USF Libraries
USF Digital Collections

Mechanisms of [beta]-amyloid clearance by anti-a[beta] antibody therapy

MISSING IMAGE

Material Information

Title:
Mechanisms of beta-amyloid clearance by anti-abeta antibody therapy
Physical Description:
Book
Language:
English
Creator:
Wilcock, Donna Marie
Publisher:
University of South Florida
Place of Publication:
Tampa, Fla.
Publication Date:

Subjects

Subjects / Keywords:
Immunization
Transgenic mice
Inflammation
Alzheimers disease
Dissertations, Academic -- Pharmacology and Therapeutics -- Doctoral -- USF   ( lcsh )
Genre:
government publication (state, provincial, terriorial, dependent)   ( marcgt )
bibliography   ( marcgt )
theses   ( marcgt )
non-fiction   ( marcgt )

Notes

Summary:
ABSTRACT: Alzheimers disease (AD) is defined as a progressive neurodegenerative disorder that gradually destroys a persons memory and ability to learn. There are three pathological hallmarks of the disease which are necessary for diagnosis of AD, these are; extracellular amyloid plaques composed of beta-amyloid (Abeta) protein, intracellular neurofibrillary tangles and neuronal loss. Amyloid plaques exist as both compact deposits which stain with Congo red and more numerous diffuse deposits. Active immunization against Abeta 1-42 or passive immunization with monoclonal anti-Abeta antibodies reduces amyloid deposition and improves cognition in APP transgenic mice.Over several studies of active immunization in APP+PS1 transgenic mice we showed a strong correlation between reduction of compact amyloid deposits and the degree of microglial activation suggesting a potential role of microglia in the removal of Abeta.Injection of anti-Abeta antibodies into the frontal cortex and hippocampus of aged APP transgenic mice revealed an early phase of Abeta removal which was removal of only diffuse amyloid deposits with no associated activation of microglia. A later phase was the removal of compact amyloid deposits. This was associated with significant activation of microglia. Prevention of this microglial activation by anti-Abeta F(ab)2 fragments or its inhibition by dexamethasone also precluded the removal of compact amyloid deposits but did not affect the removal of the diffuse deposits. Systemic injection of anti-Abeta antibodies weekly over a period of 1, 2, 3 and 5 months transiently activated microglia associated with the removal of compact amyloid deposits and elevated plasma Abeta, suggesting a peripheral mechanism contributes to removal of brain Abeta.
Thesis:
Thesis (Ph.D.)--University of South Florida, 2004.
Bibliography:
Includes bibliographical references.
System Details:
System requirements: World Wide Web browser and PDF reader.
System Details:
Mode of access: World Wide Web.
Statement of Responsibility:
by Donna Marie Wilcock.
General Note:
Title from PDF of title page.
General Note:
Document formatted into pages; contains 242 pages.
General Note:
Includes vita.

Record Information

Source Institution:
University of South Florida Library
Holding Location:
University of South Florida
Rights Management:
All applicable rights reserved by the source institution and holding location.
Resource Identifier:
aleph - 001681138
oclc - 62866021
usfldc doi - E14-SFE0001169
usfldc handle - e14.1169
System ID:
SFS0025490:00001


This item is only available as the following downloads:


Full Text

PAGE 1

Mechanisms of -Amyloid Clearance by Anti-A Antibody Therapy by Donna Marie Wilcock A dissertation submitted in partial fulfillment of the requirements for the degree of Doctor of Philosophy Department of Pharmacology and Therapeutics College of Medicine University of South Florida Major Professor: David Morgan, Ph.D. Marcia N. Gordon Ph.D. Keith Pennypacker Ph.D. Amyn Rojiani MD. Ph.D. Paula Bickford Ph.D. Date of Approval: January 21, 2004 Keywords: immunization, tran sgenic mice, inflammati on, Alzheimers disease Copyright 2005 Donna Marie Wilcock

PAGE 2

i TABLE OF CONTENTS LIST OF TABLES iii LIST OF FIGURES iv ABSTRACT vii INTRODUCTION Alzheimers disease 1 Cerebral amyloid angiopathy and AD 4 Transgenic mouse models of AD 6 Behavioral analyses of transgenic mice 9 Inflammation and AD 11 Immunotherapy for AD 14 PAPER 1: INTRACRANIALLY ADMI NISTERED ANTI-A ANTIBODIES REDUCE -AMYLOID DEPOSITION BY MECHANISMS BOTH INDEPENDENT OF AND ASSOCI ATED WITH MICROGLIAL ACTIVATION. 21 PAPER 2: MICROGLIAL ACTIVATI ON FACILITATES A PLAQUE REMOVAL FOLLOWING INTRACRA NIAL ANI-A ANTIBODY ADMINISTRATION. 59 PAPER 3: PASSIVE AMYLOID IMM UNOTHERAPY CLEARS AMYLOID AND TRANSIENTLY ACTIVIATES MICROGLIA IN A TRANSGENIC MOUSE MODEL OF AMYLOID DEPOSITION. 94 PAPER 4: PASSIVE IMMUNOTHERA PY AGAINST A IN AGED APP TRANSGENIC MICE REVERSES COGNITIVE DEFICITS AND REDUCES PARENCHYMAL AMYLOID DEPOSITS IN SPITE OF INCREASED VASCULAR AMYLOID AN D MICROHEMORRHAGE. 131 CONCLUSIONS 161 REFERENCES 176

PAGE 3

ii APENDICES 194 APPENDIX A 195 APPENDIX B 221 APPENDIX C 225 ABOUT THE AUTHOR END PAGE

PAGE 4

iii LIST OF TABLES PAPER 4 1: A loads after 5 months of immunotherapy. 162 CONCLUSIONS 1:Summary of the evidence found for the different mechanisms of -amyloid removal. 182 APPENDIX A 1: Effect of number of inoculati ons on Congo red and CD45 staining in the frontal cortex. 207

PAGE 5

iv LIST OF FIGURES PAPER 1 1: Time course of injected anti-A antibody distribution in the hippocampus from 4 hours to seven days. 36 2: Reduction in A immunohistochemist ry one day after anti-A antibody injections. 38 3: Quantification of reduced A load after anti-A antibody injections. 40 4: Reduction in thioflavine-S staining thr ee days after anti-A antibody injections. 42 5: Anti-A antibody injections results in a reduction of thioflavine-S positive plaques. 44 6: CD45 immunohistochemistry is increased three days following anti-A antibody injections. 46 7: Anti-A antibody injections results in an increased CD45 immunohistochemistry three days following injection. 48 8: MHC-II immunohistochemistry is incr eased three days following anti-A antibody injection. 50 9: Anti-A antibody injections resu lt in an increase in MHC-II immunohistochemistry three days following injection. 52 10: Anti-A antibody injections results in phagocytic rounded microglia in association with remaining Congophilic amyloid deposits thre e days after injection. 54 PAPER 2 1: Anti-inflammatory drugs impaired fibril lar amyloid removal to roughly the same extent as they decreased microglial activation following anti-A antibody injections. 84 2: Quantification of CD45, total A a nd thioflavine-S following inhibition of microglial activation by anti-inflammatory compounds. 86

PAGE 6

v 3: Anti-A F(ab) 2 fragments do not activate microglia, nor do they remove compact amyloid deposits as effectivel y as the complete anti-A IgG. 88 4: Quantification of CD45 a nd total A immunohistochemistry and thioflavine-S staining following intracranial injection of anti-A IgG and anti-A F(ab) 2 fragments. 90 PAPER 3 1: Y-maze behavioral improvement after systemic anti-A antibody administration. 119 2: Increased serum levels of anti-A antibody and A after anti-A antibody administration. 121 3: Mouse IgG immunohistochemistry shows antibody binding to Congophilic plaques in anti-A antibody tr eated mice but not in control antibody treated mice. 123 4: Total A immunohistochemistry is reduced following two months of systemic anti-A antibody administration. 125 5: Congophilic compact amyloid deposits are reduced following two months of anti-A antibody administration. 127 6: Fc receptor expression on microglia is increased following one month of anti-A antibody administration and remain s increased following two months of treatment. 129 7: CD45 expression on microglia is increas ed following two months of anti-A antibody treatment. 131 PAPER 4 1: Spatial learning deficits in APP transgenic mice were reversed following three and five months of immunization. 153 2: Passive immunization with anti-A anti bodies decreases total and parenchymal amyloid loads while increasing vascular amyloid in frontal cortex and hippocampus of APP transgenic mice. 155 3: Increased Congo red staining of bl ood vessels following anti-A antibody administration is associated with activated microglia. 157 4: Microhemorrhage associated with C AA following systemic administration of anti-A antibodies. 159

PAGE 7

vi 5: Number of Prussian blue positive pr ofiles increases with duration of anti-A antibody exposure. 161 APPENDIX A: 1: Antibody titer averages as a function of number of inoculations in transgenic and non-transgenic mice. 198 2: Congo red and CD45 staining in the hippoc ampus of mice receiving either A inoculations or control for five months. 200 3: Congo red levels in the hippocampus re lative to number of inoculations for vaccinated and control mice. 202 4: CD45 expression in hippocampus relativ e to number of inoculations for vaccinated and control mice. 204 5: Correlation of Congo red levels and CD45 expression both shown as percent control in hippocampus. 206 APPENDIX B 1: Total A is significantly reduced by anti-A antibodies regardless of epitope. 212 2: Thioflavine-S staining is significan tly reduced by anti-A antibodies regardless of their epitope. 214 3: CD45 immunohistochemistry is increased by anti-A 28-40 antibodies. 216 APPENDIX C 1: Phospho-p38 MAPK expression is decr eased in microglia with increasing duration of anti-A antibody treatment. 218 2: Phospho-p38 MAPK immunohistochemist ry shows microglial expression around amyloid plaques which is decreased with increasing duration of anti-A antibody treatment. 220 3: Phospho-p44/42 MAPK expression is incr eased in microglia with increasing duration of anti-A antibody exposure. 222 4: Phospho-p44/42 MAPK immunohistoche mistry shows microglial expression around amyloid plaques which is increased with increasing duration of anti-A antibody treatment. 224

PAGE 8

vii Mechanisms of -Amyloid Clearance by Anti-A Antibody Therapy. Donna Marie Wilcock ABSTRACT Alzheimers disease (AD) is defined as a progressive neurodegenerative disorder that gradually destroys a persons memory and ability to learn. There are three pathological hallmarks of the disease which ar e necessary for diagnosis of AD, these are; extracellular amyloid plaques composed of -amyloid (A) protein, intracellular neurofibrillary tangles and neuronal loss. Amyloid plaques exist as both compact deposits which stain with Congo red and more numer ous diffuse deposits. Active immunization against 1-42 or passive immunization with mono clonal anti-A antibodies reduces amyloid deposition and improves cogn ition in APP transgenic mice. Over several studies of active immuni zation in APP+PS1 transgenic mice we showed a strong correlation between reducti on of compact amyloid deposits and the degree of microglial activation suggesting a pote ntial role of microglia in the removal of A. Injection of anti-A antibodies into the frontal cortex and hippocampus of aged APP transgenic mice revealed an early phase of A removal which was removal of only diffuse amyloid deposits with no associated act ivation of microglia. A later phase was the removal of compact amyloid deposits. This wa s associated with si gnificant activation of microglia. Prevention of this microglial activation by anti-A F(ab) 2 fragments or its

PAGE 9

viii inhibition by dexamethasone also precluded th e removal of compact amyloid deposits but did not affect the removal of the diffuse deposits. Systemic injection of anti-A antibodie s weekly over a period of 1, 2, 3 and 5 months transiently activated microglia associated with th e removal of compact amyloid deposits and elevated plasma A, suggesti ng a peripheral mechanism contributes to removal of brain A. This systemic administration also dramatically improved cognitive performance in the Y-maze and in the radial-a rm water maze. These studies also showed a significant increase in vasc ular amyloid dependent on the number of antibody injections the mice received. Associated with this in crease in vascular amyloid was a dramatic increase in the numbers of microhemorrhages in the brain. Despite this pathology the mice showed cognitive improvement with the treatment. These effects could have major ramifications in humans and should be further investigated prior to any human clinical trials.

PAGE 10

1 INTRODUCTION Alzheimers Disease : Alzheimers disease (AD) is a prog ressive, neurodegenerative disorder characterized by a devastating mental decline. There are three pathological characteristics of AD. These are amyloid plaques, neurofibri llary tangles and neuron loss characterized by dystrophic neurites. The amyloid plaque is a microscopic focus of extracellular amyloid deposition. The plaque contains extracellular deposits of amyloidprotein ( ) that occurs primarily in fi lamentous form. Much of the found in the plaque is the 42 amino acid species ( 1-42 ) which is slightly more hydr ophobic than the shorter, 40 amino acid, species ( 1-40 ). 1-40 is normally the more abundant form of produced by cells and it does colocalize with 1-42 in the plaque. can also be deposited as diffuse deposits, often thought to be an intermediate step in the formation of a compact amyloid plaque. Amyloid plaques are often term ed neuritic plaques, due to the presence of dystrophic neurites within and immediat ely surrounding the plaque. The size of the deposit can vary greatly, rangi ng from 10 to more than 120 m (Selkoe, 2001). Neurofibrillary tangles are intraneuronal inclusions of nonmembrane bound bundles of abnormal fibers consisting of pairs of heli cal filaments. The filaments consist of hyperphosphorylated microtubule-associated pr otein tau. It is unknown what causes this hyperphosphorylation although studies have imp licated the cyclin-dependent kinase 5 (cdk5) (Patrick et al, 1999). Neuron loss is thought to result from toxicity of amyloid plaques, the inflammatory response which re sults in cytokine release and acute phase

PAGE 11

2 protein release, oxidative injury which causes disruption of neuronal metabolic and ionic homeostasis, and impaired neuronal transpor t due to the presence of hyperphosphorylated tau filaments in the neuron. Despite the many pathological characterist ics of AD, the most favored hypothesis of the disease process is the amyloid hypothe sis. This hypothesis suggests that deposition of as both diffuse and compact plaques ca n directly, and indirectly via an inflammatory cascade, result in progressive syna ptic and neuritic injury. This injury is then thought to result in altered kinase/phophatase activity which leads to hyperphosphorylation of tau and the formation of neurofibrillary tangl es. This cascade of events is ultimately thought to result in widespread neuron dysfunction and loss which will cause the dementia characteristic of AD (Hardy et al, 2002). The problem with this hypothesis has been that rese archers have been unable to show a strong correlation between neuron loss / dysfunction and levels of amyloid deposits, however, recent data suggests that it is actually small, soluble oligomers that cause the neurotoxicity in Alzheimers disease (Lue et al, 1999, Klein et al, 2001). is a product of cleavage of larger precursor protein, the amyloid pr ecursor protein (APP). APP is a single transmembrane polypeptide consisting of be tween 695 and 770 amino acid residues and is cleaved by enzymes named secretases. The three secretases have specific cleavage sites, using the 770 numbering -secretase cleavage occurs at amino acid 687, secretase cleavage occurs at amino acid 671, while -secretase cleavage can occur at amino acid 711 or 713. If a -secretase cleavage occurs along with a -secretase cleavage then will be produced, th e length of the is dependent on whether the -secretase

PAGE 12

3 cleaves at residue 711 ( 1-40 ) or 713 ( 1-42 ). Most mutations which have been found to cause early-onset familial AD are in the APP protein at the three secretase sites, primarily biasing cleavage to the secretase site or C-terminal mutations increasing the length of A to the more fibrillogenic 42 lengt h. Other mutations were found in the presenilin proteins 1 and 2, later it was discovered that these proteins alter APP metabolism and have a direct effect on -secretase by increasing the length of the A produced (Hardy 1997). It has recently been suggested by several studies that the presenilin protein is actually the -secretase. Researchers continue to investigate geneti c links to susceptibility for Alzheimers disease since only 5% of all AD cases are li nked to APP, PS1 or PS2 mutations and all are early-onset forms of th e disease. Apolipoprotein E (ApoE) genotype has been found to be a significant risk factor for the deve lopment of AD. ApoE is a plasma protein involved in the transport of cholesterol and other lipids. ApoE ha s been shown to be present in amyloid deposits and neurofibri llary tangles and has been implicated in neuronal growth and regeneration during deve lopment and following injury. All humans carry two alleles for ApoE, of which there are three types; 2, 3 and 4. A person can be ApoE 4/4, ApoE 4/3, ApoE 3/3, ApoE 2/3 or, rarely, ApoE 2/2 (Soininen and Riekkinen Sr, 1996). It has been shown that onset of AD is earliest in those pa tients carrying both ApoE4 alleles while one ApoE4 allele is later but still earlier than other alleles with the rare ApoE2 allele possibly inferring protection from the disease. ApoE3 is the most common allele. Although only 16% of the population has ApoE4 as an allele approximately 40% of sporadic AD cases ha ve been found to carry an ApoE4 allele (Strittmatter and Roses, 1995). The role of ApoE in the brain is not fully understood, nor

PAGE 13

4 is the role ApoE4 plays in increasing risk of AD. In vitro it has been shown that ApoE2 and ApoE3 can bind to tau and microtubule asso ciated protein 2c but ApoE4 cannot. This could suggest that binding of ApoE may stab ilize the tau protein and possibly prevent the aggregation into neurofibrillar y tangles (Strittmatter et al, 1994). However, this may be one of many functions of ApoE as it has been shown to be localized in neurons as well as the extracellular space of the brain and has been shown to have many metabolic functions. It is also importan t to note that an ApoE4/4 ge notype does not guarantee that Alzheimers disease will occur, only that th e person has an increased risk for developing the disease. It is highly likely that ma ny more genes like ApoE will be found in the coming years given that there are still many AD cases without an obvious genetic cause. Cerebral Amyloid Angiopathy and AD: Cerebral amyloid angiopathy (CAA) is a common term used to define the deposition of amyloid in the walls of blood vessels, primarily small and medium sized arteries and arterioles, of th e brain. In humans, CAA primarily occurs in leptominingeal and cortical vessels and is rarely observed in other brain regions such as the hippocampus or the striatum (Rensink et al, 2003). The protein accumulating in the vessels causing CAA has excessive -pleated sheet folding and also has a tendency to form fibrils which are highly insoluble. In order to be defi ned as CAA the deposits must be stained by Congo red, a dye which stains fibrils compri sed of -pleated sheet folding. There are approximately seven proteins known to cause CAA, these are A, cystatin C, transthyretin, Gelsolin, prion protein, Abri (familial British dementia) and ADan (familial Danish dementia) (Castellani et al, 2004). In AD the protein causing CAA is the A protein.

PAGE 14

5 The most severe consequence of CAA is cerebral hemorrhage, known as CAAassociated hemorrhage (CAAH). There are seve ral types of hemorrhages occurring with CAA ranging from microhemorrhages (small leaks in the vessel wall) to aneurysm (a blood-filled dilation of the blood vess el) (McCarron and Nicoll, 2004). These hemorrhages can result in further cognitive decline or, if severe, even death. CAA can occur alone or in conjunction with parenchymal amyloid deposits and neurofibrillary tangles in AD. When CAA occurs alone it can cause extensive dementia. There are several forms of hereditary CAA su ch as the Dutch type (HCHWA-D) (Natte et al, 2001) and the Iowa pedigree (D694N) (Grabowski et al, 2001), both mutations causing these hereditary CAAs lie in the APP molecule and result in A formation. Unlike those mutations of APP occurring in some familial AD cases which produce excess A 1-42 mutations causing CAA result in excessive production of A 1-40 which appears to be excessively fibrillogenic in human cerebrovascular smooth muscle (HCSM) (Grabowski et al, 2001). The role of CAA in AD is not yet fu lly known. The reported incidence of CAA in Alzheimers cases has ranged from 78% to 98% (Kallaria and Ballard, 1999) which suggests and important role for CAA in the pathogenesis of AD although not necessary for diagnosis of AD. Of those cases, approxi mately 35-40% are associated with some form of hemorrhage (Jellinger et al, 2002). Support for the contribution of CAA to the cognitive decline observed in AD arises from observations that CAA produces ischemia and hemorrhage that in other disease pr ocesses is known to result in cognitive dysfunction (Cadavid et al, 2000) and also th at the frequency and severity of CAA is increased in AD (Yamada, 2002). Howver, ther e is yet no evidence to suggest that the

PAGE 15

6 rate of cognitive declin e in AD patients withou t CAA is any dfferent compared to those AD patients with CAA. Also, in non-demen ted individuals CAA appears to have no effect on cognitive ability (Castellani et al, 2004). Transgenic mouse models of AD: Transgenic mouse models for AD became an aim for researchers following the discovery of many genetic mutations in th e APP and PS1 genes thought to cause earlyonset familial AD. An ideal AD animal mode l would develop all pathological hallmarks of AD, as well as the cognitive and memory deficits characteristic of AD. This mouse model would then be the clos est thing to an AD patient to allow testing of potential therapies. A mouse carrying the M146L mutatio n in the PS1 gene (methionine to leucine at 146) showed increased production of 1-42 / 1-40 compared to widtype littermates as measured by ELISA, however, these mice did no t deposit amyloid, either as diffuse or compact plaques (Duff et al, 1996). Mice expr essing mutations in the APP gene showed more promise and as a result several APP transgenic mice were produced. The PDAPP mouse carries the V717F mutation under the control of the plat elet derived growth factor promoter and expresses APP695, APP751 a nd APP770. This mouse begins amyloid deposition between 4 and 6 months of age, accel erates rapidly at 7 to 9 months of age with significant numbers of both diffuse and compact amyloid deposits in the frontal cortex and hippocampus by 1 year of ag e. The Tg2576 mouse carries the Swedish mutation of KM670/671NL under the control of the hamster prion protein promoter and express APP695 (Hsiao et al, 1996). These mice have detectable diffuse and compact amyloid deposits by 6 months of age and con tinue to deposit in an age-dependent manner showing an acceleration between 8 and 12 months of age (Kawarabayashi et al, 2001).

PAGE 16

7 Despite the ability to produce transgenic mice that develop amyl oid deposits in an age-dependent manner, very few of these APP transgenics demonstrated reproducible cognitive and memory deficits. Crossing the M146L PS1 mouse with the Tg2576 APP mouse not only showed accelerated amyloid deposition (Holcomb et al, 1998, Gordon et al, 2002) but also developed cognitive and memory deficits detectable by several behavior paradigms which were reproducible at defined ages (Gordon et al, 2001). The benefits of this mouse model ar e that therapies can be tested to show not only their effect on amyloid deposition but also whether they may have any clinical benefit by showing whether the treatment improves cognitive function. Although this doubly transgenic mouse is a good model in which to test ther apies, it does still lack two of the three pathological hallmarks of AD, those of neurof ibrillary tangles and ne uronal loss, despite the abundance of amyloid deposits. Recently, there have been several tau tr ansgenic mice developed, which provide a further step toward the ideal mouse mode l for AD. The P301L mutation on chromosome 17 expressed in mice results in devel opment of hyperphosphorylated tau and neurofibrillary tangle s detectable by Gallyas silver staining. The disadvantage to this mouse is that it also develops motor deficits due to expression of mutated tau in the brain stem and spinal cord, the animals are completely paralyzed by 12 months of age (Lewis et al, 2000). The mice also show differential expression between males and females, with females having 3 to 4 times more expression of the mutated tau than males. Despite the disadvantages of this mouse model it has been shown that crossing the P301L mouse with the Tg2576 mouse enhances forebrain neurofib rillary tangle formation, suggesting that the presence of influences the extent of neurofibri llary pathology (Lewis et al, 2001).

PAGE 17

8 Transgenic mice expressing V337M mutant human tau show hyperphosphorylated tau and neurofibrillary tangles in the hippo campus resulting in behavioral abnormality (Tanemura et al, 2002). The ideal mouse model would have all pathological hallmarks of AD; amyloid plaques, neurofibrillary tangles and neuron lo ss. The closest mouse model to date is a triple transgenic developed by Frank LaFe rla and colleagues (Oddo et al, 2003). This group showed a mouse with the M146L PS1 mutation knock-in as well as mutant tau and APP which both have the Th1 promoter and bo th co-integrate at the same site, this produces a closer model of AD than had previo usly been shown. All three transgenes are expressed to homozygocity and are expresse d at the same levels. The mice develop amyloid plaques and neurofibrillary tangles. Interestingly the mie develop the amyloid plaques prior to any tau pathology being observed. The mice also demonstrate agedependent LTP impairment although this oc curs prior to any AD-like pathology being present. At 6 months of age the mice have impaired long-term potentiation (LTP) suggesting that the mice may develop cognitive deficits. However, to this date no neuron loss has been reported in this transgenic model. A better mouse model for AD may be possi ble thanks to a new mouse model of tau pathology. In a recent report from Peter Davies and colleagues a mouse, known as the htau mouse, undergoes age related accumula tion of hyperphosphorylat ed tau like those observed in AD and the presence of neurofib rillary tangles (Andorfer et al, 2003). These htau mice are a cross of two existing mouse lines One is a tau transgenic known as the 8c mouse which expresses all human tau isoforms but alone does not demonstrate any evidence of tau pathology (Duff et al, 2000) The other is a tau knockout mouse which

PAGE 18

9 again, alone does not develop any tau pathology (Tucker et al, 2001). When crossed, a mouse with all human isoforms but no m ouse tau is produced which demonstrates extensive and age-related ta u pathology. This htau mouse has also importantly been shown to develop age related neuron loss in cortical and hippocampal areas. By the age of 18 months there is a 50% re duction of neurons in the pirifo rm cortex. There is also an apparent shrinkage of the cortex and enlargem ent of the ventricles (Andorfer et al, 2004). Since this mouse appears to show extens ive tau pathology and neuron loss, it is hoped that crossing this mouse with an APP tran sgenic mouse may yield a more perfect mouse model in whch to test potential treatments for AD. Behavioral Analyses of transgenic mice: The major clinical symptom of AD is cogni tive decline so therefore any effective clinical therapy must act to improve cognitive function of patients so it is critical that potential therapies are shown not only to affect the pathol ogical characteristics of AD such as amyloid plaques or ne urofibrillary tangles but must also affect memory. To test memory impairment in mice there have been several behavioral paradigms developed. The Morris water maze was first described in 1982 by Richard Morris and colleagues (Morris et al, 1982) where he s howed impairment in the task following hippocampal lesions. This task is consistently used to assess memory retention. It consists of a water pool with a hidden escape platform where the mouse must learn the location of the platform using either contextual or lo cal cues. The mouses aversion to water and swimming force it to look for an escape and therefore search out the platform. This task has been shown to be heavily hippocampal dependent, where lesions to the hippocampus or its cholinergic input signi ficantly impair performance. Time taken to locate the

PAGE 19

10 platform is measured which is known as escap e latency. Also measured is the time spent in the quadrant where the platform was once removed, these trials are known as probe trials. A modification of the Morris water maze is the radial arm water maze (RAWM) which is a circular pool with six swim alleys (arms) radiated out from an open central area with a submerged escape platform located at the end of one of the arms which the animal must find. There are several spatial cu es present on the walls and ceiling of the testing room. The platform remains in the same arm during testing for each mouse, however, the arm in which the mouse starts e ach time is different requiring the mouse to use the visual cues in order to remember where the platform is (Diamond et al, 1999). The number of wrong arms entered is measured as errors and also time to find the platform is measured. Again, this is a heavily hippocampal dependent task and performance has been shown to be impaired in some transgenic mouse models of AD (Arendash et al, 2001; Morgan et al, 2000). Another memory task is contextual fear conditioning which uses an aversive stimulus coupled to sound. The animal learns to freeze when the sound is heard as it is associated with the aversive stimulus wh ich is commonly a small electric shock. The amount of freezing is measured and lack of fr eezing is associated with impaired memory of the preceeding events (Gerla i, 2001). This task is highl y dependent upon the integrity of the amygdale however is also sensitive to disruptions in hippocampal function. The Y-maze is not dependent upon learni ng a new behavior but depends upon the tendancy of a mouse to explore new environm ents. The Y maze is a three arm maze with equal angles between all arms. Mice are init ially placed within one arm and the sequence and number of arm entries is recorded for each mouse over set period of time (usually

PAGE 20

11 between five and ten minutes). The percenta ge of triads in which all three arms are represented is recorded as an alternation to estimate short-term memory of the last arms entered. The total number of possible alternat ions is the number of arm entries minus two. Additionally, the number of arm entries serves as an indicator of activity. Inflammation and AD : Along with the three primary characteristi cs of AD there is also an extensive inflammatory response in the brain. Microg lia and astrocytes are the two primary inflammatory cells in the brain and these res pond to damage and foreign material in much the same way as do the immune cells of th e periphery. Microglial cells are from the monocytic lineage and are the resident macrophage in the brain. They have the ability to produce complement proteins in vitro, potentially contributing to the complement cascade. They can produce and secrete IL1, a cytokine with ma ny immune functions. Microglial cells can also enter into a phagocyt ic state, at which poi nt they are almost indistinguishable from a macrophage (Streit, 2002, Liu et al, 2003). Astrocytes are cells native to the CNS and have many normal functions such as inducing the blood brain barrier and contributing to th e local homeostasis of the s ynapse by expressing reuptake proteins on their membrane. Astrocytes also have the ability to produce inflammatory mediators when activated and are thought to communicate with mi croglial cells through these mediators. In AD activated microglia cluster at s ites of amyloid deposition, surrounding the deposit. Activation can be ini tially detected by an increase d expression of the leukocyte common antigen CD45 (Aloisi, 2001), a functional transmembrane protein-tyrosine phophatase (Justement, 1997). In later stages of activation there is further increase in

PAGE 21

12 expression of CD45 along with expression of the major histocompatibility complex class II (MHC-II). Much of the literature in the past decade has focused on the deleterious consequences of microglial activation (Akiya ma et al, 2000). Micr oglia are capable of producing many cytokines, reactive oxygen inte rmediates, excitatory amino acids and nitric oxide (NO) (Streit et al, 1999), all of which coul d significantly contribute to neuronal death seen in AD. In vitro, can stimulate release of IL-1, IL-6, TNFand superoxide free radicals from microglia (McGeer and McGeer 2001). IL-1 has an autocrine induction in microg lia and also enhances micr oglia proliferation, it causes direct neurotoxicity and apoptosis. IL-6 cause s astrogliosis but can be both a survival factor and a neurotoxic factor depending on its levels. TNFcan cause nitric oxide production and MHC-II expressi on in microglia (Wilson et al, 2002). All of this data led to the hypothesis that inflammation in th e AD brain, particularly the activation of microglia, contributes negatively to the di sease process, and inhibition of this inflammation was the target of AD therapies. It has been shown that the glucocorticoid anti-inflammatories are capable of inhibiting microglia activation as detected by nitric oxide production (Chang and Liu, 2000) a nd by measurement of the extent of proliferation (Tanaka et al, 1997). Epidemiologi cal studies have shown a beneficial effect of NSAIDs in the prevention of AD. It was thought that this beneficial effect was due to inhibition of inflammation, however, a repor t in 2001 showed that the effects of nonsteroidal anti-inflammatory drugs (NSAIDs) in AD may actually be independent of cyclooxygenase (COX) activity and may in fact be due to an effect at the -secretase enzyme. This study showed that in culture d cells following treatment with ibuprofen

PAGE 22

13 there is a decrease in 42 production but an increase in (1-38), which is not amyloidogenic, although the concentrations of ibuprofen used in this study are much greater than those requir ed for COX inhibition (We ggen et al, 2001). A novel nonsteroidal anti-inflammatory drugs (NSAID) with a nitric oxide donor group and an antioxidant group showed an unexpected increase in microglial activation when administered to APP+PS1 mice, associated with a significant re duction in diffuse and compact amyloid deposits (Jantzen et al, 2002 ). Recent data suggests that inflammation in AD is much more complex than origina lly thought and that microglia may have a beneficial role to play in AD. Microglia have been shown to phagocytose both in vitro and in vivo through several different mechanisms involving ops onization through the complement cascade (Rogers et al, 2002) or the scavenger receptor (Paresce et al, 1996). Curiously, however, 3D reconstruction of the microglia by electron microscopy in untreated transgenic mice was unable to detect intracellular amyloid despite amyloid fibrils being completely engulfed by microglia (Stalder et al, 2001). In terestingly, it has been shown that coculture of microglia wi th astrocytes suppresses microglia phagocytosis of senile plaques (DeWitt et al, 1998). Removal of by microglia has led to the hypothesis that there exists a dynamic equilibrium between deposition and removal, and that inhibition of microglia activation may, in fact, result in greater deposition of and a more rapid progression of the disease. An interesting finding to support the be neficial role of microglia is that when APP transgenic mice express soluble complement receptor related protein y (sCrry), a complement inhi bitor, a 2 to 3-fold increase in deposition is seen as well as a prominent accumulation of dege nerating neurons not normally seen in the

PAGE 23

14 APP transgenic mouse alone. Microglial activatio n is reduced significantly in this mouse model, suggesting that inhibiting compleme nt, therefore inhibiting opsonization and phagocytosis, will have a detrimental effect on the disease process (Wyss-Coray et al, 2002). Another transgenic model that has supporte d the beneficial role of microglia is a mouse that overproduces transf orming growth factor (TGF)1 crossed with an APP transgenic mouse. In this m odel, the overexpression of TGF1 results in a vigorous microglial activation along with a 50% reduction in deposition (Wyss-Coray et al, 2001). TGF1 has also been shown to result in clearance of by microglial cells in culture. Further support fo r microglial removal of was shown when lipopolysaccharide (LPS) was intracranially injected into the hi ppocampus of APP+PS1 mice. One week following the injecti on, there was a significant removal of associated with significant microglial activation; how ever, compact plaques were not removed (DiCarlo et al, 2001; He rber et al, 2004). Immunotherapy for AD : Using the amyloid hypothesis as the basi s for the development of AD therapies, Dale Schenk and colleagues at El an pharmaceuticals reported the use of 1-42 the amyloidogenic protein in AD, as an imm unogen. They immunized PDAPP transgenic mice with 1-42 in an aggregated / fibrillar preparat ion which is emulsified in Freunds adjuvant to increase the immune response to the antigen. Each mouse received 100 g This was repeated 2 weeks later and then monthly thereafter. It was shown that immunization reduced and/or prevented accumulation in this mouse model and associated with this reduction was an activation of microglia suggesting that part of the

PAGE 24

15 mechanism of clearance involved these infl ammatory cells. Subsequent work by our group demonstrated that not only did immunization modestly reduce amyloid burden but more importantly it prevented cognitive impairment in the doubly transgenic APP + PS1 mice (Morgan et al, 2000). This finding wa s also shown by Janus and colleagues at the same time in a different mouse model, the TgCRND8 mouse which is transgenic for APP only (Janus et al, 2000). Following this our group conducted several more immunization studies which le d to the finding that follo wing immunization there is a strong correlation betw een microglial activation and re duction in the congophilic, compact amyloid deposits (Wilcock et al, 2001). There has also been data to show that antiantibodies can dissolve fibrils in vitro (Solomon et al, 1997). More recent data to support this dissolution of plaques showed that antiF (ab)2 fragments directly applied to the brains of Tg2576 or PDAPP mice results in reduction of and thioflavine-S comparable with the reduction seen when a whole antiIgG is applied (Bacskai et al, 2002). The vaccine, now known as AN1792, advanced to human clinical trials, however, during phase II trials there were several patients found to be suffering from cerebral inflammation and meningoencephalitis (Bow ers and Federoff, 2002, Munch et al, 2002). It is important to note that the occurance of the meningoencep halitis did not correlate at all with antibody titers in t hose patients (Orgogozo et al 2003). Interestingly, it was shown that the antibodies ge nerated by humans in respons e to the AN1792 immunization recognized both diffuse and compact amyloid deposits in transgenic mouse and human brain tissue, however, it di d not cross react with full length APP or vascular (Hock et al, 2002).

PAGE 25

16 Since the original report of meningoence phalitis in some participants of the AN1792 trial there have been several reports on pathology from some of the participants following their death. The first was from James Nicoll and colleagues (2003) who reported on a 72 year old female who had been clinically diagnosed with moderate AD and had developed meningoencephalitis following the fifth injection. Some features of AD were detected such as cortical atrophy, ventricle enlargement and the presence of neurofibrillary tangles and amyloid plaques. However, it was observed that this patient that amyloid plaques were sp arse in comparison with c ontrol AD cases throughout most of the neocortex. Reactive microglia were also observed and appeared to colocalize with immunostaining for A. Importantly, it was shown that vascular amyloid deposits persisted and appeared not to be reduced by immunization. In a re port published a year later showing the pathology from another patient the findings were very similar (Ferrrer et al, 2004). This report shows the brain pathology from a 76 year old male who developed meningoencephalitis following th e second immunization. Tau pathology was comparable to that found in AD control brai n, however, very low numbers of amyloid plaques were observed and remaining plaques were associated with high numbers of activated microglia. Also observed were se vere small blood vessel disease and multiple cortical hemorrhages. Hock et al (2003) reported that antibody titers measured by ELISA were not a good indicator of cognitive performance following immunization, however a new method for measuring antibody reactivity was deve loped called the tissue amyloid plaque immunoreactivity (TAPIR) assay. The TAPIR assa y measures the binding capacity of the circulating antibodies to amyloid plaques in transgenic mouse tissue and human AD brain

PAGE 26

17 tissue. Those patients showing strong TAPI R results also showed stabilization of cognitive performance while those patients without TAPIR reactivity results showed a normal cognitive decline for AD. Together with the pathology reports from two patients in the trial these data suggest that immunot herapy may be a promising approach to the treatment of AD if the meningoe ncephalitis can be avoided. Subsequent to the initial studies of immunization using 1-42 in 1999 and 2000, researchers began to look at passive immuniza tion as a potentially sa fer approach to an immunotherapy since it had been shown that active immunization may be less effective in people with a significant amyloid burde n. A study suggests that the vaccine may be much more effective at preventing amyloid de position as opposed to removal of existing deposits (Das et al, 2001). It has also been found by work in our laboratory that the ability to produce sufficient antibody titers against 1-42 may decline with age. The first report of passive immunization came in 2000 from Elan pharmaceuticals, who administered antibodies against via an intraperitoneal injection in PDAPP mice (Bard et al, 2000). The antibodies were administered weekly for a period of six months. This resulted in a significant reduc tion in plaque burden in both the cortex and hippocampus. Importantly they showed by immunohistoche mical methods that the injected antibody does cross the blood-brain barrie r and enter the brain. They al so estimated the amount of IgG entering the brain to be 0.1% as calc ulated by examining endogenous IgG levels in brain parenchyma. The amount of IgG en tering the CNS was confirmed by a study specifically aimed at detection of peripherally injected antiIgG in the CNS of SAMP8 mice, which overexpress APP but do not deposit amyloid. The authors found that 0.11% of injected IgG ente rs the brain within 1 hour, IgG was still detect able in the

PAGE 27

18 brain 72 hours following injection (Banks et al 2002). In the Bard et al study (2000) the authors demonstrated that although the anti bodies they had used for this immunization were capable of triggering mi croglial-mediated phagocytosis of amyloid in culture, F (ab)2 fragments were unable to activate microglial removal despite retaining the full ability to bind to This result suggests that clearance of fibrillar amyloid is via Fc receptor mediated phagocytosis. This proposed m echanism was further supported by a study involving direct imaging of amyloid deposits in living mice using multiphoton microscopy (Bacskai et al, 2001). Three days following direct a pplication of antiantibodies to the brain of PDAPP mice th ere was significant removal of amyloid deposits accompanied by activ ation of microglia surroundi ng the remaining deposits. Human postmortem microglia have b een shown to phagocytose opsonized which is inhibited in the presence of excess non-specific IgG, suggesting the phagocytosis is Fc receptor mediated (Lue et al, 2002). More recently, studies involving passive immunization have suggested that the primary mechanism for clearance is peripheral and is not due to the antibodies entering the CNS. It has been shown that following intraperitoneal injection of antiantibodies in the PDAPP mouse there is a rapid 1,000-fold increase in circulating plasma suggesting that circulating antibodies bind to plasma and thus cause a disruption in the equilib rium between the brain and plasma removing from the brain. This study importantly used anti bodies that have been shown not to bind to plaques in the brain, so their effect on brain amyloid burden is not due to bi nding fibrils (DeMattos et al, 2001). This same group also demonstrated that administration of this same antibody to PDAPP mice is capable of reversing memory de ficits in only one da y without a reduction

PAGE 28

19 in amyloid burden in the brain (Dodart et al 2002). The authors suggest that this rapid reversal of cognitive deficits is due to removal of soluble from the CNS as opposed to reducing brain amyloid plaque burden. Cognitive improvement following passive immunization has also been shown in the Tg2576 mouse with an antibody recognizing A-12 which did not reduce brain A levels but did reverse memory deficits (Kotilinek et al, 2002). Application of an ti-A antibodies to the surface of the brain has been shown to not only reduce the size and number of amyl oid deposits but also to recover dystrophic neuritis from a curvy, distorted appearance to a straighter, more normal appearance (Lombardo et al, 2003). It appears that an importa nt issue of passive immunizat ion is the antibody isotype. It has been shown that IgG2a antibodies clear A from PDAP P brain sections in an ex vivo assay much more effectively than eith er IgG1 or IgG2b antibodies despite all antibodies having the same epitope (Bard et al, 2003). This data also supports the hypothesis that microglia are re sponsible for the clearance of A by immunotherapy since Fc RI and III bind with the greatest affinity to murine IgG2a antibodies (Radaev and Sun, 2001). The fact that IgG2a anti-A antibodies ap pear to be the most effective indicate that A clearance may be mediated through microglial Fc receptors. However, conflicting data suggests that effective cl earance of A by anti-A antib odies can be obtained in the absence of Fc receptors. Das et al (2003) showed that when they actively immunized APP transgenic mice crossed with Fc receptor knockout mice they showed the same amount of A reductions as immunize d, age-matched APP transgenic mice. A concerning effect of passi ve immunization is a report sh owing an increase in cerebral microhemorrhage in very old APP23 mice following passive antiimmunotherapy

PAGE 29

20 (Pfeifer et al, 2002). This effect has not b een shown in any other study, but will need to be investigated if passive immuni zation is to enter human trials. To summarize, there are three main proposed mechanisms of action of immunotherapy for AD. The first is the binding of antibody to and resulting in Fc receptor mediated phagocytosis. The second is that antibodies binding to cause a disggregation of the plaque a nd result in a dissolution. The th ird is that the effects are primarily peripheral, with ci rculating antbodies binding to in plasma and causing a disruption in equilibrium between brain and plasma causing the to be drawn out of the brain. It would be nave to think that only one of these could occur, as the data for all three is convincing. However, the main quest ion is which mechanism is going to produce the most beneficial effect with the least adverse effects and whether immunotherapy can be manipulated to take advantage of this mechanism.

PAGE 30

21 PAPER 1: INTRACRANIALLY ADM INISTERED ANTIANTIBODIES REDUCE AMYLOID DEPOSITION BY MECHANISMS BOTH INDEPENDENT OF AND ASSOCIATED WITH MICROGLIAL ACTIVATION. 1 Donna M Wilcock, 1 Giovanni DiCarlo, 1 Debbi Henderson, 1 Jennifer Jackson, 1 Keisha Clarke, 2 Kenneth E. Ugen, 1 Marcia N. Gordon and 1 Dave Morgan*. Departments of 1 Pharmacology and 2 Medical Microbiology and Immunology, Alzheimers Research Laboratory, University of South Florida, Tampa, Florida, 33612. This work was published in Journal of Neuroscience 2003 May 1 23(9): 3745-3751. ACKNOWLEDGEMENTS: This work was supp orted by National Institutes of Aging / NIH grants AG15490 (MNG), AG18478 (DM) and AG20227( KEU). DW is the Benjamin Scholar in Alzheimer's Disease Research.

PAGE 31

22 Abstract Active immunization against 1-42 with vaccines or passive immunization with systemic monoclonal anti-A antibodies reduces amyloid deposition and improves cognition in APP transgenic mice In this report, intracranial administration of antiantibodies into frontal cortex and hippo campus of Tg2576 transgenic APP mice is described. The antibody injections initially results in a broad distri bution of staining for the antibody which diminishes over 7 da ys. While no loss of immunostaining for deposited A is apparent at 4 hours, a dramatic reduction in the load is discernable at 24 hours and maintained at 3 and 7 days. A reduction in thioflavine-S positive compact plaque load is delayed until 3 days, at which time microglial activation also becomes apparent. At one week after the injection, the microglial act ivation returns to control levels, while the and thioflavine-S stai ning remains reduced. The results from this study suggest a two phase mechanism of antiantibody action. The first phase occurs between 4 and 24 hours, clears primarily diffuse deposits and is not associated with observable microglial activation. The second phase occurs between 1 and 3 days, is responsible for clearance of compact amyloid deposits and is associated with microglial activation. The results are discussed in the c ontext of other studies identifying coincident microglial activation and amyloid removal in APP transgenic animals. Introduction Alzheimers disease (AD) is a neurode generative disorder characterized by progressive cognitive deficits. There are several pathological characteristics to the disease process, including congophilic amyloid pla ques containing the be ta-amyloid peptide

PAGE 32

23 ( ), and intracellular incl usions of neurofibrillary tangles consisting of hyperphosphorylated tau protein. Another char acteristic of AD is the initiation and proliferation of a brain-specific inflammatory response consisting of activated microglia and astrocytes. Amyloid deposition is thought to be the key step in the pathogenesis of AD (Selkoe, 1991; Hardy and Selkoe, 2002); th is is the reason why development of potential therapies focuses on clearance of amyloid. Vaccination using 1-42 was first described by Schenk et al. (1999). This report demonstrated that active immunization using 1-42 in the PDAPP transgenic mouse dramatically reduced levels of deposits. This immuniza tion protected APP+PS1 transgenic mice (Morgan et al., 2000) and Tg CRND8 transgenic mice (Janus et al., 2000) from memory deficits. More recent studies showed that treatment with a passive immunization regimen consisting of antiantibodies resulted in a dramatic reduction in (Bard et al., 2000;DeMattos et al., 2001) a nd reversal of memory deficits (Dodart et al., 2002;Kotilinek et al., 2002) in the PDAPP mouse. In this experiment, we show that intracranially administered antiantibodies have both an early, microglia independent, and a later, possibly microglia dependent mechanism of action. levels are dramatically reduced 24 hours following administration in the absence of microglial activation. However, 72 hours after antibody administration thioflavine-S positive compact plaques are reduced concomitant with a striking activation of microglia.

PAGE 33

24 Materials and Methods. Transgenic Tg2576 APP mice (Hsiao et al., 1996) were obtained following breeding of Tg2576 APP mice with line 5.1 PS1 mice (Duff et al., 1996) which yields four different genotypes; nont ransgenic, transgenic APP, transgenic PS1 and doubly transgenic APP+PS1 mice. Animals were provided food and water ad libitum and were kept on a 12-hour light/dark cycle; they were housed in groups where possible until prior to the surgery when they were all singly hous ed until kill. We used 2 cohorts of mice in this study, the first cohort of 19 mo old A PP mice (n=16) and the second cohort of 16 mo old APP mice (n=22). Mice from the first cohort all received antiantibodies (Biosource, Camarillo CA, mouse antiIgG 1 recognizing AA 1-16). Mice from the second group were assigned to groups receiving either antiantibodies, control antibody (anti-HIV, ID6, K. Ugen, Dept. Med. Micro. USF) (N=5), or vehicle (0.02% thimerosal in PBS, SigmaAldrich, St Louis, MO) (N=5). All mice were injected in both the frontal cortex and hippocampus of the right hemisphere while the left hemisphere remained untreated as an internal control. Those mice receiving antiantibodies were assigned survival times of 4 (N=5), 24 (N=7), 72 (N=8) or 168 (N=6) hr. Mice receiving either co ntrol antibody or vehicle were examined after a 72 hr survival time. A third group of unreated 17 mo old APP mice (N=5) were killed uninjected and unmanipulated to assess differences between the right and left sides of the brain. On the day of surgery the mice were wei ghed, anesthetized with isoflurane and placed in a stereotaxic apparatus (51603 dua l manipulator lab standard, Stoelting, Wood

PAGE 34

25 Dale, IL). A midsagittal incision was made to expose the cranium and two burr holes were drilled using a dental drill over the right frontal cortex and hippocampus to the following coordinates: Cortex: AP +1.5mm L .0mm, hippocam pus: AP .7mm, L 2.5mm, all taken from bregma. A 26 gauge needle attached to a 10 l Hamilton (Reno, NV) syringe was lowered 3mm ventral to bregma and a 2 l injection was made over a 2 minute period. The incision was cleaned with sa line and closed with surgical staples. On the day of kill the mice were overdosed with 100mg/kg of pentobarbital (Nembutal sodium solution, Abbott laborator ies, North Chicago IL) and perfused intracardially with 25ml of 0.9% sodium ch loride and 50ml of freshly prepared 4% paraformaldehyde (pH=7.4). The brains were co llected and post fixed for 24 hours in 4% paraformaldehyde. The brains were then incubated for 24 hours in 10, 20 and 30% sucrose sequentially to cyroprotect them. Horizontal sections of 25 m thickness were then collected using a sliding microtome and stored at 4 o C in DPBS buffer with sodium azide to prevent microbial growth. Si x to eight sections approximately 100 m apart were selected spanning the injection site and stai ned using free-floating immunohistochemistry methods for total (rabbit antiserum primarily reacting with the N-terminal of the peptide 1:10000), CD45 (Serotec, Raleigh NC, 1:3000) and major histocompatibility complex class II (MHC-II, BD Pharmingen, Palo Alto CA, 1:3000) as previously described (Gordon et al., 2002). For immunosta ining, some sections were omitted from the primary antibody to assess non-specifi c immunohistochemical reactions. Also, immunohistochemical methods were used to stain for the injected antibody using antimouse IgG conjugated to horse radish peroxidase (Sigma-A ldrich, St Louis MO, 1:1000). Adjacent sections were mounted on slides a nd stained using 4% thioflavine-S (Sigma-

PAGE 35

26 Aldrich, St Louis MO) for 10 minutes. Sele cted sections stained for CD45 were counterstained for Congo red (Sigma-Aldrich, St Louis MO) to detect amyloid deposits on these sections. The immunohistochemical reaction product on all stained sections was measured using using a videometric V150 image analys is system (Oncor, San Diego, CA) in the injected area of cortex and hippocampus and corresponding regions on the contralateral side of the brain. Data are presented as the av erage ratio of injected side to non-injected side for thioflavine-S and CD45, while data for MHC-II are expressed as area occupied by positive stain since many values on the contralateral side were close to zero. To assess possible treatment-related differences, the measurement for either cortex or hippocampus of each subjec t were analyzed by ANOVA using StatView software version 5.0.1 (SAS Institute Inc., NC) followed by Fischers LSD means comparisons. Results Immunohistochemistry against mouse IgG wa s performed to trace the diffusion of anti-A antibodies after injection into the hilus of the de ntate gyrus. The injected antiantibody showed diffuse distribution throughout the entire hippocampus at 4 hours with a focal concentration in the outer molecular ar eas of the dentate and ammons' horn near the hippocampal fissure (Fig. 1A). By 24 hours, th e diffuse pattern remained broad, but the focal concentration began shifti ng towards the granule cell layers of the dentate (Fig. 1B). At 72 hours, staining for the injected antibody was lighter and became concentrated at the granule cell layer of the dent ate gyrus (Fig. 1C). Interestingly, by the one week time-

PAGE 36

27 point the injected antibody stai ning has largely cleared with some residual staining in the outer molecular layer of the ventral (ventricul ar) blade of the dentate gyrus and the glial limitans. A similar time course of staining wa s seen in the frontal cortex (data not shown). immunohistochemistry in APP transg enic mice resembled that reported by others and ourselves earlier (Hsiao et al., 1996; Gordon et al., 2002). In both cortex (Fig. 2A) and hippocampus (Fig. 2C) there were a fe w intensely stained de posits and a number of smaller, less intensely stained deposits. In prior work, we found the intensely stained A deposits were usually also stained with thioflavine-S or Congo red (Holcomb et al., 1998; Gordon et al., 2001) indicating they we re analogous to compact deposits containing fibrillar amyloid, while the less intense depos its were analogous to diffuse, nonfibrillar deposits commonly observed in AD tissue. Wh ile the deposits were fairly uniformly distributed within the corte x, in the hippocampus they were concentrated in the outer molecular layers of the dentate gyrus and ammons horn (Fig. 2C). The subiculum also appeared more rich in deposits than other areas. The injection of antiantibody into brain did not re sult in a rapid loss of signal in postmortem immunohistochemical reactions since we did not observe a change in staining 4 hours post-injection in either co rtex (Fig 3A) or hippocampus (Fig 3B). However, staining was reduced at the injection sites in frontal cortex and hippocampus 24 hours after administration of antiantibody (Fig 2B and D respectively) and remained reduced to r oughly the same extent through the one week time-point (Fig 3). The reduction in the frontal cortex was over 60% as compared to both the 4 hour time-points and the two control gr oups of vehicle and an ti-HIV antibody (Fig

PAGE 37

28 3A, P<0.001). The reduction in the hippocampus was over 50% as compared to the 4 hour time-points and the control groups (Fig 3B, P<0.005). An interesting phenomena was that the rati o of A staining on the right to left sides in untreated mice was greater than 1, indicating more deposition on the right side than the left (Figure 3). It appears that this pattern of deposition is a consistent property of the APP mice. The A distributi on seen in the mice administered control injections at 3 days and antiantibody at the 4 hour tim e point is the typical distribution found in APP transgenic mice of this age. As expected, the number of deposits staine d with thioflavine-S were considerably fewer than those stained by immunohistochemistry. Nonetheless, the regional distribution of these deposits roughly paralleled that of positive deposits in the cortex and hippocampus (Fig. 4A and C). In contrast to the thioflavine-S positive staining at the injection site was not reduced unt il 72 hours after administration of antiantibody (Fig 4B and D) and remained reduced at th e one week time point (Fig. 5). The reduction in frontal cortex was over 80% compared to the 4 and 24 hour time points as well as the control groups (Fig 5A, P<0.001). The reduction in hippocampus was over 60% compared to both the 4 and 24 hour time points and the control groups (Fig 5B, P<0.005). In untreated mice, activated microglia stained with CD45 or MHC-II antibodies are found only in the immediate periphery of co mpacted plaques. In the injected control groups, some microglial activation was detect ed at the 72 hour survival time by CD45 antibodies and this was restricted primarily to the injection site (arrows, Fig 6A, C; quantified in Fig 7). It should be noted that very little st aining for CD45 was detected on the uninjected side of the br ain, leading to inflated R/L ratios with relatively small

PAGE 38

29 increases in staining. MHC-II had a lower overa ll level of expression than that of CD45, and was largely unaffected in mice administer ed control injections (Fig 8A and C; quantified in Fig 9). In contrast, 72 hours after the injection of anti-A antibodies, the microglial activation detected with CD45 antibodies was more widespread, detected not only at the injection site but also away from the injec tion site in the frontal cortex (Fig 6B) and throughout the dentate gyrus, with a concentrat ion within the granule cell layer at the 72 h time point (Fig 6D). MHC-II staining reve aled a similar pattern, although not as extensive as that found with CD45 staining (Fig 8B, C). Quantification of these results indicated that the injection of antiantibodies significantly increased expression of the microglial marker CD45 only at the 72 hour time point as compared to all other time point s and control groups in both cortex (Fig 7A; P < 0.005) and hippocampus (Fig 7B; P < 0.005). Also, the injection of antiantibodies increased the expression of the microglial marker MHC-II at the 72 hour timepoint as compared to all other time points a nd control groups in both cortex (Fig 9A; P < 0.01) and hippocampus (Fig 9B; P 0.005). The expression of CD45 and MHC-II in the frontal cortex at the antiinjection site increased more than 8-fold over that of all other time points including one week and both of the control groups The expression of CD45 in the hippocampus at the antiinjection site increased more than 2-fold while the increase in expression of MHC-II is over 8 fold. As in our prior work, there is considerable variability am ong samples with both microglial markers, however, all antiinjected animals were higher than the means for the control groups.

PAGE 39

30 There are few remaining amyloid deposits near the injection sites in the antiantibody injected mice at 72 hours (Fig 4 and 5) These residual deposits are relatively faint when stained with Congo red and ar e can be found contacted by rounded, CD45 positive microglial cells (Fig 10A). In contra st, the more abundant amyloid deposits on the contralateral side and in the control animals are contacted by microglia with long processes which are stained for CD45 while the cell body only stai ns faintly for this marker of microglial activation (Fig 10B). Discussion We report here that intrac ranial anti-A antibody inje ctions substantially reduce load in the vicinity of th e injection in both anterior cortex and hippocampus over a 7 day time frame. By 4 hours after the injecti on there is a broad dist ribution of injected antibody filling a volume of roughly 0.5 mm 3 as estimated from anti-IgG immunohistochemistry. In addition to the broad pattern of diffusion, the antibody is concentrated in the outer molecular layers of Ammon's horn and the dentate gyrus, a zone which largely overlaps with the distribution of A staining in transgenic mice of this age (see fig 2C). Thus, it appears the injected anti body is binding to in si tu A at this early time point, but is also spread throughout the hippocampus. By 24 hours there is a reduction in the A immunostaining in the vicinity of the antibody injection in both cortex and hippocampus. This reduction in load is unlikely to be an artifact caused by the injected antibody masking the epit ope of the primary antibody used for immunohistochemistry because the reduced load was not detected 4 hours after administration, and by 24 hours the injected IgG a ppears to be concentrated closer to the

PAGE 40

31 granule cell region than the outer molecular layer in the hippocampus. Furthermore, the stoichiometry of injected antibody (13 pmol) to in deposits (estimated at 250 pmol in 0.5 mg, Chapman et al., 1999) is likely too low to interfere subs tantially with the histochemical reaction. This ear ly reduction in A load occu rs in the absence of the expression of microglial activation marker s CD45 and MHC-II. Although this does not preclude some rapid response of the microglia, it does suggest th at the role of microglia is qualitatively different at this early post-survival time than when markers of activation are being expressed. Between 24 and 72 hours after injection of anti-A antibodies, there were parallel reductions in fibrillar amyloid deposits de tected by thioflavine-S and increases in microglial activation, evaluated by CD45 and MHC-II staining. Although the control injections of anti-HIV antibody and vehicle caused some elevation of the CD45 marker, the activation was restricted to the immediat e vicinity of the injection site and likely caused by mechanical injury associated with the needle insertion and fluid compression of the tissue. Occasionally, in the anti-A antibody injected mice, some remaining wisps of amyloid could be found in the vicinity of the antibody injection at 72 hours and these were in contact with rounded CD45 i mmunopositive cells suggestive of phagocytic microglia/macrophages. Also at the 72 hour time point, there is a concentration of staining for both the injected an tibody and the microglia near the granule cell layer of the dentate gyrus. The temporal association of fibrillar amyloid loss with microglial activation suggests some causal role for mi croglial activation in this process. One possibility is that between 1 and 3 days activated microglia near the deposits in the outer molecular layer phagocytose opsonized amyloid via Fc receptor or complement mediated

PAGE 41

32 mechanisms and migrate towards the granul e cell layer. CD45 positive microglia can be readily detected in the outer molecular layer near the fissure at 3 days, although they are most heavily concentrated near the granule cell layer at this time point (Fig 6D). A second option is that after di ssolution of the A deposits, the antibodies diffuse to the granule cell region independent of the microglia. possibly, th e fibrillar deposits simply require more time to dissolve than the more diffuse material. More detailed time course studies of the period between 1 and 3 days coupled with immunoelectron microscopy will likely be required to resolve between these options. Remarkably, the microglial activation is terminated rapidly, and retu rns to normal levels by the 1 week time point in parallel with a significant reduction in staining for the injected IgG and A. An accumulating body of evidence finds an association between microglial activation and amyloid reductions in transgen ic mouse models of amyloid deposition. Schenk et al (1999) noted in the first study ev aluating A vaccines th at the clearance of amyloid was associated with enhanced microg lial activity around the remaining deposits. Wilcock et al (2001) largely confirmed this observation in a different transgenic model. Nakagawa et al (2000) unexpectedly found that fluid pe rcussion injury activates microglia and results in reduced amyloid deposit ion as mice grow older. Lim et al (2001) noted that transgenic mice treated with curcumin had a reduced amyloid load, but an increase in the activation stat e of microglia surrounding plaque s. Similarly, Jantzen et al (2002) found a reduced amyloid load in tran sgenic mice treated with a nitro-NSAID, NCX-2216, which was also associated with in creased microglial act ivation. Wyss-Coray (2001) found that crossing APP transgenic mi ce with mice over-expressing TGFled to increased microglial activation and reduced amyloid loads. Conversely, these same

PAGE 42

33 authors (Wyss-Coray et al 2002) found that blocking complement activation with sCRRY overexpression diminished the microglia l reaction in APP transgenic mice, and led to elevated amyloid loads. DiCarlo et al (2001) attempted to directly activate microglia by injecting LPS and found this wa s associated with clearance of A in the vicinity of the injection. However, note that Qiao et al, (2001) injected LPS chronically into young transgenic mice prior to normal amyloid deposition and found it could stimulate A deposition. It is also the case that careful serial s ection electron microscopy failed to detect internalized amyloid in microglia associat ed with amyloid deposits in untreated APP23 transgenic mice (Stalder et al, 2001), although mice treated to provoke microglial activation have yet to be examine d. Nonetheless, there is a growing literature associating the activatio n of microglia with a reduction in A deposition in the transgenic mouse models. A number of studies have demonstrated that cultured microglial cells are capable of internalizing 1-42 aggregates (Paresce et al., 1996;Webster et al., 2001). A can also be cleared from unfixed brain sec tions by anti-A antibodies in a microglia dependent manner (Bard et al., 2000). Direct imaging of amyloid deposits in vivo by multiphoton microscopy has show n clearance of plaque follo wing application of an antiantibody in association with an upregulati on of activated microg lia (Bacskai et al., 2001). Suggested alternative mechanisms to microglial phagocytosis include a physical interaction between antibody and resulting in disaggrega tion of deposits, which was demonstrated in vitro using monoclonal anti-A anti bodies (Solomon, 2001). Consistent with this idea, Backsai et al ( 2002) recently demonstrated that F (ab)2 fragments, prepared from an anti-A antibody reduce amyloid depos its as effectively as the intact antibody

PAGE 43

34 when applied topically to the cortex of transgenic mice through a craniotomy. Although there is no measurement of microglial activation in this study, it is plausible this occurred in the absence of mi croglial involvement. This anti body mediated dissolution hypothesis is consistent with the early phase of A reduction described here, and may still be found responsible for the second phase of fibrillar deposit reduction.. A major unresolved question is how this antibody mediated cl earance of A might apply to the human condition. Alzheimer's disease has increasingly been argued to involve inflammation as a component of its pathogenesis (McGeer and McGeer, 2001). The early stages of the A vaccine trials resulted in a small fraction of patients developing adverse reactions c onsistent with inflammation of the central nervous system, presumably including microglial activation (Schenk et al, 2002; Hock et al., 2002). Although adverse reactions to immunotherapy have been ra re in the transgenic models (Pfeifer et al, 2002), it remains the best experimental system in which to understand the different components of the immune reac tions to vaccines and to identify those components causing adverse outcomes. Cert ainly identification of immunotherapies which avoid the problem of deleterious CNS inflammation will be necessary if this treatment approach is to ever find use in th e clinic. Better understanding the mechanisms of antibody-mediated clearance of A in th e transgenic models of amyloid deposition should benefit this effort.

PAGE 44

35 Figure 1: Time course of injected antiantibody distribution in the hippocampus from 4 hours to seven days. Immunohistochemical staining for the injected antibody in the hippocampus at 4 hours (A), 24 hours (B), 72 hours (C) and 168 hours (D). Orientation and locations of hippocampal subregions as in figure 2 D. Magnification = 40X. Scale bar = 120 m.

PAGE 45

36

PAGE 46

37 Figure 2: Reduction in immunohistochemistry one day after antiantibody injections. Immunohistochemi cal staining is shown for in the frontal cortex (A and B) and hippocampus (C and D). A and C are from an animal injected with control antibody while B and D received the antiantibody. Magnification = 40X. Scale bar = 120 m. Panel B: FCX: frontal corte x, STR: striatum. Panel D: CA1: cornu ammonis 1, CA3: cornu ammonis 3, DG: dentate gyrus

PAGE 47

38

PAGE 48

39 Figure 3: Quantification of reduced load after antiantibody injections. Data are expressed as the ratio of A staining in the injected hemisphere: control hemisphere. The three bars on the left indicate the load in the untreated group (none) and the vehicle (VEH) and anti-HIV antibody (Cont-Ab) groups at 72 hr. The line shows the ratio of immunohistochemical staining at 4, 24, 72 and 168 hr survival times. Reduced load was observed in the frontal co rtex (panel A) and hippocam pus (panel B) at 24, 72 and 168 hours compared with 4 hours and both control groups (** indicates P<0.005).

PAGE 49

40

PAGE 50

41 Figure 4: Reduction in thioflavineS staining three days after antiantibody injections. Thioflavine-S staining is shown in frontal cortex (A and B) and hippocampus (C and D). A and C received control antibody while B and D received antiantibody. Magnification = 40X. Scale bar = 120 m. Orientation and locations of major subregions as in figure 2B and 2D.

PAGE 51

42

PAGE 52

43 Figure 5: Antiantibody injections results in a reduction of thioflavine-S positive plaques. Data are expressed as ratio of thio flavine-S staining in the injected hemisphere: control hemisphere. The three bars show th e thioflavine-S positive staining in the untreated group (none) and the vehicle (VEH ) and anti-HIV anti body (Cont-Ab) groups at 72hr. The line shows the ratio of thio flavine-S staining at 4, 24, 72 and 168 hour survival times. Reduced thioflavine-S staining was observed in th e frontal cortex (panel A) and hippocampus (panel B) at 72 and 168 hours compared with 4 and 24 hours and both control groups (** indicates P<0.005).

PAGE 53

44

PAGE 54

45 Figure 6: CD45 immunohistochemistry is increased three days following antiantibody injections. CD45 immunohistochemistry is shown in frontal cortex (A and B) and hippocampus (C and D). A and C receive d control antibody while B and D received antiantibody. Magnification = 40X. Scale bar = 120 m. Arrows indicat e the site of injection identified from the needle tract.

PAGE 55

46

PAGE 56

47 Figure 7: Antiantibody injections results in an increased CD45 immunohistochemistry three days following inje ction. Data are expre ssed as the ratio of CD45 staining in the injected hemisphere: cont rol hemisphere. The three bars indicate the CD45 expression in the untreated group (none ) and the vehicle (V EH) and anti-HIV antibody (Cont-Ab) groups at 72hr. The line sh ows the ratio of CD45 staining at 4, 24, 72 and 168 hour survival times. Increased CD45 st aining was observed in the frontal cortex (panel A) and hippocampus (panel B) at 72 hours compared with 4, 24 and 168 hours and both control groups (** indicates P<0.005).

PAGE 57

48

PAGE 58

49 Figure 8: MHC-II immunhistochemistry is increased three days following antiantibody injections. MHC-II immunohistochemistry is shown in frontal cortex (A and B) and hippocampus (C and D). A and C receive d control antibody, B and D received antiantibody. Magnification = 40X. Scale bar = 120 m.

PAGE 59

50

PAGE 60

51 Figure 9: Antiantibody injections results in an increase in MHC-II immunohistochemistry three days following inje ction. Data are expressed as percent area occupied by MHC-II positive staining in the in jected hemisphere. The three bars indicate the MHC-II expression in the untreated (none ) group and the vehicle (VEH) and antiHIV antibody (Cont-Ab) groups at 72hr. Th e line shows the amount of MHC-II staining at 4, 24, 72 and 168 hour survival times. Incr eased MHC-II staining was observed in the frontal cortex (panel A) a nd hippocampus (panel B) at 72 hours compared with 4, 24 and 168 hours and both control groups (** indicates P<0.01).

PAGE 61

52

PAGE 62

53 Figure 10: Antiantibody injections resu lts in rounded microglia in association with remaining Congophilic amyloid deposits three days afte r injection. CD45 immunostaining counterstained with Congo red is shown in the hippocampus at the 72 hour time-point. Panel A shows a typical intensely stained Congophilic deposit surrounded by CD45 immunostained microglial pr ocesses, with faintly stained somata (arrow). Panel B shows a faintly stained Congophilic deposit in the antiantibody injected hippocampus. Note the two rounded intensely CD45 positive cells in contact with the faintly stained deposit (arrow). Magnification = 600X. Scale bar = 8.33 m

PAGE 63

54

PAGE 64

55 References Bacskai BJ, Kajdasz ST, McLellan ME, Game s D, Seubert P, Schenk D, and Hyman BT. 2002. Non-Fc-mediated mechanisms are i nvolved in clearance of amyloidin vivo by immunotherapy. J Neurosci 22, 78737878. Bacskai BJ, Kajdasz ST, Christie RH, Carter C, Games D, Seubert P, Schenk D, and Hyman BT. 2001. Imaging of amyloid-beta de posits in brains of living mice permits direct observation of clearance of plaque s with immunotherapy. Na ture Medicine 7, 369372. Bard F, Cannon C, Barbour R, Burke RL, Games D, Grajeda H, Guido T, Hu K, Huang J, Johnson-Wood K, Khan K, Kholodenko D, Lee M, Lieberberg I, Motter R, Nguyen M, Soriano F, Vasquez N, Weiss K, Welch B, Seubert P, Schenk D, and Yednock T. 2000. Peripherally administered anti bodies against amyloid -peptid e enter the central nervous system and reduce pathology in a mouse model of Alzheimer's disease. Nature Medicine 6, 916-919. Chapman PF, White GL, Jones MW, Cooper-Bla cketer D, Marshall VJ, Irizarry M, Younkin L, Good MA, Bliss TVP, Hyman BT, Younkin SG, and Hsiao KK. 1999. Impaired synaptic plasticity and learning in aged amyloid precursor protein transgenic mice. Nature Neuroscience 2: 271-276. DeMattos RB, Bales KR, Cummins DJ, Doda rt JC, Paul SM, and Holtzman DM. 2001. Peripheral anti-A antibody alters CNS and pl asma A clearance and decreases brain A burden in a mouse model of Alzehim er's disease. PNAS 98: 8850-8855. DiCarlo G, Wilcock D, Henderson D, Gor don M, and Morgan D. 2001. Intrahippocampal LPS injections reduce A load in APP+PS1 transgenic mice. Neurobiology of Aging 22, 1007-1012. Dodart JC, Bales KR, Gannon KS, Greene SJ, DeMattos RB, Mathis C, DeLong CA, Wu S, Wu X, Holtzman DM, Paul SM. 2002. Immuni zation reverses memory deficits without reducing brain Abeta burden in Alzheimer' s disease model. Nat Neurosci 5: 452-457. Duff K, Eckman C, Zehr C, Yu X, Prada CM Perez-tur J, Hutton M, Buee L, Harigaya Y, Yager D, Morgan D, Gordon MN, Holcomb L, Refolo L, Zenk B, Hardy J, Younkin S. 1996 Increased amyloid-beta42(43) in brains of mice expressing mutant presenilin 1. Nature 383: 710-713. Gordon MN, Holcomb LA, Jantzen PT, DiCarl o G, Wilcock D, Boyett KW, Connor K, Melachrino J, O'Callaghan JP Morgan D. 2002 Time course of the development of Alzheimer-like pathology in the doubly tr ansgenic PS1+APP mouse. Exp Neurol 173: 183-195. Gordon MN, King DL, Diamond DM, Jantzen PT, Boyett KV, Hope CE, Hatcher JM, DiCarlo G, Gottschall WP, Morgan D, Arendash GW. 2001. Correlation between

PAGE 65

56 cognitive deficits and Abeta deposits in tr ansgenic APP+PS1 mice. Neurobiol Aging 22: 377-385. Hardy J, Selkoe DJ. 2002. The amyloid hypothesi s of Alzheimer's disease: progress and problems on the road to ther apeutics. Science 297: 353-356. Hauss-Wegrzyniak B, Vraniak PD, Wenk GL. 2000. LPS-induced neuroinflammatory effects do not recover with time. Neuroreport 11: 1759-1763. Holcomb L, Gordon MN, McGowan E, Yu X, Be nkovic S, Jantzen P, Wright K, Saad I, Mueller R, Morgan D, Sanders S, Zehr C, O'Campo K, Hardy J, Prada CM, Eckman C, Younkin S, Hsiao K, Duff K. 1998. Accelerated Alzheimer-type phenotype in transgenic mice carrying both mutant amyloid precursor protein and presenilin 1 transgenes. Nat Med 4: 97-100. Hsiao K, Chapman P, Nilsen S, Eckman C, Harigaya Y, Younkin S, Yang F, Cole G. 1996 Correlative memory deficits, Abeta elevat ion, and amyloid plaques in transgenic mice. Science 274: 99-102. Janus C, Pearson J, McLaurin J, Mathews PM, Jiang Y, Schmidt SD, Chishti MA, Horne P, Heslin D, French J, Mount HT, Nixon RA, Mercken M, Bergeron C, Fraser PE, George-Hyslop P, Westaway D. 2000. A beta peptide immunization reduces behavioural impairment and plaques in a model of Alzheimer's disease. Nature 408: 979-982. Kotilinek LA, Bacskai B, Westerman M, Kawarabayashi T, Younkin L, Hyman BT, Younkin S, Ashe KH. 2002. Reversible memory loss in a mouse tr ansgenic model of Alzheimer's disease. J Neurosci 22: 6331-6335. Lue LF, Kuo YM, Roher AE, Brachova L, Shen Y, Sue L, Beach T, Kurth JH, Rydel RE, Rogers J. 1999. Soluble amyloid beta peptide concentration as a predictor of synaptic change in Alzheimer's disease. Am J Pathol 155: 853-862. Morgan D, Diamond DM, Gottschall PE, Ugen KE, Dickey C, Hardy J, Duff K, Jantzen P, DiCarlo G, Wilcock D, Connor K, Hatcher J, Hope C, Gordon M, Arendash GW. 2000. A beta peptide vaccination prevents memory loss in an animal model of Alzheimer's disease. Nature 408: 982-985. Paresce DM, Ghosh RN, Maxfield FR. 1996. Microglial cells internalize aggregates of the Alzheimer's disease amyloid beta-protein via a scavenger receptor. Neuron 17: 553565. 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, Games D, Seubert P. 1999. Immunization with amyloid-beta attenuates Alzheimerdisease-like pathology in the PDAPP mouse. Nature 400: 173-177.

PAGE 66

57 Selkoe DJ. 1991. The molecular pathology of Alzheimer's disease. Neuron 6: 487-498. Solomon B. 2001. Immunotherapeutic strate gies for prevention and treatment of Alzheimer's disease. DN A Cell Biol 20: 697-703. Webster SD, Galvan MD, Ferran E, Garz on-Rodriguez W, Glabe CG, Tenner AJ. 2001. Antibody-mediated phagocytosis of the am yloid beta-peptide in microglia is differentially modulated by C1q. J Immunol 166: 7496-7503. Wilcock DM, Gordon MN, Ugen KE, Gottschall PE, DiCarlo G, Dickey C, Boyett KW, Jantzen PT, Connor KE, Melachrino J, Har dy J, Morgan D. 2001. Number of Abeta inoculations in APP+PS1 transgenic mi ce influences antibody titers, microglial activation, and congophilic plaque levels. DNA Cell Biol 20: 731-736.

PAGE 67

58 PAPER 2: MICROGLIAL ACTIVATION FACILITATES A PLAQUE REMOVAL FOLLOWING INTRA CRANIAL ANTI-A ANTIBODY ADMINISTRATION Donna M.Wilcock 1 Sanjay K. Munireddy 1 Arnon Rosenthal 3 Kenneth E. Ugen 2 Marcia N. Gordon 1 Dave Morgan 1 *. Departments of 1 Pharmacology and 2 Medical Microbiology and Immunology, Alzheimers Research Laboratory, University of South Florida, Tampa, Florida, 33612. 3 Rinat Neuroscience Corp. 3155 Porter Drive, Palo Alto, California, 94304. This work was published in Neurobiology of Disease 2004 Feb 15(1): 11-20. ACKNOWLEDGEMENTS: This work was supp orted by National Institutes of Aging / NIH grants AG15490 (MNG), AG18478 (DM) and AG20227 (KEU). DMW is the Benjamin Scholar in Alzheimers Disease Research.

PAGE 68

59 Abstract The mechanisms by which anti-A antibodies clear amyloid plaques in A depositing transgenic mice are unclear. In the current study we demonstrate that inhibition of anti-A antibody-induced microglial activ ation with anti-inflammatory drugs, such as dexamethasone, inhibits removal of fibrillar amyloid deposits. We also show that anti-A F(ab) 2 fragments fail to activate microg lia and are less efficient in removing fibrillar amyloid than the co rresponding complete IgG. Diffuse A deposits are cleared by antibodies under all circumstan ces. These data suggest that microglial activation is necessary for efficient re moval of compact amyloid deposits with immunotherapy. Inhibition of this activation may result in an impaired clinical response to vaccination against A Introduction Alzheimers disease (AD) is characterized clinically by progressive cognitive decline and characterized pat hologically by amyloid plaques, neurofibrillary tangles and neuron loss (Hardy and Selkoe, 2002). Another pathological event in AD is an inflammatory response which involves the activ ation and proliferati on of microglia and astrocytes (Akiyama et al, 2000). The amyloid hypothesis has targeted the A peptide as the primary focus for therapeutic interv entions in AD (Hardy and Selkoe, 2002). Amyloid plaques consist of amyloidprotein fibrils which are positively stained by Congo red and thioflavine-S. In addition, diffuse amyloid deposits can be identified using immunohistochemistry.

PAGE 69

60 Vaccination using A 1-42 was first described by Schenk et al (Schenk et al, 1999). That report showed th at immunization with A 1-42 in the PDAPP transgenic mouse dramatically reduced A deposit accumulation, both diffuse and compact. The vaccination was later shown to prevent cogni tive decline in APP+PS1 (Morgan et al, 2000) and TgCRND8 (Janus et al, 2000) tran sgenic mice. Passive immunization with anti-A antibodies was also demonstrated to ha ve benefit pathologically (Bard et al, 2000) and cognitively (Dodart et al, 2002 and Kotilinek et al, 2002). The A vaccine advanced quickly to human clin ical trials where, in Phase II, several patients developed cerebral inflammation, leading to a halt in further inoculations (Schenk 2002). The exact mechanism by which immunotherapy reduces A deposition remains unknown; suggested mechanisms include Fc receptor mediated phagocytosis via microglia (Schenk et al, 1999, Wilcock et al 2001 and Wilcock et al, 2003), dissolution of amyloid fibrils (Solomon et al, 1997 and Frenkel et al, 1999) and sequestration of circulating A resulting in an increased net efflux of A from brain and plasma (DeMattos et al, 2001). These competing hypotheses have led to disputes regarding the accessibility of circulating antibodies to the CNS, the role of systemic A content in this process and the degree of requirement for specific A epitopes to be targeted by the antibodies (Bard et al, 2003 and Holtzman et al, 2002). Moreover, in AD patients the blood-brain barrier is variably leaky (Hock et al, 2002) Bacskai et al (2001) were the first to demonstrate antiA antibody removal of amyloid deposits follo wing direct application into the brain, therefore bypassing the blood-brai n barrier. To identify the pote ntial role of microglia in

PAGE 70

61 antibody-mediated removal of A deposits, we have opted to avoid some of the complications regarding brain penetration and apply antibodies dire ctly to the CNS by intracranial injections. We have recently reported that following intracranial anti-A antibody administration there is a biphasic clearance of A deposits (Wilcock et al, 2003). The first is a rapid removal of diffuse A deposits occurring between 4 and 24 hours after injection. The second is the removal of comp act, thioflavine-S positive amyloid deposits between 24 and 72 hours following injection. Th is removal of fibrillar deposits is associated with a transient activation of micr oglia, detectable at 72 hours, but not 7 days after the injection. Remarkably, by 7 days both diffuse and compacted A deposits are largely cleared, the microglial reaction has resolved, and the injected anti-A antibody is almost completely removed. In the current study we further investig ate the relationship between microglial activation and fibrillar amyloid removal. First, we test th e capacity of several antiinflammatory agents to impair the microgl ial response and monito r their effect on A clearance. We also investigate whether an tibody fragments lacki ng the Fc domain can clear the fibrillar deposits, and monitor the effects on microglial activation. The results are consistent with the argument that micr oglial activation and Fc receptor mediated phagocytosis are important steps in the rapid clearance of A deposits by intracranially administered anti-A antibodies.

PAGE 71

62 Materials and Methods Anti-inflammatory drug study Singly transgenic APP Tg 2576 mice were obtained from our breeding program at USF started in 1996 (Holcomb et al, 1998). In the first experiment, 39 APP transgenic mice aged 16 months were assigned to one of 5 experimental groups. Four of these groups received intracranial anti-A antibody injections (44-352; Mouse monoclonal anti-human A 1-16 IgG1; Biosource, Camarillo, CA) into the frontal cortex and hippocampus at a concentration of 2 g/2 l in each region. The remaining group received intracranial anti-HIV monoclonal antibody directed against gp120 (from Ken Ugen, Univ. South Florida) into frontal cortex and hippocampus at a concentration of 2 g/2 l in each region (N=7) as a control for potential nons pecific activity associated with injecting IgG into the brain. Of the 4 groups receiving anti-A antibody one group received no further treatment (N=8), one group received tw ice daily intraperitoneal injections of dexamethasone (Sigma-Aldrich, St Louis MS) at a dose of 5mg/kg (N=9), one group received twice daily intraperitoneal injecti ons of minocycline (Sigma-Aldrich, St Louis MS) at a dose of 45mg/kg (N=7) and one group received once daily subcutaneous injections of NCX-2216 (nitro-ferulo-flur biprofen; NiCox, S.A., Sophia-Antipolis, France) at a dose of 7.5mg/kg (N=8). All trea tments following the in tracranial injection were commenced immediately following a 30 minute recovery from surgery. All mice were killed 72 hours following surgery a nd treatments were continued through the morning of kill.

PAGE 72

63 Antibody Fragment Study Twenty Tg2576 APP transgenic mice aged 19.5 months were assigned to one of four groups, all groups received intracranial injections into the frontal cortex and hippocampus. The first group received anti-A antibody (2286; Mouse monoclonal antihuman A 28-40 IgG1; Rinat Neurosciences, Palo A lto, CA) at a concentration of 2 g /2 l in each region. The second group received anti-A F(ab) 2 fragments prepared from the anti-A antibody at 2.2 g /2 l in each region. The third group received IgG directed against drosphila amnesiac protein (Rinat neur osciences, Palo Alto, CA) as a control for nonspecific aspects of intact IgG injec tion. The final group received control F(ab) 2 fragments prepared from the IgG directed ag ainst drosophila amnesiac protein to control for nonspecific effects of F(ab) 2 injection. All mice surviv ed for 72 hours after surgery. Preparation of F(ab) 2 fragments The Immunopure IgG1 Fab and F(ab) 2 preparation kit (P ierce Biotechnology, Rockford, IL) was used to prepare the F(ab) 2 fragments from the anti-A IgG and the control IgG against drosophila protein. The instructions provided with the kit were followed (http://www.piercenet.com/files/0465 jm5.pdf). Briefly, 0.5ml of 1mg/ml IgG was added to 0.5ml mouse IgG1 mild elution bu ffer. This was applied to an equilibrated immobilized ficin column, allowed to enter the column and digested at 37 o C for 20 hours. A 4ml elution was obtained and applied to an equilibrated immobilized protein A column for separation of the F(ab) 2 from Fc fragments and undigested IgG. Four 1ml fractions of product were obtained. As determined by r unning a gel electrophoresis only the 2 nd and

PAGE 73

64 3 rd elutions were found to contain F(ab) 2 fragments and appeared of similar intensities on the gel. The two elutions containing F(ab) 2 fragments were pooled and concentrated using centricon centrifugal filter devices (Millpore Corp. Bedford, MA) to a volume of approximately 200 l. Preliminary experime nts found that injections of the F(ab) 2 fractions concentrated directly from the colu mn caused seizures when injected into some mice. Thus the initial concentrate was diluted in 4 ml of fresh PBS and reconcentrated to dilute residual proprietary elution buffe r components which may cause seizures. No seizures or neurotoxicity were found in the mice included here. The concentrated product was run on an SDS-PAGE. A Bradford a ssay was also performed to establish concentrations of the F(ab) 2 fragments using Bradford prot ein assay reagent concentrate (Bio-Rad, Hercules, CA). Surgical procedure On the day of surgery the mice were wei ghed, anesthetized with isoflurane and placed in a stereotaxic apparatus (51603 dua l manipulator lab standard, Stoelting, Wood Dale, IL). A midsagittal incision was made to expose the cranium and two burr holes were drilled using a dental drill over the right frontal cortex and hippocampus to the following coordinates: Cortex: AP +1.5mm L .0mm, hippocam pus: AP .7mm, L 2.5mm, all taken from bregma. A 26 gauge needle attached to a 10 l Hamilton (Reno, NV) syringe was lowered 3mm ventral to bregma and a 2 l injection was made over a 2 minute period. The incision was cleaned with sa line and closed with surgical staples. Tissue Preparation On the day of kill mice were weighe d, overdosed with 100mg/kg pentobarbital (Nembutal sodium solution, Abbott laboratorie s, North Chicago IL) and intracardially

PAGE 74

65 perfused with 25ml 0.9% sodium chloride followed by 50ml freshly prepared 4% paraformaldehyde (pH=7.4). Brains were rapidly removed and immersion fixed for 24 hours in freshly prepared 4% paraformaldehyde The brains were then incubated for 24 hours in 10, 20 and 30% sucrose sequentially to cyroprotect them. Hori zontal sections of 25 m thickness were then collected using a sliding microtome and stored at 4 o C in DPBS buffer with sodium azide to prevent microbial growth. Immunohistochemical methods Six to eight sections approximately 100 m apart were selected spanning the injection site and stained using free-floating immunohistochemistry methods for total (rabbit antiserum primarily reacti ng with the N-terminal of the peptide 1:10000) and CD45 (Serotec, Raleigh NC, 1:3000) as prev iously described (G ordon et al, 2002). For immunostaining, some sections were omitted from the primary antibody to assess nonspecific immunohistochemical r eactions. Adjacent sections were mounted on slides and stained using 4% thioflavine-S (Sigma-Aldri ch, St Louis MO) for 10 minutes. It should be noted that there are a limite d number of sections that in clude the injection volume. We have opted to measure a few markers reliab ly rather than a larg er number of markers with fewer sections each. Data analysis The immunohistochemical reaction product on all stained sections was measured using using a videometric V150 image analys is system (Oncor, San Diego, CA) in the injected area of cortex and hippocampus and corresponding regions on the contralateral side of the brain. Data are presented as the ra tio of injected side to non-injected side for thioflavine-S and CD45. Normalizing each injection site to the corresponding

PAGE 75

66 contralateral site diminishes the influence of inter-animal variability and permits reliable measurements of drug effects with a smalle r number of mice. Importantly, there is no injected antibody detectable in the contralateral si de. To assess possible treatment-related differences, the ratio values for each trea tment group were analyzed by ANOVA using StatView software version 5.0.1 (SAS Inst itute Inc., NC) followed by Fischers LSD means comparisons. Results Following intracranial injection of anti-A antibody 44-352 into the hippocampus and frontal cortex there was a significant activation of microglia detectable by CD45 immunohistochemistry. In the hippocampus, the most intense area of activation appeared in the granule cell layer of the dentate gyrus close to the site of injection within the hilus/CA4 region. However, there was a much more diffuse activation which filled the remainder of the dentate gyrus (Fig. 1A). In the frontal cortex, the activation formed a gradient surrounding the injection site wit hout a clear laminar pr ofile (not shown). Following the intracranial injection of anti-A antibody, treatment with the steroidal antiinflammatory agent dexamethasone completely inhibited the microglial activation with only several small cells faintly stained fo r CD45 in hippocampus (Fig 1B; P < 0.05 Fig 2A) and in frontal cortex (Fig 2A, P < 0.001). The staining pattern and values observed in this group matched that of th e group administered a control IgG directed against an HIV protein in both brain regions (Fig. 1E, Fig. 2A). Minocycline, a drug previously shown to inhibit microglial activation in several CNS inflammation models, appeared relativel y ineffective at inhibiting the microglial

PAGE 76

67 activation observed as a result of intracranial anti-A antibody administration. In the hippocampus the intense area of activation in the granule cell layer was still present, as was the more diffuse activation in the remai nder of the dentate gyrus (Fig 1C). In the frontal cortex, although there was a signi ficant difference between the minocyclinetreated mice and the untreated mice (P < 0.01), the microglial activation in the minocycline treated mice was still significantly greater than in the dexamethasone treated mice (p<0.05; Fig 2A). NCX-2216 combines a nitric oxide generating moiety with the typical NSAID drug flurbiprofen. In the hippocampus, NC X-2216 treatment following the intracranial injection of anti-A antibody partially inhibited the activat ion of microglia (Fig 1D). This drug did not inhibit the intense activation observed in the gra nule cell layer of the dentate gyrus but did diminish the more diffuse act ivation. The quantification from the frontal cortex found a significant i nhibition of microglial activati on (P < 0.01; Fig 2A). Thus, with respect to inhibiting microglial activ ation following anti-A antibody injection, dexamethasone was the most effective drug with NCX-2216 having a partial inhibition followed by an even weaker inhibition caused by minocycline. Total A immunohistochemistry in mice ad ministered the control antibody directed against human immunodeficiency vi rus (HIV) protein gp120 was similar to that described previously in the APP transgenic mouse (Hsiao et al, 1996 and Gordon et al, 2002). The ratio of A in the right: left sides was also the same as that observed previously in unmanipulated APP transg enic mice (Wilcock et al, 2003). The A immunohistochemistry showed a few large, intensely stained deposits, which are normally also stained by Congo red or thioflav ine-S, indicating fibrillar compact amyloid

PAGE 77

68 deposits. There were also a large number of smaller, less intensely stained deposits analogous to diffuse amyloid deposits obser ved in human AD brai n tissue. In the hippocampus, A deposition was localized primarily to the molecular layers of the dentate gyrus and Ammons horn adjacent to the hippocampal fissure, as well as a large concentration in the subiculum (Fig. 1J). Anti-A antibody administration into frontal cortex and hippocampus resulted in a reduction of total A immunohistochemistry 72 hours following injection (Fig 1F). This reduction was approximately 80% in frontal cortex and 65% in hippocampus compared to APP transgenic mice administ ered HIV antibody (Fig. 2B). The antiinflammatory agents dexamethasone, NCX-2216 and minocycline had no effect on the removal of this largely diffuse A staining (Fig. 1G, H and I and Fig 2B). Thioflavine-S staining showed a different response to anti-inflammatory drug treatment than A immunostaining. Although fewer in number, the subregional distribution of thioflavin e-S positive plaques matched that observed with A immunohistochemistry in APP transgenic mice administered control IgG (Fig 1O). AntiA antibody injected into the frontal cortex and hippocampus result ed in a virtually complete removal of thioflavine-S positive plaques 72 hours following injection, reaching 90% in frontal cortex and 85% in the hippocampus (Fig 1K; Fig 2C). Administration of dexamethasone resu lted in complete arrest of anti-A antibody mediated clearance of thioflavine-S compact amyloid deposits in the hippocampus (Fig. 1L) or the frontal cortex, with values equiva lent to those mice given anti-HIV control antibody (Fig. 2C). In contrast, administrati on of minocycline had no detectable effects

PAGE 78

69 on anti-A antibody mediated clearance of thioflavin e-S deposits in eith er frontal cortex (Fig. 2C) or hippocampus (Fi g. 1M, Fig. 2C). NCX-2216 treatment had a partial effect on thioflavine-S reduction following intracranial anti-A antibodies. In the frontal cortex the levels of thioflavine-S were almost equivalent to those seen following anti-HIV control antibody administration, and were significan tly different from the levels observed following anti-A antibody administration alone (Fig. 2C ). The levels of thioflavine-S in the hippocampus were not significantly different from mice administered anti-A antibody or mice administered anti-HIV control antibody, and their values were in between these two groups (Fig. 2C), suggesting a partial impairment of antibody mediated clearance by this drug. A second series of studies investigated th e potential role of the Fc domain of the antibody in microglial activation and amyloid clearance. F(ab) 2 fragments prepared from anti-A monoclonal antibody 2286, and a cont rol monoclonal antibody directed against the drosophila protein amnesiac were analyzed via SDS-polyacrylamide-gel electrophoresis (PAGE). The gel showed ve ry pure product, with a single band at approximately 105kDa, the molecular weight for F(ab) 2 fragments. The intact IgG molecule produced one intense band at a pproximately 150kDa, the correct molecular weight for IgG molecules and a less intens e band at approximately 110 kDa. Following confirmation of purity via SDS-PAGE we then performed a Bradford assay to assess the recovery of F(ab) 2 in the purified fraction. Because we dissolved the anti-A F(ab) 2 fragments in a smaller volume than was used for the starting material the concentration of F(ab) 2 fragments injected intracranially was 1.2 g/ l, while the complete IgG

PAGE 79

70 concentration was 1 g/l, resulting in an excess of anti-A Fv domains in the F(ab) 2 solutions. The only antibody which activated microglia 72 hours follo wing intracranial injection into frontal cortex and hippocampus was the intact anti-A antibody. The frontal cortex shows a greater degree of activation than the hippocampus, however, in both regions the activation is significantly greater than th at in the groups receiving control anti-amnesiac protein IgG, F(ab) 2 or anti-A F(ab) 2 (Fig 3A, C and D, Fig. 4A; P < 0.01 or greater in all comparisons). The pattern of activation in the hippocampus following the anti-A antibody 2286 injection resembled that shown in Fig 1A when using the anti-A antibody 44-352. There is a very in tense area of activation in the granule cell layer of the dentate gyrus, with a much more diffuse activation filling the remainder of the dentate gyrus (F ig. 3A). Interest ingly, the anti-A F(ab) 2 fragments produced no microglial activation in either the frontal cortex and hippocampus (Fig. 3B, Fig. 4A). A immunohistochemistry in the two anti -amnesiac protein control groups shows the typical staining pattern observed in APP transgenic mice at 19.5 months (Fig. 3G and H). This pattern was qualitatively the same as observed at 16 months (Fig 1J), although quantitatively greater as the mice were 3.5 months older. Both the anti-A antibody and the anti-A F(ab) 2 groups significantly reduced total A immunohistochemistry to a similar extent 72 hours following injection in to frontal cortex and hippocampus. In the frontal cortex there was a re duction of approximately 60% (F ig. 4B). In the hippocampus the reduction was approximately 65% (Fig. 3E and F, Fig. 4B).

PAGE 80

71 Thioflavine-S staining detects only comp act fibrillar amyloid deposits. The mice receiving intracranial injections of either control anti-amnesiac pr otein IgG or control F(ab) 2 resembled the typical staining observed in the APP transgenic mouse at this age. In the hippocampus the majority of thioflavine-S positive plaques were located in the outer molecular layer of Ammons horn a nd the dentate gyrus near the hippocampal fissure (Fig. 3K and L). Anti-A antibody IgG significantly reduced thioflavine-S positive compact plaque by approximately 90% in the frontal cortex and hippocampus (Fig. 4C). There were no, or very few, re maining thioflavine-S positive deposits in the hippocampus (Fig. 3I). In contrast, the anti-A F(ab) 2 fragments did not remove compact amyloid plaques as eff ectively as the whole IgG molecule. In the frontal cortex there was no significant reduction in thioflav ine-S staining when compared to either control antibody group (Fig. 4C). In the hi ppocampus there was a significant difference between the anti-A F(ab) 2 group and the control groups (P < 0.05), however, this reduction was also significantly less than th e reduction observed with the whole IgG molecule (Fig. 3J, Fig. 4C; P < 0.02 or greater). Discussion The data presented here support the argum ent that activation of microglia in APP transgenic mice facilitates the removal of compact amyloid plaques. The first experiment, using several anti-inflammatory agents to re gulate the microglial response, showed that the extent of fibrillar amyloid removal rough ly corresponds to the extent of microglial activation 3 days after intracranially applied anti-A antibody. The second study identified that anti-A F(ab) 2 fragments were less capable of activating microglia

PAGE 81

72 (presumably because they lacked the Fc do main) and were significantly less effective than the corresponding whole IgG in removing fibrillar A despite the presence of excess anti-A Fv in the F(ab) 2 injections. Our earlier work in this system demonstr ated that intracrani al administration of anti-A antibodies into APP Tg2576 mice re sulted in diffusion of the antibody throughout most of the hippocampus by 4 hour s, however, no specifi c localization to amyloid plaques was noted. This cau sed a rapid removal of diffuse A deposits between 4 hours (when no reduction is detected) and 24 hours (when removal of diffuse deposits appeared complete; Wilcock et al, 2003). Th is removal was not associated with any apparent activation of microglia using mark ers such as MHC-II or CD45 at 24 hours, nor was there any reduction of the fibrillar amyl oid deposits measured with thioflavine-S staining. However, by 72 hours following the inj ection there was a dram atic reduction of thioflavine-S positive compact amyloid deposits associated with a florid microglial activation as detected by CD45 immunohistoc hemistry. One week following the anti-A antibody injection, the injection site remain ed devoid of most forms of amyloid, the microglial reaction had terminated and the in jected antibody had been fully cleared from the area. An issue of concern with these studies is whether the A epitope utilized for immunohistochemistry is masked by the inj ected antibody. This issue was addressed in our previous work where it was shown that a lthough there is a broa d distribution of the injected antibody 4 hours followi ng injection, no reduction in A immunohistochemistry is apparent (Wilcock et al, 2003). Also, if the reduction observed is simply an artifact of

PAGE 82

73 masking the A epitope, a reduction in thioflavineS staining would not be observed since this is a conformation dependent stain and not epit ope dependent. Further, the stoichiometry of injected antibody (13 pmol) to in deposits (estimated at 250 pmol in 0.5 mg, Chapman et al., 1999) is likely too low to interfere subs tantially with the histochemical reaction. Finally, 7 days after in jection, the injected antibody is no longer detectable, yet the A immunostaining remains absent. The present report further investigated the relationship between activation of microglia and the clearance of the fibrilla r amyloid plaques associated with anti-A antibody injections. We used several distinct pharmacological agents in an attempt to inhibit the microglia activation observed 72 hour s following intracranial injection of antiA antibodies. Dexamethasone is a gluc ocorticosteroid wh ich inhibits the cyclooxygenase and lipoxygenase inflammatory pathways as well as inducing a general state of immunosuppression. It has been show n that microglia respond differently to mineralocorticoid and glucocorticoid recep tor stimulation. Mineralocorticoid receptor activation stimulates the microglia while gl ucocorticoid receptor activation inhibits microglia (Tanaka et al, 1997). All pharmaco logical glucocorticosteroids possess some degree of mineralocorticoid act ion also. For the present study we selected dexamethasone as it has the maximum glucocorticoid receptor activity with the minimum mineralocorticoid receptor activity detect able among all available pharmacological glucocorticosteroids (Schimmer and Parker, 2001 ). It was found that dexamethasone was the most efficacious compound for the inhib ition of microglial act ivation among those used in this study. Dexamethasone administer ed immediately following intracranial antiA antibody administration completely inhib ited the microglial act ivation caused by the

PAGE 83

74 injection. Associated with this profound arre st of microglial activation was a complete inhibition of the antibody's ability to remove compact amyloid deposits detected with thioflavine-S staining, strongly suggesting a role for microglial involvement in the removal of compact amyloid deposits. The novel non-steroidal anti-inflamm atory (NSAID) NCX-2216, which is a flurbiprofen molecule conjuga ted to an antioxidant and a nitric oxide releasing group, was moderately effective at inhibiting micr oglial activation. Intere stingly, this compound has previously been shown to cause the activ ation of microglia a nd removal of amyloid from the brains of otherwise untreated doubl y transgenic APP+PS1 mice (Jantzen et al, 2002). NCX-2216 has also been shown to inhibit the microglial activation caused by intracranial infusion of lipopolysaccharide (L PS), a proinflammatory agent in young rats, but to increase microglial act ivation in old rats (Hauss-We grzyniak et al, 1999). In the present study, NCX-2216 partially reduced the activation of microglia caused by antibody injection. Associated w ith this partial inhi bition of microglial activation is also a partial impairment of the anti-A antibodys capacity to re move the compact amyloid plaques. The discrepancy between the eff ects observed in the current study and the effects previously observed by Jantzen et al (2002) may be explained by the fact that NCX-2216 essentially is three drugs. In a si tuation where there is intense, local microglial activation the anti-inflammatory prope rties of the drug appears to dominate. In a situation where there is diffuse microglial activation it appears that maybe the nitric oxide release dominates to enhance microglia l activation and aid in the clearance of A Although not quantified, the microglial re action immediately adjacent to amyloid deposits appeared more intense in NCX-2216 treated mice in regions distant from the

PAGE 84

75 injection (e.g. contralateral ante rior cortex). A similar bidirectional effect of this drug was found by Hauss-Wegrzyniak et al, 1999, w ho found reduced microglial activation in young rats, but enhanced activation in old ra ts treated with the NCX-2216 relative, nitroflurbiprofen. Minocycline is a tetracycline derivative which has been shown to have a novel action independent of its an tibiotic property. This agent has been shown to inhibit microglial activation following excitotoxicity (Tikka et al, 2001), ischemia (Yrjanheikki et al, 1998) and 6-hydroxydopamine lesions (H e et al, 2001). In the present study we demonstrate that minocycline is capable of a modest inhibition of microglial activation following antibody injection. Associated with this is no difference in compact plaque removal in either the frontal cortex or hippocampus. To further investigate whether the activation of microglia, which appears to be specific to the anti-A antibody, is due to Fc receptor activation we administered F(ab) 2 fragments intracranially into APP transgenic mice as well as the whole IgG, control IgG and F(ab) 2 fragments from the control IgG. We found that of the four experimental groups only the animals receiving anti-A IgG showed significant activation of microglia. The fact that anti-A F(ab) 2 fragments were unable to activate microglia, strongly suggests that Fc receptor activation is required for the significant activation of microglia following intracranial administration of anti-A antibodies. Associated with this inability to activ ate microglia was a significantly impaired capacity to remove fibrillar amyloid deposits. Still, at least in hippocampus, there was some residual capacity for fibrilla r amyloid removal using the F(ab) 2 fraction. However, with respect to diffuse A clearance, the F(ab) 2 are just as effectiv e as the corresponding

PAGE 85

76 intact IgG. These results s uggest that there may be an e quilibrium between the fibrillar deposits and the diffuse deposits, and the F(ab) 2 antibody fragments can whittle away at the fibrillar plaques without requiring Fc recep tor mediated phagocystosis. This is also the likely explanation why Bacskai et al (2002) found topically applied F(ab) 2 fragments equally effective to intact anti bodies in the clearance of thioflavine-S labeled material. It is plausible that had we inj ected a greater amount of F(ab) 2 fragments, or extended the post-injection interval, we also might have found complete clearance of fibrillar amyloid with the F(ab) 2 material. However, even though it may be capable of clearing fibrillar amyloid, the results presented here demonstrate F(ab) 2 fragments were much less efficient than the intact IgG molecule in me diating clearance associated with microglial activation. It is plausible that this also expl ains the recent observations from Bard et al (2003) that the ability of different monoclonal anti-A antibodies to clear brain amyloid when administered systemically was correla ted better with their capacity to bind Fc receptors than with their affinity for A The amounts of antibody entering the brain are roughly 0.1% of the injected amount per hour (Banks et al, 2002). Thus, a 500 g injection of a monoclonal antibody should result in 0.5 g entering the CNS within 60 minutes, somewhat less than the amounts inj ected directly in our system (2 g). Therefore, with systemically injected antibodies, the faci litation of amyloid removal by Fc receptor mediated phagocytosis is likely to be even greater than that observed here with intracranially ad ministered antibodies. Certainly, it will be useful to measure content by means other than histochemistry. Increasingly complex methods of fractionating homogenates are being used to identify pools linked most closely to neuropathology and cognitive disruption

PAGE 86

77 (Lue et al, 1999 and Golde et al, 2000) but consensus regarding the specific fractions corresponding to the most toxic form of amyl oid has not yet been achieved (Walsh et al, 2002 and Kayed et al, 2003). Additionally, th e relatively small portions of the brain affected by the intracranial injections and di fficulty dissecting these regions consistently limits the ability to use solution methods to evaluate A loads in our studies. As consensus emerges regarding the relationships between soluble, oligomeric and fibrillary forms of A in the solution domain and the histochemical domain, and larger portions of the brain become involved with either intraventricular or systemic antibody injections, we will investigate the effects of anti-A antibodies on these different A pools in both domains. It has been shown previously that vaccination using A 1-42 results in activation of microglia, which is associated with a reduction in A accumulation in PDAPP transgenic mice (Schenk et al, 1999) and APP+PS1 tran sgenic mice (Wilcock et al, 2001). It has also been shown that followi ng direct application of anti-A antibodies to the brains of PDAPP mice there is an activation of micr oglia and a reduction in amyloid deposits (Bacskai et al, 2001). In v itro studies using F(ab) 2 fragments demonstrated that they were unable to activate microglia despite re taining full ability to bind to A (Bard et al, 2000). These fragments also failed to remove fibrillar A in an ex vivo assay. Human postmortem microglia have been shown to phagocytose opsonized A which is inhibited by excess non-specific IgG, suggesting this phago cytosis is Fc receptor mediated (Lue et al, 2002). All of these data suggest that one likely mechanism of antibody action in removing amyloid deposits from transgenic m ouse brains is via bind ing to microglial Fc

PAGE 87

78 receptors and triggering activation of the mi croglia, possibly including phagocytosis of the opsonized amyloid. This by no means precludes other possible mechanisms, such as catalytic dissolution of amyl oid fibrils (Solomon et al, 1997), or sequestration of A in the periphery, effectively drawing A out of the brain (D eMattos et al, 2001). Recently, the first pathology report from a patient receiving the A 1-42 vaccination (AN1792) was published (Nicoll et al, 2003). Th is report showed that the patient had considerably fewer amyloid deposits than would have been predicted from other AD cases. Interestingly, it is repor ted that in those regions de void of amyloid plaques, the remaining A -immunoreactivity was associated with ac tivated microglia. This patient did develop meningoencephalitis and other symptoms of CNS inflammation, as did several others in the trial. The treatment chosen for the CNS inflammation in this case was dexamethasone. Assuming that the mechanisms of A vaccination in clearing amyloid is similar to that demonstrated in the present work, it mi ght be anticipated that dexamethasone would counter the amyloid removing effects of the vaccine. Although th e case described by Nicoll et al above did show evidence of rem oval, the patient had received 5 inoculations before developing adverse reactions and bei ng administered dexamethasone. It was also not indicated what the antibody titer was in the patient, nor how long the dexamethasone treatment had continued. The data presen ted here suggest that administration of glucocorticoids to vaccinated patients may counteract any be nefit the vaccine has with respect to amyloid clearance a nd possibly cognitive f unction. It is also conceivable that removal of only the soluble and diffuse A may provide cognitive improvement and inhibition of microglial activation by anti-i nflamnmatory drugs or administration of

PAGE 88

79 F(ab) 2 fragments may avoid some of the inflammatory adverse effects observed in the human clinical trial (Orgogozo et al, 2003) The inhibition of microglial mediated amyloid clearance may also have been a factor in the failure of the prednisone clinical trial for AD (Aisen et al, 2000). The adverse reactions in the human vaccine trial demonstrates a need to more fully investigate the mechanisms involved in beneficial and detrimental effects of immunotherapy. Encouraging data recently publis hed by Hock et al (2003) showed that the a subset of patient s administered the A vaccine remained cognitively stable for one year after treatment while the control patients declined at a normal rate, some vaccinated patients actually improved. While the immunot herapeutic approach may hold promise for the treatment of AD, it would appear ve ry important to better understand both the mechanisms of vaccine action, and how the tools to effectively modulate the immune reaction interact with these mechanisms. Th e data presented here make some headway toward determining what effects modulation of the immunotherapeutic approach mechanisms would have pathologically. Future studies will extend these investigations to a more clinically relevant, systemically ad ministered, passive im munization regimen to determine the importance of the mechan isms discussed in the current study.

PAGE 89

80 Figure 1: Anti-inflammatory drugs impaired fibrillar amyloid removal to roughly the same extent as they decreased micr oglial activation following anti-A antibody injections. Panels A-E s how CD45 immunohistochemistry in the hippocampus. Panels FJ show A immunohistochemistry in the hippocam pus. Panels K-O show thioflavine-S staining in hippocampus. Mice were injected intracranially with anti-A antibody followed by no treatment (A, F, K), dexamethasone treatment (B, G and L), minocycline treatment (C, H and M) or NCX-2216 treatment (D, I and N). Mice shown in panels E, J, and O were injected with anti-HIV anti body as a control for nonsp ecific effects of IgG injection. Magnification = 40X. Scale bar =120 m.

PAGE 90

81

PAGE 91

82 Figure 2: Quantifica tion of CD45, total A and thioflavine-S following inhibition of microglial activation by anti-inflammatory co mpounds. Panel A shows the ratio of right to left sides for CD45 immunohistochemistry. Panel B shows the ratio of right to left sides for total A immunohistochemistry. Panel C shows th e ratio of right to left sides for thioflavine-S staining. The solid bars indica te values for frontal cortex, the open bars indicate values for hippocampus. On the xaxis the type of an tibody injected (anti-A antibody; Abeta, or control an tibody; HIV) is shown. The post-injection treatment the mice received is also shown; Dex: dexameth asone Min: minocycline treatment, NCX: NCX-2216 treatment. *** indicates P<0.001, ** indicates P < 0.01, indicates P<0.05 as compared to the anti-A antibody alone.

PAGE 92

83

PAGE 93

84 Figure 3: Anti-A F(ab) 2 fragments do not activate microglia, nor do they remove compact amyloid deposits as effectively as the complete anti-A IgG. Panels A-D show CD45 immunohistochemistry in the hi ppocampus. Panels E-H show total A immunohistochemistry in the hippocampus. Panels I-L show thioflavine-S staining in the hippocampus. Mice were injected with in tact anti-A IgG (A, E and I), anti-A F( ab) 2 fragments (B, F and J), control (anti-amne siac) IgG (C, G and K), or control (antiamensiac) F(ab) 2 fragments (D, H and L). Magnif ication = 40X. Scale bar=120 m.

PAGE 94

85

PAGE 95

86 Figure 4: Quantification of CD45 and total A immunohistochemistry and thioflavine-S staining following intracranial injection of anti-A antibodies and anti-A F(ab) 2 fragments. Panel A shows the ratio of right to left sides for CD45 immunohistochemistry. Panel B shows the ratio of right to left sides for total A immunohistochemistry. Panel C shows the ratio of right to left sides for thioflavine-S staining. The solid bars indicate values for frontal cortex, the open bars indi cate values for hippocampus. On the x-axis IgG-Cont= control (anti-amnesi ac) intact IgG,; F(ab)2-Cont = Control (anti-amnesiac) F(ab) 2 fragments; IgG-Abeta = anti-A intact IgG; F(ab)2-Abeta= anti-A F(ab) 2 fragments.*** indicates P<0.001, indicates P<0.05 as compared to both control antibody groups. Lines over bars indicates P values for comp arisons between the specific pair of groups indicated.

PAGE 96

87

PAGE 97

88 References Aisen P.S., Davis K.L., Berg J.D., Schafer K., Campbell K., Thomas R.G., Weiner M.F., Farlow M.R., Sano M., Grundman M., Thal L. J., 2000. A randomized controlled trial of prednisone in Alzheimer's disease. Alzhei mer's Disease Cooperat ive Study. Neurology 54(3): 588-593. Akiyama H., Barger S., Barnum S., Bradt B., Bauer J., Cole G.M., Cooper N.R., Eikelenboom P., Fiebich B.L., Finch C.E., Frau tschy S., Griffin W.S.T., Hampel H., Hull M., Landreth G,. Lue L.F., Mrak M., Mackenzi e I.R., McGeer P.L., OBanion K., Pachter J., Pasinetti G., Plata-Salaman C., Rogers J., Rydel R., Shen Y., Streit W., Strohmeyer R., Tooyoma I., Van Muiswinkel F.L., Veerhuis R. Walker D., Webster S., Wegrzyniak B., Wenk G., Wyss-Coray T., 2000. Inflammation and Alzheimers disease. Neurobiol. Aging 21: 383-421. Bacskai B.J., Kajdasz S.T., McLellan M.E ., Games D., Seubert P., Schenk D., Hyman B.T., 2002. Non-Fc-mediated mechanisms are involved in clearance of amyloidin vivo by immunotherapy. J. Neurosci. 22: 2223-2226. Bacskai B.J., Kajdasz S.T., Christie R.H., Carter C., Games D., Seubert P., Schenk D., Hyman B.T., 2001. Imaging of amyloiddeposits in brains of living mice permits direct observation of clearance of plaques with immunotherapy. Nat. Med. 7: 369-372. Banks W.A., Terrell B., Farr S.A., Robins on S.M., Nonaka N., Morley J.E., 2002. Passage of amyloid protein antibody across the blood-brain barrier in a mouse model of Alzheimers disease. Peptides 23: 2223-2226. Bard F., Barbour R.Cannon C., Carretto R., Fox M., Games D., Guido T., Hoenow K., Hu K., Johnson-Wood K., Khan K., Kholode nko D., Lee C., Lee M., Motter R., Nguyen M., Reed A., Schenk D., Tang P., Vasquez N., Seubert P., Yednock T., 2003. Epitope and isotype specificities of anti bodies to beta -amyloid peptide for protection against Alzheimer's disease-like neuropathol ogy. Proc Natl Acad Sci USA 100: 2023-2028. Bard F., Cannon C., Barbour R., Burke R.L., Games D., Grajeda H., Guido T., Hu K., Huang J., Johnson-Wood K., Khan K., Kholode nko D., Lee M., Lieberburg I., Motter R., Nguyen M., Soriano F., Vasquez N., Weiss K., Welch B., Seubert P., Schenk D., Yednock T., 2000. Peripherally administered antibodies against amyloid beta-peptide enter the central nervous system and reduce pathology in a mouse model of Alzheimer disease. Nat. Med. 6: 916-919. Chapman P.F., White G.L., Jones M.W., Coope r-Blacketer D., Marshall V.J., Irizarry M., Younkin L., Good M.A., Bliss T.V.P., Hyman B.T., Younkin S.G., and Hsiao K.K.,

PAGE 98

89 1999. Impaired synaptic plasticity and lear ning in aged amyloi d precursor protein transgenic mice. Nature Neuroscience 2: 271-276. DeMattos R.B., Bales K.R., Cummins D.J., Dodart J.C., Paul S.M., Holtzmann D.M., 2001. Peripheral anti-A antibody alters CNS and plasma A clearance and decreases brain A burden in a mouse model of Alzheimers disease. Proc. Natl. Acad. Sci. USA. 98: 8850-8855. Dodart J.C., Bales K.R., Gannon K.S., Greene S.J., DeMattos R.B., Mathis C., DeLong C.A., Wu S., Wu X., Holtzman D.M., Paul S.M., 2002. Immunization reverses memory deficits without reducing brain Abeta burde n in Alzheimer's disease model. Nat. Neurosci. 5: 452-457. Frenkel D., Balass M., Katchalski-Katzir E ., Solomon B., 1999. High affinity binding of monoclonal antibodies to the sequential epitope EFRH of beta-amyloid peptide is essential for modulation of fibrillar aggregation. J. Neuroimmunol. 95: 136-142. Golde T.E., Eckman C.B., Younkin S.G ., 2000. Biochemical detection of Abeta isoforms: implications for pathogenesis, dia gnosis, and treatment of Alzheimer's disease. Biochim Biophys Acta 1502: 172-187. Gordon M.N., Holcomb L.A., Jantzen P.T ., DiCarlo G., Wilcock D., Boyett K.W., Connor K., Melachrino J., O'Callaghan J.P ., Morgan D., 2002. Time course of the development of Alzheimer-like pathology in the doubly transgen ic PS1+APP mouse. Exp. Neurol. 173: 183-195. Hardy J., Selkoe D.J., 2002. The amyloid hypothe sis of Alzheimers disease: Progress on the road to therapeutics. Science 297: 353-356. Hauss-Wegrzyniak B., Willard L.B., Del Soldato P., Pepeu G., Wenk G.L., 1999. Peripheral administration of novel anti-infla mmatories can attenua te the effects of chronic inflammation within th e CNS. Brain Res. 815: 36-43. He Y., Appel S., Le W., 2001. Minocycline inhi bits microglial activ ation and protects nigral cells after 6-hydroxydopamine injection into mouse striatum. Brain Res. 909: 187193. Hock C., Konietzko U., Streffer J.R., Tracy J., Signorell A., Muller-Tillmanns B., Lemke U., Henke K., Moritz E., Garcia E., Wollmer M.A., Umbricht D., de Quervain D.J., Hofmann M., Maddalena A., Papassotiropou los A., Nitsch R.M., 2003. Antibodies against beta-amyloid slow cognitive decline in Alzheimer's disease. Neuron 38: 547-554. Hock C., Konietzko U., Papassotiropoulos A., Wollmer A., Streffer J., von Rotz R.C., Davey G., Moritz E., Nitsch R.M., 2002. Gene ration of antibodies specific for betaamyloid by vaccination of patients with Alzheimer disease. Nat Med 8: 1270-1275.

PAGE 99

90 Holcomb L., Gordon M.N., McGowan E., Yu X., Benkovic S., Jantzen P., Wright K., Saad I., Mueller R., Morgan D., Sanders S., Zehr C., O'Campo K., Hardy J., Prada C.M., Eckman C., Younkin S., Hsiao K., Duff K., 1998. Accelerated Alzheimer-type phenotype in transgenic mice carrying both mutant amyloid precursor protein and presenilin 1 transgenes. Nat Med 4: 97-100. Holtzman D.M., Bales K.R., Paul S.M., De Mattos R.B., 2002. Abeta immunization and anti-Abeta antibodies: potentia l therapies for the prevention and treatment of Alzheimer's disease. Adv Drug Deliv Rev 7: 1603-1613. Hsiao K., Chapman P., Nilsen S., Eckman C., Harigaya Y., Younkin S., Yang F., Cole G., 1996. Correlative memory deficits, Ab eta elevation, and amyloid plaques in transgenic mice. Science 274: 99-102. Jantzen P.T., Connor K.E., DiCarlo G., Wenk G.L., Wallace J.L., Ro jiani A.M., Coppola D., Morgan D., Gordon M.N., 2002. Microglial activation and beta -amyloid deposit reduction caused by a nitric oxide-releasi ng nonsteroidal anti-inflammatory drug in amyloid precursor protein plus presenilin-1 transgenic mice. J. Neurosci. 22: 2246-2254. 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., West away D., 2000. A beta peptide immunization reduces behavioural impairment and plaques in a model of Alzheimer's disease. Nature 408: 979-982. Kayed R., Head E., Thompson J.L., McIntire T.M., Milton S.C., Cotman C.W., Glabe C.G., 2003. Common structure of soluble amyloid oligomers implies common mechanism of pathogenesis. Science 300: 486-489. Kotilinek L.A., Bacskai B., Westerman M., Kawarabayashi T., Younkin L., Hyman B.T., Younkin S., Ashe K.H., 2002. Reversible memory loss in a mouse transgenic model of Alzheimer's disease. J Neurosci. 22: 6331-6335. Lue L.F., Walker D.G., 2002. Modeling Alzheimer's disease immune therapy mechanisms: interactions of human pos tmortem microglia with antibody-opsonized amyloid beta peptide. J Neurosci. Res. 70: 599-610. Lue L.F., Kuo Y.M., Roher A.E., Brachova L., Shen Y., Sue L., Beach T., Kurth J.H., Rydel R.E., Rogers J., 1999. Soluble amyloid be ta peptide concentrati on as a predictor of synaptic change in Alzheimer's disease. Am J Path 155: 853-862.

PAGE 100

91 Morgan D., Diamond D.M., Gottschall P.E., Ugen K.E., Dickey C., Hardy J., Duff K., Jantzen P., DiCarlo G., Wilcock D., Connor K., Hatcher J., Hope C., Gordon M., Arendash G.W., 2000. A beta peptide vaccination prevents memory loss in an animal model of Alzheimer's disease. Nature 408: 982-985. Nicoll J.A., Wilkinson D., Holmes C., St eart P., Markham H., Weller R.O. (2003). Neuropathology of human Alzheimer disease after immunization with amyloid-beta peptide: a case report. Nat. Med. In press. Orgogozo J.M., Gilman S., Dartigues J.F., La urent B., Puel M., Kirby L.C., Jouanny P., Dubois B., Eisner L., Flitman S., Michel B.F., Boada M., Frank A., Hock C., 2003. Subacute meningoencephalitis in a subset of patients with AD after Abeta42 immunization. Neurology 61: 46-54. Schenk D., 2002. Amyloid-beta immunotherapy fo r Alzheimer's disease: the end of the beginning. Nat Rev Neurosci. 3: 824-828. 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., Games D., Seubert P., 1999. Immu nization with amyloid-beta attenuates Alzheimer-disease-like pathology in the PDAPP mouse. Nature 400: 173-177. Schimmer B.P., Parker K.L., 2001. Adrenoc orticotrophic hormone; adrenocortical steroids and their synthetic analogs; inhi bitors of the synthesis and actions of adrenocortical hormones. In: Goodman & G ilmans The Pharmaocological Basis of Therapeutics. (Hardman J.G., Limbird L.E., Gilman A.G. ed.). pp1649-1677. McGraw Hill Medical Publishing Division. Solomon B., Koppel R., Frenkel D., Ha nan-Aharon E., 1997. Disaggregation of Alzheimer beta-amyloid by site-directed m onoclonal antibodies. Proc. Natl. Acad. Sci. USA 94: 4109-4112. Tanaka J., Fujita H., Matsuda S., Toku K., Sakanaka M., Maeda N., 1997. Glucocorticoidand mineralocorticoid recepto rs in microglial cells: The two receptors mediate differential effects of co rticosteroids. Gl ia 20: 23-37. Tikka T., Fiebich B.L., Goldsteins G., Ke inanen R., Koistinaho J., 2001. Minocycline, a tetracycline derivative, is neuroprotective agai nst excitotoxicity by inhibiting activation and proliferation of microgl ia. J. Neurosci. 21: 2580-2588. Walsh D.M., Klyubin I., Fadeeva J.V., Cullen W.K., Anwyl R., Wolfe M.S., Rowan M.J., Selkoe D.J., 2002. Naturally secreted oligomers of amyloid beta protein potently inhibit hippocampal long-term potentiati on in vivo. Nature 416: 535-539.

PAGE 101

92 Wilcock D.M., DiCarlo G., Henderson D., J ackson J., Clarke K., Ugen K.E., Gordon M.N., Morgan D., 2003. Intracranially administered anti-A antibodies reduce -amyloid deposition by mechanisms independent of a nd associated with mi croglial activation. J. Neurosci. 213: 3745-3751. Wilcock D.M., Gordon M.N., Ugen K.E., Gottschall P.E., DiCarlo G., Dickey C., Boyett K.W., Jantzen P.T., Connor K.E., Melachrino J., Hardy J., Morgan D., 2001. Number of Abeta inoculations in APP+PS1 transgenic mice influences antibody titers, microglial activation, and congophilic plaque levels. DNA Cell Biol. 20: 731-736. Yrjanheikki J., Keinanen R ., Pelikka M., Hokfelt T., Ko istinaho J., 1998. Tetracyclines inhibit microglial activation a nd are neuroprotective in global brain ischemia. Proc. Natl. Acad. Sci. USA. 99: 10837-10842.

PAGE 102

93 PAPER 3: PASSIVE AMYLOID IMMUNOTH ERAPY CLEARS AMYLOID AND TRANSIENTLY ACTIVATES MICROGLIA IN A TRANSGENIC MOUSE MODEL OF AMYLOID DEPOSITION. Donna M. Wilcock 1 Amyn Rojiani 2 Arnon Rosenthal 3 Gil Levkowitz 3 Sangeetha Subbarao 3 Jennifer Alamed 1 David Wilson 1 Nedda Wilson 1 Melissa J. Freeman 1 , Marcia N. Gordon 1 Dave Morgan 1 1,2 : Alzheimers Research Laboratory, 1 Department of Pharmacology, 2 Departments of Interdisciplinary Oncology and Pathology, Univ ersity of South Florida, 12901 Bruce B Downs Blvd, Tampa, FL 33612, USA. 3 : Rinat Neuroscience Corp. 3155 Porter Driv e, Palo Alto, California, 94304, USA. This work was published in Journal of Neuroscience 2004 July 7 24(27): 6144-6151. ACKNOWLEDGEMENTS: This work was supp orted by National Institutes of Aging / NIH grants AG15490 (MNG) and AG18478 (DM). DMW is the Benjamin Scholar in Alzheimers Disease Research.

PAGE 103

94 Abstract The role of microglia in the removal of amyloid deposits following systemically administered anti-A antibodies remains unclear. In the current study we injected Tg2576 APP transgenic mice weekly with anti-A antibody for a period of one, two or three months such that all mice were 22 months at the end of the study. In mice immunized for three months we found an improvement in alternation performa nce in the Y maze. Histologically, we were able to detect mouse IgG bound to congophilic amyloid deposits in those mice treated with anti-A antibody but not in those tr eated with control antibody. We found that Fc receptor expression on microglia was increased following one month of treatment while CD45 was increased follo wing two months of treatment. Associated with these microglial changes was a reduc tion in both diffuse and compact amyloid deposits following two months of treatment. Interestingly, the microglia markers were reduced to control levels fo llowing three months of treatment while amyloid levels remained reduced. Serum A levels and anti A antibody levels were elevated to similar levels at all three survival times in mice given anti-A injections rather than control antibody injections. These data show that anti body is able to enter the brain and bind to the amyloid deposits, likely opsonizing the A and resulting in Fc receptor mediated phagocytosis. Together with our earlier wo rk, our data argue that all proposed mechanisms of anti-A antibody mediated amyloid removal can be simultaneously active.

PAGE 104

95 Introduction Reduction of brain amyloid following anti-A immunotherapy was first demonstrated by Schenk and colleagues (1999). Their report showed that vaccination with A 1-42 in the PDAPP transgenic mouse model of Alzheimers disease dramatically reduced levels of A deposits in the brain. Later it was shown that using the same vaccination protocol in APP+ PS1 doubly transgenic mice (Morgan et al, 2000) and in TgCRND8 transgenic mice (Janus et al, 2000) not only reduced A levels in the brain but also protected these mice from memory defici ts. More recent studies have demonstrated that passive immunization c onsisting of direct anti-A antibody injections not only results in dramatic reduction of A levels (Bard et al, 2000; DeMattos et al, 2001) in the brain but also reverses memory deficits in tr ansgenic mouse models of AD (Dodart et al, 2002; Kotilinek et al, 2002). The mechanism(s) by which immunotherapy acts remain unclear. Suggested mechanisms include microglial mediated phagoc ytosis (Schenk et al, 1999, Wilcock et al, 2001, 2003, 2004), disaggregation of amyloid deposits (Solomon 1997, Wilcock et al, 2003, 2004), and removal of A from the brain by binding of circulating A in plasma with the anti-A antibodies, resulting in a concentrat ion gradient from brain to plasma. This latter mechanism is also known as the peripheral sink hypothe sis (DeMattos et al, 2001, Dodart et al, 2002, Lemere et al, 2003, Das et al, 2003). We have previously reported th at following intracranial anti-A antibody injections into APP transgenic mice there is a rapid removal of di ffuse amyloid deposits

PAGE 105

96 apparently independent of microglial activation and also a later removal of compact amyloid deposits which appears to require microglial activation (Wilc ock et al, 2003). In fact, in a later study using the same mode l, administration of dexamethasone, which suppresses microglial activation, anti-A antibody administration inhibits the removal of compact, thioflavine-S positive, amyloid deposits (Wilcock et al, 2004). In this report we show that weekly systemic administration of anti-A antibodies for a period of one, two or three months resu lts in a dramatic re duction of both diffuse and compact amyloid deposits. Associated with this reduction is a behavioral improvement using the Y-maze task. Following one month of treatment there is a large induction of Fc receptor expression on microglia and following two months of administration there is an increase in CD45 expression indicative of microglial activation. We have detected antibody binding to congophilic plaque in APP transgenic mice treated with anti-A antibody. We also observe a drama tic increase in circulating A levels following one month of administration. Two mo nths following administration we observe a dramatic reduction in compact and di ffuse deposits. After three months of administration the microglia markers are down to control levels whilst the compact and diffuse amyloid deposits remain reduced. These results demonstrate systemically administered anti-A antibodies are accessing the brain, binding to amyloid deposits and activating microglia. The data also show an increase in circulating A in plasma, consistent with the peripheral sink hypothesis.

PAGE 106

97 Materials and Methods. Experiment design: Singly transgenic APP Tg 2576 mice were obtained from our breeding program at USF started in 1996 (Holcomb et al, 1998). Twenty two APP transgenic mice aged 19 months were assigned to one of four experi mental groups. The first three groups received weekly intraperitoneal anti-A antibody injections (antibody 2286; Mouse monoclonal anti-human A 28-40 IgG1; Rinat Neurosciences, Palo Alto, CA) for 1 month (n=6), 2 months (n=9) or 3 months (n=4). The fourth group received weekly intraperitoneal antiAMN antibody injections (2906; Mouse monoc lonal anti-drosophila amnesiac protein IgG1; (Rinat Neurosciences, Palo Alto, CA) for 3 months (n=3). Twelve nontransgenic mice were assigned to one of two experimental groups. The first group received intraperitoneal anti-A antibody injections for 3 mont hs (n=4). The second group received no treatment (n=3). Treatment of 1 month and 2 month groups was delayed to insure the mice were killed at the same age (22mo). One week prior to kill and one day following the 5 th 9 th or 13 th injection mice were tested behaviorally using the Y maze task. Behavioral testing: The Y maze is a three arm maze with e qual angles between all arms. Mice were initially placed within one arm and the seque nce and number of arm entries was recorded for each mouse over an 8 minute period. The per centage of triads in which all three arms were represented (ABC, CAB or BCA but not BAB) was recorded as an alternation to estimate short-term memory of the last ar ms entered. The total number of possible

PAGE 107

98 alternations is the number of arm entrie s minus two. Additionally, the number of arm entries serves as an indicator of activity. Antibody Purification: Antibody 2286 (mouse monoclonal anti-human A 28-40 IgG1) and the antibody 2906 (mouse monoclonal anti-drosophila amne siac protein IgG1) was purified from mouse ascites on AKTA instrumentation usin g protein A beads (MabSelect, Amersham Biosciences). Briefly, ascites was filtered in pyrogen-free .22m filter system (Corning) and applied to a 20ml bed volume in a XK16/ 20 column (Amersham Biosciences) after equilibrating the beads with 5 vol of bindi ng buffer (0.6M NaCl, 0.3M glycine, pH 8.0). The column was washed with 3 vol of bindi ng buffer and the antibody was eluted in 4 vol of elution buffer (0.1M Na Citrate pH 3.0), and held at low pH for 30 min for viral inactivation. The resulting elua nt was neutralized with 1/10 th vol of 1.0M Tris, pH 9.5. The antibody was dialyzed into sterile PBS, pH 7.4 and the concentration was determined by reading absorbance at 280. All buffers were made in pyrogen-free water. Tissue preparation: On the day of kill mice were wei ghed, overdosed with 100mg/kg pentobarbital (Nembutal sodium solution, Abbott laboratories, North Chicago IL). Blood was collected and allowed to coagulate at 4 o C for at least one hour before being centrifuged and the serum removed and stored at -80 o C until required. The mice were then intracardially perfused with 25ml of 0.9% sodium chloride. Brains were rapidly removed and the right half of the brain was dissected and frozen for biochemistry while the le ft half of the brain was immersion fixed for 24 hours in freshly prepared 4% paraformaldehyde in 100mM PO 4 (pH 7.2) for histopathology. The latter hemibrains were then incubated for 24 hours

PAGE 108

99 in 10, 20 and 30% sucrose sequentially to cyr oprotect them. Horizontal sections of 25 m thickness were collected using a sl iding microtome and stored at 4 o C in Dulbeccos phosphate buffered saline with sodium azide (pH 7.2) to prevent microbial growth. ELISA methods; A and anti-A antibody: For the A assay serum was diluted and incubated in 96-well microtiter plates (NUNC MaxiSorp, Denmark), which were pr e-coated with antibody 6E10 (Signet, Dedham, MA) at 5g/ml in PBS buffer, pH 7.4. The secondary antibody was biotinylated 4G8 (Signet, Dedham, MA) at 1:5000 diluti on. Detection was done using streptavidin horseradish peroxidase conj ugate (Amersham Biosciences), followed by TMB substrate (KPL, Maryland). Standard curves of A -40 (Global Peptide, Ft. Collins, CO) scaling from 400 6pm were used. Anti-A antibody was dissociated from endogenous A in serum as described previously (Li et al, 2004). Briefly, serum was diluted in dissociation buffer (0.2M glycine HCl, 1.5% BSA pH 2.5) and incubated at room temp erature for 20 min. The sera were pipetted into the sample reservoir of Microcon centrifugal device, YM-10 (10,000 MW cut-off, Millipore) and centrifuged at 8,000 x g for 20 min. at RT. The sample reservoir was then separated from the flow through, placed inverted into a second tube and centrifuged at 1000 x g for 3 min. The collected solution co ntaining the antibody dissociated from the A peptide was neutralized to pH 7.0 with 1M Tris buffer pH 9.5. The dissociated sera were assayed by ELISA for antibody titer. A 1-40 (Global Peptide, Ft. Collins, CO) coated 96-well microtiter plates (NUNC MaxiSorp, Denmark) were incubated with dissociated serum samples. A biotinylated goat-anti mouse IgG (H+L) (Vector, Burlingame, CA) at

PAGE 109

100 1:5000 dilution, followed by peroxidase -c onjugated streptavidin (Amersham Biosciences) was used to detect serum anti-A binding activity. Immunohistochemical methods: A series of eight equally spaced tissu e sections 2.4mm apart were randomly selected spanning the entire brain and staine d using free-floating im munohistochemistry methods for total (rabbit polyclonal anti-pan A Bisource, Camarillo, CA, 1:10000), CD45 (rat anti-mouse CD45, Serotec, Raleigh NC, 1:3000), Fc receptors II /III (rat antimouse CD16/ CD32, BD Pharmingen, San Die go, CA, 1:3000) as previously described (Gordon et al, 2002). Briefly, tissue was incubated in primary antibody overnight at room temperature. Sections were then washed a nd incubated in the appropriate biotinylated secondary antibody (for A : goat anti-rabbit 1:3000; for CD45 and Fc R: goat anti-rat 1:1000. All Vector Laboratories, Berlingame CA ) for two hours. Following multiple washes tissue was incubated in ABC (Vector Laboratories, Berlingame, CA) for a period of one hour. Color development was performed using 3,3-Diaminobenzidine (DAB, Sigma-Aldrich, St Louis, MO) enhanced with nickelous ammonium sulfate (J.T. Baker Chemical Co., Phillipsburg, NJ) for CD45 and Fc R, or without enhancement for A For immunostaining, some sections were omitted from the primary antibody to assess nonspecific immunohistochemical reactions. Additional sections were also staine d for mouse IgG using imunohistochemical methods similar to that described above. Brie fly, sections were incubated overnight in a 1:3000 concentration of anti-mouse IgG conjuga ted to horseradish peroxidase (SigmaAldrich, St Louis, MO). The sections were then washed and incubated for 5 minutes in 100ml TBS (tris-buffered sa line) containing 50mg DAB (3,3-diaminobenzidine, Sigma-

PAGE 110

101 Aldrich, St Louis, MO), 500mg nickelous ammonium sulfate (J.T. Baker Chemical Co. Phillipsburg, NJ) and 100 l 30% hydrogen peroxide to pr oduce a purple/ black color reaction product. The sections were then mounted on slides and counterstained with a 0.2% Congo red solution in 80% ethanol to asse ss the localization of positive mouse IgG stain with compact amyloid deposits. A set of sections also mounted and stained using 0.2% Congo red solution in NaCl saturated 80% ethanol. Another set of s ections was also mounted and stained using 4% thioflavine-S (Sigma-Aldrich, St Louis MO) for 10 minutes. The immunohistochemical reaction product on all sections was measured using the Image-Pro Plus version 4.5 software (Med ia Cybernetics, Silver Spring, MD). One region of the frontal cortex for all sections from each animal was analyzed and the average of 6-7 sections was taken to give a value for each animal. Three regions of the hippocampus were analyzed on approximately 4-5 sections where hippocampus was present; the CA1, CA3 and dentate gyrus. Thes e regions were analyzed both individually to yield an average per region and also combined to give an overall value for hippocampus for each animal. This ensured that there was no regional bias in the hippocampal values. These same analysis methods were used to evaluate the Congo red stain also. To assess possible treatment-relate d differences, the values for each treatment group were analyzed by one-way ANOV A followed by Fischers LSD means comparisons. Nontransgenic mice showed no treatment related differences in any histological analyses and so these groups were pooled.

PAGE 111

102 Data analysis: Percent alternation and arm entry numbers from the Y-maze behavior task were analyzed using a one-way analysis of variance (ANOVA) followed by Fischers LSD means comparisons using StatView software version 5.0.1 (SAS Institute Inc., NC). Nontransgenic mice showed no treatment relate d differences in any behavioral analyses and so these groups were pooled. ELISA values for serum A levels and circulating antibody levels were analyzed using a one ANOVA followed by Fischers LSD means. Results. Transgenic APP mice given control antibody injections showed significantly reduced Y-maze alternation when compared to the nontransgenic mice (Fig. 1A). This reduced alternation was reversed in the APP transgenic mice receiving weekly anti-A antibody injections for three months. This gr oup of mice was indistinguishable from the nontransgenic animals and showed significantly increased alternation compared to the APP transgenic mice receiving control antibody (Fig. 1A). The APP transgenic mice given weekly anti-A antibody injections for either one or two mont hs were intermediate between nontransgenic and tran sgenic mice given control anti bodies and not significantly different from either group. Nontransgenic mice also made significantly fewer arm entries than the APP transgenic mice receiv ing control antibody injections indicating hyperactivity in the APP transgenic mice. The APP transgenic mice receiving anti-A antibody injections for two and three months did not exhibit this hyperactivity and were not significantly different from any other treatment groups (Fig 1B).

PAGE 112

103 One day following anti-A antibody administration anti-A antibodies were detected in serum at high levels (400nM) following one month of administration. This level of antibody in the serum was the same af ter two or three mont hs of administration with no apparent accumulation of antibody (F ig 2A). Associated with high anti-A antibody levels in serum at one month was a dramatic increase in circulating A levels in serum. APP transgenic mice receiving c ontrol antibody had only 1.5nM circulating A in plasma compared to APP transgenic mice receiving A antibody for one month which had 130nM circulating A in plasma; an almost 100-fold increase (Fig 2B). Despite similar levels of anti-A antibody at one, two or thr ee months of administration, circulating A levels declined between one and tw o month. They also showed a slight decline between two and three months of administration although w ith both two and three months of administration circulating A levels were still significantly elevated compared to APP transgenic mice receiving control antibody (Fig 2B). Following systemic administration of anti-A antibodies weekly for one month staining for mouse IgG could be detected on plaques throughout the brains of APP transgenic mice (Fig 3B). The staining was the most intense where plaque load is greatest; the hippocampus and frontal corte x. This staining was not observed in APP transgenic mice receiving control antibody (Fig 3A ). It should be noted that staining with higher concentrations of anti-mouse IgG-HR P did show staining of plaques in both control treated and anti-A treated APP transgenic mice. Staining for mouse IgG was still present, and s lightly more intense, around the plaques that remain following two (Fig. 3C) and three (Fig. 3D ) months of treatment.

PAGE 113

104 Total A immunohistochemistry in the APP transgenic mice receiving control antibody (Fig. 4A) showed a few intensely stained deposits suggesting compacted amyloid deposits along with more numerous di ffuse deposits. There was a concentration of deposits around the hilus of the hippocam pus as well as the molecular layers of Ammons horn. This was a typical amount and distribution of A for APP transgenic mice of this age, as previously describe d (Hsiao et al, 1996 and Gordon et al, 2002). Following one month of weekly anti-A antibody injections there appeared to be a slight reduction in A immunohistochemistry in the hippocampus (Fig 4B) although this was not statistically significant (Fig 4E). The reduction appeared to be primarily diffuse deposits, with most of the compact amyloid deposits remaining (F ig 4B). After two months of weekly anti-A antibody injections we observe d a dramatic reduction in A immunohistochemistry which appeared to be both compact and diffuse amyloid deposits from the hilus and dentate gyrus regions of the hippocampus as well as the pyramidal cell regions, with only a few deposits remaining, often in the vicinity of the hippocampal fissure and outer molecular layers (Fig 4C). This reduction in A load at two months was approximately 60% in the hippocampus and approximately 55% in the frontal cortex (Fig 4E, hippocampus P<0.001, frontal cortex P<0.005). Total A levels remained reduced after three months of treatment but did not appear to decrease any further (Fig 4D and E). Congo red staining detects only compact am yloid deposits in the beta pleated sheet structure. There were far fewer Congo red positive amyloid deposits than A deposits detected by total A immunohistochemistry. Congo red positive deposits were located primarily along the fissure of the hippocampus as well as the CA1/ subiculum

PAGE 114

105 region in APP transgenic mice receiving control anti body (Fig 5A). There was no reduction in Congophilic deposits 1 month follo wing treatment in either the hippocampus (Fig. 5B and E) or the frontal cortex (Fig 5E). Following two months of treatment there was a significant reduction in both number a nd size of congophilic de posits in both the hippocampus (Fig. 5C and E) and frontal cortex (Fig. 5E). This reduction was approximately 60% in the frontal cortex and approximately 50% in the hippocampus (Fig. 5E, P<0.005 frontal cortex P<0.01 hippocampus,). There was a small further reduction between two months and three months which is approximately 30% in hippocampus and frontal cortex (Fig. 5E). Th ioflavine-S staining was also measured and confirmed the Congo red data showing the same reductions in stained area as did Congo red (data not shown). Immunohistochemical staining for Fc receptors II and III in APP transgenic mice receiving control antibody treatment for three months showed only very faint staining of microglia in close association with amyloid deposits (Fig 6A). Following one month of anti-A antibody administration there was a dramatic induction of Fc receptors II and III on microglia. The microglia expressing the Fc receptors after 1 month of treatment were not only associated with amyloid deposits but ar e also diffusely distributed (Fig 6B). This induction averaged 100-fold in the hippocampus (Fig. 6B and E, P<0.05) and frontal cortex (Fig. 6E, P<0.05). Fc receptor expression levels fell only slightly between one month and two months of treatment although th is expression was once again concentrated on microglia around remaining amyloid de posits (Fig. 6D). Induction remained approximately 100-fold in hippocampus (Fig. 6E, P<0.05) and frontal cortex (Fig. 6E, P<0.05). Following three m onths of treatment Fc receptor expression was reduced to

PAGE 115

106 levels observed in APP transgenic mice r eceiving control antibody (Fig. 6E) although it appeared to be increased in microglia around the few remaining amyloid deposits (Fig. 6D). CD45, a protein tyrosine phophatase, is normally moderately expressed on microglia around amyloid deposits in aged APP transgenic mice and is a commonly used marker for microglial activation. This mode rate expression was observed in the APP transgenic mice receiving control antibody trea tment for three months (Fig 7A and E). Following one month of treatment we obser ved an increase in CD45 expression on microglia surrounding amyloid deposits in both the hippocampus (Fig. 7B, F and I) and frontal cortex (Fig. 7I). While the expr ession in hippocampus was approximately 2.5 times that observed in control treated APP transgenic mice in the hippocampus (Fig. 7I, not significant) and twice the va lues found in the frontal cortex of control animals, the elevation was not statistically significant (Fig. 7I, not significant). Following two months of anti-A antibody treatment there was a furt her increase in CD45 expression on microglia not only surrounding the amyloid de posits but also di ffusely distributed throughout the amyloid containing br ain regions (Fig. 7C and G). It is possible that this more widespread activation is in associat ion with diffuse amyloid deposits although we cannot confirm this. The increased expressi on was approximately 3.5 times that observed in control treated mice in the hippocampus (Fig. 7I, P<0.05) and 3 times in the frontal cortex (Fig. 7I, P<0.01). After three months of anti-A antibody treatment CD45 expression remained at the same levels as that observed after one month of treatment (Fig 7I),however, the microglia were still diffuse ly distributed with fewer microglia around deposits compared to one or two mont hs of treatment (Fig 7D and H).

PAGE 116

107 Discussion. The data presented here suggest th at peripherally administered anti-A antibodies entered the brain, bound to congophilic amyloid plaques, and led to removal of deposited amyloid. In support of the argument that anti-A antibodies entered the brain, we found mouse IgG decorating the remaining congophilic amyloid plaques of APP transgenic mice administered anti-A antibody, but no IgG in APP transgenic mice administered control antibody. This difference was best di scerned when low titers of the anti-mouse IgG-HRP were used. Lemere et al (2004) al so reported immunohistoc hemical labeling of amyloid deposits for mouse IgG after passive immunization, but detected signals in both immunized and non-immunized mice. It is unclear whether lower anti-mouse IgG concentrations might have reveal ed selective staining in anti-A treated animals. These data confirm in parafomaldehyde fixed tissue the observations of Bard et al (2000), who used unfixed cryostat sections. Associated with the presence of antibody in the brain after one month of treatment was a dramatic activation of Fc receptor expression on microglia, further arguing that anti-A antibodies entered the brain and opsonize d the amyloid deposits. Later, following two months of treatment, we observed an increase in CD45 expr ession on microglia, indicating activation of these cells beyond th e level normally associated with amyloid deposits. It has previously been shown that following active immunization with A 1-42 in humans that anti-A antibodies are present in cerebros pinal fluid, in some instance equal to serum concentration, suggesting some pene tration into the brain from the periphery

PAGE 117

108 (Hock et al, 2002). It has also been shown th at 0.1% of an intravenous injection of radiolabeled anti-A antibody crosses the blood-brain ba rrier of SAMP8 mice (Banks et al, 2002). Thus, accumulating data indicates that circulating antibodies can access the brain parenchyma, which has important implications not only for the use of immunotherapy in Alzheimers disease but also for other diseases in which immunotherapy is being pursued such as Creutzfeldt-Jakob disease (Manuelidis, 1998; Sigurdsson et al, 2003) and neur al infections associated with human immunodeficiency virus (McMichael and Hanke, 2003). Associated with the changes in microglia l markers was a significant reduction in both compact and diffuse amyloid deposits fo llowing two months of treatment, these remained reduced following three m onths of treatment. Removal of A deposits from the brain appeared to be a gradual process. We did not observe significant reductions in either diffuse or compact amyloid deposits following one month of weekly anti-A antibody treatment. Following two months of tr eatment there was a dramatic reduction in total A immunohistochemistry, Congo red st aining and thioflavine-S staining, suggesting removal of both diffuse and comp act amyloid deposits. There appeared to be no accumulation of the injected antibody, since serum anti-A antibody levels were the same regardless of duration of treatment. This would suggest that this time-dependent removal of amyloid deposits was not occu rring because of increasing antibody levels, rather, it appears that some mobilization of removal mechanisms must be present for some time before significant removal is apparent. An early feature we observe d was the increase in Fc receptors II and III (CD16 / CD32) expression on microglia, which was a pparent following one and two months of

PAGE 118

109 treatment. The murine Fc receptors II and III share a high affinity for IgG1 antibodies (the isotype used in the curre nt study) as well as IgG2a (G essner et al, 1998). Following this increased Fc receptor expression was an increase in CD45 expression on microglia following two months of treatment. CD45 is a protein tyrosine phosphatase which is elevated with microglial activation. In this study, it appears that the increase in CD45 expression represents a further activation step from that seen after one month of treatment where we observe the increased Fc receptor expression. Following three months of treatment, both Fc receptor and CD45 expression on mi croglia were reduced to control levels, possibly due to the s ubstantial reduction in amyloid deposits. It is important to note that if we had looked at only the three month time-point we would not have detected the activation of the microglia by CD45 or, likel y by other markers such as Mac-1 (Das et al, 2003). We have previously observed a similar loss of microglia activation following intracranial antibody administration (Wilcock et al, 2003) and active immunization (Wilcock et al, 2001). Three days following a single injection of anti-A antibody in the frontal cortex and hippocampus we observed an increase in CD45 expression, however, seven days following injection the CD45 expr ession was reduced to control levels, in parallel with clearance of the A deposits (Wilcock et al, 2003) This suggests that the reduced microglial activation could possibly be due to the clearance of most amyloid plaques. It is also conceivabl e that the microglia could be undergoing apoptosis due to the robust activation as has been described previ ously by Liu et al (2001) when microglia are overactivated by LPS. An alternative explana tion could be tolerance of the microglia to antibody opsonized A We have previously shown a re duction in microglial reaction in

PAGE 119

110 an active immunization model using doubly tr ansgenic APP +PS1 mice. Following five monthly inoculations we observed a significant increase in CD45 expression, however, following nine monthly inoculations, CD45 leve ls were comparable to control animals despite continued high antibody titer levels (Wilcock et al, 2001). Due to the inflammatory adverse effects s een in the human clinical trial of the active immunization by Elan pharmaceuticals it could be suggested that the microglial activation observed in this study was due to an immune response unrelated to opsonization of A by the antibody. To evaluate whethe r this was the case we examined the thalamus and cerebellum, which do not contain any amyloid deposits, for any increase in CD45 or Fc receptor expression and did not ob serve any such increase. Thus it appears that the microglial activation is sp ecific to amyloid c ontaining brain regions and is likely a specific response to opsonized A as opposed to a general non-specific inflammatory reaction. The data presented here extend our earlie r observations of the benefits of active anti-A immunization on learning and memory (Morgan et al, 2000). We show that passive immunization with anti-A antibodies for a period of three months reduced amyloid deposits and improved behavioral performance as indi cated by a significant increase in alternation in th e y-maze as well as a decrease in the number of arm entries. The arm entry data suggests that there is not a complete reversal of the increased activity. There is a trend towards some improvement in alternation at the 1 month time point (although not significant) despite no reduction in total A immunohistochemistry. Such improvements may reflect rapid reductions of an A pool (oligomeric?) closely linked to memory impairments yet not easily dete cted by immunohistochemistry. This phenomena

PAGE 120

111 was previously described by D odart et al (2002) and Kotilin ek et al (2002) who showed rapid reversal of memory deficits in transgenic mice following passive immunization without significant reduction in brain A The results described above indicating entry of anti-A antibody into brain and activation of microglia suggest s that some opsonization of A is likely stimulating microglial involvement in the clearance of A deposits. This is consistent with the phagocytosis mechanisms of amyloid removal put forward by the Elan group (Schenk et al, 1999; Bard et al, 2000; Bard et al, 2003). Our earlier work with direct injection of anti-A antibody into brain suggests two mech anisms; one not requiring an Fc component nor activatio n of microglia which can clear diffuse A and a second that requires the Fc domain and ac tivation of microglia (Wilcock et al, 2003; Wilcock et al, 2004). It is conceivable that the first non-Fc requiring mechanism is analogous to the catalytic dissolution mechanism described by So lomon et al (1996). The diffuse material, whatever its state of oligom erization may be more accessibl e to this action of anti-A antibodies. Finally, at all dur ations of antibody exposure we observe a dramatic increase in circulating A levels in plasma. This is consistent with a role for the peripheral sink mechanism (DeMattos et al, 2001, Dodart et al, 2002; Lemere 2003) in the reduction of CNS A after passive immunization. We conclude that our studies usi ng antibody 2286, in aggregate, provide support for all three major pr oposed mechanisms of anti-A antibody action in lowering brain amyloid. It is essential to recognize that these mechanisms are not mutually exclusive, and are likely to be synergistic if multiple mechanisms are elicted by a single

PAGE 121

112 antibody/serum. It is also important to rec ognize that not all monoc lonal antibodies need work via all three mechanisms. Both isotype and epitope selectivity could regulate which anti-A action is dominant for a specific antibody. These studies also do not speak towards other immune system related actions that might underlie the benefits (or adverse effects) of active immunizati on. Nonetheless, given the pr eliminary data that anti-A immunotherapy may stabilize cognitive function in Alzheimer patients (Hock et al, 2003) and the consistent reversal of the phenot ype found in APP transgenic mice by such approaches, these results support further develo pment of the optimal strategies for using anti-A immunotherapy as a treatment for Alzheimer's dementia.

PAGE 122

113 Figure 1: Y-maze behavioral improvement after systemic anti-A antibody administration. Panel A shows percent alternation for nontransgenic (NTg) mice receiving no treatment (), APP transgenic mice (APP) receiving control antibody (Cont) for three months and APP transg enic mice (APP) receiving anti-A (A ) antibody for 1, 2 or 3 months. indicates P<0.05 when comp ared to nontransgenic untreated mice and APP transgenic mice receiving anti-A antibody for 3 months. Panel B shows number of arm entries for nontransgenic (NTg) mice r eceiving no treatment (), APP transgenic mice (APP) receiving control antibody (Cont) for three mont hs and APP transgenic mice (APP) receiving anti-A (A ) antibody for 1, 2 or 3 mont hs. indicates P<0.05 when compared to nontransgenic untreated mice.

PAGE 123

114

PAGE 124

115 Figure 2: Increased se rum levels of anti-A antibody and A after anti-A antibody administration. Panel A shows amounts of circulating anti-A antibodies in APP transgenic mice (APP) receiving either cont rol antibody (Cont) for 3 months or anti-A antibody (A ) for 1, 2 or 3 months, nontransgenic (NTg) mice receiving either control antibody (Cont) or anti-A antibody (A ) for 3 months and nontransgenic mice receiving no treatment. ** indicates P<0.001 compar ed to APP mice given control antibody injections. Panel B shows amounts of circulating A in sera in APP transgenic mice (APP) receiving either control antib ody (Cont) for 3 months or anti-A antibody (A ) for 1, 2 or 3 months, nontransgenic (NTg) mice r eceiving either contro l antibody (Cont) or anti-A antibody (A ) for 3 months and nontransgenic mice receiving no treatment. ** indicates P<0.01, indicates P<0.05.

PAGE 125

116

PAGE 126

117 Figure 3: Mouse IgG immunohistochemist ry shows antibody binding to congophilic plaques in anti-A antibody treated mice but not contro l antibody treated mice. Panels AD show anti-mouse IgG-HRP immunohistochemi stry counterstained with Congo red to detect compact amyloid deposits. Panel A shows a representative amyloid deposit and associated anti-mouse IgG immunostaining (black) in the hippocampus of a mouse injected with control antibody for three months. Panels B-D shows a representative amyloid deposit (red) associated with anti -mouse IgG immunostaining (black) in the hippocampus of a mouse injected with anti-A antibody for one month (B), two months (C) or three months (D). Ma gnification = 200X, scale bar = 25 m.

PAGE 127

118 118

PAGE 128

119 Figure 4: Total A immunohistochemistry is reduced fo llowing two months of systemic anti-A antibody administration. Panels A-D show total A immunohistochemistry in the hippocampus of APP transgenic mice receivi ng control antibody for th ree months (Panel A. Percent area for this section was 9.12%), anti-A antibody for one month (Panel B. Percent area for this section was 6.84%), anti-A antibody for two months (Panel C. Percent area for this se ction was 3.23%) or anti-A antibody for three months (Panel D. Percent area for this section was 2.49%). Magnification = 40 X, scale bar = 120 m. In panel D, CA1 indicates cornu ammonis 1, CA 3 indicates cornu ammonis 3, F indicates the hippocampal fissure and DG indicates the de ntate gyrus. Panel E shows quantification of the percent area occupied by A positive stain in the fron tal cortex and hippocampus. The single bar shows the value for APP transgenic mice receiving control antibody for three months. The line shows the values for APP transgenic mice receiving anti-A antibody for a period of one, two a nd three months. ** indicates P<0.01.

PAGE 129

120 120

PAGE 130

121

PAGE 131

122 Figure 5: Congophilic compact amyloid plaque s are reduced following two months of anti-A antibody administration. Panels A-D show Congo red staining in the hippocampus of APP transgenic mice receivi ng control antibody for th ree months (Panel A), anti-A antibody for one month (Panel B), anti-A antibody for two months (Panel C) or anti-A antibody for three months (Panel D). Magnification = 40X, scale bar = 120 m. In panel D, CA1 indicates cornu am monis 1, CA3 indicates cornu ammonis 3, F indicates the hippocampal fissure and DG i ndicates the dentate gyrus. Panel E shows quantification of the percen t area occupied by Congo red pos itive stain in the frontal cortex and hippocampus. The single bar shows the value for APP transgenic mice receiving control antibody for three months. The line shows the values for APP transgenic mice receiving anti-A antibody for a period of one, two and three months. ** indicates P<0.01.

PAGE 132

123 123

PAGE 133

124

PAGE 134

125 Figure 6: Fc receptor expression on microglia is increased following one month of antiA antibody treatment and remains increased following two months of treatment. Panels A-D show Fc receptor immunohistochemistry in th e hippocampus of APP transgenic mice receiving control antibody for three months (Panel A), anti-A antibody for one month (Panel B), anti-A antibody for two months (Panel C) or anti-A antibody for three months (Panel D). In panel A, F indi cates the hippocampal fissure, DG indicates the dentate gyrus. Magnificati on = 100X, scale bar = 50 m. Panel E shows quantification of the percent area occupied by Fc receptor positive stain in the frontal cortex and hippocampus. The single bar shows the value for APP transgenic mice receiving control antibody for three months. The line shows the values for APP transgenic mice receiving anti-A antibody for a period of one, two and three months. indicates P<0.05.

PAGE 135

126 126

PAGE 136

127 Figure 7: CD45 expression on microglia is in creased following two months of anti-A antibody treatment. Panels A-D show CD45 immunohistochemistry in the hippocampus of APP transgenic mice receiving control an tibody for three months (Panel A), anti-A antibody for one month (Panel B), anti-A antibody for two months (Panel C) or anti-A antibody for three months (Panel D). In pane l A, F indicates the hippocampal fissure, DG indicates the dentate gyrus. Magni fication = 100X, scale bar = 50 m. Panels E-H are magnified images of non-amyloid containing areas from panels A-D. Panels E-H show CD45 immunohistochemistry in the hippocampus of APP transgenic mice receiving control antibody for three months (Panel E), anti-A antibody for one month (Panel F), anti-A antibody for two months (Panel G) or anti-A antibody for three months (Panel H). Panel I shows quantification of the pe rcent area occupied by CD45 positive stain in the frontal cortex and hippocampus. The singl e bar shows the value for APP transgenic mice receiving control antibody for three months. The line shows the values for APP transgenic mice receiving anti-A antibody for a period of one, two and three months. indicates P<0.05.

PAGE 137

128 128

PAGE 138

129 References. Arendash GW, Gordon MN, Diamond DM, Aus tin LA, Hatcher JM, Jantzen P, DiCarlo G, Wilcock D, Morgan D. 2001. Behavioral as sessment of Alzheimers transgenic mice following long-term A vaccination: Task specificity and correlations between A deposition and spatial memor y. DNA Cell Biol 20: 737-744. Banks WA, Terrell B, Farr SA, Robinson SM, Nonaka N, Morley JE. 2002. Passage of amyloid protein antibody across the blood-br ain barrier in a mouse model of Alzheimers disease. Peptides 23: 2223-2226. Bard F, Cannon C, Barbour R, Burke RL, Games D, Grajeda H, Guido T, Hu K, Huang J, Johnson-Wood K, Khan K, Kholodenko D, Lee M, Lieberberg I, Motter R, Nguyen M, Soriano F, Vasquez N, Weiss K, Welch B, Seubert P, Schenk D, and Yednock T. 2000. Peripherally administered anti bodies against amyloid -peptid e enter the central nervous system and reduce pathology in a mouse model of Alzheimer's disease. Nature Medicine 6: 916-919. Das P, Howard V, Loosbrock N, Dickson D, Murphy MP, Golde TE. 2003. Amyloid-beta immunization effectively redu ces amyloid deposition in Fc Rgamma-/knock-out mice. J Neurosci 23: 8532-8538. DeMattos RB, Bales KR, Cummins DJ, D odart JC, Paul SM, and Holtzman DM. 2001. Peripheral anti-A antibody alters CNS and plasma A clearance and decreases brain A burden in a mouse model of Alzehimer's di sease. Proc. Natl. Acad. Sci. USA 98: 88508855. DiCarlo G, Wilcock D, Henderson D, Gor don M, and Morgan D. 2001. Intrahippocampal LPS injections reduce A load in APP+PS1 transgenic mice. Neurobiology of Aging 22, 1007-1012. Dodart JC, Bales KR, Gannon KS, Greene SJ, DeMattos RB, Mathis C, DeLong CA, Wu S, Wu X, Holtzman DM, Paul SM. 2002. Immuni zation reverses memory deficits without reducing brain Abeta burden in Alzheimer' s disease model. Nat Neurosci 5: 452-457. Gessner JE, Heiken H, Tamm A, Schmidt RE. 1998. The IgG Fc receptor family. Ann Hematol 76: 231-248. Gordon MN, Holcomb LA, Jantzen PT, DiCarl o G, Wilcock D, Boyett KW, Connor K, Melachrino J, O'Callaghan JP Morgan D. 2002. Time course of the development of Alzheimer-like pathology in the doubly tran sgenic PS1+APP mouse. Exp. Neurol. 173: 183-195.

PAGE 139

130 Hauss-Wegrzyniak B, Willard LB, Del So ldato P, Pepeu G, Wenk GL. 1999. Peripheral administration of novel anti-inflammatorie s can attenuate the effects of chronic inflammation within the CNS. Brain Res. 815: 36-43. Holcomb L, Gordon MN, 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 CM, Eckman C, Younkin S, Hsiao K, Duff K. 1998. Accelerated Alzheimer-type phenotype in transgenic mice carrying both mutant amyloid precursor protein and presenilin 1 transgenes. Nat Med 4: 97-100. Hock C, Konietzko U, Papassotiropoulos A, Wollmer A, Streffer J, Von Rotz RC, Davey G, Moritz E, Nitsch RM. 2002. Generation of antibodies specific for beta-amyloid by vaccination of patients with Alzhei mer disease. Nat Med 8(11): 1270-1275. Hsiao K, Chapman P, Nilsen S, Eckman C, Harigaya Y, Younkin S, Yang F, Cole G. 1996. Correlative memory deficits, Abeta elevat ion, and amyloid plaques in transgenic mice. Science 274: 99-102. Jantzen PT, Connor KE, DiCarlo G, We nk GL, Wallace JL, Rojiani AM, Coppola D, Morgan D, Gordon MN. 2002. Microglial activa tion and beta -amyloid deposit reduction caused by a nitric oxide-releasing nonstero idal anti-inflammatory drug in amyloid precursor protein plus presenilin-1 tr ansgenic mice. J Neurosci 22: 2246-2254. Janus C, Pearson J, McLaurin J, Mathews PM, Jiang Y, Schmidt SD, Chishti MA, Horne P, Heslin D, French J, Mount HT, Nixon RA, Mercken M, Bergeron C, Fraser PE, George-Hyslop P, Westaway D. 2000. A beta peptide immunization reduces behavioural impairment and plaques in a model of Alzheimer's disease. Nature 408: 979-982. Kotilinek LA, Bacskai B, Westerman M, Kawarabayashi T, Younkin L, Hyman BT, Younkin S, Ashe KH. 2002. Reversible memory loss in a mouse tr ansgenic model of Alzheimer's disease. J Neurosci 22: 6331-6335. Liu B, Wang K, Gao HM, Mandavilli B, Wang JY, Hong JS (2001). Molecular consequences of activated microglia in th e brain: overactivation induces apoptosis. J Neurochem 77: 182-189. Seabrook TJ, Bloom JK, Iglesias M, Spooner ET, Walsh DM, Lemere CA. 2004. Species-specific immune response to i mmunization with human vs. rodent A peptide. Neurobiol Aging. In Press. Lemere CA, Spooner ET, LaFrancois J, Malest er B, Mori C, Leverone JF, Matsuoka Y, Taylor JW, DeMattos RB, Holtzman DM, Cl ements JD, Selkoe DJ, Duff KE. 2003. Evidence for peripheral clearance of cereb ral Abeta protein following chronic, active Abeta immunization in PSA PP mice. Neurobiol Dis 10-18.

PAGE 140

131 Li Q, Cao C, Chackerian B, Schiller J, Gordon M, Morgan D. 2004. Overcoming antigen masking of anti-A antibodies reveals breaking of B cell tolerance by virus-like particles in A immunized amyloid precursor protein tran sgenic mice. BMC Neurosci. In Press. Manuelidis L. 1998. Vaccination with an at tenuated Creutzfeldt-Jakob disease strain prevents expression of a virulent agent. Proc. Natl. Acad. Sci. USA 95: 2520-2525. McMichael AJ, Hanke T. 2003. HIV vaccines 1983-2003. Nat Med 9: 874-880. Morgan D, Diamond DM, Gottschall PE, Ugen KE, Dickey C, Hardy J, Duff K, Jantzen P, DiCarlo G, Wilcock D, Connor K, Hatcher J, Hope C, Gordon M, Arendash GW. 2000. A beta peptide vaccination prevents memory loss in an animal model of Alzheimer's disease. Nature 408: 982-985. 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, Games D, Seubert P. 1999. Immunization with amyloid-beta attenuates Alzheimerdisease-like pathology in the PDAPP mouse. Nature 400: 173-177. Sigurdsson EM, Sy MS, Li R, Scholtzova H, Kascsak RJ, Kascsak R, Carp R, Meeker HC, Frangione B, Wisniewski T. 2003. Antiprion antibodies for prophylaxis following prion exposure in mice. Neurosci Lett 336: 185-187. Solomon B, Koppel R, Frenkel D, Hanan-Ah aron E. 1997. Disaggregation of Alzheimer beta-amyloid by site-directed monoclonal an tibodies. Proc. Natl. Acad. Sci. USA 94: 4109-4112. Wilcock DM, Muniredddy SK, Rosenthal A, Ugen KE, Gordon MN, Morgan D. 2004. Microglial Activation Facilitates A Plaque Removal Following Intracranial Anti-A Antibody Administration. Neurobiol. Dis. 15: 11-20. Wilcock DM, DiCarlo G, Henderson D, Jack son J, Clarke K, Ugen KE, Gordon MN, Morgan D. 2003. Intracranially administered anti-A antibodies reduce -amyloid deposition by mechanisms independent of a nd associated with mi croglial activation. J. Neurosci. 213: 3745-3751. Wilcock DM, Gordon MN, Ugen KE, Gottschall PE, DiCarlo G, Dickey C, Boyett KW, Jantzen PT, Connor KE, Melachrino J, Har dy J, Morgan D. 2001. Number of Abeta inoculations in APP+PS1 transgenic mi ce influences antibody titers, microglial activation, and congophilic plaque levels. DNA Cell Biol 20: 731-736.

PAGE 141

132 PAPER 4: PASSIVE IMMUNOTHERAPY AGAINST A IN AGED APP TRANSGENIC MICE REVERSES COGNITIVE DEFICITS AND DEPLETES PARENCHYMAL AMYLOID DEPOSITS IN SPITE OF INCREASED VASCULAR AMYLOID AND MICROHEMORRHAGE. Donna M Wilcock 1 Amyn Rojiani 2 Arnon Rosenthal 3 Sangeetha Subbarao 3 Melissa J Freeman 1 Marcia N Gordon 1 Dave Morgan 1 1, 2 : Alzheimers Research Laboratory, Univers ity of South Florida, Departments of 1 Pharmacology and 2 Interdisciplinary Oncology, 12901 Bruce B Downs Blvd, Tampa, FL 33612. 3: Rinat Neuroscience Corp. 3155 Porter Dr ive, Palo Alto, California, 94304, USA. This work was published in the Journal of Neuroinflammation 2004 December 8 1:24. Acknowledgements: This work was supported by National Inst itutes of Aging / NIH grants AG15490 (MNG) and AG18478 (DM). DMW is the Benjam in Scholar in Alzheimers Disease Research. We would like to thank Keis ha Symmonds who aided in histological processing of the tissue and Nedda Wilson who was responsible for animal husbandry during the study. We would also like to thank Lori Lutz fo r assisting in editing the manuscript.

PAGE 142

133 Abstract Background: Anti-A immunotherapy in transgenic mice reduces both diffuse and compact amyloid deposits, improves memory functi on and clears earl y-stage phospho-tau aggregates. As most AD cases occur well past midlife, the current study examined adoptive transfer of anti-A antibodies to 19 and 23 month old APP transgenic mice. Results: After three months of weekly injecti ons, this passive immunization protocol completely reversed learning and memory de ficits in these mice, a benefit which was undiminished after 5 months of treatmen t. Dramatic reductions of diffuse A immunostaining and parenchymal Congophilic am yloid deposits were observed after 5 months, indicating even we ll established amyloid depos its are susceptible to immunotherapy. However, cerebr al amyloid angiopathy increased substantially with immunotherapy and some deposits were associat ed with microhemorrhage. Reanalysis of results collected from an earlier time course study demonstrated that these increases in vascular deposits were dependent on the duration of immunotherapy. Conclusions: The cognitive benefits of pa ssive immunotherapy persist in spite of the presence of vascular amyloid and small hemorrhages. These data suggest that clinical trials evaluating such treatments will require precau tions to minimize potential adverse events associated with microhemorrhage.

PAGE 143

134 Introduction Alzheimers disease (AD) is character ized not only by the presence of parenchymal amyloid deposits and intracellu lar tangles but also by the presence of amyloid deposits in the vasc ulature, a condition referre d to as cerebral amyloid angiopathy (CAA). The CAA observed in both Al zheimers disease patients (Iwatsubo et al, 1994) and some of the transgenic mouse models (Gor don et al, 2002) is primarily composed of the shorter form of amyloid beta (A A 1-40 while the majority of amyloid deposits in the parenchyma are composed of A 1-42 although the compact amyloid deposits also contain A 1-40 Anti-A immunotherapy has been consider ed as a potential treatment for AD for some time (Solomon et al, 1996; Schenk et al, 1999). Active immunization with a vaccine including A 1-42 fibrils progressed to human clinical trials where its administration was suspended due to meningoe ncephalitits in a subset of patients (Orgogozo et al, 2003). To date there have been pathology reports on two patients who participated in the trial and subsequently di ed (Nicoll et al, 2003, Ferrer et al, 2004). Both reports note that while the numbers of pa renchymal amyloid deposits appeared lower than expected in these cases, the CAA in these patients did not appear outside the normal range In addition, one report mentioned multip le cortical hemorrhages and the presence of hemosiderin around the CAA ve ssels (Ferrer et al, 2004). Given the adverse reactions to the active immunization, th e irreversibility of such procedures and the variable antibody response to vaccines in older individuals (Weksler et al, 1997), passive immunizat ion against the A peptide emerged as an alternative immunotherapeutic strategy. Studies in young and middle aged APP transgenic mice

PAGE 144

135 have reported significant amyloid reductions with passive immuniza tion (DeMattos et al, 2001, Bard et al, 2000, Wilcock et al, 2003). Such treatments also demonstrate rapid improvements of memory function in A PP mice, sometimes without detectable reductions in amyloid (Dodart et al, 2002, Kotilinek et al, 2002, Wilcock et al, 2004b). Most recently, intracranial administration of anti-A antibodies has been shown to not only remove A but also clear early-stage hyperphosphorylated tau aggregates (Oddo et al, 2004). Importantly, in the only prior st udy evaluating adoptive an tibody transfer in older APP mice, Pfeifer et al (2002) repor ted a doubling of cerebral microhemorrhages associated with significant reductions in amyloid burden after administration of an Nterminal specific anti-A antibody. Materials and Methods: Experiment design: Mice derived from APP Tg2576 mice were obtained from our breeding program at USF started in 1996 (Holcomb et al, 1998). For the five month treatment study thirteen APP transgenic mice aged 23 months were assi gned to one of two groups. The first group received weekly intraperitoneal anti-A antibody injections (antibody 2286; Mouse monoclonal anti-human A 28-40 IgG1; Rinat Neurosciences, Palo Alto, CA) for a period of five months ( n=6). The second group received weekly intraperitoneal anti-AMN antibody (2906; Mouse monoclonal anti-dros ophila amnesiac protein IgG1; (Rinat Neurosciences, Palo Alto, CA) injec tions for a period of five months ( n=7). Seven nontransgenic mice were also assigned to one of two groups. The first group received weekly intraperitoneal anti-A antibody injections for a period of five months ( n=4). The

PAGE 145

136 second group received weekly intraperitoneal anti-AMN an tibody injections for a period of five months ( n=3). For the time course study of 1, 2 or 3 mo treatment, twenty two APP transgenic mice aged 19 months were assigned to one of four experimental groups, as described previously (Wilcock et al, 2004b). The first th ree groups received weekly intraperitoneal anti-A antibody injections for 3 months, 2 months or 1 month, ending when all mice were 22 mo of age. The fourth group receiv ed weekly intraperitoneal anti-AMN antibody injections for 3 months. Behavioral analysis: Following three and five months of tr eatment the mice from the 5 month study were subjected to a two day radial-arm wa ter maze paradigm. The apparatus was a 6 arm maze as described previously (Morgan et al, 2000). On day 1, 15 trials were run in 3 blocks of five. A cohort of 4 mice were run sequentially for each block (i.e. each of 4 mice get trial one, then the same mice get tr ial two, etc). After each 5 trial block, a second cohort of mice were run permitting an extended rest period before mice were exposed to the second block of 5 trials. The goal arm was different for each mouse in a cohort to minimize odor cues. The start arm was varied for each trial, with the goal arm remaining constant for both days. For the first 11 trials the platform was alternately visible then hidden (hidden for the last 4 trials). On da y two, the mice were run in exactly the same manner as day 1 except that the platform was hidden for all trials. Th e number of errors (incorrect arm entries) were measured in a one minute time frame. As in prior studies, mice failing to make an arm choice in 20 seconds were assigned 1 e rror (no mice in this

PAGE 146

137 study had to be assigned an error in this ma nner). The same individual administered the antibody treatments and placed mice in the radial arm water maze. Due to the numbers of mice in the study the researcher was unaware of treatment group identity of each mouse. Also, the dependent measures in the radial -arm water maze task are quantitative, not evaluative, so the potential for tester bias is reduced. In order to minimize the influence of individual trial variability, each mouse's errors for three consecutive trials were averaged producing 5 data points for each day which were analyzed statistically by ANOVA using StatView (SAS Institute Inc., NC). Tissue preparation and histology: On the day of kill mice were weighed, overdosed with 100mg/kg Nembutal sodium solution (Abbott laboratories, No rth Chicago IL). The mice were then intracardially perfused with 25ml of 0.9% sodi um chloride. Brains were rapidly removed and the left half of the brain was immersi on fixed for 24 hours in freshly prepared 4% paraformaldehyde in 100mM KPO 4 (pH 7.2) for histopatholo gy. The hemibrains were then incubated for 24 hours in 10, 20 and 30% sucrose sequentially to cyroprotect them. Horizontal sections of 25 m thickness were collected us ing a sliding microtome and stored at 4 o C in Dulbeccos phosphate-buffered sa line with sodium azide (pH 7.2) to prevent microbial growth. A series of eight equally spaced tissue sections 0600 m apart were randomly selected spanning the enti re brain and stained using free-floating immunohistochemistry for total (rabbit polyclonal anti-pan A Bisource, Camarillo, CA, 1:10000) as previously described (G ordon et al, 2002, Wilcock et al, 2004b). A second series of tissue sections 0.6mm apart were stained using 0.2% Congo red solution in NaCl saturated 80% ethanol. Another set of sections were also mounted and stained for

PAGE 147

138 hemosiderin using 2% potassium ferrocyanid e in 2% hydrochloric acid for 15 minutes followed by a counterstain in a 1% neutral re d solution for 10 minutes Quantification of Congo red staining and A immunohistochemistry was performed using the Image-Pro Plus (Media Cybernetics, S ilver Spring, MD) to analyze the percent area occupied by positive stain. One region of the frontal cortex and three regions of the hippocampus were analyzed (to ensure that there was no regional bias in the hippocampal values). The initial analysis of Congo red was performed to gi ve a total value. A second analysis was performed after manually editing out all of th e parenchymal amyloid deposits to yield a percent area restriced to va scular Congo red staining. To estimate the parenchymal area of Congo red we subtracted the vascular amyl oid values from the total percentage. For the hemosiderin stain the number of Prussian blue positive sites were counted on all sections and the average number of sites per section calculated. Looking at the sections at a low magnification we were able to observe a qualitative difference between animals, however, the percent area was so low that ma ny fields contained no positive stain. Eight equally spaced sections were examined a nd number of positive profiles were counted and averaged to a per section value. To asse ss possible treatment-related differences, the values for each treatment group were analyzed by one-way ANOVA followed by Fischers LSD means comparisons. Results Reversal of cognitive deficits by passive amyloid immunotherapy: The radial-arm water maze task detects spatial learning and memory deficits in transgenic mouse models (Morgan et al, 2000 ; Gordon et al, 2001). We treated 23 mo old mice for 5 mo with anti-A antibody 2286 or control antibody 2906 (against a

PAGE 148

139 drospohila-specific protein) a nd tested them for spatial navigation learning in a 2 day version of the radial arm water maze after 3 mo of treatment and, using a new platform location, again after 5 mo of treatment. At both testing times we found that APP mice treated with the control antibody failed to lear n platform location over two days of testing and were significantly impaired compared to the nontransgenic mice treated with either antibody (Fig 1). However, APP mice administ ered the anti-A antibodies demonstrated a complete reversal of the impairment obs erved in the control APP transgenic mice, ending day 2 with performance near 0.5 errors per trial (Fig. 1). Although learning at the later time point, when the mice were 28 mo of age, may have been slightly slower for all groups, there was no impairment of the anti-A antibody treated APP. Passive amyloid immunotherapy clears parenc hymal A deposits, but increases vascular amyloid: In a prior experiment examining the e ffects of passive anti-A immunotherapy for 1, 2 or 3 mo in APP mice killed at 21 mo of age (Wilcock et al, 2004b), we found a time dependent reduction of both A immunostaining of diffuse and fibrillar deposits and Congo red staining of fibrillar amyloid depos its. In the current study we found a similar reduction in both A immunostaining (Table 1) and total Congo red staining (Fig 2A, left panel; P<0.001 frontal cortex and P< 0.01 hippocampus) after 5 months of immunotherapy. We noted that the bulk of what remained was vascular amyloid. We then separately analyzed vascular and parenc hymal deposits which revealed a near 90% reduction in parenchymal deposits (P<0.001), but a 3-4 fold elevation of vascular Congo red staining (P<0.0001; Fig 2A, cen ter and right panels resepec tively). We also separately analyzed vascular and parenchymal Congo re d staining on mice from our earlier study

PAGE 149

140 (14) treated passively for 1, 2 or 3 months with anti-A or cont rol antibody, and found a similar result. There was a graded reduction in overall Congo red staining nearing 75% as duration of antibody exposure increased (as re ported previously; Fig 2B). However, when separated into vascular Congo red deposits and parenchymal deposits, there was an antibody exposure time dependent increase in vascular deposition in both hippocampus and frontal cortex (Fig 2C; P<0.05 frontal cortex and hippocampus ) and a corresponding near 90% decrease in parenchymal deposits (Fig 2D; P<0.001 in frontal cortex and hippocampus). These differences were readily observed examining micrographs of sections from these mice. Mice treated with control antibod ies revealed occasional cortical vascular amyloid deposits (22 mo Fig 3A, 28 mo Fig 3C), while mice administered anti-A antibodies had increased amounts of vascular amyloid staining (3 mo treatment Fig 3B; 5 mo treatment Fig 3D). Those vessels cont aining amyloid following treatment with antiA antibody also exhibited appa rent increases in microglia l activation as measured by CD45 expression (Fig 3F) compared to mice treated with control antibody (Fig 3E). Unfortunately, the shifting numbers and si zes of vascular and parenchymal deposits caused by the antibody therapy greatly complicat ed measurement of microglial activation per vascular deposit area. This apparent in crease in staining intensity could not be quantified accurately. Passive amyloid immunotherapy causes increased microhemorrhage: We used the Prussian blue histological stain to label hemosiderin, a ferric oxide material produced in the breakdown of hem oglobin. Extravenous blood in the brain leads to microglial phagocytosis of the erythroc ytes and breakdown of hemoglobin within

PAGE 150

141 them. These ferric oxide containing microglia are thus markers of past hemorrhage. In untreated, aged APP transgenic mice we obser ved very few profiles positive for Prussian blue staining in the frontal cortex (section counterstained with neutral red; Fig. 4A). However, following anti-A antibody treatment for five months we observed an increase in the number of Prussian blue profiles in th e frontal cortex which we re readily detectable at a low power in the microscope (Fig 4B). In the absence of anti-A treatment, or even when treated with antbody for 1 month, most ve ssels did not stain with Prussian blue, and could be identified only using the red counterstain (Fig 4C). However, even with 3 months of anti-A antibody tr eatment we observed frequent vessels with associated Prussian blue staining (Fig 4D) Using adjacent sections stained for Congo red, we confirmed that all vessels showing microhemo rrhage contained amyloid (Fig 4E and 4F; we were unable to double-label Prussian blue with either Congo red or thioflavine-S). However, only a minority of vessels cont aining amyloid demonstrated hemorrhage. When we counted the number of Prussian blue positive profiles in those animals receiving control antibody there was an averag e of one profile per every 2 sections (Fig 5) and this number remained the same in both control groups (aged 22 or 28 mo). Following treatment with anti-A antibody for a period of two months we observed a striking increase in Prussian blue staining, approximately fi ve times that observed in either the control group or the mice immunized for one month (Fig 5, P<0.001). Following this initial increase in Prussian bl ue staining we observed a linear increase in staining associated with increasing duration of anti-A antibody treatment (Fig 5). Five months of anti-A antibody treatment demonstrated a 6fold increase in Prussian blue staining when compared the control groups (Fig 5).

PAGE 151

142 Discussion Earlier studies with vaccines against the A peptide demonstrated protection from the learning and memory deficits associ ated with amyloid accumulation in APP transgenic mice (Morgan et al, 2000, Janus et al, 2000). Passive immunization protocols with anti-A antibodies also produced cogniti ve benefits, in some cases even in the absence of significant reduction in amyloid burden (Dodart et al, 2002, Kotilinek et al, 2002). Our recent work found that 3 months of anti-A treatment of 18 mo old APP mice improved spontaneous alternation performan ce on the Y-maze (Wilcock et al, 2004b). In the present work we confirmed that passive anti-amyloid immunotherapy can reverse spatial learning deficits in APP transgenic mice and that this benefit of immunotherapy is retained even in aged mice (26 and 28 mo at testing) with long established amyloid pathology. Additionally, we describe a more rapid means of testing spatial reference memory to reveal learning and memory deficits in APP transgenic mice. This 2 day version of the radial arm water maze included greater spacing of individual trials (mice spent time in their home cage after every trial), combined with less spacing of aggregate trials (15 trials per day rather than 4 or 5) to facilitate le arning of platform loca tion in the nontransgenic mice with a clear absence of learning in the age-matched transgenic mice. A substantial reduction in total Congophilic amyloid deposits was observed in old APP mice treated with anti-A antibodies for 2 or more months. This measurement of total Congo red staining included both parenc hymal and vascular am yloid staining. When we analyzed the sections for only CAA we found that this measure was significantly increased following two, three and five months of anti-A antibody treatment. The

PAGE 152

143 remaining parenchymal amyloid load was almost completely eliminated with this antibody approach. Clearly, sin ce total amyloid load is si gnificantly reduced not all amyloid was shifted into the vessels but it a ppears that at least some of the Congophilic material was redistributed to the vasculature. At the present time the mechanism for this redistribution is unclear. However, one possibi lity is that the microglia associated with the antibody-opsonized amyloid, either by phagocytosis or surface binding, and transported the material to the vasculature, possibly in an attempt to expel it. We and others have shown evidence for microglial i nvolvement in the removal of amyloid using both intracranial anti-A antibody injections (Wilcock et al, 2003, 2004a) and systemically administered anti-A antibody treatment (Wilcock et al, 2004b) as well as ex vivo studies (Bard et al 2000, 2003). Here we also report our impression that microglia surrounding CAA vessels in imm unized mice expressed more CD45 than control transgenic mice. This increased expression could be due to either increased expression in the same number of microglial cells or an increased number of microglial cells in these animals. It is feasible that this microglial activation was simply in reaction to the presence of increased amyloid in th e blood vessels. However, it is equally likely that microglia activated by the opsonized materi al migrated to the ve ssels for disposal of the amyloid. CAA is defined as the deposition of congophilic material in meningeal and cerebral arteries and arteriol es (capillaries and veins can also show CAA but less frequently) and it occurs to some extent in nearly all Alzheimers disease patients (Jellinger, 2002). Severe CAA affecting about 15% of cases, can be a ssociated with both infarction and hemorrhagic injury (Olichney et al, 1997, Maurino et al, 2001). It has also

PAGE 153

144 been shown that the severity of CAA can be directly linked to the severity of dementia in Alzheimers disease patients (Thal et al, 2003). In the current study we show a significantly increased number of microhemorrhages in the brain as detected by Prussian blue staining associated with the increase in CAA following pa ssive immunization. Another transgenic mouse model of amyloid deposition, the APP23 mice, have b een shown to deposit amyloid in both brain parenchyma and blood vessels and show a C AA associated increase in spontaneous cerebral hemorrhages (Winkler et al, 2001). More over, Pfeifer et al (2002) showed that these spontaneous hemorrhages were signifi cantly increased following 5 months of passive immunization of 21 mo old APP23 mice using an anti-A antibody with an Nterminal epitope, similar to those typically developed against ac tive immunization with vaccines (Schenk et al, 1999, Dickey et al, 2001, McLaurin et al, 2002) When young mice (6 mo) were immunized following the same protocol no hemorrhages were observed. More recently DeMattos et al (9 th International Confer ence on Alzheimers disease and related disorders, 2004) showed that passive imm unization with an Nterminal antibody (3D6: directed against aa 1-5 of A ) of PDAPP transgenic mice also resulted in significantly increased microhe morrhage. They were unable to detect increased microhemorrhage with a mid-domain antibody (266: directed against aa 13-28 of A ). Notably, antibody 266 fails to bind A deposited in CAA vessels or amyloid plaques (9 th International Conference on Alzheime rs disease and related disorders, 2004). Importantly, Ferrer et al (2004) noted the presence of CAA and microhemorrhage in the brain of one patient th at participated in the A vaccine trial, even though the parenchymal amyloid appeared lower than ex pected. Also, Nicoll et al [6] noted that

PAGE 154

145 CAA appeared unaffected in the brain of anot her patient that participated in the A vaccine trial. It remains to be determined whether these observations regarding increased CAA and microhemorrhage in transgenic mice are re levant to trials of passive immunotherapy in humans. It should be noted that in spite of extending the period of immunotherapy to 5 months, there was no discernable loss of the cognitive benefits of immunotherapy in the transgenic mice, all of whom showed incr eased microhemorrhage. While the observation that antibody 266 does not result in vascular leakage encour ages testing of this idiotype, data from the Zurich cohort of the A vacci ne trial argue that br ain reactive antibodies may be important for cognitive benefits (Hoc k et al, 2003). Our opinion is that these results suggest that passive immunotherapy against A should proceed with appropriate precautions taken to minimize the risk of hemorrhage (e.g. by excluding patients taking anticoagulants), and instituting measures to detect such hemorrhages if they do occur irrespective of the antibody specificity or proclivity for microhemorrhage in aged APP transgenic mice.

PAGE 155

146 Author contributions. DMW treated the mice, performed the beha vioral analysis, processed the tissue and performed pathological analyses, and dr afted the manuscript. AR evaluated slides and provided expert opinion regardi ng CAA and microhemorrhage. AR and SS developed, produced and purified the antibod ies used in the studies. KS aided in histological processing of the tissue. NW was responsible for animal husbandry during the study. MJF performed DNA extraction and PCR for genotyping of the mice. MNG oversees the breeding colony gene rating mice for the studies, collected samples from the mice and assisted in editing the manuscri pt. DM conceived the design of the study, guided data interpretation and assi sted in editing the manuscript.

PAGE 156

147 Figure 1: Spatial learning deficits in APP transgenic mice were reversed following three and five months of immunizati on. Mice were tested in a 2 da y version of the radial arm water maze. Solid lines represent APP transgenic mice while dashed lines represent nontransgenic mice. Open symbols indicat e anti-AMN control antibody treatment ( : APP control antibody, : Nontransgenic control antibody) while closed symbols indicate anti-A antibody treatment ( : APP A antibody, : Nontransgenic A antibody). Panel A shows mean number of errors made over the two day trial period following three months of immunization. Each da tapoint is the average of th ree trials. Panel B shows the mean number of errors made over the two day trial period following five months of immunization. For both graphs indicates P<0.05, ** indicates P<0.001 when the APP mice receiving control antibody are compared with the remaining groups.

PAGE 157

148

PAGE 158

149 Figure 2: Passive immunization with anti-A antibodies decr eases total and parenchymal amyloid loads while increasing vascular amyl oid in frontal cortex and hippocampus of APP transgenic mice. Panel A shows total amyloid load measured with Congo red, vascular amyloid load and parenchymal am yloid load from APP transgenic mice administered control IgG (C) or anti-A IgG (A ) for a period of five months. Panels BD show total amyloid load (Panel B), vascul ar amyloid load (Panel C) and parenchymal amyloid load (Panel D) from APP transgenic mice administered control IgG for 3 months (Cont IgG) or anti-A IgG for a period of one, two or three months (Anti-A IgG). For all panels, the solid bar and solid line repres ent values from the frontal cortex while the open bar and dashed line represent values from the hippocampus. ** P<0.01.

PAGE 159

150

PAGE 160

151 Figure 3: Increased Congo red staining of blood vessels following anti-A antibody administration is associated with activated mi croglia. Panels A and B are from the frontal cortex of 22 month old APP transgenic mice i mmunized for 3 months with either control antibody (3A) or anti-A antibody (3B). Panels C and D are from the frontal cortex of 28 month old APP transgenic mice immunized for five months with e ither control antibody (3C) or anti-A antibody (3D). Panels E and F show a high magnification image of CD45 immunohistochemistry (black) counterstaine d with Congo red (red) from 28 month old APP transgenic mice immunized for five mont hs with either cont rol antibody (Panel E) or anti-A antibody (Panel F). Panels A-D magnification = 100X Scale bar in Panel B = 50 m for panels A-D. Panels E-F magnifi cation = 200X. Scale bar in Panel E = 25 m for panels E-F.

PAGE 161

152

PAGE 162

153 Figure 4: Microhemorrhage associated with CAA following systemic administration of anti-A antibodies. Panels A and B are low magnifi cation images of the frontal cortex of APP transgenic mice receiving either c ontrol antibodies (Panel A) or anti-A antibodies (Panel B) for a period of five months. Pane ls C and D show representative images of amyloid containing vessels stained for Prussian blue (blue) counter stained with neutral red (red) from APP transgenic mice receiving e ither control antibodies (Panel C) or antiA antibodies (Panel D) for a period of thr ee months. Panel E shows a blood vessel in the frontal cortex stained for Prussian blue (blu e) counterstained with neutral red from an APP transgenic mouse administered anti-A antibodies for five months. Panel F shows the same blood vessel on an adjacent section stained for Congo red indicating that the blood vessel does in fact contain amyloid. Scale bar panel A = 120 m for panels A-B. Scale bar panel C = 25 m for panels C-D. Scale bar in panel F = 25 m for panels E-F.

PAGE 164

155 Figure 5: Number of Prussian blue positive profiles increases with duration of anti-A antibody exposure. The graph shows quantificat ion of the average number of Prussian blue positive profiles per sect ion from mice administered co ntrol IgG for 3 or 5 months (Control) or anti-A IgG for 1, 2, 3 or 5 months (anti-A ). ** P<0.01.

PAGE 165

156

PAGE 166

157 Table 1. A Loads after 5 months of Immunotherapy Region % area control antibody treated % area anti-A antibody treated % reduction following anti-A antibody treatment Frontal Cortex 34.855.265 9.681.754 72 Hippocampus 23.994.985 8.212.596 66

PAGE 167

158 References. Bard, F. Cannon C., Barbour R., Burke R.L., Games D., Grajeda H., Guido T., Hu K., Huang J., Johnson-Wood K., Khan K., Kholode nko D., Lee M., Lieberberg I., Motter R., Nguyen M., Soriano F., Vasquez N., Weiss K., Welch B., Seubert P., Schenk D., Yednock T. 2000. Peripherally administered an tibodies against amyloid -peptide enter the central nervous system and reduce pa thology in a mouse model of Alzheimer's disease. Nature Medicine 6: 916-919. Bard F. Barbour R., Cannon C., Carretto R., Fox M., Games D., Guido T., Hoenow K., Hu K., Johnson-Wood K., Khan K., Kholode nko D., Lee C., Lee M., Motter R., Nguyen M., Reed A., Schenk D., Tang P., Vasquez N., Seubert P., Yednock T. 2003. Epitope and isotype specificities of anti bodies to beta -amyloid peptide for protection against Alzheimer's disease-like neuropathol ogy. Proc Natl Acad Sci USA 100: 2023-2028. DeMattos, R.B. Bales K.R., Cummins D.J., Dodart J.C., Paul S.M., Holtzman D.M. 2001. Peripheral anti-A antibody alters CNS and plasma A clearance and decreases brain A burden in a mouse model of Alzheimer' s disease. Proc. Natl. Acad. Sci. USA 98: 8850-8855. Dodart, J.C. Bales K.R., Gannon K.S., Greene S.J., DeMattos R.B., Mathis C., DeLong C.A., Wu S., Wu X., Holtzman D.M., Paul S.M. 2002. Immunization reverses memory deficits without reducing brain Abeta burden in Alzheimer's disease model. Nat Neurosci 5: 452-457. Dickey C.A. Morgan D.G., Kudchodkar S., Weiner D.B., Bai Y., Cao C., Gordon M.N., Ugen K.E. 2001. Duration and specificity of humoral immune responses in mice vaccinated with the Alzheimer' s disease-associated beta-amyloid 1-42 peptide. DNA Cell Biol. 20: 723-729. Ferrer, I., Boada Rovira, M., Sanchez Gue rra, M.L., Rey, M.J., Costa-Jussa, F. 2004. Neuropathology and pathogenesis of encephali tis following amyloidbeta immunization in Alzheimer's disease. Brain Pathol. 14:11-20. Gordon, M.N. King D.L., Diamond D.M., Jantze n P.T., Boyett K.V., Hope C.E., Hatcher J.M., DiCarlo G., Gottschall W.P., Morg an D., Arendash G.W. 2001. Correlation between cognitive deficits and Abeta deposits in transgenic APP+PS1 mice. Neurobiol. Aging 22:377-385. Gordon, M.N. Holcomb L.A., Jantzen P.T., DiCarlo G., Wilcock D., Boyett K.W., Connor K., Melachrino J., O'Callaghan J.P ., Morgan D. 2002. Time course of the development of Alzheimer-like pathology in the doubly transgen ic PS1+APP mouse. Exp. Neurol. 173: 183-195.

PAGE 168

159 Hock C. Konietzko U., Streffer J.R., Tracy J., Signorell A., Muller-Tillmanns B., Lemke U., Henke K., Moritz E., Garcia E., Wollmer M.A., Umbricht D., de Quervain D.J., Hofmann M., Maddalena A., Papassotiropoulos A., Nitsch R.M. 2003. Antibodies against beta-amyloid slow cognitiv e decline in Alzheimer's disease. Neuron 38: 547-554. Holcomb, L. Gordon M.N., McGowan E., Yu X., Benkovic S., Jantzen P., Wright K., Saad I., Mueller R., Morgan D., Sanders S., Zehr C., O'Campo K., Hardy J., Prada C.M., Eckman C., Younkin S., Hsiao K., Duff K. 1998. Accelerated Alzheimer-type phenotype in transgenic mice carrying both mutant am yloid precursor protein and presenilin 1 transgenes. Nat Med 4: 97-100. Iwatsubo, T., Odaka A., Suzuki N., Mizusawa H., Nukina N., Ihara Y. 1994. Visualization of A beta 42(43) and A beta 40 in senile plaques with end-specific A beta monoclonals: evidence that an initially deposited species is A beta 42(43). Neuron 13: 45-53. Jellinger, K.A. 2002. Alzheimer disease a nd cerebrovascular pathology: an update. J Neural Transm. 109: 813-836. 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., George-Hyslop P., Westaway D. 2000. A beta peptide immunization reduces behavioural impairment and plaques in a mode l of Alzheimer's disease. Nature 408: 979982. Kotilinek, L.A. Bacskai B., Westerman M., Kawarabayashi T., Younkin L., Hyman B.T., Younkin S., Ashe K.H. 2002. Reversible memory loss in a mouse transgenic model of Alzheimer's disease. J Neurosci 22: 6331-6335. Maurino, J., Saposnik, G., Lepera, S., Rey, R.C., Sica, R.E. 2001. Multiple simultaneous intracerbral hemorrhages: clin ical features and outcome. Arch Neurol. 58: 629-632. McLaurin J Cecal R., Kierstead M.E., Tian X., Phinney A.L., Manea M., French J.E., Lambermon M.H., Darabie A.A., Brown M.E., Janus C., Chishti M.A., Horne P., Westaway D., Fraser P.E., Mount H.T., Przybylski M., St George-Hyslop P. 2002. Therapeutically effective antibodies against amyloid-beta peptide target amyloid-beta residues 4-10 and inhibit cytotoxicity and fibrillogenesis. Nat Med. 8: 1263-1269. Morgan, D., Diamond D.M., Gottschall P.E., Ug en K.E., Dickey C., Hardy J., Duff K., Jantzen P., DiCarlo G., Wilcock D., Connor K., Hatcher J., Hope C., Gordon M., Arendash G.W. 2000. A beta peptide vaccinat ion prevents memory loss in an animal model of Alzheimer's disease. Nature 408: 982-985.

PAGE 169

160 Nicoll J.A. Wilkinson D., Holmes C., Steart P., Markham H., Weller R.O. 2003. Neuropathology of human Alzheimer disease after immunization with amyloid-beta peptide: a case report. Nat Med 9: 448-452. Oddo, S., Billings, L., Kesslak, J.P., Cr ibbs, D.H., LaFerla, F.M. 2004. Abeta Immunotherapy Leads to Clearance of Earl y, but Not Late, Hype rphosphorylated Tau Aggregates via the Proteasome. Neuron 43: 321-33. Olichney, J.M., Ellis, R.J., Katzman, R., Sabbagh, M.N., Hansen, L. 1997. Types of cerebrovascular lesions associated with severe cerebral amyloid angiopathy. Ann NY Acad Sci 826: 493-497. Orgogozo J. M. Gilman S., Dartigues J.F., La urent B., Puel M., Kirby L.C., Jouanny P., Dubois B., Eisner L., Flitman S., Michel B.F., Boada M., Frank A., Hock C. 2003. Subacute meningoencephalitis in a subset of patients with AD after Abeta42 immunization. Neurology 61: 46-54. Pfeifer, M. Boncristiano S., Bondolfi L., St alder A., Deller T., Staufenbiel M., Mathews P.M., Jucker M. 2002. Cerebral hemorrhag e after passive anti-A immunotherapy. Science 298: 1379. 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., Games D., Seubert P. 1999. Immu nization with amyloid-beta attenuates Alzheimer-disease-like pathology in the PDAPP mouse. Nature 400: 173-177. Solomon B., Koppel R., Hanan E., Katzav T. 1996. Monoclonal antibodies inhibit in vitro fibrillar aggregation of the Alzheimer beta-amyloid peptide. Proc Natl Acad Sci USA.93: 452-455. Thal, D.R., Ghebremedhin, E., Orantes, M., Wiestler, O.D. 2003. Vascular pathology in Alzheimer disease: correlation of cerebral amyloid angiopathy and arteriosclerosis/lipohyalinosis with cognitive decline. J Neuropathol Exp Neurol 62: 1287-1301. Weksler M.E. 1997. Immunology and the elderly: an historical perspective for future international action. Mech Ageing Dev. 93: 1-6. Wilcock, D.M. DiCarlo G., Henderson D., J ackson J., Clarke K., Ugen K.E., Gordon M.N., Morgan D. 2003. Intracranially administered anti-A antibodies reduce -amyloid deposition by mechanisms independent of a nd associated with mi croglial activation. J. Neurosci. 213: 3745-3751.

PAGE 170

161 Wilcock, D.M., Muniredddy S.K., Rosenthal A., Ugen K.E., Gordon M.N., Morgan D. 2004a. Microglial Activation Facilitates A Plaque Removal Following Intracranial AntiA Antibody Administration. Neurobiol. Dis. 15: 11-20. Wilcock, D. M. Rojiani A., Rosenthal A., Levkowitz G., Subbarao S., Alamed J., Wilson D., Wilson N., Freeman M.J., Gordon M. N., Morgan D. 2004b. Passive amyloid immunotherapy clears amyloid and transiently activates microglia in a transgenic mouse model of amyloid depositi on. J. Neurosci. 24: 6144-6151. Winkler, D.T., Bondolfi L., Herzig M.C ., Jann L., Calhoun M.E., Wiederhold K.H., Tolnay M., Staufenbiel M., Jucker M. 2000. Spontaneous hemorrhagic stroke in amouse model of cerebral amyloid angiopa thy. J. Neurosci. 21: 1619-1627.

PAGE 171

162 CONCLUSIONS Alzheimers disease (AD) is a progressive neurodegenerative disorder that slowly robs sufferers of their ability to remember reason, make judgements and carry out daily activities; it is the most common cause of deme ntia with a duration of anywhere from 3 to 20 years. It is estimated that in 2004 ther e are currently 4.5 million Americans have the disease, with a projected number of 11.6 to 16 million sufferers by the year 2050 (Hebert et al, 2003). Several genes have been f ound to cause AD in humans; these are the amyloid precursor protein (APP) and preseni lin 1 and 2 (PS1 and PS2). Mutations in these genes result in a rare, aggressive form of the disease known as familial early onset Alzheimers disease (FAD) with the disease commonly occurring by age 60 and sometimes in the 30s and 40s. The discovery of these mutations led to the development of transgenic mouse models, the first of which carried normal beta-APP751 with no mutations and demonstrated some extrac ellular A deposits (Quon D et al, 1991). The first mouse overexpressing mutated human APP which developed extensive amyloid pathology but no tau pathology or ne uron loss was the PDAPP transgenic mouse (Games et al, 1995). The Tg2576 APP transgenic mouse was the first transgenic model to show learning and memory deficits associated with the amyloid pathology (Hsiao et al, 1996). Mice overexpressing human PS1 mutations were less successful models on their own, not developing any amyloid deposits (Du ff et al, 1996). However, it was shown that crossing an APP transgenic mouse with a PS1 transgenic mouse (APP+PS1) results in an accelerated model of amyloi d deposition (Borchelt et al, 1997; Citron et al, 1998;

PAGE 172

163 Holcomb et al, 1998). It was also later show n that an APP+PS1 transgenic mouse model developed reliable cognitive dysfunction which correlated with the extent of amyloid pathology (Gordon et al, 2001). These transgen ic mouse models of amyloid pathology also develop another pathological charac teristic of AD which is activation of inflammatory cells in the brain; namely mi croglia and astrocytes, which has been shown to be associated with the extent of amyloid deposition (Gordon et al, 2002). The amyloid hypothesis for AD is curren tly the most favored hypothesis for the cause of the disease which states that the pr ecipitating event in the development of AD is the over-production and deposition of A in the form of diffuse and compact amyloid deposits which in turn results in hyperphos phorylation of tau and neuronal death (Hardy et al, 1992). This hypothesis has been the focus of therapeutic intervention in AD including the development of immunother apy as a potential treatment for AD. Immunotherapy was first shown in 1999 by Schenk et al (1999) of Elan pharmaceuticals. In this report the authors showed that immunization of young PDAPP transgenic mice with A 1-42 fibrils in Freunds adjuvant over a period of 11 months prevented amyloid deposition while immunization of older PDA PP mice for 4 and 7 months resulted in significant reductions in amyloid burden. Followi ng this initial report it was later shown that immunization not only ameliorated am yloid pathology but also in resulted in improved cognition in both APP+PS1 tran sgenic mice (Morgan et al, 2000) and TgCRND8 APP transgenic mice (Janus et al, 2000). We later reporte d that over a series of several immunization studies in APP+ PS1 mice that the degree of microglial activation strongly correlated with th e reduction in compact, Congophilic amyloid deposits (Appendix A) suggesting a critical role of microglia in the removal of amyloid

PAGE 173

164 by this active immunization approach. Since th e original report in 1999 there has been a plethora of data published regarding the use of both active and passive immunization for the treatment of AD. Following the failure of active immunization in phase II clinical trials we decided that discovery of the mechanism(s) by which immunotherapy acts to reduce amyloid pathology would be critical if immunotherapy was to be successful in humans. Here we show evidence for three mech anisms for A removal by immunotherapy as well as some potentially unwanted effects of passive immunization which will likely need to be overcome if immunotherapy is to be successful in humans. By 2001 there had been three suggested mechanisms of action. The first was microglial phagocytosis via the Fc receptor which was first suggested by Schenk et al (1999) in their active immunization report and later by Bard et al (2000) in a repor t showing the benefits of passive immunization in the PDAPP mouse. We also show evidence for this mechanism in Appendix A where we observed a corre lation between microglial activation and amyloid reduction. Another mechanism was shown before immunotherapy for AD was suggested; this showed that anti-A antibod ies are capable of i nhibiting amyloid fibril formation (Solomon et al, 1996). The same gr oup also showed that anti-A antibodies could disaggregate already formed amyloi d fibrils (Solomon et al, 1997). A third suggested mechanism stated that anti-A antibodies need not en ter the CNS but act peripherally by binding to circulat ing A in the plasma and therefore resulting in a shift in the concentration gradient between CNS and plasma causing A to exit the brain and enter the serum where it would be bound by an tibodies and removed. This was evidenced by a rapid and dramatic increase in circ ulating A levels hours following anti-A

PAGE 174

165 antibody administration. Importantly, this st udy used an anti-A antibody shown not to bind to amyloid deposits in the brain (DeMattos et al, 2001). To approach the question of which m echanism was important for A removal by immunotherapy we decided to bypass the issue of blood-brain barrier penetration and work on finding out what happens when antibody is in the brain. To answer this question we injected anti-A antibodies (anti-A 1-16 ) into the right frontal cortex and hippocampus of aged Tg2576 APP transgenic mice with sign ificant amyloid burdens, leaving the left side of the brain untreated as an internal c ontrol for each mouse. We then examined what effects the antibody had after 4, 24, 72 and 168 hours. We found that diffuse amyloid deposits, stained immunohistochemically for A but not stained by thioflavine-S, were significantly reduced following 24 hours of treatment (Paper 1; Figures 2 and 3) but there was no associated microglial reaction or any reduction in compact, thioflavine-S positive, amyloid deposits. These compact deposits we re found to be significantly reduced 72 hours following injection of anti-A antibodies (Paper 1; Figures 4 and 5) and this was associated with a significant activation of microglia as detected by CD45 (Paper 1; Figures 6 and 7) and MHC-II (Paper 1; Figures 8 and 9) immunohistochemistry. 168 hours following injection we found that bot h diffuse and compact amyloid deposits remained reduced with no further reduction and microglial activation had returned to control levels. These data suggest two phases of removal of A by anti-A antibodies which may be occurring through two differe nt mechanisms. The first phase is the removal of diffuse amyloid deposits by a m echanism independent of microglial activation, possibly via a direct dissolu tion of the deposits. The second phase is the removal of compact amyloid deposits associated with activation of microglia.

PAGE 175

166 Since the time-course data was perfor med using an anti-A antibody recognizing the 1-16 portion of the A peptide we decide d to examine the differences in anti-A epitopes on removal of A from the brains of APP transgenic mice seven days following intrahippocampal and intracortic al injections. We injected antibodies recognizing 1-16, 12-28 and 28-40 portions of the A peptide as well as a control antibody directed against drosophila amnesiac protein. We found that bo th total A and compact amyloid deposits were removed equally as effectively by all three antibodies directed against A (Appendix B; Figures 1 and 2). We also found that anti-A 28-40 was slightly more efficacious in microglial activation than either anti-A 1-16 or anti-A 12-28 (Appendix B; Figure 3) although no microglial activation would have been expected given that the mice were killed seven days following injection and the time-course data indicated that microglial activation peaks at three days and is over by seven days. Additional evidence for for microglial involvement in removal of compact amyloid deposits was found when we injected anti-A 1-16 antibodies into the frontal cortex and hippocampus of aged Tg2576 APP tr ansgenic mice again but this time some mice were treated with anti-inflammatory dr ugs for the three days post-injection while others remained untreated. We treat ed mice with NCX-2216, minocycline or dexamethasone immediately following surger y and continued treatment until the morning of killing. NCX-2216 is the non-steroidal anti -inflammatory drug (NSAID) flurbiprofen with a nitric oxide donor group and a ferulic acid group attach ed. In APP+PS1 transgenic mice this drug has been shown to cause an activation of microglia and reduction of amyloid burden (Jantzen et al, 2002) but has also been shown to inhibit microglial activation following lipopolysaccharide (LPS) in jection (Hauss-Wegrzyniak et al, 1999).

PAGE 176

167 Minocycline, a tetracycline derivative, was chosen as it had been shown to suppress microglial activation following global br ain ischemia (Tikka et al, 2001), 6hydroxydopamine administration (He et al, 2001) a nd excitotoxicity (Yrjanheikki et al, 1998). Dexamethasone was used as it is a pot ent glucocorticosteroid which is known to inhibit the cyclooxygenase and lipoxygenase infl ammatory pathways as well as induce a general state of immunosuppression (Schimmer and Parker, 2001). Dexamethasone was found to be the mo st efficacious anti-inflammatory for inhibition of microglial activation due to anti-A antibodies; NCX-2216 had a moderate effect while minocycline appeared to have very little effect on the activation of microglia (Paper 2; Figure 1 and Figure 2A). Interestin gly, inhibition of mi croglial activation had no effect on removal of diffuse amyloid deposits, in all cases where anti-A antibodies were injected diffuse deposits were reduced to the same extent (Paper 2; Figure 1 and Figure 2B). In dexamethasone treated mice, however, there was no apparent reduction in compact, thioflaine-S positive, amyloid depos its demonstrating that the inhibition of microglial activation had also inhibited the removal of compact amyloid deposits. Mice treated with NCX-2216, which ha d a moderate effect on microglial activation showed modest reductions in compact amyloid deposits while those mice treated with minocycline which failed to affect microglia l activation showed reductions in compact amyloid deposits comparable to those obser ved in mice receiving no anti-inflammatory treatment (Paper 2; Figure 1 and Figure 2C). These data strongly suggest that microglial activation is necessary for the removal of compact amyloid deposits by anti-A antibodies but is not necessary for removal of diffuse amyloid deposits.

PAGE 177

168 To determine whether the microglial activ ation by anti-A antibodies occurred via the Fc receptor and to also determine whethe r it was this receptor responsible for removal of compact amyloid deposits we made F(ab) 2 fragemnts from an anti-A 28-40 IgG. These F(ab) 2 fragments lack the Fc portion of Ig G, therefore rendering them unable to interact with Fc receptors on effector cells like micr oglia yet fully capabale of binding to A in the same way as the complete IgG. We injected these anti-A F(ab) 2 fragments into the frontal cortex an d hippocampus of aged Tg2576 AP P transgenic mice and killed them 72 hours following injection. We found that the F(ab) 2 fragments were unable to activate microglia as effectively as the complete IgG is and the activation levels were comparable to that observed in the mice recei ving either control IgG or control F(ab) 2 fragments (Paper 2; Figure 3 and Figure 4A). We also found that anti-A F(ab) 2 fragments were capable of reducing diffuse amyloid depos its as effectively as the complete IgG (Paper 2; Figure 3 and Figure 4B). However, anti-A F(ab) 2 fragments were significantly worse in removing compact amyloid deposits than the complete IgG although there was small reductions observed in the hippocampus (Paper 2; Figure 3 and Figure 4C). These data suggest that although some removal of compact amyloid deposits may be possible with anti-A F(ab) 2 fragments, removal is much more efficient when the Fc portion of IgG is presen t. This further demonstrates that not only is microglial activation necessary for compact amyloid r eeduction by anti-A an tibody treatment but also it appears that the Fc receptor is the co mponent of microglia mediating this effect. Knowing that once anti-A antibodies en ter the brain parenchyma they remove A by mechanisms independent of micr oglial activation and also dependent upon microglial activation via the Fc receptor we sy stemically injected an ti-A antibodies as a

PAGE 178

169 form of passive immunization since this is mo re clinically relevant. We injected antiA 28-40 antibodies weekly for 1, 2 or 3 months such that all the mice were 22 months of age at killing. Mice received Y-maze testing dur ing the week prior to killing to test for any cognitive benefit. We found signifi cant improvement in Y-maze performance following three months of treatment by both an increase in alternations and a decrease in the number of arm entries (Paper 3; Figure 1) suggesting that the antibody treatment had improved cognitive function. We had previous ly demonstrated cognitive improvement following active immunization (Morgan et al, 2000) but this is the first time we had shown improvement following passive immunization. We also found evidence for peripheral action of anti-A antibodies. Following intraperitoneal anti-A antibody injection we found that circulating A leve ls in the serum were increased 100-fold following 1 month of treatment and remained significantly elevated following 2 and 3 months of treatment despite a slight decline compared to 1 month (Paper 3; Figure 2B). Importantly, we showed that when we stained brain tissue for mouse IgG and counterstained with Congo red to detect am yloid deposits we found that plaques were decorated with mouse IgG following anti -A antibody treatment but no staining was observed in those mice receiving control IgG (Pap er 3; Figure 3). This suggests that antiA antibodies administered systemically are able to cross the blood-brain barrier and bind to A in amyloid plaques in the brain parenchyma. We also found that A was significantly reduced following two months of anti-A antib ody treatment with a further slight reduction following three months of treat ment (Paper 3; Figure 4) as were compact amyloid deposits as detected by Congo red staining (Paper 3; Figure 5).

PAGE 179

170 It was apparent in this passive immuni zation study that there is an associated complex microglial response. The first microg lial marker to be increased was the Fcg receptors II and III, which were dramatically increased following one month of anti-A antibody treatment by approximately 100-fold. Th e expression of these receptors fell only slightly between one and two months but were reduced to control levels following three months of treatment (Paper 3; Figure 6). CD45 expression was not significantly increased until two months of anti-A antibody treatment and was reduced to control levels following three months of treatment (Paper 3; Figure 7). We later examined expression of phospho-p38 MAPK and phospho-p44/42 MAPK (also known as ERK1/2), which have been shown to be increased in microglia during activation. Ph ospho-p38 MAPK was high in APP transgenic mice receiving control anti body for three months indicating activation of microglia due to the presence of amyloid plaques. However, following treatment with anti-A antibodies for two and three months we observed a decrease in expression in microglia despite observing increases in ot her microglial activation markers at these time-points (Appendix C Figures 3 and 4) The phospho-p44/42 MAPK showed an opposite effect from the phospho-p38 MAPK which was that it was low in APP transgenic mice receiving control antibody for 3 months but was increased significantly following two and three months of anti-A an tibody treatment, in f act it was the only microglial marker which was still significantl y increased following three months of antiA antibody treatment (Appendi x C; Figures 1 and 2). This complex microglial reaction observe d following systemic administration of anti-A antibodies suggests that activation of microglia may not be a simple on-off phenomena but rather there may exist multiple states of activation dependent upon the

PAGE 180

171 stimulus and these states may result in diffe rent functional conse quences. One potential hypothesis for different activation states is that microglia may exist in two states; antigen presenting and phagocytic, as exists in peripheral macrophages. In peripheral macrophages it has been shown that phosphoryl ation of p38 MAPK results in production of IL-1 (Baldassare et al, 1999), IL-6 (Vanden Berghe et al, 1998) and IL-8 (Hobbie et al, 1999) and cuases upregulation of other proinflammatory molecules such as COX-2, iNOS and TNFby upregulating NFB driven gene expressi on (Carter et al, 1999; Kostinaho and Kostinaho, 2002). All of thes e processes associated with phospho-p38 MAPK are proinflammatory in nature and ar e associated with an antigen-presentation phenotype. Phagocytosis of apoptotic cells by peripheral macrophages results in an antiinflammatory phenotype with down regulation of pro-inflammatory chemokines, possibly via phosphorylation of p44/42-MAPK which has been shown to inhibit phosphorylation of p38-MAPK and therefore cont ribute to the anti-inflammato ry characteristics of this state (Xiao et al, 2002). Using human polym orphonuclear neutrophils it has been shown that phagocytosis of microbes is inhibited when phosphorylation of p44/42 MAPK is inhibited (Zhong et al, 2003). Since microglia are known to be derived from peripheral macrophages it is feasible that this same scenario is occurring in microglial cells. Little is currently known about the function of p38 and p44/42 MAPK s in microglia. In human AD postmortem tissue it has been shown that phopsho-p38 MAPK is highly expressed in microglia aound amyloid plaques (Hensley et al, 1999). This upregulation of phopshop38 MAPK has also been shown in the tran sgenic mouse model of amyloid deposition APP751 (Koistinaho et al, 2002). In mi croglia it has been shown that Fc receptor mediated phagocytosis is inhibited by PP 2, a Src inhibitor and piceatannol, a Syk

PAGE 181

172 inhibitor which both ultimate ly inhibit phosphorylation of p44/42, further suggesting that phosphorylation of p44/42 is necessary fo r phagocytosis (Song et al, 2004). The inflammation hypothesis of AD proposed that inflammation via activation of microglia and astrocytes in response to amyl oid deposits in the brai n actually contributes to the disease process. Activated microglia are capable of produci ng numerous cytokines, chemokines, complement components and other inflammatory mediators. Activated astrocytes are also capable of producing similar molecules. All of these inflammatory molecules are capable of eliciting numerous effects in the brain such as neuronal dysfunction and death as well as casuing furt her inflammation via activation of additional microglia and astrocytes. Despite all of the detrimental effects of inflammation in AD it has been shown that microglia are capable of eliciting beneficial e ffects in the AD brain as well. In APP transgenic mice crossed with TGF1 overexpressing mice there was significant activation of microglia and also a reduced amyloid burden when compared to the APP transgenic mice alone, however, vascul ar amyloid levels were increased (WyssCoray et al, 2001). The same group also showed that mice expressing soluble complement receptor-related protein y (sCrry), a complement inhibitor, crossed with APP transgenic mice showed reduced microglial activati on, increased amyloid plaque load and neurodegeneration (Wyss-Coray et al, 2002). Li popolysaccharide (LPS) injection into the hippocampus of aged APP transgenic mice has been shown to result in significant microglial activation and dramatic reductions in A (DiCarlo et al 2001; Herber et al, 2004). Finally there is the evidence from papers 1-3 here that microglia are necessary for effective removal of compact amyloid deposit s by microglia and also are activated in response to systemic anti-A injection.

PAGE 182

173 Following the three month time-course study in which we observed an improvement in Y-maze performance following three months of anti-A antibody treatment we took an older cohort of APP transgenic mice and treated them weekly for a period of five months and examined their c ognitive function using the much more robust method of the radial-arm water maze. Following three months of treatment mice were tested in a rapid two-day radial-arm water maze task and number of errors was measured. A significant impairment was observed in the APP transgenic mice receiving control antibody while the APP transgenic mice receivi ng anti-A antibody performed as well as the nontransgenic mice and were significantly better than the control antibody treated group (Paper 4; Figure 1A). Mice were tested two months later having received a total of five months of treatment and the benefit of the anti-A antibody treatment was still present to the same degree (Paper 4; Figur e 1B). This data suggests that complete reversal of cognitive deficits in APP transg enic mice is possible following systemic antiA antibody treatment. Histopathological analysis of Congo red and A showed the same result as had been observed in the previous time course st udy; dramatic reductions in both compact and diffuse deposits were observed followi ng anti-A antibody treatment (Paper 4; Figure 2A and Table 1). During image analysis of the Congo red and A it was observed that some of the animals appeared to have hi gh levels of amyloid in the vasculature. To quantify this we took the images used for quantification of tota l Congo red and ran the analysis again manually deselecting the pa renchymal amyloid deposits and therefore quantifying only the amyloid in the vasculat ure. There was a dramatic increase in vascular amyloid load followi ng five months of anti-A an tibody treatment. When these

PAGE 183

174 values were subtracted from the total amyl oid load the result was parenchymal amyloid load which now showed an even more dr amatic decrease due to anti-A antibody treatment (Paper 4; Figure 2A). Once we obt ained this data we decided to perform the same analyses on the three month time-cour se tissue. The same observation was made; vascular amyloid load increas ed in concert with the decreases in parenchymal amyloid. Following two months of anti-A antibody trea tment vascular amyloid burden showed a significant increase with a slight further in crease following three months of treatment. Parenchymal amyloid burden showed a significant decrease following two months of treatment with a further slight decrease fo llowing three months of treatment (Paper 4; Figure 2 B, C and D). Despite this increase in cerebral amyloid a ngiopathy (CAA) due to anti-A antibody treatment the mice showed tremendous cognitiv e benefit and so at this point the CAA is not compromising their cogni tive ability, however, if the in crease in CAA were to continue it may be predicted that at some poi nt this would cause a decline in cognition as it does in the human condition. It is important to note that not all amyloid is redistributed to the vessel, as there is still a significant decr ease in total Congo red staining, although some of the amyloid is clea rly redistributed into the vasculature. Due to the increase in CAA in our passive immunization studies and the observation by Pfeifer and colleagues (Pfeifer et al, 2002) that anti -A antibody treatment in hemorrhage prone APP transgenic mice resulte d in a significant increase in the number of microhemorrhages present we decided that we should stain the tissue for hemosiderin using a Prussian blue stain to examine for microhemorrhages. We found that there was a significant increase in the number of Prussian blue positive profiles following two, three and five months of anti-A antibody treatm ent. There was a dramatic increase between

PAGE 184

175 one and two months of anti-A antibody treatment after wh ich the number of profiles increased linearly with incr easing duration of antibody treatm ent (Paper 4; Figures 4 and 5). Importantly, all vessels showing micr ohemorrhage were also positive for CAA, although not all CAA positive vessels showed microhemorrhage. The mechanism by which anti-A anti body therapy results in increased CAA remains unclear. One hypothesis is that the microglia are capable of phagocytosing the amyloid plaques but are unable to degrade it so transport the amyloid to the vessels where they dispose of it into th e vessel wall. Support for this hypothesis comes from CD45 immunohistochemistry which showed that despite overall microglial activation being reduced to control levels following three months of anti-A antibody treatment there appeared to be high levels of microglia l activation around those vessels containing amyloid (Paper 4; Figure 3E and F). We were unable to quantify this observation due to the diffuse nature of the microglial staini ng; we could not manua lly deselect all nonvessel associated staining. It is also plausible that anti -A immunotherapy is removing more A 1-42 since this is the more prevelant species in both human AD and APP transgenic mice, and therefore sh ifting the ratio towards more A 1-40 thus resulting in amyloid deposition in the vasculature. It has previously been shown that APPDutch transgenic mice and human hereditary cerebr al hemorrhage with amyloidosis-Dutch type (HCHWA-D) have a significantly higher ratio of A to A than that observed in APP transgenic mice or human AD brain. A 1-40 predominates in vascular amyloid in AD and APP transgenic mice and thus both the APPDutch transgenic mice and the human HCHWA-D have high levels of CAA, microhemorrhage and a perivascular microglial reaction. When APPDutch transgenic mice were crossed with PS1 transgenic mice

PAGE 185

176 known to increase A 1-42 production it was observed that now the mice developed parenchymal amyloid deposits with much less prominent CAA (Herzig et al, 2004). These data suggest that a sh ift in the ratio of A:A may influence the location of deposition. Table 1 summarizes the evidence we found for each mechanism of -amyloid removal by anti-A antibody therapy. We ha ve demonstrated here by intracranial injection that anti-A antibodies in the brains of aged APP transgenic mice removes A by two distinct mechanisms. Diffuse amyloi d deposits are removed rapidly, within 24 hours, and this removal is independent of microglial activa tion. Compact amyloid deposits are removed between 24 and 72 hour s, this removal is dependent upon microglial activation and Fc receptor activation. When anti-A antibodies were administered systemically as a passive im munization the antibodies crossed the blood brain barrier and bound to amyloid plaques in the parenchyma. A removal was associated with microglial activation, Fc receptor upregulation and increased A in the serum. Also, systemically administered anti -A antibodies provided complete reversal of cognitive deficits following just three months of treatment. It was also observed that there were increased levels of vascular amyloid and multiple microhemorrhages in the brains of APP transgenic mice administered anti-A antibodies for two or mo re months. Overall, we have shown evidence for three distinct mechanisms of A removal by immunotherapy which appear to occur in concert to produce robust pathological effects. These mechanisms are direct dissolution of amyloi d fibrils, microglial phagocytosis via the Fc receptor and a peripheral sink mechanism which results from a shift in the concentration gradient of A between brain and plasma.

PAGE 186

177 There are many more questions to be answered if anti-A antibody therapy is to be successful in the clinic gi ven the pathological effects obs erved in the current studies. The main issue that needs to be overcome is the increased vascular amyloid and associated microhemorrhage. The first qu estion is whether dose of antibody and frequency of dosing would alter the pathol ogical consequences of anti-A antibody therapy. It is plausible th at a lower dose may not produce such a robust microglial response and as such, if indeed the microglia are responsible for the increase in vascular amyloid, the increased microhemorrhage and CAA may be avoided. The first way to approach this would be to establish a deta iled dose response with doses ranging from 1 to10 mg/kg for a period of three months with radial-arm water maze testing at the end. An optimal dose would provide cognitive benefit as well as reductions in parenchymal amyloid loads without an increase in CAA or microhemorrhage. In the current studies mice were injected weekly with 10mg/kg antiA antibody. It is conceivable that dosing as little as once every four weeks may be sufficient for amyloid reductions and cognitive benefit if the study were to be extended to six and nine months as opposed to three months. A study to address this would use a 10mg/kg dose and inject at intervals ranging from two to four weeks with radial-arm water maze testi ng at six and nine months. Ideally, there would be an interval which produces significant cognitive benefit and amyloid reductions with minimal incr eases in CAA and microhemorrhage. What contribution each of the three m echanisms make to the improved cognition, amyloid reductions and increases in C AA and microhemorrhage is currently unknown. Conjugating a large, polar molecule to an ti-A antibodies woul d prevent blood-brain barrier passage but would stil l permit binding of A in the plasma and therefore only the

PAGE 187

178 peripheral sink mechanism woul d be functioning. Treating the mice in exactly the same way as with the standard antibody would perm it comparisons of efficacy to be made. If the peripheral sink mechanism alone is sufficient to provide cognitive benefit and amyloid reductions it is likely that this w ould be a safe altern ative if microglial phagocytosis is responsible for the increased CAA and microhemorrhage. If the peripheral sink mechanism alone is found not to be sufficient to obtain optimal effects another approach would be to m odify antibodies so that they still cross the blood-brain barrier, can still act peripherally but are unabl e to effectively activate microglia. One such antibody is a deglycosyl ated anti-A antibody. IgG molecules have carbohydrates attached and thes e are critical to the recognition of IgGs by the Fc receptor (Radaev and Sun, 2001). If these car bohydrates are removed the IgG molecule would maintain its pharmacokinetics as the ca rbohydrates contribute very little to the overall molecular weight and woul d also not interact with the Fc receptor. Another option would be to systemica lly administer anti-A F(ab) 2 fragments similar to those used in paper 2. These, again would not interact with the Fc receptor however, due to their reduced molecular weight, their half-life in serum would be much lower as they now would be filtered out by the kidney. This means that they would need to be administered more frequently. It has been shown that at tachment of polyethylene-glycol groups to F(ab) 2 fragments, a process called PEGylation, re sults in comparable half-lives to the whole IgG molecule (Weir et al, 2002). It has been shown that PEGylation of an antiinterleukin-8 F(ab) 2 results in longer serum half-lif e and retention of comparable bioactivity (Koumenis et al, 2000). It is unclear whether pegylated F(ab) 2 fragments will

PAGE 188

179 cross the blood-brain barrier how ever, if they do not, this wo uld be an alternative way to test the peripheral sink mechanism. It is possible that microgl ial phagocytosis via the Fc receptor is necessary for effective amyloid removal and cognitive improvement and that only targeting the peripheral sink and the direct dissolution m echanisms is insufficient for clinically relevant benefits. If this is the case, attenuating the microglia l resonse using antiinflammatory drugs may minimize redistribut ion of amyloid to the vasculature and therefore reduce the inciden ce of microhemorrhage. We showed in paper 2 that an NSAID such as NCX-2216 partially inhibits microglial activation and yet significant reductions in compact amyloid deposits were observed. Co-adminis tration of an NSAID with the anti-A antibody therapy may attenua te the microglial response sufficiently to prevent increases in CAA and microhemorrhag e while providing sufficient removal of compact amyloid deposits to demonstrate cogniti ve benefit. Since the addition of a nitricoxide donor group to flurbiprofen in the NC X-2216 compound provides gastrointestinal protection this drug would be safe for l ong-term daily administration to elderly AD patients receiving the imm unotherapy. NCX-2216 has also been shown to inhibit microglial reaction to intracranial infusion of lipopolysaccharide (Hauss-Wegrzniak et al, 1999). Administration of glucocorticosteroids su ch as dexamethasone would likely not be effective for this approach as we showed in paper 2 that these completely inhibit microglial reaction and also completely inhibit the ability of anti-A antibodies to remove compact amyloid deposits. The isotype of the antibody may be critical in determining clinical efficacy and prevelance of increased CAA and microhemorrhage. In papers 3 and 4 we used an IgG1

PAGE 189

180 antibody with an epitope recognizing amino acids 28-40 of the A molecule. Microglia express three different classes of Fc receptors; Fc RI is a high affinity receptor while Fc RII and Fc RIII are low affinity receptors (Ravet ch and Kinet, 1991). It is known that in mice IgG2a binds FcgRI and III with the highest affinity while IgG1 and IgG2b bind with a lower affinity (Radaev and Sun, 2001). In fact, in a study using both in vivo and ex vivo methods, it was shown that IgG2a anti-A antibodies are more effective in removing amyloid than are IgG1 or IgG2b antibodies of the same epitope (Bard et al, 2003). In the ex vivo study the authors examined the e ffects of isotype on plaque removal from PDAPP brain sections by primary culture d microglial cells. In this study IgG2a antibodies were much more effective than Ig G1 or IgG2b antibodies of the same epitope. If microglial phagoc ytosis via the Fc receptor is responsible for not only removal of parenchymal amyloid but is also re sponsible for the increased CAA and microhemorrhage then it is possible that an IgG2a isotype antibody may in fact cause more CAA and microhemorrhage as this has a much higher affinity for the Fc receptors than the IgG1 antibody used in paper 4. The epitope may be another important i ssue to be addressed in determining efficacy. In a study using in vivo and ex vi vo methods it was shown that antibodies directed against the N-terminal of A are most effective in A removal (Bard et al, 2003). However, we show in papers 3 and 4 that Cterminal antibodies are highly effective in the removal of A. Also, a mid-domain antibody has been shown to significantly reduce brain amyloid (DeMattos et al 2001) and reverse cognitive de ficits (Dodart et al, 2002). Since mid-domain antibodies are able to bi nd soluble A but are unable to bind A in amyloid plaques these antibodies are highly effective for the peri pheral sink mechanism

PAGE 190

181 but would be unable to trigger effective mi croglial phagocytosis. It was recently shown that mid-domain antibodies do not cause mi crohemorrhage while N-terminal antibodies do (DeMattos et al, 2004). Overall, we have shown evidence that three different mechanisms are acting in concert to reduce amyloid burden in transg enic mice following anti-A antibody therapy. We have also shown that one undesirable c onsequence of anti-A antibody therapy is increased CAA and microhemorrhage. This will need to be overcome if anti-A immunotherapy is to be successful clinically.

PAGE 191

182 Table 1: Summary of the evidence found fo r the different mechanisms of -amyloid removal. Mechanism Paper 1 Paper 2 Paper 3 Paper 4 Microglial phagocytosis via Fc receptors Microglial activation peaks at the same timepoint as reductions in compact amyloid deposits are observed. Inhibition of microglial activation inhibits removal of compact amyloid deposits. F(ab) 2 fragments are unable to produce effective reductions in compact amyloid deposits. Microglial expression of Fc receptors and CD45 is increased following systemic antibody administration. Phospho-p44/42 expression is also increased. Microglia are activated around vascular amyloid deposits. Direct dissolution Diffuse amyloid deposits are reduced early with no associated microglial activation. Diffuse amyloid deposits are removed regardless of microglial inhibition or F(ab) 2 fragments. Peripheral sink 100-fold increase in circulating A levels following systemic anti-A antibody administration. Increased vascular amyloid could potentially be a result of the peripheral sink mechanism.

PAGE 192

183 REFERENCES Akiyama H, Barger S, Barnum S, Bradt B, Bauer J, Cole GM, Cooper NR, Eikelenboom P, Emmerling M, Fiebich BL, Finch CE, Frauts chy S, Griffin WST, Hampel H, Hull M, Landreth G, Lue LF, Mrak M, Mackenzie IR, McGeer PL, OBanion K, Pachter J, Pasinetti G, Plata-Salaman C, Rogers J, R ydel R, Shen Y, Streit W, Strohmeyer R, Tooyoma I, Van Muiswinkel FL, Veerhuis R, Walker D, Webster S, Wegrzyniak B, Wenk G, Wyss-Coray T. 2000. Inflammation and Alzheimers disease. Neurobiol. Aging 21: 383-421. Aloisi F. 2001. Immune function of microglia. Glia 36: 165-179. Andorfer CA, Acker CM, Kress Y, Duff K, Davies P. 2004. Neuronal deatch acoompanied by re-expression of cell cycle re gulatory proteins in a mouse model of neurofibrillary tangle formation. Progr am No. 602.3. Society for Neuroscience 34 th Annual Meeting, San Diego, CA. Andorfer C, Kress Y, Espinoza M, de Silva R, Tucker KL, Barde YA, Duff K, Davies P. 2003. Hyperphosphorylation and aggregation of tau in mice expressing normal human tau isoforms. J Neurochem 86: 582-590. Arendash GW, King DL, Gordon MN, Morgan D, Hatcher JM, Hope CE, Diamond DM. 2001. Progressive, age-related behavioral impairments in transgenic mice carrying both mutant amyloid precursor protein and pres enilin-1 transgenes. Brain Res. 891:42-53. Bacskai BJ, Kajdasz ST, McLellan ME, Game s D, Seubert P, Schenk D, Hyman BT. 2002. Non-Fc-mediated mechanisms are i nvolved in cleara nce of amyloidin vivo by immunotherapy. J Neurosci 22: 7873-7878 Bacskai BJ, Kajdasz ST, Christie RH, Carter C, Games D, Seubert P, Schenk D, Hyman BT. 2001. Imaging of amyloiddeposits in brains of living mice permits direct observation of clearance of plaques with immunotherapy. Nat Med 7: 369-372. Banks WA, Terrell B, Farr SA, Robinson SM, Nonaka N, Morley JE. 2002. Passage of amyloid protein antibody across the blood-br ain barrier in a mouse model of Alzheimers disease. Peptides 23: 2223-2226.

PAGE 193

184 Bard F, Barbour R, Cannon C, Carretto R, Fo x M, Games D, Guido T, Hoenow K, Hu K, Johnson-Wood K, Khan K, Kholodenko D, Lee C, Lee M, Motter R, Nguyen M, Reed A, Schenk D, Tang P, Vasquez N, Seubert P, Yednock T. 2003. Epitope and isotype specificities of antibodies to beta -amyloid peptide for protection against Alzheimer's disease-like neuropathology. Proc Natl Acad Sci USA 100:2023-2028. Bard F, Cannon C, Barbour R, Burke RL, Games D, Grajeda H, Guido T, Hu K, Huang J, Johnson-Wood K, Khan K, Kholofenko D, Lee M, Lieberburg I, Motter R, Nguyen M, Soriano F, Vasquez N, Weiss K, Welch B, Seubert P, Schenk D, Yednock T. 2000. Peripherally administered antibodies against amyloid -peptide enter the central nervous system and reduce pathology in a mouse model of Alzheimers disease. Nat Med 6: 916919. Borchelt DR, Ratovitski T, van Lare J, Lee MK, Gonzales V, Jenkins NA, Copeland NG, Price DL, Sisodia SS. 1997. Accelerated amyloi d deposition in the brains of transgenic mice coexpressing mutant presenilin 1 and amyloid precursor proteins. Neuron 19: 939945. Bowers WJ, Federoff HJ. 2002. Amyl oid immunotherapy-engendered CNS inflammation. Neurobiol Aging 23: 675-676. Cadavid D, Mena H, Koeller K, Frommelt RA. 2000. Cerebral beta amyloid angiopathy is a risk factor for cerebral ischemic in farction. A case contro l study in human brain biopsies. J Neuropathol Exp Neurol 59: 768-773. Castellani RJ, Smith MA, Perry G, Friedl and RP. 2004. Cerebral amyloid angiopathy: major contributor or decorative response to Alzheimer's disease pathogenesis. Neurobiol Aging 25:599-602. Chang JY, Liu LZ. 2000. Inhibition of microglial nitric oxide production by hydrocortisone and glucocorticoid pr ecursors. Neurochem Res 25(7): 903-908. Citron M, Eckman CB, Diehl TS, Corcoran C, Ostaszewski BL, Xia W, Levesque G, St George Hyslop P, Younkin SG, Selkoe DJ. 1998. Additive effects of PS1 and APP mutations on secretion of the 42-residue am yloid beta-protein. Ne urobiol Dis 5: 107-116. Das P, Howard V, Loosbrock N, Dickson D, Murphy MP, Golde TE. 2003. Amyloid-beta immunization effectively reduces amyloid de position in FcRgamma-/knock-out mice. J Neurosci 23:8532-5838. Das P, Murphy MP, Younkin LH, Younkin SG Golde TE. 2001. Reduced effectiveness of Abeta1-42 immunization in APP transgenic mice with significant amyloid deposition. Neurobiol Aging 22:721-727.

PAGE 194

185 DeMattos RB, Boone LI, Hepburn DL, Parsadan ian M, Bryan MT, Ness DK, Piroozi KS, Holzman DM, Bales KR, Gitter BD, Paul SM Racke M. 2004. In vitro and in vivo characterization of beta-amyloid antibodies binding to cerebral amyloid angiopathy (CAA) and the selective exacerbation of C AA-associated microhemorrhage. Neurobiol Aging 25: S577. DeMattos RB, Bales KR, Cummins DJ, D odart JC, Paul SM, Holtzman DM. 2001. Peripheral anti-A antibody alters CNS and plasma A clearance and decreases brain A burden in a mouse model of Alzheimers dis ease. Proc. Natl. Acad. Sci. USA 98: 88508855. DeWitt DA, Perry G, Cohen M, Doller C, S ilver J. 1998. Astrocytes regulate microglial phagocytosis of senile plaque cores of Alzh eimers disease. Expt Neurol 149: 329-340. Diamond DM, Park CR, Heman KL, Rose GM. 1999. Exposing rats to a predator impairs spatial working memory in the radial arm water maze. Hippocampus 9: 542-552. DiCarlo G, Wilcock D, Henderson D, Gor don M, and Morgan D. 2001. Intrahippocampal LPS injections reduce A load in APP+PS1 transgenic mice. Neurobiology of Aging 22, 1007-1012. Dodart JC, Bales KR, Gannon KS, Greene SJ, DeMattos RB, Mathis C, DeLong CA, Wu S, Wu X, Holtzman DM, Paul SM. 2002. Immuni zation reverses memory deficits without reducing brain burden in Alzheimers disease model. Nat Neuro 5(5): 452-457. Duff K, Knight H, Refolo LM, Sanders S, Yu X, Picciano M, Malester B, Hutton M, Adamson J, Goedert M, Burki K, Davies P. 2000. Characterization of pathology in transgenic mice over-expressing human genomi c and cDNA tau transgenes. Neurobiol Dis 7:87-98. Duff K, Eckman C, Zehr C, Yu X, Prada CM Perez-tur J, Hutton M, Buee L, Harigaya Y, Yager D, Morgan D, Gordon MN, Holcomb L, Refolo L, Zenk B, Hardy J, Younkin S. 1996. Increased amyloid-beta42(43) in brains of mice expressing mu tant presenilin 1. Nature 383: 710-713. Ferrer I, Boada Rovira M, Sanchez Gu erra ML, Rey MJ, Costa-Jussa F. 2004. Neuropathology and pathogenesis of encephali tis following amyloidbeta immunization in Alzheimer's disease. Brain Pathol 14:11-20. Games D, Adams D, Alessandrini R, Barbour R, Berthelette P, Blackwell C, Carr T, Clemens J, Donaldson T, Gillespie F, Guido T, Hagopian S, Johnson-Wood K, Khan K, Mike Lee, Paul Leibowitz, Ivan Lieberburg, Sheila Little, Eliezer Masliah, Lisa McConlogue, Montoya-Zavala M, Mucke L, Paganini L, Penniman E, Power M, Schenk D, Seubert P, Snyder B, Soriano F, Tan H, Vitale J, Wadsworth S, Wolozin

PAGE 195

186 B, Zhao J. 1995. Alzheimer-type neuropathology in transgenic mice overexpressing V717F beta-amyloid precursor protein. Nature 373: 523-527. Gerlai R. 2001. Behavioral tests of hippocampal function: simple paradigms complex problems. Behav Brain Res. 125:269-277. Gordon, M.N. King D.L., Diamond D.M., Jantze n P.T., Boyett K.V., Hope C.E., Hatcher J.M., DiCarlo G., Gottschall W.P., Morg an D., Arendash G.W. 2001. Correlation between cognitive deficits and Abeta deposits in transgenic APP+PS1 mice. Neurobiol. Aging 22:377-385. Gordon, M.N. Holcomb L.A., Jantzen P.T., DiCarlo G., Wilcock D., Boyett K.W., Connor K., Melachrino J., OCallaghan J.P., Morgan D. 2002. Time course of the development of Alzheimer-like pathology in the doubly transgen ic PS1+APP mouse. Exp. Neurol. 173: 183-195. Grabowski TJ, Cho HS, Vonsattel JP, Reb eck GW, Greenberg SM. 2001. Novel amyloid precursor protein mutation in an Iowa family with dementia and severe cerebral amyloid angiopathy. Ann Neurol 49:697-705. Hardy J. 1997. Amyloid, the presenilins and Alzheimers disease. TiNS 4: 154-159. Hardy JA, Higgins GA. 1992. Alzheimers di sease: the amyloid cascade hypothesis. Science 256: 184-185. Hauss-Wegrzyniak B., Willard L.B., Del Soldato P., Pepeu G., Wenk G.L. 1999. Peripheral administration of novel anti-infla mmatories can attenua te the effects of chronic inflammation within th e CNS. Brain Res. 815: 36-43. He Y., Appel S., Le W. 2001. Minocycline i nhibits microglial activation and protects nigral cells after 6-hydroxydopamine injection into mouse striatum. Brain Res. 909: 187193. Hebert, LE; Scherr, PA; Bienias, JL; Benne tt, DA; Evans, DA. 2003. Alzheimer Disease in the U.S. Population: Prevalen ce Estimates Using the 2000 Census. Archives of Neurology 60: 1119 1122. Hensley K, Floyd RA, Zheng NY, Nael R, Robinson KA, Nguyen X, Pye QN, Stewart CA, Geddes J, Markesbery WR, Patel E, Johnson GV, Bing G. 1999. p38 kinase is activated in the Alzheimers di sease brain. J Neurochem 72: 2053-2058. Herber DL, Roth LM, Wilson D, Mason JE, Morgan D, Gordon MN. 2004. Timedependent reduction in A levels after intracranial LPS administration in APP transgenic mice. Exp Neurol 190: 245-253.

PAGE 196

187 Herzig MC, Winkler DT, Burgermeister P, Pf eifer M, Kohler E, Schmidt SD, Danner S, Abramowski D, Sturchler-Pierrat C, Burk i K, van Duinen SG, Maat-Schieman ML, Staufenbiel M, Mathews PM, Ju cker M. 2004. Abeta is targeted to the vasculature in a mouse model of hereditary cer ebral hemorrhage with amyloidosis. Nat Neurosci 7: 954960. Hobbie S, Chen LM, Davis RJ, Galan JE 1997. Involvement of mitogen-activated protein kinase pathways in the nuclear re sponses and cytokine production induced by Salmonella typhimurium in cultured intest inal epithelial cell s. J Immunol 159: 55505559. Hock C, Konietzko U, Streffer JR, Tracy J, Signorell A, Muller-Tillmanns B, Lemke U, Henke K, Moritz E, Garcia E, Wollmer MA, Umbricht D, de Quervain DJ, Hofmann M, Maddalena A, Papassotiropoulos A, Nitsch RM. 2003. Antibodies against beta-amyloid slow cognitive decline in Alzheimer's disease. Neuron 38: 547-554. Hock C, Konietzko U, Papassotiropoulos A, Wollmer A, Streffer J, von Rotz RC, Davey G, Moritz E, Nitsch RM. 2002. Generation of antibodies specific fo r beta-amyloid by vaccination of patients with Alzh eimer disease. Nat Med. 8:1270-1275. Holcomb L, Gordon MN, McGowan E, Yu X, Be nkovic S, Jantzen P, Wright K, Saad I, Mueller R, Morgan D, Sanders S, Zehr C, OCampo K, Hardy J, Prada CM, Eckman C, Younkin S, Hsiao K, Duff K. 1998. Accelerated Alzheimer-type phenotype in transgenic mice carrying both mutant amyloid precursor protein and presenilin 1 transgenes. Nat Med 4(1): 97-100. Hsiao K, Chapman P, Nilsen S, Eckman C, Harigaya Y, Younkin S, Yang F, Cole G. 1996. Correlative memory deficits, Abeta elevat ion, and amyloid plaques in transgenic mice. Science 274: 99-102. Jantzen P.T., Connor K.E., DiCarlo G., Wenk G.L., Wallace J.L., Rojiani A.M., Coppola D., Morgan D., Gordon M.N. 2002. Microglia l activation and beta amyloid deposit reduction caused by a nitric oxide-releasi ng nonsteroidal anti-inflammatory drug in amyloid precursor protein plus presenilin-1 transgenic mice. J. Neurosci. 22: 2246-2254. 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., George-Hyslop P., Westaway D. 2000. A beta peptide immunization reduces behavioural impairment and plaques in a mode l of Alzheimers disease. Nature 408: 979982. Jellinger KA. 2002. Alzheimer disease and cerebrovascular pathology: an update. J Neural Transm. 109: 813-836. Justement LB. 1997. The role of CD45 in signal transduction. Adv Immunol 66: 1-65.

PAGE 197

188 Kalaria RN, Ballard C. 1999. Overlap betwee n pathology of Alzheimer disease and vascular dementia. Alzheimer Dis Assoc Disord. 13: S115-S123. Kawarabayashi T, Younkin LH, Saido TC, Shoji M, Hsiao-Ashe K, Younkin SG. 2001. Age-dependent changes in brai n, CSF, and plasma amyloid protein in the Tg2576 transgenic mouse model of Alzheimers disease. J Neur osci 21(2): 372-381. Klein WL, Krafft GA, Finch CE. 2001. Targeting small oligomers: the solution to an Alzheimers disease con undrum. TiNS 24(4): 219-224. Koistinaho M, Kettunen MI, Goldsteins G, Kein anen R, Salminen A, Ort M, Bures J, Liu D, Kauppinen RA, Higgins LS, Koistinaho J. 2002. Beta-amyloid precursor protein transgenic mice that harbor diffuse A beta deposits but do not form plaques show increased ischemic vulnerability: role of inflammation. Proc Natl Acad Sci U S A 99: 1610-1615. Koistinaho M and Koistinaho J. 2002. Role of p38 and p44/42 Mitogen-Activated Protein Kinases in microglia. Glia 40: 175-183. Kotilinek LA, Bacskai B, Westerman M, Kawarabayashi T, Younkin L, Hyman BT, Younkin S, Ashe KH. 2002. Reversible memory loss in a mouse transgenic model of Alzheimer's disease. J Neurosci 22:6331-6335. Koumenis IL, Shahrokh Z, Leong S, Hsei V, Deforge L, Zapata G. 2000. Modulating pharmacokinetics of an an ti-interleukin-8 F(ab) 2 by amine-specific PEGylation with preserved bioactivity. Int J Pharm 198: 83-95. Lewis J, Dickson DW, Lin WL, Chisholm L, Corral A, Jones G, Yen SH, Sahara N, Skipper L, Yager D, Eckman C, Hardy J, Hutton M, McGowan E. 2001. Enhanced neurofibrillary degeneration in transgenic mice expressing mutant tau and APP. Science 293: 1487-1491. Lewis J, McGowan E, Rockwood J, Melros e H, Nacharaju P, Van Slegtenhorst M, Gwinn-Hardy K, Paul Murphy M, Baker M, Yu X, Duff K, Hardy J, Corral A, Lin WL, Yen SH, Dickson DW, Davies P, Hutton M. 2000. Neurofibrillary tangles, amyotrophy and progressive motor disturbances in mice expressing mutant (P301L) tau protein. Nat Genet 25(4): 402-405. Liu B, Hong JS. 2003. Role of microglia in inflammation-mediated neurodegenerative diseases: Mechanisms and strategies for ther apeutic intervention. J Pharm Expt Ther 304: 1-7.

PAGE 198

189 Lombardo JA, Stern EA, McLellan ME, Kajdasz ST, Hickey GA, Bacskai BJ, Hyman BT. 2003. Amyloid-beta antibody treatment l eads to rapid normalization of plaqueinduced neuritic alterations. J Neurosci 23:10879-10883. Lue LF, Kuo YM, Roher AE, Brachova L, Shen Y, Sue L, Beach T, Kurth JH, Rydel RE, Rogers J. 1999. Soluble amyloid peptide concentration as a predictor of synaptic change in Alzheimers diseas e. Am J Pathol 155(3): 853-862. Quon D, Wang Y, Catalano R, Scardina JM, Murakami K, Cordell B. 1991. Formation of beta-amyloid protein deposits in brains of transgenic mice. Nature. 352: 239-41. McCarron MO, Nicoll JA. 2004. Cerebral amyloi d angiopathy and thrombolysis-related intracerebral haemorrhage. Lancet Neurol 3:484-492. McGeer PL, McGeer EG. 2001. Inflammation, autotoxicity and Alzheimer disease. Neurobiol Aging 22: 799-809. Morgan, D., Diamond D.M., Gottschall P.E., Ug en K.E., Dickey C., Hardy J., Duff K., Jantzen P., DiCarlo G., Wilcock D., Connor K., Hatcher J., Hope C., Gordon M., Arendash G.W. 2000. A beta peptide vaccinat ion prevents memory loss in an animal model of Alzheimers disease. Nature 408: 982-985. Morris RG, Garrud P, Rawlins JN, O'Keefe J. 1982. Place navigation impaired in rats with hippocampal lesions. Nature 297: 681-683. Munch G, Robinson SR. 2002. Potential ne urotoxic inflammatory responses to vaccination in humans. J Neural Transm 109: 1081-1087. Natte R, Maat-Schieman ML, Haan J, Born ebroek M, Roos RA, van Duinen SG. 2001. Dementia in hereditary cerebral hemorrhage with amyloidosis-Dutch type is associated with cerebral amyloid angiopathy but is i ndependent of plaques and neurofibrillary tangles. Ann Neurol.50:765-772. Nicoll JA, Wilkinson D, Holmes C, Steart P, Markham H, Weller RO. 2003. Neuropathology of human Alzheimer disease after immunization with amyloid-beta peptide: a case report. Nat Med 9:448-452. Oddo S, Caccamo A, Shepherd JD, Murphy MP, Golde TE, Kayed R, Metherate R, Mattson MP, Akbari Y, LaFerla FM. 2003. Tr iple-transgenic model of Alzheimer's disease with plaques and tangles: intracellu lar Abeta and synaptic dysfunction. Neuron 39: 409-421. Orgogozo JM, Gilman S, Dartigues JF, Lauren t B, Puel M, Kirby LC, Jouanny P, Dubois B, Eisner L, Flitman S, Michel BF, Boada M, Frank A, Hock C. 2003. Subacute

PAGE 199

190 meningoencephalitis in a subset of patie nts with AD after Abeta42 immunization. Neurology 61: 46-54. Paresce DM, Ghosh RN, Maxfield FR. 1996. Microglial cells internal ize aggregates of the Alzheimers disease amyloid -protein via a scavenger receptor. Neuron 17: 553-565. Patrick GN, Zukerberg L, Nikolic M, de la Monte S, Dikkes P, Tsai LH. 1999. Conversion of p35 to p25 deregulates Cdk5 activity and promotes neurodegeneration. Nature 402: 615-622. Pfeifer M, Boncristiano L, Bondolfi L, Stalde r A, Deller T, Staufenbiel M, Mathews PM, Jucker M. 2002. Cerebral hemorrhage after passive anti-A immunotherapy. Science 298: 1379. Radaev S and Sun P. 2001. Recognition of immunoglobulins by Fc receptors. Mol Immunol 38: 1073-1083. Ravetch JV and Kinet JP. 1991. Fc Receptors. Annu Rev Immunol 9: 457-492. Rensink AA, de Waal RM, Kremer B, Ve rbeek MM. 2003. Pathogenesis of cerebral amyloid angiopathy. Brai n Res Rev: 43:207-223. Rogers J, Strohmeyer R, Kovelowski CJ Li R. 2002. Microglia and inflammatory mechanisms in the clearance of amyloid peptide Glia 40: 260-269. 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., Games D., Seubert P. 1999. Immu nization with amyloid-beta attenuates Alzheimer-disease-like pathology in the PDAPP mouse. Nature 400: 173-177. Schimmer B.P., Parker K.L. 2001. Adrenoc orticotrophic hormone; adrenocortical steroids and their synthetic analogs; inhi bitors of the synthesis and actions of adrenocortical hormones. In: Goodman & G ilmans The Pharmaocological Basis of Therapeutics. (Hardman J.G., Limbird L.E., Gilman A.G. ed.). pp1649-1677. McGraw Hill Medical Publishing Division. Selkoe DJ. 2001. Alzheimer's disease: genes, proteins, and therapy. Physiol Rev 81: 741766. Soininen HS, Riekkinen PJ Sr. 1996. Apolipoprot ein E, memory and Alzheimer's disease. TiNS 19:224-228

PAGE 200

191 Solomon B., Koppel R., Hanan E., Katzav T. 1996. Monoclonal antibodies inhibit in vitro fibrillar aggregation of the Alzheimer beta-amyloid peptide. Proc Natl Acad Sci USA.93: 452-455. Solomon B, Koppel R, Frenkel D, Hanan-Ah aron E. 1997. Disaggregation of Alzheimer beta-amyloid by site-directed monoclonal an tibodies. Proc. Natl. Acad. Sci. USA 94: 4109-4112. Song X, Tanaka S, Cox D, Lee SC. 2004. Fc gamma receptor signaling in primary human microglia: differential roles of PI-3K and Ra s/ERK MAPK pathways in phagocytosis and chemokine induction. J Leukoc Biol 75: 1147-1155. Stalder M, Deller T, Staufenbiel M, Jucker M. 2001. 3D-reconstructio n of microglia and amyloid in APP23 transgenic mice: no evidence of intracellular amyloid. Neurobiol Aging 22: 427-434. Streit WJ. 2002. Microglia as neuroprotective, immunocompetent cells of the CNS. Glia 40: 133-139. Streit WJ, Walter SA, Pennell NA. 1999. Reac tive microgliosis. Prog Neurobiol 57: 653681. Strittmatter WJ, Roses AD.1996. Apolipoprotei n E and Alzheimer's disease. Annu Rev Neurosci. 19: 53-77. Strittmatter WJ, Saunders AM, Goedert M, Weisgraber KH, Dong LM, Jakes R, Huang DY, Pericak-Vance M, Schmechel D, Roses AD. 1994. Isoform-specific interactions of apolipoprotein E with microtubule-associate d protein tau: implic ations for Alzheimer disease. Proc Natl Acad Sci USA 91: 11183-11186. Tanaka J, Fujita H, Matsuda S, Toku K, Sakanaka M, Maeda N. 1997. Glucocorticoidand mineralocorticoid receptors in micr oglial cells: The two receptors mediate differential effects of cortic osteroids. Glia 20: 23-37. Tikka T., Fiebich B.L., Goldsteins G., Ke inanen R., Koistinaho J. 2001. Minocycline, a tetracycline derivative, is neuroprotective agai nst excitotoxicity by inhibiting activation and proliferation of microgl ia. J. Neurosci. 21: 2580-2588. Tucker KL, Meyer M, Barde YA. 2001. Neurot rophins are required for nerve growth during development. Nat Neurosci 4:29-37. Vanden Berghe W, Plaisance S, Boone E, De Bosscher K, Schmitz ML, Fiers W, Haegeman G. 1998. p38 and extr acellular signal-re gulated kinase mitogen-activated protein kinase pathways ar e required for nuclear factor -kappaB p65 transactivation mediated by tumor necrosis factor. J Biol Chem 273: 3285-3290.

PAGE 201

192 Weggen S, Eriksen JL, Das P, Sagi SA, Wang R, Pletrzik CU, Findlay KA, Smith TE, Murphy MP, Butler T, Kang DE, Marquez-Sterling N, Golde TE, Koo EH. 2001. A subset of NSAIDs lower amyloidogenic 42 independently of cyclooxygenase activity Nature 414: 212-216. Weir ANC, Nesbitt A, Chapman AP, Popplew ell AG, Antoniw P, Lawson ADG. 2002. Formatting antibody fragments to mediate spec ific therapeutic functions. Biochem Soc Trans 30: 512-516. Wilcock DM, Gordon MN, Ugen KE, Gottschall PE, DiCarlo G, Dickey C, Boyett KW, Jantzen PT, Connor KE, Melachrino J, Har dy J, Morgan D. 2001. Number of Abeta inoculations in APP+PS1 transgenic mi ce influences antibody titers, microglial activation, and congophilic plaque levels. DNA Cell Biol 20: 731-736. Wilson CJ, Finch CE, Cohen HJ 2002. Cytokines and cognition The case for a head-totoe inflammatory paradigm. JAGS 50: 2041-2056. Wyss-Coray T, Yan F, Lin AH, Lambris JD, Alexander JJ, Quigg RJ, Masliah E. 2002. Prominent neurodegeneration and increased pl aque formation in complement-inhibited Alzheimers mice. Proc Natl Acad Sci USA 99(16): 10837-10842. Wyss-Coray T, Lin C, Yan F, Yu GQ, Rohde M, McConlogue L, Masliah E, Mucke L. 2001. TGF-beta1 promotes microglial amyloid-beta clearance and reduces plaque burden in transgenic mice. Nat Med 7(5): 612-618. Xioa YQ, Malcolm K, Worthen GS, Gardai S, Schiemann WP, Fadok VA, Bratton DL, Henson PM. 2002. Cross-talk between ER K and p38 MAPK mediates selective suppression of pro-inflammatory cyt okines by transforming growth factor J Biol Chem 277: 14884-14893. Yamada M, Sodeyama N, Itoh Y, Otomo E, Matsushita M, Mizusawa H. 2002. No association of paraoxonase genotype or athero sclerosis with cerebra l amyloid angiopathy. Stroke 33:896-900. Yrjanheikki J., Keinanen R ., Pelikka M., Hokfelt T., Ko istinaho J. 1998. Tetracyclines inhibit microglial activation a nd are neuroprotective in global brain ischemia. Proc. Natl. Acad. Sci. USA. 99: 10837-10842. Yu P, Oberto G. 2000. Alzheimers disease: Transgenic mouse models and drug assessment. Pharm Res 42(2):107-114. Zhong B, Jiang K, Gilvary DL, Epling-Burnette PK, Ritchey C, Liu J, Jackson RJ, HongGeller E, Wei S. 2003. Human neutrophils utilize a Rac/Cdc42-dependent MAPK

PAGE 202

193 pathway to direct intracellular granule mobili zation toward ingested microbial pathogens. Blood 101: 3240-3248.

PAGE 203

194 APPENDICES

PAGE 204

195 APPENDIX A: NUMBER OF A INOCULATIONS IN APP+PS1 TRANSGENIC MICEINFLUENCES ANTIBODY TITERS, MICROGLIAL ACTIVATION AND CONGOPHILIC PLAQUE LEVELS. Donna M. Wilcock 1 Marcia N. Gordon 1 Kenneth E. Ugen 2 Paul E. Gottschall 1 Giovanni DiCarlo 1 Chad Dickey 1 Kristal W. Boyett 1 Paul T. Jantzen 1 Karen E. Connor 1 Jason Melachrino 1 John Hardy 3 David Morgan 1 *. 1 Alzheimers Research Laboratory, Department of Pharmacology, University of South Florida, Tampa, Florida 33612, USA. 2 Department of Medical Microbiology and Immunology, University of South Florida, Tampa, Florida 33612, USA. 3 Mayo Clinic Jacksonville, Jacksonville, Florida. This work was published in DNA and Cell Biology 2001 20(11): 731-736. Acknowledgements: This work was supported by AG 15490 (MNG) and AG 18478 (DM). We thank Karen Hsiao for providing the APP mice and Ka ren Duff for providing the PS1 mice.

PAGE 205

196 APPENDIX A (continued) Abstract There have been several reports on the use of beta-amyloid (A ) vaccination in different mouse models of Alzheimers di sease (AD) and its e ffects on pathology and cognitive function. In this re port histopathology of the A PP+PS1 doubly transgenic mice is compared after 3, 5 or 9 A inoculations. The number of inoculations influenced the effects of vaccination on Congo red levels, microglia activation and anti-A antibody titers. After 3 inoculations, the antibody titer of transgenic mice was substantially lower than that found in nontransgenic animals. Howe ver, after 9 inoculations, the levels were considerably higher in both genotypes, and no longer disc riminable statistically. The number of inoculations in fluenced CD45 expression, an indicator of microglial activation. There was an initia l up-regulation, which was signif icant after 5 inoculations, but by 9 inoculations, microglial activati on was equivalent to mice given control vaccinations. Along with this increased CD45 expression there was a correlative reduction in Congo red staining, which stains compact plaques. When mice from all groups were combined, there was a signi ficant correlation betw een activation of microglia and Congo red levels, suggesting that microglia play a role in clearance of compact plaque. Introduction Alzheimers disease (AD) is a prog ressive, neurodegenerative disorder characterized by accumulation of senile plaques consisting of beta-amyloid (A ) protein of which there are 2 forms, A 1-40 and A 1-42 This is thought to be the key step in the pathogenesis of Alzheimers disease (Selkoe, 1 991). The disorder is also characterized by

PAGE 206

197 APPENDIX A (continued) the formation of neurofibrillar y tangles consisting of tau pr otein, and by the initiation and proliferation of a brain-specific inflammatory response (Akiyama et al, 2000). Transgenic mouse models of Alzheimers disease have b een an invaluable source of information regarding the pathological pr ogression of AD and a vehicle in which to test possible therapeutic interventions. Here we use a transgenic mouse model of AD carrying two transgenes; amyloid precursor protein (APP) a nd presenilin-1 (PS1) previously described (Duff et al, 1996, Hsiao et al, 1996, Holcom b et al, 1998, 1999, Gordon et al, 2001a, Gordon et al, 2001b). Schenk et al initially described the effects of A 1-42 immunization in the PDAPP mouse. In a report published in 1999 they demonstrated the ability of their vaccination regimen to reduce A deposits in the brain. More recently, A vaccination has also been shown to prevent the cognitive decline in some transgenic mice (including the APP+PS1) in addition to reducing A load (Morgan et al, 2000, Janus et al, 2000). The data presented in this report examines the effect of increasing numbers of A 1-42 immunizations on the pathology of the APP+PS1 mouse, specifically, anti-A antibody titers, the reduction in c ongophilic plaque load and th e activation of microglia. Of particular note in this study was the observed activation of microglial cells, which are central to the inflammatory proce ss in AD, along with a concomitant reduction in congophilic plaque.

PAGE 207

198 APPENDIX A (continued) Materials and Methods Vaccination Protocol. Tg2576 APP mice (Hsiao et al, 1996) were bred with PS1 line 5.1 mice (Duff et al 1996) to obtain the double transgenic mice. These mice were then randomly assigned to groups receiving vaccinat ion with either human A 1-42 peptide (Bachem) or with keyhole limpet hemocyanin (KLH) as describe d previously (Morgan et al, 2000). Briefly, 100 g of A 1-42 or KLH were dissolved in water at 2.2 mg/ml, mixed with PBS and incubated overnight. One day later this su spension was mixed with Freunds adjuvant (complete for the primary inoculation, incomple te for the next 4 inoculations, and mineral oil for the remaining inoculations). Three va ccination groups were used. The first group was administered 3 inoculations starting at an average age of 13 months. These mice were killed at an average of 16 months of age, 13 days after the final inoculation. A second group of mice were given 5 inoculati ons starting at an average age 14.5 months. These mice were killed at an average of 19.75 months of age, 10 days after the last inoculation. A third group given 9 inoculations started at an average age of 7.5 months. They were killed at an average of 16.25 months of age, 6 week s after the last inoculation. In addition, nontransgenic mice were vaccinated with A peptide in the 3 and 9 inoculation groups. Antibody titers were meas ured by ELISA as described previously (Morgan et al, 2000, Dickey et al, 2001). The KLH immunized mice are herein referred to as control mice for these experiments.

PAGE 208

199 APPENDIX A (continued) Histopathology. Mice were overdosed with pentobarbital, perfused with 0.9% saline and the brain removed. The right side of the brain was rapidly dissected over ice and the left side fixed for 24 hours in freshly prepared, buffered 4% paraformaldehyde. Following cryoprotection through increasing concentratio ns of sucrose solutions at 24-hour intervals, frozen sections were ta ken on a sliding microtome at a 25 m thickness and stored in DPBS (Dulbeccos phosphate buffere d saline) with sodium azide to prevent microbial growth. Sections were stained us ing floating immunohistoc hemistry for total A (rabbit antiserum primarily reacting with the N terminal of the A peptide, 1:10000) and CD45 (Serotec, 1:10,000) as described pr eviously (Holcomb et al, 1999, Gordon et al, 1997). Sections were also mounted on slides and stained for Congo red (SigmaAldrich) (Gordon et al, 2001a). The area of hippocampus and frontal cortex occupied by positive stain was measured with a Videomet ric V150 image analysis system (Oncor) on a Nikon Microphot FX microscope as descri bed in detail previously (Gordon et al, 2001a). Percentage area was measured and an alyzed. Data were collected from 8-16 equally spaced horizontal sections. The values for all sec tions from one mouse were averaged to represent a single sample for statistical analysis. Statistical Analysis. Data were first analyzed by comparing A vaccinated and control mice within each group by one-way ANOVA using the Statview software program (SAS). Because of differences in the age of kill, the results from each A vaccinated mouse were normalized to the mean value of their respective control vaccinated group (percent of control) to

PAGE 209

200 APPENDIX A (continued) correct for differences in age at kill and overall staining intensity for each inoculation group. These were then analyzed by perfor ming a simple regression analysis on the Statview program. Results Antibody titers increase with increasi ng numbers of inoculations in both transgenic and nontransgenic mice (Fig 1). Initially, doubly transgenic mice have lower anti-A titers than similarly treated nontransgenic mice. However, after the ninth vaccination the transgenic mice produced an tibody titers that were similar to nontransgenic mice. In general, the Congo red staining in hi ppocampus was reduced as a result of A vaccination (Fig 2A,B). In the group receiving five inoculati ons this reduction was almost 50% and was significantly reduced in relati on to control mice (p < 0.002; Figure 3). In the groups receiving three or nine inoculations, the reduction in Congo red was not statistically significant. The results from front al cortex also follow these trends although no significance is found at any time-points (Table 1). CD45 expression in hippocampus was incr eased almost 2 fold in the hippocampus of mice given 5 inoculations (Figure 2C,D ; Figure 4; p < 0.001). There was a similar trend for CD45 up-regulation in the group re ceiving three inocula tions although not statistically significant. However, the vaccin ated mice were equivalent to the control mice in the group receiving nine inoculations. The same trend was seen in frontal cortex although not to the same degree (Table 1).

PAGE 210

201 APPENDIX A (continued) When data from all three groups are combined, there is a correlation between CD45 expression and Congo red staining in the hippo campus as illustrated in Figure 5. This relationship shows that mice with elevated CD45 expression had less Congo red staining (r=0.691, p=0.002). This result did not occur because of a biased positioning of one of the inoculation groups. By plotting the data from each group with a different symbol one can see that in each inoculation group there are mice with high CD45 staining and other mice with little, which was less than the control average. This bimodal distribution of CD45 staining has been observed in most groups of mice we have examined including the control mice in the present study. Discussion CD45 expression is indicative of microglial activation. Here, we have shown that CD45 expression in transgenic mice administered A vaccination is up-regulated, after 5 inoculations. However, on average, the e xpression was no longer elevated after a 9 th inoculation. This suggests that the microglial activati on resulting from A vaccination is transient. This likely represents a desensitization to the circulating antibodies because; even though the interval between the last inoculation and kill after 9 inoculations was 6 weeks the antibody titers were still high at necr opsy. Perhaps the most interesting data is the high correlation between Congo red staining le vels in hippocampi of transgenic mice and activation of microglia. These data a dd to a growing body of literature suggesting that in transgenic mice activation of microglia leads to clearance of A deposits. Ongoing work from our group shows reduction in amyloi d deposits in associa tion with microglial

PAGE 211

202 APPENDIX A (continued) activation in several circumstances using the transgenic mouse model of amyloid deposition. Administration of a flurbiprofen derivative which sl owly releases nitric oxide (NCX-2216) causes dramatic activation of mi croglia and substantial reduction in Congo red staining (Jantzen et al 2001). In another study, intr ahippocampal injections of lipopolysaccharide, a pro-inflammatory agent, simultaneously reduced A load and activated microglia (Di Carlo et al, 2001). Even in the norm al time course of amyloid accumulation there is a stabi lization of the congophilic depos its in doubly transgenic mice between 12 and 15 months, the age at which microglial activation becomes most pronounced (Gordon et al, 2001b). It has been well demonstrated that micr oglia in culture are readily capable of internalizing A 1-42 aggregates (Paresce et al, 1996; Webster et al, 2001). The data reported here are consistent with several other reports regarding A vaccination and microglial activation. In the original A vaccination report, the vaccine was found to result in activated microglia around the few de posits that did remain (Schenk et al, 1999). Bard et al, 2000, also demonstrated that micr oglia can be activated to clear tissue amyloid deposits by Fc receptor mediated phagocytosis in vitro. In a very direct experiment, Bacskai et al, (2001), demonstrat ed that injection of anti-A antibodies into transgenic mouse brain induced a rapid disappearance of A associated with a florid microglial reaction. These results together with those from our research group indicate that activation of microglia can have benefit in clearing A deposits from the brains of transgenic mice. It is unclear whether exce ssive activation of micr oglia can ultimately cause autotoxic inflammatory reactions in th is model or whether the mouse brain is

PAGE 212

203 APPENDIX A (continued) somehow resistant to the development of such a reaction. It is intriguing that the microglial activation in mice with 9 inoculations had largely s ubsided. This suggests that under these circumstances the microglia can deve lop tolerance to the activating stimuli. If so, vaccination may be one mechanism to, perhaps paradoxically, reduce an autotoxic reaction in the AD brain. The observation that the doubl y transgenic mice are slower to mount an immune response after A vaccination than their non-transgenic counterparts is a significant one. There are several plausible explanations for this impaired antibody response. One explanation is that A is a self-antigen in the transgenic mice (expressing human APP), and thus, they do not mount as significant a humoral response to the injected human A 142. The murine A sequence is slightly different than the human sequence, thus nontransgenic mice would not identify the human A peptide as an au toantigen. Another possibility that the transgenic mice are by so me means immune compromised as is seen in older humans; thus, they are slower to mount a significant immune response. A third explanation is absorp tion of the serum antibodies by circulating A which interferes with antibody binding to the ELISA plate. This mi ght be most evident when the antibody titers are low and antibody concentrations are stoi chiometrically sim ilar to that of A In any event, repeated vaccinations ultimately overc ome this restriction of antibody generation in the transgenic mice. This may have si gnificance for treatment of human populations, with multiple vaccinations required to activate a vigorous antibody response.

PAGE 213

204 APPENDIX A (continued) In conclusion, the data here are cons istent with the argument that the A vaccination results in plaque clearance primarily through activation of microglial cells. Still, we continue to entert ain the possibility of antibodies dissolving plaques directly (Solomon, this volume), or an tibodies binding circulating A reducing the effective plasma A concentration, and increasing the conc entration gradient between brain and blood leading to more rapid A removal from the CNS. Finally, we believe that multiple inoculations are likely to be required if A vaccination demonstrates efficacy in the treatment or prevention of AD.

PAGE 214

205 APPENDIX A (continued) Figure 1: Antibody titer averages as a function of number of inoculations in transgenic and non-transgenic mice. Figure legends: : Non-transgenic, : Transgenic. There is a highly significant difference between transg enic and non-transgenic (p=0.0002) after 3 inoculations, indicated by **, but not by 9 inoculations.

PAGE 215

206 APPENDIX A (continued)

PAGE 216

207 APPENDIX A (continued) Figure 2: A, B: Congo red stain e mice receiving 5 A red in ing in the hippocampus of th inoculations (40X magnificatio n). C, D: CD45 staining, counterstained with Congo the hippocampus of the mice receiving 5 A inoculations (200X ma gnification). A and C: Control immunized mice. B and D: A immunized mice. A and B; scale bar: 250 m. C and D; scale bar: 50 m.

PAGE 217

APPENDIX A (continued) 208

PAGE 218

209 APPENDIX A (continued) Figure 3: Congo red levels in hippocampus relative to number of inoculations for vaccinated and control mice. All mice for each group were averaged and are shown here and are all APP+PS1 transgenic mice. ** Indicates high significance (p<0.002).

PAGE 219

APPENDIX A (continued) 210

PAGE 220

211 APPENDIX A (continued) Figure 4: CD45 expression in hippocampus re lative to number of inoculations for vaccinated and control mice. All mice for each group were averaged and are shown here and are all double transgenic. ** Indicates high significance (p<0.001).

PAGE 221

APPENDIX A (continued) 212

PAGE 222

213 APPENDIX A (continued) Figure 5: Correlation of Congo red levels and CD45 expression both shown as percent control in hippocampus. R=0.69 and p=0.002. Figure legends: : 3 inoculations, : 5 inoculations, : 9 inoculations.

PAGE 223

APPENDIX A (continued) 214

PAGE 224

215 APPENDIX A (continued) Table 1: Effect of number of inoculations on Congo red and CD45 staining in the frontal cortex. Data are percent of area occupied by reaction product, shown as mean SEM. Congo red CD45 Number of inoculations A vaccinated Control A vaccinated Control 3 2.56+/0.22 2.3+/0.18 1.94+/0.33 1.84+/0.45 5 2.44+/0.28 2.99+/0.20 7.0+/0.81 5.56+/0.96 9 1.32+/0.08 1.66+/0.12 4.63+/0.67 4.96+/0.74

PAGE 225

216 APPENDIX A (continued) References Akiyama H, Barger S, Barnum S, Bradt B, Bauer J, Cole GM, Cooper NR, Eikelenboom P, Emmerling M, Fiebich BL, Finch CE, Frau tschy S, Griffin WS, Hampel H, Hull M, Landreth G, Lue L, Mrak R, Mackenzie IR, McGeer PL, OBanion MK, Pachter J, Pasinetti G, Plata-Salaman C, Rogers J, R ydel R, Shen Y, Streit W, Strohmeyer R, Tooyama I, Van Muiswinkel FL, Veerhuis R, Walker D, Webster S, Wegrzyniak B, Wenk G, Wyss-Coray T. 2000. Inflammation a nd Alzheimers disease. Neurobiology of Aging 21: 383-421. Bacskai BJ, Kajdasz ST, Christie RH, Carter C, Games D, Seubert P, Schenk D, Hyman BT. 2001. Imaging of amyloid-beta deposits in brains of living mice permits direct observation of clearance of plaques with immunotherapy. Nature Medicine 7: 369-372. Bard F, Cannon C, Barbour R, Burke RL, Games D, Grajeda H, Guido T, Hu K, Huang J, Johnson-Wood K, Khan K, Kholodenko D, Lee M, Lieberburg I, Motter R, Nguyen M, Soriano F, Vasquez N, Weiss K, Welch B, Seubert P, Schenk D, Yednock T. 2000. Peripherally administered antibodies against amyloid -peptide enter the central nervous system and reduce pathology in a mouse model of Alzheimers disease. Nature Medicine 6: 916-919. Cotman CW, Tenner AJ, Cummings BJ. 1996. Beta-Amyloid converts an acute phase injury response to chronic injury responses. Neurobiology of Aging 17: 723-731. Di Carlo G, Wilcock D, Henderson D, Go rdon M, Morgan D. 2001. Intrahippocampal LPS injections reduce A load in APP+PS1 transgenic mice. Neurobiology of Aging: In Press. Dickey CA, Morgan DG, Kudchodkar S, Weiner DB, Gordon MN, Ugen KE. 2001. Duration and specificity of humoral immune response in mice vaccinated with the Alzheimers disease associated amyloid 1-42 peptide. This edition. Duff K, Eckman C, Zehr C, Yu X, Prada CM, Perez-Tur J, Hutton M, Buee L, Harigaya Y, Yager D, Morgan D, Gordon MN, Holcomb L, Refelo L, Zenk B, Hardy J, Younkin S. 1996. Increased amyloid-beta 42 (43) in brains of mice expressing mutant presenilin 1. Nature 383: 710-713. Gordon MN, King DL, Diamond DM, Jantzen PT Boyett KW, Hope CE, Hatcher JM, Di Carlo G, Gottschall PE, Morgan D, Arendash GW. 2001a. Correlation between cognitive deficits and A deposits in transgenic APP+PS1 mice. Neurobiology of Aging 22: 377385.

PAGE 226

217 APPENDIX A (continued) Gordon MN, Holcomb LA, Jantzen P, Di Carlo G, Wilcock D, Boyett KW, Connor K, Melachrino J, OCallaghan JP Morgan D. 2001b. Time course of the development of Alzheimer-like pathology in the doubly tr ansgenic PS1+APP mouse. Experimental Neurology: In Press. Gordon MN, Schreier WA, Ou X, Holc omb LA, Morgan DG. 1997. Exaggerated astrocyte reactivity after nigrostriatal dea fferentation in the aged rat. Journal of Comparative Neurology 388: 106-119. Holcomb L, Gordon MN, McGowan E, Yu X, Be nkovic S, Jantzen P, Wright K, Saad I, Mueller R, Morgan D, Sanders S, Zehr C, OCampo K, Hardy J, Prada CM, Eckman C, Younkin S, Hsiao K, Duff K. 1998. Accelerated Alzheimer-type phenotype in transgenic mice carrying both mutant amyloid precursor protein and presenilin 1 transgenes. Nature Medicine 4: 97-100. Holcomb LA, Gordon MN, Jantzen P, Hsiao K, Duff K, Morgan D. 1999. Behavioural changes in transgenic mice expressing both amyloid precursor protein and presenilin-1 mutations: lack of association with amyloid deposits. Behavioural Genetics 29: 177-185. Hsiao K, Chapman P, Nilsen S, Eckman C, Harigaya Y, Younkin S, Young F, Cole G. 1996. Correlative memory deficits, Abeta elevat ion, and amyloid plaques in transgenic mice. Science 274: 99-102. Jantzen P, Connor KE, Di Carlo G, Wenk GL, Morgan DG, Gordon MN. 2001. NCX2216, a nitro-NSAID, causes A reduction and microglial activation in transgenic mouse brain. In preparation. Janus C, Pearson J, McLaurin J, Matthews PM, Jiang Y, Schmidt SD, Chisti MA, Horne P, Heslin D, French J, Mount HTJ, Nixon PA, Mercken M, Bergeron C, Fraser PE, St George-Hislop P, Westaway D. 2000. A peptide immunization reduces behavioural impairment and plaques in a model of Alzheimers disease. Nature 408 : 979-982. McGeer PL, McGeer EG. 1995. The inflammatory response system of brain: implications for therapy of Alzheimer and other neurodege nerative diseases. Brain Research Reviews 21: 195-218. Morgan D, Diamond DM, Gottschall PE, Ugen KE, Dickey C, Hardy J, Duff K, Jantzen P, Di Carlo, G, Wilcock D, Connor K, Hatc her J, Hope C, Gordon MN, Arendash GW. 2000. A peptide vaccination prevents memory lo ss in an animal model of Alzheimers disease. Nature 408: 982-985.

PAGE 227

218 APPENDIX A (continued) Paresce DM, Ghosh RN, Maxfield FR. 1996. Micr oglial cells internalize aggregates of the Alzheimers disease Amyloid -protein via a scavenger receptor. Neuron 17: 553565. Rogers J, Webster S, Lue L, Brachova L, Ci rin WH, Emmerling M, Sh ivers B, Walker D, McGeer P. 1996. Inflammation and Alzheimers disease pathogenesis. Neurobiology of Aging 17: 681-686. 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, Games D, Seubert P. 1999. Immunization with amyloid-beta attenuates Alzheimerdisease-like pathology in the PDAPP mouse. Nature 400: 173-177. Selkoe DJ. 1991. The molecular pathology of Alzheimers disease. Neuron 6: 497-498. Solomon B. 2001. Immunotherapeut ic strategies towards prevention and treatment of Alzheimers Disease. This edition.

PAGE 228

219 APPENDIX B Figure 1: Total A is significantly reduced by anti-A antibodies regardless of their epitope. The graph shows the ratio of right inje cted side: left uninje cted side. Mice were injected with one of anti-drosophila amnesiac antibodies (AMN), anti-A 1-16 antibodies (A-16), anti-A 12-28 antibodies (A-28) or anti-A 28-40 antibodies (A-40). Solid bars indicate values for the frontal cortex while open bars indicate values for the hippocampus. *indicates P<0.05, **indicates P< 0.01 when compared to both uninjected and AMN groups.

PAGE 229

APPENDIX B (continued) 220

PAGE 230

221 APPENDIX B (continued) Figure 2:Thioflavine-S staining is significantly reduced by anti-A antibodies regardless of their epitope. The graph show s the ratio of right injected side: left uninjected side. Mice were injected with one of anti-dr osophila amnesiac antibodies (AMN), anti-A 1-16 antibodies (A-16), anti-A 12-28 antibodies (A-28) or anti-A 28-40 antibodies (A40). Solid bars indicate values for the frontal cortex while open bars indicate values for the hippocampus. *indicates P<0.05, **indicates P<0.01 when compared to both uninjected and AMN groups.

PAGE 231

APPENDIX B (continued)

PAGE 232

223 APPENDIX B (continued) Figure 3: CD45 immunohistochemi stry is increased by anti-A 28-40 antibodies. The graph shows the ratio of right injected side: left uninjected side. Mice were injected with one of anti-drosophila amnesiac antibodies (AMN), anti-A 1-16 antibodies (A-16), anti-A 12-28 antibodies (A-28) or anti-A 28-40 antibodies (A-40). Solid bars indicate values for the frontal cortex while open bars indicat e values for the hippocampus. **indicates P<0.01 when compared to both uninjected and AMN groups.

PAGE 233

APPENDIX B (continued) 224

PAGE 234

225 APPENDIX C Figure 1: Phospho-p38 MAPK exression is d ecreased in microglia with increasing duration of anti-A antibody tr eatment. The upper graph shows data for frontal cortex while the lower graph shows data for the hippocampus. All data are shown as percent area occupied by positive stain. The bar indicat es values for APP transgenic mice treated with control antibody anti-AMN for three months. The line indicates values for APP transgenic mice receiving anti-A 28-40 antibodies for 1, 2 and 3 months. *indicates P<0.05.

PAGE 235

APPENDIX C (continued) 226 226

PAGE 236

227 APPENDIX C (continued) Figure 2: Phospho-p38 MAPK immunohistochemistry shows microglial expression around amyloid plaques which is decreased wi th increasing duration of anti-A antibody treatment. Panels A-D show phospho-p38 MA PK immunohsitochemical staining around the hippocampal fissure (F in panel D) and in the dentate gyrus (DG in panel D). APP transgenic mice were treated for 3 months with control IgG (A ) or with anti-A 28-40 IgG for one (B), two (C) or three (D) months Magnification = 100X. Scale bar panel D = 50 m.

PAGE 237

APPENDIX C (continued) 228 228

PAGE 238

229 APPENDIX C (continued) Figure 3: Phospho-p44/42 MAPK exression is increased in microglia with increasing duration of anti-A antibody tr eatment. The upper graph shows data for frontal cortex while the lower graph shows data for the hippocampus. All data are shown as percent area occupied by positive stain. The bar indicat es values for APP transgenic mice treated with control antibody anti-AMN for three months. The line indicates values for APP transgenic mice receiving anti-A 28-40 antibodies for 1, 2 and 3 months. *indicates P<0.05 when compared to control IgG treated animals.

PAGE 239

APPENDIX C (continued) 230

PAGE 240

231 APPENDIX C (continued) Figure 4: Phospho-p44/42 MAPK immunohistoc hemistry shows microglial expression around amyloid plaques which is increased wi th increasing duration of anti-A antibody treatment. Panels A-D show phospho-p44/ 42 MAPK immunohsitochemical staining around the hippocampal fissure (F in panel A) and in the dentate gyrus (DG in panel A). APP transgenic mice were treated for 3 mont hs with control IgG (A) or with anti-A 28-40 IgG for one (B), two (C) or three (D) mont hs. Magnification = 100X. Scale bar panel A = 50 m.

PAGE 241

APPENDIX C (continued) 232

PAGE 242

ABOUT THE AUTHOR Donna M Wilcock received her Bachelors degree in Pharmacology from Cardiff University in the United Kingdom in 1999 and her Masters degree in Medical Sciences from the University of South Florida in 2003. During her undergraduate years she worked in the laboratory of Dr. John Davies st udying epilepsy and also completed her undergraduate thesis under Dr. John Wilson studying left ventricular hypertrophy. Donna worked as a research technician for Drs. Dave Morgan and Marcia N. Gordon in the Alzheimers research laboratory at the University of South Florida following the completion of her Bachelors degree wher e she was involved in animal husbandry and processing mouse brain tissue. Donna started Graduate school in the fall of 2001 at the University of South Florida and persued her Ph.D. work under the tutelage of Dr. Dave Morgan studying immunotherapy for the tr eatment of Alzheimers disease. She successfully defended her doctoral dissertation in January 2005 at the University of South Florida.


xml version 1.0 encoding UTF-8 standalone no
record xmlns http:www.loc.govMARC21slim xmlns:xsi http:www.w3.org2001XMLSchema-instance xsi:schemaLocation http:www.loc.govstandardsmarcxmlschemaMARC21slim.xsd
leader nam Ka
controlfield tag 001 001681138
003 fts
005 20060215071310.0
006 m||||e|||d||||||||
007 cr mnu|||uuuuu
008 060110s2004 flu sbm s000 0 eng d
datafield ind1 8 ind2 024
subfield code a E14-SFE0001169
035
(OCoLC)62866021
SFE0001169
040
FHM
c FHM
049
FHMM
090
RM101 (Online)
1 100
Wilcock, Donna Marie.
0 245
Mechanisms of [beta]-amyloid clearance by anti-a[beta] antibody therapy.
h [electronic resource] /
by Donna Marie Wilcock.
260
[Tampa, Fla.] :
b University of South Florida,
2004.
502
Thesis (Ph.D.)--University of South Florida, 2004.
504
Includes bibliographical references.
516
Text (Electronic thesis) in PDF format.
538
System requirements: World Wide Web browser and PDF reader.
Mode of access: World Wide Web.
500
Title from PDF of title page.
Document formatted into pages; contains 242 pages.
Includes vita.
520
ABSTRACT: Alzheimers disease (AD) is defined as a progressive neurodegenerative disorder that gradually destroys a persons memory and ability to learn. There are three pathological hallmarks of the disease which are necessary for diagnosis of AD, these are; extracellular amyloid plaques composed of [beta]-amyloid (A[beta]) protein, intracellular neurofibrillary tangles and neuronal loss. Amyloid plaques exist as both compact deposits which stain with Congo red and more numerous diffuse deposits. Active immunization against A[beta] 1-42 or passive immunization with monoclonal anti-A[beta] antibodies reduces amyloid deposition and improves cognition in APP transgenic mice.Over several studies of active immunization in APP+PS1 transgenic mice we showed a strong correlation between reduction of compact amyloid deposits and the degree of microglial activation suggesting a potential role of microglia in the removal of A[beta].Injection of anti-A[beta] antibodies into the frontal cortex and hippocampus of aged APP transgenic mice revealed an early phase of A[beta] removal which was removal of only diffuse amyloid deposits with no associated activation of microglia. A later phase was the removal of compact amyloid deposits. This was associated with significant activation of microglia. Prevention of this microglial activation by anti-A[beta] F(ab)2 fragments or its inhibition by dexamethasone also precluded the removal of compact amyloid deposits but did not affect the removal of the diffuse deposits. Systemic injection of anti-A[beta] antibodies weekly over a period of 1, 2, 3 and 5 months transiently activated microglia associated with the removal of compact amyloid deposits and elevated plasma A[beta], suggesting a peripheral mechanism contributes to removal of brain A[beta].
590
Adviser: David Morgan Ph.D.
653
Immunization.
Transgenic mice.
Inflammation.
Alzheimers disease.
690
Dissertations, Academic
z USF
x Pharmacology and Therapeutics
Doctoral.
773
t USF Electronic Theses and Dissertations.
4 856
u http://digital.lib.usf.edu/?e14.1169