dependent cognition, is 
the Morris water task (MWT; Morris, 1981). The performance on this task is 
similarly affected by hippocampal lesions (Gallagher and Holland, 1992; Morris 
et al., 1982; Sutherland et al., 1982) and old age (Gage et al., 1984, 1988; 
Pelleymounter et al., 1987; Rapp et al., 1987; van der Staay, 2002). 
The deficits seen with old age, while qualitatively similar compared to the 
deficits produced by lesions, are thought to be quantitatively less severe. Both 
aged rodents (Gallagher and Pelleymounter, 1988; Rapp et al., 1987) and those 
with hippocampal damage (Morris et al., 1982; Morris, 1983) are unimpaired in 
locating a visible platform, suggesting that the impairment is indeed in learning 
and memory, and not sensori-motor in nature. 
78 
Spatial learning and memory, in general, are affected in aged humans as well 
(Driscoll et al., 2003a; Geinisman et al., 1995; Moffat and Resnick, 2002; Moffat 
et al., 2001; Raz, 2000). More specifically, we found an age-related deficit in 
performance on the virtual Morris water task, while there were no age-related 
differences in swimming to a visible platform in the virtual environment. In 
addition, the age-related deficits in learning and memory for spatial information 
are not only present in the elderly, but already become apparent in the middle-
age group (Driscoll et al., 2005). 
We already know from Experiment 1 that FBNF1 rats, even the aged ones, have 
a good visual acuity. Hence, the use of visual tests to assess learning and 
memory here is appropriate, as it will not be biased by age-related changes in 
vision. In addition, results of Experiment 2 suggest that the hippocampal 
synaptic plasticity of our aged rats should be compromised. Given that the 
relationship between synaptic plasticity and memory is often reported, we 
expect to find compromised learning and/or memory on tasks dependent on the 
hippocampus. In order to confirm that our rats have an age-related deficit in 
learning and/or memory for hippocampus-dependent spatial information, we 
trained them on the working memory variant of the Morris water task in which 
the platform moved to a new location after every two days. By manipulating the 
location of the platform in such a manner, we were able to obtain measures of 
both acquisition and retention of new information. 
79 
8.2 Materials and Methods 
Animals. Thirty Fisher 344 Brown Norway hybrid female rats (NIA/Harlan) 
participated in the experiment. Rats were divided into 3 groups: Young (/V = 10; 
age 3 months at the start of the experiment), Middle-aged (A/ = 10; age 14 
months at the start of the experiment), and Old (A/ = 9; age 24 months at the 
start of the experiment) group. One old rat died during the MR scanning 
procedures (which were performed prior to behavioral testing) due to pre 
existing stroke. All rats were free of cataracts. Rats were maintained at the 
University of Lethbridge vivarium, housed in pairs in hanging plastic cages, in a 
room with an ambient temperature of 21° C and 35% relative humidity, 12/12/ 
light dark cycle. Food and water were available ad libitum. All procedures 
involving animal care and experimentation were carried out in accordance with 
guidelines provided by the National Institutes of Health and approved by the 
institutional animal care committees at the University of Lethbridge, University 
of Calgary, and the National Research Council of Canada, which only approve 
procedures that are in accordance with the Canadian Council on Animal Care. All 
behavioral testing procedures were carried out at the University of Lethbridge. 
The Morris Water Task. All rats were trained in a modified version of the 
Morris water task (Morris, 1981) in which the platform moved to a new location 
every second day according to a random sequence of locations. This version of 
the task is useful for observing new learning as well as 24-hour retention. For 
the first two days the platform remained in the same location. On the 3rd day the 
80 
platform moved to a new location for another pair of days and so on until 4 
locations were completed (8 consecutive days). So all rats had been exposed to 
the MWT training for 8 days and experienced 4 different platform locations 
(center of NE, SW, SE, and NW quadrant). All rats were pre-trained on the task 
for 2 days prior to data acquisition. In addition, after completion of 8 days of 
hidden platform training, we administered one day of visible platform training. 
Rats received 8 trials on each training day and were released twice from each 
cardinal compass point around the perimeter of the pool in a pseudorandom 
sequence. A maximum of 60 s was allowed to locate the platform. If a rat failed 
to locate the platform within 60 s, it was removed from the water and placed on 
the platform for 10 s before returning to the holding cage. 
The apparatus involved a circular pool with white interior, filled to within 20 cm 
of the top of the wall with water (21°C, 40 cm deep) rendered opaque with milk 
powder. The platform was constructed of clear Plexiglas (13cm diameter), and 
its top surface was 1.5 cm below the surface of the water, positioned in the 
center of the NE quadrant of the pool. Performance was recorded and analyzed 
by a video tracking system (HVS 2020 Plus system). Measures recorded 
included latency and path length to reach the platform, and the speed of 
swimming. 
81 
8.3 Results 
Latency, path length, and speed of swimming measures were averaged for each 
trial across all odd (new platform location) and all even (new platform location) 
days separately, and served as a dependent variable respectively. Repeated 
measures analysis of variance (ANOVA) was conducted with Trial (8) as a 
within-subjects variable, and Group (3; Young, Middle-aged, and Old) as a 
between-subjects variable for each of the dependent measures on odd versus 
even days. 
MWT - Hidden Platform 
Latency. For latency to locate the platform on days (odd) when the platform 
moved to a new location, we found a significant effect of Trial (F (7, 175) = 
55.729, p < .001) and Group (F (2, 25) = 18.532, p < .001). The Group X Trial 
interaction was not significant (p > .1). For days (even) when the platform 
remained in the same location as the previous day, there was a significant 
Group X Trial interaction (F (14, 175) = 3.583, p < .001). The effects of Trial (F 
(7, 175) = 18.537, p < .001) and Group (F (2, 24) = 18.541, p < .001) were 
also significant. Latency to locate the platform across trials for odd and even 
days respectively is represented in Figure 7A & B. Overall, latency across trials 
as a measure of learning the new platform location suggests that all rats, 
regardless of their age, were able to acquire the information in a similar 
manner. 
82 
Path Length. For path length traveled to reach the platform on days (odd) 
when the platform moved to a new location there was a significant Group X Trial 
interaction (F (14, 175) = 1.848, p < .035), and a significant effect of Trial (F 
(7, 175) = 45.747, p < .001) and Group (F (2, 25) = 8.075, p = .002). For days 
(even) when the platform remained in the same location as the previous day, 
there was a significant Group X Trial interaction (F (14, 175) = 2.514, p = 
.003). The effects of Trial (F (7, 175) = 13.552, p < .001) and Group (F (2, 25) 
= 9.591, p = .001) were also significant. Path length to reach the platform 
across trials for odd and even days respectively is represented in Figure 7C & 
D.As it was the case with latency, path length across trials as a measure of 
learning the new platform location suggests that all rats, regardless of their age, 
were able to acquire the information in a similar manner. 
Swim Speed. For the speed of swimming, on days (odd) when the platform 
moved to a new location, there was a significant Group X Trial interaction (F 
(14, 175) = 2.037, p = .018). Also, there was a significant effect of Trial (F (7, 
175) = 2.477, p = .019), and a non-significant trend for the Group (p > .05). 
For days (even) when the platform remained in the same location as the 
previous day, there was a significant Group X Trial interaction (F (14, 175) = 
3.957, p < .001). The effects of Trial (F (7, 175) = 12.747, p < .001) and Group 
(F (2, 25) = 10.826, p < .001) were also significant. Swimming speed across 
trials for odd and even days respectively is represented in Figure 7E & F. 
Overall, the results suggest that even though significant differences in the speed 
83 
of swimming can be observed on certain trials between groups, the old rats can 
reach the same maximum speeds as their younger counterparts. 
24-hour Retention. In order to assess the retention of newly learned platform 
location over a period of 24 hours in different age-groups, we computed 
difference scores between the last trial of the day when the platform was in a 
new location and the first trial of the next day when the platform was in the 
same location as the day before. The 24-hour difference scores were averaged 
across 4 different platform locations. This measure served as a dependent 
variable with Group membership as an independent variable. We found a 
significant difference in 24-hour retention based on age-group membership 
latency to locate the platform (F (2, 27) = 3.583, p = .043; Figure 8) and a 
non-significant trend for path length (p = .082), such that both middle-aged and 
old rats were impaired in 24-hour retention of spatial information. 
MWT-Visible Platform. Latency, path length, and speed of swimming served 
as a dependent variable respectively. Repeated measures ANOVA was conducted 
with Trial (8) as a within-subjects variable, and Group (3; Young, Middle-aged, 
and Old) as a between-subjects variable for each of the dependent measures. 
For latency, there was a significant Trial X Group interaction (F (14, 175) = 
5.362, p < .001), and a significant effect of Trial (F (7, 175) = 34.049, p < 
.001) and Group (F (2,25) = 7.624, p = .003). Similarly, there was a significant 
Trial X Group interaction (F (14, 175) = 6.008, p < .001), and a significant 
effect of Trial (F (7, 175) = 33.006, p < .001) and Group (F (2,25) = 11.660, p 
84 
< .001). For the path length to reach the visible platform, there was a 
significant Trial X Group interaction (F (14, 175) = 3.326, p < .001), and a 
significant effect of Trial (F (7, 175) = 26.865, p < .001), but the effect of Group 
(F (2,25) = 2.240, p = .127) was not significant. Overall, all rats were able to 
reach the visible platform in a similar manner, regardless of their age. The 
latency, path length, and speed of swimming to a visible platform across trials 
are presented in Figure 9. 
8.4 Discussion 
Our results suggest that all animals, regardless of age, were able to learn the 
platform location. The acquisition of spatial information was more similar for the 
young and middle-aged rats. The aged rats were able to learn about the 
platform location to an equal extent by the end of the day on which the platform 
appeared in a location not previously encountered. 
The deficit appeared in 24-hour retention of this newly acquired spatial 
information for not only the aged but also the middle-aged rats. Old and middle-
aged rats had much longer latencies and path lengths to reach the platform on 
the first trial of the day on which the platform remained in the same location as 
the previous day compared to their performance on the last trial of the previous 
day. By the end of the day where platform remained in a location in which it 
appeared the previous day, however, all rats were again performing similarly. 
85 
A similar pattern of behavior was observed for both measures of place learning, 
latency (Figure 7A & B) and path length (Figure 7C & D). Latency is thought to 
be dependent on the speed of swimming, which most believe changes over the 
lifespan of a rat, and hence is considered a biased measure of learning and 
memory in old age. Path length, on the other hand, is a measure independent of 
swim speed, and hence considered unbiased. We observe similar performance 
with both measures. Also, while not always swimming at the same speed as 
their younger counterparts, the old rats are able to do so on most trials (Figure 
7 E & F). Together, our results suggest that the memory deficit observed in 24-
hour retention of newly learned spatial information is not biased by differences 
in swim speed. 
In addition, we allowed all rats to swim to a visible platform for a day (Figure 9). 
The old rats seemed to have higher latencies at first, but their performance 
shortly became almost identical to that of middle-aged and young rats. The 
speed of swimming was more variable than expected, but again the speed was 
similar for all three groups on certain trials suggesting that the old rats can 
swim at the same speed as the young rats. As far as path length to locate the 
platform is considered, all three groups performed similarly. Again, our visible 
platform results suggest that the observed hippocampus-dependent deficit in 
retention of newly learned spatial information in relation to age was independent 
of sensori-motor or motivational biases. 
86 
We have established that the age-related deficit in spatial cognition associated 
specifically with the hippocampus is present in our rats. Our findings of 
dysfunctional memory for spatial information add to the extensive literature 
reporting age-related deficits on spatial tests of learning and memory (see 
Geinisman et al., 1995; van der Staay, 2002 for a review). What is more 
interesting, however, is the fact that the pattern of age-related memory deficits 
for spatial information in the rat is very similar to that reported for human place 
learning in the virtual Morris water task (Driscoll et al., 2005), in that the 
deficits are present in both old and middle-aged group. The next question we 
ask is whether the age-related deficit in learning and memory is present for 
other types of hippocampus-dependent information processing, or is it specific 
to spatial information? 
87 
9. Experiment 4 - The Transverse Patterning Problem 
Does the transverse patterning discrimination problem, as opposed to 
elemental discriminations, depend on the hippocampus? 
9.1 Background 
Many theorists have suggested that the pattern of spared and impaired 
performance associated with damage to the hippocampus can be explained by 
distinguishing between a configural/conjunctive memory system and an 
elemental or habit system (Hirsh, 1974; Mishkin et al., 1984; O'Reilly & Rudy, 
2001; Sutherland & Rudy, 1989; Tyler & DiScenna, 1986; Wicklegren, 1979). 
Central to these theories is that the hippocampus binds together the 
independent features of an episode into a unique representation called a 
configural or conjunctive representation. 
Based on their configural association theory of the hippocampal function, 
Sutherland and Rudy (1989) predicted that non-linear discrimination problems 
should be especially dependent on the hippocampus. Problems in this category 
do not permit solutions based on the associative strengths of the individual 
elements. One such problem is called transverse patterning (Alvarado & Rudy, 
88 
1992, 1995a,b; Spence, 1952). The transverse problem involves three problems 
comprised of only three stimuli, so that each stimulus participates in two out of 
the three problems comprising this task and each stimulus is equally 
ambiguous; A+B-, B+C-, C+A-. The transverse patterning problem requires 
attending to the relationships between the stimuli and is said to require the 
hippocampus, as learning about each stimulus separately will not lead to a 
successful solution to this problem. In contrast, elemental discriminations, while 
similarly involving three problems, are comprised of six unambiguous stimuli; 
A+B-, C+D-; E+F-. Successful solution to this problem can be obtained by 
simply learning that certain stimuli are either always rewarded or non-rewarded. 
Elemental discriminations can be solved without the intact hippocampus. 
There have been several reports of anterograde amnesia for the transverse 
patterning problem after the damage to the hippocampus (Alvarado & Rudy, 
1995a,b; Dusek & Eichenbaum, 1998; but see Bussey et al., 1998). For 
example, Alvarado and Rudy (1995a) found that rats with damage to the 
hippocampus could not solve the transverse patterning problem, but they readily 
learned to solve a control set of three elemental discrimination problems. Such 
results are consistent with the idea that the hippocampus makes a selective 
contribution to problems that require configural/conjunctive solutions. The 
finding that damage did not impair learning the set of elemental discriminations 
suggests that there is a system that does not depend on the hippocampus that 
can support a solution to these problems. 
89 
Even though rats with damage to the hippocampus are not impaired in learning 
elemental discriminations, these data do not speak to a fundamental question; 
does the hippocampus contribute to the learning of these elemental problems in 
the normal, intact subject? The answer to this question is important because 
behavioral tests following anterograde damage to the hippocampus only reveal 
the residual capacities afforded by spared, extra-hippocampal circuitry. They do 
not necessarily reveal the contribution that the hippocampus makes when it is 
intact in a normal animal. Thus, the fact that rats with damage to the 
hippocampus are not impaired on the set of elemental discrimination problems 
does not mean that the hippocampus makes no contribution to solving these 
problems in the intact animal. 
We are interested in using the transverse patterning discrimination task as a 
non-spatial test of hippocampal integrity in aged rats. Before we do so, we 
decided to test definitively whether this task, as implemented in our laboratory 
engages the hippocampus or not. The purpose of the present study is twofold. 
First, we attempt to resolve the discrepancy between the studies reporting a 
spared performance on the transverse patterning problem after hippocampal 
damage and those reporting impairments. The secondary goal is to determine if 
the hippocampus contributes at all to learning elemental discriminations. To 
implement the experimental design, we used a water escape task (Prusky et al., 
90 
2000a) to control the presentation of stimuli in the visual discrimination 
problems. 
9.2 Materials and Methods 
Animals. Subjects were 8 female Brown Norway x Fisher 344 (BNxF344) hybrid 
rats (Harlan, IN), 12-14 months of age, 250-325 g. Four rats were assigned to 
the elemental discrimination condition, and four rats were assigned to the 
transverse patterning condition. One rat in the elemental discrimination group 
died due to respiratory infection and did not complete the last two conditions of 
training (the 10-day delay test after relearning the elemental discriminations 
and learning a new set of problems post-surgery). The rats were group-housed 
in hanging plastic cages in a room with an ambient temperature of 21°C, 35% 
relative humidity, 12/12 light/dark cycle, and had access to food and water ad 
libitum. All experiments were carried out in accordance with the guidelines 
provided by the Canadian Council on Animal Care. 
Surgery. The hippocampal formation was damaged by bilateral injections of 
NMDA (10 mg/ml) dissolved in phosphate-buffered saline. Rats were 
anesthetized by isofluorane inhalation (4% with 2 l/minute of oxygen and 2% 
after a surgical plane was established). A midline incision was made in the scalp 
and periosteum. Rats received 5 injections in each hippocampus through a 30-
gauge cannulae; at the rate of 0.1 nJ/min using an infusion pump, a 0.4 ^,1 of 
91 
excitotoxin at each site (3.1, 4.1, 5.0, 5.3, and 6.0 posterior to bregma; 1.5, 
3.0, 3.0, 5.2, and 5.0 mm lateral to bregma; and 3.6,4.0, 4.0, 7.3, and 7.3 mm 
ventral to the surface of the skull) in respective order. Cannulae were left in 
place for 4 min after each injection. The scalp was sutured and the animal was 
returned to its home cage to recover. Recovery was monitored for 90 minutes 
before the animal was returned to the colony room. Animals were administered 
clordiazepoxide (1 mg/kg, i.p.) 20 minutes before surgery, and then on the first 
sign of wakefulness after surgery for 3 post-surgery injections total, 
approximately 30 min apart to suppress any seizure activity that may be 
associated with NMDA injections. Animals were allowed to recover for 10 days 
before the testing began. 
Histology. All animals were sacrificed by overdose of pentobarbital (100 mg/kg 
i.p.) and perfused transcardially with 0.9% saline followed by 4% 
paraformaldehyde. The brains were removed and 40 |u,m coronal sections were 
cut on a cryostat microtome. Every fifth slice was mounted on glass slides and 
stained with cresyl violet. The stained sections were examined under a 
microscope to quantify damage. 
Discrimination Training Procedures 
Visual Water Task. The Visual Water Task (VWT) was used to train rats on the 
discrimination problems. Prusky et al. (2000) previously described this task in 
detail. Briefly, a trapezoidal metal tank (140 cm L x 80 cm W x 25 cm W and 55 
92 
cm H) was filled with water to a depth of 15 cm. The end wall of the tank was 
transparent. A midline divider (40 cm H) was placed in the tank and extended 
46 cm into the pool from the transparent wall. This divider separated two 
computer monitors (17" VGA; Viewsonic E70F) that were positioned outside the 
transparent wall of the pool and used to present the stimuli. A transparent 
Plexiglas platform (37 cm L x 13 cm W x 14 cm H) was submerged at the end of 
one of the arms, and always paired with the correct ( + ) stimulus. A computer 
program (Vista; http://www.cerebralmechanics.com) was used to control the 
right/left randomization pattern, present discriminanda, record behavioral 
responses, and provide parameter control. The discriminanda are shown in 
Figure 10. 
Overview. Rats were shaped to associate swimming to the platform with 
escape from the water. The platform was located directly below the monitor 
displaying the rewarded ( + ) stimulus. At the beginning of each trial, animals 
were released into the pool facing the transparent wall, and allowed to swim to 
one of the stimuli. If choice was correct, rat was allowed to climb the platform 
and remain there for a few seconds before being returned to the holding cage. If 
incorrect, the rat was trapped in that arm of the pool for 10 s and then allowed 
to swim to the arm containing the platform before being removed from the pool. 
Trapping was not instituted on the test day. The choice was correct if animals 
swam to the platform without entering the arm displaying the non-rewarded 
stimulus. The rat was judged as entering the incorrect arm if the hind legs 
crossed into this arm of the pool, even if the rat had turned around before 
93 
reaching the platform. The pattern of the rewarded stimulus/platform location 
was a left (L), right (R), LLRLRR sequence. The dependent variable was 
percentage of correct choices. The intertrial interval was approximately three 
minutes. 
The overall design of the experiment is presented in Table 1. All rats were 
initially trained on their respective set of discrimination problems, as just 
described. Following the test day, each rat's hippocampus was damaged 
bilaterally and 10 day later they were retested. Each rat was then retrained and 
tested using the same procedures described above. The rats trained on the 
elemental problem set were then trained on a new set of elemental 
discrimination problems (Figure 10, Set 2: G+H-, I+J-, K+L-). 
Elemental Discriminations. Rats were trained to solve three visual 
discriminations (Figure 10, Set 1: A+B-, C + D-, E + F-) using a stepwise 
approach. Training consisted of nine days in three phases (3 days for each 
phase,), a test on day 10, followed by a novel transfer testing on Day 11. Phase 
1 consisted of 3 days of training on Problem 1, 30 trials each. In Phase 2, 
Problem 2 was introduced and intermixed with Problem 1. Days 4 and 5 
consisted of 5 trials of Problem 1, followed by 20 trials of Problem 2, and again 
by 5 trials of Problem 1. On Day 6 Problem 1 and 2 alternated every 5 trials for 
a total of 30 trials, starting with Problem 1. In Phase 3, Problem 3 was 
introduced and intermixed with Problems 1 and 2. On Day 7 rats received 5 
trials of Problem 1, 20 trials of Problem 3, and 5 trials of Problem 2. Day 8 
94 
consisted of 5 trials of Problem 2, 20 trials of Problem 3, and 5 trials of Problem 
1. Day 9 consisted of alternating blocks of 5 trials of each problem (1, 2, and 3). 
The test day, Day 10, consisted of randomly alternating problems on each trial 
(3, 2, 1, 2, 1, 3, 2, 3, 1, 3, 1, 2, 1, 2, 3...), for a total of 30 trials, 10 of each 
problem. 
Transverse Patterning. Transverse patterning (Figure 10, Set 3: X+Y-; Y+Z-; 
Z+X-) training consisted of 7 days, followed by a test on Day 8, 30 trials per 
day. Phase 1 consisted of 1 day of training on Problem 1 (X+Y-). In Phase 2, 
Problem 2 (Y+Z-) was introduced and intermixed with Problem 1 for 1 day, and 
consisted of 5 trials of Problem 1, followed by 20 trials of Problem 2, and again 
by 5 trials of Problem 1. On Day 3 Problem 1 and 2 alternated every 5 trials for 
a total of 30 trials, starting with Problem 1. In Phase 3, Problem 3 (Z+X-) was 
introduced and intermixed with Problems 1 and 2. On Day 4 rats received 5 
trials of Problem 1, 20 trials of Problem 3, and 5 trials of Problem 2. Day 5 
consisted of 5 trials of Problem 2, 20 trials of Problem 3, and 5 trials of Problem 
1. Day 6 consisted of alternating blocks of 5 trials of each problem (1, 2, and 3). 
The test day, Day 8, consisted of randomly alternating problems on each trial 
(3, 2, 1, 2, 1, 3, 2, 3, 1, 3, 1, 2, 1, 2, 3...), for a total of 30 trials, 10 of each 
problem. 
95 
9.3 Results 
Histology. The damage to the hippocampal formation was extensive bilaterally 
(Figure 11). On average, approximately 70% of the hippocampus was removed. 
The smallest lesion encompassed 50% of the hippocampal formation and the 
largest lesion involved 95% of the hippocampus. Three out of seven rats had 
50%-60%of the hippocampus damaged, while the remaining four rats had lesion 
that encompassed 85% or more of the hippocampus. Sparing of the major 
hippocampal subfields was limited almost exclusively to the postero-ventral 
regions. The damage did not extend to the thalamus, subicular region, or rhinal 
cortices. 
Elemental Discriminations. We performed a repeated measures analysis of 
variance with Problem (3) and Test Time (5 levels: pre-lesion, post-lesion re-
testing, post-lesion re-training, post-lesion re-training with a 10-day delay, and 
the new discrimination learning) as within-subjects variables, and performance 
(% correct) on each problem as a dependent variable. There was a significant 
Problem X Test Time interaction (F (8, 16) = 8.221, p < .001). By the 
completion of training rats were performing at a near perfect level on all three 
problems. Damage to the hippocampus, however, significantly impaired 
performance on these problems. Nevertheless, these rats were able to relearn 
the same set of problems and displayed perfect retention 10 days later (to 
mimic the recovery period after surgery). They were also able to learn a new set 
96 
of elemental discrimination problems. Figure 12 (A & B) presents the results for 
the various testing phases of the elemental discrimination paradigm. 
In order to assess whether the rats learned the problems at a similar rate before 
and after the damage, we performed a repeated measures ANOVA with Problem 
(3) and Learning Condition (3 levels: pre-lesion, post-lesion re-training, post-
lesion new problem) as within-subjects variables, and trials to criterion (80% 
correct, 8 out of ten consecutive trials) as a dependent variable. There was a 
significant Learning Condition X Problem interaction (F (4, 8) = 5.718, p = 
.018). Rats acquired each of the three problems at a faster rate after the lesion 
regardless of whether the discrimination set was new or encountered prior to 
lesion (Figure 13), suggesting that the lesion impaired the memory for the 
stimuli but not the procedural requirements of the task itself. 
Transverse Patterning. A repeated measures ANOVA was performed with 
Problem (3) and Testing (3 levels: pre-lesion, post-lesion test, post-lesion re 
training) as within-subjects variable, and performance (% correct) as a 
dependent variable. There was a significant Learning Condition X Problem 
interaction (F (4, 12) = 3.376, p = .045), where intact rats were able to learn 
the transverse patterning but were unable to remember or relearn the problems 
after the damage to the hippocampus. Figure 14 presents the results for the 
transverse patterning testing phases before and after the hippocampus was 
damaged. Performance on the test day prior to surgery was well above chance 
on all three problems. Damage to the hippocampus, however, significantly 
97 
impaired performance on each problem learned prior to surgery, and the rats 
were unable to re-learn them. 
9.4 Discussion 
Damage to the hippocampus produced severe retrograde amnesia for both the 
transverse patterning and the elemental problems. Equally important, however, 
was that rats with damage to the hippocampus displayed perfect retention 
through a lengthy interval following retraining on the elemental problems. Thus, 
the amnesia they displayed when the hippocampus was damaged following 
training was not displayed if the problems were learned or relearned in the 
absence of the hipppocampus. Our results also replicate those of Alvarado & 
Rudy (1995a,b) that rats with damage to the hippocampus are impaired in 
learning the transverse patterning problem, but can learn a set of elemental 
problems. 
That the damage to the hippocampus severely impaired performance on the 
transverse patterning problem in both retrograde and anterograde direction is 
consistent with the theoretical claim that the hippocampus makes a critical 
contribution to the acquisition of configural representations (Sutherland & Rudy, 
1989). Our results also suggest, however, that the need for configural solution is 
not a requirement for a problem to engage the hippocampus. The finding that 
the damage to the hippocampus produced severe retrograde amnesia for the 
98 
elemental problems implies that when the hippocampus is intact it also makes 
an essential contribution to the acquisition of discrimination problems that can 
be solved simply on the basis of the associative strengths of the individual 
elements. We are not alone in making this observation. There have been a 
number of reports, that hippocampal damage can produce retrograde amnesia 
for non-configural information even though the same information is spared in 
anterograde direction (Ross et al., 1984; Sara, 1981; Sutherland et al., 2001; 
Weisend et al., 1996). 
How does the hippocampus contribute to the acquisition of elemental 
discrimination problems? To approach this question, it should be emphasized 
that rats can acquire and retain the solution to elemental discriminations 
problems when the hippocampus is damaged prior to training (Alvarado & Rudy, 
1995a, b). Thus, as has often been proposed (Hirsh, 1974; Mishkin et al., 1984; 
Sutherland & Rudy, 1989), there is an elemental association system that can 
solve problems simply on the basis of the associative strengths of the elements 
comprising the problems and that does not depend on the hippocampus. Once 
this is accepted, it then must be the case that if the hippocampus is functioning 
during training, the independent contribution of the elemental system is 
significantly compromised. If this is not the case then there should be no 
retrograde amnesia because without a hippocampus rats learn and retain the 
solution to these problems. This implies that (a) there is competition between 
the hippocampal-dependent memory system and the system that does not 
depend on the hippocampus and (b) that hippocampal system dominates. 
99 
Hence, one contribution of the hippocampus, in addition to its role in storing 
specific memories, is to reduce the contribution of the elemental system. 
It must also be the case that the hippocampus supports a memory 
representation that is at least partly independent of the elemental system. We 
noted that a common assumption is that the hippocampus binds together 
representations of co-occurring events (e.g., Fanselow, 1999, 2000; O'Reilly & 
Rudy, 2001, Rudy & O'Reilly, 2001; Squire, 1992; Sutherland & Rudy, 1989; 
Tyler & DiScenna, 1986). There are at least two ways this mechanism could 
contribute to solving elemental discriminations. The hippocampus could store 
conjunctive representations of the individual trial outcomes (e.g. A+ or B-) 
together with contextual elements that can be used to guide choice behavior. 
Alternatively, one might assume that when confronted with a set of elemental 
discrimination problems the intact rat uses the same configural representations 
it would be forced to use to solve the transverse patterning problem. Alvarado 
and Rudy (1992), however, reported data that rule out this possibility. 
The general conclusions from this analysis are (a) transverse patterning requires 
the hippocampus to be intact for successful performance, (b) elemental 
discrimination problems can be solved by a hippocampal-dependent system and 
a system that does not depend on the hippocampus, and (c) the hippocampal-
dependent system suppresses or inhibits the extra-hippocampal elemental 
association system, in such a way that the latter does not develop an 
independent memory representation sufficient to solve the tasks. 
100 
These theoretical conclusions are remarkably consistent to those derived from 
recent studies of contextual fear conditioning. If placed into a particular context 
and shocked rats subsequently display fear of that place or context by freezing. 
The literature indicates that neurotoxic damage to dorsal hippocampus before 
conditioning has little or no effect on the acquisition of contextual fear 
conditioning (Cho et al., 1999; Frankland et al., 1998; Maren et al., 1997; 
Richmond et al., 1999) but damage to the hippocampus following learning 
produces severe retrograde amnesia (Anagnostaras et al., 1999; Maren et al., 
1997). This dissociation between anterograde and retrograde effects clearly 
parallels our current conclusions about elemental visual discriminations. 
Several theorists (Fanselow, 1999; Maren, 2001; Rudy et al., 2004; Rudy and 
O'Reilly, 2001) have proposed that that contextual fear conditioning can be 
supported by a hippocampus-dependent memory system that supports a 
configural/conjunctive representation of context and an elemental system that 
supports conditioning to the independent features that make up the context. 
This idea predicts that the damage to the hippocampus should not impair 
contextual fear in the anterograde direction because the residual elemental 
system can support conditioning. To explain why hippocampal damage 
produces severe retrograde amnesia, Maren and colleagues (1997) made the 
additional proposal that there is (a) competition for conditioning between the 
hippocampal conjunctive representation and the feature representations for 
association with shock, and (b) the conjunctive representation dominates. Thus, 
101 
when the hippocampus is damaged after conditioning the conjunctive 
representation is no longer available to support fear conditioning and, because 
conditioning to features was weak, the animal shows amnesia. 
It is tempting to conclude that results from two markedly different paradigms, 
appetitive discrimination learning and contextual fear conditioning, have 
revealed a common set of principles: (a) in both cases the learning experience 
potentially can be represented in two systems, one which depends of the 
hippocampus and one that does not, and (b) in the intact animal the 
hippocampal system dominates. Rudy et al. (2004), in fact, have recently 
suggested that these two functions represent two independent contributions of 
the hippocampus. It stores conjunctive representations and inhibits or interferes 
with the learning that can be supported by the non-hippocampal system. It is 
not known from the present data sets how general this interference is across 
types of learning or whether the inhibitory/interference process is limited to the 
time of encoding/storage of the memory. 
In summary, we asked if the hippocampus makes a contribution to solving a set 
of elemental discriminations that could be solved simply on the basis of the 
associative strengths of the individual elements. Our results indicate that the 
damage to the hippocampus produced severe retrograde amnesia in rats that 
had solved these problems with hippocampus intact. We propose that this result 
indicates that the hippocampus provides an independent representation of the 
solution and inhibits learning that could be supported by a system that does not 
102 
depend on the hippocampus. This conclusion converges with recent theoretical 
views on the contribution the hippocampus makes to contextual fear 
conditioning and is consistent with a general process whereby the hippocampus 
inhibits memory processes in other learning systems. 
In addition, we asked if the hippocampus makes a contribution to solving a set 
of transverse patterning discriminations that require configural solution. Our 
results indicate that the damage to the hippocampus produced severe 
retrograde and anterograde amnesia for this set of problems. Hence, the 
transverse patterning task will serve well as a test of hippocampus-dependent 
non-spatial learning and memory. The next question we ask is, whether 
transverse patterning performance is affected by age. 
103 
10. Experiment 5 - Non-spatial Learning and Memory 
Does old age have an effect on learning and memory of non-spatial 
hippocampus-dependent information? 
10.1 Background 
Now that we have resolved that our version of the transverse patterning 
problem tests hippocampus-dependent learning and memory, we proceed by 
employing this problem as a test of non-spatial hippocampus-dependent 
mnemonic function in aging. Previously, this problem has been successfully used 
to test age-related deficits in hippocampus-dependent cognition in humans 
(Driscoll et al., 2003a). Conversely, it has never been used to test hippocampus-
related learning and memory in the aged rat. If we presume that preserved 
integrity of the hippocampus is necessary for successful performance on this 
task, and our findings from Experiment 4 suggest that it is, and given that the 
reports of compromised hippocampal integrity with old age are numerous 
(Geinisman et al., 1995), then aged rat should be impaired in learning this 
problem. The performance on simple, two-choice discriminations with no 
overlapping components should be spared, given that such learning can proceed 
normally in the absence of the hippocampus. 
104 
10.2 Materials and Methods 
Animals. Subjects were 29 female Brown Norway x Fisher 344 (BNxF344) 
hybrid rats (NIA/Harlan). At the start of the experiment, 10 rats were 3 months 
old, 10 were 12 months old, and 9 were 24 months old, weighing between 250-
425 g. One old rat died during the MR scanning procedures (which were 
performed prior to behavioral testing) due to pre-existing stroke. These rats are 
the same ones participating in Experiment 3, in which we assessed the spatial 
hippocampus-dependent performance of rats. The rats were group-housed in 
hanging plastic cages in a room with an ambient temperature of 21°C, 35% 
relative humidity, 12/12 light/dark cycle, and had access to food and water ad 
libitum. All animals were cared for and tested at The University of Lethbridge 
vivarium according the National Institutes of Health guidelines and under the 
approval of the institutional animal care committees at the University of 
Lethbridge, University of Calgary, and the National Research Council which only 
approves studies that are in accordance with the Canadian Council on Animal 
Care guidelines. 
Visual Water Task. The Visual Water Task (VWT) was used to train rats on the 
discrimination problems. Prusky et al. (2000) previously described this task in 
detail. Briefly, a trapezoidal metal tank (140 cm L x 80 cm W x 25 cm W and 55 
cm H) was filled with water to a depth of 15 cm. The end wall of the tank was 
transparent. A midline divider (40 cm H) was placed in the tank and extended 
105 
46 cm into the pool from the transparent wall. This divider separated two 
computer monitors (17" VGA; Viewsonic E70F) that were positioned outside the 
transparent wall of the pool and used to present the stimuli. A transparent 
Plexiglas platform (37 cm L x 13 cm W x 14 cm H) was submerged at the end of 
one of the arms, and always paired with the correct (+) stimulus. A computer 
program (Vista; http://www.cerebralmechanics.com) was used to control the 
right/left randomization pattern, record behavioral responses, and provide 
parameter control. 
Discrimination Training Procedures 
Overview. Rats were shaped to associate swimming to the platform with 
escape from the water. The platform was located directly below the monitor 
displaying the rewarded ( + ) stimulus. At the beginning of each trial, animals 
were released into the pool facing the transparent wall, and allowed to swim to 
one of the stimuli. If choice was correct, rat was allowed to climb the platform 
and remain there for a few seconds before being returned to the holding cage. If 
incorrect, the rat was trapped in that arm of the pool for 10 s and then allowed 
to swim to the arm containing the platform before being removed from the pool. 
Trapping was not instituted on the test day. The choice was correct if animals 
swam to the platform without entering the arm displaying the non-rewarded 
stimulus. The rat was judged as entering the incorrect arm if the hind legs 
crossed into this arm of the pool, even if the rat had turned around before 
reaching the platform. The pattern of the rewarded stimulus/platform location 
106 
was a left (L), right (R), LLRLRR sequence. The dependent variable was 
percentage of correct choices. The intertrial interval was approximately 3 
minutes. 
The overall design of the experiment is presented in Table 1. All rats were 
initially trained on a set of elemental discrimination problems (Figure 10, Set 1) 
following by the transverse patterning problem (Figure 10, Set 3). Each rat was 
trained and tested using the same procedures described above. 
Elemental Discriminations. Rats were trained to solve three visual 
discriminations (Figure 10, Set 1: A+B-, C+D-, E + F-) using a stepwise 
approach. Training consisted of nine days in three phases (3 days for each 
phase,), a test on day 10, followed by a novel transfer testing on Day 11. Phase 
1 consisted of 3 days of training on Problem 1, 30 trials each. In Phase 2, 
Problem 2 was introduced and intermixed with Problem 1. Days 4 and 5 
consisted of 5 trials of Problem 1, followed by 20 trials of Problem 2, and again 
by 5 trials of Problem 1. On Day 6 Problem 1 and 2 alternated every 5 trials for 
a total of 30 trials, starting with Problem 1. In Phase 3, Problem 3 was 
introduced and intermixed with Problems 1 and 2. On Day 7 rats received 5 
trials of Problem 1, 20 trials of Problem 3, and 5 trials of Problem 2. Day 8 
consisted of 5 trials of Problem 2, 20 trials of Problem 3, and 5 trials of Problem 
1. Day 9 consisted of alternating blocks of 5 trials of each problem (1, 2, and 3). 
The test day, Day 10, consisted of randomly alternating problems on each trial 
107 
(3, 2, 1, 2, 1, 3, 2, 3, 1, 3, 1, 2, 1, 2, 3...), for a total of 30 trials, 10 of each 
problem. 
Transverse Patterning. Transverse patterning (Figure 10, Set 3: X+Y-; Y+Z-; 
Z+X-) training consisted of 7 days, followed by a test on Day 8, 30 trials per 
day. Phase 1 consisted of 1 day of training on Problem 1 (X+Y-). In Phase 2, 
Problem 2 (Y+Z-) was introduced and intermixed with Problem 1 for 1 day, and 
consisted of 5 trials of Problem 1, followed by 20 trials of Problem 2, and again 
by 5 trials of Problem 1. On Day 3 Problem 1 and 2 alternated every 5 trials for 
a total of 30 trials, starting with Problem 1. In Phase 3, Problem 3 (Z+X-) was 
introduced and intermixed with Problems 1 and 2. On Day 4 rats received 5 
trials of Problem 1, 20 trials of Problem 3, and 5 trials of Problem 2. Day 5 
consisted of 5 trials of Problem 2, 20 trials of Problem 3, and 5 trials of Problem 
1. Day 6 consisted of alternating blocks of 5 trials of each problem (1, 2, and 3). 
The test day, Day 8, consisted of randomly alternating problems on each trial 
(3, 2, 1, 2, 1, 3, 2, 3, 1, 3, 1, 2, 1, 2, 3...), for a total of 30 trials, 10 of each 
problem. 
10.3 Results 
Elemental Discriminations. Figure 15 presents the results for each problem of 
the elemental discrimination paradigm on the test day by group. By the 
completion of training rats were performing at a near-perfect level on all three 
108 
problems. Age had no effect on performance, all rats were able to learn a set of 
three elemental discrimination problems. We performed a repeated measures 
ANOVA with Problem (3) as a within-subjects variable, age Group (3 levels: 
Young, Middle-aged, Old) as a between-subjects variable, and performance (% 
correct) on each problem as a dependent variable respectively. There was a 
significant effect of Problem (F (2,50) = 7.545, p = .001). The Group X Problem 
interaction was not significant (F (4,50) = .514, p > .7), and there was no 
significant difference between the age groups (F (2,25) = .844, p > .4). Post 
hoc tests revealed that the performance of each group did not significantly differ 
from one another regardless of the problem at hand (p's > .1). 
Transverse Patterning. Figure 16 presents the results of the transverse 
patterning paradigm . A repeated measures ANOVA was performed with Problem 
(3) as a within-subjects variable, age Group (3 levels: Young, Middle-aged, Old) 
as a between-subjects variable, and performance (% correct) on each problem 
as a dependent variable respectively. There was a significant Group X Problem 
interaction (F(4,50) = 6.285, p < .001). Both young and middle-aged rats were 
able to learn the transverse patterning problem, but not the old rats. The effect 
of Problem was significant as well (F(2,50) = 17.156, p < .001). The difference 
in performance between the age groups was also significant (F (2,25) = 
106.677, p < .001). Post hoc tests revealed that for each of the three transverse 
patterning problems, the performance of the Young and Middle-aged groups did 
not significantly differ (p's > .6). On the other hand, the Old group performed 
109 
significantly worse from both the Young and the Middle-aged group on each of 
the three problems (p < .001). 
10.4 Discussion 
The results of the present experiment support our prediction, that old age would 
significantly and selectively affect performance on the transverse patterning 
problem. When tested simultaneously on all three problems that comprise the 
transverse patterning, old rats had a significant deficit in performance while 
young and middle-aged rats had little trouble learning the problems. In contrast, 
all age groups performed similarly when tested concurrently on the three 
problems comprising elemental discriminations. Interestingly, this is the exact 
pattern of age-related performance deficits on the transverse patterning task 
that we have previously observed in humans (Driscoll et al., 2004a); no 
problems in learning the task for the young and the middle-aged participants 
whereas elderly participants were significantly impaired. 
We tested the visual acuity of young and old FBNF1 rats in Experiment 1, the 
results of which suggest that the performance deficit on visual discriminations is 
not biased by poor vision in old age. This assumption is further supported by the 
fact that the same rats are able to discriminate between concurrent elemental 
discriminations, which can be solved in the absence of the hippocampus. In 
addition, successful performance by all rats on elemental discriminations 
110 
suggests the absence of a potential bias due to potential motor or motivational 
problems that conceivably accompany aging. 
These data, coupled with impaired retention of spatial information, point to a 
broad age-related deficit in hippocampus-dependent learning and memory that 
is not specific to the type of information or the test format employed. The same 
findings can also be used to support certain properties of hippocampal function 
in general. For example, Rudy and Sutherland (1995) posit that increasing the 
ambiguity between the stimuli should increase the reliance on 
configural/conjunctive associations to solve the problem, and in turn the 
hippocampus is the structure that allows these associations to disambiguate 
stimulus compounds. Our current results are in agreement with their position. 
Hence, the study of hippocampus-dependent learning and memory in relation to 
aging will not only help define the age-related hippocampal mnemonic 
dysfunction, but should also enhance our understanding of the contributions that 
the hippocampus makes to learning and memory. In addition, we've established 
two different hippocampus-dependent tasks, one spatial and the other non-
spatial in nature, that we can employ to test hippocampus-dependent learning 
and memory in rats and humans alike. 
Now that we've established that hippocampus-dependent learning and memory 
deficits are present in aged FBNF1 rats, we proceed with assessment of 
hippocampal integrity on several different levels. First is the structural and 
111 
neurochemical assessment of hippocampal integrity in vivo in the same rats that 
we characterized behaviorally in the current study and in Experiment 3. 
112 
11. Experiment 6 - Magnetic Resonance Imaging and Spectroscopy 
Are the age-related alterations in hippocampal structural or functional 
integrity observable in vivo? 
11.1 Background 
An in vivo magnetic resonance imaging (MRI) study employing a high field 
magnet was undertaken to help develop a rat model of aging and to evaluate 
both MRI-based volumetric measurements and metabolic profile in the rat in 
vivo. Volumetric estimations based on structural MR images are commonly used 
in human research in a wide variety of conditions. In animal research anatomical 
assessments are, on the other hand, customarily conducted post mortem. MR 
imaging, however, has an advantage over post mortem analyses. For example, 
MR allows not only the assessment of structural integrity in vivo, but allows us 
to study the changes longitudinally. Indeed, in vivo MRI has been employed to 
identify structural degradation in several experimental rodent models (Assaf et 
al., 1997; Baba et al., 1990; Bouilleret et al., 2000; Norman et al., 1991; 
Saunders et al., 1989; Wolf et al., 2002). In addition, some studies even 
reported an agreement between their MRI and histological observations 
(Allegrini and Sauer, 1992; Ben Horin et al., 1996; Minematsu et al., 1992). 
113 
To the best of our knowledge, no volumetric studies of rat brains using MRI exist 
in relation to aging, regardless of the fact that such measurements are routinely 
performed in studies of human aging. The present study has two of goals in 
relation to MRI: first, to develop a reliable rat model of normal human aging, 
and to measure hippocampal volumes in these animals in vivo using MRI 
images. 
Moreover, proton magnetic resonance spectroscopy MRS) of the brain 
provides a noninvasive way of measuring metabolic information from amino 
acids, amines, sugars, and other compounds involved in high-energy 
metabolism. In recent years neurospectroscopy of the human brain has shed 
new light on metabolic mechanisms operating during normal brain development. 
N-acetylaspartate (NAA) resonance contributes to the most prominent signal in 
:H MRS. 
NAA is almost exclusively present in the central nervous system and is 
predominantly located in pyramidal neurons, oligodendrocyte type 2 astrocyte 
progenitor cells, and in mature and immature oligodendrocytes. Given that NAA 
has been histologically localized primarily to neural tissue (Birken and Olendorf, 
1989; Moffet et al., 1991), NAA levels are often thought to either reflect 
neuronal atrophy, neuronal density, or signify viable brain cells. Given that this 
substance is synthesized in the mitochondria from aspartate and acetyl-CoA 
(Benuch and D'Adamo, 1968), and metabolized by aspartoacylase into aspartate 
and acetate in the cytosol (D'Adamo et al., 1973), recently it has been 
114 
suggested that NAA should be viewed as a marker of neuronal metabolic 
function. 
In recent years, there are reports of decrease of NAA levels in the absence of 
neuronal loss (Jenkins et al., 2000). In addition, a reversible depletion of NAA 
levels either due to treatment (Kalra et al., 1998) or spontaneously (De Stefano 
et al., 1995), suggesting that even though a decrease in NAA is observed in 
these conditions at baseline - it could not simply reflect neuronal loss, given the 
turnaround. Instead, a depression in metabolic function has been offered as an 
alternate explanation. Moreover, Canavan's disease, an autosomal recessive 
leukodystrophy due to abnormal aspartoacylas clinically characterized by 
hypotonia, head lag, and macrocephaly without hydrocephalus and 
biochemically characterized by aspartoacylase deficiency, is marked by a large 
elevation in NAA levels, an elevation that clearly does not reflect an increase in 
neuronal density. Together, these findings indicate that both high and low levels 
of NAA can denote neuronal dysfunction. 
The shift has been made from interpreting NAA as a putative neuronal marker to 
a marker for normally functioning neurons. Alternatively, because it has been 
demonstrated that mitochondrial NAA synthesis is affected by mitochondrial 
toxins (Bates et al., 1996), NAA depletion could more simply reflect a deficit in 
the mitochondrial oxidative phosphorylation processes. Dautry and colleagues 
(2000) demonstrated that NAA depletion, at least in the early stages, occurs in 
115 
the absence of cell death suggesting it to be a marker of neuronal dysfunction. 
Needless to say, the function of NAA remains a subject of considerable debate. 
Decreases in hippocampal NAA have been reported repeatedly in aged humans 
(e.g. Driscoll et al., 2003a; Schuff et al., 1999), however, it has not been 
investigated thus far in aged rodents. In order to evaluate the neurochemical 
profile of namely NAA, Choline, and Creatine, we performed 1H MRS in both right 
and left hippocampus of young, middle-aged, and old FBNF1 rats. In addition, 
post mortem assessment of the same tissue will be described in Experiment 7, 
and may shed additional light on the function of NAA in general. 
11.2 Materials and Methods 
Animals. The same animals that participated in Experiments 3 and 5 (i.e. were 
assessed on a spatial and non-spatial hippocmapus-dependent learning and 
memory tasks) are involved in the imaging studies as well. Rats were divided 
into 3 groups: Young (A/ = 10; age 3 months at the start of the experiment), 
Middle-aged (A/ = 10; age 14 months at the start of the experiment), and Old 
(A/ = 10; age 24 months at the start of the experiment) group. All animals were 
housed in pairs in hanging plastic cages, in a room with an ambient temperature 
of 21° C and 35% relative humidity, 12/12/ light dark cycle. Food and water 
were available ad libitum. All procedures were approved by the institutional 
committees on animal care at the University of Lethbridge, University of 
116 
Calgary, and the National Research Council of Canada which only approve 
procedures that are in accordance with the Canadian Council on Animal Care and 
were conducted in agreement with the National Institutes of Health guidelines. 
Procedures related to MR imaging and spectroscopy were conducted at the 
Experimental Imaging Centre, University of Calgary. Rats were temporarily 
transported to the University of Calgary vivarium where they were maintained 
and allowed to acclimate for one week. Scanning for each group of rats was also 
completed over a period of a week. One old rat died during the MR scanning 
procedures (which were performed prior to behavioral testing) due to an 
interaction between anesthesia and what turned out to be a large brain damage 
probably caused by a massive stroke. Even though the scan was completed, this 
animal was removed from analysis due to massive brain damage. 
Animal Preparation. Animals were first anesthetized with 4% isofluorane while 
breathing spontaneously. For maintenance of anesthesia during the experiment, 
isofluorane was reduced to 1.5% with slight correction for body weight. 
Respiration and heart rate were monitored throughout the experiment. The body 
temperature was also monitored throughout the scan via rectal probe, and 
maintained at 37 ± 0.5°C using heating pad. Each rat was placed in a plastic 
cradle and the head was immobilized. 
MRI Procedure. Imaging was performed using a 9.4 T/21 cm horizontal bore 
magnet with an Avance console (Bruker, Germany). A sagittal scout image was 
117 
acquired to ensure proper image alignment. A spin echo sequence (TR = 2 s, TE 
= 40 ms, 4 averages, 1 mm slice thickness, no gap, 256 X 256 resolution 
matrix, FOV = 2.5cm) was employed for acquisition of T2-maps in a coronal 
plane. A representative MRI image containing the dorsal hippocampus is 
presented in Figure 17. A total of 20 coronal sections were acquired and used for 
volumetric analyses. Anatomical images were acquired with a quadrature 
volume coil of 35 mm ID and 40 mm length (custom built by the Institute for 
Biodiagnostics/ National Reseach Council of Canada; IBD/NRC). The image 
acquisition took 17 min 4sec. 
Volumetric Analysis. The MRI scans were transferred to a desktop computer. 
Volumetric analysis was performed using in-house image analysis software 
(Marevisi, http://www.ibd.nrc.ca/english/bdo_e_marevisi.htm) developed by 
IBD/NRC. This software is a PC-based program with a graphic user interface, 
which allows the user to manually outline regions of interest (ROI) and obtain 
calculations of an area for the specific ROI. Coronal images were used for the 
volumetric analysis. For all ROIs, the measures for both hemispheres were 
combined. 
The hippocampus (cornu ammonis and dentate gyrus) was manually outlined on 
coronal sections from rostral to caudal. The starting rostral slice was 
approximately 2.30 mm posterior to bregma and was defined by the CA and DG, 
and coincided with the dorsal hippocampal commissure. The caudal boundary 
was defined by the loss of contrast between the external capsule and the 
118 
subiculum, the absence of DG, and the clear separation between the two 
cerebral hemispheres. This level corresponds to approximately - 6.72 mm from 
bregma. The hippocampal formation was measured on 6 consecutive slices in all 
animals. We also obtained a measure of whole brain volume by tracing all the 
images. Whole brain volume (or intracranial volume, ICV) was also used to 
calculate normalized hippocampal volumes. Three measures resulted from 
volumetric quantitation (all in mm3): intracranial volume (ICV), absolute 
hippocampal volume, and normalized hippocampal volume (to ICV). 
1H MRS Procedure. In vivo single voxel localized 1H MRS was performed using 
the same 9.4 T Bruker system employed for MR imaging. A STEAM sequence 
(parameters: TE = 20 ms, TR = 2 s, TM = 26ms, NEX = 1024, SW = 10000 Hz) 
was used. We tailored the voxel size (width 2mm, height 2mm, depth 3mm) 
using coronal T2-maps, to minimize CSF partial volume effects as well as the 
contribution of the tissue outside of the hippocampus. The spectra were acquired 
independently from left and right voxels centered on the hippocampi 
respectively, and the acquisition time was 34min 8sec per voxel. Voxel position 
is presented in Figure 18A. Spectroscopy was conducted with the help of a single 
turn ellipsoidal surface coil with 20x23 mm dimensions (custom built by 
IBD/NRC). 
Metabolite Quantification. A time-domain fitting strategy, AMARES (Advanced 
Method for Accurate, Robust, and Efficient Spectral Fitting), of short echo time in 
vivo XH MR spectra of normal rat brain was used to successfully quantify the 
119 
information. Time domain analyses provided consistent results with well-
resolved peaks. NAA, Cho, and Cre were quantified by jMRUI analysis (Naressi 
et al., 2001; http://www.mrui.uab.es/mrui), which incorporated the algorithms 
and graphical interface, and were expressed in institutional values. 
Representative spectra are presented in Figure 18B. 
11.3 Results 
MRI-Volumetrics. The Group differences in ICV (F (2,27) = 14.327, p < .001) 
were significant, such that the ICV was the highest in the old animals (see 
Figure 19). Post hoc tests revealed that the ICV significantly differed between 
the Young and Middle-aged (F(l, 19) = 20.226, p < .001), and the Young and 
Old (F(l, 17) =17.329, p = .001) rats, but not between the Middle-aged and 
Old group (p > .4). Hippocampal volume measurements were averaged for the 
left and right hippocampus, as there were no hemispheric differences. There was 
a significant difference between the age groups in both absolute (F (2,27) = 
4.018, p = .031; Figure 20) and normalized (F (2,27) =7.274, p = .003; Figure 
21) hippocampal volumes, such that the aged group had smallest hippocampi. 
Absolute hippocampal volume differed between the Young and Middle-aged rats 
(F(l, 19) = 5.61, p = .029), and the Young and Old (F(l, 17) =7.042, p = 
.017) rats, but not between the Middle-aged and Old group (p > .4). Similarly, 
normalized hippocampal volume differed between the Young and Middle-aged 
groups (F(l, 19) ••= 11.545, p = .003), and the Young and Old (F (1, 17) 
120 
= 11.497, p = .004) rats, but not between the Middle-aged and Old group (p > 
.4). 
Metabolite Quantification. Metabolites, namely NAA, Cho, and Cre, have been 
quantified separately for the left and right voxel centered on the hippocampus. 
In the left voxel, we found no age-related differences in NAA (p > .1; M = 9.45), 
Cho (p > .8; M = 1.67), or Cre (p > .1; M = 7.16). Similarly, there were no age-
related differences in NAA (p > .5; M = 9.41), Cho (p > .6; M = 1.85), or Cre (p 
> .8; M = 7.25) in the right voxel either. 
11.4 Discussion 
We imaged rat brain with a high field-strength magnet. This provided us with 
scans of sufficient anatomical contrast to determine the volume of the 
hippocampus, as well as the total intracranial volume. Our results suggest that 
both absolute and normalized hippocampal volume, obtained by quantifying 
structural MR images, was significantly lower in both the middle-aged and old 
rats in vivo. These findings suggest alterations in structural integrity of the 
hippocampus that begin early on in adulthood and continue into old age. These 
results are in line with our previous findings in humans (Driscoll et al., 2003a), 
that hippocampal volumes observed by MRI are significantly lower in elderly 
people compared to the young ones. 
121 
Our results suggest that neurochemical profile in the hippocampus, namely the 
levels of NAA, Cho, and Cre remain unaltered with age in the hippocampus of 
FBNF1 rats. More specifically, the fact that NAA remains stable suggests that 
neuronal metabolic function in the hippocampus is not compromised with age, at 
least not in this strain of rat. Although lH MRS is routinely used to study aging in 
humans, to the best of our knowledge, the present study is the first application 
of this technique to study hippocampal aging in the rat. Our findings of stable 
levels of metabolites under investigation could simply reflect the cross-species 
differences. Another possibility is that there is variability in chemical profiles 
between the strains of rat. Both issues remain to be resolved in the future. In 
addition, one must consider the possibility that the tissue other than the 
hippocampus may be contributing to the MR spectrum (given the lower 
hippocampal volumes with age) and hence biasing the metabolite quantitation. 
It is important to note that the fact that NAA remained stable in the aged rat 
hippocampus, despite the compromised structural integrity and behavioral 
deficits, is in direct contradiction to the human literature. This finding yet again 
brings into question the interpretation of NAA as a marker of neuronal viability 
and the interpretations of numerous studies linked to a decrease in NAA in the 
human brain related to various conditions and disease states. For example, in 
the human hippocampus, lower levels of NAA are routinely observed with age 
(e.g. Schuff et al., 1999), including our in own work (Driscoll et al., 2003a). In 
addition, studies in different patient cohorts (Brooks et al., 1997, 1999; Jung et 
al., 1998) and normal populations (Jung et al., 1999) using quantitative 
122 
spectroscopy and an extensive neurocognitive battery suggest a significant 
relationship between neurochemical markers of brain integrity and brain 
function. 
Overall, our results suggest that age-related structural deficits in the integrity of 
the hippocampal formation are apparent in vivo, as reflected by lower 
hippocampal volumes with age. Given that there were no differences in NAA, the 
overall functional integrity of neurons that remain in the old hippocampus seems 
to be preserved. Together, our findings lead us to believe that while the neurons 
that reside in the aged hippocampus retain their ability for normal metabolic 
functioning, the loss in the infrastructure is such that it is unable to provide the 
needed support for normal functioning, and we see such changes reflected in 
age-related deficits in hippocampus-dependent learning and memory. 
Allthewhile, magnetic resonance imaging has proven useful in tracking age-
related changes in the rat in vivo and is in line with human findings. 
A couple of questions still remain: (1) what exactly does a decrease in structural 
volume signify, and (2) what does the change in NAA levels denote. In an 
attempt to answer the above-mentioned controversies, we next turn to electron 
microscopy procedures in order to assess the same hippocampal tissue post 
mortem. We will not only be able to assess the number of neurons, synapses, 
and mitochondria in the volume of interest, but in the end, we will be able to 
relate the investigation on a more cellular and ultrastructural level to our 
findings that were obtained using magnetic resonance technology. 
123 
12. Experiment 7 - Light and Electron Microscopy 
Are the age-related alterations in hippocampal structural integrity 
observable post mortem? 
12.1 Background 
According to the review by Geinisman et al. (1995), the most frequently 
observed structural alterations during normal aging involve the following: (1) 
loss of synapses; (2) neuronal atrophy; (3) astroglial hypertrophy, and (4) 
accumulation of lipofuscin pigment in neurons. For a long time, degeneration of 
neuronal and synaptic populations has been deemed a probable cause of age-
related learning and memory problems. To date, it is not known whether and 
what kind of an effect such changes have on memory and neuronal function in 
general. 
Morphological changes in the number of neurons and synapses have been 
extensively investigated in search of structural substrates of memory 
dysfunction in normal aging (see Geinisman et al., 1995 for a review). Many of 
these studies, but not all, have reported an age-related loss of neurons and 
synapses in the hippocampus. More importantly, it has been demonstrated that 
the extent of the age-related morphological changes correlates with the severity 
124 
of memory impairment. Morphological studies on neuron and synapse 
quantification in the hippocampus, in both humans and non-human animals, are 
plagued by a remarkable inconsistency in the results. Many of these reports 
have not used unbiased stereological techniques. 
The advent of modern unbiased stereological techniques sidestepped the pitfalls 
of traditional stereological methodology and allowed for an unambiguous 
identification of section profiles of the objects being counted. The current state 
of the literature suggests that, using unbiased stereology and impeccable 
methodology, it has been convincingly demonstrated that neuronal loss does 
occur with chronological age in the human hippocampus (West, 1993). These 
results are however preliminary, and it remains to be characterized whether it is 
so for only certain regions of the hippocampal formation or if this process is 
more of a generalized one. In non-human animals, it remains to be resolved 
with unbiased stereology, whether there is a loss of hippocampal neurons with 
age. The loss of synapses in the non-human animals, however, has been 
confirmed using an unbiased stereological methodology, suggesting it as a 
candidate marker of age-related memory dysfunction. 
Metabolic correlates of hippocampal structural changes related to aging are not 
well investigated and remain largely unknown. Metabolic function is receiving 
increased attention in aging research and can be easily assessed in vivo using 
magnetic resonance spectroscopy, but this imaging modality is most commonly 
used in human research. Mitochondria are important determinants of 
125 
metabolism and can be easily quantified on an ultrastructural level using 
electron microscopy on post mortem tissue and readily applied to non-human 
animal research. 
In Experiment 6, MRI and MRS allowed us to obtain an estimate of structural 
and functional neuronal integrity in the hippocampus in vivo. The use of such 
techniques in non-human animal research, however, is still somewhat 
unorthodox. In order to verify that structural changes are indeed present in the 
hippocampus in relation to old age we turn to a more traditional methodology 
for assessing structural integrity in non-human animals - a microscopical 
assessment of tissue post mortem. A light microscopy and the physical disector 
method (Sterio, 1984) were employed to get an estimate of the number of 
neurons in the hippocampus. In addition, we conducted an electron microscopy 
study in which serial sections were combined with the physical disector method 
(Sterio, 1984) to quantify synapses and mitochondria in the young (3 months 
old), middle-aged (14 months old), and old (26 months old) animals. 
12.2 Materials and Methods 
Animals. The same animals that participated in Experiments 3, 5, and 6 are 
involved in the light and electron microscopy studies as well. Rats were divided 
into 3 groups: Young (N = 10; age 3 months at the start of the experiment), 
Middle-aged (N = 10; age 14 months at the start of the experiment), and Old 
126 
(A/ = 8; age 24 months at the start of the experiment) group. One old rat died 
during the MR scanning procedures (which were performed prior to behavioral 
testing) due to pre-existing stroke, and another one died of old age. Hence, all 
post mortem tissue analysis was done on eight animals in the old group. 
Animals were maintained and tested at the University of Lethbridge vivarium. All 
animals were housed in pairs in hanging plastic cages, in a room with an 
ambient temperature of 21° C and 35% relative humidity, 12/12/ light dark 
cycle. Food and water were available ad libitum. All studies were conducted 
according to guidelines provided by the National Institutes of Health and 
approved by the University of Lethbridge, University of Calgary, and the National 
Research Council of Canada, which only approve procedures that are in 
accordance with the Canadian Council on Animal Care. As previously mentioned, 
one rat suffered a stroke and died during MR scanning procedure. The brain of 
this rat was not processed for microscopy procedures. 
Histological Methods. After completing behavioral testing, and magnetic 
resonance imaging and spectroscopy procedures, rats were injected with BrdU 
(150 mg/kg) and allowed to survive for 5 weeks. Following this 5 week period, 
all rats were sacrificed rats by overdose of sodium pentobarbitol pentobarbital 
(100 mg/kg i.p.) and perfused transcardially with phosphate buffered saline 
(PBS; 0.9% saline in 0.1 M phosphate buffer) followed by 4% 
paraformaldehyde. Brains were extracted, and left and right hemispheres were 
separated. One hemisphere underwent immunohistochemical procedures while 
the other hemisphere underwent tissue preparation and embedding for electron 
127 
microscopy. The hemisphere undergoing immunohistochemical procedures was 
stored in the perfusate solution overnight and then transferred to phosphate 
buffered saline solution. Hemisphere undergoing electron microscopy procedures 
was immersed in a fixative solution (0.1M cacodylate buffer, 4% 
paraformaldehyde, 2.5% glutaraldehyde, and 2.5% dimethyl sulfoxide (DMSO) 
heated to 37°C for 5 minutes. The tissue was then transferred to the same 
solution at a temperature of 4°C and remained in it until dissected. The 
assignment of left or right hemisphere to either immunohistochemical or 
electron microscopy procedures was counterbalanced across all animals. 
Tissue Preparation. The hemisphere was coronally sectioned on a vibratome 
into 300 M-m thick sections. Samples of tissue containing the dentate gyrus of 
the hippocampus were taken from each section. The tissue was then washed in 
0.1M cacodylate, postfixed in 2% osmium tetroxide/1.5% potassium 
ferrocyanide in 0.1M cacodylate buffer for 2 hours and stained en bloc with 2% 
uranyl acetate for 45 minutes. Samples were then dehydrated through a series 
of alcohols before being transferred into propylene oxide and gradually 
embedded in Eponate resin using standard resin embedding procedures. 
One block of tissue, initially containing the granule cell layer of the dentate 
gyrus, was chosen at random from each animal and 32 coronal 1 urn 'thick' 
sections were taken using a diamond knife and a microtome. The sections were 
stained with toluidine blue and used for light microscopy (Figure 22A) to 
estimates neuronal density in the molecular layer adjacent to the dorsal horn of 
128 
dentate gyrus (Figure 22B). Following the collection of 'thick' sections, a small 
pyramid was trimmed such that each ultra-thin section spanned the inner 
molecular layer as well as the granule cell layer of the dorsal blade of the 
dentate gyrus, using toluidine stained thick sections from that block as a guide. 
The inner molecular layer was identified as an innermost third, the closest to the 
granule cell layer. From the pyramid, serial silver/gray ultrathin sections (70 
nm) were taken for electron microscopy using a diamond knife and an 
ultramicrotome. Sections were collected onto Formvar-coated copper grids and 
stained with uranyl acetate followed by lead citrate. Electron-microscopic 
measurements were taken from this tissue to obtain estimates of synaptic 
(Figure 23) and mitochondrial (Figure 24) density. 
Stereological Methods 
Estimating Changes. In conditions of stable neuron numbers changes in 
neuropil volume can be detected by estimating neuronal density. Measures of 
synaptic density alone may not reveal changes in synapse number. To measure 
changes in synapse number, the number of synapses per neuron can be 
calculated given that the density of neurons and the density of synapses 
respectively per unit volume have been obtained. Similarly, mitochondrial 
density, the number of mitochondria per neuron, and mitochondria per synapse 
were calculated. Unbiased stereological methods were employed in the present 
129 
study to determine whether such structural changes within the molecular layer 
of the hippocampus exist in relation to aging. 
Neuronal Density Measures. Using a computer assisted microscope and a 
stereology software package (Stereo Investigator), the physical dissector 
method (Sterio, 1984) was used to obtain estimates of neuronal density. Briefly, 
pairs of sections within each slide were compared in succession. The first section 
was considered the 'reference' section and its pair the 'look-up' section, and so 
on. In this method, neuronal nuclei that are present in one section are counted if 
they are not present in an adjacent section. Neuronal nuclei were identified by 
the presence of a central nucleolus within a pale nucleus, and by the oval to 
pyramidal shape of the surrounding cell soma, and counted using every other 1 
\im section. Neurons were identified by multiple criteria including the presence 
of a central nucleolus, a pale nucleus surrounded by perikaryal cytoplasm and, 
frequently, by the oval to pyramidal morphology of the soma. Irregularly shaped 
small nuclei with extensive chromatin characteristics of glia were not counted. 
Within an unbiased counting frame (37 828 fxm2 at 40X), the number of neuronal 
nuclei present in the 'reference' but not the 'look-up' section (Q-) was counted. 
The disector volume of tissue (75 656 nm3) through which the cells were 
counted (Vdis) was defined as Vdis = (Aframe)(H), where Aframe is the area of the 
counting frame and H represents the thickness of the section (1 nm) multiplied 
by the number of sections. Neuronal density was then determined by NVneuron = 
Q-/ Vdis. Neuron density was subsequently used to determine the volume of 
neuropil associated with each neuron (1/ Nvneuron)-
130 
To calculate average neuron size, photographs of the region were acquired 
under a light microscope (40X) from the comparable section for each animal. 
Nine neurons were selected from each photograph and measurements of a 
diameter and an area were obtained for each neuron using NIH Image 
(http://rsb.info.nih.gov/nih-image). An average neuronal diameter and area 
were calculated for each rat in order to get a measure of neuronal size. 
Synaptic Density Measures. The area under investigation was inspected with 
the electron microscope and an area of neuropil that was as much as possible 
free of nuclei was selected. A photomicrograph montage of four pictures, such 
that each micrograph represented one quadrant of a square, was taken from 
each position (at 6000X magnification). A section was skipped and the montage 
of the same area was taken in the subsequent section. The resulting negatives 
were printed at a final magnification of ~ 23 OOOX. Synaptic density (NVsynapse) 
estimate was obtained from these micrographs by comparing the pairs of 
electron micrographs approximately 70 nm apart within the series. Synaptic 
density was measured using a physical dissector method on two sets of eight 
serially positioned electron micrographs per rat. Synapses that were present in 
the 'reference' section within an unbiased counting frame (110.5 nm2) but not in 
the 'look-up' section, were counted. Synapses were identified by the presence of 
a synaptic membrane specialization and at least three presynaptic vesicles 
present in the presynaptic bouton and a presence of a post-synaptic density. 
Synaptic density was calculated as the number of synapses counted divided by 
131 
the sample volume. From these measurement the number of synapses per 
neuron could be determined ([NVsynaPse][l/ NVneuron ], Kleim et al., 1996), given 
that neuron count was previously made on the same tissue. 
Mitochondrial Density Measures. The same methodology described above for 
measuring synaptic density was employed to obtain an estimate of 
mitochondrial density in the dentate gyrus. Neuron and synapse density were 
previously obtained on the same tissue allowing the calculation of the number of 
mitochondria per neuron and the number of mitochondria per synapse as well. 
12.3 Results 
Neuronal Density. Measures for neuronal density, neuronal diameter, and 
neuronal area were subjected to a one-way ANOVA respectively with Group as 
an independent variable. On average, a total of 19 comparisons were made to 
count 102 neurons, a total of ~ 75 mm2. There was a significant effect of Group 
on neuronal density (F (2,14) = 5.603, p = .019; Figure 25A). Follow up tests 
revealed that the young and the middle-aged group did not differ in respect to 
neuronal density (p > .8). The difference between the young and old rats (F 
(1,9) = 27.899, p = .001), and the middle-aged and old rats (F(l,9) = 6.63, p 
= .033) respectively, was significant, such that neuronal density was 
compromised only in the Old group. 
132 
Similarly, there was a significant effect of Group on the size of neuropil per 
neuron (F (2,14) = 4.728, p = .031; Figure 25B). Follow-up tests revealed that 
the young and the middle-aged group did not differ in respect to neuronal 
density (p > .85). The difference in neuropil per neuron between the young and 
old rats (F (1,9) = 22.545, p = .001) was significant, and a non-significant trend 
was present between the middle-aged and old rats (F(l,9) = 4.544, p = .066), 
such that the amount of neuropil per neuron was compromised only in the Old 
group. 
In addition, we obtained measures of neuronal area and diameter, and 
subjected them to a one-way ANOVA respectively with age Group as an 
independent variable. We found no group differences in either measure of 
neuronal size (p > .5). 
Synaptic Density. Approximately 32 dissector pairs were used to count an 
average of 135 synapses through a volume of ~ 247 nm3. Measures of synaptic 
density and the number of synapses per neuron were subjected to a one-way 
ANOVA respectively with Group as an independent variable. We found no Group 
differences in synaptic density (p > .3) or in the number of synapses per neuron 
(p > .25). 
Mitochondrial Density. Approximately 32 dissector pairs were used to count 
an average of 140 mitochondria through a volume of ~ 247 nm3. Measures of 
mitochondrial density and the number of mitochondria per neuron were 
133 
subjected to a one-way ANOVA respectively, with Group as an independent 
variable. We found no Group differences in mitochondrial density (p > .8), in the 
number of mitochondria per neuron (p > .28), or the number of mitochondria 
per synapse (p > .2). 
12.4 Discussion 
The electron microscopy literature involving age-related brain changes is 
marked by contradictions, which are often attributed to the biased methodology 
employed, to the individual differences in aging, to a post-mortem delay, and to 
species and strain differences. We employed unbiased stereological methodology 
and the tissue was processed and post-fixed immediately after the animals were 
sacrificed. Measurements of neuronal, synaptic, and mitochondrial density were 
obtained from the same resin-embedded samples, therefore the effects of tissue 
shrinkage contributed equally to all density estimates and were unlikely to bias 
the data. Given that it has been widely accepted that there is neuronal atrophy 
in the aged human hippocampus, establishing whether aged rats exhibit cell loss 
of hippocampal neurons with unbiased stereological techniques is essential for 
the development of reliable animal models of human aging. 
Employing unbiased stereological techniques, our results show a lower numeric 
neuronal density per sample volume and a lower amount of neuropil per neuron 
in the molecular layer adjacent to the dentate gyrus of the hippocampus in aged 
134 
FBNF1 rat compared to both middle-aged and young rats. Neuronal count and 
the amount of neuropil per neuron seem preserved in middle-aged compared to 
young rats. At the same time, the neurons did not differ in their length or area 
in relation to age. Moreover, our results are in line with the finding that neuronal 
count is lower in the aged human hippocampi (West, 1993). 
On the ultrastructural level, we found no age-related alterations in the number 
of synapses or the number of synapses per neuron in the dentate gyrus of the 
hippocampus of FBNF1 rats. Age-related loss of hippocampal synapses has been 
confirmed using unbiased stereology in the dentate gyrus, but not other regions 
of the hippocampal formation, and in some instances correlated with spatial 
memory impairment (see Geinisman et al., 1995). Our findings are inconsistent 
with the view that aging is related to synaptic loss, at least in this specific region 
of interest. 
Also, synaptic efficacy has often been considered a basic mechanism underlying 
learning and memory. We have already described altered synaptic efficacy at 
least as measured by long term potentiation in Experiment 2. The already 
observed alterations in synaptic plasticity do not seem to be accompanied by a 
diminishing number of synapses, at least not in this region of interest. Previous 
work has found an age-related decline in synaptic numerical density per unit 
volume to involve only non-perforate but not perforated, axospinous synapses 
(see Geinisman et al., 1995 for a review). We did not make such a distinction in 
our current work. We can only assert with confidence that there is no change 
135 
with age in overall synaptic density in the molecular layer directly adjacent to 
the granule cell layer of the dorsal horn of the dentate gyrus of the 
hippocampus. 
Both quantity and size of mitochondria are important determinants of 
metabolism (Mjaatvedt and Wong-Riley, 1991; Wong-Riley, 1989). Here we 
quantified mitochondria using the same electron micrographs produced for 
estimating synaptic density. We found no age-related alterations in 
mitochondrial density or the number of mitochondria per synapse in the dentate 
gyrus of the hippocampus of the FBNF1 rats. The number of mitochondria is 
thought to be directly related to metabolic capacity (Wong-Riley, 1989). Hence, 
our results suggest that metabolic activity is preserved in the hippocampus of 
aged FBNF1 rats. Furthermore, because the number of mitochondria per 
synapse was preserved as well, synaptic activity in this region should be 
preserved. 
136 
13. Experiment 8 - Neurogenesis 
Does age affect neurogenesis, and if yes, do such changes further affect 
hippocampal function? 
13.1 Background 
Historically, it was commonly accepted that neurogenesis occurs only early on in 
development. Recently, however, it has been confirmed that neurogenesis 
within the hippocampal formation continues throughout adulthood (Altman and 
Das, 1965; Kaplan and Hinds, 1977; Lie et al., 2004). In addition, it turns out 
that most of the newly born cells express neuronal markers, receive synaptic 
inputs, extend axons along the mossy fibers, and possess similar 
electrophysiological properties to those of mature neurons in the granule layer 
of the dentate gyrus (see Kempermann et al., 2004; Lie et al., 2004 for 
reviews). 
The neurogenesis levels in the dentate gyrus of the hippocampus are reported to 
decline with age in the rat (Gould et al., 1999a; Kempermann et al., 1998; Kuhn 
et al., 1996). Neurogenesis has also been implicated as a potential contributor 
to normal learning and memory processes (Gould et al., 1999c). Efforts have 
been made to correlate age-related neurogenic and cognitive dysfunction related 
to the hippocampus, but these efforts did not prove to be as straightforward or 
137 
as fruitful as might be hoped. Recent studies, while consistently reporting a 
reduction in neurogenesis and behavioral impairments in the context of aging, 
fail to detect a clear relationship between the two (Bizon et al., 2004; Merrill et 
al., 2003). Most have assessed, however, only the rate of proliferation. Thus, it 
has been suggested that possibly survival, rather than the proliferation, is more 
important when it comes to contributing to hippocampus-dependent functioning 
(Drapeau et al., 2003). 
Hippocampal neurogenesis generates new neurons that become incorporated 
into the dentate gyrus throughout adulthood. The precursor cells in the dentate 
gyrus initiate adult hippocampal neurogenesis and it results in new granule cell 
neurons. Several developmental milestones take place between these two end 
points. The term 'neurogenesis' is used as a term that encompasses the entire 
multi-step process, and not just the division of neuronal cell progenitors. The 
present experiment was designed to assess three aspects of hippocampal 
neurogenesis: first, cell proliferation in the subgranular zone using Ki67 labeling, 
a marker for cycling cells; second, neuronal differentiation using doublecortin, a 
marker of immature and migrating neuroblasts; and third, survival of recently 
born hippocampal neurons using a long time interval between BrdU injection and 
labeling. Finally, the hypothesis that the deficit in hippocampus-dependent tasks 
is associated with changes in neurogenesis within the dentate gyrus in the 
context of aging will be evaluated. 
138 
13.2 Materials and Methods 
Animals. The same animals that participated in Experiments 3, 5, 6, and 7 also 
participated in the current study. Rats were divided into 3 groups: Young (age 3 
months at the start of the experiment), Middle-aged (age 14 months at the 
start of the experiment), and Old (age 24 months at the start of the experiment) 
group. Animals were maintained and tested at the University of Lethbridge 
vivarium. All animals were housed in pairs in hanging plastic cages, in a room 
with an ambient temperature of 21° C and 35% relative humidity, 12/12/ light 
dark cycle. Food and water were available ad libitum. All procedures were 
approved by the University of Lethbridge, University of Calgary, and the National 
Research Council of Canada, which only approve procedures that are in 
accordance with the Canadian Council on Animal Care, and were carried out in 
accordance with the NIH guidelines. One animal in the Old group died following 
the MR procedures, and another one died of old age during the 5 week period 
between the BrdU administration and the rats being sacrificed. The 
immunohistochemistry reported involves seven rats in the Old group. 
Injection Regiment. Four weeks after the completion of all testing procedures 
rats were injected intraperitoneal^ with Bromodeoxyuridine (BrdU; Sigma, 150 
mg/kg) dissolved in saline for three consecutive days, twice per day, 10 hrs 
apart. Rats were allowed to survive for five weeks following the final BrdU 
injection. 
139 
Tissue Processing. For perfusion, all rats were sacrificed rats by overdose of 
sodium pentobarbitol pentobarbital (100 mg/kg i.p.) and perfused transcardially 
with phosphate buffered saline (PBS; 0.9% saline in 0.1 M phosphate buffer) 
followed by 4% paraformaldehyde. Brains were extracted and left and right 
hemispheres were separated. Only one hemisphere was to undergo 
immunohistochemical procedures while the other hemisphere was to undergo 
tissue preparation and embedding for electron microscopy. The assignment of 
left or right hemisphere to either immunohistochemical or electron microscopy 
procedures was counterbalanced across all animals. The hemisphere undergoing 
immunohistochemical procedures was stored in perfusate solution overnight and 
then transferred to phosphate buffered saline. Brains were stereotaxically 
blocked in the coronal plane, and serial coronal sections were cut on a freezing 
microtome at 50 (Aim. One set of every fifth section was immuno-double-labeled 
for BrdU and neuronal marker NeuN. The second set of every fifth section was 
immuno-labeled for doublecortin (DCX), a marker for immature and migrating 
neuroblasts. The third set of every fifth section was immuno-labeled for Ki67, a 
marker for cycling cells. 
Immunohistochemistry 
BrdU/NeuN. The following DNA denaturation protocol was followed to visualize 
BrdU-labeled nuclei before immunolabeling: 2-hour incubation in 50% 
formamide/2X SSC buffer (0.3M NaCI, 0.03M sodium citrate) at 65°C, rinsed 
140 
twice for 5 min in 2x SSC, 30 min incubation in 2 M HCI at 37°C, and 5 washes 
in phosphate buffered saline (PBS) over 2 hours. Immunohistochemistry for 
BrdU was performed using a rat anti-BrdU antibody (Accurate Chemical, 
Westbury, NY) at a concentration of 1:250. 
Immunohistochemistry for NeuN was performed using a mouse anti-NeuN 
antibody (Chemicon, Temecula, CA) at a concentration of 1:500. For immuno-
lableing, sections were placed in a solution of 2% normal goat serum (NGS) and 
0.3% Triton X-100 in PBS (blocking solution). 
The primary antibodies for both BrdU and NeuN were diluted in the blocking 
solution. Sections were then placed in the primary antibodies diluted in the 
blocking solution overnight (~ 19 hrs). After three 5-minute rinses in PBS, the 
sections were incubated for 1 hour with secondary antibodies (goat anti-rat, 
1:500; and goat anti-mouse, 1:250; Molecular Probes, Eugene, OR) diluted in 
5X PBS. After three more 5-minute rinses with PBS, the sections were placed in 
streptavidin (Alexa 568; Molecular Probes, Eugene, OR) at a concentration of 
1:500 for 45 minutes. Finally, sections were rinsed with PBS, mounted on 
gelatin-coated slides, air dried, and coverslipped. Specificity of both the rat anti-
BrdU and mouse anti-NeuN antibodies was verified by omitting the primary and 
secondary antibodies respectively, which resulted in a loss of cellular labeling. 
BrdU-labeled cells that also expressed NeuN (Figure 26) throughout the dentate 
gyrus were detected with a laser-scanning confocal microscope lasers (Zeiss 
141 
LSM Microsystem; Axiovert 100 TV). Sections were optically sliced in the Z plane 
at 1 [irn intervals. The data were transferred to a desktop computer and viewed 
with Northern Eclipse software (http://www.empix.com), which also allowed a 
reconstruction of images in 3D. All double-labeled cells were rotated in 
orthogonal planes to verify double-labeling. BrdU/NeuN labeling procedures 
were performed on the tissue obtained from 10 young, 10 middle-aged, and 7 
old rats. 
Doublecortin (PCX). Sections were incubated overnight (~20 hrs) at room 
temperature with doublecortin (DCX, 1:250, Santa Cruz Biotechnology, Santa 
Cruz, CA) in 0.3% Triton X-100, 0.2% normal rabbit serum (NRS) in PBS. 
Sections were rinsed 3X in PBS, and placed in a secondary biotinylated anti-goat 
IgG (1:500) for 1 hr at room temperature. Sections were again rinsed 3X in 
PBS, and processed for immunofuorescence in streptavidin (Alexa 488, 1:500, 
Molecular Probes, Eugene, OR) for 45 min. Sections were rinsed in PBS, 
mounted on gelatin-coated slides, air dried, and coverslipped. The sections were 
viewed on a Zeiss Axioskop Z microscope under fluorescence. Cell somata were 
identified and counted under a 40X objective (Figure 27). DCX-labeling 
procedures were performed on the tissue obtained from 10 young, 10 middle-
aged, and 7 old rats. 
Ki67. Sections were incubated overnight (~20 hrs) at room temperature with 
Ki67 (1:1000; Novacastra, UK) in 0.3% Triton X-100, and 0.2% normal donkey 
serum (NDS) in PBS. Sections were rinsed 3X in PBS, and placed Cy3 
142 
Conjugated Affinipure Donkey Anti-rabbit IgG (1:250; Jackson Laboratories; 
West Baltimore Pike, PA) for 45 min at room temperature. Sections were rinsed 
in PBS, mounted on gelatin-coated slides, air dried, and coverslipped. The 
sections were viewed on a Zeiss Axioskop Z microscope under fluorescence. 
Ki67-labeled cells were identified and counted under a 40X objective (Figure 
28). Ki67 procedures were performed on the tissue obtained from 9 young, 9 
middle-aged, and 7 old rats. 
13.3 Results 
BrdU/NeuN. One-way ANOVA was performed with Group (Young, Middle-aged, 
and Old) as an independent variable and the number of BrdU-labeled cells or 
BrdU/NeuN double-labeled cells as a dependent variable respectively. We found 
a significant Group difference in the total number of BrdU-labeled cells (F(2,25) 
= 8.393, p = .002; Figure 29) five weeks after the last BrdU injection, such that 
the number of BrdU-labeled cells was the highest in the young rats, lower in the 
middle-aged, and the lowest in the old rats. Middle-aged rats had 20% of BrdU-
labeled cells, and old rats had 3% of BrdU-labeled cells compared to the number 
observed in the young rats five weeks after the last injection. Follow up tests 
revealed that this difference was significant between young and middle-aged 
rats (F(l,18) = 7.891, p = .012), young and old rats (F (1,16) = 8.989, p = 
.009), and middle-aged and old rats (F(l,15) = 7.871, p = .014). 
143 
Also, there was an age difference in the number of BrdU/NeuN double-labeled 
cells (F (2,25) = 9.522, p = .001; Figure 30) five weeks after the last BrdU 
injection, such that the number of BrdU/NeuN double-labeled cells was the 
highest in the young rats, lower in the middle-aged, and the lowest in the old 
rats. The number of double-labeled BrdU/NeuN cells in middle-aged rats was 
17%, and in old rats 6% that of the young ones five weeks after the last BrdU 
injection. Follow up tests revealed that the difference was significant between 
young and middle-aged rats (F(l,18) = 9.615, p = .006), young and old rats (F 
(1,16) = 9.789, p = .007), but not middle-aged and old rats (p > .1). 
Five weeks after BrdU injections, 47% of the BrdU-labeled cells in the young 
rats, 42% in the middle-aged rats, and 88% in the old rats were also double-
labeled with NeuN. It should be noted that in aged rats, even though we are 
seeing 88% of BrdU cells also double labeled with NeuN, this observation on 
average reflects a couple of double-labeled cells and it should not be interpreted 
as an increase in the rate of neuronal survival per se in these animals. Rather, it 
seems that the few cells that did survive after a period of five weeks in aged 
animals are primarily the ones that adopted a neuronal fate. 
PCX. One-way ANOVA was performed with Group (Young, Middle-aged, and 
Old) as an independent variable and the number of DCX-labeled cells as a 
dependent variable. There was a significant Group difference in the number of 
DCX-labeled cells (F (2,26) = 402.315, p < .001; Figure 31), such that the 
young group had the highest number of DCX-labeled cells, followed by the 
144 
middle aged group, and finally the old animals. Follow up tests revealed a 
significant difference between young and middle-aged rats (F(l, 16) = 424.164, 
p < .001), young and old rats (F (1, 16) = 446.254, p < .001), and middle-aged 
and old rats (F(l, 16) = 58.051, p < .001). The middle-aged rats had 20%, and 
old rats had 5% of the number of DCX-labeled cells seen in the young group of 
rats. 
Ki67. One-way ANOVA was performed with Group (Young, Middle-aged, and 
Old) as an independent variable and the number of Ki67-labeled cells as a 
dependent variable. There was a significant Group difference in the number of 
Ki67-labeled cells (F (2,24) = 24.188, p < .001; Figure 32), such that the young 
group had the highest number of Ki67-labeled cells, followed by the middle aged 
group, and finally the old animals. Follow up tests revealed a significant 
difference between young and middle-aged rats (F(l, 17) = 26.073, p < .001), 
young and old rats (F(l,15) = 24.847, p < .001), but not the middle-aged and 
old rats (F (1,15) = 1.265, p < .28). The middle-aged rats had 20%, and old 
rats had 14% of the number of Ki67-labeled cells seen in the young group of 
rats. 
13.4 Discussion 
Regulation of neurogenesis is considered an important indicator of 
neuroplasticity in the aging brain. In the current study, neurogenesis was 
145 
assessed using three independent measures, one of proliferation - Ki67 (and 
possibly BrdU alone, but the animals survived for 5 weeks after BrdU 
administration), one of neuronal differentiation - DCX, and one of mature 
neuronal survival - BrdU/NeuN. Overall, our results suggest that aging is 
associated with a significant neurogenic reduction in the rat hippocampus, a 
finding that is consistent with other studies of aging and neurogenesis to date. 
Our Ki67 results suggest a significant decrease in the rate of proliferation with 
age that begins sometime in adulthood, here specifically before 14 months of 
age in the FBNF1 rats. Our results are in agreement with numerous other 
studies suggesting a reduction in proliferation rate in relation to aging (e. g., 
Kempermann et al., 1998; 2002). Collectively, the reports of reduced 
proliferation rate with age would lead us to believe that the number of cells 
surviving in the aged hippocampus would be consequently reduced as well, 
given the lower pool of cells available to begin with. This seems to be the case, 
at least when it comes to neuronal differentiation and survival for a period of 5 
weeks, given the dramatic reduction in both DCX- and BrdU/NeuN- labeled cells. 
It seems plausible that the number of DCX and BrdU/NeuN cells can be 
explained by reduced proliferation alone. Markers of proliferation, BrdU alone 
and Ki67, provide us with similar estimates in the number of neurons born in the 
middle-aged and old rats respectively, compared to the young ones. For 
example, middle-aged rats had 20 %, and old rats had 3% of BrdU-labeled cells 
found in the young rats 5 weeks following BrdU administration. Similarly, 
146 
middle-aged rats had 20 %, and old rats had 14% of Ki67-labeled cells. Markers 
of immature (DCX) and mature neurons (BrdU/NeuN) provide us with similar 
estimates as well. Middle-aged rats had 20%, and old rats had 5% of DCX-
labeled cells found in the young rats. Middle-aged rats had 18 %, and old rats 
had 6% of BrdU/NeuN double-labeled cells found in the young rats. The only 
substantial discrepancy in these estimates appears in the measures of 
proliferation, BrdU vs. Ki67 (3% vs. 14% respectively). 
One reason for this discrepancy could be that BrdU as implemented here is not a 
true measure of proliferation when assessed 5 weeks after BrdU administration. 
In order to properly assess the rate of proliferation with BrdU, animals need to 
be sacrificed approximately 2-3 hours following the BrdU administration. The 
goal here, however, was to use it for assessment of neuronal survival in concert 
with NeuN in the first place. Also, it is known that BrdU is associated with 
dilution as the time goes by and the cells that originally incorporated BrdU into 
their DNA continue to divide. In addition, lipofuscin - a cellular waste product 
which accumulates in the somata of aging animals, has recently been suggested 
as a possible bias in immunohistochemical methodology, and it is possible that it 
could affect one method of labeling but not the other. 
Lipofuscin is a product of lipid peroxidation that accumulates over time in 
neuronal and non-neuronal cells. Lipofuscin accumulation in cells is thought to 
be a slow process that is taken as a sign of long-term oxidative stress. It has 
been suggested that the sustained induction of neurogenesis occurs in the 
147 
cellular microenvironment that is characterized by reduced neuronal damage. 
More specifically, (Kempermann et al., 2002) found that lipofuscin deposits in 
neurons are significantly reduced by long-term exposure to environmental 
enrichment, and the amount of lipofuscin deposited has been linked to cognitive 
deficits. Lipofuscin was not measured here, but should be considered in the 
future studies of aging and neurogenesis as it has been recently suggested as a 
possible bias in immunohistochemical procedures in aged organisms. 
In addition, in order to better describe the processes involving neurogenesis and 
elucidate a possible role for neurogenesis in the normally aging brain, we 
assessed the relationships that emerged between hippocampus-dependent 
behavior and other measures of hippocampal integrity with measures of 
hippocampal neurogenesis (Ki67-labeled cells, DCX-labeled cells, and 
BrdU/NeuN double-labeled cells). We describe these relationships in more detail 
in the Correlations section below. 
148 
14. Correlations 
In order to assess whether relationships between performance on hippocampus-
dependent tasks and measures of hippocampal structural and functional 
integrity obtained both in vivo and post mortem exist, we computed zero-order 
correlations (Table 2A). 
Measures of performance on the Morris water task in the correlation calculations 
included latency and path length respectively on the last trial on the days that 
the platform was in a new location, latency and path length, respectively, on the 
first trial on the days the platform was in the same location as the previous day, 
a difference in the latency and path length respectively between the average of 
the last trial on the days the platform was in a new location and the first trial on 
the days platform was in the same location as the previous day (a measure of 
24-hour retention), a difference in the latency and path length respectively 
between the average of the first trial and the last trial on the days the platform 
was in a new location (a more general measure of learning the new platform 
location across trials), and a difference in the latency and path length 
respectively between the average of the first trial and the third trial on the days 
the platform was in a new location (a more specific measure of new learning). 
For discrimination learning, performance measures were represented as a 
percent correct on each problem (3) for both elemental and transverse 
patterning discriminations. Additional measures that were included in correlation 
computations included absolute and normalize hippocampal volumes, neuronal 
149 
density, synaptic density, mitochondrial density, hippocampal NAA, Cho, and Cre 
levels, the number of Ki67-labeled (proliferation), DCX-labeled (differentiation), 
BrdU-labeled, and BrdU/NeuN double-labeled cells (survival). 
The performance on the two hippocampus-dependent tasks, one spatial and the 
other non-spatial in nature, correlated significantly. More specifically, success 
(percent correct) on the transverse patterning Problem 1 was significantly 
correlated with a difference score between trials 1 and 3 (learning) of the MWT 
on the day the platform was in a new location for both latency (r = .454, p = 
.015) and path length (r = .454, p = .015). Success in learning Problem 1 was 
significantly correlated with performance on Problems 2 (r = .634, p < .001) 
and 3 (r = .691, p < .001) of the transverse patterning discriminations. 
Transverse patterning Problem 2 performance significantly correlated with MWT 
place learning performance (latency) on the very first trial on days when the 
platform remained in the same location (r = - .436, p = .020), with a difference 
score between trials 1 and 3 (learning) on the day the platform was in a new 
location for both latency (r = .572, p = .001) and path length (r = .440, p = 
.019), and with performance on Problems 1 (r = .634, p < .001) and 3 (r = 
.860, p < .001) of the transverse patterning discriminations. 
Transverse patterning Problem 3 performance significantly correlated with place 
learning performance (latency) on the very last trial of the day the platform was 
in a new location (r = - .430, p = .022), with place learning performance 
150 
(latency) on the very first trial on days when the platform remained in the same 
location (r = - .501, p = .007), with a difference score between trials 1 and 3 
(learning) on the day the platform was in a new location for both latency (r = 
.561, p = .002) and path length (r = .404, p = .033), and with performance on 
Problems 1 (r = .691, p < .001) and 2 (r = .860, p < .001) of the transverse 
patterning discriminations. 
Absolute hippocampal volume (mm3) measured in vivo correlated significantly 
with a difference score between trials 1 and 3 (learning) on the day the platform 
was in a new location for both latency (r = .508, p = .006) and path length (r = 
.508, p = .006). There was a significant relationship between absolute 
hippocampal volume and Problem 1 of elemental discriminations (r = .418, p = 
.027), and Problems 2 (r = .432, p = .022) and 3 (r = .446, p = .017) of the 
transverse patterning discriminations. 
Absolute hippocampal volume also significantly correlated with the number of 
DCX-labeled cells (r = .493, p = .009), with the total number of BrdU-labeled 
cells that remained after 5 weeks (r = .408, p = .039), and with the total 
number of BrdU/NeuN double-labeled cells that survived for 5 weeks (r = .447, 
p = .022). In addition, absolute hippocampal volume significantly correlated 
with the whole brain volume (r = - .592, p = .001), and normalized hippocampal 
volume (r = .969, p < .001). 
151 
Normalized hippocampal volume (to ICV) correlated significantly with latency to 
reach the platform on the first trial on days when the platform remained in the 
same location as the previous day (r = - .378, p = .047; Figure 33), and with a 
difference score between trials 1 and 3 (place learning) on the day the platform 
was in a new location for both latency (r = .474, p = .011) and path length (r = 
.474, p = .011). Normalized hippocampal volume also significantly correlated 
with Problem 1 of elemental discriminations (r = .409, p = .031), and with 
performance on Problems 2 (r = .471, p = .011) and 3 (r = .488, p = .008; 
Figure 34) of the transverse patterning discriminations. 
Normalized hippocampal volume significantly correlated with the number of 
DCX-labeled cells (r = .605, p = .001; Figure 37), with the total number of 
BrdU-labeled cells that remained after 5 weeks (r = .390, p = .049), and with 
the total number of BrdU/NeuN double-labeled cells that survived for 5 weeks (r 
= .425, p = .031). In addition, normalized hippocampal volume significantly 
correlated with the whole brain volume (r = - .771, p < .001), and absolute 
hippocampal volume (r = .969, p < .001). 
Whole brain volume (in mm3) correlated significantly with a difference score 
between trials 1 and 8 (place learning) on the day the platform was in a new 
location for latency (r = .456, p = .015) and there was a trend for path length (r 
= .350, p = .068) to reach the platform. Whole brain volume also significantly 
correlated with both normalized (r = - .771, p < .001) and absolute 
hippocampal volume (r = - .592, p = .001). 
152 
The number of DCX-labeled cells significantly correlated with place learning 
performance (latency) on the last trial of the day the platform was in a new 
location (r = - .543, p = .003; Figure 35). This relationship was significant in the 
senescent rats (middle-aged and old combined; r = - .534, p = .027), but not 
for young group (and it is in the opposite direction; r = .506, p = .136). The 
number of DCX-labeled cells also significantly correlated with a latency 
difference score between trials 1 and 8 (learning) on the day the platform was in 
a new location (r = - .530, p = .004), with a difference score between trials 1 
and 3 (learning) on the day the platform was in a new location for both latency 
(r = .431, p = .025) and path length (r = .431, p = .025) to reach the platform, 
and with performance on Problems 1 (r = .429, p = .025), 2 (r = .443, p = 
.021), and 3 (r = .481, p = .011; Figure 36) of the transverse patterning 
discriminations. The relationship between DCX and performance on Problem 3 of 
the transverse patterning task was significant in the elderly (middle-aged and 
old together; see Figure 36; r = .805, p < .001), but not in the young rats (r = 
.180, p = .619). 
The number of DCX-labeled cells in the hippocampus also significantly correlated 
with absolute (r = .493, p = .009), and normalized (r = .605, p = .001; Figure 
37) hippocampal volumes, as well as with the total number of BrdU-labeled cells 
that remained after 5 weeks (r = .612, p = .001), and with the total number of 
BrdU/NeuN double-labeled cells that survived for 5 weeks (r = .642, p = .001; 
Figure 38). 
153 
There was a trend for the total number of BrdU-labeled cells that remained after 
5 weeks (p = .084), and with the total number of BrdU/NeuN double-labeled 
cells that survived for 5 weeks (p = .091) to be significantly correlated with a 
difference score between trials 1 and 3 (learning) on the day the platform was in 
a new location on the MWT. The total number of BrdU-labeled cells present after 
5 weeks was significantly correlated with absolute (r = .408, p = .039) and 
normalized hippocampal volume (r = .390, p = .049). The total number of 
BrdU/NeuN double-labeled cells present after 5 weeks also significantly 
correlated with absolute (r = .447, p = .022) and normalized hippocampal 
volume (r = .425, p = .031). The correlation between the total number of BrdU-
labeled cells that were present after 5 weeks and the total number of 
BrdU/NeuN double-labeled cells that survived for 5 weeks was also significant (r 
= .958, p < .001). 
The number of Ki67-labeled cells significantly correlated with performance on 
Problem 3 of the transverse patterning discriminations (r = .456, p = .022; 
Figure 39), and not with any measures of performance on MWT. The number of 
Ki67-labeled cells also significantly correlated with absolute (r = .449, p = .024) 
and normalized (r = .599, p = .002; Figure 40) hippocampal volumes. In 
addition, there was a significant relationship between the number of Ki67- and 
DCX-labeled cells (r = .869, p < .001; Figure 41), but not with BrdU-labeled 
cells. 
154 
14.1 Partial Correlations 
We also performed another set of correlations with the same variables described 
above, but this time we controlled for age. A slightly different pattern of 
relationships emerged, which are described in more detail below. 
Latency to reach the platform on the last trial on the days the platform was in a 
new location correlated with percent correct on the Problem 3 of transverse 
patterning discriminations (r = - .638, p = .047). There was a relationship 
between performance on Problem 3 of transverse patterning discriminations with 
Problem 1 (r = .744, p = .014), and a trend with Problem 2 (r = .586, p = 
.075). 
Hippocampal neuronal count was significantly correlated with performance on 
Problem 1 (r = .752, p = .012) and Problem 3 (r = .798, p = .007) of 
transverse patterning discriminations. Normalized hippocampal volume 
correlated with latency (r = - .672, p = .033) to reach the platform on the last 
trial on the days the platform was in a new location, and a non-significant trend 
for path length (r = .593, p = .071). There was also a relationship with absolute 
hippocampal volume (r = .939, p < .000) and whole brain volume (r = - .653, p 
= .041). The total number of BrdU-labeled cells that remained after 5 weeks 
was significantly correlated with BrdU/NeuN double-labeled cells (r = .919, p < 
.001). 
155 
There was a significant relationship between Choline levels in the right 
hippocampus and a difference score between trials 1 and 8 (learning) on the day 
the platform was in a new location for latency (r = .639, p = .046), and a trend 
for path length (r = .606, p = .063). Right hippocampal NAA levels significantly 
correlated with a difference score in path length to reach the platform between 
trials 1 and 8 on the days the platform moved to a new location (r = .704, p = 
.023). 
Given that we observed relationships between neurogenesis and behavior, and 
hippocampal volume and behavior respectively, but no significant relationships 
between neuronal density and behavior (unless we control for age), we wanted 
to see what was driving these relationships. Conceivably, one could imagine the 
rate of neurogenesis mediating both neuronal density and the volume of the 
hippocampus. The reverse, of either the volume or the neuronal atrophy 
predicting the rate of neurogenesis, is hard to imagine. 
Hence, another set of partial correlations was calculated including all the 
variables described above for zero-order correlations, but this time we controlled 
for the number of DCX-labeled cells. When the number of DCX-labeled cells was 
controlled for, the relationships between the measures of MWT (p > .174) and 
visual discrimination performance (p > .166) respectively and measures of 
hippocampal volume were no longer present. Similarly, there was no 
relationship between neuronal density and measures of learning and memory on 
hippocampus dependent tasks (p > .635). 
156 
Similarly, we computed additional partial correlations that included all the 
variables previously described in the Correlation section and but this time we 
controlled for neuronal density. When the effects of neuronal density were 
controlled for, the relationships between hippocampal volume and measure of 
both spatial and non-spatial hippocampus-dependent learning and memory were 
no longer present (p > .28). 
157 
15. General Discussion 
Collectively, our results suggest that aging is associated with deficits in both 
spatial and non-spatial hippocampus-dependent tasks. Behavioral deficits on 
hippocampus-dependent tasks are not altogether unexpected, given the 
observed age-related changes in hippocampal plasticity seen with age. In 
addition, aging of the hippocampus is accompanied by structural alterations on 
several different levels of description that can be observed both in vivo and post 
mortem. Furthermore, our results using a within-subjects experimental design 
show that the observed hippocampal structural changes predict the age-related 
dysfunction in learning and memory performance dependent on this brain 
region. Of equal importance is what was not observed. The marked behavioral, 
physiological, and anatomical changes were not accompanied by a deficit in 
functional metabolic integrity as measured by magnetic resonance spectroscopy. 
In considering the current results, one should not forget that the laboratory rats 
live in environments that are impoverished compared to their natural habitat. 
In considering the performance on hippocampus-dependent tasks, we observed 
deficits on both tasks employed here. Both middle-aged and old rats displayed a 
deficit in 24-hour retention of spatial information on the Morris water task. The 
age profile of performance is the same as the one we reported with humans in 
the virtual Morris water task (Driscoll et al., 2005). When it comes to learning 
and memory of non-spatial hippocampus-dependent information, a slightly 
different pattern of deficits emerges. Now, the performance of middle-aged rats 
158 
is similar to that of a younger rat, and the deficit is only present in the old 
group. Again, we have previously reported this exact pattern of performance in 
relation to age on the same task in humans (Driscoll et al., 2004a). Together, 
our previous findings in humans and and current findings in the rat suggest that 
the spatial task may be more sensitive to alterations in hippocampal functional 
integrity related to aging, as the deficits can be observed early on in adulthood. 
In respect to alterations in hippocampal anatomy, we found a reduction in the 
volume of the hippocampal formation in both middle-aged and old rats as 
observed by MRI. This is the first report of reduced hippocampal volume in the 
aging rat in vivo. Furthermore, our findings suggest that the alterations in 
hippocampal volume begin early in adulthood, sometime before 14 months of 
age in the FBNF1 rats, and seem progressive in nature. The results are also of 
importance in relation to validating our rat model of aging. Present MRI results 
are on agreement with numerous observations of volume reduction in the 
hippocampus of elderly humans (e.g., Driscoll et al., 2003a; Schuff et al., 
1999). The volume in middle-aged humans in the context of aging, however, 
has not been systematically investigated and we are unable to draw a 
comparison in this age group at this point in time. 
On a more general note, the success of our volumetric study also demonstrates 
the usefulness of MRI in rodent research, especially the possibility of 
investigating longitudinal age-related changes during the lifespan of the rat, a 
lifespan that is much shorter from that of a human and hence more conducive to 
159 
investigation. The question, however, remains as to what may underlie the 
volume decrease. Two candidates become readily apparent: (1) cell death, and 
(2) the neurogenic reduction. Here we measured both. Historically, it has been 
presumed that the reduction in volume is a direct consequence of neuronal 
atrophy that is expected to accompany aging, but recently this view has been 
brought into question. We will discuss the possible sources of volume reduction 
in more detail shortly, but first we will consider the information obtained from 
the post mortem tissue. 
In addition to structural age-related alterations observed in the hippocampus in 
vivo, we have also observed structural alterations in post mortem hippocampal 
tissue. We found a reduction in both neuronal density and the amount of 
neuropil per neuron in the hippocampal inner molecular layer of old rats, 
whereas no changes were observed in the young and middle-aged group. 
Whether the number of neurons in the human hippocampus is reduced in old 
age has been resolved to the satisfaction of many (West, 1993), and the 
resolution is that the neuronal density is indeed reduced in the elderly human 
hippocampus (see Geinisman et al., 1995). The jury is still out, however, on 
whether the same holds true for the aging rat hippocampus. Our current 
findings add to the body of literature that reports alterations in neuronal density 
in the hippocampus with old age and are in agreement with human findings. The 
studies reporting no changes in neuronal density in the aged rat hippocampus, 
however, should not be ignored. It remains to be resolved whether 
inconsistencies in this literature are dependent on the strain of rat used and the 
160 
exact area of the hippocampus under investigation, both of which have been 
suggested as a possible source of discrepancies. 
Nonetheless, it is easy to imagine how the loss in neuronal density may have a 
direct effect on hippocampal volume. The volume of the hippocampus in the 
aged rats was reduced by 11%, while the neuronal density was reduced by 
25%. This reduction in neuronal density seems substantial enough to account 
for the decrease in volume observed with old age. This idea is further supported 
by the fact that the significant relationship between volume and performance on 
hippocampus-dependent tasks is no longer significant once we controlled for 
neuronal density. Yet the reduction in hippocampal volume was also present in 
middle-aged rats in the absence of altered neuronal density, suggesting that 
decreased neuronal density contributes but does not conclusively explain the 
reduction in volume of this structure with age. 
Additional anatomical measures of the hippocampus were obtained from post 
mortem tissue in relation to neurogenic processes in order to assess 
proliferation, differentiation, and survival of recently born hippocampal neurons 
and its aftereffects, and to test the hypothesis that the deficit in hippocampus-
dependent tasks is associated with a reduction in neurogenesis within the 
dentate gyrus in the context of aging. Our results suggest that the rate of 
proliferation (Ki67), the number of newly born cells that have adopted an 
immature neuronal phenotype (DCX-positive), and the number of newly born 
cells co-expressing a mature neuronal marker NeuN five weeks after labeling are 
161 
all reduced in the aged rat hippocampus, and such changes are already apparent 
in the middle-aged animals. Neurogenesis is rarely investigated in humans, 
mainly due to practical and methodological limitations. Nonetheless, a reduction 
in the proliferation rate in the aged human hippocampus has been noted 
(Eriksson et al., 1998). 
As easy as it was to imagine how neuronal loss could account for a reduction in 
volume, it is almost as easy to imagine how the reduction in neurogenesis, in 
both proliferation and survival, could underlie both the reduction in neuronal 
density and the reduction in volume. A significant relationship between volume 
and performance on hippocampus-dependent tasks disappeared after controlling 
for DCX, suggesting the role for neurogenesis in maintaining the structural 
integrity of the hippocampus and performance on the tasks that depend on its 
integrity. 
It remains to be investigated further whether it is the case that the cell loss 
increases with age and at some point neurogenesis can no longer keep up. It 
could be the case that the cell loss is steady throughout the adulthood, but the 
neurogenesis is reduced severely in old age and can no longer keep up with 
maintaining the structural integrity of the hippocampal formation. Of course, the 
third possibility remains, that both apoptosis and neurogenesis have been 
unfavorably affected by aging. Our results suggest that this may be the case, 
given the reduction in neurogenesis, and given the increase in cell loss as 
evident by decreased neuronal density. In fact, the results of our partial 
correlation results suggest that both neuronal death and birth play a role the 
162 
structural integrity of the hippocampus. Hence, age-related neuronal atrophy 
does not fully explain decrease in hippocampal volume with age. What does 
seem to be more related to and better explains the age-related reduction in 
volume in both old and middle-aged rats is the rate of neurogenesis in the 
hippocampus. 
Although the volume of the hippocampus, as observed by MRI, decreases 
beginning sometime in adulthood, the functional metabolic integrity as 
measured by MRS stays preserved in the cells that remain in old age. Together, 
these results suggest that the structural integrity of the aged hippocampus is 
compromised but the functional metabolic integrity of neurons that remain is 
preserved. This finding, however, stands in stark contrast to consistent reports 
of alterations in levels of different metabolites in the aged human hippocampus 
(e.g., Driscoll et al., 2003a; Schuff et al., 1999). Whereas MRS is now routinely 
employed in human research, it is still a fairly new modality of MR and this is the 
first attempt to evaluate the metabolic profile of the aged rat hippocampus 
employing this methodology. Hence, we are unsure as to why this discrepancy 
between rat and human findings exists. 
A couple of possibilities are apparent. One, there could be a clear cross-species 
difference, although it is hard to speculate what the source of this difference 
might be given that thus far magnetic resonance spectroscopy has not been 
employed to study normally aging rats. When employed to study metabolic 
profiles in the rat brain after ischemia, for example, (Demougeot et al., 2001), 
163 
the findings are similar to those reported in humans. The possibility remains 
that the metabolic processes in the senescent rat brains are fundamentally 
different from those occurring in the injured brain or in the aging human brain 
for that matter. Second, as previously mentioned in Experiment 2, in addition to 
possible differences between species, the rat strain differences are mentioned as 
a possible explanation for differential findings related to synaptic plasticity, and 
the same could be the case in relation to detecting metabolic age-related 
differences in the rat brain. Given that this is the first MRS study of the normally 
aging rat, both possibilities of cross-species and strain differences contributing 
to the discrepancy in the results remain to be further investigated. In addition, 
in regard to MRS, the lack of findings in the aged rat hippocampus, although 
preliminary, does bring into question the interpretations of cognitive deficits that 
accompany NAA reduction in human aging. 
Overall, a pattern of relationships emerged from the measures of hippocampal 
integrity that were obtained on several different levels of investigation. More 
specifically, we found a significant relationship between performances on two 
hippocampus-dependent tasks, one spatial and one non-spatial in nature, a 
relationship that remained even when the age was controlled for. Further, 
performance on both hippocampus-dependent tasks was significantly correlated 
with hippocampal volume measured in the same rats in vivo. The above-
mentioned patterns of relationships resembles those previously reported in 
humans in relation to aging (Driscoll et al., 2003a). 
164 
In addition, significant relationships emerged between the hippocampus-
dependent learning and memory and the number of DCX-labeled cells and 
BrdU/NeuN-labeled cells respectively, but not to the same degree with Ki67-
labeled cells. Recent studies that have attempted to correlate neurogenesis with 
spatial performance report a reduction in both neurogenesis and behavioral 
impairments in the context of aging, but fail to detect a clear relationship 
between the two (Bizon et al., 2004; Merrill et al., 2003). These studies, 
however, have primarily assessed the rate of cell proliferation. Thus, it has been 
suggested that survival, rather than the proliferation, is more important when it 
comes to contributing to hippocampus-dependent functioning (Drapeau et al., 
2003). Our findings are in line with this idea, given the relationships between 
the measures of neuronal differentiation and survival with behavior. They are 
not independent of the reduction in the proliferation rate. Given that the rate of 
proliferation was severely reduced in the aged rats to start with, the age-related 
differences in the rate of neuronal differentiation and survival could be seen as a 
direct consequence of reduced proliferation. Such observations make sense. The 
cells that are proliferating are yet to migrate to their final destination and are 
not expected to have yet committed to neuronal faith or otherwise, hence are 
not functionally incorporated into the existing circuitry. 
With regard to 1ft MRS, even though we found no age-related alterations in the 
hippocampal metabolite levels, namely NAA, the lack of results is not necessarily 
uninteresting. To the best of our knowledge, this is the first attempt to examine 
the aging rat brain with MRS and our results are preliminary. Unfortunately, the 
165 
lengthy scan time in order to acquire structural images and in order to obtain 
quality spectra from the hippocampi precluded us from examining other regions 
of the brain with same methodology. It would certainly be interesting to know 
whether alterations in the metabolic profile are present in any other regions of 
the aged rat brain, or if this strain of rat may be immune to metabolic 
alterations with age. 
Potentially more interesting, at least to the MRS community, is our investigation 
of neuronal and mitochondrial numbers in an attempt to better explain what 
NAA levels are representing. Given that hippocampal NAA levels remain stable in 
the face of aging but neuronal count was reduced in the old group, we have to 
dismiss the view of NAA as a putative neuronal marker or more precisely as an 
indicator of neuronal atrophy or neuronal density. Given that the number of 
mitochondria also remained stable with age, our results lead us to agree with 
more recent propositions of NAA as a marker of neuronal viability or neuronal 
metabolic function. 
Together, our results define some changes that occur in the rat hippocampus 
during the process of normal aging and suggest FBNF1 rats to be a very useful 
model of normal aging in humans. The pattern of similarities in changes that 
occur with aging in both rats and humans alike is striking, with the exception of 
MRS results. Also, we have not only validated learning and memory tasks that 
are hippocampus-dependent, sensitive to the aging process, and provide a 
reasonable basis for cross-species comparisons, but we have also affirmed 
166 
methodology (MRI and MRS) that can be readily applied in a similar fashion to 
study both humans and rodents. In addition, we have described some 
alterations that occur in the hippocampus with normal aging that seem to be 
related to the hippocampus-dependent learning and memory dysfunction. 
Even though some of these alterations begin quite early in adulthood they 
should be viewed as a part of normal aging process and as distinct from age-
related pathology. It is reasonable to consider that such changes can be of 
adaptive nature to an aging organism, even though the alterations are at times 
substantial. If we buy into this argument, then it would reasonable to assume 
that both elderly humans and rats are able to remain functional, and get along 
in and adapt to their environment despite the age-related changes they might 
be experiencing. The difficulty arises, however, in detecting and describing the 
point at which normal aging crosses over to pathology and an organism suffers a 
substantial loss in the quality of life. It is the identification of this point that will 
ultimately allow us to better target our diagnosis, prevention, and treatment 
efforts directed toward alleviating the devastation that accompanies the age-
related pathology. 
167 
16. Closing Comments 
Cognitive outcomes resulting from old age are said to vary in frequency, 
direction, and especially extent of the deficits. For example, a slowing of 
function, physical and psychological, is universal in people over the age of 60, 
but there is also a good chance of reaching a very old age with no notable 
memory problems for example. Also, some cognitive processes are not very 
susceptible to the deleterious effects of age, and some (such as vocabulary) 
even improve with age. Similarly, even though e4 allele of the APOE constitutes 
a genetic risk for AD, not all E4 allele carriers develop the disease. Interestingly, 
the fact in animal aging literature is that about a third of aged animals show no 
memory impairment and perform on memory tests as well as their younger 
counterparts (Gallagher and Rapp, 1997). 
One aspect of cognitive aging that has been largely ignored is the influence of 
genetic factors on the expression of cognitive phenotypes. A current theory 
concerning the etiology of age-related cognitive decline takes a combination of 
co-factors into an account (McDonald, 2002). This theory hints at the underlying 
genetic factors as well as individual differences that may differentially affect the 
aging process and its outcomes. This theory divides co-factors into two 
categories, a passive factor category (such as senile plaques, neurofibrillary 
tangles, head trauma, cholinergic depletion, and prolonged glucocorticoid 
exposure), and active factors (such as seizures, vascular abnormalities, 
circadian dysfunction, and other unknown environmental influences unique to 
168 
each organism). The idea is that the more factors that are present from each 
category the larger the damage to the hippocampus at a faster rate, and more 
pronounced and severe the deficits. The type, number, and combination of co-
factors vary from person to person (or rather rat to rat), and could possibly 
account for individual differences in cognitive aging. 
Laboratory rodents are especially suitable as models of cognitive aging in 
humans because they have a relatively short lifespan and because many tests 
have been developed to assay their cognitive performance. Animal models may 
be helpful for gaining insight into normal aging and for determining the relative 
contribution of various factors to age-related changes. Changes occurring with 
age might be due to extrinsic factors (ecology, environment) or intrinsic factors 
(genetics), or the combination of both. In laboratory experiments extrinsic 
factors are normally well controlled for and standardized in order to eliminate 
them as sources of possible confounding effects and also allow for intrinsic 
factors to take effect. 
Both human and non-human animal studies are needed in order to further our 
understanding of the genetic and environmental effects as well as the processes 
that may be influencing variability in behavior and its functional outcome with 
aging. Future research should be directed at elucidating the nature and role of 
genes on behavior, and on genetic-environmental interplay. 
169 
17. Conclusion 
A recent revolution in neuroscience techniques has lead to an accumulation of 
converging evidence on neural, physiological, and cognitive changes with aging. 
The substrates of the age-related cognitive deficits are still unclear. Spatial 
memory in particular has been extensively studied in aged rodents, but not as 
well in aged humans because techniques for investigating brain-behavior 
relations in vivo were not well developed. Recent advancement in magnetic 
resonance has allowed visualization of hippocampal anatomy and quantification 
of changes in vivo. Such techniques, however, not only allow us to delineate the 
neuroanatomical substrates in health and disease, but are also applicable to 
both humans and non-human animals allowing for direct cross-species 
comparisons. 
The aging rat as a model for human aging may shed light on the aging process 
and the accompanying neurodegenerative disorders. A particular advantage of 
rodent aging studies is that the contribution of genetic and environmental 
factors can be investigated and strictly controlled. The newly developed 
transgenic mouse models promise to further unravel the genetic factors 
underlying aging (Vijg, 2000). It has been suggested that the age-related 
cognitive impairments are related to a decline in hippocampal plasticity, which in 
turn may dependent on alterations in gene expressions. The molecular products 
of genomic responses may lead to permanent biological and/or morphological 
170 
changes in the brain. Further inquiry is needed to identify the factors responsible 
for the susceptibility of this structure to aging. 
An important feature of animal models of aging and animal models in general, is 
that they closely mimic the characteristics of human disorders and that they are 
sensitive to individual differences. We have not investigated individual 
differences in the current study, as much larger number of animals would be 
required to be able to sufficiently detect such differences. At the same time we 
feel that our rat model of human aging does closely mimic the characteristics of 
human aging and will prove to be very useful in studying both normal and 
pathological aging in the future, especially because the tasks and methodology 
employed will allow for cross-species comparisons. 
Many intervening variables must be considered when evaluating the usefulness 
of tasks to study the effect of age on learning and memory. For example, one 
should insure that performance of both human and non-human animals is not 
influenced by sensory or motor deficits, or differences in motivational or 
problem-solving strategies. The goal for the future is to develop more sensitive 
animal and human behavioral tasks that would permit an evaluation of similar 
processes in both human and non-human animals and that would also allow for 
an evaluation of the underlying neurobiology on cognitive processes in a 
tractable manner. 
171 
Although many questions regarding the navigation network as well as the 
functional role of the hippocampus remain to be determined, the behavioral 
tasks described here may be the closest link yet in behavioral tasks 
implemented between humans and other mammals. Such research may not only 
aid in better understanding of the aging processes, but will also further our 
overall understanding of the role of the hippocampus in learning and memory. 
172 
18. References 
Aberg MAI, Aberg ND, Hedbacker H, Oscarsson J, Eriksson PS. Peripheral 
infusion of IGF-I selectively induces neurogenesis in the adult rat hippocampus. 
J Neurosci 2000; 20:2896-2903. 
Aguirre GK, Detre JA, Alsop DC, DEsposito M. The parahippocampus subserves 
topographical learning in man. Cereb Cortex 1996; 6:823-829. 
Allegrini PR, Sauer D. Application of magnetic resonance imaging to the 
measurement of neurodegeneration in rat brain: MRI data correlate strongly 
with histology and enzymatic analysis. Magn Reson Imaging 1992; 10:773-778. 
Altman J, Das GD. Autoradiographic and histological evidence of postnatal 
neurogenesis in rats. J Comput Neurol 1965; 124:319-335. 
Alvarado MC, Rudy JW. Some properties of configural learning: An investigation 
of the transverse patterning problem. J Exp Psychol Anim Behav Proc 1992; 
18:145-153. 
Alvarado MC, Rudy JW. Rats with damage to the hippocampal formation are 
impaired on the transverse-patterning problem but not on elemental 
discrimination. Behav Neurosci 1995a; 109:204-211. 
Alvarado MC, Rudy JW. A comparison of kainic acid plus colchicine and ibotenic 
acid-induced hippocampal formation damage on four configural tasks in rats. 
Behav Neurosci 
1995b; 109:1052-1062. 
Amaral DG, Witter MP. The three-dimensional organization of the hippocampal 
formation: A review of anatomical data. Neurosci 1989; 31:571-591. 
Amaral DG, Witter MP. The hippocampal formation. In: Paxinos G, ed. The Rat 
Nervous System, 2nd ed. New York: Academic Press, 1995; 443-493. 
Anagnostaras SG, Maren S, Sage JR, Goodrich S, Fanselow MS. Temporally 
graded retrograde amnesia of contextual fear after hippocampal damage in rats: 
Within-subjects examination. J Neurosci 1999; 19:1106-1114. 
Assaf Y, Beit-Yannai E, Shohami E, Berman E, Cohen Y. Diffusion and T2-
weighted MRI of closed head injury in rats: a time course study and correlation 
with histology. Magn Reson Imaging 1997; 15:77-85. 
Astur RS, Constable RT. Hippocampal dampening during a relational memory 
task. Behav Neurosci 2004; 118:667-675. 
173 
Astur RS, Ortiz ML, Sutherland RJ. A characterization of performance by men 
and women in a virtual Morris water task: a large and reliable sex difference. 
Behav Brain Res 1998a; 93:185-190. 
Astur RS, Sutherland RJ. Configural learning in humans: The transverse 
patterning problem. Psychobiology 1998b; 26:176-182. 
Astur RS, Taylor LB, Mamelak AN, Philpott L, Sutherland RJ. Humans with 
hippocampal damage display severe spatial memory impairments in a virtual 
Morris water task. Beh Brain Res 2002; 132:77-84. 
Astur RS, Tropp J, Sava S, Constable RT, Markus EJ. Sex differences and 
correlations in a virtual Morris water task, a virtual radial arm maze, and mental 
rotations. Behav Brain Res 2004; 151:103-115. 
Baba T, Moriguchi M, Natori Y, Katsuki C, Inoue T, Fukui M. Magnetic resonance 
imaging of experimental rat brain tumors: histopathological evaluation. Surg 
Neurol 1990; 34:378-382. 
Bach ME, Barad M, Son H, Zhuo M, Lu YF, Shih R, Mansuy I, Hawkins RD, 
Kandel ER. Age-related defects in spatial memory are correlated with defects in 
the late phase of hippocampal long-term potentiation in vitro and are attenuated 
by drugs that enhance the cAMP signaling pathway. Proc Natl Acad Sci USA 
1999; 96:5280-5285. 
Bannerman DM, Yee BK, Lemaire M, Wilbrecht L, Jarrard L, Iverson SD, Rawlins 
JN, Good MA. The role of entorhinal cortex in two forms of spatial learning and 
memory. Exp Brain Res 2001; 141:281-303. 
Barnea A, Nottebohm F. Seasonal recruitment of hippocampal neurons in adult 
free-ranging black-capped chickades. Proc Natl Acad Sci USA 1994; 8:11217-
11221. 
Barnes CA. Memory deficits associated with senescence: A neurophysiological 
and behavioral study in the rat. J Comp Physiol Psychol 1979; 93:74-104. 
Barnes CA. Aging and the physiology of spatial memory. Neurobiol Aging 1988; 
9:563-570. 
Barnes CA. Long-term potentiation and the ageing brain. Philos Trans R Soc 
Lond B Biol Sci 2003; 358:765-772. 
Barnes CA, Eppich C, and Rao G. Selective improvement of aged rat short-term 
memory by 3, 4-diaminopyridine. Neurobiol Aging 1989; 10:337-341. 
Barnes CA, Markowska AL, Ingram DK, Kametani H, Spangler EL, Lemken VJ, 
and Olton DS. Acetyl-l-carnitine 2: Effects on learning and memory performance 
of aged rats in simple and complex mazes. Neurobiol Aging 1990; 11:499-506. 
174 
Barnes CA, McNaughton BL. Physiological compensation for loss of afferent 
synapses in rat hippocampal granule cells during senescence. J Physiol - London 
1980; 309:473-485. 
Barnes CA, McNaughton BL. An age comparison of the rates of acquisition and 
forgetting of spatial information in long-term enhancement of hippocampal 
synapses. Behav Neurosci 1985; 99:1040-1048. 
Barnes CA, Rao G, Houston FP. LTP induction threshold change in old rats at the 
perforant path-granule cell synapse. Neurobiol Aging 2000; 21:613-620. 
Barnes CA, Suster MS, Shen J, McNaughton BL. Multistability of cognitive maps 
in the hippocampus of old rats. Nature 1997; 388: 272-275. 
Bates TE, Strangward JK, Davey GP, Munro PMG, Clark JB. Inhibition of N-
acetylaspartate production: implications for XH MRS studies in vivo. Neuroreport 
1996; 7:1397-1400. 
Bayer SA. Changes in the total number of dentate granule cells in juvenile and 
adult rats: a correlated volumetric and 3H-thymidine autoradiographic study. 
Exp Brain Res 1982; 46:315-323. 
Ben Horin N, Hazvi S, Bendel P, Shul R. The ontogeny of a neurotoxic lesion in 
rat brain revealed by combined MRI and histology. Brain Res 1996; 718:97-104. 
Benuch M, D'Adamo AF. Acetyl transport mechanism. Biochem Biophys Acta 
1968; 152:611-618. 
Berch DB, Israel M. Solution of the transverse patterning problem: Response to 
cue-cue relations. Psychon Sci 1971; 23:383-384. 
Birken DL, Olendorf WH. N-acetyl-L-aspartic acid - a literature review of a 
compound prominent in H-l-NMR spectroscopic studies of brain. Neurosci 
Biobehav Rev 1989; 13:23-31. 
Bizon JL, Lee HJ, Gallagher M. Neurogenesis in a rat model of age-related 
cognitive decline. Aging Cell 2004; 3:227-234. 
Bliss TVP, Gardner-Medwin AR. Long-lasting potentiation of synaptic 
transmission in the dentate area of the unanesthetised rabbit following 
stimulation of the perforant path. J Physiol (London) 1973; 232:357-374. 
Bliss TVP, Lomo T. Long-lasting potentiation of synaptic transmission in the 
dentate area of the anesthetized rabbit following stimulus of perforant path. J 
Physiol (London) 1973; 232:331-356. 
175 
Bohbot V, Kalina M, Stepankova J, Spackova N, Petrides M, Nadel L. Spatial 
memory deficits in patients with lesions to the right hippocampus and to the 
right parahippocampal cortex. Neuropsychologia 1998; 36:1217-38. 
Bouilleret V, Nehlig A, Marescaux C, Namer IJ. Magnetic resonance imaging 
follow-up of progressive hippocampal changes in a mouse model of mesial 
temporal lobe epilepsy. Epilepsia 2000; 41:642-650. 
Brooks WM, Sabet A, Sibbitt WL, Barker PB, van Zijl P, Moonen C, Duyn J. 
Neurochemical quantification in Systemic Lupus Erythematosus. J Rheumatol 
1997; 24:2323-2329. 
Brooks WM, Jung RE, Ford CC, Greinel EJ, Sibbitt WL. Relationship between 
neurometabolite derangement and neurocognitive dysfunction in systemic lupus 
erythematosus. J Rheumatol 1999; 26:81-85. 
Bussey TJ, Warburton EC, Aggleton JP, Muir JL. Fornix lesions can facilitate 
acquisition of the transverse patterning task: A challenge for "configural" 
theories of hippocampal function. J Neurosci 1998; 18:1622-1631. 
Cameron HA, McKay RDG. Adult neurogenesis produces a large pool of new 
granule cells in the dentate gyrus. J Comp Neurol 2001; 435:406-407. 
Cameron HA, Wooley CS, McEwen BS, Gould E. Differentiation of newly born 
neurons and glia in the dentate gyrus of the adult rat. Neuroscience 1993; 
56:337-344. 
Carlen M, Cassidy RM, Brismar H, Smith GA, Enquist LW, Frisen J. Functional 
integration of adult-born neurons. Curr Biol 2002; 12:606-608. 
Chang L, Ernst T, Poland RE, Jenden DJ. In vivo proton magnetic resonance 
spectroscopy of the normal aging human brain. Life Sci 1996; 58:2049-2056. 
Cho YH, Friedman E, Silva AJ. Ibotenate lesions of the hippocampus impair 
spatial learning but not contextual fear conditioning in mice. Behav Brain Res 
1999; 98:77-87. 
Cipolotti L, Shallice T, Chan D, Fox N, Scahill R, Harrison G, Stevens J, Rudge P. 
Long-term retrograde amnesia ...the crucial role of the hippocampus. 
Neuropsychologia 2001; 39:151-172. 
Clark RE, Broadbent NJ, Squire LR. Hippocampus and remote spatial memory in 
rats. Hippocampus in press. 
Clark RE, Broadbent NJ, Zola SM, Squire LR. Anterograde amnesia and 
temporally graded retrograde amnesia for a non-spatial memory task after 
lesions of hippocampus and subiculum. J Neurosci 2002; 22:4663-4669. 
176 
Colombo M, Fernandez T, Nakamura K, Gross CG. Functional differentiation 
along the anterior-posterior axis of the hippocampus in monkeys. J Neurophysiol 
1998; 80:1002-1005. 
Corkin S. Lasting consequences of bilateral medial temporal lobectomy: Clinical 
course and experimental findings in H.M. Semin Neurol 1984; 4:249-259. 
Curtis MA, Penney EB, Pearson AG, van Roon-Mom WMC, Butterworth NJ, 
Dragunow M, Connor B, Faull RLM. Increased cell proliferation and neurogenesis 
in the adult human Huntington's disease brain. Proc Natl Acad Sci USA 2003; 
100:9023-9027. 
D'Adamo AF, Smith JC, Woiler C. The occurrence of N-acetylaspartate 
amidohidolase (Aminoacylase II) in the developing rat. J Neurocheml973; 
20:1275-1278. 
Dautry C, Vaufrey F, Brouillet E, Bizat N, Henry PG, Cone F, Bloch G, Hantraye 
P. Early N-Acetylaspartate depletion is a marker of neuronal dysfunction in rats 
and primates chronically treated with the mitochondrial toxin 3-Nitropropionic 
acid. J Cereb Blood Flow Metabol 2000; 20:789-799. 
Davie CA, Hawkins CP, Barker GJ, Brennan A, Tofts PS, Miller DH, McDonald WI. 
Serial proton magnetic resonance spectroscopy in acute multiple sclerosis 
lesions. Brain 1994; 117:49-58. 
De Stefano N, Matthews PM, & Arnold DL. Reversible decrease in N-
acetylaspartate after acute brain injury. Mag Res Med 1995; 34:721-727. 
de Toledo-Morrell L, Morrell F. Electrophysiological markers of aging and 
memory loss in rats. Ann NY Acad Sci 1985; 444:296-311. 
de Toledo-Morrell L, Geinisman Y, Morrell F. Age-dependent alterations in 
hippocampal synaptic plasticity: Relation to memory disorders. Neurobiol Aging 
1988; 9:581-590. 
Demougeot C, Gamier P, Mossiat C, Bertrand N, Giroud M, Beley A, Marie C. N-
acetylaspartate, a marker of both cellular dysfunction and neuronal loss: its 
relevance to studies of acute brain injury. J Neurochem 2001; 77:408-415. 
Derrick BE, York AD, Martinez JL Jr. Increased granulle cell neurogenesis in the 
adult dentate gyrus following mossy fiber stimulation sufficient to induce long-
term potentiation. Brain Res 2000; 857:300-307. 
Diana G, Domenici MR, Loizzo A, Scotti de Carolis A, Sagratella S. Age and strain 
differences in rat place learning and hippocampal dentate gyrus frequency-
potentiation. Neurosci Lett 1994a; 171:113-116. 
177 
Diana G, Scotti de Carolis A, Frank C, Domenici MR, Sagratella S. Selective 
reduction of hippocampal dentate frequency potentiation in aged rats with 
impaired place learning. Brain Res Bull 1994b; 35:107-111. 
Douglas RM, Goddard GV. Long-term potentiation of the perforant path-granule 
cell synapse in the rat hippocampus. Brain Res 1975; 86:205-215. 
Drapeau E, Mayo W, Aurousseau C, Le Moal M, Piazza PV, Abrous DN. Spatial 
memory performances of aged rats in the water task predict levels of 
hippocampal neurogenesis. Proc Natl Acad Sci USA 2003; 100:14385-14390. 
Driscoll I, Epp J, Hamilton DA, Sutherland RJ. Transverse patterning 
discrimination problem: Configural learning in the young, middle-aged, and 
elderly adults. 32nd Annual Meeting of the International Neuropsychological 
Society Abstract, 2004a. 
Driscoll I, Hamilton DA, Petropoulos H, Yeo RA, Brooks WM, Sutherland RJ. The 
aging hippocampus: cognitive, biochemical, and structural findings. Cereb 
Cortex 2003a; 13:1344-1351. 
Driscoll I, Hamilton DA, Yeo RA, Brooks WM, Sutherland RJ. Virtual navigation in 
humans: the impact of age, sex, and hormones on place learning. Horm Behav 
2005; 427:326-335. 
Driscoll I, Yeo RA, Petropoulos H, Brooks WM, Sutherland RJ. The polymorphism 
of Apolipoprotein E in normal aging: Behavioral, biochemical, and structural 
measures of the hippocampus. Society for Neuroscience Abstract, vol. 29, 
2003b. 
Driscoll I, Sutherland RJ, Prusky GT, Rudy JW. Damage to the hippocampal 
formation does not disrupt representational flexibility as measured by a novelty 
transfer task. Behav Neurosci 2004b; 118:1427-1432. 
Dusek JA, Eichenbaum H. The hippocampus and transverse patterning guided by 
olfactory cues. Behav Neurosci 1998; 112:762-771. 
Eichenbaum H. The hippocampal system and declarative memory in animals. J 
Cogn Neurosci 1992; 4:217-231. 
Eichenbaum H. The hippocampus and declarative memory: cognitive 
mechanisms and neural codes. Behav Brain Res 2001; 11:709-711. 
Ende GR, Laxer KD, Knowlton RC, Matson GB, Schuff N, Fein G, Weiner MW. 
Temporal lobe epilepsy: bilateral hippocampal metabolite changes revealed at 
proton MR spectroscopic imaging. Radiology 1997; 202:809-817. 
178 
Eriksson PS, Perfilieva E, Bjork-Eriksson T, Alborn AM, Nordborg C, Peterson DA, 
Gage FH. Neurogenesis in the adult human hippocampus. Nature Med 1998; 
4:1313-1317. 
Esiri M. Dementia and normal aging: neuropathology. In: Huppert FA, Brayne C, 
O'Connor DW, eds. Dementia and normal aging. Cambridge, UK: Cambridge 
University Press, 1994; 385-429. 
Evans DA, Beckett LA, Field TS, Feng L, Albert MS, Bennett DA, Tycko B, 
Mayeux R. Apolpoprotein E epsilon 4 and incidence of Alzheimer's Disease in a 
community population of older people. J A Med Assoc 1997; 277:822-824. 
Fanselow MS. Learning theory and neuropsychology: Configuring their disparate 
elements in the hippocampus. J Exp Psychol - An Behav Proces 1999; 25:275-
283. 
Fanselow MS. Contextual fear, gestalt memories, and the hippocampus. Behav 
Brain Res 2000; 110:73-81. 
Frank MJ, Rudy JW, O'Reilly RC. Transitivity, Flexibility, Conjunctive 
Representations, and the Hippocampus. II. A Computational Analysis. 
Hippocampus 2003; 13:299-312. 
Frankland PW, Cestari V, Filipkowski RK, McDonald R, Silva A. The dorsal 
hippocampus is essential for context discrimination, but not for contextual 
conditioning. Behav Neurosci 1998; 112:863-875. 
Frick KM, Baxter MG, Markowska AL, Olton DS, Price DL. Age-related spatial 
reference and working memory deficits assessed in the water task. Neurobiol 
Aging 1995; 16:149-160. 
Fukuzako H, Hashiguchi T, Sakamoto Y, Okamura H, Doi W, Takenouchi K, 
Takigawa M. Metabolite changes with age measured by proton magnetic 
resonance spectroscopy in normal subjects. Psychiatry Clin Neurosci 1997; 
51:261-263. 
Gaffan D. Scene specific memory for objects - a model of episodic memory 
impairment in monkeys with fornix transection. J Cogn Neurosci 1994; 6:305-
320. 
Gage FH, Chen KS, Buszaki G, & Armstrong D. Experimental approaches to age-
related cognitive impairments. Neurobiol Aging 1988; 9:645-655. 
Gage FH, Dunnett SB, Bjorklund A. Spatial learning and motor deficits in aged 
rats. Neurobiol Aging 1984; 5:43-48. 
Gallagher M, Bostock E, King R. Effects of opiate antagonists on spatial memory 
in young and aged rats. Behav Neural Biol 1985; 44:374-385. 
179 
Gallagher M, Burwell R, Burchinal M. Severity of spatial impairment in aging: 
Development of a learning index for performance in the Morris water task. 
Behav Neurosci 1993; 107:618-626. 
Gallagher M, Holland PC. Preserved configural learning and spatial-learning 
impairment in rats with hippocampal damage. Hippocampus 1992; 2:81-88. 
Gallagher M, Pelleymounter MA. Spatial learning deficits in old rats: A model for 
memory deficits in the aged. Neurobiol Aging 1988; 9:549-556. 
Gallagher M, Rapp PR. The use of animal models to study the effects of aging on 
cognition. Ann Rev Psychol 1997; 48:339-370. 
Garry PJ, Goodwin JS, Hunt WC, Hooper EM, Leonard AG. Nutritional status in 
healthy elderly population: Dietary and supplemental intakes. Am J Clin Nutr 
1982; 36:319-331. 
Garry PJ, Hunt WC, Koehler KM, VanderJagt DJ, Vellas BJ. Longitudinal study of 
dietary intakes and plasma lipids in healthy elderly men and women. Am J Clin 
Nutr 1992; 55:682-688. 
Geinisman Y, deToledo-Morrell L, Morrell F, Heller RE. Hippocampal markers of 
age-related memory dysfunction: behavioral, electrophysiological and 
morphological perspectives. Prog Neurobiol 1995; 45:223-252. 
Gianoulakis C. Rats exposed prenatally to alcohol exhibit impairment in spatial 
navigation test. Behav Brain Res 1990; 36:217-228. 
Goldman SA. Adult neurogenesis: From canaries to the clinic. J Neurobiol 1998; 
36:267-286. 
Golomb J, de Leon MJ, Kluger A, George AE, Tarshish C, Ferris SH. Hippocampal 
atrophy in normal aging: An association with recent memory impairment. Arch 
Neurol 1993; 50:967-973. 
Gould E, Beylin A, Tanapat P, Reeves AJ, Shors TJ. Hippocampal-dependent 
learning enhances the survival of granule neurons generated in the dentate 
gyrus of adult rats. Integrat Physiol Behav Sci 1999a; 34:120-121. 
Gould E, McEwen BS, Tanapat P, Galea LAM, Fuchs E. Neurogenesis in the 
dentate gyrus of the adult three shrew is regulated by psychosocial stress and 
NMDA receptor activation. J Neurosci 1997; 17:2492-2498. 
Gould E, Reeves AJ, Fallah M, Tanapat P, Gross CG, Fuchs E. Hippocampal 
neurogenesis in Old World primates. Proc Natl Acad Sci USA 1999b; 96:5263-
5267. 
180 
Gould E, Tanapat P, Hastings NB, Shors TJ. Neurogenesis in adulthood: a 
possible role in learning. Trends Cogn Neurosci 1999c; 3:186-192. 
Gould E, Tanapat P, McEwen BS, Flugge G, Fuchs E. Proliferation of granule cell 
precursors in the dentate gyrus of adult monkeys is diminished by stress. Proc 
Natl Acad Sci USA 1998; 95:3168-3171. 
Griffiths DP, Clayton NS. Testing episodic memory in animals: A new approach. 
Physiol Behav 2001; 73:755-762. 
Hamilton DA, Sutherland RJ. Blocking in human place learning: evidence from 
virtual navigation. Psychobiology 1999; 27:453-461. 
Hamilton DA, Driscoll I, Sutherland RJ. Human place learning in a virtual Morris 
water task: some important constraints on the flexibility of place navigation. 
Behav Brain Res 2002; 129:159-170. 
Hamilton DA, Kodituwakku P, Sutherland RJ, and Savage DD. Children with Fetal 
Alcohol Syndrome are impaired at place learning but not cued-navigation in a 
virtual Morris water task. Behav Brain Res 2003; 143:85-94. 
Haughey NJ, Nath A, Chan SL, Borchard AC, Rao MS, Mattson MP. Disruption of 
neurogenesis by amyloid beta-peptide, and perturbed neural progenitor cell 
homeostasis, in models of Alzheimer's disease. J Neurochem 2002; 83:1509-
1524. 
Hebda-Bauer EK, Morano MI, Therrien B. Aging and corticosterone injections 
affect spatial learning in Fischer-344 x Brown Norway rats. Brain Res 1999; 
827:93-103. 
Hirsch R. The hippocampus and contextual retrieval of information from 
memory. A theory. Behav Biol 1974; 12, 421-444. 
Hugg JW, Kuzniecky RI, Gilliam FG, Morawetz RB, Fraught RE, Hetherington HP. 
Normalization of contralateral metabolic function following temporal lobectomy 
demonstrated by 1H magnetic resonance spectroscopic imaging. Ann Neurol 
1996; 40:236-239. 
Jacobs WJ, Thomas KGF, Laurance HE, Nadel L. Place learning in virtual space 
II. Topographical relations as one dimension of stimulus control. Learn Motiv 
1998; 29:288-308. 
Jarrard LE. Selective hippocampal lesions and behavior: Effects of kainic acid 
lesions on performance of place and cue tasks. Behav Neurosci 1983; 97:873-
889. 
Jenkins BG, Klivenyi P, Kustermann E, Andreassen OA, Ferrante RJ, Rosen BR, 
Beal MF. Nonlinear decrease over time in N-acetyl aspartate levels in the 
181 
absence of neuronal loss and increases in glutamine and glucose in transgenic 
Huntington's disease mice. J Neurochem 2000; 74:2108-2119. 
Jernigan TA, Archibald SL, Berhow MT, Sowell ER, Foster DS, Hesselink JR. 
Cerebral structure on MRI. Part I: Localization of age related changes. Biol 
Psychiatry 1991; 29:55-67. 
Jessberger S, Kempermann G. Adult-born hippocampal neurons mature into 
activity-dependent responsiveness. Eur J Neurosci 2003; 18:2707-2712. 
Jin KL, Peel AL, Mao XO, Xie L, Cottrell BA, Henshall DC Greenberg DA. 
Increased hippocampal neurogenesis in Alzheimer's disease. Proc Natl Acad Sci 
USA 2004; 101:343-347. 
Jin KL, Sun YJ, Xie L, Batteur S, Mao XO, Smelick C, Logvinova A, Greenberg 
DA. Neurogenesis and aging: FGF-2 and HB-EGF restore neurogenesis in 
hippocampus and subventricular zone of aged mice. Aging Cell 2003; 2:175-
183. 
Jung RE, Brooks WM, Yeo RA, Chiulli SJ, Weers DC, Sibbitt WL. Biochemical 
markers of intelligence: a proton MR study of normal human brain. Proc R Soc 
Lond B Biol Sci 1999; 266:1375-1379. 
Jung RE, Chiulli SJ, Friedman SD, Brooks WM., Yeo RA. Magnetic Resonance 
Spectroscopy of brain N-acetylaspartate predicts neuropsychological 
performance following traumatic brain injury. Arch Clin Neuropsychology 1998; 
13:144-144. 
Kalra S, Cashman N, Genge A, Arnold D. Recovery of N-acetylaspartate in 
corticomotor neurons of patients with ALS after riluzole therapy. Neuroreport 
1998; 9:1757-1761. 
Kaplan MS, Hinds JW. Neurogenesis in the adult rat: electron microscopic 
analysis of light radioautographs. Science 1977; 197:1092-1094. 
Kempermann G, Gast D, Gage FH. Neuroplasticity in old age: Sustained fivefold 
induction of hippocampal neurogenesis by long-term environmental enrichment. 
Ann Neurol 2002; 52:135-143. 
Kempermann G, Kuhn HG, Gage FH. More hippocampal neurons in adult mice 
living in an enriched environment. Nature 1997; 386:493-495. 
Kempermann G, Kuhn HG, Gage FH. Experience-induced neurogenesis in the 
senescent dentate gyrus. J Neurosci 1998; 18:3206-3212. 
Kempermann G, Gast D, Gage, FH. Neuroplasticity in old age: Sustained fivefold 
induction of hippocampal neurogenesis by long-term environmental enrichment. 
Ann Neurol 2002; 52:135-143. 
182 
Kempermann G, Wiskott L, Gage FH. Functional significance of adult 
neurogenesis. Cur Op Neurobiol 2004; 14:186-191. 
Kirasic KC, Allen GC. Aging, spatial performance and spatial competence. In: 
Clarness N, ed. Aging and Human Performance. New York: Wiley, 1985; 191-
224. 
Kleim JA, Barbay S, Cooper NR, Hogg TM, Reidel CN, Remple MS, Nudo RJ. 
Motor learning-dependent synaptogenesis is localized to functionally reorganized 
motor cortex. Neurobiol Learn Mem 1996; 77: 63-77. 
Kornack DR, Rakic P. Continuation of neurogenesis in the hippocampus of the 
adult macaque monkey. Proc Natl Acad Sci USA 1999; 96:5768-5773. 
Kuhn HG, Dickinson-Anson H, Gage FH. Neurogenesis in the dentate gyrus of 
the adult rat: age-related decrease of neuronal progenitor proliferation. J 
Neurosci 1996; 16:2027-2033. 
Laurance HE, Thomas KGF, Newman KC, Kaszniak AW, Nadel L, Jacobs JW. 
Older adults map novel environments but do not place learn: Findings from a 
computerized spatial task. Aging Neuropsychol Cog 2002; 9:85-97. 
Lee B, English JA, Paul IA. LP-BM5 infection impairs spatial working memory in 
C57BL/6 mice in the Morris water task. Brain Res 2000; 856:129-134. 
Lie DC, Song H, Colamarion SA, Ming G, Gage FH. Neurogenesis in the adult 
brain: New strategies for central nervous system diseases. Annu Rev Pharmacol 
Toxicol 2004; 44:39-421. 
Lindner MD, Gribkoff VK. Relationship between performance in the Morris water 
task, visual acuity, and thermoregulatory function in aged F-344 rats. Beh Brain 
Res 1991; 45:45-55. 
Lipton SA. Immune activation and neuronal injury in AIDS: Reply. Trends 
Neurosci 1992; 15:75-79. 
Lynch MA, Voss KL. Membrane arachidonic acid concentration correlates with 
age and induction of long-term potentiation in the dentate gyrus of the rat. Eur J 
Neurosci 1994; 6:1008-1014. 
Mackintosh NJ. A theory of attention: Variation in the associabilty of stimuli with 
reinforcement. Psychol Rev 1975: 82:276-298. 
Maguire EA, Burgess N, Donnett JG, Frackowiak RSJ, Frith CD, O'Keefe J. 
Knowing where and getting there: A human navigation network. Science 1998; 
280:921-924. 
183 
Maguire EA, Burgess N, O'Keefe J. Human spatial navigation: cognitive maps, 
sexual dimorphism, and neural substrates. Curr Opin Neurobiol 1999; 9:171-
177. 
Maguire EA, Frackowiak RS, Frith CD. Learning to find your way - a role for the 
human hippocampal region. Proc R Soc Lond B Biol Sci 1996; 263:1745-50. 
Maguire EA, Frackowiak RS, Frith CD. Recalling routes around London: activation 
of the right hippocampus in the taxi drivers. J Neurosci 1997; 17:7103-10. 
Maguire EA, Gadian DG, Johnsrude IS, Good CD, Ashburner J, Frackowiak RS, & 
Frith CD. Navigation-related structural change in the hippocampi of taxi drivers. 
Proc Natl Acad Sci USA 2000; 97:4398-403. 
Maren S. Neurobiology of Pavlovian fear conditioning. Ann Rev Neurosci 2001; 
24:87-931. 
Maren S, Aharonov G, Fanselow MS. Neurotoxic lesions of the dorsal 
hippocampus and Pavlovian fear conditioning in rats. Beh Brain Res 1997; 
88:261-274. 
Martin SJ, Grimwood PD, Morris RGM. Synaptic plasticity and memory: An 
evaluation of the hypothesis. Ann Rev Neurosci 2000; 23:649-711. 
Markowska AL. Sex dimophism in the rate of age-related decline in spatial 
memory: relevance to alterations in the estrous cycle. J Neurosci 1999; 
19:8122-33. 
McDonald RJ. Multiple combinations of co-factors produce variants of age-related 
cognitive decline: A theory. Can J Exp Psycol 2002; 56:221-239. 
McGahon BM, Clements MP, Lynch MA. The ability of aged rats to sustain long-
term potentiation is restored when the aged-related decrease in membrane 
arachidonic acid concentration is reversed. Neuroscience 1997; 81:9-16. 
Meaney MJ, Aitken DH, Bhatnagar S, Sapolsky RM. Postnatal handling 
attenuates certain neuroendocrine, anatomical, and cognitive dysfunction 
associated with aging in female rats. Neurobiol Aging 1991; 12: 31-38. 
Merrill DA, Karim R, Darraq M, Chiba A, Tuszynski MH. Hippocampal cell genesis 
does not correlate with spatial learning ability in aged rats. J Comp Neurol 
459:201-207. 
Metzler J, Shepard RN. Transformational studies of the internal representations 
of three-dimensional objects. In: Solso RL, ed. Theories of cognitive 
psychology: The Loyola symposium. Hillsdale, NJ: Earlbaum, 1974; 147-201. 
184 
Milner B. Amnesia following operation on temporal lobe. In: Whitty CWM, 
Zangwill OL, eds. Amnesia. London, UK: Butterworth Press, 1966; 109-133. 
Minematsu K, Li L, Fischer M, Sotak CH, Davis MA, Fiandaca MS. Diffusion-
weighted magnetic resonance imaging: rapid and quantitative detection of focal 
brain ischemia. Neurology 1992; 42:235-240. 
Mishkin M, Malamut B, Bachevalier J. Memories and habits: Two neural systems. 
In: Lynch G, McGaugh L, Weinberger NM, eds. Neurobiology of learning and 
memory. New York: Guilford, 1984; 65-77. 
Mjaatvedt AE, Wong-Riley MTT. Effects of unilateral climbing fiber 
deafferentation on cytochrome-oxidase activity in the developing rat cerebellum. 
Develop Biol 1991; 145:219-230. 
Moffat SD, Hampson E, Hatzipantelis M. Navigation in a "virtual" maze: sex 
differences and correlation with psychometric measures of spatial ability in 
humans. Evol Hum Behav 1998; 19:73-87. 
Moffat SD, Resnick SM. Effects of age on virtual environment place navigation 
and allocentric cognitive mapping. Behav Neurosci 2002; 116:851-859. 
Moffat SD, Zonderman AB, Resnick SM. Age differences in spatial memory in a 
virtual environment navigation task. Neurobiol Aging 2001; 22:787-796. 
Moffett JR, Namboodiri MA, Cangro CB, Neale JH. Immunohistochemical 
localization of N-acetylaspartate in rat brain. Neuroreport 1991; 2:131-134. 
Morris RGM. Spatial localization does not require the presence of local cues. 
Learn Motiv 1981; 12:239-260. 
Morris RGM. Modeling amnesia and the study of memory in animals. Trends 
Neurosci 1983; 6:479-483. 
Morris RGM, Garrud P, Rawlins JNP, O'Keefe J. Place navigation impaired in rats 
with hippocampal damage. Nature 1982; 297:681-683. 
Morris RGM, Anderson E, Lynch GS, Baudry M. Selective impairment of learning 
and blockade of long-term potentiation by an N-methyl-D-aspartate receptor 
anatgonist, AP5. Nature 1986; 319: 774-776. 
Moser E, Moser MB, Andersen P. Spatial learning impairment parallels the 
magnitude of dorsal hippocampal lesions, but is hardly present following ventral 
lesions. J Neurosci 1993; 13:3916-3925. 
Moser MB, Moser EI. Functional differences in the hippocampus. Hippocampus 
1998; 8:608-619. 
185 
Murray CA, Lynch MA. Evidence that increased hippocampal expression of 
cytokine interleukin-l|3 is a common trigger for age and stress-induced 
impairments in long-term potentiation. J Neurosci 1988a; 8:2974-2981. 
Murray CA, Lynch MA. Dietary supplementation with vitamin E reverses the age-
related deficit in long-term potentiation in dentate gyrus. J Biol Chem 1988b; 
273:12161-12168. 
Nadel L, Moscovitch M. The hippocampal complex and long-term memory 
revisited. Trends Cog Sci 2001; 5:228-230. 
Naressi A, Couturier C, Devos JM, Janssen M, Mangeat C, de Beer R, Graveron-
Demilly D. Java-based Graphical User Interface for the MRUI Quantitation 
Package. Magn Reson Material Phys Biol Med 2001; 12:141-152. 
Nicholls DG. Release of glutamate, aspartate, and _-aminobutyric acid from 
isolated nerve terminals. J Neurochem 1989; 52:331-41. 
Nilsson M, Perfilieva E, Johansson U, Orwar O, Eriksson PS. Enriched 
environment increases neurogenesis in the adult rat dentate gyrus and improves 
spatial memory. J Neurobiol 1999; 39:569-578. 
Norman AB, Bertram KJ, Thomas SR, Pratt RG, Samaratunga RC, Sanberg PR. 
Magnetic resonance imaging of rat brain following in vivo disruption of the 
cerebral vasculature. Brain Res Bull 1991; 26:593-597. 
O'Keefe J, Conway DH. Hippocampal place units in freely moving rat-why they 
fire where they fire. Exp Brain Res 1978; 31:573-590. 
O'Keefe J, Dostrovsky J. The hippocampus as a spatial map. Preliminary 
evidence from unit activity in the freely moving rat. Brain Res 1971; 34:171-
175. 
O'Keefe J, Nadel L. The hippocampus as a cognitive map. Oxford, UK: Clarendon 
Press, 1978. 
O'Reilly RC, Rudy JW. Conjunctive representations in learning and memory: 
Principles of cortical and hippocampal function. Psychol Rev 2001; 108, 311-
345. 
Olson JE, Katz-Stein A, Reo NV, Jolesz FA. Evaluation of acute brain edema 
using quantitative magnetic resonance imaging: effects of pretreatment with 
dexamethasone. Magn Reson Med 1992; 24:64-74. 
Olton DS, Samuelson RJ. Remembrance of places passed: Spatial memory in 
rats. J Exp Psychol: Anim Behav Process 1976; 2:97-116. 
186 
Osteen WK, Spencer RL, Bare DJ, McEwen BS. Analysis of severe photoreceptor 
loss and Morris water-maze performance in aged rats. Behav Brain Res 1995; 
68:151-158. 
Packard MG, White NM. Dissociation of hippocampus and caudate nucleus 
memory systems by posttraining intracerebral injections of dopamine agonists. 
Behav Neurosci 1991; 105:295-306. 
Paxinos G, Watson C. The rat brain in stereotaxic coordinates. New York: 
Academic Press, 1986. 
Pavlakis SG, Lu D, Frank Y, Wiznia A, Eidelberg D, Barrnett T, Hyman RA. Brain 
lactate and N-acetylaspartate in pediatric AIDS encepalopathy. Am J Neuroradiol 
1998; 119:383-385. 
Pelleymounter MA, Smith MY, & Gallagher M. Spatial learning impairments in 
aged rats trained with a salient configuration of stimuli. Psychobiology 1987; 
15:248-254. 
Perlmutter M, Metzger R., Nezworski T, Miller K. Spatial and temporal memory in 
20 and 60 year olds. J Gerontol 1981; 36:59-65. 
Petersen RC, Jack CR, Smith GE, Warning SC, Ivnik RJ. MRI in the diagnosis of 
Mild Cognitive Impairment and Alzheimer's disease. J Int Neuropsychol Soc 
1998; 4:22. 
Petersen RC, Jack CR, Xu YC, Waring SC, O'Brien PC, Smith GE, Ivnik RJ, 
Tangalos EG, Boeve BF, Kokmen E. Memory and MRI-based hippocampal 
volumes in aging and AD. Neurology 2000; 54:581-587. 
Pierce JM, Hall G. A model for Pavlovian learning - variations in the effectiveness 
of conditioned but not of unconditioned stimuli. Psych Rev 1980; 87:532-552. 
Poe BH, Linville C, Riddle DR, Sonntag WE, Brunso-Bechtold JK. Effects of age 
and insulin-like growth factor-1 on neuron and synapse numbers in area CA3 of 
hippocampus. Neuroscience 2001; 107:231-238. 
Prusky GT, West PWR, Douglas RM. Behavioral assessment of visual acuity in 
mice and rats. Vision Res 2000a; 40:2201-2209. 
Prusky GT, West PWR, Douglas RM. Reduced visual acuity impairs place but not 
cued learning in the Morris water task. Beh Brain Res 2000b; 116:135-140. 
Prusky GT, Harker KT, Douglas RM, Whishaw IQ. Variation in visual acuity within 
pigmented, and between pigmented and albino rat strains. Behavioral Brain 
Research 2002; 1 36:339-348. 
187 
Prusky GT, Alam NM, Beekman S, Douglas RM. Rapid quantification of adult and 
developing mouse spatial vision using a virtual optomotor system. Investig 
Ophthal Vis Sci 2004; 45:4611-4616. 
Rapp P, Rosenberg R, & Gallagher M. An evaluation of spatial information 
processing. Behav Neurosci 1987; 101:3-12. 
Rawlins JNP, Olton DS. The septohippocampal system and cognitive mapping. 
Behav Brain Res 1982; 5:331-358. 
Raz N. Aging of the Brain and Its Impact on Cognitive Performance: Integration 
of Structural and Functional Findings. In: Craik FIM, Salthouse TA, eds. 
Handbook of Aging and Cognition. New Jersey: Erlbaum, 2000; 1-90. 
Reed JM, Squire LR. Impaired transverse patterning in human amnesia is a 
special case of impaired memory for two-choice discrimination tasks. Behav 
Neurosci 1999; 113:3-9. 
Rescorla RA, Wagner AR. A theory of Pavlovian conditioning: Variations in the 
effectiveness of reinforcement and nonreinforcement. In: Black AH, Prokasy WF, 
eds. Classical conditioning II: Current research and theory. New York: Appleton-
Century-Crofts, 1972; 64-99. 
Richmond MA, Yee BK, Pouzet B, Veenman L, Rawlins JNP, Felden J, Bannerman 
DM. Dissociating context and space within the hippocampus: Effects of complete 
dorsal, and entral excitotoxic hippocampal lesions on conditioned freezing and 
spatial learning. Behav Neurosci 1999; 113:1189-1203. 
Rickard TC, Grafman J. Losing their configural mind: Amnesic patients fail on 
transverse patterning. J Cogn Neurosci 1998; 10:509-524. 
Rondi-Reig L, Libbey M, Eichenbaum H, Tonegawa S. CAl-specific N-methyl-D-
aspartate receptor knockout mice are deficient in solving non-spatial transverse 
patterning task. Proc Natl Acad Sci USA 2001; 98:3543-3548. 
Ross RT, Orr WB, Holland PC, Berger TW. Hippocampectomy disrupts acquisition 
and retention of learned conditional responding. Behav Neurosci 1984; 98:211-
225. 
Rosene DL, Van Hoesen GW. Hippocampal efferents reach widespread areas of 
cerebral cortex and amygdala in rhesus-monkey. Science 1977; 198:315-317. 
Rosenzweig ES, Barnes CA. Impact of aging on hippocampal function : 
plasticity, network dynamics, and cognition. Prog Neurobiol 2003; 69:143-179. 
Rudy JW, Sutherland RJ. Configural association theory and the hippocampal 
formation: an appraisal and reconfiguration. Hippocampus 1995; 5:375-389. 
188 
Rudy JW, Keith JR, Georgen K. The effect of age in childrens learning of 
problems that require a configural association solution. Dev Psychobiol 1993; 
26:171-184. 
Rudy JW, O'Reilly RC. Conjunctive representations, the hippocampus, and 
contextual fear conditioning. Cog Affect Behav Neurosci 2001; 1:66-82. 
Rudy JW, Huff N, Matus-Amat P. Understanding contextual fear conditioning: 
Insights from a two process model. Neurosci Biobehav Rev 2004;28, 675-686. 
Ryan L, Nadel L, Keil K, Putnam K, Schnyer D, Trouard T, Moscovitch M. The 
hippocampal complex and retrieval of recent and very remote autobiographical 
memories: Evidence from functional magnetic resonance imaging in 
neurologically intact people. Hippocampus 2001; 11:707-714. 
Sara SJ. Memory deficits in rats with hippocampal or cortical lesions: retrograde 
effects. Behav Neural Biol 1981; 32:504-509. 
Saunders JK, Smith IC, MacTavish JC, Rydzy M, Peeling J, Sutherland E, Lesiuk 
H, Sutherland GR. Forebrain ischemia studied using magnetic resonance imaging 
and spectroscopy. NMR Biomed 1989; 2:312-316. 
Schuff N, Marmar CR, Weiss DS, Neylan TC, Schoenfeld F, Fein G, Weiner MW. 
Reduced hippocampal volume and N-acetylaspartate in post-traumatic stress 
disorder. Ann N Y Acad Sci 1997a; 821:516-520. 
Schuff N, Amend D, Ezekiel F, Steinman SK, Tanabe J, Norman D, Jagust W, 
Kramer JH, Mastrianni JA, Fein G, Weiner MW. Changes of hippocampal N-
acetylaspartate and volume in Alzheimer's disease. A proton MR spectroscopic 
imaging and MR study. Neurology 1997b; 49:1513-1521. 
Schuff N, Amend DL, Knowlton R, Norman D, Fein G, Weiner MW. Age-related 
metabolite changes and volume loss in the hippocampus by magnetic resonance 
spectroscopy and imaging. Neurobiol Aging 1999; 20:279-285. 
Scoville WB, Milner B. Loss of recent memory after bilateral hippocampal lesions. 
J Neurol Neurosurg Psychiatry 1957; 20:11- 21. 
Seilhean D, Duyckaerts C, Vazeux R, Bolgert F, Brunet P, Katlama C, Gentilini M, 
Hauw JJ. HIV-l-associated cognitive motor complex - absence of neuronal loss 
in the cerebral neocortex. Neurology 1993; 43:1492-1499. 
Shingo T, Sorokan ST, Shimazaki T, Weiss S. Erythropoietin regulates the in 
vitro and in vivo production of neuronal progenitors by mammalian forebrain 
neural stem cells J Neurosci 2001; 21:9733-9743. 
189 
Shore DI, Stanford L, Maclnnes WJ, Klein RM, Brown RE. Of mice and men: 
Virtual Hebb-Williams mazes permit comparison of spatial learning across 
species. Cognitive, Affective, and Behav Neurosci 2001; 1:83-89. 
Snyder JS, Kee N, Wojtowicz JM. Effects of adult neurogenesis on synaptic 
plasticity in the rat dentate gyrus. J Neurophys 2001; 85:2423-2431. 
Sokoloff L. Energetics of functional activation in neural tissues. Neurochem Res 
1999; 24:321-29. 
Spence KW. The nature of the response in discrimination learning. Psychol Rev 
1952; 59:89-93. 
Squire LR. Memory and the hippocampus: A synthesis from findings with rats, 
monkeys, and humans. Psychol Rev 1992; 99:195-231. 
Stanfield BB, Trice JE. Evidence that granule cells generated in the dentate 
gyrus of adult rats extend axonal projections. Exp Brain Res 1988; 72:399-406. 
Sterio DC. The unbiased estimation of number and sizes of arbitrary particles 
using the disector. J Microsci 1984; 134:127-136. 
Sullivan EV, Marsh L, Mathalon DH, Lim KO, Pfefferbaum A. Age-related decline 
in MRI volumes of temporal lobe gray matter but not hippocampus. Neurobiol 
Aging 1995; 16:591-606. 
Sutherland RJ, Chew GL, Baker JC, Linggard RC. Some limitations on the use of 
distal cues in place navigation by rats. Psychobiology 1987; 15:48-57. 
Sutherland RJ, Kolb B, Whishaw IQ. Spatial mapping: definitive disruption by 
hippocampal or medial frontal cortical damage in the rat. Neurosci Lett 1982a; 
31:271-276. 
Sutherland RJ, Rudy JW. Configural association theory: the role of the 
hippocampal formation in learning, memory, and amnesia. Psychobiology 1989; 
17:129-144. 
Sutherland RJ, Weisend MP, Mumby D, Astur RS, Hanlon FM, Koerner A, Thomas 
MJ, Wu Y, Moses SN, Cole C, Hamilton DA, Hoesing JM. Retrograde amnesia 
after hippocampal damage: Recent vs. remote memories in two tasks. 
Hippocampus 2001; 11:27-42. 
Sutherland RJ, Whishaw IQ, Regehr JC. Cholinergic receptor blockade impairs 
spatial localization by use of distal cues in the rat. J Comp Physiol Psychol 
1982b; 96:563-573. 
190 
Tatebayashi Y, Lee MH, Li L, Iqbal K, Grundke-Iqbal I. The dentate gyrus 
neurogenesis: a therapeutic target for Alzheimer's disease. Acta Neuropathol 
2003; 105:225-232. 
Tattersfield AS, Croon RJ, Liu YW, Kells AP, Faull RLM, Connor B. Neurogenesis 
in the striatum of the quinolinic acid lesion model of Huntington's disease. 
Neuroscience 2004; 127:319-332. 
Tielen AM, Mollevanger WJ, Lopes de Silva FH, Hollander CF. Neuronal plasticity 
in hippocampal slices of extremely old rats. In: Gispen WH, Traber J (Eds.). 
Aging of the Brain. Elsevier, Amsterdam 1983; 73-84. 
Tulving E. Episodic and semantic memory - Where should we go from here. 
Behav Brain Sci 1986; 9:573-577. 
Tyler TJ, DiScenna P. The hippocampus memory indexing theory. Behav 
Neurosci 1986; 100:147-152. 
Urenjak J, Williams SR, Gadian DG, Noble M. Proton nuclear magnetic resonance 
spectroscopy unambiguously identifies different neural cell types. J Neurosci 
1993; 13:981-989. 
Van der Staay FJ. Assessment of age-associated cognitive deficits in rats: a 
tricky business. 
Neurosci Biobeh Rev 2002; 26:753-759. 
Van der Staay FJ, de Jonge M. Effects of age on water escape behavior and on 
repeated acquisition in rats. Behav Neural Biol 1993; 60:33-41. 
Van Elzakker M, O'Reilly RC, Rudy JW. Transitivity, flexibility, conjunctive 
representations, and the hippocampus. I. An empirical analysis. Hippocampus 
2003; 13: 334-340. 
van Praag H, Kempermann G, Gage FH. Running increases cell proliferation and 
neurogenesis in the adult mouse dentate gyrus. Nat Neurosci 1999; 2:266-270. 
van Praag H, Schnider AF, Christie BR, Toni N, Palmer TD, Gage FH. Functional 
neurogenesis in the hippocampus. Nature 2002; 415:1030-1034. 
Vandenberg S, Kuse AR. Mental rotations: A group test of three-dimensional 
spatial visualization. Percept Mot Skills 1978; 47:599-604. 
Vijg J. Somatic mutations and aging: a re-evaluation. Mutat Res 2000; 447:117-
35. 
Wagner JP, Black IB, DiCicco-Bloom E. Stimulation of neonatal and adult brain 
neurogenesis by subcutaneous injection of basic fibroblast growth factor. J 
Neurosci 1999; 19:6006-6016. 
191 
Walker JA, Olton DS. Spatial memory deficit following fimbria-fornix lesions 
independent of time for stimulus processing. Physiol Bhav 1979; 23:11-15. 
Wang R, Dineley KT, Sweatt JD, Zheng H. Presenilin 1 familial Alzheimer's 
disease mutation leads to defective associative learning and impaired adult 
neurogenesis. Neuroscience 2004; 126; 2:305-312. 
Weisend M, Astur R, Sutherland RJ. The specificity and temporal characteristics 
of retrograde amnesia after hippocampal lesions. Soc Neurosci Abstract 1996; 
22:1118. 
West MJ. Regionally specific loss of neurons in the aging hippocampus. Neurobiol 
Aging 1993; 14:287-293. 
Wickelgren WA. Chunking and consolidation: a theoretical synthesis of semantic 
networks, configuring in conditioning, S-R versus cognitive learning, normal 
forgetting, the amnesic syndrome, and the hippocampal arousal system. Psychol 
Rev 1979; 86:44-60. 
Winocur G. Anterograde and retrograde amnesia in rats with dorsal hippocampal 
or dorsomedial thalamic lesions. Behav Brain Res 1990; 38:145-154. 
Winocur G, McDonald RM, Moscovitch M. Anterograde and retrograde amnesia in 
rats with large hippocampal lesions. Hippocampus 2001; 11:18-26. 
Wolf OT, Dyakin V, Patel A, Vadasz C, de Leon MJ, McEwen BS, Bulloch K. 
Volumetric structural magnetic resonance imaging (MRI) of the rat hippocampus 
following kainic acid (KA) treatment. Brain Research 2002; 934:87-96. 
Wong-Riley MTT. Cytochrome oxidase - An endogenous metabolic marker for 
neuronal activity. Trend Neurosci 1989; 12:94-101. 
Wyss JM, Chambless BD, Kadish I, van Groen T. Age-related decline in water 
task learning and memory in rats: strain differences. Neurobiol Aging 2000; 
12:671-81. 
Zaharia MD, Kulczycki J, Shanks N, Meaney MJ, Anisman H. The effects of early 
postnatal stimulation on water-maze acquisition in adult mice: genetic and 
maternal factor. Psychopharmacology 1996; 128:227-39. 
192 
Group Condition Training Phase 
1 2 3 
ED PRE-LESION A+B- A+B- A+B-
C+D- C+D-
E+F-
POST-LESION (TEST) A+B-
C+D-
E+F-
POST-LESION (RETRAIN) A+B- A+B- A+B-
C+D- C+D-
E+F-
POST-LESION (RETEST) A+B-
(10-DAY DELAY TEST) C+D-
E+F-
POST-LESION G+H- G+H- G+H-
(NEW PROBLEM) I+J- I+J-
K+L-
TP PRE-LESION X+Y- X+Y- X+Y-
Y+Z- Y+Z-
Z+X-
POST-LESION (TEST) X+Y-
Y+Z-
Z-X+ 
POST-LESION (RETRAIN) X+Y- X+Y- X+Y-
Y+Z- Y+Z-
Z+X-
Table 1. Discrimination training procedures are shown for each group (ED = elemental 
discriminations; TP = transverse patterning), and divided by the phases they were trained 
in. 
193 
A. 
VOLUME NAA NEURON 
DENSITY 
SYNAPTIC 
DENSITY 
MITOCHONDRIAL 
DENSITY 
NEUROGENESIS 
MWT r = -.378 
p = .047 X X X X 
r = - .543 
p = .003 
TPDT r = .488 
p = .008 X X X X 
r = .481 
p = .011 
B. 
VOLUME NAA NEURON 
DENSITY 
SYNAPTIC 
DENSITY 
MITOCHONDRIAL 
DENSITY 
NEUROGENESIS 
MWT X 
(P = -174) 
X X X X r = - .543 p = .003 
TPDT X 
(p = .166) 
X X X X r = .481 p = .011 
Table 2. (A) Correlations between behavioral measures (MWT ad TPDT) and 
hippocampal volume, neuronal density, synaptic density, mitochondrial 
density, and neurogenesis. Both, reduction in volume and neurogenesis 
predict learning and memory deficits on hippocampus-dependent tasks. (B) 
Partial correlations, controlling for neurogenesis (DCX-labeled) cells, suggest 
that the age-related reduction in neurogenesis primarily accounts for 
hippocampus-dependent tasks learning and memory deficits and mediates 
the relationship with hippocampal volume. 
194 
MONITORS 
PLATFORM 
DIVIDER — 
Figure 1. The Visual Water Task. Top view showing major components, 
including a trapezoidal-shaped pool, midline divider, platform, and computer 
monitors that display either reinforced (+) or non-reinforced (-) stimuli into 
the pool through a transparent wall. 
195 
I Middle-aged 
BOId 
Figure 2. Acuity of middle-aged (14 months old) and old (26 months old) 
FBNF1 rats. The grating threshold acuity of Middle-aged FBNF1 rats clustered 
around 1.54 c/d, which was significantly higher than that of the aged rats (1.36 
c/d). 
196 
0.556 
0.554 
0.552 
0.55 
& 
^ 0.548 o 
v 0.546 
3 
£ 0.544 -
LI-
li 0.542 
J- 0.54 
0.538 
0.536 
0.534 
A. 
iijii 
• Young 
FL MIHRLLE-AGPH 
• OLD 
0.555 n 
0.55 
•o Q. 
o_ 
o 0.545 c 
3 
2 
± 0.54 
TO a 
w 
0.535 
0.53 
B. 
• Young 
B Middle-aged 
EJOId 
ZL 
Figure 3 A & B. The grating acuity as measured by the OptoMotor test. Old (26 
months) FBNF1 rats had a significantly lower acuity compared to their young (4 
moths) and middle-aged (14 months) counterparts when measured in both 
clockwise (A) and counterclockwise (B) directions. 
197 
Figure 4. Electrophysiological recordings. Rats were implanted unilaterally with 
a stimulating electrode in the perforant path (A) and a recording (B) electrode in 
the dentate gyrus. 
198 
0.4-1 r —i 1 1 i 1 1 1 i 1—i 
10123456789 10 
1 
0.8 
0.6 
0.4 
0.2 
0 
-0.2-
-0.4 
i I 
—i 1 i—i 1 1 i 1 1 i i 1 
-1 0123456789 10 
4 i 
i i 1 1 i i -1 i i—i 1 1 1 i i 1 1 1 1 
-1 0123456789 10 -1 0123456789 10 
Figure 5. Effects of 10 HFS sessions on the evoked potentials of middle-aged 
(white diamonds) and old (black squares) rats as measured 1 hour (A and C) 
and 24 hours (B and D) following stimulation. HFS resulted in a long-lasting 
increase in the population spike amplitude and the EPSP slope, an effect that 
could be observed both 1 and 24 hours following stimulation. However, middle-
aged rats showed a more pronounced increase in both the population spike 
amplitude (A & B) and the EPSP slope (C & D) compared to the old rats, (p < 
.05 for all comparisons). 
199 
A B C D E 
Baseline 1 hour 
post 
24 hours 
post 
1 hour 
post 
24 hours 
post 
First HFS session Tenth HFS session 
Figure 6. Effects of HFS on hippocampal field potentials in a representative 
middle-aged (top; A, B, C, D & E) and old (bottom; F, G, H, I & J) rat. The 
effects of the 1st and 10th session of HFS on field potentials of middle-aged and 
old rats 1 hour and 24 hours post stimulation are shown. In both groups, HFS 
induced a significant increase in population spike amplitude which was 
observable both 1 hour and 24 hours following the first HFS session. Relative to 
the old rats, the magnitude of change was enhanced in the middle-aged animals 
across the HFS sessions. (Calibration: vertical= 2 mV, horizontal 2 ms.) 
200 
A. 12- -•-Young 6- -t- Young 
10-
1 
0\ 
-o- Middle-aged 5 - -•- Middle-aged I 
r 8-
-6--OLD { 
B. 
F. 
Figure 7. Performance on the hidden platform trials of the Morris water task. 
Performance was averaged across days when the platform was in a new location 
(A, C, and E) and days when the platform remained in the same location (B, D, 
and F). Group (age) differences in latency (top) and path length (middle) to 
reach the platform, and in the speed of swimming (bottom) is presented. No 
clear pattern of deficits emerged in the acquisition of spatial information and the 
speed of swimming was comparable between groups. 
201 
Figure 8. Retention scores representing the memory for the platform location 
experienced 24 hours earlier. Both middle-aged and old rats had trouble 
retaining the information related to the platform location learned on the previous 
day compared to the young rats. 
202 
0 -I 1 1 1 i 1 1 1 
1 2 3 4 5 6 7 8 
Trial 
Figure 9. Performance on the visible platform trials of the Morris water task. 
Group (age) differences in latency (A) and path length (B) to reach the 
platform, and in the speed of swimming (C) across trials. Old rats can reach the 
same latencies and path lengths traveled to reach the visible platform as the 
young and middle-aged rats, which is especially evident on trials 4-8 as the rats 
become more familiar with the task. The speed of swimming is more variable, 
but they are capable of reaching the same speed of swimming as their younger 
counterparts. 
203 
Figure 10. The discriminanda comprising elemental and transverse patterning 
problems. Elemental problems are presented in Sets 1 and 2 respectively, and 
the transverse patterning problem is presented in Set 3. 
204 
Figure 11. A photomicrograph of representative NMDA damage through dorsal 
(A.) and ventral (B.) hippocampal formation in coronal sections. 
205 
• Pre-lesion 
• Post-lesion test 
12 3 
Problem 
• Post-lesion retraining 
• 10-day delay 
• Post lesion-New 
problem 
1 2 
Problem 
Figure 12. Performance on the elemental discrimination problems before and 
after the damage to the hippocampus. Rats with damage to the hippocampal 
formation exhibit retrograde amnesia for elemental discriminations learned prior 
to surgery (A). Rats with damage to the hippocampal formation do not exhibit 
anterograde amnesia for simple discriminations - they can re-learn the set of 
elemental problems they were trained on before the damage, learn a new set of 
elemental problems, and retain it over a 10-day delay without the hippocampus 
(B). (p's < 0.05). 
206 
c 
0 
'Z 
0) 
o 
0 
0 
re 
1. 
h 
• Pre-lesion 
• Post-lesion Retraining 
1 Post-lesion New Problem 
2 
Problem 
Figure 13. Learning of elemental discriminations occurs at a faster rate after 
the hippocampal damage. Rats require fewer trials to reach criterion when 
learning elemental discriminations after the damage to the hippocampus, 
regardless whether they are the same discriminations that were learned prior to 
damage (dark gray) or whether they are a completely new set of discrimination 
problems (diagonal gray stripes), (p's < 0.05). 
207 
1 2 3 
Problem 
Figure 14. Rats with hippocampal damage exhibit both retrograde and 
anterograde amnesia for transverse patterning discriminations. When tested 
after the damage, rats did not remember the transverse patterning problem 
they learned prior to damage (dark gray) nor were they able to re-learn the 
problems (diagonal gray stripes), (p's < 0.05). 
208 
•Young 
• Middle-aged 
iOld 
1 2 3 
Problem 
Figure 15. Discriminating between three concurrent elemental problems on the 
test day. All rats were able to learn elemental discrimination problems and 
perform with 90% accuracy or above regardless of their age. Mean and SEM of 
% correct for each problem by age group, (p > .7). 
209 
Problem 
Figure 16. Transverse patterning discrimination performance. Mean and SEM of 
the % correct on each of the problems on the test day for each age group. 
Young and middle-aged rats performed similarly and had no trouble learning the 
transverse patterning problem, while the old rats were significantly impaired, (p 
< .001). 
210 
Figure 17. A representative structural MR image (T2) of the dorsal 
hippocampus of the rat seen here in a coronal plane. Hippocampal area is 
outlined in white. Structural images were used to obtain a measure of 
hippocampal volume. 
211 
Figure 18. Magnetic Resonance Spectroscopy (MRS) acquisition. Placement 
of the left (L) and right (R) voxel (2x2x3) in the hippocampus in the (A) axial 
and (B) coronal plane. 
212 
Figure 18B. A representative raw spectrum from the hippocampus at 9.4 T. 
Three prominent metabolites. Choline, Creatine, and NAA, are of particular 
interest here. 
CO 
E 
0) 
E 
O 
> 
2250 n 
2000 -
1750 -
1500 
• Young 
• Middle-aged 
moid 
Figure 19. Age-related differences in intracranial volume (ICV). Mean and SEM 
for ICV are presented for each age group. ICV was significantly smaller in the 
young rats compared to the middle-aged and old rats (p < .001). 
214 
•Young 
• Middle-; 
• Old 
90 i 
Figure 20. Age-related differences in absolute hippocampal volume. Mean and 
SEM for the absolute volume of the hippocampus are presented for each age 
group. Absolute hippocampal volume was significantly smaller in the middle-
aged and old rats compared to the young rats (p = .031). 
215 
Figure 21. Age-related differences in the hippocampal volume normalized to 
ICV. Mean and SEM for the normalized hippocampal volume are presented for 
each age group. Normalized hippocampal volume was significantly smaller in the 
middle-aged and old rats compared to the young rats (p = .003). 
216 
Figure 22. Semi-thin (1 u,m) sections were obtained from a tissue block 
stained with toluidine blue for orientation and trimming. Originally, blocks 
contained the dentate gyrus. The dentate gyrus was then trimmed such that 
each ultra-thin section spanned the inner molecular layer as well as the very 
dorsal portion of the granule cell layer (see arrow pointing to a light blue 
rectangle). A series of ultra-thin section for electron microscopy were 
obtained from the trimmed blocks. 
217 
Figure 23. Electron 
micrograph depicting a larger 
area surveyed under an 
electron microscope (top). 
Bottom left picture is an 
enlargement of the synapse 
surrounded by a white box in 
top picture, and depicts a 
typical flat synapse (including 
post-synaptic densities). 
Figure 24. An example of 
mitochondria seen under the 
electron microscope. Top 
picture represents a larger 
area surveyed under the 
electron beam. 
Bottom left represents a close-
up view of a cluster of 
mitochondria enclosed with a 
white square in the top picture. 
219 
Figure 25. Changes in neuronal density in the inner molecular layer (IML) of 
the hippocampus. Mean and SEM for neuronal density within the IML. Neuronal 
density was similar in the young and middle-aged rats, but was significantly 
lower in the old rats (p = .019). 
220 
Figure 26. Newly born neurons in the dentate gyrus. A confocal image of 
neurons in the granule cell layer double-labeled against antibodies for BrdU 
(red) and NeuN (green). Five weeks after the last injection of BrdU, cells 
double-labeled for BrdU and NeuN can be detected. This image shows 
(consistent with previous findings) that many BrdU/NeuN-labeled cells, 
indistinguishable from the surrounding granule cells can be found 5 weeks 
after BrdU administration. White arrow is pointing to a 'gold' cell, a cell that 
is showing an immunoreaction for NeuN in a BrdU-labeled cell. 
221 
Figure 27. Doublecortin labeling in the dentate gyrus of the hippocampus 
(young animal shown). DCX-labeled cells are restricted to the subgranular 
zone (SGZ). DCX-expressing cells often have long processes that extend into 
the molecular layer. The black arrow is pointing to three DCX-labeled cells 
situated in the SGZ along the ventral horn of the dentate gyrus with long 
processes extending into the molecular layer. 
222 
Figure 28. Newly born neurons expressing Ki67 along the subgranular zone 
in the dentate gyrus of a representative young rat. Ki67 is expressed in 
cycling cells and here is fluorescing in red along the dentate gyrus. 
223 
•Young 
• Middle-aged 
•Old 
Figure 29. The number of BrdU labeled cells. Mean and SEM by age group for 
the total number of BrdU-labeled cells present 5 weeks after the BrdU 
administration. The number of BrdU labeled cells significantly differed between 
each age group (p = .002). 
224 
Figure 30. The number of BrdU labeled cells that are also co-labeled with NeuN 
(a mature neuronal marker). Mean and SEM by age group for the total number 
of BrdU/NeuN double-labeled cells present 5 weeks after the BrdU 
administration. The number of BrdU/NeuN double-labeled cells significantly 
differed between each age group and decreased with age (p = .001). 
225 
Figure 31. The number of cells labeled with DCX, a marker of migrating 
neuroblasts (immature neurons). Mean and SEM by age group for the total 
number of DCX-labeled cells. The number of DCX-labeled cells significantly 
differed between each age group and decreased with age (p < .001). 
226 
•Young 
• Middle-aged 
•Old 
Figure 32. The number of cells labeled with Ki67, a marker for cycling cells. 
Mean and SEM by age group for the total number of Ki67-labeled cells. The 
number of Ki67-labeled cells decreased with age, such that both middle-aged 
and old rats had a significantly lower number of Ki67+ cells, but the two 
groups did not differ from one another (p < .001). 
227 
2 2.5 3 3.5 4 4.5 5 5.5 
nHPC volume 
Figure 33. Correlations between performance of young (black circles), middle-
aged (open squares), and elderly (gray triangles) rats on the MWT with 
normalized hippocampal (nHPC) volume. Larger hippocampal volume was 
correlated with shorter approach latencies on the first trial across the days the 
platform remained in the same location as the previous day (p = .047). 
228 
2 2.5 3 3.5 4 4.5 5 5.5 
nHPC volume 
Figure 34. Correlations between performance of young (black circles), middle-
aged (open squares), and elderly (gray triangles) rats on the transverse 
patterning task with normalized hippocampal (nHPC) volume. Larger 
hippocampal volume was correlated with better performance on Problem 3 of 
the transverse patterning task. 
229 
Figure 35. Correlations between performance of young (black circles), middle-
aged (open squares), and elderly (gray triangles) rats on the MWT with the 
number of DCX-labeled cells. Shorter latencies in locating the platform the last 
trial across the days the platform was in a new location significantly correlated 
with a larger number of DCX-labeled cells. However, this relationship was 
significant in the senescent rats (middle-aged and old combined; r = - .534, p = 
.027), but in the young group (and it is in the opposite direction; r = .506, p = 
.136). 
230 
Figure 36. Correlations between performance of young (black circles), middle-
aged (open squares), and elderly (gray triangles) rats on the transverse 
patterning task with the number of DCX-labeled cells. Better performance on 
Problem 3 of the transverse patterning task significantly correlated with a larger 
number of DCX-labeled cells. However, the relationship between DCX and 
performance on Problem 3 of the transverse patterning task was significant in 
the elderly (middle-aged and old together; see Figure 36; r = .805, p < .001), 
but not in the young rats (r = .180, p = .619). 
231 
nHP
C volum
e 
6 i 
5 -
4 • 
3" 
2 -
1 • 
n 
A • 
r = .605 
p = .001 
• 
• • 
• Young 
• Middle-aged 
A Old 
( 100 200 300 
DCX 
400 500 600 
Figure 37. Correlations between the normalized hippocampal (nHPC) volume of 
young (black circles), middle-aged (open squares), and elderly (gray triangles) 
rats with the number of DCX-labeled cells. Higher number of DCX-labeled cells 
was correlated with larger hippocampal volume (p = .001). 
232 
Figure 38. Correlations between the number of DCX-labeled cells in young 
(black circles), middle-aged (open squares), and elderly (gray triangles) rats 
and the number of BrdU/NeuN double-labeled cells. Higher number of DCX-
labeled cells was correlated with a higher number of BrdU/NeuN double-labeled 
cells (p = .001). 
233 
Figure 39. Correlations between performance of young (black circles), middle-
aged (open squares), and elderly (gray triangles) rats on the transverse 
patterning task with the number of Ki67-labeled cells. Better performance on 
Problem 3 of the transverse patterning task significantly correlated with a higher 
number of Ki67-labeled cells (p = .022). 
234 
6 i 
r = .599 
p = .002 
• Young 
• Middie-aged 
AOId 
20 40 60 80 
Ki67 
100 120 140 
Figure 40. Correlations between the normalized hippocampal (nHPC) volume 
of young (black circles), middle-aged (open squares), and elderly (gray 
triangles) rats with the number of Ki67-labeled cells. Higher number of Ki67-
labeled cells was correlated with larger hippocampal volume (p = .002). 
235 
Figure 41. Correlations between the number of DCX-labeled cells in young 
(black circles), middle-aged (open squares), and elderly (gray triangles) rats 
and the number of Ki67-labeled cells. Rats with a higher number of DCX-labeled 
cells also had a higher number of Ki67-labeled cells (p < .001). 
236