xml version 1.0 encoding UTF-8 standalone no
record xmlns http:www.loc.govMARC21slim xmlns:xsi http:www.w3.org2001XMLSchema-instance xsi:schemaLocation http:www.loc.govstandardsmarcxmlschemaMARC21slim.xsd
leader nam Ka
controlfield tag 001 001912372
007 cr mnu|||uuuuu
008 071008s2005 flu sbm 000 0 eng d
datafield ind1 8 ind2 024
subfield code a E14-SFE0001262
Cracchiolo, Jennifer R.
Dissecting out the contribution of cognitive, social, and physical activities to environmental enrichment's ability to protect Alzheimer's mice against cognitive impairment
h [electronic resource] /
by Jennifer R. Cracchiolo.
[Tampa, Fla] :
b University of South Florida,
ABSTRACT: Retrospective studies suggest that lifestyle activities may provide protection against Alzheimer s Disease (AD). However, such studies can be inaccurate and prospective longitudinal studies investigating lifestyle protection against AD are both impractical and impossible to control for. Transgenic (Tg+) AD mice offer a model in a well controlled environment for testing the potential for environmental factors to impact AD development. In an initial study, Tg+ and non-transgenic (Tg-) mice were housed in either environmentally enriched (EE) or standard housing (SH) from 2-6 months of age, with a behavioral battery given during the last 5 weeks of housing. In the Morris maze, platform recognition, and radial arm water maze tasks, Tg+/EE mice were completely protected from cognitive impairment present in Tg+/SH mice and comparable to control Tg-/SH mice in cognitive performance. The current study utilized the same cognitive-based behavioral battery and multimetric statis tical analysis to investigate the protective effects of "complete" environment enrichment (EE) versus several of its components (physical activity, social interactions) in AD transgenic mice. The AD transgenic mice utilized develop beta-amyloid (AB) deposition and cognitive impairment by 6-7 months of age. Similar to our initial study, results show that "complete" EE (physical, social, and cognitive activities) from 2 to 8 months of age completely protected AD transgenic mice from cognitive impairment in tasks representing different cognitive domains working memory, reference learning, and search/recognition. In strong contrast, Tg+ mice reared in environments that included physical activity and social interaction, or only social interaction, were not protected from cognitive impairment in adulthood --^ enhanced cognitive activity was required over and above that present in these other environments. Through use of discriminant function analysis, EE and/or NT mice were consistently discriminated from the poorer performing other housing groups. The cognitive benefits observed in EE-housed Tg+ mice occurred without significant changes in cortical AB levels, plasma cytokine levels, or plasma corticosterone levels, suggesting involvement of mechanisms independent of these endpoints. However, EE-housed Tg+ mice did have decreased dendritic length of neurons in the parietal cortex (but not hippocampus). Noteworthy is that plasma cytokine levels and hippocampal dendritic length consistently correlated with cognitive measures, suggesting their involvement in underlying mechanisms of cognitive performance. The present work provides the first evidence that "complete" EE (including enhanced cognitive activity) is needed to provide cognitive protection against AD in a Tg+ model of the disease while the physical and social activity components of EE do not alone lead to protection. These results suggest that humans desiring to gain maximal environmental protection against AD should live a lifestyle high in cognitive, social, and physical activities together.
Thesis (M.A.)--University of South Florida, 2005.
Includes bibliographical references.
Text (Electronic thesis) in PDF format.
System requirements: World Wide Web browser and PDF reader.
Mode of access: World Wide Web.
Title from PDF of title page.
Document formatted into pages; contains 167 pages.
Adviser: Gary Arendash, Ph.D.
Radial arm water maze.
t USF Electronic Theses and Dissertations.
Dissecting Out the Contribution of Cognitive, Social, and Physical Activities to Environmental EnrichmentÂ’s Ability to Pr otect AlzheimerÂ’s Mice Against Cognitive Impairment by Jennifer R. Cracchiolo A thesis submitted in partial fulfillment of the requirements for the degree of Master of Science Department of Arts and Sciences College of Biology University of South Florida Major Professor: Gary W. Arendash, Ph.D. Fredrick Essig, Ph.D. Huntington Potter Ph.D. Date of Approval: July 29, 2005 Keywords: Memory, Radial Arm Water Maze Behavior, Neurodegenerative Diseases, Amyloid Copyright 2005, Jennifer R. Cracchiolo
i Table of Contents List of Tables................................................................................................................. ....iii List of Figures................................................................................................................ ....iv Abstract....................................................................................................................... ........v AlzheimerÂ’s Disease Background.......................................................................................1 Behavior Associated with AlzheimerÂ’s Disease.............................................................2 Pathology Associated with AlzheimerÂ’s Disease...........................................................5 -Amyloid & Neurofibrillary Tangles........................................................................5 Neuronal Loss, Synaptic Loss, Brain Atrophy.........................................................11 Genetics of AlzheimerÂ’s Disease..................................................................................14 Amyloid Precursor Protein.......................................................................................15 The Presenilins..........................................................................................................16 Apoliprotein..............................................................................................................17 Risks of Alzheimer Â’s Disease......................................................................................17 Diagnosis of AlzheimerÂ’s Disease................................................................................21 Treatment of AlzheimerÂ’s disease................................................................................27 Animal Models of AlzheimerÂ’s Disease...........................................................................31 The PDAPP Mouse Model of AlzheimerÂ’s Disease.....................................................32 PDAPP Pathology.....................................................................................................33 Behavior of PDAPP Model.......................................................................................37 APP Swedish Mutation Transgenic Mouse Line..........................................................41 APPsw Brain Pathology............................................................................................41 APPsw Behavior.......................................................................................................44 APP Swedish + PS1 Mutation Transgenic Mouse Model............................................49 APPsw +PS1 Pathology............................................................................................49 Behavior of APP+PS1 Mice.....................................................................................53 APP Mutant Mice: Imperfect Transgenic Model..........................................................58 Environmental Enrichment...............................................................................................61 Enriched Environment Protection and Treatment: Human Studies..............................61 Effects of Enriched Environment in Rodents...............................................................65 Cognitive Effects of Enriched Environment in Rodents...........................................65
ii Neurohistologic and Neurochemical Effect s of Environmental Enrichment in Rodents.....................................................................................................................69 Effects of Environmental Enrichment in Â“AlzheimerÂ’sÂ” Transgenic Mice..............75 Complete Environmental Enrichme nt versus Its Components.....................................82 Specific Aims.................................................................................................................. ..86 Material and Methods.......................................................................................................88 Animals & General Protocol.........................................................................................88 Enriched Environments.................................................................................................89 Behavioral Assessment.................................................................................................90 Tissue and Blood Collection.........................................................................................94 Corticosterone Quantification.......................................................................................95 Extraction of Brain Protein for Sandwic h Enzyme-Linked Immunosorbent Assay (ELISA)........................................................................................................................ .95 Blood Cytokine Levels.................................................................................................97 Golgi Cox Analysis.......................................................................................................98 Statistical Analysis........................................................................................................99 Results........................................................................................................................ .....101 Behavioral Analysis....................................................................................................101 Neuropathologic Measures.........................................................................................112 Factor Analysis/ Discriminant Function Analysis......................................................116 Correlation Analysis...................................................................................................122 Discussion..................................................................................................................... ..126 General Summary.......................................................................................................126 Behavioral Measures...................................................................................................131 Neuopathologic Analysis............................................................................................135 Correlations Analyses between Neuropa thology, Blood Measures and Behavior.....139 Multimetric Analyses (FA and DFA).........................................................................141 Reference...................................................................................................................... ..146
iii List of Tables Table 1. Factor loading of behavioral measures, with or without pathologic measuresÂ…Â…Â…Â…Â…Â…Â…Â…Â…Â…Â…Â…Â…Â…Â…Â…Â…Â…Â…Â…Â…Â…Â…Â…Â…Â…Â…Â…Â…Â…..118 Table 2. Summary of discriminant functi onal analyses of beha vioral measuresÂ…Â…Â…120 Table 3. For all animals, a correlation ma trix of behavioral measures and plasma cytokine levelsÂ…Â…Â…Â…Â…Â…Â…Â…Â…Â…Â…Â…Â…Â…Â…Â…Â…Â…Â…Â…Â…Â…Â…Â…Â…Â…Â…Â….123 Table 4. For all animals, a correlation matr ix of behavioral measures and dendritic length/branchingÂ…Â…Â…Â…Â…Â…Â…Â…Â…Â…Â…Â…Â…Â…Â…Â…Â…Â…Â…Â…Â…Â…Â…Â…Â…Â…Â…..125
iv List of Figures Figure 1. General protocol time line for enrichment studyÂ…Â…Â…Â…Â…Â…Â…Â…Â…...Â…...89 Figure 2. Y-maze Entries and Percent Spontanous AlternationsÂ…Â…Â…Â…Â…Â…...Â…....103 Figure 3. Circular Platform overall escape latenciesÂ…Â…Â…Â…Â…Â…Â…Â…Â…Â…...Â….Â…..104 Figure 4. Morris Water Maze acquisition overall and across 5 two-day blocksÂ…........105 Figure 5. Morris Water Maze Memory Retention, as indexed by time spend in the former platform-containing quadran t, annulus crossings, and quadrant preferenceÂ…Â…Â…Â…Â…Â…Â…Â…Â…Â…Â…Â…Â…Â…Â…Â…Â…Â…Â…Â…Â…Â…Â…Â…Â…....Â…Â….Â…...107 Figure 6. Recognition/Identification pe rformance in Platform Recognition testing overall and for each of the 4 days of testingÂ…Â…Â…Â…Â…Â…Â…Â…Â…Â…Â…Â…Â…....108 Figure 7. RAWM overall and final block esca pe latencies for Trials 1, 4 and 5Â….......110 Figure 8. RAWM escape latencies for final block Trials 4 and 5Â…Â…Â…Â…Â…Â…Â…Â…..111 Figure 9. Plasma Corticosterone leve ls by treatment (housing), genotype, and genderÂ…Â…Â…Â…Â…Â…Â…Â…Â…Â…Â…Â…Â…Â…Â…Â…Â…Â…Â…Â…Â…Â…Â…Â…Â…Â…Â…Â….Â…...113 Figure 10. Standardized mean signal in tensities for 10 plasma cytokines in APP+PS1 and NT miceÂ…Â…Â…Â…Â…Â…Â…Â…Â…Â…Â…Â…Â…Â…Â…Â…Â…Â…Â….Â…Â…Â…Â…..114 Figure 11. Quantification of Soluble A (1-40 and 1-42) and Insoluble A (1-40 and 1-42) in APP+PS1 mice within different housing environmentsÂ…Â…Â…Â…Â…Â…Â…Â…Â…Â…Â…Â…Â…Â…Â…Â…Â…Â…Â…Â…Â…Â…Â…Â…Â…..Â…Â…Â…..115 Figure 12. Housing and transgenedependent changes in dendritic length/branching of neurons from APP+PS1 and NT miceÂ…Â…Â…Â…Â…Â…Â…Â…Â…Â…Â…117
v Dissecting Out the Contribution of Cognitive, Social, and Physical Activities to Environmental EnrichmentÂ’s Ab ility to Protect AlzheimerÂ’s (transgenic) Mice Against Cognitive Impairment Jennifer R.Cracchiolo Abstract Retrospective studies suggest that life style activities may provide protection against Alzheimer s Disease (AD). However, such studies can be inaccurate and prospective longitudinal studies investigati ng lifestyle protection against AD are both impractical and impossible to control for. Transgenic (Tg+) AD mice offer a model in a well controlled environment for testing the poten tial for environmental factors to impact AD development. In an initial study, Tg+ and non-transgenic (Tg-) mice were housed in either environmentally enriched (EE) or sta ndard housing (SH) from 2-6 months of age, with a behavioral battery given during the last 5 weeks of housing. In the Morris maze, platform recognition, and radial arm water maze tasks, Tg+/EE mice were completely protected from cognitive impairment present in Tg+/SH mice and comparable to control Tg-/SH mice in cognitive performance. Th e current study utilized the same cognitivebased behavioral battery and mu ltimetric statistical analysis to investigate the protective effects of Â“completeÂ” environment enrichment (EE) versus several of its components (physical activity, social in teractions) in AD transgenic mice. The AD transgenic mice
vi utilized develop beta-amyloid (A deposition and cognitive impairment by 6-7 months of age. Similar to our ini tial study, results show that Â“comp leteÂ” EE (physical, social, and cognitive activities) from 2 to 8 months of age completely protected AD transgenic mice from cognitive impairment in tasks representing different cognitive domains working memory, reference learning, and search/recognition. In strong contrast, Tg+ mice reared in environments that include d physical activity and social interaction, or only social interaction, were not protected from cogni tive impairment in adulthood Â– enhanced cognitive activity was required ove r and above that pres ent in these other environments. Through use of discriminant function analys is, EE and/or NT mice were consistently discriminated from the poorer performing other housing groups. The cognitive benefits observed in EE-housed Tg+ mice occurred with out significant changes in cortical A levels, plasma cytokine levels, or plasma co rticosterone levels, s uggesting involvement of mechanisms independent of these endpoints. However, EE-housed Tg+ mice did have decreased dendritic length of neurons in the parietal co rtex (but not hippocampus). Noteworthy is that plasma cytokine levels and hippocampal dendritic length consistently correlated with cognitive measures, sugge sting their involvement in underlying mechanisms of cognitive performance. The present work provides the first evidence that Â“completeÂ” EE (including enhanced cognitive activity) is needed to provide cognitive protection against AD in a Tg+ model of th e disease, while the physical and social activity components of EE do not alone lead to protection. These re sults suggest that humans desiring to gain maximal envir onmental protection against AD should live a lifestyle high in cognitive, social and physical activities together.
1 AlzheimerÂ’s Disease Background With projected life expectancy hypo thesized to reach 80 by 2050, emphasis on health of the aging population has become a centralized area of research. With achievement of such great ages comes an in crease in health burdens, in particular neurodegenerative disorders. Of these numerous neurodegenerative disorders, AlzheimerÂ’s disease (AD) has provoked imme nse interest in biomedical, media, and public health arenas. Prevalence of AD, wh ich rises linearly with age and accounts for 81% of dementia, effects 10% of all individua ls over the age of 65 and 40% over 85 years of age (Evans et al., 1989). AlzheimerÂ’s disease has grand implications including an enormous monetary burden on health care systems as well as the loss of livelihood of patients and their family members. Statistics suggest that cases of AD will increase significantly in the future, solidifying the importance of research into viable prevention and treatment options for the aging population. The first AD patient, Auguste D., descri bed by Alois Alzheimer, displayed what are now considered characteristic symptoms of AD including progressive memory loss, loss of social skills, and a decrease in functional language. While Alzheimer observed the cognitive manifestations and basic pathol ogical features of th e disease, it was not until advances in microscopy that hallmark pathological features including neuritic amyloid plaques and neurofibrillary tangles were described in detail. In a time course of 2 to 20 years patients first dwindle cognitively while motor and sensory function remain
2 relatively intact. However, as pathologi cal constituents ravage neuronal systems, phenotypic characteristics of motor diseases begi n to appear in the fo rm of irregular gaits and loss of coordination. At present time only postmortem autopsy reveals absolute diagnosis of AD. The potential effects of interventions to de lay the onset of AD are immense. It is estimated that that there are approximate ly 360,000 new case of AD each year. Perhaps more striking is the implication of delaying ons et of the disease by 5 years. If research findings successfully can delay onset by 5 y ears, a reduction of 1.5 million cases after 10 years and 4.04 million cases after 50 years co uld be witnessed (Brookmeyer et al., 1998). This delay could translate into b illions of savings economically. Enriched lifestyles may offer the me dium to provide this delay of onset. Longitudinal studies state that engagement in mental, social, or pr oductive activities was inversely related to incidences of dementia (Wang et al., 2002). The literature would suggest that stimulating mental and physic al activities may offer protection from dementia. Research into the mechanism of these cognitive be nefits still needs clarification. Future inve stigation into the Â“enriched environment phenomenonÂ” may offer quality of life benefits to patients and caretakers as well as a decrease in financial burden for health care systems. Behavior Associated with AlzheimerÂ’s Disease AD is the most common form of dementia and often takes an archetypical clinical course which reflects the underlying expandi ng neuropathology. The clinical phase of AD, a period of 5-10 years (Locascio et al., 199 5), is characterized by short and long term
3 memory loss, paranoia, delusions, language deficit and cognitive impairment. The clinical sequence of de generative loss in AD has been described as the progression of functional loss of capacity. Th e stages of this degeneration in many ways mirror the normal maturation of human development (Reisberg et al., 1999). Mild cognitive impairment (MCI) has been described as a transition state between typical aging and dementia. Ten to fifteen percent of patients diagnosed with MCI convert to AD annually (Peterse n et al., 2001). When diagnos ing MCI, important criteria include the memory complaint being corr oborated by an informant and second, the memory impairment is appropriate when age and education is cons idered. MCI may be useful in identifying individuals functioning at normal levels who are at risk for later developing AD. Such patients display normal actives of daily living and normal general cognitive function, however, complain of a s hort-term memory (Petersen et al., 1997). MCI has been shown to exclusivity involve short-term memory impairment (Shankle et al,. 2005). While many studies have shown in creased rates of progression of AD with individuals that display MCI, there have been conflicting conclusions regarding why not all individuals identified with MCI progress to AD. One hypothesis suggested by Petersen et al. 1997 is that di agnosed MCI patientÂ’s behavior is more heterogeneous than thought. Patients diagnosed with MCI encompa sses a group of individuals that may have a progressive form of MCI, leading to deme ntia, while others may not progress, simply remaining stagnate within the criteria for MC I. Petersen et al. (2001) find that such heterogeneity requires a subclassification. Sub-groups may include amnestic MCI, MCI with slight memory impairment in multiple do mains, or MCI with impairment of a single
4 non-memory domain. Subdividing MCI may result in a more reliable future diagnosis to patients displaying such symptoms. In the early stages of AD, a significant impairment of working (short-term) and memory is the outstanding clinical feature (F orstl & Kurz, 1999). While memory of the patientÂ’s early years and long term memory remains relatively intact, working memory impairments displayed in early stages of AD l ead to difficulties in daily activities. The loss of a patientÂ’s ability to pl an, judge, and organize leads to a decrease in the patientÂ’s ability to perform household chores. Upon cau sal inspection language may still appear to be intact, however, communication begins to de teriorate due to an attenuated vocabulary. Other impairments may be observed in objec t naming and semantic memory (Chobor et al., 1990). Frequently early AD patients will di splay difficulty with driving due to the loss of spatial orientation. Ea rly, non-cognitive disturbances appear to be variable and mild. AD patients often show signs of depr ession and apathy (Forstl & Kurz, 1999). In early stages of AD motor function remains relatively intact. The moderate stage of AD brings a further deterioration of working memory, forcing the patient to Â“live in the past (Bea tty et al., 1988).Â” Langua ge, logical reasoning, long term memory planning, reading and organizi ng significantly decline in this stage. Sequencing becomes impaired, leading to loss of the ability to feed and dress oneself. Many AD patients develop prosopagnosia making once familiar faces foreign. Restlessness, aggression, disorientation, and incontinence are also often found in the moderate stage of the illness. Psychiatric symptoms include hallucination, wandering and hoarding (Devanand et al., 1997). As motor func tion declines at this moderate AD stage, risk of falls increases due to hesitant gait and stooped pos ture (Forstl & Kurz, 1999).
5 Most all cognitive functions are severely impaired in the late stages of AD (Forstl & Kurz, 1999). Long term memory, presen t in mild and moderate stages of AD, dissipates in the final stages of the illness. Language is reduced to single phrases or words. Many patients are bedridden and en ter a vegetative state. Pneumonia followed by myocardial infarction and se pticaemia are the most frequent cause of death in AD (Forstl & Kurz, 1999) Pathology Associated wi th AlzheimerÂ’s Disease Many major pathological processe s contribute to the widespread brain destruction presented clinically in AD. These processes comprise the two hallmark pathological features of beta-amyloid plaques (A ) and neurofibrillary tangles (NFT), as well as the resultant processe s of inflammation, synaptic loss, neuronal cell loss, and brain atrophy. Each of these components co ntributes to and interre lates with each other in a maze of cause and effect that is s till being elucidated (Rosenthal & Khotianov, 2003). -Amyloid & Neurofibrillary Tangles The Â“ -amyloid hypothesisÂ” of AD is s upported by a substantial amount of research confirming the significance detriment A plaques have on neurons as well as the beneficial effects seen upon prevention and clearance of these pl aques (Loo et al., 1993, Morgan et al., 2000). Loo et al. (1993) reported that apoptosis is induced by A in cultured central nervous system neurons while in an animal model of AD, Morgan et al. (2000) found that by decreasing formation of plaques with an A vaccine, memory loss is
6 prevented. A is generated from the abnormal cleav age of a much larger transmembrane protein called amyloid precursor protein (APP) discussed later. The product of this abnormal cleavage by and -secretase results in A formation within the cytoplasm which is taken up vesicles and released into the extracellular space (Busciglio et al., 1993). Once released, star-shaped compact masses of A plaques begin to form (Selkoe, 2001). Such plaques are often associated with glia cells. Microglia are found within and adjacent to the central amyloid core while the astrocytes surround the outside of the plaque. The time course for the produc tion of the detrimental, dense core A plaques is unknown however it is assumed that the clinical phase of AD is preceded by a 15Â–30 year period of time in which continuous deposition of A plaques takes place. Plaques in the AD brain accumulate in the parietal, tem poral cortex, and most notably in the hippocampus. Immunohistochemical stains have revealed that the dense A plaques described above are not the sole A species found in the AD brain. Such staining using antibodies show a far more extensive number of A deposits than originally observed with congophilic stains. Yamaguchi et al. (1988) described thes e diffuse type of senile plaques not stained by congo red in the br ains of AD type dementia patients. Examination of the A peptide revealed that the A fragment either ends at amino acid 40 or 42. Fragments ending in A 42 and lacking in A 40 comprise the diffuse plaques. In contrast, the compact va riety contains primarily A 40 (Iwatsubo et al., 1995). It has been hypotheisisized that th e diffuse, Â“preamyloid,Â” A 42 plaque is a precursor to the compact dense plaques seen later in the AD brain (Selkoe, 2001). Examining normal
7 aged patients as well as transgenic animal models of AD provides evidence for this theory. The brains of aged, cognitively nor mal humans often display diffuse amyloid deposition however often lack extensive co mpact barnacle like plaques found in AD (Wang et al., 1999). A second support for this theory is that transgenic animals over expressing mutant APP first de velop diffuse plaque before the later development of compact plaques. These observations suggest a preclinical pathology that may be a possible therapeutic target to alter the c ourse of AD progression before behavioral symptoms appear. A is not confined to the neuronal extrac ellular space. Plaques are also found in the vasculature of the AD patient. Cerebral amyloid angiopathy (CAA) is defined as deposition of a congophilic material in mening eal and cerebral arteries and arterioles within the brain. While CAA is commonly pres ent in the aged brai n, 98% of AD patientÂ’s brains exhibit vascular A plaques. CAA is formed when smooth muscle cells synthesize A -40 intracellularly, which aggregates extracellularly into fibrils that then induces smooth muscle degeneration, making the tissues susceptible to hemorrhage. Small infarcts in AD patients may have limited imp act on the global cogniti ve decline observed clinically in moderate to late AD. Ho wever they may influence early stages by promoting an earlier development of de mentia. A second concern of compromised vasoregulation caused by CAA is the cascade of events followed by a decrease in blood flow. Vasoconstriction causing in a decrea se in blood flow will result in capillary damage, therefore impeding glucose transport. A decrease in the brains primary source of energy translates to decrease metabolism and dysfunction of neurons. It has been suggested that beneficial cogni tive effects observed from ad ministration of vasodilators
8 and blood thinners such as such as ginkgo biloba and NSAIDs, respectively, may work through reversing this mechanism (Jellinger, 2002). The presence of A results in an inflammatory cascad e. In an attempt to clear the diffuse plaques, a robust microgl ia activation takes place. Mi croglia cells make an effort to clear debris and cellular damage associated with AD. However, in the process they release free radicals and cytoki nes, the latter of which ac tivates astrocytes (McGeer & McGeer, 1995). The primary effect of microg lia cells is phagocytos is and release of cytokines. Cytokine release at tracts reactive astrocytes. Microglia cells also release dangerous free radicals resulting in oxidat ive damage to surrounding tissue. Both the astrocyteÂ’s generation of reactive oxyge n species (Abramov et al., 2004) and the microglia release of free radicals contribute to an oxidative imbalanc e. Astrocytes also secrete inflammatory proteins, found in pla ques, such as APOE and ACT (Abraham et al., 1988). These inflammatory protei ns encourage the conversion of A from the soluble to aggregated species (Ma et al., 1994). An increase in compact A results in a mechanism of neuronal and syna ptic loss (Yankner et al., 1995). The result of this inflammatory cascade is a primary source of cognitive impairment observed in the clinical symptoms of AD behavior. Braak and Braak (1991a) have desc ribed three stages in the development of amyloid deposition. The first development of pl aques is seen in the basal neocortex and the poorly myelinated temporal areas such as the perirhinal and ectorhinal cortex. A second stage is reveled when the plaques begi n to branch out into adjoining neocortical areas as well as the hippocampus. In the fina l stage, dense deposits decorate all areas of the cortex and extend deep into the inne r and outer pyramidal cell layers of the
9 hippocampal CA1 region. Prevalence of the final stage increases with age, however presence of early stages can be observed in young adults. At a late point a maximum density of A plaques is reached, however gray ma tter still remains relatively free of plaques (Braak & Braak, 1991a). Neurofibrillary tangles (NFT) are a second pathologic feature found in AD. The category of NFTs encompass both the NFTs wh ich develop within th e nerve cell body as well as the neuropil threads (NT) which are found in the distal portions of dendritic processes (Braak & Braak, 1986). In the co rtex, NFTs and NTs are selective for pyramidal cells (Braak & Braak, 1991a). Th ese tangles result when the multifunctional tau protein is hyperphosphorylat ed. Tau proteins are a member of the microtubuleassociated proteins (Weingart en et al., 1974). While these proteins are found mainly in the brain, they are seen in trace amounts in the periphery. The tau protein is involved in the assembly and stabilization of microtubu les; however when th is normal occurring protein is hyperphosphorylated, dimers of th ese phosphp-tau units form insoluble paired helical filament structures th at impair axonal transport, a nd ultimately lead to neuronal death. This cascade of events leads to loss of signal across the axon and neuronal death. The kinase involved in the abnormal phos phorilation is unknow n, however, possible players include Cdlk-5 and Gsk-3 which are overly active in AD (Cruz & Tsai, 2004, Maccioni et al., 2001). Tangle bearing neurons are found in large numbers in the frontal, temporal, and parietal cortex as well as in the hippocampus and amygdale (Selkoe, 2000). Braak and Braak 1991a describe th e evolution of NFTÂ’s as a Â“gradual development of brain destruction which begi ns in a few limbic areas of the cerebral cortex, which then spreads in a predicta ble, nonrandom manner across the hippocampus,
10 the neocortex, and a number of sub cortical nu clei.Â” In a pre-lesi ons AD brain, void of both A plaques and NFTs, abnormally phosphorylat ed tau protein is observed in the transentornhinal cortex (Braak and Braak, 1991a). As neurons develop NFTs in the transentorhinal cortex, criteria for stages I and II of AD neuropat hological changes are meet. It is important to not e that at this stage brain ch anges are not significant enough to produce clinical behavioral symptoms. As NFT formation progresses into stages III and IV, there is striking neurodege neration with in both the entorhinal and transentorhinal regions. While the neocortex remains intact there are mild changes in the hippocampus. It is this stage that first cognitive and persona lity changes are observed. It is important to note that it is not the destruction of the en torhinal and transentorhinal cortices alone which produce the behavioral changes, but instead it is loss of communication between the neocortex and the hippocampus. Personality ch anges can be attribut ed to disruption of the limbic circuit, also seen in stages III and IV. In end stages of AD, NFTs can be observed decorating virtually all subdivisions of the cerebral cortex. Braak and Braak 1991a make particular note of Â“severe destru ction of neocortical a ssociation areas.Â” The complex interaction between A and NFT formation is still being elucidated. NFTs are not solely found in AD. Other neurodegenerative diseases lacking A pathology have NFTs present that are i ndistinguishable from those found in the AD brain. While these two hallmark pathological features can occur independently, Gotz et al., (2001) described a cause and effect inter action between the two. Gotz et al., (2001) observed that injection of A 42 fibrils into the somatosens ory cortex and hippocampus of 5to 6-month-old P301L tau transgenic mice caused a fivefold increases in the numbers of NFTs in cell bodies within the amygdale, from where neurons proj ect to the injection
11 sites. This data would su ggest that presence of A induces NFT formation or at least accelerates its production. Recent work by Oddo et al., (2004) showed that A immunotherapy leads to clearance of both A plaques as well as hyperphosphorylated tau aggregates. The clearance of the two ha llmark lesions were clea red in a hierarchal fashion with the clearance of A plaques preceding the cl earance of hyperphosphorylated tau. Upon return of the lesions, the A plaques returned before the tau associated pathology (Oddo, 2004). This data would provide evidence in support for the Â“amyloid cascade hypothesesÂ” stating that Â“accumulation of A in the brain is the primary influence driving AD pathogenesis, and that th e rest of the diseas e process, including formation of neurofibrillary ta ngles containing tau protein, is proposed to result from an imbalance between A production and A clearance (Hardy, 2002).Â” Neuronal Loss, Synaptic Loss, Brain Atrophy Neuronal loss in the entorhinal co rtex, hippocampus, front al, parietal and temporal cortices has been observed in AD brai ns. Neurons in layer II of the entorhinal cortex and the CA1 region of the hippocam pus are particularly vulnerable to the AD process (Mattson, 2004). The patte rn of neuronal death in AD is similar, to but different from, the pattern see in typical aging (West et al., 1994, Fukutani et al., 1995 ) Degenerative changes found in AD in the CA2, CA3, CA4 and presubiculum overlap changes observed in normal aging however, ne uronal loss seen in entorhinal cortex, CA1, and para and post subiculum, which are generally preserved in the aged hippocampus, die in AD (Fukutani et al., 1995) This would sugge st that AD is not simply an exaggerated aging process. Neurona l death in AD is said to occur in a spatio-
12 temporal pattern (Mattson, 2004). Neuronal death occurs over a long period of time, suggesting that neurons drop off in small nu mbers over the course of the lengthy disease process. The spatio-temporal pattern combined with the relatively small number of neurons dying at any one time would suggest th at apoptosis is invol ved in the neuronal death present in AD. In contrast to apoptos is, necrosis of neuron occurs in large number over a short period of time. Due to the long pre-clinical and clini cal phase and of the disease, death by apoptosis is a likely mechanism. Furthe r evidence includes altered expression of apoptosis-related genes such as Bcl-2 family members, prostate apoptosis response-4 (par-4), DNA response genes, a nd p53, within neurons associated with plaques in AD brains (Mattson, 2000) Othe r evidence for apoptosis in AD includes the presence of APP mutations, A and formation of NFTs are all possible trigger of the programmed cell death cascade (Chan et al., 2002 Mattson, 2004). While apoptosis may play a significant role in neuronal death, a combination of both necrosis and apoptosis most likely contributes to the profile of the AD brain. Synaptic loss is established as a reliable neurobiological correlate of the cognitive deficits associated with AD (DeKosky & Scheff, 1990). Synapses are extremely vulnerable in AD due to there high content of disease-related protein, APP and Presenilins, as well as their high metabo lic and oxidative loads (Mattson, 2004). Emerging evidence suggests that synapses are not functioning optimally even before structural deterioration. While synaptic loss is a conseque nce of neuronal death, Coleman and Yao (2003) have shown evidence that comp romised synapses are observed in living intact neurons in the AD brain. Specifically the function of synapses appears to be compromised. Coleman and colleagues hypothe sis impaired synaptic function may be
13 due disruption in synaptic vesicle traffi cking. Their recent work shows a reduced expression of genes related to synaptic vesi cle trafficking in AD (Coleman & Yao, 2003). Early memory loss in AD patients has been hypothesized to be a synapse dysfunction caused by soluble A oligomers (Lacor et al., 2004). A oligomers, present in AD brains as well as transgenic mouse models of AD, rapidly block LTP (Chen et al., 2000; Walsh et al., 2002). The hypothesis that the AD brain contains functionally altered, however, structurally sound synapses offers hope th at early impairment found in AD may be reversible with new therapeutics ta rgeting this synapse dysfunction. Calcium plays a fundamental role in learning and memory and is also involved in neuronal and synaptic hea lth and function. Disruption to calcium homeostasis in closely involved in the neuronal death obs erved in AD (Verkhratsky and Toescu, 2003). Bathing neurons in excess calcium induces neurotoxicity. Improper regulation of intracellular calcium ions is associated with A (Mattson, 1997) and NFTÂ’s (Saito et al., 1993). A Â’s oxidative stress characte ristics have the ability to impair calcium pumps and increase influx through voltage-dependent Ca++ channels while NFT containing neurons show high amounts of calcium as well as hype ractivity of calcium-d ependent proteases (Mattson, 2004). The disruption in calcium, leading to neuronal and synaptic loss in AD, is the targeted mechanism by which a new class of NMDA receptor antagonist exploits. Cerebral atrophy is observed in the AD brain at autopsy as well as through imaging techniques. Widening sulci and dwi ndling gyri comprise the gross anatomical features of the AD brain. Ventricular dilati on is also observed. Annual rates of global brain atrophy in AD patients is about 2Â–3% co mpared with 02Â–05% in healthy controls
14 (Fox & Schott, 2004). Pennane n et al. (2004) observed brain volume changes in the hippocampus and entorhinal cortex of patie nts with MCI and early stages of AD. Â“Volumes of the hippocampus and entorhinal cortex were significantly reduced in the following order: control > MCI > AD (Pennane n 2004).Â” In the final stages of the disease, A and NFTÂ’s combined with synaptic a nd neuronal loss can be accountable for a substantial loss in total brain volume. Genetics of AlzheimerÂ’s Disease Ten percent of cases of AlzheimerÂ’s disease suggest a familial mode of transmission (Selkoe, 2001). The other 90% of cases have been described as nonfamilial or sporadic cases. Through phenotypi c analysis, using identified AD related genes, the majority of cases appear in the absence of a genetic link. Many investigators hypothesize that a lager percentage of cases will be linked to genetic predisposition as interrelated genes are found. Ph enotypically, familial and sporadic cases of AD are often indistinguishable (Selkoe, 2001). Often th e autosomal dominant form of AD will result in an early age of onset, however progresses as does the late onset sporadic form. The genes contributing to the 10 percent of familial cases have added a great deal to the understanding of AD. Mutations in 4 gene s have been affirmed to increase risk of AD. These include mutations in the amyloid precursor proteins (APP), presenilins 1 and 2 genes (PS1 and PS2) as well as alteration in the apolipoprotein E alleles (ApoE).
15 Amyloid Precursor Protein The finding that both AD and Down Syndrome brains shared A plaques with an identical sequence narrowed the search for a gene involved in familial AD to chromosome 21 (Glenner & Wong, 1984). Th ese findings combined with the later cloning of the APP gene (Tanzi et al., 1987) se t the stage for a new er a in AD research. APP is a transmembrane polypeptide that is translated in the ER and then posttranslationally modified through a secret ory pathway (Selkoe, 2001). The N-terminal moiety of APP is projected towards the extracellular domain while the C-terminal projects into the cytoplasm (Neve et al., 2000). Three majo r isoforms exist including 695, 751, and 770. APP is processed by cellular proteases known as and secretases. The benign cleavage by -secretase at 688 re sults in the release of a large soluble fragment into the extracellular space and in turn prevents the formation of A In contrast, the -secretase cuts at residue 671 rele asing an ectodomain derivative and exposing residue 1 of the A peptide. Previ ous to cleavage by -secretase a constitute cleavage at 711 or 713 by -secretase results in the formation of Ab40 or Ab42 respectively. These detrimental A forming cleavages increase with age and or a genetic mutation (discussed below). The missense mutations of the APP gene are rare causes of FAD. While rare, these mutations represent the first findings of a genetic cause of AD. Mutations of the
16 APP gene are strategically located near know n cleavage sights indicating their effect on proteolytic processing of AD. Families that carry APP missense mutations have AD onset before 65 and as early as 30 (Selkoe 2001). Mullan et al (1992) identified a double mutation at codons 670 and 671 in exon 16 which co-segregated with AD in two large early-onset AD families from Sweden. Ba se pair transversion from G to T and A to C results in amino acid substitutions of Lys to Asn and Met to Leu at 670 and 671 respectively. These mutations located at the N-terminus of the peptide increase the cutting of -secretase resulting in the generation of more A 40 and A 42. Mutations at the C-terminus end of APP selectively increase Ab42. The APP 717 mutation is an example of a single missense mutation found at the C-terminus. This mutation replaces a Val for a Leu and results in the increase of -secretase. An increase in -secretase increases release of A peptide from the larger APP protein, resulting in an increase in production of A 1-42. The Presenilins Due to the genetically heterogeneou s nature of AD, the finding of other autosomal dominant mutations resulting in FAD was a clear objective. Missense mutations of the presenilins provide the most common cause of autosomal dominant cases of AD to date (Selkoe, 2001). The pr esenilin mutations selectively increase secretase cleavage of C99 a nd C83, yielding peptides ending in Ab42. When examining families harboring these mutations, an appr oximately two-fold increase of Ab42 was observed (Scheuner et al., 1996). Currently two genes of inte rest make up the presenilin family. The PS1 missense mutations, found on chromosome 14, have been linked to
17 families with clinical onset of AD in their 40s and 50s and as early as 30s. A possible 75 missense mutations have been found on the PS1 gene (Hardy, 1997). A homologous PS2 gene linked to AD pathogenesis was found in a series of German families originally from the Volga valley in Russia (Levy-La had, 1995). The 3 mutati ons identified on the PS2 gene result in AD with a la ter and more variable age of onset than mutations in PS1 (Hutton, 1997) Apoliprotein While mutations of the APP gene and the presenilins make up a small percentage of AD cases, the presences of the Apoliprotein E allele (ApoE) represents a major genetic predisposition to AD. ApoE, located on chro mosome 21, exists in three variants: ApoE2, ApoE3 and ApoE4. ApoE is involved in the tr ansport of cholesterol and is important in local circuits of lipids turnove r (Veurink et al., 2003). The pr esence of 1 or 2 copies of the ApoE4 species increase the risk of AD (to be discussed later) and decrease age of onset. In contrast to APP and PS mutations, ApoE represents a genetic player of AD not inherited in an autosomal dominant fashion. Presence of the ApoE allele is a genetic factor linked to late-onset AD cases. Thus the presence of ApoE 4 precipitates the disorder primarily in subjects in their 60s and 70s (Selkoe, 2001). Risks of Alzheimer Â’s Disease There are several generally accepted risk factors of AD, while there are many others still under investigation. Unmodifiab le risk factors incl ude genetic mutations (discussed above), family history, and being a female. Head injur y, diet, and education
18 levels are examples of potentially modifiable risk factors. Rising age is the principle risk factor in the development of AD. Of the population aged 65 years and older, the proportion of individuals with AD roughly doubles with each additional 10 years of age. By 85, over 50 % of the population has develope d AD ( NIH, 1995). This increased risk with age peaks at 90 and then proceeds to drop off. First degree re latives of sporadic AD patients, despite no FAD association, have a lifetime risk increased by 3 or 4 fold (Liddell et al., 2001). Females are also twice as likely to develop AD as males (Jorm et al. 1987). Another major unmodifiable risk factor in the development of AD is inheritance of one or two copies of the ApoE4 gene. Po ssession of ApoE 4 alleles increases risk of developing AD in a dose dependa nt fashion. Individuals with two copies have earlier onset than those possessing one allele (Corder et al., 1993). While presence of the ApoE4 allele is not diagnostic of AD, their pres ence accounts for 50% of early-onset AD cases and 20% of late-onset cases (Rosenthal & Khotianov, 2003). The presence of ApoE4 alone causes increase risk in the white popul ation while both ApoE 2 and 4 increase the risk in African Americans (Maestre et al., 1995). While ApoE genotyping is commercially available, it offers little a dvantage as a prognostic indicator. Females carrying the ApoE 4 allele have a 45% probability of developing AD by age 73, while male have only a 25 % probability of de veloping AD (Breitner et al.,1999). The mechanism by which presence of ApoE4 may be increasing risk of AD is still in need of clarification; however several hypo theses have been expl ored. Gearing et al. (1996) showed that the extent of A deposition positively correlates with the number of ApoE4 alleles. ApoE4 has also been show n to increase the size and density of A plaques
19 (Yamaguchi et al., 2001). Many suggest th at theses findings are due to ApoEÂ’s involvement in the clearance of soluble A from the brain. A lack of clearance may promote the aggregation of A peptide therefore increasing insoluble A form. The ApoE isoforms have differential binding to A (ApoE2>ApoE3>ApoE4) (Yang et al,. 1997). This would suggest that risk may be due to the absence of the high affinity binding ApoE2 form rather than the presence of the ApoE4 form. Further evidence for this hypothesis is ApoEÂ’s association with anti oxidant activity. In a step down paradigm, ApoE2 is the most effective antioxidant wh ile ApoE4 is the least effective (Miyata & Smith, 1996). Carefully designed in vitro and in vivo experiments are still needed to ultimately clarify ApoEÂ’s role in the increase risk of developing AD. There are also a number of potentiall y modifiable risk f actors of AD. Head trauma has been linked to an increase risk of developing AD. After a significant trauma to the human brain, both A (Roberts et al., 1994) and tau pathology (Smith et al., 2003) appear even in young patients (Emmer ling et al., 2000). Levels of A in the CSF are also elevated after injury. Repeated mild traumas have been shown to accelerate A production and induce cognitive impairment (Geddes et al., 1999). This correlation between head trauma and increase in A offers a mechanism by which head trauma increase risk of developing AD (Jellinger, 2004). Many findings support that environmental f actors (including diet), are important variables in increasing or decreas ing the risk of AD. People of Japanese origin living in the United States have a 6.24% chance of deve loping dementia, while their counterparts living in the Japanese homeland have a less than 2% risk (Grave s et al., 1996). A possible reason for this discrepa ncy has been attributed to di fferences in diet. Two lines
20 of research connect the high leve ls of total cholesterol with increase risk of AD. The first direct correlation is an association between elevated midlife total cholesterol levels and late-life Alzheimer dise ase (Notkola et al., 1998). A second point of indirect evidence is work showing a reduced rate of the disease in people who used statins to reduce their blood total cholesterol level (Jick et al., 2000). These findings combined with ApoEÂ’s involvement in cholesterol transport makes for a causal web between environmental and genetic interaction. Kalmijn et al. (1997) s uggested that a diet high in saturated fat increases the risk of dementia, whereas fish consumption may decreased this risk. Along with an increase in fish (omega-3 fatty aci ds) consumption of other dietary supplements have also been correlated with a decrease in risk of AD. Reactive oxygen species are associated with neuronal damage in AD, maki ng the addition of antioxi dants into the diet a possible risk reducer. Results from studies trying to link increase in antioxidant vitamin consumption and risk of developing AD are in consistent. A study in 5395 people age 55 years and older found that dietary intake of v itamins E and C, but not supplement intake, was associated with a low risk of AD (Enge lhart et al., 2002). Morris et al. (1998) showed that use of higher-dose vitamin E a nd vitamin C supplements may lower the risk of AD. In contrast, another 4 year study i nvolving 980 individuals age 65 years and older did not find any associations between dietary, supplement, or total intake of carotenes, vitamin E, or vitamin C with a low risk of AD (Luchsinger et al., 2003). Due to the relative safety of dietary and supplemental v itamins, the addition of these vitamins may be seen as a benign measure, not prove n, in trials of primary prevention. Another medical risk factor asso ciated with AD is rais ed blood pressure. Kivipelto et al. (2001) colleagues examined midlife vascul ar risk factors and AD. This
21 study showed high systolic blood pressure in mi dlife was a significant risk for AD in later life. In contrast, high diastolic pressure in mi dlife did not translate in to an increase risk. While diastolic pressure, in this study appeared to be less of a risk f actor, previous studies observed an increased risk with elevated diastolic pressure (Skoog et al., 1996; Launer et al., 2000). Kivipelto et al (2001) noted Â“our data should not be interpreted as discounting the potential risk of Alzheime r's disease related to raised diastolic blood pressure but rather as emphasizing the importance of raised systolic blood pressure, even in people with normal diastolic blood pressure.Â” Si milar to the indirect support for decreased cholesterol with use of statins, discussed above, patient s treated with antihypertensive drugs may be at decreased risk for de mentia (Forette et al., 1998). A plethora of research has been conduc ted on education leve ls and risk of developing AD. Several studies have repor ted an increased risk of dementia and AlzheimerÂ’s disease among less-educated pe rsons (Schmand et al., 1997; Evans et al., 1997; Katzman, 1993). Due to the intimate relationship between education and the cognitive component of environmental enrich ment, levels of education and risk of AD will be discussed in detail in a later section. Diagnosis of AlzheimerÂ’s Disease While no cure is currently available, early diagnosis allows for patients to take advantage of current pharmaceutical treatments as well as prepare for future care while the patient is still cognoscente. A definiti ve diagnosis of AD can not be made until autopsy, however using a combination of ne urological exams, la boratory and genetic tests and brain imaging techniques 90% accu racy of diagnosis can be reached.
22 Neurological exams (NEs) enable clin icians to evaluate a patientÂ’s cognitive status, emotional, psychological, motor and sensory function. Evaluation of these measures provides information about the presen ce and progression of the disease. Firstly (NEs) offer a relatively sensitive diagnostic t ool for AD. NEs also distinguish AD from other neurological illnesses. After initial diagnosis, periodi cal NEs offer a way in which to track the diseaseÂ’s progression. NEs al so offers a way in which to customize treatments. Due to the variability in sy mptoms associated with AD, NEs offer a systematic description of symptoms which can then be addressed on an individual basis. The Mini Mental State Exam (MMSE) is a widely used NE used to identify dementia. Folstein, the developer of the MMSE, described it as "practical method for grading the cognitive state" (Folstein et al, 1975). The test includes questions which evaluate orientation, memory, attention, cal culation ability, language, and writing. There is a maximum score of 30 attainable, while a patient with AD will score a 26 or below. It is important to note that the education of th e patient should be take n into account when using the MMSE as an indicator of demen tia. Once a score belo w 26 is obtained and a diagnosis of dementia is given, laboratory te sts (discussed below) need to be performed to rule out other common causes of dementia. There are many other NEs available, including tests which measur e cognitive symptoms such as clock draw test and the Blessed test. Tests for non-cognitive sy mptoms include Behavioral Pathology in Alzheimer's Disease Rating Scale (BEHAVEAD) and the Neuropsychiatric Inventory (NPI). Brain imaging through radiolog ical techniques are used as diagnostic tools of AD. This diagnostic area can be broken up in to structural radiography and functional
23 radiography. Structural ra diography implements the use of MRIs, and CT scans to identify gross anatomical changes due to AD or other forms of dementia. Functional radiography, fMRIÂ’s and PET scans, can dete ct focal metabolic changes in the brain. Both of these subcategories can be useful addi tions in the differential diagnosis of AD. Magnetic resonance imaging (MRI) allows for volumetric measurements of anatomically relevant structure in the AD brain. As discussed above, the entorhinal cortex and the hippocampus show early neuropat hological changes with the onset of AD. Numerous studies have shown MRI-based vol umetric measurements of the entorhinal cortex and the hippocampus to be reliable di agnostic tools of early AD (Huesgen., et al 1993; Juottonen et al., 1998). Pennanen et al. (2004) have found MRI imaging to be advantageous in the classification of patie nts with mild cognitive impairment (MCI), early AD and cognitively normal elderly su bjects. MRIs showed volumes of the hippocampus and entorhinal were significantl y reduced in the following order: control > MCI > AD. Computerized tomography (CT) sc ans are a second tool of structural radiography. While the CT scan has less resolution than a MRI, they are readily available and can contribute as a component of AD diagnostic s. Like MRIs, CT scans give the ability to detect, in vivo, anat omical changes including cerebral atrophy and ventricular enlargement. Condefer et al,. (2004) examined the CTÂ’s impacted on diagnosis of dementia. The data collected s howed that the addition of a CT scan in the diagnostic process of AD may be expected to impact diagnosis and treatment of dementia in 10-15% of cases. The MRI and CT brain imaging techniques are not sufficient alone to diagnose AD and continue to be a source to differentiate AD from other causes of dementia.
24 Functional brain imaging techniques offer the ability to indirectly measure neuronal activity in individua l patients. Positron Emmition Tomography (PET) and functional magnetic resonance imaging (f MRI) are two methods implemented in functional brain imaging. The most comm on methodology of PET imaging used in the characterization of AD is the measurement of cerebral metabolic rate using glucose analogue 18-fluorodeoxyglucose (FDG). After intravenous delivery, FDG is taken up by cells via a glucose transpor ter and becomes trapped afte r phosphorylation by hexokinase (Lee et al., 2003). The resultant PET imag e reveals the distribution of FDG uptake, which is a measure of cerebral metabolic rate in regions of interest Individuals with a dementing illness other than AD often exhibit different abnormal metabolic patterns, thus making it possible to differentiate AD from ot her dementias. Bila teral temporo-parital hypometabolism is a consistently found patte rn reported on PET scans in the AD brain (Lee et al., 2003). Nestor et al. (2004) found patients susp ected of having AD to show dysfunction in temporo-parieto-oc cipital cortices. As the disease progresses, PET scans show metabolism in sensory and motor areas to remain relatively intact; however, severe AD patients often show frontal lobe hypo-metabolism (Lee et al., 2003). Hoffman et al., (2000) confirmed that bilateral temporo-pari etal hypo-metabloism highly correlates with the pathological diagnosis of AD, making PET-FDG a likel y diagnostic tool. A second novel use of PET scanning currently being exam ined as a diagnostic tool of AD, is the imaging of A plaques in vivo. Klunk et al., (2 004) implemented a tracer called Pittsburgh Compound-B (PIB) to visualize am yloid deposition. The PIB is a neutral benzothiazole that binds to A and crosses the blood brain barrier. This compound also readily clears from the brain. PET sca nning with PIB in AD patients decorated
25 association cortex region, a re gion commonly linked to high A deposition in the AD brain. This new technique c ould offer advancements in di agnosis as well as improved monitoring during clinical trials of new AD pharmaceuticals. Like the PET scan fMRI offers a method to indirectly measure functional changes associated with neuronal activity. The most common fMRI methodology used is termed the blood oxygen level-dependent method (BOLD) (Lee et al., 2003). This method measures signal change associated with oxygenation in cerebral blood vessels. Changes in neuronal activity results in a decreased oxygenated to deoxygenated blood ratio. The presence of oxygenated blood resu lts in a constant signal while deoxygenated blood results in a loss of signal. When a cognitive task is pe rformed, a surplus of oxygenated blood results in a bri ghter signal. fMRIs have several benefits in comparison to PET scans and therefore are th e preferred method in functional radiography AD diagnostic techniques. Primarily fMRI doe s not involve the inva sive injection of radioactive dye. This benef it allows for repeated proce dure during the course of the disease with no danger to the patient. A sec ond benefit is that fMRI may be performed on any clinical MRI scanner a nd is therefore accessible to many patients. Numerous approaches using fMRI have emerged in th e study of AD diagnostics. One approach examines fMRIs conducted during cognitive tasks that are performed easily by non-AD individuals, but which are more challenging for patients at ri sk of developing AD. When comparing fMRIs of control individuals a nd asymptomatic ApoE4 carriers, Brookheimer et al,. (2000) observed that patterns of brain activation during tasks requiring memory differed depending on the genetic risk (presence of ApoE4 allele). fMRIs may help to identify patients with neurochemical cha nges in the early stages of AD by studying
26 signal reactivity to pharmacol ogical challenge. The data coll ected showed that impaired cholinergic systems can be identified w ith fMRI (Goekoop et al., 2004), making this imaging technique a viable to ol of early detection. Many other causes of dementia can mi mic AD. A standard laboratory workup is conducted to rule out nutriti onal deficiencies, infection, metabolic disorders, and drug effects as possible culprits of dementia. While these laboratory tests rule out AD there are other tests under investigati on as diagnostic tools of AD. Cerebrospinal fluid tests for the presence of proteins associated w ith AD pathology. This extremely invasive procedure measures total ta u, phosphorylated tau and A 42. A patient with AD would show high levels of tau protein and phos phorylated tau, however a low level of A 42 (Blennow & Vanmechelen, 2003). Tau protein represents neuronal damage while phosphorylated tau specifically reflects the phos phorylation state of tau, which correlates to the formation of tangles. A 42 levels are decreased, due in part to trapping of brain A into plaques, as well as other unknow n reasons (Blennow & Vanmechelen, 2003). Serum levels of iron binding protein p97 are another possible biochemical marker involved in AD. Kennard et al (1996) provided evidence that the soluble form of iron binding protein p97 is found in elevated amount s in the serum of AD patients. While still in its infancy, this test represents the first blood test, aside from genetic testing, that may have potential for identifying subjects afflicted with AD. A final diagnostic tool to be discussed is the contr oversial genetic testing option for patient with a possible predisposition to AD. As discussed above many mutations of the APP, PS1, PS2 have been found and linked to genetic causation of AD. Presence of ApoE4 has also been linked with increase pr evalence of AD. Many clinicians raise the
27 question if testing for these abnormalities meet the criteria for a diagnostic test. van der Cammen et al. (2004) states Â“I n general, a medical test shou ld be useful in such a way that it raises diagnostic certai nty and has positive implicati ons for the treatment and wellbeing of the individual patie nt.Â” Many have questioned if genetic testing of AD meets these criteria. The test for ApoE, commercially available, is the mo st important genetic determinate for late-onset AD. Numer ous studies have constantly found ApoE genotyping alone does not have sufficient sens itivity in diagnosing susceptibility to developing AD (Mayeux et al., 1998; Slooter et al., 1996). When causal genes such as APP, PS1 and PS2 were examined in earlyonset AD cases, less then 10% of cases were linked to relevant AD genes, leaving 90% of cases with no association (van der Cammen et al., 2004). While genetic testing in the res earch world may offe r insight into the molecular mechanism of AD and allows for th e generation of useful transgenic animal models, its place as a clinical diagnostic tool of AD is still in need of maturation. Treatment of AlzheimerÂ’s disease Although there is currently no cure for AD, the FDA has approved two classes of drugs that have shown to be safe and effectiv e in the treatment of AD. The first class to be approved include the acetylcholinestear e inhibitors Aricept, Exelon & Reminyl, Cognex. These palliative treatments increase the compromised brain acetylcholine levels found in AD by decreasing the breakdown of acet ylcholine by acetylcholinesterase in the synaptic cleft. While this class of drugs does not slow the progression of AD, it does provide cognitive benefits for patient in the early stages of AD (Sonkusare et al., 2005). As the disease progresses th e acetylcholinesterase inhi bitors class of drugs loss
28 efficiency. The window of effectiveness is typically 6 months to 2 years (Delagarza et al., 2003). A second class of drugs recently approved for moderate to severe AD are NMDAreceptor blockers. While intact NM DA function correlated with cognition and learning, over-stimulation of glutamatergic sy stem may result in neuronal damage. An over-excited glutamatergic system will result in an abundance of intracellular calcium, resulting in a toxic environment for neurons Memantine (Namenda) is an uncompetitive NMDA-receptor antagonist that has been show to be effective in the treatment of moderate to severe AD (Reisb erg et al., 2003). The blocki ng of NMDA-receptors leads to suppression of over-excited glutamatergi c systems and therefore removes neuronal calcium overload. Implementing this mechanism, Memantine treatment reduced clinical deterioration in moderate to severe AD (Rei sberg et al., 2003). In severely demented patients, treatment with memantine improve d function, decreased care dependence and decreased behavioral symptoms associated with severe AD (Winblad & Poritis, 1999). It has been suggested that a combination th erapy including an a cetylcholinesterase inhibiters and memantine may offer th e most benefits to AD patients. Many alternative non-FDA approved compounds have emerged as possible treatments of AD. Nonsterodial antiflamma tory drugs (NSAIDs) are the most widely used drugs for management of pain, fever, and inflammation (Gas parini et al., 2004). There mechanism of action is largely attri buted to the inhibition of cyclooxgenase (COX 1-2), leading to the suppression of prostagla ndins synthesis. Epidemiological studies have shown that prolonged use of NSAIDs leads to a decrease risk of developing AD (Zandi et al., 2002; 2004). While NSAIDs as a prevention have shown positive results
29 NSAIDs as treatments are faced with medioc re beneficial findings and controversial mechanisms of action. Aisen et al. (2003) found no beneficial cognitive effects of treatment with a Cox-2 inhib itor (rofecoxib) or a non-sele ctive NSAID (naproxen) in patients with mild to moderate AD. Reines et al. (2004) provide fu rther support for this finding by showing Â“Rofecoxib had no effect on Alzheimer's disease in a 1-year, randomized, blinded, controlled study.Â” So me suggest NSAIDs work by managing the chronic inflammation associated with AD pathogenesis, while others suggest a mechanism through which NSAIDs alter the activity of secretase and therefore alter the production of A (Lle et al., 2004). Recently the Na tional Institute on Aging halted the AlzheimerÂ’s Disease Anti-inflammatory Prev ention Trial (ADAPT), in part because of the recall of cox-2 inhibitor ro fecoxib (Vioxx) due to increased risk of cardiovascular events and health concerns surrounding ot her NSAIDs including celecoxib (Celebrex), and naproxen (Aleve). Antioxidants have also been noted as possible treatments of AD. As discussed above, the amyloid cascadesis associated wi th an increase in oxidative stress. Antioxidants such as alpha-tocopherol (v itamin E) and Ginkgo Biloba have been investigated as possible treatments for AD. Vitamin E (2000 IU/day) has been shown to slow the progression of AD by about 7 months when taken over a two year period of time (Berman & Brodaty, 2004). Ginkgo Biloba has been shown to increase waves in AD patients as well as stabilize and improve cogni tive function (Maurer et al., 1997). Larger clinical trials are ne eded to provide further support fo r these antioxidants as treatments for AD.
30 Aside from treating the cognitive symptoms of AD, psychi atric and behavioral disturbances associated with AD are often treated with FDA-appr oved anti-psychotic and anti-depression drugs. Risperidone (Risperd al), a serotonin-dopamine antagonist, has been shown to significantly improve symptoms of psychosis and aggr essive behavior in patients with severe dementia (Katz et al., 199 9). Selective serotonin reuptake inhibitors and benzodiazepines have also been used in treating depression, insomnia and anxiety in patients with AD. New treatments for AD seeking FDA a pproval are on the horizon. Research into developing drugs which stop progr ession of AD rather than treat symptoms are needed in the clinical arena. Compounds that inhib it the enzymes responsib le for abnormal APP cleavage are under heavy investigation. Vaccinations against A have been meet with problems during phase 2 clin ical trials, however, passi ve rather than active A vaccination appears promising. A new ther apeutic currently involved in phase III clinical trials has been show to have potenti al as a new therapeutic for AD. Alzehemed is expected to act on two levels: firstl y to prevent and stop the formation A fibrils in the brain and secondly to inhibit the infl ammatory response associated with A build-up in AD (Geerts, 2004). The developments of transg enic animal models have catalyzed the development of new therapeutics and will continue to be a medium for which novel treatments can be developed.
31 Animal Models of AlzheimerÂ’s Disease Genetically altered mice offer a tool to de fine in vivo functions that have been observed in vitro. The ability to model a disease in a transgenic mouse enables researchers to study mechanisms of diseas e progression, examine possible preventions and devise viable treatments. In order for th e creation of a transgenic animal to become feasible, genes responsible for the disease mu st be identified through linkage studies. Once identified, the generation of a transgenic mouse model can proceed. Generating a transgenic mouse model begins with the selection of a transgene (Picciotto & Wickman, 1998). A transgene is defined as the segment of DNA that is transcribed into a mouse model. The cl assical transgenic construction requires a segment of DNA to be transcribed then s ubcloned down-stream of a neuron-specific promoter. The promoter dictates the level of expression, tissue specificity and temporal pattern of the gene involved. The gene and promoter are th en excised from the plasmid, resulting in the isolated transgene. Multiple copies of the isolated segment are injected into the pronucleus of a single fertilized egg and then implanted into a foster mother. At time of injection into the pronucleus, the gene is either integrated or degraded. Those eggs that integrate the gene become viable offspring, and carry the transgene in their gametes, go on to comprise the transgenic fonder mouse. On ce the fonder line is screened for number of transgenes integrat ed, sight of integration, and the transgeneÂ’s
32 interaction with neighboring endogenous genes, the fonder line goes on to generate a colony of transgenic mice. An important element in the creation of a transgenic mouse model is the background strain of the embryos chosen. A blend of cognitive, motivational, and sensorimotor elements are needed to cons truct a transgenic animal that can be implemented in behavior. A strain which is cognitively superior may over shadow a behaviorally detrimental transg ene. When modeling the pa thology of a disease, strain background plays an equally important role. Mice expressing the APP gene, generated on a C57BL/6 background, presents with A plaques and corresponding behavioral impairment (Picciotto &Wickman, 1998). In contrast, when the APP transgene is inserted into an embryo of a mouse with a FVB background, plaques are absent (Picciotto &Wickman, 1998). Transgenic mice with hybrid strain backgrounds offer a more practical model that demonstrates pathol ogy and behavior which closely mimics the disease modeled. The PDAPP Mouse Model of AlzheimerÂ’s Disease The PDAPP mouse model of AD is one of several models implementing the missense mutations of the APP gene. These mu tations have been linked to early onset familial AD. PDAPP mice were generated usi ng the platelet-derived growth factor directing the human APP gene encoding th e 717 mutation (Games et al., 1995). The mice were derived from several hybrid bac kgrounds, resulting in a mixed strain of C57BL/6 +DBA+ Swiss-Webster. Games et al. (1995) confirmed the successful insertion of the altered human APP gene and showed the PDAPP mouse had a 10 fold
33 overexpresion of APP independent of endoge nous mouse APP levels. This expression was observed in multiple tissues and most notably in the brain. In contrast to previously generated transgenic models containing human non-mutated APP, the PDAPP model exhibited A -production & depostion. The success in generating a transgenic model with substantial A deposition may have been due to nume rous factors. Primarily, the higher levels of the human APP expression may have lead to the generation of reproducible A pathology. Secondly, as discussed above, the promoter drives the amount of expression and tissue specificity. Implementing the platel et-derived growth factor promotor allowed for targeted expression preferentially in neurons of the cortex, hippocampus, hypothalamus and cerebellum. PDAPP Pathology The A peptides observed in the PDAPP model ranged from diffuse soluble forms to compact plaques stainable with th ioflavin S, Bielschowsky silver stain and Congo red. The outskirts of the plaques were intimately surrounded with reactive astrocytes, replicating plaques found in the brai ns of AD patients. Also closely related to the AD brain, A plaques were associated with dys trophic neurites. A final observation noted by Games et al. (1995) was the reduced synaptic and dendritic density in the hippocampus of PDAPP mice. Games et al. (1995) characterized th e progression of plaque development in the heterozygous PDAPP mice. At 4 months no pl aques were present a nd transgenic animals were indistinguishable from nontransgenic litte rmates. By 6-9 months plaques began to appear in the hippocampus, corpus callousm an d cerebral cortex. Animals greater than 9
34 months showed Â“numerousÂ” dense deposits. This progression closely mimics progression in the AD patient. Masliah et al. (1996), Irizarry et al. (1997), Dodart et al. (2000a), and Chen et al (2000) went on to further anal yze the neurodegenerative path ology in the PDAPP model. Using the same line discussed above by Games et al. (1995), an exte nsive analysis of pathology was done through use of electron microscopy by Masliah et al. (1996). Masliah et al. (1996) describe d the extracellular plaques obs erved in PDAPP mice to be Â“strikingly similar to those observed in AD brains.Â” Confirming previous findings by Games et al. (1995), Masliah et al. (1996) described dense ne uritic plaques in frontal cortex and hippocampus. Associated with these plaques were dys trophic neurites and reactive astrocytes. Further analysis of th is same line of PDAPP mice by Irizarry et al. (1997) described levels of APP expression, regional pattern of deposition, and neuron counts in relevant regions. Across all 3 tim e points investigated (4, 11 and 18 months), APP expression remained stagnant. Despite the absence of difference in expression across time, progression of A deposition in a regionally specific manner was observed. At 8 months of age A deposition was isolated to the cingulated cortex. By 12 months, A spread into surrounding CA1 of the hi ppocampus and entorhinal cortex. At 18 months, a substantial increase in A was observed in the order of 20-50%. Implementing stereological techniques, neur ons in the CA1, cingulated a nd entorhinal cortex were counted. In sharp contrast to human AD brains, no neuronal loss was observed in any region at any time point in the PDAPP mice. This lack of neuronal death in PDAPP mice strongly disputes theories that A plaques directly cause neuronal death by neurotoxic mechanisms. Irizarr y et al. (1997) point out that absence of neural death may
35 have been due to substandard methodol ogical protocols not identifying small subpopulations of neuronal death. A second hypo thesis suggested th at greater time of exposure to toxic A is needed to produce neuronal de ath. Later, Dodart et al. (2000a) provided work supporting previous findi ngs of Â“age-dependent region specificÂ” deposition of A in heterozygous PDAPP transgenic mi ce. Additional data of a direct comparison between homozygous and heterozy gous PDAPP mice enabled the analysis of dose dependent expression of APP on the resulting A pathology. At a 10-12 month time point, a three-four fold great er total number of plaques in homozygous vs heterozygous animals was seen. While this impressive di stinction was observed in cortical regions, less separation was observed in the hippocampus. In contrast to early work by Games et al. (1995), Dodart et al. ( 2000a) observe first mature A deposits in 3-4 old month animals. These plaques were exclusively lo cated in the CA1 region of the hippocampus, the medial part of the cingulated cortex, a nd corpus callosum at th e level of the dorsal hippocampus. This contradiction to Game s et al. (1995)Â’s findings of Â“no obvious pathology detected at 3-4 mont hs,Â” may be due to extendin g pathological examination past Â“obviousÂ” pathological changes to more obscure region specific changes. At 6-7 month and 10-12 month time points, while D odart et al. (2000a) provides a more extensive exploration of region specific di stribution of pathology, the findings closely mimic previous abbreviated findings by Games et al. (1995). In behaviorally tested animals, Chen et al. (2000) describe re latively low levels of deposition in the hippocampus of 9 month old animals with an increase in density at the 16 and 22 month time points. The findings at the 9 month time point are difficult to compare to previous work (Games et al., 1995; Maslish et al., 1996; Irizarry et al., 1997) due to the possible
36 confounding variable of behavioral testing. However, Dodart et al. (2000a) shows Â“numerous mature depositsÂ” at 6-7 months in similarly behaviorally tested PDAPP mice. This difference may be due to Dodart et alÂ’s counting of plaques (definition of Â“numerous plaquesÂ” unknown) versus Chen et al. (2000) an alysis of plaque dens ity. In summary of A associated pathology in PDAPP transgen ic model, these mice exhibit regional and age-dependent A deposition, independent of a constant over expression of APP, as well as lack of neuronal loss a ssociated with the AD brain. The earliest change in the AD br ain which correlates with cognitive function is synaptic loss. Alterations in synapses in the PDAPP model have al so been observed. Dodart et al. (2000a) observe d an increase in synaptic density in young 3-4 month old animals, however, saw a decrease in mice olde r than 6 months. The decrease in later life was suggested to be related to the neurotoxic effects of A The increase early in life was observed in the absence of A plaques and correlated with APP expression. This would suggest a beneficial effect of overexpression of APP. APP expression has been linked to neurite outgrowth and synaptotrophic proper ties (Morimoto et al., 1998; Wallace et al., 1997). Larson et al. (1999) st udied synaptic transmission a nd long term potentiation in hippocampal slices from young (4-5 months) and old (27-29 months) PDAPP mice. As discussed above, young PDAPP mi ce have high expression of APP but lack plaques. Aged animals have the same level of high APP expression; however, have a high burden of A plaques. In vivo electrophysioloical testing revealed that young PDAPP mice had enhanced paired-pulse facilita tion, a significant decrease in maximal size of field EPSPs, and an increase in decay rate of LTPs. These results suggest synaptic changes occur independent of plaque deposition. To state synaptic changes are completely independent
37 of A is misleading due to the absence of a measurement of soluble A forms in this work. In the old animals, the most st riking difference observed was that maximal synaptic field potentials were greatly reduced. In cont rast to young animals, aged PDAPP mice showed significantly reduced pair ed-pulse facilitation. Despite a decrease in the number of functional s ynapses, LTP mechanisms were reported to be intact in these aged animals. Larson et al. (1999) concluded PDAPP mice have altered synaptic communication preceding presence of plaque s. As well, disturbance in synaptic transmission in the presence of plaq ues are not linked to altered LTP. In 2000, Dodart et al. noted neuroanatomical abnormalities that had been overlooked in previous studies discussed above. Dodart et al. (2000) described Â“rather marked hippocampal atrophy as early as 3 months of age. A 20 to 40 percent decrease that did not progress with age was observed. The early presence a nd lack of progression of this neuroanatomical abnormality suggest s that overexpression of APP may effect neurodevelopment. Behavior of PDAPP Model The pathological changes found in the P DAPP mice translated into cognitivelybased behavioral changes. In sensorimot or tasks, PDAPP mice performed identical to nontransgenic mice at both early and late time points (Dodart, 1999; Nilsson, 2004). Dodart et al. (1999) found no motor impairments at 3, 6, or 9 months in a motor-based radial maze. Motor activity in the radial maze was recorded and an increase in freezing behavior was noted in the PDAPP mice. This stereotypical behavior was only observed when the task was novel and soon ceased after the first testing sessi on. Nilsson et al.
38 (2004) confirmed intact sensory-motor ability of PDAPP mice by showing no differences in open field task, balance beam, string agility and Y-maze entries compared to Tgmice. The findings that PDAPP mice were not impaired in sensorimotor tasks provide evidence that any impairments in cognitive functions are not due to physical, motivational, or exploratory behavioral differences. While the sensory-motor tasks perf ormed at numerous time points extracted no differences in PDAPP mice from wild type mice at numerous time points, cognitive based tasks reveal impairment that in seve ral studies correlated with the pathological abnormalities described above. Both Dodart et al. (1999, 2000b) and Chen et al. (2000) analyzed recognition memory in PDAPP mice cross-sectionally in an object-recognition task. Dodart et al. (1999) found no rec ognition impairment in young three month old animals. However, at 6 months, homozygous animals showed impairment while heterozygous animals remained cognitively intact. By 9 months both homozygous and heterozygous PDAPP mice had sign ificantly less recognition index then wild type mice. In sharp contrast, Chen et al. (2000) found no impairment in object-recognition tasks in heterozygous PDAPP mice at 69, 13-15 and 18-21 month time points. These different findings could be attributed to task protoc ol differences. Chen et al. (2000) tested animals for 5 days with a 5 minute habitu ation period while D odart et al. (1999) administrated a single day of te sting with a single 50 minute habituation one day prior to testing. Analysis of 5 days of testing may have diluted observable working memory impairments that may have been present in the first day of testing. In the same animals analyzed pathological ly by Dodart et al. (2000a) (disscused above), the Radial Arm Maze (RAM) was implemented, as a test of working memory.
39 The RAM is a more challenging task than object-recognition, as well as includes a reference memory component (which was not present in the object -recognition task). Dodart et al. (1999) found working memory impairment in the RAM task in both homozygous and heterozygous mice at 3, 6 a nd 10 month time points. Due to the hippocampal-dependent nature of this task it is hard to decipher if the early impairment observed at 3 months, in the absence significant plaque form ation, was due to AD related pathologies or the contribution of the hippo campal atrophy observed in this transgenic line at this time. In the reference memory component of the RAM task, homozygous and heterozygous mice made significantl y more reference memory errors than wild type at all three time points. While the defects in objection recognition task (discussed above) appeared to follow the A deposition time course, both working memory and reference memory impairment observed in the RAM appear to be present, at least at the 3 month time point, independent of A deposition. Chen et al (2000) later implemented at new water maze training protocol which forced the memory retrieval function to se lect for a most recently encoded platform location. Testing PDAPP mice at 6-9, 13-15, and 18-21 months in this new Â“training-tocriterionÂ” uncover two key findi ngs. The first of the findi ngs included an age-dependent deficit in trials to reach criterion. The 69 month old mice reached criterion in a number of trails that were non-significant when compar ed to the wild type animals. The 13-15 and 18-21 month old animals took significantly more trials to reach criterion when compared to the age matched wild type mice. The second finding is age independent. Despite the age differences in the PDAPP an imals, all PDAPP mice were significantly impaired in finding the first platform locat ion in a series. In comparing the age
40 dependent and independent impa irments of PDAPP mice with A burden using a correlation analysis, Chen et al. (2000) f ound there to be a significant, age-related, inverse correlation between learning capacity an d plaque burden. It is important to note that all animals (including the low pathol ogy bearing young animals) were included in the A correlation analysis, which may have el evated the relationship between pathology and behavior. In an extensive cognitive examinati on of PDAPP mice as well as other transgenic mice lacking ApoE or overexpressing huma n ACT, Nilsson et al. (2004) found no impairment in both Morris Water Maze (MWM) or Radial Arm Water Maze (RAWM) at the 2 month time point in PDAPP mice. The fi nding that these mice we re intact in both a spatial memory retention task (MWM) a nd a working memory task (RAWM) may indicate that the hippocampa l atrophy, which correlated with poor performance in a spatial discrimination task (Dodart et al 2000), was no longer present after numerous generations of crossbreeding. At a later time point, Nilsson et al. (2004) found cognitive impairment associated with both diffuse and compact A deposition in the hippocampus. At 18 months, PDAPP mice were found to be impaired in spatial learning tasks including both a long term memory (Morris Water M aze) and working memory task (RAWM). The working memory impairment observed in the RAWM correlated with both diffuse plaques (6E10 staining) and mature comp act plaques (congo red staining). This correlation suggests a cause and effect rela tionship between extent of deposition and cognitive dysfunction.
41 APP Swedish Mutation Transgenic Mouse Line The APPsw (Tg2576) mouse model of AD re presents a widely analyzed and implemented model of AD. These transg enic mice overexpress the 695-amino acid isoform of human amyloid precursor protein (APP) with a double missense mutation at 670 and 671. These mutations mimic those found in a large Swedish family which exhibit early onset AD. The transgene is e xpressed in C57B6/SJL mice and is controlled by the hamster prion protein promoter. APPsw Brain Pathology The Tg2576 APP mouse expresses the APP transgenic at unchanged levels between 2 and 14 months (Hsiao et al., 1996). Irizarry et al.Â’s (1997) observations agreed with, as well as added, a regional specificity to APP expression. Results from transgene expression revealed APP is expressed pre dominately in the neuronal layers of the hippocampal formation and in lesser amounts in cortical regions (Iriz arry et al. 1997). Despite constant expression of the APP transgene, formation of A is observed in an age specific manner. Hsiao et al.Â’s (1996) measurements of A 40 and 42 concentrations increased drastically until 13 m onths of age. Compact classic senile plaques were observed by 11-13 months in th e frontal, temporal, entorhinal cortex, hippocampus, presubiculum, subiculum and cerebellum (Hsiao et al. 1996). The observation of plaques in the cerebellum is incongruent with the absence of APP
42 expression in the cerebellum. At 16 months Irizarry et al. (1997) observed age-related A deposition in the cingulate co rtex, entorhinal cortex, dent ate gyrus and the CA1. In very old, high burdened 23 month old mice, brain A inversely correlated with levels of A in CSF (Kawarabayashi, 2001). Also associated with the Tg2576 tr ansgenic mouse is oxidative stress and impaired synaptic plasticity. Smith et al. (1998) describe a Â“tight associationÂ” between oxidative stress and A Tissues from 13-27 month ol d animals exhibited a global increase in oxidative stress (Smith et al ,. 1998). This finding makes the Tg2576 mouse a valuable model in the development of antioxida nt therapy for AD. Chapman et al. (1999) analyzed the synaptic communication in 12-15 month old mice Tg2576 mice. While short term potentiation was intact, the extremely memory dependent long term potentiation (LTP) was impaired (Chapman et al,. 1999). A group of slices from young 2-8 month old animals were also analyzed. Chapman et al. (1999) found no impairment in any synaptic plasticity m easures in young animals, indica ting that defects observed in old animals were related to presence of A Chapman et al., 1999) n contradiction with the Chapman et al.Â’s (1999) work, Fitzjohn et al. (2001) found age-related impairment in synaptic transmission; however, they observed no impairment in LTP at both 12 and 18 months of age. Synaptic transmission wa s restored with the a ddition of a glutamate receptor antagonist, kynurenate, in slices fr om mice at 12 months; however, slices from the 18 month old time point were unable to be rescued (Fitzjohn et al. (2001). This finding is of particular importance because it provides evidence for the discrepancies between the Chapman et al. (1999) study and th e Fitzjohn et al. (2001) study. The former
43 study used kynurenate treated s lices to measure synaptic tr ansmission and LTP, while the latter study showed that kynurenate rescues synaptic transmission impairment. While the Tg2576 mouse mimics AD pathology in many ways, it lacks the neuronal loss observed in the AD brain. Irizarry et al. (1997) assessed A neurotoxicity by evaluating neuronal loss in the Tg2576 m ouse. At 16 months Tg2576 mice showed no significant difference in ne uron cell count in the CA1 region when compared to nontransgenic mice. While neuronal loss is abse nt in this model, at this time point evidence of axonal changes and gl iosis are apparent (Irizarry, et al. 1997). This finding of insignificant neuronal loss indicates A deposits are not acut ely neurotoxic in the mouse brain (Irizarry, et al. 1997). A subclass of the APPsw mice, the APP23 mutant mouse exhibits some pathologic characteristics not found in the Tg2576 mouse. The first widely documented finding is presence of cerebrovascular amyl oid in the APP23 mouse. Calhoun et al. (1999) described a mouse, at 14-21 months of age, that in a ddition to having the commonly observed amyloid plaques also shows accumulation of A in vasculature. A second characteristic observed, however not as widely duplicated, is presence of neuronal loss. In 1998, Calhoun et al. published work showing the 14-18 month old APP23 mouse exhibited neuronal loss in the CA1 region. Neocortex was also analyzed and despite thinning of pyramidal cell layers, no quantit ative evidence of neuronal loss was observed.
44 APPsw Behavior The pathological characte ristics discussed above often translate into behavioral impairment in the APPsw mouse. Correlati ons have been made between cognition and A deposition, synaptic plasticity an d synaptophysin immunoreativity. Activity levels, sensory-motor measures a nd anxiety levels have been measured in the Tg2576 mouse. King & Arendash (2002) found the Tg2576 mice to have increased activity in the open field task only at the 3 month time point. This increase in activity at 3 months was absent at later time points suggesting an A independent mechanism. In an analysis of sensorimotor ab ilities (tested by open field ta sk, balance beam task, string agility task and Y-maze entries), the Tg2576 mo use showed balance beam impairment at 3, 14 and 19 months (King & Arendash, 2002) This sensorimotor impairment was confirmed in an observation of impairment at 14 and 19 months in the string agility task (King & Arendash, 2002). Despite numerous generations of Tg2576 breedings and slight differences in background strain, early im pairment in the balance beam was also observed at 5 and 6.5 months in this transgenic line (Arendash et al., 2004). In a test of anxiety the Tg2576 mouse showed no overt hyper anixety behaviors in the elevated plus maze throughout life (King &Arendash, 2002; Arenda sh, 2004). In contradiction to this finding, Lalonde et al. (2003) describe hypera ctivity in 17 month old Tg2576 mice in an open field Â“speed of movement toward cente rÂ” measure. This impairment was short
45 lived, only lasting 2 out of 3 da ys of testing and can not be ju stly compared to findings of the former work due to the differences in the two tasks measuring anxiety. The Tg2576 transgenic mouse shows robust cognitive impairment across a broad range of cognitive domains. Hsiao et al (1996) tested the Tg2576 mouse in a spontaneous spatial alternati on task (Y-maze) at 2 and 10 months. At 2 months both transgenic and non-transgenic mice tended to alternate their choices similarly; however, at 10 months a decrease in al ternations in Tg+ mice was observed. Impairment in Ymaze alternation was observed as early as 3 m onths as well as at 5, 8.5 and 19 months of age in Tg2576 mice (King &Arenda sh, 2002; Arendash, 2004). In the circular platform, a task of spatial learning/memory, transgen ic Tg2576 mice exhibited no impairment at 3, 9, 14 and 19 months (King & Arendash, 2002). However, in 7 month old animals, when learned cues were placed 180 degrees from thei r original location (e.g. reversal learning), the transgenic mice were unable to adapt a nd were found to be significantly impaired (Pompl, 1999). This finding showed that the Tg2576 mouse was not impaired in learning, however was impaired in the reversal learning aspect of the task. This task relies on the animals innate aversion toward unlikable stimulus, however it is often difficult to measure each animals different aversion to the stimulus making results from the circular platform task some what challenging to interpreted. n a sensitive task of spatial memor y, Hsiao et al. (1996) found 9-10 month old Tg2576 mice to be cognitively impaired in Morris Water Maze (MWM) when compared to age-matched controls. This impairment was absent in non-A burdened Tg+ mice at 2 months of age, indicating an association between A burden and cognitive impairment. At 6 months Tg+ mice were marginally impa ired, differing only in escape latency on the
46 last day of testing. These findings show a nice progression of impairment which mirrors increased A load. Similar age-related impairme nts were demonstrated by Chapman et al. (1999) and Westerman et al. (2002). Both groups observed a lack of cognitive impairment in young Tg2576, with emerging impair ment in (MWM) later in life. In a retention trial of MWM, West erman et al. (2002) observed a total quadrant preference for the goal arm in a subpopulation of young (4-5 ) Tg2576 mice, however spatial reference memory deteriorated in 6-11 month old mice. Examination of escape latencies in MWM acquisition showed age-related impairment in Tg2576 mice that was not observed until middle age (12-18 months). Very old 20-25 month old mice were significantly impaired in both MWM acquisition and retention when compared to Tglittermates, showing no ability to acquire or retain any spatial information. Note worthy is that a group of mice labeled Â“performance-incompetent miceÂ”, were eliminated, which generated a cohort of mice in which sensorimotor performance de ficits could be factored out of the interpretation of behavioral data. King and Arendash et al. (2002) found that in 19 month old Tg+ mice, higher levels of syna ptophysin immunostaining was correlated with impaired spatial reference memory as test ed by the MWM task. Interestingly, 19 month old transgenic mice had signifi cantly higher levels of syna ptophysin immunostaining than age-matched Tgmice. King & Arendash (2002) suggest compensatory changes in synaptic morphology and staining of dys trophic neuritics as sociated with A deposition. Taken collectively these findings woul d suggest that increased hippocampal synaptophysin levels are a manifestation of pathophysiological synaptic processing.
47 In contradiction to the findings of Westerman et al. (2002), King & Arendash (2002) showed no impairment through 19 months in the MWM. Neither MWM acquisition nor retention was able to separa te Tg2576 mice from non-transgenics, suggest intact long term and spatial memory in signi ficantly aged Tg+ mice. While these findings are consistent with the findings by Holcomb et al. (1999) who also showed no impairment in Tg2576 mice through 9 months, th ey are not consistent with findings by Hsiao et al. (1996), Chapman et al. (1999), and Westerman et al. (2002). Procedural differences may account for these differen ces. Other points of consideration are behavioral changes due to numerous generati ons of inbreeding, as we ll as the point that King & Arendash (2002) eliminated animals which were non-performers; if included such animals may have contributed to a cogni tively impaired effect. Recently Arendash et al. (2004) observed impairment in MW M in APPsw mice at 5, 6.5 and 8.5 months. These results are surprisingly different from their early research showing no impairment through 19 months. Several fact ors could contribute to these discrepancies. The most likely explanation is changes in background stra in. Evidence for this is provided by work from Savonenko et al. (2003), who showed that expression of APP at 3-fold times over endogenous levels was not enough to induce cognitive impairment in 24-26 month old transgenic mice that were on a full C57BL/ 6J background. The unimpaired APPsw mice had a more homogenous C57 background (an excellent performing background) in King & Arendash (2002), while their impaired count erparts in the later study (Arendash et al. 2004) had a mixed background. A second possibl y is that through crossbreeding over several generations, the Arendash APPsw tr ansgene line became behaviorally more sensitive to mutant APP expression, which would result in earlier impairment.
48 Recent work by Arendash et al. (2004) s howed that APPsw mice at 6.5 months of age had significantly impaired working memory, as measur ed by the Radial Arm Water Maze (RAWM) task. Compared to non-transgenic mice, APP transgenic mice made significantly more errors during block 2, Trial 4 (final acqui sition trial), as well as made significantly more working memory errors on Tr ial 5 of the last 3 blocks of testing These recent findings are not consistent with earlier work (Morgan et al. 2000) showing that APPsw mice were unimpaired in the RAWM at 5-7 months of age. Arendash et al., (2004) suggest that several ye ars of breeding have apparent ly resulted in the APP line becoming more behaviorally sensitive to mu tant APP expression and/or the process of A deposition at an earlier age. Taking 15 measures from 9 different activity, anxiety sensorimotor, and cognitive tasks, Arendash et al. (2004) used discrimi nant function analysis to separate APPsw mice from non-transgenic mice based on behavior. These findings show that APP mice are impaired globally over multiple behavioral measures. The APP23 transgenic mouse has ma ny similarities, behaviorally, to that of the closely generated Tg2576. As observed in Tg2576 mice, APP23 mi ce exhibit no anxiety compared to the non-transgenic mice. Howe ver, unlike the former model, APP23 mice were not impaired on any motor coordination ta sk (Lalonde et al. 2002). In MWM, the APP23 mice were impaired as early as 3 mont hs as well as at 6, and 24 months (Lalonde et al. 2002; Dumont et al. 2004).
49 APP Swedish + PS1 Mutation Transgenic Mouse Model An apparent continuation of the of the transgenic mouse lines was the creation of a mouse expressing human PS1. This model lacked impressive A pathology despite an increase in A 1-42/-1-40 ratio (Duff, 1996). Ho wever, co-expressing human PS1and APPsw mutations lead to a tr ansgenic mouse with impressive pathology observed months earlier than the single Tg2576. APPsw +PS1 Pathology The addition of mutant PS1 through cr ossbreeding with the APPsw mouse has been shown to accelerate amyloid pathology in the brains of transgenic mice (Borchelt et al, 1997; Holcomb, 1998). Borche lt et al. (1996) de scribed a 50% increase in the ratio of Ab42/40 in double transgenic (APPsw + PS1) mice when compared to singles at a 2-3 month time point. This finding tran slates to an acceleration of A deposition in the double transgenic mice. In a direct comp arison between single and double transgenic mice, single transgenics were plaque free until 18 months of age. In contrast, APPsw +PS1 mice exhibited plaques as early as 9 mont hs (Borchelt et al, 1997). This finding that the addition of the PS1 mutation dr amatically accelerates the rate of A deposition demonstrates PS1 acting in a synergistic re lationship with the mutated APP gene. The Borchelt mouse line is slightly different fr om a more widely used APPsw+PS1 mouse, which exhibits plaques much earlier. In the more widely implemented APPsw +PS1
50 mouse, Holcomb et al. (1998) showed that in 6-16 week double tr ansgenic mice there was a 41% increase in transgenic human A 1-42. This was quite different from the unchanged A 40 and 42 levels observed in the 616 week single APPsw mice. In the double transgenic mice, levels of A 40 and 42 increased in 12-16 week mice and showed further increase at 24-32 weeks of ag e. This increase was not observed for the most part in the single transgenic mice. Histological analysis showed that 13-16 week old double transgenic mice had compact plaques, measured by thioflavin S staining, that were absent in single transgenic littermates. Thioflavin S staining increased substantially in the doubly transgenic brains of 24-32 w eek old mice. This work was continued by Gordon et al. (2002), whom describe a regional specificity of deposition in a time course manner. At 6 months of age, doubly transgenic mice had multiple deposits particularly in the frontal, entorhinal cortices and hippocampus. Through 9, 12 and 16 month, the number of deposits increased and began to in fringe on the striatum, thalamus and brain stem. Closely associated with the dens e plaques were dystrophic neuritis, reactive astrocytes, and microglia activation (Borchelt et al, 1997; Gordon et al, 2002). Gordon et al. (2002) observed A deposits invested with dystrophic neurites as early as 6 months. Using GFAP immunostaining, Gordon et al. (20 02) measured astrocyte reactivity. At 3 months, enhanced GFAP staining was restrict ed to the hippocampus. This would suggest the hippocampus plaques were the most mature and therefore where the A pathology had begun. As animals aged, increased GFAP de nsity in the striatum, and cerebral cortex was observed. In a direct comparison of si ngle and doubly transgenic mice, as would be intuitive, the high burden doubly transgenic mice had significantly more GFAP staining
51 than less-burdened single APPsw mice. Using MHC-II immunostaining techniques, Gordon et al. (2002) showed increased microg ila activation with in creased age in APPsw + PS1 mice. Consistence with work prev iously reported, activated microglia and astrocytes increased synchronously with A burden, and were closely associated with plaques (Matsuoka, et at. 2001) Take n together, the time course of A deposition, the characteristics of the A plaque, and the inflammatory res ponse which they instigate, the doubly transgenic APPsw + PS1 mouse closely resembles A pathology found in the AD brain. While the doubly transgen ic mice develop significant A burdens, no neuron loss was initially seen with the APP + PS1 tran sgenic mouse line (Takeuchi et al, 2000). However, a recent study by Sadowski et al., ( 2004) showed significant neuronal loss in the CA1 region of 22-23 month old APP+PS1 mi ce. While there is no global neuron loss in this model (such as the extensive cort ical and hippocampal loss seen in AD), many findings suggest a lack of neuronal health in the APPsw +PS1 model. Changes in glucose metabolism (Sadowski, et al, 2004), de ndrites, dendritic spines (Moolman et al, 2004), reorganization of cholinergic term inals (Wong et al. 1999) and resultant impairments in synaptic plasticity (Dickey, et al. 2003) have all been documented. In an assessment of brain metabolism, Sadowski et al. (2004) used 14C-2DG and showed no impairment at 2 months. However, at 22 months the brain glucose utilization (BGU) index in the hippocampus of APP+PS1 mice was decrease by 26.6%. This defect correlated with spatial memory deficits (Sa dowski et al, 2004). Impairment in glucose utilization is also observed in the human AD brain, which is associated with early cognitive impairment (De Santi et al, 2001).
52 Along with impairment in neuronal metabolism, APP+PS1 mice also have alterations in dendrites and dendritic spines At 11 months, APP+ PS1 had significantly less spines and total dendrite area in the hippocampus (Moolman et al, 2004). Other characteristics observed included swollen bul bous dystrophic neurites. Moolman et al. (2004) went on to characterize human AD hi ppocampal tissue and compare the findings in APP + PS1 mice. Both quantitatively and qualitatively similarities were observed. Moolman et al. (2004) reported that Â“images of neurons from the AD brain were remarkably similar to those from the 11 m onth-old APP+PS1 mice.Â” Also similar was the 50% loss of dendritic spines in the AD brain and transgenic mice, when compared to controls. As well, cholinergic neurons are affected by the introduced mutant APPsw and PS1 transgenes. Wong et al. (1999) found there to be reorganization of cholinergic terminals in the cortex and hippocampus of double transgenic mice. These mice had prominent cholinergic synaptic deficits as measured by staining of vesicular acetylcholine transporter (VAChT) boutons (Wo ng et al, 1999). As one would anticipate the deficits observed in synaps es translated into impairment in LTP, Trinchese et al. (2004) observed abnormal LTP as early as 3 m onths that paralleled plaque appearance. In 17-18 month APP+PS1 mice, Dickey et al. (2003) showed selectively reduced expression of synaptic plasti city-related genes including Arc, Zif268, NR2B and GluR1. In summary, APP + PS1 first show A deposits at 4-6 months of age, which increase with age. This increase occurs despite a constant APP expression level throughout life. Similar to the AD brain, plaq ues are associated w ith dystrophic neurites, reactive astrocytes, and microg lia activation. While there is no global neuron loss, such
53 as the extensive cor tical and hippocampal loss seen in AD, neuronal loss is observed in the CA1 region of very old 22-23 month old APP + PS1 mice. Behavior of APP+PS1 Mice Just as in the single tr ansgenic APP (Tg2576) mice, th e double transgenic mice have been widely behaviorally characte rized. These mice often show behavioral correlations with their ADlike pathology described above. In sensorimotor tasks, double transgenic mice are often not signi ficantly different from non-transgenic counterparts; however there are selec tive tasks and time points where the double transgenic mice show impairment. In a ta sk of open field activ ity, double transgenic mice were shown to have increased line crosse s (activity) at 15-17 months, but not at 5-7 months (Arendash et al., 2001b). This in creased activity at 15-17 months was not reproducible in a later study (Jensen et al, 2005). In c ontradiction with the above work (Arendash et al., 2001b), decreased line crossi ng was observed at 4.5 to 6 months (Jensen et al. 2005), but not 15-16.5 mont hs. Liu et al. (2002) observed a decrease in activity in doubly transgenic mice at 7 months. Other te sting in open field has shown no significant differences in activity (Roach et al, 2004). The variability in findi ngs across there studies in open field activity may indicate that activity levels are extremely sensitive to background strains. While all animals discus sed have the addition of the APP and PS1 transgene, multiple generations of inbreeding may cause differences in activity across numerous studies. In balance beam, double transg enic mice showed early impairment at 5-7 months and late impairment at 15-17 months (Are ndash et al, 2001a; Are ndash et al 2001b).
54 After numerous generations of breedings, impairment in balance beam was no longer apparent at 4.5-6 or 15-16.5 months of age, although strong trends were present (Jensen et al. 2005). It could be ar gued that animals lacking sens ory motor impairment at any time point are a better model to test cognition by eliminating any sensorimotor impairment that would potentially confound any observed cognitive effect. However, no correlations between activity, balance b eam and cognition were observed by the Arendash laboratory, indicating that changes in activity or impaired balance beam performance do not deleteriously affect cogni tion. Arendash et al. (2001) showed no effects of transgenicity in the elevated pl us maze (test of anxiety) at 5-7 and 15-17 months. In contrast, Jensen et al. (2005) describe an increase in anxiety at 15-16.5 months. Though this finding was observed in one of 3 measures of anxiety (% time in open arms), it is this measure that is mo st linked to anxiety (Jensen et al., 2005). Similarly to single APP mice, doubl e transgenic mice are impaired cognitively over a multitude of cognitive domains. While one would anticipate the increased A burden earlier in the double transgenic mi ce would translate into earlier cognitive impairment, this is often not observed. In 3 month old animals, Holcomb et al. (1998) showed significant differences in Ymaze alternations be tween single and double transgenic mice when compared to non-transg enic. Both single and double transgenic were equally impaired when compared to nontransgenics. Later, in 3 and 6 month old animals, Holcomb et al. (1999) showed e qually significant differences in Ymaze alternations between single and double transgenic mice when compared to nontransgenic. This impairment in Y-maze was only present in the double transgenic mice at the 9 month time-point However, at no time point we re the single transgenic mice
55 (lacking compact A ) significantly from the double transgenic mice (Holcomb et al, 1998; Holcomb et al, 1999). This work uncovers impairment not associated with compact plaque deposition because APPsw mice lack A deposits at all test points evaluated, leading to the hypothesis that early soluble (oligomeric) forms of A may be the primary cause of impairment in this task. Later work showed no transgenic impairment in Y-maze at any time point in the doubly transgenic mouse line (Arendash et al., 2001a; Arendash et al., 2001b; Jensen et al., 2005). The authors suggest that the Ymaze task is relatively insensitive to mu tant APP-transgenic associated cognitive impairment. In an early behavioral characterizatio n of the APP + PS1 mouse Holcomb et al. (1999) observed no significant impairment at 3, 6 and 9 months in MWM, a task of spatial reference learning and memory. Wh ile there was no impairment in this study, a substantial body of literature shows that APP+PS1 mice become impaired in MWM (Arendash et al., 2001; Liu et al .,2002; Trinchese et al., 2004; Je nsen et al., 2005). At 57 months, double transgenic mice showed no impairment over 10 days of testing, however by 15-17 months spatial learning impa irment was observed (Arendash et al. 2001a). Arendash et al. (2001a) also observed no effects of transgen city on the percent of time spent in the former platform-containing quadrant at either 5-7 or 15-17 months. In more recent work by Jensen et al. (2005) from the same lab, and using the 10 day MWM protocol, double transgenic mice were signif icantly impaired at both 4.5-6 and 15-16.5 time points in acquisition. Probe trial testing for reference memory indicated that 4.5-6 and 15-16.5 month old mice non-transgenic mice showed an exclusive quadrant preference for the platform containing quadrant, but not APP+PS1 mice at the same ages.
56 These early acquisitional and memory retention impairments seen by Jensen et al. (2005) are in contradiction to previously menti oned work by the same lab (Arendash et al. 2001a). Since both studies used the same water maze protocol, variation in protocol is a moot point in this comparison. A likely reason for the observed early impairment involves the numerous generati ons of cross breeding between studies. This resulted in the APP+PS1 mice becoming more behaviora lly sensitive to mutant APP expression and/or the process of A aggregation. Thus, APPsw + PS1 mice appear to have become more impaired earlier as seen by Jensen et al. (2005). Similar to Je nsen et al (2005), Trinchese et al. (2004) noted no impairment in MWM until 6 months. It should be noted that differences in water maze protocols o ccurred. The Arendash protocol included 4 trails for 10 days of acquisition while Trinchese only tested in 3 trials, for 3 days 3 times a day. Changes in behavioral sensitivity are al so observed when analyzing results of the platform recognition task. In this task of recognition and id entification, Arendash et al. (2001a) observed no differences between non-tr ansgenic mice and APP+PS1 mice at 5-7 or 15-17 months of age. In contrast, Jens en et al. (2005), saw no impairment at 4.5-6 months, but did observe impairment at the 15 -16.5 month time point. This impairment is further evidence of increased behavioral se nsitivity after numer ous generations of crossbreeding. An early symptom of AD is the loss of wo rking memory. Using a task of spatial working memory (Radial Arm Water M aze RAWM), doubly transgenic mice 15-16 months were observed to be impaired in th e final block of testing on the final memory retention trial (trial 5) (Gordon et al., 2001; Morgan et al., 2000; Arendash et al., 2001a).
57 Similarly to findings in formerly discusse d cognitive tasks, Y-maze and Morris water maze, the doubly transgenic mice were not significantly more impaired than single transgenic mice (Morgan et al., 2000). This observation provides more evidence that the higher burden found in the APP + PS1 mice does not translate into intensified impairment. Gordon et al, (2001) also s howed that the working memory sensitive RAWM correlated with A deposition in the frontal cortex and the hippocampus. It is important to note that young mice free of A burden (5-6 months) did not show impairment in RAWM in early studies (Arenda sh et al., 2001a). In this initial work, APP+ PS1 mice showed impairment only in the T5 memory retention trial, with no impairment in T4 (Gordon et al., 2001; Are ndash et al., 2001a). In more recent work, Jensen et al. (2005) also describes a somewhat earlier and more profound working memory impairment than had been reported by the same group (Arendash, 2001a). Jensen et al. (2005) observed RAWM impairme nt not only at the later 15-16.5 time-point, but also months at 4.5-6 months. At both 4.5-6 & 15-16.5 month time-points, doubly transgenic mice showed robust impairment on T4 and T5 across all 3 blocks. This early RAWM impairment, taken collectively with ea rlier impairment in the MWM task, shows that, for the Arendash labÂ’s results, numerous generations of crossbreeding lead these transgenic mice to become more behaviorally sensitive to mutant APP expression and/or the process of A aggregation. In another study from the same group, RAWM working memory impairments were observed during T4 in very old 18 month APPsw + PS1 animals (Austin et al., 2003), although no impairment was observed on T5 in 18 month old animals. Finally, Trinchese et al. (2004) observed another charac teristic seen in AD patients. At 2 months, doubly transgenic mice were not impaired in working (RAWM) or
58 long term memory (Morris water maze). At 3-4 months, the mice still exhibited intact long term memory, however th ey showed impairment in RAWM working memory. By 6-7 months Trinchese et al. (2004) show ed that the double transgenic mice show impairment in long term and working memory tasks. While this impairment profile includes earlier impairment than that of fo rmer work, progression of memory impairment closely mimics the early impairment of work ing memory followed by later impairment in long term memory seen in AD patients. In summary, the APP+PS1 transgenic mice from the Arendash colony first show reference memory and working memory impairme nt at 5-6 months. Jensen et al. (2005) showed this impairment in Morris Water Maze Acquisition/ Retention and RAWM. This impairment is also observed in 15-16.5 m onth old animals. Present at 15-16.5 months, that was absent in young mice, is an iden tification/strategy switching impairment, as measured by platform recognition (Jensen et al., 2005). While in creased activity has been observed in APP+PS1 mice at 5-6 mont hs of age, no changes in sensorimotor function or anxiety have been observed at this time point (Jensen et al. 2005). APP Mutant Mice: Imperfect Transgenic Model While there are many characteristic s of transgenic mouse models that closely emulate characteristics of AD, APP transgenic mice are not a perfect model of the disease progression, pathology or behavi or. The behavioral characte ristics of AD are broad and often inconsistent between patients. The co mplex layered behavior of the disease cannot be broken down into specific non-overlappi ng categories such as exploration, anxiety, long-term and short-term memory. Despite these complications, cognitive impairment
59 involving spatial working me mory and reference memory impairment is nicely reproduced in APP transgenic lines. Trinch ese et al. (2004) show ed the progression of typical AD-like cognitive impairment in A PP+PS1 mice, first showing working memory impairment and later reference memory impairment. The age-related platform recognition impairment observed in APP + PS1 mice by Jensen et al. (2005) nicely mimics the strategy switching impairment obser ved in AD patients. APP transgenics also show the characteristic anxi ety often observed in people w ith AD (Jensen et al., 2005). There are some behavioral characteristics of AD that can not be reproduced in transgenic mice including, hallucinations, depression, and pa ranoia. While not perfect, behavioral testing of transgenic mice to analysis therap eutics and mechanisms is a noteworthy tool which has been paramount in exploring the disease. As for pathology, while at first gl ance the APP transgenic mice do appear to closely mimic AD brain pat hology there are some notable differences. The first significant difference is the lack of overt neur on loss in the transgenic mice. Despite high concentration of the neurotoxic A in the mouse brain, this do es not translate into the neuronal loss observed in AD brai ns. This could be due to A being less toxic in the mouse brain when compared to human brains or the short life-span of the mice relative to humans. Lack of neuronal loss could also be due to the much decreased inflammatory response A has in the mouse brain. Evidence for this hypothesis was observed by Schwab et al. (2004) in a direct comparison of lesions in the neocortex and hippocampus in elderly APP23 transgenic mice and lesi ons from the AD brain. Despite similar staining for A protein, positive ApoE staining, and comparable levels of reactive astrocytes, a largely reduced level of mi croglia activation was observed in aged
60 transgenics. In contrast to lesions of the AD brain, Tg+ mice had weakly activated microglia, which expressed low levels of comp lement receptor activation. Schwab et al. suggest that this weak immune response in Tg+ mice compared to the very strong immune response in human AD could be why A vaccination in mice was useful and resulted in the clearance of A whereas in AD, stimulation to an already strongly activated immune system had gr ave results. Differences in microglia activation make it difficult to anticipate if immune-activating th erapies that result in positive effects in mouse models, will result in similar effects in humans. Another distinct difference in APP transgenic mice observed by Schwab et al (2004) was the absence of NFTs in the Tg+ mice. In the AD brain, A deposits were surrounded with the presence of ghost tangles, while the neurons around the A deposits in the Tg+ mice appeared to just be displaced (Schwab et al. 2004). The A hypothesis theorizes that presence of A induces production of NFTs. APP transg enic mice are free of NFTs despite high plaque burden. Overall, it is hard to decipher if the mechanism leading to im pairment in transgenic mice is the same as causes of impairment in hum an AD. While flaws of the APP models are present, the development of mice with Â“AD likeÂ” behavior and pathology has been monumental in advancing AD research.
61 Environmental Enrichment Enriched Environment Protection and Treatment: Human Studies A large body of literature suggests that life experiences including education, occupation, physical activity and social inte ractions may provid e protection against dementia later in life. These findings woul d suggest environmental factors could impact the development of AD. Cognitive stimulati on in the forms of educational attainment, occupations, participation in non-occupa tional cognitive activities, Â“environmental complexity,Â” Â“favorable life experiences,Â” a nd leisure activities have been examined in association with decreased AD risk. Much of this work is the pr oduct of retrospective case-control studies. Low educa tion attainment has been associ ated with an increase risk of AD (Stern et al., 1994, Ott et al., 1995, Le tenneur et al., 1999). Why education is protective against AD is not clear. Possible explanations include that education provides a cognitive reserve, while others suggest that higher levels of education enables patients to simply preformed better on neurological tests. Some have suggested that low education results in earlier ons et of AD (Friedland et al., 19 93). Stern et al. (1994) found that low levels of education resulted in an increase risk of AD. However, they were unable to decipher if lower levels of risk, a ssociated with higher education, was due to Â“decreasing ease of clinical det ection or by imparting a reserv e that delays the onset of clinical manifestations.Â” In a 5 year longitudinal study Le tenneur et al. (1999) supported
62 the above work, showing low educational attain ment is associated with higher risk of AD. While many have found this correlation, se veral studies have failed to confirm this association. After adjusting for age and examining subtypes of dementia, Cobb et al. (1995) found no association betw een risk of AD and education. While education is participated in during early-lif e, work by Friedland et al. ( 2001) would suggest mid-life cognitive activity is associated with lower ri sk of AD. Occupa tional attainment falls into the category of mi d-life cognitive stimulation. Both St ern et al. (1994) and Smyth et al. (2004) observed delay of AD development if subjects had held mentally stimulating occupations. Other mid-life activities, such as social and leis ure activities, physical activity and cognitive stimulation have also be en investigated and f ound to be associated with decrease risk of AD (Scarmeas et al., 2001; Wang et al., 2002). Verghese et al. (2003) showed that individuals that participated in leisure activities such as reading, playing board games, playing a musical instru ment and dancing had a decrease risk for developing AD. Work by Schooler & Mula tu (2001) suggests that carrying out substantively complex tasks late in life st ill has the capacity to improve intellectual functioning. Moreover, physical activity perfor med in mid-life has also been linked to a decrease in AD (Friedland et al., 2001; Churchill et al., 2002 ). As a component of the Â“Nun Study,Â” Wilson et al. (2002) provided evidence that cognitively stimulating activities participated later in life was associ ated with decreased risk of AD. A large group of non-demented clergy was followed an nually for up to 7 years. The data collected showed that particip ation in cognitively stimula ting activities wa s associated with a decrease of global cognitive impairmen t, working memory impairment as well as perceptual speed impairment (Wilson et al ., 2002). Taken together, these findings
63 suggest that lifestyle in both early and mid-life may influe nce the development of AD. More work is needed to show if each component of enriched lifestyles (social, physical, cognitive activity) are contribu ting equally to the effects obser ved or if one weighs more heavily as a protector. Such a dissection would, however, be most difficult to achieve in human studies. A substantial body of literature has s hown EE as a protective entity. This has lead to the investigation of EE as a possible treat ment option. Many treatment studies focus on the potential of Â“cognitive rehabilitative interventionÂ” as a treatment for AD. Ball et al. (2002) showed that inte rventions conducted in small group settings lasting 60-72 minutes over a 5-6 week period of time resu lted in improvement in Â“targeted cognitive abilitiesÂ” in older adults. Quayhagen et al (1995) extended these findings to AD patients by showing improved overall cognitive /memory function following cognitive rehabilitative intervention. Weekly cognitive sessions were performed by caretakers. Those in the treatment group initially show ed improvement in behavior as well as memory. However, this benefit declined to baseline after 9 months. In contrast, the placebo group declined in cognitive test scores over the course of the clinical trial. While relatively short lived, the 9 month treatmen t benefit observed closely competes with current pharmaceutical treatments that are ofte n effective up to one year. It is also important to note that intervention was only ad ministered in a short term (5-12 week) biweekly manner, suggesting an even greater potential of cogni tive rehabilitative intervention in a chronic and more rigorous fram ework. In more recent work Davis et al. (2001) and Farina et al. ( 2002) observed modest cognitive benefits from cognitive rehabilitative intervention. For Davis et al. (2001), cognitive intervention included 5
64 weeks of training in face-name associations, sp aced retrieval, and cognitive stimulation. When patients were analyzed postcognitiv e intervention, there was improvement in tasks which were included in cognitive in tervention sections (face-name associations, spaced retrieval). However, these benefits did not translate into improved overall neuropsychological testing or improved quality of life. Similarly unimpressive cognitive benefits after cognitive interv ention were observed by Farina et al. (2002). Two types of cognitive intervention were te sted in this study. Test s ubjects were given Â“procedural memory interventionÂ” or training of Â“partially spared cognitive functions.Â” In a test of Â“functional living skills,Â” both groups showed improvement. However, only the group given Â“procedural memory interventionÂ” showed improvement on the Attentional Matrices and Verbal Fluency for Letters tasks. More recent work by Loewenstein et al. (2004) showed that mildly impaired AD patients that were enrolled in a cognitive rehabilitation (CR) program maintained pe rformance on specific cognitive and functional tasks. CR training consisted of two 45-mi nute training sessions twice per week for 14 weeks. Similar to other studi es, Davis et al. (2001), perfor med CR training in specific tasks (face-name association tasks, object recall training, functiona l tasks) and observed benefits within these trained tasks. As disc ussed above, these more recent studies also involved acute and sparse cognitive intervention that may not be as beneficial as a longterm intensity therapy. More work i nvolving chronic administration of cognitive intervention is needed in order to better gauge the potential of cogni tive intervention as a viable treatment option for AD.
65 Effects of Enriched Environment in Rodents The effects of EE in rodent are often compared between standard group housed rodents and rodents living in larger cages containing toys, wheels and tunnels. Most work is done with animals living in the a ugmented cage (Kempermann et al., 1997) while other studies have acute exposur e for a few hours a day (Frick et al., 2003). Some studies enhance their EE by administering intra-weekly cognitive sessions in which animals are taken from their cages and exposed to new environments (Arendash et al. 2004; Teather et al. 2002). Many studies will add an Â“im poverishedÂ” test group including singly housed mouse in standard cages (Mohammed et al., 1990) Do to the complexity of the enriched environments, it is hard to piece out if co mplete enrichment (cognitive, physical, and social stimulation) or if one or several components of EE are leading to the changes observed. Cognitive Effects of Enriched Environment in Rodents While there is a limited amount of work involving EE and cognitive benefits in AD transgenic mice, a substantial body of literature offers insight into the effects of EE on cognition in wild type mice. While it is unknown if the cognitive benefits often observed in wild type mice will translate into beneficial effects in AD transgenic mice, the work done in typical rodents offers an e xploitable resource to predict possible effects of EE in AD transgenic mice.
66 The effects of EE have been te sted in young, middle and aged rodents. Furthermore, studies have al so tested effects of enrich ed environment on genetically manipulated, dietary altered, and lesioned rode nts. Locomotor activity is among many of the behaviors explored in animals exposed to EE. Van Waas and Soffie (1996) tested if EE changed activity levels in rats. In a ymaze task, despite age, young (4 months) and old mice (22 months) reared in an EE did not have increases in Â“arms visited.Â” This work was corroborated by Wolfer et al. (2004), who found no significant differences in enriched Â“adultÂ” mice versus standard housed mi ce in an open field task. This absence of differences in activity in enriched animals suggests that differences in performance in exploration and cognitive tasks are not due to underlying differences in activity. EE mice also show no differences in anxiety levels, as measured by the elevated plus maze (Tsai et al. 2003). Hippocampal forms of learning and memory have been shown to be widely affected by EE ( Kempermann et al., 1997, 1998, 2002; Wincur et al. 1999; Teather et al. 2002; Frick et al. 2003; Arna iz et al. 2004). Kemperma nn et al. (1997, 1998) have constantly showed that mice exposed to an EE from weaning to adulthood ( 2, 6, and 18 months) perform superior in Morris Water Maze when compared to standard housed mice. Work by Teather et al. (2002) showed 6 month ol d rats living in an EE had superior performance in hippocampal dependent tasks (standard water maze when compared to Â“restrictedÂ” housed (RC) rats In contrast, EE mice and RC rats were identical on tasks that were hi ppocampal independent (Teather et al. 2002) Frick et al. (2003) showed that EE reduces age-related impairment in spatial memory. In the Morris Water Maze (MWM) task, 18 month old mice liv ing in an EE 25-29 days prior to testing
67 were not impaired while those in sta ndard housing showed impairment in both acquisition and retention portions of the MWM. It is important to note that no treatment differences were observed in the cued vers ion of the MWM, showing that EE is not simply improving vision of the aged mice. Equally noteworth y, after statistical separation of males and females, both benefite d from EE. This would suggest that both males and female could potentially benefit fr om age-related cognitive protection provided by enrichment. Kempermann et al. (2002) show ed that EE late in life benefited aged mice. In this study mice entered EE at 10 months and were observed to be cognitively superior in MWM when compared to SH mice at 20 months. Benefits in rats were also observed by Arnaiz et al. (2004). This work showed that rats at 27 months of age (Arnaiz et al., 2004) benefited in MWM from life time EE. In contrast to this work, Wolfer et al. (2004) observe no differences in MWM betw een Â“adult miceÂ” living in an EE when compared to standard house mice. This di screpancy could be due to the age of animals tested. No definition of adult mice was gi ven, raising the possibility that the animals tested were too young to have any impairment to protect against. The work of Wolfer et al. (2004) would suggest that EE may not be useful for augmenting baseline cognition. Duration and starting time of EE on cognition was investigated by Williams et al. (2001) and Kobayashi et al. (2002). Williams et al. (2001) tested mice at numerous time points in the MWM after exposure to enrichment in ear ly life (35 to 94 days of age) or later in life (100 to 159 days of age). Three rounds of (MWM) testing were included over the extent of the study. A baseline test, follo wed by a first and second test period was conducted. This extensive testing can lead to over testing and therefore mask impairments or cognitive benefits from EE. While the authors suggest that EE at the
68 earlier period of time offered benefits in MW M performance, these benefits were modest at best. No difference amongst any of the groups (individually housed, group housed, enrichment early, enrichment late) was observed at the first test point In the second test period, differences were observed only betw een individually housed mice and all other groups. This lack of robust differences in the EE mice could be due to the 3 rounds of 5 day MWM testing. Kobayashi et al. (2002) te sted if long-term exposure offered greater benefits than short-term exposure. In th e cognitive based Hebb-Williams Maze task, rats reared in an EE from weani ng until 25 months benefited more from EE than those given short term (3 months) of EE right before te sting (e.g. EE between 22-25 months of age) However, both long term and short term EE benefited 15 month old mice equally. The authors concluded that short-term enrichment has potential to improve aging animals; however, long-term enrichment is superior in protecting cognition in ag ed animals. This work offers data that would suggest living an enriched life may lead to successful aging, as well as showing that while not as impressive as long-term EE, short-term enrichment later in life may also be beneficial. Until this point, experiments conducted w ith EE as the sole variable have been discussed; however, EE has been explored in the presence of other variables. Winocur and Greenwood (1999) showed high dietary fat translated into negative cognitive effects that can be overcome with EE. Kalmijn et al (1997) showed that high fat intake leads to increased risk of dementia in humans. Taken together, these studies suggest EE may reduce risk of AD by reversing mechanisms of cognitive impairment that fat intake exacerbates. Traumatic brain injury and ischemic stroke often result in cognitive impairment. Wagner et al. (2002) and Dahlqvist et al. (2004) showed positive cognitive
69 effects after EE treatment in male rats after traumatic brain injury and rats after focal cerebral ischemia, respectively. Neurotrans mitter abnormities are associated with and targeted by pharmaceuticals in AD. Degroot et al. (2005) found increased hippocampal acetylcholine (ACh) efflux when actively mani pulating a novel object. These effluxes lead to an enhancement in cognition when te sted in a radial arm maze. This would suggest that if EE works prim arily through a mechanism of in creased ACh, its therapeutic value for late stages of AD may not be significant due to the substantial loss of cholinergic neurons in the AD brain. This po ssibility appears remote, however, in view of multiple other neurochemical, histologic, a nd genetic mechanisms that are impacted by EE, as will be discussed in the next secti on. Overall, EE has been shown effective in healthily aging, as well as in reversing effects from non-heal thy insults such as high fat diets and traumatic brain injury. Neurohistologic and Neurochemical Effects of Environmental Enrichment in Rodents A plethora of changes in ne urohistology and neurochemistry have been observed in rodents exposed to EE. One widely doc umented change is increase neurogenesis. Contrary to past thought, th e brain has the capacity to produce new nerve cells in adulthood. Two areas have been shown to give rise to new neurons. These areas are the subventricular zone of the anteri or lateral ventricles (which gi ve rise to cells that become neurons in the olfactory bulb) and the subgranular zone in the dentate gyrus (which generates new granule cell neurons in the hippocampus). EE has been shown to induce hippocampal but not olfactory bu lb neurogenesis in rodents (Brown et al., 2003). This
70 selective neurogensis by EE in the hippocampus an area incongruently destroyed in AD, makes it a noteworthy area of explor ation in treatment of AD. In 1997 Kempermann et al. showed more hippocampal neurons in adult rodents living in an enriched environment. Th is work was consistently reproduced in Kempermann et al., 1998a, 1998b, 1999, 2002 and Nilsson et al., 1999. Implementing bromodexyuridine (BrdU) (which is incorpor ated into dividing cells and their progeny) and stereology techniques, Kempermann et al. (1997) was able to measure a 57% percent increase in BrdU labeled cells and a significant increase of 15% in volume of the granule cell layer in 3-4 month old mice that had lived in EE for 40 days. Positive BrdU staining could indicate increase neuronal proliferation or simply increase survival of neurons. To explore these options Kempermann et al. ( 1997) sacrificed mice one day after BrdU injections and 4 weeks af ter injection. In th e group sacrificed one day after injection, no significant differences were observed in EE mice when compared to control mice (Kempermann et al., 1997, 1998a). However, in animals sacrificed 4 weeks after injections, a significant increas e in labeled cells was obser ved. Take together these findings show that EE alone does not increase proliferation of neurons, but rather is increasing their survival. This survival may be impart due to the inhibition of spontaneous apoptosis observed in rats re ared in EE (Young, 1999). Young et al. (1999) described a 45% reduction in apoptotic death in the rat hippocampus. This EE-induced prevention of apoptosis could be a mechanis m by which EE is improving survival of new neurons. It is important to note that increased survival with no changes in proliferation is background strain dependent. In contrast to work above performed on behaviorally superior C57BL/6 mice, 129/SvJ mice [(whic h have significantly less hippocampal
71 neurogensis and do not perform well in le arned tasks (Kempermann et al, 1997b)] had increased survival and proliferation of ne urons (Kempermann et al. 1998b). This is different from previous st udies using C57BL mice, whic h reported only increased survival, not proliferation. Along with st rain background dependence (Kempermann et al., 1998b), EE-induced neurogens is is also dependent on temporal variance. Kempermann & Gage (1999) found that animal s exposed to EE and then removed from EE had twice as many proliferating cells in th e dentate gyrus when compared to standard housed and long-term exposed animals. Ke mpermann and Gage (1999) hypothesize that this increase of neurogenesis in short-term enrichment could be due to early stimulation preserving neurogenic potential, combined w ith novelty (changing envi ronments),both of which contributed to the EE-i nduced neurogenesis. While no velty may be contributing to the short Â–term enrichment effect, both groups had exposure to EE early in life, making this hypothesis that short-term enrichment effects were due to early life exposure somewhat weak. The work discussed thus far has been done in adult animals. However, does EE have the potential to induce neurogenesis in aged mice? Kempermann et al. (2002) observed a five-fold induction of dentate gyru s neurons in 20 month old mice. Animals were placed in enrichment or standard housi ng at 10 months old and then maintained for 10 additional months. At first analysis, no si gnificant differences we re observed in BrdU staining. However, when neuronal phenot ype was taken into account, it was observed that the control group had increased astrocyt es in a compensatory manner. When looking just at BrdU staini ng, which over lapped with Ne uN staining for neurons, a 5fold increase in neurogensis of EE mice emer ged. This work shows that beginning EE in
72 middle age may be just as beneficial as st arting early. This work also highlights the ability of EEÂ’s benefits to extend into late life. The work above exemplifies th e power of environmental enrichment in inducing neurogenesis in adult and aged wild type mice, but what are the implications of EE induced neurogenesis in AD? Haughey et al. (2002) showed disruption of neurogensis by A in the dentate gyrus of AD transgenic models. In 12-14 month old APP mice Haughey et al. (2002) described a reduction in neural progen itor cell when compared to age matched controls. Twelve days after a finial Brdu injection there was a greater decrease, 55%, in APP mutant mice compar ed to the 25% decrease in control mice (Haughey et al., 2002). In contra st to what is observed in transgenic models, Jin et al. (2003) described an increase in neurogensis at autopsy in the dentate gyrus of AD brains when compared to brains of individuals w ithout neurological diso rders. The authors suggest a compensatory theme which occurs wit hout cognitive benefit. This brings into question if increased neurogensis, through EE, w ill translate into benefits in AD patients. Jin et al. (2003) suggest po ssible reasons for the limited repair capacity of increased neurogenesis in AD. Authors note the sma ll amount of dentate gyrus region-limited neurogenesis is not enough to compensate for the mass destruction present in the AD brain. A second point is that the microenvironment of AD br ains may be too toxic for new neurons, thus keeping them from becomi ng fully functional, mature neurons that integrant into surviving brain circuitry. While these obstacles are present, the evidence that the AD brain has the capacity to maintain neurogenesis offers the possibility that measures which increase neurogenesis, such as EE, may have therapeutic value in AD.
73 While neurogenesis has been the most widely documented effect of EE, other pathological changes are observed after exposur e to an enriched environment. Synaptic loss correlates with cognitive deficits asso ciated with AD ( DeKosky, 1990). Mice 27-28 months that were exposed to EE for 3 hours a day for 14 days prior to behavioral testing showed increase synaptophysin levels comb ined with cognitive benefits (Frick & Fernandez, 2002). While increased synap tic area was associated with improved cognition, no correlation analysis was perfor med. Increased synaptophysin, indicative of synaptic area, was increased in the frontalparietal cortex and the hippocampus. Further evidenced of increased synaptic area induced by EE was observed by Saito et al. (1994). Saito et al. (1994) showed a decrease in synap tic content in aged mice that was prevented if reared in an EE. Noteworthy for AD is the finding that mice with an ApoE4 transgene do not show EE-induced synaptogenesis (Lev i et al, 2003). This may suggest that presence of ApoE4 blocks some effects of EE exposure. While synaptophysin is a measure of synaptic area, changes in de ndritic morphology offer another measure of neuronal health and synaptic plasticity. Using golgi-cox morphological analysis, Faherty et al. (2003) found increases in dendr itic length in the CA1 region and the dentate gyrus of mice reared in an EE until 4-5 months of age. Spine density was analyzed by Comery et al. (1995). These investigators found a 30 percent incr ease in spine density within the corpus striatum of 2 month old rats exposed to EE, compared to individually reared littermates. More recently in 3 month old deer mice, Turner et al. (2003) observed increases in average number of spines in la yer V pyramidal neurons of the motor cortex and in medium spiny neurons of the dorsolate ral striatum. No differences were observed in the granule cells of the dentate gyrus.
74 EE has been shown to change growth fact or levels in the brain ( Pham, et al., 1999; Ickes et al., 2000). Work by Pham et al. (1999) showed that NGF levels and NGF receptors are both increased in 14 month old rats housed in an EE. These increases occurred in the hippocampus, visual and ento rhinal cortices afte r 12 months of EE and were linked to improvement in spatial lear ning, although no correlation analysis was run. Along with increases in NGF, Ickes et al. (2000) observed regional increases in brainderived neurotrophic factor (BDNF) and neur otrophin-3 (NT-3) levels in 14 month old rats that had lived in EE from weaning. Both BDNF and NGF were increased in the very AD relevant hippocampal formation. In add ition to changes at the protein level, it is important to note that EE has been show to effect gene expression. Genes linked to neuronal structure, neuronal plasticitly, a nd transmission have been linked to EE (Rampon et al., 2000, Keyvani et al. 2004, Pinaud et al. 2004). Pinaud et al. (2001) showed rats exposed to EE 1 hour a day for 3 weeks had marked up-regulation of arc in the cerebral cortex. This ge ne, associated with plasticity, was also up-regulated in the CA1, CA2, and CA3 hippocampal subfields. Rampon et al. (2000) analyzed various time exposures to EE and changes in gene expression. Four month old mice were exposed to EE for 3 hours, 6 hours, 2 days or 14 days. Early gene expression changes were measured in a comparison of mice expos ed to EE for 3 or 6 hours. Seventy-seven percent of genes were consiste ntly altered at bo th 3 and 6 hours. Some up-regulated genes included DNA methyltransferase (maint enance of DNA during replication), myelin gene expression factor and es trogen-responsive finger protei n (transcription factors). Genes that were down regulated include d caspase-6 (pro-apoptotic) and proyl oligopeptidase (regulated degrada tion of neuropeptides). For animals that were in EE for
75 2 and 14 days, a different late gene expr ession was observed. For these time points, changes in genes involved in neuronal tr ansmission and structure were the common theme. X-box binding protein (involved in the cAMP pathway) and postsynaptic density 95 (involved in NMDA receptor anchoring) are two example of up-regulated gene expression at the 2 and 14 day time points. EE has also been shown to effect stre ss-related organs and hormones Despite beneficial effects of EE, Mon cek et al (2004)Â’s work woul d suggest EE to be a chronic stress environment. Changes indicative of st ress, such as increased adrenal weight and increased stress hormones such as corticor sterone and adrenocorticotropic hormone (ACTH), were observed in EE mice. In contrast Belz et al. (2003) showed a decrease in base line stress hormones and a decrease in stress hormone response after mild stress induction in EE mice. The differences in these studies can be attribut ed to differences in housing protocol. In the study that found incr eases in stress, 10 male rats were group housed in an enriched environment. In c ontrast, Belz et al. (2003) measured stress responses in singly housed mice with toys a ugmenting their home cage. The increased stress measurement found in the Moncek et al. (2004) paper could be the result of a hierarchy competition in the EE cages. In summary, EE has been shown to effect measures at the gene to pr otein level, which often tran slate into morphological and behavior benefits. Effects of Environmental Enrichment in Â“AlzheimerÂ’sÂ” Transgenic Mice While limited, some work involving th e effects of EE on Â“AlzheimerÂ’sÂ” transgenic mice has been done. The literature thus far is inconstant from a pathological
76 standpoint and sparse behaviorally. The curre nt literature can be broken down into EE as a protective entity and EE as a treatment. Fo r protective studies, mice are either exposed or reared in an EE from weaning while trea tment studies involve putting animals together after behavioral impairment and pathology is present. Jankowsy et al. (2003) offered the first work in preventative EE and AD. APP/ PS1 females were used and empha sis was placed on EE effects on A Mice were keep in an EE until 8.5 months of age at which time they were sacrificed. The methods state that animals from different cages were combined midway through the study and new young animals were added as older animals were removed. Despite 64 animals used, results were most impressive in 3 sibling pairs, which were hi ghlighted in the results. For these 3 sibling pairs, Jankowsky et al. (2003) found increases in A burden. Results were more profound in hippocampus than cortex in numerous measures. In a measurement of aggregated A by size-exclusion filter trap assay, EE mice had a 30% increase in A in the hippocampus with only a 16 % in crease in the cortex. Total A measured by ELSIA, showed a 52% increase in the hippocampus vers es a 31 % increase in the cortex. These results came as a surprise, taking into acc ount the numerous positive effects of EE that have been documented. Rather than hi ghlighting possible mechanisms by which EE exacerbates A deposition, it may be more appropriate to discuses why this EE protocol exacerbated A deposition. The paradigm described above may have lead to a more stressful than enriching environment for th e mice measured. The adding and subtracting of animals throughout the study removed a stable social base that could be considered crucial to the EE effect. This lack of a st able social environment may have also added stress. While not as extreme among females, a social hierarchy is still present and an
77 always changing hierarchy may have added to this stress. Also it is important to note that the majority of the results showing an increase in A were seen in 6 out of 64 mice. In direct contradiction to Jankowsky et al. (2003), recent work by Lazarov et al. (2005) showed that EE reduces A pathology in transgenic mice. Similarly to the protection protocol of Jankowsky et al. ( 2003), APP/PS1 transgenic animals entered EE or SH at weaning and were exposed to an EE for 5 months. Unlike the former study which used females, only males were used in the latter study. Probing with 3D6 antibodies, EE-housed mice had significantl y less amyloid burden in the hippocampus and the cortex. This theme continued for abundance and size of A deposits using thioflavine S staining. Reduction of non-deposited A levels was also measured. SH mice has significantly more detergent soluble and formic acid soluble A 40 and 42. Taken together, these result s how a global decrease in A pathology that spans various species of A Lazarov et al. (2005) also included an anal ysis of gene expression in EE mice. DNA microarray analysis revealed selective upregulation in le vels of transcripts encoded by genes associated with learning and me mory, vasculogenesis, neurogenesis, cell survival pathways, A sequestration, and prostaglandin synthesis. The upregulation of these genes appear to be constant with e ffects observed in wild type mice. The 11.244.3-fold increase in A sequestering genes may be respons ible for the robust decrease of A in many forms. It is also important to note that the physical ac tivity component of EE was dissected and analyzed in this paper. Results of this subsection will be highlighted below. Overall Lazarov et al. (200 5) showed a robust decrease in brain A levels
78 coupled with positive upregulation of genes th at may have lead to the decreases in A in transgenic mice. Work by Jankowsky et al. (2005) is the firs t to show that EE ha s the potential to protect against cognitive impairme nt in a transgenic model of AD. Mice expressing APP and/or PS1 were placed in EE at 2 months of age and tested cognitively at 8 months of age. Mice were tested in standard Morri s Water Maze (MWM), repeated reversal MWM, and radial arm water maze (RAWM). Analysis of behavioral data demonstrated that mice with APP alone or in combination with PS1, which lived in a standard house, were impaired in hippocampal-dependent learni ng and memory. In MWM acquisition, both the APP and APP+PS1 mice significantly benefited from enrichment when compared to standard housed mice. Compared with stan dard housed mice of the same genotype, both APP and APP+PS1 in enrichment housing mice swam significantly shorter distances in MWM acquisition. While an improvement was observed, enrichment did not enable the APP+PS1 transgenic mice to perform as well as the enriched non-tr ansgenics or the PS1 mice in MWM acquisition. MWM retention bene fits were only observed in APP single transgenic mice. Time spend in the goal quadrant for MWM retention was significantly increased in APP mice living in an EE. In c ontrast, an improvement in retention was not observed for APP+PS1 mice living in EE. Doub le transgenic EE mice spent significantly less time in the goal quadrant than all other genotypes given EE. Similar results were observed in the repeated-reversal version of MWM and the RAWM. APP and APP+PS1 mice given EE swam shorter distances in repeated reversal MWM and made fewer errors in the RAWM than standard housed mice of the same genotype. This was shown by their ability to decrease Â“average distance traveledÂ” in the
79 MWM repeated reversal task to levels co mparable to standard housed non-transgenic mice. In the repeated-reversal version of the MWM, both enriched APP and enriched APP+PS1 did not decrease their latency as quickly as non-transgenic and PS1 enriched mice. Noteworthy for both the MWM repeated reversal task and RAWM task is that, following EE exposure, both APP and APP+PS1 mice overcame a genotype effect and were not significantly different from standard housed or even enriched non-transgenic mice. This suggests that APP mutant mice (A PP and APP+PS1) benefited more from EE than PS1 or non-transgenic mice. This improvement in cognition was accompanied by increases in A levels. This finding is consistent with earlier work by Jankowsky et al. (2004) showing EE leads to increased levels of A n contrast to the earlier protocol used by Jankowsky et al. (2004), mice were maintained in a consistent social environment throughout the study. While not clear, Jankowsky et al (2005) suggest increased synaptic activity leading to increases in A production as a mechanism by which A levels are elevated w ith EE. Taking together, the review by van Praag (2000) showing enhanced synaptic strength and connectivity in EE mice and the findings by Kamenetz et al. (2003) that increased s ynaptic activity enhances APP processing by BACE1, the mechanism offered by Jankowsky et al (2005) appears reasonable. Their results would suggest that EE can strongly modulate A 's impact on the brain and make it a less relevant impairing factor. While there has been some work show ing pathological changes in AD Tg+ mice reared in EE, there is limited literature showing if the pathological changes translate into cognitive protection. Costa et al. (2005) found that PDAPP+PS1 mice placed in a Â“completeÂ” EE (including social, physical and cognitive stimulation) from weaning are
80 protected from the cognitive deficits observe d at 7.5 months in PDAPP+PS1 transgenic mice. In the platform recognition tas k, a measurement of object recognition and identification, EE mice were identical to th e non-transgenic mice and were completely protected from the cognitive impairment observed in their standard housed (SH) litter mates. Equally impressive was the protecti on in working memory impairment, measured by the RAWM task. For overall latency on trial 4 and trial 5 of the RAWM, SH Tg+ mice were significantly impaired. In contrast the Tg+ mice th at had lived in an EE were indistinguishable from non-transgenic mice and significantly better than Tg+ SH mice. In order to decipher a mechanism through which cognitive protection was achieved, AD pathology and gene micro-array was done by Costa et al. (2005). In order to separate out any effects that could have been due to the intensive behavioral test battery used, a second cohort of mice that di d not undergo behavioral testing was also analyzed. Two measures of A pathology was performed, including measurement of both diffuse and compact A deposition. Diffuse A immunoreactivity, using the 6E10 monoclonal antibody, revealed that th ere was no difference in total A load between Tg+/EE and Tg+/SH mice in either cereb ral cortex or hippocampus for the Â“nonbehaviorally testedÂ” group. Though not signi ficant, a trend towards decreases in A deposition was observed. This same trend wa s seen for compact plaque deposits, as measured by Thioflavin S staining. Importa ntly, Costa et al. (2005) did observe that Â“behaviorally-testedÂ” EE mice had significantly lower A deposition than those mice that underwent only EE. This shows the EE alone was not enough to decrease A however EE and behavioral testing (itself a fo rm of EE) did significantly decrease A levels, suggesting an additive effect.
81 In a micro-array analysis performed in Costa et al. (2005), roughly 70 genes showed statistically significant changes by a f actor of 2.0 or more, in response to EE. Some of the more robust changes that have implications in the AD process included upregulation of A sequestering genes, and neuropr otective genes, as well as downregulation of genes involved in memory impair ment. For example, insulin-like growth factor ( IGF-2 ), which has been shown to play a neuroprotective role against A was upregulated in EE mice by 2.5 fold. A 10fold increase in transthyretin ( TTR ), an A binding protein shown to sequester A and inhibit amyloid form ation, was also observed in the EE mice. Furthermore, phosphodiesterase 4 was down-regulated 2.2 fold and the cholecystokinin receptor was down-regulated 3.2 fold, both of which when experimentally inhibited, are known to improve memory. Overall Costa et al. (submitted) have shown that Â“completeÂ” EE has the potential to protect PDAPP + PS1 mice from cognitive impairment through mechanisms involving reductions in A deposition and beneficial ch anges in gene expression. EE as a treatment in transgenic mice has not been widely researched. However in a single study, Arendash et al. (2004) showed improved c ognition in aged transgenic mice despite stable levels of A Moderately A -burdened APPsw mice entered EE at 16 months of age and lived in EE until 22 mont hs at sacrifice. At 20 months, animals were taken through multiple behavioral tasks, testing a wide range of cognitive domains. In Morris Water Maze (MWM) acquisition and retention EE mice outperformed SH mice. In platform recognition task, whic h requires mice to switch from a reference learning/memory strategy to an identify/r ecognize strategy, EE mi ce easily transitioned
82 while the SH mice lagged behi nd in their effort to conve rt. Discriminant function analysis using 8 behavioral measures was able to separate the EE mice from the SH mice revealing significantly better overall performance. Despite these behavioral benefits obser ved in the EE treated mice, no differences in A were observed. In both the parietal cortex and hippocampus no differences were seen in diffuse or compact plaques meas ured by 6E10 staining and Thioflavin S, respectively. A cognitive benefit wi thout a large decr ease in brain A deposition lead the authors to the possibility that an A -independent mechanism is driving the positive treatment effects of EE. Complete Environmental Enrichment versus Its Components As discussed, the effects of EE are wi de ranging and are in some case quite robust. However, are these effects due to the complete EE experience or are there components of enrichment (social, physical, cognitive) that contribute more than others. While challenging to separate out the compone nts of EE, some have attempted to and have observed EE component dependent results. Recent work observed that walking alone was associated with a reduced risk of dementia (Abbott, et al., 2004). This pros pective cohort study assessed the distance walked per day for 2 years of 2257 men aged 71-93. Men that walked the least had a 1.8 fold increase of dementia. This work woul d suggest that physical activity as modest as walking 2 miles a day could be enough physical activity and enrichment to decrease the risk of dementia. It is important to note th at prospective studies such as these are often challenging to control. It is also a potent ial confound that those who walk more probably
83 also lead healthy lives in other domains of life such as diet. Exercise has been tested as a treatment for AD (Mahendra &Arkin, 2003; Teri et al. 2003). Patients that were enrolled in an exercise and caregiver training program had improved neurological exams (Teri et al. 2003). These improvements lead to a decr ease in institutionaliz ing combined with less behavioral disturbances. Improvement with physical activity was also observed by Mahendra & Arkin (2003) whom describe d a comprehensive c ognitive-linguistic intervention program for mild to moderate AD. Patients that participated in the 4 year program maintained or improved performan ce on numerous measures. While these studies offer some evidence of increased physical activity being beneficial as a prevention and treatment of AD, these studi es did not look at the physical effects independently they were paired with other activities such as supervised volunteer work and behavioral management techniques. In rodent studies, physical activity has been show to enhance genes involved in brain health (metabolic, anti-aging, immunity genes) and plasticity (neurotrophic factor trafficking and vesicle recycling) (Cotman &Berchtold, 2002). Increase BDNF levels (Johnson & Mitchell, 2003), and enhance neurog enesis (van Praag et al., 1999; Kitamura et al, 2003; Ehninger & Kempermann, 2003) were also observed. Recent work by Lazarov et al. (2005) found that EE mice with the highest physical activity had the greatest reduction in A It is hard to decipher if the animals with less A were more active due to less impairment or if the phys ical activity lead to the decrease in A In a transgenic model (TgCRND8) of AD, Adlard et al. (2005) s howed that physical activity decreased amyloid load. Five month old transgenic mice reared in an EE had a 38% decrease in A in the frontal cortex and a 40% decrease in A in the hippocampus when
84 compared to transgenic standardhoused mice. This change in A was not associated with differences in APP expression, or secretase expression. Overall, physical activity appears to have positive effects on brai n health and could be a leading contributor to EE effects observed in humans and rodents. Social activity is another compone nt of EE that has been explored. An epidemiological study in humans showed that socially-oriented ac tivities may protect against dementia (Wang et al., 2002). While so cial activity appeared to decrease risk, a connection between physi cal activities and decreased risk of dementia was not observed. However, mentally stimulating activities did appear to correl ate with decreased risk of AD. Work in rodents has not shown social in teraction alone as a benefit variable. Early work by Rosenzeig et al. (1978) showed that social grouping alone could not account for cerebral effects of EE. Male rats were expos ed to several types of environments, some including social interaction a nd others lacking social intera ction. After living 30 days in their designated environments, animals with so cial interaction were not different in brain region weight, RNA and DNA contents, or ac etylcholinesterase activity from singly housed rats. Social interac tion alone has been shown to not be effective in improving MWM performance in adult mice (Williams et al., 2001). This contradiction between humans and rodents could be due the hierarc hy stress present in mice that is absent in human social activities; alte rnatively other unknown variable s may have contributed to the benefit against dementia seen with social activity in humans. van Praag et al. (2000) stat es that voluntary exercise and EE have Â“notably similar resultsÂ” while social interactions alone cannot elicit the effects obser ved in EE. At this point it appears that no single variable can account for the changes observed in EE mice.
85 Ehninger et al. (2003) showed th at EE increased proliferation of astrocytes in layer 1 of the motor cortex, while voluntary wheel runn ing caused an induction in proliferation of microglia in superficial cortical layers. Th is would suggest that different components of EE could be eliciting different anatomical and electrophysiol ogical changes resulting in a global behavioral benefit. van Praag et al. (1999) showed that both complete EE and voluntary wheel running enhanced the survival of newborn neurons in the dentate gyrus. Taking into account this work, it is not unreasonable to suggest that isolated elements of EE may result in similar anatomical and physiological changes by a common pathway (van Praag et al., 2000). Di rect comparisons of several morphological and biochemical measures, as well as comparisons in a wide spectrum of behavioral tasks are needed to draw definite conclusions regarding the c ontributions that individual components of EE have.
86 Specific Aims Costa et al. (2005) has shown that Â“comple teÂ” environmental enrichment protects against behavioral impairment in PDAPP+ PS1 transgenic mice, however it is unknown which component(s) of complete EE are re sponsible for that protection and what mechanisms are involved therein. Therefore, the specific aims of this thesis are: Specific Aim 1 : To identify the contribution of each component (cognitive, social, physical) of Â“completeÂ” environmental enrich ment that lead to protective effects on cognitive performance in AD transgenic mice. This will be achieved through implementation of our AD Tg+ mice in a cont roled longitudinal investigation of the components of environm ental enrichment. Specific Aim 2 : To elucidate the mechanisms through which environmental enrichment, and its various components, protect against cognitive impairment in AD transgenic mice. Brain -Amyloid levels, plasma cytokines levels, dendritic branching and spine density (all implicated to be involved with the protective ability of environmental enrichment) will be analyzed in order to provide a mechanisms by which enrichment (and its component activities) ma y be providing cogniti ve protection in AD Tg+ mice.
87 Specific Aim 3 : To clarify the stress associated wi th living in a Â“completeÂ” enriched environment as well as living in variable housing environments that contribute to Â“completeÂ” enrichment. With stress levels potentia lly confounding behavioral and pathological effects of enrichment (and its co mponent activities), corticosterone levels will be measured. Specific Aim 4: To determine the inter-relationships between behavioral, neurohistologic, and biochemical meas ures taken from the same animals. This unique, controlled study enables the analysis of interactions between behavioral and pathological findings within the same anim al. Through correlation analysis, identification of these interactions could ai d in the clarification of the mechanisms whereby enrichment benefits cognitive performance in AD transgenic mice.
88 Material and Methods Animals & General Protocol All mice contained a mixed backgr ound of 56.25% C57B6, 12.5% B6D2F1, 18.75 % SJL, and 12.5% SW. Mice were genera ted from a cross between male mice, heterozygous for the mutant APPK670N, M671L gene, and mutant PS1 (6.2 line) females. Mice were initially genotyped at the time of weaning and then had a confirmatory genotyping at 4 months of age. A total of 38 mice were used in three genotypic categories: non-tran sgenics, APP/PS1 double transgenic mice, and APP single transgenic mice. Mice were randomly divi ded into the various housing groups, where they lived through testing. Ad libitum access to rodent chow & water was provided, with mice maintained on a 14/10 light/dark cycle. Figure 1 presents a timeline for this st udyÂ’s procedures. Beginning at 6 weeks of age, transgenic mice and non-transgenic litter mates were moved from standard social cages to one of the following housing envir onments: an impoverished environment (n= 8), a social housed environment (n=9), a phys ical activity environment (n=5), or into a Â“completeÂ” enrichment environment (n=6). At 6 months into th eir respective housing conditions (and at 7.5-8.5 months of age), all mice were test ed in a 5-week behavioral battery, while still living in their designated housing. Following completion of the behavioral battery at 8.5-9.5 mont hs of age, all animals were euthanized and brains of the APP+PS1 mice and non-transgenic controls were removed for histopathology and
89 neurochemistry. All procedures used were reviewed and approved by the USF Institutional Animal Care a nd Use Committee (IACUC). Enriched Environments At 6 weeks of age, transgenic mice were placed into one of five test environments. In a step-wise ascension towards the Â“completeÂ” enrichment paradigm, transgenic mice were put into an impoveri shed environment (n= 8; 6 APP+PS1 and 2 APP), a social housed environment (n=8; 4 APP+PS1 and 4 APP ), a physical activity environment (n=5; 5 APP+PS1), or into a Â“completeÂ” enrichment environment (n=6; 6 APP+PS1). Animals of the same gender, how ever of different ge notypes, were housed together. Both groups of transgenic mice (APP+PS1 and APP) in all conditions performed identically, so their behavioral data were combined for statistical analysis. However, because APP+PS1 and APP mice were significantly different pathologically at euthanasia (e.g., no A plaques burdens in APP mice) the APP mice were eliminated from neuropathological and neurochemical analysis. A group of socially housed non1 4 37 6 5 2 8 9 0 Animals Born Entered Housing Environment Behavioral Testing Animals Euthanatized MonthsFigure 1. General protocol time line for enrichment study.
90 transgenic mice were included as a standard ho used control (n=12). Animals living in an Â“impoverished environmentÂ” were housed individually and had access to only food and water within their st andard mouse cage (6.5Â”wide, 10.5 Â” long, 5.5 Â”high). Socially housed mice lived in standard mouse cages with other mice (2 4 mice/cage) of the same gender, thus constituting a Â“social activit yÂ” group. Another group of socially housed mice had both a larger rat cage and access to running wheels. This Â“physical activityÂ” group had cages that were 7Â” wide, 11Â” long, and 5Â” high, with each cage equipped with 2 running wheels. Â“CompleteÂ” enrichment mice had social, physical and cognitive stimulation. A 110 liter Sterilite container (19Â” wide, 32 Â” long, 13.5 Â”high), with an inner Â“CritterTrailTWOÂ” rodent house, was used as housing for the Â“CompleteÂ” enrichment group. Housing for this group (5-7 mice/ cage) also contained runni ng wheels and toys in the courtyard surrounding the rodent house; these items were changed weekly for novelty. Â“CompleteÂ” enrichment mice were also placed in novel, complex environments for at least 1 hour 3 times a week over the course of the study. Mice lived in their designated environment from 6 weeks of age unt il euthanasia at about 9 months of age. All cages contained an igloo and neslets. Behavioral Assessment While still living in their selected housi ng environment, mice were tested in Ymaze, standard water maze, circular platform platform recognition, and radial arm water maze tasks, in that order. All behavioral testing was conducted during the light cycle. Y-maze In this single day task, each anim al was placed in a walled Y-maze for a single 5 minute trial to test both general ac tivity (entries) and ba sic mnemonic processing
91 (percent alternations). Each of the 3 Y-maze arms was 21 by 4 cms with 40 cm high walls. The total number of arm entries and the sequence of arm choices were recorded. Basic mnemonic function was measured as perc entage of spontaneous alternation (ratio of arm choices differing from the previous two choices divided by the total number of entries). For example, the sequence of arm entries (2,3,1,3,2,1,2,3) has six alternation opportunities (total entries-2) and the percent alternati on would be 67%. Standard Water Maze. A 100-cm circular pool was di vided into 4 equal quadrants by drawing lines on the bottom of the pool. Quadrant 2 (goal quadrant) contained a clear, 9-cm diameter submerged platform 1.5 cms below the water. Surrounding the pool were visual cues, which were placed pr oximal and distal to the pools edge. Visual/spatial cues consisted of large, bri ghtly colored 2D and 3D objects, including a beach ball, poster, and inflatable pool toys. During testing, the pool water was maintained at 23-27 C. For each of the 10 days of acqui sition, mice were given 4 trials. For each of the four successive 60-sec trials per day, mice were started from a different quadrant; the same quadrant start pattern was used acros s all 10 days of acquisition. Latency to find the platform (maximum of 60s) was record ed, and the average late ncy of the 4 trials was calculated for use in statistical eval uation. Once the mouse found the platform, it was allowed to stay for 30s. If after 60s the mouse did not find the platform, it was gently lead to the platform and given a 30s stay. After 10 days of acquisition testing, memory retention was evaluated in a single 60s probe trial the following day. For this trial, the submerged platform was removed a nd animals were released from the quadrant directly opposite the goal quadran t. This single trial was vi deo recorded and the percent time spent in each quadrant, as well as the number of annulus crossings, was analyzed.
92 Circular Platform Task The circular platform maze consisted of a 69-cm circular platform with 16 holes equally placed ar ound the periphery of the walled maze. Surrounding the maze were two different colore d shower curtains that had 2-dimensional visual cues attached to them for use by the mi ce in navigating the maze. One of the holes contained an escape box that remanded in the same location for all 8 days of testing. Animals were encouraged to find the escape by adding aversive stimuli of bright lights and fan wind to the maze. The aversive stimuli included two 150-W flood lamps hung 76cm above the platform and one high-speed fan 15 cm above the platform. During a single 5 minute daily trial, both the number of errors (head pokes into non-escape holes) and latency to find the escape hole (up to 300 sec.) were recorded. Although the escape hole remained the same for any given animal over 8 days of testing, the escape box was relocated after each animalÂ’s tr ial to one of 3 other hole locations to control for olfactory cues. The maze surface was cl eaned with dilute vinegar after each animal for added olfactory cue control. Platform Recognition. The platform recognition ta sk measures the ability to search for and identify/recognize a variably-placed elevated pl atform. It requires animals to switch strategies and ignore the spatial cu es present around a circular pool, which was the same pool used in earlier Morris water maze testing. A visible platform, 9-cm in diameter with an attached 10 x 40 cm ensign, was placed into the same 100-cm pool in which standard water maze was conducted. The visible platform was elevated 0.8 cm above the waterÂ’s surface. For 4 days of te sting, animals were placed in the pool at the same start position for each of 4 trials, with the platform being moved to a different one of the 4 quadrants for each trial. The latenc y to ascend the platform was recorded (60s
93 maximum) and the daily 4 trials were aver aged. A 30s stay was given when the mice found the platform. Mice that did not find the pl atform within the 60s were gently guided to the platform by the experimenter and allowed to stay for 30s. Radial Arm Water Maze Spatial working memory was assessed in a Â“win-stayÂ” version of the radial arm water maze (RAWM) task. In the same 100-cm pool utilized for Morris Water Maze and Platform Recognition tes ting, an aluminum insert was introduced in order to divide the pool into 6 equa lly spaced swim arms (30.5 cm length x 19 cm width) radiating from a central circular swim area (40 cm diameter). The insert extended 5 cm above the surface of the water, allowi ng the mice to easily view surrounding visual cues, which were generously placed outside of the pool. Visual/spatial cues consisted of large brightly colored 2D and 3D objects, including a beach ball, poster, and inflatable pool toys. During testing, the pool water was maintained at 23-27 C. In one of the arms, a transparent 9cm submerged escape platform was placed 1.5cm below the water near the wall end. Each mouse was given five 1 minute tr ials per day for nine days. The last of the four consecutive acquisition trials (T rial 4, T4) and a 30 minute delayed retention trial (Trial 5, T5) are indices of working me mory. On any given day, the escape platform location was placed at the end of one of th e 6 arms, with the platform moved to a different arm in a semi-random fashion for each day of testing. In contrast to the stationary platform of standard water m aze, moving the escape platform forced the animal to learn a new platform location da ily, therefore evaluati ng working memory. On each day, different start arms for each of the five daily trials were selected from the remaining five swim arms in a semi-random sequence that involved all five arms. For any given trial, the mouse was placed into that trialÂ’s start arm, facing the center swim
94 area, and given 60s to find the platform. When the mouse made an incorrect choice, it was gently pulled back to that trialÂ’s start arm and an error was recorded. An error was also recorded if the mouse failed to make a choice in 20s (in which case it was returned to that trails start arm), or if the animal en tered the platform-contai ning arm, but failed to locate the platform. A 30s stay was given on ce the mouse had found the platform. If the mouse did not find the platform within a 60s trial, it was guided by the experimenter to the platform, allowed to stay for 30s, and was assigned a latency of 60s. For animals that did not make at least 3 choices an error valu e of 5.6 was assigned as a penalty. This number was calculated by averaging errors for all animals that did not locate the platform for Block 1 (day1-day3) on Trial 1 (T1). Both errors (incorrect ar m choices) and escape latency were recorded for each daily trial. Tissue and Blood Collection Following the final day of behavioral te sting, a blood sample (.5 mL) was taken from the submandibular vein, plasma was separated, and stored at -80 C for later analysis of corticosterone levels. Two days afte r plasma collection, animals were deeply anesthetized with pentobarbital and a second bloo d sample was collected intra-cardially, and stored at -80 C for later analysis of cytokines. Animals were then perfused with 100ml of 0.9% saline. Po st mortem brains were immediately removed and bisected sagitally. The left hemisphere was placed in 4% paraformaldehyde over night and then transferred to a graded series of sucrose solutions (10%, 20%, and finally 30%) wherein tissues remained until histologic sectioning. The right hemisphere was chilled in cold saline and then dissected into 4 major areas : 1) a 2-3mm thick coronal slice through the
95 posterior cortex/hippocampus, 2) striatum, 3) anterior cortex, and 4) cerebellum. The posterior cortex/hippocampus s lice was drop fixed in 10% ne utral buffered formalin for later Golgi Cox staining. The remaining brai n regions were transf erred into individual 1.5 ml Eppendorf tubes, immediately fro zen on dry ice, and stored at -80 C for later neurochemical analysis. Corticosterone Quantification Corticosterone (CS) levels were meas ured using a radioimmunoassay (RIA). Serum levels of CS were determined usi ng a double-antibody RIA kit purchased from ICN Biomedicals (Costa Mesa, CA). Samples from all mice were assayed in duplicate. Approximately 5ng/ml was the minimum detectable concentration. Extraction of Brain Protein for Sandwich En zyme-Linked Immunosorbent Assay (ELISA) A 5% sucrose homogenate (wet weight of tissue/ volume) from frozen mouse anterior cortex was prepared and extracted as described by Schmidt et al. (2004). This procedure began with the weighing of each ti ssue in an empty microcentrifuge tube. Once weighed, 2mls of tissue homogenizati on buffer (THB) per 100 mg of tissue was added to the sample. Immediately prior to homogenization, a proteas e inhibitor cocktail (20% of wet tissue weight to volume ratio) wa s added to prevent degr adation of proteins. While the sample was on ice, the calculated am ount of THB + inhibitors was added to the sample and fully homogenized. After homogeniza tion, samples were frozen and kept at 80 C until Diethyl Amine (DEA) and Formic Acid (FA) extraction was done.
96 DEA extraction was used to separate soluble A from the brain homogenate. First, 100 L of 5 % homogenate was mixed with 0.4% DEA (diluted in 100mM NaCl) on ice. The mixture was then transferred in to a thick-walled pol ycarbonate tube and centrifuged at 100,000 x g for 1 hour at 4 C. After centrifuging, 170 l of supernatant was removed and added to a second tube containing 17 l of 0.5M Tris Base, pH 6.8 (1 l per 10 l of supernatant). After briefly vortexi ng, samples were frozen on dry ice and stored at -80 C. For the FA extr action of insoluble A 100 l of THB was added to the pellet from the DEA extraction to return the mixture to the original volume. Following the addition of 220 l of 95% FA, pellets were sonicated for 1 minute on ice. The mixture was centrifuged at 100,000 x g for 1 hour at 4 C. After centrifugation, 52.5 l of the intermediate phase of the FA extracted mixt ure was added to 1 ml of FA neut ralization solution. Samples were vortexed, immediatel y frozen on dry ice, and stored at -80 C. A Sandwich ELSIA Sandwich ELSIA kits for A 1-40 and 1-42 analysis were purchased from Signet Laboratories and the instructions provided were followed. Briefly, for the 1-40 kit, a standard curve was generated with the highe st point set at 2000 ng/ml and the lowest point set at 0 ng/ml. Before loading the 96 well plate, tissue samples were diluted in provided wash/sample diluent (FA 1:500, DEA 1: 200). Either diluted tissue or standard curve samples were loaded in duplicate ont o the 96 well plate. The sample and the standard curve wells were allowe d to incubate ov ernight at 4-8 C. After incubation, plates were washed, diluted primary anti body was added, and incubated for 2 hours. The plates were then washed again and the s econdary antibodyHRP complex was added for
97 2 hours. Wells were again washed, or thophenylenediamine dihydrochloride (OPD) substrate solution was added, and incubated in a dark room for 45 minutes. Stop solution was added and the optical density was read at 490nm. For the A 1-42 ELISA, all steps were the same as for A 1-40, except that samples were diluted to 1:100 for DEA extracted samples. To account for background, the average optical densities of the 0 ng/ml wells on the standard curve were subtra cted across all 96 wells After generation of the standard curve, optical densities were corrected to reflect protein to wet brain concentration. Blood Cytokine Levels Relative cytokine levels were de termined through the us e of a custom Mouse Cytokine Antibody Array purchased from RayBio tech Incorporated. Ten separate antimouse-cytokine antibodies were provided on membranes. Cytokines analyzed included; IL1, IL1 IL-2, IL-4, IL-6, IL-10, IL-12, IL-12 (p70), TFN IFN and GMCSF. Membranes were first treated with blocking buffer and then incubated with 1 ml of diluted (1:10) plasma for 1 hour. Sample were decanted and 1x secondary biotinylated antibodies were added and incubated for one hour. A second two hour incubation was conducted with diluted (1:1000) labeledstrepavidin, completing the conjugated secondary antibody complex. The final detect ion step involved incubation for 1 min. with provided detection buffers. Using Fujif ilm AR x-ray film, signals from membranes were detected and developed using ba ck-lit photography. For duplicated samples, analysis of signal mean intensities minus background signal intensity were determined using Kodak 1D Image Analysis Software. In order to account for the large variability in
98 signal intensities among the various cytokines, signals were standardi zed on a zero to one scale based on minimum and ma ximum mean intensity readi ngs for each cytokine. Standardized values were then used to co mpare cytokine levels between animals. Golgi Cox Analysis The 2-3 mm thick hippocampus/posterior cortex slices that had been stored in 10% neutral buffered formalin were stained en bloc using the Rapid Golgi Method (Valverde, 1993). Briefly, blocks of tissue were immers ed in a mixture of osmium tetroxide and potassium dichromate for 5-6 days. After 5-6 days, tissues was then subsequently immersed in a silver nitrate solution for 36-38 hours. Tissue blocks were then dehydrated and embedded in nitrocellulose. Stained blocks were cut on a sliding microtome at 120 m, cleared in alpha-terpineol, rinsed in xylene, and cover-slipped under Permount. For analysis of dendritic bran ching, camera lucida drawings of basilar dendritic arbors from randomly selected neurons were generate d. For both parietal cortex Layer V and hippocampus CA1 region, six neurons were drawn per animal and were then subsequently quantified for the amount and distribution of dendritic arbors. Quantification of dendritic distribution was accomplished by the using Sholl analysis, Method of Concentric Ci rcles (Sholl, 1953). Neuronal dendritic spine densities were also analyzed. Dendritic spines were counted directly on a Zeiss research micr oscope using 100x longworking distance oilimmersion objectives. Both th e terminal tips of layer V pyr amidal cells and hippocampal CA1 pyramidal cells were used for the spin e analysis. Spines were counted on 3-5
99 terminal tip segments per neuron, 30 micr ons in length, from 6 neurons per brain (neurons were different from the 6 used in dendritic branching analysis). Statistical Analysis Behavioral Analysis Behavioral performance wa s statistically evaluated to determine any group difference based on housing c ondition or transgenicity. For both of the single day tasks ( Y-maze and water m aze retention) one-way ANOVAs were used. For multi-day tasks (standard water maze ac quisition, circular platform, platform recognition and RAWM) both one-way ANO VAs and two-way repeated measure ANOVAs were performed. Prior to analysis MWM data was broken down into five-2 day blocks and the RAWM data was divided into three-3 day blocks, to aid in data presentation and analysis. After ANOVA analysis, post hoc pair-by-pair differences between groups (planned comparisons) were re solved using the Fisher LSD test. All group comparisons were considered signifi cant at P<0.05. While few in number, any outliers or non-performers (e.g. repeated circulars, consistent floaters) were eliminated from behavioral statistical analysis. Histological/Biochemical Analysis Pathological data an alysis including both histological and biochemical portions of the experiment were performed using ANOVA. Statistical analysis of the spine counts and dendrictic branching we re carried out using ANOVA with a post-hoc tukey test. In order to test if relationships were present between pathologic, biochemical, and behavioral meas ures, correlation analysis was performed using the Systat analytical software package. Correlation values s hown are the results of a pair, uncorrected correlation anlysis.
100 FA and DFAs: To group behavioral, biochemical, and histologic measures by common factors, Factor Analyses (FA) were performed using Systat software. Regardless of genotype or housing, FA used all collected data to relate measures into individual factors. In an initial FA, all 15 behavioral measures were included. This enabled the determination of how different behavioral tasks related to one another, as well as how performance of one task might pr edict performance in other tasks. Follow-up FAÂ’s included not only the 15 behavioral measures, but also the histologic and biochemical measures as well. These FAÂ’s extracted inter-relations within behavioral, histologic and biochemical data. To determ ine if the 5 experimental groups (NT, IMP, SH, PA, and EE) were distinguishable fr om one another based on the behavioral measures set, DFA was performed using Syst at software. DFA was performed using all 15 measures as well as with only the behavioral measures that loaded in factor 1. Both a direct entry and stepwise-for ward DFA was conducted in each case. The direct entry method uses all measures available, while th e step-wise forward method selects measures based on their variance contribution and adds them in a step-wise fashion to best discriminate between groups. All of the above DFAÂ’s were then repeated with inclusion of histologic and biochemical measures.
101 Results Behavioral Analysis Of the five cognitive-based tasks evaluated, the Y-maze and circular platform tasks revealed no impairment in control Tg +/SH mice when compared to NT mice, and thus no protective effect of environmental enri chment (EE) could be present. In Y-maze testing for spontaneous alternat ions behavior, no group differe nces in percent alternation or arm entries were evident (Fig. 2). As well, no group differences were evident in escape latencies over 8 days of circular plat form testing for spatial reference memory (Fig. 3). Animals in all groups collectively improved their performance during circular platform testing, as shown by overall reduced escape latencies across days [F(7, 238) = 2.50, p<0.02]. In Morris water maze (MWM) acquisition, escape latency data was divided into five 2 day blocks to facilitate statistical an alysis and presentation (Fig. 4). Over all 5 blocks of testing, EE mice exhibited significantly lower escape latencies in comparison to both IMP and PA mice (Fig. 4A), although no effect of transgen icity was evident between NT and control SH mice. Evalua ting acquisition across individual blocks revealed that EE mice had significantly lowe r escape latencies on Block 2 compared to all other groups (Fig. 4B), indicating an ab ility of EE mice to improve their learning of this task faster than other groups. For a ll animals collec tively across the 5 blocks of testing, there was an overall learning e ffect [F(4,32) = 16.80; p< 0.0001]. As well, a
102 blocks by treatment interaction was presen t [F(4,128) = 1.92; P< 0.025], apparently due to the worsened performance of NT mice during the last two blocks of testing (Fig. 4B).
103 Figure 2. Y-maze Entries and Percent Spontanous Alternations. No group differences were noted for either entries or percent alternation in Y-maze performance. Abbreviations: NT = non-transgenic socially housed mice, IMP = transgenic impoverish housed mice, SH= transgenic socially housed mice, PA= transgenic physical activity mice, EE= transgenic Â“completeÂ” environment enrichment. Y-Maze Entries NT IMP SH PA EE Enrties 0 10 20 30 40 50 Y-Maze % Alternations NT IMP SH PA EE % Alternations 0 20 40 60 80
104 Figure 3. Circular Platform overall escape latencies. No group differences were evident over 8 days of testing. Abbreviations: as in Figure 2. Circular Platform Overall Latency NT IMP SH PA EE Latency (sec) 0 50 100 150 200 250
105 Figure 4. Morris Water Maze acquisition overall (A) and across 5 two-day blocks (B). EE mice showed better spat ial learning than several other Tg+ groups overall and were able to improve their performa nce sooner than all other groups. = EE mice significantly lower latencies vs all other groups (p<0.05 or hi gher level of significance). Â‡= EE and NT significantly lowe r latencies from IMP(p<0.05). Â† = EE mice significantly lower latencies than NT (p<0.02). ** = EE significantly lower latencies than IMP and PA (p 0.05). Abbreviations: as in Figure 2. Morris Water Ma ze Acqusition Overall 10 Days NT IMP SH PA EE Latency (sec) 20 30 40 50 60 ** ** Morris Water Maze Acquisition Blocks 012345Latency (sec) 20 30 40 50 60 NT IMP SH PA EE Â† Â‡ A) B)
106 Results from the MWM probe trial are presented in Fig. 5. No housing group spent significantly more time searching in the quadrant formerly containing the submerged platform (Q2) than any other gr oup (Fig 5A). As well, there were no group differences in annulus crossings (Fig 5B). While none of the 5 groups showed an exclusive quadrant preference for the former platform-containing qua drant (Q2), both NT and SH groups exhibited a partia l quadrant preference for Q2 (Fig. 5C). In contrast, PA, IMP and EE mice showed no quadrant pref erence (e.g., only one or no quadrants significantly less in % time vs. Q2). Thus, across 3 indices of reference memory in Morris maze testing, EE mice did not perform sign ificantly better than other Tg+ groups. In platform recognition testing, an overal l groups effects was present across all 4 days of testing [F(4,28) = 5.04; p<0.005]. Post hoc planned comparisons of overall escape latencies revealed that IMP, SH, a nd PA groups were impaired overall vs. NT controls, whereas EE mice performed identical to NT controls (Fig. 6A). Analysis of performance on individual days indicated th at both NT and EE groups quickly reduced their escape latencies across th e 4 days of testing (Fig. 6B ). This rapid reduction in escape latency was not observe for IMP, SH and PA mice, indicating that EE mice were much better at changing from the spatia l (cued) strategy of the MWM to the recognition/identification strategy of platfo rm recognition. By Day 4, however, there were no group differences in escape latency, indicating that all housing groups were able to eventually reduce their escape latencies to levels comparable to NT controls. Indeed, all animals collectively improved their perf ormance across the 4 days of testing, as evidenced by a strong overall effect of training [P(3,84)=25.91; p<0.00001].
107 Figure 5. Morris Water Maze Memory Retentio n, as indexed by time spend in the former platform-containing quadrant (A), annulus crossings (B), and quadrant preference (C). No group differences were observed for percent time spent in the former platform-containing quadrant (Q2) or annulus crossings. For percent time spent in individual quadrants, NT and SH groups showed a partial quadrant preference. = significantly percent time compared to Q2 (p<0.05). Abbreviations: as in Figure 2. NT IMP SH PA EE % Time in Quadrant 0 10 20 30 40 50 Q1 Q2 Q3 Q4 * * * NT IMP SH PA EE % Time in Q2 0 10 20 30 40 50 NT IMP SH PA EE Annulus Crossings 0 1 2 3 4 5 A) B) C) Morris Water Maze Retention Platform Quadrant Preference Annulus Crossings Quadrant Preference
108 Figure 6. Recognition/Identi fication performance in Pl atform Recognition testing overall (A) and for each of th e 4 days of testing (B). Over all 4 days of testing (A), all Tg+ groups except EE mice were significantly poorer in performance compared to NT controls. For individual test days, NT mice and EE mice rapidly reduce their escape latency, while IMP, SH, and PA mice have a slowed reduction in latency. *= significantly higher escape late ncies than NT (p<0.05 or hi gher level of significance). Abbreviations: as in Figure 2. Platform Recognition Over Days Latency (sec) 0 10 20 30 40 50 60 NT IMP SH PA EE 1 2 3 4 12 3 4 12 3 4 12 3 41 2 3 4 Days Platform Recognition Overall Latency NT IMP SH PA EE Latency (sec) 0 10 20 30 40 50 * A) B)
109 These results show that Â“completeÂ” enrich ment is needed to protect Tg+ mice against impaired ability to switch from a sp atial to a recognition/identification strategy and that physical activity or so cial interaction are not enough to provide this protection. Figures 7 and 8 present escape latencies in RAWM testing. Data were evaluated across three 3-day blocks for T1 (randomized in itial trial), T4 (final acquisition trial) and T5 (delayed retention trial); T4 and T5 are indices of working memory. An overall groups effect was present for both T4 [F(4,28)=4.91; p<0.005] and T5 [F(4,28)=4.89; p<0.005]. Post hoc analysis of overall T4 and T5 late ncies revealed that IMP, SH, and PA groups were significantly impaired vs. NT controls, whereas performance of EE mice was not different from NT cont rols and significantly better th an the SH transgenic control group (Fig. 7A). By the final block of tes ting, complete separation of EE and NT mice from all other groups was observed (Fig. 7B). On T5 of this final block, NT control mice had significantly lower escape latencies than IMP, SH, and PA mi ce (P< 0.01). In sharp contrast, EE mice were identical to NT mice and significantly better than all other Tg+ groups (P<0.025). Figure 8 echoes the fi nal block findings shown in Figure 7, highlighting the indistinguishable escape late ncies between NT and EE mice for T4 of the final block, as well as the complete separa tion of NT and EE mice from all other Tg+ groups (IMP, SH, PA) in T5. These RAWM data underscore findings from platform recognition in showing that Â“completeÂ” EE pr otects against cognitive (working memory) impairment while physical activity and social housing alone are not protective.
110 Figure 7. RAWM overall and final block escape latencies for Trials 1, 4 and 5. In overall working memory performance (A), NT mice had significantly lower escape latencies than IMP, PA, and SH on both overa ll T4 and T5. In contrast, EE mice were indistinguishable from the NT group. On the final block of testing (B), NT mice and EE mice had significantly lower escape latencies th an all other groups (IMP, SH, PA) for T5. = significant difference between NT and IM P, SH, and PA (p<0.02 or higher level of significance). Â†= NT significantly difference from PA and SH (p<0.02). ** = Both NT and EE significantly different from IMP, SH and PA (p<0.025 or higher level of significance). Abbrevia tions: as in Figure 2. RAWM Overall T1 T4 T5 Trials Latency (sec) 20 30 40 50 60 NT IMP SH PA EE * Final Block T1 T4 T5 Trials Latency (sec) 0 10 20 30 40 50 60 ** Â† A) B)
111 Figure 8. RAWM escape latencies for final block Trials 4 and 5. During T4, NT mice achieved significantly lower escape latenc ies when compared to SH and PA mice, while EE mice were no different from NT c ontrols. During T5, both NT and EE groups had significantly lower latencies then all other groups (IMP, SH, PA). *= significantly different from NT group at p<0.02. **= significantly different from both NT and EE at p<0.025 or higher level of significance. Abbreviations: as in Figure 2. RAWM Final Block Trial 5 NT IMP SH PA EE Latency (sec) 0 10 20 30 40 50 60 ** ** ** NT IMP SH PA EE Latency (sec) 0 10 20 30 40 50 * B) A) RAWM Final Block Trial 4
112 Neuropathologic Measures On the day following completion of be havioral testing, a blood sample was collected to measure plasma corticosterone levels of mi ce living in different housing environments. Statistical analysis revealed no differences in corticosterone levels among animals in the different housing environmen ts (Fig. 9A) and no overall genotype effect (Fig. 9B). The finding that di fferent housing of the four tran sgenic groups did not effect corticosterone levels suggests that stress did not play a role in the differences in cognitive performance observed among these groups. In contrast, male mice collectively had significantly lower levels of corticosterone levels when compared to females [P(1,37)=8.44; p< 0.01]. Levels were almost 3fold lower in male mice vs. females (Fig. 9C). At euthanasia, 3 days following behavior al testing (and approximately 6.5 months into enrichment), a second blood sample wa s collected from APP+PS1 and NT mice for measurement of plasma cytokine levels. As shown in Fig. 10, no significant differences were observed among the housing groups for a ny of the 10 proa nd anti-inflammatory cytokines measured. Analysis by genotype revealed no di fferences in any of these cytokines between APP+PS1 mice collectively and NT mice (data not shown). For APP+PS1 in each housing environment, A analysis by ELISA was performed on anterior cortex tissues. As shown in Fig.11, no significant group differences were observed in A 1-40 or 1-42 levels for either soluble or insoluble species. Housing and transgene-dependent ch anges in dendritic length/branching of neurons from APP+PS1 and NT mice were anal yzed by Golgi staining. For neurons in
113 Plasma Corticosterone Levels Figure 9. Plasma Corticosterone levels by A) treatment (housing), B) genotype, and C) gender. No treatment or transgenic differenc es were observed for corticosterone levels. In contrast,saa male mice had significa ntly lower levels of corticosterone than female mice. *= significantly lower levels of corticosterone (p<0.01). Abbreviations: as in Fig. 2 Treatment NT IMP SH PA EE ng/ml 0 50 100 150 200 Genotype NT Tg+ ng/ml 0 20 40 60 80 100 120 140 Gender M F ng/ml 0 20 40 60 80 100 120 140 160 A) C) B) *
114 Figure 10. Standardized mean signal in tensities for 10 plasma cytokines in APP+PS1 and NT mice. No group (housing) differences were observed for any cytokine measured. Abbreviations: as in Fig. 2 Plasma Cytokine LevelsIL-1a IL-1b IL-2 IL-4 IL-6 Standardized Mean Intensity 0.0 0.2 0.4 0.6 0.8 NT IMP SH PA EE IL-10 IL-12 TNF-a IFN GM-CSF Standardized Mean Intensity 0.0 0.2 0.4 0.6 0.8 NT IMP SH PA EE
115 Figure 11. Quantification of Soluble A (1-40 and 1-42) and Insoluble A (1-40 and 1-42) in APP+PS1 mice within different housing environments. No group differences between housing groups were noted for either soluble or insoluble 140 or 1-42 species. Abbreviations: as in Fig. 2 IMP SH PA EE mol/g 0 20 40 60 80 100 120 140 160 IMP SH PA EE mol/g 0 20 40 60 80 100 120 140 160 180 IMP SH PA EE pmol/g 0 2000 4000 6000 8000 10000 IMP SH PA EE mol/g 0 20 40 60 80 100 120 140 160 Soluble A 1-40 Soluble A 1-42 Insoluble A 1-40 Insoluble A 1-42
116 both the hippocampal CA 1 region and overly ing parietal cortex, there were no differences in total dendritic length or dendri tic branching between NT and transgenic SH control groups (Fig. 12). Thus, APP+PS1 c ontrol mice had a dendrit ic arbor similar to NT mice. There also was no effect of housing on total dendritic length of neurons in the CA1 region or parietal cortex (Fig. 12, upper), as well as fo r dendritic branching in the CA1 region (Fig. 12, lower). Although there al so was generally no effect of housing on dendritic branching in the parietal cortex, EE mice did have significantly less dendritic branching of neurons in parietal cortex compared to both NT and IMP groups (Fig. 12, lower). Factor Analysis/ Discriminant Function Analysis FA of behavioral measures with a nd without neuropathology/neurochemistry measures was performed to determine the und erlying relationships between behavioral tasks and pathology (Table 1). Including all NT and Tg+ mice, FA involving all 15 behavioral measures resulted in 12 of those measur es loading on four principle factors. Collectively, these four factors accounted for over 65% of the total variance. A measure was considered Â“significantÂ” for loading on a factor if its component loading value exceeded 0.600 for that factor. All measures for RAWM and platform recognition loaded heavily under factor 1, which provided more va riance (32.5%) than any of the other three factors. An activity measure, Y-maze entrie s, loaded under Factor 2 along with circular platform final latency. The remaining measur es (two circular platform measures, Morris water maze retention, and Y-maze alternations ) were distributed be tween the remaining two factors (Table 1).
117 Total Branching Length ( m) Total Branching Points Figure 12. Housing and transgene-dependent changes in dendritic length/branching of neurons from APP+PS1 and NT mice No group differences were observed for either the hippocampal CA 1 region and overlyi ng parietal cortex fo r dendritic length or dendritic branching between NT and transg enic SH control groups. EE mice did have significantly less dendritic branch ing of neurons in parietal cortex compared to both NT and IMP groups. *= significantly less dendritic bran ching vs. NT and IMP (p<0.05). Abbreviations: as in Fig. 2 Parietal Cortex NT IMP SH PA EE 0 200 400 600 800 1000 1200 1400 1600 CA1 NT IMP SH PA EE 0 500 1000 1500 2000 2500 CA1 NT IMP SH PA EE 0 5 10 15 20 25 30 Parietal Cortex NT IMP SH PA EE 0 5 10 15 20 25 30 *
118 SOL-40 soluble A ; INSOL-40 insoluble A ; SOL-42 soluble A ; CORT levels of corticosterone; CA1-DL total dendritic length within CA1 region; CA1-TBP total branching points within the CA1; PC-DL total dendritic length within parietal cortex; PC-TBP total branching points w ithin parietal cortex; YM-Ent Y-maze entries; YM-Alt Y-Maze % alternations; WM-Fin water maze latency on last day; WM-Avg water maze latency over all days; WM-Ret water maze % time spent in Q2 during probe trial; CPE-Fin circular platform errors on last day; CPE-Avg circular platform errors over all days; CPL-Fin circular platform latency on last day; CPL-Fin circular platform latency over all days ; PR-Fin platform recognition latency on final day; PR-Avg platform recognition latency over all days; RM-T4-Fin RAWM latency for trial 4 of final block; RM-T5-Fin RAWM latency for tria l 5 of final block; RM-T4Avg RAWM latency over all blocks for trial 4; RM-T5-Avg RAWM latency over all blocks for trial 5. BehavioralBehavioral and FactorMeasuresPathological Measures (Only Tg+) 1 (32.55)(29.94) RM-T5-FinRM-T4 RM-T5RM-T5 RM-T4-FinRM-T5-Fin RM-T4WM-Ret PR-FinPCTBP PR-AvgRM-T4-Fin CA1TBP CA1DL PR-Avg PCDL 2 (15.47)(16.85) YM-EntYM-Ent CPL-FinCPE-Avg CPL-Avg INSOL-40 3 (10.51)(14.10) No SignificantYM-Alt CORT CPE-Fin 4 (9.85)(13.71) CPE-FinSOL-40 WM-RetSOL-42 CPL-Fin 5 (8.26)(9.51) CPE-FinCORT YM-Alt 6 (7.02) INSOL-42aPercent of total variance explained by a given factor is indicated in bold type within parentheses. Table 1 Factor loadings of behavioral measures, with and without pathologic measuresaLoadings (Both Tg+ and NT)
119 When the 8 pathological measures (4 A deposition measures, 4 dendritic measures) and corticosterone levels were included in FA involving only Tg+ mice, all four measures of dendritic le ngth/dendritic branching loaded in factor 1 (Table 1). Similar to the FA without pathologic measures, RAWM and platform recognition measures also loaded in factor 1. In add ition, SWM retention also loaded on factor 1. The loading of neuro-morphologic and behavi oral measures together indicates an underlying relationship between the two, whic h will be further elucidated in the Correlation Analysis section below. The 4 A deposition measures loaded either alone or in factors somewhat independent of behavioral measures. Pl asma corticosterone levels loaded alone in factor 5, as well as with one circular platform measure and Y-maze alternations in factor 3. In a final FA th at involved behavioral measures and plasma cytokine levels in all APP+PS1 mice, 4 factor s results (data not show n). Interestingly, factor 1 loaded all 10 cytokines and both Mo rris maze acquisition measures, while factor 2 contained all RAWM and PR measures DFA was utilized to determine if beha vioral performance of the five housing groups (NT, IMP, SH, PA and EE) or the three main housing groups (NT, SH, and EE) could be distinguished from one another (Table 2). Two DFA methods were utilized: the Â“direct entryÂ” method (which includes all behavioral measures evaluated) and the Â“stepwise-forwardÂ” method (which selects be havioral measures from the total group of measures based on their contribution to the variance). For both the 5 and the 3 group analyses, direct entry DFAs could not disc riminate between the housing groups based on their behavioral performance. In sharp contrast, a stepwise-f orward DFA could
120 pvalues are from WilksÂ’s N.S, not significant; all ot her abbreviations defined in Table 1. Direct Entry Measures # of GroupsMethodSignigicance All 155N.S.p<0.005 NT vs IMP, SH, PA All 153N.S.p<0.01 NT, EE vs SH Factor 15N.S.p<0.05 EE vs IMP, SH, PA Factor 13N.S.p= 0.07 EE vs SH RM-T5-Fin RM-T5-Fin Table 2. Summary of discriminant functional analyses of behavioral measures. Stepwise-forward Method Measures Reatained RM-T5 WM-Avg WM-Fin WM-Ret WM-Fin RM-T5-Fin (6 cognitive measures) (6 cognitive measures)
121 effectively discriminate NT from IMP, SH, and PA groups (p<0.005), but not the NT and EE groups. Four behavioral measures (incl uding RAWM overall T5 latency and all 3 Morris water maze measures) were retained as providing maximal discrimination. When only 3 groups were included, stepwise -forward DFA distinguished NT and EE from the SH group (p<0.01). Two measures (one from RAWM and one from MWM) provided maximal discrimination. Additional DFAs were performed utilizi ng only behavioral measures that had loaded on factor 1 in FA (see Table 1). For all five groups or the main three groups, the direct entry method was again unsuccessful in discriminating between groups (Table 2). In contrast, stepwise-forward DFAÂ’s nicely discriminated between housing groups. With all five groups included, th e stepwise-forward method separated EE mice from IMP, SH and PA groups (p<0.01), but could not distingui sh EE from NT mice (Table 2). When stepwise-forward DFA was re peated using only the three main groups, EE and SH groups were nearly separated (p=0.07), but not EE and SH groups. Trial 5 latency on the final block of RAWM testing was the sole measur e retained for of these stepwise-forward DFAs. Thus, for all behavioral measures incl uded, or with inclusion of the cognitive measures in factor 1, the four stepwise-for ward DFAÂ’s were able to discriminate the comparably-performing NT and EE groups from the poorer performing IMP, SH, and PA groups.
122 Correlation Analysis Behavior vs. Plasma Cytokine Levels Table 3 shows a correlation matrix between plasma levels of the 10 cytokines m easured and the 15 behavior measures for all Tg+ mice collectively. A total of 18 correlations were pres ent, 14 of which involved the three Morris water maze measures. Three correlations were evident between plasma cytokines levels and Y-maze a lternation percentage. Correlations between behavior and plasma levels of both TNFand IFN were most prevalent. For all 18 correlations, higher plasma cytokine levels were associated with better cognitive performance. A very similar pattern of correlations was pr esent when all animals (NT and Tg+) were included in the correl ation analysis. Behavior vs. Plasma Co rticosterone Levels. With all animals (NT and Tg+) included, no correlations were present between any of the 15 behavioral measures and plasma corticosterone levels. Similarly, for all Tg+ mice or individual Tg+ mouse housing groups, no correlations were evident. Behavior vs. Cortical A Levels. Correlation analyses we re performed between the 15 behavioral measures and the 4 cortical A measures (A 1-40 and A 1-42, both soluble and insoluble forms) for all APP+PS1 mice collectively. Only one correlations was significant: higher Y-maze alternations (bet ter performance) was inversely correlated with soluble A (1-40) levels [r = -517; p<0.05]. Even when correlation analysis was performed for each housing group separately, few and inconsistent correlations were present between behavioral measures and cortical A levels. These correlations involved
123 Table 3. For all animals, a correlation matrix of behavioral measures a nd plasma cytokine levels. Positive correlations ( shaded boxes) were noted between many cytokines and both Y-maze and Morris water maze performance r= Pearson product-moment correlation coefficient. p= pr obability. All behavioral measure abbreviations are defined in T able 1. YM-AltYM-EWM-FinWM-OAWM-RetCPE-FinCPE-AvgCPL-FinCPL-AvgPR-FinPR-OARM-T4-FinRM-T5-FinRM-T4-AvgRM-T5-Avg TNFr0.408-0.143-0.548-0.5640.5200.110-0.3010.2700.219-0.404-0.037-0.231-0.315-0.356-0.191 p0.5830.5830.0150.0120.0470.6520.2100.2630.3680.0870.1160.3420.1890.1340.434 IFNr 0.589-0.288-0.469-0.3940.7040.108-0.2730.3840.379-0.383-0.345-0.226-0.366-0.344-0.247 p 0.0130.2620.0430.0950.0030.6600.2580.1050.1100.1050.1490.3520.1590.1490.307 IL-1 r 0.422-0.256-0.542-0.5210.4600.165-0.1620.2410.253-0.383-0.269-0.274-0.378-0.485-0.272 p 0.0910.3210.0170.0220.0850.5010.5070.3190.2960.1060.2660.2560.1110.0350.261 IL-1 r 0.385-0.008-0.362-0.4530.417-0.078-0.4230.2160.096-0.193-0.002-0.288-0.414-0.397-0.197 p 0.9750.9750.1280.0510.1220.7510.0710.3750.6970.4280.9930.2330.0780.0920.419 IL-2 r -0.027-0.082-0.424-0.6330.0760.199-0.1770.077-0.029-0.368-0.147-0.248-0.352-0.402-0.173 p 0.7550.7550.0700.0040.8820.4140.4690.7530.9050.1220.5480.3050.1390.0880.478 IL-4 r 0.328-0.180-0.404-0.4390.471-0.102-0.2040.2010.321-0.395-0.199-0.241-0.266-0.414-0.289 p0.4890.4890.0860.0600.0760.6780.4020.4090.1800.0940.4130.3190.2710.0780.230 IL-6 r0.325-0.086-0.494-0.5610.4160.044-0.3300.3050.238-0.364-0.261-0.111-0.222-0.333-0.183 p0.7440.7440.0320.0130.1230.8590.1680.2050.3270.1260.2800.6500.3620.1630.454 IL-10 r 0.448-0.106-0.500-0.5430.498-0.007-0.1980.3850.349-0.249-0.174-0.187-0.312-0.327-0.249 p 0.0720.6850.0290.0160.0590.9770.4160.1030.1430.3050.4770.4430.1930.1720.303 IL-12 r 0.503-0.185-0.421-0.4340.6030.103-0.3230.4140.355-0.355-0.272-0.152-0.380-0.388-0.336 p 0.0400.4780.0730.0640.0170.6730.1770.0780.1350.1360.2590.5340.1080.1010.160 GM-CSF r 0.632-0.329-0.273-0.2860.5030.200-0.0860.3660.342-0.520-0.421-0.109-0.294-0.195-0.273 p 0.0070.1970.2570.2360.0560.4110.7260.1230.1510.0220.0730.6570.2220.4230.258
124 soluble and insoluble A levels to the same extent. For all APP+PS1 mice collectively or for individual housing groups, surprisingly few correlations involved RAWM measures. Behavior vs. Dendritric Morphology. For all APP+PS1 and NT mice combined, correlation analysis between behavior a nd dendritic morphology revealed strong and consistent correlations between total dendritic length of neurons in the hippocampal CA1 region and both platform rec ognition and RAWM measures (Table 4). Both platform recognition final day and overall latency, as we ll as RAWM Trial 4 (final block and over all) and Trial 5 (final block) correlated positiv ely with dendritic leng th in the CA1 region. Therefore, lesser dendritic length of CA1 ne urons was associated with better performance in these tasks. Importantl y, 4 of these five correlations were present when only APP+PS1 mice were analyzed, while only one wa s present when NT mice were analyzed separately. Thus, APP+PS1 mice were driv ing these correlations involving dendritic morphology in the hippocampal CA1 region. No correlations were evident between behavior and dendritic morphology in the parietal cortex (Table 4). Brain A Levels vs. Dendritic Morphology For all Tg+ mice collectively, there were no correlations pr esent between brain A levels and dendritic length/branching in either hippocampus or parietal cortex.
125 Table 4. For all animals, a correlation matrix of behavioral measures and de ndritic length/branching. Positive correlations (bold font in shaded boxes) were noted between dendritic length of neurons in the CA1 region and both pla tform recognition and RAWM measures. r= Pearson product-moment corre lation coefficient. p= probability. Abbreviations: CA1-DL, tota l dendritic length within CA1 regi on; CA1-TBP, total branch ing points within the CA1; PC-DL, to tal dendritic length within pariet al cortex; PC-TBP, total branching points with in parietal cortex; All behavioral meas ure abbreviations are defined in Table 1. YM-AltYM-EWM-FinWM-OAWM-RetCPE-FinCPE-AvgCPL-FinCPL-AvgPR-FinPR-OARM-T4-FinRM-T5-FinRM-T4-AvgRM-T5-Avg CA1-DL r 0.059 0.0150.0990.243-0.1570.2170.0080.1220.2460.4480.4750.3950.3610.3620.317 p 0.771 0.9420.6030.1870.4450.240.9650.5130.1830.0120.0070.0280.0460.0450.083 CA1-TBP r -0.102 0.167 -0.096-0.2060.0010.233-0.073-0.012-0.2350.0320.1230.063-0.004-0.016-0.022 p 0.6140.4060.6130.2660.9970.2060.6950.9510.2020.8650.510.7370.9820.9320.908 CP-DL r -0.237 -0.263 0.0590.0620.1110.2560.0170.3610.2460.0640.0750.0110.1990.0590.083 p 0.243 0.1950.7610.7430.5980.1720.9290.0500.1900.7350.6950.9530.2930.7570.663 CP-TBP r -0.230 -0.168 0.0270.1110.1620.294-0.1570.3210.1720.1410.104-0.0110.154-0.092-0.044 p 0.2590.4110.8910.560.4390.1150.4080.0840.3640.4570.5860.9530.4160.6290.816
126 Discussion General Summary The present study utilized an elaborate cognitive-based behavi oral battery and multimetric statistical analysis to inves tigate the protective effects of Â“completeÂ” environment enrichment (EE) versus several of its components (physical activity, social interactions) in AD transgenic mice. Our results show that Â“completeÂ” EE (physical, social, and cognitive activities) entirely pr otected AD transgenic mice from cognitive impairment in tasks representi ng different cognitive domains working memory, reference learning, and search/recognition. In strong contrast, transgenic (Tg+) mice reared in environments that included physical activity and social in teraction, or only the addition of social interaction, were not protect ed from cognitive impa irment in adulthood. Noteworthy is that there were never any di fferences between IMP and SH transgenic groups in any task, indicating that individually-housed mice ar e not at a disadvantage for cognitive performance and that either IMP OR SH housing is a suitable Tg control. Through use of discriminant function analysis to determine if housing groups could be distinguished from one another based on mu ltiple behavioral measures, EE and/or NT mice were consistently discriminated from the poorer performing ot her housing groups. Thus, the importance of Â“completeÂ” EE as the protective paradigm is again underscored. Importantly, EE mice were protected from c ognitive impairment through 8.5 months of age, suggesting its potential to preven t or at least delay onset of AD.
127 The cognitive benefits observed in EE-housed APP+PS1 mutant mice occurred without significant ch anges in cortical A levels, plasma cytokine levels, or plasma corticosterone levels, sugge sting involvement of mechanisms independent of these endpoints. However, the EE-housed mice did ha ve decreased dendritic length of neurons in the parietal cortex (but not hippocampus ), suggesting that some extent of dendritic Â“pruningÂ” may be involved in the cognitive bene fits observed. Despite the lack of robust changes in pathologic measures, correlat ion analysis offered possible underlining mechanisms involved in superior cognitiv e performance. Although the lack of correlations between behavioral measures and cortical A measures in APP+PS1 mice offers additional evidence of an A -independent mechanism involved in EEÂ’s protective effects, it is important to note that only anterior cortex was analysed. It is possible that measurement of A levels in hippocampal tissue woul d have elucidated EE-associated changes and more numerous correlations with cognitive performance. As well, plasma cytokine levels and hippocampal dendritic length constantly correlated with cognitive measures, suggesting their involvement in underlying mechanisms of cognitive performance. In factor analysis, moreover, all four dendritic lengt h/branching measures loaded in factor 1, along side key cognitive measures. This finding emphasizes dendritic morphologyÂ’s underlying involvement in cognitive performance. A large body of literature in epidemiological research suggests enriching life experiences (including educati on, occupation, physical activity and social interactions) may provide protection against dementia later in life (Scarmeas et al., 2001; Friedland et al. 2001; Verghese et al. 2003.) Many such retrospective studies suggest cognitive protection from an enriched life style. In addi tion, cognitive benefit, as well as a plethora
128 of neurochemical/neurohistologic changes have also been noted in wild type rodents exposed to an enriched environment (van Pr aag et al., 2000). Both young adult and aged rodents show cognitive benefit following several months of environment enrichment (Kemperman et al., 1997; 1998; 2002; Teather et al., 2002; Frick and Fernandez, 2003). Morris water maze learning has been the ta sk most used to demonstrate improved cognitive performance in enriched mice a nd rats, although delayed alternation, visual discrimination, and food preference tasks ha ve also been utilized (Winocur and Greenwood, 1999; Teather et al. 2002; Need et al. 2003). As for neurochemical/neurohistologic changes, ne urogenesis within th e hippocampal dentate gyrus is increased by environmental enri chment in both young adult and aged mice (Kempermann et al., 1997; 1998; 2002; Van Pr aag et al., 1999). Additionally, there is evidence that synaptogenesis occurs followi ng enrichment in rodents, as evidence by studies showing 1): greater synaptophysin le vels in hippocampus and neocortex (Frick and Fernandez, 2003), 2) an increased number of Golgi-stained dendritic branches and dendritic spines in neocortex/neostriatum (C omery et al., 1995; Tu rner et al., 2003), and 3) greater length of Golgistained dendrites in hippocampus (Faherty et al., 2003). In addition, brain levels of NGF, BDNF, NT-3, a nd GDNF have been shown to be increased by environmental enrichment in rodents (Y oung et al., 1999; Ickes et al., 2000;Johnson et al., 2003). Moreover, gene microarray an alysis indicates that genes involving synaptogenesis, NMDA receptor function, and neuronal growth are all up-regulated in the brain following environmental enrichme nt (Rampon et al., 2000). It has been proposed that the interaction between cognitive, social, and physical activities in the Â“completeÂ” enriched environment may be essential for cognitive benefit as well as the
129 neurochemical/neurohistologic changes and th at no single one of th ese activities alone leads to these beneficial alte rations (van Praag et al., 2000). However, no direct evidence for this premise has appeared in the scientific literature. In our initial report, we found that compete EE (e.g., physical, social, and cognitive activity combined) improves cognition in aged Alzheimer's transgenic (Tg2576) mice despite stable beta-amyloid de position (Arendash et al. 2004), suggesting a therapeutic value of cognitive stimulation for AD patients that is independent of A deposition. In our more recent submitted wo rk (Costa, 2005), we have shown extensive cognitive protection from Â“completeÂ” EE in another APP mouse model of AD. This improvement was observed without changes in A deposition in non-behaviorally-tested PDAPP+PS1 transgenic mice; however, exte nsive changes in gene expression were evident. Specifically, while PDAPP+PS1 mice exposed to EE alone showed no significant decreases in A mice given both EE and behavioral testing showed a 50% reduction in brain A Microarray analysis using hi ppocampal tissue revealed large EEinduced changes in the expression of genes/ proteins related to memory, neuroprotection, and A sequestration (Costa, 2005). The robust cognitive protection provided by Â“completeÂ” EE calls into question if the components of EE (physical, social, and cognitive activity) contribute equally to the behavioral benefits observed through su ch an Â“enrichedÂ” lifestyle. Recent epidemiologic/retrospective studies have tried to separate these components of enrichment and have reported that enhan ced physical activity (F riedland et al., 2001; Churchill et al., 2002), and social activity (Wang et al., 2002) are associated with decreased incidence of AD. This has led to the conclusion that physi cal or social activity
130 alone can provide substantive protection agai nst AD. However, in the present closely controlled animal model study, no such protectio n of physical activity or social activity alone was observed in APP mutant transgenic mice. Retrospective human data can be challenging to decipher a nd/or misleading, due to the many confounding variables inherently present and the inaccuracies associ ated with recall of life experiences over decades. Indeed, the complete separation of physical or social activity in either a retrospective or longitudinal prospective hum an study would be impossible to accomplish in that varying degrees of a ll three components are always pr esent. For example, those individuals that lead a life style involving high physical act ivity or high social activity may also lead cognitively stimulating lives. In any event, the present work shows the first evidence in an AD transgenic model that Â“completeÂ” EE is needed to provide cognitive protection against AD, while the phys ical and social activity components of EE do not alone lead to protection. Thus, it appear s that cognitiv e activity is central to the protection observed. Our results are in contrast to several anim al studies that inve stigated effects of physical activity and social activity on cogni tive function. In bot h wild type and AD transgenic mice there is liter ature that suggests physical activity alone can result in cognitive benefit and/or neur ochemical/neurohistologic cha nges. Physical activity has been show to enhance genes involved in br ain health (metabolic, anti-aging, immunity genes) and plasticity (neurotrophic factor trafficking and vesicl e recycling) (Cotman &Berchtold, 2002). Increase BDNF levels (Johnson & Mitchell, 2003), and enhance neurogenesis (van Praag et al., 1999; Kita mura et al, 2003; Ehninger & Kempermann, 2003) were also observed in r odents through increased physical activity. Several studies
131 indicate that physical activity alone ca n improve Morris water maze performance (Fordyce and Farrar, 1991; Fordyce et al., 1993). Recent work, directly relevant to the current study involved APP mutant mice and found that physical activity was associated with decreases in brain A .deposition (?) (Lazarov et al., 20 05; Adlard et al., 2005). In both studies, however, the extent of wheel running (physical activity ) associated with reduced A and/or cognitive benefit was at the level of a marathon runner, a level of physical activity rarely attainable in humans. The contribution of social activity to EE benefits has not been investigated in contro lled studies involving w ild-type rodents, much less in transgenic mice. Through itÂ’s step -wise design of housing conditions, the present study is the first to dissect out the contri bution of each of EEÂ’s 3 components to the protection afforded against cognitive impairment in AD transgenic mice. Behavioral Measures Our past work has shown the RAWM task to be an extremely sensitive test of working memory (Arendash et al., 2001a; Austin et al., 2003; Nilsson et al., 2004; Jensen et al., 2005), one of the earlie st and most prominent symptoms of AD. In our laboratory, APP+PS1 mice are impaired in this task by 56 months of age (Jen sen et al., 2005), but not at 3 months of age (unpublished observations). Consistent with that premise, Tg+ SH control mice in the present study showed impaired RAWM working memory at 8.5 months of age, most notable in the last block and in overall performance across all blocks. EE-housed Tg+ mice, by contrast, were able to reduce their latencies to levels equivalent to NT mice for both working memory Trials 4 and 5 on the final block, as well as over all blocks. This protec tion is consistent with our ea rlier work showing protection
132 of RAWM performance in APP transgenic mice through Â“completeÂ” EE housing (Costa et al., 2005). Importantly, th is working memory protection was not observed in IMP, PA or SH groups. In fact, complete separati on between EE and NT groups versus all other groups was observed on Trial 5 of the final block. Our complete EE results are consistent with findings from Jankowsky et al. (2005), who reported superior RAWM performance of Tg+ mice reared in a complete EE vers us SH Tg+ mice. Also noteworthy in the present study is that Tg+ mice living in an impoverished envir onment appear to not be at risk for enhanced cognitive impairment dur ing aging, as shown by their identical RAWM performance to the SH transgenic group. Ou r RAWM results suggest that Â“completeÂ” EE may have significant protective potential ag ainst the working-memory impairment associated with AD. Despite the lack of a transgenic eff ect on Morris maze acquisition when Tg+ SH control mice were compared to NT mice, Â“completeÂ” EE mice showed superior performance when compared to other transg enic housing groups on both individual block data and overall acquisition. The absence of a transgenic effect may be attributed to the relatively poor performance of NT mice, rather than the superior performance of the transgenic SH mice. We have previously reported acquisitional impairment in the Morris maze as early as 4.5-6 months of age (Jensen et al., 2005). In contrast, the transgenic SH mice in the present study were not significantly different from NT mice at 8.5 months of age. During the last several bl ocks of testing, escape latencies of NT mice reported in Jensen et al. (2005) were around 15 seconds ve rsus around 40 seconds for NT mice in the current study. These differences could be the result of different strains between colonies used or different F generati ons between the two colonies. In any event,
133 EE mice were able to improve their learning acqui sition at a faster rate than other groups in the present study. A congruent effect was observed in recent work by Jankowsky et al. (2005), whom showed significantly decreased es cape distances in APP+PS1 mice reared in Â“completeÂ” EE when compared to Tg+ control mice. In Morris maze reference memory retenti on, a transgenic effect was absent in all three measures (platform quadrant preference, annulus crossings, and quadrant preference). As well, no group differences were present for zone 2 pr eference or annulus crossings. NT and SH mice showed partia l quadrant preference for the goal quadrant while the other groups did not. Although there was no effect of transgenicity in Morris maze retention, complete EE did not provide bene fit to transgenic mice relevant to the other transgenic housing groups (IMP, SH, PA ). Similarity, Jankowsky et al. (2005) found that APP+PS1 mice did not benefit fr om EE during memory retention testing compared to transgenic SH controls. By performing the platform recognition task after the Morris maze task, mice must switch from the spatial strategy of the Morris maze to a recognition/identification strategy. (Arendash et al. 2001a; Austin et al., 2003; Jensen et al. 2005 ). Over 4 days of testing, the NT mice and EE mice quickly reduced their escape latencies. In contrast, the IMP, SH, and PA transgenic groups lagged in th eir effort to convert from a spatial to a recognition strategy. Despite a ll groups being identical by day 4, the strategy switching ability of NT and EE Tg+ mice was clearly supe rior to the other groups. In this regard, overall latencies revealed IMP, SH and PA mice to be significantly impaired in comparison to NT mice; however, EE mice were identical to the NT group. It should be noted that EE protected against an Â“age-relat edÂ” impairment in platform recognition since
134 our prior work showed that APP+PS1 transgenic mice are not impaired in this task at 5-6 months of age (Jensen et al., 2005). The above platform recognition results showing complete strategy-switching protection in EE mice is consistent with our earlier EE study, wher ein a different APP transgenic line was used (Cos ta et al. (2005). Results from both studies again underscore the unprotective nature of the individual en richment components. Similar to RAWM findings, deleterious cognitive effects were not observed in transgenic mice living in an impoverished environment (IMP) versus SH mice, again showing that either housing condition is an appropriate c ontrol for interventio n strategies. Wide ly documented in AD patients is impairment in a variety of attentionrelated tasks, resulting in an inability to Â“searchÂ” or to shift attention from one item to another (Tales et al., 2004). Our platform recognition results involving such strategy sw itching could thus be of considerable clinical significance. EE could offer significant prophylactic value for search/identification if introduced early and if all com ponents of enrichment (social, physical and cognitive) are included. SH transgenic control mice were not im paired in either Y-maze spontaneous alternation or circular platform tasks, nega ting any potential for EE or itÂ’s components to provide protection. This work is consistent with our previous findings, which showed limited sensitivity of APP+PS1 transgenic mice to Y-maze basic mnemonic function and circular platform reference memory impair ment (Arendash et al ., 2001a; Jensen et al., 2005). As anticipated, measures from Y-maze and circular platform did not load with measures in the prime cognitive factor in factor analysis. This finding offers further evidence that measures from these two task s do not relate well to, and are measuring
135 something different from, the three wate r-based tasks (Morris maze, platform recognition, and RAWM). Neuopathologic Analysis A Neurochemistry. The robust cognitive benefits observed from environmental enrichment occurred without a significant d ecrease in soluble or insoluble levels of A in anterior cortex; indeed, there were no differences in cortical A deposition among any of the 4 transgenic housing groups. Thus it app ears that cognitive benefits from EE occur in the presence of high levels of brain A It should be noted, ho wever, that technical problems prevented analysis of A levels in the hippocampus, wherein EE-induced changes may have been more likely to occur. Past literature rega rding the effects of protection-based EE on brain A levels are inconsistent. Early work by Jankowsky et al. (2003) showed that EE exacerbated A levels in APP+PS1 mice. This was observed in the hippocampus and/or cortex as meas ured by histological staining, and A ELISA. These results are consistent with recent work from the same laboratory showing improved cognitive performance with EE is associated with increases in hippocampal A levels, as measured by histologic staining and A ELISA (Jankowsky et al., 2005). By contrast, Lazarov et al. (2005) showed significant decreases in cortex and hippocampal A deposition after housing APP+PS1 mice in EE. Additionally, decreases in brain A levels, measured by A ELISA, were observed in EE mice. In our submitted work (Costa et al.; 2005), no change in A deposition resulted from housing PDAPP+PS1 mice in EE; however, when EE wa s combined with extensive behavioral
136 testing (a form of EE itself), decreases in A deposition occurred in both hippocampus and cortex. These differences in findings be tween the three laboratories may be the result of different strain backgrounds and/or differe nt EE protocols. Interestingly, Lazarov et al. (2005) found that enriched mice e xhibiting high wheel running activity had significantly less brain A deposition than those with low activity. Several events, other than wheel running-induced decreases in brain A may have lead to these findings. Firstly, mice were not randomly selected to test for physical activity, leading to the question if animals with less A deposition were simply more active than those with higher A burdens. A second possible explanation is that the wheel to mouse ratio was 1 to 2, leading to establishment of a hierarchy for wheel use. Mice that gained more access to wheels may have been the Â“bullyÂ” animals while those that did not have wheel access were lower in the hierarchy, translating into more stressful living. Despite the incongruent findings in APP transgenic models regarding EEÂ’s effects on brain A the current work and past studies (Costa et al., 2005; Jankowsky, et al., 2005) show cognitive protection despite steady state and even increases in A suggesting possible A independent mechanisms contri buting to the EE effect. Golgi Analysis of Dendritic Morphology. The finding that dendritic branching in hippocampus of wild-type mice was increased by complete EE but not physical activity alone (Faherty et al. 2003) offered a possi ble mechanism by which EE mice in the present study were cognitively protect, while PA mice were not. However, no group differences in dendritic branching or dendri tic length were seen, both in te rms of a transgenic effect or a housing effect. Thus, EE did not affect dendritic morphology in either hippocampus or parietal cortex of APP+PS1 mice. Dissi milarly, Moolman et al. (2004) have shown
137 APP+PS1 mice to have significantly less tota l dendrite area in hippocampus than NT mice. One major difference between Moolman et al. (2004) and the current study is the confounding variable of behavioral testing. Our extensive behavi oral testing over 5 weeks may have standardized all groups a nd therefore disguised any effects on dendritic morphology due to trangenicity or housing. In our prior study (Costa et al., 2005), which used the same behavioral battery, we obser ved a genotype effect in that EE induced a decrease in dendritic branchi ng in the cortex; however, no effect of complete EE was on dendritic branching was evident. It is important to note th at our prior study involved a different AD transgenic model (PDAPP+PS1 mi ce) rather than the APP +PS1 mice used in the current study. In any event, no gr oup differences were obser ved at the dendritic level in the present work. Exanimation of dendritic spines in hippocampus and cortex may reveal group differences among the 5 groups of the present study, and such an examination is in progress. Along this line, recent work by Leggio et al. (2005) in wildtype mice showed an EE-induced increase in dend ritic spines in the parietal cortex, with cognitive benefits in both the Morris wa ter maze and RAWM. Whether or not such dendritic changes can be induced by EE in APP transgenic lines, generating high brain levels of A will be important to determine. Furt her work at the synaptic (spine) level may offer a mechanism by which Â“complet eÂ” environmental is providing cognitive protection to AD transgenic mice. Blood Measurements Cytokine Analysis Across all 10 plasma cytokines measured, there were no effects of housing condition or of APP+PS1 tr ansgenicity. These re sults suggest that
138 neither housing environment nor the APP transg ene induce a global, sustained change in the immune system. Thus, cognitive defic its observed in APP transgenics and the cognitive protection seen in EE mice occurred independently of a change in the systemic immune profile. These findings are consis tent with work by Marashi et al. (2004) showing no significant differences in spleen cytokine levels (IL2, IL-4, IL-10, and IFN) of mice raised in EE. Despite the absen ce of significant changes in cytokine levels, strong correlations were evident between pl asma cytokines and cognitive performance (see next section on correlations). For the pr esent study, it should be noted that cytokine measurements were for plasma only, and may not be indicative of regional brain cytokine profiles. In work submitted for publicati on, we found no differences in brain cytokine levels between NT and APP+PS1 mice of the same age as animals in the present study, suggesting no overt immunological response in the brain to the mutant APP transgene would be expected in this studyÂ’s Tg+ mice. The absence of brain cytokine differences in Tg+ mice suggests that these mice may be Â“imm une tolerantÂ” to the presence of human A This response would be expected for any peptide produced endogenously at high levels throughout life. Future studies must be conducted to analyze brain cytokine levels following EE in Tg+ mice to determine if immunologic changes occurring in the CNS could be contributing to the cogni tive protection provided by EE. Corticosterone Levels A stressful environment has the potential to mask or exacerbate effects independent of the variable s tested. In order to control for this variable, plasma corticorsterone levels were measured for the different housing groups following completion of behavioral testing. No group differences were observed, based on either housing or transgenicity. Thus, it is unlikely that different levels of stress were
139 associated with the four housing environments therefore eliminating the possibility that stress may have contributed to the differ ences in cognitive performance between housing groups. In contrast to our findings, Belz et al. (2003) has shown that wild-type mice housed in EE had increased adrenal weight a nd increased blood stress hormones (such as corticosterone and ACTH), suggesting that EE can be a stressful environment in wildtype mice. Despite no housing differences in corticosterone levels in the present study, gender differences were observed, with male mice having almost 3-fold lower plasma corticosterone levels compared to females. This is consistent with work showing female wild-type mice have significantl y higher basal levels of cor ticosterone than male mice (Finn et al., 2004, Gal ea et al., 1997). Correlations Analyses between Neur opathology, Blood Measures and Behavior The cognitive benefits of 6 months of Â“completeÂ” EE occurred in APP transgenic mice without significant decreases in sol uble or insoluble levels of brain A changes in blood cytokines, or changes in dendritric morphology. However, correlation analysis between 15 behavioral measures and pathologic /blood measures revealed some consistent associations, possibly linking the la ter to the former measures. Surprisingly, correlation analyses pe rformed between cognitive measures and brain levels of both so luble and insoluble A revealed limited and inconsistent correlations. This is in sharp contrast to the strong correlations we have generally found between cognition and various brain A measures in our earlier work (Arendash et al., 2001a; Gordon et al., 2001; Leighty et al., 2004 ). This apparent discrepancy may be explained by the measurement of A levels by ELISA in the current study, while our
140 prior studies involved histologic staining for A burdens. As well, the anterior cortex was analyzed in the present study (due to unavailability of the hippocampus), while hippocampus was most often utilized in our prior studies. Since the hippocampus is critical to working memoryand spatial le arning-based tasks such as the RAWM and Morris maze tasks, respectively, closer associations between hippocampal A levels and cognitive performance would have been an ticipated. Finally, the various housing conditions themselves (particularly EE) may have neutralized the association between brain A and cognition. The lack of association between A and behavior in the present study adds evidence to the findings of Ja nkowsky et al. (2005) wherein EE-induced cognitive protection occurred de spite increases in brain A suggesting that EE mitigates cognitive impairment in the presence of amyloid burden. In contrast to the absenc e of correlations between A and cognition, consistent correlations were observed for both dendritic morphology and cytokines levels versus cognition. Higher plasma cytoki ne levels correlated with better cognitive performance. These correlations were limited to Morris water maze and Y-maze alternation measures. Interestingly these tasks were not among those which highlighted the be neficial effects of Â“completeÂ” EE and the non-protective nature of so cial and physical inte raction. In a prior study involving APP+PS1 mice given adoptivel y-transferred T cells we also observed the same positive association be tween higher levels of plas ma cytokines and cognitive performance (unpublished observations). T ogether, results from these two studies suggest that higher levels of individual cytokines may pl ay a role, or are at least indicative of, improved cognitive performance. Regarding correlations involving dendritic morphology and cognitive performa nce, dendritic length in the hippocampus
141 CA1 region (but not parietal cortex) consis tently correlated with RAWM and platform recognition measures. Decreased dendritic le ngth was associationed with better cognitive performance suggesting a Â“pruning backÂ” of extraneous branching as a possible mechanism for enhanced cognitive performan ce. Also noteworthy is that the two strongly correlated tasks (RAW M and platform recognition) we re also the tasks in which Â“completeÂ” EE was most protective. Despite th e lack of any significant group differences in any of the dendritic morphology measures, the presence of strong correlations between CA1 dendritic length and key cognitive tasks suggests hippocampal dendritic morphology as an underlying determinant of cogn itive performance. It will be important to follow these findings with analysis of dendri tic spines from the same animals. Such a study is now ongoing. Multimetric Analyses (FA and DFA) A benefit to utilizing an extensive be havioral battery which tests multiple cognitive domains is the ability to employ higher level statisti cal analysis to characterize and distinguish between genotype or treatment groups. Using this novel approach in the present study, we were able to determine underlying relationships between behavioral and pathologic measures (via factor analysis ; FA), as well as to compare experimental groups across multiple behavioral tasks/measures (via discriminant function analysis; DFA). In past studies we have utilized thes e advanced statistics to discriminate between the behavioral performance of: 1) mutant APP tr ansgenic lines that vary in their extent of A deposition (Leighty et al., 2004; Nilsson et al., 2004), 2) mutant APPsw vs. mutant Tau trangenic lines (Arendash et al. 2004), 3) A vaccinated mice vs. controls (Jensen et
142 al. 2005), and 4) environmentally enriched vs. non-enriched APPsw transgenic mice (Arendash et al., 2004). In the present study, we used these sensitive multi-metric statistics to try to distinguish between Â“completeÂ” enrichment and several of its components. Factor analysis performed on both beha vioral measures alone revealed one primary factor (factor 1) which was compri sed of all measures from the RAWM and platform recognition tasks. Other behavioral measures loaded under three lower factors. Noteworthy is that factor 1 was comprised of measures from the two tasks wherein EEÂ’s effects on cognitive performance were most ev ident. This inclusion of all RAWM and platform recognition measures into the primary factor is very consis tent with our prior studies, in which the same loading occurred (L eighty et al., 2004; Jensen et al., 2005). When pathologic measures were included with behavioral measures in FA, all 4 dendritic length/branching measures loaded in fact or 1 along with all RAWM and platform recognition measures. This suggests an intimate relationship between dendritic morphology and cognitive performance. These findings underscore the need for further analysis of dendritic structur e in the same brains, specifica lly dendritic spine analysis (which is in progress). In contrast to the loadin g of dendritic morphology measures in factor 1 along side key cognitive measures, brain A levels and blood levels of both corticosterone and cytokines loaded in other factors. It is not surprising that A levels did not load in factor 1 since there were few correlations between cognitive performance and brain A levels. In our prior studies wherein strong correlati ons were present between cognition and brain A deposition, A deposition measures loaded in f actor 1 along with key cognitive
143 measures (Leighty et al., 2004; Jensen et al ., 2005). Interestingly, in the FA involving behavioral measures and plasma cytokine le vels, factor 1 contai ned all 10 cytokines along side both Morris maze acquisition measur es. In that acquisitional performance in Morris maze was highly correlated with plasma cytokine levels, there clearly is some underlying relationship between performance in this task and blood cytokine levels. DFA was used to determine if behavior al performance alone could distinguish between the 5 experimental groups of this st udy. In addition to the DFA on all 5 groups, further DFA analysis involving 3 groups (NT, EE, SH) was performed to clarify any Â“completeÂ” EE effects that may have been confounded by interactions between all 5 experimental groups. Although the Â“direct entr yÂ” DFA method of using all behavioral measures was largely unsuccessful in di scriminating between groups, the Â“stepwiseforwardÂ” DFA method was consistently able to discriminant between groups in either the 5or the 3-group analysis, irrespectively of whether all behavioral measures were included or whether only fact 1 behavioral measures we re included. NT and/or EE groups were always discriminated from all other groups (IMP, SH, PA). For example, utilizing factor 1 measures with all 5 groups included, EE mice were clearly discriminated from IMP, SH, and PA transgen ic groups. This was consistent with our analysis of individual measur es wherein the EE mice were identical to NT mice in having excellent cognitive performance, while the IM P, SH, and PA groups were impaired in RAWM and platform recognition. The one cognitive measure consistently retained for providing this discrimination was RAWM perf ormance in Trial 5 during the final block of testing. Thus, this measure alone pr ovided maximal discrimination between the groups of this study, although multiple measur es and several cognitive domains were
144 involved in the discrimination when the DF A was initiated with all 15 behavioral measures. Importantly, our stepwise-for ward DFA results extend and complement traditional single-measure assessments by indicating that the beneficial effects of Â“completeÂ” EE occurred across multiple cognitive measures from multiple tasks evaluated as a group. Much like the indexing of multiple behavioral measures/domains of elderly humans or AD patients (Altepeter et al., 1990, Whelihan et al., 1997), DFA of AD transgenic models provides a valuable tool in assessing cognitive performance across various cognitive domains. In summary, this study clearly show s the effectiveness of Â“completeÂ” environmental enrichment in protecting cogniti on AD transgenic mice. In contrast, the physical and social components of enrich ment alone did not provide protection, suggesting it is the cognitive or the additive effect of all components that leads to the protective effects of EE. Despite robust behavioral changes, limited changes were observed in the pathologic variab les measured. Taken together, A quantification, FA and correlations strongly suggest an A independent mechanism in the current study. However, morphological changes do correlate with behavior and offer a potential underlying mechanism. Although the exact mechanisms whereby Â“complete enrichmentÂ” offers protection while enrich ment components do not are unclear, other possible mechanism are under consideration. Firstly, changes in neuronal morphology at the dendritic spine level, re sulting in enhanced cognition, have been observed in EEraised mice (Leggio et al., 2005). This increase in synaptic material, potentially leading to enhanced synaptic transmission, may offer a mechanism by which Â“completeÂ” EE mice are protected from cognitive impairment. The finding in the present study that decreased
145 dendritic length correlated with cognitiv e performance underscore s the potential of a morphological dendritic pruning-based mechan ism of action. A second prospective mechanism is that Â“completeÂ” EE is rendering A irrelevant as an impairing factor through increasing neuronal health by enhanc ing neurotrophic fact ors and/or enhancing neuro-protective genes (Lazarov et al. 2005; Co sta et al., 2005). This mechanism nicely fits with the findings of JankowskyÂ’s group, who showed that despite increases in A deposition, cognitive protection was present. Ov er all increases in neuronal health which renders neurons less susceptible to the toxic effects of A as well as enhanced synaptic matter leading to a reserve of synaptic transm ission may be working in concert to provide the protective effects of EE. In any even t the results of the cu rrent study suggest that environment may be a powerful, modifiable va riable in the development of AD later in life.
146 Reference Abbott RD, White LR, Ross GW, Masaki KH, Curb JD, Petrovitch H. Walking and dementia in physically capable elde rly men. JAMA. 2004 Sep 22;292(12):1447-53. Abraham CR, Selkoe DJ, Potter H. Immunochemi cal identification of the serine protease inhibitor alpha 1-antichymotrypsin in the brai n amyloid deposits of Alzheimer's disease. Cell. 1988 Feb 26;52(4):487-501. Abramov AY, Canevari L, Duchen MR. Ca lcium signals induced by amyloid beta peptide and their consequences in neurons and astrocytes in cu lture. Biochim Biophys Acta. 2004 Dec 6;1742(1-3):81-7. Adlard PA, Perreau VM, Pop V, Cotman CW. Voluntary exercise decreases amyloid load in a transgenic model of Alzheimer's di sease. J Neurosci. 2005 Apr 27;25(17):4217-21. Altepeter TS, Adams RL, Buchanan WL, Bu ck P. Luria Memory Words Test and Wechsler Memory Scale: comparison of utility in discriminating neurologically impaired from controls. J Clin Psychol. 1990 Mar;46(2):190-3. Arendash GW, King DL, Gordon MN, Morgan D, Hatcher JM, Hope CE, Diamond DM. Progressive, age-related behavi oral impairments in transgenic mice carrying both mutant amyloid precursor protein and presenilin-1 transgenes. Brain Res. 2001a Feb 9;891(12):42-53 Arendash GW, Gordon MN, Diamond DM, Aus tin LA, Hatcher JM, Jantzen P, DiCarlo G, Wilcock D, Morgan D. Behavioral a ssessment of Alzheimer's transgenic mice following long-term Abeta vaccination: task specificity and correlations between Abeta deposition and spatial memory. DNA Cell Biol. 2001b Nov ;20(11):737-44. Arendash GW, Lewis J, Leighty RE, McGowa n E, Cracchiolo JR, Hutton M, Garcia MF. Multi-metric behavioral comparison of A PPsw and P301L models for Alzheimer's Disease: linkage of poorer cognitive performance to tau pathology in forebrain. Brain Res. 2004 Jun 25;1012(1-2):29-41. Arendash GW, Garcia MF, Costa DA, Cracchiolo JR, Wefes IM, Potter H. Environmental enrichment improves cogniti on in aged Alzheimer's transgenic mice despite stable beta-amyloid deposit ion. Neuroreport. 2004 Aug 6;15(11):1751-4.
147 Arnaiz SL, D'Amico G, Paglia N, Arismendi M, Basso N, del Rosario Lores Arnaiz M. Enriched environment, nitric oxide production and synaptic plastic ity prevent the agingdependent impairment of spatial cogniti on. Mol Aspects Med. 2004 Feb-Apr;25(1-2):91101. Aisen PS, Schafer KA, Grundman M, Pfeiffer E, Sano M, Davis KL, Farlow MR, Jin S, Thomas RG, Thal LJ; Alzheimer's Disease C ooperative Study. Effects of rofecoxib or naproxen vs placebo on Alzheimer disease progr ession: a randomized controlled trial. JAMA. 2003 Jun 4;289(21):2819-26. Belz EE, Kennell JS, Czambel RK, Rubin RT Rhodes ME. Environmental enrichment lowers stress-responsive hormones in singl y housed male and female rats. Pharmacol Biochem Behav. 2003 Dec;76(3-4):481-6. Berman K, Brodaty H. Tocopherol (vitamin E) in Alzheimer's disease and other neurodegenerative disorders. CNS Drugs. 2004 ;18(12):807-25. Beatty WW, Salmon DP, Butters N, Heinde l WC, Granholm EL. Retrograde amnesia in patients with Alzheimer's disease or Hun tington's disease. Neurobiol Aging. 1988 MarApr;9(2):181-6. Blennow K, Hampel H. CSF markers for inci pient Alzheimer's disease. Lancet Neurol. 2003 Oct;2(10):605-13. Braak, H Braak, E. Neuropathological staging of Alzheimer-related changes. Acta Neuropathol. 82:239-259;1991a Braak H, Braak E. Demonstration of amyloi d deposits and neurofibrillary changes in whole brain sections. Brain Pathol. 1991b Apr;1(3):213-6. Braak H, Braak E, Grundke-Iqbal I, Iqbal K. O ccurrence of neuropil th reads in the senile human brain and in Alzheimer's disease: a third location of paired helical filaments outside of neurofibrillary tangles and ne uritic plaques. Neurosci Lett. 1986 Apr 24;65(3):351-5. Breitner JC, Wyse BW, Anthony JC, Wels h-Bohmer KA, Steffens DC, Norton MC, Tschanz JT, Plassman BL, Meyer MR, Skoog I, Khachaturian A. APOE-epsilon4 count predicts age when prevalence of AD increase s, then declines: th e Cache County Study. Neurology. 1999 Jul 22;53(2):321-31. Brown J, Cooper-Kuhn CM, Kempermann G, Va n Praag H, Winkler J, Gage FH, Kuhn HG. Enriched environment and physical activity stimulate hippocampal but not olfactory bulb neurogenesis. Eur J Neur osci. 2003 May;17(10):2042-6.
148 Bookheimer SY, Strojwas MH, Cohen MS, Sa unders AM, Pericak-Vance MA, Mazziotta JC, Small GW. Patterns of brain activation in people at risk for Alzheimer's disease. N Engl J Med. 2000 Aug 17;343(7):450-6. Brookmeyer R, Gray S, Kawas C. Projections of Alzheimer's disease in the United States and the public health impact of delayi ng disease onset. Am J Public Health. 1998 Sep;88(9):1337-42. Borchelt DR, Thinakaran G, Eckman CB, Lee MK, Davenport F, Ratovitsky T, Prada CM, Kim G, Seekins S, Yager D, Slunt HH, Wang R, Seeger M, Levey AI, Gandy SE, Copeland NG, Jenkins NA, Price DL, Younki n SG, Sisodia SS. Familial Alzheimer's disease-linked presenilin 1 variants elevate Abeta1-42/140 ratio in vitro and in vivo. Neuron. 1996 Nov ;17(5):1005-13. Borchelt DR, Ratovitski T, van Lare J, L ee MK, Gonzales V, Jenkins NA, Copeland NG, Price DL, Sisodia SS. Accelerated amyloid de position in the brains of transgenic mice coexpressing mutant presenilin 1 and amyloid precursor proteins. Neuron. 1997 Oct ;19(4):939-45. Busciglio J, Gabuzda DH, Matsudaira P, Yankne r BA. Generation of beta-amyloid in the secretory pathway in neuronal and nonneur onal cells. Proc Natl Acad Sci U S A. 1993 Mar 1;90(5):2092-6. Costa D, Cracchiolo J, Bachstetter A, Hughes T, Bales K, Paul S, Mervis R Arendash G, Potter H. Submitted. 2005. Chan SL, Furukawa K, Mattson MP. Presen ilins and APP in neuritic and synaptic plasticity: implications for the pathogenesis of Alzheimer's disease. Neuromolecular Med. 2002;2(2):167-96. Chapman PF, White GL, Jones MW, Cooper-Bla cketer D, Marshall VJ, Irizarry M, Younkin L, Good MA, Bliss TV, Hyman BT, Y ounkin SG, Hsiao KK. Impaired synaptic plasticity and learning in aged amyloid precu rsor protein transgenic mice. Nat Neurosci. 1999 Mar ;2(3):271-6. Chen QS, Kagan BL, Hirakura Y, Xie CW. Impairment of hippocampal long-term potentiation by Alzheimer amyloid beta-pep tides. J Neurosci Re s. 2000 Apr 1;60(1):6572. Chen G, Chen KS, Knox J, Inglis J, Bern ard A, Martin SJ, Justice A, McConlogue L, Games D, Freedman SB, Morris RG. A learning deficit related to age and beta-amyloid plaques in a mouse model of Alzhei mer's disease. Nature. 2000 Dec 21-28 ;408(6815):975-9.
149 Chobor KL, Brown JW. Semantic deterioration in Alzheimer's: the patterns to expect. Geriatrics. 1990 Oct;45(10):68-70, 75. Cobb JL, Wolf PA, Au R, White R, D'Agos tino RB. The effect of education on the incidence of dementia and Alzheimer's disease in the Framingham Study. Neurology. 1995 Sep;45(9):1707-12. Comery TA, Shah R, Greenough WT. Differe ntial rearing alters spine density on medium-sized spiny neurons in the rat cor pus striatum: evidence for association of morphological plasticity with early respons e gene expression. Ne urobiol Learn Mem. 1995 May;63(3):217-9. Corder, E.H., Saunders, A.M., Strittmatte r, W.J., Schmechel, D.E., Gaskell, Condefer KA, Haworth J, Wilcock GK. Clinical utility of computed tomography in the assessment of dementia: a memory clin ic study. Int J Geriatr Psychiatry. 2004 May;19(5):414-21. Coleman PD, Yao PJ. Synaptic slaughter in Alzheimer's disease. Neurobiol Aging. 2003 Dec;24(8):1023-7. Cotman CW, Berchtold NC. Exercise: a behavi oral intervention to enhance brain health and plasticity. Trends Neur osci. 2002 Jun;25(6):295-301. Cruz JC, Tsai LH. Cdk5 deregulation in the pa thogenesis of Alzheimer's disease. Trends Mol Med. 2004 Sep;10(9):452-8. Dahlqvist P, Ronnback A, Bergstrom SA, Soderstrom I, Olsson T. Environmental enrichment reverses learning impairment in the Morris water maze after focal cerebral ischemia in rats. Eur J Ne urosci. 2004 Apr;19(8):2288-98. Davis RN, Massman PJ, Doody RS. Cognitive intervention in Alzheimer disease: a randomized placebo-controlled study. Al zheimer Dis Assoc Disord. 2001 JanMar;15(1):1-9. Degroot A, Wolff MC, Nomikos GG. Ac ute exposure to a novel object during consolidation enhances cognition. Ne uroreport. 2005 Jan 19;16(1):63-7. DeKosky ST, Scheff SW. Synapse loss in frontal cortex biopsies in AlzheimerÂ’s disease: correlation with cognitive se verity. Ann Neurol 1990;27:457Â–64. Delagarza VW. Pharmacologic treatment of Alzheimer's di sease: an update. Am Fam Physician. 2003 Oct 1;68(7):1365-72. De Santi S, de Leon MJ, Rusinek H, Convit A, Tarshish CY, Roche A, Tsui WH, Kandil E, Boppana M, Daisley K, Wang GJ, Schl yer D, Fowler J. Hippocampal formation
150 glucose metabolism and volume losses in MCI and AD. Neurobiol Aging. 2001 Jul-Aug ;22(4):529-39. Devanand DP, Jacobs DM, Tang MX, Del Casti llo-Castaneda C, Sano M, Marder K, Bell K, Bylsma FW, Brandt J, Albert M, Stern Y. The course of psychopa thologic features in mild to moderate Alzheimer disease. Arch Gen Psychiatry. 1997 Mar;54(3):257-63. Dickey CA, Loring JF, Montgomery J, Gordon MN, Eastman PS, Morgan D. Selectively reduced expression of synaptic plasticity-related genes in amyloid precursor protein + presenilin-1 transgenic mice. J Neurosci. 2003 Jun 15;23(12):5219-26. Dodart JC, Meziane H, Mathis C, Bale s KR, Paul SM, Ungerer A. Behavioral disturbances in transgenic mice overexpr essing the V717F beta-amyloid precursor protein. Behav Neurosci 1999 Oct ;113(5):982-90. Dodart JC, Mathis C, Saura J, Bales KR Paul SM, Ungerer A. Neuroanatomical abnormalities in behaviorally characterized A PP(V717F) transgenic mice. Neurobiol Dis. 2000a Apr ;7(2):71-85. Dodart JC, Mathis C, Bales KR, Paul SM, U ngerer A. Behavioral deficits in APP(V717F) transgenic mice deficient for the apoli poprotein E gene. Neuroreport. 2000b Feb 28;11(3):603-7. Double KL, Halliday GM, Kril JJ, Harasty JA Cullen K, Brooks WS, Creasey H, Broe GA. Neurobiol Aging. Topography of brain atro phy during normal aging and Alzheimer's disease. 1996 Jul-Aug;17(4):513-21. Duff K, Eckman C, Zehr C, Yu X, Prada CM Perez-Tur J, Hutton M, Buee L, Harigaya Y, Yager D, Morgan D, Gordon MN, Holcomb L, Refolo L, Zenk B, Hardy J, Younkin S. Increased amyloid-beta42(43) in brains of mice expressing mu tant presenilin 1. Nature. 1996 Oct 24;383(6602):710-3. Dumont M, Strazielle C, Stau fenbiel M, Lalonde R, Dumont M, Strazie lle C, Staufenbiel M, Lalonde R. Spatial learni ng and exploration of environm ental stimuli in 24-month-old female APP23 transgenic mice with th e Swedish mutation. Brain Res. 2004 Oct 22;1024(1-2):113-21. Ehninger D, Kempermann G. Regional eff ects of wheel running and environmental enrichment on cell genesis and microglia pro liferation in the adu lt murine neocortex. Cereb Cortex. 2003 Aug;13(8):845-51. Emmerling MR, Morganti-Kossmann MC, Kossm ann T, Stahel PF, Watson MD, Evans LM, Mehta PD, Spiegel K, Kuo YM, Roher AE, Raby CA. Traumatic brain injury elevates the AlzheimerÂ’s amyloi d peptide A beta 42 in human CSF. A possible role for nerv e cell injury. Ann NY Acad Sci 2000; 903:118Â–122.
151 Engelhart MJ, Geerlings MI, Ruitenberg A, van Swieten JC, Hofman A, Witteman JC, Breteler MM.Dietary intake of antioxidants and risk of Alzheimer disease. JAMA 2002; 287: 3223Â–29. Evans DA, Funkenstein HH, Albert MS, Sc herr PA, Cook NR, Chown MJ, Hebert LE, Hennekens CH, Taylor JO. Prevalence of Al zheimer's disease in a community population of older persons. Higher than previously reported. JAMA. 1989 Nov 10;262(18):2551-6. Evans DA, Hebert LE, Beckett LA, Scherr PA, Albert MS, Chown MJ, Pilgrim DM, Taylor JO. Education and other measures of socioeconomic status and risk of incident Alzheimer disease in a defined population of older persons. Arch Neurol 1997;54:1399Â– 405. Faherty CJ, Kerley D, Smeyne RJ. A Golg i-Cox morphological analysis of neuronal changes induced by environmental enrichment. Brain Res Dev Brain Res. 2003 Mar 14;141(1-2):55-61. Fitzjohn SM, Morton RA, Kuenzi F, Rosahl TW, Shearman M, Lewis H, Smith D, Reynolds DS, Davies CH, Collingridge GL, Seabrook GR. Age-related impairment of synaptic transmission but normal long-term potentiation in transgenic mice that overexpress the human APP695SWE mutant form of amyloid precursor protein. J Neurosci. 2001 Jul 1;21(13):4691-8. Folstein, M., Folstein, S., & McHugh, P. R. (1975). Mini-Mental State: a practical method for grading the cognitive state of patients for the clin ician. Journal of Psychiatric Research, 12, 189-198. Fordyce DE, Farrar RP. Physical activity e ffects on hippocampal and parietal cortical cholinergic function and spatial learning in F344 rats. Behav Brain Res. 1991 May 15;43(2):115-23. Fordyce DE, Wehner JM. Physical activity e nhances spatial learni ng performance with an associated alteration in hippocampal protei n kinase C activity in C57BL/6 and DBA/2 mice. Brain Res. 1993 Aug 13;619(1-2):111-9. Forette F, Seux M-L, Staessen JA, Th ijs L, Birkenhger WH, Babarskiene M-R, et al. Prevention of dementia in randomised double-blind placebocontrolled systolic hypertension in Europe (Syst-Eur) trial. Lancet 1998;352:1347-51. Frstl H, Kurz A. Clinical features of Al zheimerÂ’s disease. Eur Arch Psychiatry Clin Neurosci (1999) 249 : 288Â–290.
152 Fox NC, Schott JM. Imaging cerebral atrophy: normal ageing to Alzheimer's disease. Lancet. 2004 Jan 31;363(9406):392-4. Frick KM, Fernandez SM. Enrichment enhances spatial memory and increases synaptophysin levels in aged female mi ce. Neurobiol Aging. 2003 Jul-Aug;24(4):615-26. Frick KM, Stearns NA, Pan JY, Berger-Sweeney J. Effects of envi ronmental enrichment on spatial memory and neurochemistry in middle-aged mice. Learn Mem. 2003 MayJun;10(3):187-98. Friedland R. Epidemiology, education, a nd the ecology of Alzheimer's disease. Neurology. 1993;43:262-249. Friedland RP, Fritsch T, Smyth KA, Koss E, Lerner AJ, Chen CH, Petot GJ, Debanne SM. Patients with Alzheimer's disease have re duced activities in midlife compared with healthy control-group members. Proc Natl Acad Sci U S A. 2001 Mar 13;98(6):3440-5. Epub 2001 Mar 6. Fukutani Y, Kobayashi K, Nakamura I, Wa tanabe K, Isaki K, Cairns NJ. Neurons, intracellular and extracel lular neurofibrillary tangles in subdivisions of the hippocampal cortex in normal ageing and Alzheimer's di sease. Neurosci Lett. 1995 Nov 10;200(1):5760. Games D, Adams D, Alessandrini R, Barbour R, Berthelette P, Blackwell C, Carr T, Clemens J, Donaldson T, Gillespie F, Guido T, Hagopian S, Johnson-Wood K, Khan K, Lee M, Leibowitz P, Lieberburg I, Little S, Masliah E, McConlogue L, Montoya-Zavala M, Mucke L, Paganini L, Penniman E, Power M, Schenk D, Seubert P, Snyder B, Soriano F, Tan H, Vitale J, Wadsworth S, Wolozin B, Zhao J. Alzheimer-type neuropathology in transgenic mice overexpressing V717F beta-amyloid precursor protein.Nature. 1995 Feb 9;373(6514):523-7. Gasparini L, Ongini E, Wenk G. Non-steroi dal anti-inflammatory drugs (NSAIDs) in Alzheimer's disease: old and new mech anisms of action. J Neurochem. 2004 Nov;91(3):521-36. Gearing M, Mori H, Mirra SS. Abeta-peptid e length and apolipoprotein E genotype in Alzheimer's disease. Ann Neurol. 1996 Mar;39(3):395-9. Geddes JF, Vowles GH, Nicoll JA, Revesz T. Neuronal cytoskeletal changes are an early consequence of repetitive head injury. Acta Neuropathol (Berl) 1999; 98:171Â–178. Geerts H. NC-531 (Neurochem). Curr Op in Investig Drugs. 2004 Jan;5(1):95-100.
153 Glenner GG, Wong CW. Alzheimer's disease a nd Down's syndrome: sharing of a unique cerebrovascular amyloid fibril prot ein. Biochem Biophys Res Commun. 1984 Aug 16;122(3):1131-5. Goekoop R, Rombouts SA, Jonker C, Hibbel A, Knol DL, Truyen L, Barkhof F, Scheltens P. Challenging the cholinergic system in mild cognitive impairment: a pharmacological fMRI study. Ne uroimage. 2004 Dec;23(4):1450-9. Gmez-Isla T, Hollister R, West H, Mui S, Growdon JH, Petersen RC, Parisi JE, Hyman BT. Neuronal loss correlates with but exceed s neurofibrillary tangles in Alzheimer's disease. Ann Neurol. 1997 Jan ;41(1):17-24. Gordon MN, Holcomb LA, Jantzen PT, DiCarl o G, Wilcock D, Boyett KW, Connor K, Melachrino J, O'Callaghan JP, Morgan D. Time course of the development of Alzheimerlike pathology in the doubly transgenic PS1+APP mouse. Exp Neurol. 2002 Feb ;173(2):183-95. Gotz J, Chen F, van Dorpe J, Nitsch RM. Fo rmation of neurofibrillary tangles in P301l tau transgenic mice induced by Abeta 42 fi brils. Science. 2001 Aug 24;293(5534):14915. Graves AB, Larson EB, Edland SD, Bowen JD, McCormick WC, McCurry SM, Rice MM, Wenzlow A, Uomoto JM. Prevalence of de mentia and its subtypes in the Japanese American population of King County, Washi ngton state. The Kame Project. Am J Epidemiol. 1996 Oct 15;144(8):760-71. Geerts H. NC-531 (Neurochem). Curr Op in Investig Drugs. 2004 Jan;5(1):95-100. Giacchino J, Criado JR, Games D, Henriksen S. In vivo synaptic transmission in young and aged amyloid precursor protein tran sgenic mice. Brain Res. 2000 Sep 8;876(12):185-90. Gordon MN, King DL, Diamond DM, Jantzen PT, Boyett KV, Hope CE, Hatcher JM, DiCarlo G, Gottschall WP, Morgan D, Are ndash GW. Correlation between cognitive deficits and Abeta deposits in transgen ic APP+PS1 mice. Neurobiol Aging. 2001 MayJun ;22(3):377-85. Hardy J. The Alzheimer family of diseases: many etiologies, one pa thogenesis? Proc Natl Acad Sci U S A. 1997 Mar 18;94(6):2095-7. Hardy J, Selkoe DJ. The amyloid hypothesi s of Alzheimer's disease: progress and problems on the road to therapeutics. Science. 2002 Ju l 19;297(5580):353-6.
154 Haughey NJ, Nath A, Chan SL, Borchard AC, Rao MS, Mattson MP. Disruption of neurogenesis by amyloid beta-peptide, and pe rturbed neural progenitor cell homeostasis, in models of Alzheimer's disease. J Neurochem. 2002 Dec;83(6):1509-24. Holcomb L, Gordon MN, McGowan E, Yu X, Be nkovic S, Jantzen P, Wright K, Saad I, Mueller R, Morgan D, Sanders S, Zehr C, O'Campo K, Hardy J, Prada CM, Eckman C, Younkin S, Hsiao K, Duff K. Accelerated Al zheimer-type phenotype in transgenic mice carrying both mutant amyloid pr ecursor protein and presenilin 1 transgenes. Nat Med. 1998 Jan;4(1):97-100. Holcomb LA, Gordon MN, Jantzen P, Hsiao K, Duff K, Morgan D. Behavioral changes in transgenic mice expressing both amyloid precursor protein and presenilin-1 mutations: lack of association with amyloid de posits. Behav Genet. 1999 May ;29(3):177-85. Hoffman JM, Welsh-Bohmer KA, Hanson M, Cr ain B, Hulette C, Earl N, Coleman RE. FDG PET imaging in patients with pathologi cally verified dementia. J Nucl Med. 2000 Nov;41(11):1920-8. Hsiao K, Chapman P, Nilsen S, Eckman C, Harigaya Y, Younkin S, Yang F, Cole G. Correlative memory deficits, Abeta elevation, and amyloid plaques in transgenic mice. Science. 1996 Oct 4;274(5284):99-102. Huesgen CT, Burger PC, Crain BJ, Johnson GA. In vitro MR microscopy of the hippocampus in AlzheimerÂ’s disease. Neurology 1993;43:145Â–52. Hutton M, Hardy J. The presenilins and Alzheimer's disease. Hum Mol Genet. 1997;6(10):1639-46. Ickes BR, Pham TM, Sanders LA, Albeck DS, Mohammed AH, Granholm AC. Longterm environmental enrichment leads to regiona l increases in neurotrophin levels in rat brain. Exp Neurol. 2000 Jul;164(1):45-52. Irizarry MC, Soriano F, McNamara M, Page KJ, Schenk D, Games D, Hyman BT. Abeta deposition is associated with neuropil cha nges, but not with overt neuronal loss in the human amyloid precursor protein V717F (PDAP P) transgenic mouse. J Neurosci. 1997 Sep 15;17(18):7053-9. Irizarry MC, McNamara M, Fedorchak K, Hs iao K, Hyman BT. APPSw transgenic mice develop age-related A beta deposits and ne uropil abnormalities, but no neuronal loss in CA1. J Neuropathol Exp Neur ol. 1997 Sep ;56(9):965-73. Iwatsubo T, Mann DM, Odaka A, Suzuki N, Ihara Y. Amyloid beta protein (A beta) deposition: A beta 42(43) pr ecedes A beta 40 in Down syndrome. Ann Neurol. 1995 Mar;37(3):294-9.
155 Jankowsky JL, Melnikova T, Fadale DJ, Xu GM, Slunt HH, Gonzales V, Younkin LH, Younkin SG, Borchelt DR, Savonenko AV. Environmental enrichment mitigates cognitive deficits in a mouse model of Alzheimer's disease. J Neurosci. 2005 May 25;25(21):5217-24. Jankowsky JL, Xu G, Fromholt D, Gonzales V, Borchelt DR. Environmental enrichment exacerbates amyloid plaque formation in a tran sgenic mouse model of Alzheimer disease. J Neuropathol Exp Neurol. 2003 Dec;62(12):1220-7. Jellinger KA. Alzheimer disease and cerebr ovascular pathology: an update. J Neural Transm (2002) 109: 813Â–836. Jellinger KA. Head injury and dementia Curr Opin Neurol 2004 Dec;17(6):719-23. Jensen MT, Mottin MD, Cracchiolo JR, Leighty RE, Arendash GW. Lifelong immunization with human beta-amyloid (1-4 2) protects Alzheimer's transgenic mice against cognitive impairment throughout aging. Neuroscience. 2005 ;130(3):667-84. Jick H, Zornberg GL, Jick SS, Seshadri S, Drachman DA. Statins and the risk of dementia. Lancet. 2000;356:1627-31. Jin K, Peel AL, Mao XO, Xie L, Cottrell BA, Henshall DC, Greenberg DA. Increased hippocampal neurogenesis in Alzheimer's dis ease. Proc Natl Acad Sci U S A. 2004 Jan 6;101(1):343-7. Epub 2003 Dec 5. Johnson RA, Mitchell GS. Exercise-induced changes in hippocampal brain-derived neurotrophic factor and neurotrophin-3: e ffects of rat strain. Brain Res. 2003 Sep 5;983(1-2):108-14. Jorm AF, Korten AE, Henderson AS. The prevalence of dementia: a quantitative integration of the literature. Acta Psychiatr Scand. 1987 Nov;76(5):465-79. Juottonen K, Laakso MP, Insausti R, Lehtovirta M, Pitkanen A, Partanen K, et al. Volumes of the entorhinal and perirhinal cortices in AlzheimerÂ’s disease. Neurobiol Aging 1998;19:15Â–22. Kalmijn S, Launer LJ, Ott A, Witteman JC, Ho fman A, Breteler MM. Dietary fat intake and the risk of incident dementia in the Rotterdam Study. Ann Neurol. 1997 Nov;42(5):776-82. Katz IR, Jeste DV, Mintzer JE, Clyde C, Napolitano J, Brecher M. Comparison of risperidone and placebo for psychosis and be havioral disturbances associated with
156 dementia: a randomized, double-blind trial. Ri speridone Study Group. J Clin Psychiatry. 1999 Feb;60(2):107-15. Katzman R. Education and the prevalence of dementia and AlzheimerÂ’s disease. Neurology 1993;43:13Â–20. Kawarabayashi T, Younkin LH, Saido TC, Shoji M, Ashe KH, Younkin SG. Agedependent changes in brain, CSF, and plas ma amyloid (beta) protein in the Tg2576 transgenic mouse model of Alzheimer's di sease. J Neurosci. 2001 Jan 15;21(2):372-81. Kempermann G, Gast D, Gage FH. Neuroplas ticity in old age: sustained fivefold induction of hippocampal neurogenesis by l ong-term environmental enrichment. Ann Neurol. 2002 Aug;52(2):135-43. Kempermann G, Gage FH. Experience-depe ndent regulation of adult hippocampal neurogenesis: effects of long-term stimul ation and stimulus withdrawal. Hippocampus. 1999;9(3):321-32. Kempermann G, Kuhn HG, Gage FH. Experien ce-induced neurogenesis in the senescent dentate gyrus. J Neurosci. 1998a May 1;18(9):3206-12. Kempermann G, Brandon EP, Gage FH. Environmental stimulation of 129/SvJ mice causes increased cell prolifera tion and neurogenesis in the ad ult dentate gyrus. Curr Biol. 1998b Jul 30-Aug 13;8(16):939-42. Kempermann G, Kuhn HG, Gage FH. More hip pocampal neurons in adult mice living in an enriched environment. Na ture. 1997 Apr 3;386(6624):493-5. Kempermann G, Kuhn HG, Gage FH. Genetic influence on neurogenesis in the dentate gyrus of adult mice. Proc Natl Acad Sci U S A. 1997 Sep 16;94(19):10409-14. Kennard ML, Feldman H, Yamada T, Jefferi es WA. Serum levels of the iron binding protein p97 are elevated in Alzheimer' s disease. Nat Med. 1996 Nov;2(11):1230-5. Keyvani K, Sachser N, Witte OW, Paulus W. Gene expression profiling in the intact and injured brain following environmental en richment. J Neuropathol Exp Neurol. 2004 Jun;63(6):598-609. King DL, Arendash GW. Behavioral characte rization of the Tg2576 transgenic model of Alzheimer's disease through 19 months Physiol Behav. 2002 Apr ;75(5):627-42. King DL, Arendash GW. Maintained synaptophysin immunoreactivity in Tg2576 transgenic mice during aging: correlations with cognitive impairment. Brain Res. 2002 Feb 1;926(1-2):58-68.
157 Kitamura T, Mishina M, Sugiyama H. E nhancement of neurogenesis by running wheel exercises is suppressed in mice lacking NMDA receptor epsilon 1 subunit. Neurosci Res. 2003 Sep;47(1):55-63. Kivipelto M, Helkala EL, Laakso MP, Hanninen T, Hallikainen M, Alhainen K, Soininen H, Tuomilehto J, Nissinen A. Midlife vascular risk factors and Alzheimer's disease in later life: longitudinal, population ba sed study. BMJ. 2001 Jun 16;322(7300):1447-51. Klunk WE, Engler H, Nordbe rg A, Wang Y, Blomqvist G, Holt DP, Bergstrom M, Savitcheva I, Huang GF, Estrada S, Ausen B, Debnath ML, Barletta J, Price JC, Sandell J, Lopresti BJ, Wall A, Koivisto P, Antoni G, Mathis CA, Langstrom B. Imaging brain amyloid in Alzheimer's disease with Pittsburgh Compound-B. Ann Neurol. 2004 Mar;55(3):306-19. Kobayashi S, Ohashi Y, Ando S. Effects of enriched environments with different durations and starting times on learning capacity during agin g in rats assessed by a refined procedure of the Hebb-Williams maze task. J Neurosci Res. 2002 Nov 1;70(3):340-6. Lacor PN, Buniel MC, Chang L, Fernandez SJ, Gong Y, Viola KL, Lambert MP, Velasco PT, Bigio EH, Finch CE, Kra fft GA, Klein WL. Synaptic targeting by Alzheimer'srelated amyloid beta oligomers. J Neurosci. 2004 Nov 10;24(45):10191-200. LaFerla FM. Calcium dyshomeostasis and intr acellular signalling in Alzheimer's disease. Nat Rev Neurosci. 2002 Nov;3(11):862-72. Lalonde R, Dumont M, Staufenbi el M, Sturchler-Pierrat C, St razielle C. Spatial learning, exploration, anxiety, and motor coordination in female APP23 transgenic mice with the Swedish mutation. Brain Res. 2002 Nov 22;956(1):36-44. Lalonde R, Lewis TL, Straziel le C, Kim H, Fukuchi K. Transgenic mice expressing the betaAPP695SWE mutation: effects on e xploratory activity, anxiety, and motor coordination. Brain Res. 2003 Jul 4;977(1):38-45. Larson J, Lynch G, Games D, Seubert P. Alte rations in synaptic transmission and longterm potentiation in hippocampal slices fr om young and aged PDAP P mice. Brain Res. 1999 Sep 4;840(1-2):23-35. Launer LJ, Ross GW, Petrovitch H, Masaki K, Foley D, White LR, Havlik RJ. Midlife blood pressure and dementia: th e Honolulu-Asia aging study. Neurobiol Aging 2000;21:49-55. Lazarov O, Robinson J, Tang YP, Hairston IS, Korade-Mirnics Z, Lee VM, Hersh LB, Sapolsky RM, Mirnics K, Sisodia SS. Envir onmental enrichment reduces Abeta levels and amyloid deposition in transgenic mice. Cell. 2005 Mar 11;120(5):701-13.
158 Lee BC, Mintun M, Buckner RL, Morris JC. Imaging of Alzheimer's disease. J Neuroimaging. 2003 Jul;13(3):199-214. Leggio MG, Mandolesi L, Federico F, Spirito F, Ricci B, Gelfo F, Petrosini L. Environmental enrichment promotes improve d spatial abilities a nd enhanced dendritic growth in the rat. Behav Brain Res. 2005 May 20. Letenneur L, Gilleron V, Commenges D, He lmer C, Orgogozo JM, Dartigues JF. Are sex and educational level independent predictors of dementia and Alzheimer's disease? Incidence data from the PAQUID proj ect. J Neurol Neurosurg Psychiatry. 1999 Feb;66(2):177-83. Levi O, Jongen-Relo AL, Feldon J, Roses AD, Michaelson DM. ApoE4 impairs hippocampal plasticity isoform-sp ecifically and blocks the en vironmental stimulation of synaptogenesis and memory. Neur obiol Dis. 2003 Aug;13(3):273-82. Levy-Lahad, E., Wasco. W., Poorkaj, P., Romano, D.M., Oshima, J., Pettingell, W.H., Yu, C.E., Jondro, P.D ., Schmidt, S.D., Wang, K., Crowley, A.C., Ying-Hui, F., Guenette, S.Y., Galas, D., Nemens, E., Wijsman, E.M., Bird, T.D., Schellenberg, G.D. and Tanzi, R.E. (1995) Candidate gene for the chromosome 1 familial AlzheimerÂ’s disease locus. Science 269, 973Â–977. Liddell MB, Lovestone S, Owen MJ. Genetic risk of Alzheimer's disease: advising relatives. Br J Psychiat ry. 2001 Jan;178(1):7-11. Liu L, Ikonen S, Heikkinen T, Heikkil M, Puol ivli J, van Groen T, Tanila H. Effects of fimbria-fornix lesion and amyloid pat hology on spatial learning and memory in transgenic APP+PS1 mice. Behav Brain Res. 2002 Aug 21;134(1-2):433. Lle A, Berezovska O, Herl L, Raju S, Deng A, Bacskai BJ, Frosch MP, Irizarry M, Hyman BT. Nonsteroidal anti-inflammatory dr ugs lower Abeta(42) a nd change presenilin 1 conformation. Nat Med. 2004 Oct ;10(10):1065-6. Locascio JJ, Growdon JH, Corkin S.Cognitive te st performance in detecting, staging, and tracking AlzheimerÂ’s disease. 1995; Arch Neurol 52(11) : 1087Â–1099 Loo DT, Copani A, Pike CJ, Whittemo re ER, Walencewi cz AJ, Cotman CW. Apoptosis is induced by beta-amyloid in cult ured central nervous system neurons. Proc Natl Acad Sci U S A. 1993 Sep 1;90(17):7951-5. Loewenstein DA, Acevedo A, Czaja SJ, Duar a R. Cognitive rehabilitation of mildly impaired Alzheimer disease patients on c holinesterase inhibi tors. Am J Geriatr Psychiatry. 2004 Jul-Aug;12(4):395-402.
159 Luchsinger JA, Tang MX, Shea S, Mayeux R. An tioxidant vitamin intake and risk of Alzheimer disease. Arch Neurol 2003; 60: 203Â–08. Ma J, Yee A, Brewer HB, Das S, Potter H. Amyloid-associated proteins alpha 1antichymotrypsin and apolipoprotein E promote assembly of Alzheime r beta-protein into filaments. Nature. 1994 Nov 3;372(6501):92-4. Maccioni RB, Munoz JP, Barbeito L. The mo lecular bases of Alzheimer's disease and other neurodegenerative disorders. Arch Med Res. 2001 Sep-Oct;32(5):367-81. Maestre G, Ottman R, Stern Y, Gurland B, Chun M, Tang MX, Shelanski M, Tycko B, Mayeux R. (1995). Apolipoprotein E and Alzh eimer's disease: Ethnic variation in genotypic risks. Annals of Neurology. 37(2): 254-259. Mahendra N, Arkin S. Effects of four years of exercise, language, and social interventions on Alzheimer discourse. J Commun Disord. 2003 Se p-Oct;36(5):395-422. Masliah E, Sisk A, Mallory M, Mucke L, Schenk D, Games D. Comparison of neurodegenerative pathology in transgenic mice overexpressing V717F beta-amyloid precursor protein and Alzheimer's dis ease. J Neurosci. 1996 Sep 15;16(18):5795-811. Matsuoka Y, Picciano M, Malester B, LaFran cois J, Zehr C, Daeschner JM, Olschowka JA, Fonseca MI, O'Banion MK, Tenner AJ, Lemere CA, Duff K. Inflammatory responses to amyloidosis in a transgenic mouse model of Alzheimer's disease. Am J Pathol. 2001 Apr ;158(4):1345-54. Mattson, M. P. Apoptosis in ne urodegenerative disorders. Na ture Rev. Mol. Cell Biol. 1, 120Â–129 (2000). Mattson. Pathways towards and away from Alzheimer's disease.. Nature. 2004 Aug 5;430(7000):631-9. Mattson, M. P. Cellular actions of beta-amyloid precursor protein and its soluble and fibrillogenic derivatives. Physiol. Rev. 77, 1081Â–1132 (1997). Maurer K, Ihl R, Dierks T, Frolich L. Clin ical efficacy of Ginkgo biloba special extract EGb 761 in dementia of the Alzheimer type J Psychiatr Res. 1997 Nov-Dec;31(6):64555. Mayeux R, Saunders AM, Shea S, Mirra S, Evans D, Roses AD, Hyman BT, Crain B, Tang MX, Phelps CH. Utility of the apoli poprotein E genotype in the diagnosis of Alzheimer's disease. Alzheimer's Disease Centers Consortium on Apolipoprotein E and Alzheimer's Disease. N Engl J Med. 1998 Feb 19;338(8):506-11.
160 McGeer PL, McGeer EG. The inflammatory re sponse system of brain: implications for therapy of Alzheimer and other neurodegene rative diseases. Brain Res Brain Res Rev. 1995 Sep;21(2):195-218. Miyata M, Smith JD. Apolipoprotein E allele -specific antioxidant activity and effects on cytotoxicity by oxidative insults and be ta-amyloid peptides. Nat Genet. 1996 Sep;14(1):55-61. Mohammed AK, Winblad B, Ebendal T, La rkfors L. Environmental influence on behaviour and nerve growth factor in th e brain. Brain Res. 1990 Sep 24;528(1):62-72. Moncek F, Duncko R, Johansson BB, Jezova D. Effect of environmental enrichment on stress related systems in rats. J Neuroendocrinol. 2004 May;16(5):423-31. Moolman DL, Vitolo OV, Vonsattel JP, Sh elanski ML. Dendrite and dendritic spine alterations in Alzheimer models. J Neurocytol. 2004 May ;33(3):377-87. Morgan D, Diamond DM, Gottschall PE, Ugen KE, Dickey C, Hardy J, Duff K, Jantzen P, DiCarlo G, Wilcock D, Connor K, Hatcher J, Hope C, Gordon M, Arendash GW. A beta peptide vaccination prevents memory loss in an animal model of Alzheimer's disease. Nature. 2000 Dec 21-28 ;408(6815):982-5. Morimoto T, Ohsawa I, Takamura C, Ishi guro M, Kohsaka S. Involvement of amyloid precursor protein in functional synapse fo rmation in cultured hippocampal neurons. J Neurosci Res. 1998 Jan 15;51(2):185-95. Morris MC, Beckett LA, Scherr PA, Hebert LE, Bennett DA, Field TS, Evans DA. Vitamin E and Vitamin C supplement use and risk of incident Alzheimer disease. Alzheimer Dis Assoc Disord. 1998 Sep ;12(3):121-6. Mullan M, Crawford F, Axelman K, Houlde n H, Lilius L, Winblad B, Lannfelt L. A pathogenic mutation for probable Alzheimer's di sease in the APP gene at the N-terminus of beta-amyloid. Nat Genet. 1992 Aug;1(5):345-7. Need AC, Irvine EE, Giese KP. Learning and memory impairments in Kv beta 1.1-null mutants are rescued by environmental en richment or ageing. Eur J Neurosci. 2003 Sep;18(6):1640-4. Neve RL, McPhie DL, Chen Y. Alzheimer' s disease: a dysfunction of the amyloid precursor protein(1). Brain Res. 2000 Dec 15;886(1-2):54-66. Nestor PJ, Scheltens P, Hodges JR. Advances in the early detect ion of Alzheimer's disease. Nat Med. 2004 Jul;10 Suppl:S34-41.
161 Nilsson M, Perfilieva E, Johansson U, Or war O, Eriksson PS. Enriched environment increases neurogenesis in the adult rat de ntate gyrus and improves spatial memory. J Neurobiol. 1999 Jun 15;39(4):569-78. Notkola IL, Sulkava R, Pekkanen J, Erkinjuntti T, Ehnholm C, Kivinen P, et al. Serum total cholesterol, apolipopro tein E4 allele, and AlzheimerÂ’s disease. Neuroepidemiology. 1998;17:14-20. Oddo S, Billings L, Kesslak JP, Cribbs DH, LaFerla FM. Abeta immunotherapy leads to clearance of early, bu t not late, hyperphosphorylated tau aggregates via the proteasome. Neuron. 2004 Aug 5;43(3):321-32. Ott A, Breteler MM, van Harskamp F, Claus JJ, van der Cammen TJ, Grobbee DE, Hofman A. Prevalence of Alzheimer's diseas e and vascular demen tia: association with education. The Rotterdam study. BMJ. 1995 Apr 15;310(6985):970-3. Pennanen C, Kivipelto M, Tuomainen S, Hartikainen P, Hanninen T, Laakso MP, Hallikainen M, Vanhanen M, Nissinen A, Helk ala EL, Vainio P, Vanninen R, Partanen K, Soininen H. Hippocampus and entorhinal cortex in mild cognitive impairment and early AD. Neurobiol Aging. 2004 Mar;25(3):303-10. Petersen RC, Doody R, Kurz A, Mohs RC, Morris JC, Rabins PV, Ritchie K, Rossor M, Thal L, Winblad B. Current concepts in m ild contivie impairment. Arch Neurol. 2001 Dec;58(12):1985-92. Petersen RC, Smith GE, Waring SC, Ivnik RJ, Tangalos EG. 1997 Aging, memory and mild cognitive impairment. Int Ps ychogeriatrics 9 (Suppl. 1): 65-69. Pham TM, Ickes B, Albeck D, Soderstrom S, Granholm AC, Mohammed AH. Changes in brain nerve growth factor le vels and nerve growth factor receptors in rats exposed to environmental enrichment for one ye ar. Neuroscience. 1999;94(1):279-86. Picciotto MR, Wickman K. Using knoc kout and transgenic mice to study neurophysiology and behavior. P hysiol Rev. 1998 Oct;78(4):1131-63. Pinaud R. Experience-dependent immediate ear ly gene expression in the adult central nervous system: evidence from enriched-environment studies. Int J Neurosci. 2004 Mar;114(3):321-33. Pinaud R, Penner MR, Robertson HA, Currie RW. Upregulation of the immediate early gene arc in the brains of rats exposed to environmental enrichme nt: implications for molecular plasticity. Brain Res Mol Brain Res. 2001 Jul 13;91(1-2):50-6. Pompl PN, Mullan MJ, Bjugstad K, Arendash GW. Adaptation of the circular platform spatial memory task for mice: use in detecting cognitive impairment in the APP(SW)
162 transgenic mouse model for Alzheimer's disease. J Neurosci Methods. 1999 Feb 1;87(1):87-95. Rampon C, Jiang CH, Dong H, Tang YP, Loc khart DJ, Schultz PG, Tsien JZ, Hu Y. Effects of environmental enrichment on gene expression in the brai n. Proc Natl Acad Sci U S A. 2000 Nov 7;97(23):12880-4. Reines SA, Block GA, Morris JC, Liu G, Nessly ML, Lines CR, Norman BA, Baranak CC, Rofecoxib: no effect on Alzheimer's disease in a 1-year, randomized, blinded, controlled study. Neurology. 2004 Jan 13;62(1):66-71. Reisberg B, Franssen EH, Hasan SM, Monteiro I, Boksay I, Souren LE, Kenowsky S, Auer SR, Elahi S, Kluger A. Retrogenesi s: clinical, physiolo gic, and pathologic mechanisms in brain aging, Alzheimer's and other dementing processes. Eur Arch Psychiatry Clin Neuros ci. 1999;249 Suppl 3:28-36. Reisberg B, Doody R, Stoffler A, Schmitt F, Ferris S, Mobius HJ; Memantine Study Group. Memantine in moderate-to-severe Alzh eimer's disease. N Engl J Med. 2003 Apr 3;348(14):1333-41. Roberts GW, Gentleman SM, Lynch A, Murray L, Landon M, Graham DI. Beta amyloid protein deposition in the brain after severe head injury: implications for the pathogenesis of AlzheimerÂ’s disease. J Ne urol Neurosurg Psyc hiatry 1994; 57:419Â–425. Rosenthal TC, Khotianov N. Managing Alzheimer dementia tomorrow. J Am Board Fam Pract. 2003 Sep-Oct;16(5):423-34. Sadowski M, Pankiewicz J, Scholtzova H, Ji Y, Quartermain D, Jensen CH, Duff K, Nixon RA, Gruen RJ, Wisniewski T. Amyloi d-beta deposition is associated with decreased hippocampal glucose metabolism and spatial memory impairment in APP/PS1 mice. J Neuropathol Exp Ne urol. 2004 May ;63(5):418-28. Saito K, Elce JS, Hamos JE, Nixon RA.Wid espread activation of calcium-activated neutral proteinase (calpain) in the brain in Alzheimer disease: a potential molecular basis for neuronal degeneration. Proc. Na tl Acad. Sci. USA 90, 2628Â–2632 (1993). Saito S, Kobayashi S, Ohashi Y, Igarashi M, Komiya Y, Ando S. Decreased synaptic density in aged brains and its preventi on by rearing under enriched environment as revealed by synaptophysin contents. J Neurosci Res. 1994 Sep 1;39(1):57-62. Scarmeas N, Levy G, Tang MX, Manly J, Ster n Y. Influence of le isure activity on the incidence of Alzheimer's disease. Neurology. 2001 D ec 26;57(12):2236-42.
163 Scheuner D, Eckman C, Jensen M, Song X, Citron M, Suzuki N, Bird TD, Hardy J, Hutton M, Kukull W, Larson E, Levy-Lahad E, Viitanen M, Peskind E, Poorkaj P, Schellenberg G, Tanzi R, Wasco W, Lannfelt L, Selkoe D, Younkin S. Secreted amyloid beta-protein similar to that in the senile pl aques of Alzheimer's disease is increased in vivo by the presenilin 1 and 2 and APP mutatio ns linked to familial Alzheimer's disease. Nat Med. 1996 Aug;2(8):864-70. Schmand B, Smit J, Lindeboom J, et al. Lo w education is a genuine risk factor for accelerated memory decline and dementia. J Clin Epidemiol 1997;50:1025Â–33. Schooler C, Mulatu MS. The reci procal effects of leisure ti me activities an d intellectual functioning in older people: a longitudina l analysis. Psychol Aging. 2001 Sep;16(3):46682. Selkoe DJ. Alzheimer's disease: genes, proteins, and therapy. Physiol Rev. 2001 Apr;81(2):741-66. Selkoe DJ. The genetics and molecular pat hology of Alzheimer's disease: roles of amyloid and the presenilins. Ne urol Clin. 2000 Nov;18(4):903-22. Shankle WR, Romney AK, Hara J, Fortier D, Dick MB, Chen JM, Chan T, Sun X. Methods to improve the detection of mild c ognitive impairment. Proc Natl Acad Sci U S A. 2005 Mar 29;102(13):4919-24. Epub 2005 Mar 21. Sholl DA. Dendritic organization in the neurons of the visual and motor cortices of the cat. J Anat 1953;87 (4):387-406. Skoog I, Lernfelt B, Landahl S, Palmertz B, Andreasson L-A, Nilsson L, et al. 15-year longitudinal study of bl ood pressure and dementia. Lancet 1996;347:1141-5. Slooter AJ, Breteler MB, Ott A, Van Bro eckhoven C, van Duijn CM. APOE genotyping in differential diagnosis of Alzheimer' s disease. Lancet. 1996 Aug 3;348(9023):334. Small, Roses, Haines, and Pericak-Vance. Ge ne dose of apolipoprotein E type 4 allele and the risk of AlzheimerÂ’s disease in late onset families. Science. (2004) 261, 921Â–923. Smith MA, Hirai K, Hsiao K, Pappolla MA, Harris PL, Siedlak SL, Tabaton M, Perry G. Amyloid-beta deposition in Alzheimer transg enic mice is associated with oxidative stress. J Neurochem. 1998 May 1;70(5):2212-5. Smith C, Graham DI, Murray LS, Nicoll JA. Tau immunohistochemistry in acute brain injury. Neuropathol Appl Neurobiol 2003; 29:496Â–502.
164 Sonkusare SK, Kaul CL, Ramarao P. Dementia of Alzheimer's disease and other neurodegenerative disorders--memantine, a ne w hope. Pharmacol Res. 2005 Jan ;51(1):117 Stern Y, Gurland B, Tatemichi TK, Tang MX, Wilder D, Mayeux R. Influence of education and occupation on the inciden ce of Alzheimer's disease. JAMA. 1994 Apr 6;271(13):1004-10. Tales A, Muir J, Jones R, Bayer A, Snow den RJ. The effects of saliency and task difficulty on visual search performan ce in ageing and Alzheimer's disease. Neuropsychologia. 2004;42(3):335-45. Tanzi RE, Gusella JF, Watkins PC, Bruns GA, St George-Hyslop P, Van Keuren ML, Patterson D, Pagan S, Kurnit DM, Neve RL. Amyloid beta protein gene: cDNA, mRNA distribution, and genetic linka ge near the Alzheimer locus. Science. 1987 Feb 20 ; 235(4791):880-4 Teather LA, Magnusson JE, Chow CM, Wurtman RJ. Environmental conditions influence hippocampus-dependent behaviours and brain levels of amyloid precursor protein in rats. Eur J Ne urosci. 2002 Dec;16(12):2405-15. Teri L, Gibbons LE, McCurry SM, Logsdon RG, Buchner DM, Barlow WE, Kukull WA, LaCroix AZ, McCormick W, Larson EB. Exer cise plus behavioral management in patients with Alzheimer disease: a random ized controlled trial. JAMA. 2003 Oct 15;290(15):2015-22. Trinchese F, Liu S, Battaglia F, Walter S, Mathews PM, Arancio O. Progressive agerelated development of Alzheimer-like pat hology in APP/PS1 mice. Ann Neurol. 2004 Jun 1;55(6):801-14. Turner CA, Lewis MH, King MA. Environmen tal enrichment: effects on stereotyped behavior and dendritic morphology. Dev Psychobiol. 2003 Jul;43(1):20-7. Tsai PP, Stelzer HD, Hedrich HJ, Hackbarth H. Are the effects of different enrichment designs on the physiology and behaviour of DBA/2 mice consistent? Lab Anim. 2003 Oct;37(4):314-27. Valverde F. The rapid Golgi technique for staining CNS neurons: Light microscopy. Neuroscience Pr otocols 1993;1:1-9. van der Cammen TJ, Croes EA, Dermaut B, de Jager MC, Cruts M, Van Broeckhoven C, van Duijn CM. Genetic testing has no place as a routine diagnostic test in sporadic and familial cases of Alzheimer's disease. J Am Geriatr Soc. 2004 Dec;52(12):2110-3. van Praag H, Kempermann G, Gage FH. Ne ural consequences of environmental enrichment. Nat Rev Neurosci. 2000 Dec;1(3):191-8.
165 van Praag H, Christie BR, Sejnowski TJ, Gage FH. Running enhances neurogenesis, learning, and long-term potentiation in mi ce. Proc Natl Acad Sci U S A. 1999 Nov 9;96(23):13427-31. van Waas M, Soffie M. Differential envir onmental modulations on locomotor activity, exploration and spatial behaviour in young and old rats. Physiol Behav. 1996 Feb;59(2):265-71. Verghese J, Lipton RB, Katz MJ, Hall CB, Derby CA, Kuslansky G, Ambrose AF, Sliwinski M, Buschke H. Leisure activities and the risk of dementia in the elderly. N Engl J Med. 2003 Jun 19;348(25):2508-16. Verkhratsky A, Toescu EC. Endoplasmic retic ulum Ca(2+) homeostasis and neuronal death. J Cell Mol Med. 2003 Oct-Dec;7(4):351-61. Veurink G, Fuller SJ, Atwood CS, Martins RN Genetics, lifestyle and the roles of amyloid beta and oxidative stress in Al zheimer's disease. Ann Hum Biol. 2003 NovDec;30(6):639-67. Wagner AK, Kline AE, Sokoloski J, Zafont e RD, Capulong E, Dixon CE. Intervention with environmental enrichment after experi mental brain trauma enhances cognitive recovery in male but not female ra ts. Neurosci Lett. 2002 Dec 16;334(3):165-8. Wallace WC, Akar CA, Lyons WE. Amyloi d precursor protein potentiates the neurotrophic activity of NGF. Brain Res Mol Brain Res. 1997 Dec 15;52(2):201-12. Wang J, Dickson DW, Trojanowski JQ, Lee VM. The levels of soluble versus insoluble brain Abeta distinguish Alzheimer's diseas e from normal and pathologic aging. Exp Neurol. 1999 Aug;158(2):328-37. Wang HX, Karp A, Winblad B, Fratiglioni L. Late-life engagement in social and leisure activities is associated with a decreased risk of dementia : a longitudinal study from the Kungsholmen project. Am J Epidemiol. 2002 Jun 15;155(12):1081-7. Walsh DM, Klyubin I, Fadeeva JV, Cullen WK Anwyl R, Wolfe MS, Rowan MJ, Selkoe DJ. Naturally secreted oligomers of amyloi d beta protein potently inhibit hippocampal long-term potentiation in vi vo. Nature. 2002 Apr 4;416(6880):535-9. Weingarten MD, Lockwood AH, Hwo SY, Kirsc hner MW. A protein factor essential for microtubule assembly. Proc Natl Acad Sci U S A. 1975 May;72(5):1858-62. West, M. J. et al. Differences in the pa ttern of hippocampal neuronal loss in normal ageing and AlzheimerÂ’s disease. Lancet 344, 769Â–772 (1994). Westerman MA, Cooper-Blacketer D, Mari ash A, Kotilinek L, Kawarabayashi T, Younkin LH, Carlson GA, Younkin SG, Ashe KH. The relationship between Abeta and
166 memory in the Tg2576 mouse model of Al zheimer's disease. J Neurosci. 2002 Mar 1;22(5):1858-67. Whelihan, W.M., Thompson, J.A., Piatt, A.L., Caron, M.D., & Chung, T. (1997) The relationship of neuropsychological measures to levels of cognitive functioning in elderly subjects: A discriminant analysis ap proach. Applied Neuropsychology, 4, 160-164. Williams BM, Luo Y, Ward C, Redd K, Gibson R, Kuczaj SA, McCoy JG. Environmental enrichment: effects on spatial memory and hippocampal CREB immunoreactivity. Physiol Behav. 2001 Jul;73(4):649-58. Wilson RS, Mendes De Leon CF, Barnes LL, Schneider JA, Bienias JL, Evans DA, Bennett DA. Participation in cognitively stimulati ng activities and risk of incident Alzheimer disease. JAMA. 2002 Feb 13;287(6):742-8. Winblad B, Poritis N. Memantine in seve re dementia: results of the 9M-Best Study (Benefit and efficacy in severe ly demented patients during treatment with memantine). Int J Geriatr Psychiat ry. 1999 Feb;14(2):135-46. Winocur G. Environmental influences on c ognitive decline in aged rats. Neurobiol Aging. 1998 Nov-Dec;19(6):589-97. Winocur G, Greenwood CE. The effects of hi gh fat diets and environmental influences on cognitive performance in rats. Behav Brain Res. 1999 Jun;101(2):153-61. Wolfer DP, Litvin O, Morf S, Nitsch RM, Lipp HP, Wurbel H. Laboratory animal welfare: cage enrichment and mouse be haviour. Nature. 2004 Dec 16;432(7019):821-2. Wong TP, Debeir T, Duff K, Cuello AC. Reor ganization of choliner gic terminals in the cerebral cortex and hippocampus in transgenic mice carrying mutated presenilin-1 and amyloid precursor protein transgenes J Neurosci. 1999 Apr 1;19(7):2706-16. Yankner BA, Dawes LR, Fisher S, Villa-Kom aroff L, Oster-Granite ML, Neve RL. Neurotoxicity of a fragment of the amyloid precursor as sociated with Alzheimer's disease. Science. 1989 Jul 28;245(4916):417-20. Yamaguchi H, Hirai S, Morimatsu M, Shoji M, Harigaya Y. Diffuse type of senile plaques in the brains of Alzheimer-type dementia. Acta Neuropathol (Berl). 1988;77(2):113-9. Yamaguchi H, Sugihara S, Ogawa A, Oshi ma N, Ihara Y. Alzheimer beta amyloid deposition enhanced by apoE epsilon4 gene precedes neurofibrillary pathology in the frontal association cortex of nondemented se nior subjects. J Neur opathol Exp Neurol. 2001 Jul;60(7):731-9.
167 Yang DS, Smith JD, Zhou Z, Gandy SE, Martins RN. Characterization of the binding of amyloid-beta peptide to cell culture-derived native apolipoprotein E2, E3, and E4 isoforms and to isoforms from human plasma. J Neurochem. 1997 Feb;68(2):721-5. Young D, Lawlor PA, Leone P, Dragunow M, During MJ. Environmental enrichment inhibits spontaneous apoptosis, prevents se izures and is neuropr otective. Nat Med. 1999 Apr;5(4):448-53. Zandi PP, Anthony JC, Hayden KM, Mehta K, Mayer L, Breitner JC; Cache County Study Investigators. Reduced incidence of AD with NSAID but not H2 receptor antagonists: the Cache Count y study. Neurology 2002;59:880-6. Zandi PP, Anthony JC, Khachaturian AS, Stone SV, Gustafson D, Tschanz JT, Norton MC, Welsh-Bohmer KA, Breitner JC; C ache County Study Group. Reduced risk of Alzheimer disease in users of antioxidant vitamin supplements: the Cache County Study. Arch Neurol 2004; 61: 82Â–88.