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 001681155
007 cr mnu|||uuuuu
008 051012s2005 flu sbm s000 0 eng d
datafield ind1 8 ind2 024
subfield code a E14-SFE0001033
Effects of hyperoxia in alzheimers transgenic mice
h [electronic resource] /
by April Cox.
[Tampa, Fla.] :
b University of South Florida,
Thesis (M.S.)--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 115 pages.
ABSTRACT: An association between major surgery in the elderly and precipitation of Alzheimers disease (AD) has been reported. Hyperoxia (100%) oxygen is commonly administered after surgery to increase the oxygen content of blood. However, hyperoxia is a potent cerebral vasoconstrictor and generator of free radicals, as is [beta]amyloid (A[beta];). This study was aimed at examining behavioral, neuropathological, and neurochemical effects of hyperoxia treatments in APPsw transgenic mice (Tg+), which have elevated brain A[beta]; levels by 3-4 months of age but are not yet cognitively-impaired. At 3 months of age, Tg+ mice were pre-tested in the radial arm water maze (RAWM) task of working memory and found to be unimpaired. At 4.5 months of age, half of the Tg+ mice received the first of 3 equally-spaced hyperoxia sessions (3 hrs each) given over the ensuing 3 months. The other half of the Tg+ mice were exposed to compressed air during these 3 sessions.RAWM testing performed immediately following the final gas session at 7.5 months of age revealed significant working memory impairment in Tg+ mice exposed to hyperoxia. The Tg+ group that was exposed to placebo treatment showed a trend towards impairment, however, was not significantly different from the non-transgenic group. Hyperoxia-induced memory impairment in Tg+ mice did not involve changes in brain A[beta] deposition, degenerative cell numbers in hippocampus, neocortical lipid peroxidation, or hippocampal levels of APP, ApoE, COX-2, or GFAP. The combination of excess A[beta] and hyperoxia could have induced greater oxidative stress and cerebral vasoconstriction than either one alone, resulting in a pathologic cerebral hypoperfusion that triggered subsequent cognitive impairment.
Adviser: Dr. Gary Arendash.
t USF Electronic Theses and Dissertations.
i Effects of Hyperoxia in Al zheimerÂ’s Transgenic Mice by April Ann Cox A thesis submitted in partial fulfillment of the requirements for the degree of Master of Science Department of Biology College of Arts and Sciences University of South Florida Major Professor: Gary Arendash, Ph.D. Dr. Brian Livingston Dr. Huntington Potter Date of Approval: April 15, 2005 Keywords: hyperoxia, transgenic mouse, oxidative stress, vasoconstriction Copyright 2005, April Cox
i Table of Contents List of Tables................................................................................................................. ....iii List of Figures................................................................................................................ ....iv AlzheimerÂ’s Disease Background.......................................................................................1 Behavioral Characterization ..........................................................................................1 Pathology ........................................................................................................................3 Genetics .........................................................................................................................10 Diagnosis ......................................................................................................................12 Risk Factors ..................................................................................................................16 Treatments .....................................................................................................................18 Animal Models.................................................................................................................2 2 Transgenic Mice ............................................................................................................22 PDAPP Model ...............................................................................................................24 Psw(APP23) & APPsw Models ........................................................................................26 APPsw + PS1 Models .....................................................................................................32 Hyperoxia...................................................................................................................... ....40 Behavior ........................................................................................................................40 Pathological and Physiologica l Effects of Hyperoxia ..................................................43 Linking Hyperoxia and Precipitation of AD .................................................................50 Specific Aims.................................................................................................................. ..53 Materials and Methods......................................................................................................55 Animals .........................................................................................................................55 General Protocol ..........................................................................................................57 Hyperoxia ......................................................................................................................58 Behavioral Assessment ..................................................................................................58 Histological Analysis ....................................................................................................60 Neurochemical Analysis ................................................................................................63 Statistical Analysis ........................................................................................................64 Results........................................................................................................................ .......65 Behavior ........................................................................................................................65 Neuropathology and Neurochemistry Â…Â…Â…Â…Â…Â…Â…Â…Â…Â…Â…Â…Â…Â…Â…Â…Â…Â…Â…75 Discussion .........................................................................................................................82
ii General 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…Â…Â…83 Behavior 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…Â…...83 Pathology 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…Â….85 Vasoactive role of A 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…...93 Clinical Implications of the Hyperoxia Findings Â…Â…Â…Â…Â…Â…Â…Â…Â…Â…Â…Â…Â…Â…Â….94 References..................................................................................................................... ....95
iii List of Tables Table 1. Animals included in the behavior a nd biochemical portions of the study 58 Table 2. Lipid Peroxidation Markers 80
iv List of Figures Figure 1. General protocol time line for the hyperoxia study 59 Figure 2. RAWM pre-treatment acquisi tion (T1-T4) and memory retention (T5) in Tg+ mice and Tgmice over five 3-day blocks. 69 Figure 3. Pre-Treatm ent acquisition (T1-T4) and memory retention (T5) overall errors (5 block average) in Tgand Tg+ mice. 70 Figure 4. RAWM post-treatment working memory in Tg-, Tg+Con and Tg+O2 mice over three 3-day blocks. 71 Figure 5. Overall RAWM post-treatme nt working memory performance across 9 days of testing. 72 Figure 6. RAWM prevs. post-treatm ent overall T5 Errors for the Tg-, Tg+Con, and Tg+O2 groups. 73 Figure 7. Individual pl ots of prevs. post-treatmen t overall T5 errors for all animals in each group. 74 Figure 8. The mean number of degene rative neurons in various hippocampal regions using the acid fuchsin and touilidine blue method. 78 Figure 9. A series of correlation scatte r plots illustrating th at the degree of neuronal degeneration in the dentate gyrus and the number of degenerative neurons in, CA3, and CA4 region of the hippocampus were positively correlated with working memory impairment in Tg+ mice, irrespective of gas treatment. 79 Figure 10. A pair of correlation scatterplo ts illustrating that an inverse correlation was present between iso-furans in anterior cortex and working memory for Tg+ mice (left), as well as for combined Tgand Tg+ mice (right). 80 Figure 11. A loads in neocortex and hippocampus of Tg+ mice were not affected by hyperoxia treatments. 81
v Hyperoxia Treatment Triggers Cognitive Impa irment in AlzheimerÂ’s Transgenic Mice April A Cox ABSTRACT An association between major surgery in th e elderly and precipitation of AlzheimerÂ’s disease (AD) has been reported. Hyper oxia (100%) oxygen is commonly administered after surgery to increase the oxygen content of blood. However, hyperoxia is a potent cerebral vasoconstricto r and generator of free radicals, as is -amyloid (A ). This study was aimed at examining behavioral, neuropa thological, and neuroc hemical effects of hyperoxia treatments in APPsw transgenic mi ce (Tg+), which have elevated brain A levels by 3-4 months of age but are not yet cognitively-impaired. At 3 months of age, Tg+ mice were pre-tested in the radial arm water maze (RAWM) task of working memory and found to be unimpaired. At 4.5 months of age, half of the Tg+ mice received the first of 3 equally-spaced hype roxia sessions (3 hrs each) given over the ensuing 3 months. The other half of th e Tg+ mice were exposed to compressed air during these 3 sessions. RAWM testing perf ormed immediately following the final gas session at 7.5 months of age revealed significant working memory impairment in Tg+ mice exposed to hyperoxia. The Tg+ group that was exposed to placebo treatment showed a trend towards impairment, however, was not significantly different from the non-transgenic group. Hyperoxia-induced me mory impairment in Tg+ mice did not involve changes in brain A deposition, degenerative cell numbers in hippocampus, neocortical lipid peroxidation, or hippocampal levels of A PP, ApoE, COX-2, or GFAP. The combination of excess A and hyperoxia could have in duced greater oxidative stress and cerebral vasoconstriction than either one alone, resulting in a pathologic cerebral
vi hypoperfusion that triggered subs equent cognitive impairment. These results suggest that humans genetically pre-disposed to AD and those with increased brain A levels have increased risk of developing cognitive im pairment following hyperoxia treatment and cast doubt on the wide spread use of hyperoxia in aged individuals at risk for developing AD.
1 AlzheimerÂ’s Disease Background AlzheimerÂ’s disease can be described as a Â“disease of the brain that results in an illness of the mindÂ” (Martin et al., 2002). The term AlzheimerÂ’s Disease (AD), first discovered by Alois Alzheimer in 1906, is marked by severe memory loss as well as an array of both psychological and physical symptoms. The primary risk factor for developing AD is age. As humans progres sively have longer life spans it is not, therefore unexpected to also s ee a dramatic rise in the preval ence of the disease. In the year 2000 there were approximately 4.5 million diagnosed persons with AD in the US, by the year 2050 it is estimated that there will be 13.2 million persons with AD (Herbert et al., 2003). AD has been under study for over a century, yet, there is still a profound lack of understanding as to the exact molecular mechanisms underlying the disease. Research has lead to the identification of genetic muta tions responsible for th e early onset form of AD. These mutations have provided a great de al of insight into some of the underlying causes of the disease as well as provide res earchers with various an imal models. While there are currently several di fferent treatments designed to slow the progression of the disease, none of the current treatments have been able to provide a cure for this debilitating disease. Hopefu lly, with further research a nd a better understanding of the disease a cure will soon be in sight. Behavioral Characterization AD has become synonymous with the gene ral public as a disease primarily marked by memory loss. However, the dis ease encompasses an array of behavioral
2 symptoms such as paranoia, confusion, diso rientation, loss of mental competence, and difficulty in understanding language (Martin et al., 2002). The symptoms of AD are highly dynamic. Symptoms emerge and evolve at different rates and at different stages of the disease. While there is a large degree of va riation in the sympto ms an individual might be faced with, the disease does fo llow a predictable sequence of progression. Currently, it is commonly believed that pathological changes associated with AD begin many years prior to the manifestation of clinical symptoms, this time is termed the latent period. The earliest obs ervable behavioral symptom is an impairment in short-term (working) memory. Short-term memory is defined as the memory used to recall information that persists for only a few mi nutes after being exposed to the memory encoding stimuli. This early loss of memory coincides with studi es that suggest the hippocampus and entorhinal cortex are the first brain areas to be affected by the neurofibrillary pathology, areas that are believed to be important in new memory formation (Braak & Braak, 1991). A patient who only suffers from short-term (working) memory impairment and who retains the ab ility to perform everyday tasks can be classified as being in the pre-dementia stag e (Nestor et al., 2004). This stage has also been termed mild cognitive impairment (MCI). The underlying concept of MCI is that it represents the earliest perceptible stage of AD (Nestor et al., 2004). Thus, a patient with a diagnosis of MCI would be considered at high risk fo r the development of AD. However, currently due to a lack of homogene ity in the clinical testing and evaluation of MCI, a diagnosis does not fully correlate with an eventual development of AD. There is an obvious need for researchers to more precis ely define this stage of AD as it has been
3 suggested that the MCI stage might be an op timal time to intervene with drug treatment (Voisin et al., 2003). The time course of progression of MCI to AD is extremely variable. As mentioned earlier the first symp tom to appear is a deficit in short-term memory that can begin at least 5 years prior to di sease diagnosis. Deficits in semantic memory (ability to recognize words, faces, and objects), atten tional processes, and executive functions slowly begin to emerge in a patient who is in the pre-diagnosis stage (N estor et al., 2004). When these deficits become widespread and severe enough to interfere with a patients normal life, and the criteria for a diagnosis of AD are met as set by various groups ( for example, American Psychiatric Associat ionÂ’s Diagnostic and St atistical Manual of Mental Disorders), a patient would be then officially diagnosed with early stage AD (Nestor, 2004). Neuropsychiat ric symptoms also begin to emerge during the early and moderate stages of AD. The most comm on symptoms include depression, apathy, and irritability (Hart et al., 2003). Als o, but less frequently patients experience hallucinations, elation, and disinhibition (Har t et al., 2003). A dia gnosis of moderate stage of AD is normally given to a patient w hose symptoms have progressively worsened and who also starts to suffer from language deficits and a long term memory loss. The time frame of progression varies among indi viduals with some patients progressing from early to end stage AD in only a few years and others living with the diagnosis for 10-20 years. The late or advanced stage of AD is marked by a very severe loss of memory and cognitive functions, and a severe loss of moto r skills. In end stage AD a patient would be in a semi-vegetative bedridden state, death would most likely result from illness. Pathology
4 The complexity of the underlying pathological mechanisms causing AD can be attested to the fact that after nearly 100 years of re search into the disease there still is not a complete understanding. Currently, the path ological hallmarks of the disease include: neuritic plaques, neurofibrill ary tangles (NFTÂ’s), loss of cholinergic function, widespread neuronal loss and synaptic modifications in the cortex, hippocampus and other brain areas responsible for memory and cognition (Par ihar et al., 2004). There is much debate over the exact relationship among these pathologies. It has been firmly established that neuritic plaque formation precedes neurofibrilla ry formation in cortical areas (Vickers et al., 2000). Also, it has been established that pathologies begin developing as far as twenty years before the onset of clini cal symptoms during the latent period. One of the first changes the brain wo uld undergo during the latent period is a reduction in viable acetycholine receptors (Kih ara et al., 2004). More specifically, the M1, M2 and M4 subtypes of muscarinic acetych oline receptors have all been reported to be decreased in AD (Volpicelli et al., 2004) The functioning of the M1 receptors subtype is thought to be down-regulated in AD and treatment with M1 receptor agonists has shown positive results (Fisher et al., 2000) The loss of activity in the cholinergic system is a target point for many pharmaceutical therapies. Second to a loss of acetycholine receptors, the brain begins to develop and accumulate -amyloid (A ) plaques. The plaques consist of extracellular deposits of A arranged in a star-shaped mass with a compact core. The A that comprises the plaque is a byproduct that results from cleavage events to the larger parent protein; amyloid precurs or protein (APP). APP is a transmembrane protein that is found in the endoplasmic reticulum and is posttranslationally modified through the secr etory pathway to yield several different
5 peptides (Selkoe, 2001). The major secreted derivative of APP is -APPs a result of cleavage by -secretase 12 amino acids away from the NH2 terminal (Selkoe, 2001). There have been a number of suggestions ma de as to the exact functioning of the -APPs in a normal physiological state. Researchers have proposed that the -APPs peptide may act as an autocrine factor or as a neuritotr ophic factor. The deletion of the APP gene does not confer early mortality or high morbidit y suggesting that other homologous proteins may assume whatever role the -APPs plays (Selkoe, 2001). In an alternate toxic form of APP processing, the toxic A protein can be formed. First, the enzyme -secretase cleaves the APP protein at the N-terminus amino acid 671. Then a second enzyme, -secretase cuts the remaining larg er fragment of the APP protein at amino acid 711 or 713 to produce the A 40 or A 42 fragment (Tamagno et al., 2002). The A fragments then begin to aggregate toge ther to form first into dimers and oligomeric isoforms. Then as aggregation continues the fragments form diffuse plaques and then finally compact plaques. Activated glial cells such as mi croglia and astrocytes will surround the amyloid plaques in response to their formation and produce toxins such as inflammatory cytokines that lead to furt her neurodegeneration (Mat tson et al., 2004). There is also some evidence that activated microglia can clear A deposits (Jantzen et al., 2002). A 1-40 is the major species produced, however, the A 1-42 is less soluble and aggregates more readily (Veu rink et al., 2003). These plaque s are found most notably in the hippocampus and the association cortices Another key prot ein, apolipoprotein E (apoE) plays a vital role in the formation of neuritic plaques. While the exact mechanism is still unknown, studies done w ith transgenic mice have found that the presence of apoE is necessary in the formation of plaques (M arques et al., 2004). Current research is
6 focused on the role of proteolytic fragments in the formation of pla ques. The exact time course for the formation of a neuritic pla que is unknown. However, it is widely accepted that the process occurs rather slowly. The accumulated A 42 in plaques then begins to release diffusible oligomers of the A 42 and protofibrils that ar e thought to have adverse affects on neighboring cells (Selkoe, 2001). Dendritic loss, neuronal de ath, and dystrophic neurit es are often found within and surrounding the plaques. This cellula r damage is thought to be a result of the increased oxidative stress seen around plaques. Studies have shown that A can induce the formation of oxygen dependent free radicals that lead to lipid and protein oxidation (Veurink et al., 2003). The formati on of free radicals centered around A plaques results in areas of increased oxidative stress. The term oxidative stress is a general term used to describe the level of oxidative damage being caused by oxidants, such as reactive oxygen species (ROS) and reactive nitrogen species (RNS), to the surrounding cells and ti ssues (Emerit et al., 2004). Oxidative stress occurs when the producti on of oxidants exceeds the antioxidant capabilities of the cell (Macci oni et al., 2001). ROS, such as the superoxide (O-) free radical are formed as normal byproducts of metabolic reactions, such as the energy generating reactions that occur in the mitoc hondria. Oxygen is the primary source of free radicals in aerobic organism s, and exposure to hyperoxia (>21% Oxygen) increases the production of free radicals (Reite r et al., 1998). The role that oxidative stress plays in the pathogenesis of AD has been widely explor ed. The AD brain has been found to overproduce ROS and NOS, both of which are implic ated in cellular damage and apoptosis (Maccioni et al., 2001). Studi es have found that the AD brain has increases in lipid
7 peroxidation, decreases in polyunsaturated fatty acids, increases in 4-hydroxynonenal (an aldehyde product of lipid peroxidation), and increases in protein and DNA oxidation; all of which are evidence of oxidative stress (Markesbery et al., 1997). Oxidative stress plays a pivotal role in disr upting cellular functions such as, ion transport, calcium mobilization, and excitotoxicity and induc ing apoptosis in neurons, astrocytes and microglia (Emerit et al., 2004). Excitotoxic ity, the process by which excess glutamate over-activates glutamate receptors resulti ng in intracellular calcium overload with subsequent neural cell death, is seen in the AD brain and is closely linked with oxidative stress (Facheris et al., 2004). The exact s ource of the increased oxi dative stress seen in the AD brain is not yet fully known. Howeve r, one study suggested that it may result from the mitochondrial dysfuncti on induced by the presence of A and by glial activation (Emerit et al., 2004). It is widely accepted that oxidative stress is an early event in AD and most likely plays an active role in the pathogenesis of the disease (Practico et al, 2002). The damage caused by A accumulation and oxidative stress to neighboring cells triggers an inflammatory response. One of the first inflammatory responses seen is activated microglia and their cytokine rel ease (Selkoe, 2001). Sp ecifically, microglia exposed to A secrete pro-inflammatory cytokines interleukin-1 (IL-1 ), interleukin-6 (IL-6), and tumor necrosis factor (TNF ) (Tuppo et al., 2005). The release of these cytokines in turn attracts astrocytes which in turn release the additional inflammatory proteins ACT and ApoE (Selkoe, 2001) Astrocytes clustered around A plaques have been shown to secrete chemokines, cytoki nes, and reactive oxyge n species that may contribute to the neuronal damage seen surrounding A plaques (Johnstone et al., 1999).
8 In addition to secreting cytokines and stimul ating astrocytes, microglia also appear to play a role in A clearance. Activated microglia have been shown to actively phagocytize and clear A away from deposition sites (Rogers et al., 2001). The inflammatory response seen in the brain after the accumulation of A plaques is thought to play a pivotal ro le in the cascade of patholog ical changes seen in the AD brain, including accentuation of NFT formation. This idea has been coined the amyloid cascade hypothesis and has been overwhelmi ngly the most popular current concept for the explanation of AD. However, there are researchers who challenge the amyloid hypothesis. One of the biggest pieces of eviden ce that researchers who oppose the amyloid theory use is the concept that A levels have little predictive value for the cognitive state of the patient (Lee et al., 2004). Th ey also use the premise that studies used to formulate the amyloid theory were done in cell culture mode ls and that they are not good indicators of in vivo processes. As an alternate theory re searchers have been s uggesting for years that A does not function as the disease-caus ing pathogenic agent, but rather A plays a protective role in the brain (Lee et al., 2004). A recent study found that A functions as an endogenous regulator of ne uronal activity (Esteban et al., 2004). There has been growing evidence that points toward s the antioxidant capabilities of A as demonstrated by its ability to prev ent lipoprotein oxidati on in the CSF (Kontush et al., 2001). A Â’s ability to bind and chelate Cu is the mo st likely mechanism through which it reduces oxidative stress (Atwood et al., 2003). Thes e findings are in accord with earlier work that has found A plaque load to be inversely corre lated with oxidative stress (Nunomura et al., 2001). Thus, oxidative stress triggers A generation in an attempt to decrease the
9 oxidative stress load (Atwood et al ., 2003). Individuals with high A plaque loads that remain cognitively normal may be regarded as showing a healthy response to the oxidative stress encountered during ageing. Researcher s in favor of viewing A as a side effect, rather than causative agent of the di sease also acknowledge th e toxic properties of the protein. They propose that at high enough oxidative stress burdens A begins to accumulate and results in uncontrollable plaque growth with the subsequent toxic effects mentioned above (Atwood et al., 2003). This theory, in sharp contrast to the amyloid cascade hypothesis, calls into question the wide spread use of therapeutic strategies aimed at reducing A loads. NFTs, the second hallmark lesion of AD, ar e intracellular aggreg ates composed of the microtubule associated tau protein (Dicks on et al., 2004). NFTÂ’s form as a result of hyperphosphorlyation of the tau protein at several different residues. The phosphorylation of the tau protein leads to a dissociation from the microtubules and eventual aggregation into paired helical f ilaments (Rosenberg et al., 2000). These tangles can be found in numerous brain regions including: hippocampus, parahippocampal gyrus, amygdala, frontal, temp oral and occipital a ssociation cortices (Selkoe, 2001). NFTs are also found in numerous other neurodegenerative disorders (KufÂ’s disease, subacute sclerosing panence phalitis, etc), and in th e brains of aged nondemented individuals (Gomez-Ramos et al., 1998). There is evidence that suggests A plays a causative role in the formation of ne urofibrillary tangles (NFTs). However, the exact relationship is still uncerta in. The fact that NFTs can arise in the absence of betaamyloid deposition suggests that tangles can form as a result of a variety of neuronal conditions.
10 Currently, there are many different theori es as to the exact etiopathology that induces the formation of NFTÂ’s in AD. On e of these theories of NFT formation has arisen by examining the brains from pre-di agnosed AD patients. Researchers have found that the pathologies associated with A plaques very closely resembles structural damage to the axons of neurons (Vickers et al., 2000). The structural damage to the axons caused by the accumulation of A early in the disease may trigge r an inflammatory response in the brain that becomes prolonged and results in the formation of NF TÂ’s (Vickers et al., 2000). The degree of cognitive impairment in an AD patient correlates more closely with the amount and distribution of NF TÂ’s than with the amount of A deposited in the brain (Bennett et al., 2004). This finding puts an increased emphasis on th e role that NFTÂ’s play in the progression of AD. During the course of AD progression a patient would also expect to see a substantial decrease in brain volume. This global volume loss is due to many factors; death of neurons, gyri shrinking, widene d sulci, and loss of synapses. Genetics AlzheimerÂ’s disease can be subdivided into two main groups: familial and sporadic. Familial AD (FAD) accounts for approximately 10% of all cases and is characterized by known genetic mutations and a much earlier age of onset. Symptoms can arise in patients with FAD as early as 40 years of age but generally occur no later than 60 years of age. Phenotypically, familial AD is almost indistinguishable from sporadic AD, except for the fact that some forms of familial AD have a more rapid progression than the sporadic forms. Sporad ic AD accounts for approximately 90% of all cases and is distinguished by a late age of ons et. Sporadic AD also has an associated
11 genetic mutation that has been discovered, th e ApoE4 allele, that will be discussed in detail later. This mutation does not definitively confer development of AD but persons carrying the gene are considered to be at a mu ch higher risk. Age of onset varies greatly in the sporadic form of AD, ranging from 65-90 years old. Once a person reaches 90 their chances of developing AD drops dramatically. The first genetic mutation that was discovered in the familial form of AD was a mutation in the gene coding for APP (Sel koe, 2001). There are 16 known pathogenic mutations and 4 nonpathogenic mutations in th e APP gene (Zekanowski et al., 2004). Mutations result in changes in the peptide sequence located next to the -secretase and secretase cleavage sites. These changes in sequence, although subtle are thought to have a major affect on the proteolytic cleavage of the protein. The double Swedish mutation (K670N/M671L), a mutation isolated from a family from Sweden, affects -secretase activity and results in elevated levels of A 40 and A 42 (Zekanowski et al., 2004). The London mutation affects the activity of -secretase, resulting in elevated levels of both A 40 and A 42 with the greatest increase seen in A 42 In general, families carrying AD pathogenic mutations of APP develop symptoms in their fifties. APP mutations only account for 4-6% of the familial form of AD (Zekanowski, 2004). These mutations have helped provide more evidence for the amyloid cascade hypothesis. APP gene dosage has proven to be anot her key piece of evidence for researchers supporting the amyloid hypothesis. The effects of the location of the APP gene on allele 21 is clearly seen in patients with downÂ’s syndrome. An individual with trisomy 21 can expect to see identical neuropathology as co mpared with AD due to the increased gene
12 expression seen with an extra copy of the A PP gene. Individuals identified with this mutation often have AD onset during their 50Â’s (Selkoe, 2001). The most common genetic mutation resulting in FAD are those found in presenilin 1 genes (PS1) and presenilin 2 genes (PS2) (Zek anowski et al., 2004). While the exact functions of the PS1 and PS2 genes ar e not clear, there is growing evidence that suggests they both may serve as the active sites of the -secretase complex (Zekanowski et al., 2004). The mutations directly a ffect APP processing, resulting in increased amounts of A 42, and providing even more evidence for the amyloid cascade hypothesis. Currently, there have been more than 100 muta tions identified in the PS1 gene and only 9 in the PS2, with nearly a ll being missense mutations located around the transmembrane domains of the proteins (Zekanowski et al ., 2004). Animal models of APP mutations crossed with animals models carrying the PS1 mutation yield animals with dramatically increased levels of A plaques. The fact that presenil in mutations result in increased A levels further supports the amyloid cascade hypothesis. Diagnosis The tools and techniques us ed in the diagnosis of AD are advancing as fast as technology and knowledge of the disease permits. For detecting the ear ly form of AD, a test must be able to accurately discrimina te between pathological processes and normal ageing, as well as other neurological disord ers that may cause memory loss (Nestor, 2004). The diagnosis of AD is commonly based on the criteria put forward by the National Institute of Neurologic and Communi cative Disorders and Stroke-AlzheimerÂ’s Disease and Related Disorders Associat ion (NINCDS-ADRDA) (Cummings, 2004). According to the NINCDS-ADRDA there are thr ee levels of diagnosis: definite (clinical
13 diagnosis with histological confirmation), probable (typical clinical syndrome without histological confirmation) or possible (atypical clinical syndrome but no alternative diagnosis apparent withou t histological confirmati on) (Cummings, 2004). A patient who initially presents themselves with dementia as a candidate for early AD would first go through a battery of tests ruling out other dementia causing diseases such as syphilis, AIDS, inflammatory diseases or exposure to dementia-inducing toxins (Cummings, 2004). After ruling out other ca uses of dementia, a patient would most likely begin cognitive testing. The most common cognitive te st given is the Mini-Mental State Exam (MMSE). Psychological testing is required for a clin ical diagnosis of AD and can help the physician determine the progression of the disease. Up until very recently, psychological testing alone was considered adequate to provide enough evidence for a diagnosis of AD. Psychologica l testing does have its weaknesses. One weakness is the high variability in perform ance seen in the normal human population, which can result in test scores not indicative of cognitive status. The tests themselves can be biased, and norms are established from pools of people not necessarily indicative of the population at large (Zamrini et al., 2004) Other approaches using biological markers or neuroimaging eliminate the pitfal ls seen in psychological screening of AD patients. The most researched biological ma rkers used to diagnose AD are A proteins and levels of total and phosphorylated tau. A is generated in all individuals from normal APP metabolism and is secreted in the extr acellular space allowing for its detection in both the cerebral spinal fl uid (CSF) and blood plasma (Seubert et al., 2002). The majority of studies have found a decrease in the amount of A 42 in the CSF of AD
14 patients (Sobow et al., 2004). Howeve r, the use of decreased levels of A 42 is not a definitive diagnosis of AD alone since decreased levels of A 42 have also been found in patients with depression and other dementia s (Andreasen et al., 2001) Numerous studies have found increased amounts of both normal and phosphorylated tau in the CSF of AD patients (Sobow et al., 2004). The combina tion of analyzing the ratio of normal tau to A 42 has proven to be the most effective tool in using biological markers in the diagnosis of AD (Gomez-Tortosa et al., 2003). The accuracy of using these biological markers alone in diagnosing AD is widely questioned. However, the analysis of biological markers can greatly help in an AD diagnos is when paired with other diagnostic techniques such as brain imaging. A landmark announcement by Medicare in June, 2004 stated th at it will be providing coverage for brain scans in order to help confirm diagnosis of AD. Scientists have known for years that many different im aging techniques including positron emission tomography (PET), magnetic resonance imag ing (MRI) and computer tomography (CT) scans can help tremendously in confirming AD Also, due to the fact that AD begins decades prior to psychological symptoms, im aging studies may provide doctors with a tool for diagnosing AD in the very beginning stages. Neuroimaging can provide doctors with not only preclinical diseas e state, but also rate of di sease progression if imaging studies are done in series (Z amrini et al, 2004). The implications of being able to diagnose a patient 20 years prior to disease ons et are astounding in th at it opens the door for possible prevention therap eutics to be utilized. Both structural and functional neuroimagi ng techniques offer insight into disease progress and diagnosis. Stru ctural imaging uncovers anat omical changes that are
15 occurring in the brain prior or during dis ease pathogenesis. Imaging studies recently have shown that atrophy in the medial tempor al lobe shows very high predicted cognitive decline (Fox et al., 2004). Longitudinal studies have s hown that the severity of AD correlated with rates of hippocampal volume loss (Zamrini et al ., 2004). Neuroimaging of structural changes that occur not only give researchers insight into the disease pathogenesis, but also may provi de a tool that can accurately assess an individualÂ’s risk for developing AD. Functional imaging may offer an even mo re accurate way to diagnose AD. The most common way researchers measure cere bral activity is through Positron Emission Tomography (PET). PET scanning works by in jecting the patient with a small amount of tracer drug that attaches itself to a variety of energy sources, most often 2deoxyglucose (2DG). The 2DG, tagged with a positron emitting isotope, will accumulate in areas that are more active than others. Th e tracer isotope emits positrons, that collide with electrons nearby, resulting in gamma rays that are picked up by sensors and mapped onto different brain regions showing areas of high activity, and also more importantly in the case of diagnosing AD, areas of low activ ity. A patient suspected of having AD will show a dysfunction or reduced brain activity (low 2DG levels) in temporo-parietooccipital cortices on a PET scan (Nestor et al., 2004). Another interesting imaging technique that has emerged in the last year is a technique using a novel amyloid-imaging PET tracer, called Pittsburgh Compound-B (PIB) (Klunk et al., 2004). A study done r ecently involved tracing amyloid deposition in 16 patients showed that PIB retention was incr eased primarily in the association cortex, an area known to contain large amyloid load in AD patients (Klunk et al., 2004). This
16 new imaging technique may provi de researchers with a tool to quantify amyloid load in living patients. Also, the technique has th e ability to monitor the effects of drugs designed to limit amyloid deposition. In addition to PET scanning, another technique, Single-Photon Emission Computed Tomography (SPECT) is also used in functional im aging studies. The basic idea behind SPECT scanning is the same to PET scanning. Only SPECT scanning utilizes a radioactive compound th at is injected into the pa tient. Regions with higher activity show an increased signal that is ma pped onto brain regions. SPECT scans have also helped to provide more evidence for the argument that MCI is an antecedent to AD. A study using SPECT scanning techniques reve aled that patients diagnosed with MCI who showed a hypoperfusion in the posterior cingulated cortex la ter converted to AD (Nestor et al., 2004). Findings such as thes e are greatly improving the accuracy of AD diagnosis. Scanning techniques offer a promising tool for a more accurate, earlier detection of AD. However, due to the nature of the di sease, it is still essential to perform cognitive testing. Brain imaging techniques, alongsid e cognitive testing and biological marker analysis can provide doctors with a po werful tool in the diagnosis of AD. Risk Factors The only known genetic inheritance that is considered to be a risk factor for developing sporadic AD is the 4 allele of the apolipoprotei n gene. Apoliprotein (ApoE) has three alleles, 2, 3, and 4. ApoE is involved in a number of functi ons including cholesterol transport, dendr itic growth, neuronal repair and it has possible antiinflammatory functions (Lahiri et al., 2004). Individuals who are homozygous for the
17 ApoE4 allele of the protein would be consid ered at a higher risk for the eventual development of AD and account for 10-15% of AD cases (Zekanowski et al., 2004). However, only one third of people who inherit these alleles develop AD illustrating further the complexity of factors involved with the disease. Inheritance of the 4 allele of ApoE results in increased amyloid accumula tion, neurotoxicity, and increased oxidative stress (Lahiri et al., 2004). While this is the only widely documented genetic inheritance associated with an increased occurence of sporadic AD, researchers believe that most likely mutations in as many as 50 other genes co uld act as risk factors (Zekanowski et al., 2004). Although these genes have yet to be disc overed, the fact that first degree relatives of sporadic AD patients are considered to be at a higher risk helps provide evidence that genetics may play a larger role than what is currently known. The primary risk factor for AD is aging, which is unavoidable. Interestingly, an individualsÂ’ risk for developing AD will peak at around the age of 90 and then decline in the mid 90Â’s. Even though an individual ha s no way of controlling the risk factor of aging, there are many other risk factors that fall within a personsÂ’ ability to control. Environmental factors including lifestyle choices have been investigated using identical twin studies. Using twins that are discordant for AD, resear chers have examined varying environmental factors that may have played a role in the development of the disease. One such study found that a higher level of schooling was correlated with a decreased risk of developing AD (Raiha et al., 1998). Also, researchers from the same twin study found that a reduced risk was associated with ambidextrousness, and an increased risk was associated with both marriage an d widowhood (Raiha et al., 1998).
18 Vascular risk factors, such as hypertensi on, have also been shown to be associated with AD (Skoog et al., 2003). In order for hypert ension to be considered a risk factor for AD, an individual would have experienced it during mid life, decades prior to developing AD. Cholesterol levels have been examined as possible risk factors. Studies have shown that individuals with higher total cholesterol levels had nearly tripled their risk in developing AD (Wellington et al., 2 004). One risk factor that can be controlled is an individualÂ’s diet. A di et high in vitamins C, B6, E, B12 and folate have been associated with a lower risk of AD (Luchsinger et al., 2004). Studies have reported that diets low in fish and cereals are associated with an increa sed risk for the development of AD (Grant et al., 1999). The wide array of risk factors seen to play a role in AD, is further testament to the complex nature of the disease in term s of both its development and progression. Treatments While a cure has not yet been discover ed for AD, many treatments are available that have been proven to lessen symptoms. The treatments can be broken down into two broad categories: drug therapies and altern ative treatments. Drug therapies include; antioxidants, cholinesterase inhibitors, NMDA receptor antagonists (Memantine), antiinflammatory agents, neuropsychiatric drugs, herbal supplements, and statins. Also, there are several drug therapies currently under clinical investigation including the antiamyloid drug Alzhemed, and estrogen replacem ent therapies. Alternative treatments include psychotherapy, music therapy, exercise and other environmentally stimulating activities. Alpha-tocopherol (vitamin E) has been i nvestigated as a potential therapeutic for AD due to its antioxidant capabilities. Vitami n EÂ’s primary function is to defend tissues
19 against oxidative damage by reducing lipid peroxidation in membranes (Conte et al., 2004). It has been shown to successfully cr oss the blood brain barrier and significantly reduce brain lipid peroxidation rates (Conte et al., 2004). In the Tg2576 mouse model for AD, vitamin E has been shown to reduce A loads (Sung et al., 2004). The increased oxidative stress seen in AD due to deposition of A and resultant neuroinflammation is thought to play a central role in the disease. By administering a potent antioxidant such as vitamin E, the disease progress may be reta rded. Several epidemiological studies have pointed towards the protective effects that v itamin E plays in AD, and current clinical practice is in favor of using it as a treatment for AD (B erman et al., 2004). Cholinesterase inhibitors ha ve become the standard of treatment for patients with AD (Cummings, 2004). The brain of an AD patie nt is marked by an extreme decrease in cholinergic activity. A cholines terase inhibitor wo rks to prevent the enzyme responsible for breaking down acetycholine, thus, incr easing the amount available for neuronal signaling. Four cholinesterase inhibitors ar e currently FDA approved in the treatment of AD: tacrine, donepezil, rivastigmine, and ga lantamine (Cummings, 2004). Of the four, tacrine (Cognex) is rarely used due to its pot ential liver toxicity (Del agarza et al., 2003). Cinical trials have shown that treatment with cholinestera se inhibitors delays nursing home placement and stabilizes or improves cognitive function (Delagarza et al., 2003). The effects of cholinesterase inhibitors are however, modest. The drugs improve cognitive function in mild dementia only and cannot prevent or slow the disease progression (Sonkusare et al., 2005). Studies have indicated that treatment with these drugs will delay further cognitive deteriora tion by about one year (Delagarza et al., 2003).
20 The FDA has also approved another drug, Memantine, for the treatment of AD. Memantine is a partial NMDA receptor antagonist. NMDA-receptor-mediated glutamate excitotoxicity plays a pivotal role in the neuronal death seen as a result of A deposition (Sonkusare, 2005). By blocking the function of NMDA receptors, the drug prohibits the toxic effects of the receptors that are seen in AD. Memantine has been approved for the treatment of moderate to severe AD, and ha s been found to improve cognitive, social and physical impairments. Antiflammatory drugs (NSAIDs) have been considered as a possible therapeutic in AD. The inflammation seen around neuritic plaques is thought to play a role in the subsequent neuronal damage s een surrounding the plaques (Del agarza et al., 2003). By reducing the inflammatory response, the am ount of neuronal damage could also be limited. Large clinical trials have not yet b een able to show an improvement in patients undergoing anti-inflammatory drug therapies (Imbimbo et al., 2004). Current research has shown the ability of sele ctive anti-inflammatory drugs to decrease production of A (Imbimbo et al., 2004). New research is underway to develop new NSAID analogues capable of providing both antiflammatory protec tion as well as anti-amyloid activity. Patients suffering from AD often expe rience psychiatric and behavioral disturbances. In order to alleviate these sy mptoms on both the patient and caregiver, it has become fairly common to administer a variety of drugs used to alleviate the individual symptoms of the patient. Anti-psychotic drugs such as risperidone are commonly used in patients experiencing ps ychosis (Weiser et al., 2002). Selective serotonin reuptake inhibitors (SSRIs) are used in patients experiencing depression. Also, mood stabilizers and sedatives are used to help stabilize behavior in AD patients. These
21 drugs are all used to manage the symptoms of AD only and do not confer any protective or therapeutic effects to the disease itself. Ginkgo biloba has been the most popular herbal supplement used in both the prevention and treatment of AD. Studies have shown that AD patients treated with ginkgo biloba have had modest improveme nts as compared with control groups (Cummings, 2004). The exact function that gi nkgo biloba serves is not yet clear, however, it is thought that it confers its benefits through antioxidant and/or vasodilatory.properties Another herbal suppl ement, huperzine A is also used in the treatment of AD (Zangara et al ., 2003). Huperzine A serves as a cholinesterase inhibitor, it crosses the blood brain barrier easily, and has a prolonged bi ological half-life (Jiang et al., 2003). Although it is not yet FDA approved in the United St ates for treatment of AD, other countries such as China have been usi ng this supplement for a number of years. Cholesterol lowering drugs (statins) have been shown to have protective effects against AD. Recent evidence suggests th at cholesterol metabolism modulates A production and that drugs designe d to inhibit cholesterol meta bolism may be beneficial in the treatment of AD (Wolozin et al., 2004). Th ere have been several studies that have reported preliminary evidence that chronic treatment with statins is linked to a significantly decreased risk of deve loping AD (Wolozin et al., 2000). There are numerous current clinical tria ls investigating various drug therapies currently. A current phase III trial being c onducted at the National Institutes of Health (NIH) involves the new drug Alzehemed, a drug designed to prevent the aggregation of A and lower oxidative stress (Neurochem, Inc. 2004). Another phase III trial at the NIH
22 is using a neurotrophic agent Cerebrol ysin that has been found to decrease A and protect synaptic terminals in a rodent model of AD (Rockenstein et al., 2003). In addition to the numerous pharmaceutical therapies available for AD patients there are also several non-pharmacological tr eatments commonly use d. Caregivers have found that by enriching the patients envir onment through adding activities such as light exercise or by allowing them to listen to mu sic or view videotapes of family members that their neuropsychiatric symptoms are greatly alleviated (Cummings, 2004). While these treatments are designed to lessen the severity of sy mptoms during the progression of the disease, a recent study points towards the therapeutic potential that environmental enrichment can have. A study using a mouse model of AD found that long-term environmental enrichment greatly improved cognitive function in animals that otherwise are cognitively impaired (Arendash et al., 2004 ). This study suggests that environmental enrichment can provide therapeutic cogniti ve benefits to patients with AD. Animal Models Transgenic Mice Animal models have proven to be very useful in the resear ch of many diseases such as ParkinsonÂ’s, diabetes, and ALS. The creation of an animal model allows researchers to test potential drug therapie s as well as study in further detail the pathologies underlying the diseas e. The designs and creatio ns of animal models are being modified as further technology and knowledge of diseases advances. While the genes inserted or methods of gene inserti on may change, the basic concept underlying the creation of an animal model remains the same.
23 In order to create an animal model fo r a disease with a known genetic mutation the gene encoding the mutation must be first identified and then cl oned. Next, embryonic stem cells are removed from the donor mother a nd the cloned gene of interest is inserted into the embryonic stem cells in a random inse rtion through the use of a vector. The cells that have incorporated the mutated gene are th en selected and reinserted into the embryo. The embryo is reinjected into the mother, and the resultant offspring will be a chimeric animal. A chimeric animal contains both nor mal cells and cells containing the genetic insertion or transgene. The chimeric animal will then be crossed with a wild-type mouse to create a heterozygote transgenic animal that can be bred with another heterozygote animal to create a homozygous transgenic animal. The resultant animal will be a transgenic animal expressing the gene produc t inserted into its genome with hopes of mimicking the human disease. An important element of designing a transgenic animal is choosing an appropriate promot er. Promoters determine the level, tissue specificity, and temporal pattern of the transgene being expressed (Picciotto & Wickman, 1998). The process of creating the AD specific tran sgenic animals is modified slightly from the process described above. The fert ilized egg is removed from the pregnant mother and the transgene, in cDNA form, is introduced into the single-cell mouse embryo through pronuclear injection (Picciotto & Wickman, 1998). The transgene injected embryo will be re-implanted into a pse udopregnant mother. The inserted DNA will become integrated into the mouseÂ’s genom e through a random insertion event. Most often multiple copies of the gene are integr ated into the genome. The copy number can have an affect on the transcrip tional level of the gene. Howe ver, the site of integration seems to have a larger effect on transcrip tion. Assuming the insertion event occurred at
24 the one cell stage all of the cel ls that compose the resultant offspring should contain the transgene. If the insertion is done in an embryo that has already undergone multiple rounds of cell division, then a chimeric animal will be produced. Researchers involved in th e field of AD have been successful in creating several different mouse models of AD including the PDAPP, Psw (APP23), and the APPsw(Tg2576) + PS1 models. All of these an imal models utilize the genetic mutations discovered in FAD in the APP and Presenilin genes. The mouse models currently used have proven to be very beneficial to research ers in helping to elucid ate further the disease mechanisms involved in AD. There is, however much criticism centered on the fact that the mouse models for AD are incomplete mode ls that do not mirror perfectly the disease as seen in humans. PDAPP Model One of the first mouse models develo ped for AD is the PDAPP model. The animal model was generated by a single subs titution in the APP ge ne associated with familial AD (P717V) being driven by a pl atelet-derived growth factor (PDGF)promoter (Games et al, 1995). The PDAP P model expresses high levels of human mutant APP that result in the formation of neuritic plaques, synap tic loss, astrocytosis and microgliosis (Games et al, 1995). Plaque formation is seen between 6-9 months of age, and is found in the hippocampus, corpus callosum and cerebral cortex (Games et al, 1995). Also, it has been reported that age-de pendent changes in s ynaptic densities has been seen in this animal m odel (Dodart et al, 2000). This animal model has failed to exhibit any neuronal loss in the entorhin al cortex, CA1 hippocampal subfield or cingulated cortex through 18 months of age (Ir izarry et al, 1997). Synaptic transmission
25 has been studied in the PDAPP model, a nd one study reported that there is abnormal neurotransmission in the hippocampal circuits prior to A deposition (Giacchino et al, 2000). The generation of the PDAPP mouse mo del has shown that the overproduction of the mutant APP gene is sufficient to cause plaque formation with subsequent activation of astro-glial cells that is similar to wh at is seen in AD (Masliah et al, 1996). The PDAPP model also exhibits numer ous behavioral impairments associated with AD like behavior. A study re ported that the model shows deficits in the radial arm maze task (an eight-arm dry radial maze designe d to test spatial discrimination) as early as 3 months of age (Dodart et al, 1999). The cognitive deficits in spatial learning may be due to hippocampal atrophy and modifications in synaptic density (D odart et al, 2000). Impairments in an object recognition task (t ested by analyzing a r odentsÂ’ propensity to explore a novel object as opposed to a familiar on e in an open field) are seen at 6 months of age in homozygous animals, and in 9-10 months of age in he terozygous animals (Dodart et al, 1999). Another study testing P DAPP animals across a full test battery of cognitive and sensorimotor tasks (at an ear ly and late time point) reported that at 2 months there were no cognitive deficits in th e PDAPP model (Nilsson et al, 2004). At the late time point (16 months), however, te sting revealed an impairment in the final block of Morris water maze acquisition and in overall radial arm water maze performance (Nilsson et al, 2004). It is noteworthy that the mixed background of mice in the Nilsson et al. (2004) study resulted in eliminati on of hippocampal atro phy present in earlier studies (Dodart et al., 1999). Studies have also been done examining the effect of amyloid plaque formation on behavior. One study reported that in the modified water maze (the escape platform is moved across 5 su ccessive locations and animals are given 8
26 trials a day designed to test working memory ) that both age-related and age-independent working memory impairments are seen in the PDAPP model (Chen et al, 2000). This study also reported that the impairments in this modified water-maze correlated with an age-related increase in plaque burden (Chen et al, 2000). An additional more recent study has also reported deficits in spatial le arning in the circular maze task in both young (3-5 months) and aged (20-26 months) PDAPP animals (Huitron-Resen diz et al, 2002). The authors suggest that glucose hypometa bolism, hippocampal atrophy, and age-related increases in A deposition may be the cause of the impairments seen. A study focusing on changes in emotionality in this mouse model found abnormalities in fear (tested by analyzing posture patterns in a fear conditioning paradigm) and exploratory activity (tested by analyzing motor patterns in the fear conditioning paradigm) in 11 month old females (Gerlai et al, 2002). Changes in emo tionality are prominent effects of AD, and it is therefore, important to have an animal model that incorporates not only pathological and cognitive changes but also em otional changes that mimic AD. Psw(APP23) & APPsw Models The Â“Swedish MutationÂ” was discovered in the gene encoding the APP protein in a Swedish family with a familial form of AD. The mutation has been used in the development of two animal models: the Psw (APP23) and the APPsw (Tg2576). Both of these animal models contain the same tr ansgene with a double mutation (K670N-M671L) found in the 695 amino acid long human APP gene (Mullan et al, 1995). These two AD animal models differ in the promoters used to drive expression of the mutant APP and the insertion sites of the transgene (both inser tions are random events). These differences result in two animal models showing unique pathologies.
27 The APPsw (Tg2576) mouse model for AD contains the 695 amino acid long human APP gene encoding the double mutation found in a Swedish form of familial AD (K670N-M671L) driven by a hamster prion protei n promoter (Irizarry et al, 1997). The Tg2576 animal model exhibits soluble A formation at 6 months of age and plaque formation by 11-13 months of age accompan ied by neuritic dystrophy and activated astrocytes and microglia (Iriz arry et al, 1997). Plaques ar e found predominately in the cortical and limbic regions of the brain. Th is model expresses the mutant APP only in neurons, presumably due to the actions of the prion promoter (Hsiao et al, 1995). Glialmediated inflammatory response has been stud ied in detail in the Tg2576 model. It has been found that the inflammato ry cytokine interleukin-1 and the tumor necrosis factor are localized to A plaques (Benzing et al, 1999). Al so this animal model has been shown to express increased glial fibrillary ac idic protein, a protein found to be increased in the AD brain (Lim et al, 2000). This model does not have any neuronal death in the CA1 region of the hippocampus, a region that undergoes profound loss of neurons in human AD (West, 1994). The behavioral defi cits seen in this animal model maybe attributed to impairments in synapt ic plasticity (Chapman et al, 1999). Sensorimotor analysis of the Tg2576 model has yielded conflicting results. One study reported deficiencies in the balance b eam and string tests at 3-, 14-, and 19-month old animals (King & Arendash, 2002). Anot her study also found impairments in balance beam at 5 and 6.5 months of age (Arendash et al, 2004). In cont radiction to these findings, a study by Lalonde et al (2003) re ported that between 15-20 months of age (average age 17 months), the Tg2576 model did not display any deficits on the balance beam and rotorod test. The discrepancy may be due to differences in equipment used to
28 test the animals or, less likely ; it may be due to differences in the strain background of the mice used. The Tg2576 model has also been sh own to exhibit increased activity in the open field test at 17 months of age (Lalonde et al, 2003). Simila r findings regarding activity level have been reported. A study done reported that at 3 months of age there was a significant increase in activity in the open field task, and overall increased activity through 19 months of age (King & Arendash, 2 002). In disagreement to these findings of increased activity a study done on 5 mont h old mice reported no differences in activity/exploratory activity (Arendash et al 2004). Changes in anxiety have also been reported. A study done using Tg2576 animals at 17 months of ag e reported finding reduced anxiety in the elevated plus maze (Lalonde et al, 2003). A study done in younger mice, 5 months of age, reported no differen ces in anxiety in the elevated plus maze (Arendash et al, 2004). The seeming disc repancies found in the Tg2576 model may be attributed to a number of va riables including differences in background strain, alterations in procedures used during testing, and us e of animals at various ages. There have also been conflicting reports regarding the cognitive abilities of the Tg2576 model. The Y-maze task for spontaneous alternation has been utilized by several researchers in characterizing the Tg2576 m odel. One study reported that at 3 and 19 months of age Tg2576 mice had significantly re duced spontaneous alternations; however at 9 and 14 months of age there was a modest impairment only (King & Arendash, 2002). The authors suggest that the Y-maze task ma y be relatively insensitive to cognitive impairment associated with the transgene, possibly accounting for the differences in ages at which researchers report finding impairments. Another study tested spontaneous alternation at 3 and 10 months of age and re ported that only at 10 months of age was
29 there significantly less spontaneous altern ations in transgenics compared to nontransgenic controls (Hsiao et al, 199 6). A study done on 17 month old Tg2576 mice reported impaired spontaneous alternation (Lalo nde et al, 2003). In agreement with these results another study tested spontaneous altern ation at 5 and 8.5 months and reported that overall there was a significant impairment (A rendash et al, 2004). Regardless of the exact time point in which impairments ar e seen in the Y-maze, impairments are consistently found in the performance of the Tg2576 mouse model for this test of cognitive function. The circular platform (a spatial referen ce memory task) is also used to test cognitive function in rodent models. A st udy done in 7 month old Tg2576 mice revealed an impairment in circular platform Â“rev ersal learningÂ” (modified methodology in which animals must relearn a new escape location) (Pompl et al, 1999). A study done on 3, 9, 14, and 19 month old Tg2576 mice found no signifi cant decreases in performance in the standard circular platform task (7 days of testing omitting the Â“reversal learningÂ” phase) (Arendash, 2002). A more recent study using the standard protocol fo r circular platform testing also reported no impairment at 6 m onths of age (Arendash et al, 2004). The Morris water maze is another task th at is commonly used by researchers and is designed to test referen ce/spatial learning and memory. A study was done on 2-, 6-, 910, and 12-15 month Tg2576 mice investigating performance in the Morris water maze (Hsiao et al, 1996). The 12-15 month old mi ce were tested in a follow up study that retested a subset of the original 2 and 6 mont h old mice. The study revealed that at the 2 and 6 month time points there was not a signif icant difference between the transgenic and non-transgenic matched controls in the Morris water maze task. At the 9to 10 month
30 age, the escape latency of transgenic mi ce during acquisitional testing was significantly increased; during the probe trial, the tran sgenic animals had a significantly reduced number of center crossings (Hsiao et al, 1996). The 12-15 month old mice showed impaired latencies after the 5th trial block and on probe tria ls. Another study was done using the Morris water maze on 3, 9, 14 and 19 month old Tg2576 mice. Compared to non-transgenic controls, Tg2576 mice were unimp aired collectively and at separate time points in both the learning and memory rete ntion phase of testing (King & Arendash, 2002). A recent study was done on 5.5 month old Tg2576 mice and found impairment in escape latency (time taken to find the submerged escape platform) across all 9 days of testing, the earliest time point at which impa irment has been observed (Arendash et al, 2004). This study also reported an impairment in the memory retention trial (final probe trial day 10) and an overall impairment ove r all three probe tria ls. The apparent inconsistent Morris water maze results in the later two studies from the same laboratory (King & Arendash, 2002; Arendash et al., 2004) may be due to differences in the strain background. The colonyÂ’s crossbreeding over several years has apparently resulted in both their Tg2576 and APP/PS1 lines becoming behaviorally more sensitive to mutant APP expression and/or the process of A deposition at an earlier age. The Tg2576 model has also been evaluated in the radial arm water maze (RAWM), an excellent working memory task showing a high degree of sensitivity to A deposition in numerous AD transgenic line s (Arendash et al, 2001). At 6.5-7 months Tg2576 mice have been reported to show wo rking memory impairments as assessed through Trials 4 and 5 of the RAWM (Arendash et al, 2004). This study also revealed especially prominent impairments in trial 5 of the RAWM, designed to mimic
31 Â“registration-recallÂ” tesing of AD patients, providing added evidence for the working memory impairment seen. It is widely re ported that cognitive impa irments in the Tg2576 model precedes the appearance of A plaque formation at around 9-11 months, suggesting that the soluble form may be causi ng the early phase of cognitive impairment (King et al, 1999). Also preceding plaque formation is evidence of increased oxidative stress in the form of lipid peroxidation. A study done exam ining the urine, plasma, and brain tissues for the lipid peroxida tion by product 8,12-iso-iPF2 -VI, found that at 8 months of age all of the samples showed increased lipid peroxi dation (Pratico et al, 2001). Another study examining the oxidative stress markers NHE and HO-1 also found that the Tg2576 model exhibits oxidative damage associated with A accumulation (Smith et al, 1998). The primary pathological differen ce between the APP23 and Tg2576 mouse model is found in the vasculature. Th e APP23 model exhibits deposition of A in the cerebral vasculature (CAA) that is remarkab ly similar to that observed in human AD (Calhoun et al, 1999). The CAA found in the APP23 model is associated with local neuronal death, dysfunctional synapses, activ ation of microglia and microhemorrhage (Calhoun et al, 1999). The APP23 model can al so serve as a mouse model of cerebral amyloid angiopathy. The accumulation of A in the arterioles a nd capillaries leads to the death of vascular smooth muscle cells, an eurismal vasodilatation, and a weakening of the vessel wall that can lead to rupture and severe hemorrhage (Winkler et al, 2001). CAA is one of the prominent features of AD thus, making the APP23 animal model a valuable tool in the study of AD (Mueggler et al, 2004).
32 The APP23 mouse model displays impaired spatial learning abilities at 2 years of age in the Morris water maze, and hyperac tivity assessed through the open field task (Dumont et al, 2004). Learning and memory defici ts in the APP23 mode l, as examined in the Morris water maze, have been reported as early as 3-months of age (Van Dam et al, 2003). An additional study found an impairment in latency and swim distance beginning at 3 months of age and persisting through 18 mo nths (Kelly et al, 2003). A study looking at 16 month old female APP23 mice found an impairment of spatial learning in the Morris water maze (Lalonde et al, 2002). Studi es have also reported disturbed activity patterns similar to the disturba nces in circadiam rhythms seen in AD patients (Van Dam et al, 2003). APPsw + PS1 Models Given the knowledge that mutations in the presenilin 1 (PS1) and presenilin 2 (PS2) gene results in the familial form of AD, the creation of an animal model expressing these mutations was created. As discussed earlier the missense mutations in PS1 and PS2 alters APP processing resulting in the formation of the toxic A 1-42 (Borchelt et al, 1997). Single transgenic animals expressing the PS1 mutant transgene are not seen to produce significant amounts of A pathology, however, there is a drastic increase in the ratio of A 1-42/ A 1-40 levels (Duff et al, 1996). By crossing the PS1 (M146L) animal model with the Tg2576 mouse, a double transg enic animal (APP/PS1) is produced that exhibits earlier and more pronounced AD like pathology (Borchelt et al, 1997; Holcomb et al, 1998). A deposition in the APP/PS1 is seen as early as 12 weeks in the cortex and hippocampus, far earlier than what is seen in the Tg2576 model (Holcomb et al, 1998).
33 By 12 months of age there is a 20 fold increase in A deposition in the frontal cortex and a 40 fold increase in A deposition in the CA1 region of the hippocampus as compared with the Tg2576 model (Takeuchi et al, 2000). The diffuse A plaque burden (detected through immunostaining with 4G8) peaks at ar ound 1 year of age, however, the fibrillar form of A (detected through thioflavin s staining ) increases through 2.5 years of age, the latest time point tested (M atsuoka et al, 2001). The A plaque formation activates astrocytes and microglia. The numbers of activ ated astrocytes and microglia increase in parallel with A plaque loads (Matsuoko et al, 2001). This robust A deposition alongside activated astrocytes and microglia in the APP/PS1 model has resulted in the model becoming very popular currently amongs t researchers. Recently, it has been reported that the APP/PS1 model exhibits abnormal LTP at 3 months of age, and impairments in basal synaptic transmission (B ST) at 6 months of age (Trinchese et al, 2004). Both LTP and BST are underlying processes that occur during memory consolidation and abnormalities in their functi oning result in memory deficits. Another study also looking at LTP and synapse func tion used microarray and quantitative RTPCR to profile synaptic plasticity genes in the APP/PS1 model. These researchers found reduced mRNA expression of several genes necessary for LTP (Zif268, NR2B, GluR1, etc.) (Dickey et al, 2003). Neuronal morphology in the APP/PS1 model has also been studied using labeling techniques. It is repo rted that at 11 months of age, the APP/PS1 model displayed swollen bulbous dystrophic ne urites alongside si gnificantly reduced spine numbers and reduced total spine area (Moolman et al, 2004). A decrease in the density of cholinergic synapses in the front al cortex and a decrease in the size of cholinergic synapses in the frontal cortex a nd hippocampus has been reported at 8 months
34 of age (Wong et al, 1999). While the exact ca use of the synaptic dysfunction is not yet known, it is quite evident that the APP/PS1 model exhibits dy sfunctional synaptic function as evidenced by multiple studies. In aged (22 months) APP/PS1 mice, researchers have found a reduced level of glucose utilization in the hippocampus accompanied by a 35.8% dropout of neurons in th e CA1 region (Sadowski et al, 2004). The report of neuronal dropout is novel. Prio r to this study no findings of neuronal death in the APP/PS1 model had been re ported (Takeuchi et al, 2000). Behavioral testing of th e APP/PS1 model has been done including testing in sensorimotor, anxiety and cognitive tasks. Sensorimotor testing has revealed impairments in this mouse model. An impairment in sensorimotor function assessed by the balance beam task has been found to emerge as early as 5 months of age and persist through 16 months of age (Arendash et al, 2001 ). In contrast to this study a study reported that no sensorimotor function im pairment was found in 4.5-6 and 15-16.5 month old animals in the balance beam task (Jensen et al, 2005). The string agility test (a task measuring sensorimotor ability) revealed no differences at 5-7 months of age, and impairments were found at 15-17 months of ag e as compared to non-transgenic controls (Arendash et al, 2001). Anxiety and activity testing have also b een performed on this animal model. The Y-maze alternation task is also used to dete rmine an activity index through measuring the total number of arm entries. In this ta sk 3-3.5 month old APP/PS1 mice had increased activity as compared with singly transgenic or non-transgenic litterm ates (Holcomb et al, 1998). Another study found similar results re porting that in 5-7 and 15-17 month old APP/PS1 mice there was a signi ficant increase in arm entr ies (Arendash et al, 2001).
35 Open field testing (a task measuring activ ity and exploratory le vels), has revealed increased activity levels at 15-17 months of age (Arendash et al 2001). A study done on 8 and 22 month old APP/PS1 mice showed no signi ficant changes in activity levels tested in an open-field task (Sadowski et al, 2004). Open field testing has been shown to reveal decreased activity (measured through horizon tal locomotor movements) in 7 month old APP/PS1 mice (Liu et al, 2002). However, in contrast, a study has re ported that at 4.5-6 months of age APP/PS1 mice had increased open field activity (Jensen et al, 2005). Anxiety testing on the elevated plus maze ha s revealed increased anxiety in animals 1516.5 months of age (Jensen et al, 2005). Overal l, the majority of studies (Holcomb et al., 1998; Arendash et al., 2001; Jensen et al., 2005) report increased activity in the APP/PS1 model. This increase in activity begins to em erge as early as 3 months of age and is still apparent at 15-17 months of age. The studies that reported either no significant changes in activity (Sadowski et al., 2004) or decrease s in activity (Liu et al., 2002) both used photoactometers to assess movement. The use of a photoactometer is a more sensitive technique as it utili zes interruptions in photobeams cap able of sensing horizontal and vertical motion to assess movement; whereas the open field testing employed in the other studies mentioned above utili zed the number of line crossi ngs (horizontal movement) in the open field as the only meas ure of activity. Thus, the conf licting results may be due to differences in equipment used. Cognitive testing including tasks such as the Y-maze alternation task, Morris water maze, and radial arm water maze (RAWM) have shown conflicting results. The Y-maze task of general memory function, revealed an impairment in spontaneous alternations in APP/PS1 mice at 3-3.5 months (Holcomb et al, 1998). However, another
36 study has reported no differences in Y-maze perf ormance at 5-7 months of age (Arendash et al, 2001). A recent study reported finding no differences in spontaneous alternation in the Y-maze task at both 4.5-6 months and at 15-16.5 months (Jensen et al., 2005). As mentioned prior the incongruity of findi ngs in the Y-maze task for spontaneous alternations may be due to th e non-specificity of the task to the impairments conferred by the APP/PS1 genes. Spatial memory and learning assessed in the Morris water maze revealed no differences in escape latency, time spent in goal arm, or number of pl atform crossings in animals at 9 months of age (Holcomb et al, 1999). However, a l ongitudinal study testing at 4.5-6 months and again at 15-16.5 months found a significant impairment in escape latencies at both time points (Jensen et al 2005). The Jensen et al., 2005 study also reported that in the probe trial at 4.5-6 m onths there was s sign ificant reduction in annulus crossings as compared with the non-transgenic group; and at 15-16.5 months there was no difference in performance. Work ing memory has been tested in the radial arm water maze task in the APP/PS1 model at several different ages. A longitudinal study using 5-7 month old APP/PS1 found no differences in the RAWM and at 15-17 months reported that the APP/PS1 animals ma de significantly more errors during trial 5, the memory retention trial in the RAWM (A rendash et al., 2001). An impairment in working memory during radial arm water maze (RAWM) testing has been found in the APP/PS1 model at 16 months of age, as meas ured by the number of errors animals made during trial 5 of the task (Austin et al., 2003). Two recent studies have reported impairments in working memory at a much earlier age. A study done on 6-8 month old APP/PS1 mice reported finding significantl y more working memory errors in the
37 APP/PS1 animals vs. the non-transgenic controls in the RAWM (Trinchese et al, 2004). An even more recent study reported that A PP/PS1 animals at 4.5-6 months of age made significantly more working memory errors in the RAWM task, revealing cognitive deficits at a much earlier age than previously reported (Jensen et al 2005). Interestingly, the same transgenic colony was utilized in bot h Arendash et al., (2001) and Jensen et al. (2005) for RAWM testing. The much earlier impairments in working memory found by Jensen et al., 2005 may be attributable to the fact that the colony utilized during the test battery had been cross-breed over several gene rations and the resultant transgenic mice may have become more behaviorally sensit ive to mutant APP expression and/or the process of A deposition. Using the most recent and thorough behavi oral evaluation of this model by Jensen et al (2005), an overview of the APP/PS1 model can be formed. The APP/PS1 model was reported by Jensen et al, ( 2005) to be cognitively impaired at 4.5-6 months of age in the RAWM task (assessed thr ough number of errors, impairme nt seen overall and in last block), showing impairments in working memo ry. From this data it is reasonable to assume that any treatment/study that aimed at utilizing animals that are trending towards impairment should begin around 3 months of age. Jensen et al., ( 2005) reported that at 4.5-6 months of age this model had a significant impairment in Morris water maze acquisition and retenti on. These findings show an impairment in working and spatial reference memory in the APP/PS1 model at th is early time point. Jensen et al, (2005) reported that at the late 15-16.5 month time point an additional cognitive impairment was found in the platform recognition task. Thus, in the APP/PS1 model, one could expect to find a significant cognitive impairment in wo rking, reference and long-term memory to
38 emerge at 4.5-6 months of age (tested by e ither the Morris water maze or RAWM) and expect the impairment to pe rsist through 15-16.5 months. Transgenic animal models have provided researchers with a powerful tool in the study of AD. Animal models are, however, an incomplete mode l of AD due to their lacking of the complete set of pathologies seen in AD. All of the mutant transgenic APP lines (PDAPP, APP23, Tg2576, APP/PS1) comple tely lack the formation of NFTÂ’s (Loeffler, 2004). As discussed prior, NFTÂ’s ar e one of the pathologica l hallmarks of AD. It is, therefore, desirable that an animal model would form NFTÂ’s. The APP mouse model also differs from AD in regards to th eir immune and inflammatory response. The APP transgenic mouse displays only weakly act ivated microglia expressing low levels of complement factors that were gathered onl y around the periphery of plaques (Schwab, 2004). This is in stark comparison with the human response which involves microglia found within the core of the plaques that are highly activated and e xpress high levels of complement factors (Schwab, 2004). The r eason for this diminished immune and inflammatory response in unknown, although it has been hypothesized that it is due to low recognition of mouse comp lement factors to human A (Webster, 1997). The plaques themselves are more fibrillar in state, with fibrils radiating out from a dense core, whereas human AD exhibits plaques in a more homogenous state (Schwab, 2004). The A deposits in the APP23 model are soluble in SDS-containing buffer, however, human AD deposits are insoluble (Schwab, 2004). These differences may be due to posttranslational modification differe nces in the mouse model. The more fibrillar and toxic nature of the animal model plaques has not b een definitively shown to result in increased neuronal death. One study reported findi ng neuronal loss in a limited region around A
39 plaques in the Tg2576 model (Tomidokoro, 2001) A study using 16 month old Tg2576 mice found no neuronal loss (Stein, 2002). The debate is centered on whether there is Â“limitedÂ” neuronal loss; however, it is appare nt that the animal model does not display the profound neuronal losses seen in AD. Th e mutant APP transgenic animal model is therefore only a limited model for AD. Th e lack of a robust immune and inflammatory response may suggest that this animal m odel be limited to studies involving the prevention of A accumulation (Schwab, 2004). There are also limitations in the mouse model regarding behavior. Ideally, an animal model for a disease mimics that found in humans, but due to the highly psychologically-based nature of AD it is impossi ble to create a complete mouse model. One of the first striking deficits in using a mouse model is the inab ility to test verbal skills. Also, it is very difficult to test an animal model for any of the psychoses experiences by AD patients such as hallucina tions, depression, and apathy. While there is extensive memory testing in animal models, it is impossible to assess semantic memory. Most behavioral studies use mazes with varying escapes and measures the ability of an animal to lear n the escape, thus giving the researcher insight into the animalsÂ’ short and long-term spatial memory skills. Semantic memory, ability to recognize people, places, etc. is difficult to m easure in an animal model. Often an AD patient will forget a loved onesÂ’ face or be unable to recogniz e voices of their loved ones. This type of deficit in semantic memory is readily apparent in AD. However, animal testing is limited to testing only object recognition in memory tasks or identif ication of an escape platform. Nonetheless, animal mode ls have proven to be highly useful in AD
40 research. It is however, important to understa nd the limitations that are faced when using animal models in AD research. Hyperoxia Behavior It is estimated that the human brain consumes approximately 20-30% of total energy usage for the entire body and is consider ed to be the most metabolically active organ (Roland, 1993). Oxygen is regularly consumed by brain cells during normal cellular respiration (Benton et al, 1996). Given this know ledge it is reason able to assume that varying oxygen saturation levels (hyperoxia or hypoxia) would affect cognitive functioning as well as other metabolic paramete rs. Researchers have been investigating the effects of administering hyperoxia (oxyge n level above 20.9 %) treatments in humans for several years. A recent study done by Andersson et al (2002) found no differences in cognitive performance after subjects ( 48 participants with a mean age of 21.1 6.49 years) inhaled 100% oxygen for 1 minute prior to the start of testing. The cognitive testing given during this study was designed to assess work ing memory, prospective memory, attention and long-term memory. This study also exam ined physical changes that occurred during the treatment. They found that individuals who underwent the hype roxia treatment had significantly higher oxygen blood saturati on, lasting for 1 minute post treatment (Andersson et al, 2002). Heart rate was not affected by the treatment. This study is in contradiction to several others (Moss et al, 1998; Scholey et al 1998;1999), who have found significant cognitive improvements following treatment with hyperoxia.
41 An older study done by Moss et al (1998) reported a significant improvement in immediate and delayed word recall in test su bjects (20 participants aged 21-48 years) exposed to 100% oxygen for a duration of 1 mi nute or 3 minutes pr ior to the start of testing. This study used 4 different sc hedules of hyperoxia treatment; 30 seconds, 1 minute, 3 minutes, and constant oxygen admini stration throughout the test session. The constant oxygen schedule did not show impr ovements in the cognitive tasks. The researchers speculated that there is a point in which an enhancing factor becomes deleterious. Working memory did not show any enhancement in any of the hyperoxia treatment groups. The researchers suggest that the placement of the working memory task at the end of the test battery may acc ount for their finding. Overall, the authors reported that transient administration (1 and 3 minute exoposures) of 100% oxygen improved attention, vigilance and long term memory. They hypothesized that the transient increase in oxygen saturation may re sult in a global upgrade in metabolites that enhance cognition. Two studies done by Scholey et al (1998 & 1999) also reported cognitive improvements after exposure to hyperoxia. The study done in 1998 (20 participants with a mean age of 21 years) found enhanced wo rd recall in test s ubjects exposed to 100% oxygen 5 minutes prior to, immediately before or immediately after word presentation. This study used multiple schedules of treatment given 5 or 10 minutes prior to word presentation, or 5 or 10 minutes after word presentation. The study s howed that subjects given oxygen 10 minutes before or 10 minutes after word presentation did not show any improvements. The cognitive testing was limited to word recall. The researchers examined oxygen saturation and reported that the effects of the hyperoxia treatment on
42 saturation were apparent immediately before during, or after word presentation where word recall was enhanced. This supports their hypothe sis that the additional blood oxygen resulting from the hyperoxia treatments is utilized through the neural mechanisms responsible for memory formation. The st udy done in 1999 by Sholey et al. had similar findings. Subjects (34 participants with a mean age of 21 years) were exposed to hyperoxia for 1 minute and then tested for react ion time and word recall. The subjects who received the hyperoxia treatments recalled mo re words and had faster reaction times. Alongside cognitive testing, oxygen saturation was also measured. A significant increase in blood oxygen saturation was seen in the subjects exposed to the hyperoxia that persisted from the gas administration phase th rough the word presentation phase of the experiment. This finding, in conjunction with the improvements seen in word recall, led the authors to conclude that the elevated bl ood oxygen is utilized by task-sensitive neural substrates during period of cognitive processing. In all of the studies listed above it is important to note that they utilized an acu te administration of oxygen. The effects of longer-term oxygen administration (for a period of several hours rath er than minutes) on cognitive performance could, therefore, have dr astically different resu lts. It is also important to note that hyperoxia has not b een found to have deleterious effects on the cognitive function of healt hy individuals. A study done by Prior & Chander (1982) found that hyperoxia exposure ( 12 hour treatment) post-surger y to elderly patients did not show any deleterious cognitive effects. It is important to note that cognitive testing was performed shortly after hyperoxia exposure (not months thereafter). There has only been limited research done on the behavioral effects of hyperoxia in rodent models. A study done by Fukui et al (2001) examined learning and memory in
43 a rat model for oxidative stress. The study i nvolved three different ages of rats; 3 months, 15 months, and 25 months. An additio nal group of animals were fed a vitamin-E deficient diet for 9 weeks prior to testing. The animals were given time to learn the Morris water-maze task prior to exposure to hyperoxia. After the animals had adequately learned the task they were exposed to 100% oxygen for a lengthy 48 hours prior to the memory retention phase of Morris water-maze testing. Performing retention trials daily for up to 14 days after hyperoxia exposure, th ey claim that not all of the animals could relearn the task suggesting that the damage done during the treatment was long lasting. The methods used to conclude that the animal s could not relearn the task were somewhat confusing. By performing repeated probe tria ls the researchers were actually assessing Â“memory-extinctionÂ” (long-term memory) rather than the an imalsÂ’ ability to relearn. Also the study lacked a full explanation of the statistical measures used in their behavioral analysis. However incomplete this study may be, these findings do point towards the toxic nature of long-term exposure to hyperoxia in a rodent model Pathological and Physiologica l Effects of Hyperoxia Hyperoxia is commonly used in hospitals to treat asthma, chronic obstructive pulmonary disease (COPD), premature babies patients with severe head injury, etc (Pagano & Barazzone-Argiroffo 2003). The use of hyperoxia in severe head injury has been controversial. A study done by Tolias, et al (2004) revealed benefits of using hyperoxia in patients with head injuries. Using patients with se vere head injuries exposed to 100% oxygen for 24 hours the resear chers examined five cerebral metabolic markers. They found increased glucose leve ls, decreased glutamat e and lactate levels, and reduced intracranial pressure in the gr oup exposed to hyperoxia. Another study done
44 by Magnoni, et al (2003) also found decreases in lactate. However, they reported no differences in glucose and glutamate levels. A decrease in lactate has also been reported by Schaffranietz, et al (2000) in patients expos ed to hyperoxia. These differences may be attributed to the fact that the patients used in the Magnoni, et al ( 2003) study only received 3 hours of hyperoxia. Of course, treatments with hyper oxia also increase a patientsÂ’ oxygen saturation making it benefici al to anyone suffering from a disease in which low blood oxygen saturation is a sy mptom (Kergoat & Faucher, 1999). While hyperoxia does have some limited be nefits there is growing evidence that the treatment has many harmful side effects re sulting in permanent damage. One of the most profound and widely documented effects of hyperoxia is found in the vasculature. Hyperoxia is a potent cerebral vasoconstric tor (Ouattara et al, 2004). A study done in humans using MRI to analyze the influence of hyperoxia on cerebral blood flow reported that during hyperoxia, patients had diminished regional cerebral blood flow in all regions except in the parietal and left frontal gray matter (Kolbits ch et al, 2002). The exact mechanisms underlying oxygen induced vasocons triction are still unde r investigation. The enzyme super-oxide dismutase (SOD) alongs ide the signaling molecule nitric oxide (NO) is thought to regulate vasoreactivity. SOD catalyzes the dismut ation of superoxide radical (O2 -) to hydrogen peroxide. Approximately 5% of inspired oxygen is converted to the dangerous superoxide radi cal (Fridovich, 1983). The SOD class of enzymes plays an important role in antioxidant defense m echanisms. SOD enzymes are implicated in diseases involving oxidative stress such as AD. NO is a diffusible gas that is found in neurons, macrophages and epithelial tissues. In epithelial tissue it functions to relax smooth muscle in the arteriole walls resul ting in vasodilation. A study by Johnson (2001)
45 has shown that O2 reacts with NO to form the toxic peroxynitrite (ONOO-). Peroxynitrite is a potent oxidant that can react with DNA, pr oteins, and lipids leading to cellular damage. Johnson (2001) also sugge sted that increased levels of brain ONOOcan further worsen the damage caused by the overactive microglia, fu rther advancing the progression of AD. A study done by Demchenko, et al (2002) investig ated the role of SOD in oxygen-induced cerebral vasoconstr iction. The study revealed that SOD promotes cerebral vasodilation by scavenging free O2 in a normal state system. During hyperoxia, they found that the effects of br ain NO were decreased due to the increased O2 production, resulting in vasoconstricti on. Thus, SOD works by regulating the availability of O2 to act as a vasoconstrictor by degr ading the radical, allowing for NO induced vasodilation. In a study done by Park, et al (2005) th e effects of NADPH oxidase-derived ROS and A were investigated in re gards to cerebrovascular dysfunction. The authors also came to the co nclusion that the effects of ROS mediating cerebrovascular dysfunction invo lve reduced bioavailability of the vasodilator nitric oxide. A study done by Sjoberg, et al (1999) quantified the vasocons triction seen in the CNS during hyperoxia in a pig model for hyper oxaemia (condition in which the arterial partial pressure of oxygen exceeds 80 mm Hg). They reported that after 25 minutes of administration of 70% oxygen they found an 11% reduction in cap illary blood flow, alongside a significant increase in cerebrocor tical tissue oxygenation. This decrease of blood flow may be a protective mechanism. By reducing the brains exposure to the toxic high levels of oxygen that are generated duri ng hyperoxia, the brain may be reducing the possible damage incurred.
46 As mentioned prior, the AD brain contains elevated levels of A The increased brain A levels seen in the APPsw mouse model may play a contributory role towards their increased susceptibility to brain isch emic injury (Xu et al, 1998). Alongside these findings it has also been reported in a st udy by Iadelcola, et al. (1999) that APP717 transgenic mice have an alteration in cerebrova scular regulation; spec ifically, there is a reduced vasodilatory response to acet ycholine and an enhanced response to vasoconstrictors. This suggests that th e impairment seen in cerebrovascular responsiveness results in vasculature more prone to vaso constriction. Several additional studies have been done further investigating the effects of A on endothelial dysfunction. An in vitro study was done by Thomas, et al (1997) revealed that A increased cerebral vasoconstriction, and decreased vasodilation. The authors also reported that the A induced endothelial cell damage, apparently caused by reactive oxygen radicals produced by A Also, the authors suggest that the vascular damage done by A may be an early event in AD. Another in vitro study done by Crawford, et al (1998) further supports the finding that A enhances vasoconstriction in rat aortae Specifically, the authors looked at the effects of A 1-40 & A 1-42 on vasoconstriction. They found that A 1-40 had a more profound vasoconstrictiv e effect than A 1-42 Interestingly, the authors also reported that the endothelium is not required for vasoactivity. A -induced vasoactivity is seen immediately after expos ure to solubilized A and the lack of requirement of the endothelium suggest that the imabalance between NO and O2 works alongside A to further increase vasoconstriction. Thes e authors also suggested that chronic vasoconstriction would result in subclinical ischemia that would in turn stimulate increased A formation around the vasculature. In vivo studies have also been done
47 providing additional evidence that A serves as a vasoconstrictor. A study done by Arendash, et al (1999) reveal ed that spontaneously hypoten sive rats infused with A 1-40 experienced substantial increases in mean arterial blood pres sure (MAP). Hypertension has been recognized as a risk factor fo r AD (Kokmen, et al, 1991), and longitudinal studies have shown that elevat ed blood pressure is associat ed with development of the disease 10-15 years later (Skoog, et al, 1996). The authors suggest that the disease process itself may induce the onset of hyperten sion during the 10-15 years before clinical onset. Also the author suggests that the A -induced vasoconstriction demonstrated in the rat model may play a contribu tory role to the AD process in humans. Another study done by Suo, et al. (1998) reporte d that rats infused with A 1-40 had decreased cerebral blood flow and increased cerebro-vascula r resistance. This study found that A specifically affected cerebral vasculature, s uggesting that the cereb ral hypoperfusion that is observed in early AD may be the result of A induced vasoconstriction. Consistent with the findings in murine models, human st udies using PET have al so revealed that in the AD brain there is reduced vascular activity to vasodilatory stimuli (Mentis et al, 1996; Warkentin & Passant, 1997). The pathological effects of hyperoxia in rodent models have been much more extensively studied. A study done by Urano, et al (1997) using a rat model investigated the morphological changes through electron mi croscopy in the brai n associated with exposure to 100% oxygen for 48 hours followed by immediate sacrifice. They found swollen astrocytes around vessels, deformed nerve cell nuclei, swo llen mitochondria, and abnormal accumulation of synaptic vesicles in swollen nerve terminals. Also, they reported changes in the plasma membrane including decreased membrane fluidity,
48 increases in membrane permeability to sucrose, and an increased cholesterol/phospholipids ratio of the membrane. The authors suggest that the increased amount of free radicals generated may damage nerve terminals and peroxidize the plasma membrane. Several studies have been done examin ing the cellular and biochemical effects that hyperoxia has on the brain. The effect s of hyperoxia on inducible nitric oxide synthase (iNOS) expression ha s been studied in rat pups by Hoehn, et al (2003). Seven day old rat pups were exposed to >80% oxyge n for 24 hours, immediately sacrificed, and then the amount and distribu tion of iNOS was examined. Biologically, iNOS is an enzyme responsible for synthesizing NO (a vasodilator). The study revealed that animals exposed to hyperoxia ha d increased total brain iNOS levels. The increased iNOS lead to the formation of peroxynitrite (NO+ O2 ONOO-), a toxic molecule that can cause oxidative cellular damage. The re searchers concluded that the increase in iNOS and resultant increase in peroxynitrite may lead to subsequent damage to brain structures. Another study done by Mamdouha, et al (1986) also look ed into the effects of hyperoxia on brain structures and found neuronal necrosis. The study utilized a series of different schedules of alternat ing hypoxia/hyperoxia treatments (100% O2 for 3 hours) in both young and adult rats th at were immediately sacrifi ced following treatment. The study revealed extensive neuronal karyorrhex is in the subiculum, cingulated cortex, thalamus, and reticular formation in newbor n rats exposed to hyperoxia. The more mature and the adult rats did not develop a ny neuronal karyorrhexis demonstrating their ability to cope with the treat ment better. The researchers hypothesized that the selective vulernability of the immature brain was due to lack of full development of the antioxidant
49 defense mechanisms seen in the adult animal. Also, the researchers hypothesized that the neuronal necrosis seen in the newborn ra ts may be the result of lipid peroxidation of the cell membranes due to the antioxidant defense mechanisms becoming overwhelmed. An additional study done by Taglialatela, et al (1998) was done examining effects of antioxidants in rats exposed to hyperoxia. The study used two groups of newborn rats; one control group and one group given Buthi onine sulfoximine (BSO) a glutathione synthesis inhibitor. Glutathi one is a powerful antioxidant and can provide protection for the mitochondria against oxygen radicals. The animals were exposed to 95% oxygen for 5 days, sacrificed immediately, and measured for nerve growth f actor protein (NGF), glutathione, and the extent of apoptosis. The study revealed that hyperoxia decreased the amount of NGF protein, and that the animal s treated with both the BSO and hyperoxia had a substantial increase in brain apoptosis. The authors suggest that the oxidative stress of hyperoxia in conjunction with limited glutat hione resulted in neuronal damage. Thus, several studies have revealed th at in a brain that is limited in its antioxidant capabilities hyperoxia can induce neuronal damage. It is important to note that the above studies (Hoehn et al, 2003; Mamdouha et al, 1986; Taglia latela et al, 1998) sacrificed the animals immediately following the hyperoxia treatment li miting the scope of effects to only the immediate and short-term effects. Numerous studies examining the effect s of hyperoxia in lung tissue have also been conducted. Hyperoxia has been widely documented to cau se extensive alveolar cell death through mechanisms still under debate (Pagano & Barazzone-Argiroffo, 2003). A study done by Buccellato, et al (2004) exam ined the role of reactive oxygen species (ROS) such as O2 in the induction of cell death. Th e researchers exposed rat epithelial
50 cells to hyperoxia and found th at the treatment resulted in activation of Bax in the mitochondrial membrane with subsequent cyto chrome c release and cell death. Bax is a pro-apoptotic protein that is involved in re gulating programmed cell death (De Smet et al, 2004). The researchers hypothesized that Bax activation was dependent on the generation of ROS. Hyperoxia has been show n to generate intrac ellular production of O2 and H2O2 by the mitochondria (Freeman & Crapo, 1981). Another study done by Pagano, et al (2004) focused on the role of cytochro me c (a cell death promoting factor) in the alveolar cell death seen during hyperoxia in a rat mode l. Hyperoxia induces high amounts of cytochrome c to be released from the mitochondria into the surrounding cytosol. Their study reveal ed that by blocking the m itochondria from releasing cytochrome c through administration of cycl osporine A, lung tissue could be protected from damage during hyperoxia treatments. Wh ile the exact biological events are not yet fully understood it is apparent that mitochondr ia play an integral role in hyperoxia induced cell death. It is witnessed by seve ral studies that hyperoxia has an exacerbating effect on antioxidant defense systems. The toxic nature of excess free radicals can result in a multitude of damaging effects such as apoptosis, mitochondrial dysfunction, increases in harmful enzymes, and structural changes. Linking Hyperoxia and Precipitation of AD A study done by Shua-Haim, et al (1998) suggested a link between surgery in the elderly and precipitation of AD. The study repo rted an acute onset of AD post-surgery. The acute onset is in constradiction to th e definition by the NINCDS-ADRDA that states the onset of AD is gradual. This suggests that the patients may have been suffering from a Â“pre-clinicalÂ” form of AD prior to the surgery. The surgery including all of the
51 treatments surrounding surgery resulted in a rapid neurological degeneration and subsequent diagnosis of AD. This study suggests that some aspect of surgery or postsurgical care may act as a risk factor in aged individuals. Along this line, a study was done by Newma n, et al (1995) inve stigating a number of factors surrounding surgery and their predic tive value for cognitive decline. The study revealed that the strongest predictor of cogn itive decline post surgery is age. The mean arterial pressure (MAP) was also examined, and the study revealed th at in older patients MAP may play a role in post-surgery cogniti ve decline. Also they looked at the apolipoprotein E4 allele, the gene variant associated with the sporadic form of AD, as a genetic predictor. They found a significant association between the presence of the apolipoprotein E4 allele in patients who had declines in cognitive performance postsurgery. This evidence further point towards the exacerbating effects that surgery has on patients who may be at a highe r risk for developing AD. Surgery involves numerous components including anesthesia, administration of numerous drugs, changes in blood pressure and possible exposure to hyperoxia postsurgery. Studies trying to tease out one vari able as being the causative agent for the cognitive deterioration are di fficult as the typical surgery incorporates many complicating factors. Several studies have attempted to isolate the effects of anesthesia. A recent retrospective hospital-based case-control study done by Gasparini, et al (2002) reported finding no association between exposure to anesthesia and development of AD. The study examined 115 AD patientsÂ’ hospital reco rds for exposure to anesthesia and found that the exposure to anesthesia 1 and 5 years prior to disease onset was not correlated to future development of AD. An olde r study done by Bohnen, et al (1994) further
52 supports their findings. In a retrospectiv e, population based, cas e-control study they looked at the exposure of 208 AD patients and f ound that it is unlikely that even multiple exposures to anesthesia increas es the risk of developing AD. However, the same authors also reported that the age at which an indivi dual was exposed to and extent of anesthesia given may play a more important role. The study investigated the cumulative exposure to anesthesia for more than 40 years prior to onset of AD, and fo und that the age of AD onset was inversely related to the cumulative exposure to anesthesia before the age of 50. The authors suggest that the ons et of AD may be related to exposure to anesthesia at a relatively earlier age. Blood pressure has also been investigat ed with conflicting findings. At study done by Moller, et al (1998) reported that hypotension (abnormally low blood pressure) was not a significant risk factor in predicting cognitive dysfunction post-surgery. However, recent work by Dr. T. Monk (per sonal communication) has suggested that hypotension can be extremely dangerous to a patient during surgery. Dr. Monk conducted a study in which he found that for ev ery minute that a patientsÂ’ systolic blood pressure dropped below 80 mmHg there was a 4% increase in the chance of death within one year post-surgery. These findings are very alarming, as they point towards the extremely dangerous nature of hypotension dur ing surgery. It also points towards the irreversible effects that hypot ension exerts on the body. Hypot ension can result in a state of hypoperfusion in the brain. The steep drop in blood pressure results in a decrease of oxygenated blood being available to adequately sustain the metabolic demands of the brain. Patients that exhibit hypotension dur ing surgery are given treatments with hyperoxia post-surgery in order to boost a patientsÂ’ blood oxygen saturation (J. Robert,
53 personal communication). He hypothesizes that hypotension a lters the bodies inflammatory response. Treatment with antiinflammatory agents post surgery has been found to reduce the number of deaths seen 1 year post surgery. Hyperoxia treatments are frequently given to patients post-surgery to increase the oxygen content of blood, but it also acts as a potent cerebrovasocons trictor resu lting in decreased cortical blood flow (Rostrup et al, 95 ). As a vasoconstrictor, it is plausible that treatments with hyperoxia could also result in a hypoperfusion of the brain. Hyperoxia exposure alone has not been shown to have deleterious cognitive effects as discussed in several studies above. However, the studies conducted l ooking at its effects were done on either young individuals or in dividuals not proven to be at risk for developing AD. The evidence presented a bove linking surgery a nd precipitation of AD alongside the knowledge th at hyperoxia has the potential to cause a variety of toxic side effects has provoked a study to be conducted investigating hyperoxia as a potential causative agent of cognitive impairment in aged individuals post-surgery, who have subclinical AD. Specific Aims Hyperoxia has been proven to generate free radicals (O2 -) thereby inducing oxidative stress, confer struct ural damage in nerves, upregulate enzymes responsible for cellular death, peroxidize membranes, a nd decrease cerebral blood flow. These deleterious effects have been discovered using both human and animal models that had compromised/immature antioxidant defense sy stems such as those seen in the studies using infant models or in the human studies using patients with trau matic brain injuries.
54 Oxidative stress is one of the primary pa thological mechanisms underlying AD. As mentioned earlier, oxidative stress plays a critical role in the early stages of the disease; perhaps, paving the way for more severe neuro-degeneration (Pratico & Sung, 2004). An increased level of A (both soluble and insoluble) is also one of the primary pathological mechanisms underlying AD. A has been shown in the numerous studies mentioned above to act as a vasoconstrictor a nd as a generator of free radicals. Both A and hyperoxia are capable of generating supe roxide radicals, causing cerebrovascular constriction, thus decreasing cerebral blood flow (brain hypo perfusion). I propose that hyperoxia treatments given to cognitivelnor mal APP/PS1 transgenic mice will work in combination with their brain A to markedly increase product ion of superoxide radicals, causing oxidative stress damage to brain tissue and result in cerebrovascular constriction with subsequent cerebral hypoperfusion. I pr opose that the damage done by the increase in oxidative stress and cerebral hypoperfusion will result in cognitive impairments. In this study I specifically propose to expos e AlzheimerÂ’s transgenic mice to several hyperoxia treatments and to: 1) Assess the cognitive function of the mice before and after hyperoxia treatment in the Radial Arm Water Maze task of working memory 2) Measure the amount of oxidat ive stress incurred through examination of markers of lip id peroxidation in specific brain regions 3) Quantify brain A loads, both diffuse and compact
55 4) Perform neurodegenerative cel l staining in specific hippocampal regions 5) Perform correlation analyses examining any possible relationships between pathological, neurochemical, and behavioral measures. Materials and Methods Animals A total of 20 mice were utilized during the course of this study. All mice contained a mixed background of 56.25% C57, 12.5% B6, 18.75%SJL, and 25.5% Swiss Webster. The animals genotypically fall into one of three categor ies; nontransgenic, double transgenic (containing both the APP/ PS1 mutations), and single transgenic animals (containing the APP mutation only). All of the mice were bred from a cross between a P (parental generation) hetero zygous male mouse carrying the mutant APPK670N, M671L gene with a F1 PS1 (transgenic line 6.2) female mouse to obtain the F2 generation including APP/PS1, APP, PS1, and non-transgenic mice. After weaning, the mice were genotyped and singly housed in cages with rodent chow and water ad libum Mice were maintained in a 10 hour dark 14 hour light cycle, with al l behavioral testing done during the light cycle. After the init ial round of pre-trea tment RAWM testing, transgenic animals were broken into two behaviorally-balance d groups; the hyperoxia group (O2) and the control group (air). As show n in Table 1, the group of transgenic mice that were exposed to hyperoxia c onsisted of 3 APP/PS1 mice and 2 single
56 transgenic APP mice. Pre-treatment perf ormance of the 3 double transgenics and the 2 single transgenic mice was found to be nearly identical, and thus clearly not significant different (p= 0.7 for overall T4 and T5). The control group of animals that were exposed to compressed air consisted of 6 APP/PS1 mice, whose behavior was statistically similar to the hyperoxia group (determined by analysis of pre-treatment behavior). A total of 8 non-transgenic mice were used as behavioral controls and were not exposed to either hyperoxia or air treatment. For the bioche mical portion of the experiment, the same 8 non-transgenic mice were used as controls (T able 1). Also the same group of 6 APP/PS1 control (air exposed) mice were used as the control group in the bi ochemical portion of the experiment. The hyperoxia group of an imals consisted of 4 APP/PS1 mice. The addition of the extra mouse as compared w ith the group of 3 APP/ PS1 mice used in the behavioral portion is explained due to the fact the fourth mouse was deleted from the behavioral portion of the study. The animal was deleted due to the fact that it was statistically determined to be an outlie r after completing the mid-point portion of behavioral testing. The 2 APP animals expos ed to hyperoxia were not included in the biochemical portion of the study due to genotype differences. Genotype & Treatment Be havior Biochemical Non-Transgenics 8 8 APP/PS1 Controls (Air) 6 6 APP/PS1 O2 3 4 APP O2 2 0 Table 1. Animals included in the behavi or and biochemical portions of the study.
57 General Protocol At 3 months of age, 9 APPsw+PS1, 2 A PP transgenic mice and 8 non-transgenic littermates were pre-tested in the radial arm water maze (RAWM) task of working memory for 15 days. The transgenic (Tg+ ) mice were then divided into two groups balanced in cognitive performance from the RAWM pre-test. One group of Tg+ mice were exposed to 100% oxygen for 3 hours. The remaining Tg+ mice received the same treatment, except that normal air flowed th rough the gas chamber. The non-transgenic mice were not exposed to any treatment. Be ginning three days following gas treatment, all animals were re-tested in the RAWM task for 15 days, with the two groups of Tg+ mice showing no differences in performance. At 5 and 7 months of age, all animals received a second and third gas treatment, re spectively. Final RAWM testing for 9 days was begun three days after the 3rd gas treatment. At the completion of testing, all mice were euthanized and their br ains processed for: 1) A staining with 6E10 antibody and Thioflavin S, 2) degenerative neuronal numbers (acid fuchsin/toulidine blue staining), 3) lipid peroxidation markers (Iso-Furans and 8-IsoProstane), and 4) hippocampal APP, APOE, COX-2, and GFAP levels. Fig.1. General protocol time line for the hyperoxia study. 0 1 3 2 4567 8 9Animals Born 1st RAWM Testing 1st gas treatment 2nd RAWM Testing 2nd gas treatment 3rd gas treatment 3rd RAWM Testing Animals Sacrificed
58 Hyperoxia All animals were food-deprived for 24 hours prior to O2 gas insult. For gas treatments, a multi-chambered apparatus was co nstructed of plexiglass with input valves allowing for the flow of 100 % oxygen/compresse d air into the chamber. For all gas insults, airflow per minute was regulated by a flow meter and maintained at 1.5x chamber volume. Treatments were conducted under isobar ic conditions. Bara-Lyme was used in each chamber to control for excess CO2. Each gas treatment was administered for 3 hours. The control gas treatments were c onducted in the same manner as the hyperoxia with the exception that compressed air (20% O2) was utilized. All animals received a total of three treatments, administered at a pproximately 6 week inte rvals (Fig. 1). The first treatment was given when the animals we re 4 months of age. The second treatment was given at 5.5 months of age and the last tr eatment was given at 7 months of age. Behavioral Assessment Spatial working (short-term) memory was assessed in the radial arm water maze task (RAWM). The task is conducted in a 100-cm diameter inflatable pool with an aluminum insert. The aluminum insert crea tes 6 radial swim arms (30.5 cm length x 19 cm width), with each swim arm radially distri buted in the pool from a central circular swim area 40 cm in diameter. The insert exte nded 5 cm above the surface of the water. The water was maintained at 23-27 Celsius. A transparent escape platform (9 cm diameter) was placed in one of the arms submerged 1.5 cm below the surface of the water. Around the perimeter of the pool and on an adjacent wall visual/spatial cues were placed. The visual cues consiste d of large brightly colored objects distinct in shape. For example, the visual cue located at the end of arm 2 was a suspended beach ball, and the
59 visual cue at the end of arm 6 was a small clos ed umbrella. The visual cues were used by animals to orient themselves within the maze and facilitate finding the escape platform. On each day of testing, the animals were given fi ve 1 min. trials. Each trial lasted for 1 minute, with a 30 second delay between tria ls 1-4 (during which the mouse remains on the platform), and a 30 minute delay between tr ials 4 and 5. Trials 1-4 are acquisition trials in which the animals are learning the loca tion of the escape platform for that day. The last of four consecutive acquisition trials (trial 4, T4) and the delayed retention trial (trial 5, T5) are indices of working memor y. On any given day, the escape platform location is placed at the end of one of th e six swim arms. The platform location was moved daily to a different arm in a se mi-random fashion. By moving the escape platform, the animal must learn a new locati on of the platform daily and rely on working memory rather than long-term memory. On each day, different start arms for each of the five daily trials were also selected in a se mi-random fashion that incorporated all five arms. At the beginning of any given trial, the mouse was placed into the start arm facing the center of the pool and given 60 sec. to fi nd the platform, with a 30 sec stay. For each trial, the latency (amount of time in sec) and number of errors to find the submerged platform in the goal arm were recorded. An error was recorded when an animal swam into an arm that did not contai n the escape platform. Each time an animal made an error, the researcher gently pulled the animal out of the wrong ar m and guided the animal back into the start arm. If the animal failed to make an arm choice for 20 sec, or if the animal entered the platform-containing arm but failed to locate the platform, then an error was also recorded and the animal was brought back in to the start arm for that trial. An error of 4.4 was given to any animal that did not ma ke at least three choices during a given trial
60 for the pre-treatment test point. An error of 7.4 was given to any animal that did not make at least three choices during a given tr ial for the post-treatment test point. The numbers 4.4 and 7.4 were calculated by averagin g errors for all animals that did not locate the platform for block 1 (day 1-day 3) on trial 1 (T1). The animals were tested at three different time points: before treatment ( 15 days), mid-treatment (15 days), and posttreatment (9 days). Brain Collection and Dissection Following the third and final behavior al testing in the RAWM, animals were deeply anesthetized with pe ntobarbital (100mg/kg) and pe rfused with 100ml of 0.9% saline. Post mortem brains were immediatel y removed and bisected sagitally. The left hemisphere was fixed in 4% paraformaldehyde. The right hemisphere was chilled in cold saline for 1 minute and then dissected into 5 major brain regions; cerebellum, anterior cortex, striatum, posterior cortex, and hippo campus. The individual brain regions were transferred into individual 1.5 ml Eppendorf t ubes and then were quick frozen at -80C on dry ice for biochemical analyses. Th e left hemisphere was stored in 4% paraformaldehyde overnight (12 hrs) and then transferred to a graded series of sucrose solutions (stored continually at 4C) begi nning at 10% and finishing at 30%, where tissues remained until sectioning. Tissues were coronally sectioned on a sliding microtome at 25 m for A immunohistochemistry and histology. Histological Analysis Acid Fuchsin/Toulidine Blue Staining. Degenerative neurons were stained using a combination of acid fuchsin and toulidine blue staining. Acid fuchsi n is an anionic dye that strongly stains the nuclei and cytoplas m of necrotic neurons due to high levels of
61 proteins rich in arginine and lysine (Kiernan et al., 1998; Victorov et al., 2000). The acidophilia, or strong attraction to acidic dyes, is considered to be a hallmark of neuronal damage and death that may be due specifically to brain ischemia. Toulidine blue is a basic metachromatic stain that stains nuclei da rk blue and cytoplasm light blue and serves to intensify staining. Degenerative cells ha ve an irregular atrophied shape and a much darker stain, denoting the aci dophilia associated with degene ration. To begin staining, 25 m coronal sections of brains stored at 4 C in PBS (including hippocampal regions) were mounted on gelatin prepared slides. The slides were brought through a standard rehydration scheme beginning with xylene and progressing from 100% EtOH to 40% EtOH and finishing in two water baths. Rehydr ated sections were then dipped in 1% acid fuchsin solution for approximately 35sec. The acid fuchsin stained slides were washed twice in water and then dipped in toluidin e blue solution for approximately 45 sec and rewashed in two water baths. After air dr ying, slides were coverslipped for microscope analysis. The first analysis performed (semi-quant itatively) scored the degree of cell degeneration in the dentate gyrus associated with the dorsal hippocampus. For any given animal, 5 sections (spaced at least 75 m ap art) were scored usi ng a scale of 0-4 and averaged. In the scale, 0 represents an an imal with only 1-2 dege nerative cells visible along the superficial gra nule cell layer, 1 represents an animal with 3-6 sporadic cells along the superficial gr anule cell layer, 2 represents 7-15 degenerative cells along the superficial layer, 3 represents lines of degene rating cells along the s uperficial layer, and 4 represents an animal with lines of degene rating cells along the s uperficial granule cell layer alongside clusters of dege nerating cells penetrating deep er layers. Analyses were
62 conducted blind to genotype and treatmen t at 20x magnification on a Zeiss MC-63A microscope. Degenerative cell counts were performed in the CA1&2, CA3, and CA4 regions of the dorsal hippo campus. All counts were done on 5 representa tive sections (spaced at least 75 m apart) per animal and averaged. 6E-10 & Thioflavin S staining. Staining with the 6E10 antibody detects both compact and diffuse A plaques. In brief, 25m s ections were mounted on pre-treated slides and processed through standard heat induced epitope retrie val steps beginning in 25mM citrate buffer (pH 7.3). Sections we re incubated overnight at 4C with the primary antibody; an antiA antibody (6E-10 purchased from Signet) diluted 1:2500. A secondary antibody, anti-mouse IgG was used and sections were developed with a NovaRed (Vector) substrate kit. Slides we re then brought through a dehydration scheme (beginning with water proceeding through graded series of alcohols and ending in a 5 sec xylene dip). Finally slides were coversli pped with a xylene-based mounting media for microscope analysis. Thioflavin S staining was used to only detect compact (dense) A deposits. Sections (25 m) mounted on gelatin dipped slid es were immersed in 1% Thioflavin S in 50% EtOH for 5 minutes. Sections were then immersed in graded alcohols, followed by xylene, and coverslipped fo r microscope analysis. Image Analysis All data collected from the 6E -10 and Thioflavin-S staining was analyzed on a Nikon Eclipse E1000 microscope using either 4x (Thioflavin S) or 10x (6E-10) Plan Flour objective lenses. Im ages were captured using a Retiga 1300 CCD with a QImaging RGB LCD-slider. For th e thioflavin S staining, a Nikon BV-2B fluorescence filter cube was used. Data from both of these stains was obtained from three
63 equally spaced coronal sec tions through the dorsal hippocam pus including the overlying parietal cortex. Image analysis was perfor med using customized software written in Visual Basic 6.0 (Microsoft) that used Auto -Pro function calls to segment and quantify images according to the established prot ocols used by Costa et al (2004). A deposition was quantified as a percent of area of interest (=Area stained total/Area Measuredtotal). Neurochemical Analysis Lipid Peroxidation Measures. Isofurans and 8-isopros tanes are stable byproducts of lipid peroxidation and were assayed in post-mortem brain tissue of the Tgand double transgenic (APP/PS1) mice. The anteri or and posterior cort ices were sent to Dr. J. Roberts at Vanderbilt University for analysis. Briefly, isofurans and 8-isoprostanes were quantified by stable is otope dilution gas chromatogr aphy/negative ion chemical ionization mass spectrometry as decr ibed by Fessel et al. (2003). Protein Markers. Protein level expression was assessed using western blot analysis for several proteins including: A PP, APOE, COX-2, and GFAP. Tissue samples from the hippocampus were homogenized in 10mM sodium acetate buffer pH 7.2 containing 0.1% triton X-100 and mammalia n protease inhibitor cocktail (Sigma Chemical). Samples were then centrifuged and adjusted to cont ain identical protein concentration based on the Lowry protein assay. The remaining homogenate was electrophoresed over 4-20% gradient poly acrylamide gels and transferred to polyvinylidene difluoride membranes. Blots were visualized using anti bodies specific to each protein, and developed with chemilumiscence detection using horseradish peroxidase-conjugated secondary IgG. Gro ss differences in expression level were determined by visually comparing blot sizes.
64 Statistical Analysis Behavioral Analysis. A total of three sets of behavi oral data were collected from the three rounds of RAWM testi ng. Before statistical analysis of the behavioral data, the data was divided into 3 day blocks to ai d in data presentation. Any outliers or nonperformers (animals who display consistent behavior inhibiting proper performance) were eliminated from the behavioral statistica l analysis. The statistical analysis included only those animals with a full data set (e.g., completed all behavioral testing); as the animals that died during the course of the study were eliminated from all behavioral analyses. The RAWM behavioral measures were analyzed with both one-way ANOVAs and two-way repeated measure ANOVAs. Following ANOVA analysis, post-hoc pairby-pair differences between groups were de termined through the Fisher LSD test. Differences between groups were considered significant at p<0.05. Paired T-tests were performed to determine if there were any ch anges in each groupÂ’s prevs. post-treatment behavior. Histological/Biochemical. Pathological data analysis from the histological and biochemical portion of the experiment was performed using ANOVA followed by FisherÂ’s LSD post-hoc test. Correlationa l analyses were conducted using the Systat analytical software package. Correlations were done between beha vioral data and the histological/biochemical data to determine if any relationships exist.
65 Results Behavior Pre-Treatment Testing. In pre-treatment RAWM testing (Figs. 2 and 3), both Tg+ and Tgmice performed similarly as ev idenced by no overall group effect across all 5 blocks for working memory Trials 4 [F(1,17)=0.21, p=n.s.] and Trial 5 [F(1,17)=0.27, p=n.s.]. Animals in both groups collectively show ed improved working memory across all 5 blocks of testing, as indicated by hi ghly significant block effects for both Trial 4 [F(4,68)=3.86, p<0.01] and Trial 5 [F(4,68)= 10.83, p<0.00001). At individual blocks, the only significant group difference occurred during Block 2, Trial 5 of testing. During the last three blocks of te sting, however, Tgand Tg+ mice were identical in working memory performance (Fig. 2). Across all 15 days of testing, Tg+ and Tgmice were both able to reduce their Â“overallÂ” number of errors between the first semi-random trial (Trial 1) and working memory Trials 4 and 5 (F ig. 3). Thus, Tg+ and Tgmice exhibited similar working memory performance during pr e-treatment testing, with both groups able to improve performance across trials and acro ss blocks of testing. When animals in the Tg+ group were assigned to receive hyperoxi a or air treatments, the pre-treatment performance of these two sub-groups was id entical. Even comparing pre-treatment performance between the 3 APPsw+PS1 a nd 2 APPsw mice comprising the future hyperoxia group, there were no differences in working memory. Mid-Treatment Testing. There were no differences in working memory performance between Tg+Con and Tg+O2 treatment groups overal l or during individual blocks of performance for either T4 or T5 (data not shown).
66 Post-Treatment Testing. Over the 3 blocks of post-tr eatment testing (Figs. 4 and 5), there was a significant group e ffect for Trial 5 [F(2,16)=3.56; p 0.05], with post hoc analysis indicating that Tg+O2 mice (but not Tg+Con mice) had impaired T5 working memory vs. Tgcontrols (p<0.02). The gr oup effect for T4 approached significance [F(2,16)=2.45, p=0.11] with post hoc analysis again showing impairment of Tg+O2 mice vs. Tgcontrols (p 0.05). Closer inspection of indivi dual blocks (Fig. 4) revealed a consistent trend for Tg+O2 mice to have the greatest number of T4 and T5 errors such that, when overall T4 and T5 performance wa s analyzed (Fig. 5), significantly greater numbers of working memory T4 and T5 errors were evident for the Tg+O2 group vs. Tgcontrols. Performance of Tg+Con mice consistently fell between the Tg+O2 and Tggroups, indicating a non-signi ficant trend for impairment. The impaired working memory of Tg+O2 mice is underscored by comp aring overall T1 vs. T5 for each group (Fig. 5). Both the Tgand Tg+Con groups show ed a significant redu ction in errors from T1 to T5 (p<0.00001 and p<0.005, respectively), demonstrating good working memory. By contrast, Tg+O2 mice could not significan tly reduce their overall number of errors between T1 and T5 (p=n.s.). As was the case for pre-treatment testing, there was no difference in post-infusion performance of the 3 APP+PS1 mice vs. the 2 APP mice that comprised the hyperoxia treatment group (data not shown). Swim speed analysis (the number of seconds per error made) revealed that there were no significant differences between groups in swim speed (data not shown. Pre Â– vs. Post-Treatment To further elucidate any changes in wo rking memory performance resulting from gas treatment, each groupÂ’s pre-treatment RA WM performance (overall T5 errors) was
67 compared to their post-treatment performa nce (Fig. 6). Tgmice showed a nearlysignificant (p<0.09; paired t-te st) decrease in post-treatment T5 errors, while Tg+Con mice exhibited stable prevs. post-treatmen t performance. In sharp contrast, Tg+O2 mice made significantly more T5 errors dur ing post-treatment testing (p<0.05). The poorer working memory performance of Tg +O2 mice during post-treatment testing is further exemplified by examining the prevs. post-treatment performance of individual animals for all three groups (Fig. 7). The va st majority of animal s in the Tgand Tg+Con groups showed improved or stabilized perfor mance during post-treatment testing, while four of the 5 Tg+O2 mice exhibited poorer post-treatment perf ormance.
68 T1 T4 T5 B1 Pre-Treatment Errors 0 1 2 3 4 5 6 T1 T4 T5 B2 Tg Tg + T1 T4 T5 B3 T1 T4 T 5 B4 T1 T 4 T 5 B5 Fig. 2. RAWM pre-treatment acquisition (T1T4) and memory retention (T5) in Tg+ mice and Tgmice over five 3-day blocks. = significant difference between Tgand Tg+ (p<.025) indicating an impa irment selectively in Block 2 (B2) in working memory (T4 & T5) for the Tggroup. Otherwise, no ot her group differences were seen, especially in the last 3 blocks (B4 & B5) indicating similar working memory in the Tgand Tg+ groups during pre-trea tment testing.
69 Pre-Treatment Overall Errors 1 2 3 4 5 * * T1 T4 T5 TgT1 T4 T5 Tg+ Fig. 3. Pre-Treatment acquisition (T1-T4) and memory retention (T5) overall errors (5 block average) in Tgand Tg+ mice. Both groups of mice display the ability to significantly reduce their number of errors from T1 to T4 and T5. p<.04 or higher level of significance for both groups vs. T1.
70 T1 T4 T5 B1 Post-Treatment Errors 0 1 2 3 4 5 6 7 TgTg+Con Tg+O2 T1 T4 T5 B2 T1 T4 T5 B3 Fig.4. RAWM post-treatment working memory in Tg-, Tg+Con and Tg+O2 mice over three 3-day blocks. = significan t difference between Tgand Tg+O2 (p<.02). During working memory trials 4 & 5 a trend for the Tg+O2 group to make more errors than Tg+Con mice vs Tgis particularly evident during the first two blocks.
71 Post-TreatmentT1 T4 T5 Trials Overall Errors 0 1 2 3 4 5 6 7 TgTg+Con Tg+O2 ** Fig.5. Overall RAWM post-treatment worki ng memory performance across 9 days of testing. Tg+O2 mice were significantly impaired in comparison to Tgmice during both T4 overall (* p .05) and T5 overall (** p< .02). Tg+Con mice remained unimpaired.
72 Pre Post Prevs. Post-Treatment Pre Post Overall T5 Errors 1 2 3 4 5 TgTg+Con Tg+O2 Pre Post Fig.6. RAWM prevs. post-treatment overall T5 Errors for the Tg-, Tg+Con, and Tg+O2 groups. p< .05 (paired t-test) pre-treatment vs. post-treatment overall T5 errors in the Tg+O2 group. Both the Tgand Tg+Con group displayed a trend to reduce their number of errors from the preto post-treatment time points.
73 TgOverall T5 Errors 0.0 1.0 2.0 3.0 4.0 5.0 Pre Post Pre Post Tg+O2 Pre Post Prevs. Post-Treatment Tg+Con Fig.7. Individual plots of prevs. post-treatment overall T5 errors for all animals in each group. All of the animals in the Tg+O2 group (with the exception of one that stabilized) are shown to increase their number of overall T5 errors.
74 Neuropathology and Neurochemistry Degenerative Neuronal Counts. To determine any effects of transgenicity and/or treatment on the number of neurodegenera tive neurons in hippocampus, the acid fuschin/toluine blue method for identifying su ch neurons was used. As shown in Figure 8, there were no effects of transgenic or hyperoxia treatment on mean numbers of degenerative neurons in CA1&2, CA3 and CA 4 hippocampal regions. Utilizing a scale of 0 to 4, semi-quantitative analysis of neurodegenative neuron numbers in dentate gyrus also yielded no effects of transgen ic or hyperoxia treatment (Tg-, 1.6 0.3; Tg+Con, 1.1 0.4; Tg+O2, 1.8 0.4). However, correlation anal ysis revealed a number of significant associations be tween hippocampal neurodege nerative neurons and posttreatment working memory in Tg+ mice collectiv ely, irrespective of gas treatment. These correlations generally involved overall Trial 5 performance, such as the three positive correlations shown in Fig. 9 between T5 errors and neurodegenerative neuronal counts/rating in dentate gyrus (p<0. 02), CA3 (p<0.05), and CA4 (p<0.02). The correlation involving CA3 was even more st riking if overall T5 Â“latencyÂ” was used (p<0.005, r=0.861), rather than T5 errors. Thus, high numbers of neurodegenerative neurons in these hippocampal brain regions were associ ated with more overall T5 errors. Neurochemistry. Two novel products of lipid peroxida tion, 8-isoprostane and isofurans, were measured in brain tissue from all three groups. As shown in Table 2, analysis of 8isoprostane revealed no transgen icity or hyperoxia effects in e ither anterior or posterior cortex. Although no effect of hyperoxia treatment was also seen for isofurans in either cortical area, anterior cortex isofuran le vels were significantly reduced in both Tg+ groups (Table 2). Correlati on analysis between lipid peroxidation measures and working
75 memory revealed significant negative correla tions between final block T5 performance and iso-furan levels in anterior cortex (Fig. 9). Animals with highe r iso-furan levels in anterior cortex exhibited be tter working memory (less errors). This association was present for Tg+ mice alone, as well as fo r Tg+ & Tgmice combined. Although there were no correlations between lipid per oxidation markers and degenerative neuronal counts in hippocampus for combined Tg+ & Tgmice or Tg+ mice alone, several correlations were present for Tg + controls alone (n=6). Highe r posterior cortex levels of isofurans were correlated w ith the number of degenera tive neurons in CA1/2 (r=0.858, p<0.05) and higher posterior cortex levels of 8-isoprostane correlated with the number of degenerative neurons in the CA3 region. Thus increased lipid per oxidation in posterior cortex was associated with higher numb ers of hippocampal degenerative neurons. Preliminary western blots were done investigating the levels of expression of APP, ApoE, COX-2, and GFAP. Visual analysis of the bl ots for each of these proteins revealed no differences in protein ex pression levels between groups (data not shown). Brain A Deposition. Immunostaining for diffuse A deposits using the 6E10 antibody indicated that diffuse A loads in neocortex and hippocampus of Tg+ mice were unaffected by hyperoxia treatments (Fig. 10, uppe r). Similarly, Thioflavin S fluorescent staining for compact A deposits also revealed no eff ect of hyperoxia treatments for either brain area in Tg+ mice (Fig. 10, lower). Surprisingly few correlations were found between these two A neuropathology measures vs. working memory, degenerative neuron counts, and lipid peroxi dation markers. Indeed, no correlations were evident between either A neuropathology measure and any working memory measure or degenerative neuronal counts in hippocampus. Although no co rrelations were also found
76 between 6E10 staining and lipid peroxidati on markers, there was a single significant correlation involving Thioflavin S: Higher cortical Thioflavin S staining correlated with higher anterior cortical is ofuran levels (r=0.627, p=0.05)
77 CA1 & 2 CA3 CA4 Mean Number of Degenerative Neurons 0 2 4 6 8 10 12 14 TgTg+Con Tg+O2 Hippocampal Region Fig.8. The mean number of degenerative ne urons in various hippocampal regions using the acid fuchsin and touilidine blue method. No significant differe nces between groups were found.
78 Fig.9. A series of correlation s catter plots illustrating th at the degree of neuronal degeneration in the dentate gyrus and the number of degenerative neurons in, CA3, and CA4 region of the hippocampus were posit ively correlated with working memory impairment in Tg+ mice, irrespective of gas treatment. More degenerative neurons were seen in Tg+ mice making high numbers of wo rking memory (Trial 5) errors overall. Dentate Gyrus 0.41.22.02.8Degenerative Neuron Rating 1.0 2.5 4.0 5.5Overall T5 Errorsp= 0.017 r= 0.764 CA3 0246Mean # of Degenerative Neurons 1 2 3 4 5 p= 0.044 r= 0.679 CA4 -241016Mean # of Degenerative Neurons 1.0 2.5 4.0 5.5 p= 0.016 r= 0.765
79 Table 2. The mean amount (results presen ted as ng/g wet weight of tissue) of 8IsoProstane and Iso-Furan, both markers of lip id peroxidation, in anterior cortex and posterior cortex from each group of animal s. Lipid peroxidation in neocortex was unaffected by hyperoxia treatments in Tg+ mice, although anterior cortex isofuran levels were significantly reduced in both groups of Tg+ mice. = p<0.05 vs. Tgfor iso-furan measurement in the anterior cortex. Lipid Peroxidation Markers TgTg+Con Tg+O2 8-IsoProstane(ng/g) Anterior Cortex 3. 4 0.3 3.1 0.3 3.2 0.4 Posterior Cortex 3. 9 0.9 3.4 1.0 5.0 1.3 Iso-Furans(ng/g) Anterior Cortex 6.5 1.0 3.2 1.1* 2.5 1.4* Posterior Cortex 8.7 3.1 4.5 3.6 9.4 4.4
80 Fig.10. A pair of correlation scatterplots illustrating that an inverse correlation was present between iso-furans in anterior cortex and working memory for Tg+ mice (left), as well as for combined Tgand Tg+ mice (rig ht). Mice with higher iso-furan levels actually had better working memo ry (made less errors) than those with lower iso-furan levels. f Tg+ Mice Only -1135Iso-Furans in Anterior Cortex (ng/g) 0 1 2 3 4 5 6T5 Errors (Last Block)p= 0.025 r= -0.73 All Mice (Tg& Tg+) -2261014IsoFurans in Anterior Cortex (ng/g) -10 0 10 20 30 40 50 60 70T5 Latency (Last Block)p= 0.046 r= -0.490
81 Neocortex Hippocampus A Load (% area) 0.0 0.2 0.4 0.6 0.8 1.0 Tg+Con Tg+O2 6E10 Immunostaining Thioflavin S Neocortex Hippocampus A Load (% area) 0.0 0.4 0.8 1.2 1.6 Fig.11. A loads in neocortex and hippocampus of Tg+ mice were not affected by hyperoxia treatments. Hyperoxia had no significant effect on A deposition in either brain region, as determined by both 6E10 i mmunostaining and Thioflavin-S staining. Brains from Tgmice exhibited no staining with either method.
82 Discussion General Summary. In the present study, we evalua ted the effects of hyperoxia on AD transgenic mice in the RAWM. We de termined that exposures to 100% oxygen (hyperoxia) trigger working memory impairme nt in Tg+ mice that otherwise would have been unimpaired. Hyperoxia induced memory impairment in Tg+ mice did not involve changes in brain A deposition, degenerative cell nu mbers in hippocampus, neocortical lipid peroxidation, or hippocampal levels of APP, ApoE, COX-2, or GFAP. The combination of excess A and hyperoxia could have induced greater cerebral vasoconstriction than either one alone, resu lting in a pathologic cerebral hypoperfusion that triggered subsequent cognitive impairm ent. These results suggest that humans genetically pre-disposed to AD a nd those with increased brain A levels have increased risk of developing cognitive impairment follo wing hyperoxia treatment. This is a novel finding that calls into question the wide use of 100% oxygen treatments in aged individuals at high risk for deve loping AD following major surgery. Behavior. In pre-treatment RAWM testing at 3 months of age, Tgand Tg+ mice were identical in work ing memory performance. Both groups were able to reduce errors from T1 to T4 and T5. Therefore, the Tg+ animals were not impaired in working memory at 3 months of age (no overt plaque deposition). Post-treatment RAWM testing at 7.5 months of age revealed a cognit ive impairment selectively in the Tg+O2 group vs. the Tggroup in overall errors for both T4 and T5. The Tg+Con group trended towards impairment, however, remained non-significant. The data shows this consistent trend through all 3 blocks of post-treatment te sting, where the Tg+C on group consistently made more errors than the Tgand less errors than the Tg+O2 group. The three
83 hyperoxia treatments given to the Tg+ mice i nduced an earlier cognitive impairment in animals that would have otherwise remained unimpaired. This is evidenced by the fact that the Tg+Con group of animal s (littermates of the Tg+O2 group) were not impaired (as compared with the Tggroup) at the post-treatment test point. In comparing prevs. post-treatment beha vior in the three gr oups of animals the effect of hyperoxia becomes even more apparent. While bot h the Tgand Tg+Con groups either stabilize or improve their perf ormance in overall T5 errors from preto post-treatment testing the Tg+O2 group significantly increase d their number of errors from preto post-treatment testing. Th e behavioral finding, that hyperoxia induces impairment in rodents has been reported on pr eviously in a publication by Fukui, et al (1999). In that study the author s exposed a rat model for oxida tive stress (rats were fed a vitamin E deficient diet) to 48 hours of hyper oxia and tested cognition and memory in the Morris water-maze. They reported that th e animals exposed to hyperoxia were not capable of Â“relearningÂ”, and although the te sting methods were somewhat questionable, their findings that hyperoxia ha d a negative effect on behavi or corroborates the findings of the present study. As far as the knowledge of this laboratory permits this is the only study investigating the effects of hyperoxi a in a rodent model for behavior. Studies done with humans and hyperoxia have reported that hyperoxia has a beneficial effect on short-term memory in normal adults. In 1998 a study done by Moss et al., (1998) reported significant improvement s in immediate and delayed word recall in individuals exposed to hyperoxia for 1 or 3 minutes prior to th e start of testing. A study done by Scholey et al., (1999) reported that individuals exposed to hyperoxia for 1 minute prior to testing recalled more words a nd had faster reaction times. These studies,
84 alongside other similar studies (Prior & Cha nder 1982 & Sholey et al ., 1998) suggest that hyperoxia may be beneficial to short-term me mory when administered directly prior to testing. It is important to note that thes e studies are only tangen tially related to the current study. The long term effect of hyper oxia was not investigated in any of these studies, and more importantly the studies utilized young healthy a dults as the test subjects. Only one study done by Prior & Chander (1982) investig ated the effects of hyperoxia in elderly patients. In that study the researchers exposed elderly patients to 12 hours of post-surgery hyper oxia and then performed cognitive testing immediately following the treatment. This study reported that the hyperoxia exposure did not have a deleterious effect on cognition. It is important to note that only a single test was given directly following treatment, with no follow up studies investigati ng behavioral changes once the patient returned home. Thus, while previous literature has suggested that hyperoxia may be either benign or even be neficial under specifie d narrow conditions, the long-term effects of hyperoxia given to patien ts at high risk for developing AD has not yet been established. Pathology. Insoluble A was quantified using two staining techniques; antibody staining for compact and diffuse A deposits was done using the 6E-10 antibody from Signet, and Thioflavin-S st aining was done for compact A deposits. Both staining techniques yielded no differences between the amount of insoluble A in the Tg+Con group and the Tg+O2 group. Therefore, hype roxia does not increase A deposition in the mouse model utilized in this study. Prio r studies using APP models have reported finding a correlation between A deposition and RAWM performance (Arendash et al., 2001; Leighty et al., 2004). However, this study did not yield any correlations between
85 A deposition and RAWM performance. While the data from the two staining techniques mentioned above did not show a ny differences it would be interesting in a follow up study to quantify soluble A in an ELISA assay. Some evidence has linked hyperoxia treatment to potentially increasd A production in rodents. A study done by Wen, et al (2004) found that cerebral ischem ia in rats caused a 30% increase in secretase activity. The cerebral vaso -constriction caused by hyperoxia would hypothetically induce mild ischemia, or hypoperfusion in the brain. -secretase, as mentioned prior, is the enzyme responsible for the pathological cleavage of APP that results in the toxic A fragment. Therefore, an increase in production of -secretase caused by severe restriction of blood flow to the brain could lead to an increased amount of deposited A The expression level of -secretase would also be interesting to look at in any possible follow up studies. The APP/PS1 mouse model utilized in this study has been shown to exhibit the AD hallmarks of elevated soluble A levels alongside A deposition into compact plaques by 6 months of age (Takeuchi et al, 2000). A deposition is seen as early as 12 weeks in the cortex and hippocampus of this animal model (Holcomb et al, 1998). The hippocampus plays a strong role in spatial and memory related tasks, such as the RAWM. The sensitivity of the task to the hippocampus and A deposition makes the task extremely fitting for the present study. The present study aimed to use APP transgenic mice at an age when they were on the verge of displaying cognitive impairments, yet remained unimpaired. The hyperoxia treatme nts were designed to push such animals into cognitive impairment at an earlier age than when the animals naturally would have developed an impairment. An animal model that is in a state just prior to the onset of
86 cognitive impairment models an elderly pati ent that is in the la tent or possibly the prodromal stage of AD. As mentioned previously, A accumulation and deposition is thought to occur many years prior to disease di agnosis. Thus, aged individuals at a high risk for developing AD, have increased amounts of A in their brain. Therefore, the animal model used exhibits both an imminent decline in cognition al ongside the presence of an abundance of the toxic A protein. The marked decline in working memory in the Tg+O2 group may therefore, be predictive of what an elderly patient may experience after being treated with hyperoxia. Prior literature has suggested a specific vulnerability of the neurons of the hippocampus to ischemia (Pulsinelli et al, 1985 ; Kirino et al, 1985). Both Pulsinelli and Kirino exposed a rat model to 10-30 min. of ischemia (t hrough a 4 vessel occlusion technique) and found that the neurons in the CA1 region of the hippocampus were selectively vulnerable to the treatment. Ki rino et al, 1985 related tw o theories explaining this selective vulnerability. The first theory has been coined the Â“vascular theoryÂ” and stated that the design of the local vasculat ure and the location of the hippocampus in the watershed area between the carotid and verteb robasilar territories lends the hippocampus to selective vulnerability due to ischemia. The second theory hypothesized that the neurons of the hippocampus were selectively vuln erable to ischemia due to differences in their physical and chemical characteristics. Prior literature has suggested that the hippocampus is also selectively vulnerable to injury induced by hypoxia (Gorgias et al, 1996). In a study by Gorgias et al. (1996), rats exposed to extreme hypoxia (3%O2) for 6 minutes displayed selective injury to the ne urons of the CA1 region of the hippocampus.
87 This study suggested that the neurons of th e hippocampus, specifica lly in the CA1 region, are particularly vulnerable to cha nges in oxygen availability. The treatment given in the present study, hyperoxia, is a stro ng generator of free radicals as well as a potent vasoconstrictor of cerebral vessels that may result in mild ischemia. It was therefore, hypothesized that in the present study there would be evidence of increased neuronal degene ration in the hippocampus in the Tg+O2 group due to oxidative damage and decreased cerebr al blood flow (at least during hyperoxia treatments). However, acid fuchsin and touili dine blue staining for degenerative neurons did not reveal any group diffe rences at euthanasia (10 days following final hyperoxia treatment). Cell counts performed in the CA1/2, CA3 and CA4 regions of the hippocampus, as well as in the dentate gyrus, did not reveal any di fferences between the three groups. Therefore, there was no effect of transgenicity or hyperoxia on hippocampal neuronal degeneration. Current lit erature reports no evidence of significant cell loss in the APP/PS1 model in cortical a nd hippocampal areas at 3 to 12 months of age (Takeuchi et al, 2000). Therefore, it was not expected that there should be a transgenic effect for this histological m easure at the 8 month sacrifice timepoint. Nonetheless, a recent study by Sadowski et al. (2004) reported that in 22 month old APP/PS1 mice there was a 35.8% dropout of neurons on the CA1 region of the hippocampus that was detected using stereol ogical techniques. Also, a study by Schmitz et al. (2004) reported signif icant hippocampal neuronal loss in 17 month old APP/PS1 transgenic mice bearing both the Swedish and London APP mutations. Perhaps, the absence of significant hyperoxi a effects on neuronal degenera tion was due to three 3-hour hyperoxia treatments being insufficient to i nduce degeneration detect able by the methods
88 used. Or there may have been an incr ease in degenerative neurons immediately following hyperoxia that was missed due to the 10 day delay in sacrifice. In future studies, a more lengthy treatment time and im mediate sacrifice shoul d be considered. Clearly, the hyperoxia-induced cognitive im pairment found in this study was not attributable to hippocampal neuronal degenera tion that was detected by acid fuchsin and touilidine blue staini ng. However, a more thorough invest igation of the cellular integrity in the hippocampus, or stereol ogical analysis should be performed before this can be ruled out as a probable cause of the impairment. It is possible that, with more sensitive t echniques a treatment or transgenic effect on hippocampal neurons may be detectable. Fo r example, it is possible that by using scanning electron microscopy, changes in cellular morphology (including organelle integrity) could be identified. A study done by Urano, et al (1997) using a rat model investigated the morphological changes through electron microscopy in the brain associated with exposure to 100% oxygen fo r 48 hours followed by immediate sacrifice. They found swollen astrocytes around vessels, deformed nerve cell nuclei, swollen mitochondria, and an abnormal accumulation of synaptic vesicles in swollen nerve terminals. These cellular changes may be the underlying causes of the cognitive impairment reported on here. Correlational analyses of cell count data w ith RAWM data revealed an interesting correlation for all Tg+ mice combined. Th e number of overall T5 working memory errors was highly correlated w ith the degenerative cell scorin g in the dentate gyrus, and the mean number of degenerative neurons found in the CA3 and CA4 region of Tg+ animals. Tg+ animals that had a high nu mber of degenerative neurons were found to
89 make more overall T5 errors. This correlati onal analysis provides limited evidence that the amount of neuronal degeneration in the hi ppocampus is linked to working memory in Tg+ animals. While there is no single measure used to assess the overall level of oxidative stress in an organism, an assay of iso-fura ns and 8-isoprostanes (stable by-products of lipid peroxidation uniquely regulated by oxygen tension) is used to assess the amount of lipid peroxidation tissue has undergone. Isofurans, in particular, are increased in conditions of elevated oxygen concentrations such as that induced by hyperoxia (Roberts L & Fessel J, 2004). The current study investigated levels of iso-furans and 8isoprostanes in both the anteri or and posterior cortices of all animals. There was no effect of hyperoxia treatment for either marker As well, there were largely no effects of transgenicity on these lipi d peroxidation markers with one exception. There was significantly less iso-furan in th e anterior cortices of Tg+ animals. This reduction was somewhat surprising as previous literature has reported increased lipid peroxidation in AD transgenic animals (Schuessel et al, 2005; Pratico et al, 2001). The study done by Schuessel et al, (2005) found that in APP tran sgenic mice, the lipid peroxidation marker 4-hydroxynenal was significantly increased by 3 months of age. The study done by Pratico et al, 2001 examined is oprostanes (specifically, 8,12iso -iPF2 Â–VI) in urine, plasma, and brain tissue in the APPsw mouse mo del. The authors reported increases in isoprostanes universally by 8 months of age, months before overt A deposition in such mice. Increases in lipid peroxidation in AD tr ansgenic mice may be due in part to the presence of A A study by Matsuoka et al, 2001 found that the lipid peroxidation marker 4-hydroxy-2-noneal (HNE) increases in APP/PS1 mice in relation to an age-
90 associated increase in amyloid load between 7 and 30 months of age. This study suggests a significant role for A in modulating oxidative stress. A has also been found to generate free radicals (Hensley, 1994). Increases in free radicals lead to increased oxidative stre ss with subsequent oxidative damage such as lipid peroxidation. This fundamental idea coupled alongside the previous literature provided the foundation for the hypothesis that the Tg+ animals would have increased levels of the lipid peroxidation markers ex amined. Our hypothesis that the treatment group (Tg+O2 ) would have the highest amount of lipid peroxidation incorporates the increases in oxidative stress previously reported in AD Tg+ animals with the addition of hyperoxia. Hyperoxia is also a known producer of free radicals, Â“about 5% of inspired oxygen is converted to the dangerous superoxi de radicalÂ” (Fridovi ch, 1983). Superoxide as mentioned above was shown in a study by Johnson (2001) to react with NO to form the toxic peroxynitrite (ONOO-) radical. Peroxynitrite is a pot ent oxidant that attacks DNA, proteins, and lipids leading to cellular damage. Thus, the Tg+O2 animals, being exposed to two sources of increased oxidative stress were hypothesized to exhibit a heightened amount of lipid peroxidation markers. Howe ver, as mentioned previously, the data presented here does not suppor t this hypothesis. Our fi nding that Tg+ animals had significantly decreased amounts of iso-furan in the anteri or cortex may be due to decreases in brain metabolism. A study done by Sadowski et al, 2004 reported finding significant decreases in hippocampal glucose me tabolism in 22 month old APP/PS1 mice. Given the fact that glucose metabolism can ge nerate free radicals, a significant decrease in metabolism could easily explain a decrease in oxidative stress markers. Another possible explanation for the decrease in lipid peroxidation in Tg+ animals might involve
91 the role of A itself. As mentioned previously, th ere is growing evidence that points towards the antioxidant capabilities of A as demonstrated by its ability to prevent lipoprotein oxidation in the C SF (Kontush, 2001). These findings are in accord with earlier work that has found A plaque load in humans to be inversely correlated with oxidative stress, where oxidative damage is qua ntitatively greatest in the early stages of the disease and reduces with disease progr ession (Nunomura, 2001). Thus, decreased lipid peroxidation in AD brains wa s associated with increased A deposition. The Tg+ animals may have had decreas ed lipid peroxidation as co mpared with non-transgenic littermates due to the presence of A in their brains. The major ity of literature does not support this idea. However, as the exact role of A has yet to be fully elucidated, it is relevant to mention this hypothesis that A is acting as an antioxidant as it has not yet been disproven. In any event, a possible explanation for the lack of a treatment effect may be simply due to the fact that the anim als were sacrificed 10 days post-treatment. The effects of hyperoxia on lipid peroxidation may therefore be transient. A follow-up study in which animals are sacrificed im mediately following hyperoxia treatment is necessary to answer this question. The lipid peroxidation data yielded severa l interesting correlations when analyzed alongside the behavioral and cell count data. In Tg+ mice alone, as well as for combined Tg+ and Tgmice, higher iso-furan levels co rrelated with lower T5 errors. Thus, an animal that performed well in the last bl ock of RAWM testing (exhibited by a low number of errors) had a higher amount of isofuran in their anterior cortex. Perhaps, more cognitively active animals utilize more oxygen resulting in increased iso-furan. This data suggests that animals with incr eased iso-furan have better cognition than
92 animals with lower levels. In Tg+ cont rol animals alone, a correlation was found between higher posterior cortex levels of iso-furans and the number of degenerative neurons in CA1/2; as well, higher posterior co rtex levels of 8-isoprostane correlated with degenerative neurons in CA3. Thus, increased lipid peroxidation in posterior cortex was associated with higher numbers of hippocam pal degenerative neurons in Tg+ untreated animals. In a yet broader picture, incr eased degenerative cell numbers in the hippocampus correlated with bot h poorer cognitive performance and increased lipid peroxidation. The exact relationship between these markers is still unclear. More research needs to be done on these novel ma rkers to discover th eir underlying function with regards to memory and cognitive functions. Vasoactive role of A Another hypothesis pres ented in this study focuses on the vasoactive role of A Numerous studies (Thomas et al, 1997; Crawford et al, 1998; Arendash et al, 1999) have re ported that freshly solubilized and vascularly injected A enhances cerebral vasoconstriction in rat models. This enhanced cerebral vasoconstriction results in decreased cerebral blood flow (Suo et al., 1998). The study by Suo et al., (1998) uti lized fluorescent micros pheres to monitor both cerebral blood flow and cerebrovascular resist ance in a rat. This technique c ould be utilized in a future study to monitor the effects of hyperoxia on cereb ral blood flow. In conjunction with these murine studies, a human study done by Mentis et al., (1996) reporte d that PET scanning techniques show decreases in resting cerebra l blood flow in patients with AD. Thus, patients with an increased amount of A in their brains may have chronic decreased cerebral blood flow resulting in a hypoperfusion of the brain. It has also been suggested by Aliyev et al., (2005) that chronic cerebral hypoperfusion may be an initiator of AD.
93 Hyperoxia is also a potent cerebral vasocons trictor (mentioned above). In the present study, hyperoxia and A may have worked in concert with one another to promote a greater hypoperfusion than either one alone, resulting in subsequent cognitive impairment. It would be interesting in a follow-up study to collect data regarding this hypothesis. During hyperoxia tr eatment, cerebral blood flow can be monitored using laser-Doppler flowmetry; this data would be able to provi de evidence that the hyperoxia treatment induces vasoconstriction in AD transgenic mice. In post-mortem tissue, white matter lesions and the extent of activated mi croglia and astroglia could be analyzed as markers of hypoperfusion (S hibata et al., 2004). Clinical Implications of the Hyperoxia Findings. Currently, two main risk factors that are widely accep ted for developing AD are age and the inheritance of the ApoE4 allele. Of course, many other risk factors have been put forth involving an individualsÂ’ health, health history, and lifestyle (discussed above). It has been previously suggested by Ahua-Haim et al., (1998) that a link between surgery in the elderly and precipitation of AD exists. Su rgery incorporates many potentially harmful components including administration of va rious drugs, anesthesia, changes in blood pressure, and possible periopera tive hyperoxia treatments. Results from the present study suggest that perioperative hyperoxia tr eatment and a pre-disposition to AD are additional and synergistic risk factors for AD -related postoperative cognitive impairment. These findings will hopefully trigger subsequent studies that further elucidate the role of hyperoxia in patients at risk fo r developing AD, as well as call into question the wide use of this potentially toxic treatment.
94 References Aliyev A, Chen SG, Seyidova D, Smith MA, Perry G, de la Torre J, Aliev G.. Mitochondria DNA deletions in atheroscleroti c hypoperfused brain microvessels as a primary target for the developm ent of AlzheimerÂ’s disease. Journal of Neurological Science 1:229-92. 2005 Andersson J, Berggren P, Gronkvist M, Magnus son S, Svensson E. Oxygen saturation and cognitive performance. Psychopharmacology 162: 119-128. 2002 Andreasen N, Minthon L, Davidsson P, Vanm echelen E, Vanderstichele H, Winblad B, Blennow K. Evaluation of CSF-tau and CSF-A 42 as diagnostic markers for AlzheimerÂ’s disease in clinical practice. Archives of Neurology 58: 373-9. 2001 Arendash GW, Su GC, Crawford FC, Bjugstad KB, Mullan M. Intravascular -amyloid infusion increases blood pressure: imp lications for a vasoactive role of -amyloid in the pathogenesis of AlzheimerÂ’s disease. Neuroscience Letters 268:17-20. 1999 Arendash GW, Garcia MF, Costa DA, Cr acchiolo JR, Wefes IM, Potter H. Environmental enrichment improves cogniti on in aged AlzheimerÂ’s transgenic mice despite stable beta-amyloid deposition. Neuroreport 15(11):1751-4. 2004 Arendash GW, Lewis J, Leighty RE, McGowa n E, Cracchiolo JR, Hutton M, Garcia MF. Multi-metric behavioral comparison of APPs w and P301L models for AD: linkage of poorer cognitive performance to tau pathology in forebrain. Brain Research 1012:29-41. 2004 Arendash GW, Gordon MN, Diamond DM, Aus tin LA, Hatcher JM, Jantzen P, DiCarlo G, Wilcock D, Morgan D. Behavioral a ssessment of AlzheimerÂ’s transgenic mice following long-term A vaccination: task specificity and correlations between A deposition and spatial memory. DNA and Cell Biology 20(11):737-44. 2001 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 transgenes. Brain Research 891:42-53. 2001 Austin L, Arendash GW, Gordon MN, Diamond DM, DiCarlo G, Dickey C, Ugen K, Morgan D. Short-term beta-amyloid vacci nations do not improve cognitive performance in cognitively impaired APP + PS1 mice. Behavioral Neuroscience 117(3): 478-84. 2003 Benton D, Parker PY, Donohoe RT. The suppl y of glucose to the brain and cognitive functioning. Journal of Biosocial Sciences 28:463-79. 1996
95 Benzing WC, Wujek JR, Ward EK, Shaffe r D, Ashe KH, Younkin SG, Brunden KR. Evidence for glial-mediated inflammation in aged APPsw transgenic mice. Neurobiology of Aging 20: 581-89. 1999 Berman K & Brodaty H. Tocopherol (vitami n E) in AlzheimerÂ’s disease and other neurodegenerative disorders. CNS Drugs 18(12):807-25. 2004 Bohnen NI, Warner MA, Kokmen E, Beard CM Kurland LT. AlzheimerÂ’s disease and cumulative exposure to anesthesia: a case control study. Journal of American Geriatrics Society 42(2):198-201. 1994 Bohnen N, Warner MA, Kokmen E, Kurla nd LT. Early and midlife exposure to anesthesia and age of onset of AlzheimerÂ’s disease. International Journal of Neuroscience 77(3-4):181-5. 1994 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 19:939-945. 1997 Braak, H. & Braak, E. Neuropa thological staging of Alzheimer-related changes. Buccellato LJ, Tso M, Akinci OI, Chande l NS, Budinger GR. Reactive oxygen species are required for hyperoxia-induced Bax activatio n and cell death in alveolar epithelial cells. Journal Biological Chemistry 279(8):6753-60. 2003 Calhoun ME, Burgermeister P, Phinney AL, Stalder M, Tolnay M, Wiederhold KH, Abramowski D, Sturchler-Pierrat C, Sommer B, Staufenbiel M, Jucker M. Neuronal overexpression of mutant amyloid precursor pr otein results in prominent deposition of cerebrovascular amyloid. PNAS 96(24): 14088-93. 1999 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 amyloi d precursor protein transgenic mice. Nature Neuroscience 2(3): 271-76. 1999 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 -amyloid plaques in a mouse model of AlzheimerÂ’s disease. Nature 408: 975-79. 2000 Conte V, Uryu K, Fujimoto S, Yao Y, Rokach J, Longhi L, Trojanowski JQ, Lee VM, McIntosh TK, Pratico D. Vitamin E re duces amyloidosis and improves cognitive function in Tg2576 mice following repeti tive concussive brain injury. Journal of Neurochemistry 90:758-64. 2004
96 Costa DA, Nilsson LN, Bales KR, Paul SM, Po tter H. Apolipoprotein is required for the formation of filamentous amyloid, but not for amorphous A deposition, in an A PP/PS double transgenic mouse model of AlzheimerÂ’s disease. Journal of AlzheimerÂ’s Disease 6:509-514. 2004 Crawford F, Suo Z, Fang C, Mullan M. Characteristic of the in Vitro vasoactivity of Amyloid Peptides. Experimental Neurology 150:159-68. 1998 Cummings JL. AlzheimerÂ’s Disease. New England Journal of Medicine 351(1):56-67. 2004 Delagarza VW. Pharmacologic treatment of AlzheimerÂ’s disease: an update. American Family Physician 68(1) 1365. 2003 De Smet K, Eberhardt I, Reekmans R, Cont reras R. Bax-induced cell death in Candida albicans. Yeast 21(16):1325-34. 2004 Demchenko IT, Oury TD, Crapo JD, Piantadosi CA. Regulation of the BrainÂ’s Vascular Responses to Oxygen. Circulation Research 91:1030-37. 2002 Dodart JC, Meziane H, Mathis C, Bales KR, Paul SM, Ungerer A. Behavioral disturbances in transgenic mice overexpressing the V717F -Amyloid precursor protein. Behavioral Neuroscience 113(5): 982-990. 1999 Dodart JC, Mathis C, Saura J, Bales KR Paul SM, Ungerer A. Neuroanatomical abnormalities in behaviorally character ized APPv717F Transgenic Mice. Neurobiology of Disease 7: 71-85. 2000 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. Journal of Neuroscience 23(12):5219-26. 2003 Dickson, Dennis W. Apoptotic mechanisms in Alzheimer neurofibrillary degeneration: cause or effect? The Journal of Clinical Investigation 114(1) 23-27. 2004 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-beta 42(43) in brains of mice expressing mutant presenilin 1. Nature 383(6602):710-3. 1996 Dumont M, Strazielle C, Stau fenbiel M, Lalonde R. Spatia l learning and exploration of environmental stimuli in 24-month-old female APP23 transgenic mice with the Swedish mutation. Brain Research 1024(1-2):113-21. 2004 Esteban JA. Living with the enem y: a physiological role for the -amyloid peptide. Trends in Neuroscience 27(1). 2004
97 Facheris M, Beretta S, Ferrarese C. Peripheral markers of oxidative stress and excitotoxicity in neurodegenerative disord ers: tools for diagnosis and therapy? Journal of AlzheimerÂ’s Disease 6(2): 177-84. 2004 Fessel JP, Hulette C, Powell S, Roberts LJ 2nd, Zhang J. Isofurans, but not F2 Â– isoprostanes, are increased in the substantia nigra of patients with ParkinsonÂ’s disease and with dementia with lewy body disease. Journal of Neurochemistry 85: 645-650. 2003 Fisher A. Therapeutic strategies in Al zheimerÂ’s disease: M1 muscarinic agonists. Journal of Pharmacology 84(2): 101-12. 2000 Freeman BA & Crapo JD. Hyperoxia increas es oxygen radical production in rat lungs and lung mitochondria. The Journal of Biological Chemistry 256(21):10986-992. 1981. Fridovich I. Superoxide ra dical: an endogenous toxicant. Annual Reviews Pharmacology and Toxicology 23:239-57. 1983. Fox, NC & Schott JM. Imaging cerebral atroph y: normal ageing to AlzheimerÂ’s disease. Lancet 363(9406): 392-4. 2004 Fukui K, Onodera K, Shinkai T, Suzuki S, Ur ano S. Impairment of learning and memory In rats caused by oxidative stress and agi ng, and changes in antioxidative defense systems. Annals of the New York Academy of Sciences 928:168-75. 2001 Gasparini M, Vanacore N, Schi affini C, Brusa L, Panella M, Talarico G, Bruno G, Meco G, Lenzi GL. A case-control study on Alzhei merÂ’s disease and exposure to anesthesia. Neurological Science 23:11-14. 2002 Games D, Adams D, Alessandrini R, Barbour R, Berthelette P, Blackwell C, Carr T, Clemens J, Donaldson T, Gillespie F, et al Alzheimer-type neuropathology in transgenic mice overexpressing V717F -amyloid precursor protein. Nature 373: 523-27. 1995 Gerlai R, Fitch T, Bales KR, Gitter BD. Behavioral impairment of APPv717F mice in fear conditioning: is it only cognition? Behavioural Brain Research 136: 503-9. 2002 Giacchino J, Criado JR, Games D, Henriksen S. In vivo synaptic transmission in young and aged amyloid precursor protein transgenic mice. Brain Research 876:185-190. 2000 Gomez-Tortosa E, Gonzalo I, Fanjul S, Sain z MJ, Cantarero S, Cemillan C, Yebenes JG, del Ser T. Cerebrospinal fluid markers in dementia with lewy bodies compared with Alzheimer disease. Archives of Neurology 60(9): 1218-22. 2003
98 Gomez-Ramos P, Moran MA. Ultrastructural as pects of neurofibrill ary tangle formation in ageing and AlzheimerÂ’s disease. Microscopy Research and Technique 43(1): 49-58. 1999 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 A deposits in transgenic APP+PS1 mice. Neurobiology of Aging 22(3):37785. 2001 Gorgias N, Maidatsi P, Tsolaki M, Alvanou A, Kiriazis G, Kai doglou K, Giala M. Hypoxic pretreatment protects against neurona l damage of the rat hippocampus induced by severe hypoxia. Brain Research 714:215-225. 1996 Grant WB. Dietary Links to Al zheimerÂ’s disease: 1999 update. Journal of AlzheimerÂ’s Disease 1(4-5):197-201. 1999 Hart DJ, Craig D, Compton SA, Critchlow S, Kerrigan BM, McIlroy SP, Passmore AP. A retrospective study of the beha vioural and psychological symp toms of mid and late phase AlzheimerÂ’s disease. International Journal of Geriatric Psychaitry 18 (11): 1037-42. 2003 Herbert LE, Scherr PA, Bienias JL, Bennett DA, Evans DA. Alzheimer disease in the US population: prevalence estimates using the 2000 census. Arch Neurology. 60(8):1119-22. 2003 Hoehn T, Felderhoff-Mueser U, Maschewski K, Stadelmann C, Sifringer M, Bittigau P, Koehne P, Hoppenz M, Obladen M, Buhrer C. Hyperoxia causes inducible nitroc oxide synthase-mediated cellular damage to the immature rat brain. Pediatric Research 54(2):179-184. 2003 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 precursor protein and presenilin transgenes. Nature Medicine 1:97-100. 1998 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 deposits. Behavior Genetics 29(3):177-85. 1999 Hsiao KK, Borchelt DR, Olson K, Johannsdottir R, Kitt C, Yunis W, Xu S, Eckman C, Younkin S, Price D, et al. Age-related CN S disorder and early death in transgenic FVB/N mice overexpressing Alzheime r amyloid precursor proteins. Neuron 15:1203-18. 1996
99 Hughes LF, Perkins K, Wright BD, Westrick H. Using a Rasch schale to characterize the clinical features of patients with a clinical diagnosis of uncertain, probable, or possible Alzheimer disease at intake. Journal of Alzh eimerÂ’s Disease 5(5): 367-73. 2003 Huitron-Resendiz S, Sanchez-Alavez M, Gallego s R, Berg G, Crawford E, Giacchino JL, Games D, Henriksen SJ, Criado JR. Age-i ndependent and age-related deficits in visuospatial learning, sleep-wake states, thermoregulat ion and motor activity in PDAPP mice. Brain Research 928(1-2): 126-137. 2002 Imbimbo BP. The potential role of non-ster oidal anti-inflammatory drugs in treating AlzheimerÂ’s disease. Expert Opinions Investigating Drugs 13(11):1469-81. 2004 Irizarry MC, McNamara M, Fedorchak K, Hsiao K, Hyman BT. APPsw Trasngenic Mice Develop Age-Related A Deposits and Neuropil Abnorma lities, but no neuronal loss in CA1. Journal of Neuropathology an d Experimental Neurology 56(9): 965-973. 1997 Jantzen PT, Connor KE, DiCarlo G, Wenk GL, Wallace JL, Rojiani AM, Coppola D, Morgan D, Gordon MN. Microglial ac tivation and -Amyloid Deposit Reduction Caused by a nitric oxide releasing nonsteroidal anti-inflammatory drug in amyloid precursor protein plus presenilin -1 transgenic mice. The Journal of Neuroscience 22(6):2246-54. 2002 Jensen MT, Mottin MD, Cracchiolo JR, Leighty RE, Arendash GW. Lifelong immunization with human -amyloid (1-42) protects alzheimerÂ’s transgenic mice against cognitive impairment throughout aging. Neuroscience 130:667-684. 2005 Jiang H, Luo X, Bai D. Progress in clinical pharmacological, chemical and structural biological studies of huperzine A: a drug of traditional Chin ese medicine origin for the treatment of AlzheimerÂ’s disease. Current Medical Chemistry 10(21): 2231-52. 2003 Johnson, S. Gradual micronutrient accumulation and depletion in AlzheimerÂ’s disease. Medical Hypotheses 56(6):595-597. 2001 Johnstone M, Gearing AJ, Miller KM. A central role for astrocytes in the inflammatory response to beta-amyloid: chemokines, cytokines and reactive oxygen species are produced. Journal of Neuroimmunology 93(1-2): 182-93. 1999 Iadecola C, Zhang F, Niwa K, Eckman C, Turner SK, Fischer E, Younkin S, Borchelt DR, Hsiao KK, Carlson GA. SOD1 rescues cerebral endothelial dysfunction in mice overexpressing amyloid precursor protein. Nature Neuroscience 2(2):157-161. 1999 Irizarry MC, Soriano F, McNamara M, Page KJ, Schenk D, Games D, Hyman BT. Abeta deposition is associated with neuropil change s, but not overt neuronal loss in the human amyloid precursor protein V717F (PDAPP) transgenic mouse. Journal of Neuroscience 17(18):7053-9. 1997
100 Kelly PH, Bondolfi L, Hunziker D, Schlecht HP, Carver K, Maguire E, Abramowski D, Wiederhold KH, Sturchler-Pierrat C, Jucker M, Bergmann R, Staufenbiel M, Sommer B. Progressive age-related impairment of cognitiv e behavior in APP23 transgenic mice. Neurobiology of Aging 24(2): 365-78. 2003 Kergoat H & Faucher C. Effects of oxygen and carbogen breathing on choroidal hemodynamics in humans. Investigational Opthalmo logy Visual Sciences 40(12):290611. 1999 Kihara T, Shimohama S. AlzheimerÂ’s disease and acetycholine receptors. Acta Neurobiologiae Experimentalis 64 (1): 99-105. 2004 King, DE & GW Arendash. Behavioral char acterization of the Tg2576 transgenic model of AlzheimerÂ’s disease through 19 months. Physiology & Behavior 75: 627-642. 2002 King DL, Arendash GW, Crawford F, Sterk T, Menendez J, Mullan MJ. Progressive and gender-dependent cognitive impairment in the APP (sw) transgenic mouse model for AD. Behavioral Brain Research 103(2): 145-62. 1999 Kirino T, Tamura A, Sano K. Selective vulnerability of the hippocampus to ischemia Â– reversible and irreversible t ypes of ischemic cell damage. Progress in Brain Research 63: 39-57. 1985 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 AlzheimerÂ’s disease with Pittsburgh Compound-B. Annals of Neurology 55 (3): 306-319. 2004 Kokmen E, Beard CM, Chandra V, Offord KP, Schoenberg BS, Ballard DJ. Clinical risk factors for AlzheimerÂ’s disease; a population-based case-control study. Neurology 41:1393-1397. 1991 Kolbitsch C, Lorenz IH, Hormann C, Hinter egger M, Lockinger A, Moser PL, Kremser C, Schocke M, Felber S, Pfeiffer KP, Benzer A. The influence of hyperoxia on regional cerebral blood flow (RCBF), regional cerebral blood volume (rCBV) and cerebral blood flow velocity in the middle cerebral artery (CBFVMCA) in human volunteers. Magnetic Resonance Imaging 20(7):535-41. 2002 Kontush A, Berndt C, Weber W, Akopyan V, Arlt S, Schippling S, Beisiegel U. Amyloid-beta is an antioxidant for lipoprot eins in cerebrospinal fluid and plasms. Free Radical Biology & Medicine 30(1): 119-28. 2001 Lalonde R, Lewis TL, Strazielle C, Kim H, Fukuchi K. Transgenic mice expressing the APP695SWE mutation: effects on e xploratory activity, anxiet y, and motor coordination. Brain Research 977:38-45. 2003
101 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 Research 956:36-44. 2002 Lahiri DK. Apolipoprotein E as a target fo r developing new therapeutics for AlzheimerÂ’s disease based on studies from protein, RNA, and regulatory region of the gene. Journal of Molecular Neuroscience 23(3):225-33. 2004 Lee HG, Casadesus G, Zhu X, Takeda A, Pe rry G, Smith MA. Challenging the Amyloid Cascade Hypothesis Annals of the New York Academy of Sciences 1019: 1-4. 2004 Leighty RE, Nilsson LN, Potter H, Costa DA, Low MA, Bales KR, Paul SM, Arendash GW. Use of multimetric statistical analysis to characterize and discriminate between four AlzheimerÂ’s transgenic mouse lin es differing in Abeta deposition. Behavioral Brain Research 153(1): 107-21. 2001 Lim GP, Yang F, Chu T, Chen P, Beech W, Teter B, Tran T, Ubeda O, Ashe KH, Frautschy SA, Cole GM. Ibuprofen Suppre sses Plaque Pathology and Inflammation in a moude model for AD. The Journal of Neuroscience 20(15): 5709-5714. 2000 Liu L, Ikonen S, Heikkinen T, Tapiola T, va n Groen T, Tanila H. The effects of longterm treatment with metrifonate, a cholines terase inhibitor, on cholinergic activity, amyloid pathology, and cogniti ve function in APP and PS1 doubly transgenic mice. Experimental Neurology 173:196-204. 2002 Loeffler, David A. Using animal models to determine the significance of complement activation in AlzheimerÂ’s disease. Journal of Neuroinflammation 18. 2004 Luchsinger JA & R. Mayeux. Dietary Factors and AlzheimerÂ’s Disease. The Lancet: Neurology 3: 579-587. 2004 Maccioni RB, Munoz JP, Barbeito L. The Molecular Bases of AD and other neurodegenerative disorders. Archives of Medical Research 32:367-81. 2001 Magnoni S, Ghisoni L, Locatelli M, Caimi M, Colombo A, Valerian i V, Stocchetti N. Lack of improvement in cerebral metabolism after hyperoxia in severe head injury: a microdialysis study. Journal of Neurosurgery 98:952-58. 2003 Mamdouha AB, Moossy J, Nemoto EM, Lin MR Hyperoxia produces neuronal necrosis in the rat. Journal of Neuropathology and Experimental Neurology 45(3):233-46. 1986 Markesbery WR. Oxidative stress hy pothesis in AlzheimerÂ’s disease. Free Radical Biology and Medicine 23(1):134-47. 1997.
102 Martin, Joseph B. The Integration of Neur ology, Psychiatry, and Ne uroscience in the 21st Century. American Journal of Psychiatry 159: 695-704. 2002 Marques MA, Owens PA, Crutcher KA. Pr ogress towards identification of protease activity involved in proteolysis of apolipoprotein e in human brain. Journal of Molecular Neuroscience 24(1): 73-80. 2004 Masliah E, Sisk A, Mallory M, Mucke L, Schenk D, Games D. Comparison of Neurodegenerative Pathology in Tran sgenic Mice Overexpressing V717F -Amyloid Precursor Protein and AlzheimerÂ’s Disease. The Journal of Neuroscience 16(18): 57955811. 1996 Mattson MP. Pathways towards and aw ay from AlzheimerÂ’s disease. Nature 430: 631639. 2004 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 reponses to amyloidosis in a transgenic mouse model of AlzheimerÂ’s disease. American Journal of Pathology 158:1345-1354. 2001 Matsuoka Y, Picciano M, La Fran cois J, Duff K. Fibrillar -Amyloid evokes oxidative damage in a transgenic mouse m odel of AlzheimerÂ’s disease. Neuroscience 104(3):09613. 2001 Mentis MJ, Horwitz B, Grady CL, Alexander GE, VanMeter JW, Maisog JM, Pietrini P, Schapiro MB, Rapoport SI. Visual cortical dys function in AlzheimerÂ’s disease evaluated with a temporally graded Â“stress testÂ” during PET. American Journal of Psychiatry 153(1):32-40. 1996 Moller JT, Cluitmans P, Rasmussen LS, Houx P, Rasmussen H, Canet J, Rabbitt P, Jolles J, Larsen K, Hanning CD, Langeron O, Johns on T, Lauven PM, Kristensen PA, Biedler A, van Beem H, Fraidakis O, Silverstein JH, Beneken JE, Gravenstein JS. Long-term postoperative cognitive dysfunction in the elderly: ISPOCD1 study. The Lancet 351:857861. 1998 Moolman DL, Vitolo OV, Vonsattel JP, Shel anski ML. Dendrite and dendritic spine alterations in Alzheimer models. Journal of Neurocytology 33(3): 377-87. 2004 Moss MC, Scholey AB, Wesnes K. Oxyge n administration selectively enhances cognitive performance in healthy young adults: a placebo-controlled double-blind crossover study. Psychopharmacology 138:27-33. 1998 Mueggler T, Meyer-Luehmann M, Rausch M, Staufenbiel M, Jucker M, Rudin M. Restricted diffusion in the brain of transg enic mice with cerebral amyloidosis. European Journal of Neuroscience 20(3): 811-7. 2004
103 Mullan M, Bennett C, Figueredo C, Hughes D, Mant R, Owen M, Warren A, McInnis M, Marshall A, Lantos P, et al. Clinical f eatures of early onset, familial AD linked to chromosome 14. American Journal of Medical Genetics 60(1): 44-52. 1995 Nilsson LN, Arendash GW, Leighty RE, Cost a DA, Low MA, Garcia MF, Cracciolo JR, Rojiani A, Wu X, Bales KR, Paul SM, Po tter H. Cognitive impairment in PDAPP mice depends on ApoE and ACT-catal yzed amyloid formation. Neurobiology of Aging 25: 1153-1167. 2004 Nestor PJ, Scheltens P, Hodges JR. Advan ces in the early dete ction of AlzheimerÂ’s disease. Nature & Medicine Suppl: S34-41. 2004 Neurochem, Inc. http://www.clinicaltrials.gov/ct/show/NCT00088673?order=1 2004 Ouattara A, Boccara G, Lecomte P, Souktani R, Le Cosquer P, Mouren S, Coriat P, Riou B. Amplification by hyperoxia of cor onary vasodilation induced by propofol. Anesthesia and Analgesia 98(3): 595-603. 2004 Pagano A, Donati Y, Metrailler I, Barazzone Argiroffo C. Mitochondrial cytochrome c release is a key event in hype roxia-induced lung injury: pr otection by cyclosporin A. American Journal of Physiology Lung Cell Molecular Physiology 286:275-83. 2004 Pagano A & Barazzone-Argiroffo C. Alveol ar cell death in hyperoxia-induced lung injury. Annals of the NY Academy of Sciences 1010:405-16. 2003 Parihar, MS & T Hemnani. AlzheimerÂ’s disease pathogenesis and therapeutic interventions. Journal of Clinical Neuroscience 11(5): 456-467. 200 Park L, Anrather J, Zhou P, Frys K, Pits tick R, Younkin S, Carlson GA, Iadecola C. NADPH oxidase-derived reactive oxygen species mediate the cerebrovascular dysfunction by the amyloid peptide. Neurobiology of Disease 25(7): 1769-1777. 2005 Picciotto MR & Wickman K. Using K nockout and Transgenic Mice to Study Neurophysiology and Behavior. Physiological Reviews 78(4):1131-64. 1998 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) transgenic mouse model for Alzheimer's disease. Journal of Neuroscience Methods 87(1):87-95. 1999 Pratico D, Uryu K, Leight S, Trojanoswki JQ, Lee VM. Increased lipid peroxidation precedes amyloid plaque formation in an animal model of alzheimer amyloidosis. The Journal of Neuroscience 21(12): 4183-87. 2001 Pratico Domenico. AlzheimerÂ’s disease and oxygen radicals: new insights. Biochemical Pharmacology 63 563-67. 2002
104 Pratico D & Sung S. Lipid peroxidation and oxidative imbalance: early functional events in AlzheimerÂ’s disease. Journal of AlzheimerÂ’s Disease 6(2):171-5. 2004 Prior FN & Chander P. Air as a vaporizing ga s. Cognitive functions in elderly patients undergoing anaesthesia. British Journal of Anaesthesia 54(11):1207-12. 1982. Pulsinelli WA. Selective neuronal vuln erability: morphological and molecular characteristics. Progressive Brain Research 63:29-37. 1985 Raiha I, Kaprio J, Koskenvuo M, Rajala T, Sourander L. Environmental differences in twin pairs discordant for AlzheimerÂ’s disease. Journal of Neurology, Neurosurgery & Psychiatry 65:785-787. 1998 Roland PE. Brain Activation. Wiley-Liss 1993. Reiter, Russel J. Oxidative damage in the central nervous system: protection by melatonin. Progress in Neurobiology 56:359-384. 1998. Roberts LJ & Fessel JP. The biochemistry of the isoprostane, neuroprostane, and isofuran pathways of lipid peroxidation. Chemistry and Physics of Lipids 128: 173-186. 2004 Rockenstein E, Adame A, Mante M, Moe ssler H, Windisch M, Masliah E. The neuroprotective effects of Cerebroylsin in a tr ansgenic model of AD are associated with improved behavioral performance. J Neural Transm 110(111): 1313-27. 2003 Rogers J & Lue L. Microg lial chemotaxis, activ ation, and phagocytosis of amyloid peptide as linked phenomena in AlzheimerÂ’s disease. Neurochemistry International 39:333-40. 2001 Rosenburg RN. The molecular and genetic ba sis of AD: the end of the beginning. Neurology 54:2045-54. 2000 Rostrup E, Larsson HB, Toft PB, Garde K, He nriksen O. Signal changes in gradient echo images of human brain induced by hypoand hyperoxia. NMR Biomedical 8(1):41-7. 1995 Sadowski M, Pankiewicz J, Scholtzova H, Ji Y, Quartermain D, Jensen CH, Duff K, Nixon RA, Gruen RJ, Wisniewski T. Amyloiddeposition is associated with decreased hippocampal glucose metabolism and spatial memory impairment in APP/PS1 mice. Journal of neuropathology and experimental neurology 63(5):418-428. 2004 Schaffranietz L, Heinke W, Rudolph C, Olthoff D. Effect of normobaric hyperoxia on parameters of brain metabolism. Anaesthesiologia Reanimie 25(3):68-73. 2000
105 Schmitz C, Rutten BP, Pielen A, Schafer S, Wirths O, Tremp G, Czech C, Blanchard V, Multhaup G, Rezaie P, Korr H, Steinbusch HW, Pradier L, Bayer TA. Hippocampal neuron loss exceeds amyloid plaque load in a transgenic mouse model of AlzheimerÂ’s disease. American Journal of Pathology 164(4): 1495-1503. 2004 Scholey AB, Moss MC, Wesnes K. Oxygen a nd cognitive performance: the temporal relationship between hyperoxi a and enhanced memory. Psychopharmacology 140:123126. 1998 Scholey AB, Moss MC, Neave N, Wesnes K. Cognitive performance, hyperoxia, and heart rate following oxygen administ ration in healthy young adults. Physiology & Behavior 67(5): 783-89. 1999 Schwab C, Hosokawa M, McGeer PL. Tr ansgenic mice overexpressing amyloid beta protein are an incomplete model of Alzheimer disease. Experimental Neurology 188: 5264. 2004 Shua-Haim JR, Shua-Haim V, Ross JS. Surger y as a trigger factor for AD: A case of listen to the family. MSJ 1(5):291-94. 1998 Selkoe, Dennis J. AlzheimerÂ’s Diseas e: Genes, Proteins and Therapy. Physiological Review 81(2):741-766. 2001 Seubert P, Vigo-Pelfrey C, Esch F, Lee M, Dovey H, Davis D, Sinha S, Schlossmacher M, Whaley J, Swindlehurst C, et al. Isolation and quantif ication of soluble Alzheimer's beta-peptide from biological fluids. Nature 359: 325-7. 2002 Shibata M, Ohtani R, Ihara M, Tomimoto H. White matter lesions and glial activation in a novel mouse model of chronic cerebral hypoperfusion. Stroke 2598-603. 2004 Sjoberg F, Gustafsson U, Eintrei C. Specific blood flow reducing effects of hyperoxaemia on high flow capill aries in the pig brain. Acta Physiologia Scandanavia 165:33-38. 1999 Skoog I, Gustafson D. Hypertension, hyperten sion-clustering factors and AlzheimerÂ’s disease. Neurological Research 25(6):675-80. 2003 Smith MA, Hirai K, Hsiao K, Pappolla MA, Harris PL, Siedlak SL, Tabaton M, Perry G. Amyloiddeposition in Alzheimer transgenic mice is associated with oxidative stress. Journal of Neurochemistry 70(5): 2212-15. 1998 Sobow T, Flirski M, Liberski PP. Amyloid-be ta and tau proteins as biochemical markers of AD. Acta Neurobiologiae Experimentalis 64(1): 53-57. 2004
106 Sonkusare SK, Kaul CL, Ramarao P. Deme ntia of AlzheimerÂ’s disease and other neurodegenerative disorders-memantine, a new hope. Pharmacological Research 51(1): 1-17. 2005 Stein TD & Johnson JA. Lack of neurodegene ration in transgenic mice overexpressing mutant amyloid precursor protei n is associated with increased levels of transthyretin and the activation of cell survival pathways. The Journal of Neuroscience 22(17): 73807388. 2002 Suo Z, Humphrey J, Kundtz A, Sethi F, Pl aczek A, Crawford F, Mullan M. Soluble AlzheimerÂ’s -amyloid constricts the cer ebral vasculature in vivo. Neuorscience Letters 257:77-80. 1998. Taglialatela G, Perez-Polo JR, Rassin DK. Induction of apoptosis in the CNS during development by the combination of hyperoxia an d inhibition of glutat hione synthesis. Free Radical Biology & Medicine 25(8): 936-42. 1998 Takeuchi A, Irizarry MC, Duff K, Saido TC, Hsiao Ashe K, Hasegawa M, Mann DM, Hyman BT, Iwatsubo T. Age-related amyloid deposition in transgenic mice overexpressing both Alzheimer muta nt presenilin 1 and amyloid precursor protein Swedish mutant is not associat ed with global neuronal loss. American Journal of Pathology 157(1): 331-9. 2000 Tamagno E, Bardini P, Obbili A, Vitali A, Bo rghi R, Zaccheo D, Pronzato MA, Danni O, Smith MA, Perry G, Tabaton M. Oxidative stress Expression and Activity of BACE in NT2 Neurons. Neurobiology of Disease 10: 279-288. 2002 Thomas T, McLendon C, Sutton ET, Thomas G. Beta-Amyloid-induced cerebrovascular endothelial dysfunction. Annals of the NY Academy of Sciences 826:447-51. 1997 Tolias CM, Reinert M, Seiler R, Gilman C, Scharf A, Bullock MR. Normobaric hyperoxiaÂ—induced improvement in cerebral metabolism and reduction in intracranial pressure in patients with severe head inju ry: a prospective historical cohort-matched study. Journal of Neurosurgery 101(3):435-44. 2004 Tomidokoro Y, Harigaya Y, Matsubara E, Ikeda M, Kawarabayashi T, Shirao T, Ishiguro K, Okamoto K, Younkin SG, Shoji M. Brain abeta amyloidosis in APPsw mice induces accumulation of presenilin-1 and tau. Journal of Pathology 194: 500-506. 2001 Trinchese F, Liu S, Battaglia F, Walter S, Mathews PM, Arancio O. Progressive AgeRelated Development of Alzheimer-l ike Pathology in APP/PS1 Mice. Annals of Neurology 55:801-14. 2004 Tuppo EE & Arias HR. The role of imflammation in AlzheimerÂ’s disease. The International Journal of Biochemistry & Cell Biology 37: 289-305. 2005
107 Van Dam D, D'Hooge R, Stau fenbiel M, Van Ginneken C, Van Meir F, De Deyn PP. Age-dependent cognitive decl ine in the APP23 model pr ecedes amyloid deposition. European Journal of Neuroscience 17:388-96. 2003 Veurink G, Fuller SJ, Atwood, Martins RN. Ge netics lifestyle and th e roles of amyloid beta and oxidative stress in AD. Annals of Human Biology 30(6):639-667. 2003 Vickers JC, Dickson TC, Adlard PA, Saunders HL, King CE, McCormack G. The cause of neuronal degeneration in AlzheimerÂ’s disease. Progressive Neurobiology 60(2): 139165. 2000 Voisin T, Touchon J, Vellas B. Mild c ognitive impairment: a nosological entity?. Current Opinions in Neurology Suppl 2:S43-5. 2003 Volpicelli LA, Levey AI. Muscarinic acetylc holine receptor subtypes in cerebral cortex and hippocampus. Progressive Brain Research 145: 59-66. 2004 Warkentin S & Passant U. Functional imaging of the frontal lobes in organic dementia. Regional cerebral blood flow findings in nor mals, in patients with frontotemporal dementia and in patients with AlzheimerÂ’s disease, performing a word fluency test. Dementia Geriatrics Cognitive Disorders 8(2): 105-9. 1997 Webster S, Lue LF, Brachova L, Tenner AJ, McGeer PL, Terai K, Walker DG, Bradt B, Cooper NR, Rogers J. Molecular and cellula r characterization of the membrane attack complex, C5b-9, in AlzheimerÂ’s disease. Neurobiology of Aging 18:415-21. 1997 Weiser M, Rotmensch HH, Korczyn AD, Hartman R, Cicin-Sain A, Anand R; Rivastigmine-Risperidone Study Group. A pilo t, randomized, open-label trial assessing safety and pharmacokinetic parameters of co-administration of rivastigmine with risperidone in dementia patients with behavioral disturbances. International Journal of Geriatric Psychiatry 17(4):343-6. 2002 Welsh, KA. Detection and staging of demen tia in AlzheimerÂ’s disease. Use of the neuropsychological measures developed for th e consortium to Establish a Registry for AlzheimerÂ’s Disease. Archives of Neurology 49(5):448-52. 1992 Wellington CL. Cholesterol at the crossroads: AlzheimerÂ’s di sease and lipid metabolism. Clinical Genetics 66:1-16. 2004 Wen Y, Onyewuchi O, Yang S, Liu R, Simpkins JW. Increased -secretase activity and expression in rats following transient cerebral ischemia. Brain Research 1009:1-8. 2004 West MJ. Differences in the pattern of hippocampal neuron loss in normal ageing and AD. Lancet 344:769-72. 1994
108 Winkler DT, Bondolfi L, Herzig MC, Jann L, Calhoun ME, Wiederhold KH, Tolnay M, Staufenbiel M, Jucker M. Spontaneous Hemorrhagic Stroke in a Mouse model of Cerebral Amyloid Angiopathy. The Journal of Neuroscience 21(5): 1619-27. 2001 Wolozin, B. Cholesterol, statins and dementia. Current Opinions in Lipidology 15(6): 667-72. 2004 Wolozin B, Kellman W, Ruosseau P, Celesia GG, Siegel G. Decreased prevalence of Alzheimer disease associated with 3-hydr oxy-3-methyglutarul coenzyme A reductase inhibitors. Archives of Neurology 57(10): 1439-43. 2000 Wong TP, Debeir T, Duff K, Cuello AC. Reor ganization of choliner gic terminals in the cerebral cortex and hippocampus in transgen ic mice carrying mutate d presenilin-1 and amyloid precursor protein transgenes. The Journal of Neuroscience 19(7): 2706-2716. 1999 Xu C, Qian C, Zhang Z, Wu C, Zhou P, Li ang X. Effects of beta-amyloid peptide on transient outward potassium current of acute ly dissociated hippocampal neurons in CA1 sector in rats. Chinese Medical Journal 111(6):492-5. 1998 Zamrini, Edward. Imaging is superior to cognitive testing for early diagnosis of AlzheimerÂ’s disease. Neurobiology of Aging 25: 685-691. 2004 Zangara A. The psychopharmacology of huperz ine A: an alkaloid with cognitive enhancing and neuroprotective properties of interest in the treatment of AlzheimerÂ’s disease. Pharmacol Biochem Behav 75(3): 675-86. 2003 Zekanowski C, Religa D, Graff C, Filipek S, Kuznicki J. Genetic Aspects of AlzheimerÂ’s disease. Acts Neurobiologiae Experimentalis 64: 19-31. 2004. Zhilyaev SY, Moskvin AN, Platonova TF, Gu tsaeva DR, Churilina IV, Demchenko IT. Hyperoxic vasoconstriction in the brain is mediated by inactivation of nitric oxide by superoxide anions. Neuroscience Behavioral Physiology 33(8): 783-7. 2003