|USFDC Home | USF Electronic Theses and Dissertations||| RSS|
This item is only available as the following downloads:
Neuroinflammation in Alzheimers Disease: Characterization and Modification of the Response of Transgenic Mice to Intrahippocampal Lipopolysaccharide Administration by Donna Lorraine Herber A dissertation submitted in partial fulfillment of the requirements for the degree of Doctor of Philosophy Department of Pharmacology and Therapeutics College of Medicine University of South Florida Major Professor: Marcia N. Gordon, Ph.D. Dave Morgan, Ph.D. Javier Cuevas, Ph.D. Keith Pennypacker, Ph.D. Amyn Rojiani, MD, Ph.D. Date of Approval: December 10, 2004 Keywords: microglia, astrocyte, nicotin ic receptor, amyloid, dexamethasone Copyright 2005, Donna Lorraine Herber
Dedication This work is dedicated to the memory of Leo William Herber, Jr. We miss you Dad.
Acknowledgements All of the studies described he rein were a collaborative effo rt. I greatly appreciate the training I received upon entering the program from Mary Pacheco, Javier Cuevas, Giovanni DiCarlo, Donna Wilock, Chad Dick ey, and Emily Severance. Excellent technical support was also provided by Jessi ca Maloney, Lisa Roth, Menchu Barcenas, Keisha Symmonds, and Jennifer Alamed. Th e mouse colony was maintained by Nedda and David Wilson, Jeri Mason, a nd Melissa Freeman. I would like to say a special thank you to Marcia Gordon and Dave Morgan for pr oviding me with resources, support, and independence to conduct my research. This work was supported by NIH grants AG15490 and AG18478.
i Table of Contents List of Tables iii List of Figures iv Abstract vi Introduction 1 Chapter One: Biochemical and Histochemi cal Evidence of Nonspecific Binding of 7 nAChR Antibodies to Mouse Brain Tissue 10 Abstract 10 Introduction 11 Materials and Methods 13 Results 18 Discussion 31 Chapter Two: Acute and Chronic Microglial Infl ammatory Responses After Intrahippocampal Administra tion of Lipopolysaccharide 35 Abstract 35 Introduction 36 Materials and Methods 38 Results 45 Discussion 60 Chapter Three: Time-Dependent Reduction in A Levels After Intracranial LPS Administration in APP Transgenic Mice 65 Abstract 65 Introduction 66 Materials and Methods 68 Results 71 Discussion 82
ii Chapter Four: Dexamethasone Suppresses L PS-Induced Microglial Activation and Amyloid Clearance in APP Transgenic Mice 85 Abstract 85 Introduction 86 Materials and Methods 87 Results 93 Discussion 105 Conclusions 108 References 144 About the Author End Page
iii List of Tables Table 1 Primary Antibodies Used for Immunodetection (Ch. 1) 18 Table 2 Reverse Transcrip tion Master Mix and Cyc ling Conditions (Ch. 2) 41 Table 3 Primer Sequences for RNA Analysis (Ch. 2) 42 Table 4 Primary Antibodies Used fo r Immunohistochemistry (Ch. 2) 45 Table 5 Primary Antibodies Used fo r Immunohistochemistry (Ch. 3) 70 Table 6 Primary Antibodies Used fo r Immunohistochemistry (Ch. 4) 90 Table 7 Primer Sequences for RNA Analysis (Ch. 4) 92
iv List of Figures Figure 1. Immunolabeling by 7 antibodies in nontransgenic mouse brain 20 Figure 2. Reactive astrocytes colo calize with Congophilic plaques in APP+PS1 mice 22 Figure 3. Lipopolysaccharide stimulates astrogliosis and immunolabeling by anti7 antibodies 24 Figure 4. Genotyping a nd RNA analysis of 7+ and 7mice 26 Figure 5. Immunohistochemical analysis of 7+ and 7tissue 28 Figure 6. Western blot analysis of 7+ and 7tissue 30 Figure 7. RNA analysis by qRT-PCR 47 Figure 8. Time course of RNA expression 50 Figure 9. Time course of microglia l protein expression and morphology 53 Figure 10. Quantitation of immunohistochemical results for microglial markers 55 Figure 11. Time course of astr ogliosis in response to LPS 57 Figure 12. Time course of TLR4 levels after LPS injection 59 Figure 13. Dose response of LPS stimulated A removal 72 Figure 14. Glial response to LPS 74 Figure 15. Time-dependent removal of A by LPS injection 76 Figure 16. Time-dependent glia l reaction to LPS injection 79 Figure 17. Altered microglial mo rphology in response to LPS 81 Figure 18. LPS injection reduced diffuse but not compact amyloid deposits and was reversed by dexamethasone 96
v Figure 19. CD45 and CR3 are induced by LPS and inhibited by co-treatment with dexamethasone 99 Figure 20. Fc RII/III and SRA are induced by LPS but not inhibited by co-treatment with dexamethasone 102 Figure 21. LPS stimulated gene transcri pts have a pattern similar to protein expression 104 Figure 22. Autotoxic mechanisms in Alzheimers disease 114 Figure 23. Microglial activation states 121 Figure 24. Lipopolysaccharid e signaling cascades 123 Figure 25. Mechanism of LPS st imulated inflammation and A removal 131 Figure 26. Mechanism of LPS st imulated inflammation and A removal, and inhibition by dexamethasone 134
vi Neuroinflammation in Alzheimers Disease: Characterization and Modification of the Response of Transgenic Mice to Intrahippocampal Lipopolysaccharide Administration Donna Lorraine Herber ABSTRACT Alzheimers disease (AD) is pathological ly characterized by amyloid plaques, neurofibrillary tangles, inflammation, and neur odegeneration. According to the amyloid hypothesis of AD, the central mediating event of the disease is the deposition of amyloid. The inflammation hypothesis of AD states that it is the inflammatory response to plaques and tangles, rather than the actual lesions, wh ich causes the disease. Studies described here combine the two approaches into a single model. Four studies are presented using a basic protocol of intrahippocampal lipopolys accharide (LPS) injection to stimulate inflammation in transgenic mice. The firs t study looked at alpha 7 nicotinic receptors during the glial response to Abeta deposits and LPS. Reactive astrocytes which immunolabeled for alpha7 were co-locali zed with Congophilic de posits in APP and APP+PS1 mice, and increased after LPS injection. Unfortunately, LPS injection into alpha7 knock out mice revealed the alpha7 labe ling to be nonspecific. The second study evaluated the time course of protein and gene expression after LPS injection into nontransgenic mice. This experiment identified both a transient and chronic microglial inflammatory response, with changes in cell morphology. The third study evaluated a similar time course in APP mice. Concurrent with the inflammatory response, transient
vii reductions in Abeta burden were seen, though compact plaque load was unaffected. The fourth and final study used dexamethasone to inhibit LPS-induced inflammation in APP mice. LPS injection reduced Abeta burden, but was completely blocked by dexamethasone co-treatment. Though dexa methasone inhibited LPS-induced CD45 and complement receptor 3 levels (markers of general microglial activation), dexamethasone had no effect on scavenger receptor A or Fc gamma receptor II/III levels. An overall hypothesis regarding LPS mediated reductions in Abeta can be proposed: It is not the presence of the LPS molecule, nor the upregulation of receptors involved in phagocytosis, but rather general glial cell activation that medi ates Abeta removal. Thus, a phagocytic cell must not only bind Abeta ( by various receptors) but must also be capable of engulfing the materi al (via general cell activatio n). Taken together, these studies suggest that some leve l of inflammation in AD is be neficial and responsible for maintaining a balance between am yloid deposition and removal.
1 Introduction In the United States, there are two approved therapeutic approaches to Alzheimers disease (AD) therapy. The neurodegeneration seen in AD is mainly cholinergic, thus anticholinesterase drugs were developed to incr ease the half-life of acetylcholine in the synapse. These drugs temporarily overcome cholinergic dysfunction, but have no effect on neurodegeneration. Exci totoxicity has also been linked to the neurodegeneration of AD, thus n-methyl-d-asp artate (NMDA) receptor blockers have been developed. These two courses of thera py are only modestly e ffective long term, and other targets have been identified. At the ce nter of new developments are various means of amyloid removal coupled with anti-infl ammatory therapies. Amyloid-depositing transgenic mice are useful in developing work ing models of AD as well as targeting new therapies. Though amyloid deposition, inflammation, and cognitive impairment have been shown in various transgenics, there are limitations to the models, including a lack of neurofibrillary tangles, no apparent neuron lo ss, and minimal cholinergic deficits. It is possible that enhancing the inflammatory st ate in the brain of these mice would more closely resemble the pathology of AD. Li popolysaccharide (LPS, endotoxin) stimulates the innate immune response, leading to gliosis in the central nervous system and has been used historically to induce inflammation. The goals of the current studies are to evaluate and modify the response of transgenic mi ce to LPS administration. Inflammation would
2 potentially alter the glial response, amyloid load, c holinergic function, and neuronal survival of these mice. Alzheimers Disease Described by Alois Alzheimer almost a century ago, AD is pathologically distinguished by amyloid plaques, neurofibri llary tangles, and neuron loss (Alzheimer, 1907). The roles of neurofibrillary tangles and amyloid deposits in neurodegeneration and dementia of AD are controversial. Amyloid formation follows a series of st eps beginning with the amyloid precursor protein (APP), a membrane associated prot ein (reviewed in Nunan & Small, 2002). Proteolytic cleavage of APP by secretases yi elds a variety of peptide products, some harmless, others neurotoxic. Activity by gamma and beta secretases yields a 40 or 42 amino acid product known as the amyloid peptide (A 1-40 and A 1-42) that is released extracellularly. The A 1-42 is the more insoluble form, which can form oligomers, fibrils, and ultimately diffuse and compact plaques. The biochemical pathways associated with A neurotoxicity are diverse, including di rect interaction with neuronal cell surface proteins, and indirect actions of free radical production and glial activation (Canevari et al., 2004; Small et al., 2001; Walsh et al., 2002). Compelling evidence supporting the key role of A in the pathogenesis of AD are the genetic aberrations found in familial cases of the disease. Mutations in APP a nd the presenilins (PS1 and PS2) have been identified in early onset AD, leading to the development of animal models. Mice transgenic for human mutations of APP and PS1 have accelerated amyloid deposition and measurable cognitive deficits (Gordon et al., 2001). In these mice, anti-A antibody and nonsteroidal anti-inflammatory drug (NSA ID) therapies led to reductions in A and
3 behavioral improvements (Jantzen et al., 2001; Morgan et al., 2000; Wilcock et al., 2003, 2004a,b). Human trials using an active vaccination with A 1-42 are also showing promising results for halting the cognitive decl ine process, providing further evidence for the amyloid hypothesis of AD (H ock et al., 2003; Schenk, 2004). Neurofibrillary tangles are also abnormal deposits of insoluble fibrils which are formed in a similar aggregation cascade (Friedhoff et al., 2000). Tangle accumulation inside neurons begins with tau, a microtubule associated protein. Phosphorylation leads to dissociation of tau from the microtubules Hyperphosphorylated tau can polymerize, ultimately forming characteristic tangles whic h prevail even after the host neuron dies. Tauopathy is not exclusive to AD. Frontot emporal dementias and Pick disease also exhibit tangles, as well as memory loss, personality changes, and extrapyramidal symptoms (Morishima-Kaw ashima & Ihara, 2002). Neuronal loss in AD is grossly seen in the overall atrophy of the cortical and limbic systems of the human brain. Specifically the basal forebrain cholinergic system is affected with overall loss of nicotinic and muscarinic receptors, as well as decreased acetylcholine release and cholin e acetyltransferase activity (A uld et al., 2002). The exact mechanism of cell death is probably a comb ination of factors. Accumulation of A may lead to a localized autotoxic innate immunoreaction (McGeer & McGeer, 2002). Gliosis generates reactive oxyge n species, cytokines, and ch emokines, all potentially contributing to neurotoxicity (Akiyama et al., 2000). Direct action of A on neurons has also been hypothesized to cause excitotoxi city, followed by neuronal death (Canevari et al., 2004). The preferential cholinergic loss in the disease may be attributable to the ability of A to interact with alpha 7 nicotinic acetylcholine receptors (Bourin et al.,
4 2003). Current approved therapies for human use target cholinergic deficits with anticholinesterase action (Frisoni, 2001). More recent therapies aim to prevent neuronal death via NMDA receptor blockade, preven ting glutamate induced excitotoxicity (Reisberg et al., 2003). The role of inflammation in AD is complicat ed. At the center of the controversy is whether inflammation results from the pathol ogy of AD or contribute s to it. Complex interactions between the complement cas cade, cytokine and chemokine pathways, prostanoid mediated inflammation, and reac tive oxygen species may all contribute to neurotoxicity (Akiya ma et al., 2000). Cell mediators of the inflammatory process include the neurons themselves, microglia, and astrocytes. Microglia, the re sident macrophages of the brain (Kreutzberg, 1996), and astrocytes are activat ed in AD brain compared to non-demented controls, and are associated with amyloid deposits (Lue et al., 1996; Vehmas et al., 2003). In vitro experiments have demonstrated the ability of microglia to produce cytokines (interleukins IL1 and IL6, tumor necrosis factor TNF transforming growth factor TGF ) and reactive oxygen species when stimulated with A (Colton et al., 2000; Lue et al., 2001; Small et al., 2001). Immunohistochemical studi es in human AD brain verify the presence of inflammatory cytokines associated with amyloid deposits (reviewed in McGeer & McGeer, 1995). Supporting the role of inflammation in AD related cognitive impairment are epidemiological studies indicating decreased risk of AD associated with chronic nonsteroidal anti-inflammatory (N SAID) therapy (Gasparini et al., 2004; In t Veld et al., 1998) Animal models concur, demonstra ting decreased amyloi d load after NSAID
5 treatment (Eriksen et al., 2003; Jantzen et al ., 2002; Yan et al., 2003). Human trials have been less promising, possibly due to the shor t duration (6-12 months ) of such trials (Gasparini et al., 2004; Mc Geer et al., 1996; Szekel y et al., 2004). Alpha 7 Nicotinic Acetylcholine Receptors in AD Pathology Cholinergic dysfunction is a hallmark of AD neurodegene ration. Specifically, the basal forebrain cholinergic system is affected with overall loss of nicotinic receptors, as well as decreased acetylcholine release and choline acetyl tran sferase activity (Auld et al., 2002). The two main nicotinic acetylcholine receptors (nAC hRs) in the human brain consist of 4 2 or 7 subunits, with smaller populations of 2-9 and 3-4 (Paterson & Nordberg, 2000). 7 nAChRs are ligand gated ion channe ls consisting of five identical subunits, which conduct both sodium and calcium ions. They are distinguishable from other nAChRs by affinity for -bungarotoxin. In human AD brain, nicotinic receptor popul ations have been extensively studied. 7 nAChR levels are generally decreased in AD, though there are variations in reports (Perry et al., 2001). Declines in receptors may be due to actual cell loss, or synaptic decline might account for the abnormalities. Age related changes in 7 nAChRs expression in APP transgenic mice have also been reported (Dineley et al., 2001, 2002a,b). Focus has been on 7 nAChRs due to the potential interaction of the receptor with A peptides. Co-immunoprecipitation and receptor binding assays indicated A 1-42 binds 7 nAChRs with high affinity, but not 4 2 nAChRs, in both human and rat tissue (Wang et al., 2000a,b). Studies conducted with wild type 7 nAChRs indicate that A 1-42 blocks the receptor (Grassi et al., 2003; Lee et al., 2003; Liu et al., 2001; Pettit et al., 2001). Glial cells may also be i nvolved in the interactions of A 1-42 and 7 nAChRs as
6 recent reports have shown expression of thes e receptors by reactive astrocytes in human AD brain in association with amyl oid deposits (Teaktong et al., 2003). Transgenic Mice The pathological markers of AD are amyl oid plaques, neurofibrillary tangles, inflammation, and neuron loss. Although genetic aberrations can account for only 5-15% of all AD cases, transgenic models have been developed that provi de insight into the pathology. Four genes are asso ciated with amyloid deposition (reviewed in MorishimaKawashima & Ihara, 2002). Early onset familial AD has been associated with genes coding for APP, PS1, and PS2 (Goate et al .,1991; Levy-Lahad et al., 1995; Sherrington et al., 1995). Many APP mutations have been desc ribed, resulting in pr eferential cleavage by beta and gamma secretases, or by increasing formation of A 1-42, the more insoluble form of A The presenilins have been identified as components of the gamma secretase complex (De Strooper, 2003). In contrast, mu tations in PS1 or PS2 lead to enhanced activity of gamma secretase, t hus increasing production of A A susceptibility to AD has been associated with the apolipoprotein E4 allele, leading to increased amyloid deposition through an unknown mechanism (Cor der et al., 1993). Mutations coding for tau have been identified in frontal temporal dementia and Parkinsons disease, but not in AD. In humans these mutations result in the hyperphosphorylated state of tau, neurofibrillary tangles, and neuronal loss. Transgenic mouse models of AD are limited in their pathology, but are useful for studying amyloid deposition. Mutated form s of human APP when expressed in transgenic mice leads to an age related depos ition of amyloid and inflammatory gliosis. Mice transgenic for PS1 mutations do not show significant pathology. However, double
7 cross of the APP mouse line with PS1 yields accelerated amyloid deposition, inflammatory markers, down regulation of genes associated with learning and memory, and cognitive deficits, wit hout neurotoxicity (Dickey et al., 2003; Duff et al., 1996; Gordon et al., 2001, 2002; Holcomb et al., 1998 ). The discrepancy between these transgenics and human AD brain lies obviously in the lack of neuron loss and tangles. The specific role of the cholinergic system in AD may be explored using nicotinic receptor knockout mice. As previously discussed, the ability of A to interact with 7 nAChRs makes this ion channel of particular interest. Mice transgenic for a homozygous null mutation of the 7 nAChR gene are viable and fertil e, without any apparent gross or cellular nervous system disturbances, neur ological or behavior al dysfunction (OrrUrtreger et al., 1997; Paylor et al., 1998). A role for 7 nAChRs in learning and memory has been presumed, thus the lack of a beha vioral phenotype in these mice may be due to compensation by other populations. In vivo examination of 7 nAChRs during inflammationand A induced gliosis and neurotoxicity might be accomplished with these mice. Lipopolysaccharide as a Neuroinflammatory Agent Lipopolysaccharide (LPS) is a Gram-negat ive bacterial cell su rface proteoglycan, also known as bacterial endotoxin, which tri ggers an inflammatory response by the host (Palsson-McDermott & ONeil, 2004). LPS bi nds to circulating LPS binding protein, which subsequently transfers the LPS to soluble or cell membrane bound cluster differentiation marker CD14, a common recep tor for bacterial components (Heumann & Roger, 2002). The CD14 protein does not cont ain an intracellular domain, so interaction with additional components is likely. Toll like receptor 4 (TLR4), as well as myeloid
8 differentiation protein MD2, form a comple x through which LPS:CD14 can transduce a signal (Thomas et al., 2002). Multiple cascades are involved in the signal transduction, with nuclear factor NF B as a key player (Sen & Baltimore, 1986). Downstream gene activation leads to th e production of cytokines, reactiv e oxygen species, an d prostanoids stimulating the cellular inflammatory response. Historically, LPS has been used to mimic the inflammatory conditions seen in neurodegenerative diseases. Acute and ch ronic application of LPS into the brain ventricles of rats resulted in gliosis, cytokine production, cognitive deficits and neurotoxicity (Hauss-Wegrzyniak et al ., 1998a,b, 2000, 2002; Willard et al., 1999). Pharmacological intervention with steroids NSAIDs, antioxidants, or NMDA receptor antagonists attenuated some of the eff ects of LPS (Castano et al., 2002; HaussWegrzyniak et al., 1999a,b; Jain et al., 2002; Kheir-Eldin et al., 2001; Szczepanik et al., 2003; Wenk et al., 2000, 2002; Willard et al., 2000). With the introduction of amyloid de positing transgenics, LPS has been administered to accelerate A deposition, activate glia, and trigger neurotoxicity (the latter a key component of AD that is lack ing in these models). The experimental conditions of reports vary including the type and age of transgenic animals used; type, dose and route of administration of LPS; and pos t injection survival time. In some cases amyloid deposition was triggered (Qaio et al., 2001; Sheng et al., 2003; Sly et al., 2001), while in others amyloid clearance resulted (DiCarlo et al., 2001; Quinn et al., 2003). When considering the species of A 1-40 and A 1-42, the main indicator of A reduction seems to be closely related to plaque load and survival time. Aged animals with significant amyloid deposits, su rviving several days post injection, show ed clearance of
9 A deposits after LPS treatment. Compact plaques detected by either Congo red or thioflavin-S staining were inconclusively affected. In vitro studies show that microglia and as well as astrocytes react to LPS with phagocytic activity, as we ll as prostaglandin, cytokine, and nitric oxide produc tion (Kalmar et al., 2001). The ability of LPS to cause neuron deat h appears to be duration and region specific. Chronic infusion of LPS into the brai n ventricles of rats resulted in neuron loss limited to the forebrain cholinergic nucl ei (Hauss-Wegrzyniak et al., 1998a, 1998b, 2000, 2002; Willard et al., 1999). However, acute LPS injection into the brain parenchyma such as the anterior cortex and hippocampus did not triggered ce ll death (Kim et al., 2000; Herber et al., 2004a,b). An exception is the dopaminergic system which does show susceptibility to the degenerative effects of L PS (Kim et al., 2000; Castano et al., 2002). Taken together, stimulating an inflammato ry response in the br ain of transgenic mice should lead to a better understanding of the pathogenesis of AD. Studies described herein address the effects of inflammation on gliosis and amyloid burden in these mice.
10 Chapter 1 Biochemical and Histochemical Evidence of Nonspecific Binding of 7 nAChR Antibodies to Mouse Brain Tissue Abstract Alpha 7 nicotinic acetylcholine receptors ( 7 nAChRs) are involved in learning and memory, and are implicated in the pathology of Alzheimers disease and schizophrenia. Detection of 7 subunits can be accomplished via immunodetection or bungarotoxin binding techniques. In studies described here, sta ndard protocols for immunohistochemistry and West ern blotting were followed using several commercially available antibodies. Various mice were ev aluated including nontransgenics, APP, PS1, APP+PS1, and 7 knockouts. Initial results with amyloid depositing mice revealed 7 immunolabeled astrocytes, in addition to expe cted neuronal staining. Subsequent studies with intrahippocampal injecti ons of lipopolysaccharide into 7 knockout mice showed that both neuronal and astrocytic labeling by 7 antibodies was nonspecific. On Western blots of mouse brain proteins, none of the bands detected with antibodies directed against 7 subunits diminished in the 7 knockout mice. Although LPS related changes in the expression of some bands was found, these also were unaffected by the 7 genotype of the mice. In general, the Western staini ng patterns for these antibodies revealed few overlapping bands. These immunodetection data are in contrast to genotyping results and mRNA analyses which confir med the disruption of the 7 allele, and lack of 7 message
11 in the knockouts. These findings suggest cau tion in interpreting results using several commercially available 7 nicotinic receptor antibodies. Introduction Cholinergic dysfunction is a hallmar k of Alzheimers disease (AD) neurodegeneration. Specifically, the basal forebrain cholinergi c system is affected with overall loss of nicotinic receptors, as well as decreased acetylcholin e release and choline acetyltransferase activity (Auld et al., 2002). In vertebrates, the two main nicotinic acetylcholine receptors (nAChRs) in the brain are 4 2 and 7, with smaller populations of 2-10 and 3-4. 7 is a ligand-gated ion channel cons isting of five identical subunits, with a high relative calcium permeability, and distinguishable from other central nervous system nAChRs by its affinity for -bungarotoxin (McGehee & Role, 1995; Seguela et al., 1993). In human AD brain, nicotinic receptor popul ations have been extensively studied. 7 levels are generally decrea sed in AD, though there are vari ations in reports (reviewed in Perry et al., 2001). Declines in receptors ma y be due to cell loss or synaptic decline. Age-related changes in 7 expression in mice transgenic for a human mutated Alzheimer precursor protein (APP) have also been reported (Dineley et al., 2001; 2002a,b). Particular focus has been on 7 due to the potential interact ion of the receptor with A peptides. Co-immunoprecipitation and receptor binding assays indicated A 1-42 binds 7 with high affinity, but not 4 2, in both human and rat tissue (Wang et al., 2000). Studies conducted with wild type 7 receptors indicate that A 1-42 blocks the receptor (Grassi et al., 2003; Lee et al ., 2003; Liu et al., 2001; Pettit et al., 2001). Glial cells may also be involved in these interactions as recent reports have shown expression of 7 by
12 reactive astrocytes in human AD brain, particul arly in association with amyloid deposits (Graham et al., 2002; Wevers et al ., 1999; Teaktong et al., 2003). In these experiments, we used sta ndard protocols for immunostaining and Western blot analysis of 7 nAChR subunits in murine models of Alzheimers disease and neuroinflammation. Several commercially available antibodies were evaluated, and the results compared to genot yping and RNA analyses. Init ially, four genotypes resulting from breeding transgenic mice carrying eith er mutant APP or presenilin 1 (PS1) transgenics were examined for 7 expression. Various reports in the literature led us to expect decreases in levels of nicotinic recep tors in amyloid-depositing mice, such as the APP and APP+PS1 transgenics. Though no d ecreases in neuronal staining was seen, a surprising finding was 7-immunopositive astroglia in apposition with compact plaques in APP and APP+PS1 mice, but not PS1 or nont ransgenics. In order to determine if amyloid was causing the astrocytic 7 expression, or if it was part of a more general inflammatory response, we injected LPS in trahippocampally into APP and nontransgenic mice. Immunohistochemical analysis revealed many 7 positive astrocytes in the injected animals, leading us to believe the expression was part of a general inflammatory response. In a final experiment using 7 null mice, the specificity of the antibodies was tested under both control and LPS stimulated co nditions. No discernible differences were seen between 7+/+ versus 7-/mice with any antibody used, regardless of procedure. Genotyping and RNA analyses conf irmed the disruption of the 7 allele and lack of 7 message in the knockouts. We therefore conclu de that commercially available antibodies against 7 as used in the methods detailed here fa il to specifically detect the subunits.
13 Materials and Methods Mouse Strains Transgenic mice carrying an APP (Tg2576) and/or PS1 mutation were bred as described previously (Holcomb et al., 1998). Nontransgenic littermates were used as controls. Alpha-7-null mutant mice, origina lly described by Orr-Urtreger and colleagues in 1997, were derived from heterozygous breed ings of animals purchased from Jackson Laboratories (Bar Harbor, ME). Nontrans genic and homozygous null mice from this cross were used in this study. Animals were group-housed under a 12 hr light-dark cycle with free access to chow and water. Experi mental groups were balanced regarding age and gender. Genotyping APP and PS1 lines were anal yzed as previously repor ted (Gordon et al., 2001). Mice from the 7 null mutation line were genotyped as follows: 2 mm diameter ear clips from the 7 heterozygous breedings we re digested and the DNA extracted using Qiagens DNeasy Kit (Valencia, CA), per the manufactur ers instructions. Jackson Laboratories supplied the sequence of primers used to id entify either the neo-cassette of the null mutation or the wild type allele, for us e with the polymerase chain reaction (PCR): Forward 5'CCTGGTCCTGCTGTGTTAAACTGCTTC3'; Reverse-WT ( 7+) 5'CTGCTGGGAAATCCTAGGCACACTTGAG3'; Reverse-Neo ( 7-) 5'GACAAGACCGGCTTCCATCC3'. Thermocyc ling conditions were as follows: 95 C for 4 minutes; 30 cycles of 95 C for 15 seconds, 60 C for 60 seconds, 72 C for
14 60 seconds; 72 C for 7.5 minutes; store at 4 C. PCR products were run on a 1% agarose gel, using ethidium bromide ultraviole t (UV) detection of bands at 440 bp ( 7+) or 750 bp ( 7-). Intrahippocampal Injections APP mice aged 16 months, 7 null mice aged 11 months, and their nontransgenic littermates were used in these studies. Mi ce were anesthetized using isoflurane and immobilized in a stereotaxic apparatus. One microliter injections of either saline or 4 g/ l lipopolysaccharide ( Salmonella abortus equi Sigma, Saint Louis, MO) were delivered over a two minute period into the hippocampus (stereotaxic coordinates from bregma: -2.7 mm posterior; +/-2.5 mm lateral; -3.0 mm ventral). This procedure had been previously demonstrated in our lab to induce a neuroinflamma tory response without adversely affecting animal survival (DiC arlo et al., 2001). All animal work was conducted under National Institute of Health guidelines, and approved by the University of South Floridas institutional animal car e and use committee. Animals were singly housed for the 7 day post treatment survival period under standard vivarium conditions. Tissue Preparation Mice were anesthetized with pentobar bital (200 mg/kg, ip), then perfused transcardially with 25 ml of saline. The righ t hemisphere of the brain was dissected into regions and stored at 80 C for subsequent biochemical analyses. For immunohistochemistry, left hemispheres were transferred into a 4% paraformaldehyde solution for 24 hr, then processed through a cryprotection schedule of 10, 20, then 30% sucrose (24 hr in each solution). The ti ssue was sectioned horiz ontally on a sliding
15 microtome at 25 m. Sections were then stored in Dulbeccos phosphate buffered saline pH 7.4 (DPBS) with sodium azide (100 mM) at 4 C. RNA Analysis Mice obtained from the 7 heterozygous breedings were analyzed for mRNA using reverse transcription followed by PCR. The procedure was originally described in detail elsewhere (Dickey et al., 2003). RNA was extracted from the injected hippocampus of 7+ and 7mice, using Qiagens Rneasy procedure (Valencia, CA). RNA concentration was determined with Molecular Probes RiboGreen RNA quantitation kit (Molecular Probes, Eugene, OR). Reverse transcription with mMLV (Invitrogen, Carlsbad, CA) was performed, a nd the resulting cDNA subjected to PCR using Amplitaq Gold (Applied Biosystems, Foster City, CA). Primers used to identify 7 were directed towards portions of ex ons 9-10: F5GTGGGCCTCTCAGTGGTCGT3; R5GTCCCCATCAGAGGGGTGTG3. Thermocy cling conditions were as follows: 95 C for 3 minutes; 45 cycles of 95 C for 15 seconds, 60 C for 60 seconds, 72 C for 60 seconds; 72 C for 7.5 minutes; store at 4 C. PCR products were run on a 1% agarose gel, using ethidium bromide UV detection of bands at 381 bp. Histology Immunohistochemical analysis of 7 nAChRs was performed using 25 m free floating sections spaced 300 m apart through the hippocampus. Details of this procedure were originally described else where (Gordon et al., 1997). All steps were performed on a rotating platform at approxi mately 40 rpm, room temperature, unless otherwise stated. Sections were blocked for endogenous pe roxidases (10% methanol,
16 3% hydrogen peroxide in DPBS) for 15 minutes, then washed 3 x 5 minutes with DPBS. The tissue was then permeabil ized in a solution of 100 mM lysine, 0.2% triton x-100, and 4% normal goat or horse serum (Pel Freeze, Rogers, AK) in DPBS for 30 minutes. Sections were then incubate d overnight in the appropriate primary antibody in DPBS and 4% serum, without shaking (Table 1). The following day, sections were incubated with shaking in primary antibody for one hour. Sectio ns were then washed 3 x 5 minutes with DPBS, and then incubated in appropriate biotinylated second ary antibody (Vector Laboratories, Burlingame, CA) at a concentration of 0.5 g/ml in DPBS and 4% serum, for two hours. Sections were washed 3 x 5 minutes with DPBS, and then incubated for one hour in Vectastain Elite ABC solution (Vector Labora tories, Burlingame, CA), using 8 drops each of components A and B pe r 100 ml of DPBS. Sections were then washed 2 x 5 minutes with DPBS, followed by a single wash in tris buffered saline, pH 7.6 (TBS) for 5 minutes. The tissue was then incubated with a solution of 0.5% nickelous ammonium sulfate hexahydrate and 0.05% diam inobenzidine in TBS for 5 minutes. Color development was achieved by the addition of 0.03% hydrogen peroxide, and incubation for an additional 5 minutes, followed by three final washes. Controls for nonspecific binding of the secondary anti body were performed by excluding primary antibodies. Stained sections were mounted onto slides and air dried overnight. Slides were then processed through a dehydration sche dule of 10 dips in water, followed by 2 x 3 minutes in each of 25%, 50%, and 75% ethanol, then 3 x 5 minutes in each of 95% ethanol, 100% ethanol, and Hist o-Clear (National Diagnostics, Atlanta, GA). Slides were cover slipped with DPX (E.M. Sciences, Fort Washington, PA) and allowed to dry overnight.
17 In some cases, tissue stained with 7 antibodies was count erstained with Congo red to verify location of compact plaque de posits. For this counterstain, immunostained sections were slide mounted and allowed to dr y overnight. The slides were then hydrated for 30 seconds in water. Slides were then submerged for 20 minutes in a solution of 80% ethanol supersaturated with sodium chlo ride, then made alkaline with a final concentration of 0.01% sodium hydroxide. Slid es were then incuba ted for 30 minutes in a separate portion of alkaline alcoholic sa turated sodium chloride containing 0.2% Congo red dye (solution filtered prior to use). Slid es were then dehydrated with 8 dips in 95% ethanol, then 8 dips in two ba ths of 100% alcohol. The tissue was finally run through 3 x 5 minutes of xylene and c over slipped with DPX. Western Blotting Mice obtained from the 7 heterozygous breedings we re analyzed for protein using SDS-PAGE followed by Western blot. Ce rebral cortex previ ously stored at 80 C was homogenized in 10 mM HEPES buffer pH 7.4 containing a protease inhibitor cocktail (Roche, Indianapolis, IN). Crude pr otein concentrations were determined by the Bradford method using Bio-Rad Protein Assa y Dye Reagent (Hercules, CA). Samples were denatured with Bio-Rad Laemmli sample buffer by boiling for 5 minutes. 10 g of protein was loaded per well, and proteins separated using 10% polyacrylamide Bio-Rad Ready Gels. Bio-Rad Precision-Plus Protein All Blue molecular weight standards were run for band identification. The separated proteins were transferred to Immobilon -P polyvinylidine fluoride membranes and imm unoblotted (Millipore, Bedford, MA). Membranes were first blocked with 5% nonfat dry milk in borate-buffered saline pH 8.5 and 0.05% Tween-20 (BST) for one hour on a rock ing platform at room temperature.
18 Membranes were then incubated with prim ary antibody in 2.5% nonfat dry milk in BST for one hour (Table 1). After washing 3 x 5 mi nutes with BST, blots were incubated with appropriate horseradish pe roxidase-conjugated seconda ry antibody (Santa Cruz Biotechnology, Santa Cruz, CA) at a 1:10000 d ilution in 2.5% milk-BST for 30 minutes. Finally, membranes were triple washed in BS T. Bands were identified using Western Blotting Luminol Reagent (Santa Cruz Biotechnology, Santa Cruz, CA) for chemiluminescent detection and subsequent film exposure for 0.5-5 minutes. The presence of 7 protein was verified by comparing the protein bands to the molecular weight standard markers. The expected molecular weight of the 7 subunit was 56 kDa. Table 1. Primary Antibodies Used for Immunodetection Clone Type Source Immunohistochemistry Titer Western Blot Titer 319 Rat monoclonal Sigma 1:5000 1:1000 306 Mouse monoclonal Covance 1:3000 1:50 H-302 Rabbit polyclonal Santa Cruz 1:200 1:100 Results Standard protocols for immunohistochemi cal and Western blot detection of 7 nAChR subunits were followed using several commercially available antibodies. These results were ultimately compared to genotyping and mRNA analyses of the 7 subunit. Initially, four genotypes resulting from br eeding of APP and PS1 transgenics were
19 examined for 7 expression. Previous data had been published demonstrating the amyloid burden, gliosis, and behavioral phenotype of these mice (Gordon et al., 2001, 2002). The antibodies used for the curre nt study included a rat monoclonal antibody (mAb 319), a mouse monoclonal (mAb 306) a nd a rabbit polyclonal (pAb H-302). In nontransgenic mice (Fig. 1), distinct ubiqu itous positive immunostaining was seen throughout the brain with mAb 319. Regions analyzed included the dentate gyrus of the hippocampus, frontal cortex, stri atum, and cerebellum. These results are consistent with previous reports (Dominguez del Toro et al., 1994). Figure 1. Immunolabeling by 7 antibodies in nontransgenic mouse brain. Horizontal sections were immunostained using mAb 319 at a 1:5000 dilution. Ubiquitous neuronal stain was seen in all regions examined in cluding the dentate gyrus of the hippocampus (A), cerebral cortex (B), striatum (C), and cerebel lum (D). Scale bar = 100 m.
20 Figure 1. Immunolabeling by 7 antibodies in nontransgenic mouse brain.
21 Mice transgenic for APP+PS1 were next an alyzed. Distinct neuronal staining was noted with mAb 319 (not shown), similar to th at seen in nontransgenics. A comparison of nontransgenic versus APP+PS1 mice reveal ed no differences in the intensity of neuronal staining in any region. In amyloid bearing regions such as the hippocampus and cortex, immunolabeling of reactive astroc ytes was also noted. The cells were deemed astrocytes based on their size, mor phology, and distribution within the laminated regions of the hippocampus. The 7 immunopositive astrocytes seemed to cluster around amyloid deposits, which was confirmed when br ain sections were c ounter stained with Congo red (Fig. 2). This micrograph shows mAb 319 positive cells associated with a Congophilic plaque. APP mice showed similar patterns of astrocytic staining, but to a lesser degree, possibly due to the lower amyl oid burden in these mice (data not shown). PS1 mice resembled nontransgenics; rarely were 7 immunolabeled astrocytes seen in those mice.
22 Figure 2. Reactive astrocytes co-loca lize with Congophilic pl aques in APP+PS1 mice. Horizontal sections were immunostained using mAb 319 at a 1:5000 dilution, and then counterstained with Congo red. Micrographs sh ow the molecular layer of the CA3 region of the hippocampus. Black staining of the ce ll bodies is clearly defined, with faint staining of processes, all surrounding the red plaque core. Scale bar = 50 m.
23 To determine if 7 expression by reactive astrocyt es was due to a generalized inflammatory response to amyloid, we inje cted LPS into the hippocampus of APP and nontransgenic mice. LPS treatment resu lted in a diffuse pattern of mAb 319 immunolabeling of astrocytes, cells known to react to LPS injections (Hauss-Wegrzyniak et al., 1998). No changes were seen in neuronal staining with mAb 319 (Fig. 3). Shown are the CA3 regions of saline injected (panel s A,C) versus LPS injected mice (panels B,D). A few immunopositive astrocytes were s een in saline treated APP mice (panel C), but not in the nontransgenics (panel A). Both nontransgeni cs (Fig. 3B) and APP (Fig. 3D) mice showed reactive 7 immunopositive astrocytes upon LPS treatment. Figure 3. LPS stimulates astrogliosis and immunolabeling by anti-7 antibodies Nontransgenic (A,B) and APP (C,D) mice were injected with saline (A,C) or LPS (B,D) 7 days prior to sacrifice. Horizontal sections were im munostained using mAb 319 at a 1:5000 dilution. Micrographs show the CA 3 region of the injected hippocampus. Pyramidal cells were intensely stained, located in the upper left corner of each panel. LPS treatment resulted in widespr ead activation of astrocytes which were immunolabeled by mAb 319 (B,D). Scale bar = 50 m. Abbreviations: NT = nontransgenic, APP = mutant amyloid precursor protein transgenic, LPS = lipopolysaccharide.
24 Figure 3. LPS stimulates astrogliosis and immunolabeling by anti-7 antibodies
25 In a final set of experiments, LPS was injected into 7 null mice, thus testing the specificity of the anti7 antibodies. Mice carry ing a null deletion for 7 were genotyped and selected to be either nontransgenic ( 7+) or homozygous null ( 7-). Genotyping results indicated a single band at 440 bp ( 7+) or 750 bp ( 7-). Figure 4A shows typical PCR product banding for nontransgenic and nul l genotypes. All mice were genotyped and sorted based on these results. Su bsequently, RNA was extracted from the hippocampus, reverse transcribed, and s ubjected to PCR to determine whether 7 mRNA was expressed. Mice that had been genotyped as 7-, did not express 7 mRNA. Typical samples are shown in Figure 4B, with a positive signal at 381 bp in lane 1 (nontransgenic), as well as a smaller prime r-dimer product running at the end of the gel (present even in the absence of cDNA). All animals genotypically 7+ also had 7 mRNA; none of the 7 null mutant mice had 7 mRNA.
26 Figure 4. Genotyping and RNA analysis of 7+ and 7mice. Panel A shows a typical PCR product with ethidium bromide/UV det ection of genomic DNA. Lanes 1-2 are from an 7+ mouse, showing only a 440 bp band in la ne 1 indicating the wild type allele, but no band for the neo-cassette primers in la ne 2. Lanes 3-4 are from a homozygous null mutant 7mouse, lacking the band for the wild type primers in lane 3, showing only a 750 bp band indicating the neo-ca ssette disrupted allele in lane 4. Panel B shows a typical RT-PCR product derived from mRNA and detected with ethidium bromide/UV. Both lanes show a small product due to prime r-dimer formation. Lane 1 shows a single specific band of the expected size of 381 bp from an 7+ mouse. Lane 2, from a homozygous null mutant 7mouse, shows no specific bands.
27 Immunohistochemistry was perfor med using tissue from both 7+ and 7mice, with or without LPS injection. Three diffe rent commercially avai lable antibodies were used (Table 1). Shown in Figure 5 are three anti7 antibodies (mAb 319 in panels A-B; mAb 306 in panels C-D; pAb H-302 in panels E-F). Each antibody showed intense, ubiquitous labeling of neurons throughout the br ain; cortical neurons are shown in this figure. No difference in the staini ng patterns was seen between untreated 7+ (Panels A,C,E) and 7mice (Panels B,D,F). All three antibodies immunolabeled neurons in both genotypes. LPS injection into the hi ppocampus caused reactive astrogliosis similar to that seen in Figure 3 (d ata not shown). Again, mAb 319 labeled astrocytes in both 7+ and 7mice. Astrocytes were not labeled by mAb 306 nor pAb H-302 in either genotype or treatment. Thus, regard less of treatment or antibody used, both 7+ and 7mice showed distinct nonspecific labeli ng of neurons (mAbs 306, 319, pAb H-302) as well as astrocytes (mAb 319).
28 Figure 5. Immunohistochemical analysis of 7+ and 7tissue. Horizontal sections were immunohistochemically an alyzed using various anti7 antibodies. Micrographs show neurons in the frontal cortex. Ubiquit ous staining of cell bodies was seen with all antibodies. Some cytoplasm and proce sses are stained with mAb 306 (C-D, 1:3000 dilution) and pAb H-302 (E-F, 1:200 dilution) mAb 319 primarily stained neuronal cell bodies and nuclei (A-B, 1:5000 dilution). Neurons in both 7+ (A,C,E) and 7tissue (B,D,F) were immunolabeled by these anti7 antibodies. Scale bar = 50 m.
29 Antibody staining patterns were also exam ined by Western blot, using cortical tissue from both genotypes, with or without LPS injection. The exposure times were increased to maximize detection of faint ba nds. Although most antibodies showed many nonspecific bands, no antibody revealed differe nces in protein bands in untreated 7+ versus 7mice at the expected molecular weig ht of 56 kDa (Fig. 6). LPS injection caused upregulation of an unknown protein in both 7+ and 7mice at approximately 45 kDa (mAb 306) and 75 kDa (mAb 306, pAb H-302) No additional bands were seen with mAb 319 after LPS treatment. However, none of the antibodies detected differences between LPS treated 7+ versus 7mice.
30 Figure 6. Western blot analysis of 7+ and 7tissue Cortical tissue was homogenized, subjected to SDS-PAGE, tran sferred and immunoblot ted. ECL detection was captured with subsequent film exposure. The scanned image shown contains typical samples of alternating lanes of 7+ tissue and 7tissue. Tissues from both untreated control mice and LPS injected mice were run for each genotype and antibody. A primary band at 45 kDa, and a secondary band at 30 kDa was seen with mAb 319 (lanes 1-4). Strong bands at 30, 50, and 75 kDa were seen w ith mAb 306 (lanes 5-8). Strong bands at 30, 45, 75, and 100 kDa were evident with pAb H-302 (lanes 9-12). Both nontransgenic 7+ and null mutant 7tissue were immunolabeled by these anti7 antibodies. LPS treatment resulted in an increase in immunolabeling by mAb 306 (bands at 50 and 75 kDa) and pAb H-302 (bands at 75 kDa). Th is increase in staining was seen in both 7+ and 7tissue.
31 Discussion Three different commercially available antibodies intended to label 7 nAChR subunits were used in standard immunohist ochemical and immunoblotting protocols. The antibodies were generated using diffe rent host species, and different 7 subunit sequences, though all products are direct ed towards portions of exon 10. Previous reports have shown the utiliza tion of monoclonal antibodies from clones 306 and 319 for immunohistochemical analysis of rat brain (Dominguez del Toro et al., 1994). The results reported by Dominguez del Toro indicated ubiquitous neuronal expression, similar to what we saw in our studies. Both the Dominguez del Toro procedure and ours used paraformaldehyde fixation of the brain prior to sucrose cryoprotection. Both procedures used 25 m free floating sections, an avidin-biotinperoxidase protocol, and DAB color development. Th e immunohistochemical protocols used in our experiments were developed to yield significant positive stain while minimizing background. Various concentratio ns and combinations of primary (1:1001:10000) and secondary antibodies (1:1000-1:10000) were tested in an effort to identify conditions that stained the 7+/+, but not the 7 null mice, without success. Tissue perfusion (saline versus paraformaldehyde) a nd post fixation times (2-24 hr) were also evaluated separately in nontransgenic 7+ mice. No differences in the immunolabeling patterns were seen with any of these variations. It is co nceivable that the antigenic determinant(s) was modified by the paraformaldehyde treatment (Montero, 2003). Another possibility is that the levels of 7 may be so low, or the amount of antibody needed to yield a signal so high, that cross-reacting protein binding masked any 7 positive cells. Other researchers that have worked with the 7 null mice did not use
32 immunohistochemistry to demonstrate the absen ce of this receptor, but rather showed the absence of -bungarotoxin binding (Orr-Urtreger et al. 1997; Franceschini et al., 2002; Wang et al., 2003). Using our paraformaldehyde treated tissue, we failed to detect specific labeling using fluorophore conjugated -bungarotoxin (data not shown). The immunolabeling of reactive astrocyt es in amyloid depositing mice by mAb 319 was an interesting finding of these studies. Subsequent studies wi th LPS injection in 7 knockout mice revealed this labeling to be nonspecific. Recent reports by Teaktong and colleagues (2003) showed immunolabeled 7+ astrocytes in human AD brain. Cholinergic signaling by rat astrocytes ha s also been demonstrated (Sharma and Vijayaraghavan, 2001). Activation of these cel ls produced calcium flux that was blocked by -bungarotoxin, indicating the presence of 7 nAChRs. The potential for some of the astrocyte 7 labeling to be specifi c should not be ruled ou t. Thus, in designing experiments to evaluate 7 expression an alternative approach would be the use of bungarotoxin instead of anti7 antibodies. However, we were unable to develop adequate -bungarotoxin labeling in formaldehyde fixed sections. Several reports have demonstrat ed Western blot analysis of 7 in rodent brain extracts using monoclonal antibodies from clones 306 and 319 (Schoepfer et al., 1990; Dominguez del Toro et al., 1994; Orr-Urtreger et al., 1997; Dineley et al., 2001; FabianFine et al., 2001). These reports have lis ted the apparent mol ecular weight of the 7 subunit ranging from 48-72 kDa, though the calc ulated molecular weight is 56 kDa. Some of the investigators used standa rd homogenization procedures, SDS-PAGE separation, transfer, and subs equent immunoblotting, similar to the procedure described here. Other investigator s affinity purified the 7 nAChRs with cobratoxin or
33 bungarotoxin prior to SDS-PAGE, resulting in a primary band at approximately 56 kDa (Dominguez del Toro et al., 1994; Orr-Urt reger et al., 1997). The immunoblotting protocols used in our laboratory were developed to minimize nonspecific bands and background. Various concentrations and combinations of primary and secondary antibodies, ranging from 1:50-1:10000, were te sted in an effort to come up with conditions that differentiated between the 7+ and the 7 null mice, without success. Separately, various homogeniza tion buffers (10 mM HEPES 1-2% triton-x 100, 50-150 mM PBS 1-2% triton-x 100, 50 mM TBS 1-2% triton-x 100), vendors of secondary antibodies (Vector, Santa Cruz), blocking solutions (BST + 2.5-5% milk, TBST + 2.55% milk), and subcellular fractionation prep arations ( crude, membrane, and solubilized fractions) were tried, but failed to reveal differences in the st aining pattern between 7+ and 7 null mice. Other researchers that have performed Western blot analysis of 7 null tissue did not use commercially-available an tibodies, and also affinity purified the 7 subunit prior to analysis (Orr-Urtreger et al 1997; Franceschini et al., 2002; Wang et al., 2003). Again, an alternative approach in designing experiments to evaluate 7 expression would be the use of -bungarotoxin affinity purifi cation. However, the large amounts of tissue needed for affinity purific ation complicate the pr ocedure and preclude analysis of small brain regions (such as the hippocampus) in individual rodents. There have been reports of a partial duplication of the human 7 nAChR gene, which has four novel N-terminus exons a nd conserved exons 5-10 (Gault et al., 1998; Villiger et al., 2002). Such duplication has not been reported in rodents, but cannot be completely ruled out. If there were such duplication, it might account for the 7 protein detection in the absence of 7 gene expression. Still, one would expect at least a
34 quantitative difference in the amounts of st ained material. Alternatively spliced 7 mRNAs have also been show n in mice (Saragoza et al., 2003). The resultant mRNA includes a novel exon 9b. Additionally, an other splice variant with novel exon 4b has been reported in rat (Sever ance et al., 2004). The Sarago za variant could potentially interfere with our RNA analysis, as the primer s are designed to prime to exons 9-10. In contrast, the Severance variant does contain exons 9-10 and should be eliminated in the knock out mouse. However, alternatively spliced mRNAs would ha ve no effect on the genotyping results as exons 8-10 are interru pted in the knock out. Moreover, the complete absence of 7 mRNA in the null mice would require some mutation in any duplicated 7 gene which would disrupt the primer pairs from annealing. We find such circumstances unlikely to account for the results we have obtained here. In conclusion, careful examination of protocols will be required in order to draw any conclusions made from immunodetection studies of 7 nAChRs. Localization of the 7 subunit with immunohistochemistry must be interpreted with cau tion. Confirmation with -bungarotoxin binding experiments is recommended, as well as RNA analysis where applicable.
35 Chapter 2 Acute and Chronic Microglial Inflammatory Responses After Intrahippocampal Administra tion of Lipopolysaccharide Abstract Inflammation has been argued to play a primary role in the pathogenesis of neurodegeneration in Alzheimers disease. The inflammatory response can be either beneficial or harmful, depending partic ularly upon the dura tion of the reaction. Lipopolysaccharide (LPS) activates the innate immune response and triggers gliosis when injected into the central nervous system. In studies describe d here, we evaluated the time course of microgliosis after acute intraparenchymal admini stration of LPS. Mice were injected bilaterally into the hippocampus with 4 g of LPS. Post injection survival times were 1, 6, and 24 hr, as well as 3, 7, 14, and 28 days. Protein and RNA analyses were performed for inflammatory markers. Signi ficant elevations of cluster differentiation marker CD45, glial fibrillary acidic protein (GFAP), scave nger receptor A (SRA), and Fc receptor mRNA were seen after 24 hr. Immunohistochemistry revealed a complex pattern of protein expression by microglia coupled with changes in morphology. RNA and protein for Fc receptor, GFAP, and SRA were transiently elevated, peaking at 3 days, and returned to basal levels after a week. In contrast, microglia remained significantly activated through the 28 day time point as determined by CD45 and complement receptor 3 levels. These findings indicated that the inflammatory response
36 had both an acute and chronic response to L PS that was mediated primarily by microglia in the central nervous system. Introduction Neuroinflammation is a consistent pat hological event in many neurodegenerative diseases including Alzheimers di sease (AD). AD is of particul ar interest due to several epidemiological studies implicating NSAID use with lower levels of AD risk (McGeer & McGeer, 1996; In t Veld et al., 1999). The inflammatory response of the periphery is well characterized, but until recently, the brain was considered a privileged organ, without much of an immuno-inflammatory system. We now know that the resident macrophages of the brain, microglia, coordinate local inflammation as well as the innate immune response (Streit, 2002). The inflammatory response in AD encompasses components of the complement system, cytokines and chemokines, reactive oxygen species, prostaglandins, and a host of others (Akiyama et al., 2000). Lipopolysaccharide (LPS) has traditionally been used to stimulate the innate immune response. LPS is a Gram-negativ e bacterial cell surface proteoglycan, also known as bacterial endotoxin, which triggers an inflammatory response by the host. LPS binds to circulating LPS binding protein, wh ich subsequently transfers the LPS to membrane bound cluster differentiation (C D) marker CD14, a common receptor for bacterial components (Heumann & Roger, 2002). The CD14 protein does not contain an intracellular domain, so interaction with ad ditional components is likely. Toll like receptor 4 (TLR4), as well as myeloid diffe rentiation protein (MD2), form a complex through which LPS:CD14 can transduce a signal (Thomas et al., 2002). Multiple cascades are involved in the signal transduction, with nuclear factor NF B as a key
37 player. Downstream gene activation triggers the production of cytoki nes, nitric oxide, and prostanoids leading to the cellular inflammatory response. The role of inflammation in the brain is controversial. The phagocytic capabilities of microglia have been exploited in amyloid depositing transgenic mice, aiding in the removal of A deposits during passive imm unization (Wilcock et al., 2003, 2004a,b) as well as after activation of the innate immune system (DiCarlo et al. 2001, Herber et al., 2004b). However, the ability of activated glia to produce cytokines, chemokines, and other noxious products may lead to neuronal damage. Many in vitro studies have demonstrated that LPS-stimulat ed microglia cause neuron death (Chao et al., 1992; Kim et al., 2000; Lehnard t et al., 2002). However, only under chronic activation in vivo has LPS been shown to induce enough in flammation to cause basal forebrain neurodegeneration (Hauss-Wegrzniak et al., 2002). An exception is th e acute injection of LPS into the substantia nigra, which caused cell death, possibly due to the high ratio of microglial per neuron in that brain region (Kim et al., 2000). In our studies, we have combined traditional immunohistochemistry with quantitative real time polymerase chain re action (qRT-PCR) to better understand the inflammatory response to LPS in the brain. We injected LPS into the hippocampus of young mice and subsequently measured mRNA a nd protein for markers of inflammation and phagocytosis. We demonstrate the time course of this response, and the involvement in particular of microglia.
38 Materials and Methods Mouse Strains Nontransgenic mice obtained during the breeding of our APP+PS1 transgenic mouse colony (Holcomb et al, 1998) aged 5-6 months were used in this study, with 6 animals per experimental time point, for a to tal of 48 mice. Animals were group-housed under a 12 hr light-dark cycle with free access to chow and water. Experimental groups were balanced regarding gender. Intrahippocampal Injections Mice were anesthetized using isoflura ne and immobilized in a stereotaxic apparatus. One microl iter injections of 4 g/ l lipopolysaccharide (Salmonella abortus equi, Sigma, Saint Louis, MO) were deliv ered over a two minute period into both hippocampi (stereotaxic coordinates from br egma: -2.7 mm posterior; +/-2.5 mm lateral; -3.0 mm ventral). This procedure had been prev iously demonstrated in our lab to induce a neuroinflammatory response without adversel y affecting animal survival or causing neurotoxicity (DiCarlo et al., 2001). The c ontrol group consisted of untreated mice. LPS treated mice survived 1, 6, 24, 72, 168, 336, or 672 hr post injection. Animals were singly housed for the post treatment survival period under standard vivarium conditions. All animal work was conducted under National Institu te of Health guidelines, and approved by the University of South Fl oridas institutional animal care and use committee. Tissue Preparation Mice were anesthetized with pentobar bital (200 mg/kg, ip), then perfused transcardially with 25 ml of saline. Th e brain was removed, the right hippocampus was dissected, and the tissue was then stored at 80 C for subsequent bioc hemical analyses.
39 For immunohistochemistry, left hemispheres we re transferred into a 4% neutral buffered paraformaldehyde solution for 24 hr, then pr ocessed through a cryopr otection schedule of 10, 20, then 30% sucrose (24 hr in each soluti on). The tissue was sectioned horizontally on a sliding microtome at 25 m. Sections were then stored in Dulbeccos phosphate buffered saline pH 7.4 (DPBS) with sodium azide (100 mM) at 4 C. RNA Analysis Tissue from the right hippocampus wa s analyzed for mRNA expression via reverse transcription followed by qRT-PCR, as described by Dickey and colleagues (2003). RNA was extracted from the inje cted hippocampus using Qiagens Rneasy procedure (Valencia, CA). The hippocampus was homogenized in 700 l of RLT buffer for 30 seconds using an electr onic rotor-stator. The manuf acturers protocol for RNA preparation was followed as indicated, including the optional on-column DNAse digestion. Sample RNA was then as sayed using Molecular Probes RiboGreen RNA quantitation kit (Molecular Probes, Eugene, OR ). A series of standard RNA solutions were prepared ranging from 20-1000 ng/l. Samples and standards were prepared in triplicate, using 1 l of sample per 200 l reaction volume. Fluorescence was monitored at 485 nm excitation and 538 nm emission wa velengths. Each sample RNA was then diluted in water to a final concentration of 50 ng/l and 5 ng/l. The 50 ng/l samples were run on a 1% agarose gel using ethi dium bromide/ultraviolet detection. A composite of all RNA samples was made in order to prepare a standard curve for qRT-PCR analysis. A pool consisting of 2 l of each 50 ng/l sample was made. Serial dilutions of this composite standard were made down to 0.2 ng/l to construct a standard curve. Reverse transcription (RT) with mMLV (Invitrogen, Carlsbad, CA) was
40 then performed on the standard dilutions and 5 ng of each sample RNA solution, using the master mix and conditions detailed in Table 2. The resulting cDNA was then subjected to qRT-PCR using SYBR Green PCR Master Mix (Molecular Probes, Eugene, OR). Table 3 contains the primer sequences and cycling conditions used in the PCR reaction. The sequence for scavenger recept or B is identical to that listed for CD36 as listed in GenBank; note that the same se quence is listed for SR-B1 and various other pseudonyms. For each reaction, 12.5 l of SYBR Green, 7.5 l water, 1.5 l of each primer, and 2 l of the 5 ng sample dilu tions were run usi ng an MJ Research DNA Engine Cycler with Opticon Detector (MJ Research, Inc., Waltham, MA). Quantitation of several mRNAs was perfor med using the cDNA generated by reverse transcription.
41 Table 2. Reverse Transcription Ma ster Mix and Cycling Conditions Master mix reagent Stock concentration Source l per reaction Water 3.875 MgCl2 25 mM Sigma 4 Random hexamers 500 g/ml Integrated DNA Technologies (IDT) 0.5 Oligo DT 500 g/ml IDT 0.5 dNTPs 10 mM Invitrogen 1 Buffer 5x Invitrogen 1 DTT 0.1 M Invitrogen 2 Betaine 5 M Sigma 6 Reverse transcriptase 200 U/l Invitrogen 0.125 Sample RNA 5 ng/l 1 Cycling Conditions 1) 25 C for 10 min 2) 42 C for 30 min 3) 60 C for 30 min 4) 95 C for 5 min 5) 4 C for storage
42 Table 3. Primer Sequences for qRT-PCR Target Primer Pairs Cluster differentiation marker (CD14) F: 5-GGAACATTTGCATCCTCCTG-3 R: 5-TGAGTTTTCCCCTTCCGTGT-3 Cluster differentiation marker (CD45) F: 5-CAGAGCATTCCACGGGTATT-3 R: 5-GGACCCTGCATCTCCATTTA-3 Fc gamma receptor IIb (Fc RIIb) F: 5-GGAAGAAGCTGCCAAAACTG-3 R: 5-CCAATGCCAAGGGAGACTAA-3 Glial fibrillary acidic protein (GFAP) F: 5-GATCGCCACCTACAGGAAAT-3 R: 5-GTTTCTCGGATCTGGAGGTT-3 Interleukin 1 beta (IL1 ) F: 5-CTCATTGTGGCTGTGGAGAA-3 R: 5-GCTGTCTAATGGGAACGTCA-3 Ribosomal RNA 18S subunit (18S rRNA) F: 5-GTAACCCG TTGAACCCCATT-3 R: 5-CCATCCAATCGGTAGTAGCG-3 Scavenger receptor A (SRA) F: 5-GACGCTTCCAGAATTTCAGC-3 R: 5-ATGTCCTCCTGTTGCTTTGC-3 Scavenger receptor B (SRB) F: 5-AAGTGGTCAACCCAAACGAG-3 R: 5-ACTTGTCAGGCTGGAAATGG-3 Synaptotagmin F: 5-CATCGACCAGATCCACTTGT-3 R: 5-TCGTTTCCTACTTGGCACAC-3 Toll like receptor 4 (TLR4) F: 5-GCCGGAAGGTTATTGTGGTA-3 R: 5-AGGCGATACAATTCCACCTG-3 Tumor necrosis factor alpha (TNF ) F: 5-CTGTGAAGG GAATGGGTGTT-3 R: 5-CCCAGCATCTTGTGTTTCTG-3 Cycling Conditions 1) 95 C for 15 min 2) 95 C for 15 sec 3) 60 C for 60 sec 4) Plate read 5) Go to line 2, 39 more times
43 The SYBR Green reaction occurs when th e dye binds to double stranded DNA. Maximum absorbance is 497 nm, with emissi on at 520 nm. The Opticon instrument is compatible with a 96 well format, and uses individual excitation a nd detection of each well. For each mRNA of interest, a me lt curve analysis of the PCR product was performed after cycling was complete. Fluorescence was read in 1 C increments from 56-99 C. Primer pairs were designed to yield a single, specific product, with no evidence of primer-dimer formation using Primer 3 software from the MIT web site http://frodo.wi.mit.edu/cg i-bin/primer3/primer3 The following rules were helpful in primer design: 50% GC content, 18-30 base s long (20 optimal), identical predicted melting temperatures of bot h primers (optimal is 60 C), verify no self complementarity or primer-dimer formation (additional inform ation on potential primer-dimer is available from Qiagen on the Web at http://oligos.qiagen.com/oligos/toolkit.php no run of more than 3 bases annealing between primers), avoi d 3 or more G or C bases at 3 end of primers, and target a product size between 100-300 bp. All primers were checked for target specificity using the NCBI homepage at http://www.ncbi.nlm.nih.gov/BLAST/ The Quantity Calculations derived by the Opticon MONITOR Analysis Software (version 1.07) yeild the cycle numbe r at which the fluorescence exceeds the background level, called the C(T) value. In ge neral, the lower the C(T) value, the higher number of cDNA copies in the sample. For ea ch mRNA of interest, the C(T) values from the PCR were determined. To create a linear curve, the log of ng RNA in each standard was plotted against the average C(T) value ( each sample and standard was analyzed in triplicate). Log ng RNA were calculated for ea ch sample, and then converted to ng RNA. For a given reverse transcription, all targets of interest were normalized to results for 18S
44 rRNA for the same sample to control for vari ations in the starting RNA concentration and enzyme efficiency. The ratio of the ng ta rget RNA to ng 18S rRNA was calculated, and the results subjected to one way analysis of variance (ANOVA), followed by Fishers protected least squares difference (P LSD) with significance taken at p < .05. The final results are presented as fold change versus control mice. Histology Immunohistochemical analysis of reactive glial markers was performed using four 25 m free-floating sections spaced 300 m apart through the hippocampus. Details of this procedure were originally described el sewhere (Gordon et al., 1997 ). Sections were first blocked for endogenous peroxidases (10% methanol, 3% hydrogen peroxide in DPBS) for 15 minutes, washed with DPBS, then permeabilized in a solution of 100 mM lysine, 0.2% triton x-100, and 4% normal goat or horse serum (Pel Freeze, Rogers, AK) in DPBS for 30 minutes. Sections were th en incubated overnight in the appropriate primary antibody in DPBS and 4% serum (Table 4). The following day, sections were washed with DPBS, and then incubated in appropriate biotinylat ed secondary antibody (Vector Laboratories, Burlingame, CA) at a concentration of 0.5 g/ml in DPBS and 4% serum, for two hours. Sections were washed with DPBS, and then incubated for one hour in Vectastain Elite ABC solution (Vector Laboratories, Burlingame, CA). Sections were then washed and incubated with a so lution of 0.5% nickelous ammonium sulfate hexahydrate and 0.05% diaminobenz idine in tris buffered saline for 5 minutes. Color development was achieved by the addition of 0.03% hydrogen peroxide, and incubation for an additional 5 minutes, followed by washes Controls for nonspecific binding of the secondary antibody were performed by ex cluding primary antibodies. GFAP
45 immunostaining was performed using horsera dish peroxidase-streptavidin (Vector Laboratories, Burlingame, CA) at 1 g/ml, ra ther than ABC and nickel enhancement was not used. Sections were slide mounted and cover slipped, then imaged at 100x magnification, focusing on the dentate gyrus of the hippocampus. Images were then analyzed for area percent positive stain using Image-Pro Plus software (MediaCybernetics, Silver Spring, MD). The results were subjected to one way ANOVA, followed by Fishers PLSD with significance taken at p < .05. The final results are presented as fold cha nge versus control mice. Table 4. Primary Antibodies Used for Immunohistochemistry Antigen Antibody Type Source Catalog # Titer CD45 Rat monoclonal Serotec MCA1031G 1:3000 CR3 (CD11b) Rat monoclonal Serotec MCA711 1:3000 Fc RII/III Rat monoclonal PharMingen 553141 1:3000 GFAP Rat monoclonal Zymed 13-0300 1:3000 TLR4 Rabbit polyclonal Imgenex IMG-579 1:3000 Results LPS was injected into the hippocampus of nontransgenic mice and a time course established for both RNA and protein expressi on of inflammatory markers. Animal health overall was mildly affected by LPS. Some weight loss was seen, which was significant after 7 days, but re turned to normal beginning at 2 weeks. Malaise was only notable at the 24 hr time point, primarily evidenced by hypo-locomotion.
46 Advantages of RNA analysis by qRTPCR include accurate quantitation of several markers from very small samples and assessment of specific isoforms, even when suitable antibodies are not av ailable for immunohistochemi stry. Figure 7A shows a typical standard curve generated for Ribogreen assay of sample RNA concentrations. All samples were within the linear range of the assay, averaging 65 ng of RNA per l of sample extract. Figure 7B shows typical et hidium bromide/UV detection of the RNA. All samples were intact, with no evidence of RNA degradation. Samples at a concentration of 5 ng/l in water, as well as standards spanning 0.2-50 ng/l in water, were subjected to qRT-PCR. Melt curve an alysis of the PCR pr oduct indicated a single specific reaction product. An example of a t ypical melt curve is shown in Figure 7C. A single peak without shoulders or evidence of multiple peaks indicated specific primer annealing. The standard curves for the qR T-PCR were then generated, as shown in Figure 7D. All targets had correlation coeffi cients of not less than 0.95, with less than 25% variability in the slope s between the curves for different target mRNAs (for example, comparing 18S to IL1 ). The slope represents the efficiency of the PCR reaction for the particular primer pair used. Less than 10% variati on would be necessary to detect more subtle changes in gene expres sion as variations in th e efficiency of the PCR of the target compared to 18s could artificially bias the result.
47 Figure 7. mRNA analysis by qRT-PCR. RNA from the hippocampi of LPS injected mice was extracted and assayed. Panel A shows the standard curve used for RiboGreen estimation of total RNA levels. Panel B s hows the typical ribosomal banding pattern of 50 ng total RNA samples, indi cating the absence of signifi cant degradation. After reverse transcription, cDNA was subjected to qRT-PCR. Panel C shows the melt curve for 18s rRNA qRT-PCR product, typical of al l primers used. A single peak without shoulders indicated specific priming. Panel D shows a standard curve generated for the 18s rRNA qRT-PCR, plotting mass of the RNA a dded to the reverse tr anscription versus the threshold cycle C(T). All primers used in this work generated a standard curve with slopes varying by less than 25%.
48 A time course was established for LPS stim ulated gene transcription. Figure 8A contains cell-specific markers. The glia l markers CD45 and GFAP both changed their expression over time, whereas sy naptotagmin did not change ( F(8,45) = 25, p < .0001; F(8,45) = 35, p < .0001; F(8,45) = 1.2, p = .30, respectively). Post hoc analysis revealed significant differences compared to untreated control mice, with significance taken at p < .05. The microglial marker CD45 was si gnificantly elevated by 6 hr, peaked at 3 days (72 hr), and remained elevated throughout the time course to 28 days (672 hr) post injection. GFAP followed a similar time course and increased significantly from 6 hr to 7 days (168 hr), but returned to baseline by 14 days (336 hr). Figure 8B shows results for three receptors involved in phagocytosis. Scavenger receptor A (SRA), scavenger receptor B (SRB), and Fc receptor IIb (Fc RIIb) all changed expression over time ( F(8,45) = 9.5, p < .0001; F(8,45) = 6.4, p < .0001; F(8,45) = 25, p < .0001 respectively). Post hoc analysis revealed significant differences compared to untreated controls, with significance taken at p < .05. SRA was significantly elevat ed by 6 hr, peaked at 3 days, and then dramatically declined by 14 days. In contrast, SRB was briefly and modestly elevated only at the 6 hr time point otherwise remaining unchanged. The Fc RIIb followed a similar time course to SRA, but remained modestly elevated out to 28 days.
49 Figure 8. Time course of mRNA expression. RNA from the hippocampus of mice injected with LPS and allowed to survive 1 hr, 6 hr, 24 hr, 72 hr (3 days), 168 hr (7 days), 336 hr (14 days), and 672 hr (28 days), was s ubjected to qRT-PCR. Results are presented as mean SEM of fold change compared to un treated control mice (n=6 each group). Note that the x-axis is not linear in order to permit evaluation of both early and late time points. Panel A shows the cell specific ma rkers CD45 (microglia), GFAP (astrocytes), and synaptotagmin (neurons). Both CD45 and GFAP levels peak around 3 days; only CD45 remained elevated at 28 days. Synaptot agmin levels were unaffected by treatment. Panel B shows three receptors capable of mediating phagocytosis, SRA, SRB, and Fc RIIb. SRA was transiently elevated, peaking from1-3 days Fc RIIb also peaked around 3 days, but remained modestly elevated at 28 days. SRB had a small increase, significant only at 6 hr. Pa nels C and D show results of RNA from the hippocampus of mice injected with LPS and allowed to survive 72 hr (3 days). Panel C shows significant increases in both TLR4 and CD14 messages. Panel D shows dramatic increases in TNF and IL1 messages; note greater than 30 fold increases of IL1 p < .05 for all panels versus untreated controls.
50 Figure 8. Time course of mRNA expression.
51 Based on the findings that most of the ta rget RNAs peaked at 3 days, several additional markers were examined at that time point. Figure 8C shows the relative expression of two key components of the LPS signaling cascade, compared to untreated control mice. Both the LPS receptor cluster differentiation marker (CD14) and toll like receptor (TLR4) were significantly elevated. Figure 8D shows the relative contributions of interleukin (IL1 ) and tumor necrosis factor (TNF ). IL1 message was increased by greater than 30-fold. TNF was also significantly elevat ed, but to a lesser extent. Immunohistochemical analysis of horizo ntal brain sections was conducted to verify the cell-specific expression of severa l inflammatory markers identified by RNA analysis. Figure 9 shows the time course of changes in expression for four microglial markers. Representative micrographs from th e site of LPS injection (the hilus of the dentate gyrus subregion of the hippocampus ) are presented. Untreated mice demonstrated resting, ramified microglia, with fine delicate processes which were marked by CD45 and complement receptor CR3 (Fi g. 9A-D). LPS treatment resulted in a widespread activation of microglia with a co ncurrent shift in mor phology. After 1 hr (Fig. 9E-H), cells increased expr ession of CD45, CR3, and Fc receptor II/III (Fc RII/III) and developed thicker, shorter proc esses; this increase was more obvious at 6 hr (Fig. 9I-L). By 24 hr, the increases were observable fo r all four markers CD45, CR3, SRA, and Fc RII/III (Fig. 9M-P). The ramified microgl ia predominated at this time point, but small, rounded cells were also seen. After 72 hr, the staining intensity was at its peak for all four markers. Multiple morphologically distinct cells were visible including hyperramified (bushy, indicated by arrow in Fig. 9S), round, and amorphous/ameboid cells with smaller numbers of thick processes (Fig. 9Q-T). After a week, SRA and Fc RII/III
52 expression had subsided, but CD45 and CR3 remained high. The bushy microglial morphology predominated (Fig. 9U-X). An interesting finding was both CD45 and CR3 continued to mark bushy, reac tive microglial after 2 weeks (Fig. 9Y-BB) and 4 weeks (Fig. 9CC-FF). Figure 9. Time course of micr oglial protein expression and morphology Brains from LPS injected mice were sectioned horizontal ly at 25 m and immunostained. Each row represents a different time point: untreated mi ce (A-D); LPS treated mice (E-FF) survived 1 hr (E-H), 6 hr (I-L), 24 hr (M-P), 3 days (Q-T), 7 days (U-X), 14 days (Y-BB), and 28 days (CC-FF). Each column represents a diffe rent marker: the first column is CD45, the second column is CR3, the third column is SRA, and the fourth column is Fc RII/III. Micrographs are from the hilus of the de ntate gyrus of the hippocampus. Resting microglia are seen after sa line treatment (A-D), expres sing CD45 (A) and CR3 (B). From 1 hr (E-H) to 6 hr (I-L) cells became sw ollen and increased levels of CD45 (E,I), CR3 (F, J) and Fc RII/III (H,L) are appare nt. By 24 hr (M-P), all four markers are expressed, on both round and ramified cells. At 3 days, expression leve ls are peaking, as round, ameboid, and ramified/bus hy cells stain intensely for CD45 (Q), CR3 (R), SRA (S), and Fc RII/III (T). The arrow in Panel S indicates a bushy morphology. At 7 days (U-X), 14 days (Y-BB), and 28 days (CC-FF), SRA (W AA, EE) and Fc RII/III (X,BB,FF) decrease to control levels; CD 45 (U,Y,CC) and CR3 (V, Z, DD) remained elevated. Scale bar = 50 m.
53 Figure 9: Time course of micr oglial protein expression and morphology.
54 Semi-quantitative results for microglia l immunohistochemical analyses in the dentate gyrus are shown in Fi gure 10. CD45, CR3, SRA, and Fc RII/III all changed significantly over time ( F(8,45) = 11.4, p < .0001 F(8,45) = 12.3, p < .0001 F(8,45) = 3.4, p = .0037 F(8,45) = 4.4, p = .0006, respectively). Post hoc analysis revealed significant differences compared to untreated cont rol mice, with significance taken at p < .05. Results are expressed as fold ch ange versus untreated controls. Up to three fold increases in CD45 levels were seen, beginning at 1 hr and lasting to 28 days (Fig. 10A). A similar profile for CR3 was obtained, with up to four-fold increases from 1-28 days (Fig. 10B). SRA, though significantly incr eased only at 1 and 3 days, showed over 30 fold change relative to the virtual absence of stai ning in control animals (Fig. 10C). Fc RII/III staining was also transiently and significan tly elevated at 1 and 3 days by 25 fold (Fig. 10D). Figure 10. Quantitation of immunohistoche mical results for microglial markers Brains from LPS injected mice were sectioned hor izontally at 25 m, and immunostained. Images were collected from the dentate gyrus of the hippocampus at 100x magnification. The percent area occupied by positive stain was calculated. Results shown are fold change compared to uninjected mice. Note th at the x-axis is not linear in order to permit evaluation of both early and late time points. CD45 was significantly elevated from 1 hr out to 28 days. CR3 displayed a similar time course, with significant increases from 1-28 days. In contrast SRA and Fc RII/III were transiently activated, with significant increases only at 1-3 days. p < .05.
55 Figure 10. Quantitation of immunohistoche mical results for microglial markers.
56 Astrogliosis was also evaluated in res ponse to LPS and is shown in Figure 11. Untreated mice showed detectable levels of GFAP in the hi ppocampus (Fig. 11A). Three days after treatment, astroc ytes throughout the injected hippocampus increase GFAP expression and had thickened processes (Fig. 11B). Figure 11C shows the dentate gyrus results expressed as fold change of stained area compared to untreated controls. A main effect of time was determined ( F(8,45) = 8.2, p < .0001). Post hoc analysis revealed significant increases beginning at 1 day, peaki ng from 3-7 days, and remaining modestly elevated at 28 days. Figure 11. Time course of astr ogliosis in response to LPS GFAP levels in the brains of LPS injected mice were demonstrated by im munohistochemistry. Panel A depicts the hippocampus of an untreated mouse showing lo w, basal levels of GFAP. Panel B shows the heightened astrocyte reaction 3 days after LPS injection. Statisti cal analysis of the dentate gyrus subregion, shown in Panel C, revealed significant activation beginning at 1 day, peaking from 3-7 days, then remaining moderately elevated out to 28 days. The x-axis is nonlinear to show the ch anges at early time points. Scale bar = 500 m. p < .05.
57 Figure 11. Time course of astr ogliosis in response to LPS
58 Finally, Figure 12 presents the i mmunohistochemical results for TLR4. Representative micrographs of the dentat e gyrus subregion of the hippocampus are shown. No cellular staining for TLR4 was det ectable in untreated animals (Fig. 12A). As early as 1 hr after injection, TLR4 imm unostaining was evident as a band of staining along the inner extent of the gr anule cell layer of the dentate gyrus (Fig. 12B) as well as along the edge of the hippocampal fissure (not shown). Beginning at 24 hr, punctuate staining was evident in the hilus (Fig. 12C). Figure 12D shows the dentate gyrus fold change of stained area compared to controls Post hoc analysis revealed significant increases from 1-6 hr. This is due to the inte nse layer of stain seen in Figure 12B. This layer of stain decreases over the first week, as punctuate st ain starts to appear. The increase in the punctuate st ain, though visually discerni ble, was not statistically significant using the area measurement ( p = .055 at 3 days; p = .077 at 7 days). Figure 12. Time course of TLR4 levels after LPS injection Immunohistochemical analysis of TLR4 was conducted using 25m hori zontal brain sections from LPS treated mice. Micrographs are represen tative of the dentate gyrus subregion of the hippocampus. Panel A shows an untreated mouse with lit tle to no positive stain. Panel B shows an intense layer of stain along th e inner dentate granul e cell layer, 1 hr af ter LPS injection. Panel C shows punctate and more dispersed st ain in the hilus, be ginning 3 days after injection. Statistical an alysis, shown in Panel D, revealed significant increases at 1 and 6 hr, as well as smaller increases at 3 and 7 days. The x-axis is nonlinear to show the changes at early time points. Scale bar = 200 m. p < .05.
59 Figure 12. Time course of TLR4 levels after LPS injection.
60 Discussion Lipopolysaccharide is a prot otypical inflammatory agent used in both the periphery and in the brain. Here, we describe the time course of the microglial reaction to LPS in vivo RNA and protein data c oncur that there is an ac ute and chronic response to a single intrahippocampal injection of LPS. The acute reaction to LPS began as early as 1 hr post injection. This short term response peaked at 3 days for both RNA and pr otein levels, returning to basal levels by about 7 days for several markers. The re sponse of astrocytes was demonstrated by increases in GFAP. Several markers of the microglial response were also evaluated. Two key receptors that ca n mediate phagocytosis, Fc RII/III and SRA, were only briefly up-regulated, perhaps priming the system for either mounting an acquired immune response or continuing the innate re sponse. The endogenous ligand for Fc receptor is typically the Fc region of IgG. However, ther e is no evidence in the current studies that anti-LPS antibodies were produced in the LPSinjected mice, although such a possibility cannot be ruled out. On the other hand, it has been demonstrated that LPS can induce Fc expression in the absence of IgG, as part of cell activation. In addition, Fc receptor expression by macrophages can be regulated by cytokines as well as by LPS (Amigorena et al., 1989; Lynch et al., 1990; Laszlo & Di ckler, 1990; Loughlin et al., 1992; Keller et al., 1994). Thus, it is likely that the induction of Fc reflects another facet of microglial activation and not the presence of specific immunoglobulin. Scavenger receptors play a key role in mediating innate immunity. They are pattern recognition rece ptors, with bacterial and vira l products as typical ligands,
61 including LPS. It is in teresting that fibrillar A is also a ligand for SRA and SRB (Huseman et al., 2002). Our lab has prev iously shown decrea sed levels of A after intrahippocampal injection of LPS in amyl oid depositing mice (Chapter 3 of this dissertation, Herber et al., 2004b; DiCarlo et al, 2001). It is possible that LPS increased expression of SRA which promoted the removal of A via phagocytosis in that study. We were surprised that we did not see larg e increases in SRB in the current study, which has been implicated in Alzheimers disease (Bamberger et al., 2003). We only detected a 33% increase in mRNA, and only at the 6 hr time point. Appropriate antibodies were unavailable for immunohistoc hemical analysis, thus no pr otein data were collected. The chronic response to LPS was seen ma inly in microglia, as demonstrated by both RNA and protein expression of CD45. This was further confirmed by CR3 immunohistochemistry. Our data showed signi ficant microgliosis 28 days after a single injection of LPS. This confirms our earlier report using a similar protocol in amyloid depositing transgenic mice (Herbe r et al., 2004b). Our results fr om the direct injection of LPS into the central nervous system are in co ntrast to L5 nerve transection, where CR3 expression peaked at 14 days, but returned to baseline by 28 days (Tanga et al., 2004). Tanga and colleagues also showed a sustained response by GFAP to both nerve transection and to peripheral administration of LPS (Tanga et al., 2004; Raghavendra et al., 2004). In our model, the GFAP respons e was largely resolved after 7 days. CD45 is a protein tyrosine phosphatase (review ed in Irie-Sasaki et al., 2003). It is thought that activation of CD45 leads to dephos phorylation of key pr oteins, allowing for a quick response upon subsequent stimulation. CR3 is an integrin that mediates cell migration, adhesion, and phagocytosis (reviewed in Ehlers, 2000). CR3 can also bind
62 LPS and is involved with CD14 in the activation of NF B. The sustained elevation of CD45 and CR3 by microglia may reflect a hype r-reactive system, remaining primed for subsequent stimuli. These same cells ma y also continue to produce cytokines and chemokines, creating a potentially hostile environment. Changes in microglial morphology and pr otein expression after LPS injection were complex. Early in the time course (1-6 hr), we saw thickeni ng and shortening of the cell processes as CD45, CR3 and Fc receptor expression increased. By 24 hr, SRA expression was also seen on these bushy, hypertrophied cells. Small round cells expressing CD45 and CR3 appeared at 24 hr and an amorphous/ameboid cell form was at 3 days. The three cell shapes, bushy, r ound, and ameboid coexisted at the 3 day time point and expressed all four markers. On ly the bushy phenotype persisted and only in cells expressing CD45 and CR3. The signifi cance of these cell shapes and their contribution to cell function has yet to be dete rmined. It is possible that the small round cells seen at 24 hr are actua lly peripheral monocytes that had entered the CNS after injection. Montero-Menei a nd colleagues showed that as much as 80% of OX-42 (rat equivalent of CR3) and ED1 (CD68) positive cells present in the rat brain 24 hr after intracerebral LPS injection were attributable to recruited monocytes (1996). Similar MHCII positive round cells have been reported 48 hr after intraventr icular LPS injection in rats (Hauss-Wegrzyniak et al., 1998a,b). This same report showed MHCII positive bushy cells 4-7 days post injection. It is impossible to different iate between brain and peripheral macrophages using curr ent labeling techniques. One is tempted to assume the round and/or ameboid cells are phagocytic. Our own results in amyloid depositing
63 transgenic mice, showing decreased A burden 3-14 days after LPS injection, suggests phagocytosis by some cell populati on (Herber et al., 2004b). Several additional markers of inflammation were investigated. Increases in the RNA for cytokines IL1 and TNF were expected after LPS administration (Kim et al., 2000; Nadeau & Rivest, 2002; Raghavendra et al., 2004). These cytokines not only prime the system but can also create a da maging environment. Neurodegeneration was not detectable in our model usi ng cresyl violet or fluorojade stains (data not shown). Our observations of the level of inflammation in the brain indicated a spreading wave of microgliosis, using CR3 as a marker. Untreat ed mice showed low levels of CR3 even in resting microglia. However, at 1 hr, the entire hippocampus showed increased microgliosis. By 6 hr, the hippocampal fimbria, rhinal cortices, thalamus, brainstem, and posterior striatum were similarly affected. At 24 hr, the entire br ain, including temporal, parietal, frontal, and occipita l cortices showed microgliosis. This continued through 72 hr with a marked increase in the striatum. By 168 hr (7 days), the effect began to wane and only areas very near the hippocampus show ed gliosis (striatum, fimbria, rhinal cortices). By 28 days, only the hippocampus was still experiencing microgliosis. Note that the very bushy, round, and amorphous/ameboi d microglia described here were only seen in the hippocampus. Outside that re gion, microglia increased CR3 expression, but did not have marked changes in morphology. Thus, although a widespread inflammatory response was mounted in response to LPS, the neurons survived. Here we report that both TLR4 and CD14 mRNA are induced following LPS injection. TLR4 mRNA has b een localized by others in m ouse microglial cell cultures, but not in astrocytes, oli godendroglia, or neurons (Le hnardt et al., 2002, 2003). In
64 contrast, in situ hybridization of rat brain dem onstrated basal TLR4 and CD14 mRNA expression in the circumventricular organs, le ptomeninges, and chor oids plexi (Laflamme & Rivest, 2001). Intrastriatal LPS injec tion similarly increase d CD14 mRNA, but not TLR4 as detected by in situ hybrid ization (Nadeau & Rivest, 2002). Immunohistochemical analysis of TLR4 levels in the central nervous system (with or without LPS) has not been reported to our knowledge. Singh and Jiang recently reported that peripheral administration of LPS did not cross the blood brain barrier, but rather bound endothelial cells which then signaled micr oglial to respond (Singh & Jiang, 2004). Although LPS clearly increased TLR4 RNA and immunostaining, the TLR4 immunohistochemical analysis did not reve al a population of highly ramified, bushy, reactive microglia similar to those stained by CD45. Possibly, the staining pattern we saw was endothelial cells, es pecially at the early time points. Conversely, the TLR4 staining pattern at 72 hr was reminiscent of the round and ameboid cells seen with CD45 staining at 72 hr. The ability of micr oglial to express TLR4 in response to LPS in vivo remains to be adequately demonstrated by dual label immunocytochemistry. In conclusion, we have examined the time c ourse of the response of the brain to a single intrahippocampal injecti on of LPS. The response had both a transient, acute phase as well as a chronic phase a ssociated with the induced in flammation. Evaluation of the phenotype of reactive microglia may lead to a better understanding of neuroinflammatory diseases.
65 Chapter 3 Time-Dependent Reduction in A Levels After Intracranial LPS Administration in APP Transgenic Mice Abstract Inflammation has been argued to play a primary role in the pathogenesis of Alzheimers disease. Lipopolysaccharide (LPS) activates the innate immune system, triggering gliosis and inflammation when inje cted in the central nervous system. In studies described here, APP transgenic mice we re injected intrahi ppocampally with 4 or 10 g of LPS and evaluated 1, 3, 7, 14, or 28 days later. A load was significantly reduced at 3, 7, and 14 days, but quickly returned near baseline 28 days after the injection. No effects of LPS on Congophilic amyloid deposits could be detected. LPS also activated both microglia and astrocytes in a time-dependent manner. The GFAP astrocyte reaction and the Fc receptor microglial reaction peaked at 7 days after LPS injection, returning to base line by 2 weeks post injection. When stained for CD45, microglial activation was detected at all time points, although the morphology of these cells transitioned from an ameboid to a ra mified and bushy appearance between 7 and 14 days post injection. These resu lts indicate that activation of brain glia can rapidly and transiently clear diffuse A deposits, but has no effect on co mpacted fibrillar amyloid.
66 Introduction Chronic neuroinflammation is a hallmark of many neurodegenerative diseases, including Alzheimers disease (AD, Akiyama et al., 2000). The role of inflammation in AD is controversial, with two sides to the response. Acute inflammation can be neuroprotective, aiding in removal of pathogens and protecting cells (reviewed in Streit, 2002). Conversely, excessive, chronic inflamma tion is thought to enhance cell death via an autotoxic mechanism (McGeer & McGeer, 20 02). Current transgenic animal models replicate the amyloid pathology, glial response, and cognitive impairment of AD, without apparent neurodegeneration. One approach to more accurately reflect the conditions in AD brain is to inject lipopolysaccharide (L PS) into transgenic mice to enhance the inflammatory state of the brain. It was hypothesized that LPS dosing might also alter amyloid deposition while tri ggering neurotoxicity. LPS is a Gram-negative bacterial cell surface proteoglycan, also known as bacterial endotoxin, which triggers an inflam matory response by the host. LPS binds to circulating LPS binding protei n, which subsequently tran sfers the LPS to the cell membrane bound cluster differentiation mark er CD14, a common receptor for bacterial components (Heumann & Roger, 2002). The CD14 protein does not contain an intracellular domain, so interaction with ad ditional components is likely. Toll like receptor 4 (TLR4), as well as myeloid di fferentiation protein MD2, form a complex through which LPS:CD14 can transduce a signal (Thomas et al., 2002). Multiple cascades are involved in the signal transduction, with nuclear factor NF B as a key player. Downstream gene activation triggers the production of cytoki nes, nitric oxide, and prostanoids leading to the cellular inflammatory response.
67 LPS has been used to mimic the inflammatory conditions seen in human AD brain. Acute and chronic application of LPS into the brain ventricles of rats resulted in gliosis, cytokine production, increased APP le vels, cognitive defic its, and neurotoxicity limited to the forebrain cholinergic nucl ei (Hauss-Wegrzyniak et al., 1998a,b, 2000, 2002; Willard et al., 1999). LPS has also been administered to amyloid-depositing transgenic mice to cause g lial activation. However, so me investigators reported increased A levels in response to LPS (Sly et al ., 2001; Qaio et al., 2001; Sheng et al., 2003), while others observed A clearance (DiCarlo et al ., 2001; Quinn et al., 2003). The experimental conditions of these previous reports vary, including the type and age of transgenic mice used; type, dose and route of administration of L PS; and post injection survival time. A potential mechanism of A removal includes phagocytosis. In vitro studies show that both microglia and astrocytes reac t to LPS with phagocytosis, as well as with production of potentially cytotoxic agents such as prostaglandins, cyt okines, and reactive oxygen species (Kalmar et al., 2001). Thus, this model of neuroinflammation may be useful in triggering the innate immune system to clear A deposits, and/or kill neurons. The experiments described herein will addre ss the overall hypothesis that neuroglia play a key role in the inflammation seen in AD, and may represent an innate mechanism of A removal. In an effort to clarify the mechanism of A clearance, dose-response and timedependence experiments were conducted using intrahippocampal inject ions of LPS into aged APP mice.
68 Materials and Methods Mouse Strains Transgenic mice were bred to devel op Alzheimers like pathology using Tg2576 APP mice as described previously (Holco mb et al., 1998). APP mice and their nontransgenic littermates aged 16-17 months were used in this study. Animals were group housed prior to surgery under a 12 hr light -dark cycle with free access to chow and water. Intrahippocampal Injections Mice were anesthetized using isoflura ne and immobilized in a stereotaxic apparatus. A single, one microliter injection of either saline, 4 g/ l LPS, or 10 g/ l LPS ( Salmonella abortus equi Sigma, St. Louis, MO) was delivered over a two minute period into the right hippocam pus (coordinates from bregma : 2.7 mm posterior, -2. 5 mm lateral, and -3.0 mm ventral). The incisi on was closed with wound clips, isoflurane was discontinued, and the animal revived on a heated pad. All mice completely recovered within five minutes. Animals were singly housed for the post treatment survival period under standard vivarium conditio ns. We used at least four mice (4-7) for each condition, balanced for gender. The dos e response study had three conditions: APP mice injected with either saline, 4 g LPS, or 10 g LPS; all survived 7 days. The time course utilized two genotypes for a total of 10 conditions. Nontransgenics were injected with saline and survived 7 da ys, or were injected with 4 g of LPS and survived 1, 3, or 7 days. APP mice were injected with saline and survived 7 days or were injected with 4 g of LPS and survived 1, 3, 7, 14, or 28 days.
69 Tissue Preparation Mice were anesthetized with pentobarbital, and then perfused transcardially with 25 ml of normal saline, followed by 50 ml of freshly prepared neutral buffered 4% paraformaldehyde. The brain was postfixed in 4% paraformaldehyde solution for 24 hr, then processed through a cryprotection schedul e of 10, 20, then 30% sucrose. Frozen brains were sectioned horizontal ly on a sliding microtome at 25 m. Sections were then stored in Dulbeccos phosphate buffered sa line pH 7.4 (DPBS) with 100 mM sodium azide at 4 C. Histology Immunohistochemical analysis was perf ormed for each marker using six, 25 m free-floating sections spaced 300 m apart through the hippocam pus. Details of this procedure are described elsewhere (Gordon et al., 1997). Briefly, sections were blocked for endogenous peroxidases (10% methanol and 3% hydrogen peroxide in 80% DPBS), washed with DPBS, then permeabilized ( 100 mM lysine, 0.2% triton x-100, 4% normal serum in DPBS). Sections were then inc ubated overnight in the appropriate primary antibody (Table 5). The following day, sections were washed, and then incubated in appropriate biotinylated secondary antibody. After another cycle of washes, the tissue was incubated with Vectastain Elite ABC kit (Vector Laboratories, Burlingame, CA). In the case of glial fibrillary acidic prot ein (GFAP), streptavidin-peroxidase was used rather than ABC. The tissue was then wa shed and stained with a diaminobenzidine: peroxide system, followed by final washes. In the case of CD45 and Fc gamma receptor II/III (Fc RII/III), nickel enhancement of the colo r development was used. The extent of nonspecific binding was assessed in the absen ce of primary antibodies for all assays.
70 Sections were mounted onto slides, dehydr ated, and cover slipped with DPX (E.M. Sciences, Fort Washington, PA). Separate ly, compact plaques were evaluated after sections were slide mounted, by incubating in alkaline alcoholic saturated sodium chloride (AASSC), followed by 0.2% C ongo red in AASSC. Tissue damage was evaluated using a 0.05% cresyl violet (pH 3.3) solution. Table 5. Primary Antibodies Used for Immunohistochemistry Antigen Type Source Catalog # Titer A Rabbit polyclonal Biosource 44-136 1:10000 CD45 Rat monoclonal Serotec MCA1031G 1:3000 Fc RII/III Rat monoclonal PharMingen 553141 1:3000 GFAP Rat monoclonal Zymed 13-0300 1:3000 Data Analysis Stained sections were imaged at 100x magnification, focusing on the molecular layer of CA1, CA3, and dentate gyrus (DG) re gions of the injected hippocampus. Images were then analyzed for area percent positive stain using Image-Pro Plus software (MediaCybernetics, Silver Spring, MD). Area percent stained was calculated for each individual region of the hippocampus, as well as an average of the three regions. Results were averaged for each animal, and the treatment groups compared by one way analysis of variance (ANOVA) followed by post hoc analysis using Fishers protected least squares difference (PLSD, significance at p < .05). Results shown are for the CA3
71 region, which showed the greatest effect. Similar results were seen in all of the hippocampal regions analyzed, as well as for the average of the three. Results Dose Response Intrahippocampal injections of LPS were made into APP transgenic mice in order to optimize the A reductions reported prev iously (DiCarlo et al ., 2001). Dose response studies were done us ing saline, 4, or 10 g/ l of LPS. Animals were evaluated 7 days after the injection by immunohi stochemistry for total A burden using an anti-A antibody which recognizes both A 1-40 and A 1-42. Micrographs represent the injected hippocampus of a saline injected control (Fig. 13A) versus a 10 g LPS-dosed APP mouse (Fig. 13B) are shown. In the control mouse, both lightly stained, diffuse A deposits and darkly stained, compacted deposit s were seen. LPS injection resulted in removal of a major portion of the diffuse de posits in the CA1 and CA3 regions of the hippocampus. The quantification are from the CA3 region (the region indicated by the box in Fig. 13D). Statistical analysis of the data shown in Figure 13C indicated a significant removal of A in LPS-injected APP mice ( F(2,13) = 6.6, p < .05). The effect of LPS injection on compact amyloid plaque burden in APP mice was determined with Congo red staining. Compact plaques were detected typically near the hippocampal fissure in the saline injected mice (Fig. 13D). No detectable changes in the amyloid load were seen with LPS treatmen t in any hippocampal subfield (Fig. 13E). Quantitation also failed to reveal a significant change in amyloid burden in CA3
72 (Fig. 13F), CA1, or dentate gyru s subfields, or in the average of all fields from the hippocampus. Similarly, LPS treatment wa s not accompanied by reductions in amyloid detected by thioflavin-S (results not shown). Figure 13. Dose response of LPS stimulated A removal. APP mice were injected with saline (A,D) or 10 g of LPS (B,E) 7 days pr ior to sacrifice. I mmunostaining for total A burden is shown in panels A and B. A significant reduction at both dosages is shown in panel C. Histochemical st aining of compact plaques with Congo red dye is shown in panels D and E. No reduction was detected as a result of LPS dosing (F). The CA3 subregion enclosed by a box in Panel D repres ents the area used for calculating percent positive stain in data analyses (C,F). Abbreviations: CA1, CA3: subregions of Ammons horn; DG: dentate gyrus. Scale bar = 250 m. p < .05.
73 Gliosis, a possible link to the mechanism of A reduction in APP mice, was analyzed using CD45 staining for reactive microglia. Salineinjected APP mice (controls) are shown in Figure 14A. These mi ce have CD45+ microglia largely restricted to the vicinity of putative amyloid plaques a nd the injection site within the hilus of the dentate gyrus. We use the term putative amyloid plaques because, even though such aggregates of microglia have previously been demonstrated to surround Congophilic cores (Gordon et al.; 2002), no Congo red count erstain was performed on these sections. LPS treatment of APP mice caused a widesp read eruption in the number and staining intensity of CD45+ microglia throughout the hippocampus (Fig. 14B). Both doses caused significant increases in CD45 imm unoreactivity (mean area % for saline dose = 1.1 0.3; 4 g dose = 12.3 4.0; 10 g dose = 9.1 2.6). The 10 g dose caused increased reactivity to extend outside the injected hippocampus, spreading into the adjacent cortices and even to the cont ralateral hippocampus (data not shown). Astrogliosis was detected using the astr ocyte specific marker GFAP. Salineinjected APP mice had diffuse distribution of GFAP+ astroglial cells, as well as plaque associated clusters of astroglia (Fig. 14C) LPS treatment caused an increase in the staining intensity of GFAP staining throughout the hippocampus (Fig. 14D). Both doses caused a mild, but not significant, increase in GFAP immunoreactivity, (mean area % for saline dose = 17.1 4.2; 4 g dose = 19.7 3.9; 10 g dose = 18.6 3.1). The highest dose triggered a notable glial response outside the inject ed area (data not shown).
74 Figure 14. Glial response to LPS. APP mice injected with saline or with 10 g of LPS, 7 days prior to sacrifice, showed increased microglial activity due to LPS (B) versus saline (A). GFAP immunoreactivity showed incr eased astrocyte activit y in LPSinjected (D) versus salineinjected (C ) APP mice. Scale bar = 250 m.
75 The potential for LPS injection to cause tissue damage was examined histochemically with cresyl violet stain. Sections were evaluated for evidence of neurodegeneration, which included pyknotic nuc lei as well as gross cell loss in the pyramidal and granule cell layers of the hippocampus. Some animals had evidence of mechanical trauma in the injection vicinit y, but this was observed in both salineand LPS-injected mice to the same degree. No widespread evidence of neuronal damage caused by LPS was detected. Overall, the dose response study demons trated that LPS injection into the hippocampus of APP mice led to clearance of A within one week, without altering compact plaque load or tri ggering neurodegeneration. Concu rrently, treatment enhanced gliosis, suggesting a glialmediated mechanism of A reduction. The 4 g LPS dose was chosen for use in subsequent experiments as adequate clearance of A was seen at that dose. Time Course To further evaluate the L PS-induced clearance of A a time course was conducted. APP mice were injected with 4 g of LPS, and allowed to survive 1, 3, 7, 14, or 28 days. Nontransgenic mice were also inje cted with LPS, and survived 1, 3, or 7 days post injection. Control groups for both genotypes received saline injections, with 7 day survival periods. A burden in APP mice was determined immunohistochemically using an anti-A antibody, and analyzed in the CA1, CA3, and DG subfields of the hippocampus. Analysis of variance showed a time-dependent effect on A levels in LPSinjected mice ( F(5,30) = 2.7, p < .05), particularly in the CA3 subregion (Fig. 15). Post hoc analysis using Fishers PLSD indicated a statistically significant removal of A in LPS
76 injected APP mice from 3-14 days, which return ed to baseline by 28 days. As in the dose response study at 7 days, no effect of LPS injection on Congo red stained plaques in APP mice was seen in any subregion of the hi ppocampus, or in the hippocampus overall, at any time point (data not shown). Figure 15. Time dependent removal of A by LPS injections APP mice were injected with saline or 4 g of LPS and immunostained for A Mice injected with LPS were killed at 1, 3, 7, 14, or 28 days post injection. Control mice were injected with saline and killed 7 days later. Results are mean SEM of percent area stained in the CA3 region of the hippocampus, indicating signific ant reductions in diffuse A from 3-14 days. X-axis label C = saline injected contro l mice killed 7 days post injection; p < .05.
77 Gliosis, a possible link to the mechanism of A reduction in APP mice, was analyzed by looking at reactive glial markers. Representative micrographs from the CA3 subregion of the hippocampus of APP mi ce are presented in Figure 16. Reactive microglia were immunohistochemically anal yzed for CD45 expression in both APP and nontransgenic mice. Saline-injected APP mice demonstrated intense reactivity in conjunction with putative amyloid deposits (F ig. 16A). LPS treatment resulted in a widespread activation of microglia (Fig. 16B ). LPS similarly activated microglia in nontransgenics (not shown). As shown in Pa nel C, the increase in CD45 immunostaining was time dependent ( F(9,48) = 8.6, p < .0001). Post hoc analysis indicated a statistically significant increase beginning at 3 days, peak ing at 7 days, and remaining somewhat elevated throughout the time course to 28 days Notably, the APP mice were significantly more reactive than th eir nontransgenic littermates at 7 days with CD45 immunostaining ( p < .01). Astrocytes in both APP and nontransgenic mice were detected using the astrocyte specific marker GFAP, as shown for the CA3 subregion in Figure 16. The hippocampus of saline-injected mice showed many GFAP+ astrocytes (APP mouse shown in Fig.16D). LPS treatment resulted in a widespread activatio n of astrocytes (Fig. 16E). The increase in GFAP immunostaining was time dependent ( F(9,48) = 3.7, p < 0.005), shown in Panel F. Post hoc analysis showed GFAP expression p eaking at 7 days, then returning to control levels by 14 days. This is in contrast to CD45 which remained elevated through the 28 day time point. LPS injection into nontransge nic mice resulted in a mild increase in GFAP immunoreactivity compared to saline injection, which was not statistically significant.
78 Fc RII/III immunostaining of microglia was in vestigated following LPS injection. Figure 16G-H shows the CA3 subregion analys is of APP mice. Saline-injected mice showed no notable immunoreactivity. LPS treat ment led to a widespread reaction in the injected hippocampus of APP mice (Fig. 16H ). The increase in immunostaining was time dependent ( F(9,48) = 1.8, p < .05), shown in Figure 16I. Post hoc analysis indicated significant elevation of Fc RII/III at 3 and 7 days post in jection in APP mice. Levels returned to baseline by 14 days. Nontransge nic mice showed elevated reactivity as well, which was significant at 3 days when the data were analyzed separately from the transgenics. Figure 16. Time dependent glia l reaction to LPS injection. APP mice were injected with saline (A,D,G) or 4 g of L PS (B,E,H). Micrographs ar e representative of a 7 day survival for both groups, imaging the CA3 region of the injected hippocampus. Immunostaining for CD45 (A,B) showed intens e microglial reaction to LPS, which was significant from 3-28 days post injection (C ). Immunostaining for GFAP (D,E) showed reactive astrocytes responding to LPS, which was significant at 7 days post injection (F). Immunostaining for Fc RII/III (G,H) showed upregulation of this receptor in response to LPS, which was significant from 3-7 days (I). Scale bar = 50 m; x-axis label C = saline injected control mice; NT = nontransgenic; p < .05 versus control for that genotype; ** p < .05 versus APP controls as well as co mpared to NT at the same time point.
79 Figure 16. Time dependent glia l reaction to LPS injection.
80 The morphology of the microglia ac ross time was examined with CD45 immunostaining (Fig. 17). Saline-injected APP mice showed predominantly lightly stained, resting, ramified microglia with thin delicate processes (Fig. 17A). Treatment with LPS resulted in a production of rounded, ameboid microglia, staining intensely with CD45, appearing 1-7 days post inje ction (Fig. 17B). Starting at 7 days and lasting though 28 days, microglia acquired a hyper-ramifi ed, bushy appearance, with short thick processes stained heavily with CD45 (Fig. 17C). CD45-positive microglia, associated with putative amyloid deposits in saline-treated APP mice, have a bushy appearance and are heavily stained by CD45 antibodies (F ig. 17D). After LPS injection, ameboid microglia surrounded these putative plaques at 1, 3, and 7 days (Fig. 17E). The microglia resumed their bushy appearance at 7, 14, and 28 days, and remained widely distributed throughout the hippocampus as well as associat ed with putative amyloid plaques (Fig. 17F). Particularly at the 7 day time point, there was overlap with multiple morphologically different mi croglia being detected.
81 Figure 17. Altered microglial mo rphology in response to LPS. CD45 immunostaining was examined at each time point. Micrographs are from the CA1 molecular layer of APP mice injected with saline 7 days earlier (pan els A,D), LPS 3 days earlier (panels B,E) or LPS 28 days earlier (panels C,F). Microgr aphs were chosen from a region free of putative amyloid plaques (A-C) or a field centered on a putative amyloid plaque (D-F). Scale bar = 50 m.
82 Discussion These findings provide evidence of an innate mechanism of diffuse A removal present in APP transgenic mice. LPS injec tions into the hippocampus led to a timedependent clearance of parenchymal A These results confirm those reported earlier by our lab and others (Dicarlo et al., 2001; Quinn et al., 2003). Congophilic plaque load was unaffected, suggesting these different pools of amyloid are selectivel y accessible by glial cells or other clearance mechanisms. Pote ntial mechanisms for removal of diffuse A include phagocytosis by glial cells, secretion of degradative enzymes, or facilitation of A clearance from the brain by mechanisms other than phagocytosis. Scavenger receptors may represent the link between LPS and A as these receptors have been shown to bind both liga nds, and can mediate A phagocytosis in vitro (Husemann et al., 2002). Though both microglia and astr oglia have phagocytic capabilities in vitro the time course of A reduction observed in these experime nts most closely mirrors that of microglial activation. Furthermore, early tim e points were associated with a shift in microglial morphology to an ameboid shape that is associated with phagocytic capability in other systems. Another marker for a phagocytic state of microglia is Fc receptor expression. Fc RII/III were up-regulated on microglial cells in response to LPS from 3-7 days post injection. Previous studies performed by our laboratory have demonstrated that anti-A antibodies injected into the brain of A PP mice stimulated microglia to remove A in a process that appears at least partially Fc receptor mediated (Wilcock et al., 2003, 2004a,b). The endogenous ligand for Fc receptor is the Fc region of IgG. However, there is no evidence in the current studies that anti-LPS or anti-A antibodies have been
83 produced in the LPS-injected mice, although su ch a possibility cannot be ruled out. On the other hand, it has been demonstrated that LPS can induce Fc expression in the absence of IgG, possibly as part of cell activa tion. In addition, Fc receptor expression by macrophages can be regulated by cytokines as well as by LPS (Amigorena et al., 1989; Lynch et al., 1990; Laszlo and Dickler, 1990; Loughlin et al., 1992; Keller et al., 1994). Thus, it is possible that the induction of Fc RII/III reflects another facet of microglial activation and not the specifi c presence of immunoglobulin. The microglial response, though decreasing with time, remained significantly activated throughout the time course. At 28 days post injection, CD 45+ microglia were still significantly activated, and had adopt ed a hyper-ramified, bushy appearance. The glia were both plaque-associated as we ll as widespread thro ughout the injected hippocampus. It was interesting that the a ppearance of microglia at 28 days was very similar to those associated with plaques in our control, saline treated APP animals. Our experiments also demonstrated exacerbated gliosis after 7 days in the LPS dosed APP transgenics compared to their nontransgeni c littermates. Both CD45+ microglia and GFAP+ astrocytes were more intensely immunol abeled in the transgen ics. These results suggest a role for A in the inflammatory response. An intriguing observation occurred at the 28 day time point. A levels, which had been significantly reduced from 3-14 days post injection accumulated rapidly between 2 and 4 weeks after LPS administration, and returned to pre-in jection levels after one month. The original A burden in these mice developed over their lifespan of 16-17 months. This finding is highly significant because it suggests that amyloid pools are more labile than previously assumed. It is unclear whether this rapid re-accumulation
84 resulted from cessation of A removal mechanisms, enhancement of A deposition or both. If the LPS molecule is required for sufficient microglial activation and A removal, complete clearance of LPS might signal cessation of removal mechanisms as the immune response is no longer stimulated. On the other hand, evidence of enhanced APP processing in response to insult is well documented. Chr onic neuroinflammation and trauma-induced injury models both show ed increases in APP and its processing (Siman et al., 1989; Kawarabayashi et al., 1991 ; Griffin et al., 1994; Banati et al., 1995; Hauss-Wegrzyniak et al., 1998; Sugaya et al ., 1998). Kainate lesions also increased APP expression in astrocytes (Siman et al., 1989; Wright et al., 1999). Enhanced alphasecretase as well as presenilin activity were shown in other injury models (Brugg et al., 1995; Pennypacker et al., 1999). Future e xperiments will examine expression and proteolytic processing of APP after LPS admini stration to ascertain whether they play a role in this rapid re-accumulation of AB burden.
85 Chapter 4 Dexamethasone Suppresses LPS -Induced Microglial Activation and Amyloid Clearance in APP Transgenic Mice Abstract Inflammation has been argued to play a primary role in the pathogenesis of Alzheimers disease (AD). Mice transgenic for mutant human amyloid precursor protein (APP) have amyloid deposits, gliosis, a nd cognitive impairment, without apparent neurodegeneration. Previous efforts by our group to use lipopolysaccharide (LPS) to elicit neurodegeneration in APP mice were unsuccessful. Instead we saw a surprising reduction in A burden concurrent with the inflammato ry response. In studies described here, we clarify the mechanisms invol ved in LPS mediated removal of A using dexamethasone to inhibit the microglia re sponse. APP mice were intrahippocampally injected with LPS and survived 3 or 7 days with or without dexamethasone co-treatment. Brain tissue was then analyzed by immunohistochemistry and qua ntitative real time PCR. Total A burden was reduced 7 days after LPS injection; this was prevented by cotreatment with dexamethasone. Markers of general microglial activation, CD45 and complement receptor 3, were increased by LPS and inhibited by dexamethasone. In contrast, the Fc receptors II/III and scavenger re ceptor A were increased by LPS but unaffected by dexamethasone treatment. The implications of the glial response to LPS and dexamethasone are two-fold. First, cells do not require general activation in order to
86 upregulate their expression of receptors capable of mediating phagocytosis. However, in order to effectively clear out foreign or toxic materials (such as LPS, A etc.) general cell activation is necessary. Introduction Alzheimers disease (AD) is one of ma ny neurodegenerative disorders marked by chronic inflammation (Akiyama et al., 2000) Although amyloid plaques in human AD brain are surrounded by reactive astrocytes a nd microglia, the cellular response is not well understood (Itagaki et al.,1989 ). Acute inflammation can be neuroprotective, aiding in removal of pathogens and protecting cel ls (Streit, 2002). Conversely, excessive, chronic inflammation is thought to cause neuron death via an autotoxic mechanism (McGeer & McGeer, 2002). Current transgen ic mouse models replicate the amyloid pathology, glial response, and cognitive im pairment of AD, without apparent neurodegeneration (Gordon et al., 2001, 2002). Based on the inflammation hypothesis of AD, we have created a model using LPS intracranial injections into mice carrying a mutant human amyloid precursor protein (APP) transgene. LPS is expressed on the cell wall of Gram-nega tive bacteria, and can cause an inflammatory response by the hos t (Palsson-McDermott & ONeil, 2004). Once in the host, LPS binds to circulating LPS binding protein, which subsequently transfers the LPS to cell membrane bound cluster differentiation marker CD14 (Heumann & Roger, 2002). Toll like receptor 4 (TLR4), as well as myeloid differentiation protein MD2, form a complex through which LPS:CD14 can transduce a sign al (Thomas et al., 2002).
87 We have reported previously that direct injection of lipopolysaccharide (LPS) into the hippocampus of APP mice resulted in reductions in A burden (DiCarlo et al., 2001; Herber et al., 2004b). Similar findings were s een after intraperitoneal injections of LPS (Quinn et al., 2003) although systemic ad ministration to young mice may have the opposite effect and precipitate amyloid deposit s where none previously existed (Qiao et al., 2001). A potential mechanism of A removal includes phagocytosis (D'Andrea et al., 2004). In vitro studies showed that both microglia and astrocytes respond to LPS with phagocytosis, as well as with production of poten tially cytotoxic agents such as cytokines and reactive oxygen species (Kalmar et al., 2001 ; Shaffer et al., 1995). Wyss-Coray and colleagues also showed reductions in amyloi d deposits when brain slices from APP mice were coated with astrocytes (2003). Here we describe experiments that e xplore the glial contribution to the LPS associated A reductions. Intrahippocampa l LPS injections reduced A burden and stimulated microgliosis. The microglia res ponse was then inhibited with dexamethasone, resulting in A burdens at control levels. Materials and Methods Mouse Strains Transgenic mice were bred to devel op Alzheimers like pathology using Tg2576 APP mice as described previously (Holco mb et al., 1998). APP mice and their nontransgenic littermates aged 17 months were used in this study. Animals were grouphoused prior to surgery under a 12 hr light-dar k cycle with free access to chow and water.
88 Intrahippocampal Injections Mice were anesthetized using isoflurane and immobilized in a stereotaxic apparatus. A single, one microliter injection of either saline or 4 g/ l of LPS ( Salmonella abortus equi Sigma, St. Louis, MO) was de livered over a two minute period into the hippocampus (coordinates from bregma : 2.7 mm posterior, +/-2. 5 mm lateral, and -3.0 mm ventral). The incision was cl osed with wound clips, isoflurane was discontinued, and the animal was revived unde r ambient conditions. All mice completely recovered within five minutes. Animals were singly housed for the post treatment survival period under standard vivarium cond itions. We used at least 5 mice (5-10) for each genotype and treatment, balanced for ge nder. Post LPS injection, the survival period was either 3 or 7 days. Drug Administration Two different time points were evalua ted. A 3 day study was conducted after intrahippocampal LPS injection. Within one hour of injection, mice were administered dexamethasone (5 mg/kg, ip), followed by twice daily administration for a total of 3 days. A 7 day study was conducted by pre-treating w ith dexamethasone for 24 hours, injecting with LPS, then continuing twice daily dexamethasone for 7 days. Tissue Preparation Mice were overdosed with pentobarbital, a nd then perfused transcardially with 25 ml of normal saline. Tissue was collected for RNA analysis by removing the hippocampus and storing at -80 C. For immunohistochemistry, the brain was post fixed in 4% paraformaldehyde solution for 24 hour s, then processed through a cryprotection schedule of 10, 20, and 30% sucrose. Brai ns were sectioned horizontally at 25 m on a
89 freezing stage using a sliding microtome. Sections were then stored in Dulbeccos phosphate buffered saline pH 7.4 (DPBS) with 100 mM sodium azide at 4 C. Histology Immunohistochemical analysis was perf ormed for each marker using six, 25 m free-floating sections spaced 200 m apart through the hippocam pus. Details of this procedure are described elsewhere (Gordon et al., 1997). Briefly, sections were blocked for endogenous peroxidases (10% methanol and 3% hydrogen peroxide in 80% DPBS), washed with DPBS, and then permeabili zed (100 mM lysine, 0.2% triton x-100, 4% normal serum in DPBS). Sections were th en incubated overnight in the appropriate primary antibody (Table 6). The following day, se ctions were washed, and then incubated in appropriate biotinylated s econdary antibody. After another cycle of washes, the tissue was incubated with Vectastain Elite ABC kit (Vector Laboratories, Burlingame, CA). The tissue was then washed and stained w ith a nickel: diaminobenzidine: peroxide system, followed by final washes. In the case of A immunostaining, nickel enhancement of the color development was not used. The extent of nonspecific binding was assessed in the absence of primary antibodi es for all assays. Sections were mounted onto slides, dehydrated, and c over-slipped with DPX (E.M. Sciences, Fort Washington, PA). Slide-mounted untreated sections were st ained with Congo red to detect compact amyloid plaques. Slides were briefly hydrated in water, and then incubated in alkaline alcoholic saturated sodium chloride (AAS SC) for 20 minutes, followed by 30 minutes in 0.2% Congo red in AASSC. Slides were then quickly dehydrated through a graded ethanol series, cleared in xylene, and cover-slipped.
90 Table 6. Primary Antibodies Used for Immunohistochemistry Antigen Antibody Type Source Catalog # Titer A Rabbit polyclonal Kind gift from Dr. Paul Gottschall N/A 1:10000 CD45 Rat monoclonal Serotec MCA1031G 1:10000 CR3 Rat monoclonal Serotec MCA711 1:3000 Fc RII/III Rat monoclonal PharMingen 553141 1:3000 SRA Rat monoclonal Serotec MCA1322 1:3000 Immunostained sections were imaged at 100x magnification, focusing on the CA1, CA3, dentate gyrus, and hippocampal fissu re subregions of the hippocampus, as well as the anterior cortex. Images were then analyzed for area percent positive stain using Image-Pro Plus software (MediaCybernetics, Silver Spring, MD). Area percent stained was calculated for each individual regi on of the hippocampus, and averaged to calculate the staining for the entire hippocampus Results were averaged for all sections from each animal, and the treatment groups compared by one way analysis of variance (ANOVA) followed by post hoc anal ysis using Fishers protec ted least squares difference (PLSD, significance at p < .05). In some cases, results are reported in terms of fold change from control levels. In other instan ces, the ratio of staining in the hippocampus to that found in the anterior cortex was us ed to normalize for the variable amyloid deposition in individual mice.
91 RNA Analysis Tissue from the hippocampus was analy zed for mRNA expression by reverse transcription followed by quantitative real time polymerase chain reaction (qRT-PCR), as previously described (Dickey et al., 2003) RNA was extracted from the injected hippocampus using Qiagens Rneasy procedure (Valencia, CA). The purified RNA was then assayed using Molecular Probes RiboGreen RNA quantitation kit (Molecular Probes, Eugene, OR). Each sample was then diluted in water to a final concentration of 50 ng/l or 5 ng/l. The 50 ng/ l samples were run on a 1% agarose gel using ethidium bromide/ultraviolet detection to verify RNA integrity. A composite of all RNA samples was made in order to prepare a standard curve ranging from 0.2 50 ng/l. Reverse transc ription with mMLV (Invitrogen, Carlsbad, CA) was then performed on the standard dilutions and 5 ng of each sample RNA solution. The resulting cDNA was then s ubjected to qRT-PCR using SYBR Green PCR Master Mix (Molecular Probes, Eugene, OR). Tabl e 7 contains the primer sequences used in the qRT-PCR reaction. For each mRNA of interest, a melt curve analysis of the qRT-PCR product was pe rformed after cycling was complete. Fluorescence was read in 1 C increments from 56-99 C. Primer pairs were designed to yield a single, specific product, with no evidence of primer-dimer formation.
92 Table 7. Primer Sequences for RNA Analysis Transcript Primer Pairs 18S rRNA F: 5-GTAACCCG TTGAACCCCATT-3 R: 5-CCATCCAATCGGTAGTAGCG-3 CD45 F: 5-CAGAGCATTCCACGGGTATT-3 R: 5-GGACCCTGCATCTCCATTTA-3 IL1 F: 5-CTCATTGTGGCTGTGGAGAA-3 R: 5-GCTGTCTAATGGGAACGTCA-3 Fc RIIb F: 5-GGAAGAAGCTGCCAAAACTG-3 R: 5-CCAATGCCAAGGGAGACTAA-3 SRA F: 5-GACGCTTCCAGAATTTCAGC-3 R: 5-ATGTCCTCCTGTTGCTTTGC-3 For each mRNA of interest, the threshold cycle [C(T)] values from the qRT-PCR were determined. To create a linear curve, the log of the ng of RNA in each standard was plotted against the average C(T) value (each sample and standard was analyzed in triplicate). Next, the log of the ng of RNA in each sample were calculated, and then converted to ng of RNA. For a given reverse transcription, all targets of interest were normalized to results for 18S ribosomal RNA ( 18S rRNA) for the same sample to control for variations in the starting RNA concentrati on and reaction efficienc y. The ratio of the ng of target RNA to ng of 18S rRNA was calcu lated, and the results subjected to one way ANOVA, followed by Fishers PLSD with significance taken at p < .05. The final results are presented as fold change vers us saline injected control mice.
93 Results Intrahippocampal injections of LPS were made into APP transgenic mice and the subsequent neuroinflammatory effects on amyl oid load evaluated. We first looked at brain amyloid burden, considering both total A and Congophilic amyloid deposits. Figure 18A-D shows the changes in total A burden after treatment. The micrographs are representative of the injected hippocam pus of mice which survived 7 days. Control mice showed substantial amounts of A immunostaining, includi ng focal aggregations scattered throughout the molecular layer of the CA subfields as well as laminar accumulations along the hippocampal fissure and outer molecular layer of the dentate gyrus (Fig. 18A). In contrast LPS treated mice showed less A staining throughout the hippocampus (Fig. 18B). Reductions in the laminar accumulations were particularly striking. Mice both injected with LPS and co -treated with dexamethasone showed many deposits and laminar accumulation (Fig. 18C), similar to control mice. The area percent positive stain was calculated for each region of the hippocampus, with the greatest effects seen at the hippocampal fissure. The average area percent of the entire hippocampus was compared to the amount of A staining in the ipsilateral anterior cortex to control for inter-animal variability. The mean area percent standard error of the mean for the anterior cortex was not significantly different among the groups: 18.5 3.6 saline, 18.8 2.9 LPS, 14.8 1.3 LPS-dexamethasone at 3 days; 12.5 2.9 saline, 19.4 2.2 LPS, 15.4 2.2 LPS-dexamethasone at 7 days. Statis tical analysis revealed that the LPS treated group surviving 7 days had significant reductions in A load in the hippocampus compared to saline injected control mice. This effect was completely blocked by
94 concurrent administration of de xamethasone. No changes in A load were found at 3 days in any of the groups (Fig. 18D). The effect of LPS injection on compact plaques was evaluated using Congo red staining, shown in Figure 18E-H. The microgra phs are representative of the injected hippocampus of mice which survived 7 days. Saline injected mice showed several focal deposits along the hippocampal fissure (Fig. 18E). LPS treated mice showed similar numbers and sizes of deposits in the hippocam pus in the absence (Fig. 18F) and presence (Fig. 18G) of dexamethasone at both 3 and 7 days. These results were analyzed in the same manner as for A immunostaining and are presente d in Figure 18H. There were no statistically significan t effects of LPS or dexamethasone on Congo red levels at either time point.
95 Figure 18. LPS injection reduced diffuse but not compact amyloid deposits and was reversed by dexamethasone APP mice were injected with saline, LPS, or LPS plus concurrent dexamethasone treatment for 3 or 7 days. Micrographs ar e representative of animals survived 7 days. Immunostaining for A in saline injected mice showed many deposits throughout the hippocampus (A). LPS injection reduced the total amount of A staining in the hippocampus (B). LPS inj ection plus dexamethasone co-treatment resulted in many A deposits (C), similar to saline injected mice. The LPS-induced reduction in A was significant at the 7 day time point and blocked by dexamethasone (D). In contrast, similar amounts of Congoph ilic deposits were seen in saline injected (E), LPS injected (F), and LPS plus dexame thasone co-treated groups (G). There was no significant difference in Congo red staining be tween any of the treatment groups. Scale bar = 500 m, p < .05.
96 Figure 18. LPS injection reduced diffuse but not compact amyloid deposits and was reversed by dexamethasone
97 Two markers of general microglial activ ation, CD45 and CR3 were analyzed by immunohistochemistry and are shown in Fi gure 19. All micrographs are from the injected hippocampus of mice which surviv ed 7 days. CD45 (Fig. 19A) and CR3 (Fig. 19E) stained reactive microglia in association with putative amyloid deposits in saline injected mice. LPS injection resulted in a widespread activation of both CD45-positive (Fig. 19B) and CR3-positive (Fig. 19F) microglia, throughout the hippocampus and nearby cortices. The insets in these pane ls show higher power micrographs of the microglial morphology. Dexamethasone treatme nt of LPS -injected mice blocked the elevation of CD45 (Fig. 19C) and CR3 (Fig. 19G) staining, similar to that observed in saline injected mice. Quantitation of the fold change of the area percent stained in the dentate gyrus compared to saline injected control mice is shown in Figure 19D for CD45 and Figure 19H for CR3. Similar effects we re seen on all regions of the hippocampus, but the injection site was typically in the hilus of the dentate gyrus. Two way ANOVA for CD45 revealed significant effects of treatment ( F2,46 = 39, p < .0001), time ( F1,46 = 5.8, p = .02), and interaction ( F2,46 = 14, p < .0001). Similar results for CR3 showed significant effects of treatment ( F2,46 = 17, p < .0001), time ( F1,46 = 76, p < .0001), and interaction ( F2,46 = 3.5, p = .04). Post hoc analysis conf irmed that both the CD45 and CR3 responses were significantly elevated at 3 days and peaked at 7 days after LPS injection. Dexamethasone inhibited this r eaction at both time points; levels were no different from controls.
98 Figure 19. CD45 and CR3 are induced by LP S and inhibited by co-treatment with dexamethasone. APP mice were injected with sali ne, LPS, or LPS with concurrent dexamethasone treatment for 3 or 7 days. Micrographs are representative of animals survived 7 days. Saline injected animals s howed detectable levels of CD45 (A) and CR3 (E). Both CD45 (B) and CR3 (F) showed wide spread increases after LPS injection; the insets depict higher magnification of reactiv e microglia. CD45 and CR3 also showed an effect of time; both were signi ficantly elevated at 7 days compared to 3 days after LPS injection. Dexamethasone co-treatment in hibited CD45 (C) and CR3 (G) at both time points. Quantitation of the fold change versus saline injected controls is presented in Panel D (CD45) and Panel H (CR3). Scale bar = 500 m, p < .05 compared to saline injected controls, # p < .05 compared to the 3 day time point.
99 Figure 19. CD45 and CR3 are induced by LPS and inhibited by co-treatment with dexamethasone.
100 Two markers of the phagocytic capabilitie s of microglia, the low affinity Fc gamma receptors Fc RII/III and scavenger receptor SRA, were also examined by immunohistochemistry and are shown in Fi gure 20. All micrographs are from the injected hippocampus of mice which survived 7 days. Fc RII/III staining was associated with putative amyloid deposits, particularly along the hippocampal fissure in saline injected mice (Fig. 20A). LPS injection resulted in a marked response throughout the hippocampus (Fig. 20B). Dexamethasone treatment of LPS injected mice had no effect, and many Fc RII/III-positive microglia were seen (Fig. 20C), similar to mice who had received LPS injection only. The insets in Figure 20C and G show higher power micrographs of the microglia. Quantitation of the fold change of the area percent stained in the dentate gyrus compared to salin e injected control mice is reported. Fc RII/III staining in LPS injected and LPS-dexamethasone co-treated mice after 3 and 7 days were significantly elevated compared to controls when analyzed by ANOVA (Fig. 20D: F( 2,46) = 7.8, p = .001). There was no significant effect of time, with similar increases at both 3 and 7 days. SRA immunostaini ng results were similar to Fc RII/III. Saline injected mice showed little SRA immunoreactivity (Fig. 20E). LPS injection resulted in a widespread reaction throughout the hippocampus (Fig. 20F ). Dexamethasone treatment of LPS injected mice had no effect, and many SRA-positive microglia were seen (Fig. 20G), similar to mice who had received LPS injec tion only. Statisti cal analysis of SRA staining using ANOVA revealed a significan t effect of LPS treatment (Fig. 20H: F( 2,46) = 4.0, p = .03). Note that because immunostainin g in the control condition was very low, fold inductions for this marker are extremely large, and highly variable. SRA levels in LPSinjected and LPS plus dexamethasone co-treated mice were increased by 200 fold
101 after 3 days, and 100 fold after 7 days. An insignificant effect of time was noted, with similar inductions of SRA at 3 and 7 days. Figure 20. FcRII/III and SRA are induced by LPS but not inhibited by co-treatment with dexamethasone. APP mice were injected with sali ne, LPS, or LPS with concurrent dexamethasone treatment for 3 or 7 days. Micrographs are representative of animals survived 7 days. Low levels of Fc RII/III were seen in saline injected mice (A). No SRA staining is evident after saline injection (E). Both Fc RII/III (B) and SRA (F) showed widespread increases af ter LPS injection. Only a m ild, insignificant effect of time was noted. Dexamethasone co -treatment did not inhibit Fc RII/III (C) or SRA (G) levels at either time point; the insets depict higher magnification of microglia. Quantitation of the fold change versus saline injected controls is presented in Panel D (Fc RII/III) and Panel H (SRA). Scale bar = 500 m, p < .05 compared to saline injected controls.
102 Figure 20. FcRII/III and SRA are induced by LPS but not inhibited by co-treatment with dexamethasone
103 Based on previous findings that most R NAs peaked at 3 days (Herber et al., 2004a), several markers were examined at th at time point, as shown in Figure 21. The relative expression of CD45 af ter LPS injection or LPS-dexamethasone co-treatment (compared to saline injected control mice) is shown in Figure 21A. ANOVA revealed a significant effect of dose ( F2,17 = 4.5, p = .03). Post hoc analysis showed LPS significantly increased CD45 mRNA levels, which was inhibited by dexamethasone cotreatment. Results from post hoc analysis of the cytokine interleukin 1 beta (IL1 ) levels also showed a significant increase due to LPS, which was ameliorated by dexamethasone (Fig. 21B). In contrast, Fc RIIb (Fig. 21C) and SRA (Fi g. 21D) mRNA were induced by LPS and unaffected by dexamethasone co-t reatment. ANOVA rev ealed a significant effect of treatment for both markers (Fc RIIb: F2,17 = 4.8, p = .02; SRA: F2,17 = 8.4, p = .003). Post hoc analysis showed LPS and L PS plus dexamethasone co-treatment groups were both significantly elevated compared to controls for Fc RIIb and SRA. Figure 21. LPS stimulated gene transcripts have a pattern si milar to protein expression. APP mice were injected with saline, LPS, or LPS with concurrent dexamethasone treatment for 3 days. Graphs depict RNA anal ysis of the injected hippocampus compared to saline injected controls. CD45 (A) and IL1 (B) transcripts were induced by LPS and inhibited by dexamethasone. In contrast, Fc RIIb (C) and SRA (D) transcripts were induced by LPS but not inhibited by conc urrent dexamethasone treatment. p < .05.
104 Figure 21. LPS stimulated gene transcripts have a pattern si milar to protein expression.
105 Discussion Here we have shown the ability of th e innate immune system to remove A from the brains of APP transgenic mice. Intrahi ppocampal injections of LPS reduced the total A burden without affecting Congophilic plaque s, confirming previous reports (DiCarlo et al., 2001; Herber et al., 2004b). Thes e findings indicate that the diffuse A material is more sensitive to whatever mechanisms are activated by the LPS injection compared to compacted material. As a first attempt to identify such mechanisms, we partially inhibited the inflammatory reaction to LPS injection with systemic dexamethasone administration. This resulted in attenuati on of the amyloid clearance by LPS, and a significant reduction in most mi croglial markers which were elevated by LPS. Therefore it is likely that the glial reaction to LPS wa s responsible for the removal of diffuse A In an effort to characterize the glial re sponse to LPS, we evaluated both general markers of activation (CD45 and CR3) as we ll as receptors involved in phagocytosis (Fc RII/III and SRA). Though CD45, CR3, Fc RII/III, and SRA were all induced by LPS injection, only CD45 and CR3 were i nhibited by subsequent dexamethasone treatment. RNA analysis from the hippocampus of LPS injected versus LPS plus dexamethasone co-treated mice confirmed the immunohistochemistry results. The inhibition of CD45 by dexamethasone ha s also been reported after anti-A antibody treatment, thereby preventing an tibody-mediated removal of A (Wilcock et al., 2004a). Similarly, in another report, dexamethasone in hibited CR3 levels after LPS injection into the substantia nigra (Castano et al., 2002). In contrast, glucocorticoids actually enhanced certain Fc receptor subtypes in some mode ls (Kizaki et al., 19 96; Sivo et al., 1993;
106 Yamaguchi et al., 2001). We have previously shown that Fc RII/III and SRA expression after LPS injection was transient, peaking from 3-7 days and returning to baseline by 14 days post injection (Herber et al., 2004a,b). This was in st ark contrast to CD45 and CR3 which remained elevated out to 28 days post injection. Promoter an alysis of all four genes shows little similarity which may acc ount for their differential regulation in response to LPS and glucocor ticoid treatment (Grewal et al., 2001; Nishimura et al., 2001; Timon & Beverley, 2001). Both microglia and astrocytes are capable of phagocytosing A in vitro and in slice preparations (Shaffer et al., 1995; Wyss-Coray et al., 2003). The mechanism by which this is accomplished in vivo is unclear and as yet unprov en. Scavenger receptors, -integrins, complement proteins lipoprotein receptor relate d proteins, and receptors of advanced glycation end products can all bind A and mediate endocytosis (Herz & Strickland, 2001; Huseman et al., 2002; Koeh igsknecht & Landreth, 2004; Schmidt et al., 2001). Although Fc RII/III are increased in our LPS m odel, we have no direct evidence of antibody production, and upregulation of th ese receptors was probably due to the general activation of the cell in preparation for initiation of an acquired immune response (Amigorena et al., 1989; Iwasaki & Medzhi tov, 2004; Keller et al., 1994; Laszlo & Dickler, 1990). Though scavenger receptor B (SRB) has been implicated in AD (Coraci et al.., 2002), no information is available fo r SRA. We previously examined SRB mRNA following LPS injection and saw only a sma ll increase at 6 hr post injection, and therefore have shifted focus to SRA (Herber et al., 2004a). SRA seems to be at least partially responsible for astrocytemediate amyloid removal, though similar results with microglia are not available (Wyss-Coray et al., 2003).
107 The implications of the glial response to LPS and dexamethasone are two-fold. First, cells do not require general activati on in order to upregulat e their expression of receptors capable of mediating phagocytosis. However, in order to effectively clear out foreign or toxic materials (such as LPS, A etc.) general cell activation is necessary.
108 Conclusions Alzheimers disease was first described by Alois Alzheimer in 1907 in an article entitled, On an Unusual Illness of the Cerebral Cortex. In this seminal work, Alzheimer describes a 51 year old female patient w ho suffered from severe dementia. Upon autopsy, her brain was examined and found to be grossly atrophied. Histopathology revealed, the nucleus and the cell itself disintegrate and only a tangle of fibrils indicates the place where a neuron was previously located Many neurons, especially in the upper layer, have completely disappear ed. Distributed all over the cortex, but especially numerous in the upper layers, there are minute military foci which are caused by the deposition of a speci al substance in the cortex. The terms neurofibrillary tangles (NFT) and plaques or senile plaques, have since been coined for the typical pathology of AD. Additional aspects of the disease include inflammation and alterations in th e cerebrovasculature (Akiyama et al., 2000; Rhodin & Thomas, 2001; de la Torre, 2002). Several hypotheses have been proposed over the past century as to the cause of the disease. Recent therapeutic developments aim to remove or prevent amyloid deposition, ba sed on the amyloid hypothesis of AD. This hypothesis proposes amyloid deposition as the central event in the aetiology of Alzheimers disease (Hardy & Allsop, 1991). Such a view presumes neurofibrillary tangles, neurodegeneration, and inflammation (t he other pathological components of AD) to be a consequence of amyloid deposition.
109 The major component of plaques is the A peptide, which can aggregate to form amyloid deposits. This peptide originates fr om the amyloid precursor protein (APP), as a result of proteolytic cleavage by betaa nd gamma -secretases (Nunan & Small, 2002). The A peptide varies in length from 37-43 amino acids. A 1-40 is the predominant species, but A 1-42 is more hydrophobic and therefore more readily forms fibrils. In support of the amyloid hypothesis are the discoveries of severa l genetic mutations, all of which cause enhanced deposition of amyloid and ultimately lead to autosomal dominant AD, particularly at a young age (pre-senile). These mutations make up the portion of the population with familial Alzheimers disease, and account for approximately 5-10% of all AD cases. In 1991 Goate and colleagues reported mutations located on human chromosome 21 which ultimately mapped to the APP gene. A single amino acid substitution of valine to isoleucine, cl ose to the carboxy terminus of the A peptide, results in a more hydrophobic, and possibly more fibrillogenic, species with the added potential for enhanced gamma secretase cl eavage. Over the past 15 years, many mutations have been describe d near the c-terminus of A such as the Austrian, French, Florida, London, and Australian varieties (N unan & Small, 2002). Mutations located in the middle of the A peptide cause cerebral amyloid angiopathy (Van Nostrand et al., 2002). A double substitution at the N-terminus of A (K670N, M671L), known as the Swedish mutation, leads to enhanced cleavag e by beta secretase, and is the genetic mutation used in the Tg2576 transgenic mice desc ribed in our studies (H siao et al., 1996). These mice show ageassociated amyloid de position, reactive gliosis, and cognitive impairments, much like AD. Trisomy 21 leads to a triple dose of APP, and AD pathology can be seen in older Downs pati ents, even though the gene is not mutated.
110 In 1995, mutations in two additional gene s were identified as causal for AD. Sherrington and colleagues found missense muta tions in the gene encoding presenilin 1 (PS1), located on chromosome 14, which led to autosomal dominant AD. That same year Levy-Lehad and colleagues repor ted mutations in a similar gene, presenilin 2 (PS2), located on chromosome 1, which also led to autosomal dominant AD. Greater than 85 mutations in the presenilins have been identified in familial AD. Presenilin is part of the gamma -secretase complex, which proteolytically cleaves APP in conjunction with beta secretase to form A (De Strooper, 2003). Transgenic mice carrying a PS1 mutation (M146L, line 5.1) show little pathology, with no amyloid deposition (Duff et al., 1996). However, breeding the Tg2576 APP mice with PS1 mice leads to accelerated amyloid deposition and cognitive dysfunction in mice bearing both mutations compared to mice with mutations in APP only (Holcomb et al., 1998; Gordon et al, 2001; 2002). Associated with increased risk for AD are allelic forms of apolipoprotein E (APOE; Mahley & Huang, 2004). There are three APOE alleles expressed in humans (APOE2, 3, and 4) which are invo lved in lipid and cholestero l transport and homeostasis. The predominant allele is APOE3. Corder a nd colleagues identified a genedosing effect of the APOE alleles in 1993. An increased risk of late onset (>65 years old) Alzheimers disease was associated with th e E4 allele. In cases of AD with APOE4, there is increased plaque and tangle density as well as an earlier age of onset compared with the E3 allele (Marz et al., 1996). Although the mechanism remains uncertain, Colton and colleagues have demonstrated a link between the APOE4 allele and exaggerated innate immune activation (Colton et al., 2002a ,b, 2004; Czapiga & Colton, 2003).
111 The genetic mutations associated with familial AD as well as the gene dosing effects of APOE4 all increase the amyloid burden of the brain and are causal or associated with increased risk of AD. Though there are identifiable mutations in tau which cause tangle formation and lead to fr ontotemporal dementia or Picks disease, none of them are associated with AD (Frie dhoff et al., 2000). Thus, the genetic linkage to amyloid deposition has further stre ngthened the amyloid hypothesis of AD. Another approach to the study of AD is the inflammation hypothesis (McGeer & McGeer, 1995). According to this line of reasoning, it is not the lesions in AD brain (plaques, tangles, and neuron loss) but rather the inflammatory response to such lesions that leads to neuron loss and the clinical ma nifestations of the disease. Originally described by Celsus in the fi rst century A.D., peripheral in flammation is characterized by rubor (redness), calor (heat), dolor (pain), and tumor (swell ing). These responses are secondary to the primary reacti on of the tissue. The primary process is characterized by both acute and chronic phases (Hardman et al., 2001). During the acute phase, tissue damage causes the release of cellular conten ts which triggers changes in the local vasculature as well as migration of le ukocytes and macrophages to the site. The infiltrating immune cells can then release chemical mediators such as histamine, prostaglandins, leukotrienes, and cytokines. These chemical and cellular mediators of inflammation fight off infection as well as de stroy damaged and infected cells. There are three outcomes to injury: the host wins and inflammation is resolved, the host loses and dies, or there is a stalemate where chronic inflammation persis ts. This chronic phase is characterized by macrophage activation and re lease of chemical mediators which can
112 ultimately cause tissue damage. Typical examples of peripheral chronic inflammation are arthritis and inflammatory bowel disease. In the brain, swelling is the only seco ndary inflammatory response seen (no redness, pain, or heat). Swe lling is seen only under extreme circumstances such as after traumatic brain injury or during encephali tis. However swelling is not seen in neurodegenerative conditions such as Alzhei mers or Parkinsons diseases, or multiple sclerosis. Instead, a more subtle inflammato ry process can predominate in the brain, as characterized by reactive glia and the products they produce. A classic example of this type of silent brain infla mmation is Alzheimers disease. This process is mainly mediated by the innate immune system, w ith the endogenous microglia as the tissue specific macrophages. B and T cells are largel y absent from the brain and thus adaptive immune activation does not play a major role in the inflammation of AD (Akiyama et al., 2000). Similar to peripheral inflammation, there are acute and chronic inflammatory responses in the brain, resulting in microglial activation, release of chemical mediators, and the potential for neuron death. Cell death in the CNS is problematic due to the fact that neurons do not divide/rep licate, and there is little repopulation by stem cells. If a neuron dies as a result of an inflammatory process, the lesion is permanent. The exact mechanism of neuron death due to inflammation is probably a combination of factors. Reactive micr oglia, like all macrophages can produce a respiratory burst as well as cytotoxic elements such as complement proteins, cytokines, and prostaglandins (Colton et al., 19 92; 2000; Colton & Gilbert, 1993; Colton & Chernyshev, 1996; Czapiga & Colton, 1996; Akiyama et al., 2000). In vitro release of this chemical milieu can be triggered by phagocytic activities and/or exposure to
113 complement proteins, cytokines, prostaglandi ns, and microbial products (such as LPS), as well as the A peptide (McGeer & McGeer, 1995). The respiratory burst (leading to release of oxygen free radicals a nd nitric oxide) causes oxidati ve damage to proteins and the subsequent stress leads to apoptotic ca scades and neuron death. The complement cascade, both classical and alternative, can ul timately lead to formation of the membrane attack complex which participates in ce ll lysis (Shen & Meri, 2003). Cytokines and prostaglandins reinforce the re action of microglia in a self perpetuating cascade as the inflammation remains unresolved. The accumu lation of amyloid deposits suggests that microglia are unable to remove the deposists Microglia are associated with amyloid plaques and may be experiencing frustrated pha gocytosis as they are unable to effective clear the material (Colto n et al., 2000; Fonsesca et al., 2004; Henson, 1971). The microglia therefore remain act ivated in a chronic inflamma tory response. Though direct toxicity to neurons due to amyloid deposits a nd tangle formation is probable (Canevari et al., 2004; Davies, 2000), the progression ma y be slow without the additional inflammatory stimulus to intensify the pr ocess (McGeer & McGeer 1995). Bystander lysis may also occur whereby undamaged ne urons are killed as a part of this uncontrolled/chronic inflammatory response. Though reactive microglia have never been shown to kill healthy neurons in vivo a chronic inflammatory cascade may be neurotoxic in AD. McGeer and colleagues (1995) propose an autotoxic loop which could lead to neuron death, shown in Figure 22.
114 Figure 22. Autotoxic mechanism s in Alzheimers Disease. Supporting the inflammation hypothes is are studies showing multiple inflammatory cascades active in the AD br ain including complement, cytokines and chemokines, prostanoids, acute phase proteins and free radicals (Akiyama et al., 2000). Plaques and tangles are both inert, insoluble aggregates in vivo and glia mount an inflammatory response to these aggregates. Studies in the 1980s of human post mortem AD brain showed reactive microglia and inf iltrating T-cells (McGeer et al., 1987; 1988 a,b; Itagaki et al., 1988). Subsequently, the brain inflam matory response was correlated to AD and cognitive decline. Lue et al. (1996) measured plaque and tangle load as well as inflammatory markers such as comple ment proteins and MH CII in three groups: normal controls, high pathology controls (HPC), and overt AD patients. Both the HPC and AD brains had pathological diagnosis of AD with significant amounts of plaques and tangles. However, the HPC group had no s ynaptic loss, no overa ll brain atrophy, and significantly less inflammatory reaction. Th ese findings led the authors to conclude:
115 of all the pathological variables exam ined, the best predictors of synaptic changes (i.e. the variables that accounted for the highest proportions of synapse variance) are those related to inflammation, C5b9 immunoreactivity, a nd activated LN3+ microglia. Taken together, these data suggest that elderly patients ma y present at autopsy with profuse cortical plaques and entorhinal cortex NFTs but may not evidence synaptic loss unless these changes are accompanied by inflammatory reac tions. Inflammation may therefore be one of the fina l common pathways through which A deposits and NFTs manifest their ne urodegenerative effects A similar study in 2003 by Vehmas et al. demons trated a change in gliosis associated with the transition from probable to defin ite AD. Microgliosis was elevated in both groups compared to normal controls, with higher levels in the definite AD group. Astrogliosis was an even better correlate of th e disease, with significantly elevated GFAP levels in definite AD compared to possible AD. Inflammatory gliosis has proven to be a robust finding in AD brain, and is also seen in amyloid depositing transgenic mice (Gordon et al., 2002). In support of the inflammation hypothesis are several epidemiological studies showing a decreased risk for AD with anti-i nflammatory drug use (reviewed in McGeer et al., 1996). Early studies examined arthritis as a factor in AD and established an overall 0.556 (osteoarthritis) and 0.194 (rheumatoid arthritis) odds ratio (OR) indicating decreased risk for AD and possible protectiv e effects of the anti-inflammatory drugs commonly used by people with arthritis. When drug use was considered as a factor in AD, both steroids (OR = 0.656) and NSAIDS (OR = 0.496) decreased AD risk. A recent study conducted in the Netherlands considered exact drug and thera py duration as factors in AD (In t Veld et al., 1998). The odds rati o overall was 1.0. However, when the data was stratified for age, a protec tive effect was seen in persons less than 85 years old (OR =
116 0.53). Additionally, a greater protective effect was seen in patients with more than 6 months NSAID use. Thus, it seems that inhi biting inflammation coul d be beneficial in Alzheimers disease. Based on these findings, NSAIDs have b een administered to amyloid depositing transgenic mice, with impr essive results. Ibuprofen and NCX-2216 (a nitric oxide donating flurbiprofen derivative) signif icantly reduced amyloid burden in doubly transgenic APP:PS1 mice (Jantzen et al., 2002). Similar results for ibuprofen have been reported (Lim et al., 2000; Yan et al., 2003), a nd Eriksen et al. (2003) demonstrated that multiple NSAIDs reduced soluble A in singly transgenic APP mice. The actual A lowering mechanism of NSAIDs is a matter of controversy (reviewed in Gasparini et al., 2004). In vitro NSAIDS can inhibit A aggregation, and decrease A production. However, the A lowering effects seem to be i ndependent of cyclooxygenase (COX) activity as doses required to alter A levels are much high than those necessary to inactivate the COX enzyme (Weggen et al., 2001). The A lowering effects fo NSAIDs also seem to be independent of PPAR activity. Ciglitazone, a PPAR agonist, did not alter A levels in vitro (Sagi et al., 2003), nor did piogl itazone affect amyloid deposition in vivo (unpublished data by Paul Jantzen in our lab). Several NSAIDs have been used in hu man trials to establish effects on AD, though results to date have been discourag ing. Rofecoxib, naproxyn, and a combination of diclofenac/misoprostol have all been eliminated as potential AD therapies in clinical trials (Aisen et al., 2003; Re ines et al., 2004; Scharf et al., 1999). Currently under clinical investigation ar e ibuprofen, indomethacin, a nd celecoxib, listed on the US governments web site of ac tive clinical drug trials ( http://clinicaltrials.gov ).
117 However, not all inflammation is regarded as detrimental. As mentioned previously there are three potential outcom es to injury/insult: the host wins and inflammation is resolved, the host loses and di es, or a stalemate is reached and chronic inflammation ensues. The first outcome (t he host wins, tissue is repaired, and inflammation is resolved) reve als the protective side of inflammation. In the brain, reactive glia can aid in remova l of pathogens and cellular debr is, as well as supply trophic factors. In AD, a key therapeutic strategy uses vaccination against the A peptide and stimulates the immune system resulting in decreased amyloid burden (reviewed later in this discussion). Therefore, stimulati ng the immune system, including microglia, represents a potential means of removing brain amyloid in AD. The model we used for the studies descri bed in this dissertation combined both the amyloid and inflammation hypotheses of AD in an effort to create a system that more closely resembled the human condition. We injected the inflammatory agent lipopolysaccharide (LPS) into the hippocampus of various transgenic mice, in particular amyloid -depositing APP mice. LPS triggers the innate immune response leading to inflammation. The expected outcome was enhanced amyloid burden and/or neuron death, similar to that seen in human AD brain. In the first study, we examined 7 nAChRs in amyloiddepositing mice and during neuroinflammatory conditions. Our desi re to evaluate nico tinic receptors was based on several observations. First, cho linergic signaling is impaired in AD with declines in 7 nAChRs among other populations (Perry et al., 2001). Second, there have been numerous reports of nicotines protective effect for AD (reviewed in Rusted et al., 2000), and nicotine administration reduced th e amyloid burden in APP transgenic mice
118 (Hellstrom-Lindahl et al., 2004; Nordberg et al., 2002). Third, a subset of nicotinic receptors, those containing 7 subunits, can interact with the A peptide. In vitro A binds to 7 nAChRs and blocks the receptor altering calcium homeostasis and potentially impairing neuronal signaling (Dinel ey et al., 2001, 2002a,b; Grassi et al., 2003; Pettit et al., 2001; Wang et al., 2000a,b). In APP transgenic mice, there are reports of decreased 7 nAChRs in older animals, which correlated with amyloid burden and cognitive impairments (Dineley et al., 2001, 2002a,b). Lastly, 7 nAChRs have been implicated in the inflammatory response. In human AD brain, re active astrocytes colocalize with amyloid plaques and label for 7 nAChRs (Teaktong et al., 2003). There is also evidence that microglia can expresss 7 nAChRs (Shytle et al., 2004). These nicotinic receptors have also been studied in the periphe ral immune system. Wang and colleagues demonstrated that 7 negatively regulates inflammation, as 7 knock out mice showed an enhanced inflammatory respons e to LPS challenge versus wild type mice (2003). Based on the potential of 7 nAChRs to interact with A we further investigated this relationship in our model of LPS induced neuroinflammation (Herber et al., 2004c). We used several commercially available antibodies to detect 7 nAChRs by immunohistochemistry and Western blotting. We saw no changes in immunohistochemical labeling of neurons for 7 when comparing amyloid depositing mice (APP, APP:PS1) to nontransgenic mice, in c ontrast to previous reports (Dineley et al., 2001, 2002a,b). However, we confirmed the findings of Teaktong and colleagues by showing 7 nAChR immunoreactive astrocytes wh ich co-localized with Congophilic
119 deposits in APP and APP:PS1 mice. In orde r to determine if this was a result of 7 interacting with A or part of a more general infla mmatory response, we injected LPS into the hippocampus of se veral mouse models (nontransgenic, APP transgenic, and 7 knock out mice). This treatment resulted in substantial astrogliosi s and immunolabeling by 7 antibodies as determined by immunohistoche mistry, in all genotypes tested. These results were compared to Western blot, DNA, and RNA analysis and established the nonspecificity of the 7 antibodies. Recent reports have confirmed the nonspecificity of not only 7 but also 4 nAChR antibodies (Moser et al., 2004). Current work by Dineley and colleagues showed decreased soluble A in APP mice crossed with the 7 knockouts, thus there may be some role for 7 in Alzheimers disease, though determining the exact mechanisms will require ri gorous testing protocols (Dineley et al., 2004). The remaining chapters of this dissertation looked at the contribut ion of glia to the brain inflammatory response in order to gain insight from reactive gliosis and apply that information to the pathological processes of AD. Neurons, astrocyt es, and microglia all contribute to the inflammatory response seen in AD, though the precise role of each cell type in this reaction is unclear. In the earlier part of the twentieth cen tury, Rio-Hortega proposed that microglia are the resident macrophages of the centra l nervous system (1932). Microglia are immunocompetent and act as mediators of inna te immunity, but can also present antigen as part of the adaptive immune respons e (Streit, 2002). Sim ilar to peripheral macrophages, microglial activation has both toxi c and protective roles, thus these cells
120 are likely candidates for mediating inflammation, neurotoxicity, and/or A removal in AD brain. Activated microglia can produ ce many neurotoxic species such as complement proteins, cytokines, prostaglan dins, and reactive oxygen species (Akiyama et al., 2000). In 1996 Kreutzberg described vari ous microglial activati on states. Activated microglia undergo dramatic changes in morphology, and these changes are possibly linked to phagocytic versus antigen pres enting functions. Hauss-Wegrzyniak and colleagues have also described a bushy phenotype that arises during chronic LPS infusion into brain ventricles (1998). St reit and colleagues (1999) have proposed the following phenotypes for reactive microglia, shown in Figure 23.
121 Figure 23. Microglial activation states. Neuronal injury can trigger injury signals which activate resting (ramified) microglia. The br anches of these cells can become swollen (reactive/activated) or hyper-ram ified (an intermediate stage). Microglia can also further activate and become phagocytic.
122 In general, the microglial inflammatory re sponse to amyloid deposits is viewed as harmful (Akiyama et al., 2000). In vitro studies have demonstr ated the neurotoxic properties of both A peptides as well as inflammatory mediators such as cytokines and reactive oxygen species (Canevar i et al., 2004; Small et al., 2001; Walsh et al., 2002). However, in vivo neurodegeneration due to either A or inflammation has been difficult to demonstrate. Intracranial administration of LPS ha s been used historically as an in vivo model of neuroinflammation, activating both micr oglia and astrocytes. LPS is bound by the LPS binding protein which can th en associate with CD14 and TLR4 to transduce a signal through the cell (Heumann & Roge r, 2002; Thomas et al., 2002). Several pathways are activated including NF B, resulting in gene transcrip tion and the ultimate production of inflammatory mediators (Sen & Baltimo re, 1986). The LPS signaling cascade is illustrated in Figure 24.
123 Figure 24. Lipopolysaccharid e signaling cascades. Lipopolysaccharide (LPS) is bound in the circulation by LPS binding protein (LBP ). This complex then interacts with cell membrane associated CD14 and toll like recepto r 4 (TLR4) to transduce an intracellular signal. Various kinase cascades are induced (such as p38MAPK and I B) leading to the activation of transcription fact ors including STAT, AP1, and NF B. Down stream targets include cyclooxygenase (COX), nitric oxide synthase (iNOS), and superoxide dysmutase (SOD). Ultimately inflammatory me diators are produced such as prostanoids, nitric oxide (NO), tumor necrosis fa ctor (TNF), and interleukins (IL). Extracellular Intracellular LBP
124 Chapters 2 and 3 of this dissertation looke d at the time course of glial activation after LPS administration into either nontrans genic mice or APP transg enic mice. Several markers of gliosis were examined incl uding CD45, complement receptor 3 (CR3), Fc receptors, scavenger receptors, and GFAP. Both sides of the inflammatory response toxic versus protective were evaluated. Brain inflammation begins with the glial response to insult. Once cells are activated, they can produce inflammatory agents which can be cytotoxic. LPS administration into both transgenic APP and nontransgenic mice led to acute and chronic microglial activation (Herber et al., 2004a ,b). Two proteins, CD45 and CR3 were upregulated beginning as early as 24 hours after injecti on of LPS, and remained upregulated after 28 days. Resting and reactiv e microglia express these two proteins, and thus CD45 and CR3 are considered marker s of general microg lial activation. CD45 (leucocyte common antigen, LCA) is a pr otein tyrosine phosphatase. The endogenous ligand and the exact functioning of this prot ein in inflammation are unknown (Irie-Sasaki et al., 2003). Complement recept or 3 (CR3, CD11b, MAC-1) is a 2 integrin which mediates cell adhesion (Ehlers, 2000). In addition to the ex tracellular matrix, a prominent endogenous ligand for CR3 is complement component C3b1. Downstream activation pathways include MAP kinase activ ation and ultimately actin reorganization, promoting migration and phagocytosis. It is noteworthy that CR3 may also bind to LPS, thus activating the alternative pathway. In our studies, large increases in the levels of CD45, CR3, and the cytokines IL1 and TNF were seen after a single injection of LPS, indicating a significant inflammatory response. A spreading wave of inflammation was noted with the injected hippocampus respondi ng first, then adjacent cortices and the
125 striatum, and ultimately the entire brain (inclu ding frontal cortex and brainstem) showed microglia expressing high levels of CD 45 and CR3. A surprising finding was the continued expression of these two proteins in the hippocampus by microglial cells 28 days after LPS injection. This may indicate a state of vigilance by the microglia, with the cells primed for future action. Changes in cell morphology at various time points after LPS injection was examined with CD45 and CR3 immunostaini ng. Untreated nontransgenic mice had microglia which were faintly stained for CD 45/CR3 and had a ramified appearance with fine delicate processes. In untreated amyloid depositing mice (APP, APP:PS1), the microglia showed increased staining and ha d a bushy, reactive phenotype in association with deposits. After LPS administration, the microglia increased both CD45 and CR3 expression. Some had shorter, thicker, branched process, with a bushy shape, and were seen throughout the time course after LPS injection. Other cells were round or ameboid and were only transiently seen from 1-7 days after LPS. Though resting microglia can be distinguished from peripheral macrophages by their branched shape, round and/or ameboid microglia are impossible to discrimina te from peripheral macrophages. Thus, in vivo it is difficult to determine which exact cell type is being evaluated because all macrophages express similar markers. It is also difficult to draw correlations from studies conducted in vitro It is currently not possible to culture adult mouse microglia, thus neonatal or immortalized microglial cell lines are typi cally used (Yao et al., 1990; Colton et al., 1991; 1992a,b; Colton & Gilbert, 1993; Kopec & Carro ll, 1998). Studies using adult versus neonatal mouse astrocytes however show vast differences between the two populations, and similar disparities can be assumed for microglia (Wyss coray et al.,
126 2003; Paul et al., 2004). Thus, though micr oglia clearly become activated, produce inflammatory mediators, migrate, proliferate, phagocytose, and present antigen in vitro, the same functions are not as clear in vivo Though a widespread, significant inflammato ry response was seen, neuron death was not observed in any of our studies. This is different from the brains response to chronic LPS administration. Several public ations by Wenk and Hauss-Wegrzyniak and colleagues over the past 15 years have demons trated the effects of chronic infusion of LPS into the fourth ventricle of rats (H auss-Wegrzyniak et al., 1998a, b; 2000; 2002; Willard et al., 1999). Their protocol led to si gnificant astroand microgliosis as well as forebrain cholinergic neuron death. In cont rast, acute LPS injection into the brain parenchyma such as the anterior cortex and hippocampus does not triggered cell death, as demonstrated by our lab and others (DiCarlo et al., 2001; Herber et al, 2004a,b; Kim et al., 2000). However the substantia nigra, a part of the dopaminergic system implicated in the pathogenesis of Parkinsons disease, is susceptible to the degenerative inflammatory effects of acute LPS administration (Castano et al., 2002; Kim et al ., 2000). In our own studies, acute intrahippocampal LPS admi nistration did not cause detectable neurodegeneration as determined by cresyl violet and fluoroj ade staining, nor was synaptic dysfunction detected by synaptotag min levels, regardless of dose, time, or genotype examined (Herber et al., 2004a,b). LPS injection induced NF B expression (unpublished data) which has been linked to neuronal survival, and thus may be protective under the conditions of our protocol (Kassed et al., 2004). The only conditions which caused neuron death were co-inj ections of LPS and interferon gamma
127 (unpublished data). It is possible that the presence of both endogenous and exogenous inflammatory mediators are necessary to kill cells. The protective aspects of microglial activ ation include neurotrophic support, as well as clearance of cellular debris, forei gn substances, and microbial components. Microglia have demonstrable phagocytic f unctions, similar to peripheral macrophages (Kalmar et al., 2001; Yao et al., 1990; C zapiga & Colton, 1999). In our studies, stimulation of the innate immune system by intrahippocampal injection of LPS led to time-dependent reductions in A levels by as much as 70% (Herber et al., 2004b). The reductions in A were transient, occurring between 3 and 14 days after LPS injection, then returning to baseline leve ls after 28 days. Diffuse A material, but not compacted Congophilic deposits, were decreased after LPS tr eatment (it is interesting to note that stimulating microglia using an A antibody reduced both diffuse and compact deposits in a similar protocol; Wilcock et al., 2003; 2004a,b). The time course of A reductions after LPS injection was closely related to th e expression of microglial phagocytic markers and changes in cell morphology. We therefore believe that the microglia are removing the A thus performing their prot ective phagocytic function. Other research groups have administered LPS to accelerate A deposition, activate glia, and trigger neur otoxicity. The experimental conditions of reports vary including the type and age of transgenic animals used; type, dose, and route of administration of LPS; and postinjection survival time. In some cases amyloid deposition was triggered (Qaio et al., 2001; Sheng et al., 2003; Sly et al., 2001), while in others amyloid clearance resulted (DiCarlo et al., 2001; Quinn et al., 2003). When considering the species of A 1-40 and A 1-42, the main indicator of A reduction seems to
128 be closely related to plaque load and surv ival time. Aged animals with significant amyloid deposits, surviving several days post injection, showed clearance of A deposits after LPS treatment. Compact plaques dete cted by either Congo red or thioflavin-S staining were inconclusively affected. In our studies, we examined marker s of a phagocytic microglial phenotype including Fc and scavenger receptors at various ti me points after LPS injection. Resting microglia do not express appreciable levels of Fc receptors, though some plaque associated microglia in APP mice do stai n positively for the receptors. After LPS injection, both protein (Fc RII/III) and mRNA (Fc RII) levels increased, peaked between 24-72 hours, and then returned to low basal levels after a week. Fc receptors are traditionally linked to adaptive immune responses, with the Fc portion of the antibody:antigen complex as the ligand (reviewed in Gessn er et al., 1998). The bound complex can then be internalized, processe d, and antigen presented via MHCII. The ligand for Fc receptors is IgG and there are three cl asses, with class I a high affinity, and classes II and III low affinity receptors. Class I and III are activating; class II is inhibitory. The vari ous functions of Fc receptors as listed in the Gessner review include, clearance of antigen/antibody immune co mplexes, regulation of antibody production, enhancement of antigen presentation, anti body-dependent cell-mediated cytotoxicity, phagocytosis, degranulation, and activation of inflammatory cells. There is evidence that peritoneal macrophage s and B-cells express Fc RII after treatment with LPS, even in the absence of antibody:antigen complexes (Amigorena et al ., 1989). Microglia and bone marrow derived macrophages can also be induced to express Fc receptors after LPS or interferon gamma treatment (Keller et al ., 1994; Loughlin et al., 1993). In the
129 experiments described herein, the innate im mune system is involved, but we would not expect antibody production as part of an acqui red immune, thus Fc expression might be part of a general inflammatory response, priming the cell for phagocytosis should an adaptive response be mounted. Scavenger receptors A and B are also i nvolved in the recognition of LPS by the innate immune system and can mediate clearan ce of bacteria and vi ral components. In our studies, only a brief, mild increase in SRB mRNA was seen at 6 hr after LPS injection. We were unable to confirm SRB expression by immunohistochemistry as available antibodies showed nonspecific staini ng patterns. In contrast, a significant, thirty fold increase in SRA was seen both in mRNA and protein levels. Resting microglia did not express appreciable levels SRA, nor di d plaque associated microglia in untreated APP mice. After LPS injection, SRA increas ed with widespread expression, peaked between 24-72 hours, and then returned to low basal leve ls after a week. The cell population staining for SRA appeared to be microglia/macrophages based on their morphology, rather than astrocytes. Scavenger receptors are traditionally li nked to lipid metabolism and innate immune responses (Husemann et al., 2002; Kr eiger et al., 2001; Febbraio et al., 2001; Platt and Gordon, 2001). They are also calle d pattern recognition receptors as they nonspecifically recognize many bacterial a nd viral components, usually based on the presence of polyanionic ligands. Both mi croglia and astrocytes can express various scavenger receptors, depending upon age and ac tivation state of the cell. Scavenger receptors A and B have been implicated in AD with increased expression in glial cells surrounding amyloid deposits (Coraci et al., 2002). We were interested in SRA
130 specifically because it has been shown to bind both LPS and A though other scavenger receptors can also bind A (Husemann et al., 2002). Base d on our results, we propose that LPS induced expression of scavenger receptors that could then bind A and mediate removal. A diagram constructing the results from chapters 2 and 3 is found in Figure 25.
131 Figure 25. Mechanisms of LPS stimulated inflammation and A removal. This cartoon is a composite of various stained sections. A cresyl violet stain of the neurons in the dentate gyrus (upper left corner ) and portions of CA3 (lower left corner) are the backdrop of the figure (the cells are stained blue). A deposits are stained brown, with compacted, Congophilic deposits stained red. A reactive as trocyte (stained for GFAP) is shown in the upper right corner. Microg lia are stained for CD45 and are black, shown transitioning among various activation states. LPS injecti on increases levels of the cytokines IL1 and TNF as well as cell surface receptors CD45, CR3, SRA, and Fc RII/III. The graph in the upper left corner illustra tes the changes over time in CD45/CR3 (purple), SRA and Fc RII/III (green) and A (orange). Astrocyte MICROGLIA resting ameboid hyper ramified SR FcR CD45 CR3 IL1 TNF 3d 7d 14d 28d LPS
132 The final chapter of this di ssertation sought to further confirm that microglia are involved in the LPS-stimulated removal of A We used dexamethasone co-treatment with LPS intrahippocampal injection, thereby blocking microglial activation. Dexamethasone is a potent glucocorticoi d anti-inflammatory agent with similar physiological activities as cort isone (Hardman et al., 2001). Dexamethasone is bound in the circulation by corticosteroid binding gl obulin. Dexamethasone can then transverse the cell membrane as it is hydrophobic. Once in side the cell, dexamethasone binds to the glucocorticoid receptor in the cytosol, releas ing inhibitory heat shock proteins from the glucocorticoid receptor allo wing it to translocate to the nucleus. Glucocorticoid receptor:dexamethasone complexes can then antagonize NF B response elements (as well as others), thereby inhibi ting the inflammatory response. In this final study, LPS induced the protot ypical inflammatory response discussed previously. CD45, CR3, and IL1 levels were increased, as well as SRA and Fc receptors. Concurrent with this response, we demonstrated the expected decrease in the diffuse A load, with no effect on compact de posits. Dexamethasone co-treatment significantly inhibited CD45, CR3 and IL1 levels. The reductions in diffuse A caused by LPS were also inhibited by dexamethas one co-treatment, thus implicating the microglial response in the removal process. However, neither Fc R nor SRA were inhibited by dexamethasone co-treatment. These findings were confirmed by both immunostaining and mRNA analysis where possible. Dexamethasones ability to inhibit reac tive microgliosis is well documented. Our lab has previously demonstrated that CD45 le vels (induced by intr acranial injection of A antibodies into APP mice) were significantly inhibited after dexamethasone treatment
133 (Wilcock et al., 2004a). Others have s hown that dexamethasone inhibits IL1 after either LPS or A intraventricular injection into nontransgenic mice (Szczepanik & Ringheim, 2003). Thus, our findings that dexame thasone inhibited CD45, CR3, and IL1 and also subsequently prevented the removal of A were expected. These data provide evidence that general microglial activation, as demons trated by CD45 and CR3 levels, is required for the removal of A in our protocol. The lack of inhibition of Fc and scavenger receptors in the LPS-dexamethasone co-treated mice was an unexpected result. There is evidence to support these findings. Of all the markers examined, only Fc R class II have an identified a glucocorticoid response elements (GRE) in the promoter (Hoga rth et al., 1991). In this case, the Fc receptor is inhibitory and the GRE is stimulat ory, thus we would expect an increase in the levels of the class II receptor under de xamethasonetreated conditions. Sivo and colleagues reported similar results where interferon gammastimulated Fc RII/III expression, and was not inhibi ted by dexamethasone treatment (1993). A review of the literature did not reveal any data concerning dexamethasones effects on SRA expression. Taken together, the continued expression of Fc R and SRA after dexamethasone treatment indicates that expression of mark ers associated with phagocytosis are not sufficient to effect removal of A in our protocol. Figure 26 illustrates our findings to date.
134 Figure 26. Mechanisms of LPS stimulated inflammation and A removal, and inhibition by dexamethasone. This cartoon is a composite of various stained sections (details provided in Figure 25). LPS injection in creases levels of the cytokines IL1 and TNF as well as cell surf ace receptors CD45, CR3, SRA, and Fc RII/III. Dexamethasone cotreatment inhibited IL1 CD45, and CR3, and also prevented A reductions, but had no effect on SRA or Fc R levels. The graph in the upper left corner illustrates the changes over time in CD45/CR3 (purple), SRA and Fc RII/III (green), and A (orange). Astrocyte MICROGLIA resting ameboid hyper ramified SR FcR CD45 CR3 IL1 TNF 3-7d D D e e x x a a m m e e t t h h a a s s o o n n e e LPS
135 Our current hypothesis is that micr oglia are removing the diffuse A in our in vivo model. This idea is supported by the finding that dexamethasone inhibited microglial activation and A removal. There are othe r possible explanations for dexamethasones action in our model. The three experimental groups evaluated in our final study were a) saline inject ed controls, b) LPS injected, and c) LPS injected plus dexamethasone treated APP mice. We did not evaluate dexamethasone-only treated mice; we were looking for the drugs effect on LPS-induced inflammation, not dexamethasones effect on amyloid burden. There are no published studies looking at glucocorticoid administration to amyloid depositing mice in the absence of other stimulating agents. Human trials with predni sone did not ameliorate symptoms (Aisen et al., 2000). In vitro neither prednisone nor dexamethasone lowered A (Sagi et al., 2003). NSAIDS, which can lower A in vitro and in long term studies with mouse models, do not have reproducible effects after sh ort term treatment (3-5 days; reviewed in Gasparini et al., 2004). Therefor e, it is unlikely that in th e short time frames studied herein (3 and 7 days) dexamethasone would ex ert an appreciable effect on brain amyloid burden. However, it cannot be ruled out that the effects we saw here were due to the potential for dexamethasone to increase A burden thus negating the effects of LPS. Alternatively, dexamethasone can decrease th e permeability of the blood brain barrier, which may have prevented the flux of A from the brain to the periphery (Romero et al., 2003). Future studies could evaluate dexame thasones effect on amyloid deposition in the absence of inflammation, as well as the effects of othe r anti-inflammatory drugs on microglial activation and A reductions.
136 Our primary hypothesis regarding the removal of A centered on the scavenger receptors. Based on our results, we proposed that LPS induced expression of scavenger receptors that could then bind A and mediate removal. Ho wever, we found that SRB was not highly upregulated, and that SRA expression pers isted in the absence of A removal. Therefore, a working hypothesis regarding LPS-stimulated reductions in A can be proposed: It is not the mere presence of the LPS molecule, nor is the upregulation of receptors involved in pha gocytosis sufficient for A removal. Rather, general cell activation is required. Thus, a pha gocytic cell must not only bind A (by various receptors) but must also be capable of engulfing the material (via general cell activation). There are several cell surf ace receptors grouped under the general classification of scavenger receptors which can bind A including SRB, CD36, SRA, CD68, receptor of advanced glycation end products, and lipoprotei n receptor related protein (Peiser et al., 2002; Huseman et al., 2002). S RB clearly interacts with A and can mediate microglial activation (Coraci et al., 2002; Moore et al., 2002; El Khoury et al., 2003). There is some confusion as to the identity of SRB and CD36 (a class B type receptor). They are the focus of separate lines of research, but the sequences listed in Ge nbank are identical, and the nomenclature overlaps in the literature. For the work described herein, we used the sequence for murine CD36, and have referred to it as SRB throughout. Wyss-Coray and colleagues have shown that blocking SRB1 had no effect on astrocyte-mediated A clearance (2003). Whether there is truly a difference between CD36 and SRB, and the potential roles for either receptor in LPS stimulated A removal remains to be elucidated. Future studies could block scavenger receptor s with fucoidan, or sp ecific antibodies, to
137 further evaluate their role in our model. Scavenger receptor knockout mice for both CD36 and SRA are also available and could be bred with APP mice to test subsequent effects on LPS mediated A removal. CD68 (macrosialin) is another scavenge r receptor implicated in the pathogenesis of AD. The endogenous ligands for CD68 are modified lipoproteins and it is commonly referred to as a late endosomal or lysoso mal protein, though cell surface expression also occurs (Ramprasad et al., 1996; Holness et al., 1993). It has been shown to be upregulated in AD brain and in APP mice (Bor nemann et al., 2001; Sasaki et al., 2002). We have briefly examined CD68 immunosta ining in our model and found both plaqueassociated microglia as well as LPSinduced widespread expression of CD68 (unpublished data). This protein may be marking a phagocytic population and deserves further investigation. The receptor for advanced glycation end products (RAGE) has been intensely investigated in AD (Zlokovic et al, 2004) Endogenous ligands for RAGE include nonenzymatically glycated adducts, beta -pleated sheet structures, and A Rather than triggering A removal via endocytosis, stimulation of RAGE leads to prolonged cell activation (Schmidt et al., 2001). RAGE is upregulated in AD brai n, particularly around plaques, and has been proposed to mediate the flux of A from the periphery into the brain, thereby altering the balance between de position and removal. Lipoprotein receptor related protein (LRP) is thought to act as c ounterpoint to RAGE by mediating flux of A out of the brain into the pe riphery with APOE mediating binding to the receptor. The main ligands for LRP are modified lipoproteins but this receptor can also bind bacterial and complement components In vitro models have shown that LPS decreased
138 macrophage LRP expression, whereas dexamethas one increased LRP levels (Laithwait et al., 1999; Marzolo et al ., 2000). The effects in vivo remain to be elucidated thus LRP is an attractive target for investigation in our model of LPS mediated removal of A Future studies to fu rther elucidate the A removal described herein would evaluate each of these proteins under both LPS and dexamethasone treatment conditions. An alternative explanation is that scave nger receptors or pha gocytosis in general may not be involved at all. Other mechanisms of A removal include non scavenger receptor-mediated phagocytosis, decreased APP processing, increased proteolytic degradation of A or contributions from the periphery in clearance. Integrins mediate cell adhesion to both the extracellular matrix and to other cells, triggering intracellu lar signaling. The 6 1 integrin is expressed by macrophages and binds A in vitro as part of a trimolecular co mplex and mediates phagocytosis (Bamberger et al., 2003; Koenigsknecht & Landreth, 2004). The trimolecular complex includes 61, CD36, and CD47 (an integrin associated protein). Though we investigated CD36 (SRB) in our model, we have not yet eval uated the roles of 61 or CD47 and this remains a potenti al route of LPS mediated A removal. CR3 is another integrin ( 2) and is the receptor for complement component C3. CR3 is a pattern recognition receptor which can bind LPS (E hlers et al., 2000), though no data on its ability to bind A was found after a literature search. Complement protein C3 is a major component of the alternativ e pathway. Wyss-Coray and co lleagues have shown that inhibiting C3 led to increased A deposition in transgenic mi ce (2002). In our studies, CR3 was upregulated by LPS, and inhibite d by dexamethasone, mirroring the effects on
139 A removal. Therefore CR3 coul d be directly involved in A reductions in our model. Future studies could inhibit th e integrins using invasin and ev aluate subsequent effects on A levels. Several metalloproteinases have been shown to degrade A including neprilysin (NEP), insulin degrading enzyme (IDE), and matrix metalloproteinase 9 (MMP-9). NEP is capable of degrading many circulating pe ptides including enkepha lin, substance P, and A NEP polymorphisms have been linked to increased risk of AD (Sakai et al., 2004), and transgenic mouse models have confirmed that overe xpression of NEP can reduce brain A levels (Mohajeri et al, 2004; Iwata et al., 2004). We conducted a preliminary investigation to determine NEP levels after L PS injection in our mode l. We did not find any significant changes in th e immunostaining patterns (unp ublished findings). Kaneko and colleagues have evidence that LPS decr eases NEP levels in neutrophils, though effects on other macrophage populations ar e unknown (2003). Another enzyme, IDE, degrades both insulin and A Based on the relationship between diabetes mellitus and AD, an increased risk for AD was found with certain variants of the IDE gene, and IDE activity, levels, and mRNA are decreased in AD brain (reviewed in Zhao et al., 2004). A search of the literature did not reveal data reporting the e ffects of LPS on IDE levels, and we have not looked in our studies. A separate class of metalloprotei nases is associated with the extracellular matri x, and has been implicated in plaque formation in cardiovascular and Alzheimers diseases. MMP -1, -3, and -9 are capable of degrading A and the levels of these enzymes are altered in AD brain (Leake et al., 2000; Helbecque et al., 2003; Yoshiyama et al., 2000 ). Woo and colleagues recently showed that LPS can induce expression of MMP-9 in macrophages (2004), and thus may be
140 partially responsible fo r the reductions in A in our model. Future studies should include evaluation of these enzymes in our model. One simple explanation for the decreased A load in our model is decreased levels of APP or its processing. In an atte mpt to answer this question, we immunostained for APP and also measured mRNA levels afte r LPS administration (data not shown). The immunostaining pattern was found to be nonspe cific, and no signifi cant changes in APP mRNA levels were found in response to LPS. Thus we do not believe there were large changes in APP levels which could acc ount for the up to 70% decreases in A seen in our studies. In general, most rese arch regarding the brains res ponse to insult indicates that APP levels increase, and presenilin is also induced (Pennypacker et al., 1999; Brugg et al., 1995; Siman et al., 1989; Wright et al., 1999 ; Sheng et al., 2003). In other studies, LPS was shown to increase neuronal levels of A species, though the researchers did not specifically examine the secretases (Sheng et al., 2003; Sly et al., 2001). The activity or levels of alpha, beta and/or gamma secretas es may be altered after LPS injection, which could account for the decreased levels of A in our model. We did not measure the secretases, thus such studies represen t future lines of investigation. Another unexplored option in the LPSmediated A removal is the transcription factor peroxisome proliferat or-activated receptor (PPAR ). PPAR is traditionally associated with lipid and insulin metabolism and is targeted in therapies for Type II diabetes, but also has potent anti-infla mmatory actions. Camacho and colleagues recently reported in an in vitro assay that the PPAR agonists rosiglitazone and pioglitazone reduced A secretion from APP transfected HEK cells (2004). However, our
141 own lab failed to demonstrate reductions in amyloid burden after pioglitazone administration to transgenic mice (Paul Jant zen et al., unpublished re sults.) In another study using rats, PPAR was downregulated by periphera l LPS treatment (Ohata et al., 2004). However, the effects of intracranial administration of LPS are unknown and thus PPAR represents another potential mechanism mediating A reductions. Contributions of the peripheral immune system to the reductions in A in our model cannot be ruled out. There is evid ence of peripheral bloodborne cells in the brains of our LPS injected mice as seen in the small round cells s een between 24 and 72 hours after LPS injection. As much as 80% of these cells may be recruited monocytes (Montero-Menei et al., 1996). In order to ru le out peripheral cells as mediators of A removal, we could irradiate the mice or ch emically ablate their bone marrow cells, and then inject LPS and monito r subsequent effects on A burden. Another potential mechanism of A reductions can be found in the pe ripheral sink hypothesis (Matsuoka et al., 2003). According to this line of thinking, A levels can be reduced by creating a concentration gradient down which A would move out of the brain and into the periphery. LPS has been shown to disrupt the blood brain barrier and may therefore alter A trafficking (Nonaka et al., 2005). We did not ev aluate the levels of A in the periphery and therefore do not know the contri bution of any potential sink in our model. Future studies could collect blood from the mice after treatment and evaluate the levels of A The majority of the markers we used to evaluate gliosis in our model are macrophage specific. Astrocytes may also contribute to the removal of A seen herein.
142 Astrocytes normally perform homeostatic func tions but also react to amyloid deposits in the brain. Astrocytes are capable of re leasing inflammatory mediators including complement proteins, cytokines, prostaglan dins, and reactive oxygen species (Akiyama et al., 2000). These cells are spa tially associated with amyloid and may be forming a glial scar around the deposit, though astrocyte phagoc ytic functions are also possible (DeWitt et al., 1998; Kurt et al., 1999; Mandybur et al., 1990; Wegiel et al., 2000). One of the most compelling studies linki ng astrocytes to amyloid re moval was demonstrated by Wyss-Coray and colleagues in 2003. They coated brain slices from APP transgenic mice with adult mouse astrocytes. A significant 40 % reduction in the total amyloid burden was seen after 24 hour incubation. Fucoidan and polyinosinic acid significantly blocked A removal, implicating scavenger receptors in the process. In our studies, the only true astrocyte marker evaluated was GFAP. LPS i nduced a significant increase in GFAP in all of our studies, though the react ion seemed to be transient. In vivo it is difficult to separate individual populations of cells and th eir activities, thus some dual labeling may be necessary once the receptors/mechanism of A removal is discovered, in order to confirm whether it is microglia and/or astrocytes involved in this process. Assuming that the amyloid hypothe sis is correct, removing A should ameliorate the symptoms and/or progression of AD. Stra tegies aimed at stimulating the immune system to remove A are at the forefront of AD research. There is evidence from the human vaccine trial that treat ment successfully decreased amyloid load (Schenk et al., 2004). An immune/inflammatory reaction wa s seen during the tr ial, which led to meningoencephalitis in rare cases, ev en though many other patients experienced cognitive stability and possible improvement due to the vaccine. In transgenic mice, both
143 passive and active immunization against A led to microglial activa tion in the process of lowering A burden and improving cognitive performance (Wilcock et al., 2001, 2003, 2004a,b). Our own results with stimulation of the innate immune system with LPS yielded similar results: activation of microglia resulting in reductions in A In none of the laboratory trials was neurodegenerati on triggered by the inflammatory response. Taken together, one might conclude that some level of inflammation in AD is actually beneficial and is responsible for maintain ing a balance between amyloid deposition and removal. Future studies would aim to characterize the activation state of microglia by either stimulating them with various agents or inhi biting specific aspects of the inflammatory response. Though LPS primarily signals thr ough TLR4, an interes ting approach would be to activate the cells via other tolllike receptors, a nd then examine differential microglial activation while monitoring effect s on amyloid burden. In our studies, the inflammatory reaction seen after innate immunity stimulation was blocked by dexamethasone, and also prevented A removal. Targeted anti-inflammatory or antioxidant treatments using compounds such as nitro-flurbiprofen, cel ecoxib, pioglitazone, or melatonin could be useful in identify ing the degree of cell activation required for phagocytosis. It is increasingly evident that the inflammatory response, particularly modulated by microglia, is very complex. Th erapeutic opportunities await analysis of the different stages of microglial activat ion, their function, morphology, and protein expression during both the chronic response to amyloid as well as during phagocytic removal of the deposits.
144 References Aisen, P. S., Davis, K. L., Berg, J. D., Scha fer, K., Campbell, K., Thomas, R. G., Weiner, M. F., Farlow, M. R., Sano, M., Grundman, M., & Thal, L. J. (2000). A randomized controlled trial of prednisone in Alzheimer's disease. Alzheimer's Disease Cooperative Study. Neurology, 54, 588-593. Aisen, P. S., Schafer, K. A., Grundman, M., Pf eiffer, E., Sano, M., Davis, K. L., Farlow, M. R., Jin, S., Thomas, R. G., & Thal, L. J. (2003). Effects of rofecoxib or naproxen vs placebo on Alzheimer disease progression: a randomized controlled trial. JAMA 289, 2819-2826. Akiyama, H., Barger, S., Barnum, S., Bradt, B., Bauer, J., Cole, G. M., Cooper, N. R., Eikelenboom, P., Emmerling, M., Fiebich, B. L., Finch, C. E., Frautschy, S., Griffin, W. S., Hampel, H., Hull, M., Landr eth, G., Lue, L., Mrak, R., Mackenzie, I. R., McGeer, P. L., O'Banion, M. K., P achter, J., Pasinetti, G., Plata-Salaman, C., Rogers, J., Rydel, R., Shen, Y., Stre it, W., Strohmeyer, R., Tooyoma, I., Van Muiswinkel, F. L., Veerhuis, R., Walker D., Webster, S., Wegrzyniak, B., Wenk, G., & Wyss-Coray, T. (2000). Inflam mation and Alzheimer's disease. Neurobiol.Aging, 21, 383-421. Alzheimer, A., Stelzman, R. A., Schnitzlein, H. N., & Murtagh, F. R. (1907). Uber eine eigenartige erkankung der hirnrinde. Allgemeine Zeitschrift fr Psychiatrie und psychisch-gerichtlich e Medizin, Berlin, 64 146-148. Amigorena, S., Bonnerot, C., Choquet, D., Fr idman, W. H., & Teillaud, J. L. (1989). Fc gamma RII expression in resti ng and activated B lymphocytes. Eur.J.Immunol., 19 1379-1385. Auld, D. S., Kornecook, T. J., Bastianetto, S ., & Quirion, R. (2002). Alzheimer's disease and the basal forebrain cholinergic system : relations to beta-amyloid peptides, cognition, and treatment strategies. Prog.Neurobiol., 68, 209-245. Bamberger, M. E., Harris, M. E., McDona ld, D. R., Husemann, J., & Landreth, G. E. (2003). A cell surface recepto r complex for fibrillar beta-amyloid mediates microglial activation. J.Neurosci., 23 2665-2674.
145 Banati, R. B., Gehrmann, J., Wiessner, C ., Hossmann, K. A., & Kreutzberg, G. W. (1995). Glial expression of the beta-amyloid precursor protein (APP) in global ischemia. J.Cereb.Blood Flow Metab, 15, 647-654. Bornemann, K. D., Wiederhold, K. H., Pauli, C., Ermini, F., Stalder, M., Schnell, L., Sommer, B., Jucker, M., & Staufenbiel, M. (2001). Abeta-induced inflammatory processes in microglia cells of APP23 transgenic mice. Am.J.Pathol., 158 63-73. Bourin, M., Ripoll, N., & Dail ly, E. (2003). Nicotinic receptors and Alzheimer's disease. Curr.Med.Res.Opin., 19, 169-177. Brugg, B., Dubreuil, Y. L., Huber, G., Wollman, E. E., Delhaye-Bouchaud, N., & Mariani, J. (1995). Inflammatory processe s induce beta-amyloid precursor protein changes in mouse brain. Proc.Natl.Acad.Sci.U.S.A, 92 3032-3035. Camacho, I. E., Serneels, L., Spittaels, K., Merchiers, P., Dominguez, D., & De Strooper, B. (2004). Peroxisome proliferator-activat ed receptor gamma induces a clearance mechanism for the amyloid-beta peptide. J.Neurosci., 24 10908-10917. Canevari, L., Abramov, A. Y., & Duchen, M. R. (2004). Toxicity of amyloid beta peptide: tales of calcium, m itochondria, and oxidative stress. Neurochem.Res., 29 637-650. Castano, A., Herrera, A. J., Cano, J., & Macha do, A. (2002). The degenerative effect of a single intranigral in jection of LPS on the dopaminerg ic system is prevented by dexamethasone, and not mimicked by rh-TNF-alpha, IL-1beta and IFN-gamma. J.Neurochem., 81 150-157. Chao, C. C., Hu, S., Molitor, T. W., Shaskan, E. G., & Peterson, P. K. (1992). Activated microglia mediate neuronal cell inju ry via a nitric oxide mechanism. J.Immunol., 149, 2736-2741. Colton, C. A., Yao, J., Taffs, R. E., Keri, J. E., & Oster-Granite, M. L. (1991). Abnormal production of interleukin-1 by mi croglia from trisomy 16 mice. Neurosci.Lett., 132 270-274. Colton, C. A., Yao, J., Keri, J. E., & Gilb ert, D. (1992a). Regul ation of microglial function by interferons. J.Neuroimmunol., 40 89-98. Colton, C. A., Abel, C., Patchett, J., Keri J., & Yao, J. (1992b). Lectin staining of cultured CNS microglia. J.Histochem.Cytochem., 40, 505-512. Colton, C. A. & Gilbert, D. L. (1993). Micr oglia, an in vivo sour ce of reactive oxygen species in the brain. Adv.Neurol., 59, 321-326.
146 Colton, C. A., Snell, J., Chernyshev, O., & Gilb ert, D. L. (1994). Induction of superoxide anion and nitric oxide produc tion in cultured microglia. Ann.N.Y.Acad.Sci., 738, 54-63. Colton, C. A. & Chernyshev, O. N. (1996). I nhibition of microglial superoxide anion production by isoproterenol and dexamethasone. Neurochem.Int., 29 43-53. Colton, C. A., Chernyshev, O. N., Gilbert, D. L., & Vitek, M. P. (2000). Microglial contribution to oxidative stress in Alzheimer's disease. Ann.N.Y.Acad.Sci., 899 292-307. Colton, C. A., Brown, C. M., Cook, D., Needha m, L. K., Xu, Q., Czapiga, M., Saunders, A. M., Schmechel, D. E., Rasheed, K., & Vitek, M. P. (2002a). APOE and the regulation of microglial nitric oxide pr oduction: a link between genetic risk and oxidative stress. Neurobiol.Aging, 23, 777-785. Colton, C. A., Brown, C. M., Czapiga, M ., & Vitek, M. P. (2002b). Apolipoprotein-E allele-specific regulation of nitric oxide production. Ann.N.Y.Acad.Sci., 962 212225. Colton, C. A., Needham, L. K., Brown, C ., Cook, D., Rasheed, K., Burke, J. R., Strittmatter, W. J., Schmechel, D. E. & Vitek, M. P. (2004). APOE genotypespecific differences in human and m ouse macrophage nitric oxide production. J.Neuroimmunol., 147 62-67. Combs, C. K., Johnson, D. E., Karlo, J. C ., Cannady, S. B., & Landreth, G. E. (2000). Inflammatory mechanisms in Alzheimer's disease: inhibition of beta-amyloidstimulated proinflammatory respons es and neurotoxicity by PPARgamma agonists. J.Neurosci., 20 558-567. Coraci, I. S., Husemann, J., Berman, J. W., Hule tte, C., Dufour, J. H., Campanella, G. K., Luster, A. D., Silverstein, S. C., & El Khoury, J. B. (2002). CD36, a class B scavenger receptor, is expr essed on microglia in Alzhei mer's disease brains and can mediate production of reactive oxygen species in response to beta-amyloid fibrils. Am.J.Pathol., 160 101-112. Corder, E. H., Saunders, A. M., Strittmatter, W. J., Schmechel, D. E., Gaskell, P. C., Small, G. W., Roses, A. D., Haines, J. L., & Pericak-Vance, M. A. (1993). Gene dose of apolipoprotein E type 4 allele and the risk of Alzheimer's disease in late onset families. Science, 261 921-923. Czapiga, M. & Colton, C. A. (1999). Function of microglia in organotypic slice cultures. J.Neurosci.Res., 56 644-651. D'Andrea, M. R., Cole, G. M., & Ard, M. D. (2004). The microglial phagocytic role with specific plaque types in the Alzheimer disease brain. Neurobiol.Aging, 25, 675683.
147 de la Torre, J. C. (2002). Vascular basis of Alzheimer's pathogenesis. Ann.N.Y.Acad.Sci., 977, 196-215. De Strooper, B. (2003). Aph-1, Pen-2, and Nicas trin with Presenilin generate an active gamma-Secretase complex. Neuron, 38, 9-12. DeWitt, D. A., Perry, G., Cohen, M., Doller, C., & Silver, J. (1998). Astrocytes regulate microglial phagocytosis of senile pl aque cores of Alzheimer's disease. Exp.Neurol., 149, 329-340. DiCarlo, G., Wilcock, D., Henderson, D., Gordon, M., & Morgan, D. (2001). Intrahippocampal LPS injections reduce Ab eta load in APP+PS1 transgenic mice. Neurobiol.Aging, 22, 1007-1012. Dickey, C. A., Loring, J. F., Montgomery, J ., Gordon, M. N., Eastman, P. S., & Morgan, D. (2003). Selectively reduced expression of synaptic plasticit y-related genes in amyloid precursor protein + presenilin-1 transgenic mice. J.Neurosci., 23, 52195226. Dickey, C. A., Gordon, M. N., Mason, J. E., Wilson, N. J., Diamond, D. M., Guzowski, J. F., & Morgan, D. (2004). Amyloid suppre sses induction of ge nes critical for memory consolidation in APP + PS1 transgenic mice J.Neurochem., 88 434442. Dineley, K. T., Westerman, M., Bui, D., Bell, K., Ashe, K. H., & Sweatt, J. D. (2001). Beta-amyloid activates the mitogen-activated protein kinase cascade via hippocampal alpha7 nicotinic acetylcholine receptors: In vitro and in vivo mechanisms related to Alzheimer's disease. J.Neurosci., 21 4125-4133. Dineley, K. T., Bell, K. A., Bui, D., & Sw eatt, J. D. (2002). beta -Amyloid peptide activates alpha 7 nicotinic acetylcholine receptors expressed in Xenopus oocytes. J.Biol.Chem., 277, 25056-25061. Dineley, K. T., Xia, X., Bui, D., Sweatt, J. D., & Zheng, H. (2002). Accelerated plaque accumulation, associative learning deficits, and up-regulation of alpha 7 nicotinic receptor protein in transgenic mice co-expressing mutant human presenilin 1 and amyloid precursor proteins. J.Biol.Chem., 277, 22768-22780. Dineley, K. T., Zheng, H., & Sweatt, J. D. (2004). TG2576 Cognitive deficit worsens in the absence of alpha 7 nico tinic acetylcholine receptors. Podium Presentation at the Society for Neuroscience Annual Meeting. Dominguez del Toro, E., Juiz, J. M., Pe ng, X., Lindstrom, J., & Criado, M. (1994). Immunocytochemical localization of th e alpha 7 subunit of the nicotinic acetylcholine receptor in the rat central nervous system. J.Comp Neurol., 349, 325-342.
148 Duff, K., Eckman, C., Zehr, C., Yu, X., Prada, C. M., Perez-tur, J., Hutton, M., Buee, L., Harigaya, Y., Yager, D., Morgan, D., Gordon, M. N., Holcomb, L., Refolo, L., Zenk, B., Hardy, J., & Younkin, S. (1996) Increased amyloid-beta42(43) in brains of mice expressing mutant presenilin 1. Nature, 383, 710-713. Ehlers, M. R. (2000). CR3: a general purpose adhesion-recognition receptor essential for innate immunity. Microbes.Infect., 2 289-294. El Khoury, J. B., Moore, K. J., Means, T. K., Leung, J., Terada, K., Toft, M., Freeman, M. W., & Luster, A. D. (2003). CD36 mediat es the innate host response to betaamyloid. J.Exp.Med., 197 1657-1666. Eriksen, J. L., Sagi, S. A., Smith, T. E., We ggen, S., Das, P., McLendon, D. C., Ozols, V. V., Jessing, K. W., Zavitz, K. H., Koo, E. H., & Golde, T. E. (2003). NSAIDs and enantiomers of flurbiprofen target ga mma-secretase and lower Abeta 42 in vivo. J.Clin.Invest, 112 440-449. Fabian-Fine, R., Skehel, P., E rrington, M. L., Davies, H. A., Sher, E., Stewart, M. G., & Fine, A. (2001). Ultrastructural distribu tion of the alpha7 nicotinic acetylcholine receptor subunit in rat hippocampus. J.Neurosci., 21 7993-8003. Febbraio, M., Hajjar, D. P., & Silverstein, R. L. (2001). CD36: a class B scavenger receptor involved in angiogenesis, atherosclerosis, inflammation, and lipid metabolism. J.Clin.Invest, 108, 785-791. Fonseca, M. I., Zhou, J., Botto, M., & Tenner, A. J. (2004). Absence of C1q leads to less neuropathology in transgenic mouse models of Alzheimer's disease. J.Neurosci., 24 6457-6465. Franceschini, D., Paylor, R., Broide, R., Salas, R., Bassetto, L., Gotti, C., & De Biasi, M. (2002). Absence of alpha7-containing neur onal nicotinic acetylcholine receptors does not prevent nicotine-induced seizures. Brain Res.Mol.Brain Res., 98, 29-40. Friedhoff, P., von Bergen, M., Mandelkow, E. M., & Mandelkow, E. (2000). Structure of tau protein and assembly into paired helical filaments. Biochim.Biophys.Acta, 1502 122-132. Frisoni, G. B. (2001). Treatment of Alzhei mer's disease with acetylcholinesterase inhibitors: bridging the gap be tween evidence and practice. J.Neurol., 248 551557. Gasparini, L., Ongini, E., & Wenk, G. (2004) Non-steroidal anti-inflammatory drugs (NSAIDs) in Alzheimer's disease: old and new mechanisms of action. J.Neurochem., 91, 521-536.
149 Gault, J., Robinson, M., Berger, R., Drebi ng, C., Logel, J., Hopkins, J., Moore, T., Jacobs, S., Meriwether, J., Choi, M. J., Kim, E. J., Walton, K., Buiting, K., Davis, A., Breese, C., Freedman, R., & Leonard, S. (1998). Genomic organization and partial duplication of the human alpha7 ne uronal nicotinic acetylcholine receptor gene (CHRNA7). Genomics, 52 173-185. Gessner, J. E., Heiken, H., Tamm, A., & Schmidt, R. E. (1998). The IgG Fc receptor family. Ann.Hematol., 76 231-248. Goate, A., Chartier-Harlin, M. C., Mullan, M., Brown, J., Crawford, F., Fidani, L., Giuffra, L., Haynes, A., Irving, N., & Jame s, L. (1991). Segregation of a missense mutation in the amyloid precursor protein gene with familial Alzheimer's disease. Nature, 349 704-706. Gordon, M. N., Schreier, W. A., Ou, X., Holcomb, L. A., & Morgan, D. G. (1997). Exaggerated astrocyte reactivity after nigrostriatal deaffere ntation in the aged rat. J.Comp Neurol., 388 106-119. Gordon, M. N., King, D. L., Diamond, D. M., Ja ntzen, P. T., Boyett, K. V., Hope, C. E., Hatcher, J. M., DiCarlo, G., Gottschall, W. P., Morgan, D., & Arendash, G. W. (2001). Correlation between cognitive defici ts and Abeta deposits in transgenic APP+PS1 mice. Neurobiol.Aging, 22 377-385. Gordon, M. N., Holcomb, L. A., Jantzen, P. T ., DiCarlo, G., Wilcock, D., Boyett, K. W., Connor, K., Melachrino, J., O'Callaghan, J. P., & Morgan, D. (2002). Time course of the development of Alzheimer-lik e pathology in the doubly transgenic PS1+APP mouse. Exp.Neurol., 173, 183-195. Graham, A. J., Martin-Ruiz, C. M., Teakt ong, T., Ray, M. A., & Court JA (2002). Human brain nicotinic receptors, their distributi on and participation in neuropsychiatric disorders. Curr.Drug Targets.CNS.Neurol.Disord., 1, 387-397. Grassi, F., Palma, E., Tonini, R., Amici, M ., Ballivet, M., & Eusebi, F. (2003). Amyloid beta(1-42) peptide alters the gating of human and mouse alpha-bungarotoxinsensitive nicotinic receptors. J.Physiol, 547 147-157. Grewal, T., Priceputu, E., Davignon, J., & Bern ier, L. (2001). Identification of a gammainterferon-responsive element in the promoter of the human macrophage scavenger receptor A gene. Arterioscler.Thromb.Vasc.Biol., 21, 825-831. Griffin, W. S., Sheng, J. G., Gentleman, S. M., Graham, D. I., Mrak, R. E., & Roberts, G. W. (1994). Microglial interleukin-1 alpha expression in human head injury: correlations with neuronal and neuritic beta-amyloid precursor protein expression. Neurosci.Lett., 176 133-136. Hardman, J. G., Limbird, L. E., eds. Goodman & Gilman's the phar macological basis of therapeutics 10th Ed. (2001) New York: McGraw-Hill.
150 Hardy, J. & Allsop, D. (1991). Amyloid depositi on as the central event in the aetiology of Alzheimer's disease. Trends Pharmacol.Sci., 12, 383-388. Hauss-Wegrzyniak, B., Dobrzanski, P., Stoehr J. D., & Wenk, G. L. (1998a). Chronic neuroinflammation in rats reproduces components of the neurobiology of Alzheimer's disease. Brain Res., 780 294-303. Hauss-Wegrzyniak, B., Lukovic, L., Biga ud, M., & Stoeckel, M. E. (1998b). Brain inflammatory response induced by in tracerebroventricu lar infusion of lipopolysaccharide: an imm unohistochemical study. Brain Res., 794 211-224. Hauss-Wegrzyniak, B., Vraniak, P., & Wenk, G. L. (1999a). The effects of a novel NSAID on chronic neuroinfla mmation are age dependent. Neurobiol.Aging, 20, 305-313. Hauss-Wegrzyniak, B., Willard, L. B., Del Soldato, P., Pepeu, G., & Wenk, G. L. (1999b). Peripheral administration of novel anti-inflammatories can attenuate the effects of chronic inflam mation within the CNS. Brain Res., 815, 36-43. Hauss-Wegrzyniak, B., Vannucchi, M. G ., & Wenk, G. L. (2000). Behavioral and ultrastructural changes induced by ch ronic neuroinflammation in young rats. Brain Res., 859 157-166. Hauss-Wegrzyniak, B., Lynch, M. A., Vran iak, P. D., & Wenk, G. L. (2002). Chronic brain inflammation results in cell loss in the entorhinal cortex and impaired LTP in perforant path-granule cell synapses Exp.Neurol., 176, 336-341. Helbecque, N., Hermant, X., Cottel, D., & Am ouyel, P. (2003). The role of matrix metalloproteinase-9 in dementia. Neurosci.Lett., 350 181-183. Hellstrom-Lindahl, E., Court, J., Keverne, J., Svedberg, M., Lee, M., Marutle, A., Thomas, A., Perry, E., Bednar, I., & No rdberg, A. (2004). Nicotine reduces A beta in the brain and cereb ral vessels of APPsw mice. Eur.J.Neurosci., 19 27032710. Henson, P. M. (1971). The immunologic re lease of constituents from neutrophil leukocytes. II. Mechanisms of release during phagocyt osis, and adherence to nonphagocytosable surfaces. J.Immunol., 107 1547-1557. Herber, D. L., Maloney, J. L., Roth, L. M., Freeman, M. J., Morgan, D., & Gordon, M. N. (2004a). Acute and chronic micropglia l inflammatory responses after intrahippocampal administra tion of lipopolysaccharide. Unpublished work Herber, D. L., Roth, L. M., Wilson, D., W ilson, N., Mason, J. E., Morgan, D., & Gordon, M. N. (2004b). Time-dependent reduction in Abeta levels afte r intracranial LPS administration in APP transgenic mice. Exp.Neurol., 190 245-253.
151 Herber, D. L., Severance, E. G., Cuevas, J., Morgan, D., & Gordon, M. N. (2004c). Biochemical and histochemical evidence of nonspecific binding of alpha7nAChR antibodies to mouse brain tissue. J.Histochem.Cytochem., 52 1367-1376. Herber, D. L., Symmonds, K. A., Roth, L. M ., Maloney, J. L., Barcenas, M., Freeman, M. J., Morgan, D., & Gordon, M. N. (2004d). Dexamethasone suppresses LPSinduced microglial activation and amyloid clearance in APP transgenic mice. Unpublished Work Herz, J. & Strickland, D. K. (2001). LRP: a multifunctional scavenger and signaling receptor. J.Clin.Invest, 108, 779-784. Heumann, D. & Roger, T. (2002). Initial re sponses to endotoxins and Gram-negative bacteria. Clin.Chim.Acta, 323 59-72. Hock, C., Konietzko, U., Streffer, J. R., Tracy, J., Signorell, A., Muller-Tillmanns, B., Lemke, U., Henke, K., Moritz, E., Garcia E., Wollmer, M. A., Umbricht, D., de Quervain, D. J., Hofmann, M., Maddalena, A., Papassotiropoulos, A., & Nitsch, R. M. (2003). Antibodies against beta-amyloid slow cognitive decline in Alzheimer's disease. Neuron, 38 547-554. Hogarth, P. M., Witort, E., Hulett, M. D., Bonnerot, C., Even, J., Fridman, W. H., & McKenzie, I. F. (1991). Structure of th e mouse beta Fc gamma receptor II gene. J.Immunol., 146 369-376. Holcomb, L., Gordon, M. N., McGowan, E., Yu X., Benkovic, S., Jantzen, P., Wright, K., Saad, I., Mueller, R., Morgan, D., Sanders, S., Zehr, C., O'Campo, K., Hardy, J., Prada, C. M., Eckman, C., Younkin, S., Hsiao, K., & Duff, K. (1998). Accelerated Alzheimer-type phenotype in transgenic mice carrying both mutant amyloid precursor protein and presenilin 1 transgenes. Nat.Med., 4, 97-100. Holness, C. L., da Silva, R. P., Fawcett, J., Gordon, S., & Simmons, D. L. (1993). Macrosialin, a mouse macr ophage-restricted glycoprotei n, is a member of the lamp/lgp family. J.Biol.Chem., 268, 9661-9666. Hsiao, K., Chapman, P., Nilsen, S., Eckma n, C., Harigaya, Y., Younkin, S., Yang, F., & Cole, G. (1996). Correlative memory de ficits, Abeta elevation, and amyloid plaques in transgenic mice. Science, 274, 99-102. Husemann, J., Loike, J. D., Anankov, R., Febbraio, M., & Silverstein, S. C. (2002). Scavenger receptors in neurobiology and neuropathology: their role on microglia and other cells of the nervous system. Glia, 40 195-205. In 't Veld, B. A., Launer, L. J., Hoes, A. W., Ott, A., Hofman, A., Breteler, M. M., & Stricker, B. H. (1998). NSAI Ds and incident Alzheimer's disease. The Rotterdam Study. Neurobiol.Aging, 19 607-611.
152 Irie-Sasaki, J., Sasaki, T., & Penninger, J. M. (2003). CD45 regulated signaling pathways. Curr.Top.Med.Chem., 3 783-796. Itagaki, S., McGeer, P. L., Akiyama, H., Zhu, S., & Selkoe, D. (1989). Relationship of microglia and astrocytes to amyloi d deposits of Alzheimer disease. J.Neuroimmunol., 24 173-182. Iwasaki, A. & Medzhitov, R. (2004). Toll-like receptor cont rol of the adaptive immune responses. Nat.Immunol., 5 987-995. Iwata, N., Mizukami, H., Shirotani, K., Taka ki, Y., Muramatsu, S., Lu, B., Gerard, N. P., Gerard, C., Ozawa, K., & Saido, T. C. (2004). Presynaptic localization of neprilysin contributes to efficient clearance of amyl oid-beta peptide in mouse brain. J.Neurosci., 24 991-998. Jain, N. K., Patil, C. S., Kulkarni, S. K., & Singh, A. (2002). Modulatory role of cyclooxygenase inhibitors in agingand scopolamine or lipopolysaccharideinduced cognitive dysfunction in mice. Behav.Brain Res., 133, 369-376. Jantzen, P. T., Connor, K. E., DiCarlo, G., We nk, G. L., Wallace, J. L., Rojiani, A. M., Coppola, D., Morgan, D., & Gordon, M. N. (2002). Microglial activation and beta -amyloid deposit reduction caused by a nitr ic oxide-releasing nonsteroidal antiinflammatory drug in amyloid precursor prot ein plus presenilin-1 transgenic mice. J.Neurosci., 22, 2246-2254. Kalmar, B., Kittel, A., Lemmens, R., Kornyei, Z., & Madarasz, E. (2001). Cultured astrocytes react to LPS with increased cyclooxygenase activity and phagocytosis. Neurochem.Int., 38 453-461. Kaneko, T., Stearns-Kurosawa, D. J., Taylor, F., Jr., Twigg, M., Osaki, K., Kinasewitz, G. T., Peer, G., & Kurosawa, S. (2003) Reduced neutrophil CD10 expression in nonhuman primates and humans after in vivo challenge with E. coli or lipopolysaccharide. Shock, 20 130-137. Kassed, C. A., Butler, T. L., Patton, G. W ., Demesquita, D. D., Navidomskis, M. T., Memet, S., Israel, A., & Pennypacker, K. R. (2004). Injury-induced NF-kappaB activation in the hippocampus: imp lications for neuronal survival. FASEB J., 18, 723-724. Kawarabayashi, T., Shoji, M., Harigaya Y., Yamaguchi, H., & Hirai, S. (1991). Expression of APP in the ear ly stage of brain damage. Brain Res., 563 334-338. Keller, R., Keist, R., & Joller, P. W. (1994). Macrophage response to bacteria and bacterial products: modulation of Fc ga mma receptors and secretory and cellular activities. Immunology, 81, 161-166.
153 Kheir-Eldin, A. A., Motawi, T. K., Gad, M. Z., & Abd-ElGawad, H. M. (2001). Protective effect of vitamin E, beta-car otene and N-acetylcysteine from the brain oxidative stress induced in rats by lipopolysaccharide. Int.J.Biochem.Cell Biol., 33, 475-482. Kim, W. G., Mohney, R. P., Wilson, B., Je ohn, G. H., Liu, B., & Hong, J. S. (2000). Regional difference in susceptibility to lipopolysaccharide-induced neurotoxicity in the rat brain: role of microglia J.Neurosci., 20 6309-6316. Kizaki, T., Oh-ishi, S., Ookawara, T., Ya mamoto, M., Izawa, T., & Ohno, H. (1996). Glucocorticoid-mediated generation of suppressor macrophages with high density Fc gamma RII during acute cold stress. Endocrinology, 137, 4260-4267. Koenigsknecht, J. & Landreth, G. (2004). Mi croglial phagocytosis of fibrillar betaamyloid through a beta1 integrin-dependent mechanism. J.Neurosci., 24, 98389846. Kopec, K. K. & Carroll, R. T. (1998). Alzh eimer's beta-amyloid peptide 1-42 induces a phagocytic response in murine microglia. J.Neurochem., 71 2123-2131. Kreutzberg, G. W. (1996). Mi croglia: a sensor for patholog ical events in the CNS. Trends Neurosci., 19, 312-318. Krieger, M. (2001). Scavenger receptor class B type I is a multiligand HDL receptor that influences diverse physiologic systems. J.Clin.Invest, 108 793-797. Kurt, M. A., Davies, D. C., & Kidd, M. (1999). beta-Amyloid immunoreactivity in astrocytes in Alzheimer's disease brai n biopsies: an electron microscope study. Exp.Neurol., 158 221-228. Laflamme, N. & Rivest, S. (2001). Toll-like r eceptor 4: the missing link of the cerebral innate immune response triggered by circul ating gram-negative bacterial cell wall components. FASEB J., 15, 155-163. Laithwaite, J. E., Benn, S. J., Yamate, J., FitzGerald, D. J., & LaMarre, J. (1999). Enhanced macrophage resistance to Pse udomonas exotoxin A is correlated with decreased expression of the low-density lipoprotein receptor-related protein Infect.Immun., 67, 5827-5833. Laszlo, G. & Dickler, H. B. (1990). Re gulation of Fc gamma R II expression and function by B lymphocyte activators. Mol.Immunol., 27, 1189-1193. Leake, A., Morris, C. M., & Whateley, J. (200 0). Brain matrix metalloproteinase 1 levels are elevated in Alzheimer's disease. Neurosci.Lett., 291 201-203.
154 Lee, D. H. & Wang, H. Y. (2003). Differentia l physiologic responses of alpha7 nicotinic acetylcholine receptors to beta-amyloid1-40 and beta-amyloid1-42. J.Neurobiol., 55, 25-30. Lehnardt, S., Lachance, C., Patrizi, S., Le febvre, S., Follett, P. L., Jensen, F. E., Rosenberg, P. A., Volpe, J. J., & Vart anian, T. (2002). The toll-like receptor TLR4 is necessary for lipopolysaccharide-induced oligodendrocyte injury in the CNS. J.Neurosci., 22, 2478-2486. Levy-Lahad, E., Wasco, W., Poorkaj, P., Roma no, D. M., Oshima, J., Pettingell, W. H., Yu, C. E., Jondro, P. D., Schmidt, S. D ., & Wang, K. (1995). Candidate gene for the chromosome 1 familial Alzheimer's disease locus. Science, 269 973-977. Lim, G. P., Yang, F., Chu, T., Chen, P., Beech, W., Teter, B., Tran, T., Ubeda, O., Ashe, K. H., Frautschy, S. A., & Cole, G. M. (2000). Ibuprofen suppresses plaque pathology and inflammation in a mouse model for Alzheimer's disease. J.Neurosci., 20 5709-5714. Liu, Q., Kawai, H., & Berg, D. K. (2001). beta -Amyloid peptide blocks the response of alpha 7-containing nicotinic re ceptors on hippocampal neurons. Proc.Natl.Acad.Sci.U.S.A, 98, 4734-4739. Loughlin, A. J., Woodroofe, M. N., & Cuzner M. L. (1993). Modul ation of interferongamma-induced major histocompatibility complex class II and Fc receptor expression on isolated microglia by transforming growth factor-beta 1, interleukin-4, noradrenaline and glucocorticoids. Immunology, 79 125-130. Lue, L. F., Brachova, L., Civin, W. H., & Rogers, J. (1996). Inflammation, A beta deposition, and neurofibrillary tangle fo rmation as correlates of Alzheimer's disease neurodegeneration. J.Neuropathol.Exp.Neurol., 55 1083-1088. Lynch, R. G., Sandor, M., Waldschmidt, T. J., Mathur, A., Schaiff, W. T., Berg, D. J., Snapp, K., Mueller, A., Robinson, M. G., & Noben, N. (1990). Lymphocyte Fc receptors: expression, re gulation and function. Mol.Immunol., 27, 1167-1179. Mahley, R. W. & Huang, Y. (1999). Apoli poprotein E: from atherosclerosis to Alzheimer's disease and beyond. Curr.Opin.Lipidol., 10, 207-217. Mandybur, T. I. & Chuirazzi, C. C. (1990). As trocytes and the plaques of Alzheimer's disease. Neurology, 40, 635-639. Marz, W., Scharnagl, H., Kirca, M., Bohl J., Gross, W., & Ohm, T. G. (1996). Apolipoprotein E polymorphism is associat ed with both senile plaque load and Alzheimer-type neurofibri llary tangle formation. Ann.N.Y.Acad.Sci., 777, 276280.
155 Marzolo, M. P., von Bernhardi, R., Bu, G., & Inestrosa, N. C. (2000). Expression of alpha(2)-macroglobulin receptor/low density lipoprotein receptor-related protein (LRP) in rat microglial cells J.Neurosci.Res., 60 401-411. Matsuoka, Y., Kitamura, Y., Takahashi, H., Tooyama, I., Kimura, H., Gebicke-Haerter, P. J., Nomura, Y., & Taniguchi, T. (1999). Interferon-gamma plus lipopolysaccharide induction of delayed neuronal apoptosis in rat hippocampus. Neurochem.Int., 34 91-99. Matsuoka, Y., Saito, M., LaFrancois, J., Saito, M., Gaynor, K., Olm, V., Wang, L., Casey, E., Lu, Y., Shiratori, C., Lemere, C., & Duff, K. (2003). Novel therapeutic approach for the treatment of Alzheimer' s disease by peripheral administration of agents with an affinity to beta-amyloid. J.Neurosci., 23, 29-33. McGeer, P. L., Itagaki, S., Tago, H., & Mc Geer, E. G. (1987). Reactive microglia in patients with senile dementia of th e Alzheimer type are positive for the histocompatibility glycoprotein HLA-DR. Neurosci.Lett., 79 195-200. McGeer, P. L., Itagaki, S., Boyes, B. E., & McGeer, E. G. (1988). Reactive microglia are positive for HLA-DR in the substantia nigra of Parkinson's and Alzheimer's disease brains. Neurology, 38 1285-1291. McGeer, P. L. & McGeer, E. G. (1995). The inflammatory response system of brain: implications for therapy of Alzheime r and other neurodegenerative diseases. Brain Res.Brain Res.Rev., 21 195-218. McGeer, P. L., Schulzer, M., & McGeer, E. G. (1996). Arthritis and anti-inflammatory agents as possible protective factors fo r Alzheimer's disease: a review of 17 epidemiologic studies. Neurology, 47, 425-432. McGeer, P. L. & McGeer, E. G. (2002). Innate immunity, local inflammation, and degenerative disease. Sci.Aging Knowledge.Environ., 2002 re3. McGehee, D. S. & Role, L. W. (1995). Physio logical diversity of ni cotinic acetylcholine receptors expressed by vertebrate neurons. Annu.Rev.Physiol, 57 521-546. Mohajeri, M. H., Kuehnle, K., Li, H., Poirie r, R., Tracy, J., & Nitsch, R. M. (2004). Antiamyloid activity of neprilysin in plaq ue-bearing mouse models of Alzheimer's disease. FEBS Lett., 562, 16-21. Montero-Menei, C. N., Sindji, L., Garcion, E., Mege, M., Couez, D., Gamelin, E., & Darcy, F. (1996). Early events of the infl ammatory reaction induced in rat brain by lipopolysaccharide intracerebral injecti on: relative contribution of peripheral monocytes and activated microglia. Brain Res., 724 55-66. Montero, C. (2003). The antigen-antibody reaction in immunohistochemistry. J.Histochem.Cytochem., 51 1-4.
156 Moore, K. J., El Khoury, J., Medeiros, L. A ., Terada, K., Geula, C., Luster, A. D., & Freeman, M. W. (2002). A CD36-in itiated signaling cascade mediates inflammatory effects of beta-amyloid. J.Biol.Chem., 277 47373-47379. Morgan, D., Diamond, D. M., Gottschall, P. E., Ugen, K. E., Dickey, C., Hardy, J., Duff, K., Jantzen, P., DiCarlo, G., Wilcock, D ., Connor, K., Hatcher, J., Hope, C., Gordon, M., & Arendash, G. W. (2000). A beta peptide vaccination prevents memory loss in an animal m odel of Alzheimer's disease. Nature, 408 982-985. Morishima-Kawashima, M. & Ihara, Y. ( 2002). Alzheimer's disease: beta-Amyloid protein and tau J.Neurosci.Res., 70 392-401. Moser, N., Wevers, A., Plomann, M., Roth, U., Dramiga, J., et al. (2004). Transmitterreceptor deficient mice as tools in speci ficity testing of receptor antibodies. Poster Presentation at th e annual meeting for the Society for Neuroscience Nadeau, S. & Rivest, S. (2002). Endotoxemia prevents the cerebral inflammatory wave induced by intraparenchymal lipopolysaccharid e injection: role of glucocorticoids and CD14. J.Immunol., 169 3370-3381. Nishimura, T., Narita, T., Miyazaki, E., Ito, T ., Nishimoto, N., Yoshizaki, K., Martial, J. A., Bellfroid, E. J., Vissing, H., & Taniya ma, T. (2001). Characterization of the human Fc gamma RIIB gene promoter: human zinc-finger proteins (ZNF140 and ZNF91) that bind to different regions function as transcri ption repressors. Int.Immunol., 13, 1075-1084. Nonaka, N., Shioda, S., & Banks, W. A. (2005). Effect of lipopolysaccharide on the transport of pituitary adenylate cycl ase activating polypeptide across the bloodbrain barrier. Exp.Neurol., 191, 137-144. Nordberg, A., Hellstrom-Lindahl, E., Lee, M ., Johnson, M., Mousavi, M., Hall, R., Perry, E., Bednar, I., & Court, J. (2002). Chronic nicotine treatment reduces betaamyloidosis in the brain of a mouse model of Alzheimer's disease (APPsw). J.Neurochem., 81 655-658. Nunan, J. & Small, D. H. (2002). Proteolyti c processing of the amyloid-beta protein precursor of Alzheimer's disease Essays Biochem., 38, 37-49. Ohata, M., Suzuki, H., Sakamoto, K., Hash imoto, K., Nakajima, H., Yamauchi, M., Hokkyo, K., Yamada, H., & Toda, G. (2004) Pioglitazone prevents acute liver injury induced by ethanol and lipopol ysaccharide through the suppression of tumor necrosis factor-alpha. Alcohol Clin.Exp.Res., 28 139S-144S. Olson, J. K. & Miller, S. D. (2004). Microglia initiate central nervous system innate and adaptive immune responses through multiple TLRs. J.Immunol., 173, 3916-3924.
157 Orr-Urtreger, A., Goldner, F. M., Saeki, M., Lorenzo, I., Goldberg, L., De Biasi, M., Dani, J. A., Patrick, J. W., & Beaudet, A. L. (1997). Mice deficient in the alpha7 neuronal nicotinic acetylcholine receptor lack alpha-bungarotoxin binding sites and hippocampal fast nicotinic currents. J.Neurosci., 17, 9165-9171. Palsson-McDermott, E. M. & O'Neill, L. A. (2004). Signal transduction by the lipopolysaccharide receptor, Toll-like receptor-4. Immunology, 113, 153-162. Paterson, D. & Nordberg, A. (2000). Neuronal nicotinic receptors in the human brain. Prog.Neurobiol., 61, 75-111. Paul, S. M., Koistanaho, M., Lin., S., Es terman, E., Koger, D., et al. (2004) Apolipoprotein E promotes astrocyte colo calization and degradation of deposited amyloid-beta peptides. Presente d at the annual meeting of the Society for Neuroscience Paylor, R., Nguyen, M., Crawley, J. N., Patr ick, J., Beaudet, A., & Orr-Urtreger, A. (1998). Alpha7 nicotinic receptor subunits are not necessary for hippocampaldependent learning or sens orimotor gating: a behavi oral characterization of Acra7-deficient mice. Learn.Mem., 5 302-316. Peiser, L., Mukhopadhyay, S., & Gordon, S. (2 002). Scavenger receptors in innate immunity. Curr.Opin.Immunol., 14, 123-128. Pennypacker, K. R., Hernandez, H., Benkovic, S., Morgan, D. G., Willing, A. E., & Sanberg, P. R. (1999). Induction of pres enilins in the rat brain after middle cerebral arterial occlusion. Brain Res.Bull., 48 539-543. Perry, E. K., Martin-Ruiz, C. M., & Cour t JA (2001). Nicotinic receptor subtypes in human brain related to aging and dementia. Alcohol, 24, 63-68. Pettit, D. L., Shao, Z., & Yakel, J. L. (2001). beta-Amyloid(142) peptide directly modulates nicotinic receptors in the rat hippocampal slice. J.Neurosci., 21, RC120. Platt, N. & Gordon, S. (2001). Is the cla ss A macrophage scavenger receptor (SR-A) multifunctional? The mouse's tale. J.Clin.Invest, 108 649-654. Qiao, X., Cummins, D. J., & Paul, S. M. (2001). Neuroinflammation-induced acceleration of amyloid deposition in the APPV717F transgenic mouse Eur.J.Neurosci., 14 474-482. Quinn, J., Montine, T., Morrow, J., Woodwar d, W. R., Kulhanek, D., & Eckenstein, F. (2003). Inflammation and cerebral amyloidos is are disconnected in an animal model of Alzheimer's disease J.Neuroimmunol., 137 32-41.
158 Raghavendra, V., Tanga, F. Y., & DeLeo, J. A. (2004). Complete Freunds adjuvantinduced peripheral inflammation evokes glial activation and proinflammatory cytokine expression in the CNS. Eur.J.Neurosci., 20 467-473. Ramprasad, M. P., Terpstra, V., Kondrate nko, N., Quehenberger, O., & Steinberg, D. (1996). Cell surface expression of mouse macrosialin and human CD68 and their role as macrophage receptors for oxidized low density lipoprotein. Proc.Natl.Acad.Sci.U.S.A, 93, 14833-14838. Reines, S. A., Block, G. A., Morris, J. C., Liu, G., Nessly, M. L., Lines, C. R., Norman, B. A., & Baranak, C. C. (2004). Rofecoxib: no effect on Alzheimer's disease in a 1-year, randomized, b linded, controlled study. Neurology, 62, 66-71. Rhodin, J. A. & Thomas, T. (2001). A vascul ar connection to Alzheimer's disease. Microcirculation., 8, 207-220. Del Rio-Hortega, P. (1936) in Cytology a nd Cellular Pathology of the Nervous System (Penfield, W., ed.) pp. 481-494. Heuber. Romero, I. A., Radewicz, K., Jubin, E., Mich el, C. C., Greenwood, J., Couraud, P. O., & Adamson, P. (2003). Changes in cytoskel etal and tight j unctional proteins correlate with decreased permeability indu ced by dexamethasone in cultured rat brain endothelial cells. Neurosci.Lett., 344, 112-116. Rusted, J. M., Newhouse, P. A., & Levi n, E. D. (2000). Nicotinic treatment for degenerative neuropsychiatric disorder s such as Alzheimer's disease and Parkinson's disease. Behav.Brain Res., 113, 121-129. Sagi, S. A., Weggen, S., Eriksen, J., Golde, T. E., & Koo, E. H. (2003). The noncyclooxygenase targets of non-steroidal an ti-inflammatory drugs, lipoxygenases, peroxisome proliferator-activated recept or, inhibitor of kappa B kinase, and NF kappa B, do not reduce am yloid beta 42 production J.Biol.Chem., 278 3182531830. Sakai, A., Ujike, H., Nakata, K., Takehisa, Y., Imamura, T., Uchida, N., Kanzaki, A., Yamamoto, M., Fujisawa, Y., Okumura, K ., & Kuroda, S. (2004). Association of the Neprilysin gene with susceptibilit y to late-onset Alzheimer's disease. Dement.Geriatr.Cogn Disord., 17 164-169. Saragoza, P. A., Modir, J. G., Goel, N., French, K. L., Li, L., Nowak, M. W., & Stitzel, J. A. (2003). Identification of an alterna tively processed nico tinic receptor alpha7 subunit RNA in mouse brain. Brain Res.Mol.Brain Res., 117, 15-26.
159 Sasaki, A., Shoji, M., Harigaya, Y., Kawaraba yashi, T., Ikeda, M., Naito, M., Matsubara, E., Abe, K., & Nakazato, Y. (2002). Amyl oid cored plaques in Tg2576 transgenic mice are characterized by giant plaques, s lightly activated microglia, and the lack of paired helical filame nt-typed, dystrophic neurites. Virchows Arch., 441 358367. Scharf, S., Mander, A., Ugoni, A., Vajda, F., & Christophidis, N. (1999). A double-blind, placebo-controlled trial of diclofenac/ misoprostol in Alzheimer's disease. Neurology, 53 197-201. Schenk, D., Hagen, M., & Seubert, P. (2004) Current progress in beta-amyloid immunotherapy. Curr.Opin.Immunol., 16 599-606. Schmidt, A. M., Yan, S. D., Yan, S. F., & Stern, D. M. (2001). The multiligand receptor RAGE as a progression factor amplifyi ng immune and inflammatory responses. J.Clin.Invest, 108, 949-955. Schoepfer, R., Conroy, W. G., Whiting, P., Go re, M., & Lindstrom, J. (1990). Brain alpha-bungarotoxin binding pr otein cDNAs and MAbs reveal subtypes of this branch of the ligand-gated ion channel gene superfamily. Neuron, 5, 35-48. Seguela, P., Wadiche, J., Dineley-Miller, K., Dani, J. A., & Patrick, J. W. (1993). Molecular cloning, functional pr operties, and distribution of rat brain alpha 7: a nicotinic cation channel hi ghly permeable to calcium. J.Neurosci., 13 596-604. Sen, R. & Baltimore, D. (1986). Inducibility of kappa immunoglobulin enhancer-binding protein Nf-kappa B by a posttranslational mechanism. Cell, 47, 921-928. Severance, E. G., Zhang, H., Cruz, Y., Pakhl evaniants, S., Hadley, S. H., Amin, J., Wecker, L., Reed, C., & Cuevas, J. ( 2004). The alpha7 nicotinic acetylcholine receptor subunit exists in two isoforms that contribute to functional ligand-gated ion channels. Mol.Pharmacol., 66, 420-429. Shaffer, L. M., Dority, M. D., Gupta-Bans al, R., Frederickson, R. C., Younkin, S. G., & Brunden, K. R. (1995). Amyloid beta protei n (A beta) removal by neuroglial cells in culture. Neurobiol.Aging, 16, 737-745. Sheng, J. G., Bora, S. H., Xu, G., Borchelt, D. R., Price, D. L., & Koliatsos, V. E. (2003). Lipopolysaccharide-induced-neuroinfl ammation increases intracellular accumulation of amyloid precursor protein and amyloid beta peptide in APPswe transgenic mice. Neurobiol.Dis., 14 133-145. Sherrington, R., Rogaev, E. I., Liang, Y., R ogaeva, E. A., Levesque, G., Ikeda, M., Chi, H., Lin, C., Li, G., & Holman, K. (1995) Cloning of a gene bearing missense mutations in early-onset familial Alzheimer's disease. Nature, 375 754-760.
160 Shytle, R. D., Mori, T., Townsend, K., Vendrame, M., Sun, N., Zeng, J., Ehrhart, J., Silver, A. A., Sanberg, P. R., & Tan, J. (2004). Cholinergic modulation of microglial activation by alpha 7 nicotinic receptors. J.Neurochem., 89 337-343. Siman, R., Card, J. P., Nelson, R. B., & Davis, L. G. (1989). Expression of beta-amyloid precursor protein in reactive astr ocytes following neuronal damage. Neuron, 3, 275-285. Singh, A. K. & Jiang, Y. (2004). How does pe ripheral lipopolysaccharide induce gene expression in the brain of rats? Toxicology, 201 197-207. Sivo, J., Politis, A. D., & Vogel, S. N. ( 1993). Differential effects of interferon-gamma and glucocorticoids on Fc gamma R gene expression in murine macrophages. J.Leukoc.Biol., 54, 451-457. Sly, L. M., Krzesicki, R. F., Brashler, J. R., Buhl, A. E., McKinley, D. D., Carter, D. B., & Chin, J. E. (2001). Endogenous brain cytokine mRNA and inflammatory responses to lipopolysacchar ide are elevated in th e Tg2576 transgenic mouse model of Alzheimer's disease. Brain Res.Bull., 56 581-588. Small, D. H., Mok, S. S., & Bornstein, J. C. (2001). Alzheimer's disease and Abeta toxicity: from top to bottom. Nat.Rev.Neurosci., 2, 595-598. Streit, W. J., Walter, S. A., & Pennell, N. A. (1999). Reactive microgliosis. Prog.Neurobiol., 57 563-581. Streit, W. J. (2002). Microglia as neuroprot ective, immunocompetent cells of the CNS. Glia, 40, 133-139. Sugaya, K., Chou, S., Xu, S. J., & McKinney, M. (1998). Indicators of glial activation and brain oxidative stress after in traventricular infusion of endotoxin. Brain Res.Mol.Brain Res., 58, 1-9. Szczepanik, A. M. & Ringheim, G. E. (2003). IL-10 and glucocorticoids inhibit Abeta(142)and lipopolysaccharide-induced pro-in flammatory cytokine and chemokine induction in the centr al nervous system. J.Alzheimers.Dis., 5, 105-117. Szekely, C. A., Thorne, J. E., Zandi, P. P ., Ek, M., Messias, E., Breitner, J. C., & Goodman, S. N. (2004). Nonsteroidal anti-inflammatory drugs for the prevention of Alzheimer's disease: a systematic review. Neuroepidemiology, 23, 159-169. Tanga, F. Y., Raghavendra, V., & DeLeo, J. A. (2004). Quantitative real-time RT-PCR assessment of spinal microglial and astrocyt ic activation markers in a rat model of neuropathic pain. Neurochem.Int., 45 397-407.
161 Teaktong, T., Graham, A., Court, J., Perry, R., Jaros, E., Johnson, M., Hall, R., & Perry, E. (2003). Alzheimer's disease is associat ed with a selective increase in alpha7 nicotinic acetylcholine receptor immunoreactivity in astrocytes. Glia, 41, 207211. Thomas, C. J., Kapoor, M., Sharma, S., Bausi nger, H., Zyilan, U., Lipsker, D., Hanau, D., & Surolia, A. (2002). Evidence of a trimolecular complex involving LPS, LPS binding protein and soluble CD14 as an effector of LPS response. FEBS Lett., 531 184-188. Timon, M. & Beverley, P. C. (2001). Structur al and functional analysis of the human CD45 gene (PTPRC) upstream region: evid ence for a functional promoter within the first intron of the gene. Immunology, 102 180-189. Van Nostrand, W. E., Melchor, J. P., Roma nov, G., Zeigler, K., & Davis, J. (2002). Pathogenic effects of cerebral amyloid a ngiopathy mutations in the amyloid betaprotein precursor. Ann.N.Y.Acad.Sci., 977 258-265. Vehmas, A. K., Kawas, C. H., Stewart, W. F., & Troncoso, J. C. (2003). Immune reactive cells in senile plaques and cognitive decline in Alzheimer's disease. Neurobiol.Aging, 24 321-331. Villiger, Y., Szanto, I., Jaconi, S., Blanchet, C., Buisson, B., Krause, K. H., Bertrand, D., & Romand, J. A. (2002). Expression of an alpha7 duplicate nicotinic acetylcholine receptor-related protein in human leukocytes J.Neuroimmunol., 126, 86-98. Walsh, D. M., Klyubin, I., Fadeeva, J. V ., Rowan, M. J., & Selkoe, D. J. (2002). Amyloid-beta oligomers: their producti on, toxicity and therapeutic inhibition Biochem.Soc.Trans., 30, 552-557. Wang, H., Yu, M., Ochani, M., Amella, C. A., Tanovic, M., Susarla, S., Li, J. H., Wang, H., Yang, H., Ulloa, L., Al Abed, Y., Czura, C. J., & Tracey, K. J. (2003). Nicotinic acetylcholine receptor alpha7 subunit is an essential regulator of inflammation. Nature, 421, 384-388. Wang, H. Y., Lee, D. H., D'Andrea, M. R., Peterson, P. A., Shank, R. P., & Reitz, A. B. (2000a). beta-Amyloid(1-42) binds to alpha 7 nicotinic acetylcholine receptor with high affinity. Implications for Alzheimer's disease pathology. J.Biol.Chem., 275, 5626-5632. Wang, H. Y., Lee, D. H., Davis, C. B., & Shank, R. P. (2000b). Amyloid peptide Abeta(1-42) binds selectively and with picomolar affinity to alpha7 nicotinic acetylcholine receptors. J.Neurochem., 75 1155-1161.
162 Weggen, S., Eriksen, J. L., Das, P., Sagi, S. A., Wang, R., Pietrzik, C. U., Findlay, K. A., Smith, T. E., Murphy, M. P., Bulter, T., Kang, D. E., Marquez-Sterling, N., Golde, T. E., & Koo, E. H. (2001). A subset of NSAIDs lower amyloidogenic Abeta42 independently of cyclooxygenase activity. Nature, 414, 212-216. Wegiel, J., Wang, K. C., Tarnawski, M., & Lach, B. (2000). Micr oglia cells are the driving force in fibrillar pl aque formation, whereas astroc ytes are a leading factor in plague degradation. Acta Neuropathol.(Berl), 100 356-364. Wenk, G. L., McGann, K., Mencarelli, A., Ha uss-Wegrzyniak, B., Del Soldato, P., & Fiorucci, S. (2000). Mechanisms to prevent the toxici ty of chronic neuroinflammation on forebr ain cholinergic neurons Eur.J.Pharmacol., 402, 7785. Wenk, G. L., Rosi, S., McGann, K., & HaussWegrzyniak, B. (2002). A nitric oxidedonating flurbiprofen deriva tive reduces neuroinflamma tion without interacting with galantamine in the rat. Eur.J.Pharmacol., 453 319-324. Wevers, A., Monteggia, L., Nowacki, S., Bloc h, W., Schutz, U., Lindstrom, J., Pereira, E. F., Eisenberg, H., Giacobini, E., de Vos, R. A., Steur, E. N., Maelicke, A., Albuquerque, E. X., & Schroder, H. (1999). Expression of nicotinic acetylcholine receptor subunits in the cerebral cortex in Alzheimer's disease: histotopographical correlation with amyloid plaques and hyperphosphorylated-tau protein. Eur.J.Neurosci., 11, 2551-2565. Wilcock, D. M., Gordon, M. N., Ugen, K. E., Gottschall, P. E., DiCarlo, G., Dickey, C., Boyett, K. W., Jantzen, P. T., Connor, K. E., Melachrino, J., Hardy, J., & Morgan, D. (2001). Number of Abeta inoculations in APP+PS1 transgenic mice influences antibody titers, microglial activati on, and congophilic plaque levels. DNA Cell Biol., 20, 731-736. Wilcock, D. M., DiCarlo, G., Henderson, D ., Jackson, J., Clarke, K., Ugen, K. E., Gordon, M. N., & Morgan, D. (2003). Intr acranially administered anti-Abeta antibodies reduce beta-amyloid depositi on by mechanisms both independent of and associated with microglial activation J.Neurosci., 23 3745-3751. Wilcock, D. M., Rojiani, A., Rosenthal, A., Levkowitz, G., Subbarao, S., Alamed, J., Wilson, D., Wilson, N., Freeman, M. J., Gordon, M. N., & Morgan, D. (2004a). Passive amyloid immunotherapy clears amyloid and transiently activates microglia in a transgenic mous e model of amyloid deposition. J.Neurosci., 24 6144-6151. Wilcock, D. M., Munireddy, S. K., Rosent hal, A., Ugen, K. E., Gordon, M. N., & Morgan, D. (2004b). Microglial activati on facilitates Abeta plaque removal following intracranial antiAbeta antibody administration. Neurobiol.Dis., 15, 1120.
163 Willard, L. B., Hauss-Wegrzyniak, B., & Wenk, G. L. (1999). Pathological and biochemical consequences of acute a nd chronic neuroinflammation within the basal forebrain cholinergic system of rats. Neuroscience, 88 193-200. Willard, L. B., Hauss-Wegrzyniak, B., Danysz, W., & Wenk, G. L. (2000). The cytotoxicity of chronic neuroinfla mmation upon basal forebrain cholinergic neurons of rats can be attenuate d by glutamatergic antagonism or cyclooxygenase-2 inhibition. Exp.Brain Res., 134 58-65. Woo, C. H., Lim, J. H., & Kim, J. H. (2004). Lipopolysaccharide induces matrix metalloproteinase-9 expression via a mitochondrial reactive oxygen species-p38 kinase-activator protein1 pathway in Raw 264.7 cells J.Immunol., 173, 69736980. Wright, K. L., Morgan, D. G., Yu, X., Goss, J. R., Salbaum, J. M., Duff, K., & Gordon, M. N. (1999). Mice transgenic for a huma n amyloid precursor protein promoterlacZ reporter construct. J.Mol.Neurosci., 13 111-120. Wyss-Coray, T., Yan, F., Lin, A. H., Lambris, J. D., Alexander, J. J., Quigg, R. J., & Masliah, E. (2002). Prominent neurodegene ration and increased plaque formation in complement-inhibited Alzheimer's mice Proc.Natl.Acad.Sci.U.S.A, 99, 1083710842. Wyss-Coray, T., Loike, J. D., Brionne, T. C., Lu, E., Anankov, R., Yan, F., Silverstein, S. C., & Husemann, J. (2003). Adult mouse astr ocytes degrade amyloid-beta in vitro and in situ. Nat.Med., 9 453-457. Yamaguchi, M., Hirai, K., Komiya, A., Miyama su, M., Furumoto, Y., Teshima, R., Ohta, K., Morita, Y., Galli, S. J., Ra, C., & Yamamoto, K. (2001). Regulation of mouse mast cell surface Fc epsilon RI expression by dexamethasone. Int.Immunol., 13 843-851. Yan, Q., Zhang, J., Liu, H., Babu-Khan, S., Vassar, R., Biere, A. L., Citron, M., & Landreth, G. (2003). Anti-inflammatory drug therapy alters beta-amyloid processing and deposition in an anim al model of Alzheimer's disease. J.Neurosci., 23 7504-7509. Yao, J., Harvath, L., Gilbert, D. L., & Colton, C. A. (1990). Chemotaxis by a CNS macrophage, the microglia. J.Neurosci.Res., 27 36-42. Yoshiyama, Y., Asahina, M., & Hattori, T. (2000). Selective di stribution of matrix metalloproteinase-3 (MMP-3) in Alzheimer's disease brain. Acta Neuropathol.(Berl), 99, 91-95.
164 Zhao, L., Teter, B., Morihara, T., Lim, G. P ., Ambegaokar, S. S., Ubeda, O. J., Frautschy, S. A., & Cole, G. M. (2004). Insulin-degr ading enzyme as a downstream target of insulin receptor signaling cascade: imp lications for Alzheimer's disease intervention J.Neurosci., 24 11120-11126. Zlokovic, B. V. (2004). Clearing amyloi d through the blood -brain barrier J.Neurochem., 89 807-811.
About the Author Donna Lorraine Herber, ne McCleary, gr aduated from Jacks onville University with a degree in biology in 1991. She then spent almost 10 years in the pharmaceutical industry as an analytical chemist. Her i ndustry career culminated at Bausch & Lomb Pharmaceuticals as a team leader in research and development. In 2001, Donna returned to academia in pursuit of a Ph.D. Over th e course of three and a half years, Donna presented at four international conferences successfully published two peer reviewed research articles, and submitted a third for consideration which is expected to be published in 2005.
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 001681047
007 cr mnu|||uuuuu
008 051220s2004 flu sbm s000 0 eng d
datafield ind1 8 ind2 024
subfield code a E14-SFE0001075
Herber, Donna Lorraine.
Neuroinflammation in Alzheimers disease
h [electronic resource] :
b characterization and modification of the response of transgenic mice to intrahippocampal lipopolysaccharide administration /
by Donna Lorraine Herber.
[Tampa, Fla.] :
University of South Florida,
Thesis (Ph.D.)--University of South Florida, 2004.
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 175 pages.
ABSTRACT: Alzheimers disease (AD) is pathologically characterized by amyloid plaques, neurofibrillary tangles, inflammation, and neurodegeneration. According to the amyloid hypothesis of AD, the central mediating event of the disease is the deposition of amyloid. The inflammation hypothesis of AD states that it is the inflammatory response to plaques and tangles, rather than the actual lesions, which causes the disease. Studies described here combine the two approaches into a single model. Four studies are presented using a basic protocol of intrahippocampal lipopolysaccharide (LPS) injection to stimulate inflammation in transgenic mice. The first study looked at alpha7 nicotinic receptors during the glial response to Abeta deposits and LPS. Reactive astrocytes which immunolabeled for alpha7 were co-localized with Congophilic deposits in APP and APP+PS1 mice, and increased after LPS injection.Unfortunately, LPS injection into alpha7 knock out mice revealed the alpha7 labeling to be nonspecific. The second study evaluated the time course of protein and gene expression after LPS injection into nontransgenic mice. This experiment identified both a transient and chronic microglial inflammatory response, with changes in cell morphology. The third study evaluated a similar time course in APP mice. Concurrent with the inflammatory response, transient reductions in Abeta burden were seen, though compact plaque load was unaffected. The fourth and final study used dexamethasone to inhibit LPS-induced inflammation in APP mice. LPS injection reduced Abeta burden, but was completely blocked by dexamethasone co-treatment. Though dexamethasone inhibited LPS-induced CD45 and complement receptor 3 levels (markers of general microglial activation), dexamethasone had no effect on scavenger receptor A or Fc gamma receptor II/III levels.
Adviser: Marcia Gordon.
t USF Electronic Theses and Dissertations.