<%BANNER%>

Peripheral Expression of Plasma Gelsolin as a Treatment for Alzheimer's Disease


PAGE 1

PERIPHERAL EXPRESSION OF PLASMA GELSOLIN AS A TREATMENT FOR ALZHEIMER’S DISEASE By AARON HIRKO A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL OF THE UNIVERSITY OF FLOR IDA IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY UNIVERSITY OF FLORIDA 2006

PAGE 2

Copyright 2006 by Aaron Hirko

PAGE 3

To my daughter, Ava Wrenn Hirko; and to her brother, Elliott Todd Hirko

PAGE 4

iv ACKNOWLEDGMENTS I would like to thank my parents, Robert and Theresa Hirko, for all of their love and support. I also need to thank my wife, Laura; daughter, Ava; and my son Elliott for their inspiration, love, and support. I owe a deep debt of gratit ude to my mentor, Dr. Jeffrey Hughes, for giving me the opportunity to study in the pharmaceutics progr am. I also need to thank Dr. Edwin Meyer for taking me in as an undergrad, in troducing me to Dr. Hughes, and giving me space in his laboratory. I thank my supervisory committee members (Dr. Michael King, Dr. Guenther Hochhaus, and Dr. Sihong Song) fo r all the excellent advice and insight. I acknowledge Dr. Raj Rao, Dr. Ke Wu, Dr. Ron Klein, Dr. Yan Gong, Dr. Ke Ren, and Dr. Preeti Yadava for their friendship, help, and advice. Finally, I thank all of the students and staff from the Department of Pharmaceutics.

PAGE 5

v TABLE OF CONTENTS page ACKNOWLEDGMENTS.................................................................................................iv LIST OF FIGURES.........................................................................................................viii ABSTRACT....................................................................................................................... xi CHAPTER 1 INTRODUCTION........................................................................................................1 Dementia....................................................................................................................... 1 Discovery of Alzheimer’s disease................................................................................2 Current Therapies.........................................................................................................3 Inherited Alzheimer’s...................................................................................................3 Amyloid Precursor Protein...........................................................................................4 Amyloid Cascade Hypothesis.......................................................................................5 Evidence Supporting the Amyloid Cascade Hypothesis.......................................6 Critics of the Amyloid Cascade Hypothesis..........................................................6 Elan and Wyeth Trial.............................................................................................8 Sink-Hypothesis....................................................................................................8 Gelsolin................................................................................................................10 2 MATERIALS AND METHODS...............................................................................14 Reagents......................................................................................................................1 4 Subcloning Vectors.....................................................................................................14 Large Scale Plasmid Preparation................................................................................15 Cell Culture.................................................................................................................17 Animals and Procedures.............................................................................................17 Immunoprecipitation and Western Blot......................................................................19 Detection of Message.................................................................................................20 Enzyme Linked Immunosorbent Assay......................................................................21 Histochemistry............................................................................................................22 Immunohistochemistry...............................................................................................23 Image Analysis...........................................................................................................24

PAGE 6

vi 3 THE INTERACTION OF PLAMA GELSOLIN AND AMYLOID ......................26 Introduction.................................................................................................................26 Experimental Methods and Results............................................................................27 Measurement of 125I Labeled A (1-42) Binding to Recombinant Human Plasma Gelsolin...............................................................................................27 Measuring Binding of HiLyte Fluor 488 Labeled Amyloid (1-42) to Bovine Plasma Gelsolin Using Fluorescence Anisotropy............................................27 HiLyte Fluor 488 Labeled A (1-42) Fibril Formation.......................................30 HiLyte Fluor 488 Labeled A (1-42) Fibril Disassembly with Gelsolin............30 Measuring Binding Amyloid to Plasma Gelsolin Using Surface Plasmon Resonance........................................................................................................32 Conclusions.................................................................................................................34 4 EXPRESSING PLASMA GELSOLIN AND EFFECTS IN TRANSGENIC MICE48 Introduction.................................................................................................................48 Results and Discussion:..............................................................................................50 Conformation of Vector Product and Activity....................................................50 Hydrodynamic Gene Delivery in Mice...............................................................50 Hematoxylin and Eosin Staining in Mi ce After Hydrodynamic Gene Transfer.51 Conclusions.................................................................................................................52 5 EFFECT OF GELSOLIN EXPRES SION ON AMYLOID DEPOSITION...............57 Introduction.................................................................................................................57 Results........................................................................................................................ .59 Mice Expressing Swedish Mutant Amyl oid Precusor Protein (Mouse/Human Hybrid) and Exon 9 Deleted Mutant Presenilin-1...........................................59 Message Detection.......................................................................................59 Total Brain Amyloid (1-42) Concentrations.............................................60 Plasma Amyloid Concentrations...............................................................61 Dense Cored Amyloid Deposits...................................................................61 Diffuse Amyloid Deposits............................................................................62 Soluble Amyloid Oligomers.........................................................................63 Microglia......................................................................................................64 Astrocytes.....................................................................................................66 Mice Expressing Swedish Mutant Amyl oid Precursor Protein (Human) and M146L Mutant Presenilin-1.............................................................................66 Message Detection.......................................................................................66 Amyloid (1-42) Quantification by ELISA................................................67 Plasma Amyloid Concentrations...............................................................68 Dense Cored Amyloid Deposits...................................................................69 Diffuse Amyloid Deposits............................................................................71 Microglia, Soluble Oligomers, and Astrocytes............................................72 Regression Analysis.....................................................................................74

PAGE 7

vii Conclusion...........................................................................................................75 6 DISCUSSION AND FUTURE DIRECTIONS........................................................102 LIST OF REFERENCES.................................................................................................109 BIOGRAPHICAL SKETCH...........................................................................................124

PAGE 8

viii LIST OF FIGURES Figure page 1-1 Amyloid precursor protein processing........................................................................13 2-1 Vectors.................................................................................................................... .....25 2-2 Restriction digest to c onfirm orientation of clones.....................................................25 3-1 Binding of 125I labeled A 1-42 to Whatman GF/C filters..........................................37 3-2 Binding of amyloid (1-42) to the antibody 6E10.....................................................37 3-3 Binding of plasma gelsolin to fluorescently labeled amyloid (1-42).......................38 3-4 Amyloid (1-42) binding to plasma gelsolin after concentration by centrifugal filtration....................................................................................................................3 8 3-5 Fibril formation of HiLyte Fluor 488 Labeled A (1-42)...........................................39 3-6 Dialysis experiment.....................................................................................................40 3-7 Amino-coupling of amyloid 1-40 to CM5 chip........................................................41 3-8 Amino-coupling of amyloid 1-42 to CM5 chip........................................................42 3-9 Antibody binding to am yloid coupled CM5 chip........................................................43 3-10 Human plasma gelsolin bindi ng to amyloid coupled CM5 chip...............................44 3-11 Amino-coupling of human plasma gelsolin to CM5 chip.........................................45 3-12 Binding of GS-2C4 to human plasma gelsolin coupled CM5 chip...........................46 3-13 Binding of amyloid 1-40 to human plasma gelsolin coupled CM5 chip................47 3-14 Binding of amyloid 1-42 to human plasma gelsolin coupled CM5 chip................47 4-1 Immunoblot of transfected media................................................................................53 4-2 Immunoprecipitation...................................................................................................53 4-3 Bioluminescence resulting from hydrodynamic gene transfer....................................54

PAGE 9

ix 4-4 Green fluorescent protein............................................................................................55 4-5 Western blot of plasma samples..................................................................................55 4-6 Hematoxylin and eosin staining..................................................................................56 5-1 Detecting message in mo/huAPP/PS1 E9 mice...............................................................78 5-2 Amyloid (1-42) concentrations in mo/huAPP/PS1 E9 brains.......................................78 5-3 Plasma concentrations of Amyloid in mo/huAPP/PS1 E9 mice......................................79 5-4 Analysis of dense-core amyloid deposits in mo/huAPP/PS1 E9 mice.............................80 5-5 Analysis of diffuse amyloid deposits in mo/huAPP/PS1 E9 mice...................................81 5-6 Dense-core amyloid deposits in mo/huAPP/PS1 E9 mice...............................................82 5-7 Diffuse amyloid deposits in mo/huAPP/PS1 E9 mice.....................................................83 5-8 Side by side comp arison of staining............................................................................84 5-9 Staining intensity of so luble amyloid oligomers in mo/huAPP/PS1 E9 mice.................85 5-10 High magnification tiled images of amyloid oligomer staining................................85 5-11 Cell types stained for soluble amyloi d oligomers in pUFGL-injected mice.............86 5-12 Staining intensity of microglia in mo/huAPP/PS1 E9 mice...........................................87 5-13 Low magnification images of microglia....................................................................88 5-14 High magnification images of microglia staining.....................................................89 5-15 Average GFAP percent stained area in mo/huAPP/PS1 E9 cortex and hippocampus..89 5-16 High magnification images of astrocyt es surrounding congo red positive amyloid deposits.....................................................................................................................90 5-17 Message detection in hAPP/PS1M146L mice...............................................................90 5-18 Amyloid (1-42) concentrations in huAPP/PS1M146L brains.....................................91 5-19 Plasma concentrations of amyloid in huAPP/PS1M146L mice.....................................92 5-20 Dense-core amyloid deposits in huAPP/PS1 L mice..............................................93 5-21 Analysis of dense-core amyloid deposits in huAPP/PS1 L mice...........................94

PAGE 10

x 5-22 Analysis of dense-core amyloid depos its in terms of fractions compared to untreated littermates in huAPP/PS1 L mice..........................................................95 5-23 Diffuse amyloid deposits in huAPP/PS1 L mice....................................................96 5-24 Analysis of diffuse amyloid deposits in huAPP/PS1 L mice.................................97 5-25 Analysis of diffuse amyloid deposits in terms of fractions compared to untreated littermates in huAPP/PS1 L mice..........................................................................98 5-26 Side by side comparison of staining in. huAPP/PS1M146L mice..................................99 5-27 Analysis of microglia, soluble am yloid oligomers, and astrocytes in huAPP/PS1M146L mice..............................................................................................100 5-28 Linear regression analysis.......................................................................................101

PAGE 11

xi Abstract of Dissertation Pres ented to the Graduate School of the University of Florida in Partial Fulfillment of the Requirements for the Degree of Doctor of Philosophy PERIPHERAL EXPRESSION OF PLASMA GELSOLIN AS A TREATMENT FOR ALZHEIMER’S DISEASE By Aaron Hirko August 2006 Chair: Jeffrey Hughes Major Department: Pharmacy Alzheimer’s disease (AD) is a progressive neurodegenerative disorder affecting memory, thinking, behavior, and emotion. It is characterized by a progressive accumulation of extracellular amyloid plaques and intracellular neurof ibrillary tangles. Evidence suggests that the deposition of amyloi d triggers a cascade that ultimately leads to Alzheimer’s pathology, making amyloid a pr omising target for the treatment of AD. Amyloid plaques are composed mainly of the 4.5 kD peptide fragment amyloid One strategy targeting A is to deliver an A binding agent outside the brain, creating a peripheral sink that causes efflux of A across the blood-brain barrier. One such agent is the 89 kD protein plasma gelsolin. However, administering such a large compound poses formidable formulations challenges, and prot eins generally have poor pharmacokinetic properties. Taking a gene-therapy approach by delivering a DNA vector coding for plasma gelsolin offers an alte rnative to repeated inje ctions of protein.

PAGE 12

xii We developed a plasmid vector for human plasma gelsolin. We determined that plasma gelsolin may have enzymatic-like functions toward A shifting the equilibrium from fibrillization and deposition to sol ubilization and elimination. We obtained expression of our plasmid vector for plasma ge lsolin in the periphery of 2 different mouse models of Alzheimer’s, and showed that it re sults in a significant reduction in the amount of A in the brain. We also showed that this reduction of A in the brain may occur along with an increase in microglia activity. These results show the validity of using plasma gelsolin as a peripheral gene therapy of Alzheimer’s disease.

PAGE 13

1 CHAPTER 1 INTRODUCTION Alzheimer’s disease (AD) is the most co mmon form of dementia. In the next 45 years, the number of Amer icans afflicted with AD is e xpected to quadruple: from about 2.5 million cases today, to nearly 10 million in 2050 (Sloane et al., 2002). This increased prevalence of AD can be attributed to aging of the population: in the year 2000, 5.9% of the population was over age 75; in 2050, this is expected to be 11.4% (Kawas and Brookmeyer, 2001). Age is the single strongest risk factor for AD. It afflicts 10% of people over 65, and almost 50% of pe ople older than 85 (Evans et al., 1989). Alzheimer’s disease has a huge impact on a pe rson’s quality of life (e.g., memory loss, impaired activities of daily liv ing, depression, and behavioral disturbances) (Sloane et al., 2002). As the disease progresses, independen ce decreases, placing an increased financial and psychological burden on family caregivers. Koppel estimated that in 2001, the total economic impact of AD in the US was between $183 and 207 billion (2002). With the expected increased prevalence of AD, finding efficacious treatments will be critical to ease the social burden of this disease. Dementia In Latin “dementia” is defined as irra tionality. Medically speaking, dementia describes a collection of symptoms that robs an individual of his/her cognitive functions, resulting in the loss of the abil ity to carry out normal daily activities eventually requiring the full-time care of family or professionals These symptoms can be caused by a number of different diseases that affect the brain. Typically the diagnosis of dementia requires

PAGE 14

2 significant deficits in at leas t two or more brain functions, such as reasoning, judgment, perception, language, and memory. Discovery of Alzheimer’s disease In March 1901 the husband of a 50-yea r-old woman (Auguste D) noticed a paranoid symptomalogy in his wife, which rapi dly progressed to include sleep disorders, aggressiveness, crying, confus ion, and disturbances of memory. By November, the deteriorating mental state of Auguste D forced her husband to admit her for inpatient treatment at the Community Psychiatric Hospital at Frankfurt. A senior assistant at the hospital, Dr. Alois Alzheimer, thoroughly documented the progression of August D’s symptoms. On her death on April 8, 1906, Alzh eimer was able to examine her brain both histologically and morphologically. His exam ination showed that Auguste D’s brain was atrophied, and included histological abnormalitie s later considered the hallmarks of AD, known as neurofibrillary tangles (NFT) and senile plaques. He described these findings (along with their relationship to more than 4 years of clinical observations) at the 37th meeting of South-West German Psychiat rists in Tbingen on November 3, 1906. Although Alzheimer’s lecture was not well r eceived, his observation would be later recognized as the first demonstrated relati onship between clinical history of specific cognitive changes and neurolog ical lesions at autopsy. After reports of the case of Auguste D, a number of other patients with similar ailments were described (Moller and Graeber, 1998). The term Alzheimer’s disease was first coined to describe the condition by a colleague of Alzheimer in Munich, Dr. Emil Kraepelin (1910). Since then Alzheimer’s disease has been recognized as the most common form of dementia worldwide.

PAGE 15

3 Current Therapies To date, the only FDA-approved treatmen ts for AD are acetylcholinesterase inhibitors (tacrine, donepezil, rivastigmine, and galantam ine) and an NMDA antagonist (memantine). The aim of using acetylcholineste rase inhibitors is to enhance selective cholinergic transmission in the brain by decr easing the catabolism of acetylcholine. Basal forebrain cholinergic neurons, critical for memory and learning, are diminished in AD, resulting in a reduction of choline acetyl transferase and acetylcholine (Coyle et al., 1983; Terry and Katzman, 1983). Increasing the levels of acetylcholine can help ameliorate deficits in memory and learni ng (Weinstock, 1995). The aim of using NMDA antagonist is to block the effects of elevat ed levels of glutamate which may lead to neuronal dysfunction (Mattson et al., 1992). Both treatment strategies have shown modest improvements in maintaining indepe ndence, function, and decreasing cost to society (Trinh et al., 2003; Wimo et al., 2003) However, these modest improvements are far from ideal and only delay the onset of the inevitable dependency of care by a short period of time. These treatments only addres s biochemical symptoms of AD rather than preventing progression of the underlying cause. Inherited Alzheimer’s It was not until 75 years after Alzheimer de scribed the case of August D that the main constituent of the senile plaques was biochemically identified. Allsop and coworkers (1983) identified that these pla ques consisted mainly of a 40-42 amino acid peptide named amyloid This peptide was later discovered to originate from a larger precursor given the name amyloid pr ecursor protein (APP) (Kang et al., 1987). The subsequent discovery that a single mi ssense mutation on the APP gene corresponded with an inheritable form of Alzheimer’s disease (Goate et al., 1991) led to the

PAGE 16

4 formulation of the amyloid cascade hypothesi s, which states the underlying cause of AD is the result of an increased c oncentration and deposition of A (Hardy and Allsop, 1991). This hypothesis was bolster ed by the later discovery th at every inheritable form of AD results from mutations i nvolved in the processing of A from APP: either on APP itself, Presenilin-1, or Presenilin-2 (Cla rk and Goate, 1993; Levy-Lahad et al., 1995; Sherrington et al., 1996). A lthough inherited forms of AD co mprise fewer than 10% of all Alzheimer’s cases, every inheritable form involves the facilitation of the oligomerization and la ter precipitation of A Amyloid Precursor Protein In humans, amyloid precursor protein is a large transmembrane glycoprotein that exists as three major isofor ms (APP695, APP751, and APP770) that are all the result of alternative processing of premRNA generated from the APP gene on Chromosome 21. The function of APP is poorly understood. However, evidence suggests it may have cell adhesive, intrace llular communication, membrane to nucleus communication, neurotrophic, or neuroprolifer ative activity (Turner et al ., 2003). Even though APP may play a role in many biological functions, compensatory mechanisms allow for the viability of APP knockout mice (Zheng et al ., 1996). These mice show reductions in body weight and synaptic transmission, impair ed locomotor activity and grip strength, and a hypersensitivity to epileptic seizures and forebrain commissural defects (Zheng et al., 1995). Three proteolytic cleavage sites have been identified on APP: two near the plasma membrane on the extracellular side, and one w ithin the plasma membrane. The protease complexes responsible for the cleavage are known as and secretase. Cleavage by

PAGE 17

5 secretase and secretase releases large aminoterminal fragments known as APPsand APPsrespectively. These fragments differ in size by 17 amino acids at the carboxy-terminus of the fragments (APPsis larger than APPs. The remaining fragments of APP stay anchored to the plas ma membrane, and are referred to as C99 for the secretase product and C83 for the secretase product. Both C83 and C99 are substrates for secretase which cleaves within the plasma membrane. Amyloid (A ) is formed after C99 is cleaved by secretase. This cleavage us ually results with a 40 amino acid length peptide A 1-40. However, secretase cleavage of C99 can also result in the more hydrophobic 42 amino acid peptide product A 1-42. The A 1-42 is more prone to oligomerzation and fibril formation than A 1-40 (Hasegawa et al., 1999). The inheritable forms of AD invariably increase the relative amounts of A 1-42 as compared to A 1-40 (Scheuner et al., 1996; Sinha and Lieberburg, 1999; Su zuki et al., 1994; Tamaoka et al., 1994). Amyloid is first released from neurons as a soluble monomer which has an helical secondary structure. During the process of oligomeri zation this undergoes a series of conformational changes to form cross -sheet structures as oligomers. These soluble oligomeric forms of A may have a protofibular-like stru cture (Lashuel et al., 2002) or an amorphous micellular-like structure (Hoshi et al., 2003). In the fibrillar model for amyloid deposition the soluble oligomers begin to aggregate forming first protofibrils, and then fibrils that finally come togeth er to make up the plaques (Figure 1-1). Amyloid Cascade Hypothesis One of the central controversies in th e AD research community is whether A is the cause or result of the pathoge nic process. The hypothesis that A is central to the

PAGE 18

6 pathogenesis of AD is known as the am yloid cascade hypothesis (Selkoe, 1989, 1990). This hypothesis states that something cause s either an overproduction or a decreased clearance of A The increased levels of A result in the formation of oligomeric forms of A These then aggregate and deposit as plaques. This deposition of plaques causes the activation of microglia and astrocytes, resultin g in the release of pro-inflammatory cytokines and reactive oxyge n species (Akama et al., 1998; Hoozemans et al., 2005; Johnstone et al., 1999). Together microgl ia activation, astrocytic activation, and oligomeric A can all cause synaptic and neuriti c injury, including neurofibrillary tangles, which then lead to dementia. Evidence Supporting the Amyloid Cascade Hypothesis The main evidence supporting the amyloid cas cade hypothesis is that every form of familial Alzheimer’s disease (FAD) involve s mutations on either APP itself or the enzymes that cleave APP, resulting in an overproduction of A The presence of an extra copy of chromosome 21, in which the gene for APP is located, is found in Down’s syndrome; this inevitably lead s to an early onset of Alzh eimer’s-like pathology. In the more common sporadic form of AD, the presen ce of the apolipoprotein E4 (apoE4) allele is considered a risk factor for the dis ease (Corder et al., 199 3); evidence suggests that apoE is involved with the clearance of A (Brendza et al., 2002). Further evidence supporting the amyloid cascade hyp othesis is the fact that, in vitro, A itself has been found to be neurotoxic (Dahlgren et al., 2002; Pike et al., 1991). Critics of the Amyloid Cascade Hypothesis Critics of the amyloid cascade hypothesis argue that A accumulation may occur secondary to other pathological events and act ually play a role in neuroprotection. They

PAGE 19

7 point out that although specific forms of A can be toxic in vitro ; this toxicity is less reliable in animal models, providing the argument that the in vitro toxicity may be an artifact. They make the point that amyloid deposition is poorly corr elated with cognitive deficits, and that neurofibri llary tangles (NFT) and neur on number are much better indicators of cognitive decline (Giannakopoulos et al., 2003). It is true that when A deposits in humans are meas ured histochemically, they do not correlate well with cognitive decline. However, soluble forms of A measured biochemically from brain extracts correlates very well with synaptic density and can be used to discriminate between AD patients and non-AD controls that do have a high degree of amyloid deposit pat hology (Lue et al., 1999; Nasl und et al., 2000). Total A 40 and 42 levels of nursing home resident brai n extracts measured biochemically has also been correlated to cognitive decline as m easured by the Clinical Dementia Rating (CDR) scale (Naslund et al., 2000). There is evidence that A may play a role as an antioxi dant (Curtain et al., 2001) or a neurotrophin (Yankner et al ., 1990). In fact, Lopez-To lendano and Shelanski (2004) recently found that A was neurogenic in a dose-depende nt manner when treating neural stem cells. However, Liu et al (2004) demonstrated that A ’s toxicity in primary neuronal cultures was dependent on the expres sion of the microtubule-associated protein tau, cyclin-dependent kinase 5 (Cdk5), a nd the cell’s state of differentiation. This is an interesting ob servation, providing evidence for a mechanism of how A can cause NFTs. Hyperphosphorylated tau is the main component of NFTs and Cdk5 is one of the enzymes thought responsible for tau’s phosphorylation (C ruz and Tsai, 2004; Noble et al., 2003). Although both sides can make strong arguments supporting their

PAGE 20

8 point of view, it is unlikely that there will be any consensus over what role A plays in the progression of AD until human trials are completed that target A Elan and Wyeth Trial The observations that active immuniza tion in transgenic mice dramatically reduced the accumulation of A (Schenk et al., 1999), and showed protection against memory deficits (Janus et al., 2000; Morgan et al., 2000) led to the phase II trial undertaken by the pharmaceutical companies Elan and Wyeth in late 2001. They actively immunized patients against A hoping to trigger an immune response that would increase clearance of A from the CNS. These trials we re halted early because about 6% of the patients developed meningoencepha litis (Orgogozo et al., 2003). Follow-up studies on the participants i ndicated that antibody responders had significantly improved memory function as measured by the neurops ychological test ba ttery and decreased cerebral spinal fluid levels of ta u protein (Gilman et al., 2005). Sink-Hypothesis One proposed mechanism for how immunization works is that anti-A antibodies enter the CNS and stimulate microglial phagocytosis of A -antibody complexes. This has been demonstrated by Bacskai et. al. (2001). Another possible mechanism for how anti-A antibodies can clear plaques is by shifting the equilibrium of A from the CNS to the periphery. The so called “sink hypothesis” is supported by the fi nding that less than 0.1% of antibodies in the serum gain access acr oss the blood brain barrier (BBB) (Bard et al., 2000) and studies performed by se parate groups using different A binding agents administered peripherally(Deane et al., 2003; Matsuoka et al., 2003).

PAGE 21

9 For the sink-hypothesis to be valid, carrier or receptor mediated transport of A must occur, because the BBB normally prevents free exchange of polar solutes between blood and brain or brain and blood. The main transporter identified being responsible for transport out of the CNS is low-density li poprotein receptor-related protein-1 (LRP-1) (Shibata et al., 2000). LRP-1 is a large endocyt ic receptor responsible for the transport of apoE and cholesterol-containing lipoproteins. Likewise the receptor for advanced glycation end products (RAGE) has been identified as a membrane-bound receptor that transports A from the circulation into the CNS (Deane et al., 2003). RAGE is a multiligand receptor in the immunoglobin superfamily. Generally there is little expres sion of RAGE in most tissues. However the accumulation of RAGE ligands, such as A triggers RAGE expre ssion, in contrast to a decrease of LRP-1 expression seen in an A rich environment (Shibata et al., 2000). Exacerbating this effect, RAGE transport of A results in the increased expression of proinflammatory cytokines and endothelin -1 at the BBB causing decreased cerebral blood flow (Deane et al., 2003). Deane et al. (2003) demonstrated that when PD-hAPP mice were treated with intraperitoneal injections of a truncated soluble fo rm of RAGE (sRAGE), A transport into the CNS was interfered with and significant increases in plasma A levels along with a decrease in brain A levels and plaque loads was obs erved. Likewise Matsuoka et al. saw similar results when they treated PS/APP mice with the A binding agents GM1 and plasma gelsolin (Matsuoka et al ., 2003; Morgan et al., 2000).

PAGE 22

10 Gelsolin Plasma gelsolin is a highly conserved 93 kD actin-binding protein, also known as brevin or actin-depolymerizing factor, normally found in the plasma at concentrations of about 179 mg/L (Chauhan et al., 1999). Its ma in function is thought to be part of the actin-scavenging system, to prot ect the microcirculation from the effects of long F-actin polymers released during cell death (Lee and Galbraith, 1992). However recent evidence points to the possibility that plasma gelsol in may play a variety of roles in the body, including mediating inflammatory responses by binding to pro-inflammatory compounds (Bucki et al., 2005; Bucki et al., 2004; Chauhan et al., 1999; Lind and Janmey, 1984; Smith et al., 1987), or by altering cell motility and endocytosis(Witke et al., 2001). There are three known forms of gelsolin [cytoplasmic (Yin and Stossel, 1979), plasma (Nodes et al., 1987), and gelso lin-3 (Vouyiouklis and Brophy, 1997)] all coded for by the same gene, resulting from altern ative post-transcriptional processing. The cytoplasmic form of gelsolin was first describe d as a factor able to solubilize gels formed by macrophage extracts, hence the name gel sol -in (Yin and Stossel, 1979). Plasma gelsolin differs from the other two by the pres ence of an N-terminal 23 amino acid signal peptide, which causes plasma gelsolin to be secreted outside the cell producing it. Gelsolin is regulated by polyphosphoinositide and Ca+2. Gelsolin severs and caps F-actin in response to Ca+2, and phosphoinositides block th e capping function (Bucki et al., 2004; Kwiatkowski, 1999). It alters cell shape by remodeling actin filaments and is involved with cell motility (Cooper et al., 1987; Janmey et al., 1987; McLaughlin et al., 1993). Data from gelsolin knockout mice indicat es gelsolin is necessary for rapid motile responses in cell types invol ved in responding to stress such as hemostasis, wound healing, and inflammation (Witke et al., 1995) A mutated form of gelsolin (either

PAGE 23

11 D187N or D187Y) results in aberrant pr oteolytic cleavage by furin causing a 68kD gelsolin fragment to be secreted and then deposited as amyloid in the Finish type of amyloidosis (Chen et al., 2001; Kazmirski et al., 2000). Gelsolin has been shown to be an effector of apoptosis through its interaction with the cysteinyl-protease caspase3. A gelsolin cleavage frag ment of caspace-3 has been shown to cause numerous cell types to “round up, detach from the plate, and undergo nuclear fragmentation”(Kothakot a et al., 1997), most likely th e result of the N-terminal gelsolin fragment’s ability to activate DNase-1 (Chhabra et al., 2005). Gelsolin has also been shown to be prot ective against excitotoxic induced apoptosis by altering the actin cytoskeleton in response to Ca+2 influx, preventing the reduction of the mitochondrial permeability transition pore opening and membrane potential loss, and preventing caspase-3 activation (Harms et al., 2004). Plasma gelsolin has been shown to pr otect against inflammatory reactions associated with injury (Christofidou-Solom idou et al., 2002a; Rothe nbach et al., 2004). Vasconcellos et al. showed that plasma gelso lin reduced the viscosity of cystic fibrosis sputum (1994). In fact Biogen Inc. evalua ted recombinant plasma gelsolin in phase 2 clinical trials as a treatment for cystic fibrosis. Chauhan et al. showed that human plasma gelsolin binds to A prevents fibrillization, and disa ssembles preformed A fibrils, suggesting a possible role for gelsolin in the clearance of amyloid (Chauhan et al., 1999; Ray et al., 2000). Plasma gelsolin is found in the cerebral spinal flui d (CSF), and produced in the choroid plexus (Matsumoto et al., 2003). At the 2004 International Conference on Alzheimer’s and Related Disorders Chauhan et al. reported that gelsolin levels in AD patients’ CSF was

PAGE 24

12 significantly reduced as compared to nondement ed age matched controls, suggesting that decreased gelsolin may play a role in the increased amyloid content seen in AD (2004). Matsuoka et al. showed injections with bovine plasma gelsolin can prevent deposition of gelsolin in younger huAPP K670N,M671L/ PS-1 M146L (2003). The focus of our study will be on the effects plasma gelsolin gene expre ssion has on amyloid deposition in transgenic mouse models of Alzheimer’s disease.

PAGE 25

13 Figure 1-1. Amyloid precurso r protein processing. cut site COOH NH2 cut site secretase NH2 COOH APPssecretase APPsNH2 COOH COOH NH2 C99 COOH NH2 C83 cut sites secretase NH2 COOH Amyloid helix) oligomerization Soluble Amyloid Oligomers ( sheet) fibrillization -COOHPlaque formation

PAGE 26

14 CHAPTER 2 MATERIALS AND METHODS Reagents Except where noted, all chemicals used we re purchased from Fisher Scientific (Hampton, NH). Molecular biology reagents and enzymes were purchased from New England Biolabs (Ipswich, MA). Amyloid pe ptides were purchased from Anaspec (San Jose, CA). ThePCR primers were ordered from Sigma Genosys (The Woodlands, TX). The ELISA kits were purchased from Biosource (Camarillo, CA). Precast polyacrilamide gels and PVDF membranes were purchased from Bio-Rad Laboratories (Hercules, CA) Subcloning Vectors An expression plasmid for plasma gelsoli n, pPGLE (Figure 2-1A), that is based on the commercially available plasmid (pCD M8) was kindly provided by Dr. Hisakazu Fujita (Kwiatkowski et al., 1989). The coding sequence for plasma gelsolin was removed from the pCDM8 backbone by a HindIII and XbaI digest, followed by separation on a 1% agarose gel, and purification us ing a Qiagen gel purification k it. Blunt ends were made by treating with T4 DNA polymerase in the pr esence of dNTPs. HindIII linkers were ligated to the blunted ends using T4 DNA ligase. Samples were then run on a 1% agarose gel. The blurred band containing the gelsolin insert, along with different amounts of linkers, was purified from the gel us ing a Qiagen gel purif ication kit. This purified band was then subject to a HindIII digest.

PAGE 27

15 A plasmid backbone containing the cytome galovirus/chicken be ta-actin hybrid (CBA) promoter and the woodchuck hepatitis virus post-transcriptional regulatory element (WPRE) was prepared from our pGFP vector (Figure 2-1B)((Klein et al., 2002) by excising the Green Fluorescent Protein (G FP) coding sequence with a HindIII digest followed by separation on a 1% agarose gel; and was then purified using a Qiagen gel purification kit. The 5’p hosphate groups were removed with calf intestine alkaline phosphatase in order to pr event self-ligation. The purified backbone along with the plasma gelsolin insert were ligated together overnight with T4 DNA ligase. Electrocompetent SURE cells (Stratagene, Garden Grove, CA) were transformed with the re sultant ligated product, using a Bio-Rad electroporator, with the resistance set at 400 the capacitance at 25 F, and the voltage at 2.2kv. Transformed bacteria were grown for an hour in 1mL of NZY broth at 37C, followed by plating on NZY agar plates contai ning ampicillin (50mg/L), and then grown overnight at 37C. Several colonies were selected for screening, and each was grown overnight in 5 mL of ampicillin-containing NZY broth. Plasmids were purified from the cultures, using Qiagen mini plasmid prep kits Plasmids were then subjected to a BglII digest to confirm orientation of the insert (Figure 2-2). A clone (W16) with the forward insert was given the name pUFGL (Figure 2-1C). Large Scale Plasmid Preparation For large-scale preps, plasmids were pr opagated overnight in 5 mL of ampicillin containing NZY broth. This was used to i noculate 2 L of ampicillin containing NZY broth, and then grown overnight again. Th e cultures were pelleted by centrifugation. They were then resuspended in a lysozyme buffer (80 mL/L of cultu re) and treated with lysozyme (2 mg/mL, Sigma, St. Louis, MO). Next the cultures were subject to alkaline

PAGE 28

16 lysis by adding a 1% SDS 0.2N NaOH solution at 196 mL/L of culture. The mixture was neutralized with a 3M NaAc (pH4.8) 0.6% ch loroform solution (144 mL/L of culture). The resultant chromosomal and protein precipit ates were separated by centrifugation. Plasmid DNA in the supernatant was then precipitated by bringing the solution to 10% polyetheleneglycol (PEG). The precipitates were separated by centrifugation, and then resuspended in distilled water (40 mL/L of culture). The RNA was precipitated and separated by adding 5.5 M LiCl (40 mL/L of culture), followed by centrifugation. Plasmid DNA was precipitated from the s upernatant by bringing the solution to 36% isopropanol. Precip itated plasmid DNA was resuspended in a 5.3M CsCl solution containing 1 mM ethidium bromide. The resultant solution was centrifuged in a Beckman 70.1 Ti rotor at 55,000 rpm for 19 hours. The lower band containing the plasmid DNA was removed using an 18 g syri nge. Ethidium bromide was removed by performing four extractions with isoamyl alcohol. The plasmid DNA was precipitated from the aqueous layer by bringing the solution to a 40% Ethanol concentration. The plasmid precipitate was pelleted by centrifugation, and resuspended in TE (10 mM Tris-HCl 1 mM EDTA pH 8.0). Any residual protein contamination was remove d from this soluti on by performing four extractions with phenol-chloroform, followed by one extraction with chloroform alone. Plasmid DNA was precipitated once again by adding one tenth volume of 3 M NaAC pH 4.8 and 2.5 volumes of 100% ethanol. Prec ipitates were pelleted by centrifugation, followed by washing excess salt with a 75% ethanol solution. The plasmids were resuspended in sterile TE buffer. Con centration and purity of the samples were determined by UV absorbance at 260/280 nm.

PAGE 29

17 Cell Culture Human embryonic kidney (HEK) 293 cells were cultured in DMEM with 10% fetal bovine serum (FBS), 100 units/mL penicillin and 100 g/mL streptomycin (Gibco, Invitrogen, CA) in a 5% CO2 incubator at 37C. Cells were grown on 10cm dishes to 50% and 80% confluency. One half hour before transfection using the CaPO4 precipitation method, culture media was repl aced with fresh media. To prepare transfection complexes 20 g of either pUFGL or pGFP in 700 l of 250 mM CaCl2 solution was added in a dropwise fashion to 700 l 2X HEPES Buffered Saline while vortexing slowly (for 2X HB S, 300 mM NaCl, 1.8 mM Na2HPO4, 11 mM dextrose, and 40 mM HEPES, pH 7.12). The transfection solutions were then in cubated at room temperature for 20 minutes. Following incubation, th e transfection solutions were mixed gently then added to cell culture dishes in a drop wise fash ion. Culture dishes were incubated with transfection media for 12 hours then replaced with fresh media. Forty-eight hours later, media was collected, a proteinase inhibitor cocktail was added (Sigma P8340), and it was either used for immunoprecipitation, or con centrated ten fold with Centricon 50,000 nmw cutoff centrifugal filters for western blot analysis. Animals and Procedures All procedures were done w ith prior approval and oversig ht of the University of Florida’s institutional animal use and care committee. Double transgenic mice expressing both Swedish mutant mouse /human APP695 K594N,M595L and exon 9 deleted mutant presenilin-1 (mo/huAPP /PS1 E9) were supplied by Jackson Labo ratories (Bar Harbor, ME) (Jankowsky et al., 2004). Double transgen ic mice expressing both Swedish mutant

PAGE 30

18 human APPK670N,M671L (huAPP K670N,M671L, Tg2576)(Hsiao et al., 1996) and mutant presenilin-1 M146L (PS-1 M146L)(Duff et al., 1996), and tran sgenic mice expressing only mutant PS-1 M146L were supplied by The Nathan Kline Institute (NY,NY). One litter of mo/huAPP /PS1 E9 mice was aged until 36 weeks, at which time the mice were either injected with our test plas mid pUFGL (n=3) or left untreated (n=3). Mice expressing huAPP K670N,M671L/ PS-1 M146L were aged until 32 weeks, at which time they were either injected with plasmid DNA, pUFGL (n=5), pGFP (n=3), or left untreated (n=7). PS-1 mice used for western blot detection of plasma gelsolin were injected at the age of 36 weeks with either pUFGL (n=2), pGFP (n=2), or left untreated (n=2). For injections, plasmid DNA was diluted in lactated Ringer’s solution to a concentration such that there was 25 g of plasmid DNA/10% of body weight volume of ringers (e.g. a 30 g mouse received a 3 mL in jection). DNA solutions were warmed to a temperature of 37C. Animals were warmed briefly under a heat lamp, mildly anesthetized with isoflourane, and restrained in a custom-made harne ss. A three milliliter syringe was used with a 27 g half inch needle The injection solution was injected in a time period of 5-10 seconds. Animals were then recovered on a heating pad, and returned to their cages. Blood samples were taken from PS-1 mice from the retro-orbital sinus using heparinized capillary tubes. Animals were mild ly anesthesized with isoflourane, and then the capillary tube was used to puncture the retro-orbital sinus The tube was allowed to fill with blood, and then plugged on one e nd with clay. Plasma was separated

PAGE 31

19 immediately by centrifugation, snap frozen in liquid nitrogen, and stor ed at -80 C for analysis by western blot. Two and one half weeks following injections, double transgenic mice were sacrificed along with age-matched untreated co ntrols. Animals were deeply anesthetized with isoflourane and perfused with PBS. Li vers and brains were excised, hemi-brains and a sample of liver were snap frozen in liqui d nitrogen and stored at -80 C for analysis by ELISA and RT-PCR. The remaining hemi-brain and liver tissue was fixed for 48 hours in a 4% paraformaldehyde in PBS solutio n, and then equilibrated in 30% sucrose in PBS solution for cryoprotection. Immunoprecipitation and Western Blot An immunoprecipitation kit from Sigma (IP-50, St. Louis, MO) was used for the immunoprecipitation reactions. 600 L of media from cell culture s with or without being spiked with A 1-42 to 5 M along with 2 l anti-Gelsolin monoclonal antibody clone GS-2C4 (Sigma G-4896, St. Louis, MO) was a dded to the spin columns provided with the kit and incubated overnight at 4 C. 30 L/column of protein-G agarose was washed in 1X IP buffer, then resuspended in 50 L of 1X IP buffer and added to the samples in the columns. These samples were then incubate d overnight at 4C. The tips were broken from the columns, then the columns were centrifuged and the effl uent was discarded. The beads in the columns were washed five times with 1X IP buffer, followed by a sixth wash in 0.1X IP buffer. 50 L of Laemmli sample buffer containing 5% mercaptoethanol was then added to the bead s and incubated at 95 C for ten minutes. Samples were spun through the columns and then 50 l loaded onto a precast SDS 4-20% PAGE Tris-HCl gel.

PAGE 32

20 For Western blots of concentrated media 25 L of sample was mixed with 25 L of 2X Laemmli sample buffer with 5% -mercaptoethanol, boiled for 5 minutes, and then loaded onto a precast SDS 10% PAGE Tris-HCl ge l. For western blot analysis of plasma samples 4 L of plasma was diluted in 21 L of distilled water and then mixed with 25 L of 2X Laemmli sample buffer with 5% -mercaptoethanol, boiled for 5 minutes, and then loaded onto a precast SDS 10% PAGE Tris-H Cl gel. Gels were run using a Biorad power supply set at 100 V for one hour. Separated proteins were then transferre d to a PVDF membrane at 75V for 2 hours on ice. Membranes were incubated overnight in a blocking soluti on (5% nonfat dry milk and 0.05% Tween 20 in PBS) at 4C. Prim ary antibodies were then added [for anti-A 6E10 from Chemicon(Temecula, CA) was used at 1:1000 dilution, for anti-gelsolin GS2C4 from Sigma was used at 1:1000 dilution] and incubated at room temperature for 2 hours. Membranes were washed three times in PBS with 0.05% Tween 20, and incubated with horseradish peroxida se (HRP)-conjugate d anti-mouse antibody [Amersham (Piscataway, NJ) at 1:5000] in blocking so lution for 1 hr at room temperature. Following three more washes they were incubated with substrate [electrochemiluminescence (ECL), Amersham (Piscataway, NJ)] for 1 min and exposed (Kodak, Rochester, NY). Detection of Message A Qiagen (Valencia, CA) RNeasy mini-kit was used for RNA extractions. 30 mg of frozen liver or brain tissue was homogenized on ice using a Polytron homogenizer (Brinkmann Instruments, Westbury, NY) in 600 mL of buffer RLT containing mercaptoethanol. RNA was purified and wa shed using the columns and reagents

PAGE 33

21 provided by the Qiagen kit as recommended by the manufacturer. RNA was eluted from the column to a final volume of 60 L in RNase-free water provided with the kit. Primers were designed to yield ~900 b.p. product from mRNA transcribed from our vector pUFGL or ~1800b.p. product from unprocessed RNA or DNA contamination, by having the forward primer (GGC TCT GAC TGA CCG CGT TTA C, Tm = 68.7C) anneal to sequence from Exon 1 in the vect or, and reverse primer (CTG TTG GAA CCA CAC CAC TGG, Tm = 67.7C) a nneal to sequence from the coding region of gelsolin. Primers for -Actin (ATG AGG TAG TCT GTC AGG T, Tm = 52.9C, & ATG GAT GAC GAT ATC GCT G, Tm = 52.7C) were used as a positive control. A Qiagen one-step RT-PCR kit was used for the RT-PCR reaction. One microliter of RNA was used in each 25 L reaction with final prim er concentrations of 0.6 M, and Q solution was included in the mixture. A MJ Research PTC-200 Peltier Thermal Cycler was used for the RT-PCR reaction. Reverse transcription was done for 30 minutes at 50C, followed by PCR activation at 95C for 15 minutes. Next came thirty cycles that consisted of: denaturation for one minute at 94C, annealing for 30 seconds at 50C, and extension for two minutes at 72C. There was one final extension for 10 minutes at 72C, and then samples were held at 4C. Samples were then loaded onto a 2% agarose gel and run at 85 volts. Bands were th en imaged by ethidium bromide staining. Enzyme Linked Immunosorbent Assay For enzyme linked immunosorbent as say (ELISA) Biosource colorimetric immunoassay kits were used for both amyloid 1-40 and 1-42. Frozen hemi-brains were weighed. Eight times the mass of 5 M guani dine HCl, 50 mM Tris HCl, pH 8.0 was added to the brains then homogenized using a Poly tron homogenizer (Brinkmann

PAGE 34

22 Instruments). The homogenate was mixed at room temperature for four hours, and then aliquotted and stored at -80C. The guani dine extracted homoge nates were diluted 3000 times in BSAT-DPBS [Dulbecco’s phos phate buffered saline with 5% bovine serum albumin, 0.03% Tween-20 and 1 mM 4-(2-Aminoethyl)benzenesulfonylfluoride (AEBSF)]. Mixtures were centrifuged at 16,000 G and 4C for twenty minutes. The supernatants were diluted f our fold in the standard diluent buffer provided with the Biosource kit, with AEBSF at 1mM. Samples were then incubated on a shaking platform at room temperature for two hours in the well s provided with the Bi osource kit with an equal volume of primary antibody solution. Sa mples were then washed four times, and incubated in HRP solution four one half hour Samples were washed four times again and then incubated for a half hour with HRP substrate, in a box to protect the samples from light. Stop solution was then added and absorbance at 450 nm was measured using a Dynex Technologies MRX microplate r eader. Concentrations of Amyloid (1-42) were determined from standards provided with the kit. Histochemistry Coronal sections (50 m thick) were cut from the hemi-brains on a sliding microtome with a freezing stag e. Four sections, six secti ons apart each, were mounted on slides for thioflavine S staining. Sections were allowed to dry on the slides for 15 minutes. The slides were then placed in de ionized water for five minutes. They were then placed in filtered Mayer’s Hematoxylin for five minutes. Next, the slides were rinsed under running tap water for five mi nutes, followed by a five minute rinse in deionized water. The slides were then placed in a 1% thioflavine S solution (in dH2O, filtered, Sigma) for five minutes. The slides were differentiated in 70% ethanol for five

PAGE 35

23 minutes, given short rinses in deionized wate r followed by PBS, and cover slipped with glycerol gelatin (Sigma). Immunohistochemistry Antigen detection on 50 m thick coronal hemibrain sections was conducted on free-floating sections by inc ubating the sections overnight at 4C in blocking solution (3% goat serum, 0.3% Triton X-100, 0.05% azide in PBS) Endogenous peroxidase activity was quenched by incubating the sections for 10 minutes in 0.5% H2O2 in PBS at room temperature prior to blocking. Primary antibodies used were: 6E 10 (1:1000, Signet, Dedham, MA), OX-42 (1:200, Serotec, Raleigh, NC), and anti-amyl oid oligomer (1:250 Chemicon). Sections were incubated with primary antibodies diluted in blocking solution at 4C for three days. Sections were then washed with PBS thr ee times for 5 minutes each wash. Then the sections were incubated overn ight at 4C with secondary antibody (biotinylated antimouse IgG or biotinylated anti-rabbit IgG, 1:1000, Dako, CA) diluted in blocking solution. Sections were again washed three times in PBS. Next the sections were incubated for two hours at room temperature in PBS with ExtrAvidin peroxidase (HRP) conjugate (1:1000, Sigma). Washing was pe rformed again, and then development of tissue labeled with HRP was performed with a solution of 0.67 mg/mL diaminobenzidine (DAB, Sigma) and 0.13 L of 30% H2O2 per mL of 80 mM sodium acetate buffer containing 8 mM imid azole and 2% NiSO4. The sections were mounted on Superfrost pl us microscopic slides (Fisher, NH), air dried and dehydrated by passing through wate r, followed by 70%, then 95%, and 2

PAGE 36

24 changes of 100% ethanol. Then they we re passed through two changes of xylene and coverslipped with Eukitt (C alibrated Instruments, NY). For immunofluorescence Alexa Flour 555 goat anti-rabbit IgG (H+L) (1:2000, Molecular Probes, OR), Alexa Flour 488 goat anti-mouse IgG (H+L) (1:1000, Molecular Probes), AMCA conjugated F(ab') 2 fragment goat anti-mouse IgG, F(ab')2 fragment specific (1:100, Jackson Imm unoResearch Laboratories, PA) secondary antibodies were used diluted in bloc king solution following the primary antibody incubation. Nuclear counter st aining was performed by incuba ting the sections in 4’,6diamidino-2-phenylindole (DAPI, 1 g/mL, Sigma) for 15 minutes at room temperature. Fluorescent slides were cover slipped with glycerol gelatin mounting medium (Sigma). Image Analysis For percent amyloid burden measurements (both dense cored and diffuse) sections were analyzed in a blinded manner using the NIH Image J software. Regions of interest (ROI) were created encompassing both th e hippocampus and neocortex of digital micrographs of each stained section. The ROI’s area was measured in pixels2. The number of plaques staine d, plaque sizes (in pixels2), and total stained areas in the hippocampus and cortex (in pixels2) were determined by th resholding segmentation. Total stained areas were divided by total area, and then multiplied by 100% to give the percent amyloid burden.

PAGE 37

25 Figure 2-1 Vectors, pPGLE kindly provided by Dr. Hisakazu Fujita (A), The coding sequence for plasma gelsolin was excised and then inserted into the CBA promoter and WPRE containing ba ckbone from the control plasmid pGFP (B) to make our test plasmid pUFGL (C). Figure 2-2 Restriction digest to confirm orientation of clones. W8, W9, and W12 are all clones representing antisense orie ntation, having bands of 3173, 3013, 1137, & 1088 base pairs long. W16 is a sense clone given the name pUFGL having bands of 3173, 2103, 1998, & 1137 base pairs long. f1(+) origin pPGLE (6515bp) CBA promoter TR CMV enhancer Exon 1 Intron WPRE Poly A ColE1 ori ApR f1(+) origin TR GFP pGFP (6531bp) pUFGL (8468bp) CBA promoter TR CMV enhancer Exon 1 Intron WPRE Poly A ColE1 ori ApR TR Plasma Gelsolin M13 Ori CMV promoter T7 Pro Plasma Gelsolin Splice & poly A SV 40 Ori A B C

PAGE 38

26 CHAPTER 3 THE INTERACTION OF PLAMA GELSOLIN AND AMYLOID Introduction Studies by Chauhan et al. showed that hu man plasma gelsolin binds to amyloid prevents fibrillization, and di sassembles preformed amyloid fibrils, suggesting a possible role for gelsolin in the clearance of amyloid (Chauhan et al., 1999; Ray et al., 2000). Chauhan used a solid phase binding assay to measure the dissociation rate constants (Kd) for two binding sites on human gelsolin, and found them to be 1.38 and 2.55 M. Matsuoka et al. showed injections with bovine plasma gelsolin can prevent deposition of amyloid in young huAPP K670N,M671L/ PS-1 M146L (2003). Our hypothesis for how gelsolin prevents amyloid deposition was based on its ability to bind amyloid in the periphery shifting its e quilibrium from deposition in the CNS to clearance in the periphery. Chauhan’s measurement of human plasma gelsolin’s Kd indicates human plasma gelsolin does not have a very high affinity for amyloid On the other hand Matsouka’s use of bovine plasma gelsolin in mice showed encouraging results. The different species forms of gelsolin used in Chauhan’s studies and Matsuoka’s study triggered us to ask the question of whethe r bovine gelsolin has a higher affinity for amyloid than the human form of gelsolin. In this chapter we furt her characterize the interaction between amyloid and human and bovine plasma gelsolin.

PAGE 39

27 Experimental Methods and Results Measurement of 125I Labeled A (1-42) Binding to Recombinant Human Plasma Gelsolin An attempt to measure the binding affinity of 125I labeled A 1-42 to human plasma gelsolin was carried out using stan dard binding protocol. Labeled A at concentrations of 2.0, 5.0, and 10.0 M were incubated with or without 2 M recombinant human plasma gelsolin in PBS for one day. The following day the samples were run through Whatman GF/C filters that had been preincubated for 30 min with 0.5% polyethylenimine, followed by three washes with cold Krebs Ringer buffer (KRB; 118 mM NaCl, 5 mM KCl, 10 mM glucose, 1 mM MgCl2, 2.5 mM CaCl2, 20 mM HEPES; pH 7.5). The filters were counted for radioactivity, and it was found that the samples with plasma gelsolin ha d lower counts (Figure 3-1). Because the molecular weight of plasma gels olin is ~ twenty times that of amyloid we were expecting that plasma gelsolin w ould bind the glass filters and in the presence of amyloid more radioactivity would be detected and be representative of how much amyloid is binding to plasma gelsolin. However it is apparent from our results that a significant amount of amyloid itself binds to the glass fi lters and in the presence of plasma gelsolin less amyloid binds to the filters. This experimental approach was subsequently abandoned; however theses results are discussed further later. Measuring Binding of HiLyt e Fluor 488 Labeled Amyloid (1-42) to Bovine Plasma Gelsolin Using Fluorescence Anisotropy Another approach at measuring the binding of amyloid (1-42) to plasma gelsolin is to use the property of fluorescence anis otropy. The change of orientation of a population of fluorophores, from that of a sp ecific orientation, isotropy, to a random

PAGE 40

28 orientation, anisotropy can be measured by monitoring the rate at which fluorescence decays in a population of fluorophores when obs erved through polarized filters and is termed the fluorescence anisotropic decay. This change occurs via Brownian rotational diffusion. This property can be expressed us ing Equation 3-1 as the molecular diffusion coefficient (Dr) and is dependent on absolute temperat ure (T), the viscosity of the solution ( and the molecular volume (V); R is the gas constant (Weber, 1953). V RT Dr6 (3-1) Protein-protein interactions can be measured by fluorescence anisotropy by observing changes in the rotati onal molecular motion due to the increase in molecular volume when two or more species associ ate with each other. The fluorescence anisotropy can be expressed as a functi on of molecular volume using Equation 3-2 (Perrin’s equation) (Lakowicz, 2002). The value of anisotropy in the absence of rotational diffusion is defined as A0; and is the fluorescence lifetime of the fluorophores. As the molecular volume increases ( V ), as when two or more prot eins bind to one another, the numerator of Equation 3-2 will decrease, increasing the value of the anisotropy. V RT A A 10(3-2) Understanding the principles of fluorescen t anisotropy, we decided use the change in anisotropy to measure the bindi ng of fluorescently labeled amyloid (1-42) ( HiLyte Fluor 488 labeled amyloid (1-42) from Anaspec). We first used the monoclonal antibody 6E10 (Chemicon) as a positive cont rol for our binding study. The fluorescent amyloid concentration was kept between 547516 nM the fluorescence measurements through out the assay maintained values of 90-100 relative units (RU). The 6E10 was

PAGE 41

29 added stepwise with a starting concentrati on of 0 nM and a final concentration of 369 nM. Results show a clear sigmoidal relatio nship (Figure 3-2) indicating a saturatable binding of amyloid 42) to 6E10, validating our appr oach of using anisotropy to measure the binding of amyloid (1-42). The same experiment was then repeated with bovine plasma gelsolin (Sigma) substituted for the 6E10. The fluorescent amyloid concentration was kept between 547-526 nM and the plasma gelsolin was added stepwise with a star ting concentration of 0 nM and a final concentration of 423 nM. As the gelsolin was added, instead of rising, the anisotropy fell from about 0.060 to about 0.048, at the same time the fluorescence rose dramatically from 51 RU and then le veled off at about 210 RU(Figure 3-3). A hypothesis was formulated as a result of th e data: the anisotropy decreased with an increasing concentration of gelsolin was re presentative of gelsolin disassembling oligomeric forms of amyloid and the fluorescence increase is representative of more fluorophores interacting with the aqueous phase of the solution, rather than being tied up into hydrophobic areas of the oligomers. Using a sample from the experiment an attempt was made to separate the free amyloid from the amyloid bound to gelsolin, to determine if a Scatchard analysis was possible, using ultra-filtration using Centricon filters with a 30 kD nmw cutoff. Prior to the spin the volume was brought up to 1 mL by adding distilled water, making the concentration of gelsolin 418 nM. To control for A adsorbing to the filter material, the same amount of amyloid was added to 1 mL of water g making the final amyloid concentration 82 nM) with no gelsolin. The fluorescence of the samples was

PAGE 42

30 measured prior to centrifugation and afterwar ds the retentate and the flow through were measured (Table 3-1). After filtration the retentate was taken from the sample containing only amyloid and gelsolin was added in a stepwise fashion, bringing the concentrati on of gelsolin from 0-1758 nM. As the gelsolin was added the fluorescence increased from about 1 to 9, while the anisotropy declined from 0.363 to 0.118 (Figure 3-4). HiLyte Fluor 488 Labeled A (1-42) Fibril Formation The fibrillation of amyloid is thought to be a major ev ent in the pathology of AD. As A is formed it is soluble and has an -helical confirmation. Fibril formation involves a conformational change to a cross -pleated sheet structure, oligomerization, followed by aggregation. In vitro, both synthetic 1-40 and synthetic amyloid (1-42) at 100 M form fibrils spontaneously within 48 hours(Wegiel et al., 1996). In this experiment we monitor the fibril formation of HiLyte Fluor 488 labeled amyloid (1-42) as a decrease in fluorescence. A 3.2 M solution of HiLyte Fluor 488 labeled amyloid (1-42) (chosen due to availability of labeled peptide) in distilled water was incubated at room temperature. A rapid decrease in fl uorescence of the solution was seen along with the appearance of visible precipitates (Figure 3-5). HiLyte Fluor 488 Labeled A (1-42) Fibril Disassembly with Gelsolin It has been well demonstrated that huma n plasma gelsolin has the ability to disassemble preformed A fibrils (Ray et al., 2000). The administration of bovine plasma gelsolin has prevented amyloid depos ition in transgenic mice (Matsuoka et al., 2003). Due to theses observations we deci ded to determine if there are species’

PAGE 43

31 differences between bovine plasma gelsolin a nd human plasma gelsolin in the ability to disassemble preformed A fibrils. Under sterile conditions, dialysis tubes w ith a 10 kD molecular weight cutoff were used as a membrane in order to separate the monomer or dimer forms of labeled amyloid 1-42 from more aggregated oligomeric forms. Inside the dialysis tubes100 l of 1.6 M fibrillized HiLyte Fluor 488-Labeled amyloid (1-42), from the previous experiment, was added. On the outsi de of the dialysis tubes 200 l of distilled water was added. The experimental groups included adding 0.5, 1.0, 2.0, 3.0, 5.0, and 8.0 g of human plasma gelsolin or 1.0, 2.0, 3.0, 5.0, and 8.0 g of bovine plasma gelsolin on the inside of the dialysis tubes, with the amyloid As a control, only amyloid inside the dialysis tube and no plasma gelsolin, was used. To determine the maximum amount of diffused labeled amyloid possible, a group with only labeled amyloid on the inside of the dialysis tube was used with 5M urea in distilled water on the outside of the tube for complete disassembly of amyloid fibers to amyloid monomers, The dialysis reactions were set up in triplicate under st erile conditions and allowed to incubate at room temperature for 1 week. At this point the fluorescence was measured outside the dialysis tubes, revealing the am ount of amyloid that was diffusible across the filters. This experiment was repeated twice using triplicate samples each time. The results are displayed in Figure 3-6.

PAGE 44

32 Measuring Binding Amyloid to Plasma Gelsolin Using Surface Plasmon Resonance Surface plasmon resonance ( SPR) is method that can be employed to observe interactions between macromolecules by measur ing local changes in the refractive index of a solution containing a substrate flowing across a metal surface to which a ligand has been attached. Amyloid 1-40 and 1-42 were coupled to a gold chip modified with carboxymethylated dextran layer, Biacore CM 5 chip, using amino-coupling chemistry. Reactive sucinimide esters were produced on the surface of the CM5 chip by using a 1:1 mixture of 1-ethyl-3-[3-(dimethylamino) propyl]carbodiimide hydrochloride (EDC) and N-hydroxysuccinimide (NHS). Then an acetate solution, pH 4, containing 50 g/ml amyloid 1-40 (Figure 3-7) or 1-42 (Figure 3-8) was passed acr oss the activated surface of the CM5 chip. Free amino groups in the am yloid are able to reac t with the activated surface of the CM5 chip becoming covalent ly bound. Following attachment of the amyloid a high concentration of ethanolamine wa s passed over the surface to block any unreacted carboxymethyl groups on the surface of the chip. For amyloid 1-40 an increase of 650 resonance units (R U) was observed, and for amyloid 1-42 an increase of 3880 RU was observed. Following coupling of amyloid to the CM5 chip a solution containing 50 g/ml of the antibody 6E10 was injected across both surfaces bound with amyloid 1-40 and 142 for 240 seconds, and also a blank surface of the chip, as a positive control for binding. The resultant sensorgram is displayed in Fi gure 3-9. This shows th at the 6E10 binds to both surfaces coupled with amyloid 1-40 or 1-42 showing a response of 4000 RU and

PAGE 45

33 5000 RU respectively, and no binding to the blank surface. The off-rate for 6E10 appears quite slow with no loss of response occurring after cessation of inject ion. The surface of the chip was then regenerated with a so lution of 4M guanidine HCL removing any unconjugated protein or pep tide. Following regenera tion a solution containing 50 g/ml of human plasma gelsolin was injected across the surfaces for 240 seconds. The resultant sensorgram (Figure 3-10) s hows response increases of a bout 750 RU for the amyloid 140, 1-42, and the blank channel. This indicates that the ob served response results from bulk flow changes resulting from differences in buffer composition that cause changes in the refractive index. Speci fic binding of gelsolin to the coupled amyloid is difficult to detect in this case. Another CM5 chip was then coupled with human plasma gelsolin using the same amino coupling chemistry (EDC/NHS ) as described for the amyloid coupling. The coupling procedure resulted in a final incr ease of 1650 RU for the surface of the chip (Figure 3-11). Following the coupling proce dure with human plasma gelsolin a solution containing 30 second injection of 50 mg/m l of GS-2C4 antibody (recognizes human plasma gelsolin) was injected across the su rface of the human plasma gelsolin coupled CM5 chip (Figure 3-12). There was a bulk in crease in response of 225 RU, with only a small amount of response increase due to GS -2C4 binding (~40 RU). After binding with GS-2C4 antibody the surface of the chip was regenerated with 4M guanidine HCL. A solution of 50 mg/ml of amyloid 1-40 was then injected across the surface of the chip. After a bulk increase in re sponse of 8000 RU, the increase in response attributable to amyloid b bi nding was ~ 20 RU (Figure 3-13). In order to confirm the presence of bound amyloid 1-40, 50 g/ml of 6E10 was then injected across the

PAGE 46

34 surface. 6E10 resulted in an increase in re sponse of ~ 1200 RU. The surface was then regenerated with 4M guanidine HCL. A solution containing 50 g/ml of amyloid 1-42 was then injected across the surface of the chip again resulting in a bulk response of ~8000 RU, and a response increase a ttributable to binding of amyloid 1-42 of ~ 30 RU (Figure 3-14). 6E10 was again used to confirm the presence of bound amyloid 1-42, resulting in a response increase of ~1020 RU. Conclusions The preceding experiments were undertaken in an effort to determine if there were species differences in the binding affinities of bovine and human plasma gelsolin. Human gelsolin had been shown to bind to a nd disassemble amyloid fibrils (Chauhan et al., 1999) and bovine gelsolin had been shown to have an effect In Vivo in mice that deposit a human form of amyloid (Matsuoka et al., 2003). We hypothesized that plasma gelsolin may be an effective periphe rally expressed gene th erapy for Alzheimer’s disease based on a sink-hypothesi s, that binding amyloid peripherally will shift the equilibrium of amyloid from depositing in the CNS to the periphery where it can be cleared. So we set out to meas ure binding affinities in order to determine if one species form would be advantageous over the other. Although we were unable to measure sp ecific on and off rates of the binding interaction between plasma gelsolin and amyloid our results demonstrate that the presence of plasma gelsolin alte rs the binding activity of amyloid The data from the 125I labeled amyloid (1-42) binding experiment (Figur e 3-1) and centrifugal filtration experiment (Table 3-1) both demonstrate th at in the presence of gelsolin amyloid has an improved ability to pass through either 30 kD f ilter or glass filter paper. This is most

PAGE 47

35 likely due to gelsolin’s abili ty to disassemble amyloid fibrils. Our data shows that HiLyte Fluor 488 Labeled amyloid (1-42) forms insoluble fi bril precipitates (Figure 35B). While this occurs the fluorescence activity decreases (F igure 3-5A), most likely due to fluorophores being sequestered in the non-aqueous phase of the precipitates. As gelsolin is added to the amyloid fibril suspension there is a fluorescence increase that is accompanied by a decrease in the anisotropy (Figures 3-3 and 3-4 ) This decrease in anisotropy can be attributed to an increase in rotationa l diffusion caused by a decrease in size as amyloid fibrils are disassembled. The fact that gelsolin can do this across a membrane (Figure 3-6) suggest that gelsolin may have enzyme-like activity shifting the equilibrium from amyloid fibril formation to soluble -helical amyloid monomers. Gelsolin does have similar enzyme-like activ ity with its interaction with actin. It disassembles actin filaments and can cap act in monomers in preparation for actin filament elongation. If gelsolin ’s interaction with amyloid is enzyme-like, it may have a high affinity for oligomeric amyloid that has a -sheet secondary st ructure, and a low affinity for than soluble monomeric amyloid with -helical secondary structure. This may lead to quick off rates, which can be di fficult to measure. Supporting the idea that gelsolin may have fast off rates are meas urements for the interaction (Kd) between amyloid (1-40) and human plasma gelsolin repo rted by Chauhan et al. to be in the M range (1999). To conclude our data suggest th at gelsolin may in fact have an enzymaticlike property that makes it an even more at tractive agent for the treatment of amyloid related disorders than just an amyloid binding agent, because one molecule of gelsolin potentially have an effect on a large population of amyloid molecules.

PAGE 48

36 Table 3-1 Measurements of Fluorescence and Anisotropy before and after centrifugal filtration through 30,000 NMW cutoff filters. Anisotropy is reported as SEM, values > 0.400 were excluded. Fluorescence Anisotropy A (82nM) prefilter 7.30 0.060 0.002 A flow through 0.94 0.400 na A retentate 1.00 0.363 0.025 A (82nM) + Gelsolin (418nM) prefilter 26.50 0.050 0.001 A + Gelsolin flow through 4.80 0.047 0.002 A + Gelsolin retentate 16.46 0.076 0.001

PAGE 49

37 Figure 3-1 Binding of 125I labeled A 1-42 to Whatman GF/C filters. P values from ttests performed show there is consistently lower binding of A 1-42 (at 2, 5, and 10 M) in the presence of 2 M recombinant gelsolin. 0 100 200 300 40 0 0.000 0.025 0.050 0.075 0.100 0.125 0.150 0.1756E10 ( nM ) Anisotrop y Figure 3-2 Binding of amyloid (1-42) to the antibody 6E10. Blocks represent means SEM of triplicate measurements. Line repr esents best fit with GraphPad Prism software with a minimum anisotropy of 0.05035, a maximum of 0.1453, a log EC50 of 140.7 nM, and a hillslope of 0.01037. The fit had an r2 of 0.9938, and an absolute sum of squares of 2.34210-4.

PAGE 50

38 0 100 200 300 400 500 0 50 100 150 200 250Anisotropy Fluorescence 0.00 0.01 0.02 0.03 0.04 0.05 0.06 0.07 0.08 0.09 0.10Bovine Plasma Gelsolin ( nM ) FluorescenceAnisitropy Figure 3-3 Binding of plasma gelsolin to fluorescently labeled amyloid (1-42) 0 250 500 750 1000 1250 1500 1750 2000 0.0 2.5 5.0 7.5 10.0Anisotropy Fluorescence 0.0 0.1 0.2 0.3 0.4Plasma Gelsolin ( nM ) FluorescenceAnisotropy Figure 3-4 Amyloid (1-42) binding to plasma gelsol in after concentr ation by centrifugal filtration

PAGE 51

39 Figure 3-5 Fibril formation of HiLyte Fluor 488 Labeled A (1-42) fluorescence rapidly decreases (A), while visible precipitates are formed (B). Ph oto taken at day 7. A (1-42) Fibrillization 0 1 2 3 4 5 6 7 8 0 50 100 150 200 250Da y FluorescenceA B

PAGE 52

40 Figure 3-6 Dialysis experiment. Amyloid fibril disassembly by gelsolin across a membrane Human and bovine refer to the speci es of gelsolin added. Outside and inside refers to whether the gelso lin was added outside the dialysis tubing or inside the dialysis tubing. Fibrillized A was added to the inside of dialysis tubes in all samples. Fluorescence was m easured outside the dialysis tubes for all samples. The urea samples contain no gelsolin, just added as a line to represent the maximum fluorescence obtainable. 0 1 2 3 4 5 6 7 8 9 0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0 1.1 1.2 g Gelsoli n Fluorescence (fraction of control)Human Urea Bovine

PAGE 53

41 Figure 3-7 Amino-coupling of amyloid 1-40 to CM5 chip. A ri se in the response is observed at around 300 seconds when th e injection of EDC/NHS reagents occurs, likewise a drop back to baseline is observed when the injections ends at 700 seconds this change in response is due to bulk differences in buffer composition. At 850 seconds the injection of amyloid 1-40 begins corresponding with anothe r increase in response. Upon completion of the amyloid injection the res ponse does not fall completely back to base line indicative of covalently attached amyl oid. Ethanolamine is then injected between 1400 to 1800 seconds to block a ny unreacted carboxymethyl groups. Another rise in response is observe d attributed to bulk changes buffer composition again. At the completion of a ttachment there is a rise in baseline of about 650 RU.

PAGE 54

42 Figure 3-8 Amino-coupling of amyloid 1-42 to CM5 chip. A ri se in the response is observed at around 300 seconds when th e injection of EDC/NHS reagents occurs, likewise a drop back to baseline is observed when the injections ends at 700 seconds this change in response is due to bulk differences in buffer composition. At 850 seconds the injection of amyloid 1-42 begins corresponding with anothe r increase in response. Upon completion of the amyloid injection the res ponse does not fall completely back to base line indicative of covalently attached amyl oid. Ethanolamine is then injected between 1400 to 1800 seconds to block a ny unreacted carboxymethyl groups. Another rise in response is observe d attributed to bulk changes buffer composition again. At the completion of a ttachment there is a rise in baseline of about 3880 RU.

PAGE 55

43 Figure 3-9 Antibody binding to amyloid couple d CM5 chip. Response increases of 4000 RU and 5000 RU, for amyloid 1-40 and 1-42 respectively, are observed after a 240 second injection of a solution containing 50 g/ml of the antibody 6E10. No response is observed when th e solution is injected across the blank channel of the chip.

PAGE 56

44 Figure 3-10 Human plasma gelsolin binding to amyloid coupled CM5 chip. Response increases of about 750 RU for the amyloid 1-40, 1-42, and the blank channel are observed during a 240 second injection of 50 g/mL solution of human plasma gelsolin. This indicates that the observed response results from bulk flow changes caused from differences in buffer composition resulting in changes in the refractive index.

PAGE 57

45 Figure 3-11 Amino-coupling of human plasma gelsolin to CM5 chip. A rise in the response is observed at around 300 seconds when the injection of EDC/NHS reagents occurs, likewise a drop back to baseline is observed when the injections ends at 700 seconds this change in response is due to bulk differences in buffer composition. At 850 seconds the inje ction of plasma gelsolin begins corresponding with a nother increase in response. Upon completion of the amyloid injection th e response does not fall completely back to base line indicative of covale ntly attached amyloid. Ethanolamine is then injected between 1400 to 1800 seconds to block any unreacted carboxymethyl groups. Another rise in re sponse is observed attributed to bulk changes buffer composition again. At the completion of attachment there is a rise in baseline of about 1650 RU

PAGE 58

46 Figure 3-12 Binding of GS-2C4 to human plasma gelsolin coupled CM5 chip.

PAGE 59

47 Figure 3-13 Binding of amyloid 1-40 to human plasma gelsolin coupled CM5 chip. Figure 3-14 Binding of amyloid 1-42 to human plasma gelsolin coupled CM5 chip.

PAGE 60

48 CHAPTER 4 EXPRESSING PLASMA GELSOLIN AND EFFECTS IN TRANSGENIC MICE Introduction Gene therapy is a novel approach, whic h utilizes specific sequences of DNA to treat, cure, or ultimately prevent disease. There are major hurdles to overcome for it to be effective; the delivery of such large molecules to the ta rget tissue offers a challenge because of vulnerability of degradation, due to endogenous nucleases. Another challenge is having the cells at the ta rget tissue interna lize the DNA and transport it into the nucleus. Finally, having the DNA transcribe d and translated to produce a therapeutic protein that is transported to the proper site of action offers another hurdle to overcome. Our target is the amyloid that accumulates as senile pl aques in the brains of those suffering from Alzheimer’s disease. In the pr evious chapter we have demonstrated that human plasma gelsolin holds promise as an agent that can disassemble preformed amyloid fibrils. Because plasma gelsolin contains an amino-terminal 23 amino acid signal peptide which signals the cell produci ng it to secrete it to wards the bloodstream, the site of action for our target can be in the bloodstream rather than the brain, simplifying delivery. Therefore expressing pl asma gelsolin in any peripheral tissue should be enough to increase plasma gelsolin in the bloodstream. The hydrodynamic gene delivery method offers an efficient technique for testing if peripheral expression of plasma gelsolin can have an effect on am yloid distribution and clearance in transgenic mice. The h ydrodynamic gene delivery method was first developed by Dr. Dexi Liu (1999). This me thod involves injecting a large volume of a

PAGE 61

49 DNA solution in a short period of time, via the tail vein of a mouse, and results in a high level of transgene expression in the liver. The mechanism for how this works seems to be that the initial rapid increase in blood volume causes an increas e in venous pressure, which forces an enlargement of the liver fenestrae, and causes the formation of transient pores on the membranes of hepatocytes allowing the plasmid DNA to enter the cells (Zhang et al., 2004). There is typically a high level of tr ansgene expression following hydrodynamic gene delivery, followed by a quick drop off to a lower stable level of expression (Liu et al., 1999). Alino, Crespo, and Da si showed that when the full length hAAT promoter was used to drive expression, after hydrodynamic gene delivery a stable therapeutic level of human alpha-1-antitrypsin (hAAT) is detected in the blood for up to 120 days, (2003). The majority of plasmids delivered du ring hydrodynamic gene delivery studies are driven by human cytomegalovirus immediate-ear ly promoter (CMV). Song et al showed that the CMV-chicken beta actin hybrid pr omoter (CBA) had well over 100 times the activity in the mouse liver than the CMV prom oter (2001). We chose to incorporate the CBA promoter in our expression plasmid (Fi gure 2-1). We also included the woodchuck hepatitis virus post-transcri ptional regulatory element (WPRE) in our plasmid. The WPRE functions to stabilize mRNA having th e effect of increasi ng the half-life for mRNA and ultimately increasing the amount of gene product produced. We have previously described that incorporating the WPRE into vectors increased green fluorescent protein (GFP) and nerve growth factor (NGF) expression by more than ten fold in rats (Klein et al., 2002).

PAGE 62

50 Results and Discussion Conformation of Vector Product and Activity To test that the vector we constructe d pUFGL (Figure 2-1C) does in fact produce plasma gelsolin, we transf ected 293 cells using the Ca3(PO4)2 precipitation method, collected and concentrated media, and then performed a western blot to confirm size and identity of plasma gelsolin secreted into th e media (Figure 4-1). There was a protein that ran at ~ 91kD that was immunoreactive with the anti-human gelsolin antibody, GS-2C4 (Sigma, St. Louis, MO), which was absent in the samples that had been transfected with the control plasmid pGFP. To confirm that our gene product retained its amyloid binding activity we spiked some of the unconcentrated media with A (1-42). Performed an immunoprecipitation using the GS-2C4 anti-human gelsolin anti body and then ran a Western blot with the precipitates. We found that A 1-42 coprecipitated (Figure 42) with gelsolin in the pUFGL transfected, A 1-42-spiked media and not w ith the pGFP transfected. A 1-42spiked media. Amyloid 1-42 was not found in the precipitates when the media was not spiked with A 1-42. Hydrodynamic Gene Delivery in Mice To determine the distribution and level of gene expression we should expect from using the hydrodynamic gene delivery techni que mice were injected either pGFP or GWIZ luciferase expression plasmi d via the hydrodynamic technique. One day following injections with the GWIZ vector the two animals injected were imaged by the University of Florida’s biomed ical engineering depart ment (Figure 4-3). Enough visible light was produced by the gene product luciferase that the liver was fully

PAGE 63

51 illuminated for at least 30 minutes after an intraperitoneal injection of luciferin (luciferase’s natural substrate). Two and a half weeks afte r injecting mice with pGFP plasmid the animals were sacrificed and their livers were excised. Sections of the liver were made, 30 m thick, and the distribution of fluorescence wa s examined by fluorescence microscopy (Figure 4-4). At the two and a half w eek post-injection timepoint GFP fluorescence distribution was detected widely throughout the liver. Transgenic mice expressing mutant presen ilin-1M146L (mutant PS-1)(Duff et al., 1996) were injected via the hydrodynamic gene delivery method either pUFGL or pGFP. Plasma samples were taken at 24, 48 and 96 hours. These samples were used for a western blot (Figure 4-5). Plasma samples from pUFGL injected an imals had clear bands corresponding to human plasma gelsolin that were not present in the pGFP injected animal. Hematoxylin and Eosin Staining in Mi ce after Hydrodynamic Gene Transfer Hematoxylin and eosin (H & E) staining is a routine stain that takes advantage of two separate dyes. Hematoxylin stains nucle ar material a purplish color, while eosin stains membranes and connectiv e tissue an orange-pinkish color. While unable to identify specific chemical markers for in flammation, H & E staining is useful at identifying, abnormal growth, division in the nucl eus, or cellular death in tissues that may be related to disease or injury. To determine if damage or inflammation a ffected the liver tissu e 18 days after gene delivery via the hydrodynamic injections, liver s were examined using H & E staining. Eighteen days after gene delivery of pGFP or pUFGL in double transgenic mice

PAGE 64

52 expressing both human mutant APP695K 670N,M671L (Tg2576)(Hsiao et al., 1996) and mutant presenilin-1M146L(Duff et al., 1996) (huAPP/PS1M146L mice). H & E staining revealed no differences among livers from mice injected with pGFP, pUFGL or untreated mice (Figure 4-6). Conclusions In this chapter we demonstrate that our vector, pUFGL, does in fact produce human plasma gelsolin immunoreactivity, verified by we stern blot. We also show that our gene product maintains its A binding activity verified by the co-immunoprecipitation of A (1-42) with plasma gelsolin. We also show that by delivering vectors by the hydrodynamic technique we are able to obtain high levels of gene expression for at least two and a half weeks, and this expression doe s not result in damage or inflammation to the liver, detectable by H & E staining. Finally we are able to find detectable levels of human plasma gelsolin in the plasma of mice up to 96 hours after gene delivery by the hydrodynamic technique.

PAGE 65

53 Figure 4-1 Immunoblot of transfected media. 293 cells at 50% confluency (lane B and D) or 80% confluency (lane C and lane E) were transfected using the Ca3(PO4)2 precipitation method with either pGFP (lane B and lane C) or pUFGL (lane D and lane E). 48 hours af ter transfection, media was collected, concentrated, and then separated on a 7.5% PAGE and immunoblotted with a monoclonal anti-gel solin antibody. Lane A contains 1 g of human plasma gelsolin. Figure 4-2 Immunoprecipitation Amyloid (1-42) co-immunopr ecipitates with 293 expressed plasma gelsolin. The first la ne contains immunopr ecipitates from pGFP transfected media with A 1-42. The second lane contains immunoprecipitates from pGFP transfected media with no A The third lane contains immunoprecipitates from pUFGL transfected media with A 1-42. The fourth lane contains immunoprecipi tates from pUFGL transfected media with no A + _ Amyloid (1-42) 4.5kD pGFP pUFGL Am y loid 1-42+ + + + + _ Plasma Gelsolin

PAGE 66

54 Figure 4-3 Bioluminescence resulting from hydrodynamic gene transfer. Imaging was done by the University of Florida’s Depa rtment of Biomedical Engineering, as collaboration with fellow pharmaceutics graduate student Natalie Toussaint. The white rectangular shows the region imaged with a th ermoelectrically cooled (-70 C), back illuminated CCD array (Roper Scientific Instrumentation, Trenton, NJ) coupled w ith an optical lens subsystem (Zoom 7000, Navitar, Rochester, NY).

PAGE 67

55 Figure 4-4 Green fluorescent pr otein. Expression observed in the liver of a mouse two and a half weeks after pGFP delivery via the hydrodynamic gene delivery technique. GFP expressing hepatocytes, green cells, are wi dely distributed throughout the liver. The red color resu lted from background stained with Alexa Flour 488 goat anti-mouse IgG (H+L ). Dark empty spots are hepatic sinusoids. Figure 4-5 Western blot of plasma samples ta ken from PS-1 mice injected with pUFGL at 24 hours (Lanes D and E), 48 hours (Lane F), and 96 hours post injection (Lane G); or plasma taken from pGFP injected mice at 24 hours (Lane K), 48 hours (Lane L), and 96 hours post injection (Lane M). Lane N is plasma from a non injected mouse and Lane O is plas ma from a non injected mouse spiked with 500ng of human plasma gelsolin. Lanes A and H contain 1500ng of human plasma gelsolin, Lanes B a nd I contain 1000ng of human plasma gelsolin, and Lanes C and J contai n 500ng of human plasma gelsolin.

PAGE 68

56 Figure 4-6 Hematoxylin and eo sin staining of livers 18 da ys after hydrodynamic gene delivery. Upper image represents an unt reated liver section, and lower image represents a pUFGL injected liver section. Scale bar represents 50 m for both images.

PAGE 69

57 CHAPTER 5 EFFECT OF GELSOLIN EXPRESSION ON AMYLOID DEPOSITION Introduction Since the discovery that senile deposits in Alzheimer’s disease are composed mainly of the fibrillar amyloid peptide (Glenner and Wo ng, 1984), researchers have discovered a number of muta tions on either amyloid ’s parent protein (APP), or proteins that process APP (PS1), that l ead to inheritable forms of th e disease. These discoveries have been quite useful in the development of transgenic mouse models of Alzheimer’s disease pathology. Achieving elevated levels of transgene expression was a critical step in the development of transgenic mice to model neur odegenerative diseases. It was recognized in the early 1990s that the gene encoding fo r mammalian prion protein (PrP) would make an effective expression package to produce fore ign proteins in the cen tral and peripheral nervous systems of mice (Hsiao et al., 1995; Scott et al., 1992; Telling et al., 1994). A 42 kb cosmid clone of the Syrian hamster Pr P gene was made and it was noted that the entire open reading frame is contained in a si ngle exon. (Basler et al ., 1986). This exon can be excised and exchanged with the cDNA of a gene of interest, which can then be used for a pronuclear injection into mouse embr yos to generate a transgenic line of mice expressing the protein of intere st at high levels in the ner vous system and heart of the mice (Borchelt et al., 1996). In 1996, using this technique, Hsiao et al developed a transgenic mouse model expressing the Swedish double mutant APPK670N,M671L (mutant APP, Tg2576) with a

PAGE 70

58 C57/BI6 and SJL mouse background. Thes e mice expressed the mutant APP about 5.5 times that of the endogenous murine APP. After 11 months of age amyloid plaquelike deposits are found throughout the cortex, hippocampus, presubiculum, subiculum, and the cerebellum (1996). These mice also demonstrated a deficit in memory-related behavior that correlated to th e levels of insoluble amyloid in the brain (Westerman et al., 2002). Duff and coworkers developed mice that express a mutant form of presenilin-1M146L (mutant PS-1). These mice had no detectable histopathology of Alzheimer’s disease, however they did have elevated levels of amyloid (1-42) (1996). When the mutant PS1 mice are bred with the Tg2576 mice the resu ltant double transgenic progeny (APP/PS1) have an accelerated rate of amyloid deposition, about 3-5 times that of the singly transgenic Tg2576, with a age of onset of be tween 3 and 6 months of age (Holcomb et al., 1998; Holcomb et al., 1999). Using a similar strategy Borchelt et al. de veloped a transgenic model expressing a humanized version of murine APP695. Th is humanized gene was controlled by the mouse PrP promoter that drove expressi on of cDNA containing all murine sequence except for the amyloid domain and the mutations (K595N, M596L) that are linked to the human Swedish form of familiar Alzheimer’s disease. The level of transgene expression of these mice was about 2-3 times that over the endoge nous APP expression in nontransgenic littermates (1996). These mice develop amyloid deposits around 18 months of age (1997). Jankowsky and coworkers developed a line of mice that express the mouse human chimeric Swedish APP695 that Borchelt create d along with a form of human presenilin-1

PAGE 71

59 with exon nine deletion (PS1 E9). They showed that these mice produce about 2.5 times the level of amyloid (1-42) while amyloid (1-40) levels remain constant (2004). The elevated levels of amyloid (1-42) result in deposits occu rring at a much accelerated rate as compared to the singly transgenic mice. These mice begin to develop deposits by the age of 6 months as compared to 18 mont hs in the mice which don’t co express PS1 E9. APP/PS1 mice have been useful as mode ls to study treatments that target amyloid Morgan et al. successfully vaccinated APP/PS1 mice against A which had a dramatic effect on amyloid deposition that protected against memory and learning deficits (2000). Deane et al. also sa w dramatic effects on amyloid deposition by administering a soluble form of the recep tor for advanced glyclation end products (RAGE)(2003). Matsuoka et al. treated young APP/PS1 mice with the amyloid binding agents GM1 and plasma gelsolin, and sa w significant reductions in amyloid levels in the brain (2003). This is why we belie ve APP/PS1 mice will make a good model to determine if peripheral expression of plasma gelsolin can effect amyloid deposition. Results Mice Expressing Swedish Mutant Amyl oid Precusor Protein (Mouse/Human Hybrid) and Exon 9 Deleted Mutant Presenilin-1 Message Detection Nine month old double transgenic mice e xpressing both Swedish mutant mouse /human hybrid APP695 K594N,M595L and mutant presenilin-1 E9 (mo/huAPP /PS1 E9) (Jankowsky et al., 2004) were injected with pUFGL, via the hydrodynamic gene delivery method. Two and one half weeks following inje ctions the animals were sacrificed, along with three untreated littermates. RNA was purified from the liver and brain tissue, as described in the methods section. RT-PCR wa s performed using vector specific primers

PAGE 72

60 designed to yield a 900 bp pr oduct from processed mRNA coded from pUFGL or an 1800bp fragment from unprocessed RNA or DNA. All samples from animals that received an injection of pUFGL showed vector specific mRNA hybridization. In samples from animals that did not receive injections vector specific mRNA hybridi zation was not detected (Figur e 5-1). All of the samples did show mRNA hybridization when reactions were run with primers specific for -actin (data not shown) indicating that mRNA is pres ent in all of the samples. Both samples from RNA purified from 293 cells transf ected with pUFGL had positive bands. RNA samples purified from brain tissue did not show vector specific mRNA hybridization (data not shown) suggesting if there is vector gene expressi on in the brain it is below detectable quantities. Total Brain Amyloid (1-42) Concentrations Enzyme linked immunosorbent assays ( ELISA) were performed in order to measure the concentration of both solubl e and insoluble fractions of amyloid (1-42) in the hemi-brains from both pUFGL-injected a nd noninjected mice. All hemibrains were subjected to a guanidine ex traction in order to obta in the total amount of A contained within the brain tissue (Johns on-Wood et al., 1997; Masliah et al., 2001). The samples were run in duplicate on two separate occasi ons for a total of four samples assayed per animal. Untreated controls had a mean SEM amyloid (1-42) concentration of 2,306 202.6 picomoles per gram of brain tissue and the injected animals had 1,174 334.7 picomoles per gram of brain tissue. A onetailed t-test was performed and showed that these groups differ significan tly with a P value = 0.0222 (Figure 5-2).

PAGE 73

61 Plasma Amyloid Concentrations Blood was collected via the retr o-orbital sinus using hepari nized capillary tubes just prior to sacrifice. Samples were immediat ely centrifuged, followed by plasma collection which was flash frozen in liquid nitrogen. Samp les were then thawed at a later time point and analyzed in duplicate by ELISA to determine amyloid 1-40 and 1-42 concentrations. Untreated mice had a mean SEM plasma A 40 concentration of 315 116.3 fmol/mL, while pUFGL-injected mice had 200.3 86.5 fmol/mL (Figure 5-3A). Untreated mice had a mean SEM plasma A 42 concentration of 95.6 23.2 fmol/mL, while pUFGL-injected mice had 150.9 53.97 fmol/mL (Figure 5-3B). Trends but not significant changes in either A 42 or A 40 concentrations were observed; however a significa nt decrease in the ratio of A 40/ A 42 was observed in the pUFGL-injected mice (Figure 53C). Untreated mice had a mean A 40/ A 42 ratio of 3.0 0.65, while pUFGL-injected mice had a mean A 40/ A 42 ratio of 1.24 0.49 (P = 0.045). Dense Cored Amyloid Deposits Thioflavin S staining was used to examine the extent of dense cored amyloid Deposit pathology in injected and un treated mice (Sun et al., 2002). 50 m thick coronal sections were made. Four s ections 0.3 mm apart were st ained with thioflavin S and digital micrographs were made of epifluores cence images (Figure 5-6). Images were analyzed in a blinded manner using NIH Imag e J software. The area of the hippocampus and cortex, total stained area, area of each individual stained deposit, and the number of stained deposits was determined by thres holding segmentation. The amyloid burden was

PAGE 74

62 determined by dividing the total area staine d by the total area of the hippocampus and cortex. Untreated animals had mean amyloid burden of 1.03 0.13%, while pUFGLinjected animals had a mean amyloid burden of 0.39 0.09%. A one-tailed t-test showed that these groups differ significantly with a P value = 0.0085 (Figure 5-4A). Untreated mice had a mean total stained area of 82,660 10,150 pixels2, while the pUFGL-injected mice had a mean total stained area of 33,740 8429 pixels2. These measurements were also determined to be significantly different by an unpaired one tailed t-test, P value = 0.0103 (Figure 5-4B). The average deposit size for untreated mice was determined to be 42.2 4.8 pixels2, while the average size fo r pUFGL was 34.2 1.2 pixels2. These were not found to be statistically different (P value = 0.0905, by an unpaired one-tailed t-test) (Figure 5-4C). Untreated mice had an average of 503 84 deposits per section while pUFGL-injected mice averaged 244 51 deposits per section. These were found to differ statistically (P value = 0.029, unpaired one-tailed t-test) (Figure 5-4D). Diffuse Amyloid Deposits Diffuse amyloid deposits were visuali zed by immunohistochemistry using the antibody 6E10, which recognizes the firs t 17 amino acids of human amyloid 50 m thick coronal sections were made. Three se ctions 0.3 mm apart were stained. Ditgital micrographs were made using light microsc opy (Figure 5-7). Stained sections were analyzed in a blinded manner using NIH Imag e J software. The area of the hippocampus and cortex, total stained area, area of each individual stained deposit, and the number of stained deposits was determined by thres holding segmentation. The amyloid burden was

PAGE 75

63 determined by dividing the total area staine d by the total area of the hippocampus and cortex. Untreated animals had a mean amyloid burden of 1.48 0.19%, while pUFGLinjected animals had a mean amyloid burde n of 1.12 0.05%. An unpaired one-tailed ttest showed that these groups were not st atistically different, however a strong trend towards significance was present (P value = 0.074) (Figure 5-5A). Untreated mice had a mean total stained area of 67,700 10,590 pixels2, while the pUFGL-injected mice had a mean total stained area of 48,570 2643 pixels2. These measurements were also not significantly different by an unpaired one tailed t-test, but again there was a trend (P value = 0.078) (Figure 5-5B). The average deposit size for untreated mice was determined to be 16.9 2.0 pixels2, while the average size fo r pUFGL was 16.2 1.2 pixels2. These were not found to be statistically different (P value = 0.387, by an unpaired one-tailed t-test) (Figure 5-5C). Untreated mice had an av erage of 1326 53 deposits per section while pUFGL-injected mice averaged 997 123 depos its per section. These were found to differ statistically (P value = 0.035, by an unpa ired one-tailed t-te st) (Figure 5-5D). Soluble Amyloid Oligomers The distribution and relative quantities of soluble amyloid oligomers were examined by immunostaining with the A11 antibody. The A11 antibody recognizes an epitope that is common to so luble amyloid oligomers, but is not found in amyloidogenic monomers or mature amyloid fibrils(Kayed et al., 2003). Three sections from each animal were stained, all sections were in cubated for equal amount s of time during the labeling procedure, and DAB reactions were carried out on slides containing one section from each animal, in order to minimize differen ces in reaction times or reagents. Digital

PAGE 76

64 micrographs were prepared of the whole slid es under the same light source using an Olympus BH-2 light microscope equipped w ith a motorized stage and focus control system (Prior Scientific). Im age Pro plus version 4.0 software was used for tiling images together using a 4 nosepiece objective, a 1.25 internal magnification changer, and a 2.5 camera tube objective. To determine st aining intensity and the relative amount of antigen present grey values were measured from negative images of each section using NIH Image J software. Soluble amyloid oligomers appear to be widely distributed among both the pUFGL-injected mice and the untreated mice. Untreated mice had significantly lower negative grey values (127.6 4.6) compar ed to pUFGL-injected mice (142 1.8, P = 0.0220), which may indicate an increase in the concentration of soluble amyloid oligomers in pUFGL-injected mice (Figure 59). However, variab ility in fixation and section can also result in hol es in the section that may co rrespond to less staining of the section. At high magnificati on the distribution of amyloi d oligomer staining in both untreated mice and pUFGL-injected mice is mainly confined to the neuronal soma, axons, and dendritic processes (Figure 5-10) However in pUFGL-injected animals a number of glia cells and cel ls associated with the vasc ulature show an apparent immunoreactivity to the A11 antibody also (Figure 5-11). Microglia Microglia are a population of dendritic cells in the brain, thought to be of the same origin as monocytes and macrophages. Microg lia can reside in the brain in a number of different states. Resting microglia (ramif ied) have long finely branched processes extending from all directions from the perinu clear cytoplasm (Giulian and Baker, 1986). Resting microglia, in response to a number of different insults to the nervous system,

PAGE 77

65 begin to proliferate while still in non-phagocyt ic state, and are referred to as activated microglia. Activated microglia are recognizab le by retracted cellula r processes and mild hypertrophy resulting in a stout configuration. Further transf ormed microglia, with their processes even more retracte d, are known as reactive or phagocytic microglia and are considered to be like “brain macrophages”(Str eit et al., 1988). Reactive microglia have a round shape and can appear ruffled, due to s hort cytoplasmic projec tions, and are capable of releasing growth factors, cytokines, and free radicals. The state of microglia cells in mo/huAPP/PS1 E9 brain sections was examined by immunostaining using the OX42 antibody. OX 42 recognizes complement receptor 3 (CR3) which is expressed by microglia during al l stages. Three sections per animal were examined and all sections were incubated for equal amounts of time during the labeling procedure. DAB reactions were carried out on slides containing one section from each animal, in order to minimize differences in reaction times. Calibrate d digital micrographs were prepared of the whole slides with a c onstant light source, and relative grey values were measured from negative images for each section using NIH Image J software, to determine staining intensity. Untreated mice had significantly lower rela tive negative grey values (130.7 10.3) compared to pUFGL-injected mice (163.1 11.1, P = 0.0496), indicating higher concentration of CR3 receptor in the pUFGLinjected animals (Figure 5-12 & 5-13). At higher magnification it is qualitatively clea r that the increase in staining intensity observed in pUFGL treated mice is due to an increase in the number of activated and reactive microglia, distinguished by their re tracted processes and condensed cell bodies (Figure 5-14).

PAGE 78

66 Astrocytes Astrocytes are a type of glial cell recogni zable by numerous arms which give them a star shaped appearance. They play a numbe r of roles in the brai n including structuring the brain, providing neurons with nutrients contributing to the blood brain barrier, altering cerebral blood flow, clearing neurotra nsmitters and regulating ion concentrations in the extracellular space (P ellerin, 2005; Sofroniew, 2005; Volterra and Meldolesi, 2005). Astrocytes become activated in res ponse to disease or injury. One of the pathological features in Alzheimer’s disease is the presence of activated astrocytes in and around amyloid deposits (McGeer and McGeer, 2003). Astrocytic activation was examined in mo/huAPP/PS1 E9 brain sections by immunostaining for glial fibrillary acidic pr otein (GFAP). Two 50 mm thick sections per animal were analyzed for total percent stai ned area of the hippocampus and cortex using NIH Image J software. Untreated mice had a mean SEM GFAP positive percent area of 11.06 1.59 %, while pUFGL-injected mice had a mean SEM GFAP positive percent area of 14.15 2.42 % (Figure 5-15). These values do not differ significantly however upon examination at a high magnificat ion GFAP positive astrocytes in pUFGLinjected mice appear to have thicker and more numerous processes around areas corresponding to Congo Red positive amyloid deposits (Figure 5-16). Mice Expressing Swedish Mutant Amyloi d Precursor Protein (Human) and M146L Mutant Presenilin-1 Message Detection Eight-month old double transgenic mice ex pressing both Swedish mutant human APP695 (huAPP K670N,M671L, Tg2576) (Hsiao et al., 1996) and mutant presenilin-1 M146L (PS –1M146L) (Duff et al., 1996) were injected vi a the hydrodynamic gene delivery method

PAGE 79

67 either pUFGL (5) or pGFP (3). Two and one half weeks following injections animals were sacrificed along with age-matched untre ated mice (7). RNA was purified from the liver and brain tissue, as described in the methods section. RT-PCR was performed with primers designed to yield a 900 bp product from processed mRNA coded from our vector pUFGL. All samples from an imals that received an inje ction of pUFGL had positive bands. Samples from animals that did not rece ive injections or received injections of pGFP did not have bands (Figure 5-16 not a complete data set, a representative gel). All samples did produce bands when reac tions were run with primers for -actin (data not shown) indicating that mRNA is present in the all of the samples. Sample from RNA purified from 293 cells tran sfected pUFGL had a positiv e band. None of samples purified from brain tissue had positive bands (data not shown) suggesting there is no vector gene expres sion in the brain. Amyloid (1-42) Quantification by ELISA ELISA was used to quantify the concentr ation of guanidine extractable amyloid 42) in brain homogenates of the huAPP/PS1M146L mice. The resultant mean A 42 concentrations were 9179 916 picomoles pe r gram of brain tissu e for the untreated group, 10,928 731 picomoles per gram of brai n tissue for the pGFPinjected group, and 6740 998 picomoles per gram of brain tissue for the pUFGL group. A one-way ANOVA showed that these means do not diffe r significantly, P = 0.069 (Figure 5-18A), however a trend toward si gnificance was present. In order to account for inte r-litter variability, amyloid 42 concentrations from pGFP and pUFGL-injected mice were divided by the amyloid 42 concentrations obtained from their respective untreated litter mates. These values were reported as the

PAGE 80

68 fraction of amyloid 42 of untreated littermates. The pGFP group had a mean fraction of 1.00 0.06, while the pUFGL has a mean fraction of 0.797 0.05. A unpaired one tailed t-test showed that these groups di ffer significantly with a P value = 0.022 (Figure 5-18B). Plasma Amyloid Concentrations Blood was collected either via the retro-or bital sinus using he parinized capillary tubes or by intra cardiac puncture with an EDTA treated syringe just prior to sacrifice. Samples were immediately centrifuged; plas ma collected and flash frozen in liquid nitrogen. Samples were then thawed and analyzed by ELISA in duplicate to determine amyloid 1-40 and 1-42 concentrations. Untreated mice had a mean SEM plasma A 40 concentration of 1039 244.6 fmol/mL, pGFP-injected mice had 971.7 209 fmol/mL, and pUFGL-injected mice had 930 252 fmol/mL (Figure 5-19A). Untreated mice had a mean SEM plasma A 42 concentration of 484.9 143 fmol/mL, pGFP-injected mice had 453.5 45.7 fmol/mL, and pUFGL-injected mice had 467 43.3 fmol/mL (Figure 5-19B). There we re no significant changes in either A 42 or A 40 concentration. There were also no si gnificant changes measured in the A 40/ A 42 ratio; untreated mice had a mean con centration of 2.29 0.22, pGFP-injected mice had a mean ratio of 2.1 0.26, and pUFGL-injected mice had a mean ratio of 1.9 0.46 (Figure 5-19C). In light of the observation that amyloid binds to heparins (Brunden et al., 1993; Leveugle et al., 1994), it is possible that the recovery of amyloid may not have been the same between the two methods used for sample collection,. The data were examined in terms of collection method and differences were observed. Plasma collected retro-

PAGE 81

69 orbitally had a mean amyloid 40 concentration of 560.3 164 fmol/mL and a mean amyloid 42 concentration of 344.8 59.1 fmol/mL; when collected via intra-cardiac puncture the mean amyloid 40 concentration was 1185 128 and a mean amyloid 42 concentration of 531 53.0 fmol/mL (P = 0.016 and 0.057 respectively, by unpaired two tailed t-test, Figure 5-19D &E ). We concluded c oncentrations obtained from plasma samples collected with hepari nized tubes are probably not accurate and cation should be used when analyzing such da ta. When two different collection methods are used, the samples are definitely not comparable. Dense Cored Amyloid Deposits Thioflavin S staining was used to examine the extent of dense cored amyloid deposit pathology in injected (pGFP and pUFG L) and untreated mice (Sun et al., 2002). 50 m thick coronal sections were made. Four sections 0.3 mm apart were stained with thioflavin S and digital micrographs were ma de (Figure 5-20) as de scribed previously. Images were analyzed in a blinded manner us ing NIH Image J software. The area of the hippocampus and cortex, total stained area, area of each individual stained deposit, and the number of stained deposits was determined by thresholding segmentation. The amyloid burden was determined by dividing the total area stained by the total area of the hippocampus and cortex. Untreated animals had a mean amyloid burden of 1.48 0.16%, pGFP-injected animals had a mean amyloid burden of 1.59 0.10%, pUFGL-injected animals had a mean amyloid burden of 1.09 0.14%. A one-way ANOVA was performed, and the medians did not differ significantly with a P-value = 0.069 (Figure 5-21A).

PAGE 82

70 Untreated mice had a mean total stained area of 115,329 18,081 pixels2, pGFPinjected mice had a mean total stained area of 140,952 14,898 pixels2, while the pUFGL-injected mice had a mean total stained area of 87,503 14,622 pixels2. These measurements were also found not to differ statistically (Figure 5-21B, P-value = 0.127, by one-way ANOVA). There was also no difference observed in deposit size among the groups. Untreated mice had a mean deposit size of 75.62 15.4 pixels2, pGFP-injected mice had a mean Deposit size of 56.03 5.6 pixels2, and pUFGL-injected mice had a mean deposit size of 44.78 6.3 pixels2 (Figure 5-21C, P-value = 0.247, by one-way ANOVA). Deposit numbers per section did not change between the treatment groups either. Untreated mice had an average of 458 107 de posits per section; pGFP-injected mice had an average of 632 53 deposits per se ction, and pUFGL-injected mice had an average of 513 32 deposits per section. Again to account for inter-lit ter variability, dense-cored amyloid burdens, average deposit sizes, and average number of de posits per section from pGFP and pUFGLinjected mice were divided by the values obtained from their respective untreated littermates. No differences were detected in dense-cored amyloi d burden fractions or dense-cored deposit size fractions; pGFP-inject ed mice had a mean dense-cored amyloid burden fraction of 0.88 0.05 and a mean dense-cored deposit size fraction of 1.06 0.11, pUFGL-injected mice had a mean de nsecored amyloid burden fraction of 0.89 0.16 and a mean dense-cored deposit size fraction of 0.86 0.28 (Figure 5-22A &B).

PAGE 83

71 An increase in the average number of depos its per section fraction was detected in pUFGL-injected mice; pGFP-injected mice had a mean average dense-cored deposit number per section fraction of 1.06 0.11, pU FGL-injected mice had an average densecored deposit number per section frac tion of 1.38 0.13; P = 0.048 by unpaired twotailed t-test (Figure 5-22C). Diffuse Amyloid Deposits Diffuse amyloid deposits were visuali zed by immunohistochemistry using the antibody 6E10, which recognizes the firs t 17 amino acids of human amyloid 50 m thick coronal sections were made. Three sections 0.3 mm apart were stained. Digital micrographs were made (Figure 5-23). Stai ned sections were analyzed in a blinded manner using NIH Image J software. The area of the hippocampus and cortex, total stained area, area of each individual stai ned deposit, and the number of stained deposits was determined. The amyloid burden was determined by dividing the total area stained by the total area of the hippocampus and cortex. Untreated animals had mean amyloid burden of 4.25 0.46%, pGFP-injected mice had a mean diffuse amyloid burden of 4.04 0.46%, and pUFGL-injected animals had a mean amyloid burden of 2.83 0.05%. A one -way ANOVA showed that these groups were not statistically different, (P value = 0.099) (Figure 5-24A). The average diffuse deposit size for untreated mice was determined to be 20.3 1.5 pixels2, the average size for pGFP-in jected mice was 26.3 2.3 pixels2, and the average size for pUFGL-injected mice was 20.9 3.2 pixels2. These were not found to be statistically different (P value = 0.27 1, by one-way ANOVA) (Figure 5-24B).

PAGE 84

72 The average number of diffuse deposits per section for untreated mice was determined to be 3990 481 deposits/sec tion, 3437 513 deposits/section for pGFPinjected mice, and 3623 729 deposits/section for pUFGL-injected mice. These were not found to be statistically differe nt (P value = 0.812, by one-way ANOVA) (Figure 5-23C). Again to account for inter-litter variab ility, diffuse amyloid burdens, average diffuse deposit sizes, and average number of diffuse deposits per section from pGFP and pUFGL-injected mice were divided by the values obtained from their respective untreated littermates (Figure 5-25). No di fferences were detected; however a trend toward a reduction in diffuse amyloid burden fractions was seen, pGFP-injected mice had a mean diffuse amyloid burden fraction of 1.00 0.13, and pUFGL-injected mice had a mean diffuse amyloid burden fraction of 0.72 0.10 (P = 0.071, by unpaired one-tailed ttest). No trends were detected in diffuse deposit size fraction (pGFP-injected mice had a mean diffuse deposit size fraction of 1.06 0.06, pUFGL-injected mice had a mean diffuse deposit size fraction of 1.06 0.16) a nd average number of diffuse deposits per section fraction (pGFP-injected mice had an average number of diffuse deposits per section fraction of 1.13 0.18, pUFGL-injected mice had had an average number of diffuse deposits per section fraction of 0.86 0.11). Microglia, Soluble Oli gomers, and Astrocytes The state of microglia and the degree of soluble amyloid oligomers in huAPP/PS1M146L brain sections were examined by i mmunohistochemistry (Figure 5-26). Three sections per animal were examined and all sections were incubated for equal amounts of time during the labeling procedure. DAB reactions were carried out on slides containing one section from each animal, in or der to minimize differences in reaction

PAGE 85

73 times. Digital micrographs were prepared of the whole slides w ith a constant light source, and relative grey values were meas ured from negative images for each section using NIH Image J software, to determine stai ning intensity. Astr ocytic activation was examined in huAPP/PS1M146L brain sections by immunostaining for glial fibrillary acidic protein (GFAP). Two 50mm thick sections pe r animal were analyzed for total percent stained area of the hippocampus and co rtex using NIH Image J software. For microglia, untreated mice had a mean SEM relative negative grey value of 27.5 4.8, pGFP-injected mice had 29.2 3.3, and pUFGL-injected mice had 46.53 16.0. These values did not differ significantly (Figure 5-27A). When examined in terms of fractions of their untreated littermates, pGFP-injected mi ce had a mean microglia stain fraction of 1.32 0.05 and pUFGL-injected mi ce had a mean microglia stain fraction of 1.84 0.60. These values values were also found not to be different statistically (Figure 5-27B) When soluble amyloid oligomers were examined, untreated mice had a mean SEM relative negative grey value of 85.8 4.7, pGFP-injected mice had 85.9 1.3, and pUFGL-injected mice had 87.1 8.5. These values did not differ significantly (Figure 5-27C). When examined in terms of fractions of their untreated littermates, pGFP-injected mice had a mean soluble am yloid oligomer stain fraction of 0.93 0.01 and pUFGL-injected mice had a mean soluble amyloid oligomer stain fraction of 1.23 0.18. These values were also found not to be different statistically although a trend was evident (Figure 5-27D). Staining for astrocytes using GFAP rev ealed that untreated mice had 16.7 1.4 % of their hippocampus and neocortex st ained, pGFP-injected mice had 13.5 0.89 %

PAGE 86

74 stained, and pUFGL had 18.2 2.8 % stained. These values di d not differ statistically (Figure 27E). When examined in terms of fr actions of their untreated littermates, pGFPinjected mice had a mean soluble amyloid oligomer stain fraction of 0.96 0.04 and pUFGL-injected mice had a mean soluble am yloid oligomer stain fraction of 1.26 0.16. These values were also found not to be diffe rent statistically although a trend was evident (Figure 5-27F). Regression Analysis Linear regression was employed to exam ine possible correlations in pUFGLinjected mice between the fractions of untr eated littermate soluble amyloid oligomer staining, microglia staining, or % GFAP stai ning, and fraction of untreated littermate A 42 concentrations, diffuse or dense-core amyloid burdens. Slopes trended toward being significantly different from zero when soluble oligomers fractions (P = 0.079) and microglia fractions (P = 0.062) were comp ared with values for fractions of A 42 concentrations (Figure 5-28A & B). Slopes also showed a strong trend toward being significantly different from zero when percent GFAP stain fractions were compared with dense-cored amyloid burden fractio ns (P = 0.084, Figure 28C). When percent GFAP stain fractions were compared with fractions of diffuse amyloid burden, slope was significantly diffe rent than zero (P = 0.008, Figure 28D), indicating a strong correlation fo r an increase in the % area stained for GFAP compared to untreated littermates, with a decrease in diffuse amyloid burden compared to untreated littermates in pUFGL-injected mice.

PAGE 87

75 Conclusion In this study we examined the effects of peripheral plasma gelsolin expression on two different transgenic mouse models of Al zheimer’s disease. Two different models were used due to a failure in our huAPP/PS1 L breeding colony. However, testing two separate models has advantages. Two separate models lets us examine gelsolin expression in animals that produce amyloid at different rates (huAPP/PS1 L, faster; mo/huAPP/PS1 E9, slower). They also allow us to te st effects of gelsolin expression with mice of different genetic backgrounds. Studying the differences in response to gelsolin expression between the different models can allow us to make inferences on the efficiency and mechanisms of actions of plasma gelsolin expression may have. These results demonstrate that both mo/huAPP/PS1 E9 mice and huAPP/PS1 L mice express pUFGL in the liver for at least tw o and a half weeks after gene delivery by hydrodynamic injection (Fi gures 5-1 & 5-16). In mo/huAPP/PS1 E9 mice expression of pUFGL results in a significant decrease in the concentration of total brain amyloid 1-42 (Figure 5-2). The decreased concentration of A 42 is accompanied with a signifi cant decrease in the dense-cored amyloid deposit load when compared to untreated mice in mo/huAPP/PS1 E9 (Figure 5-4A). This reduction of dense-cored deposit load is accompanied with a significant decrease in number of both dense-cored and diffuse depos its observed per secti on of pUFGL-injected mice (Figure 5-4D & 5-5D). Surprisingly, th e reduction in dense cored deposit load in mo/huAPP/PS1 E9 mice was not accompanied by signifi cant reductions in diffuse deposit loads (Figure 5-5A). However, a strong trend toward significance was observed.

PAGE 88

76 Our data indicates that the reduction of amyloid deposits may be associated with an increase in the apparent st aining of soluble amyloid oligomers (Figure 5-9) and an increase in the apparent activation st ate of microglia in the brains of mo/huAPP/PS1 E9 mice injected with pUFGL (Figures 5-12 & 13). Although statistically significant, our results may just be an artifact due to va riability of tissue treatment post-mortem. However an increase in oligomer staining coul d explain an increase in the activation state of microglia. It has been reported that soluble amyloid and not insoluble amyloid activates microglia (Floden and Combs, 2006; Lindberg et al., 2005). Activated microglia have been implicated of playi ng a role in antibody-mediated clearance of dense-cored amyloid deposits in transgenic mice (Wilcock et al., 2003). So it would not be surprising if microglia play a similar ro le in gelsolin mediated clearance of densecored deposits in our study also. This woul d be a significant observation because before clearance by peripheral amyloi d binding agents was thought to be a result of mass action diffusion according to the pe ripheral sink hypothesis. A decrease in total brain amyloid 1-42 was also observed in huAPP/PS1 L mice when inter-litter variability is accounte d for (Figure 5-18). In contrast to the mo/huAPP/PS1 E9 group, huAPP/PS1 L did not show a significant reduction in dense cored amyloid deposit load (Figure 5-21), nor was there a difference in diffuse amyloid deposit load (Figure 5-24). However both di splayed a strong trend towards a reduction in amyloid load. No differences were eviden t among deposit size, but strangely when interlitter variability was accounted for there wa s a significant increase in the dense-cored deposit number fraction (Figure 5-21).

PAGE 89

77 There were also no differences observed in amyloid oligomers staining intensity, microglia staining intensity, or % GFAP stained. Only slight trends were seen when inter-litter variability were accounted for (Fi gure 5-26). The possibility of variability existing to the response of peripheral human plasma gene expression motivated us to examine if there were any correlations amyloid pathology and glial response. When correlations were examined (Figure 5-27) str ong trends were eviden t indicating a possible correlation between decreasing total amyloid 42 concentrations and increasing soluble oligomer or microglia staining intens ity compared to untreated littermates. A strong correlation was observed between decreases in diffuse amyloid deposits and increases in the % area stained for GFAP indicating astrocytes may be involved in the clearance of diffuse amyloid pathology. Evidence supporting this hypothesis has been observed in post-mortem studies of non-demented humans (Funato et al., 1998; Yamaguchi et al., 1998). Adult mouse astroc ytes have also been observed to migrate toward and degrade immobilized amyloid in vitro and in situ in transgenic mouse brain sections (Wyss-Coray et al., 2003).

PAGE 90

78 Figure 5-1 Detecting message in mo/huAPP/PS1 E9 mice. RT-PCR using primers specific for mRNA coded from pUFGL vector, us ing RNA purified from liver tissue from non-injected mice (– symbols), liver tissue from pUFGL-injected mice (+ symbols), and 293 cells transfected with pUFGL at a high or low confluency. No RNA refers to a pr oduct from a reaction mixture with no RNA. 500bp marker fr om Biorad was used. Untreate d p UFG L 0 1000 2000 300 0 P = 0.0222picomole/g of brain tissu e Figure 5-2 Amyloid (1-42) concentrations in mo/huAPP/PS1 E9 brains. Diamonds represent individual animal means, bars represent means and standard error. Untreated mice had a mean SEM A (1-42) concentration of 2,306 202.6 pmoles/g of brain tissue, and the in jected animals had 1,174 334.7 pmoles/g of brain tissue. A one-t ailed t-test was performe d and showed that these groups differ significantly with a P value = 0.0222. _ + + + pUFGL Low pUFGL High No RNA 500 b.p. marker

PAGE 91

79 Figure 5-3 Plasma concentrations of amyloid in mo/huAPP/PS1 E9 mice. A) Amyloid 40 Concentrations in fmol/mL. B) Amyloid 42 Concentrations in fmol/mL. C) The Ratio of Amyloid 40 to Amyloid 42 Concentrations.

PAGE 92

80 Untreated pUFGL 0.00 0.25 0.50 0.75 1.00 1.25A P = 0.0085Percent Amyloid Burden (Thioflavin S) Untreated pUFGL 0 25000 50000 75000 100000B P = 0.010Pixels2 Untreated pUFGL 20 25 30 35 40 45 50C P = 0.091Pixels2 Untreated pUFGL 0 100 200 300 400 500 600 700D P = 0.029Number of Plaques Per Section Figure 5-4 Analysis of dens e-core amyloid deposits in mo/huAPP/PS1 E9 mice. Deposits visualized by thioflavin S staining of brain sections. A) Percent Amyloid Burden. B) Total Stained Area. C) Average Deposit Size. D) Average Number of Deposits per section. Di amonds represent individual animal means. Bars represent group means the standard error.

PAGE 93

81 Untreated pUFGL 0.00 0.25 0.50 0.75 1.00 1.25 1.50 1.75 2.00A P = 0.074% Amyloid Burden (6E10) Untreated pUFGL 0 25000 50000 75000 10000 0 B P = 0.078Pixels2 Untreated p UFG L 0.0 2.5 5.0 7.5 10.0 12.5 15.0 17.5 20.0 22.5 25.0C P = 0.387Pixels2 Untreated pUFGL 250 500 750 1000 1250 1500D P = 0.035Number of Plaques/Section Figure 5-5 Analysis of di ffuse amyloid deposits in mo/huAPP/PS1 E9 mice. Deposits visualized by immunostaining of brain se ctions. A) Percent Amyloid Burden. B) Total Stained Area. C) Average Deposit Size. D) Average Number of Deposits per section. Diamonds represent individual animal means. Bars represent group means the standard error.

PAGE 94

82 Figure 5-6 Dense-core amyloid deposits in mo/huAPP/PS1 E9 mice. Deposits visualized by thioflavin S Staining. Negative digital micrographs of untreated mice (left) and pUFGL-injected mice (right). One section is included from each animal.

PAGE 95

83 Figure 5-7 Diffuse amyloid deposits in mo/huAPP/PS1 E9 mice. Deposits visualized by immunostaining with 6E10, digital microgr aphs of untreated mice (left) and pUFGL-injected mice (right). One se ction is included from each animal.

PAGE 96

84 Figure 5-8 Side by side comparison of staini ng for dense cored amyloid deposits, diffuse amyloid deposits, soluble amyloid o ligomers, microglia, and GFAP in mo/huAPP/PS1 E9 mice. Untreated are lower sections, pUFGL are upper.

PAGE 97

85 Figure 5-9 Staining intensity of soluble amyloid oligomers in mo/huAPP/PS1 E9 mice. Increasing negative grey values corre spond to darker sections. Diamonds represent individual animal means determ ined from measurements from three sections, while bars represent group means SEM. Figure 5-10 High magnification tile d images of amyloid oligom er staining. Untreated on the left, and pUFGL-injected on th e right. Scale bar represents 100 m for both images. p UFGL Untreate d

PAGE 98

86 Figure 5-11 Cell types stained for soluble amyl oid oligomers in pUFGL-injected mice. Upper photo displays an as trocyte containing soluble amyloid oligomers. Lower photo displays vasculature staine d for soluble amyloid oligomers. Deposits in the lower photo were counterstained with congo red and visualized through a polarized light source. Scale bar represents 25 m for both images.

PAGE 99

87 Figure 5-12 Staining intensity of microglia in mo/huAPP/PS1 E9 mice. P value calculated from an unpaired one-tailed t test.

PAGE 100

88 Figure 5-13 Low magnification images of mi croglia. Untreated (left), and pUFGLinjected (right) one image represented from each animal in the treatment groups (scale bar represents 100 m).

PAGE 101

89 Figure 5-14 High magnification images of microglia staining. Untreated animal (right)and pUFGL-injected animal (lef t) are counterstain ed with Congo Red. (scale bar represents 25 m). Untreate d p UFG L 0 2 4 6 8 10 12 14 16 18P = 0.17 2 % Area stained for GFAP Figure 5-15 Average GFAP percent stained area in mo/huAPP/PS1 E9 cortex and hippocampus. P value calculated from an unpaired one-tailed t test.

PAGE 102

90 Figure 5-16 High magnification images of astrocytes surrounding congo red positive amyloid deposits. Untreated mice on the left and pUFGL-injected mice on the right. Scale bar represents 30 m for all images. Figure 5-17 Message detection in huAPP/PS1M146L mice. Specific for mRNA coded from pUFGL vector, using RNA purified from liver tissue from non-injected mice (– symbols), liver tissue from pGFP-inj ected mice (+ symbols), or liver tissue from pUFGL-injected mice (+ symbol s), and 293 cells transfected with pUFGL 500 basepair marker from Biorad was used. pGFP pUFGL – + – – – – – – – – + + + – + + + – – – – + – – 293Cells 500 b.p. Marker

PAGE 103

91 Untreated pGFP pUFGL 0 2500 5000 7500 10000 12500A P = 0.069picomole/g of brain tissu e p GFP p UFG L 0.35 0.55 0.75 0.95 1.15B P = 0.022 Figure 5-18 Amyloid (1-42) cincentrations in huAPP/PS1M146L brains. Amyloid (1-42) concentrations (A). Fraction of Amyloid (1-42) concentrations of untreated littermates (B). Diamonds represent i ndividual animal means, bars represent means and standard error, P values calculated from one-way ANOVA (A) and an unpaired one tailed t-test (B).

PAGE 104

92 Untreated pGFP pUFGL 0 500 1000 1500 2000 A P = 0.960A 1-40 (fmol/ml) Untreated pGFP pUFGL 0 100 200 300 400 500 600 700 800 900 B P = 0.731A 1-42 (fmol/ml) Untreated pGFP pUFGL 0 1 2 3 C P = 0.933A 40/42 ratio Retro -orbital Cardiac Puncture 0 500 1000 1500 2000 D P = 0.016A 1-40 (fmol/ml) Retro -orbital Cardiac Puncture 0 100 200 300 400 500 600 700 800 900 E P = 0.057A 1-42 (fmol/ml) Figure 5-19 Plasma concentrations of amyloid in huAPP/PS1M146L mice. A) Amyloid 40 Concentrations in fmol/mL. B) Amyloid 42 Concentrations in fmol/mL. C) The Ratio of Amyloid 40 to Amyloid 42 Concentrations. D) Amyloid 40 and E) Amyloid 42 Concentrations in fmol/mL in terms of sample collection method. P values determined oneway ANOVA or unpaired two-taile d t-test where appropriate.

PAGE 105

93 Figure 5-20 Dense-core amyloid deposits in huAPP/PS1 L mice. Deposits visualized by thioflavin S Staining. Negative di gital micrographs of untreated mouse (top), pGFP-injected mouse (middle) and pUFGL-injected mouse (bottom). Untreated pGFP pUFGL

PAGE 106

94 Untreated pGFP pUFGL 0.0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 1.6 1.8A P = 0.069Percent Amyloid Burden Untreated pGFP pUFGL 0 25000 50000 75000 100000 125000 150000 175000B P = 0.12 4 Pixels2 Untreate d p GFP p UFGL 0 25 50 75 100 125 150C P = 0.247Pixels2 Untreated pGFP pUFGL 0 250 500 750 1000D P = 0.499Number of Plaques Per Section Figure 5-21 Analysis of densecore amyloid deposits in huAPP/PS1 L mice. Deposits visualized by thioflavin S staining. A) Percent Amyloid Burden. B) Total Stained Area. C) Average Deposit Size. D) Average Number of Deposits per section. Diamonds represent individual animal means. Bars represent group means the standard error. P-values determined by ANOVA

PAGE 107

95 pGFP pUFGL 0.5 0.7 0.9 1.1 1.3 A P = 0.968Fraction of Dense-Cored Amyloid Burden Compared to Untreated Littermates pGFP pUFGL 0.00 0.25 0.50 0.75 1.00 1.25 1.50B P = 0.590Fraction of Dense-Cored Amyloid Burden Compared to Untreated Littermates pGFP pUFGL 0.30 0.55 0.80 1.05 1.30 1.55C P = 0.048Fraction of Dense-Cored Deposit Number Compared to Untreated Littermates Figure 5-22 Analysis of dense-core amyloid de posits in terms of fractions compared to untreated littermates in huAPP/PS1 L mice. Deposits visualized by thioflavin S staining. A) Percent Amyl oid Burden. B) Average Deposit Size. C) Average Number of Deposits per s ection. Diamonds represent individual animal means. Bars represent group means the standard error. P-value determined by two-tailed t-test.

PAGE 108

96 Figure 5-23 Diffuse amyloid deposits in huAPP/PS1 L mice. Deposits visualized by immunostaining with 6E10. Digital mi crographs of untreated mouse (top), pGFP-injected mouse (middle) and pUFGL-injected mouse (bottom).

PAGE 109

97 Untreated pGFP pUFGL 0 1 2 3 4 5 6 7 A P = 0.099Percent Diffus e Amyloid Burden Untreated pGFP pUFGL 12.5 17.5 22.5 27.5 32.5 E P = 0.271Pixels2 Untreated pGFP pUFGL 0 1000 2000 3000 4000 5000 6000 C P = 0.812Number of Diffuse Plaques Per Section Figure 5-24 Analysis of di ffuse amyloid deposits in huAPP/PS1 L mice. Deposits visualized by immunostaining. A) Pe rcent Amyloid Burden. B) Average Deposit Size. C) Average Number of De posits per section. Circles represent individual animal means. Bars repres ent group means the standard error. Pvalues determined by one-way ANOVA.

PAGE 110

98 pGFP pUFGL 0.25 0.50 0.75 1.00 1.25 A P = 0.071Fraction of Untreated Littermate Diffuse Plaque Burden pGFP pUFGL 0.50 0.75 1.00 1.25 1.50B P = 0.498Fraction of Untreated Littermate Diffuse Plaque Size pGFP pUFGL 0.50 0.75 1.00 1.25 1.50Fraction of Untreated Littermate Diffuse Plaque NumberC P = 0.110 Figure 5-25 Analysis of diffuse amyloid depos its in terms of fract ions compared to untreated littermates in huAPP/PS1 L mice. Deposits visualized by immunostaining. A) Percent Amyloid Bu rden. B) Average Deposit Size. C) Average Number of Deposits per section. Diamonds represent individual animal means. Bars represent group means the standard error. P-values determined by one-tailed t-test.

PAGE 111

99 Figure 5-26 Side by side co mparison of staining in huAPP/PS1M146L mice for dense cored amyloid deposits, soluble amyloid oligomers, microglia, and GFAP positive cells in huAPP/PS1 L Mice. Untreateds represent littermates for both pGFP-injected animal (upper) and pU FGL-injected animal (lower).

PAGE 112

100 Untreated pGFP pUFGL 0 25 50 75 100 A P =0.357Relative Negative Grey Values for Microglia pGFP pUFGL 0 1 2 3 4 5 B P = 0.248Fraction of Microglia Staining Untreated pGFP pUFGL 50 60 70 80 90 100 110 C P =0.985Relative Negative Grey Values For Soluble Amyloid Oligomers pGFP pUFGL 0.9 1.0 1.1 1.2 1.3 1.4 1.5 1.6 1.7 1.8 D P = 0.104Fraction Soluble Amyloid OligomerStaining Untreated pGFP pUFGL 10.0 12.5 15.0 17.5 20.0 22.5 25.0 27.5 E P = 0.375% GFAP Stained pGFP pUFGL 0.8 0.9 1.0 1.1 1.2 1.3 1.4 1.5 1.6 1.7 F P = 0.087GFAP Fraction Figure 5-27 Analysis of microglia, soluble amyloid oligomers, and astrocytes in huAPP/PS1M146L mice. A) Microglia. B) Mi croglia examined in terms of fractions of untreated litte rmates. C) Soluble amyloid oligomers. D) Soluble amyloid oligomers in terms of fractions of untreated littermates E) Astrocytes. F) Astrocytes in terms of fractions of untreated littermates. Diamonds represent individual animal means, and bars represent group means SEM. P-values determined one-way ANOVA or unpaired one-tailed t-test where appropriate.

PAGE 113

101 Figure 5-28 Linear regression analysis. Correlating solubl e amyloid olig omer (A) and microglia stain fractions (B) to A 42 concentration fractions, and % GFAP stain fractions with dense-cored (C) a nd diffuse (D) amyloid burden fractions for pUFGL-injected huAPP/PS1M146L mice. Dashed lines indicate a 95% confidence interval for the fitted lines.

PAGE 114

102 CHAPTER 6 DISCUSSION AND FUTURE DIRECTIONS Previous studies have demons trated that human gelsolin purified from plasma has the ability bind to amyloid (Chauhan et al., 1999). It has also been reported that human plasma gelsolin disassembles and prevents the assembly of amyloid fibrils (Ray et al., 2000). Matsouka et al. showed that administ ration of bovine plasma gelsolin reduces amyloid levels in huAPP/PS1 L mice (2003). We attempted to compare bovine and human plasma gelsolin’s interaction with amyloid due to these findings. In our attempt to measure binding affinities for both the human and bovine forms of plasma gelsolin, by three se parate experimental approaches we discovered that in the presence of plasma gelsolin (both human and bovine) amyloid is less ‘sticky’. This is an important finding indicating that rather than gelsolin just having binding activity towards amyloid it appears to have enzymatic type activity, perhaps by maintaining amyloid in a soluble -helix state, rather than oligomerization prone -sheet state. We compared the relative activity for both huma n and bovine plasma gelsolin to disassemble preformed amyloid fibrils, and our results confirm that both huma n and bovine plasma gelsolin have fibril disassembly activity, and th at the human form is as efficient if not more so, as the bovine form. We hypothesized that human plasma gelsol in would make an excellent candidate for a peripherally expressed gene therapy fo r AD based on several observations. Plasma gelsolin contains an internal signal peptide allowing for cellular export into the plasma

PAGE 115

103 from peripheral expressing tissues. Plasma gelsolin has been shown to bind to and disassemble preformed A fibers and prevent fibril lization of soluble amyloid (Chauhan et al., 1999; Ray et al., 2000). Gelsolin concentrations appear to be reduced in the CSF of Alzheimer’s patients (Chauhan, 2004). Administration of plasma gelsolin at a dose of 0.6 mg/kg every two days for three w eeks significantly reduced the accumulation of deposits in young amyloid-depositing tran sgenic mice (Matsuoka et al., 2003). To test this hypothesis we construc ted a plasmid DNA mammalian expression vector, pUFGL. Our studies provide eviden ce that pUFGL does produce a human plasma gelsolin immunoreactive protein of the same mo lecular weight of plasma gelsolin when expressed by human embryonic kidney cells in vitro We also demonstrate that the gene product recombinant human plasma gelsolin will bind to amyloid 1-42. We are the first to demonstrate that recombinant plasma gelsolin maintains this activity. We choose to use the hydrodynamic gene deli very technique to deliver our plasmid DNA. Our results confirm that delivering vect ors by this method results in high levels of gene expression, which is wide ly distributed throughout the li ver for at least two and a half weeks. Little if any damage or infl ammation is observed as a result of the gene delivery and expression results in the liver. We tested our vector in two separate m ouse models of Alzheimer’s disease, and saw significant reductions of amyloid 1-42 concentrations in the brains of mo/huAPP/PS1 E9 mice, and a trend toward significant reductions in huAPP/PS1 L mice that was significant when inter-litter variability was accounted for. Only mo/huAPP/PS1 E9 mice had significant reductions in dense cored amyloid deposit load. However, huAPP/PS1 L mice did trend toward a decrease in dense cored amyloid deposit load; it

PAGE 116

104 may be that a longer treatment regimen will result in a significant decrease in amyloid 1-42 concentrations. This difference between the two lines of transgenic mice may be accounted for by the fact that in huAPP/PS1 L mice there was only about a 20% reduction in amyloid 1-42 concentration, while the mo/huAPP/PS1 E9 mice had nearly a 50% reduction. This difference in plasma gelsolin’s apparent abil ity to clear amyloid from the brains of the two lines of mice may not be a function of gels olin working better in one line rather than the other, but result from th e fact that eight month old huAPP/PS1 L mice have an amyloid 1-42 concentration that is nearly five times that of nine month old mo/huAPP/PS1 E9 mice, which is reflective of a much higher rate of amyloid synthesis in huAPP/PS1 L mice as compared to mo/huAPP/PS1 E9 mice. Likewise the huAPP/PS1 L mice have a much higher dense cored amyloid deposit burden at the beginning of treatment, and our data is cons istent with the commonly held belief that it should be harder to remove deposits that alr eady exist, rather than prevent new deposits from forming (Levites et al., 2006). The mechanism for this brain amyloi d reduction is not clear. Peripheral sequestration is a possible explanati on; however, increases in plasma A often observed with immunization (DeMattos et al., 2001; De Mattos et al., 2002; Lemere et al., 2003) or other A binding strategies (Deane et al., 2003; Ma tsuoka et al., 2005) are not detected in this study, agreeing with the previously reported study by Matsuoka and colleagues (Matsuoka et al., 2003). However, it is possi ble that there were increases in plasma A at earlier time points that were not de tected in this experiment.

PAGE 117

105 Unlike Matsouka’s study we did observe a decrease in the ratio of amyloid 140/1-42. Normally amyloid depositing transgen ic mice have much higher concentrations of amyloid 1-40 than amyloid 1-42 in the plasma (DeMatto s et al., 2002; Lanz et al., 2005; Matsuoka et al., 2003). It has also b een reported that in hu mans a decrease the ratio of plasma amyloid 1-40/1-42 is assoiciated with an decreased risk for Alzheimer’s disease (van Oijen et al., 2006) This observation may indicat e that higher concentrations of amyloid 1-42 in the plasma are a reflection of amyloid 1-42’s ability to be cleared from the CNS. Our observation that amyloid 1-40/1-42 ratio is reduced in mo/huAPP/PS1 E9 mice may be associated with the act ivity described in this report of gelsolin’s ability to solubilize amyloid It is possible that plasma gelsolin en ters the parenchyma across the blood brain barrier and directly solubili zes amyloid deposits. Transport mechanisms for gelsolin across the BBB are not known, though reports in dicate that the BBB is compromised in mouse models of AD (Dickstein et al., 2006; Marco and Skaper, 2006). Gelsolin efficiently disassembles preformed amyloid fi brils (Ray et al., 2000), so this seems a feasible hypothesis. The possible increase of soluble amyloid oligomers observed in gelsolin expressing mice in this study suppor ts this hypothesis. The increase in soluble amyloid could be an explanation for the in crease in reactive and activated microglia observed in our study. It has been reporte d that soluble amyloi d and not insoluble amyloid activates microglia (Floden and Combs, 2006; Lindberg et al., 2005). An alternative to gelsolin entering the pa renchyma and solubilizing amyloid is the activity we observed for gelsolin to shift the equilibrium of amyloid from forming fibrillar insoluble forms to monomeric or dimeric soluble forms of amyloid across a

PAGE 118

106 membrane (Figure 3-6). Our data supports both the hypothesis of gelsolin entering the CNS and disassembling aggregates directly or the hypothesis gelsolin is not entering the CNS and just shifting the equilibrium of amyloid to a more soluble state. A third possible explanation for our results is hypothesis that plasma gelsolin may have immunomodulatory functions based on its ability to bind a number of immunomodulatory chemicals (Bucki et al ., 2005; Bucki et al., 2004; Chauhan et al., 1999; Lee et al., 2006; Lind and Janmey, 1984; Smith et al., 1987), since it has been shown to decrease the toxicity associated with inflammation after injury or trauma (Candiano et al., 2005; Christ ofidou-Solomidou et al., 2002a; Christofidou-Solomidou et al., 2002b; DiNubile et al., 2002; Mounzer et al., 1999). The roles of microglial and astrocytic activation during Alzheimer’s are ac tively being studied. It has been proposed that in the presence of A glia become activated and s ecrete neurotoxic cytokines and chemicals (Dheen et al., 2005; Haas et al., 2002; von Bernhardi and Eugenin, 2004; Walker and Lue, 2005). It has also been proposed that glia cells can serve neuroprotective roles. They have been shown to produce neurotrophins in response to injury or disease (Dougherty et al., 2000; Nakajima et al., 2001). Glia and immune cell production of brain derived neurotrophic factor (BDNF) has been shown to be important in hippocampal neurogenesis and spatial memory related be havior (Ziv et al., 2006). Astrocytes, microglia and macrophages have also been shown to internalize A by phagocytosis; this function appears to be deficient in AD (Fiala et al., 2005) and amyloid depositing transgenic mice (Alarcon et al ., 2005; Fiala et al., 2005; Roge rs et al., 2002). Stimulating the immune system either by active or passive immunization against A or

PAGE 119

107 lipopolysaccharide administration restores phagocytosis of A by microglia (Das et al., 2003; DiCarlo et al., 2001; Herb er et al., 2004; Wilc ock et al., 2003). If human gelsolin expression does affect amyloid by stimulati ng the immune system, it is most likely not the result of a global resp onse to the expression of a foreign protein. If this were the case, GFP expression would cause similar effects and none are observed. Also, hematoxylin and eosin staining of liver se ctions reveals no differences between any of the treatment groups at the 18 day time point. Gene therapy intends to treat, cure, or prevent diseases with the use of nucleic acid sequences. The vast majority of gene th erapy approaches for AD involve delivery of vectors coding for either neurotrophic f actors (nerve growth factor, NGF)(Bradbury, 2005; Klein et al., 2000; Wu et al., 2004), or amyloid degrad ing enzymes (neprilysin or insulin degrading enzyme) (Eckman and Eckm an, 2005; Marr et al., 2003). The rationale behind delivering neurotrophic fact ors is to protect the neuron al populations that are lost due to AD pathology, and the rationale behind using amyloid degrading enzymes is to increase the clearance of amyloid. Thes e approaches are promising; however both involve invasive surgery that requires an injection directly into the brain, which does not come without risks. Treating AD through a peripherally admi nistered gene therapy would be advantageous in this regard. NGF is also a secreted protein; however, peripheral expression is not a suitable treatment strate gy due to significant side effects observed after systemic administration (P etty et al., 1994). In contrast because IDE activity is predominantly cytosolic (Aff holter et al., 1990; Gao et al., 2004), and NEP is bound to the plasma membrane (Back and Gorenstein, 1990), these are not lik ely to be suitable

PAGE 120

108 agents for peripheral therapy, however may prove to be effective if expressed locally in the brain (Eckman and Eckman, 20 05; Leissring et al., 2003). Future directions for this study will in clude determining if plasma gelsolin expressed in the brains has similar effects on amyloid and glia in the CNS of transgenic mice; also determine how much gelsolin can cross the blood brai n barrier from the periphery. These studies will also examine behavior associated with memory and learning. Studies with multiple treatments at di fferent time points will be important for learning how fast gelsolin works and for how long. In vitro studies will be carried out to determine if plasma gelsolin affects the ab ility of primary cultures of microglia and astrocytes to phagocytose amyloid and determine by what cellular mechanism gelsolin has this effect. In conclusion our study demonstrates that a peripherally delivered and expressed gene therapeutic can affect amyloid dynamics in the central nervous system. Peripheral delivery offers significant advantages over the traditional gene therapeutic approaches for AD in that no invasive surgeries are requi red, thus reducing cost s and complications. Plasma gelsolin is well suited as a peripherally expressed therapeutic in that it contains an internal signal peptide directi ng gelsolin to be secreted from the cells that make it at any peripheral location. Care should be taken using plasma gels olin as a treatment strategy as we observed an increase in microglial activ ation and soluble oligomeric forms of A both of which are thought to be a possible source of neurotoxicity.

PAGE 121

109 LIST OF REFERENCES Affholter, J. A., Hsieh, C. L., Francke, U ., and Roth, R. A. (1990). Insulin-degrading enzyme: stable expression of the human complementary DNA, characterization of its protein product, and chromosomal ma pping of the human and mouse genes. Mol Endocrinol 4 1125-35. Akama, K. T., Albanese, C., Pestell, R. G ., and Van Eldik, L. J. (1998). Amyloid betapeptide stimulates nitric oxide produc tion in astrocytes through an NFkappaBdependent mechanism. Proc Natl Acad Sci U S A 95 5795-800. Alarcon, R., Fuenzalida, C., Santibanez, M., an d von Bernhardi, R. (2005). Expression of scavenger receptors in glial cells. Co mparing the adhesion of astrocytes and microglia from neonatal rats to surface-bound beta-amyloid. J Biol Chem 280 30406-15. Alino, S. F., Crespo, A., and Dasi, F. (2003) Long-term therapeutic levels of human alpha-1 antitrypsin in plasma afte r hydrodynamic injection of nonviral DNA. Gene Ther 10 1672-9. Allsop, D., Landon, M., and Kidd, M. (1983). The isolation and amino acid composition of senile plaque core protein. Brain Res 259 348-52. Back, S. A., and Gorenstein, C. (1990) Localization of ne utral endopeptidase-24.11 (enkephalinase) to N-methyl -D-aspartate sensitive cel l populations in the rat forebrain. Prog Clin Biol Res 328 199-202. Bacskai, B. J., Kajdasz, S. T., Christie, R. H., Carter, C., Games, D., Seubert, P., Schenk, D., and Hyman, B. T. (2001). Imaging of amyl oid-beta deposits in brains of living mice permits direct observation of cl earance of plaques with immunotherapy. Nat Med 7 369-72. Bard, F., Cannon, C., Barbour, R., Burke, R. L ., Games, D., Grajeda, H., Guido, T., Hu, K., Huang, J., Johnson-Wood, K., Kh an, K., Kholodenko, D., Lee, M., Lieberburg, I., Motter, R., Nguyen, M., So riano, F., Vasquez, N., Weiss, K., Welch, B., Seubert, P., Schenk, D., and Yednock, T. (2000). Peripherally administered antibodies against amyloid beta-peptide enter th e central nervous system and reduce pathology in a mouse model of Alzheimer disease. Nat Med 6 916-9. Basler, K., Oesch, B., Scott, M., Westaway, D ., Walchli, M., Groth, D. F., McKinley, M. P., Prusiner, S. B., and Weissmann, C. (1986). Scrapie and cellular PrP isoforms are encoded by the same chromosomal gene. Cell 46 417-28.

PAGE 122

110 Borchelt, D. R., Davis, J., Fischer, M., Lee, M. K., Slunt, H. H., Ratovitsky, T., Regard, J., Copeland, N. G., Jenkins, N. A., Sis odia, S. S., and Price, D. L. (1996). A vector for expressing foreign genes in the brains and hearts of transgenic mice. Genet Anal 13 159-63. Borchelt, D. R., Ratovitski, T., van Lare, J., Lee, M. K., Gonzales, V., Jenkins, N. A., Copeland, N. G., Price, D. L., and Sis odia, S. S. (1997). Accelerated amyloid deposition in the brains of transgenic mice coexpressing mutant presenilin 1 and amyloid precursor proteins. Neuron 19 939-45. Bradbury, J. (2005). Hope for AD with NGF gene-therapy trial. Lancet Neurol 4 335. Brendza, R. P., Bales, K. R., Paul, S. M., and Holtzman, D. M. (2002). Role of apoE/Abeta interactions in Alzheimer's disease: insights from transgenic mouse models. Mol Psychiatry 7 132-5. Brunden, K. R., Richter-Cook, N. J., Chaturve di, N., and Frederickson, R. C. (1993). pHdependent binding of synthetic beta-amylo id peptides to glycosaminoglycans. J Neurochem 61 2147-54. Bucki, R., Georges, P. C., Espinassous, Q., Funaki, M., Pastore, J. J., Chaby, R., and Janmey, P. A. (2005). Inactivation of endotoxin by human plasma gelsolin. Biochemistry 44 9590-7. Bucki, R., Pastore, J. J., Randhawa, P., Vegne rs, R., Weiner, D. J., and Janmey, P. A. (2004). Antibacterial activities of rh odamine B-conjugated gelsolin-derived peptides compared to those of the an timicrobial peptides cathelicidin LL37, magainin II, and melittin. Antimicrob Agents Chemother 48 1526-33. Candiano, G., Bruschi, M., Pedemonte, N., Caci E., Liberatori, S., Bini, L., Pellegrini, C., Vigano, M., O'Connor, B. J., Lee, T. H ., Galietta, L. J., and Zegarra-Moran, O. (2005). Gelsolin secretion in interleukin-4-treated br onchial epithelia and in asthmatic airways. Am J Respir Crit Care Med 172 1090-6. Chauhan, V. P. (2004). Anti-amyloidogenic Role of Gelsolin in Alzheimer's Disease. In The 9th International Conference on Alzhei mer's Disease and Related Disorders." (A. Chauhan, ed.). Chauhan, V. P., Ray, I., Chauhan, A., and Wisn iewski, H. M. (1999). Binding of gelsolin, a secretory protein, to amyloid beta-protein. Biochem Biophys Res Commun 258 241-6. Chen, C. D., Huff, M. E., Matteson, J., Page, L., Phillips, R., Kelly, J. W., and Balch, W. E. (2001). Furin initiates gelsolin fam ilial amyloidosis in the Golgi through a defect in Ca(2+) stabilization. Embo J 20 6277-87.

PAGE 123

111 Chhabra, D., Nosworthy, N. J., and dos Remedios, C. G. (2005). The N-terminal fragment of gelsolin inhibits the interac tion of DNase I with isolated actin, but not with the cofilin-actin complex. Proteomics 5 3131-6. Christofidou-Solomidou, M., Scherpereel, A., Solo mides, C. C., Christie, J. D., Stossel, T. P., Goelz, S., and DiNubile, M. J. (2002a). Recombinant plasma gelsolin diminishes the acute inflammatory response to hyperoxia in mice. J Investig Med 50 54-60. Christofidou-Solomidou, M., Scherpereel, A ., Solomides, C. C., Muzykantov, V. R., Machtay, M., Albelda, S. M., and DiNubile, M. J. (2002b). Changes in plasma gelsolin concentration during acute oxidant lung injury in mice. Lung 180 91104. Clark, R. F., and Goate, A. M. (1993). Mol ecular genetics of Alzheimer's disease. Arch Neurol 50 1164-72. Cooper, J. A., Bryan, J., Schwab, B., 3rd, Frie den, C., Loftus, D. J., and Elson, E. L. (1987). Microinjection of ge lsolin into living cells. J Cell Biol 104 491-501. Corder, E. H., Saunders, A. M., Strittmatter, W. J., Schmechel, D. E., Gaskell, P. C., Small, G. W., Roses, A. D., Haines, J. L., and 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 261 921-3. Coyle, J. T., Price, D. L., and DeLong, M. R. (1983). Alzheimer's disease: a disorder of cortical cholinergic innervation. Science 219 1184-90. Cruz, J. C., and Tsai, L. H. (2004). A Jeky ll and Hyde kinase: roles for Cdk5 in brain development and disease. Curr Opin Neurobiol 14 390-4. Curtain, C. C., Ali, F., Volitakis, I., Cherny, R. A., Norton, R. S., Beyreuther, K., Barrow, C. J., Masters, C. L., Bush, A. I., and Barnham, K. J. (2001). Alzheimer's disease amyloid-beta binds copper and zinc to generate an allosterically ordered membrane-penetrating structure containi ng superoxide dismutase-like subunits. J Biol Chem 276 20466-73. Dahlgren, K. N., Manelli, A. M., Stine, W. B ., Jr., Baker, L. K., Krafft, G. A., and LaDu, M. J. (2002). Oligomeric and fibrilla r species of amyloid-beta peptides differentially affect neuronal viability. J Biol Chem 277 32046-53. Das, P., Howard, V., Loosbrock, N., Dickson, D., Murphy, M. P., and Golde, T. E. (2003). Amyloid-beta immuni zation effectively reduces amyloid deposition in FcRgamma-/knock-out mice. J Neurosci 23 8532-8.

PAGE 124

112 Deane, R., Du Yan, S., Submamaryan, R. K ., LaRue, B., Jovanovic, S., Hogg, E., Welch, D., Manness, L., Lin, C., Yu, J., Zhu, H., Ghiso, J., Frangione, B., Stern, A., Schmidt, A. M., Armstrong, D. L., Arnold, B., Liliensiek, B., Nawroth, P., Hofman, F., Kindy, M., Stern, D., and Zlokovic, B. (2003). RAGE mediates amyloid-beta peptide transport across the blood-brain barrier and accumulation in brain. Nat Med 9 907-13. DeMattos, R. B., Bales, K. R., Cummins, D. J., Dodart, J. C., Paul, S. M., and Holtzman, D. M. (2001). Peripheral anti-A beta antibody alters CNS and plasma A beta clearance and decreases brain A beta bur den in a mouse model of Alzheimer's disease. Proc Natl Acad Sci U S A 98 8850-5. DeMattos, R. B., Bales, K. R., Cummins, D. J., Paul, S. M., and Holtzman, D. M. (2002). Brain to plasma amyloid-beta efflux: a measure of brain amyloid burden in a mouse model of Alzheimer's disease. Science 295 2264-7. Dheen, S. T., Jun, Y., Yan, Z., Tay, S. S., a nd Ling, E. A. (2005). Retinoic acid inhibits expression of TNF-alpha and iNOS in activated rat microglia. Glia 50 21-31. DiCarlo, G., Wilcock, D., Henderson, D ., Gordon, M., and Morgan, D. (2001). Intrahippocampal LPS injections reduce Ab eta load in APP+PS1 transgenic mice. Neurobiol Aging 22 1007-12. Dickstein, D. L., Biron, K. E., Ujiie, M., Pfeife r, C. G., Jeffries, A. R., and Jefferies, W. A. (2006). Abeta peptide immunization rest ores blood-brain barrier integrity in Alzheimer disease. Faseb J 20 426-33. DiNubile, M. J., Stossel, T. P., Ljunghusen, O. C., Ferrara, J. L., and Antin, J. H. (2002). Prognostic implications of declining plasma gelsolin levels after allogeneic stem cell transplantation. Blood 100 4367-71. Dougherty, K. D., Dreyfus, C. F., and Black, I. B. (2000). Brain-de rived neurotrophic factor in astrocytes, oligodendrocytes and microglia/macrophages after spinal cord injury. Neurobiol Dis 7 574-85. 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., and Younkin, S. (1996) Increased amyloid-beta42(43) in brains of mice expressing mutant presenilin 1. Nature 383 710-3. Eckman, E. A., and Eckman, C. B. (2005). Abeta-degrading enzymes: modulators of Alzheimer's disease pathogenesis and targets for therapeutic intervention. Biochem Soc Trans 33 1101-5. Evans, D. A., Funkenstein, H. H., Albert, M. S., Scherr, P. A., Cook, N. R., Chown, M. J., Hebert, L. E., Hennekens, C. H., a nd Taylor, J. O. (1989). Prevalence of Alzheimer's disease in a community popul ation of older persons. Higher than previously reported. JAMA 262 2551-6.

PAGE 125

113 Fiala, M., Lin, J., Ringman, J., Kermani-Arab, V., Tsao, G., Patel, A., Lossinsky, A. S., Graves, M. C., Gustavson, A., Sayre, J., Sofroni, E., Suarez, T., Chiappelli, F., and Bernard, G. (2005). Ineffective pha gocytosis of amyloid-beta by macrophages of Alzheimer's disease patients. J Alzheimers Dis 7 221-32; discussion 255-62. Floden, A. M., and Combs, C. K. (2006). Beta -amyloid stimulates murine postnatal and adult microglia culture s in a unique manner. J Neurosci 26 4644-8. Funato, H., Yoshimura, M., Yamazaki, T., Saido, T. C., Ito, Y., Yokofujita, J., Okeda, R., and Ihara, Y. (1998). Astrocytes containi ng amyloid beta-protein (Abeta)-positive granules are associated with Abeta40-pos itive diffuse plaques in the aged human brain. Am J Pathol 152 983-92. Gao, W., Eisenhauer, P. B., Conn, K., Lynch, J. A., Wells, J. M., Ullman, M. D., McKee, A., Thatte, H. S., and Fine, R. E. (2004). Insulin degrading enzyme is expressed in the human cerebrovascular endothelium and in cultured human cerebrovascular endothelial cells. Neurosci Lett 371 6-11. Giannakopoulos, P., Herrmann, F. R., Bussiere, T., Bouras, C., Kovari, E., Perl, D. P., Morrison, J. H., Gold, G., and Hof, P. R. (2003). Tangle and neuron numbers, but not amyloid load, predict cognitive status in Alzheimer's disease. Neurology 60 1495-500. Gilman, S., Koller, M., Black, R. S., Jenkins, L., Griffith, S. G., Fox, N. C., Eisner, L., Kirby, L., Boada Rovira, M., Forette, F., and Orgogozo, J. M. (2005). Clinical effects of A{beta} immunization (AN1792) in patients with AD in an interrupted trial. Neurology Giulian, D., and Baker, T. J. (1986). Characterization of ame boid microglia isolated from developing mammalian brain. J Neurosci 6 2163-78. Glenner, G. G., and Wong, C. W. (1984). Alzheimer's disease and Down's syndrome: sharing of a unique cerebrovasc ular amyloid fibril protein. Biochem Biophys Res Commun 122 1131-5. Goate, A., Chartier-Harlin, M. C., Mullan, M., Brown, J., Crawford, F., Fidani, L., Giuffra, L., Haynes, A., Irving, N., James, L., and et al. (1991) Segregation of a missense mutation in the amyloid precursor protein gene with familial Alzheimer's disease. Nature 349 704-6. Haas, J., Storch-Hagenlocher, B., Biessma nn, A., and Wildemann, B. (2002). Inducible nitric oxide synthase and argininosucci nate synthetase: co -induction in brain tissue of patients with Alzheimer's deme ntia and following stimulation with betaamyloid 1-42 in vitro. Neurosci Lett 322 121-5. Hardy, J., and Allsop, D. (1991). Amyloid deposi tion as the central ev ent in the aetiology of Alzheimer's disease. Trends Pharmacol Sci 12 383-8.

PAGE 126

114 Harms, C., Bosel, J., Lautenschlager, M., Harm s, U., Braun, J. S., Hortnagl, H., Dirnagl, U., Kwiatkowski, D. J., Fink, K., and Endres, M. (2004). Neuronal gelsolin prevents apoptosis by enhanc ing actin depolymerization. Mol Cell Neurosci 25 69-82. Hasegawa, K., Yamaguchi, I., Omata, S., Ge jyo, F., and Naiki, H. (1999). Interaction between A beta(1-42) and A beta(1-40) in Alzheimer's beta-amyloid fibril formation in vitro. Biochemistry 38 15514-21. Herber, D. L., Roth, L. M., Wilson, D., Wilson, N., Mason, J. E., Morgan, D., and Gordon, M. N. (2004). Time-dependent reduction in Abeta levels after intracranial LPS administration in APP transgenic mice. Exp Neurol 190 245-53. 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., and 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. Holcomb, L. A., Gordon, M. N., Jantzen, P., Hsiao, K., Duff, K., and Morgan, D. (1999). Behavioral changes in transgenic mice e xpressing both amyloi d precursor protein and presenilin-1 mutations: lack of association with amyloid deposits. Behav Genet 29 177-85. Hoozemans, J. J., van Haastert, E. S., Veerhuis, R., Arendt T., Scheper, W., Eikelenboom, P., and Rozemuller, A. J. (2005). Maximal COX-2 and ppRb expression in neurons occurs during early Braak stages prior to the maximal activation of astrocytes and micr oglia in Alzheimer's disease. J Neuroinflammation 2 27. Hoshi, M., Sato, M., Matsumoto, S., Noguchi, A., Yasutake, K., Yoshida, N., and Sato, K. (2003). Spherical aggregates of beta-amyloid (amylospheroid) show high neurotoxicity and activate tau protein ki nase I/glycogen synthase kinase-3beta. Proc Natl Acad Sci U S A 100 6370-5. Hsiao, K., Chapman, P., Nilsen, S., Eckman, C., Harigaya, Y., Younkin, S., Yang, F., and Cole, G. (1996). Correlative memory de ficits, Abeta elevation, and amyloid plaques in transgenic mice. Science 274 99-102. Hsiao, K. K., Borchelt, D. R., Olson, K., J ohannsdottir, R., Kitt, C ., Yunis, W., Xu, S., Eckman, C., Younkin, S., Price, D., and et al. (1995). Age-related CNS disorder and early death in transgenic FVB/ N mice overexpressing Alzheimer amyloid precursor proteins. Neuron 15 1203-18.

PAGE 127

115 Jankowsky, J. L., Fadale, D. J., Anderson, J., Xu, G. M., Gonzales, V., Jenkins, N. A., Copeland, N. G., Lee, M. K., Younkin, L. H., Wagner, S. L., Younkin, S. G., and Borchelt, D. R. (2004). Mutant presenilins specifically elevate the levels of the 42 residue beta-amyloid peptide in vivo: evidence for augmentation of a 42-specific gamma secretase. Hum Mol Genet 13 159-70. Janmey, P. A., Iida, K., Yin, H. L., and Stossel, T. P. (1987). Polyphosphoinositide micelles and polyphosphoinositide-contai ning vesicles dissociate endogenous gelsolin-actin complexes and promote actin assembly from the fast-growing end of actin filaments blocked by gelsolin. J Biol Chem 262 12228-36. Janus, C., Pearson, J., McLaurin, J., Mathews, P. M., Jiang, Y., Schmidt, S. D., Chishti, M. A., Horne, P., Heslin, D., French, J., Mount, H. T., Nixon, R. A., Mercken, M., Bergeron, C., Fraser, P. E., St George -Hyslop, P., and Westaway, D. (2000). A beta peptide immunization reduces behavioural impair ment and plaques in a model of Alzheimer's disease. Nature 408 979-82. Johnson-Wood, K., Lee, M., Motter, R., H u, K., Gordon, G., Barbour, R., Khan, K., Gordon, M., Tan, H., Games, D., Lieber burg, I., Schenk, D., Seubert, P., and McConlogue, L. (1997). Amyloid precurs or protein processing and A beta42 deposition in a transgenic mouse model of Alzheimer disease. Proc Natl Acad Sci U S A 94 1550-5. Johnstone, M., Gearing, A. J., and Miller, K. M. (1999). A central role for astrocytes in the inflammatory response to beta-amy loid; chemokines, cytokines and reactive oxygen species are produced. J Neuroimmunol 93 182-93. Kang, J., Lemaire, H. G., Unterbeck, A., Salbau m, J. M., Masters, C. L., Grzeschik, K. H., Multhaup, G., Beyreuther, K., and Mu ller-Hill, B. (1987). The precursor of Alzheimer's disease amyloid A4 protein resembles a cell-surface receptor. Nature 325 733-6. Kawas, C. H., and Brookmeyer, R. (2001). Aging and the public health effects of dementia. N Engl J Med 344 1160-1. Kayed, R., Head, E., Thompson, J. L., McIntire, T. M., Milton, S. C., Cotman, C. W., and Glabe, C. G. (2003). Common structure of soluble amyloid oligomers implies common mechanism of pathogenesis. Science 300 486-9. Kazmirski, S. L., Howard, M. J., Isaacson, R. L., and Fersht, A. R. (2000). Elucidating the mechanism of familial amyloidosisFinnish type: NMR studies of human gelsolin domain 2. Proc Natl Acad Sci U S A 97 10706-11. Klein, R. L., Hamby, M. E., Gong, Y., Hirko, A. C., Wang, S., Hughes, J. A., King, M. A., and Meyer, E. M. (2002). Dose and pr omoter effects of adeno-associated viral vector for green fluorescent prot ein expression in the rat brain. Exp Neurol 176 66-74.

PAGE 128

116 Klein, R. L., Hirko, A. C., Meyers, C. A., Grimes, J. R., Muzyczka, N., and Meyer, E. M. (2000). NGF gene transfer to intrinsi c basal forebrain neurons increases cholinergic cell size and protects from ag e-related, spatial memory deficits in middle-aged rats. Brain Res 875 144-51. Koppel, R. (2002). Alzheimer's Disease: The Costs to U.S. Businesses in 2002. Alzheimer's Associati on, Washington, D.C. Kothakota, S., Azuma, T., Reinhard, C., K lippel, A., Tang, J., Chu, K., McGarry, T. J., Kirschner, M. W., Koths, K., Kwiatkowsk i, D. J., and Williams, L. T. (1997). Caspase-3-generated fragment of gelsolin : effector of morphological change in apoptosis. Science 278 294-8. Kraepelin, E. (1910). "Psychiatrie : ein Le hrbuch fr Studierende und rzte," 8., vollst. umgearb. Aufl./Ed. Barth, Leipzig, Germany. Kwiatkowski, D. J. (1999). Functions of gelsolin: motility, signaling, apoptosis, cancer. Curr Opin Cell Biol 11 103-8. Kwiatkowski, D. J., Janmey, P. A., and Yi n, H. L. (1989). Identif ication of critical functional and regulatory domains in gelsolin. J Cell Biol 108 1717-26. Lakowicz, J. R. (2002). "Topics in fluore scence spectroscopy," Kluwer Academic, New York. Lanz, T. A., Fici, G. J., and Merchant, K. M. (2005). Lack of specific amyloid-beta(1-42) suppression by nonsteroidal anti-infla mmatory drugs in young, plaque-free Tg2576 mice and in guinea pig neuronal cultures. J Pharmacol Exp Ther 312 399-406. Lashuel, H. A., Hartley, D., Petre, B. M ., Walz, T., and Lansbury, P. T., Jr. (2002). Neurodegenerative disease: amyloi d pores from pathogenic mutations. Nature 418 291. Lee, P. S., Drager, L. R., Stossel, T. P ., Moore, F. D., and Rogers, S. O. (2006). Relationship of plasma gelsolin levels to outcomes in critically ill surgical patients. Ann Surg 243 399-403. Lee, W. M., and Galbraith, R. M. (1992). The extracellular actin-scavenger system and actin toxicity. N Engl J Med 326 1335-41. Leissring, M. A., Farris, W., Chang, A. Y., Walsh, D. M., W u, X., Sun, X., Frosch, M. P., and Selkoe, D. J. (2003). Enhanced proteoly sis of beta-amyloid in APP transgenic mice prevents plaque formation, secondary pathology, and premature death. Neuron 40 1087-93.

PAGE 129

117 Lemere, C. A., Spooner, E. T., LaFrancois, J ., Malester, B., Mori, C., Leverone, J. F., Matsuoka, Y., Taylor, J. W., DeMattos, R. B., Holtzman, D. M., Clements, J. D., Selkoe, D. J., and Duff, K. E. (2003). Evidence for peripheral clearance of cerebral Abeta protein following chronic, active Abeta immunization in PSAPP mice. Neurobiol Dis 14 10-8. Leveugle, B., Scanameo, A., Ding, W., and F illit, H. (1994). Binding of heparan sulfate glycosaminoglycan to beta-amyloid peptide: inhibition by potentially therapeutic polysulfated compounds. Neuroreport 5 1389-92. Levites, Y., Das, P., Price, R. W., Rochet te, M. J., Kostura, L. A., McGowan, E. M., Murphy, M. P., and Golde, T. E. (2006). Anti-Abeta42and an ti-Abeta40-specific mAbs attenuate amyloid deposition in an Alzheimer disease mouse model. J Clin Invest 116 193-201. Levy-Lahad, E., Wijsman, E. M., Nemens, E ., Anderson, L., Goddard, K. A., Weber, J. L., Bird, T. D., and Schellenberg, G. D. (1995). A familial Alzheimer's disease locus on chromosome 1. Science 269 970-3. Lind, S. E., and Janmey, P. A. (1984). Huma n plasma gelsolin binds to fibronectin. J Biol Chem 259 13262-6. Lindberg, C., Selenica, M. L., Westlind-Danielsson, A., and Schultzberg, M. (2005). Beta-amyloid protein structure determines the nature of cytokine release from rat microglia. J Mol Neurosci 27 1-12. Liu, F., Song, Y., and Liu, D. (1999). Hydrodynamics-based transfection in animals by systemic administration of plasmid DNA. Gene Ther 6 1258-66. Liu, T., Perry, G., Chan, H. W., Verdile, G., Martins, R. N., Smith, M. A., and Atwood, C. S. (2004). Amyloid-beta-induced toxic ity of primary neurons is dependent upon differentiation-associated increases in tau and cyclin-dependent kinase 5 expression. J Neurochem 88 554-63. Lopez-Toledano, M. A., and Shelanski, M. L. (2004). Neurogenic effect of beta-amyloid peptide in the development of neural stem cells. J Neurosci 24 5439-44. Lue, L. F., Kuo, Y. M., Roher, A. E., Brachova, L., Shen, Y., Sue, L., Beach, T., Kurth, J. H., Rydel, R. E., and Rogers, J. ( 1999). Soluble amyloid beta peptide concentration as a predic tor of synaptic change in Alzheimer's disease. Am J Pathol 155 853-62. Marco, S., and Skaper, S. D. (2006). Amyloid beta-peptide(1-42) al ters tight junction protein distribution and e xpression in brain microvessel endothelial cells. Neurosci Lett

PAGE 130

118 Marr, R. A., Rockenstein, E., Mukherjee, A., Kindy, M. S., Hersh, L. B., Gage, F. H., Verma, I. M., and Masliah, E. (2003). Ne prilysin gene transfer reduces human amyloid pathology in transgenic mice. J Neurosci 23 1992-6. Masliah, E., Rockenstein, E., Veinbergs, I ., Sagara, Y., Mallory, M., Hashimoto, M., and Mucke, L. (2001). beta-amyloid peptid es enhance alpha-synuclein accumulation and neuronal deficits in a transgenic mouse model linking Alzheimer's disease and Parkinson's disease. Proc Natl Acad Sci U S A 98 12245-50. Matsumoto, N., Kitayama, H., Kitada, M., Ki mura, K., Noda, M., and Ide, C. (2003). Isolation of a set of genes expressed in the choroid plexus of the mouse using suppression subtractive hybridization. Neuroscience 117 405-15. Matsuoka, Y., Saito, M., LaFrancois, J., Ga ynor, K., Olm, V., Wang, L., Casey, E., Lu, Y., Shiratori, C., Lemere, C., and Duff, K. (2003). Novel therapeutic approach for the treatment of Alzheimer's disease by pe ripheral administration of agents with an affinity to beta-amyloid. J Neurosci 23 29-33. Matsuoka, Y., Shao, L., Debnath, M., Lafranco is, J., Becker, A., Gray, A., Aisen, P., Mathis, C., Klunk, W., and Duff, K. ( 2005). An Abeta sequestration approach using non-antibody Abeta binding agents. Curr Alzheimer Res 2 265-8. Mattson, M. P., Cheng, B., Davis, D., Bryant, K ., Lieberburg, I., and Rydel, R. E. (1992). beta-Amyloid peptides destabilize calciu m homeostasis and render human cortical neurons vulnerable to excitotoxicity. J Neurosci 12 376-89. McGeer, E. G., and McGeer, P. L. (2003). Inflammatory processes in Alzheimer's disease. Prog Neuropsychopharmacol Biol Psychiatry 27 741-9. McLaughlin, P. J., Gooch, J. T., Mannherz, H. G., and Weeds, A. G. (1993). Structure of gelsolin segment 1-actin complex and the mechanism of filament severing. Nature 364 685-92. Moller, H. J., and Graeber, M. B. (1998). The case described by Alois Alzheimer in 1911. Historical and conceptual perspectiv es based on the clinical record and neurohistological sections. Eur Arch Psychiatry Clin Neurosci 248 111-22. 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., and Arendash, G. W. (2000). A beta peptide vaccination prevents memory loss in an animal m odel of Alzheimer's disease. Nature 408 982-5. Mounzer, K. C., Moncure, M., Smith, Y. R., and Dinubile, M. J. (1999). Relationship of admission plasma gelsolin levels to clinical outcomes in patients after major trauma. Am J Respir Crit Care Med 160, 1673-81. Nakajima, K., Honda, S., Tohyama, Y., Imai, Y., Kohsaka, S., and Kurihara, T. (2001). Neurotrophin secretion from cultured microglia. J Neurosci Res 65 322-31.

PAGE 131

119 Naslund, J., Haroutunian, V., Mohs, R., Davis, K. L., Davies, P., Greengard, P., and Buxbaum, J. D. (2000). Correlation between elevated levels of amyloid betapeptide in the brain and cognitive decline. Jama 283 1571-7. Noble, W., Olm, V., Takata, K., Case y, E., Mary, O., Meyerson, J., Gaynor, K., LaFrancois, J., Wang, L., Kondo, T., Davies P., Burns, M., Veeranna, Nixon, R., Dickson, D., Matsuoka, Y., Ahlijanian, M., Lau, L. F., and Duff, K. (2003). Cdk5 is a key factor in tau aggregat ion and tangle formation in vivo. Neuron 38 55565. Nodes, B. R., Shackelford, J. E., and Lebherz H. G. (1987). Synthe sis and secretion of serum gelsolin by smooth muscle tissue. J Biol Chem 262 5422-7. Orgogozo, J. M., Gilman, S., Dartigues, J. F., Laurent, B., Puel, M., Kirby, L. C., Jouanny, P., Dubois, B., Eisner, L., Flitman, S., Michel, B. F., Boada, M., Frank, A., and Hock, C. (2003). Subacute meningoe ncephalitis in a s ubset of patients with AD after Abeta42 immunization. Neurology 61 46-54. Pellerin, L. (2005). How astrocytes feed hungry neurons. Mol Neurobiol 32 59-72. Petty, B. G., Cornblath, D. R., Adornato, B. T., Chaudhry, V., Flexner, C., Wachsman, M., Sinicropi, D., Burton, L. E., and Pe routka, S. J. (1994). The effect of systemically administered recombinant human nerve growth factor in healthy human subjects. Ann Neurol 36 244-6. Pike, C. J., Walencewicz, A. J., Glabe, C. G., and Cotman, C. W. (1991). Aggregationrelated toxicity of synthetic beta-amy loid protein in hippocampal cultures. Eur J Pharmacol 207 367-8. Ray, I., Chauhan, A., Wegiel, J., and Chauha n, V. P. (2000). Gelsolin inhibits the fibrillization of amyloid beta-protein, and also defibrillizes its preformed fibrils. Brain Res 853 344-51. Rogers, J., Strohmeyer, R., Kovelowski, C. J., and Li, R. (2002). Microglia and inflammatory mechanisms in the cl earance of amyloid beta peptide. Glia 40 2609. Rothenbach, P. A., Dahl, B., Schwartz, J. J., O'Keefe, G. E., Yamamoto, M., Lee, W. M., Horton, J. W., Yin, H. L., and Turnage, R. H. (2004). Recombinant plasma gelsolin infusion attenuate s burn-induced pulmonary microvascular dysfunction. J Appl Physiol 96 25-31. Schenk, D., Barbour, R., Dunn, W., Gordon, G., Grajeda, H., Guido, T., Hu, K., Huang, J., Johnson-Wood, K., Khan, K., Kholodenko, D., Lee, M., Liao, Z., Lieberburg, I., Motter, R., Mutter, L., Soriano, F. Shopp, G., Vasquez, N., Vandevert, C., Walker, S., Wogulis, M., Yednock, T., Ga mes, D., and Seubert, P. (1999). Immunization with amyloidbeta attenuates Alzheimer-d isease-like pathology in the PDAPP mouse. Nature 400 173-7.

PAGE 132

120 Scheuner, D., Eckman, C., Jensen, M., Song, X., Citron, M., Suzuki, N., Bird, T. D., Hardy, J., Hutton, M., Kukull, W., Larson, E., Levy-Lahad, E., Viitanen, M., Peskind, E., Poorkaj, P., Schellenberg, G ., Tanzi, R., Wasco, W., Lannfelt, L., Selkoe, D., and Younkin, S. (1996). Secreted amyloid beta-protein similar to that in the senile plaques of Alzheimer's diseas e is increased in vivo by the presenilin 1 and 2 and APP mutations linked to familial Alzheimer's disease. Nat Med 2 864-70. Scott, M. R., Kohler, R., Foster, D., and Pr usiner, S. B. (1992). Ch imeric prion protein expression in cultured cells and transgenic mice. Protein Sci 1 986-97. Selkoe, D. J. (1989). The deposition of amyl oid proteins in the aging mammalian brain: implications for Alzheimer's disease. Ann Med 21 73-6. Selkoe, D. J. (1990). Deciphering Alzheimer' s disease: the amyloid precursor protein yields new clues. Science 248 1058-60. Sherrington, R., Froelich, S., Sorbi, S., Camp ion, D., Chi, H., Rogaeva, E. A., Levesque, G., Rogaev, E. I., Lin, C., Liang, Y., Iked a, M., Mar, L., Brice, A., Agid, Y., Percy, M. E., Clerget-Darpoux, F., Piacen tini, S., Marcon, G., Nacmias, B., Amaducci, L., Frebourg, T., Lannfelt, L., Rommens, J. M., and St GeorgeHyslop, P. H. (1996). Alzheimer's disease associated with mutations in presenilin 2 is rare and variably penetrant. Hum Mol Genet 5 985-8. Shibata, M., Yamada, S., Kumar, S. R., Ca lero, M., Bading, J., Frangione, B., Holtzman, D. M., Miller, C. A., Strickland, D. K., Ghiso, J., and Zlokovic, B. V. (2000). Clearance of Alzheimer's amyloid-ss(1-40) peptide from brain by LDL receptorrelated protein-1 at the blood-brain barrier. J Clin Invest 106 1489-99. Sinha, S., and Lieberburg, I. (1999). Cellula r mechanisms of beta-amyloid production and secretion. Proc Natl Acad Sci U S A 96 11049-53. Sloane, P. D., Zimmerman, S., Suchindran, C., Reed, P., Wang, L., Boustani, M., and Sudha, S. (2002). The public health impact of Alzheimer's disease, 2000-2050: potential implication of treatment advances. Annu Rev Public Health 23 213-31. Smith, D. B., Janmey, P. A., Herbert, T. J., and Lind, S. E. (1987). Quantitative measurement of plasma gelsolin and its incorporation into fibrin clots. J Lab Clin Med 110 189-95. Sofroniew, M. V. (2005). Reactive astrocyt es in neural repair and protection. Neuroscientist 11 400-7. Song, S., Embury, J., Laipis, P. J., Berns, K. I ., Crawford, J. M., and Flotte, T. R. (2001). Stable therapeutic serum levels of huma n alpha-1 antitrypsin (AAT) after portal vein injection of recombinant aden o-associated virus (rAAV) vectors. Gene Ther 8 1299-306.

PAGE 133

121 Streit, W. J., Graeber, M. B., and Kreutzbe rg, G. W. (1988). Functional plasticity of microglia: a review. Glia 1 301-7. Sun, A., Nguyen, X. V., and Bing, G. (2002). Comparative analysis of an improved thioflavin-s stain, Gallyas silver st ain, and immunohistochemistry for neurofibrillary tang le demonstration on the same sections. J Histochem Cytochem 50 463-72. Suzuki, N., Cheung, T. T., Cai, X. D., Odaka, A., Otvos, L., Jr., Eckman, C., Golde, T. E., and Younkin, S. G. (1994). An increas ed percentage of long amyloid beta protein secreted by familial amyloid be ta protein precursor (beta APP717) mutants. Science 264 1336-40. Tamaoka, A., Odaka, A., Ishibashi, Y., Usami, M., Sahara, N., Suzuki, N., Nukina, N., Mizusawa, H., Shoji, S., Kanazawa, I., and et al. (1994). APP717 missense mutation affects the ratio of amyloid beta protein speci es (A beta 1-42/43 and a beta 1-40) in familial Alzheimer's disease brain. J Biol Chem 269 32721-4. Telling, G. C., Scott, M., Hsiao, K. K., Foster, D., Yang, S. L., Torchia, M., Sidle, K. C., Collinge, J., DeArmond, S. J., and Prus iner, S. B. (1994). Transmission of Creutzfeldt-Jakob disease from humans to transgenic mice expressing chimeric human-mouse prion protein. Proc Natl Acad Sci U S A 91 9936-40. Terry, R. D., and Katzman, R. (1983). Sen ile dementia of the Alzheimer type. Ann Neurol 14 497-506. Trinh, N. H., Hoblyn, J., Mohanty, S., and Ya ffe, K. (2003). Efficacy of cholinesterase inhibitors in the treatment of neur opsychiatric sympto ms and functional impairment in Alzheimer disease: a meta-analysis. Jama 289 210-6. Turner, P. R., O'Connor, K., Tate, W. P., a nd Abraham, W. C. (2003). Roles of amyloid precursor protein and its fragments in regulating neural act ivity, plasticity and memory. Prog Neurobiol 70 1-32. van Oijen, M., Hofman, A., Soares, H. D., Koud staal, P. J., and Breteler, M. M. (2006). Plasma Abeta(1-40) and Abet a(1-42) and the risk of de mentia: a prospective casecohort study. Lancet Neurol 5 655-60. Vasconcellos, C. A., Allen, P. G., Wohl, M. E., Drazen, J. M., Janmey, P. A., and Stossel, T. P. (1994). Reduction in viscosity of cys tic fibrosis sputum in vitro by gelsolin. Science 263 969-71. Volterra, A., and Meldolesi, J. (2005). As trocytes, from brain glue to communication elements: the revol ution continues. Nat Rev Neurosci 6 626-40. von Bernhardi, R., and Eugenin, J. (2004). Mi croglial reactivity to beta-amyloid is modulated by astrocytes a nd proinflammatory factors. Brain Res 1025 186-93.

PAGE 134

122 Vouyiouklis, D. A., and Brophy, P. J. (1997) A novel gelsolin isoform expressed by oligodendrocytes in the central nervous system. J Neurochem 69 995-1005. Walker, D. G., and Lue, L. F. (2005). Invest igations with cultured human microglia on pathogenic mechanisms of Alzheimer' s disease and other neurodegenerative diseases. J Neurosci Res 81 412-25. Weber, G. (1953). Rotational Brownian moti on and polarization of the fluorescence of solutions. Adv Protein Chem 8 415-59. Wegiel, J., Chauhan, A., Wisniewski, H. M., Nowakowski, J., Wang, K. C., and Le Vine, H., 3rd (1996). Promotion of synthetic am yloid beta-peptide fibrillization by cell culture media and cessation of fibrillization by serum. Neurosci Lett 211 151-4. Weinstock, M. (1995). The pharmacotherapy of Alzheimer's disease based on the cholinergic hypothesis: an update. Neurodegeneration 4 349-56. Westerman, M. A., Cooper-Blacketer, D., Ma riash, A., Kotilinek, L., Kawarabayashi, T., Younkin, L. H., Carlson, G. A., Younkin, S. G., and Ashe, K. H. (2002). The relationship between Abeta and memo ry in the Tg2576 mouse model of Alzheimer's disease. J Neurosci 22 1858-67. Wilcock, D. M., DiCarlo, G., Henderson, D ., Jackson, J., Clarke, K., Ugen, K. E., Gordon, M. N., and 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 3745-51. Wimo, A., Winblad, B., Stoffler, A., Wirt h, Y., and Mobius, H. J. (2003). Resource utilisation and cost analysis of memantin e in patients with moderate to severe Alzheimer's disease. Pharmacoeconomics 21 327-40. Witke, W., Li, W., Kwiatkowski, D. J., a nd Southwick, F. S. (2001). Comparisons of CapG and gelsolin-null macrophages: demons tration of a unique role for CapG in receptor-mediated ruffling, phagocytosis, and vesicle rocketing. J Cell Biol 154 775-84. Witke, W., Sharpe, A. H., Hartwig, J. H., Azum a, T., Stossel, T. P., and Kwiatkowski, D. J. (1995). Hemostatic, inflammatory, and fibroblast responses are blunted in mice lacking gelsolin. Cell 81 41-51. Wu, K., Meyers, C. A., Guerra, N. K., King, M. A., and Meyer, E. M. (2004). The effects of rAAV2-mediated NGF gene delivery in adult and aged rats. Mol Ther 9 262-9. Wyss-Coray, T., Loike, J. D., Brionne, T. C., Lu, E., Anankov, R., Yan, F., Silverstein, S. C., and Husemann, J. (2003). Adult mouse astrocytes degrade amyloid-beta in vitro and in situ. Nat Med 9 453-7.

PAGE 135

123 Yamaguchi, H., Sugihara, S., Ogawa, A., Sai do, T. C., and Ihara, Y. (1998). Diffuse plaques associated with astroglial am yloid beta protein, possibly showing a disappearing stage of senile plaques. Acta Neuropathol (Berl) 95 217-22. Yankner, B. A., Duffy, L. K., and Kirschner, D. A. (1990). Neurotr ophic and neurotoxic effects of amyloid beta protein: reversal by t achykinin neuropeptides. Science 250 279-82. Yin, H. L., and Stossel, T. P. (1979). Contro l of cytoplasmic actin gel-sol transformation by gelsolin, a calcium-depe ndent regulatory protein. Nature 281 583-6. Zhang, G., Gao, X., Song, Y. K., Vollmer, R., Stolz, D. B., Gasiorowski, J. Z., Dean, D. A., and Liu, D. (2004). Hydroporation as the mechanism of hydrodynamic delivery. Gene Ther 11 675-82. Zheng, H., Jiang, M., Trumbauer, M. E., Hopkins R., Sirinathsinghji, D. J., Stevens, K. A., Conner, M. W., Slunt, H. H., Sisodia, S. S., Chen, H. Y., and Van der Ploeg, L. H. (1996). Mice deficient for th e amyloid precursor protein gene. Ann N Y Acad Sci 777 421-6. Zheng, H., Jiang, M., Trumbauer, M. E., Siri nathsinghji, D. J., Hopkins, R., Smith, D. W., Heavens, R. P., Dawson, G. R., Boyce, S., Conner, M. W., Stevens, K. A., Slunt, H. H., Sisoda, S. S., Chen, H. Y ., and Van der Ploeg, L. H. (1995). betaAmyloid precursor protein-deficient mice show reactive gliosis and decreased locomotor activity. Cell 81 525-31. Ziv, Y., Ron, N., Butovsky, O., Landa, G., S udai, E., Greenberg, N., Cohen, H., Kipnis, J., and Schwartz, M. (2006). Immune cel ls contribute to the maintenance of neurogenesis and spatial lear ning abilities in adulthood. Nat Neurosci 9 268-75.

PAGE 136

124 BIOGRAPHICAL SKETCH The author was born in Pittsburgh, Pe nnsylvania, on May 9, 1972. In 1980, he moved to Gainesville, Florida, where he spent the remainder of his childhood. He graduated from Buchholz High School in 1990, a nd started studies at the University of Florida immediately afterward. After 2 years of studies at th e University of Florida he left school and was trained as a commercial airc raft pilot. He returned to school in 1998, and obtained a part-time job in the laboratory of Dr. Edwin Meyer, where his interest in research first started. While working in Dr. Meyer’s lab, he contribu ted to three articles published in scientific journals. Also while working there, he was introduced to and worked with Dr. Jeffrey Hughes, who later became his mentor. He graduated with honors with a bachelor’s degr ee (food science and human nutri tion) in 2000. In 2001 he matriculated as a graduate student under the tutelage of Dr. Jeffrey Hughes, in pursuit of a Ph.D. in the pharmaceutical sciences.


Permanent Link: http://ufdc.ufl.edu/UFE0013095/00001

Material Information

Title: Peripheral Expression of Plasma Gelsolin as a Treatment for Alzheimer's Disease
Physical Description: Mixed Material
Copyright Date: 2008

Record Information

Source Institution: University of Florida
Holding Location: University of Florida
Rights Management: All rights reserved by the source institution and holding location.
System ID: UFE0013095:00001

Permanent Link: http://ufdc.ufl.edu/UFE0013095/00001

Material Information

Title: Peripheral Expression of Plasma Gelsolin as a Treatment for Alzheimer's Disease
Physical Description: Mixed Material
Copyright Date: 2008

Record Information

Source Institution: University of Florida
Holding Location: University of Florida
Rights Management: All rights reserved by the source institution and holding location.
System ID: UFE0013095:00001


This item has the following downloads:


Full Text












PERIPHERAL EXPRESSION OF PLASMA GELSOLIN AS A TREATMENT FOR
ALZHEIMER' S DISEASE















By

AARON HIRKO


A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL
OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT
OF THE REQUIREMENTS FOR THE DEGREE OF
DOCTOR OF PHILOSOPHY

UNIVERSITY OF FLORIDA


2006

































Copyright 2006

by

Aaron Hirko

































To my daughter, Ava Wrenn Hirko; and to her brother, Elliott Todd Hirko















ACKNOWLEDGMENTS

I would like to thank my parents, Robert and Theresa Hirko, for all of their love

and support. I also need to thank my wife, Laura; daughter, Ava; and my son Elliott for

their inspiration, love, and support.

I owe a deep debt of gratitude to my mentor, Dr. Jeffrey Hughes, for giving me the

opportunity to study in the pharmaceutics program. I also need to thank Dr. Edwin

Meyer for taking me in as an undergrad, introducing me to Dr. Hughes, and giving me

space in his laboratory. I thank my supervisory committee members (Dr. Michael King,

Dr. Guenther Hochhaus, and Dr. Sihong Song) for all the excellent advice and insight. I

acknowledge Dr. Raj Rao, Dr. Ke Wu, Dr. Ron Klein, Dr. Yan Gong, Dr. Ke Ren, and

Dr. Preeti Yadava for their friendship, help, and advice. Finally, I thank all of the

students and staff from the Department of Pharmaceutics.
















TABLE OF CONTENTS



A C K N O W L E D G M E N T S ................................................................................................. iv

L IST O F FIG U R E S .............. ............................ ............. ........... ... ........ viii

ABSTRACT .............. .......................................... xi

CHAPTER

1 IN TR OD U CTION ............................................... .. ......................... ..

D em entia ..................... ..... ........... ...... ..... ....... ....................... ................
D discovery of A lzheim er's disease ................................... ...........................................2
C current Therapies ................................................................ 3
Inherited Alzheim her's .................................................... .. .... ................ .3
A m yloid Precursor Protein ........................................ ................................. 4
A m yloid Cascade H ypothesis........................... ................................................ 5
Evidence Supporting the Amyloid Cascade Hypothesis.............. .................6
Critics of the Amyloid Cascade Hypothesis...................................................6
E lan and W y eth T rial............. ................................................................. .. .. ....
Sink-H hypothesis ....................................................... 8
G elsolin .................................................. 10

2 M ATERIALS AND M ETHOD S ........................................ ......................... 14

R e a g e n ts .........................................................................................14
Sub cloning V ectors............ .............................................................. .......... ...... 14
L arge Scale Plasm id Preparation .................................................................... ..... 15
C e ll C u ltu re ........................................................................................................... 1 7
A nim als and Procedures ............................................................................. .. 17
Immunoprecipitation and Western Blot..... .................... ...............19
D election of M message .............................................................................. ... ... 20
Enzyme Linked Immunosorbent Assay ....................................... ...............21
H isto c h em istry ............................................................................................2 2
Im m unohistochem istry ....................................................................... ..................23
Im age A analysis ..............2...........................24





v









3 THE INTERACTION OF PLAMA GELSOLIN AND AMYLOID 3 ......................26

Introduction .......................................... ..... ..... ... ...... .......... .. 26
Experim ental M ethods and Results ...................... .... ... .................... ....27
Measurement of 125I Labeled A3 (1-42) Binding to Recombinant Human
P lasm a G elsolin .............. ........... ...... ............ ... ... ........ ..... .. ..............27
Measuring Binding of HiLyte Fluor 488 Labeled Amyloid 3 (1-42) to Bovine
Plasma Gelsolin Using Fluorescence Anisotropy ...................................27
HiLyte Fluor 488 Labeled A3 (1-42) Fibril Formation................................. 30
HiLyte Fluor 488 Labeled A3 (1-42) Fibril Disassembly with Gelsolin ............30
Measuring Binding Amyloid 3 to Plasma Gelsolin Using Surface Plasmon
R eson an ce .......................................................................32
C o n clu sio n s..................................................... ................ 3 4

4 EXPRESSING PLASMA GELSOLIN AND EFFECTS IN TRANSGENIC MICE 48

Intro du action ...................................... ................................................ 4 8
R results and D discussion: .............. ................ .................. .. ........ .. .... 50
Conformation of Vector Product and Activity ..................................................50
Hydrodynamic Gene Delivery in Mice ..................... ... ..................50
Hematoxylin and Eosin Staining in Mice After Hydrodynamic Gene Transfer .51
C o n clu sio n s..................................................... ................ 5 2

5 EFFECT OF GELSOLIN EXPRESSION ON AMYLOID DEPOSITION ...............57

In tro d u ctio n ............. ...... ...... ... ................. ................................ 5 7
Results............... .. ... ....... .............. ......... .............. 59
Mice Expressing Swedish Mutant Amyloid Precusor Protein (Mouse/Human
Hybrid) and Exon 9 Deleted Mutant Presenilin-1 .......................................59
M message D election ..................................... ....................................... 59
Total Brain Amyloid 3 (1-42) Concentrations........... ....... ............ 60
Plasma Amyloid 3 Concentrations................... .. .................... 61
Dense Cored Amyloid Deposits........... .............................................61
D iffuse A m yloid D eposits................................... ................................... 62
Soluble A m yloid O ligom ers.................................... ......................... 63
M ic ro g lia ................................................................. ................ 6 4
A strocytes ........... ........ ...... .. .. ..... .. ......... .... ............. 66
Mice Expressing Swedish Mutant Amyloid Precursor Protein (Human) and
M 146L M utant Presenilin-1........................................................ ....... 66
M message D election ............................. ..................................... ............. 66
Amyloid 3 (1-42) Quantification by ELISA ..............................................67
Plasma Amyloid 3 Concentrations.................... ..................... 68
Dense Cored Amyloid Deposits.................... .................. ............. 69
Diffuse Amyloid Deposits.......... .. ......... ........... .. .............. 71
Microglia, Soluble Oligomers, and Astrocytes ........................................72
R egression A analysis ........................... ...... ................ ........ .. ...... .. 74









C o n c lu sio n .................................................. ................ 7 5

6 DISCUSSION AND FUTURE DIRECTIONS................................1.02

L IST O F R E FE R E N C E S ......................................................................... ................... 109

BIOGRAPH ICAL SKETCH .............................................................. ............... 124
















LIST OF FIGURES


Figure p
1-1 Amyloid precursor protein processing. ............................................ ............... 13

2-1 Vectors ........................... ........... ................. ......... 25

2-2 Restriction digest to confirm orientation of clones. .................................................25

3-1 Binding of 125I labeled AP 1-42 to Whatman GF/C filters............... ..................37

3-2 Binding of amyloid 3 (1-42) to the antibody 6E10 ................................................37

3-3 Binding of plasma gelsolin to fluorescently labeled amyloid 3 (1-42) .....................38

3-4 Amyloid 3 (1-42) binding to plasma gelsolin after concentration by centrifugal
filtration ............................................................... .... ..... ......... 38

3-5 Fibril formation of HiLyte Fluor 488 Labeled AP (1-42) .........................................39

3-6 D ialysis experim ent. ............................... ..................................... .. ....... 40

3-7 Amino-coupling of amyloid 3 1-40 to CM 5 chip......................................................41

3-8 Amino-coupling of amyloid 3 1-42 to CM 5 chip......................................................42

3-9 Antibody binding to amyloid coupled CM5 chip ...................................................43

3-10 Human plasma gelsolin binding to amyloid coupled CM5 chip. ...........................44

3-11 Amino-coupling of human plasma gelsolin to CM5 chip ........................................45

3-12 Binding of GS-2C4 to human plasma gelsolin coupled CM5 chip .........................46

3-13 Binding of amyloid 3 1-40 to human plasma gelsolin coupled CM5 chip...............47

3-14 Binding of amyloid 3 1-42 to human plasma gelsolin coupled CM5 chip...............47

4-1 Im m unoblot of transfected m edia......................................... ........................... 53

4-2 Im m unoprecipitation ......................................................................... ....................53

4-3 Bioluminescence resulting from hydrodynamic gene transfer..................................54









4-4 G reen fluorescent protein. ................................................ ............................... 55

4-5 W western blot of plasm a sam ples ...... ........ ................................... .....................55

4-6 H em atoxylin and eosin staining ............................................................................ 56

5-1 Detecting message in mo/huAPP/PSl6E9 mice ......... .......... .. ................. 78

5-2 Amyloid 3 (1-42) concentrations in mo/huAPP/PS16E9 brains................... ............... 78

5-3 Plasma concentrations of Amyloid in mo/huAPP/PS 1E9 mice.......................... 79

5-4 Analysis of dense-core amyloid deposits inmo/huAPP/PS 1E9 mice.............................80

5-5 Analysis of diffuse amyloid deposits in mo/huAPP/PSl6E9 mice ...................................81

5-6 Dense-core amyloid deposits in mo/huAPP/PSl6E9 mice ........................................ 82

5-7 Diffuse amyloid deposits in mo/huAPP/PS 16E9 mice. ............................... ...............83

5-8 Side by side com prison of staining ............................................................................84

5-9 Staining intensity of soluble amyloid oligomers in mo/huAPP/PS 1E9 mice ...............85

5-10 High magnification tiled images of amyloid oligomer staining.............................85

5-11 Cell types stained for soluble amyloid oligomers in pUFGL-injected mice.............86

5-12 Staining intensity of microglia in mo/huAPP/PSl6E9 mice..........................................87

5-13 Low m agnification images of microglia ............................. .....................88

5-14 High magnification images of microglia staining. ................ .............................89

5-15 Average GFAP percent stained area in mo/huAPP/PS 1E9 cortex and hippocampus. .89

5-16 High magnification images of astrocytes surrounding congo red positive amyloid
deposits ................ ...... ... ......... .......................................90

5-17 Message detection in hAPP/PSlM146L mice. ................................... ............... 90

5-18 Amyloid 3 (1-42) concentrations in huAPP/PS M146L brains............................... 91

5-19 Plasma concentrations of amyloid in huAPP/PSlM146L mice ...................................92

5-20 Dense-core amyloid deposits inhuAPP/PSlM146L mice............................................93

5-21 Analysis of dense-core amyloid deposits in huAPP/PS1M146L mice.............................94









5-22 Analysis of dense-core amyloid deposits in terms of fractions compared to
untreated littermates in huAPP/PSlM146L mice............................... ...............95

5-23 Diffuse amyloid deposits in huAPP/PS M146L mice................. ............... .............96

5-24 Analysis of diffuse amyloid deposits in huAPP/PSlM146L mice.............................97

5-25 Analysis of diffuse amyloid deposits in terms of fractions compared to untreated
litterm ates in huAPP/PSlM146L m ice............................................... ............... 98

5-26 Side by side comparison of staining in. huAPP/PSlM146L mice...............................99

5-27 Analysis of microglia, soluble amyloid oligomers, and astrocytes in
huAPP/PS M146L m ice .................. .......................... .. .... .. .. .. ........ .... 100

5-28 L inear regression analysis .............................................. ............................. 101















Abstract of Dissertation Presented to the Graduate School
of the University of Florida in Partial Fulfillment of the
Requirements for the Degree of Doctor of Philosophy

PERIPHERAL EXPRESSION OF PLASMA GELSOLIN AS A TREATMENT FOR
ALZHEIMER' S DISEASE

By

Aaron Hirko

August 2006

Chair: Jeffrey Hughes
Major Department: Pharmacy

Alzheimer's disease (AD) is a progressive neurodegenerative disorder affecting

memory, thinking, behavior, and emotion. It is characterized by a progressive

accumulation of extracellular amyloid plaques and intracellular neurofibrillary tangles.

Evidence suggests that the deposition of amyloid triggers a cascade that ultimately leads

to Alzheimer's pathology, making amyloid a promising target for the treatment of AD.

Amyloid plaques are composed mainly of the 4.5 kD peptide fragment amyloid 3 (A3).

One strategy targeting A3 is to deliver an A3 binding agent outside the brain, creating a

peripheral sink that causes efflux of A3 across the blood-brain barrier.

One such agent is the 89 kD protein plasma gelsolin. However, administering such

a large compound poses formidable formulations challenges, and proteins generally have

poor pharmacokinetic properties. Taking a gene-therapy approach by delivering a DNA

vector coding for plasma gelsolin offers an alternative to repeated injections of protein.









We developed a plasmid vector for human plasma gelsolin. We determined that

plasma gelsolin may have enzymatic-like functions toward A3, shifting the equilibrium

from fibrillization and deposition to solubilization and elimination. We obtained

expression of our plasmid vector for plasma gelsolin in the periphery of 2 different mouse

models of Alzheimer's, and showed that it results in a significant reduction in the amount

of A3 in the brain. We also showed that this reduction of A3 in the brain may occur

along with an increase in microglia activity. These results show the validity of using

plasma gelsolin as a peripheral gene therapy of Alzheimer's disease.














CHAPTER 1
INTRODUCTION

Alzheimer's disease (AD) is the most common form of dementia. In the next

45 years, the number of Americans afflicted with AD is expected to quadruple: from

about 2.5 million cases today, to nearly 10 million in 2050 (Sloane et al., 2002). This

increased prevalence of AD can be attributed to aging of the population: in the year

2000, 5.9% of the population was over age 75; in 2050, this is expected to be 11.4%

(Kawas and Brookmeyer, 2001). Age is the single strongest risk factor for AD. It afflicts

10% of people over 65, and almost 50% of people older than 85 (Evans et al., 1989).

Alzheimer's disease has a huge impact on a person's quality of life (e.g., memory loss,

impaired activities of daily living, depression, and behavioral disturbances) (Sloane et al.,

2002). As the disease progresses, independence decreases, placing an increased financial

and psychological burden on family caregivers. Koppel estimated that in 2001, the total

economic impact of AD in the US was between $183 and 207 billion (2002). With the

expected increased prevalence of AD, finding efficacious treatments will be critical to

ease the social burden of this disease.

Dementia

In Latin "dementia" is defined as irrationality. Medically speaking, dementia

describes a collection of symptoms that robs an individual of his/her cognitive functions,

resulting in the loss of the ability to carry out normal daily activities eventually requiring

the full-time care of family or professionals. These symptoms can be caused by a number

of different diseases that affect the brain. Typically the diagnosis of dementia requires









significant deficits in at least two or more brain functions, such as reasoning, judgment,

perception, language, and memory.

Discovery of Alzheimer's disease

In March 1901 the husband of a 50-year-old woman (Auguste D) noticed a

paranoid symptomalogy in his wife, which rapidly progressed to include sleep disorders,

aggressiveness, crying, confusion, and disturbances of memory. By November, the

deteriorating mental state of Auguste D forced her husband to admit her for inpatient

treatment at the Community Psychiatric Hospital at Frankfurt. A senior assistant at the

hospital, Dr. Alois Alzheimer, thoroughly documented the progression of August D's

symptoms. On her death on April 8, 1906, Alzheimer was able to examine her brain both

histologically and morphologically. His examination showed that Auguste D's brain was

atrophied, and included histological abnormalities later considered the hallmarks of AD,

known as neurofibrillary tangles (NFT) and senile plaques. He described these findings

(along with their relationship to more than 4 years of clinical observations) at the 37th

meeting of South-West German Psychiatrists in Tubingen on November 3, 1906.

Although Alzheimer's lecture was not well received, his observation would be later

recognized as the first demonstrated relationship between clinical history of specific

cognitive changes and neurological lesions at autopsy. After reports of the case of

Auguste D, a number of other patients with similar ailments were described (Moller and

Graeber, 1998). The term Alzheimer's disease was first coined to describe the condition

by a colleague of Alzheimer in Munich, Dr. Emil Kraepelin (1910). Since then

Alzheimer's disease has been recognized as the most common form of dementia

worldwide.









Current Therapies

To date, the only FDA-approved treatments for AD are acetylcholinesterase

inhibitors (tacrine, donepezil, rivastigmine, and galantamine) and an NMDA antagonist

(memantine). The aim of using acetylcholinesterase inhibitors is to enhance selective

cholinergic transmission in the brain by decreasing the catabolism of acetylcholine.

Basal forebrain cholinergic neurons, critical for memory and learning, are diminished in

AD, resulting in a reduction of choline acetyltransferase and acetylcholine (Coyle et al.,

1983; Terry and Katzman, 1983). Increasing the levels of acetylcholine can help

ameliorate deficits in memory and learning (Weinstock, 1995). The aim of using NMDA

antagonist is to block the effects of elevated levels of glutamate which may lead to

neuronal dysfunction (Mattson et al., 1992). Both treatment strategies have shown

modest improvements in maintaining independence, function, and decreasing cost to

society (Trinh et al., 2003; Wimo et al., 2003). However, these modest improvements are

far from ideal and only delay the onset of the inevitable dependency of care by a short

period of time. These treatments only address biochemical symptoms of AD rather than

preventing progression of the underlying cause.

Inherited Alzheimer's

It was not until 75 years after Alzheimer described the case of August D that the

main constituent of the senile plaques was biochemically identified. Allsop and

coworkers (1983) identified that these plaques consisted mainly of a 40-42 amino acid

peptide named amyloid 3 (A3). This peptide was later discovered to originate from a

larger precursor given the name amyloid precursor protein (APP) (Kang et al., 1987).

The subsequent discovery that a single missense mutation on the APP gene corresponded

with an inheritable form of Alzheimer's disease (Goate et al., 1991) led to the









formulation of the amyloid cascade hypothesis, which states the underlying cause of AD

is the result of an increased concentration and deposition of A3 (Hardy and Allsop,

1991). This hypothesis was bolstered by the later discovery that every inheritable form

of AD results from mutations involved in the processing of A3 from APP: either on APP

itself, Presenilin-1, or Presenilin-2 (Clark and Goate, 1993; Levy-Lahad et al., 1995;

Sherrington et al., 1996). Although inherited forms of AD comprise fewer than 10% of

all Alzheimer's cases, every inheritable form involves the facilitation of the

oligomerization and later precipitation of A3.

Amyloid Precursor Protein

In humans, amyloid precursor protein is a large transmembrane glycoprotein that

exists as three major isoforms (APP695, APP751, and APP770) that are all the result of

alternative processing of pre-mRNA generated from the APP gene on Chromosome 21.

The function of APP is poorly understood. However, evidence suggests it may

have cell adhesive, intracellular communication, membrane to nucleus communication,

neurotrophic, or neuroproliferative activity (Turner et al., 2003). Even though APP may

play a role in many biological functions, compensatory mechanisms allow for the

viability of APP knockout mice (Zheng et al., 1996). These mice show reductions in

body weight and synaptic transmission, impaired locomotor activity and grip strength,

and a hypersensitivity to epileptic seizures and forebrain commissural defects (Zheng et

al., 1995).

Three proteolytic cleavage sites have been identified on APP: two near the plasma

membrane on the extracellular side, and one within the plasma membrane. The protease

complexes responsible for the cleavage are known as ac, 3, and y secretase. Cleavage by









a secretase and 3 secretase releases large amino-terminal fragments known as APPs-a

and APPs-P respectively. These fragments differ in size by 17 amino acids at the

carboxy-terminus of the fragments (APPs-a is larger than APPs-P). The remaining

fragments of APP stay anchored to the plasma membrane, and are referred to as C99 for

the P secretase product and C83 for the a secretase product. Both C83 and C99 are

substrates for y secretase which cleaves within the plasma membrane. Amyloid 3 (A3) is

formed after C99 is cleaved by y secretase. This cleavage usually results with a 40 amino

acid length peptide A3 1-40. However, y secretase cleavage of C99 can also result in the

more hydrophobic 42 amino acid peptide product A3 1-42. The A3 1-42 is more prone

to oligomerzation and fibril formation than A3 1-40 (Hasegawa et al., 1999). The

inheritable forms of AD invariably increase the relative amounts of A3 1-42 as compared

to AP 1-40 (Scheuner et al., 1996; Sinha and Lieberburg, 1999; Suzuki et al., 1994;

Tamaoka et al., 1994).

Amyloid 3 is first released from neurons as a soluble monomer which has an a-

helical secondary structure. During the process of oligomerization this undergoes a series

of conformational changes to form cross P-sheet structures as oligomers. These soluble

oligomeric forms of A3 may have a protofibular-like structure (Lashuel et al., 2002) or an

amorphous micellular-like structure (Hoshi et al., 2003). In the fibrillar model for

amyloid deposition the soluble oligomers begin to aggregate forming first protofibrils,

and then fibrils that finally come together to make up the plaques (Figure 1-1).

Amyloid Cascade Hypothesis

One of the central controversies in the AD research community is whether AP is

the cause or result of the pathogenic process. The hypothesis that A3 is central to the









pathogenesis of AD is known as the amyloid cascade hypothesis (Selkoe, 1989, 1990).

This hypothesis states that something causes either an over- production or a decreased

clearance of A3. The increased levels of A3 result in the formation of oligomeric forms

of Ap. These then aggregate and deposit as plaques. This deposition of plaques causes

the activation of microglia and astrocytes, resulting in the release of pro-inflammatory

cytokines and reactive oxygen species (Akama et al., 1998; Hoozemans et al., 2005;

Johnstone et al., 1999). Together microglia activation, astrocytic activation, and

oligomeric A3 can all cause synaptic and neuritic injury, including neurofibrillary

tangles, which then lead to dementia.

Evidence Supporting the Amyloid Cascade Hypothesis

The main evidence supporting the amyloid cascade hypothesis is that every form of

familial Alzheimer's disease (FAD) involves mutations on either APP itself or the

enzymes that cleave APP, resulting in an overproduction of A3 1-42. The presence of an

extra copy of chromosome 21, in which the gene for APP is located, is found in Down's

syndrome; this inevitably leads to an early onset of Alzheimer's-like pathology. In the

more common sporadic form of AD, the presence of the apolipoprotein E4 (apoE4) allele

is considered a risk factor for the disease (Corder et al., 1993); evidence suggests

that apoE is involved with the clearance of A3 (Brendza et al., 2002). Further evidence

supporting the amyloid cascade hypothesis is the fact that, in vitro, A3 itself has been

found to be neurotoxic (Dahlgren et al., 2002; Pike et al., 1991).

Critics of the Amyloid Cascade Hypothesis

Critics of the amyloid cascade hypothesis argue that A3 accumulation may occur

secondary to other pathological events and actually play a role in neuroprotection. They









point out that although specific forms of A3 can be toxic in vitro; this toxicity is less

reliable in animal models, providing the argument that the in vitro toxicity may be an

artifact. They make the point that amyloid deposition is poorly correlated with cognitive

deficits, and that neurofibrillary tangles (NFT) and neuron number are much better

indicators of cognitive decline (Giannakopoulos et al., 2003).

It is true that when A3 deposits in humans are measured histochemically, they do

not correlate well with cognitive decline. However, soluble forms of A3 measured

biochemically from brain extracts correlates very well with synaptic density and can be

used to discriminate between AD patients and non-AD controls that do have a high

degree of amyloid deposit pathology (Lue et al., 1999; Naslund et al., 2000). Total A3

40 and 42 levels of nursing home resident brain extracts measured biochemically has also

been correlated to cognitive decline as measured by the Clinical Dementia Rating (CDR)

scale (Naslund et al., 2000).

There is evidence that A3 may play a role as an antioxidant (Curtain et al., 2001) or

a neurotrophin (Yankner et al., 1990). In fact, Lopez-Tolendano and Shelanski (2004)

recently found that A3 was neurogenic in a dose-dependent manner when treating neural

stem cells. However, Liu et al. (2004) demonstrated that A3's toxicity in primary

neuronal cultures was dependent on the expression of the microtubule-associated protein

tau, cyclin-dependent kinase 5 (Cdk5), and the cell's state of differentiation.

This is an interesting observation, providing evidence for a mechanism of how A3

can cause NFTs. Hyperphosphorylated tau is the main component of NFTs and Cdk5 is

one of the enzymes thought responsible for tau's phosphorylation (Cruz and Tsai, 2004;

Noble et al., 2003). Although both sides can make strong arguments supporting their









point of view, it is unlikely that there will be any consensus over what role A3 plays in

the progression of AD until human trials are completed that target A3.

Elan and Wyeth Trial

The observations that active immunization in transgenic mice dramatically

reduced the accumulation of A3 (Schenk et al., 1999), and showed protection against

memory deficits (Janus et al., 2000; Morgan et al., 2000) led to the phase II trial

undertaken by the pharmaceutical companies Elan and Wyeth in late 2001. They actively

immunized patients against A3, hoping to trigger an immune response that would

increase clearance of A3 from the CNS. These trials were halted early because about 6%

of the patients developed meningoencephalitis (Orgogozo et al., 2003). Follow-up

studies on the participants indicated that antibody responders had significantly improved

memory function as measured by the neuropsychological test battery and decreased

cerebral spinal fluid levels of tau protein (Gilman et al., 2005).

Sink-Hypothesis

One proposed mechanism for how immunization works is that anti-AP antibodies

enter the CNS and stimulate microglial phagocytosis of A3-antibody complexes. This

has been demonstrated by Bacskai et. al. (2001). Another possible mechanism for how

anti-AP antibodies can clear plaques is by shifting the equilibrium of A3 from the CNS to

the periphery. The so called "sink hypothesis" is supported by the finding that less than

0.1% of antibodies in the serum gain access across the blood brain barrier (BBB) (Bard et

al., 2000) and studies performed by separate groups using different A3 binding agents

administered peripherally(Deane et al., 2003; Matsuoka et al., 2003).









For the sink-hypothesis to be valid, carrier or receptor mediated transport of A3

must occur, because the BBB normally prevents free exchange of polar solutes between

blood and brain or brain and blood. The main transporter identified being responsible for

transport out of the CNS is low-density lipoprotein receptor-related protein-1 (LRP-1)

(Shibata et al., 2000). LRP-1 is a large endocytic receptor responsible for the transport of

apoE and cholesterol-containing lipoproteins.

Likewise the receptor for advanced glycation end products (RAGE) has been

identified as a membrane-bound receptor that transports A3 from the circulation into the

CNS (Deane et al., 2003). RAGE is a multiligand receptor in the immunoglobin

superfamily. Generally there is little expression of RAGE in most tissues. However the

accumulation of RAGE ligands, such as A3, triggers RAGE expression, in contrast to a

decrease of LRP-1 expression seen in an A3 rich environment (Shibata et al., 2000).

Exacerbating this effect, RAGE transport of A3 results in the increased expression of

proinflammatory cytokines and endothelin-1 at the BBB causing decreased cerebral

blood flow (Deane et al., 2003).

Deane et al. (2003) demonstrated that when PD-hAPP mice were treated with

intraperitoneal injections of a truncated soluble form of RAGE (sRAGE), A3 transport

into the CNS was interfered with and significant increases in plasma A3 levels along with

a decrease in brain A3 levels and plaque loads was observed. Likewise Matsuoka et al.

saw similar results when they treated PS/APP mice with the A3 binding agents GM1 and

plasma gelsolin (Matsuoka et al., 2003; Morgan et al., 2000).









Gelsolin

Plasma gelsolin is a highly conserved 93 kD actin-binding protein, also known as

brevin or actin-depolymerizing factor, normally found in the plasma at concentrations of

about 179 mg/L (Chauhan et al., 1999). Its main function is thought to be part of the

actin-scavenging system, to protect the microcirculation from the effects of long F-actin

polymers released during cell death (Lee and Galbraith, 1992). However recent evidence

points to the possibility that plasma gelsolin may play a variety of roles in the body,

including mediating inflammatory responses by binding to pro-inflammatory compounds

(Bucki et al., 2005; Bucki et al., 2004; Chauhan et al., 1999; Lind and Janmey, 1984;

Smith et al., 1987), or by altering cell motility and endocytosis(Witke et al., 2001).

There are three known forms of gelsolin [cytoplasmic (Yin and Stossel, 1979),

plasma (Nodes et al., 1987), and gelsolin-3 (Vouyiouklis and Brophy, 1997)] all coded

for by the same gene, resulting from alternative post-transcriptional processing. The

cytoplasmic form of gelsolin was first described as a factor able to solubilize gels formed

by macrophage extracts, hence the name gel-sol-in (Yin and Stossel, 1979). Plasma

gelsolin differs from the other two by the presence of an N-terminal 23 amino acid signal

peptide, which causes plasma gelsolin to be secreted outside the cell producing it.

Gelsolin is regulated by polyphosphoinositide and Ca+2. Gelsolin severs and caps

F-actin in response to Ca+2, and phosphoinositides block the capping function (Bucki et

al., 2004; Kwiatkowski, 1999). It alters cell shape by remodeling actin filaments and is

involved with cell motility (Cooper et al., 1987; Janmey et al., 1987; McLaughlin et al.,

1993). Data from gelsolin knockout mice indicates gelsolin is necessary for rapid motile

responses in cell types involved in responding to stress such as hemostasis, wound

healing, and inflammation (Witke et al., 1995). A mutated form of gelsolin (either









D187N or D187Y) results in aberrant proteolytic cleavage by furin causing a 68kD

gelsolin fragment to be secreted and then deposited as amyloid in the Finish type of

amyloidosis (Chen et al., 2001; Kazmirski et al., 2000).

Gelsolin has been shown to be an effector of apoptosis through its interaction with

the cysteinyl-protease caspase-3. A gelsolin cleavage fragment of caspace-3 has been

shown to cause numerous cell types to "round up, detach from the plate, and undergo

nuclear fragmentation"(Kothakota et al., 1997), most likely the result of the N-terminal

gelsolin fragment's ability to activate DNase-1 (Chhabra et al., 2005).

Gelsolin has also been shown to be protective against excitotoxic induced apoptosis

by altering the actin cytoskeleton in response to Ca+2 influx, preventing the reduction of

the mitochondrial permeability transition pore opening and membrane potential loss, and

preventing caspase-3 activation (Harms et al., 2004).

Plasma gelsolin has been shown to protect against inflammatory reactions

associated with injury (Christofidou-Solomidou et al., 2002a; Rothenbach et al., 2004).

Vasconcellos et al. showed that plasma gelsolin reduced the viscosity of cystic fibrosis

sputum (1994). In fact Biogen Inc. evaluated recombinant plasma gelsolin in phase 2

clinical trials as a treatment for cystic fibrosis.

Chauhan et al. showed that human plasma gelsolin binds to A3, prevents

fibrillization, and disassembles preformed A3 fibrils, suggesting a possible role for

gelsolin in the clearance of amyloid 3 (Chauhan et al., 1999; Ray et al., 2000). Plasma

gelsolin is found in the cerebral spinal fluid (CSF), and produced in the choroid plexus

(Matsumoto et al., 2003). At the 2004 International Conference on Alzheimer's and

Related Disorders Chauhan et al. reported that gelsolin levels in AD patients' CSF was






12


significantly reduced as compared to nondemented age matched controls, suggesting that

decreased gelsolin may play a role in the increased amyloid 3 content seen in AD (2004).

Matsuoka et al. showed injections with bovine plasma gelsolin can prevent deposition of

gelsolin in younger huAPP K670N,M671L/ PS-1 M146L (2003). The focus of our study will be

on the effects plasma gelsolin gene expression has on amyloid deposition in transgenic

mouse models of Alzheimer's disease.




















COOH p secretase




4C99
C99 I d


y cut sites


COOH


COOH r

y secretase


H2



Amyloid 3
(a helix) oligomerization

Soluble
Amyloid P
'OOH Oligomers
(P sheet)



Figure 1-1. Amyloid precursor protein processing.


Plaque formation


I T "' .















CHAPTER 2
MATERIALS AND METHODS

Reagents

Except where noted, all chemicals used were purchased from Fisher Scientific

(Hampton, NH). Molecular biology reagents and enzymes were purchased from New

England Biolabs (Ipswich, MA). Amyloid peptides were purchased from Anaspec (San

Jose, CA). ThePCR primers were ordered from Sigma Genosys (The Woodlands, TX).

The ELISA kits were purchased from Biosource (Camarillo, CA). Precast

polyacrilamide gels and PVDF membranes were purchased from Bio-Rad Laboratories

(Hercules, CA)

Subcloning Vectors

An expression plasmid for plasma gelsolin, pPGLE (Figure 2-1A), that is based

on the commercially available plasmid (pCDM8) was kindly provided by Dr. Hisakazu

Fujita (Kwiatkowski et al., 1989). The coding sequence for plasma gelsolin was removed

from the pCDM8 backbone by a HindIII and Xbal digest, followed by separation on a 1%

agarose gel, and purification using a Qiagen gel purification kit. Blunt ends were made

by treating with T4 DNA polymerase in the presence of dNTPs. HindIII linkers were

ligated to the blunted ends using T4 DNA ligase. Samples were then run on a 1%

agarose gel. The blurred band containing the gelsolin insert, along with different

amounts of linkers, was purified from the gel using a Qiagen gel purification kit. This

purified band was then subject to a HindIII digest.









A plasmid backbone containing the cytomegalovirus/chicken beta-actin hybrid

(CBA) promoter and the woodchuck hepatitis virus post-transcriptional regulatory

element (WPRE) was prepared from our pGFP vector (Figure 2-1B)((Klein et al., 2002)

by excising the Green Fluorescent Protein (GFP) coding sequence with a HindIII digest

followed by separation on a 1% agarose gel; and was then purified using a Qiagen gel

purification kit. The 5'phosphate groups were removed with calf intestine alkaline

phosphatase in order to prevent self-ligation.

The purified backbone along with the plasma gelsolin insert were ligated together

overnight with T4 DNA ligase. Electrocompetent SURE cells (Stratagene, Garden

Grove, CA) were transformed with the resultant ligated product, using a Bio-Rad

electroporator, with the resistance set at 400 Q, the capacitance at 25 pF, and the voltage

at 2.2kv. Transformed bacteria were grown for an hour in ImL of NZY broth at 370C,

followed by plating on NZY agar plates containing ampicillin (50mg/L), and then grown

overnight at 370C. Several colonies were selected for screening, and each was grown

overnight in 5 mL of ampicillin-containing NZY broth. Plasmids were purified from the

cultures, using Qiagen mini plasmid prep kits. Plasmids were then subjected to a BglII

digest to confirm orientation of the insert (Figure 2-2). A clone (W16) with the forward

insert was given the name pUFGL (Figure 2-1C).

Large Scale Plasmid Preparation

For large-scale preps, plasmids were propagated overnight in 5 mL of ampicillin

containing NZY broth. This was used to inoculate 2 L of ampicillin containing NZY

broth, and then grown overnight again. The cultures were pelleted by centrifugation.

They were then resuspended in a lysozyme buffer (80 mL/L of culture) and treated with

lysozyme (2 mg/mL, Sigma, St. Louis, MO). Next the cultures were subject to alkaline









lysis by adding a 1% SDS 0.2N NaOH solution at 196 mL/L of culture. The mixture was

neutralized with a 3M NaAc (pH4.8) 0.6% chloroform solution (144 mL/L of culture).

The resultant chromosomal and protein precipitates were separated by centrifugation.

Plasmid DNA in the supernatant was then precipitated by bringing the solution to

10% polyetheleneglycol (PEG). The precipitates were separated by centrifugation, and

then resuspended in distilled water (40 mL/L of culture). The RNA was precipitated and

separated by adding 5.5 M LiCl (40 mL/L of culture), followed by centrifugation.

Plasmid DNA was precipitated from the supernatant by bringing the solution to 36%

isopropanol. Precipitated plasmid DNA was resuspended in a 5.3M CsCl solution

containing 1 mM ethidium bromide. The resultant solution was centrifuged in a

Beckman 70.1 Ti rotor at 55,000 rpm for 19 hours. The lower band containing the

plasmid DNA was removed using an 18 g syringe. Ethidium bromide was removed by

performing four extractions with isoamyl alcohol.

The plasmid DNA was precipitated from the aqueous layer by bringing the

solution to a 40% Ethanol concentration. The plasmid precipitate was pelleted by

centrifugation, and resuspended in TE (10 mM Tris-HCl 1 mM EDTA pH 8.0). Any

residual protein contamination was removed from this solution by performing four

extractions with phenol-chloroform, followed by one extraction with chloroform alone.

Plasmid DNA was precipitated once again by adding one tenth volume of 3 M NaAC pH

4.8 and 2.5 volumes of 100% ethanol. Precipitates were pelleted by centrifugation,

followed by washing excess salt with a 75% ethanol solution. The plasmids were

resuspended in sterile TE buffer. Concentration and purity of the samples were

determined by UV absorbance at 260/280 nm.









Cell Culture

Human embryonic kidney (HEK) 293 cells were cultured in DMEM with 10%

fetal bovine serum (FBS), 100 units/mL penicillin and 100 [tg/mL streptomycin (Gibco,

Invitrogen, CA) in a 5% CO2 incubator at 370C. Cells were grown on 10cm dishes to

50% and 80% confluency. One half hour before transfection using the CaPO4

precipitation method, culture media was replaced with fresh media. To prepare

transfection complexes 20 |tg of either pUFGL or pGFP in 700 [tl of 250 mM CaC12

solution was added in a dropwise fashion to 700 [tl 2X HEPES Buffered Saline while

vortexing slowly (for 2X HBS, 300 mM NaC1, 1.8 mM Na2HPO4, 11 mM dextrose, and

40 mM HEPES, pH 7.12).

The transfection solutions were then incubated at room temperature for 20

minutes. Following incubation, the transfection solutions were mixed gently then added

to cell culture dishes in a drop wise fashion. Culture dishes were incubated with

transfection media for 12 hours then replaced with fresh media. Forty-eight hours later,

media was collected, a proteinase inhibitor cocktail was added (Sigma P8340), and it was

either used for immunoprecipitation, or concentrated ten fold with Centricon 50,000 nmw

cutoff centrifugal filters for western blot analysis.

Animals and Procedures

All procedures were done with prior approval and oversight of the University of

Florida's institutional animal use and care committee. Double transgenic mice

expressing both Swedish mutant mouse /human APP695 K594N,M595L and exon 9 deleted mutant

presenilin-1 (mo/huAPP /PS16E9) were supplied by Jackson Laboratories (Bar Harbor, ME)

(Jankowsky et al., 2004). Double transgenic mice expressing both Swedish mutant









human APPK670N,M671L (huAPP K670N,M671L, Tg2576)(Hsiao et al., 1996) and mutant

presenilin-1 M146L (PS-I M146L)(Duff et al., 1996), and transgenic mice expressing only

mutant PS-1 M146L were supplied by The Nathan Kline Institute (NY,NY).

One litter of mo/huAPP /PS1lE9 mice was aged until 36 weeks, at which time the

mice were either injected with our test plasmid pUFGL (n=3) or left untreated (n=3).

Mice expressing huAPP K670N,M671L/ PS-1 M146L were aged until 32 weeks, at which time

they were either injected with plasmid DNA, pUFGL (n=5), pGFP (n=3), or left

untreated (n=7). PS-1 mice used for western blot detection of plasma gelsolin were

injected at the age of 36 weeks with either pUFGL (n=2), pGFP (n=2), or left untreated

(n=2).

For injections, plasmid DNA was diluted in lactated Ringer's solution to a

concentration such that there was 25 |tg of plasmid DNA/10% of body weight volume of

ringers (e.g. a 30 g mouse received a 3 mL injection). DNA solutions were warmed to a

temperature of 37C. Animals were warmed briefly under a heat lamp, mildly

anesthetized with isoflourane, and restrained in a custom-made harness. A three milliliter

syringe was used with a 27 g half inch needle. The injection solution was injected in a

time period of 5-10 seconds. Animals were then recovered on a heating pad, and

returned to their cages.

Blood samples were taken from PS-1 mice from the retro-orbital sinus using

heparinized capillary tubes. Animals were mildly anesthesized with isoflourane, and then

the capillary tube was used to puncture the retro-orbital sinus. The tube was allowed to

fill with blood, and then plugged on one end with clay. Plasma was separated









immediately by centrifugation, snap frozen in liquid nitrogen, and stored at -80o C for

analysis by western blot.

Two and one half weeks following injections, double transgenic mice were

sacrificed along with age-matched untreated controls. Animals were deeply anesthetized

with isoflourane and perfused with PBS. Livers and brains were excised, hemi-brains

and a sample of liver were snap frozen in liquid nitrogen and stored at -800 C for analysis

by ELISA and RT-PCR. The remaining hemi-brain and liver tissue was fixed for 48

hours in a 4% paraformaldehyde in PBS solution, and then equilibrated in 30% sucrose in

PBS solution for cryoprotection.

Immunoprecipitation and Western Blot

An immunoprecipitation kit from Sigma (IP-50, St. Louis, MO) was used for the

immunoprecipitation reactions. 600 p.L of media from cell cultures with or without being

spiked with A3 1-42 to 5 [tM along with 2 [tl anti-Gelsolin monoclonal antibody clone

GS-2C4 (Sigma G-4896, St. Louis, MO) was added to the spin columns provided with

the kit and incubated overnight at 40 C. 30 tiL/column of protein-G agarose was washed

in 1X IP buffer, then resuspended in 50 ptL of 1X IP buffer and added to the samples in

the columns. These samples were then incubated overnight at 40C. The tips were broken

from the columns, then the columns were centrifuged and the effluent was discarded.

The beads in the columns were washed five times with 1X IP buffer, followed by a sixth

wash in 0.1X IP buffer. 50 ptL of Laemmli sample buffer containing 5% 3-

mercaptoethanol was then added to the beads and incubated at 950 C for ten minutes.

Samples were spun through the columns and then 50 [tl loaded onto a precast SDS 4-20%

PAGE Tris-HCl gel.









For Western blots of concentrated media 25[tL of sample was mixed with 25 ptL

of 2X Laemmli sample buffer with 5% P-mercaptoethanol, boiled for 5 minutes, and then

loaded onto a precast SDS 10% PAGE Tris-HCl gel. For western blot analysis of plasma

samples 4[tL of plasma was diluted in 21 ptL of distilled water and then mixed with 25

tL of 2X Laemmli sample buffer with 5% P-mercaptoethanol, boiled for 5 minutes, and

then loaded onto a precast SDS 10% PAGE Tris-HCl gel. Gels were run using a Biorad

power supply set at 100 V for one hour.

Separated proteins were then transferred to a PVDF membrane at 75V for 2 hours

on ice. Membranes were incubated overnight in a blocking solution (5% nonfat dry milk

and 0.05% Tween 20 in PBS) at 40C. Primary antibodies were then added [for anti-A3,

6E10 from Chemicon(Temecula, CA) was used at 1:1000 dilution, for anti-gelsolin GS-

2C4 from Sigma was used at 1:1000 dilution] and incubated at room temperature for 2

hours. Membranes were washed three times in PBS with 0.05% Tween 20, and

incubated with horseradish peroxidase (HRP)-conjugated anti-mouse antibody

[Amersham (Piscataway, NJ) at 1:5000] in blocking solution for 1 hr at room

temperature. Following three more washes, they were incubated with substrate

[electrochemiluminescence (ECL), Amersham(Piscataway, NJ)] for 1 min and exposed

(Kodak, Rochester, NY).

Detection of Message

A Qiagen (Valencia, CA) RNeasy mini-kit was used for RNA extractions. 30 mg

of frozen liver or brain tissue was homogenized on ice using a Polytron homogenizer

(Brinkmann Instruments, Westbury, NY) in 600 mL of buffer RLT containing 3-

mercaptoethanol. RNA was purified and washed using the columns and reagents









provided by the Qiagen kit as recommended by the manufacturer. RNA was eluted from

the column to a final volume of 60[tL in RNase-free water provided with the kit.

Primers were designed to yield -900 b.p. product from mRNA transcribed from

our vector pUFGL or ~1800b.p. product from unprocessed RNA or DNA contamination,

by having the forward primer (GGC TCT GAC TGA CCG CGT TTA C, Tm = 68.70C)

anneal to sequence from Exon 1 in the vector, and reverse primer (CTG TTG GAA CCA

CAC CAC TGG, Tm = 67.70C) anneal to sequence from the coding region of gelsolin.

Primers for P-Actin (ATG AGG TAG TCT GTC AGG T, Tm = 52.90C, & ATG GAT

GAC GAT ATC GCT G, Tm = 52.70C) were used as a positive control.

A Qiagen one-step RT-PCR kit was used for the RT-PCR reaction. One microliter

of RNA was used in each 25 ptL reaction with final primer concentrations of 0.6 [LM, and

Q solution was included in the mixture. A MJ Research PTC-200 Peltier Thermal Cycler

was used for the RT-PCR reaction. Reverse transcription was done for 30 minutes at

500C, followed by PCR activation at 950C for 15 minutes. Next came thirty cycles that

consisted of: denaturation for one minute at 940C, annealing for 30 seconds at 500C, and

extension for two minutes at 720C. There was one final extension for 10 minutes at

720C, and then samples were held at 40C. Samples were then loaded onto a 2% agarose

gel and run at 85 volts. Bands were then imaged by ethidium bromide staining.

Enzyme Linked Immunosorbent Assay

For enzyme linked immunosorbent assay (ELISA) Biosource colorimetric

immunoassay kits were used for both 3 amyloid 1-40 and 1-42. Frozen hemi-brains were

weighed. Eight times the mass of 5 M guanidine HC1, 50 mM Tris HC1, pH 8.0 was

added to the brains then homogenized using a Polytron homogenizer (Brinkmann









Instruments). The homogenate was mixed at room temperature for four hours, and then

aliquotted and stored at -800C. The guanidine extracted homogenates were diluted

3000 times in BSAT-DPBS [Dulbecco's phosphate buffered saline with 5% bovine

serum albumin, 0.03% Tween-20 and 1 mM 4-(2-Aminoethyl)benzenesulfonylfluoride

(AEBSF)]. Mixtures were centrifuged at 16,000 G and 40C for twenty minutes. The

supernatants were diluted four fold in the standard diluent buffer provided with the

Biosource kit, with AEBSF at ImM. Samples were then incubated on a shaking platform

at room temperature for two hours in the wells provided with the Biosource kit with an

equal volume of primary antibody solution. Samples were then washed four times, and

incubated in HRP solution four one half hour. Samples were washed four times again

and then incubated for a half hour with HRP substrate, in a box to protect the samples

from light. Stop solution was then added and absorbance at 450 nm was measured using

a Dynex Technologies MRX microplate reader. Concentrations of Amyloid 3 (1-42)

were determined from standards provided with the kit.

Histochemistry

Coronal sections (50 |tm thick) were cut from the hemi-brains on a sliding

microtome with a freezing stage. Four sections, six sections apart each, were mounted on

slides for thioflavine S staining. Sections were allowed to dry on the slides for 15

minutes. The slides were then placed in deionized water for five minutes. They were

then placed in filtered Mayer's Hematoxylin for five minutes. Next, the slides were

rinsed under running tap water for five minutes, followed by a five minute rinse in

deionized water. The slides were then placed in a 1% thioflavine S solution (in dH2O,

filtered, Sigma) for five minutes. The slides were differentiated in 70% ethanol for five









minutes, given short rinses in deionized water followed by PBS, and cover slipped with

glycerol gelatin (Sigma).

Immunohistochemistry

Antigen detection on 50 |tm thick coronal hemibrain sections was conducted on

free-floating sections by incubating the sections overnight at 40C in blocking solution

(3% goat serum, 0.3% Triton X-100, 0.05% azide in PBS). Endogenous peroxidase

activity was quenched by incubating the sections for 10 minutes in 0.5% H202 in PBS at

room temperature prior to blocking.

Primary antibodies used were: 6E10 (1:1000, Signet, Dedham, MA), OX-42

(1:200, Serotec, Raleigh, NC), and anti-amyloid oligomer (1:250 Chemicon). Sections

were incubated with primary antibodies diluted in blocking solution at 40C for three days.

Sections were then washed with PBS three times for 5 minutes each wash. Then the

sections were incubated overnight at 40C with secondary antibody (biotinylated anti-

mouse IgG or biotinylated anti-rabbit IgG, 1:1000, Dako, CA) diluted in blocking

solution. Sections were again washed three times in PBS. Next the sections were

incubated for two hours at room temperature in PBS with ExtrAvidin peroxidase (HRP)

conjugate (1:1000, Sigma). Washing was performed again, and then development of

tissue labeled with HRP was performed with a solution of 0.67 mg/mL diaminobenzidine

(DAB, Sigma) and 0.13 tL of 30% H202 per mL of 80 mM sodium acetate buffer

containing 8 mM imidazole and 2% NiSO4.

The sections were mounted on Superfrost plus microscopic slides (Fisher, NH), air

dried and dehydrated by passing through water, followed by 70%, then 95%, and 2









changes of 100% ethanol. Then they were passed through two changes ofxylene and

coverslipped with Eukitt (Calibrated Instruments, NY).

For immunofluorescence Alexa Flour 555 goat anti-rabbit IgG (H+L) (1:2000,

Molecular Probes, OR), Alexa Flour 488 goat anti-mouse IgG (H+L) (1:1000,

Molecular Probes), AMCA conjugated F(ab')2 fragment goat anti-mouse IgG, F(ab')2

fragment specific (1:100, Jackson ImmunoResearch Laboratories, PA) secondary

antibodies were used diluted in blocking solution following the primary antibody

incubation. Nuclear counter staining was performed by incubating the sections in 4',6-

diamidino-2-phenylindole (DAPI, 1 [tg/mL, Sigma) for 15 minutes at room temperature.

Fluorescent slides were cover slipped with glycerol gelatin mounting medium (Sigma).

Image Analysis

For percent amyloid burden measurements (both dense cored and diffuse) sections

were analyzed in a blinded manner using the NIH Image J software. Regions of interest

(ROI) were created encompassing both the hippocampus and neocortex of digital

micrographs of each stained section. The ROI's area was measured in pixels2. The

number of plaques stained, plaque sizes (in pixels2), and total stained areas in the

hippocampus and cortex (in pixels2) were determined by thresholding segmentation.

Total stained areas were divided by total area, and then multiplied by 100% to give the

percent amyloid burden.









T7 Pro Splice & poly A
promoter Plasma 40 O
M13 Ori prmoerlsolin

A

pPGLE
(6515bp)
CBA promoter
pGFP Poly A ColE1 or
CMV enhancer Exon 1 fl(+) origin
\ \ \i Int WPRE TR ApR
TR Intron
B

pGFP
(6531bp)
CBA promoter
Poly A ColEl ori
CMV enhancer Exon 1 Plasma fl(+) origin
TR Gelsolin WPRE TR o ApR



pUFGL
(8468bp)
Figure 2-1 Vectors, pPGLE kindly provided by Dr. Hisakazu Fujita (A), The coding
sequence for plasma gelsolin was excised and then inserted into the CBA
promoter and WPRE containing backbone from the control plasmid pGFP (B)
to make our test plasmid pUFGL (C).



W8 W9 W12 W16







Figure 2-2 Restriction digest to confirm orientation of clones. W8, W9, and W12 are all
clones representing antisense orientation, having bands of 3173, 3013, 1137,
& 1088 base pairs long. W16 is a sense clone given the name pUFGL having
bands of 3173, 2103, 1998, & 1137 base pairs long.














CHAPTER 3
THE INTERACTION OF PLAMA GELSOLIN AND AMYLOID 3

Introduction

Studies by Chauhan et al. showed that human plasma gelsolin binds to amyloid 3,

prevents fibrillization, and disassembles preformed amyloid 3 fibrils, suggesting a

possible role for gelsolin in the clearance of amyloid 3 (Chauhan et al., 1999; Ray et al.,

2000). Chauhan used a solid phase binding assay to measure the dissociation rate

constants (Kd) for two binding sites on human gelsolin, and found them to be 1.38 and

2.55 pM. Matsuoka et al. showed injections with bovine plasma gelsolin can prevent

deposition of amyloid 3 in young huAPP K670N,M671L/ PS-1 M146L (2003). Our hypothesis

for how gelsolin prevents amyloid 3 deposition was based on its ability to bind amyloid 3

in the periphery shifting its equilibrium from deposition in the CNS to clearance in the

periphery. Chauhan's measurement of human plasma gelsolin's Kd indicates human

plasma gelsolin does not have a very high affinity for amyloid P. On the other hand

Matsouka's use of bovine plasma gelsolin in mice showed encouraging results. The

different species forms of gelsolin used in Chauhan's studies and Matsuoka's study

triggered us to ask the question of whether bovine gelsolin has a higher affinity for

amyloid 3 than the human form of gelsolin. In this chapter we further characterize the

interaction between amyloid 3, and human and bovine plasma gelsolin.









Experimental Methods and Results

Measurement of 125I Labeled AP (1-42) Binding to Recombinant Human Plasma
Gelsolin

An attempt to measure the binding affinity of 125I labeled A3 1-42 to human plasma

gelsolin was carried out using standard binding protocol. Labeled A3 at concentrations

of 2.0, 5.0, and 10.0 tlM were incubated with or without 2 tlM recombinant human

plasma gelsolin in PBS for one day. The following day the samples were run through

Whatman GF/C filters that had been preincubated for 30 min with 0.5%

polyethylenimine, followed by three washes with cold Krebs Ringer buffer (KRB;

118 mM NaC1, 5 mM KC1, 10 mM glucose, 1 mM MgC12, 2.5 mM CaC12, 20 mM

HEPES; pH 7.5). The filters were counted for radioactivity, and it was found that the

samples with plasma gelsolin had lower counts (Figure 3-1).

Because the molecular weight of plasma gelsolin is twenty times that of amyloid

3 we were expecting that plasma gelsolin would bind the glass filters and in the presence

of amyloid 3 more radioactivity would be detected and be representative of how much

amyloid 3 is binding to plasma gelsolin. However it is apparent from our results that a

significant amount of amyloid 3 itself binds to the glass filters and in the presence of

plasma gelsolin less amyloid 3 binds to the filters. This experimental approach was

subsequently abandoned; however theses results are discussed further later.

Measuring Binding of HiLyte Fluor 488 Labeled Amyloid P (1-42) to Bovine Plasma
Gelsolin Using Fluorescence Anisotropy

Another approach at measuring the binding of amyloid 3 (1-42) to plasma gelsolin

is to use the property of fluorescence anisotropy. The change of orientation of a

population of fluorophores, from that of a specific orientation, isotropy, to a random









orientation, anisotropy can be measured by monitoring the rate at which fluorescence

decays in a population of fluorophores when observed through polarized filters and is

termed the fluorescence anisotropic decay. This change occurs via Brownian rotational

diffusion. This property can be expressed using Equation 3-1 as the molecular diffusion

coefficient (Dr) and is dependent on absolute temperature (T), the viscosity of the solution

(r), and the molecular volume (V); R is the gas constant (Weber, 1953).

RT
Dr = (3-1)
6r1V

Protein-protein interactions can be measured by fluorescence anisotropy by

observing changes in the rotational molecular motion due to the increase in molecular

volume when two or more species associate with each other. The fluorescence

anisotropy can be expressed as a function of molecular volume using Equation 3-2

(Perrin's equation) (Lakowicz, 2002). The value of anisotropy in the absence of rotational

diffusion is defined as Ao; and c is the fluorescence lifetime of the fluorophores. As the

molecular volume increases (V), as when two or more proteins bind to one another, the

numerator of Equation 3-2 will decrease, increasing the value of the anisotropy.


A = A (3-2)
1+ RT-r/V

Understanding the principles of fluorescent anisotropy, we decided use the change

in anisotropy to measure the binding of fluorescently labeled amyloid 3 (1-42) ( HiLyte

Fluor 488 labeled amyloid 3 (1-42) from Anaspec). We first used the monoclonal

antibody 6E10 (Chemicon) as a positive control for our binding study. The fluorescent

amyloid 3 concentration was kept between 547-516 nM the fluorescence measurements

through out the assay maintained values of 90-100 relative units (RU). The 6E10 was









added stepwise with a starting concentration of 0 nM and a final concentration of

369 nM. Results show a clear sigmoidal relationship (Figure 3-2) indicating a saturatable

binding of amyloid 3 (1-42) to 6E10, validating our approach of using anisotropy to

measure the binding of amyloid 3 (1-42).

The same experiment was then repeated with bovine plasma gelsolin (Sigma)

substituted for the 6E10. The fluorescent amyloid 3 concentration was kept between

547-526 nM and the plasma gelsolin was added stepwise with a starting concentration of

0 nM and a final concentration of 423 nM. As the gelsolin was added, instead of rising,

the anisotropy fell from about 0.060 to about 0.048, at the same time the fluorescence

rose dramatically from 51 RU and then leveled off at about 210 RU(Figure 3-3). A

hypothesis was formulated as a result of the data: the anisotropy decreased with an

increasing concentration of gelsolin was representative of gelsolin disassembling

oligomeric forms of amyloid 3 and the fluorescence increase is representative of more

fluorophores interacting with the aqueous phase of the solution, rather than being tied up

into hydrophobic areas of the oligomers.

Using a sample from the experiment an attempt was made to separate the free

amyloid 3 from the amyloid 3 bound to gelsolin, to determine if a Scatchard analysis was

possible, using ultra-filtration using Centricon filters with a 30 kD nmw cutoff. Prior to

the spin the volume was brought up to 1 mL by adding distilled water, making the

concentration of gelsolin 418 nM. To control for A3 adsorbing to the filter material, the

same amount of amyloid 3 was added to 1 mL of water ( 0.4 |tg making the final

amyloid 3 concentration 82 nM) with no gelsolin. The fluorescence of the samples was









measured prior to centrifugation and afterwards the retentate and the flow through were

measured (Table 3-1).

After filtration the retentate was taken from the sample containing only amyloid 3

and gelsolin was added in a stepwise fashion, bringing the concentration of gelsolin from

0-1758 nM. As the gelsolin was added the fluorescence increased from about 1 to 9,

while the anisotropy declined from 0.363 to 0.118 (Figure 3-4).

HiLyte Fluor 488 Labeled AP (1-42) Fibril Formation

The fibrillation of amyloid 3 is thought to be a major event in the pathology of AD.

As Ap is formed it is soluble and has an a-helical confirmation. Fibril formation

involves a conformational change to a cross P-pleated sheet structure, oligomerization,

followed by aggregation. In vitro, both synthetic 3 1-40 and synthetic amyloid 3 (1-42)

at 100 [tM form fibrils spontaneously within 48 hours(Wegiel et al., 1996). In this

experiment we monitor the fibril formation of HiLyte Fluor 488 labeled amyloid 3 (1-42)

as a decrease in fluorescence. A 3.2 [LM solution of HiLyte Fluor 488 labeled amyloid 3

(1-42) (chosen due to availability of labeled peptide) in distilled water was incubated at

room temperature. A rapid decrease in fluorescence of the solution was seen along with

the appearance of visible precipitates (Figure 3-5).

HiLyte Fluor 488 Labeled Ap (1-42) Fibril Disassembly with Gelsolin

It has been well demonstrated that human plasma gelsolin has the ability to

disassemble preformed Ap fibrils (Ray et al., 2000). The administration of bovine

plasma gelsolin has prevented amyloid deposition in transgenic mice (Matsuoka et al.,

2003). Due to theses observations we decided to determine if there are species'









differences between bovine plasma gelsolin and human plasma gelsolin in the ability to

disassemble preformed A3 fibrils.

Under sterile conditions, dialysis tubes with a 10 kD molecular weight cutoff were

used as a membrane in order to separate the monomer or dimer forms of labeled

amyloid 3 1-42 from more aggregated oligomeric forms. Inside the dialysis tubesl00 [tl

of 1.6 [LM fibrillized HiLyte Fluor 488-Labeled amyloid 3 (1-42), from the previous

experiment, was added. On the outside of the dialysis tubes 200[tl of distilled water was

added.

The experimental groups included adding 0.5, 1.0, 2.0, 3.0, 5.0, and 8.0 |tg of

human plasma gelsolin or 1.0, 2.0, 3.0, 5.0, and 8.0 |tg of bovine plasma gelsolin on the

inside of the dialysis tubes, with the amyloid P. As a control, only amyloid 3 inside the

dialysis tube and no plasma gelsolin, was used. To determine the maximum amount of

diffused labeled amyloid 3 possible, a group with only labeled amyloid 3 (1-42) on the

inside of the dialysis tube was used with 5M urea in distilled water on the outside of the

tube for complete disassembly of amyloid 3 fibers to amyloid 3 monomers,

The dialysis reactions were set up in triplicate under sterile conditions and allowed

to incubate at room temperature for 1 week. At this point the fluorescence was measured

outside the dialysis tubes, revealing the amount of amyloid that was diffusible across the

filters. This experiment was repeated twice using triplicate samples each time. The

results are displayed in Figure 3-6.









Measuring Binding Amyloid 3 to Plasma Gelsolin Using Surface Plasmon
Resonance

Surface plasmon resonance (SPR) is method that can be employed to observe

interactions between macromolecules by measuring local changes in the refractive index

of a solution containing a substrate flowing across a metal surface to which a ligand has

been attached.

Amyloid 3 1-40 and 1-42 were coupled to a gold chip modified with

carboxymethylated dextran layer, Biacore CM5 chip, using amino-coupling chemistry.

Reactive sucinimide esters were produced on the surface of the CM5 chip by using a 1:1

mixture of 1-ethyl-3-[3-(dimethylamino)propyl]carbodiimide hydrochloride (EDC) and

N-hydroxysuccinimide (NHS). Then an acetate solution, pH 4, containing 50[tg/ml

amyloid 3 1-40 (Figure 3-7) or 1-42 (Figure 3-8) was passed across the activated surface

of the CM5 chip. Free amino groups in the amyloid are able to react with the activated

surface of the CM5 chip becoming covalently bound. Following attachment of the

amyloid 3 a high concentration of ethanolamine was passed over the surface to block any

unreacted carboxymethyl groups on the surface of the chip. For amyloid 3 1-40 an

increase of 650 resonance units (RU) was observed, and for amyloid 3 1-42 an increase

of 3880 RU was observed.

Following coupling of amyloid 3 to the CM5 chip a solution containing 50 [tg/ml

of the antibody 6E10 was injected across both surfaces bound with amyloid 3 1-40 and 1-

42 for 240 seconds, and also a blank surface of the chip, as a positive control for binding.

The resultant sensorgram is displayed in Figure 3-9. This shows that the 6E10 binds to

both surfaces coupled with amyloid 3 1-40 or 1-42 showing a response of 4000 RU and









5000 RU respectively, and no binding to the blank surface. The off-rate for 6E10 appears

quite slow with no loss of response occurring after cessation of injection. The surface of

the chip was then regenerated with a solution of 4M guanidine HCL removing any

unconjugated protein or peptide. Following regeneration a solution containing 50 [tg/ml

of human plasma gelsolin was injected across the surfaces for 240 seconds. The resultant

sensorgram (Figure 3-10) shows response increases of about 750 RU for the amyloid 13 1-

40, 1-42, and the blank channel. This indicates that the observed response results from

bulk flow changes resulting from differences in buffer composition that cause changes in

the refractive index. Specific binding of gelsolin to the coupled amyloid 3 is difficult to

detect in this case.

Another CM5 chip was then coupled with human plasma gelsolin using the same

amino coupling chemistry (EDC/NHS) as described for the amyloid 3 coupling. The

coupling procedure resulted in a final increase of 1650 RU for the surface of the chip

(Figure 3-11). Following the coupling procedure with human plasma gelsolin a solution

containing 30 second injection of 50 mg/ml of GS-2C4 antibody (recognizes human

plasma gelsolin) was injected across the surface of the human plasma gelsolin coupled

CM5 chip (Figure 3-12). There was a bulk increase in response of 225 RU, with only a

small amount of response increase due to GS-2C4 binding (-40 RU). After binding with

GS-2C4 antibody the surface of the chip was regenerated with 4M guanidine HCL.

A solution of 50 mg/ml of amyloid 3 1-40 was then injected across the surface of

the chip. After a bulk increase in response of 8000 RU, the increase in response

attributable to amyloid b binding was 20 RU (Figure 3-13). In order to confirm the

presence of bound amyloid 3 1-40, 50 [tg/ml of 6E10 was then injected across the









surface. 6E10 resulted in an increase in response of- 1200 RU. The surface was then

regenerated with 4M guanidine HCL. A solution containing 50 [tg/ml of amyloid 3 1-42

was then injected across the surface of the chip again resulting in a bulk response of

-8000 RU, and a response increase attributable to binding of amyloid 3 1-42 of- 30 RU

(Figure 3-14). 6E10 was again used to confirm the presence of bound amyloid 3 1-42,

resulting in a response increase of -1020 RU.

Conclusions

The preceding experiments were undertaken in an effort to determine if there were

species differences in the binding affinities of bovine and human plasma gelsolin.

Human gelsolin had been shown to bind to and disassemble amyloid fibrils (Chauhan et

al., 1999) and bovine gelsolin had been shown to have an effect In Vivo in mice that

deposit a human form of amyloid 3 (Matsuoka et al., 2003). We hypothesized that

plasma gelsolin may be an effective peripherally expressed gene therapy for Alzheimer's

disease based on a sink-hypothesis, that binding amyloid 3 peripherally will shift the

equilibrium of amyloid 3 from depositing in the CNS to the periphery where it can be

cleared. So we set out to measure binding affinities in order to determine if one species

form would be advantageous over the other.

Although we were unable to measure specific on and off rates of the binding

interaction between plasma gelsolin and amyloid 3, our results demonstrate that the

presence of plasma gelsolin alters the binding activity of amyloid P. The data from the

125I labeled amyloid 3 (1-42) binding experiment (Figure 3-1) and centrifugal filtration

experiment (Table 3-1) both demonstrate that in the presence of gelsolin amyloid 3 has

an improved ability to pass through either 30 kD filter or glass filter paper. This is most









likely due to gelsolin's ability to disassemble amyloid 3 fibrils. Our data shows that

HiLyte Fluor 488 Labeled amyloid 3 (1-42) forms insoluble fibril precipitates (Figure 3-

5B). While this occurs the fluorescence activity decreases (Figure 3-5A), most likely due

to fluorophores being sequestered in the non-aqueous phase of the precipitates.

As gelsolin is added to the amyloid 3 fibril suspension there is a fluorescence

increase that is accompanied by a decrease in the anisotropy (Figures 3-3 and 3-4). This

decrease in anisotropy can be attributed to an increase in rotational diffusion caused by a

decrease in size as amyloid 3 fibrils are disassembled. The fact that gelsolin can do this

across a membrane (Figure 3-6) suggest that gelsolin may have enzyme-like activity

shifting the equilibrium from amyloid 3 fibril formation to soluble a-helical amyloid 3

monomers.

Gelsolin does have similar enzyme-like activity with its interaction with actin. It

disassembles actin filaments and can cap actin monomers in preparation for actin

filament elongation. If gelsolin's interaction with amyloid 3 is enzyme-like, it may have

a high affinity for oligomeric amyloid 3 that has a P-sheet secondary structure, and a low

affinity for than soluble monomeric amyloid with a-helical secondary structure. This

may lead to quick off rates, which can be difficult to measure. Supporting the idea that

gelsolin may have fast off rates are measurements for the interaction (Kd) between

amyloid 3 (1-40) and human plasma gelsolin reported by Chauhan et al. to be in the [LM

range (1999). To conclude our data suggest that gelsolin may in fact have an enzymatic-

like property that makes it an even more attractive agent for the treatment of amyloid

related disorders than just an amyloid 3 binding agent, because one molecule of gelsolin

potentially have an effect on a large population of amyloid 3 molecules.







36





Table 3-1 Measurements of Fluorescence and Anisotropy before and after centrifugal
filtration through 30,000 NMW cutoff filters. Anisotropy is reported as +
SEM, values > 0.400 were excluded.
Fluorescence Anisotropv


Ap (82nM) prefilter
Ap flow through
Ap retentate
Ap (82nM) + Gelsolin (418nM) prefilter
Ap + Gelsolin flow through
Ap + Gelsolin retentate


7.30
0.94
1.00
26.50
4.80
16.46


0.060 0.002
0.400 na
0.363 0.025
0.050 0.001
0.047 0.002
0.076 0.001














50000-


40000-


30000-


20X0-


10000-
1O-


P value <


0.0001


I


P value = 0.0007





















lO.pM-Gelilin 10pl -GelIlin


Figure 3-1 Binding of 125I labeled A3 1-42 to Whatman GF/C filters. P values from t-
tests performed show there is consistently lower binding of A3 1-42 (at 2, 5,
and 10 [LM) in the presence of 2[M recombinant gelsolin.


0.175-

0.150-

0.125-

0.100-

S0075-

0.050J

0.025-


nr (


U U U U U


v. ,v I I I I
0 100 200 300 40C
6E10 (nM)
Figure 3-2 Binding of amyloid 3 (1-42) to the antibody 6E10. Blocks represent means +
SEM of triplicate measurements. Line represents best fit with GraphPad Prism
software with a minimum anisotropy of 0.05035, a maximum of 0.1453, a log
EC50 of 140.7 nM, and a hillslope of 0.01037. The fit had an r2 of 0.9938,
and an absolute sum of squares of 2.342x104.


P value = 0.0154


fl ~ """' .


2ZLh-Geldlin 2plGelmlin


51pMGelsmiin


5pM-Gelmlin












250- -0.10
-0.09
200 -0.08
0.07
150C -0.06 .
-0.05 '
100- 0.04
-0.03
50. Fluorescence 0.02
Anisotropy 0.01
0 -0.00
0 100 200 300 400 500
Bovine Plasma Gelsolin (nM)
Figure 3-3 Binding of plasma gelsolin to fluorescently labeled amyloid 3 (1-42)


10.0- 0.4


7.5* -0.3


S5.0- Fluorescence 0.2
SAnisotropy
2.5- -0.1


0.0- 0.0
0 250 500 750 1000 1250 1500 1750 2000
Plasma Gelsolin (nM)
Figure 3-4 Amyloid 3 (1-42) binding to plasma gelsolin after concentration by centrifugal
filtration














Aj (1-42) Fibrillization


:)
C

u,

O1I'
U-


1 2 3 4
Day


5 6 7 8


Figure 3-5 Fibril formation ofHiLyte Fluor 488 Labeled A3 (1-42) fluorescence rapidly
decreases (A), while visible precipitates are formed (B). Photo taken at day 7.












Human

" Bovine
Urea


0 1 2 3 4 5 6
g Gelsolin


7 8 9


Figure 3-6 Dialysis experiment. Amyloid 3 fibril disassembly by gelsolin across a
membrane. Human and bovine refer to the species of gelsolin added. Outside
and inside refers to whether the gelsolin was added outside the dialysis tubing
or inside the dialysis tubing. Fibrillized A3 was added to the inside of dialysis
tubes in all samples. Fluorescence was measured outside the dialysis tubes for
all samples. The urea samples contain no gelsolin, just added as a line to
represent the maximum fluorescence obtainable.








41





RU
45000

40000

35000

| 3BODO0-. Amyloid 1-40

thDC/NHSm
EDC NHS

Baseline= 18,610 Final 19260
20000


15000
0 200 400 600 800 1000 1200 1400 1600 1800 2000
Time
Figure 3-7 Amino-coupling of amyloid 3 1-40 to CM5 chip. A rise in the response is
observed at around 300 seconds when the injection of EDC/NHS reagents
occurs, likewise a drop back to baseline is observed when the injections ends
at 700 seconds this change in response is due to bulk differences in buffer
composition. At 850 seconds the injection of amyloid 3 1-40 begins
corresponding with another increase in response. Upon completion of the
amyloid injection the response does not fall completely back to base line
indicative of covalently attached amyloid. Ethanolamine is then injected
between 1400 to 1800 seconds to block any unreacted carboxymethyl groups.
Another rise in response is observed attributed to bulk changes buffer
composition again. At the completion of attachment there is a rise in baseline
of about 650 RU.










RU
40000


35000 -Ethanolamine




observed at around 300 seconds when the injection of EDC/NS reagents
composition. At 850 seconds the injection of amyloid 1-42 beg1-42
Ba li 18,660t

----- Final= 22.540


0 200 400 600 8O 1000 1200 1400 1600 1800 2000

Figure 3-8 Amino-coupling of amyloid 3 1-42 to CM5 chip. A rise in the response is
observed at around 300 seconds when the injection ofEDC/NHS reagents
occurs, likewise a drop back to baseline is observed when the injections ends
at 700 seconds this change in response is due to bulk differences in buffer
composition. At 850 seconds the injection of amyloid en 1-42 begins
corresponding with another increase in response. Upon completion of the
amyloid injection the response does not fall completely back to base line
indicative of covalently attached amyloid. Ethanolamine is then injected
between 1400 to 1800 seconds to block any unreacted carboxymethyl groups.
Another rise in response is observed attributed to bulk changes buffer
composition again. At the completion of attachment there is a rise in baseline
of about 3880 RU.
















Amyloid j 1-42


Blank


6400 6450 6500 6550 6600 6650 6700 6750 6800
Time s
Antibody binding to amyloid coupled CM5 chip. Response increases of 4000
RU and 5000 RU, for amyloid 3 1-40 and 1-42 respectively, are observed
after a 240 second injection of a solution containing 50 [tg/ml of the antibody
6E10. No response is observed when the solution is injected across the blank
channel of the chip.


RU
27000




25000




23000


21000


19000




17000
6350

Figure 3-9


m











RU
21500-

21000 -

20500-- Amyloid 1 1-42

20000-




19000

L31anL
18500 _____l, B

18000

17500 I I I I I I
100 150 200 250 300 350 400 450
Time s
Figure 3-10 Human plasma gelsolin binding to amyloid coupled CM5 chip. Response
increases of about 750 RU for the amyloid 3 1-40, 1-42, and the blank channel
are observed during a 240 second injection of 50 [tg/mL solution of human
plasma gelsolin. This indicates that the observed response results from bulk
flow changes caused from differences in buffer composition resulting in
changes in the refractive index.











RU
40000


35000


30000 -
S Ethanolamine
2 EDC/NHS
25000

0 Baseline 18,250 Final 19,900
20000 Plasmna Gelsolin


150000 -
S 200 400 600 800 1000 1200 1400 1600 1800 2000
Time
Figure 3-11 Amino-coupling of human plasma gelsolin to CM5 chip. A rise in the
response is observed at around 300 seconds when the injection of EDC/NHS
reagents occurs, likewise a drop back to baseline is observed when the
injections ends at 700 seconds this change in response is due to bulk
differences in buffer composition. At 850 seconds the injection of plasma
gelsolin begins corresponding with another increase in response. Upon
completion of the amyloid injection the response does not fall completely
back to base line indicative of covalently attached amyloid. Ethanolamine is
then injected between 1400 to 1800 seconds to block any unreacted
carboxymethyl groups. Another rise in response is observed attributed to bulk
changes buffer composition again. At the completion of attachment there is a
rise in baseline of about 1650 RU











RU
20000



19950



199001



19850



19800



19750-



19700 ii i
630 640 650 660 670 680
Time
Figure 3-12 Binding of GS-2C4 to human plasma gelsolin coupled CM5 chip.


690
s











RU
29000 -




27000-




25000-


23000 -




21000




19000


Amyloid i 1-40







6EIO



I ) --


5300 5350 5400 5450 5500 5550 5600 5650
Time s
Figure 3-13 Binding of amyloid 13 1-40 to human plasma gelsolin coupled CM5 chip.




RU
29000 -


27000 4


25000 -



23000 -


21000 -


Amyloid 0 1-42









6E10
>--


19000 III I I I
6700 6750 6800 6850 6900 6950 7000
Time s
Figure 3-14 Binding of amyloid 3 1-42 to human plasma gelsolin coupled CM5 chip.














CHAPTER 4
EXPRESSING PLASMA GELSOLIN AND EFFECTS IN TRANSGENIC MICE

Introduction

Gene therapy is a novel approach, which utilizes specific sequences of DNA to

treat, cure, or ultimately prevent disease. There are major hurdles to overcome for it to

be effective; the delivery of such large molecules to the target tissue offers a challenge

because of vulnerability of degradation, due to endogenous nucleases. Another challenge

is having the cells at the target tissue internalize the DNA and transport it into the

nucleus. Finally, having the DNA transcribed and translated to produce a therapeutic

protein that is transported to the proper site of action offers another hurdle to overcome.

Our target is the amyloid 3 that accumulates as senile plaques in the brains of those

suffering from Alzheimer's disease. In the previous chapter we have demonstrated that

human plasma gelsolin holds promise as an agent that can disassemble preformed

amyloid fibrils. Because plasma gelsolin contains an amino-terminal 23 amino acid

signal peptide which signals the cell producing it to secrete it towards the bloodstream,

the site of action for our target can be in the bloodstream rather than the brain,

simplifying delivery. Therefore expressing plasma gelsolin in any peripheral tissue

should be enough to increase plasma gelsolin in the bloodstream.

The hydrodynamic gene delivery method offers an efficient technique for testing if

peripheral expression of plasma gelsolin can have an effect on amyloid distribution and

clearance in transgenic mice. The hydrodynamic gene delivery method was first

developed by Dr. Dexi Liu (1999). This method involves injecting a large volume of a









DNA solution in a short period of time, via the tail vein of a mouse, and results in a high

level oftransgene expression in the liver. The mechanism for how this works seems to

be that the initial rapid increase in blood volume causes an increase in venous pressure,

which forces an enlargement of the liver fenestrae, and causes the formation of transient

pores on the membranes of hepatocytes allowing the plasmid DNA to enter the cells

(Zhang et al., 2004).

There is typically a high level of transgene expression following hydrodynamic

gene delivery, followed by a quick drop off to a lower stable level of expression (Liu et

al., 1999). Alino, Crespo, and Dasi showed that when the full length hAAT promoter

was used to drive expression, after hydrodynamic gene delivery a stable therapeutic level

of human alpha-1-antitrypsin (hAAT) is detected in the blood for up to 120 days, (2003).

The majority of plasmids delivered during hydrodynamic gene delivery studies are

driven by human cytomegalovirus immediate-early promoter (CMV). Song et al showed

that the CMV-chicken beta actin hybrid promoter (CBA) had well over 100 times the

activity in the mouse liver than the CMV promoter (2001). We chose to incorporate the

CBA promoter in our expression plasmid (Figure 2-1). We also included the woodchuck

hepatitis virus post-transcriptional regulatory element (WPRE) in our plasmid. The

WPRE functions to stabilize mRNA having the effect of increasing the half-life for

mRNA and ultimately increasing the amount of gene product produced. We have

previously described that incorporating the WPRE into vectors increased green

fluorescent protein (GFP) and nerve growth factor (NGF) expression by more than ten

fold in rats (Klein et al., 2002).









Results and Discussion

Conformation of Vector Product and Activity

To test that the vector we constructed pUFGL (Figure 2-1C) does in fact produce

plasma gelsolin, we transfected 293 cells using the Ca3(PO4)2 precipitation method,

collected and concentrated media, and then performed a western blot to confirm size and

identity of plasma gelsolin secreted into the media (Figure 4-1). There was a protein that

ran at 91kD that was immunoreactive with the anti-human gelsolin antibody, GS-2C4

(Sigma, St. Louis, MO), which was absent in the samples that had been transfected with

the control plasmid pGFP.

To confirm that our gene product retained its amyloid binding activity we spiked

some of the unconcentrated media with A3 (1-42). Performed an immunoprecipitation

using the GS-2C4 anti-human gelsolin antibody and then ran a Western blot with the

precipitates. We found that A3 1-42 coprecipitated (Figure 4-2) with gelsolin in the

pUFGL transfected, A3 1-42-spiked media and not with the pGFP transfected. A3 1-42-

spiked media. Amyloid 3 1-42 was not found in the precipitates when the media was not

spiked with AP 1-42.

Hydrodynamic Gene Delivery in Mice

To determine the distribution and level of gene expression we should expect from

using the hydrodynamic gene delivery technique mice were injected either pGFP or

GWIZ luciferase expression plasmid via the hydrodynamic technique.

One day following injections with the GWIZ vector the two animals injected were

imaged by the University of Florida's biomedical engineering department (Figure 4-3).

Enough visible light was produced by the gene product luciferase that the liver was fully









illuminated for at least 30 minutes after an intraperitoneal injection of luciferin

luciferase'ss natural substrate).

Two and a half weeks after injecting mice with pGFP plasmid the animals were

sacrificed and their livers were excised. Sections of the liver were made, 30tlm thick,

and the distribution of fluorescence was examined by fluorescence microscopy

(Figure 4-4). At the two and a half week post-injection timepoint GFP fluorescence

distribution was detected widely throughout the liver.

Transgenic mice expressing mutant presenilin-1M146L (mutant PS-1)(Duff et al.,

1996) were injected via the hydrodynamic gene delivery method either pUFGL or pGFP.

Plasma samples were taken at 24, 48 and 96 hours. These samples were used for a

western blot (Figure 4-5). Plasma samples from pUFGL injected animals had clear bands

corresponding to human plasma gelsolin that were not present in the pGFP injected

animal.

Hematoxylin and Eosin Staining in Mice after Hydrodynamic Gene Transfer

Hematoxylin and eosin (H & E) staining is a routine stain that takes advantage of

two separate dyes. Hematoxylin stains nuclear material a purplish color, while eosin

stains membranes and connective tissue an orange-pinkish color. While unable to

identify specific chemical markers for inflammation, H & E staining is useful at

identifying, abnormal growth, division in the nucleus, or cellular death in tissues that may

be related to disease or injury.

To determine if damage or inflammation affected the liver tissue 18 days after gene

delivery via the hydrodynamic injections, livers were examined using H & E staining.

Eighteen days after gene delivery of pGFP or pUFGL in double transgenic mice









expressing both human mutant APP695K670N,M671L (Tg2576)(Hsiao et al., 1996) and

mutant presenilin-1M146L(Duff et al., 1996) (huAPP/PSlM146L mice). H & E staining

revealed no differences among livers from mice injected with pGFP, pUFGL or untreated

mice (Figure 4-6).

Conclusions

In this chapter we demonstrate that our vector, pUFGL, does in fact produce human

plasma gelsolin immunoreactivity, verified by western blot. We also show that our gene

product maintains its A3 binding activity verified by the co-immunoprecipitation of A3

(1-42) with plasma gelsolin. We also show that by delivering vectors by the

hydrodynamic technique we are able to obtain high levels of gene expression for at least

two and a half weeks, and this expression does not result in damage or inflammation to

the liver, detectable by H & E staining. Finally we are able to find detectable levels of

human plasma gelsolin in the plasma of mice up to 96 hours after gene delivery by the

hydrodynamic technique.











A B C D E


S! Plasma Gelsolin

Figure 4-1 Immunoblot oftransfected media. 293 cells at 50% confluency (lane B and
D) or 80% confluency (lane C and lane E) were transfected using the
Ca3(P04)2 precipitation method with either pGFP (lane B and lane C) or
pUFGL (lane D and lane E). 48 hours after transfection, media was collected,
concentrated, and then separated on a 7.5% PAGE and immunoblotted with a
monoclonal anti-gelsolin antibody. Lane A contains 1 |tg of human plasma
gelsolin.




Amyloid 3 1-42 + +
pGFP + + -

pUFGL -+ +


Amyloid 3 (1-42) 4.5kD -


Figure 4-2 Immunoprecipitation. Amyloid 3 (1-42) co-immunoprecipitates with 293
expressed plasma gelsolin. The first lane contains immunoprecipitates from
pGFP transfected media with A3 1-42. The second lane contains
immunoprecipitates from pGFP transfected media with no A3. The third lane
contains immunoprecipitates from pUFGL transfected media with A3 1-42.
The fourth lane contains immunoprecipitates from pUFGL transfected media
with no Ap.































Figure 4-3 Bioluminescence resulting from hydrodynamic gene transfer. Imaging was
done by the University of Florida's Department of Biomedical Engineering, as
collaboration with fellow pharmaceutics graduate student Natalie Toussaint.
The white rectangular shows the region imaged with a thermoelectrically
cooled (-700C), back illuminated CCD array (Roper Scientific
Instrumentation, Trenton, NJ) coupled with an optical lens subsystem (Zoom
7000, Navitar, Rochester, NY).




























Figure 4-4 Green fluorescent protein. Expression observed in the liver of a mouse two
and a half weeks after pGFP delivery via the hydrodynamic gene delivery
technique. GFP expressing hepatocytes, green cells, are widely distributed
throughout the liver. The red color resulted from background stained with
Alexa Flour 488 goat anti-mouse IgG (H+L). Dark empty spots are hepatic
sinusoids.



A B C D E F G

Saf mmm map*

H I J K L M N O




Figure 4-5 Western blot of plasma samples taken from PS-1 mice injected with pUFGL at
24 hours (Lanes D and E), 48 hours (Lane F), and 96 hours post injection
(Lane G); or plasma taken from pGFP injected mice at 24 hours (Lane K), 48
hours (Lane L), and 96 hours post injection (Lane M). Lane N is plasma from
a non injected mouse and Lane O is plasma from a non injected mouse spiked
with 500ng of human plasma gelsolin. Lanes A and H contain 1500ng of
human plasma gelsolin, Lanes B and I contain 1000ng of human plasma
gelsolin, and Lanes C and J contain 500ng of human plasma gelsolin.
































Figure 4-6 Hematoxylin and eosin staining of livers 18 days after hydrodynamic gene
delivery. Upper image represents an untreated liver section, and lower image
represents a pUFGL injected liver section. Scale bar represents 50 jtm for
both images.














CHAPTER 5
EFFECT OF GELSOLIN EXPRESSION ON AMYLOID DEPOSITION

Introduction

Since the discovery that senile deposits in Alzheimer's disease are composed

mainly of the fibrillar amyloid 3 peptide (Glenner and Wong, 1984), researchers have

discovered a number of mutations on either amyloid 3's parent protein (APP), or proteins

that process APP (PS1), that lead to inheritable forms of the disease. These discoveries

have been quite useful in the development of transgenic mouse models of Alzheimer's

disease pathology.

Achieving elevated levels of transgene expression was a critical step in the

development of transgenic mice to model neurodegenerative diseases. It was recognized

in the early 1990s that the gene encoding for mammalian prion protein (PrP) would make

an effective expression package to produce foreign proteins in the central and peripheral

nervous systems of mice (Hsiao et al., 1995; Scott et al., 1992; Telling et al., 1994). A

42 kb cosmid clone of the Syrian hamster PrP gene was made and it was noted that the

entire open reading frame is contained in a single exon. (Basler et al., 1986). This exon

can be excised and exchanged with the cDNA of a gene of interest, which can then be

used for a pronuclear injection into mouse embryos to generate a transgenic line of mice

expressing the protein of interest at high levels in the nervous system and heart of the

mice (Borchelt et al., 1996).

In 1996, using this technique, Hsiao et al. developed a transgenic mouse model

expressing the Swedish double mutant APPK670N,M671L (mutant APP, Tg2576) with a









C57/BI6 and SJL mouse background. These mice expressed the mutant APP about

5.5 times that of the endogenous murine APP. After 11 months of age amyloid plaque-

like deposits are found throughout the cortex, hippocampus, presubiculum, subiculum,

and the cerebellum (1996). These mice also demonstrated a deficit in memory-related

behavior that correlated to the levels of insoluble amyloid 3 in the brain (Westerman et

al., 2002).

Duff and coworkers developed mice that express a mutant form of presenilin-1M146L

(mutant PS-1). These mice had no detectable histopathology of Alzheimer's disease,

however they did have elevated levels of amyloid 3 (1-42) (1996). When the mutant PS-

1 mice are bred with the Tg2576 mice the resultant double transgenic progeny (APP/PS1)

have an accelerated rate of amyloid 3 deposition, about 3-5 times that of the singly

transgenic Tg2576, with a age of onset of between 3 and 6 months of age (Holcomb et

al., 1998; Holcomb et al., 1999).

Using a similar strategy Borchelt et al. developed a transgenic model expressing a

humanized version of murine APP695. This humanized gene was controlled by the

mouse PrP promoter that drove expression of cDNA containing all murine sequence

except for the amyloid 3 domain and the mutations (K595N, M596L) that are linked to

the human Swedish form of familiar Alzheimer's disease. The level of transgene

expression of these mice was about 2-3 times that over the endogenous APP expression

in nontransgenic littermates (1996). These mice develop amyloid deposits around 18

months of age (1997).

Jankowsky and coworkers developed a line of mice that express the mouse human

chimeric Swedish APP695 that Borchelt created along with a form of human presenilin-1









with exon nine deletion (PS16E9). They showed that these mice produce about 2.5 times

the level of amyloid 3 (1-42) while amyloid 3 (1-40) levels remain constant (2004). The

elevated levels of amyloid 3 (1-42) result in deposits occurring at a much accelerated rate

as compared to the singly transgenic mice. These mice begin to develop deposits by the

age of 6 months as compared to 18 months in the mice which don't co express PS16E9.

APP/PS1 mice have been useful as models to study treatments that target

amyloid P. Morgan et al. successfully vaccinated APP/PS1 mice against A3, which had a

dramatic effect on amyloid 3 deposition that protected against memory and learning

deficits (2000). Deane et al. also saw dramatic effects on amyloid deposition by

administering a soluble form of the receptor for advanced glyclation end products

(RAGE)(2003). Matsuoka et al. treated young APP/PS1 mice with the amyloid binding

agents GM1 and plasma gelsolin, and saw significant reductions in amyloid 3 levels in

the brain (2003). This is why we believe APP/PS1 mice will make a good model to

determine if peripheral expression of plasma gelsolin can effect amyloid 3 deposition.

Results

Mice Expressing Swedish Mutant Amyloid Precusor Protein (Mouse/Human
Hybrid) and Exon 9 Deleted Mutant Presenilin-1

Message Detection

Nine month old double transgenic mice expressing both Swedish mutant mouse

/human hybrid APP695 K594N,M595L and mutant presenilin-16E9 (mo/huAPP /PSl1E9)

(Jankowsky et al., 2004) were injected with pUFGL, via the hydrodynamic gene delivery

method. Two and one half weeks following injections the animals were sacrificed, along

with three untreated littermates. RNA was purified from the liver and brain tissue, as

described in the methods section. RT-PCR was performed using vector specific primers









designed to yield a 900 bp product from processed mRNA coded from pUFGL or an

1800bp fragment from unprocessed RNA or DNA.

All samples from animals that received an injection of pUFGL showed vector

specific mRNA hybridization. In samples from animals that did not receive injections

vector specific mRNA hybridization was not detected (Figure 5-1). All of the samples

did show mRNA hybridization when reactions were run with primers specific for P-actin

(data not shown) indicating that mRNA is present in all of the samples. Both samples

from RNA purified from 293 cells transfected with pUFGL had positive bands. RNA

samples purified from brain tissue did not show vector specific mRNA hybridization

(data not shown) suggesting if there is vector gene expression in the brain it is below

detectable quantities.

Total Brain Amyloid 0 (1-42) Concentrations

Enzyme linked immunosorbent assays (ELISA) were performed in order to

measure the concentration of both soluble and insoluble fractions of amyloid P (1-42) in

the hemi-brains from both pUFGL-injected and noninjected mice. All hemibrains were

subjected to a guanidine extraction in order to obtain the total amount of A3 contained

within the brain tissue (Johnson-Wood et al., 1997; Masliah et al., 2001). The samples

were run in duplicate on two separate occasions for a total of four samples assayed per

animal. Untreated controls had a mean SEM amyloid 0 (1-42) concentration of

2,306 202.6 picomoles per gram of brain tissue and the injected animals had

1,174 334.7 picomoles per gram of brain tissue. A one- tailed t-test was performed and

showed that these groups differ significantly with a P value = 0.0222 (Figure 5-2).









Plasma Amyloid 0 Concentrations

Blood was collected via the retro-orbital sinus using heparinized capillary tubes just

prior to sacrifice. Samples were immediately centrifuged, followed by plasma collection

which was flash frozen in liquid nitrogen. Samples were then thawed at a later time point

and analyzed in duplicate by ELISA to determine amyloid 3 1-40 and 1-42

concentrations. Untreated mice had a mean SEM plasma AP40 concentration of

315 116.3 fmol/mL, while pUFGL-injected mice had 200.3 86.5 fmol/mL

(Figure 5-3A). Untreated mice had a mean SEM plasma AP42 concentration of

95.6 23.2 fmol/mL, while pUFGL-injected mice had 150.9 53.97 fmol/mL

(Figure 5-3B). Trends but not significant changes in either A342 or A340 concentrations

were observed; however a significant decrease in the ratio of A340/ A342 was observed

in the pUFGL-injected mice (Figure 5-3C). Untreated mice had a mean A340/ A342

ratio of 3.0 + 0.65, while pUFGL-injected mice had a mean A340/ A342 ratio of

1.24 0.49 (P = 0.045).

Dense Cored Amyloid Deposits

Thioflavin S staining was used to examine the extent of dense cored amyloid

Deposit pathology in injected and untreated mice (Sun et al., 2002). 50[tm thick coronal

sections were made. Four sections 0.3 mm apart were stained with thioflavin S and

digital micrographs were made of epifluorescence images (Figure 5-6). Images were

analyzed in a blinded manner using NIH Image J software. The area of the hippocampus

and cortex, total stained area, area of each individual stained deposit, and the number of

stained deposits was determined by thresholding segmentation. The amyloid burden was









determined by dividing the total area stained by the total area of the hippocampus and

cortex.

Untreated animals had mean amyloid burden of 1.03 0.13%, while pUFGL-

injected animals had a mean amyloid burden of 0.39 0.09%. A one-tailed t-test showed

that these groups differ significantly with a P value = 0.0085 (Figure 5-4A). Untreated

mice had a mean total stained area of 82,660 10,150 pixels2, while the pUFGL-injected

mice had a mean total stained area of 33,740 + 8429 pixels2. These measurements were

also determined to be significantly different by an unpaired one tailed t-test, P value =

0.0103 (Figure 5-4B). The average deposit size for untreated mice was determined to be

42.2 4.8 pixels2, while the average size for pUFGL was 34.2 + 1.2 pixels2. These were

not found to be statistically different (P value = 0.0905, by an unpaired one-tailed t-test)

(Figure 5-4C). Untreated mice had an average of 503 84 deposits per section while

pUFGL-injected mice averaged 244 51 deposits per section. These were found to

differ statistically (P value = 0.029, unpaired one-tailed t-test) (Figure 5-4D).

Diffuse Amyloid Deposits

Diffuse amyloid deposits were visualized by immunohistochemistry using the

antibody 6E10, which recognizes the first 17 amino acids of human amyloid P. 50|tm

thick coronal sections were made. Three sections 0.3 mm apart were stained. Ditgital

micrographs were made using light microscopy (Figure 5-7). Stained sections were

analyzed in a blinded manner using NIH Image J software. The area of the hippocampus

and cortex, total stained area, area of each individual stained deposit, and the number of

stained deposits was determined by thresholding segmentation. The amyloid burden was









determined by dividing the total area stained by the total area of the hippocampus and

cortex.

Untreated animals had a mean amyloid burden of 1.48 0.19%, while pUFGL-

injected animals had a mean amyloid burden of 1.12 0.05%. An unpaired one-tailed t-

test showed that these groups were not statistically different, however a strong trend

towards significance was present (P value = 0.074) (Figure 5-5A). Untreated mice had a

mean total stained area of 67,700 10,590 pixels2, while the pUFGL-injected mice had a

mean total stained area of 48,570 2643 pixels2. These measurements were also not

significantly different by an unpaired one tailed t-test, but again there was a trend

(P value = 0.078) (Figure 5-5B).

The average deposit size for untreated mice was determined to be

16.9 2.0 pixels2, while the average size for pUFGL was 16.2 + 1.2 pixels2. These were

not found to be statistically different (P value = 0.387, by an unpaired one-tailed t-test)

(Figure 5-5C). Untreated mice had an average of 1326 53 deposits per section while

pUFGL-injected mice averaged 997 + 123 deposits per section. These were found to

differ statistically (P value = 0.035, by an unpaired one-tailed t-test) (Figure 5-5D).

Soluble Amyloid Oligomers

The distribution and relative quantities of soluble amyloid oligomers were

examined by immunostaining with the Al 1 antibody. The Al 1 antibody recognizes an

epitope that is common to soluble amyloid oligomers, but is not found in amyloidogenic

monomers or mature amyloid fibrils(Kayed et al., 2003). Three sections from each

animal were stained, all sections were incubated for equal amounts of time during the

labeling procedure, and DAB reactions were carried out on slides containing one section

from each animal, in order to minimize differences in reaction times or reagents. Digital









micrographs were prepared of the whole slides under the same light source using an

Olympus BH-2 light microscope equipped with a motorized stage and focus control

system (Prior Scientific). Image Pro plus version 4.0 software was used for tiling images

together using a 4x nosepiece objective, a 1.25x internal magnification changer, and a

2.5x camera tube objective. To determine staining intensity and the relative amount of

antigen present grey values were measured from negative images of each section using

NIH Image J software.

Soluble amyloid oligomers appear to be widely distributed among both the

pUFGL-injected mice and the untreated mice. Untreated mice had significantly lower

negative grey values (127.6 4.6) compared to pUFGL-injected mice (142 1.8, P =

0.0220), which may indicate an increase in the concentration of soluble amyloid

oligomers in pUFGL-injected mice (Figure 5-9). However, variability in fixation and

section can also result in holes in the section that may correspond to less staining of the

section. At high magnification the distribution of amyloid oligomer staining in both

untreated mice and pUFGL-injected mice is mainly confined to the neuronal soma,

axons, and dendritic processes (Figure 5-10). However in pUFGL-injected animals a

number of glia cells and cells associated with the vasculature show an apparent

immunoreactivity to the All antibody also (Figure 5-11).

Microglia

Microglia are a population of dendritic cells in the brain, thought to be of the same

origin as monocytes and macrophages. Microglia can reside in the brain in a number of

different states. Resting microglia ramifiedd) have long finely branched processes

extending from all directions from the perinuclear cytoplasm (Giulian and Baker, 1986).

Resting microglia, in response to a number of different insults to the nervous system,









begin to proliferate while still in non-phagocytic state, and are referred to as activated

microglia. Activated microglia are recognizable by retracted cellular processes and mild

hypertrophy resulting in a stout configuration. Further transformed microglia, with their

processes even more retracted, are known as reactive or phagocytic microglia and are

considered to be like "brain macrophages"(Streit et al., 1988). Reactive microglia have a

round shape and can appear ruffled, due to short cytoplasmic projections, and are capable

of releasing growth factors, cytokines, and free radicals.

The state of microglia cells in mo/huAPP/PS1lE9 brain sections was examined by

immunostaining using the OX42 antibody. OX42 recognizes complement receptor 3

(CR3) which is expressed by microglia during all stages. Three sections per animal were

examined and all sections were incubated for equal amounts of time during the labeling

procedure. DAB reactions were carried out on slides containing one section from each

animal, in order to minimize differences in reaction times. Calibrated digital micrographs

were prepared of the whole slides with a constant light source, and relative grey values

were measured from negative images for each section using NIH Image J software, to

determine staining intensity.

Untreated mice had significantly lower relative negative grey values (130.7 10.3)

compared to pUFGL-injected mice (163.1 + 11.1, P = 0.0496), indicating higher

concentration of CR3 receptor in the pUFGL-injected animals (Figure 5-12 & 5-13). At

higher magnification it is qualitatively clear that the increase in staining intensity

observed in pUFGL treated mice is due to an increase in the number of activated and

reactive microglia, distinguished by their retracted processes and condensed cell bodies

(Figure 5-14).









Astrocytes

Astrocytes are a type of glial cell recognizable by numerous arms which give them

a star shaped appearance. They play a number of roles in the brain including structuring

the brain, providing neurons with nutrients, contributing to the blood brain barrier,

altering cerebral blood flow, clearing neurotransmitters and regulating ion concentrations

in the extracellular space (Pellerin, 2005; Sofroniew, 2005; Volterra and Meldolesi,

2005). Astrocytes become activated in response to disease or injury. One of the

pathological features in Alzheimer's disease is the presence of activated astrocytes in and

around amyloid deposits (McGeer and McGeer, 2003).

Astrocytic activation was examined in mo/huAPP/PS 16E9 brain sections by

immunostaining for glial fibrillary acidic protein (GFAP). Two 50mm thick sections per

animal were analyzed for total percent stained area of the hippocampus and cortex using

NIH Image J software. Untreated mice had a mean SEM GFAP positive percent area

of 11.06 1.59 %, while pUFGL-injected mice had a mean SEM GFAP positive

percent area of 14.15 2.42 % (Figure 5-15). These values do not differ significantly

however upon examination at a high magnification GFAP positive astrocytes in pUFGL-

injected mice appear to have thicker and more numerous processes around areas

corresponding to Congo Red positive amyloid deposits (Figure 5-16).

Mice Expressing Swedish Mutant Amyloid Precursor Protein (Human) and M146L
Mutant Presenilin-1

Message Detection

Eight-month old double transgenic mice expressing both Swedish mutant human

APP695 (huAPP K670N,M671L, Tg2576) (Hsiao et al., 1996) and mutant presenilin-1 M146L

(PS-1M146L) (Duff et al., 1996) were injected via the hydrodynamic gene delivery method









either pUFGL (5) or pGFP (3). Two and one half weeks following injections animals

were sacrificed along with age-matched untreated mice (7). RNA was purified from the

liver and brain tissue, as described in the methods section. RT-PCR was performed with

primers designed to yield a 900 bp product from processed mRNA coded from our vector

pUFGL. All samples from animals that received an injection of pUFGL had positive

bands. Samples from animals that did not receive injections or received injections of

pGFP did not have bands (Figure 5-16, not a complete data set, a representative gel). All

samples did produce bands when reactions were run with primers for 3-actin (data not

shown) indicating that mRNA is present in the all of the samples. Sample from RNA

purified from 293 cells transfected pUFGL had a positive band. None of samples

purified from brain tissue had positive bands (data not shown) suggesting there is no

vector gene expression in the brain.

Amyloid 0 (1-42) Quantification by ELISA

ELISA was used to quantify the concentration of guanidine extractable amyloid

P (1-42) in brain homogenates of the huAPP/PSlM146L mice. The resultant mean A342

concentrations were 9179 916 picomoles per gram of brain tissue for the untreated

group, 10,928 731 picomoles per gram of brain tissue for the pGFP-injected group, and

6740 + 998 picomoles per gram of brain tissue for the pUFGL group. A one-way

ANOVA showed that these means do not differ significantly, P = 0.069 (Figure 5-18A),

however a trend toward significance was present.

In order to account for inter-litter variability, amyloid 1 1-42 concentrations from

pGFP and pUFGL-injected mice were divided by the amyloid 1 1-42 concentrations

obtained from their respective untreated littermates. These values were reported as the









fraction of amyloid 3 1-42 of untreated littermates. The pGFP group had a mean fraction

of 1.00 + 0.06, while the pUFGL has a mean fraction of 0.797 0.05. A unpaired one

tailed t-test showed that these groups differ significantly with a P value = 0.022 (Figure

5-18B).

Plasma Amyloid 0 Concentrations

Blood was collected either via the retro-orbital sinus using heparinized capillary

tubes or by intra cardiac puncture with an EDTA treated syringe just prior to sacrifice.

Samples were immediately centrifuged; plasma collected and flash frozen in liquid

nitrogen. Samples were then thawed and analyzed by ELISA in duplicate to determine

amyloid P 1-40 and 1-42 concentrations. Untreated mice had a mean SEM plasma

Ap40 concentration of 1039 244.6 fmol/mL, pGFP-injected mice had

971.7 209 fmol/mL, and pUFGL-injected mice had 930 252 fmol/mL (Figure 5-19A).

Untreated mice had a mean SEM plasma AP42 concentration of 484.9 143 fmol/mL,

pGFP-injected mice had 453.5 45.7 fmol/mL, and pUFGL-injected mice had

467 43.3 fmol/mL (Figure 5-19B). There were no significant changes in either A342

or Ap40 concentration. There were also no significant changes measured in the Ap40/

AP42 ratio; untreated mice had a mean concentration of 2.29 0.22, pGFP-injected mice

had a mean ratio of 2.1 0.26, and pUFGL-injected mice had a mean ratio of 1.9 0.46

(Figure 5-19C).

In light of the observation that amyloid P binds to heparins (Brunden et al., 1993;

Leveugle et al., 1994), it is possible that the recovery of amyloid P may not have been the

same between the two methods used for sample collection,. The data were examined in

terms of collection method and differences were observed. Plasma collected retro-









orbitally had a mean amyloid 3 1-40 concentration of 560.3 164 fmol/mL and a mean

amyloid 3 1-42 concentration of 344.8 59.1 fmol/mL; when collected via intra-cardiac

puncture the mean amyloid 3 1-40 concentration was 1185 128 and a mean

amyloid 3 1-42 concentration of 531 53.0 fmol/mL (P = 0.016 and 0.057 respectively,

by unpaired two tailed t-test, Figure 5-19D &E). We concluded concentrations obtained

from plasma samples collected with heparinized tubes are probably not accurate and

cation should be used when analyzing such data. When two different collection methods

are used, the samples are definitely not comparable.

Dense Cored Amyloid Deposits

Thioflavin S staining was used to examine the extent of dense cored amyloid

deposit pathology in injected (pGFP and pUFGL) and untreated mice (Sun et al., 2002).

50[tm thick coronal sections were made. Four sections 0.3 mm apart were stained with

thioflavin S and digital micrographs were made (Figure 5-20) as described previously.

Images were analyzed in a blinded manner using NIH Image J software. The area of the

hippocampus and cortex, total stained area, area of each individual stained deposit, and

the number of stained deposits was determined by thresholding segmentation. The

amyloid burden was determined by dividing the total area stained by the total area of the

hippocampus and cortex.

Untreated animals had a mean amyloid burden of 1.48 0.16%, pGFP-injected

animals had a mean amyloid burden of 1.59 0.10%, pUFGL-injected animals had a

mean amyloid burden of 1.09 0.14%. A one-way ANOVA was performed, and the

medians did not differ significantly with a P-value = 0.069 (Figure 5-21A).









Untreated mice had a mean total stained area of 115,329 18,081 pixels2, pGFP-

injected mice had a mean total stained area of 140,952 14,898 pixels2, while the

pUFGL-injected mice had a mean total stained area of 87,503 14,622 pixels2. These

measurements were also found not to differ statistically (Figure 5-21B, P-value = 0.127,

by one-way ANOVA).

There was also no difference observed in deposit size among the groups. Untreated

mice had a mean deposit size of 75.62 + 15.4 pixels2, pGFP-injected mice had a mean

Deposit size of 56.03 + 5.6 pixels2, and pUFGL-injected mice had a mean deposit size of

44.78 6.3 pixels2 (Figure 5-21C, P-value = 0.247, by one-way ANOVA).

Deposit numbers per section did not change between the treatment groups either.

Untreated mice had an average of 458 107 deposits per section; pGFP-injected mice

had an average of 632 53 deposits per section, and pUFGL-injected mice had an

average of 513 32 deposits per section.

Again to account for inter-litter variability, dense-cored amyloid burdens, average

deposit sizes, and average number of deposits per section from pGFP and pUFGL-

injected mice were divided by the values obtained from their respective untreated

littermates. No differences were detected in dense-cored amyloid burden fractions or

dense-cored deposit size fractions; pGFP-injected mice had a mean dense-cored amyloid

burden fraction of 0.88 0.05 and a mean dense-cored deposit size fraction of

1.06 0.11, pUFGL-injected mice had a mean dense- cored amyloid burden fraction of

0.89 0.16 and a mean dense-cored deposit size fraction of 0.86 0.28

(Figure 5-22A &B).









An increase in the average number of deposits per section fraction was detected in

pUFGL-injected mice; pGFP-injected mice had a mean average dense-cored deposit

number per section fraction of 1.06 0.11, pUFGL-injected mice had an average dense-

cored deposit number per section fraction of 1.38 0.13; P = 0.048 by unpaired two-

tailed t-test (Figure 5-22C).

Diffuse Amyloid Deposits

Diffuse amyloid deposits were visualized by immunohistochemistry using the

antibody 6E10, which recognizes the first 17 amino acids of human amyloid P. 50[tm

thick coronal sections were made. Three sections 0.3 mm apart were stained. Digital

micrographs were made (Figure 5-23). Stained sections were analyzed in a blinded

manner using NIH Image J software. The area of the hippocampus and cortex, total

stained area, area of each individual stained deposit, and the number of stained

deposits was determined. The amyloid burden was determined by dividing the total area

stained by the total area of the hippocampus and cortex.

Untreated animals had mean amyloid burden of 4.25 0.46%, pGFP-injected mice

had a mean diffuse amyloid burden of 4.04 0.46%, and pUFGL-injected animals had a

mean amyloid burden of 2.83 0.05%. A one-way ANOVA showed that these groups

were not statistically different, (P value = 0.099) (Figure 5-24A).

The average diffuse deposit size for untreated mice was determined to be 20.3 + 1.5

pixels2, the average size for pGFP-injected mice was 26.3 2.3 pixels2, and the average

size for pUFGL-injected mice was 20.9 + 3.2 pixels2. These were not found to be

statistically different (P value = 0.271, by one-way ANOVA) (Figure 5-24B).









The average number of diffuse deposits per section for untreated mice was

determined to be 3990 + 481 deposits/section, 3437 + 513 deposits/section for pGFP-

injected mice, and 3623 729 deposits/section for pUFGL-injected mice. These were

not found to be statistically different (P value = 0.812, by one-way ANOVA)

(Figure 5-23C).

Again to account for inter-litter variability, diffuse amyloid burdens, average

diffuse deposit sizes, and average number of diffuse deposits per section from pGFP and

pUFGL-injected mice were divided by the values obtained from their respective

untreated littermates (Figure 5-25). No differences were detected; however a trend

toward a reduction in diffuse amyloid burden fractions was seen, pGFP-injected mice had

a mean diffuse amyloid burden fraction of 1.00 + 0.13, and pUFGL-injected mice had a

mean diffuse amyloid burden fraction of 0.72 0.10 (P = 0.071, by unpaired one-tailed t-

test). No trends were detected in diffuse deposit size fraction (pGFP-injected mice had a

mean diffuse deposit size fraction of 1.06 0.06, pUFGL-injected mice had a mean

diffuse deposit size fraction of 1.06 + 0.16) and average number of diffuse deposits per

section fraction (pGFP-injected mice had an average number of diffuse deposits per

section fraction of 1.13 0.18, pUFGL-injected mice had had an average number of

diffuse deposits per section fraction of 0.86 + 0.11).

Microglia, Soluble Oligomers, and Astrocytes

The state of microglia and the degree of soluble amyloid oligomers in

huAPP/PSlM146L brain sections were examined by immunohistochemistry (Figure 5-26).

Three sections per animal were examined and all sections were incubated for equal

amounts of time during the labeling procedure. DAB reactions were carried out on slides

containing one section from each animal, in order to minimize differences in reaction









times. Digital micrographs were prepared of the whole slides with a constant light

source, and relative grey values were measured from negative images for each section

using NIH Image J software, to determine staining intensity. Astrocytic activation was

examined in huAPP/PS lM146L brain sections by immunostaining for glial fibrillary acidic

protein (GFAP). Two 50mm thick sections per animal were analyzed for total percent

stained area of the hippocampus and cortex using NIH Image J software.

For microglia, untreated mice had a mean SEM relative negative grey value of

27.5 4.8, pGFP-injected mice had 29.2 3.3, and pUFGL-injected mice had 46.53

16.0. These values did not differ significantly (Figure 5-27A). When examined in terms

of fractions of their untreated littermates, pGFP-injected mice had a mean microglia stain

fraction of 1.32 0.05 and pUFGL-injected mice had a mean microglia stain fraction of

1.84 0.60. These values values were also found not to be different statistically (Figure

5-27B)

When soluble amyloid oligomers were examined, untreated mice had a mean +

SEM relative negative grey value of 85.8 4.7, pGFP-injected mice had 85.9 1.3, and

pUFGL-injected mice had 87.1 8.5. These values did not differ significantly

(Figure 5-27C). When examined in terms of fractions of their untreated littermates,

pGFP-injected mice had a mean soluble amyloid oligomer stain fraction of 0.93 0.01

and pUFGL-injected mice had a mean soluble amyloid oligomer stain fraction of 1.23 +

0.18. These values were also found not to be different statistically although a trend was

evident (Figure 5-27D).

Staining for astrocytes using GFAP revealed that untreated mice had 16.7 1.4 %

of their hippocampus and neocortex stained, pGFP-injected mice had 13.5 0.89 %









stained, and pUFGL had 18.2 2.8 % stained. These values did not differ statistically

(Figure 27E). When examined in terms of fractions of their untreated littermates, pGFP-

injected mice had a mean soluble amyloid oligomer stain fraction of 0.96 0.04 and

pUFGL-injected mice had a mean soluble amyloid oligomer stain fraction of 1.26 0.16.

These values were also found not to be different statistically although a trend was evident

(Figure 5-27F).

Regression Analysis

Linear regression was employed to examine possible correlations in pUFGL-

injected mice between the fractions of untreated littermate soluble amyloid oligomer

staining, microglia staining, or % GFAP staining, and fraction of untreated littermate

AP42 concentrations, diffuse or dense-core amyloid burdens. Slopes trended toward

being significantly different from zero when soluble oligomers fractions (P = 0.079) and

microglia fractions (P = 0.062) were compared with values for fractions of A3 42

concentrations (Figure 5-28A & B). Slopes also showed a strong trend toward being

significantly different from zero when percent GFAP stain fractions were compared with

dense-cored amyloid burden fractions (P = 0.084, Figure 28C).

When percent GFAP stain fractions were compared with fractions of diffuse

amyloid burden, slope was significantly different than zero (P = 0.008, Figure 28D),

indicating a strong correlation for an increase in the % area stained for GFAP compared

to untreated littermates, with a decrease in diffuse amyloid burden compared to untreated

littermates in pUFGL-injected mice.









Conclusion

In this study we examined the effects of peripheral plasma gelsolin expression on

two different transgenic mouse models of Alzheimer's disease. Two different models

were used due to a failure in our huAPP/PSlM146L breeding colony. However, testing two

separate models has advantages. Two separate models lets us examine gelsolin

expression in animals that produce amyloid 3 at different rates (huAPP/PSlM146L, faster;

mo/huAPP/PSl6E9, slower). They also allow us to test effects of gelsolin expression with

mice of different genetic backgrounds. Studying the differences in response to gelsolin

expression between the different models can allow us to make inferences on the

efficiency and mechanisms of actions of plasma gelsolin expression may have.

These results demonstrate that both mo/huAPP/PS16E9 mice and huAPP/PSlM146L mice

express pUFGL in the liver for at least two and a half weeks after gene delivery by

hydrodynamic injection (Figures 5-1 & 5-16).

In mo/huAPP/PS16E9 mice expression of pUFGL results in a significant decrease in

the concentration of total brain amyloid 3 1-42 (Figure 5-2). The decreased

concentration of A342 is accompanied with a significant decrease in the dense-cored

amyloid deposit load when compared to untreated mice in mo/huAPP/PS1lE9 (Figure 5-4A).

This reduction of dense-cored deposit load is accompanied with a significant decrease in

number of both dense-cored and diffuse deposits observed per section of pUFGL-injected

mice (Figure 5-4D & 5-5D). Surprisingly, the reduction in dense cored deposit load in

mo/huAPP/PS 1E9 mice was not accompanied by significant reductions in diffuse deposit

loads (Figure 5-5A). However, a strong trend toward significance was observed.









Our data indicates that the reduction of amyloid deposits may be associated with an

increase in the apparent staining of soluble amyloid oligomers (Figure 5-9) and an

increase in the apparent activation state of microglia in the brains of mo/huAPP/PSl6E9

mice injected with pUFGL (Figures 5-12 & 13). Although statistically significant, our

results may just be an artifact due to variability of tissue treatment post-mortem.

However an increase in oligomer staining could explain an increase in the activation state

of microglia. It has been reported that soluble amyloid and not insoluble amyloid

activates microglia (Floden and Combs, 2006; Lindberg et al., 2005). Activated

microglia have been implicated of playing a role in antibody-mediated clearance of

dense-cored amyloid deposits in transgenic mice (Wilcock et al., 2003). So it would not

be surprising if microglia play a similar role in gelsolin mediated clearance of dense-

cored deposits in our study also. This would be a significant observation because before

clearance by peripheral amyloid binding agents was thought to be a result of mass action

diffusion according to the peripheral sink hypothesis.

A decrease in total brain amyloid 3 1-42 was also observed in huAPP/PS1M146L mice

when inter-litter variability is accounted for (Figure 5-18). In contrast to the

mo/huAPP/PSl6E9 group, huAPP/PSlM146L did not show a significant reduction in dense

cored amyloid deposit load (Figure 5-21), nor was there a difference in diffuse amyloid

deposit load (Figure 5-24). However both displayed a strong trend towards a reduction in

amyloid load. No differences were evident among deposit size, but strangely when inter-

litter variability was accounted for there was a significant increase in the dense-cored

deposit number fraction (Figure 5-21).









There were also no differences observed in amyloid oligomers staining intensity,

microglia staining intensity, or % GFAP stained. Only slight trends were seen when

inter-litter variability were accounted for (Figure 5-26). The possibility of variability

existing to the response of peripheral human plasma gene expression motivated us to

examine if there were any correlations amyloid pathology and glial response. When

correlations were examined (Figure 5-27) strong trends were evident indicating a possible

correlation between decreasing total amyloid 3 1-42 concentrations and increasing

soluble oligomer or microglia staining intensity compared to untreated littermates.

A strong correlation was observed between decreases in diffuse amyloid deposits

and increases in the % area stained for GFAP indicating astrocytes may be involved in

the clearance of diffuse amyloid pathology. Evidence supporting this hypothesis has

been observed in post-mortem studies of non-demented humans (Funato et al., 1998;

Yamaguchi et al., 1998). Adult mouse astrocytes have also been observed to migrate

toward and degrade immobilized amyloid 3 in vitro and in situ in transgenic mouse brain

sections (Wyss-Coray et al., 2003).






78






pUFGL pUFGL No 500 b.p.
+ + + Low High RNA marker











Figure 5-1 Detecting message in mo/huAPP/PSl6E9 mice. RT-PCR using primers specific
for mRNA coded from pUFGL vector, using RNA purified from liver tissue
from non-injected mice (- symbols), liver tissue from pUFGL-injected mice
(+ symbols), and 293 cells transfected with pUFGL at a high or low
confluency. No RNA refers to a product from a reaction mixture with no
RNA. 500bp marker from Biorad was used.



3000-
SP =0.0222





1000-


0 I I
Untreated pUFGL
Figure 5-2 Amyloid 3 (1-42) concentrations in mo/huAPP/PSl6E9 brains. Diamonds
represent individual animal means, bars represent means and standard error.
Untreated mice had a mean SEM AP (1-42) concentration of 2,306 202.6
pmoles/g of brain tissue, and the injected animals had 1,174 334.7 pmoles/g
of brain tissue. A one-tailed t-test was performed and showed that these
groups differ significantly with a P value = 0.0222.






















Untreated


300-

CM
F 200-

wo
El
Y 100-
(3_


IIF.


Untreated


~+7~


A
P = 0.237


pUFGL


B
P = 0.200


C
P = 0.045


U.U L J
U-U- ---- )--------) ----
Untreated pUFGL
Figure 5-3 Plasma concentrations of amyloid in mo/huAPP/PSl6E9 mice. A) Amyloid
p 1-40 Concentrations in fmol/mL. B) Amyloid 3 1-42 Concentrations in
fmol/mL. C) The Ratio of Amyloid 3 1-40 to Amyloid 3 1-42
Concentrations.


pUFGL


500-


400-

E 300-
o
200-

100-


IMF- I







80


1.25 1 B
_-c;i P=0.010
1.00 P 0.0085 75000 P 0010
71 0.75


025

Untreated pUFGL Untreated pUFCL


50-c 70o-
450 P = 0.091 600- P 0.029
4. =0 50..029
35 4D
3 2 300-
30
200-
25 10
20 -----------| --- 0 --- |-------| ---
Untrated pUFGL Untreated pUFGL

Figure 5-4 Analysis of dense-core amyloid deposits inmo/huAPP/PS1lE9 mice. Deposits
visualized by thioflavin S staining of brain sections. A) Percent Amyloid
Burden. B) Total Stained Area. C) Average Deposit Size. D) Average
Number of Deposits per section. Diamonds represent individual animal
means. Bars represent group means + the standard error.











A
P = 0.074


75000


(J*


Untreated pUFCL

25.0- C
5- P= 0.387
20.0
15.0- .-- -----

125-
10.0
7.5-
5.0
25-
0 0


Untreated


pUFCL


B
P = 0.078


Untreated


-


SJntrated


pUGL


D
P 0.035


pUGL


Figure 5-5 Analysis of diffuse amyloid deposits inmo/huAPP/PSl E9 mice. Deposits
visualized by immunostaining of brain sections. A) Percent Amyloid Burden.
B) Total Stained Area. C) Average Deposit Size. D) Average Number of
Deposits per section. Diamonds represent individual animal means. Bars
represent group means + the standard error.


T






















A.r
W?.1 ~_
w1 .


Figure 5-6 Dense-core amyloid deposits in mo/huAPP/PSlE9 mice. Deposits visualized by

thioflavin S Staining. Negative digital micrographs of untreated mice (left)

and pUFGL-injected mice (right). One section is included from each animal.


'


,,
IF






















































Figure 5-7 Diffuse amyloid deposits in mo/huAPP/PSl6E9 mice. Deposits visualized by
immunostaining with 6E10, digital micrographs of untreated mice (left) and
pUFGL-injected mice (right). One section is included from each animal.





84

Dense Soluble
Cored Diffuse Amyloid Microglia GFAP
Plaques q s ligomers



B/

U




















Figure 5-8 Side by side comparison of staining for dense cored amyloid deposits, diffuse
amyloid deposits, soluble amyloid oligomers, microglia, and GFAP in
mo/huAPP/PSlE9mice. Untreated are lower sections, pUFGL are upper.










150-

140-
(n
iD
o 130-
1M
0
120-


~I1 I *


P = 0.0220

--


-I--


Untreated


pUFGL


Figure 5-9 Staining intensity of soluble amyloid oligomers in mo/huAPP/PS 16E9 mice.
Increasing negative grey values correspond to darker sections. Diamonds
represent individual animal means determined from measurements from three
sections, while bars represent group means + SEM.


Figure 5-10 High magnification tiled images of amyloid oligomer staining. Untreated on
the left, and pUFGL-injected on the right. Scale bar represents 100 ptm for
both images.





































Figure 5-11 Cell types stained for soluble amyloid oligomers in pUFGL-injected mice.
Upper photo displays an astrocyte containing soluble amyloid oligomers.
Lower photo displays vasculature stained for soluble amyloid oligomers.
Deposits in the lower photo were counterstained with congo red and
visualized through a polarized light source. Scale bar represents 25 tm for
both images.













200-


- -i 150-
>,_,

0)
o 100-

0 50-


z LL


Ik _


P = 0.0496


t^


Untreated
Untreated


pUFGL


Figure 5-12 Staining intensity of microglia in mo/huAPP/PS16E9 mice. P value calculated
from an unpaired one-tailed t test.
























































Figure 5-13 Low magnification images of microglia. Untreated (left), and pUFGL-
injected (right) one image represented from each animal in the treatment
groups (scale bar represents 100 [tm).