USF Libraries
USF Digital Collections

Optimization of anti-Abeta antibody therapy

MISSING IMAGE

Material Information

Title:
Optimization of anti-Abeta antibody therapy
Physical Description:
Book
Language:
English
Creator:
Karlnoski, Rachel Anne
Publisher:
University of South Florida
Place of Publication:
Tampa, Fla.
Publication Date:

Subjects

Subjects / Keywords:
Alzheimer's disease
Amyloid
Angiopathy
Microglia
Transgenic mice
Dissertations, Academic -- Molecular Pharmacology and Physiology -- Doctoral -- USF   ( lcsh )
Genre:
bibliography   ( marcgt )
theses   ( marcgt )
non-fiction   ( marcgt )

Notes

Summary:
ABSTRACT: Alzheimer's disease (AD) is the most common form of dementia, a disease that gradually destroys brain cells and leads to progressive decline in mental function. The presence of high densities of neuritic plaques composed of Abeta in the cerebral cortices is a criterion for the post-mortem diagnosis of AD. The view that Abeta deposition drives the pathogenesis of AD (amyloid hypothesis) has received support from a wide range of molecular, genetic, and animal studies. This hypothesis has been the focus of therapeutic intervention leading to the development of anti-Abeta immunotherapy as a potential treatment. There is a great deal of evidence that supports the capacity of immunization against Abeta to reduce amyloid pathology and restore memory function in transgenic mouse models of amyloidogenesis.However, as a result of anti-Abeta immunotherapy, many investigators have reported increased severity of cerebral amyloid angiopathy (CAA) and increased incidences of microhemorrhage. The mechanism/s responsible for the redistribution of Abeta to the vasculature is unclear. We examine two possible mechanisms that may influence the severity of CAA following immunization; the rate of Abeta clearance with deglycosylated antibodies via a dose response study and anti-Abeta antibody epitope specificity. Dose response results with a deglycosylated antibody showed that lower doses resulted in greater clearance of amyloid and significant improvements in cognition, suggesting that clearance mechanisms become saturated with high doses of antibody.Treatment with antibodies directed against different epitopes of Abeta implied that the degree of parenchymal Abeta clearance determines the extent of vascular Abeta accumulation; epitope specificity is not critical in directing the vascular accumulation. Passive anti-Abeta immunization can prevent Abeta deposition in APP transgenic mice. We investigated amyloid accumulation after immunization was terminated, and discovered that after treatment, amyloid began to accumulate as a factor of time and gradually built up but never reached the Abeta levels in control APP mice. These data suggest that delayed deposition of amyloid leads to long term delays in AD associated pathology. These data strongly support the use of prophylactic immunotherapy treatments, and it appears that existing amyloid deposits will require interventions that actively clear amyloid as the only means to efficiently reduce brain Abeta in AD.
Thesis:
Dissertation (Ph.D.)--University of South Florida, 2007.
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 Rachel Anne Karlnoski.
General Note:
Title from PDF of title page.
General Note:
Document formatted into pages; contains 142 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 - 001921028
oclc - 190845809
usfldc doi - E14-SFE0002145
usfldc handle - e14.2145
System ID:
SFS0026463:00001


This item is only available as the following downloads:


Full Text

PAGE 1

Optimization of Anti-Abeta Antibody Therapy by Rachel Anne Karlnoski A dissertation submitted in partial fulfillment of the requirement s for the degree of Doctor of Philosophy Department of Molecular P harmacology and Physiology College of Medicine University of South Florida Major Professor: David Morgan, Ph.D. Marcia N. Gordon, Ph.D. Keith Pennypacker, Ph.D. Paula Bickford, Ph.D. Amyn Rojiani, M.D., Ph.D. Date of Approval July 16, 2007 Keywords: Alzheimer’s disease; amyloi d, angiopathy, microglia, transgenic mice Copyright 2007, Rachel Anne Karlnoski

PAGE 2

i TABLE OF CONTENTS LIST OF TABLES ii LIST OF FIGURES iii ABSTRACT v INTRODUCTION Alzheimer’s Disease 1 Transgenic Mouse Models of Amyloidogenesis 4 Immunotherapy for AD 7 Cerebral Amyl oid Angiopathy 15 IgG Isotypes and Fc-Gamma Receptors 18 PAPER 1: DEGLYCOSYLATED ANTI-A ANTIBODY DOSE-RESPONSE EFFECTS ON AMYLOID PATHOLOGY AND MEMORY PERFORMANCE IN APP TRAN SGENIC MICE 20 PAPER 2: COMPARISON OF ANTIBODY EPITOPE WITH RESPECT TO PLAQUE REMOVAL, MICROHEMORRHAGE, A ND COGNITIVE BEHAVIOR IN APP TRANSGENIC MICE 50 PAPER 3: SUPPRESSION OF AMYLOID DEPOSITION LEADS TO LONG TERM REDUCTIONS IN AL ZHEIMER’S PATHOLOGIES IN TG2576 MICE 77 CONCLUSIONS 104 REFERENCES 125 APPENDICES Appendix A Binding affinities of IgG isot ypes to Fc-ga mma receptors 139 ABOUT THE AUTHOR End Page

PAGE 3

ii LIST OF TABLES PAPER 1 Table 1: CD45 expression determined by i mmunohistochemistry in APP transgenic mice following 3 months of passive immunization. 49 APPENDIX A Table 1: Affinities (Kd) of antibody isotype for Murine Fc receptors 140 Table 2: Affinities (Kd) of 2H6, De -2H6, and 2286 for effector proteins 141

PAGE 4

iii LIST OF FIGURES PAPER 1 Figure 1: Dose related increases in serum levels of A and anti-A antibody after deglycosylated anti-A antibody administration. 42 Figure 2: Low doses of deglycosylated antibody improve cognition after 12 weeks of treatment. 43 Figure 3: Total A immunohistochem istry is reduced after systemic administration of anti-A anti bodies. 44 Figure 4: Total A immunohistoc hemistry is significantly reduc ed in the frontal cortex after systemic treatment with the 3 mg/kg dose of deglycosylated antibody 45 Figure 5: Congo red staining in the fr ontal cortex after immunotherapy. 46 Figure 6: Low doses of deglycosylated antibody cause gr eater decreases in fibrillar A. 47 Figure 7: Incidence of microhemorrhage is not changed with any dose of deglycosylated antibody. 48 PAPER 2 Figure 1: Nand C-terminal anti-A antibodies im prove cognition after 12 weeks of treatment 70 Figure 2: Total A immunohistochemistry is reduced by systemic administration of ant i-A antibodies 71 Figure 3: Diffuse A immunohistochemistry is si gnificantly reduced in the frontal cortex and hippocampus after systemic treatment with N-terminal, Mid domain, and C-terminal ant i-A antibodies. 72 Figure 4: Compact Congophilic amyloid deposits ar e reduced by Nand C-terminal anti-A antibody administration while vascular amyloid is increased. 73

PAGE 5

iv Figure 5: Congo red staining in APP transgenic mice is reduced following treatment with Nand C-terminal anti-A antibodies. 74 Figure 6: Nand C-terminal anti-A ant ibodies increase microhemorrhage levels in APP transgenic mice. 75 Figure 7: The quantification of CD45 pos itive stain reveals increas es around remaining deposits in the frontal co rtex and hippocampus of APP transgenic mice after 12 weeks of C-terminal anti-A antibody administration. 76 PAPER 3 Figure 1: Theoretical interpretation of abet a accumulation based on the percent area of total A immunohist ochemistry at 3 age points; 8, 14, and 17 months. 98 Figure 2: Total A immunohistochemis try is reduced following 6 months of systemic anti-A antibody administration. 99 Figure 3: Congo red histochem istry representing compact amyloid deposits are reduced after 6 months of systemic anti-A antibody administration and these reductions are maintained 3 months after the cessation of treatment. 100 Figure 4: CD 45 expression is reduced a fter 6 months of systemic anti-A antibody administration and these reduction s are maintained 3 months after the cessation of treatment. 101 Figure 5. Stacked ELISA measurement s of A40 and A42 including all 3 extracts TBS, SDS, and formic acid from the anterior cortex of Tg2576 mice following 6 months of passive imm unization with either control IgG or anti-A antibody 2H6. 102 Figure 6. Western blot analysis of holo APP & BACE-1 in 14 and 17 mont h old Tg2576 mice that received control IgG or anti-A 2H6 103

PAGE 6

v Optimization of Anti-Abeta Antibody Therapy Rachel Anne Karlnoski ABSTRACT Alzheimer’s disease (AD) is the most common form of dementia, a disease that gradually destroys brain cell s and leads to progressive decline in mental function. The presence of high densities of neuritic plaques composed of A in the cerebral cortices is a criteri on for the post-mortem diagnosis of AD. The view that A deposition drives the pathogen esis of AD (amyloid hypothesis) has received support from a wi de range of molecular, genetic, and animal studies. This hypothesis has been the focus of t herapeutic intervention leading to the development of anti-A immunotherapy as a potential treatment. There is a great deal of evidence that supports the capacit y of immunization against A to reduce amyloid pathology and restor e memory function in transgenic mouse models of amyloidogenesis. However, as a re sult of anti-A immunotherapy, many investigators have reported increased seve rity of cerebral amyloid angiopathy (CAA) and increased incidences of microhemorrhage. The mechanism/s responsible for the redistribution of A to the vasculature is unclear. We examine two possible mechanisms that may influence the severity of CAA following immuniza tion; the rate of A clearance with deglycosylated antibodies via a dose response study and anti-A antibody epitope specificity. Dose response re sults with a deglycosylated antibody showed

PAGE 7

vi that lower doses resulted in greater clearance of amyloid and significant improvements in cognition, suggesti ng that clearance mechanisms become saturated with high doses of antibody. Treatment with ant ibodies directed against different epitopes of A imp lied that the degree of parenchymal A clearance determines the extent of vascula r A accumulation; epitope specificity is not critical in directing the vascular accumulation. Passive anti-A immunization can prevent A deposition in APP transgenic mice. We investigated amyl oid accumulation after immunization was terminated, and discovered that after treat ment, amyloid began to accumulate as a factor of time and gradually built up but never reached the A levels in control APP mice. These data suggest that delayed deposition of amyloid leads to long term delays in AD associated pathology. These data strongly support the use of prophylactic immunotherapy treatments, and it appears that existing amyloid deposits will require interventions that acti vely clear amyloid as the only means to efficiently reduce brain A in AD.

PAGE 8

1 INTRODUCTION Alzheimer’s disease: Alzheimer’s disease (AD) is the most common form of senile dementia. AD is characterized clinically by progr essive cognitive decline and pathologically by amyloid plaques, neurofibrillary t angles, and neuron loss (Hardy and Selkoe 2002). The cause of AD is unknown, howev er, 1-2% of AD patients genetically inherit the disease through autosomaldominant inheritance. This type of inherited AD is termed familial AD (F AD) and results from a mutation on the amyloid precursor protein (APP), presenili n 1 (PS-1) or the presenilin 2 genes (PS-2). It is now clear that mutations in these genes result in APP processing changes. A common result of mutations is a shift from a production of A that is 40 amino acids long (A40) to A that is 42 amino acids long (A42). These findings support the dominant hypothesis for the cause of AD, the Amyloid cascade hypothesis; a hypothesis that pr oposes that A accumulation is the cause of AD pathogenesis. APP structure, function, and metabolism Amyloid precursor proteins are single transmembrane domain glycoproteins that span the membranes of several cell types. The APP gene encompasses ~400 kb of DNA, contai ns 19 exons, and encodes several alternatively spliced APP mRNAs (Price et al 1998). Despite intensive study, the functions of APP have yet to be defined. Some believe that the APP protein

PAGE 9

2 maintains synaptic integrity and aids in neurite outgrowth. APP undergoes metabolism by secretase enzymes that cleave the protein endoproteolytically. APP metabolism by -secretases results in the release of the large soluble ectodomain fragment ( -APPs) into the extracellular space and retention of an 83-amino acid carboxyl-terminal fragment (C TF) in the membrane. Alternatively, APP molecules can also be cleaved by secretase. -secretase cleavages generate a slightly smaller ectodomain derivative (-APPs) and retain a 99residue CTF in the membrane (Selkoe 1998) Following a -secretase cleavage, a subsequent -secretase cleavage leads to the formation and release of A into the extracellular space. Approximately 90% of secreted A peptides are A40, a soluble form of the peptide, whereas 10% of secreted A peptides are A42. A42 is highly fibrillogenic, r eadily aggregated, and deposited early and selectively in amyloid plaques (Price et al 1998). A42 has the ability to recruit or trap A40 into fibril-rich amyloid plaques. Diffuse plaques are largely comprised of A42, with little or no A40 immunoreac tivity (Iwatsubo et al 1994). A40 accumulates in the vasculature and causes a condition known as cerebral amyloid angiopathy (CAA). The firs t specific genetic cause of AD to be identified was missense mutations in APP by Goate et al 1991. These mutations result in FAD, which encompass only 1-2% of all AD cases. The mutations are located within proximity to the cleavage sites. This discovery suggests that these mutations lead to AD by altering proteolytic proc essing at the three secretase cleavage sites. Cells that express the APP mutati on do not appear to secrete higher levels

PAGE 10

3 of A (other than the Swedish mutation) but, rather, secrete a higher fraction of A42 relative to cells that express wild-type APP (Suzuki et al 1994). Presenilin structure, function, and metabolism PS1 and PS2 are highly homologous prot eins that have six to nine transmembrane domains found in the in tracellular membranes of rough endoplasmic reticulum and Golgi apparatus. The Nand C-termini are oriented toward the cytoplasm (Doan et al 1996). T he functional roles of presenilins are not fully understood. However, studies have shown that presenilins play a role in intracellular protein traffi cking, including components of -secretase, and are required for the -secretase cleavage. Presenilins are also required for Notch cleavage, which is involved in development al signaling. PS1 knockout mice show developmental abnormalities similar to those seen in mice with components of the Notch system knocked out (Wong et al 1997). Like APP, presenilins are processed and cleaved to produce an Nand C-terminal fragment. It is not clear whether the holoprotein or the protein fr agments are important physiologically or pathologically. While the mutations found in the amyloid precursor protein are with in the vicinity of cleavage sites, mutations in presenilin are f ound throughout the protein. Mutations in the presenilin prot ein result in a subtle alteration of APP processing causing an increased producti on of A42 over A40. Over 50 mutations in the presenilin proteins have been found to cause FAD (Scheuner et al., 1996).

PAGE 11

4 In summary, the two known genetic c auses of AD are mutations in the APP and PS proteins. Mutations in these proteins all cause an increase in the production of A42, the longer, more hydrophobic form of A. These genetic data strongly support the “Amyloid casc ade hypothesis”; the dominant hypothesis that proposes that an increased accumulati on of A42 initiates the pathogenic events in AD. A42 accumulates and aggr egates, disrupts synapse function, activates microglia and astrocytes whic h release cytokines and reactive oxygen species, further enhances neuritic injury, al ters ionic homeostasi s, alters kinase and phosphatase activities, causes the formation of neurofibrillary tangles, and ultimately widespread neuronal dysfunction and cell death (Hardy et al 2002). Transgenic Models of A Amyloidogenesis: A major advance in AD research has been made in the last several years with the generation of mice that develop age-dependent A accumulation and deposition in their brain and age-dependent learning deficits The first transgenic model to show numerous extracellula r amyloid deposits were the PDAPP or APPV717F mice produced by Games and colleagues in 1995 (Games et al 1995). The PDAPP mice were generated using a human APP minigene encoding the APP717V F mutation; driven by a platelet derived growth factor, PDGF-, promoter. These mice were hemi zygous for the transgene and exhibited deposits of human A at 6-9 months of age. Similar to the pathology found in human AD, these mice deposited amyloid in the hippocampus, corpus callosum, and cerebral cortex. Age-dependent in creases in amyloid deposit size and

PAGE 12

5 density were observed. GFAP-positive reac tive astrocytes, activated microglia, and distorted neurites were closely associated with the A plaques. PDAPP mice also exhibited progressive decline in cognitive function when tested with the water-maze, a behavioral paradigm for l earning and memory. This cognitive decline and inability to re call platform locations correlated with age-related increases in amyloid plaque density in t he hippocampus (Chen G. et al 2000). A transgenic model with similar phenotypic fi ndings was reported by Hsiao et al (Hsiao et al 1996) and contained a double mutation K670N and M671L in the human APP gene under the contro l of a hamster prion pr omoter (PrP). This double mutation is also known as the Swedish mutation because it was found in a large Swedish family with early onset FA D. Transgenic mice with the Swedish mutation are referred to as Tg2576 or APPSW mice. Similar to the PDAPP mice, the Tg2576 mice showed behavioral impairm ents in learning and memory that correlated with increased amounts of A (Hsiao et al., 1996). They also showed amyloid deposits surrounded by GFAPimmunoreactive astrocytes and dystrophic neurites in the cerebral co rtex and hippocampus. TgCRND8 mice encode V717F and the KM670/671NL mutation under the control of the PrP gene promoter (Chishti et al 2001). Thes e mice have robust dense-cored plaques and neuritic pathology by 3 months of age, al ong with learning impairments that were correlated with A42 levels in a referenc e memory version of the Morris water maze (Janus et al 2000). PDAPP and Tg2576 mice have been crossed with other genetically modified mice. One of these was the pr esenilin-1 (PS1) mouse from Duff and

PAGE 13

6 colleagues (Duff et al 1996). The crossing of Tg2576 with a mutant PS1M146L transgenic line generated a transgenic m odel with accelerated amyloid deposition in the cerebral cortex and hi ppocampus far earlier than the singly transgenic Tg2576 mice (Holcomb et al 1998). In these doubly transgenic mice, the diffuse deposits were primarily com posed of A42 while the compact plaques and vascular A are composed of A40. GFAP reactive astrocytes, dystrophic neurites and activated microglia were more pronounced in the doubly transgenic mice compare to the singly transgenic T g2576 or PS1 mutant mice (Gordon MN et al 2002). The doubly transgenic mice showed increased activity in the Y-maze test having significantly higher total num ber of arm entries than the Tg2576 or PS1 mutant mice (Holcomb et al 1998) Follow-up studies showed that the doubly transgenic mice were significantly impaired at 9 mont hs of age compared to non-transgenic mice in the Y-maze test (Holcomb et al 1999). The main objective in using transgenic models of amyloidogenesis is to provide insight into the role of A in Al zheimer’s disease. Memory dysfunction is a very consistent behavioral finding in APP mutant mice. Correlations with cognitive decline and age-related increases in amyloid plaque density suggest that amyloid plays an underlying role in me mory impairment. It is important to note that these models are not “complete” models of AD in that they do not develop marked neuronal loss or typical neurofibrillary tangles found consistently in human AD. However, the patterns of depos ition, regional distribution, and even the anatomical localization of the short and long variants of A mimic the human disease (Morgan 2003). For this reas on, our group and several others are

PAGE 14

7 focusing on finding therapeutic approaches to reduce A loads and reverse cognitive decline in transgenic model s of amyloid deposition. Immunotherapy for AD: Monoclonal antibodies (mAbs) were firs t shown to inhibit the aggregation of fibrillar A in an in vitro study by Solomon and colleag ues in 1996 (Solomon et al 1996). In this study mAbs were ra ised against “aggregating epitopes” of A; specifically amino acids 1-28 and 8-17. These epitopes were found to be sites where A aggregation was in itiated. The mAb 1-28 s howed the ability to prevent A aggregation and convert fibrillar A to a nonfibrillar conformation. However, these changes were not exhibited by t he mAb 8-17. In the following year, Solomon and colleagues show ed that mAb raised against the N-terminal region of A, amino acids 1-16, caused disaggre gation of fibrils, restoration of A solubility, and inhibition of A neurotoxicit y on PC 12 cells at molar ratios ranging from 10 to 100 A/mAb (Solom on et al 1997). It was also found that mAb which bind to epitopes 13-28 and the C-terminal region of A were considerably less effective. Immunizations against A in transgenic models of AD were first demonstrated by Schenk and colleagues in 1999 using the PDAPP mouse (Schenk et al 1999). In this study 2 gr oups of mice, young (6 weeks) and old (11 months) were treated with immunogens that were either synthetic A42 or peptides derived from the primary ami no acid sequence of serum amyloid-P component (SAP). Two additional groups we re either left untreated or were

PAGE 15

8 immunized with PBS as controls. Similar titer responses against A42 were seen in the older and younger animals. The A42 peptide vaccine was shown to prevent A deposition in the young mice and cause a near-total reduction of A in the older mice. The A reduction was accompanied by a virtual elimination of dystrophic neurites and astrogliosis. In contrast, the SAP peptide vaccination did not show significant reductions in A compared to control mice. It can be concluded from this study that active A42 immunization either prevents deposition and/or enhances the clearance of A from the brains of PDAPP mice. One suggested method of clearance wa s that anti-A antibodies triggered microglial cells to clear A using signals mediated by Fc receptors. This was postulated because A-immunoreactive mi croglial cells appeared in regions of remaining plaques. Bard et al (Bard et al 2000) further tested t he postulate that A clearance was mediated by microglial phagocytosis in a passive immunization study. PDAPP mice were injected intraperitoneally weekly for 6 months with anti-A mAbs. Peripherally administered mAbs were found to enter the CNS, decorate plaques, and induce clearance of preexisting am yloid. Unlike actively immunized mice, mice receiving exogenous mAbs (passively immunized) did not demonstrate a T-cell response to A. This indicates that a T-cell response is not necessary for A clearance in PDAPP mice. An ex vivo assay using microglial cells cultured with unfixed sections of ei ther PDAPP mouse or human AD brains was established to examine the effect of antibodies on pl aque clearance. Monoclonal antibodies that effect ively reduced amyloid in the in vivo studies also

PAGE 16

9 reduced amyloid deposits in the ex vivo assays. Following mAb treatment in the ex vivo assay, nearly all amyloid was contai ned in vesicles within the exogenous microglial cells. F(ab)2 fragments of an effective mAb were prepared to determine if Fc receptors were necessary for A phagocytosis. F(ab)2 fragments are antibodies with intact antigen binding regions but lack the Fc region required for Fc R activation. F(ab) 2 fragments failed to trigger microglial phagocytosis, indicating that clearance of A was Fc re ceptor-mediated. Wilcock et al (2001) found that APP+PS1 mice actively inocul ated 5 times with A1-42 showed a correlation between increased microglial activation and fibrillar A reductions in the hippocampus (Wilcock et al 2001). Ou r group further demonstrated the role of microglia in A clearance by intracr anially administering mAbs followed by the administration of anti-inflammatory drugs to impair the microglial response. The impairment of microglia resulted in a complete arrest of anti-A antibodymediated clearance of fibrillar A in the hippocampus (Wilcock et al 2004). In a second series of studies, anti-A F(ab)2 fragments were intracranially administered to APP+PS1 mice. As previously shown by Bard and colleagues in an ex vivo study, the anti-A F(ab)2 were also less effective at activating microglia and therefore less effe ctive at removing fibrillar A in vivo ; although diffuse A was reduced. On the contrary topical application of N-terminal F(ab)2 fragments and full length N-terminal antib odies showed similar reductions in diffuse A and dense-core deposits after 3 days in Tg2576 and PDAPP mice indicating that non-Fc mediated mechanisms are also involved in the clearance of A (Bacskai et al., 2002).

PAGE 17

10 Although active and passive immunizati on were shown to reduce amyloid burden and reverse dystrophic neurites in vivo the effects of vaccination caused microglial activation and inflammation. Morgan and colleagues proposed that inflammatory reactions caused by the vaccine might disrupt normal memory function and tested this theory in APP+PS1 transgenic mice that were actively vaccinated for 5 months (Morgan et al 2000). The results were, in fact, completely opposite from what they had ex pected. Eleven to twelve month old APP+PS1 mice vaccinated with A perfo rmed equivalently with nontransgenic mice in the radial arm water maze (R AWM) test of working memory. This indicated that microglial activation c aused by vaccination was insufficient to disrupt learning and memory functions. Vaccinations were continued until the mice were 15-16 months old and were retest ed in the RAWM. At this age the A vaccinated mice were slower to lear n the platform locations than the nontransgenic mice, however, by trial 5 t he A vaccinated transgenic mice and nontransgenic mice were virtually indistingu ishable. Control transgenic mice were memory deficient. In anot her behavioral study, TgCRND8 mice were vaccinated with A42 or islet-associated poly peptide (IAPP), a peptide with similar biophysical properties to A but not asso ciated with the CNS (Janus et al 2000). Sera from A42 immunized mice stained fi brillar A, and very weakly stained diffuse, non-fibrillar A deposits. This i ndicated that the A42 antibody used in this study was directed toward A in a -pleated sheet conformation. The mice were also tested in a reference memory ve rsion of the Morris wa ter maze test. A series of reversal tests confirm ed that A-immunized TgCRND8 mice had

PAGE 18

11 improved performance. A42 immunizati on caused ~50% reduction in number and size of amyloid plaques without c hanging the levels of APP metabolism, measured by amino-terminal (APPs) or carboxy-terminal (CTFs) secreted fragments in the brain. Three mechanisms of action have been presented for the removal of amyloid deposits following active or passive immunization. The first involves the catalytic disaggregation of A into fibrils leading to monomeric A and clearance from the brain (Solomon et al 1997). Mi croglial mediated phagocytosis of A is the second proposed mechanism of action and is supported by several investigators (Schenk et al., 2002), (Bar d et al., 2000), (Wilcock et al.,2004). The third mechanism of action was proposed when mAbs directed against the central domain of A were peripherally administered to 3 month old PDAPP mice, resulting in a rapid 1000 fold in crease in plasma A and reductions in amyloid burden in the brain without the mAb actually ent ering the brain. It is thought that the antibody sequestered plasma A which disrupted the equilibrium of A in the CNS and plasma, and caused an efflux of A out of the brain into the periphery (DeMattos et al 2001). Human Vaccine Trials After discovering that antibodies agai nst A could reduce AD pathology and improve cognition in APP mouse model s, Elan Pharmaceuticals and Wyeth Laboratories initiated clinic al testing with preaggregated, fibrillar, synthetic A142 under the name of AN 1792. The singl e and multiple dose Phase I trials showed that AN 1792 was well tolerated and a good immunological response

PAGE 19

12 was obtained (Bayer et al., 2005). In t he Phase IIa study, 372 patients with mild to moderate AD were enrolled, 300 of which received AN 1792 plus QS-21 adjuvant intramuscularly, and 72 patients received a saline injection (Gilman et al.,2005). Unfortunately, signs and symptoms consistent with meningoencephalitis were reported in 18 of the patients treated with the active vaccine (Orgogozo et al 2003). As a re sult, all study dosing was halted in January 2002. Although dosing was terminat ed, the study remained intact and was adjusted to monitor the patients in a blinded fashion. Biological and longterm clinical follow-up revealed indicati ons for efficacy and proof of concepteven after the occurrence of meningoencephalitis. The cognitive status of phase IIa patients continued to be monitored following the cessation of the trial. Hock and colleagues repor ted that patients who generated plaque reactive antibodies agai nst A remained cognitively stable while those who did not generate such ant ibodies worsened when tested with the Mini Mental State Examinati on (Hock et al 2003). One patient in the Phase I trials that received the AN-1792 vaccine died fr om a pulmonary embolism 20 months after the first injection and 12 months afte r the last injection (Nicoll et al 2003). The patient’s brain, examined post-mo rtemly, strongly resembled the changes seen in transgenic mice that received A immunotherapy. Focal reductions in cortical amyloid plaques, dystrophic neurites and astrocytosis along with Aimmunoreactive microglia were seen. There we re also signs of CNS infiltration by CD4+ T-cells, suggesting an inflammatory and autoimmune reaction. A immunostained aggregates were found in proximity to phagocytic microglia,

PAGE 20

13 suggesting the microglial mediated mechani sm of A clearance. Two additional case reports were published that showed si milar pathological findings in patients treated with AN-1792 with t he addition of multiple small hemorrhages and increased CAA (Ferrer et al 2004). One pat ient was immunized with 3 injections of AN-1792 and did not exhibit symptoms of encephalitis. Similar to the first patient, this patient’s post mortem ex amination showed an almost complete absence of amyloid and neuritic plaques in the cortex. In fact, Aimmunoreactivity was only found in aggr egates within CD-68-immunoreactive microglia. Interestingly, only a minimal amount of lymphocytic infiltration was observed (Masliah et al 2005). Although the clinical trials with AN 1792 were halted, the study yielded several very important insights into the c linical potential of immunization for AD. Neuropathological investigations of 4 patients immunized with AN-1792 showed reduced -amyloid pathology in certain brain regions, where tau pathology remained (Bombois et al., 2007), (Nico ll et al., 2003), (Fe rrer et al., 2004), Masliah et al,.2005). Importantly, amyl oid reductions were observed in the presence and absence of meni ngoencephalitis, suggesting t hat T cell infiltration was not required for -amyloid removal. Despite preserved tau pathology, measurements of tau in CSF samples showed decreased tau levels, trending towards values seen in normal elderly subjects thus reinforcing the critical role of A in AD pathology (Gilman et al., 2005) In addition to reductions in AD pathology, patients which generated high ant ibody titers as per the tissue amyloid plaque immunoreactivity assay (T APIR) showed cognitive improvements

PAGE 21

14 or slower rates of decline in activities of daily living (Hock et al., 2003). T-cell infiltration was confirmed in post mo rtem exams, alon g with an increased incidence of microhemorrhage and CAA. These adverse effects can hopefully be overcome with passive immunization. In order to overcome the adverse e ffects seen with active immunization, many studies have been conducted using passi ve immunization with monoclonal anti-A antibodies. Passive immunizati on has several advantages over active immunization. The first advantage is that passively administered antibodies are not expected to elicit a T-cell re sponse. Another advantage of passive immunization is control of the dosage of antibody in circulation. Antibody production resulting from acti ve vaccination requires the activation of B-cells by T helper cells. In the elderly however, this interaction is impaired and there is an age-related increase in the number of non-responders to vaccination (Murasko et al., 2002). In a small subset of pat ients treated with AN 1792, 12 out of 24 showed relatively low serum titres (< 1:1000), while 4/24 had very high titres (>1:10,000) (Hock et al., 2002). Passi ve immunization allows for the administration of a measurable and controllable amount of antibody; and overcomes the hurdles of hyporesponsiven ess and variablity. Thus, inadequate immune responses or potential adverse r eactions can be controlled. The dosage and frequency of administration of passi ve immunotherapy can be adjusted to suit the situation. Anot her advantage is that antibod ies raised against specific epitopes on A can be generated and compared for efficacy (Morgan and Gitter 2004).

PAGE 22

15 Cerebral Amyloid Angiopathy: Cerebral amyloid angiopathy (CAA) is the deposition of amyloid in the walls of medium and small sized leptomeni ngeal and cortical arteries. Severe CAA may be associated with micro aneurysm formation, fibrinoid necrosis, or microangiopathies (Rensink et al 2003). Increases in CAA have been found as a consequence of both active and passive immunization. Passively immunized APP23 mice with an N-terminal mAb showed significant reductions in amyloid load while the frequency of CAA-asso ciated cerebral hemorrhages was increased more than two-fold (Pfeifer et al 2002). Similar findings were reported by Wilcock and colleagues when Tg2576 mice were passively immunized with a C-terminal mAb. This study also ev aluated behavior and found that despite the increased CAA and microhemorrhages, Tg 2576 mice still showed cognitive improvements following passive immunization (Wilcock et al 2004b). Two out of the three post-mortem brains of pati ents treated with AN -1792 showed CAA and multiple small microhemorrhages remaini ng after treatment despite the removal of plaques (Nicoll et al 2003; Ferrer et al 2004; Masliah et al.,2005). A possibility for increased CAA is t hat immunization ma y stimulate the efflux of A from the brain through per ivascular drainage pathways (Nicoll et al 2003). Another possibility is that amyloi d clearance may disrupt the integrity of the blood brain barrier and blood vessels by removing amyloid associated with the vessels (Orgogozo et al 2003). A -containing blood vessels are commonly surrounded by inflammatory activity (act ivated microglia); raising the possibility

PAGE 23

16 that inflammation may induce abnormalit ies in blood vessel function (Greenberg 2003). Passive immunization already av oids the T-cell activation otherwise caused by active immunization. However, immunohistochemical measures of microglial activation have shown significant increases in CD45 and Fc receptors II and III (CD16b and CD32) in areas of am yloid deposition after 1 and 2 months of passive immunization (Wilcock et al 2004b). CD45 positive microglia were found closely associated with amyloid bur dened blood vessels (Wilcock et al 2004a). In order to overcome the binding of the Fc receptor to the antigenantibody complex, deglycosylated A antibodies were produced by Rinat Neurosciences Corporation. Deglyco sylation of an antibody removes the carbohydrate side chains on the Fc por tion of the antibody and significantly impairs low affinity Fc receptor recognition. Low affinity Fc receptors, mFc IIb and mFc IIIa, are the majority of Fc receptors in mice (Ravetch 1997; Gessner et al., 1998). Also impaired is the antibody Fc portions ability to bind complement (Windelhalke et al, 1980). Deglycosylated anti-A antibodies were injected into the hippocampi and cortices of Tg2576 mice and were found to reduce diffuse and fibrillar A as efficiently as the intact mAb in both regions without increasing Fc expression or CD45 expression (Carty et al 2006). Intraperitoneal administration of the deglycosylated antibody into Tg2576 mice for 4 months showed similar results to the intact antibody in that diffuse and fibrillar A was reduced, and cognitive improvements were observed when tested in the RAWM. CAA was increased

PAGE 24

17 3.5-fold in the frontal cortex and 3-fold in the hippocampus in mice that received the intact antibody. In comparison, mice treated with the deglycosylated antibody showed significantly less vascular amyloid than the mice treated with the intact antibody; 2-fold in the frontal cortex and 1.5-fold in the hippocampus when compared to the mice administered the control antibodies. The incidence of microhemorrhage in the deglycosylated tr eated mice was 67% less than that found in the intact treated mice (Wilco ck et al., 2006). These data suggest that the deglycosylated antibody uses a differ ent method of A removal than the intact antibody. It is strongly suggest ed that intact antibodies eliminate A through a microglial medi ated phagocytosis, which requires Fc receptor recognition. The deglycosylated antibody does not activate the Fc receptor on microglia yet still reduces A loads to a similar extent. A proposed mechanism of deglycosylated antibody removal of A is the disaggregation of A into more soluble forms, a method of A removal that does not invo lve microglial activation. Alternatively, the deglycosylated antibody may cause A internalization by nonFcR receptors, such as scavenger rec eptors (Brazil et al ., 2000). Hartman and colleagues recently showed that treatment with an anti-A antibody was able rescue the A-induced inhibition of long term potentiation (LTP) in the CA1 hippocampal region of PDAPP mice com pared to untreated PDAPP mice which failed to induce LTP (Hartman et al., 2005) Therefore, treat ment with an anti-A antibody may be able to reverse functi onal abnormalities caused by A by changing the microenvironment at a transcr iptional and translational level.

PAGE 25

18 IgG isotypes and Fc-Gamma Receptors (Fc R) Biological and pathological activities differ with various IgG isotypes. These differences have conventionally been a ttributed to disparities in the ability of certain isotypes to engage co mplement or one of the known Fc R. The finding that individual Fc Rs interact differently with IgG is otypes in mediating protective inflammatory responses is certainly re levant for the potential use of these receptors as therapeutic target s in the treatment of disease. Murine effector cells such as microglia within t he CNS express four different classes of IgG-specific Fc receptors: a high affinity receptor, Fc RI, two low affinity receptors, Fc RII and Fc RIII, and an intermediate affinity receptor, Fc RIV (Ravetch et al., 1998) (Nimmerjahn et al., 2005). Fc receptors I, III, and IV ar e all activating receptors characterized by an immunoreceptor tyrosine-based activation motif (ITAM). These receptors are important for triggering phagocytosis by activated macrophages. Fc RII is an inhibitory receptor c haracterized by the presence of an ITIM motif that recruits i nhibitory phosphatases that limit effective signaling. Bard et al (2003) examined the effica cy of different IgG isotypes on plaque removal from PDAPP brain sections in an ex vivo and in vivo assay and found that IgG2a antibodies against A were mo re efficacious than IgG1 or IgG2b in reducing neuropathology. IgG2a antibodies exhibit the highest affinity for Fc RI therefore the affini ty of IgG for Fc receptors may be important in the efficiency of the antibody. Chauhan et al 2005 f ound that all IgG1 antibodies, when administered in a single bolus ICV inje ction in TgCRND8 mice, cleared cerebral

PAGE 26

19 amyloid more efficiently than IgG2a and IgG 2b. IgG1 has the highest affinity for Fc RIII. Fc RIV was recently discovered an d found to have an intermediate binding affinity for IgG2a and IgG2b, a mechanism that can be proposed for the enhanced activity of IgG2 over IgG1 (Nimmerjahn et al., 2005). If microglial phagocytosis via the Fc receptor is responsible for not only removal of parenchymal amyloid but also for t he increased CAA and microhemorrhage then it is possible that the Ig G isotype of an antibody may result in more CAA and microhemorrhage. However, conflicting data suggests that effective clearance of A by antiA antibodies can be obtained in the absence of Fc receptors. Das et al (2003) showed that when they actively immuni zed APP transgenic mice crossed with Fc receptor knockout mice (lack expression of Fc RIII and Fc RI) they showed the same amount of A reductions as immunized, age-matched APP transgenic mice.

PAGE 27

20 PAPER 1: DEGLYCOSYLATED ANTI-A ANTIBO DY DOSE-RESPONSE EFFECTS ON AMYLOID PATHOLOGY AND MEMORY PERFORMANCE IN APP TRANSGENIC MICE Karlnoski, RAa, Alamed Ja, Ronan Va Gordon MNa, Gottschall PEa, Rosenthal Ab, Grimm Jb, Pons Jb, Morgan Da aAlzheimer's Research Laboratory, Universi ty of South Florid a, School of Basic Biomedical Sciences, Depar tment of Molecular Pharmacology and Physiology, 12901 Bruce B Downs Blvd, Tampa, Florida 33612, USA. bRinat Neuroscience Corp., 230 E Grand Ave, South San Francisco, CA 94080. ACKNOWLEDGEMENTS: This work was s upported by National Institutes of Aging / NIH grants AG15490 (MNG), AG18478 (DM). RK is the Thorne Scholar in Alzheimer Research.

PAGE 28

21 ABSTRACT Anti-A antibody administration to am yloid depositing transgenic mice can reverse amyloid pathology and restore memo ry function. However, in old mice, these treatments also increase vascu lar leakage and promote formation of vascular amyloid deposits. Deglycosylated ant ibodies with reduced affinity for Fcreceptors and complement are associ ated with reduced vascular amyloid and microhemorrhage, while retaining am yloid clearing and memory enhancing properties of native intact antibodies. In the current experiment we investigated the effect of 3, 10, or 30 mg/kg of deglycosylated antibody (D-2H6) on amyloid pathology and cognitive behavior in old Tg2576 mice. A control group received antibody against drosophila amnesiac pr otein (AMN) at a concentration of 10mg/kg. All mice were given weekly intr aperitoneal injections for 12 weeks. The 3 mg/kg dose of deglycosylated anti body effectively reversed learning and memory deficits and significantly reduced diffuse and compact plaque burden in these mice. In contrast, the 30 mg/kg dose failed to improve memory and minimally reduced plaque pathology. Interestingly, all groups had similar increases in CAA despite the differences in plaque clearance. The number of microhemorrhages was less than one prof ile per section for all groups, considerably less than found in prior studies with intact antibody. In conclusion, low doses of deglycosylated antibodies ap pear more efficaci ous than higher doses in reducing pathology and memo ry loss in APP transgenic mice. These

PAGE 29

22 data suggest that excess antibody unbound to antigen can interfere with antibody mediated A clearance, possibly by satu rating the FcRn antibody transporter. INTRODUCTION Passive immunization with anti-A antibodies has been shown to reduce A load and reverse cognitive declin e while increasing cerebral amyloid angiopathy (CAA) and microhemorrhage in transgenic models of amyloid deposition (Wilcock et al., 2004) (Racke et al., 2005). Although the mechanism/s responsible for the redistri bution of A is unclear, we have recently discovered that deglycosylation of the Fc portion on the anti-A anti body significantly reduces the severity of CAA and microhemorrhage compared to native antibodies when passively administered to Tg2576 mice, albeit with slightly lower clearance of parenchymal amyloid deposits (W ilcock et al., 2006) (Carty et al., 2006). Fc receptor and complement activati on are two important mechanisms that initiate phagocytosis by microglia and macrophages. We and others have shown evidence for microglial involvement in the removal of amyloid using both intracranial and systemic administration of anti-A antibodies (Bacskai et al., 2001) (Chauhan et al., 2004) (Wilcock et al., 2003) (Wilcock et al., 2004a) (Wilcock et al., 2004b) (Bard et al., 2000) (Bard et al., 2003). We have reported observations of reactive microglia surr ounding blood vessels that are both highly burdened with A plaque and positive fo r microhemorrhage following passive immunization with anti-A antibodies. Following these observations, we

PAGE 30

23 suggested that activated microglia bound op sonized amyloid and migrated to the vasculature to dispose of the material (Wilcock et al., 2004b). This evidence suggests that activated microglia may exacerbate the CAA and microhemorrhage. The deglycosylation of the carbohydrat e side chains on the Fc portion of an anti-A antibody greatly reduces the affinity of the antibody for Fc receptors, particularly murine Fc RIIb and Fc RIIIa, on effector cells like microglia yet the antibodies are fully capable of binding to A in the same way as the intact antibody (Ravetch et al., 1997) (Gess ner et al., 1998). Deglycosylation also impairs the antibody’s ability to bind complement (Winkelhake et al., 1980). Therefore, one explanation for our prior re sults is that by deglycosylating the A antibody, we are able to mitigate spec ific A clearance mechanisms that contribute to the accumulation of CAA and microhemorrhage. An alternative, however, is that our prior observations were primarily due to a slower rate of A clearance, which, by avoiding saturati on of normal vascular efflux pathways, minimized the accumulation of vascular deposits. One means of resolving the qualitative (specific mechanism) versus quantitative (slower removal) explanations of the effe cts of antibody deglycosylation on vascular amyloid accumulation would be to increase t he rate of amyloid removal with deglycosylated antibodies. If a higher rate of amyloid removal by deglycosylated antibody increased vascular deposits like nati ve antibodies, then the quantitative explanation would be the most likely explanation.

PAGE 31

24 Here we examined the rate of A clearance from the brain by using 3 doses of deglycosylated anti-A antibody in old Tg2576 mice. Based on previous studies that have shown that 0. 1% of peripherally administered antibody enters the CNS, we designed a dose resp onse study using the deglycosylated A antibody with the noti on that higher doses of peripherally administered antibody would enter the br ain and clear amyloid at a higher rate, while the smallest dose would clear amyloid at t he slowest rate (Banks et al., 2002) (Bard et al., 2000). Somewhat paradoxically we found that long term systemic administration of low doses of deglycosyl ated anti-A were more effective at reducing amyloid deposits and reversing co gnitive deficits in old APP transgenic mice when compared to animals treated wit h intermediate and higher doses of antibody. MATERIALS AND METHODS Experimental Design. All APP Tg2576-derived (Hsiao et al., 1996) mice were bred in our facility at the University of South Florida and genotyped using previously described methods (Holcomb et al., 1998) (Gordon et al., 2002). Importantly, we have intent ionally bred out the reti nal degeneration 1 mutation from this colony to avoid the inclusion of occasional mice t hat are blind due to homozygous inheritance of this muta tion contributed by the SJL/J background (Alamed et al., 2006). Twenty Tg2576 mice ( 19 months of age) were assigned to 1 of 4 groups as follows; three groups received the deglycosylated antibody (D2H6 A33-40 IgG2b; Rinat Neurosciences, Palo Al to, CA) at 3 mg/kg, 10mg/kg, or

PAGE 32

25 30mg/kg in saline vehicle. The fourth group received a control IgG antibody against drosophila amnesiac protein (A MN) at a concentration of 10mg/kg (mouse monoclonal anti-drosophila amnesiac protein IgG2b; Rinat Neurosciences, Palo Alto, CA). All mi ce were given weekly intraperitoneal injections of the appropriate antibody and dose for 12 weeks. A fifth group of agematched, nontransgenic mice were used as a control in the behavioral sector of the experiment. Deglycosylation of 2H6 A deglycosylated version of 2H6, D-2H6, was generated by enzymatic removal of N-linked glycans. 2H6 was incubated at 37C for 1 week with peptide-N-glycosi dase F (QA-Bio, San Mateo, CA, 0.05 U/mg of antibody) in 20 mM Tris-HCl pH 8.0; 0. 01% Tween. The deglycosylated antibody was purified by Protein A chromat ography and endotoxin was removed by QSepharose. Completeness of deglycosylat ion was verified by MALDI-TOF-MS and protein gel electrophoresis (Carty et al., 2006). Behavioral Paradigm. Following 3 months of treatment the mice were subjected to a two day behavioral paradigm consis ting of the radial-arm water maze paradigm (RAWM), followed by an open pool visible platform task. The radial arm water maze task was run as previously described (Wilcock et al., 2004). Briefly, the RAWM contained 6 arms radiating out of an open cent ral area, with a hidden escape platform located at the end of one of the arms (Alamed et al., 2006). On day 1, 15 trials were run in 3 blocks. The start arm was varied for each trial, with

PAGE 33

26 the goal arm remaining constant for both da ys. For the first 11 trials, the platform was alternately visible, then hidden, and then remained hidden for the last 4 trials. On day two the testing was sim ilar except that the platform was hidden for all trials. The number of errors (inc orrect arm entries) was measured in a one minute time frame. Mice failing to ma ke an arm choice in 15 seconds were assigned 1 error. In order to minimize the influence of in dividual trial variability, each mouse's errors for three consecutiv e trials were averaged producing 5 data points for each day which were analyzed statistically by ANOVA using StatView (SAS Institute Inc., NC). Next, the mice were run in a one day open pool task with a visible platform to test swim ming ability and eyesight. Any mice demonstrating impaired ability to swim or to see were excluded from behavioral analyses. ELISA analysis of serum A. Serum was diluted and incubated in 96-well microtiter plates (MaxiSorp; Nunc, Rosklide, Denmark), which were precoated with antibody 6E10 (Biosource, Camarillo, CA) at 5 g/ml in PBS buffer, pH 7.4. The detection antibody was biotinylated 4G8 (Signet, Dedham, MA) at a 1:5000 dilution. Detection was done usin g a streptavidin-horseradish peroxidase conjugate (Amersham Biosciences, Arli ngton Heights, IL), followed by tetramethylbenzidine substrate (Sigma-Ald rich, St. Louis, MO). Standard curves of A 1-40 (American Peptide, Sunnyvale, CA) scaling from 6-400 pM were used. ELISA values for serum A levels and ci rculating antibody levels were analyzed using a one-way ANOVA followe d by Fischer’s LSD means

PAGE 34

27 ELISA analysis of serum anti-A antibodies: The anti-A antibody was dissociated from endogenous A in serum as described previously (Li et al., 2004). Briefly, serum was diluted in di ssociation buffer (0.2 M glycine HCl and 1.5% BSA, pH 2.5) and incubated at room temperature for 20 min. The sera were pipetted into the sample reservoir of a Microcon centrifugal device (10,000 molecular weight cutoff; YM-10; Millipor e, Bedford, MA) and centrifuged at 8000 g for 20 min. at room temper ature. The sample reservoi r was then separated from the flow-through, placed inverted into a second tube, and centrifuged at 1000 g for 3 min. The collected solution containi ng the antibody dissoci ated from the A peptide was neutralized to pH 7.0 with 1 M Tris buffer, pH 9. 5. The dissociated sera were assayed by ELISA for antibody titer. A1–40 (Global Peptide)-coated 96-well microtiter plates (MaxiSorp; Nunc) were incubated with dissociated serum samples. A biotinylated goat-ant i mouse IgG (heavy and light chain; Vector Laboratories,Burlingam e, CA) at a 1:5000 dilution followed by peroxidaseconjugated streptavidin (Amersham Biosci ences) was used to detect serum antiA binding activity. Tissue Preparation. On the day of sacrifice the mice were overdosed with 100 mg/kg of Nembutal sodium solution (Abbo tt laboratories, North Chicago IL). The mice were perfused intracardially wit h 25ml 0.9% saline. The right brain hemisphere was dissected and stored for la ter analysis. These dissections were rapidly frozen in dry ice and stored at -80oC. The left hemisphere was removed

PAGE 35

28 and immersion fixed in freshly prepared 4% paraformaldehyde for 24 hours, then passed through 10, 20 and 30% sucrose solutions for 24 hours each. Histology Horizontal sections of 25 m thickness were collected using a sliding microtome and stored in DPBS + azide. A series of 8 sections spaced approximately 600 m apart were stained immunohistochemically for A (6E10 mouse monoclonal anti-A, Biosource, Ca marillo, CA, 1:30,000) to determine the degree of A removal and for CD45 (rat monoc lonal anti-CD45, Serotec, Raleigh, NC 1:5000) to determine the extent of microglial activation. A series of tissue sections 600m apart were stained using 0.2% Congo red solution in NaCl saturated 80% ethanol. Another set of se ctions were stained for hemosiderin using 2% potassium ferrocyanide in 2% hy drochloric acid (Prussian Blue) for 15 minutes followed by a counterstain in a 1% neutral red solution for 10 minutes. Quantification of Congo red staining, CD45 and A immunohistochemistry was performed using the Image-Pro Plus (M edia Cybernetics, Silver Spring, MD) software to analyze the percent area occ upied by positive stain. One region of the frontal cortex and three regions of the hippocampus were analyzed (to ensure that there was no r egional bias in the hippocampal values). The initial analysis of all Congo red pr ofiles provided a total C ongo red value. A second analysis was performed after manually editing out the parenchymal amyloid deposits to yield a percent area restri cted to vascular Congo red staining (Wilcock et al., 2006). To estimate t he parenchymal area of Congo red, we subtracted the vascular Congo red values from the total Congo red values. For

PAGE 36

29 the hemosiderin staining, the numbers of Prussian blue positive deposits were counted over the entire cortex and hippoc ampus on all sections and the average number of hemosiderin deposit s per section was calculated. To assess possible treatment-related differences in pathology the histochemical values for each treatment group were analyzed by one-wa y ANOVA followed by Fischer’s LSD means comparisons using StatVi ew (SAS Institute Inc, NC). RESULTS To study the dose response of deglycosy lated antibodies on fibrillar and diffuse A removal, cerebral amyl oid angiopathy (CAA), microhemorrhage, and cognition, we used transgenic mice c ontaining a double mutation, K670N and M671L, in the human APP gene under the cont rol of a hamster prion promoter (Hsiao et al., 1996). We passively i mmunized 19 month old APP mice with deglycosylated C-terminal (aa35-40) antiA antibodies (D-2H6) for 12 weeks with one of the following doses; 3mg/kg, 10mg/kg, or 30mg/kg. Dose-dependent increases in serum A and circulating anti-A antibody titer: Antibody titer in the final se rum bleeds displayed a dose-dependent response where the highest dose of degl ycosylated antibody, 30mg/kg, showed the greatest anti-A IgG titer at approx imately 2500nM, while the 10mg/kg dose was intermediate at 1500nM, and the 3mg/kg dose had the lowest titer at 250nM (Fig. 1B). Anti-A antibodies were not detected in the APP transgenic mice treated with the control ant ibody or in the nontransgenic mice. Similar to

PAGE 37

30 antibody titer, serum levels of circul ating total A displayed a dose-dependent response, with total serum A increasi ng in parallel with in creasing doses of deglycosylated antibody (Fig. 1A). Reversal of cognitive deficits with low doses of deglycosylated anti-A antibodies: After 12 weeks of antibody treatment, behavioral analysis using the radial-arm water maze (RAWM) showed spat ial reference memory deficits in the control APP transgenic group treated with the anti-dros ophila amnesiac antibody when compared to the nontransgenic mice (Fig. 1). The control APP transgenic group performed the maze wit h the greatest number of errors and the nontransgenic mice performed with the least e rrors. On the second day of testing the mice treated with 3mg/kg of D2H6 performed equivalently to the nontransgenic mice with performance errors near 0.5 per trial (Fig. 1). The APP transgenic mice treated with 10mg/kg or 30mg/kg doses of de2H6 performed at an intermediate level between the nontr ansgenic mice and APP transgenic mice treated with control antibody and were not significantly different from either group. Total A immunohistochemistry is reduced following treatment with deglycosylated antibodies: Total A consists of both compact and diffuse amyloid deposits, and can be detected using immunohistochemical me thods. In old APP transgenic mice the pattern of total A immunohistoc hemistry resembles that of human AD

PAGE 38

31 pathology. In the frontal cortex we observe a typical pattern of diffuse A staining as well as, compact, intensely-stained depos its that are positiv e for amyloid when stained with Congo red or thioflavine-S (Fig. 3A). In the hippocampus, A deposits are concentrated around the hippo campal fissure and the CA1 region (Fig. 3B) with fewer deposits throughout the dentate gyrus and remainder of hippocampal tissue. The appearance of A immunostaining after administration of different doses of D2H6 is presented in the frontal cortex (Fig. 3C, E, G) and hippocampus (Fig. 3D, F, H). Quantificat ion of the percent area occupied by positive staining showed a reverse dose response trend where the lowest dose significantly reduced total A in the front al cortex by 48% while the intermediate dose and high doses were less effective at reducing A (Fig. 4). A similar trend was observed in the hippocampus, although the reductions caused by the lowest dose were not significant (Fig. 4). Deglycosylated anti-A imm unization reduces fibrillar A: Congo red binds preferentially to the betapleated sheet conformation of fibrillar A and is the standard protocol for detecting compact amyloid plaques. Congo red histology in the frontal cort ex of mice treated with deglycosylated antibody is presented in Fig 5. Both the quantity and size of fibrillar, compact plaques in the parenchyma can be apprec iated. In many fields an apparent increase in vascular deposits could al so be observed and are indicated with arrows in Fig 5. Quantification of t he total area occupied by Congo red stain revealed significant reductions with bot h the 3mg/kg and 10mg/ kg doses in the

PAGE 39

32 frontal cortex (Fig 6A). Quantific ation of the parenchymal amyloid deposits showed that the 3 mg/kg dose resulted in the greatest reduction of compact plaque by approximately 70%, followed by the intermediate and highest doses of 10 and 30 mg/kg with a 40% reduction (Fig. 6B). There was an increase in vascular amyloid in both regions (Fig. 6C ). Compared to the c ontrol group, all 3 doses of deglycosylated antibody resulted in a significant elevation of vascular Congo red, albeit the vascular deposits were still considerably fewer than the parenchymal amyloid deposits. It is important to recognize that variability is high when the fractional stained areas are as low as found for vascular deposits in this experiment (less than 0.25% of area). The 30 mg/kg dose increased CAA to the same extent as the 3 mg/kg dose without producing a compar able reduction in compact amyloid. Deglycosylated antibody administ ration does not cause an increase in microhemorrhage: Prussian blue histochemistry is the classic method for demonstrating iron in tissues. Hemosi derin (iron storage granules) may be present in areas of hemorrhage or may be deposited in tissues with iron overload. After 3 months of passive immunization with a deglycosylated Cterminal antibody the average number of positive Prussian blue profiles per section was less than 1 for all antibody dos es, despite the increase in vascular amyloid (Fig. 7). For comparison purpos es, we have included in Fig 7 our prior data obtained with the native antibody (2H6) at 10 mg/kg, which resulted in greater than 3 positive profiles per section (Wilcock et al., 2006).

PAGE 40

33 Microglial activation was measured with CD45 immunohistochemistry. CD45 is a protein tyrosine phosphatase t hat is expressed when microglia are activated. CD45 positive microglia are observed surrounding compact, Congo red positive amyloid plaques. One method of microglial activation is through the Fc receptor, a method that is great ly diminished with the deglycosylated antibody. In a time study conducted by Wilcock et al., CD45 expression after passive administration of intact C-terminal anti-A antibodies showed that CD45 expression is significantly up-regulated at the 1 and 2 month time point and returns to baseline at the 3 month time point (Wilcock et al., 2004b). In this study, CD45 immunohistochemistry showed no differences between the 3 doses of deglycosylated antibody and the contro l group, which confirms that the deglycosylated antibody is not activating microglia at the 3 month time point (Table 1).

PAGE 41

34 DISCUSSION We have previously shown that deglycosylation of an anti-A antibody retains the antibody’s ability to reduce par enchymal amyloid deposits, to reverse cognitive deficits, and to reduce the potentially adverse changes such as microglial activation, CAA and microhemorr hage (Wilcock et al., 2006) (Carty et al., 2006). The data presented in the current experiment suggest that peripheral administration of high doses (>10 mg/kg) of deglycosylated anti-A antibodies in old APP transgenic mice are not as effe ctive at reducing A or reversing cognitive deficits as lower doses In the present study, the lowest dose (3 mg/kg) of D-2H6 reduced diffuse A, fibrillar A, and reversed cognitive deficit s to the greatest extent compared to higher doses of D-2H6 or control IgG. Similar results were presented by Gitter and colleagues at a scientific meeting (Gitte r et al.,2002). In their work, a dose response study using an intact mid-domai n antibody administered for 5 months to PDAPP mice showed that the highest dose failed to reduce A plaque burden and at the same time sequestered the most peripheral A in the plasma, while the lower doses sequestered less A in the serum and reduced total plaque burden to a greater extent. Similar to our study, plasma A and IgG levels showed a dose response relationship where the highest dose of antibody sequestered the highest levels of plasma A. However, in this study as well as in Gitter’s study, peripheral A sequestration di d not correlate to reductions in brain A levels. One possible explanation for this discordance is that much of the

PAGE 42

35 increase in plasma A results from retarded A degradation due to antibody sequestration rather than simply increas ed clearance from t he brain. Deglycosylation of anti-A antibodies significantly reduces the affinity of the antibody for Fc receptors I, II, III located on effector cells, and the complement cascade initiator, C1q (C arty et al., 2006). Immunohistochemical analyses of APP brain sections after intr acranial injections of intact or deglycosylated antibody confirmed that deglycosylation does not activate microglia as measured wit h antibodies against Fc RII/III and CD45 when compared to mice treated with control or nat ive intact anti-A IgG (Carty et al., 2006). In addition to intracranial studies our lab performed a direct comparison of native and deglycosylated anti-A antibodie s in a long-term systemic study. After 4 months of treatment reversal of cognitive deficits and reductions in fibrillar and diffuse A were found with both the deglycosylated and intact antibody, although the deglycosylated antib ody appeared slightly less active. However, the incidence of CAA and micr ohemorrhage were significantly reduced with the deglycosylated antibody compared to its native counterpart (Wilcock et al., 2006). Although deglycosylation reduces the affinity of IgG for Fc receptors, the interaction of immunoglobin with neonatal Fc receptor (FcRn), also known as the Fc transport receptor (FcTR), is not affe cted by the removal of carbohydrate side chains (Hobbs et al., 1992). The FcRn is st ructurally and functionally distinct from the Fc receptors (Ravetch et al., 2001) (Bra mbell et al., 1964). The functional

PAGE 43

36 roles of the FcRn are to recycle IgG and transport IgG bidirectionally across epithelial barriers (Lencer et al., 2005). FcRn expression is found at the br ain microvasculature and choroid plexus epithelium on the blood brain barrier (BBB) (Schlachetzki et al., 2002). The expression of the FcRn in the blood brain barrier may mediate the 'reverse transcytosis' of IgG and immune complexes (ICs) in the brain to blood direction (Schlachetzki et al., 2002). In a study by Banks et al., (Banks et al., 2002), brain uptake of anti-A antibodies and albumin enter the brain at similar rates for the first hour after i.v. injection. However in la ter time points, the in flux rate of anti-A antibody was significantly reduced to less than 1%. The reduction in influx rate was caused by antibody efflux mechanisms. In order for new antibody to enter the brain, antibody in the brain mu st depart. Zhang and Pardridge (Zhang et al., 2001) tested brain efflux mechanisms of IgG with intracranial injections of radiolabeled IgG2a and found that efflux mechanisms were saturated with the addition of excess Fc fragments or intact IgG molecules, resulting in complete suppression of efflux mechanisms. However, efflux mechanisms were not inhibited by high concentrations of F(ab)2 fragments or albumin (Zhang et al., 2001). These findings support the fact t hat the receptors necessary for IgG transport require the Fc portion on IgG to be active. Furthermore, Deane and colleagues (Deane et al., 2005) found that the FcRn transport system is the main mechanism mediating the transcytos is of A-anti-A immune complexes (ICs) from the brain to blood in old Tg2576 mice. This group also found that antibody mediated A clearance from the brain is abolished in old FcRn -/mice.

PAGE 44

37 Similar to Zhang and Pardridge, Deane and colleagues demonstrated that the addition of excess IgG inhi bited the clearance of immune complexes from the brain. Therefore, high concentrations of immunoglobin will saturate the FcRn receptors with antibody not bound to antigen, hinder the FcRn binding of A-antiA immune complexes, and ultimately result in the inhibition of A clearance from the brain. The saturation of FcRn may explain why high doses of deglycosylated antibody are not as effect ive at reducing diffuse and fibrillar A compared to antibodies administered at lowe r doses. The saturati on of this efflux system may also result in A building up along the brain microvasculature and lead to the formation of CAA. Ironically, our original purpose for conducting this study, (to contrast qualitative versus quant itative explanations for the benefits of deglycosylation) could not be addressed, as the high doses of antibody failed to clear more amyloid. In summary, we have shown that t he efficacy of passively administered deglycosylated antibodies depend upon drug dos e. Our study demonstrated that the lowest dose of deglycosylated anti-A antibody was the most efficacious at reducing total A, fibrillar A, and reversing cognitive deficits when tested in old APP transgenic mice. Antibody influx and efflux mechanisms may play a major role in A clearance and in the effica cy of an antibody. Saturation of these mechanisms with uncomplexed antibody may greatly reduce antibody efficacy, emphasizing the requirement for pr ecise titration of dose-response characteristics in human trials of anti A immunotherapy.

PAGE 45

38 REFERENCES Alamed, J., Wilcock, D., Diamond, D ., Gordon, M. and Morgan, D., Two-day radial-arm water maze learning and memory task; robust resolution of amyloid related memory deficits in transgenic mice. Nature Protocols, in press (2006). Bacskai, B.J., Kajdasz, S.T., Christie, R. H., Carter, C., Game s, D., Seubert, P., Schenk, D. and Hyman, B.T., Imaging of am yloid-beta deposits in brains of living mice permits direct observation of clearance of plaques with immunotherapy, Nat.Med., 7 (2001) 369-372. Banks, W.A., Terrell, B., Fa rr, S.A., Robinson, S.M., N onaka, N. and Morley, J.E., Passage of amyloid beta protein antibody across the blood-brain barrier in a mouse model of Alzheimer's dis ease, Peptides, 23 (2002) 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., Kholodenko, D., Lee, C., Lee, M., Motter, R., Nguyen, M., Reed, A., Schenk, D., Tang, P., Vasquez, N., Seubert, P. and Yednock, T., Epitope and isot ype specificities of antibodies to beta -amyloid peptide for protection against Alzheimer's disease-like neuropathology, Proc.Natl.Acad.Sc i.U.S.A, 100 (2003) 2023-2028. Bard, F., Cannon, C., Barbour, R., Burke, R.L., Games, D., Gr ajeda, 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. and Ye dnock, T., Peripherally administered antibodies against amyloid beta-peptide ent er the central nervous system and reduce pathology in a mouse model of Al zheimer disease, Nat.Med., 6 (2000) 916-919. Brambell, F.W., Hemmings, W.A. and Mo rris, I.G., A THEORETICAL MODEL OF GAMMA-GLOBULIN CATABOLISM Nature, 203 (1964) 1352-1354. Carty, N.C., Wilcock, D. M., Rosenthal, A., Grimm, J., Pons, J., Ronan, V., Gottschall, P.E., Gordon, M.N. and Morgan, D., Intracranial administration of deglycosylated C-terminal-specific anti-A beta antibody efficiently clears amyloid plaques without activating microglia in amyloid-depositing transgenic mice, J.Neuroinflammation., 3 (2006) 11. Chauhan, N.B., Siegel, G.J. and Lichtor, T., Effect of age on the duration and extent of amyloid plaque reduction and micr oglial activation after injection of antiAbeta antibody into the third ventricl e of TgCRND8 mice, J.Neurosci.Res., 78 (2004) 732-741. Deane, R., Sagare, A., Hamm, K., Pari si, M., LaRue, B., Guo, H., Wu, Z., Holtzman, D.M. and Zlokovic, B.V., Ig G-assisted age-depende nt clearance of

PAGE 46

39 Alzheimer's amyloid beta peptide by the blood-brain barrier neonatal Fc receptor, J.Neurosci., 25 (2005) 11495-11503. Gessner, J.E., Heiken, H., Tamm, A. and Schmidt, R.E., The IgG Fc receptor family, Ann.Hematol., 76 (1998) 231-248. Gitter, B.D., Gannon K.S., Cummins D.J., Brown-Augsburger P.L., Bales K.R., Bailey D.L., Ballard D.W., Brazelton A. D., Czilli D.L.,Greene S.J., Hepburn D.L., Schirtzinger L.M., Yue X. M., Paul S.M. Galbreath E.J. Reduction in Brain Amyloid Burden and Reversal of Memory Impairment in APP V717F Transgenic Mice Following Chronic Administra tion of the Anti-Amyloid Antibody m266.2, Neurobiol.Aging, 23 (2002) s105. Gordon, M.N., Holcomb, L.A., Jantzen, P.T., DiCarlo, G., W ilcock, D., Boyett, K.W., Connor, K., Melachrino, J., O'Calla ghan, J.P. and Morgan, D., Time course of the development of Alzheimerlike pathology in the doubly transgenic PS1+APP mouse, Exp.N eurol., 173 (2002) 183-195. Hobbs, S.M., Jackson, L. E. and Hoadley, J., Interaction of aglycosyl immunoglobulins with the IgG Fc transpor t receptor from neonatal rat gut: comparison of deglycosylation by tunicamycin treatment and genetic engineering, Mol.Imm unol., 29 (1992) 949-956. Holcomb, L., Gordon, M.N. McGowan, E., Yu, X., Benkovic, S., Jantzen, P., Wright, K., Saad, I., Mueller, R., Morgan, D., Sanders, S., Zehr, C., O'Campo, K., Hardy, J., Prada, C.M., Eckman, C., Younkin, S., Hsiao, K. and Duff, K., Accelerated Alzheimer-type phenotype in transgenic mice carrying both mutant amyloid precursor protein and preseni lin 1 transgenes, Nat.Med., 4 (1998) 97100. Hsiao, K., Chapman, P., Nilsen, S., Eckman, C., Harigaya, Y., Younkin, S., Yang, F. and Cole, G., Correlative memory def icits, Abeta elevation, and amyloid plaques in transgenic mice, Science, 274 (1996) 99-102. Lazarov, O., Robinson, J., Tang, Y.P., Ha irston, I.S., Korade-Mirnics, Z., Lee, V.M., Hersh, L.B., Sapolsky, R.M., Mirnics, K. and Sis odia, S.S., Environmental enrichment reduces Abeta le vels and amyloid deposition in transgenic mice, Cell, 120 (2005) 701-713. Lencer, W.I. and Blumberg, R. S., A passionate kiss, t hen run: exocytosis and recycling of IgG by FcRn, Trends Cell Biol., 15 (2005) 5-9. Li, Q., Cao, C., Chackerian, B., Schiller, J., Gordon, M., Ugen, K.E. and Morgan, D., Overcoming antigen masking of anti-am yloidbeta antibodies reveals breaking of B cell tolerance by virus-like part icles in amyloidbeta immunized amyloid precursor protein transgenic mice BMC.Neurosci., 5 (2004) 21.

PAGE 47

40 Pfeifer, M., Boncristiano, S. Bondolfi, L., Stalder, A., Deller, T., Staufenbiel, M., Mathews, P.M. and Jucker, M., Cerebral hemorrhage after passive anti-Abeta immunotherapy, Science, 298 (2002) 1379. Racke, M.M., Boone, L.I., Hepburn, D.L., Parsadainian, M., Br yan, M.T., Ness, D.K., Piroozi, K.S., Jordan, W.H., Brown, D.D., Hoffm an, W.P., Holtzman, D.M., Bales, K.R., Gitter, B.D., May, P. C., Paul, S.M. and DeMattos, R.B., Exacerbation of cerebral amyloid angiopathy-associ ated microhemorrhage in amyloid precursor protein transgenic mi ce by immunotherapy is dependent on antibody recognition of deposited forms of amyloid beta, J.Neurosci., 25 (2005) 629-636. Ravetch, J.V., Fc receptors, Cu rr.Opin.Immunol., 9 (1997) 121-125. Ravetch, J.V. and Bolland, S., IgG Fc receptors, Annu.Rev.Immunol., 19 (2001) 275-290. Schlachetzki, F., Zhu, C. and Pardridge, W.M., Expression of the neonatal Fc receptor (FcRn) at the blood-brain ba rrier, J.Neurochem., 81 (2002) 203-206. Wilcock, D.M., Alamed, J., Gottschall, P. E., Grimm, J., Rosent hal, A., Pons, J., Ronan, V., Symmonds, K., Gordon, M.N. and Morgan, D., Deglycosylated antiamyloid-beta antibodies eliminate cogni tive deficits and reduce parenchymal amyloid with minimal vascular consequenc es in aged amyloid precursor protein transgenic mice, J.Neurosci., 26 (2006) 5340-5346. Wilcock, D.M., DiCarlo, G. Henderson, D., Jackson, J. Clarke, K., Ugen, K.E., Gordon, M.N. and Morgan, D., Intracrani ally administered ant i-Abeta antibodies reduce beta-amyloid depos ition by mechanisms both independent of and associated with microglial activati on, J.Neurosci., 23 (2003) 3745-3751. Wilcock, D.M., Gordon, M.N. and Morgan, D., Quantificat ion of Cerebral Amyloid Angiopathy and Parenchymal Amyloid Pl aques with Congo Red Histochemical Stain, Nature Protocols, in press (2006). Wilcock, D.M., Munireddy, S.K., Rosent hal, A., Ugen, K.E., Gordon, M.N. and Morgan, D., Microglial activation faci litates Abeta plaque removal following intracranial anti-Abeta antibody administ ration, Neurobiol.Dis., 15 (2004) 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. and Mo rgan, D., Passive amyloid immunotherapy clears amyloid and tr ansiently activates microglia in a transgenic mouse model of amyloid depos ition, J.Neurosci., 24 (2004) 61446151.

PAGE 48

41 Wilcock, D.M., Rojiani, A., Rosenthal, A., Subbarao, S., Freem an, M.J., Gordon, M.N. and Morgan, D., Passive imm unotherapy against Abeta in aged APPtransgenic mice reverses cognitive def icits and depletes parenchymal amyloid deposits in spite of increased vascul ar amyloid and microhemorrhage, J.Neuroinflammation., 1 (2004) 24. Winkelhake, J.L., Kunicki, T.J., Elcombe, B.M. and Aster, R.H., Effects of pH treatments and deglycosylati on of rabbit immunoglobulin G on the binding of C1q, J.Biol.Chem., 255 (1980) 2822-2828. Zhang, Y. and Pardridge, W.M. Mediated efflux of IgG molecules from brain to blood across the blood-brain barrier J.Neuroimmunol., 114 (2001) 168-172.

PAGE 49

42 Figure 1: DOSE RELATED INCREASES IN SERUM LEVELS OF A AND ANTI-A ANTIBODY AFTER DEGLYCOSYLATED ANTI-A ANTIBODY ADMINISTRATION. Panel A shows the average amo unts of circulating A in sera and Panel B shows average amounts of circulating anti-A antibodies 24 hours following the final antibody inject ion in APP transgenic mice receiving either control antibody (Cont) or anti-A antibody at doses of 3, 10, or 30 mg/kg, nontransgenic (NTg) mice received no treat ment. ** indicates P<0.001 compared to APP mice given control antibody injections.

PAGE 50

43 Figure 2: LOW DOSES OF DEGLYCOSYLATED ANTIBODY IMPROVE COGNITION AFTER 12 WEEKS OF TREATMENT Radial Arm Water Maze performance is plotted as the mean number of errors per block of 3 trials over a two day trial period following three m onths of immunizati on with deglycosylated anti-A. The graph shows the average number of errors for APP transgenic mice treated with either 3 mg/kg degl ycosylated anti-A antibody ( dashed line), 10 mg/kg deglycosylated 2H6 anti-A antibody ( dashed line), 30 mg/kg deglycosylated anti-A antibody ( dashed line), control anti-AMN antibody ( solid line) and nontransgenic mice ( solid line). indicates P<0.05 between the 3mg/kg treated group and the control IgG treated group.

PAGE 51

44 Figure 3: TOTAL A IMMUNOHISTOCHEMISTRY IS REDUCED AFTER SYSTEMIC ADMINISTRATION OF ANTI-A ANTIBODIES Panels A-H show A immunohistochemistry from APP transgenic mice in the frontal cortex (Panels A, C, E, and G) and CA1 region of the hippocampus (Panels B, D, F, H). Mice were treated with control antibody ( anti-AMN 2908; Panels A and B), 3 mg/kg deglycosylated anti-A antibody (Panel C and D), 10 mg/kg deglycosylated 2H6 anti-A antibody (Panel E and F), or 30 mg/kg deglycosy lated anti-A antibody (Panels G and H). Both the number of par enchymal deposits and their intensity appear less after immunotherapy. Scale bar in panel A = 25m for all panels. Abbreviations: CA1: CA1 pyra midal cell layer; F: fissure

PAGE 52

45 Figure 4: TOTAL A IMMUNOHISTOCHEMISTRY IS SIGNIFICANTLY REDUCED IN THE FRONTAL CORT EX AFTER SYSTEMIC TREATMENT WITH THE 3 MG/KG DOSE OF DEGLYCOSYLATED ANTIBODY. The graph shows the quantification of per cent area of total A stai ning in the frontal cortex (A) and hippocampus (B) after 12 weeks of treatment with control IgG (CTRL), 3mg/kg, 10 mg/kg, and 30 mg/kg doses of deglycosylated anti-A antibody. indicates P<0.05 when compared to mice treated with c ontrol antibody.

PAGE 53

46 Figure 5: CONGO RED STAINING IN THE FRONTAL CORTEX AFTER AFTER IMMUNOTHERAPY. (A) anti-AMN control antibody; (B) 30 mg/kg dose D-2H6 anti-A antibody; (C) 10 mg/kg ; (D) 3 mg/kg dose. Scale bar in panel A = 50 m. Arrows shown in panels B, C, and D indi cate vessels that are positively stained for CAA.

PAGE 54

47 Figure 6: LOWER DOSES OF DEGLYCOS YLATED ANTIBODY CAUSE GREATER DECREASES IN FIBRILLAR A Panel A shows the quantification of percent area of total C ongo red staining in the frontal cortex and hippocampus after 12 weeks of anti-A antibody passive immunization in front al cortex (solid bars) and hippocampus (open bars). P anel B shows quantification of parenchymal Congo red in the frontal cort ex (solid bars) and hippocampus (open bars). Panel C shows quantification of va scular Congo red in the frontal cortex (solid bars) and hippocampus (open bars) indicates p<0.05, ** indicates p<.001 compared with control antibody.

PAGE 55

48 Figure 7: INCIDENCE OF MICROHEMORRHAG E IS NOT CHANGED WITH ANY DOSE OF DEGLYCOSYLATED ANTIBODY The graph shows quantification of positive Prussian blue pr ofiles per section in APP transgenic mice treated with control anti-AMN IgG, 3mg/kg, 10 mg/kg, or 30 mg/kg doses of deglycosylated antibody for 3 months (Mean +/SEM). Incidence of microhemorrhage after intact 2H6 antibody at 10 mg/kg observed by Wilcock et al (2006) was much higher.

PAGE 56

49 Table 1. CD45 expression determined by immunohistochemistry in APP transgenic mice following 3 months of passive immunization. Treatment Anterior Cortex Hippocampus Control anti-AMN antibody 2.912 19% 2.693 59% 3 mg/kg of D-2H6 2.491 42% 2.898 26% 10 mg/kg of D-2H6 2.736 43% 2.517 23% 30 mg/kg of D-2H6 2.331 32% 2.310 15% Values are mean SEM

PAGE 57

50 PAPER 2 COMPARISON OF ANTIBODY EPITOPE WITH RESPECT TO PLAQUE REMOVAL, MICROHEMORRHAGE, AND COGNITIVE BEHAVIOR IN APP TRANSGENIC MICE. Karlnoski RA1, Wilcock DM1, Alamed J1, Mercer M1, Gordon MN1, Gottschall PE1, Kobayashi D2, Rosenthal A2, Grimm J2, Pons, J2, Morgan D1* 1Alzheimer's Research Laboratory, Universi ty of South Florida, Department of Pharmacology, 12901 Bruce B Downs Blvd, Tampa, Florida 33612, USA. 2Rinat Neuroscience Corp. 3155 Porter Dr ive, South San Francisco, California 94304, USA ACKNOWLEDGEMENTS: This work was s upported by National Institutes of Aging / NIH grants AG15490 (MNG), AG18478 (DM). RK is the Thorne Scholar in Alzheimer Research.

PAGE 58

51 ABSTRACT Passive immunization in transgenic m ouse models of A amyloidogenesis has shown that anti-A antibodies have t he ability to prevent and reduce neuritic plaques, to reverse cogniti ve decline, and have the tendency to increase the occurrence and severity of CAA and microhemorrhage. In the current study we compare an tibodies directed against different epitopes of the A peptide for their abili ty to clear abeta deposits, mitigate or exacerbate CAA and microhemorrhages, as well as their effects on behavior. Tg2576 mice, 21-22 months of age, were treated with one of the following antiA antibodies; N-terminal antibody (A1-5 IgG2b), mid domain antibody (A16-28 IgG2b), or C-terminal antibody (A33-40 IgG2b). The fourth group received a control IgG antibody against drosophila amnesiac protein (mouse monoclonal anti-drosophila amnesiac protein IgG 2b). All mice were given weekly intraperitoneal injections of the appropriate antibody at a concentration of 10mg/kg for 12 weeks. The Nand C-terminal anti-A antibodies reversed cognitive deficits and significantly reduced diffuse and compact plaque loads, but at the same time increased CAA and microhemorr hage compared to mice that received control IgG. In contrast, the mid domain anti body significantly reduced diffuse A and showed no exacerbation of C AA or microhemorrhage but also failed to improve memory. These data suggest that the exacerbation of CAA and incidence of hemorrhage are a result of compact am yloid clearance. Furthermore, a component of compact plaques appears responsible for memory loss.

PAGE 59

52 INTRODUCTION The amyloid -protein (A) is derived from the -amyloid precursor protein (APP) by the cleavages of and -secretase (Seubert et al., 1993). Depending on the exact point of cleavage, three prin cipal forms of A, comprising 38, 40, or 42 amino acids residues are produced. T he overproduction and accumulation of A is the precipitating ev ent that cascades to the development of Alzheimer’s disease (AD) (Hardy and Selkoe, 2002) Thus, based upon this concept of amyloid accumulation, one major target of therapeutic intervention in AD has been to actively clear A from the brain with immunotherapy. Both active and passive immunotherapies against A have shown promising results in humans and in transgenic mice. Work in mouse models show remarkable efficacy in reducing am yloid loads and restoring cognitive function. Passive immunization, however, permits careful titration of antibody dosage, allows for reversibility, and enables scientists to modify and target specific epitopes of a protein. Most antibodies developed with active immunization are directed against the N-terminal amino acids of the A pept ide (Dickey et al., 2001) (Schenk et al., 1999) (Hock et al., 2002) (Lee et al., 2005). Solomon and co lleagues reported that amino acid residues 3-6 were t he minimal effective epitope in raising antibodies against A because these N-term inal residues were responsible for the aggregation of A (Solomon et al., 1996). Initial studies showed that antibodies specific for the N-terminal regi on of A could preven t the formation of

PAGE 60

53 fibrillar amyloid in vitro (Solomon et al., 1996) (Solomon et al., 1997). Peptides with this epitope specificity have been f ound effective at reducing amyloid and reversing cognitive deficits in mouse model s (Bard et al., 2000) (Kotilinek et al., 2002). However, antibodies directed agains t the mid domain and C-terminal of the A peptide have also been effective at reducing A load (DeMattos et al., 2001) and reversing cognitive deficits (Wilcock et al., 2004b). More recently, immunization against A has revealed adverse effects. Despite significant reductions in amyloi d load and improvements in cognition, the exacerbation of cerebral amyloid angiopathy (CAA) and microhemorrhage after chronic administration of monoclonal antibodies are two unwarranted effects. Pfeifer and colleagues passively immuni zed 21-month-old APP23 mice for 6 weeks with an N-terminal antibody and found a significant reduction in diffuse A42 but not A40, as well as a two fold increase in CAA-associated microhemorrhages (Pfeifer et al., 2002). Racke and colleagues found that Nterminal antibodies bound to CAA associat ed microvessels significantly better than mid-domain antibodies (Racke et al., 2005). DeMattos and colleagues used a mid domain antibody and found that although they could not detect the antibody in the brain, it was able to r educe diffuse plaque in the brain by rapidly sequestering plasma A (DeMattos et al ., 2001). Our lab passively immunized Tg2576 mice with a C-terminal antibody and found a significant reduction in both diffuse and fibrillar parenchymal A as well as improvements in cognition, despite increased CAA and microhemorrhages (Wilcock et al., 2004b).

PAGE 61

54 The purpose of this experiment is to compare antibodies directed against different epitopes of the A protein to determine whethe r they are equally effective in modulating amyloid and cognition. Furthermore, comparison of antibodies with different epitopes will ident ify if the increased vascular amyloid and hemorrhage are a general property of antibody-associated A clearance, or if antibodies with different epitope spec ificities could avoi d these potentially adverse consequences. MATERIALS AND METHODS Experimental Design. All APP Tg2576-derived mice were bred in our facility at the University of South Florida an d genotyped using previously described methods (Holcomb et al., 1998) (Gordon et al., 2002). Impor tantly, we have intentionally bred out the re tinal degeneration 1 mutation from this colony to avoid the inclusion of occasional mice that are blind due to homozygous inheritance of this mutation contributed by the SJL/J background (Alamed et al., 2006). Twenty-four Tg2576 mice (21-22 mont hs of age) were assigned to one of 4 groups. The first group received the N-terminal antibody anti-A1-5 IgG2b (2324, Rinat Neurosciences, Palo Alto, CA). The second group received the mid domain antibody anti-A16-28 IgG2b (2289, Rinat Neurosciences, Palo Alto, CA). The third group received the C-terminal antibody anti-A33-40 IgG2b (2H6, Rinat Neurosciences, Palo Alto, CA). The fourth group received the control IgG antibody against drosophila amnesiac prot ein (mouse monoclonal anti-drosophila amnesiac protein IgG2b; Rinat Neuroscienc es, Palo Alto, CA). All mice were

PAGE 62

55 given weekly intraperitoneal injections of the appropriate antibody at a concentration of 10mg/kg for 12 wee ks. A fifth group of age matched, nontransgenic mice was used as a cont rol in the behavioral sector of the experiment. Behavioral Paradigm. Following 3 months of treatment the mice were subjected to a two day behavioral paradigm consis ting of the radial-arm water maze (RAWM), followed by an open pool visible platform task. The radial arm water maze task was run as previously descri bed (Wilcock et al., 2004b). Briefly, the RAWM contained 6 arms radiating out of an open central area, with a hidden escape platform located at the end of one of the arms (Alamed et al., 2006). On day 1, 15 trials were run in 3 blocks. The start arm was varied for each trial, with the goal arm remaining constant for both da ys. For the first 11 trials, the platform was alternately visible then hidden (hidden fo r the last 4 trials). On day two, the mice were run in exactly the same manner as day 1 except that the platform was hidden for all trials. The number of errors (incorrect arm entries) were measured in a one minute time frame. Mice failing to make an arm choice in 15 seconds were assigned 1 error. In order to minimize the infl uence of individual trial variability, each mouse's errors for th ree consecutive trials were averaged producing 5 data points for each day which were analyzed statistically by ANOVA using StatView (SAS Institute Inc., NC). Next, the mice were run in a one day open pool task with a visible platform to veri fy swimming ability and eyesight. No mice demonstrated impaired ability to swim or to see.

PAGE 63

56 Tissue Preparation. On the day of sacrifice the mice were overdosed with 100 mg/kg Nembutal sodium solution (Abbott laboratories, North Chicago IL). The mice were perfused intracardially with 25 ml of 0.9% saline. The right brain hemisphere was dissected, rapidly frozen on dry ice and stored at -80oC. The left hemisphere was removed and immers ion fixed in freshly prepared 4% paraformaldehyde for 24 hours, then pa ssed through 10, 20 and 30% sucrose solutions for 24 hours each. Histology Horizontal sections of 25 m thickness were collected using a sliding microtome and stored in DPBS + azide. A series of 8 sections spaced approximately 600 m apart were stained via free-fl oating immunohistochemically for total A (rabbit polyclonal anti-A, gift from Dr. Gottschall USF, 1:10,000) to determine the degree of A removal and for CD45 (rat monoclonal anti-CD45, Serotec, Raleigh, NC 1:5000) to determine t he extent of microglial activation. The free-floating immunohistochemis try protocol was previous ly described by Gordon et. al., (Gordon et al., 2002). A series of tissue sections 600m apart were stained using 0.2% Congo red solution in NaCl saturated 80% ethanol. Another set of sections were stained for hemosider in using 2% potassium ferrocyanide in 2% hydrochloric acid (Prussian Blue) for 15 minutes followed by a counterstain in a 1% neutral red solution for 10 minutes. Quantification of Congo red staining, CD45 and A immunohistochemistry was performed using the Image-Pro Plus (M edia Cybernetics, Silver Spring, MD) software to analyze the percent area occ upied by positive stain. One region of

PAGE 64

57 the frontal cortex and three regions of the hippocampus were analyzed (to ensure that there was no r egional bias in the hippocampal values). The initial analysis of all Congo red pr ofiles provided a total Congo red value. A second analysis was performed after manually editing out the parenchymal amyloid deposits to yield a percent area restri cted to vascular Congo red staining (Wilcock et al., 2006). To estimate t he parenchymal area of Congo red, we subtracted the vascular Congo red values from the total Congo red values. For the hemosiderin staining, the numbers of Prussian blue positive deposits were counted over the entire cortex and hippoc ampus on all sections and the average number of hemosiderin deposits per sect ion calculated. To assess possible treatment-related differences in pathology the histochemical values for each treatment group were analyzed by one-wa y ANOVA followed by Fischer’s LSD means comparisons using StatView (SAS Inst itute Inc, NC). The histological data presented in the manuscript are all no rmalized to the control group. RESULTS Reversal of cognitive deficits with Nand C-terminal anti-A antibodies: After 12 weeks of antibody treatment, behavioral analysis using the radialarm water maze (RAWM) showed spatial refe rence memory deficits in the control APP transgenic group treated with the anti-drosophila amnesiac antibody when compared to the nontransgenic mice (F ig. 1). The control APP transgenic group performed the maze with the greatest num ber of errors and the non-transgenic mice performed with the fewest errors. On the second day of testing the mice

PAGE 65

58 treated with the Nand C-terminal antibod ies performed signific antly better than the APP mice treated with cont rol IgG with performance e rrors near 1 per trial (Fig. 1). The APP transgenic mice treat ed with the mid domain antibody did not improve in the behavioral paradigm. T he average number of errors for the groups treated with the mid domain antibody and control IgG were around 3 per trial at the end of the second day (Fig. 1). The one day open pool task determined that all mice had the ability to swim and see a visible platform (data not shown). Total A is reduced following treatment with all 3 antibodies: Total A consists of both compact and diffuse amyloid deposits that can be detected using immunohistochemical me thods. In old APP transgenic mice the pattern of total A immunohistoc hemistry resembles that of human AD pathology. In the frontal cortex we observe a typical pattern of diffuse A staining as well as, compact, intensely-stained deposit s that are positive for A (Fig. 2A). In the hippocampus, A deposits are concentrated around the hippocampal fissure and the CA1 region with fewer deposits throughout the dentate gyrus and remainder of hippocampal tissue (Fig. 2B). The appearance of A immunostaining in the frontal cortex and hippocampal regions are shown in Figure 2. A qualitative reduction in diffu se A with remaining compact deposits can be seen in both the cortex and hippoc ampus after administration of all 3 antibodies compared to the control IgG (F igure 2). Quantificat ion of the percent area occupied by positive staining showed signi ficant reductions in total A by all 3 antibodies in both the cortex and hippocam pus (Fig. 3). The values shown are

PAGE 66

59 the percent area of A normalized to the control group. Nand C-terminal, but not mid domai n anti-A antibodies cause reductions in fibrillar A: Congo red binds preferentially to the betapleated sheet conformation of fibrillar A and detects compact amyloid plaques. Congo red histology in the frontal cortex and hippocampal fissure of mice treated with control IgG, Nterminal, mid domain, and C-terminal antibod y are presented in figure 4. Many congophilic plaques can be detected in control IgG and mid domain antibody treated APP mice. Following anti-A treat ment, an apparent increase in vascular deposits could also be observ ed and are indicated with arrows (Fig 4). These vascular deposits were only apparent in areas devoid of parenchymal amyloid. Quantification of the tota l area occupied by Congo red stain revealed significant reductions with both the N-terminal and Cterminal anti-A antibodies in the frontal cortex and hippocampus (Fig 5A). The percent area positive for Congo red was normalized to the control values; this gave the control groups a value of one for all measures. Quantification of the pa renchymal amyloid deposits showed that the Cterminal antibody resulted in the greatest reduction of compact plaque by approximately 95%, followed by the N-te rminal antibody with a 60% reduction in the frontal cortex and hippocampus (Fig. 5B ). The mid domain antibody failed to reduce compact deposits. The C-terminal antibody caused an almost complete clearance of parenchymal deposits in both the hippocampus and cortex however;

PAGE 67

60 there was a two-fold increase in vascu lar amyloid compared to mice that received control IgG (Fig. 5C). C-terminal anti-A antibodies cause si gnificant increases in activated microglia surrounding amyloid deposits: Microglial were visualized with CD45 immunohistochemistry. CD45 is a protein tyrosine phosphatase that is expr essed when microglia are activated. The fibrillar A plaques formed by APP transgenic mice are accompanied by local activation of microglia and are decorated with CD45 pos itively stained microglia (Gordon et al., 2002). CD45 st ain in the group treated with the Cterminal antibody revealed intensely st ained microglia surrounding the remaining parenchymal, compact amyloid deposits while the pattern and intensity of microglial expression from the mid domain and N-terminal antibodies showed no differences compared to controls. Howeve r, quantification of the percent area of CD45 immunohistochemistry showed no di fferences between the 3 antibodies tested in either the frontal cortex or hippocampus (Fig. 6A & 6B). Since the majority of activated microglia are associated with compact deposits, these measurements reflect the amount of mi croglial activation per deposit. Quantification of Congo red histochemistry revealed significant reductions in compact plaque with antibody treatment; t herefore we calculated the ratio of CD45 immunohistochemistry to total C ongo red percent area to estimate the microglial activation per amyloid deposit. Calculating this ratio demonstrated increased microglial expressi on of CD45 associated with the remaining amyloid

PAGE 68

61 deposits in both the frontal cortex (Fig. 6C) and hippocampus (Fig. 6D) only after treatment with the c-te rminal antibody. No change in the ratio was found following treatment with the Nterminal or mid domain anti-A antibodies in either the frontal cortex (Fig.6C) or hippocampus (Fig. 6D). Both Nand C-terminal anti-A antibodies cause an increase in microhemorrhage: Prussian blue histochemist ry is the classic met hod for demonstrating iron in tissues. Hemosiderin (iron storage granules) may be present in areas of hemorrhage or may be deposited in tissues with iron overload. After 3 months of passive immunization with N-terminal, mid domain, or C-terminal anti-A antibodies, we find significant increases in the number of positive Prussian blue profiles per section with the N-terminal and C-terminal antibodies (Fig. 7). We found an average of 2 positive Prussian bl ue profiles per section with the Nterminal antibody and an average of 3 pos itive profiles per section with the Cterminal antibody.

PAGE 69

62 DISCUSSION In the present study, we demonstrat ed that antibodies with different epitope specificities have distinct effect s on different A pools and, ultimately, different behavioral outcomes. In addition, we investigated the extent of vascular A accumulation and microhemorrhage after systemic treatment with antibodies directed against different epitopes of A We found that increases in vascular amyloid correlate with the degree of co mpact, parenchymal A clearance and epitope specificity is not a cr itical factor in directing the vascular accumulation. After 3 months of weekly intr aperitoneal injections, cognitive improvements were seen in the mice tr eated with the N-terminal and C-terminal anti-A antibodies, while the mid domai n antibody failed to show behavioral improvements. Despite t he differences found with the behavioral data, all three antibodies resulted in similar reductions in total A immunostaining. Meanwhile, only the Nand C-terminal antibodies were able to significantly reduce fibrillar A as measured with Congo red hi stochemistry. Interestingl y, of the 3 antibodies tested, the C-terminal antibody caused t he greatest reductions in A with more than 90% of the parenchymal deposits remo ved. Although these reductions were associated with a two-fold increase in va scular A, the remaining vascular loads were still much less than the total br ain amyloid loads, suggesting both a redistribution of parenchymal amyloid into the vasculature and a dramatic clearance of A fr om the brain. It has been previously hypothesized that microglial activation via anti-A antibodies is required for the removal of fibrillar deposits (Wilcock et al., 2001)

PAGE 70

63 (Schenk et al., 1999) (Wilcock et al ., 2004) (Bard et al., 2000). Here we demonstrate that the C-te rminal antibody causes the greatest reduction in compact amyloid plaques and the remainin g amyloid plaques have the greatest number of associated microglia. Our results support previous findings that an N-terminally directed and Cterminally directed antibody could exac erbate CAA and result in increased microhemorrhage (Racke et al., 2005) (Pfeifer et al., 2002) (Carty et al., 2006). Interestingly, the N-terminal antibody wa s not as effective at reducing A loads and did not significantly increase CAA as compared to the C-terminal antibody. Despite the lack of CAA, the N-term inal antibody increased microhemorrhage more than 2-fold compared to mice treated with control IgG. It has been previously noted that Nterminal antibodies bound to deposited A in tissue sections and to CAA-bearing vessels (Ra cke et al., 2005) (Bard et al., 2003). A single application of N-terminally dire cted antibodies under a cranial window showed modest clearance of CAA while chroni c topical infusion resulted in robust clearance of CAA; determined with quantitative in vivo image analysis (Prada et al., 2007). Vascular and parenchymal depos its can be removed with N-terminally directed antibodies. However, the remo val of vascular amyloid may further disrupt the integrit y of the vessel or cause localized vascular inflammation and exacerbate microhemorrhage. The mid domain antibody did not lead to increased vascular A or incidence of microhemorrhage. Binding c haracterizations of a different mid domain antibody, 266, demonstrated binding only to soluble A but not to

PAGE 71

64 deposited A in the brain parenchyma or cerebral vessels (DeMattos et al., 2001) (Racke et al., 2005). This may ex plain why we see reductions in only the diffuse amyloid with the mid domain ant ibody. These data also suggest that clearance of the fibrillar deposits and not diffuse deposits cause the exacerbation of CAA, which in this case was only observed with the C terminal antibody. Another significant observation from this study was that behavioral improvements correlated with antibodies that reduced compact amyloid, suggesting that compact amyloid plaques are in some manner associated with the memory impairments. In support of these findings, active immunization with A of double mutant APP TgCRND8 mice par tially prevented the development of reference memory deficits in a water maze task, with only a 50% reduction in the size and number of dense core amyloi d deposits, and no effect on the total soluble pool of A in brain (Janus et al., 2000). The researchers concluded that the prevention of memory deficits wa s due to the reduced amyloid pathology seen in their immunized mice. However, compact deposits are likely to be surrounded by a number of smaller, more diffusible, oligomeric assemblies. Chronic active immunization had similar beneficial effects on memory impairment in two different strains of transgenic mi ce as assessed using a radial-arm water maze (Morgan et al., 2000). Notably, treatm ent did not affect amyloid pathology in the same way in the two strain s of mice. Immunized APP+PS1 double transgenic mice showed a reduction in diffuse (nonfibrillar) A deposits in the cerebral cortex and hippocampus, but not in amyloid (fibrillar A) deposits. In contrast, Tg2576 APP transgenic mice showed a small but statistically significant

PAGE 72

65 reduction in cortical amyloid burden, suggest ing that active immunization reduces the development of fibrillar A deposit s in this mouse strain. These authors concluded that active immunization prev ents memory deficits by altering either brain amyloid pathology or an unknow n pool of non-deposited A, perhaps a soluble pool of A. Lesne and colleagues searched for the appearance of an A species that coincided with the first obs erved changes in spatial memory in 6 month old Tg2576 mice and found that only nonamer and dodecamer (A*56) levels correlated with impai rment of spatial memory (Lesne et al., 2006). The toxic moiety responsible for synaptic dysfunction and neuronal cell loss still remains to be identified. Ta ken together with our data, these results suggest that the relationship between soluble and in soluble brain A concentrations and memory impairment in transgenic mice is complex. One possibility is that in young or mid-aged mice, simple removal of soluble A species may restore learning and memory, yet in old mice plaque reduction may be more critical. Further examination into the redistribution and reduction of different forms of A will be useful to elucidate what form/ s of A accompanies the changes in memory performance in APP transgenic mi ce following passive immunization. All in all, one antibody failed to produce substantial A clearance with dramatically less vascular amyloid accu mulation. These data suggest that the degree of parenchymal A clearance deter mines the extent of vascular A accumulation. The direct binding of Nterminal antibodies to amyloid-laden vessels may aid in the redistribution and removal of parenchymal and vascular

PAGE 73

66 A respectively but may also exacer bate microhemorrhage by aggravating the integrity of the vessel.

PAGE 74

67 Reference List Alamed J, Wilcock DM, Diamond DM, Gordon MN, Morgan D (2006) Two-day radial-arm water maze learning and memory task; robust resolution of amyloidrelated memory deficits in transgeni c mice. Nat Protoc 1: 1671-1679. Bard F, Barbour R, Cannon C, Carretto R, Fox 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-lik e neuropathology. Proc Natl Acad Sci U S A 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, Khol odenko D, Lee M, Lieberburg I, Motter R, Nguyen M, Soriano F, Vasquez N, Wei ss K, Welch B, Seubert P, Schenk D, Yednock T (2000) Peripherally administ ered antibodies against amyloid betapeptide enter the central nervous syst em and reduce pathology in a mouse model of Alzheimer disease. Nat Med 6: 916-919. Carty NC, Wilcock DM, Rosenthal A, Grim m J, Pons J, Ronan V, Gottschall PE, Gordon MN, Morgan D (2006) Intracranial administration of deglycosylated Cterminal-specific anti-Abeta antibody effi ciently clears amyloid plaques without activating microglia in amyloid-deposit ing transgenic mice. J Neuroinflammation 3: 11. DeMattos RB, Bales KR, Cummins DJ, Dodart JC, Paul SM, Holtzman DM (2001) Peripheral anti-A beta antibody al ters CNS and plasma A beta clearance and decreases brain A beta burden in a mouse model of Alzheimer's disease. Proc Natl Acad Sci U S A 98: 8850-8855. Dickey CA, Morgan DG, Kudchodkar S, We iner DB, Bai Y, Cao C, Gordon MN, Ugen KE (2001) Duration and specificity of humoral immune responses in mice vaccinated with the Alzheimer's diseaseassociated beta-amyloid 1-42 peptide. DNA Cell Biol 20: 723-729. Gordon MN, Holcomb LA, Jantzen PT, DiC arlo G, Wilcock D, Boyett KW, Connor K, Melachrino J, O'Callaghan JP, Mor gan D (2002) Time course of the development of Alzheimer-like pathol ogy in the doubly transgenic PS1+APP mouse. Exp Neurol 173: 183-195. Hardy J, Selkoe DJ (2002) The amyloid hypothesis of Alzheimer's disease: progress and problems on the road to t herapeutics. Science 297: 353-356.

PAGE 75

68 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 betaamyloid by vaccination of patients wit h Alzheimer disease. Nat Med 8: 12701275. 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, Du ff K (1998) Accelerated Alzheimer-type phenotype in transgenic mice carrying both mutant amyloid precursor protein and presenilin 1 transgenes. Nat Med 4: 97-100. Janus C, Pearson J, McLaurin J, Mathew s 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 impairm ent and plaques in a model of Alzheimer's disease. Nature 408: 979-982. Kotilinek LA, Bacskai B, Westerman M, Ka warabayashi 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. Lee M, Bard F, Johnson-Wood K, Lee C, Hu K, Griffith SG, Black RS, Schenk D, Seubert P (2005) Abeta42 immunization in Alzheimer's disease generates Abeta N-terminal antibodies. Ann Neurol 58: 430-435. Lesne S, Koh MT, Kotilinek L, Kayed R, Glabe CG, Yang A, Gallagher M, Ashe KH (2006) A specific amyloid-beta protein assembly in the brain impairs memory. Nature 440: 352-357. Morgan D, Diamond DM, Gottsch all 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 vaccina tion prevents memory loss in an animal model of Alzheimer's disease. Nature 408: 982-985. Pfeifer M, Boncristiano S, Bondolfi L, Stalder A, Delle r T, Staufenbiel M, Mathews PM, Jucker M (2002) Cerebral he morrhage after passive anti-Abeta immunotherapy. Science 298: 1379. Prada CM, Garcia-Alloza M, Beten sky RA, Zhang-Nunes SX, Greenberg SM, Bacskai BJ, Frosch MP (2007) Antibody -mediated clearance of amyloid-beta peptide from cerebral amyl oid angiopathy revealed by quantitative in vivo imaging. J Neurosci 27: 1973-1980. Racke MM, Boone LI, Hepburn DL, Parsadai nian M, Bryan MT, Ness DK, Piroozi KS, Jordan WH, Brown DD, Hoffman WP, Holtzman DM, Bales KR, Gitter BD, May PC, Paul SM, DeMattos RB (2005) Exacerbation of cerebral amyloid angiopathy-associated microhemorrhage in amyloid precursor protein transgenic

PAGE 76

69 mice by immunotherapy is dependent on ant ibody recognition of deposited forms of amyloid beta. J Neurosci 25: 629-636. 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, Lieberbu rg 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 Alzheimer-disease-like patholog y in the PDAPP mouse. Nature 400: 173-177. Seubert P, Oltersdorf T, Lee MG, Barbour R, Blomquist C, Davis DL, Bryant K, Fritz LC, Galasko D, Thal LJ, (1993) Secretion of beta-amyloid precursor protein cleaved at the amino terminus of the beta-amyloid peptide. Nature 361: 260-263. Solomon B, Koppel R, Frankel D, Hanan-Aharon E (1997) Disaggregation of Alzheimer beta-amyloid by site-direct ed mAb. Proc Natl Acad Sci U S A 94: 4109-4112. Solomon PR, Knapp MJ, Gracon SI, Groccia M, Pendlebury WW (1996) Longterm tacrine treatment in patients with Al zheimer's disease. Lancet 348: 275-276. Wilcock DM, Gordon MN, Morgan D (2006) Quantification of cerebral amyloid angiopathy and parenchymal amyloid pl aques with Congo red histochemical stain. Nat Protoc 1: 1591-1595. Wilcock DM, Rojiani A, Rosenthal A, Levkowitz G, Subbarao S, Alamed J, Wilson D, Wilson N, Freeman MJ, Gordon MN Morgan D (2004a) Passive amyloid immunotherapy clears amyloid and transient ly activates microglia in a transgenic mouse model of amyloid depositi on. J Neurosci 24: 6144-6151. Wilcock DM, Rojiani A, Rosenthal A, Subbarao S, Freeman MJ, Gordon MN, Morgan D (2004b) Passive immunotherapy against Abeta in aged APPtransgenic mice reverses cognitive def icits and depletes parenchymal amyloid deposits in spite of increased vascul ar amyloid and microhemorrhage. J Neuroinflammation 1: 24. Wilcock DM, Gordon MN, Ugen KE, Gottsc hall PE, DiCarlo G, Dickey C, Boyett KW, Jantzen PT, Connor KE, Melachrino J, Hardy J, Morgan D (2001) Number of Abeta inoculations in APP+PS1 transgeni c mice influences antibody titers, microglial activation, and congophilic plaq ue levels. DNA Cell Biol 20: 731-736.

PAGE 77

70 Figure 1: NAND C-TERMINAL ANTI-A AN TIBODIES IMPROVE COGNITION AFTER 12 WEEKS OF TREATMENT. Radial Arm Water Maze mean number of errors made over a two day trial period following three months of immunization with anti-A antibodies directed at different epitopes of the A peptide. Each data point is the average of three trials. The graph shows the average number of errors for APP transgenic mice treated with either N-terminal anti-A antibody ( dashed line), Mid-domain anti-A antibody ( dashed line), C-terminal antiA antibody ( dashed line), control anti-AMN antibody ( solid line) and nontransgenic mice ( solid line). indicates P<0.05 and ** indicates P< 0.01. Significance is indicated for both the N-terminal and C-terminal anti-A antibody treatment groups compared to the group treated with control IgG.

PAGE 78

71 Figure 2: TOTAL A IMMUNOHISTOCHEMISTRY IS REDUCED BY SYSTEMIC ADMINISTRATION OF ANTI-A ANTIBODIES Panels A-F show A immunohistochemistry from APP transgenic mice in the frontal cortex (Panels A, C. E and G) and hippocampus (Panels B, D, F, and H). Mice were treated with control anti-AMN antibody (P anels A and B), N-terminal anti-A antibody (Panel C and D), Mid-domain anti-A antibody (Panel E and F), or C-terminal anti-A antibody (Panel G and H). F represents the hippocampal fissure. Scale bar in panel A = 50m for all panels.

PAGE 79

72 Figure 3: TOTAL A IMMUNOHISTOCHEMISTRY IS SIGNIFICANTLY REDUCED IN THE FRONTAL CORTEX AND HIPPOCAMPUS AFTER SYSTEMIC TREATMENT WITH N-TERMINAL, MID-DOMAIN, AND CTERMINAL ANTI-A ANTIBODIES. The graph shows t he quantification of percent area of total A staining in t he frontal cortex and hippocampus after 12 weeks of treatment with control IgG, N-terminal (N), Mid-domain (M), and Cterminal (C) anti-A antibodies. Frontal co rtex (solid bars); hippocampus (open bars). indicates P<0.05 and ** indicates P<0.01; when compared to mice treated with control antiAMN antibody (CTRL).

PAGE 80

73 Figure 4: COMPACT CONGOPHILIC AMYLOID DEPOSITS ARE REDUCED BY NAND C-TERMINAL ANTI-A ANTIBODY ADMINISTRATION WHILE VASCULAR AMYLOID IS INCREASED. Panels A, C, E, and G show Congo red staining in the frontal cortex. Panels B, D, F, and H show Congo red staining in the hippocampus. Staining is shown fr om APP transgenic mice treated with control anti-AMN antibody (P anels A and B), N-terminal anti-A antibody (Panel C and D), Mid-domain anti-A antibody (Panel E and F), or C-terminal anti-A antibody (Panel G and H). Arrows indica te blood vessels positive for vascular amyloid, while F reveals the position of the hippocampal fissure. Scale bar in panel A = 50m for all panels.

PAGE 81

74 Figure 5: CONGO RED STAINING IN APP T RANSGENIC MICE IS REDUCED FOLLOWING TREATMENT WITH NAND C-TERMINAL ANTI-A ANTIBODIES. Panel A shows the quantification of percent area of total Congo Red staining in the frontal cortex (s olid bars on left) and hippocampus (open bars on right) after 12 weeks of anti-A antibody passive immunizati on frontal cortex (solid bars) and hippocampus (open bar s). Panel B shows quantification of parenchymal Congo red in the frontal cort ex (solid bars) and hippocampus (open bars). Panel C shows quantification of va scular Congo red in the frontal cortex (solid bars) and hippocampus (open bars). indicates p<0.05, ** indicates p<.001 verse control IgG treated mice.

PAGE 82

75 Figure 6: THE QUANTIFICATION OF CD 45 POSITIVE STAIN REVEALS INCREASES AROUND REMAINING DEPO SITS IN THE FRONTAL CORTEX AND HIPPOCAMPUS OF APP TRANSGENIC MICE AFTER 12 WEEKS OF CTERMINAL ANTI-A ANTIBODY ADMINISTRATION CTRL indicates the antiAMN control antibody, N indicates the Nterminal anti-A antibody, M indicates the mid-domain anti-A antibody, and C indicates the C-terminal anti-A antibody. Panels A and B show the total per cent area occupied with positive stain for CD45 while panels C and D show calcul ated ratios of CD45 staining to total Congo Red. Panels A and C show quantification in the frontal cort ex while panels B and D show quantification in the hippocampus. *Indica tes P<0.05 and ** indicates P<0.01 compared to control.

PAGE 83

76 Figure 7: NAND C-TERMINAL ANTI-A ANTIBODIES INCREASE MICROHEMORRHAGE LEVELS IN APP TRANSGENIC MICE. The number of positive Prussian blue profiles per sect ion was analyzed. CTRL indicates the anti-AMN control antibody, N indicates the N-terminal anti-A antibody, M indicates the mid-domain anti-A antibody and C indicates the C-terminal antiA antibody. ** indicates P<0.01 vers e APP transgenic mice treated with control IgG.

PAGE 84

77 PAPER 3: SUPPRESSION OF AMYLOID DEPOSI TION LEADS TO LONG TERM REDUCTIONS IN ALZHEIMER’S PATHOLOGIES IN TG2576 MICE Karlnoski, RAa, Kobayashi Db Alamed Ja, Mercer Ma, Gordon MNa, Gottschall PEa, Morgan Da aAlzheimer's Research Laboratory, University of South Florida, School of Basic Biomedical Sciences, Depar tment of Molecular Pharmacology and Physiology, 12901 Bruce B Downs Blvd, Tampa, Florida 33612, USA. bRinat Neuroscience Corp., 230 E Grand Ave, South San Francisco, CA 94080. ACKNOWLEDGEMENTS: This work was supported by National Institutes of Aging / NIH grants AG15490 (MNG), AG 18478 (DM) and AG204418 (PPG). RK is the Thorne Scholar in Alzheimer’s Disease Research.

PAGE 85

78 ABSTRACT The accumulation of -amyloid in Alzheimer’s disease is a key pathological event. However, the reason for the accumulation is unclear. A accumulation can be described by two hypot heses, the Accumulation hypothesis or the Equilibrium hypothesis. The a ccumulation hypothesis suggests that A deposition is a result of production s lightly exceeding clearance mechanisms leading to a gradual accumulation of the excess A in the brain. The equilibrium hypothesis suggests that the rate of A production, degradation and excess is in a steady state equilibrium. The incr ease with age results from equilibrium dynamics. This study was designed to invest igate the extent to which the level of Abeta in brain is due to passage of time versus a change in production/clearance with age. We accomplished this experim ent by passively immunizing Tg2576 mice with anti-A antibodies (2H6), starting at an age w hen amyloid deposition first begins (8 months) and ending treat ment 6 months later at an age when untreated Tg2576 mice have considerable amounts of deposited amyloid. We then monitored the progressi on of amyloid accumulation over the next 3 mo in the two groups to ascertain if the s uppressed group would “catch up” to the untreated group (equilibrium) or if it would remain suppressed (accumulation). We found that the anti-A 2H6 an tibody successfully suppressed the deposition of A after 6 mont hs of treatment. We also found that 3 months after the cessation of treatment with anti-A 2H6, diffuse and compact A deposits as well as markers of microglial activation, remained significantly reduced compared to the Tg2576 mice that received control IgG. These results suggest that A

PAGE 86

79 deposition occurs because producti on mechanisms exceed clearance mechanisms by a small amount, and the excess becomes converted to a form that deposits and builds up in a time dependent manner.

PAGE 87

80 INTRODUCTION A is continuously produced from the APP in brains of normal subjects and patients with AD. By some undefined mechanism, A deposits accumulate in the brains of patients developing AD. This observable fact is mimicked in APP transgenic mouse models (Games et al, 1995) (Hsiao et al, 1996),despite stability of transgene expression (J ohnson-Wood et al., 1997). A40 & 42 accumulates with age, first as diffuse pl aques lacking glial changes and later as fibrillar, thioflavin-S positive neuritic/g lial plaques (Selkoe, 2001). Deposition of A42 is believed to precede that of A40 in the brain, which is supported by in vitro studies that show A42 can aggregat e more readily (Jarrett et al., 1993); (Iwatsubo et al., 1994). A plaques are thought to be formed from the gradual accumulation and aggregation of secreted A in the extrac ellular space. This statement is consistent with the accumulation hypothesis which suggests that A deposition is a result of production ex ceeding clearance mechanisms by a small amount, and the excess becomes converted to a form that deposits and builds up in a timedependent manner (Maggio et al., 1992); (Prior et al., 1996). The accumulation hypothesis might indicate that treatm ents reducing A produc tion or increasing soluble A clearance might prevent further A deposition, but would not remove existing deposits (this would requi re clearing fibrillar deposits). Another possible explanation for increasing levels of A in the brain is that steady-state disequilibrium exists between anabolic and catabolic activities. The equilibrium hypothesis suggests that the rate of A production, degradation and

PAGE 88

81 removal is altered with age, and that the di fferent pools of A are to some extent exchangeable; removal of one pool will ultimate ly lead to remova l of the others. Insulin degrading enzyme (IDE) and neprilysi n, two proteases that degrade A, have age related reductions in steady-state levels in regions of the brain with amyloid deposition (Caccamo et al ., 2005), (Iwata et al., 2002). In this study, we developed a scheme to evaluate why aberrant A accumulation occurs by assessing whether A deposition is a factor of accumulation over time (Accumulation Hypot hesis) or is triggered by changes in the brain microenvironment with age (Equilibrium Hypothesis). We accomplished this by delaying deposition using passive immunization with anti-A antibodies between the ages of 7-14 months (Tg2576 mice typically start depositing A at 8 months of age.) We then released the suppression and evaluated the pathology 3 months later. MATERIALS AND METHODS Experimental Design. All APP Tg2576-derived mice were bred in our facility at the University of South Florida an d genotyped using previously described methods (Holcomb et al., 1998); (Gordon et al., 2002). Importantly, we have intentionally bred out the re tinal degeneration 1 mutation from this colony to avoid the inclusion of occasional mice that are blind due to homozygous inheritance of this mutation contributed by the SJL/J background (Alamed et al., 2006). Fourteen eight month old Tg2576 mi ce were injected with the anti-A antibody, 2H6 (A33-40 Ig2b; Rinat Neurosciences, S outh San Francisco, CA), for

PAGE 89

82 6 months at a concentration of 10mg/kg week ly (i.p.) until they reach 14 months of age, at which time we ended injections At this point we euthanatized the mice at 1 day and 12 weeks post treatment. A group of 14 Tg2576 mice received weekly injections of control antibody directed against the drosophila specific amnesiac protein (2908, m ouse monoclonal anti-dros ophila amnesiac protein IgG2b; Rinat Neurosciences, South San Francisco, CA). These control APP transgenic mice underwent the same tr eatment and euthanization schedule as the mice treated with the 2H6 antibody. Tissue Preparation. On the day of sacrifice the mice were overdosed with 100 mg/kg of Nembutal sodium solution (Abbo tt laboratories, North Chicago IL). The mice were perfused intracardially with 25 ml of 0.9% saline. The right brain hemisphere was dissected and stored for later analysis. These tissues were rapidly frozen on dry ice and stored at -80oC. The left hemisphere was removed and immersion fixed in freshly prepared 4% paraformaldehyde for 24 hours, then passed through 10, 20 and 30% sucrose solutions for 24 hours each. Histology Horizontal sections of 25 m thickness were collected using a sliding microtome and stored in DPBS + azide. A series of 8 sections spaced approximately 600 m apart were stained immunohistoc hemically using the freefloating method for total A (rabbit pol yclonal anti-A, generously provided by Paul E. Gottschall, USF, FL 1:10000) to determine the amount of diffuse A, as previously described (Gordon et al., 2002). Additional sections were

PAGE 90

83 immunostained for various microglial marker s associated with different stages of amyloid deposition. Complement rec eptor 3 (CD11b) (r at monoclonal antiCD11b, Serotec, Raleigh, NC 1:3000) is associated with the earliest deposits in young mice, CD45 (rat monoclonal anti-CD45, Serotec, Raleigh, NC 1:3000) is an intermediate marker, and MHC-II (rat monoc lonal anti-I-A/I-E, BD Pharmingen 1:3000) is a marker associated with deposits in older mice (Gordon et al., 2002). Adjacent sections were stained with 0. 2% Congo red solution in NaCl saturated 80% ethanol, which detects the compact amyloid plaques made of up fibrillar A. Quantification of the hi stological markers was performed using the ImagePro Plus (Media Cybernetics, Silver Spri ng, MD) software to analyze the percent area occupied by positive stain. Two regi ons of the cortex and three regions of the hippocampus were analyzed (to ensure that there was no regional bias in the values). To assess possible treatment-re lated differences in pathology, the histochemical values for each treatm ent group were analyzed by one-way ANOVA followed by Fischer’s LSD means comparisons using StatView (SAS Institute Inc, NC). Abeta ELISA. Tissue homogenization and extraction: Microdissected frontal cortices from one hemisphere were sequent ially extracted with buffers designed to increasingly solubilize amyloid depos its. At each step, homogenization, (30 seconds) in an appropriate buffer (150mg/ml wet weight) which contained protease inhibitors (complete mini pr otease inhibitor cocktail; Roche) was followed by centrifugation at 100,000 X g fo r 1 hr at 4C. The supernatant was

PAGE 91

84 then removed and the pellet was homogenized in the next solution used in the extraction process. The ex traction took place first in Tris-buffered saline (TBS) (20mM Tris, 137 mM NaCl, pH 7.6), and the resultant pe llet was then extracted with SDS diluted with TBS to a final c oncentration of 2%, and the resultant pellet was then extracted with 70% formic acid (FA) diluted with TBS. A1-40 and A1-42 was quantified in these samples using individual A1-40, 1-42 ELISA kits (Signet, Berkeley, CA ) in accordance with the instructions of the manufacturer, except the standards included dilutions of 2%SDS or 70% formic acid when necessary. Western Blot. Posterior cortices were plac ed in 300l ice cold cell lysis buffer with protease inhibitors (Cell Si gnaling Technology, Danvers MA). The brains were then homogenized for 30 seconds followed by a 30 second sonication. Total protein content was assessed using the Bio-Rad Protein assay kit according to the manufacturer’s instru ctions. A 50 g aliquot of total protein from each sample was diluted in 4X SD S loading buffer (each 100 ml of buffer contains: 3 g Tris; 8 g SDS; 2.5 g DT T; 0.05 g Bromopheno l blue; 40% [v:v] glycerol) and electrophoretic ally separated using 4-20% tris-glycine gels at 80v for 2 hours. Electrophoresed proteins were then transferred to polyvinylidene difluoride membranes (Bio-Rad, Richmond, CA), washed in 1x Tris-buffered saline with 20% tween (TBS-T) three time s, and blocked for 1 hour at room temperature with BlockerTM BLOTTO in TBS (Pierce, Rockford, IL). After blocking, the membranes were hybridized over night at 4C with various primary antibodies. The primary antibodies included the APP C-terminal 751-770

PAGE 92

85 antibody (Calbiochem, San Diego, CA), BACE-1 (gifted from Dr.Vassar and Dr. Binder), and GAPDH (Biodesign Internati onal, Saco, ME). The membranes were washed three times with TBS-T and incubat ed for 1 hour at room temperature with the appropriate HRP-conjugated sec ondary antibody. All antibodies were diluted with BlockerTM BLOTTO. Immunoreactive bands were detected by LumiGold ECL western blotti ng detection kit (Ver II) (Signagen Laboratories, Gaithersburg, MD) according to the manufa cturer's instruction and then exposed to film which was developed fo r later densitometric analys is. All membranes are reprobed (without stripping) with anti-GAPDH to control fo r equal protei n loading in each lane. Bands were quantified using Scion Image (NIH, Bethesda, MD) by analyzing pixel density. Semiquantitative analysis was performed by densitometry, correcting protein levels for GAPDH. RESULTS To study the deposition of A we used transgenic mice containing a double mutation, K670N and M671L, in the human APP gene under the control of a hamster prion promot er (Tg2576; Hsiao et al., 1996). The primary dependent measure in this study was A depositi on. Hypothetical out comes depicting the percent area of total A deposition accord ing to the Accumulation hypothesis or the Equilibrium hypothesis are shown in Figures 1A and B respectively. If the outcome were to support the accumulation hypothesis, we would expect to see, once suppression is released, the slope of the lines for both the control and treated groups to be the same but shift ed to the right (delayed) for the mice

PAGE 93

86 whose amyloid deposition was suppressed with 2H6 for 6 months (Fig 1A). If the outcome were to support the equilibrium hypothesis, we would expect to observe a rapid elevation of A load following t he release of suppression in the mice treated with 2H6. The rapid elevation in A load would near the control transgenic mice amyloid levels within 12 months (Figure 1B). This type of outcome would indicate that the environment of the aging brain was determining the amount of A deposited. The hypothet ical and actual results are based upon histological observations at 3 time poi nts; 8 months of age (the age at which Tg2576 mice begin to deposit), 14 months of age (the time at which suppression of A deposition with 2H6 is released), and 17 months of age (3 months posttreatment). After systemic treatment with antiA antibodies for 6 months, we observed that antibody admin istration prevented A depos ition and substantially delayed the deposition of amyloid (Figur e 2). Total A immunohistochemistry in the Tg2576 mice that received the cont rol antibody showed several intensely stained deposits along with numerous diffu se deposits in the frontal cortex compared to the age-matched transgenic mi ce that received the 2H6 antibody (Figure 2A & B). The mice that rece ived the control antibody demonstrated the typical amount and distribution of A fo r Tg2576 mice at these age points with a 5-fold increase in total A between the ages of 14 and 17 months; with an average value of 15% of the frontal cort ical area being littered with amyloid by the age of 17 months. (Figure 2E) (Hsiao et al., 1996) (Kawarabayashi et al., 2001). The mice treated with 2H6 showed a significant suppression of amyloid

PAGE 94

87 deposition at the 14 month time point (Figure 2C). The total A immunohistochemistry showed very few comp act deposits as well as very sparse diffuse A stain which encompassed less than 1.5% area of the frontal cortex (Figure 2C & E). Three months after treatment was stopped, the total percent area of A was 6% in the frontal cortex of mice t hat had been previously treated with 2H6. The A percent area from the 17 month 2H6 group resembled the percent area from the 14 m onth old control group (Figure 2E). Nonetheless, treatment with 2H6 antibody for 6 months resulted in significant reductions in total A immunohistochemistry in the frontal cortex and these significant reductions were maintained 3 mont hs after treatment was stopped. Congo red staining detects compact amyloid deposits in the -pleated sheet formation. There are far fe wer Congo red positive deposits than A deposits detected by total A immunohistochem istry. Compared to the mice that received control antibody; the mice tr eated with 2H6 showed significantly less compact deposits in the frontal cortex at the 14 month time point (Figure 3A,& C). Three months after the release of suppr ession, these mice continued to show significant reductions (approximately 50% ) in compact amyloid (Figure 3E). In addition to A amyloid histology, we immunohistochemically stained for microglial markers that correlate to vari ous levels of activation; complement receptor 3 (CD11b), CD 45, and MHC-II. Complement-receptor 3 expression was increased around the earliest Congophilic deposits in young mice. CD 45 is a protein-tyrosine phosphatase and is expr essed at intermediate stages of A deposition. MHC-II is a mark er that is expressed around mature deposits in older

PAGE 95

88 mice (Gordon et al., 2002). In this study, we did not see any differences in complement-receptor 3 expr ession between the contro l and treated groups at the 14 or 17 month age points (Data not shown). We also did not find differences in MHC-II immunohistochemistry between cont rol and treated groups. In fact we rarely observed positive MHC-II microglia at the 14 or 17 month time points (Data not shown). CD 45 expression increases at an intermediate stage of plaque maturation. At the 14 mont h time point, CD 45 expressi on is extremely low in both the treated and control group (Figur e 4A & C). However, we observed a significant increase in CD45 expre ssion on microglia surrounding amyloid deposits in the frontal cort ex of the control group at the 17 month time point (Figure 4B). Three months after the treatment was stopped, the 17 month old 2H6 group CD45 expression resembled t he 14 month control group expression levels (Figure 4E). The CD 45 expression time course resembled the total A time course, where the control group dem onstrated an accelerated increase in expression between14 and 17 months of age and the treated group showed a slight increase in this period of ti me with the endpoint resembling the 14 month untreated group levels. To confirm the histological findin gs, anterior cortex homogenates from the contralateral side of the brain were extr acted sequentially in TBS, 2% SDS, and 70% formic acid and measured for A40 and A42 levels with sandwich ELISAs. Sequential extractions allo wed measurement of differ ent pools of A. The TBS fraction represented the most soluble fr action. Tg2576 mice at 14 to 17 months

PAGE 96

89 of age have small amounts of TBS-solu ble A (Kawarabayashi et al., 2001). Most A requires SDS or fo rmic acid for solubilizati on since the most abundant form of A at these ages are the insolubl e, fibrillar forms (Kawarabayashi et al., 2001). In both A40 and A42 ELISAs, the greatest concentration of A was found in the SDS extraction. The ELI SAs confirmed the changes found with the histological markers. Significant reductions were foun d in the levels of both A40 and A42 in the transgenic mice treated wit h 2H6 at the 14 month time point (Figure 5A & B). These significant reduc tions were maintained 3 months after treatment was stopped. Inte restingly, the 2H6 antibod y directed against A33-40 was able to reduce A42. To ensure that long-term antibody tr eatment did not alter APP processing, we measured steady state levels of APP and BACE in posterior cortex homogenates by western blot analysis. We found that the holo-APP protein levels were unaltered with age or treatm ent (Figure 6A &B). BACE-1 levels, however, increased with age but treatment did not alter these levels (Figure 6A & C). A recent observation in Tg25 76 mice found that BACE-1 expression increased with age in Tg2576 mice (Zhao et al., 2007). In addition to holo-APP and BACE-1, we measured -CTFs and -CTFs and found no significant changes with treatment or age (data not shown).

PAGE 97

90 DISCUSSION This study was designed to investigat e the extent to which changes in the brain environment with aging regulate the extent of amyloid deposition. The process of amyloid deposit ion in the brain can be explained by one of two hypotheses, the accumulation hypothesis or the equilibrium hypothesis. The accumulation hypothesis suggests that deposition of A is a time-dependent process, while the equilibrium hypothesis suggests that depositi on is a factor of the brain microenviroment. We attempt ed to discriminate between these two options by passively immunizing Tg2576 mice with anti-A antibodies (2H6), starting at an age when amyl oid deposition first begins (8 months) and ending treatment 6 months later at an age when control Tg2576 mice have considerable amounts of deposited amyloid. In order to evaluate the pattern of amyloid deposition, we allowed 3 months post-treat ment for amyloid to deposit in these mice. There were two major observa tions taken from this study. The first major observation is that the data confirm that passively administered anti-A antibodies reduce or prevent A deposition in Tg2576 mice. Previous studies have shown that ant i-A immunization prevents amyloid deposition in young PDAPP mice (Schenk et al., 1999). The immunohistochemical data plotted for t he 14 month time point depicts the percent area of A pathology found in the fr ontal cortex of Tg2576 after 6 months of treatment with 2H6 anti bodies compared to control transgenic mice. We found that treatment with the 2H 6 antibody caused significant reductions in total and compact A as measured with anti-A immunohistochemistry and Congo red.

PAGE 98

91 We did not find significant reductions in CD45 expression at the 14 month time point because the total expression of CD45 is very low in both groups at 14 months of age. In addition to hi stopathological findings, biochemical quantification of A40 and A42 with ELISAs showed significant reductions in the treated group at the 14 month time point. The second and principal observation ta ken from this study is that the deposition of amyloid and microglial activa tion never reached the levels found in the control transgenic group. In fact, A pathology was delayed in the mice treated with 2H6 compared to the mice tr eated with control antibody. The data plotted for the 17 month time point por trays the percent area of AD pathology found in the frontal cortex of Tg2576 mi ce 3 months after the cessation of antibody treatment. The control transgenic mice displayed accelerating increases in both diffuse and compact amyloid patholog y, as well as microglial expression between the 14 and 17 month data points. Tg2576 mice that received control antibody exhibited a five-fold increase in total A between 14 and 17 months of age; with an average value of 15% of the frontal cortic al area being littered with amyloid by the age of 17 mont hs, while the total percent area of A was only 6% in the frontal cortex of the 17 month old mice that had been previously treated with 2H6. In addition to total A, we found that the percent area of Congo red positive plaques was two-fold less in t he mice previously treated with 2H6 at the 17 month time point compared to the c ontrol transgenic mice. At the 17 month time point, the control transgenic mice had a 3-fold higher expression of activated microglia in the frontal cortex than t he suppressed mice. A40 and A42 specific

PAGE 99

92 ELISAs confirm these findings and show up to two-fold less A in the 17 month old mice that had been treated with 2H6. Essentially, these data suggest that delayed deposition of amyloid lead to long term delays in AD associated pathology because steady state levels of br ain amyloid are controlled less by the brain microenvironment and more by deposition over time. Several reports have supported the c oncept of the equilibrium hypothesis and have demonstrated a rapid return of A levels to control transgenic levels following an intracranial treatment. Our wo rk previously showed that intracranial administration of LPS resulted in rapid cl earance of diffuse A that returned to control levels by 3 weeks (Herber et al., 2004). Intracranial injections of anti-A antibodies rapidly cleared 50% or more of deposited amyloid, yet resulted in rapid return towards control levels (Chauhan et al., 2004) (Oddo et al., 2004). In addition to these studies, there ar e several anomalies found with anti-A immunotherapy that would suggest that am yloid plaques are rather pliable and support the equilibrium hypothesis. For exampl e, antibodies directed against the 40 amino acid length A peptide result in the clearance of both A40 and A42 (Levites et al., 2006). Our data support the findings by Levites and colleagues as shown with the A ELISA results in Fig 5. The 2H6 antibody used in our study is selective for the 40 amino acid length A peptide yet significantly clears A42. Antibodies generated against A in the human Elan trials were composed almost exclusively of N-terminal anti-A anti bodies, yet ultimately the serum cleared both intact and N-terminally truncated A (Nicoll, James AD/PD Salzburg Meeting March 2007).

PAGE 100

93 Based on the results described above, we anticipated that the equilibrium hypothesis would be at least partially re sponsible for the observed accumulation of amyloid with age. However, there ar e several differences between our study design and the studies that show a rapid re turn in amyloid levels after treatment. One major difference is the age of the animals at the time of treatment. All of the mice used in these studies that suppor t the rapid return of A plaque had abundant plaque deposition at the time of treatment. We purposely chose a 6 month treatment regimen st arting on Tg2576 mice at 8 months to prevent the deposition of amyloid plaque. We assume that we were not only able to prevent the deposition of amyloid but also pr evented the production of A "seeds", analogous to nucleation center s for crystal formation. The accumulation of A is hypothesized to depend on preexisting seeds to recruit more A and form plaques. Growth of the fibrils occurs by assembly of the A seeds into intermediate protofibrils which in turn self-associate to form mature fibers. Fibrillogenesis is a two-st ep reaction involving an initial slow, lag period that reflects the thermodynamic ba rrier to the format ion of a nucleation “seed” followed by a rapid fibril pr opagation and aggregation stage (Jarrett and Lansbury, Jr., 1993). The C-terminal region of the A peptide, amino acids 2940/42, contributes to the key -sheet struct ure as well as fibril assembly via sidechain interactions (Kirschner et al., 1987). This region is also important for the structural transition and t he stability of A fibrils. The anti-A 2H6 antibody targets the C-terminal end of A and may in terfere with the precursor pool that is responsible for the aggregati on of A, thus resulting in a more effective

PAGE 101

94 prevention. It would be interesting to see if anti-A antibodies directed against other epitopes of A would result in si milar accumulation patterns after cessation of treatment. The accumulation hypothesis indica tes that treatments reducing A production or increasing soluble A cl earance might prevent further A deposition, but would not remove exis ting deposits, especially the forms composed of multimeric complexes, su ch as fibrillar deposits with associated neuritic dystrophy and glial activation. Jankowsky and colleagues (2005) created a mouse that overexpresses mutant APP fr om a vector that is regulated by doxycycline (Jankowsky et al., 2005). U nder normal conditions, this mouse model develops A pathology. Administ ration of doxycycline suppresses the mutant APP expression up to 95%, and essent ially turns off the production of A. In this particular experiment, doxycycli ne was administered after the onset of amyloid pathology. The result of the s uppression caused by doxycycline halted the progression of amyl oid pathology; however, existing plaques remained (Jankowsky et al., 2005). Based upon our data and data from Jankowsky and colleagues, amyloid deposits appe ar to require far longer to disperse than to assemble (Jankowsky et al., 2005). We interpret our findings as evidence that AD therapies that alter the production of A by inhibiti ng secretase activity or inhibiting APP expression may not quickl y reverse preexisting pathology. These data strongly support the use of prophylact ic treatments, as it appears that amyloid deposits will require interventions t hat actively clear amyloid as the only means to efficiently reduce brain A in AD.

PAGE 102

95 REFERENCE LIST Alamed J, Wilcock DM, Diamond DM, Gordon MN, Morgan D (2006) Two-day radial-arm water maze learning and memory task; robust resolution of amyloidrelated memory deficits in transgeni c mice. Nat Protoc 1: 1671-1679. Caccamo A, Oddo S, Sugarman MC, Akbar i Y, LaFerla FM (2005) Ageand region-dependent alterations in Abetadegrading enzymes: implications for Abeta-induced disorders. Neurobiol Aging 26: 645-654. Chauhan NB, Siegel GJ (2002) Reversal of amyloid beta toxicity in Alzheimer's disease model Tg2576 by intraventricular antiamyloid beta antibody. J Neurosci Res 69: 10-23. Chauhan NB, Siegel GJ, Lichtor T (2004) Effect of age on the duration and extent of amyloid plaque reduction and micr oglial activation after injection of antiAbeta antibody into the third ventricl e of TgCRND8 mice. J Neurosci Res 78: 732-741. Games D, Adams D, Alessandrini R, Bar bour R, Berthelette P, Blackwell C, Carr T, Clemens J, Donaldson T, Gillespie F, (1995) Alzheime r-type neuropathology in transgenic mice overexpressing V717F beta-amyloid precursor protein. Nature 373: 523-527. Gordon MN, Holcomb LA, Jantzen PT, DiCar lo G, Wilcock D, Boyett KW, Connor K, Melachrino J, O'Callaghan JP, Mor gan D (2002) Time course of the development of Alzheimer-like pathol ogy in the doubly transgenic PS1+APP mouse. Exp Neurol 173: 183-195. Herber DL, Roth LM, Wilson D, Wils on N, Mason JE, Morgan D, Gordon MN (2004) Time-dependent reduction in Abet a levels after intracranial LPS administration in APP transgenic mice. Exp Neurol 190: 245-253. 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, Du ff 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 elevation, and amyloid plaques in transgenic mice. Science 274: 99-102. Iwata N, Takaki Y, Fukami S, Tsubuk i S, Saido TC (2002) Region-specific reduction of A beta-degradi ng endopeptidase, neprilysin, in mouse hippocampus upon aging. J Neurosci Res 70: 493-500.

PAGE 103

96 Iwatsubo T, Odaka A, Suzuki N, Mi zusawa 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 initia lly deposited species is A beta 42(43). Neuron 13: 45-53. Jankowsky JL, Slunt HH, Gonzales V, Savonenko AV, Wen JC, Jenkins NA, Copeland NG, Younkin LH, Lester HA Younkin SG, Borchelt DR (2005) Persistent amyloidosis following suppressi on of Abeta production in a transgenic model of Alzheimer disease. PLoS Med 2: e355. Jarrett JT, Berger EP, Lansbury PT, Jr. (1 993) The carboxy te rminus of the beta amyloid protein is critical for the seeding of amyloid formation: implications for the pathogenesis of Alzheimer's diseas e. Biochemistry 32: 4693-4697. Jarrett JT, Lansbury PT, Jr. (1993) Seeding "one-dimensional crystallization" of amyloid: a pathogenic mechanism in Al zheimer's disease and scrapie? Cell 73: 1055-1058. Johnson-Wood K, Lee M, Motter R, Hu K, Gordon G, Barbour R, Khan K, Gordon M, Tan H, Games D, Lieberburg I, Sc henk D, Seubert P, McConlogue L (1997) Amyloid precursor protein processing and A beta42 deposition in a transgenic mouse model of Alzheimer disease. Pr oc Natl Acad Sci U S A 94: 1550-1555. Kawarabayashi T, Younkin LH, Saido TC, Shoji M, Ashe KH, Younkin SG (2001) Age-dependent changes in brain, CSF, and plasma amyloid (beta) protein in the Tg2576 transgenic mouse model of Alzheimer' s disease. J Neurosci 21: 372-381. Kirschner DA, Inouye H, Duffy LK, Sinclair A, Lind M, Selkoe DJ (1987) Synthetic peptide homologous to beta protein from Alzheimer disease forms amyloid-like fibrils in vitro. Proc Natl Acad Sci U S A 84: 6953-6957. Levites Y, Jansen K, Smithson LA, Daki n R, Holloway VM, Das P, Golde TE (2006) Intracranial adeno-associated viru s-mediated delivery of anti-pan amyloid beta, amyloid beta40, and amyloid bet a42 single-chain variable fragments attenuates plaque pathology in amyloid pr ecursor protein mice. J Neurosci 26: 11923-11928. Maggio JE, Stimson ER, Ghilardi JR, A llen CJ, Dahl CE, Whitcomb DC, Vigna SR, Vinters HV, Labenski ME, Mantyh PW (1 992) Reversible in vitro growth of Alzheimer disease beta-amyloid plaqu es by deposition of labeled amyloid peptide. Proc Natl Acad Sci U S A 89: 5462-5466. Oddo S, Billings L, Kesslak JP, Cr ibbs DH, LaFerla FM (2004) Abeta immunotherapy leads to clearance of early but not late, hyperphosphorylated tau aggregates via the proteas ome. Neuron 43: 321-332.

PAGE 104

97 Prior R, D'Urso D, Frank R, Prikulis I, Cleven S, Ihl R, Pavlakovic G (1996) Selective binding of sol uble Abeta1-40 and Abeta1-42 to a subset of senile plaques. Am J Pa thol 148: 1749-1756. 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, Lieberbu rg 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 Alzheimer-disease-like patholog y in the PDAPP mouse. Nature 400: 173-177. Selkoe DJ (2001) Alzheimer's disease: genes, proteins, and therapy. Physiol Rev 81: 741-766. Zhao J, Fu Y, Yasvoina M, Shao P, Hitt B, O'Connor T, Logan S, Maus E, Citron M, Berry R, Binder L, Vassar R (2007) Beta-site amyloid precursor protein cleaving enzyme 1 levels become elevat ed in neurons around amyloid plaques: implications for Alzheimer's disease pathogenesis. J Neurosci 27: 3639-3649.

PAGE 105

98 Figure 1. THEORETICAL INTERPRETATION OF ABETA ACCUMULATION BASED ON THE PERCENT AREA OF TOTAL A IMMUNOHISTOCHEMISTRY AT 3 AGE POINTS; 8, 14, AND 17 MONTHS The solid line in the figure represents the control gr oup A levels, and the dashed line represents the group treated with 2H6 for 6 months. A. Repr esents the accumulation hypothesis. B. Represents the equilibrium hypothesis.

PAGE 106

99 Figure 2. TOTAL A IMMUNOHISTOCHEMIST RY IS REDUCED FOLLOWING 6 MONTHS OF SYSTEMIC ANTI-A ANTIBODY ADMINISTRATION. Panels A & B show total A immunohistochemistry in the frontal cort ex of Tg2576 mice that received control antibody at 14 months of age and at 17 months of age respectively. Panel C & D show total A immunohistochemistry in Tg2576 mice that received anti-A antibody at 14 m onths of age and at 17 months of age. Scale bar 1cm = 50m. Panel E shows quantification of the percent area occupied by A positive stain in the front al cortex. The solid line shows the value for APP transgenic mice that received control antibody. The dotted line shows the values for APP transgenic mice that re ceived anti-A antibody. ** indicates P<0.01.

PAGE 107

100 Figure 3 CONGO RED HISTOCHEMISTRY REPRESENTING COMPACT AMYLOID DEPOSITS ARE REDUCED AFTER 6 MONTHS OF SYSTEMIC ANTI-A ANTIBODY ADMINISTRAT ION AND THESE REDUCTIONS ARE MAINTAINED 3 MONTHS AFTER THE CESSATION OF TREATMENT. Panels A & B show Congo red histochemistr y in the frontal cortex of Tg2576 mice that received control antibody at 14 months of age and at 17 months of age. Panel C & D show Congo red histochemistr y in Tg2576 mice that received antiA antibody at 14 months of age and at 17 months of age. Scale bar 1cm = 50m. Panel E shows quantific ation of the percent area occupied by Congo red positive plaques in the frontal cortex. The solid line shows the value for APP transgenic mice that received control antib ody. The dotted line shows the values for APP transgenic mice that received anti-A 2H6 antibody. ** indicates P<0.01.

PAGE 108

101 Figure 4. CD 45 EXPRESSION IS REDUCED AFTER 6 MONTHS OF SYSTEMIC ANTI-A ANTIBODY ADMINISTRATION AND THESE REDUCTIONS ARE MAINTAINED 3 MO NTHS AFTER THE CESSATION OF TREATMENT. Panels A & B show CD 45 immunohi stochemistry in the frontal cortex of Tg2576 mice that received cont rol antibody at 14 months of age and at 17 months of age respectively. Panel C & D show CD 45 immunohistochemistry in Tg2576 mice that received anti-A antibody at 14 months of age and at 17 months of age. Scale bar 1cm = 50m. Panel E show s quantification of the percent area occupied by Congo red positiv e plaques in the frontal cortex. The solid line shows the value for APP transgenic mice that received control antibody. The dotted line shows the values for APP transgenic mice that received anti-A antibody. ** indicates P<0.01.

PAGE 109

102 Figure 5. ELISA MEASUREMENTS OF A40 AND A42 AFTER EXTRACTION IN TBS, SDS, AND FORMIC ACID FROM THE ANTERIOR CORTEX OF TG2576 MICE FOLLOWING 6 MONTHS OF PASSIVE IMMUNIZATION WITH EITHER CONTROL IGG OR ANTI-A ANTIBODY 2H6. Panel A represents the stacked total of A40 in 14 and 17 mont h old Tg2576 mice treated with either control IgG or anti-A 2H6. Panel B represents the stacked total of A42 in 14 and 17 month old Tg2576 mice treated with ei ther control IgG or anti-A 2H6. represents p< 0.05 for the stacked averag es consisting of the 3 extractions in mice treated with 2H6 compared to mice treated with control IgG.

PAGE 110

103 Figure 6. WESTERN BLOT ANALYSIS OF HOLO APP & BACE-1 IN 14 AND 17 MONTH OLD TG2576 MICE THAT RECEIVED CONTROL IGG OR ANTI-A 2H6 The blots for APP and BACE-1 are shown in panel A for posterior cortex. Panel B & C show the quantification of western blots for APP and BACE-1. The levels for each time point were adjust ed to the relative GAPDH levels, which were used as a protein loading control.

PAGE 111

104 CONCLUSIONS Alzheimer’s disease (AD) is the most common cause of age-related dementia with a prevalence approaching 40 to 50% by age 80. At present, 4 million Americans are affected with AD at an estimated annual care cost of almost 100 billion dollars. The number of individuals 65 years and older is growing rapidly due to an increase in t he general average life expectancy. As a result, it is estimated that the total incidence of AD will quadruple by the year 2050 (Hebert et al., 2003). Therefore, it is urgent to find a means of preventing, delaying the onset, or reversing the course of AD. The presence of high densities of neurit ic plaques composed of A in the cerebral cortices is a criterion for the post-mortem diagnosis of AD. The view that A deposition drives the pathogenesis of AD (amyloid hypothesis) has received support from a wide range of molecular, genetic, and animal studies (Selkoe, 2001). The first form of evidence came from the localization of the APP gene to chromosome 21. AD-like neuropathology is invariably seen in Down’s syndrome, and results from increased APP expressi on and consequent higher A levels (Mann et al., 1984) (Glenner and Wong, 1984). Previous studies have shown that synthetic A peptides are neurotoxic to hippocampal and cortical neurons, both in culture and in vivo (Pike et al., 1993) (Lambert et al., 1998) (Deshpande et al., 2006). Inherited mutations in the APP gene occur within the A region and alter the amount or aggregation properties of A and are sufficient to cause early-

PAGE 112

105 onset AD (Goate et al., 1991) (Levy et al., 1990). In addition to the APP gene, inherited mutations within the preseni lin 1 and 2 genes increase the A42/40 ratio and cause early forms of AD (Kumar -Singh et al., 2006). The discovery of these mutations has allowed for the dev elopment of mice transgenic for mutant human APP. These APP mutant mice show a time-dependent increase in extracellular A and develop certain neur opathological and behavioral changes similar to those seen in AD (Hsiao et al., 1996) (Games et al., 1995). The amyloid cascade hypothesis has been a driving factor in the field of AD therapeutics and has led to the devel opment of drugs which target the production and clearance of A; one such approach is immunotherapy. Since the first report that active i mmunization with fibrillar A 42 in PDAPP transgenic mice prevented the development of amyloi d plaques, neuritic dystrophy, and astrogliosis (Schenk et al., 1999), ther e have been a plethora of data published from studies conducted on APP transgenic mice and humans regarding the use of both active and passive immunization fo r the treatment of AD. Although there is still no consensus on how A immunot herapy works, the data thus far have given us insight into possible mechani sms of action of ant ibody mediated A removal from the brain while unveiling adverse effects su ch as cerebral amyloid angiopathy (CAA) and microhemorrhage. Passive immunization has allowed sci entists to examine the in vivo binding properties, pharmacokinetics, brain penetrance, and changes in A levels after peripheral administration. The average half-life of a monoclonal antibody (mAb) in mouse plasma has been reported to be ~1 week (Vieira and

PAGE 113

106 Rajewsky, 1986). After an intraperitoneal in jection, no more than 50% of the monoclonal anti-A antibody is bioavailable in the plasma (Levites et al., 2006). Once in the plasma, the anti-A antibodi es bind circulating A. We and many others demonstrated that twent y-four hours after anti-A antibody administration, both A40 and A42 significantly increase in serum. (DeMattos et al., 2002) (Wilcock et al., 2004b) (Levites et al., 2006) (Paper 1). Antibody administration may clear brain A without the mAb actually entering the brain. When mAbs directed against the central domain of A were peripherally administered to 3 month ol d PDAPP mice, the result was a rapid 1000 fold increase in plasma A and reductions in amyloid burden in the brain. It is thought that the antibody sequeste red plasma A, which disrupted the equilibrium of A in the CNS and plasma, c ausing an efflux of A out of the brain into the periphery (DeMattos et al., 2002). However, here we show in paper 1, evidence which would appear to contradict the peripheral sink mec hanism. In a deglycosylated anti-A dose response study, we show that the highest dose of 30 mg/kg was able to sequester the most A in the serum while the intermediate dose 10 mg/kg and lowest dose of 3 mg/kg sequestered intermediary and smaller A levels, respectively (Paper 1; Figure 1). If the peripheral sink mechanism were playing a major role in A clearan ce from the brain, one wo uld predict the greatest reductions in A would occur with the hi ghest dose. Yet in fact, we found the exact opposite. The lowest dose cleared the most diffuse and compact A from the brain while the highest dose failed to significantly reduce amyloid levels

PAGE 114

107 compared to mice treated with control Ig G. Therefore, changes in serum A levels do not show a relationship with r eductions in brain A in APP transgenic mice after systemic treatment with anti-A antibodies. A possible explanation for these re sults may be that the passively administered anti-A antibodies bind already circulating A. Previous studies have established that A has a very short half-life in the plasma. When free A is injected intravenously into mice, it is cleared within 10 minutes (Ghiso et al., 2004). However, peripheral administration of anti-A antibodies creates a stable mAb:A complex in the plasma and the co mplex is cleared slowly. The antibody bound A has a half-life of 57 days (Levites et al., 2006). Thus the rapid rise in plasma A observed is attributable to pr olongation of the half-life of A bound to antibody. This phenomenon may be unique to APP transgenic mice since the over expression of APP and the accumulation of A are not specific to the CNS but also occur in several organ systems (Kawarabayashi et al., 2001). APP transgenic mice have higher endogenous circulating A levels than humans. Therefore, one might not obser ve the spike in plasma A following immunization in humans (Kawarabayashi et al., 2001), as reported by Hock et al (2002). A small fraction of anti-A antibod ies can penetrate the BBB; when quantified these levels were <0.1% of t he total administered dose (Bard et al., 2000) (DeMattos et al., 2002) (Banks et al., 2005). However, even these low levels of antibody are sufficient to elic it a profound reduction in pathology. After anti-A antibodies cross the blood brai n barrier, there were two suggested mechanisms of action for the removal of am yloid deposits. The first involves the

PAGE 115

108 catalytic disaggregation of A fibrils or neutralization of A o ligomers leading to clearance from the brain (Solomon et al., 1997) (Klyubin et al., 2005). The second mechanism relates to microglial -mediated phagocytosis or degradation of A by activated microglia. The constant region (Fc) on the anti-A antibody activates microglia via the Fc -receptor. Fc -activation stimulates microgl ial-mediated phagocytosis of A (Schenk et al., 1999) (Bard et al., 2000) (Wilcock et al., 2004a). The studies conducted by Wilcock and colleagues with monoclonal anti-A antibodies have led us to conclude that the removal of fibrillar A deposits in the APP transgenic mice is facilitated when microglial cells ar e activated. First, the time course of Congo red clearance following intracranial injections of anti-A antibodies coincides with elevation of CD45 expressi on in microglia (Wilcock et al., 2003). Second, co administration of dexamethasone with intracrani al injections of intact IgG or Fab2 fragments results in the prevention of microglia activation, and the clearance of fibrillar A deposits is block ed (Wilcock et al., 2004a). Thirdly, a time course of passively administered ant ibodies in Tg2576 mice compared the effects of 1, 2 or 3 months of tr eatment on histopathology and behavior in Tg2576 mice. The results showed a complex microglial response that correlated with amyloid reductions and cognitive im provements (Wilcock et al., 2004b). In addition to the evidence of microglial in volvement in A clearance in APP mice, neuropathological investigations of 3 patients immunized with AN-1792 showed less than expected A pathology in focal regions of the brain with remaining microglia filled with A (Masliah et al., 2005) (Nicoll et al., 2003).

PAGE 116

109 More recently, microglial activation has been shown surrounding amyloidcontaining blood vessels follo wing systemic immunization. A deposition in the cerebral vasculature is a condition refe rred to as cerebral amyloid angiopathy (CAA). Severe CAA disrupts the integrity of the blood vessels and amplifies the potential for microhemorrhages. In a th ree month passive immunization time course study with old Tg2576 mice and a C-terminal anti-A antibody, Wilcock and colleagues demonstrat ed an antibody-exposure-ti me-dependent increase in vascular deposition with a concommitant decrease in parenchymal A levels (Wilcock et al., 2004b). CD45 immunohistoc hemistry showed that these amyloidburdened blood vessels were affiliated with increased microglial activation (Wilcock et al., 2004b). Furthermore, a study with APP23 transgenic mice, a model with parenchymal and significant vascu lar A deposition, showed that the incidence and severity of microhemorr hage increased more than twofold after 6 weeks of passive immunization with an N-terminal anti-A antibody (Pfeifer et al., 2002). Similar to the result s obtained by Pfeifer et al. (2002), increased incidence of CAA and microhemorrhage were also observed in aged PDAPP mice treated for 6 weeks with an N-termi nal anti-A antibody (Racke et al., 2005). An important question following t hese observations is what mechanism caused the antibodies to exacerbate the CAA-associated microhemorrhage? The microglial association with amyloid-burdened blood vessels suggests that antibody-mediated microglial activation via the Fc receptor may play a dominant role. To answer this question, we have tested a deglycosylated C-terminal anti-

PAGE 117

110 A antibody. The deglycosylation of car bohydrate side chains on the Fc portion of an anti-A antibody greatly reduces the affinity of the antibody for Fc receptors, particularly murine Fc RIIb and Fc RIIIa, on effector cells like microglia; yet deglycosylated antibodies are fu lly capable of binding to A with an identical affinity as the intact antibody (Ravetch, 1997) (Gessner et al., 1998). The deglycosylated antibody offers a unique abi lity to ascertain which features of the microglial reaction to antibody adm inistration are caused by opsonized antigen binding to the effector molecules, by comparison to mice injected with intact antibodies. Immunohistochemical analyses of APP brain sections after intracranial injections of intact or deglycosylated antibody confirmed that deglycosylation does not activate microglia to the same extent when compared to mice treated with control or native intact anti-A IgG as measured with antibodies against Fc RII/III and CD45 (Carty et al., 2006). Furthermore, deglycosylation of the Fc portion on the anti-A antibody signi ficantly reduces the severity of CAA and microhemorrhage compared to nat ive antibodies when passively administered to Tg2576 mice, albeit with sli ghtly lower clearance of parenchymal amyloid deposits (Wilcock et al., 2006). O ne explanation for our prior results is that by deglycosylating the A antibody we are able to mitigate specific antibody-mediated mechanisms t hat contribute to the accumulation of CAA and microhemorrhage. Alternatively, we suggest that the deglycosylated antibody may remove A from the brain at a slower rate, resulting in less sa turation of vascular efflux pathways and minimized accumulation of vascular deposits. One means of

PAGE 118

111 resolving the qualitative (specific me chanism) versus quantitative (slower removal) explanations of the effects of antibody deglycosylation on vascular amyloid accumulation would be to increas e the rate of amyloid removal with higher doses of deglycosylated antibodies. If a higher rate of amyloid removal by deglycosylated antibody increased vascula r deposits like native antibodies, then the quantitative explanation would be most likely. Here we examined the rate of A cl earance from the brain and its effect on CAA and microhemorrhage by passi vely administering 3 doses of deglycosylated anti-A antibody in old Tg25 76 mice for 3 months in paper 1. We chose the 3 month treatment regimen because previous studies have shown significant reductions in parenchymal amyl oid at this time point as well as increased CAA and microhemorrhage (Wilco ck et al., 2004b). Based on prior studies which validate that 0.1% of peripherally adminis tered antibody enters the CNS, we designed a dose response study using the deglycosylated A antibody with the notion that higher doses of per ipherally administered antibody would enter the brain and clear amyloid at a hi gher rate, while the smaller doses would clear amyloid at the slower rates (B anks et al., 2002) (Bard et al., 2000). Somewhat paradoxically, we found that long term systemic administration of low doses of deglycosylated anti-A are more effective at reducing amyloid deposits and reversing cognitive deficits in old APP transgenic mice when compared to animals treated with interm ediate or higher doses of antibody (Paper 1). Interestingly, all groups had similar increas es in CAA despite the differences in plaque clearance (Paper 1; Figure 5 & 6C). The number of microhemorrhages

PAGE 119

112 was less than one profile per section for all groups, considerably less than that found in prior studies with intact ant ibody (Paper 1; Figure 7). A possible explanation for the inverse dose res ponse may be that ex cess antibody unbound to antigen can interfere with antibody-m ediated A clearance, possibly by saturating the FcRn antibody transporter. It is well known that the LRP receptor mediates the efflux of A from the brain and that RAGE mediates the influx of A across the blood brain barrier (BBB) (Deane et al., 2003) (Deane et al., 2004a). More recently, FcRn expression was found at the brain micr ovasculature and choroid plexus epithelium on the BBB (Lazarov et al ., 2005) and may mediate the 'reverse transcytosis' of IgG and immune complexe s from the brain to blood direction (Schlachetzki et al., 2002). Deglycosylat ion of antibodies does not change the interaction of immunoglobin with the FcRn (Hobbs et al ., 1992). Previous studies have shown that receptor mediated A efflux mechanisms can become saturated with high levels Fc fragments or intact Ig G but not by high concentrations of Fab2 fragments or albumin (Zhang and Pard ridge, 2001). Furthermore, the FcRn transport system was found to be t he main mechanism mediating the transcytosis of A-anti-A immune comple xes from the brain to blood in old Tg2576 mice (Deane et al., 2005). This gr oup also found that antibody-mediated A clearance from the brain is abolished in old FcRn -/mi ce (Deane et al., 2004b). Similar to Zhang and Pardridge, Deane and colleagues demonstrated that the addition of excess IgG inhibi ted the clearance of immune complexes from the brain. Therefore, high concentrations of immu noglobin will saturate the

PAGE 120

113 FcRn receptors with antibody not bound to antigen, hinder the FcRn binding of A-anti-A immune complexes, and ultimate ly result in the inhibition of A clearance from the brain. The saturation of FcRn may explain why high doses of deglycosylated antibody are not as effect ive at reducing diffuse and fibrillar A compared to antibodies administered at lowe r doses. The saturati on of this efflux system may also result in the ac cumulation of A along the brain microvasculature and lead to the formation of CAA. The 3 month time point chosen for t he dose response study allowed for us to examine changes in the distribution and levels of A, cognitive improvements, and microhemorrhage. However, based up the earlier time course study with intact anti-A antibodies, microglial ma rkers return to baseline at 3 months (Wilcock et al., 2004b). In the dose respons e study, we showed no differences in CD45 expression between the 3 doses of deglycosylated antibody and the control group at the 3 mont h time point (Paper 1; Ta ble 1). Even though the deglycosylated antibody was designed to avoid activation of the Fc receptor, the antibody may still elicit changes in micr oglial expression indirectly. Further experimentation is required to exami ne changes in microg lial phenotypes with the administration of a degl ycosylated anti-A antibody in the form of a time course study with a similar experiment al design as Wilcock et al. (2004a). Elucidating the series of changes in t he microglial phenotype that occur during the active phase of amyloid removal will be usef ul in identifying features critical to clearance of antigen-antibody complexes, and in identifying changes associated with the buildup of vascular deposits.

PAGE 121

114 In addition to the further experimentati on in the form of a time course, brain leptomeningeal vessels could be isol ated from the mice used in the time course study to examine changes in A content, vascular markers, as well as FcRn expression. The surface leptomeni ngeal vessels of t he transgenic mouse brain provide an ideal system to examine CAA, becaus e vessel segments of up to several hundred micrometers in length c an be observed (Prada et al., 2007). Besides the antibody dose and the subsequent rate of A clearance from the brain, the epitope that ant i-A antibodies recognize ma y be a critical factor in parenchymal A clearance and redistributio n. Intracranial injections of Nterminal, mid domain, and C-terminal antiA antibodies resulted in similar reductions in diffuse and fibrillar A as measured with A immunohistochemistry and thioflavin-S histochemistry, respec tively (Wilcock; data not published). Several passive immunization studie s with N-terminal, mid domain, and Cterminal antibodies have shown signific ant reductions in am yloid plaque burden (Bard et al., 2003) (DeMattos et al ., 2001) (Wilcock et al., 2004b). Independent studies have demonstrated increased CAA and microhemo rrhage following passive administration of Nand C-te rminal antibodies (Wilcock et al., 2006) (Pfeifer et al., 2002) (Racke et al., 2005). In order to determine if increas ed vascular amyloid and hemorrhage are conditions that result from epitope specif icity, we directly compared aged Tg2576 mice passively immunized with antibodies directed against the N-terminal of A15, the mid domain of A16-28, or the C-terminal of A33-40. After 3 months of weekly intraperitoneal injections, the mice were tested in a radial arm water maze

PAGE 122

115 paradigm for learning and memory. T he behavioral test revealed cognitive improvements in mice tr eated with the N-terminal and C-terminal anti-A antibodies as compared to mice treated with control Ig G, while the mice treated with mid domain antibody failed to show behavioral improvements (Paper 2; Figure 1). Despite the differences f ound with the behavioral data, all three antibodies resulted in similar reductions of diffuse A (Paper 2; Figures 2 & 3). However, only the Nand C-terminal antibodies were able to significantly reduce fibrillar A as measured with Congo red hi stochemistry (Paper 2; Figures 4 & 5). Interestingly, of the 3 antibodies te sted, the C-terminal antibody caused the greatest reductions in A with more than 90% of the parenchymal deposits removed (Paper 2; Figure 5B). These reduc tions were associated with increases in vascular A (Paper 2; Figure 5C). Another significant observation from this experiment was that behavioral improvements correlated with reductions in compact amyloid, which suggests that compact amyloid plaques contain a species of A that is critical for acquisition of memory. Both the Nand C-te rminal antibody resulted in significant incidences of microhemorrhage (Paper 2; Figure 6). We did not observe substantial A clearance with dramatically less vascular amyloid accumulation with one particular antibody. These data s uggest that the degree of parenchymal A clearance determines the extent of vascular A accumulation and hemorrhage development; epitope specificity is not critical in directing the vascular accumulation.

PAGE 123

116 The mechanism by which anti-A ant ibody therapy results in increased CAA remains unclear. One hypothesis is that the microg lia are capable of phagocytosing the amyloid plaques but are unable to degrade them so they transport the amyloid to the vessels where t hey dispose of it into the vessel wall. Support for this hypothesis comes from CD45 immunohistochemistry which showed that, despite overall microglial acti vation being reduced to control levels following three months of anti-A antibody treatment, there appeared to be high levels of microglial activation around t hose vessels containing amyloid (Wilcock et al., 2004b). Deglycosylated antibodies fu rther support this contention, as the severity of CAA and occurrence of micr ohemorrhage are significantly mitigated when Fc receptor activation is abated (Carty et al., 2006) (Wilcock et al., 2006) (Paper 1). A second hypothesis for increased CAA after immunization is that the antibody binds to A already present in the vasculature. In vitro and in vivo binding assays showed that N-terminal but not mid-domain anti-A antibodies directly bind to A in the vessels (R acke et al., 2005). The formation of this immune complex results in a local inflamma tory reaction, which might destabilize the already weakened vessel (Pfeifer et al ., 2002) (Burbach et al., 2007). It is also plausible that receptor-medi ated A clearance mechanisms become saturated or overwhelmed with increased mAb:A immune complexes, increased free antibody, or increased soluble A in the brain following administration of anti-A antibodies.

PAGE 124

117 In sporadic AD, as well as in Tg2576 mice, vascular amyloid is mainly composed of A1-40 with a lesser, vari able amount of A1-42 (Kawarabayashi et al., 2001). In sporadic AD, leptom eningeal amyloid exhibits minimal modifications while parenchymal plaqu es commonly exhibit N-terminal degradations and posttranslation modificati ons (Roher et al., 1993) (Kuo et al., 1997). Interestingly, the composition of vascular amyloid observed in the patients vaccinated with AN 1792 resembled that of the typical AD parenchymal amyloid plaque with N-terminal and chemical modifications (P atton et al., 2006). These observations further suggest a mobiliz ation of A from the parenchyma to the vasculature with immunization. Additi onally, vascular A1-42 was elevated in the individuals vaccinat ed with AN 1792 compared to nonimmunized AD controls (Nicoll et al., 2006). Comprehending the ultimate effects of plaque amyloid dispersal on brain vascular function is vital. Parenchymal A may exert deleterious effects on vascular endothelial and smooth muscle cell metabolism. Soluble A is a potent vasoconstrictor capable of altering cerebr al blood flow, (Thomas et al., 1996) and A peptides are also powerful angiogenes is inhibitors (Paris et al., 2004). Changes in vascular A content have not been closely examined in APP transgenic mice following immuniza tion with anti-A antibodies. Several important questions can be te sted using anti-A antibodies such as: what factor/s influence the deposition of A? What happens after immunotherapy is stopped? Is A deposition a fa ctor of time or age? In order to answer such questions, we first proposed two hypotheses regarding the

PAGE 125

118 deposition of A. One hypothesis assume s that A deposition is a result of production exceeding clearance mechanisms by a small amount, and the excess becomes converted to a form that deposits and builds up in a time dependent manner (Maggio et al., 1992) (Prior et al ., 1996). This statement describes the accumulation hypothesis (Paper 3; Figure 1A). The other hypothesis suggests that the rate of A pro duction, degradation and remova l is altered with age, and was termed the equilibrium hypothesis (Paper 3; Figure 1B). We tested these hypotheses by delay ing A deposition in Tg2576 mice using passive immunization with a C-terminal anti-A antibody starting at an age when amyloid deposition first begins (8 months) and ending treatment 6 months later at an age when control Tg2576 mi ce have considerable amounts of deposited amyloid. In order to evaluate t he pattern of amyloid deposition, we allowed 3 months post-tr eatment for amyloid to deposit in these mice. The variables measured in this exper iment were total A, which consists of diffuse and compact amyloid, C ongo red positive com pact amyloid, and markers of microglial activation. The per cent areas of each of these variables were plotted on a line graph to reveal a pattern of deposition that resembled the theoretical graph for the accumulation hypot hesis (Paper 3; Figures 2, 3, & 4). Following successful suppression of A deposition with anti-A immunotherapy, cessation of treatment resulted in delayed diffuse and compact A accumulation. The deposition of amyloid never reached the levels found in Tg2576 mice treated with control IgG. The cont rol transgenic mice displayed accelerated increases in

PAGE 126

119 both diffuse and compact amyloid patholog y, as well as microglial expression between the 14 and 17 month data points (P aper 3; Figures 2, 3, & 4). We also examined individual changes in A40 and A42 with ELISAs and confirmed the trends found with the histologic al markers. Significant reductions were found in the levels of both A 40 and A42 in the transgenic mice treated with 2H6 at the 14 month time point (Paper 3; Figures 5A & B). These significant reductions were maintained 3 months afte r treatment was stopped. Interestingly, the 2H6 antibody directed against A33-40 was able to reduce A42. The results from this study have provided us with the information to suggest that A accumulates as a functi on of time. A accumulation in a timedependent manner results in am yloid plaques that are less malleable, and active clearance of A may be necessary to remove existing plaques. The accumulation hypothesis indicates that tr eatments which reduce A production or increase soluble A clearanc e might prevent further A deposition, but would not remove existing deposits. We interpre t our findings as evidence that AD therapies that alter the production of A by inhibiting secretase activity or inhibiting APP expression may not quickl y reverse preexisting pathology. These data strongly support the use of prophylact ic treatments, as it appears that amyloid deposits will require interventions t hat actively clear amyloid as the only means to efficiently reduc e brain A in AD. To further explain the pattern of amyloid accumulation following immunization, we assume that the administration of anti-A mAbs, before the deposition of amyloid, altered some select pool or species of A that was present

PAGE 127

120 at low abundance and critical for deposition. It is cr itical that biochemical analysis, on the remaining tissue from t he accumulation study, be performed to determine changes in forms of A that ac t as seeds and drive A accumulation. We have found that cognitive improvements in aged Tg2576 mice accompany reductions in compact amyloi d plaques after chronic administration of anti-A antibodies (Paper 1 & 2). In support of these findings, active immunization with A of double mutant APP TgCRND8 mice partially prevented the development of reference memory defic its in a water maze task, with only a 50% reduction in the size and number of dense core amyloid deposits, and no effect on the total soluble pool of A in brain (Janus et al., 2000). The researchers concluded that the prevention of memory deficits was due to the reduced amyloid pathology seen in thei r immunized mice. However, compact deposits are likely to be surrounded by a num ber of smaller, more diffusible, oligomeric assemblies. Chronic active i mmunization had similar beneficial effects on memory impairment in two different strains of transgenic mice as assessed using a radial-arm water maze (Morgan et al., 2000). Notably, treatment did not affect amyloid pathology in the same way in the two strains of mice. Immunized APP+PS1 double transgenic mice showed a reduc tion in diffuse (nonfibrillar) A deposits in the cerebral cortex and hippoc ampus, but not in amyloid (fibrillar A) deposits. In contrast, Tg2576 APP transgenic mice showed a small but statistically significant reduction in cortic al amyloid burden, suggesting that active immunization reduces the development of fibrillar A deposits in this mouse strain. These authors conclu ded that active immuniza tion prevents memory

PAGE 128

121 deficits by altering either brain amyloid pathology or an unknown pool of nondeposited A, perhaps a soluble pool of A. Taken together with our data, these results suggest that anti-A immunotherapy is able to remove the A moiety that is responsible for cognitive decline. Ho wever, the relationship between soluble and insoluble brain A concentrations and memory impairment in transgenic mice is complex. Further examination into the redistribution and reduction of different forms of A will be useful to el ucidate what form/s of A accompanies the changes in memory performance in APP transgenic mice following passive immunization. Research on AD seeks to answer t he central question: what causes the impairment of episodic memory? It was widely believed that the fibrillar A peptide found in neuritic plaques was t he neurotoxic species and caused the neurodegeneration which accompanies the di sease. However, many apparently healthy older humans have substantial amount s of amyloid in their cerebrum at post-mortem examination, yet do not show the signs and symptoms of AD. Furthermore, there are repor ts of weak quantitative co rrelations between amyloid plaque counts in post-mortem brain se ctions and the extent of cognitive symptoms measured pre-mortem (Katzman, 1986). There is also evidence that APP transgenic mice show memory impairment before the first signs of amyloid deposition (Moechars et al., 1999). It has even been postulated that the large A aggregates play a neuroprotective role by sequestering neurotoxic A forms, although they are surrounded by dystrophic neurites. However, insoluble protein aggregates are likely to be surrounded by a number of smaller, more diffusible

PAGE 129

122 assemblies. Therefore, the toxic moiety responsible for synaptic dysfunction and neuronal cell loss still remains to be identif ied. Currently, the focus has shifted from fibrillar A as the neurotoxic spec ies to soluble A oligomers for their subsequent ability to cause neuronal injury. A large and confusing body of literatur e describes many types of assembly forms of synthetic A, including protofib rils (PFs), A-derived diffusible ligands (ADDLs), and many lengths of oligomeric A Protofibrils are intermediates of synthetic A fibrillization t hat can continue to polymeriz e to form amyloid fibrils and have a -sheet structure. They are detected with Congo red or thioflavin-S (Harper et al., 1997) (Walsh et al., 1997) ELISAs have shown that soluble oligomers correlate much better with the presence and degree of cognitive deficits than do plaque counts (Naslund et al., 2000). Small oligomeric species that might affect neural signal-transduction pathwa ys have been designated as ADDLs (Gong et al., 2003). Lesne and colleagues searched for the appearance of an A species that coincided with the first observed changes in spatial memory in Tg2576 mice (Lesne et al., 2006) They found that levels of only nonamer and dodecamer forms of A corre lated with impairm ent of spatial memory. The purified dodecamer assemb ly (A*56) was injected into the ventricle of pre-trained, wild-type rats and caused a dramatic debilitation in spatial memory performance (Lesne et al., 2006). Although levels of A*56 coincide with spatial memory deficits, several studies have shown impaired performance in hippocampal-dependent fear conditioning and long-term potentiation (LTP) in Tg2576 mice, at ages long before the detection of the

PAGE 130

123 dodecamer species (Dineley et al., 2002). Therefore, at the current stage of research, one should not conclude that eit her large, insoluble deposits or small, soluble oligomers represent the sole neurotoxic entity. However, current evidence points to the detrimental role of diffusible oligomers in the early, presymptomatic stages of AD. Numerous strategies to prevent A aggregation and accumulation are being evaluated as ways to treat or pr event AD. Compounds that decrease the production of soluble A monomers and pot entially inhibit the formation of soluble oligomers, are the or secretase inhibitors. Eli Lilly and Company is involved in Phase II trials testing the -secretase inhibitor, LY450139 dihydrate. This compound has been found to reduce the ra te of formation of A in vitro and in vivo (Siemers et al., 2006). Wyeth Research is testing another -secretase inhibitor that has previously been show n to reverse cognitive deficits in Tg2576 mice (Comery et al., 2005). Unfort unately, secretases are expressed ubiquitously, and process many important substrates which when inhibited, may result in adverse effects. In addition, recent evidence implies that existing plaques are not pliable and may requi re an active means of clearance. Therefore, the inhibition of monomeric A production may not be sufficient to reverse the disease process (Paper 3) (Jankowsky et al., 2005). Because passive administration of monoclonal anti-A antibodies works as effectively as active immunization in APP mice, it is generally acknowledged that it is the anti-A ant ibody response that mediates the effects of active immunization. In contrast to the dat a discussed above, several studies show

PAGE 131

124 reductions in amyloid levels as well as cognitive improvement s without the direct target of A. In an AD patient wit h stroke, Akiyama and McGeer (2004) observed local reductions in senile plaques in cortical areas affected by ischemia and suggested that the obser ved reductions in A were a result of an inflammatory response. Furt hermore, this group proposed t hat the effects of antiA immunotherapy may be more related to nonspecific immune activation than to actions directed against the A peptide (A kiyama and McGeer, 2004). A series of agents that cause general immune activation often including T cells, have been found to reduce amyloid loads. These includ e Borna virus infection (Stahl et al., 2006), glatiramer acetate (Butovsky et al., 2006) and a proteosome adjuvant (Monsonego et al., 2006). Nasal vaccinat ion with a proteosome-based adjuvant alone and in combination with glatiramer acetate decreased A plaques in an APP transgenic mouse model (Frenkel et al., 2005). The clearance was demonstrated in B-cell deficient mice and in mice with an intact immune system. The suggested mechanism of action wa s antibody-independent and involved the induction of T cells and activation of microglia (Frenkel et al., 2005). Tg2576 mice inoculated with the Borna virus showed a persisting, subclinical infection of cortical and limbic brain areas characterized by slight T-cell infiltrates, expression of cytokines, and massive microglia l activation in the hippocampus and neocortex (Stahl et al., 2006). Interest ingly, Stahl and colleagues revealed a decrease in thioflavin-S positive plaques companied with increased CAA following infection with the borna virus (Stahl et al., 2006).

PAGE 132

125 Overall, we have shown evidence t hat A accumulates in a timedependent manner. As a result, active clearance methods are required to successfully reverse existing pathology. Moreover, it is important to emphasize the requirement for precise titration of dose-response characteristics because saturation of influx and effl ux mechanisms may play a majo r role in the efficacy of A clearance. We also show that epitope s pecificity is not critical in directing vascular A accumulation but the degree of parenchymal A clearance determines the extent of vascula r A accumulation and hemorrhage development. A great deal of experimentati on is still required in order to elucidate why CAA is exacerbated following immunization.

PAGE 133

126 REFERENCES CITED Akiyama H, McGeer PL (2004) Specific ity of mechanisms for plaque removal after A beta immunotherapy for Alzheime r disease. Nat Med 10: 117-118. Bacskai BJ, Kajdasz ST, McLellan ME, Games D, Seubert P, Schenk D, Hyman BT (2002) Non-Fc-mediated mechanisms are involved in clearance of amyloidbeta in vivo by immunotherapy. J Neurosci 22: 7873-7878 Banks WA, Pagliari P, Nakaoke R, Morl ey JE (2005) Effects of a behaviorally active antibody on the brain uptake and clearance of amyloid beta proteins. Peptides 26: 287-294. Banks WA, Terrell B, Farr SA, Robinson SM, Nonaka N, Morley JE (2002) Passage of amyloid beta protein antibody across the blood-brain barrier in a mouse model of Alzheimer's di sease. 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, 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-lik e neuropathology. Proc Natl Acad Sci U S A 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, Khol odenko D, Lee M, Lieberburg I, Motter R, Nguyen M, Soriano F, Vasquez N, Wei ss K, Welch B, Seubert P, Schenk D, Yednock T (2000) Peripherally administ ered antibodies against amyloid betapeptide enter the central nervous syst em and reduce pathology in a mouse model of Alzheimer disease. Nat Med 6: 916-919. Brambell FW, Hemmings WA, Morris IG (1 964) A theoretical model of gammaglobulin catabolism. Nature 203: 1352-1354. Burbach GJ, Vlachos A, Ghebremedhin E, Del Turco D, Coomaraswamy J, Staufenbiel M, Jucker M, Deller T ( 2007) Vessel ultrastructure in APP23 transgenic mice after passive antiAbeta immunotherapy and subsequent intracerebral hemorrhage. N eurobiol Aging 28: 202-212. Butovsky O, Koronyo-Hamaoui M, K unis G, Ophir E, Landa G, Cohen H, Schwartz M (2006) Glatiram er acetate fights against Alzheimer's disease by inducing dendritic-like microglia expressing insulin-like growth factor 1. Proc Natl Acad Sci U S A 103: 11784-11789.

PAGE 134

127 Caccamo A, Oddo S, Sugarman MC, Akbar i Y, LaFerla FM (2005) Ageand region-dependent alterations in Abetadegrading enzymes: implications for Abeta-induced disorders. Neurobiol Aging 26: 645-654. Carty NC, Wilcock DM, Rosenthal A, Grim m J, Pons J, Ronan V, Gottschall PE, Gordon MN, Morgan D (2006) Intracranial administration of deglycosylated Cterminal-specific anti-Abeta antibody effi ciently clears amyloid plaques without activating microglia in amyloid-deposit ing transgenic mice. J Neuroinflammation 3: 11. Chauhan NB, Siegel GJ, Lichtor T (2001) Distribution of intraventricularly administered antiamyloid-beta peptide (Abe ta) antibody in the mouse brain. J Neurosci Res 66: 231-235. Chauhan NB, Siegel GJ, Lichtor T (2004) Effect of age on the duration and extent of amyloid plaque reduction and microgl ial activation after injection of antiAbeta antibody into the third ventricle of TgCRND8 mice. J Neurosci Res 78: 732-741. Chauhan NB, Siegel GJ (2005) Efficacy of anti-Abeta ant ibody isotypes used for intracerebroventricular immunization in Tg CRND8. Neurosci Lett 375: 143-147. Chen G, Chen KS, Knox J, Ingl is J, Bernard A, Martin SJ, Justice A, McConlogue L, Games D, Freedman SB, Morris RG (2000) A learning deficit related to age and beta-amyloid plaques in a mouse model of Alzheimer's disease. Nature 408: 975-979. Chen KS, Masliah E, Grajeda H, Guido T, Huang J, Khan K, Motter R, Soriano F, Games D (1998) Neurodegenerative Alzhei mer-like pathology in PDAPP 717V->F transgenic mice. Prog Brain Res 117: 327-334. Chishti MA, Yang DS, Janus C, Phinney AL, Horne P, Pearson J, Strome R, Zuker N, Loukides J, French J, Turner S, Lozza G, Grilli M, Kunicki S, Morissette C, Paquette J, Gervais F, Bergeron C, Fraser PE, Carlson GA, George-Hyslop PS, Westaway D (2001) Early-onset am yloid deposition and cognitive deficits in transgenic mice expressing a double mutant form of amyloid precursor protein 695. J Biol Chem 276: 21562-21570. Comery TA, Martone RL, Aschmies S, Atchison KP, Diamantidis G, Gong X, Zhou H, Kreft AF, Pangalos MN, Sonne nberg-Reines J, Jacobsen JS, Marquis KL (2005) Acute gamma-secretase inhi bition improves contextual fear conditioning in the Tg2576 mouse model of Alzheimer's disease. J Neurosci 25: 8898-8902. Deane R, Du YS, Submamaryan RK, LaRue B, Jovanovic S, Hogg E, Welch D, Manness L, Lin C, Yu J, Zhu H, Ghiso J, Frangione B, Stern A, Schmidt AM, Armstrong DL, Arnold B, Liliensiek B, Nawr oth P, Hofman F, Kindy M, Stern D,

PAGE 135

128 Zlokovic B (2003) RAGE mediates am yloid-beta peptide transport across the blood-brain barrier and accumulation in brain. Nat Med 9: 907-913. Deane R, Sagare A, Hamm K, Parisi M, La Rue B, Guo H, Wu Z, Holtzman DM, Zlokovic BV (2005) IgG-assisted agedependent clearance of Alzheimer's amyloid beta peptide by the blood-brain ba rrier neonatal Fc receptor. J Neurosci 25: 11495-11503. Deane R, Wu Z, Zlokovic BV (2004a) RAGE (yin) versus LRP (yang) balance regulates alzheimer amyloid beta-peptide clearance through transport across the blood-brain barrier. Stroke 35: 2628-2631. Deane R, Zheng W, Zlokovic BV (2004b) Brain capillary endothelium and choroid plexus epithelium regulate transport of transferrin-bound and free iron into the rat brain. J Neurochem 88: 813-820. DeMattos RB, Bales KR, Cummins DJ, Dodart JC, Paul SM, Holtzman DM (2001) Peripheral anti-A beta antibody al ters CNS and plasma A beta clearance and decreases brain A beta burden in a mouse model of Alzheimer's disease. Proc Natl Acad Sci U S A 98: 8850-8855. DeMattos RB, Bales KR, Cummins DJ, P aul SM, Holtzman DM (2002) Brain to plasma amyloid-beta efflux: a measure of brain amyloid burden in a mouse model of Alzheimer's disease. Science 295: 2264-2267. Deshpande A, Mina E, Glabe C, Busciglio J (2006) Diffe rent conformations of amyloid beta induce neurotoxicity by di stinct mechanisms in human cortical neurons. J Neurosci 26: 6011-6018. Dickey CA, Gordon MN, Mason JE, Wilson NJ, Diamond DM, Guzowski JF, Morgan D (2004) Amyloid suppresses induc tion of genes critical for memory consolidation in APP + PS1 transgeni c mice. J Neurochem 88: 434-442. Dineley KT, Xia X, Bui D, Sweatt JD, Zheng H (2002) Accelerated plaque accumulation, associative learning defic its, and up-regulation of alpha 7 nicotinic receptor protein in transgenic mice co -expressing mutant human presenilin 1 and amyloid precursor proteins J Biol Chem 277: 22768-22780. Doan A, Thinakaran G, Borchelt DR, Slunt HH, Ratovitsky T, Podlisny M, Selkoe DJ, Seeger M, Gandy SE, Price DL, Si sodia SS (1996) Protein topology of presenilin 1. Neuron 17: 1023-1030. Duff K, Eckman C, Zehr C, Yu X, Pr ada CM, Perez-tur J, Hutton M, Buee L, Harigaya Y, Yager D, Mor gan D, Gordon MN, Holcomb L, Refolo L, Zenk B, Hardy J, Younkin S (1996) Increased am yloid-beta42(43) in brains of mice expressing mutant presenili n 1. Nature 383: 710-713.

PAGE 136

129 Duff K (1998) Transgenic models for Alzhei mer's disease. Neuropathol Appl Neurobiol 24: 101-103. Eckman EA, Reed DK, Eckman CB (2001) Degradation of the Alzheimer's amyloid beta peptide by endothelin-conver ting enzyme. J Biol Chem 276: 2454024548. Eckman EA, Watson M, Marlow L, Sambamurti K, Eckman CB (2003) Alzheimer's disease beta-amyloid pepti de is increased in mice deficient in endothelin-converting enzyme. J Biol Chem 278: 2081-2084. Ferrer I, Boada RM, Sanchez Guerra ML, Rey MJ, Costa-Jussa F (2004) Neuropathology and pathogenesis of enc ephalitis followin g amyloid-beta immunization in Alzheimer's disease. Brain Pathol 14: 11-20. Frenkel D, Maron R, Burt DS, Wei ner HL (2005) Nasal vaccination with a proteosome-based adjuvant and glatiramer acetate cl ears beta-amyloid in a mouse model of Alzheimer disease. J Clin Invest 115: 2423-2433. Galon J, Robertson MW, Galinha A, Ma zieres N, Spagnoli R, Fridman WH, Sautes C (1997) Affinity of the intera ction between Fc gamma receptor type III (Fc gammaRIII) and monomeric human IgG s ubclasses. Role of Fc gammaRIII glycosylation. Eur J Immunol 27: 1928-1932. Games D, Adams D, Alessandrini R, Bar bour R, Berthelette P, Blackwell C, Carr T, Clemens J, Donaldson T, Gillespie F, (1995) Alzheime r-type neuropathology in transgenic mice overexpressing V717F beta-amyloid precursor protein. Nature 373: 523-527. Gessner JE, Heiken H, Tamm A, Schmidt RE (1998) The IgG Fc receptor family. Ann Hematol 76: 231-248. Ghetie V, Ward ES (1997) FcRn: the MHC cl ass I-related receptor that is more than an IgG transporter. Immunol Today 18: 592-598. Ghiso J, Shayo M, Calero M, Ng D, Tomidokoro Y, Gandy S, Rostagno A, Frangione B (2004) Systemic catabolism of Alzheimer's Abeta40 and Abeta42. J Biol Chem 279: 45897-45908. Gilman S, Koller M, Black RS, Jenkins L, Griffith SG, Fox NC, Eisner L, Kirby L, Rovira MB, Forette F, Orgogozo JM (2005) Clinical effects of Abeta immunization (AN1792) in patients with AD in an inte rrupted trial. Neurology 64: 1553-1562. Glenner GG, Wong CW (1984) Alzheimer's disease: initial report of the purification and characterization of a nov el cerebrovascular amyloid protein. Biochem Biophys Res Commun 120: 885-890.

PAGE 137

130 Goate A, Chartier-Harlin MC, Mullan M, Brow n J, Crawford F, Fidani L, Giuffra L, Haynes A, Irving N, James L, (1991) Segre gation of a missense mutation in the amyloid precursor protein gene with fam ilial Alzheimer's disease. Nature 349: 704-706. Gong Y, Chang L, Viola KL, Lacor PN, Lam bert MP, Finch CE, Krafft GA, Klein WL (2003) Alzheimer's disease-affected br ain: presence of oligomeric A beta ligands (ADDLs) suggests a molecular bas is for reversible memory loss. Proc Natl Acad Sci U S A 100: 10417-10422. 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 cognitive deficits and Abeta deposits in transgenic APP+PS1 mice. Neurobiol Agin g 22: 377-385. Gordon MN, Holcomb LA, Jantzen PT, DiCar lo G, Wilcock D, Boyett KW, Connor K, Melachrino J, O'Callaghan JP, Mo rgan D (2002) Time course of the development of Alzheimer-like pathol ogy in the doubly transgenic PS1+APP mouse. Exp Neurol 173: 183-195. Gouras GK, Almeida CG, Takahas hi RH (2005) Intraneuronal Abeta accumulation and origin of plaques in Al zheimer's disease. Neurobiol Aging. Greenberg SM, Bacskai BJ, Hyman BT (2003) Alzheimer disease's doubleedged vaccine. Nat Med 9: 389-390. Hardy J (2002) Testing times for the "a myloid cascade hypothesis". Neurobiol Aging 23: 1073-1074. Hardy J, Selkoe DJ (2002) The amyloid hypothesis of Alzheimer's disease: progress and problems on the road to t herapeutics. Science 297: 353-356. Harper JD, Lieber CM, Lansbury PT, Jr. (19 97) Atomic force microscopic imaging of seeded fibril formation and fibril branching by the Alzheimer's disease amyloidbeta protein. Chem Biol 4: 951-959. Hartman RE, Izumi Y, Bales KR, Paul SM, Wozniak DF, Holtzman DM (2005) Treatment with an amyloidbeta antibody ameliorates plaque load, learning deficits, and hippocampal long-term potentiati on in a mouse model of Alzheimer's disease. J Neurosci 25: 6213-6220. Hebert LE, Scherr PA, Bienias JL, B ennett DA, Evans DA (2003) Alzheimer disease in the US population: prevalence estimates using the 2000 census. Arch Neurol 60: 1119-1122. Herber DL, Roth LM, Wilson D, Wils on N, Mason JE, Morgan D, Gordon MN (2004) Time-dependent reduction in Abet a levels after intracranial LPS administration in APP transgenic mice. Exp Neurol 190: 245-253.

PAGE 138

131 Hobbs SM, Jackson LE, Hoadley J (1992) Interaction of aglycosyl immunoglobulins with the IgG Fc transpor t receptor from neonatal rat gut: comparison of deglycosylation by tunicamycin treatment and genetic engineering. Mol Im munol 29: 949-956. Hock C, Konietzko U, Streffer JR, Tra cy 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, Papasso tiropoulos A, Nitsch RM (2003) Antibodies against beta-amyloid slow cogn itive decline in Alzheimer's disease. Neuron 38: 547-554. 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, Du ff K (1998) Accelerated Alzheimer-type phenotype in transgenic mice carrying both mutant amyloid precursor protein and presenilin 1 transgenes. Nat Med 4: 97-100. Holcomb LA, Gordon MN, Jantzen P, Hsiao K, Duff K, Morgan D (1999) Behavioral changes in transgenic mice expressing both amyloid precursor protein and presenilin-1 mutations: lack of association wit h amyloid deposits. Behav Genet 29: 177-185. Hsiao K, Chapman P, Nilsen S, Eckman C, Harigaya Y, Younkin S, Yang F, Cole G (1996) Correlative memory deficits, Abeta elevation, and amyloid plaques in transgenic mice. Science 274: 99-102. Ishiwata A, Kitamura S, Nagazumi A, Terashi A (1998) [Cerebral blood flow of patients with age-associated memory impairment and the early stage of Alzheimer's disease. A study by SPE CT using the ARG method]. Nippon Ika Daigaku Zasshi 65: 140-147. Iwata N, Takaki Y, Fukami S, Tsubuk i S, Saido TC (2002) Region-specific reduction of A beta-degradi ng endopeptidase, neprilysin, in mouse hippocampus upon aging. J Neurosci Res 70: 493-500. Iwatsubo T, Odaka A, Suzuki N, Mi zusawa 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 initia lly deposited species is A beta 42(43). Neuron 13: 45-53. Jankowsky JL, Slunt HH, Gonzales V, Savonenko AV, Wen JC, Jenkins NA, Copeland NG, Younkin LH, Lester HA Younkin SG, Borchelt DR (2005) Persistent amyloidosis following suppressi on of Abeta production in a transgenic model of Alzheimer disease. PLoS Med 2: e355. Janus C, Pearson J, McLaurin J, Mathew s 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

PAGE 139

132 reduces behavioural impairm ent and plaques in a model of Alzheimer's disease. Nature 408: 979-982. Janus C, D'Amelio S, Amit ay O, Chishti MA, Strome R, Fraser P, Carlson GA, Roder JC, George-Hyslop P, Westaway D (2000) Spatial learning in transgenic mice expressing human presenilin 1 ( PS1) transgenes. Neurobiol Aging 21: 541549. Jarrett JT, Berger EP, Lansbury PT, Jr. (1 993) The carboxy te rminus of the beta amyloid protein is critical for the seeding of amyloid formation: implications for the pathogenesis of Alzheimer's diseas e. Biochemistry 32: 4693-4697. Johnson-Wood K, Lee M, Motter R, Hu K, Gordon G, Barbour R, Khan K, Gordon M, Tan H, Games D, Lieberburg I, Sc henk D, Seubert P, McConlogue L (1997) Amyloid precursor protein processing and A beta42 deposition in a transgenic mouse model of Alzheimer disease. Pr oc Natl Acad Sci U S A 94: 1550-1555. Katzman R (1986) Alzheimer's dis ease. N Engl J Med 314: 964-973. Kawarabayashi T, Younkin LH, Saido TC, Shoji M, Ashe KH, Younkin SG (2001) Age-dependent changes in brain, CSF, and plasma amyloid (beta) protein in the Tg2576 transgenic mouse model of Alzheimer' s disease. J Neurosci 21: 372-381. Klyubin I, Walsh DM, Lemere CA, Cull en WK, Shankar GM, Betts V, Spooner ET, Jiang L, Anwyl R, Selkoe DJ, Ro wan MJ (2005) Amyl oid beta protein immunotherapy neutralizes Abeta oligomers that disrupt synaptic plasticity in vivo. Nat Med 11: 556-561. Kumar-Singh S, Theuns J, Van Broeck B, Pirici D, Vennekens K, Corsmit E, Cruts M, Dermaut B, W ang R, Van Broeckhoven C ( 2006) Mean age-of-onset of familial alzheimer disease caused by presenilin mutations correlates with both increased Abeta42 and decreased Abeta40. Hum Mutat 27: 686-695. Kuo YM, Emmerling MR, Woods AS, Cotte r RJ, Roher AE (1997) Isolation, chemical characterization, and quantitation of A beta 3-pyroglut amyl peptide from neuritic plaques and vascular amyloid deposits. Biochem Biophys Res Commun 237: 188-191. Lambert MP, Barlow AK, Chromy BA, Edward s C, Freed R, Liosatos M, Morgan TE, Rozovsky I, Trommer B, Viola KL, Wa ls P, Zhang C, Finch CE, Krafft GA, Klein WL (1998) Diffusible, nonfibrillar li gands derived from Abeta1-42 are potent central nervous system neurotoxins. Pr oc Natl Acad Sci U S A 95: 6448-6453. Lazarov O, Robinson J, Tang YP, Hairst on IS, Korade-Mirnics Z, Lee VM, Hersh LB, Sapolsky RM, Mirnics K, Sisodia SS (2005) Environmental enrichment reduces Abeta levels a nd amyloid deposition in tr ansgenic mice. Cell 120: 701713.

PAGE 140

133 Leissring MA, Farris W, Chang AY, Walsh DM Wu X, Sun X, Frosch MP, Selkoe DJ (2003) Enhanced proteolysis of bet a-amyloid in APP transgenic mice prevents plaque formation, secondary pat hology, and premature death. Neuron 40: 1087-1093. Lemere CA, Beierschmitt A, Iglesias M, Spooner ET, Bloom JK, Leverone JF, Zheng JB, Seabrook TJ, Louard D, Li D, Selkoe DJ, Palmour RM, Ervin FR (2004) Alzheimer's disease abeta vaccine reduces central nervous system abeta levels in a non-human prim ate, the Caribbean vervet. Am J Pathol 165: 283-297. Lesne S, Koh MT, Kotilinek L, Kayed R, Glabe CG, Yang A, Gallagher M, Ashe KH (2006) A specific amyloid-beta protein assembly in the brain impairs memory. Nature 440: 352-357. Levites Y, Jansen K, Smithson LA, Daki n R, Holloway VM, Das P, Golde TE (2006) Intracranial adeno-associated viru s-mediated delivery of anti-pan amyloid beta, amyloid beta40, and amyloid bet a42 single-chain variable fragments attenuates plaque pathology in amyloid pr ecursor protein mice. J Neurosci 26: 11923-11928. Levy E, Carman MD, Fernandez-Madrid IJ, Power MD, Lieberburg I, van Duinen SG, Bots GT, Luyendijk W, Frangione B ( 1990) Mutation of the Alzheimer's disease amyloid gene in hereditary cer ebral hemorrhage, Dutch type. Science 248: 1124-1126. Lobo ED, Hansen RJ, Balthasar JP (2004) Antibody pharmacokinetics and pharmacodynamics. J Pharm Sci 93: 2645-2668. Maggio JE, Stimson ER, Ghilardi JR, A llen CJ, Dahl CE, Whitcomb DC, Vigna SR, Vinters HV, Labenski ME, Mantyh PW (1 992) Reversible in vitro growth of Alzheimer disease beta-amyloid plaqu es by deposition of labeled amyloid peptide. Proc Natl Acad Sci U S A 89: 5462-5466. Mann DM, Yates PO, Marcyniuk B (1984) Al zheimer's presenile dementia, senile dementia of Alzheimer type and Down's syndrome in middle age form an age related continuum of pat hological changes. Neuropathol Appl Neurobiol 10: 185207. Masliah E, Hansen L, Adame A, Crews L, Bard F, Lee C, Seubert P, Games D, Kirby L, Schenk D (2005) Abeta vaccinatio n effects on plaque pathology in the absence of encephalitis in Alzheimer disease. Neurology 64: 129-131. McLaurin J, Kierstead ME, Brown ME, Hawkes CA, Lambermon MH, Phinney AL, Darabie AA, Cousins JE, French JE, Lan MF, Chen F, Wong SS, Mount HT, Fraser PE, Westaway D, George-Hyslop P (2006) Cyclohe xanehexol inhibitors of Abeta aggregation prevent and reverse Alzh eimer phenotype in a mouse model. Nat Med 12: 801-808.

PAGE 141

134 Moechars D, Lorent K, Van Leuven F (1999) Premature death in transgenic mice that overexpress a mutant amyloid precursor protein is preceded by severe neurodegeneration and apoptosis. Neuroscience 91: 819-830. Monsonego A, Imitola J, Petrovic S, Zota V, Nemirovsky A, Baron R, Fisher Y, Owens T, Weiner HL (2006) Abeta-i nduced meningoencephalitis is IFN-gammadependent and is associated with T celldependent clearance of Abeta in a mouse model of Alzheimer's disease. Pr oc Natl Acad Sci U S A 103: 5048-5053. Morgan D, Diamond DM, Gottsch all 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 vaccina tion prevents memory loss in an animal model of Alzheimer's disease. Nature 408: 982-985. Morgan D, Keller RK (2002) What evidence would prove the amyloid hypothesis? Towards rational drug treatments for Alzhei mer's disease. J Alzheimers Dis 4: 257-260. Morgan D (2003) Antibody therapy for Alz heimer's disease. Expert Rev Vaccines 2: 53-59. Naslund J, Haroutunian V, Mohs R, Davis KL, Davies P, Greengard P, Buxbaum JD (2000) Correlation between elevated le vels of amyloid beta-peptide in the brain and cognitive decline. JAMA 283: 1571-1577. Nicoll JA, Barton E, Boche D, Neal JW, Ferrer I, Thompson P, Vlachouli C, Wilkinson D, Bayer A, Games D, Seuber t P, Schenk D, Holmes C (2006) Abeta species removal after abeta42 immuniza tion. J Neuropathol Exp Neurol 65: 1040-1048. Nicoll JA, Wilkinson D, Holmes C, St eart P, Markham H, Weller RO (2003) Neuropathology of human Alzheimer diseas e after immunization with amyloidbeta peptide: a case r eport. Nat Med 9: 448-452. Nimmerjahn F, Bruhns P, Horiuchi K, Ra vetch JV (2005) FcgammaRIV: a novel FcR with distinct IgG subclass sp ecificity. Immunity 23: 41-51. Orgogozo JM, Gilman S, Dartigues JF, Laurent B, Puel M, Kirby LC, Jouanny P, Dubois B, Eisner L, Flitman S, Mic hel BF, Boada M, Frank A, Hock C (2003) Subacute meningoencephalitis in a subset of patients with AD after Abeta42 immunization. Neurology 61: 46-54. Paris D, Townsend K, Quadros A, Hum phrey J, Sun J, Brem S, WotoczekObadia M, DelleDonne A, Patel N, Ob regon DF, Crescentini R, Abdullah L, Coppola D, Rojiani AM, Crawford F, Sebti SM, Mullan M ( 2004) Inhibition of angiogenesis by Abeta peptides Angiogenesis 7: 75-85.

PAGE 142

135 Patton RL, Kalback WM, Esh CL, Kokjohn TA, Van Vickle GD, Luehrs DC, Kuo YM, Lopez J, Brune D, Ferrer I, Maslia h E, Newel AJ, Beach TG, Castano EM, Roher AE (2006) Amyloidbeta peptide remnants in AN-1792-immunized Alzheimer's disease patients: a bioc hemical analysis. Am J Pathol 169: 10481063. Pfeifer M, Boncristiano S, Bondolfi L, Stalder A, Delle r T, Staufenbiel M, Mathews PM, Jucker M (2002) Cerebral he morrhage after passive anti-Abeta immunotherapy. Science 298: 1379. Pike CJ, Burdick D, Walencewi cz AJ, Glabe CG, Cotman CW (1993) Neurodegeneration induced by beta-amyloid peptides in vi tro: the role of peptide assembly state. J Neurosci 13: 1676-1687. Prior R, D'Urso D, Frank R, Prikulis I, Cleven S, Ihl R, Pavlakovic G (1996) Selective binding of sol uble Abeta1-40 and Abeta1-42 to a subset of senile plaques. Am J Pa thol 148: 1749-1756. Racke MM, Boone LI, Hepburn DL, Parsadai nian M, Bryan MT, Ness DK, Piroozi KS, Jordan WH, Brown DD, Hoffman WP, Holtzman DM, Bales KR, Gitter BD, May PC, Paul SM, DeMattos RB (2005) Exacerbation of cerebral amyloid angiopathy-associated microhemorrhage in amyloid precursor protein transgenic mice by immunotherapy is dependent on ant ibody recognition of deposited forms of amyloid beta. J Neurosci 25: 629-636. Ravetch JV (1997) Fc receptors. Curr Opin Immunol 9: 121-125. Ravetch JV, Clynes RA (1998) Divergent ro les for Fc receptors and complement in vivo. Annu Rev Immunol 16: 421-432. Rensink AA, de Waal RM, Kremer B, Verbeek MM (2003) Pathogenesis of cerebral amyloid angiopathy. Brai n Res Brain Res Rev 43: 207-223. Roher AE, Lowenson JD, Clarke S, Woods AS, Cotter RJ, Gowing E, Ball MJ (1993) beta-Amyloid-(1-42) is a major component of cerebrovascular amyloid deposits: implications for the pathology of Alzheimer disease. Proc Natl Acad Sci U S A 90: 10836-10840. 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, Lieberbu rg 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 Alzheimer-disease-like patholog y in the PDAPP mouse. Nature 400: 173-177. Schenk D (2002) Amyloid-bet a immunotherapy for Alzheimer's disease: the end of the beginning. Nat Re v Neurosci 3: 824-828.

PAGE 143

136 Schenk DB, Yednock T (2002) The role of microglia in Alzheimer's disease: friend or foe? Neurobi ol Aging 23: 677-679. Schlachetzki F, Zhu C, Pardridge WM (2002) Expression of the neonatal Fc receptor (FcRn) at the blood-brai n barrier. J Neurochem 81: 203-206. Selkoe DJ (1998) The cell biology of beta-amyloid precursor protein and presenilin in Alzheimer's disease. Trends Cell Biol 8: 447-453. Selkoe DJ (2001) Alzheimer's disease: genes, proteins, and therapy. Physiol Rev 81: 741-766. Siemers ER, Quinn JF, Kaye J, Farlow MR Porsteinsson A, Tariot P, Zoulnouni P, Galvin JE, Holtzman DM, Knopman DS, Sa tterwhite J, Gonzales C, Dean RA, May PC (2006) Effects of a gamma-secreta se inhibitor in a randomized study of patients with Alzheimer dis ease. Neurology 66: 602-604. Solomon B, Koppel R, Frankel D, Hanan-Aharon E (1997) Disaggregation of Alzheimer beta-amyloid by site-direct ed mAb. Proc Natl Acad Sci U S A 94: 4109-4112. Solomon PR, Knapp MJ, Gracon SI, Groccia M, Pendlebury WW (1996) Longterm tacrine treatment in patients with Al zheimer's disease. Lancet 348: 275-276. Stahl T, Reimers C, Johne R, Schli ebs R, Seeger J (2006) Viral-induced inflammation is accompanied by beta-am yloid plaque reduction in brains of amyloid precursor protein transgenic Tg2576 mice. Eur J Neurosci 24: 19231934. Suzuki N, Cheung TT, Cai XD, Odaka A, Otvos L, Jr., Eckman C, Golde TE, Younkin SG (1994) An increased per centage of long amyloid beta protein secreted by familial amyloid beta prot ein precursor (beta APP717) mutants. Science 264: 1336-1340. Thomas T, Thomas G, McLendon C, Su tton T, Mullan M (1996) beta-Amyloidmediated vasoactivity and vascular endot helial damage. Nature 380: 168-171. Vieira P, Rajewsky K (1986) The bulk of endogenously produced IgG2a is eliminated from the serum of adult C57BL/ 6 mice with a half-life of 6-8 days. Eur J Immunol 16: 871-874. Walsh DM, Lomakin A, Benedek GB, C ondron MM, Teplow DB (1997) Amyloid beta-protein fibrillogenesis. Detection of a protofibrillar intermediate. J Biol Chem 272: 22364-22372. Weiner HL, Selkoe DJ (2002) Inflamma tion and therapeutic vaccination in CNS diseases. Nature 420: 879-884

PAGE 144

137 Wilcock DM, Alamed J, Gottschall PE, Gri mm J, Rosenthal A, Pons J, Ronan V, Symmonds K, Gordon MN, Morgan D ( 2006) Deglycosylated anti-amyloid-beta antibodies eliminate cognitive deficit s and reduce parenchymal amyloid with minimal vascular consequences in aged am yloid precursor protein transgenic mice. J Neurosci 26: 5340-5346. Wilcock DM, DiCarlo G, Henderson D, Jackson J, Clarke K, Ugen KE, Gordon MN, Morgan D (2003) Intracranially adm inistered anti-Abeta antibodies reduce beta-amyloid deposition by mechanisms both independent of and associated with microglial activation. J Neurosci 23: 3745-3751. Wilcock DM, Munireddy SK, Rosenthal A, Ugen KE, Gordon MN, Morgan D (2004a) Microglial activation facili tates Abeta plaque removal following intracranial anti-Abeta antibody admi nistration. Neurobiol Dis 15: 11-20. Wilcock DM, Rojiani A, Rosenthal A, Subbarao S, Freeman MJ, Gordon MN, Morgan D (2004b) Passive immunotherapy against Abeta in aged APPtransgenic mice reverses cognitive def icits and depletes parenchymal amyloid deposits in spite of increased vascul ar amyloid and microhemorrhage. J Neuroinflammation 1: 24. Wong PC, Zheng H, Chen H, Becher MW Sirinathsinghji DJ, Trumbauer ME, Chen HY, Price DL, Van der Ploeg LH, Siso dia SS (1997) Presenilin 1 is required for Notch1 and DII1 expression in the paraxial mesoderm. Nature 387: 288-292. Zhang Y, Pardridge WM (2001) Mediated effl ux of IgG molecules from brain to blood across the blood-brain barri er. J Neuroimmunol 114: 168-172.

PAGE 145

APPENDICES

PAGE 146

139 APPENDIX A: BINDING AFFINITIES OF IgG ISOTYPES TO FC-GAMMA RECEPTORS Biological and pathological activities differ with various IgG isotypes. These differences have conventionally been a ttributed to disparities in the ability of certain isotypes to engage co mplement or one of the known Fc R. The finding that individual Fc Rs interact differently with IgG isotypes in mediated protective inflammatory responses is cert ainly relevant for the potential use of these receptors as therapeutic targets in the treatment of disease. Murine effector cells such as microglia within the CNS express four different classes of IgG-specific Fc receptors: a high affinity receptor, Fc RI, two low affinity receptors, Fc RII and Fc RIII, and an intermediate affinity receptor, Fc RIV (Ravetch, 1997) (Nimmerjahn et al., 2005). Fc receptors I, III, and IV are all activating receptors characterized by an immunoreceptor tyrosine-based activation motif (ITAM). These rec eptors are important for triggering phagocytosis by activated macrophages. Fc RII is an inhibitory receptor characterized by the presence of an IT IM motif that recr uits inhibitory phosphatases that limit effect ive signaling. Table 1 is a summary of previously published affinities of different is otypes of IgG for the various Fc receptors.

PAGE 147

140Table 1 Affinities (Kd) of antibody isotype for Murine Fc receptors Isotype mFcRI (M) mFcRIIb (M) mFcRIII (M) mFc RIV (M) IgG1 X 3.01 x 10-7 3.2 x 10-6 X IgG2a 1 x 10-9 2.39 x 10-6 1.46 x 10-6 3.45 x 10-8 IgG2b X 4.48 x 10-7 1.55 x 10-6 5.9 x 10-8 X indicates no detectable binding Binding affinity of 2H6 and D-2H6 antibodies to Fc receptors or complement protein C1q we re measured using BIAcore (table 2). Purified murine Fc receptors (from R&D Systems) and human C1q (from Quidel) were immobilized on BIAcore CM5 ch ip by amine chemistry: Fc receptors or C1q were diluted into 10 mM sodium acet ate pH 4.0 and injected over an EDC/NHS activated chip at a concentration of 0. 005 mg/mL. Variable flow time across the individual chip channels were used to obtain 2000–3000 response units (RU). The chip was blocked with ethanolamine. Serial dilutions of monoclonal antibodies (ranging from 2 nM to 70 m) were injected. HBS-EP (0.01 M HEPES, pH 7.4, 0.15 M NaCl, 3 mM EDTA, 0.005% Surfactant P20) was used as running and sample buffer. Regeneration studies show ed that a mixture of Pierce elution buffer (Product No. 21004, Pierce Biot echnology, Rockford, IL) and 4 M NaCl (2:1) effectively removed the bound antibody peptide while keeping the activity of Fc receptors and C1q. Binding affinities of A for the antibodies was determined similarly by immobilizing the antibodies on a CM5 chip using amine chemistry, and flowing AB1-40 over the chip at mult iple concentrations. Binding data were

PAGE 148

141 analyzed using 1:1 Langmuir interaction model for high affinity interactions, or steady state affinity model for low affi nity interactions (Carty et al., 2006). Table 2 Affinities (Kd) of 2H6, De-2H6, and 2286 for effector proteins Antibody Isotype mFcRI (nM) mFcRIIb (nM) mFcRIII (M) hC1q (M) 2H6 IgG2b 1,600 20,000 39,000 5,000 De-2H6 IgG2b 6,500 30,000 67,000 30,000 2286 IgG1 87 2,000 1,000 >100,000

PAGE 149

142 APPENDIX A REFERENCES Carty NC, Wilcock DM, Rosenthal A, Grim m J, Pons J, Ronan V, Gottschall PE, Gordon MN, Morgan D (2006) Intracranial administration of deglycosylated Cterminal-specific anti-Abeta antibody effi ciently clears amyloid plaques without activating microglia in amyloid-deposit ing transgenic mice. J Neuroinflammation 3: 11. Nimmerjahn F, Bruhns P, Horiuchi K, Ra vetch JV (2005) FcgammaRIV: a novel FcR with distinct IgG subclass sp ecificity. Immunity 23: 41-51. Ravetch JV (1997) Fc receptors. Curr Opin Immunol 9: 121-125.

PAGE 150

ABOUT THE AUTHOR Rachel Anne Karlnoski received her Ba chelor’s degree in Microbiology and Cell Science from the University of Florida in 2001 and her Master’s degree in Medical Sciences from the University of South Florida in 2006. Before entering the Ph.D program at USF, Rachel worked at the H. Lee Moffitt Cancer Center as a research associate. Enthralled by the exciting scientific observations made through molecular and histological work, her interest in scientific research intensified. In January 2004, Rachel entered the Alzheimer’s disease Research Laboratory under the mentorship of Dave Morgan Ph.D. and Marcia Gordon Ph.D. Rachel’s research focused on elucidating mechanisms responsible for amyloid removal after immunotherapy as a treatment for Alzheimer’s disease. She successfully defended her doctoral dissertat ion in July 2007 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 001921028
003 fts
005 20080116131251.0
006 m||||e|||d||||||||
007 cr mnu|||uuuuu
008 080116s2007 flu sbm 000 0 eng d
datafield ind1 8 ind2 024
subfield code a E14-SFE0002145
035
(OCoLC)190845809
040
FHM
c FHM
049
FHMM
090
RM101 (ONLINE)
1 100
Karlnoski, Rachel Anne.
0 245
Optimization of anti-Abeta antibody therapy
h [electronic resource] /
by Rachel Anne Karlnoski.
260
[Tampa, Fla.] :
b University of South Florida,
2007.
520
ABSTRACT: Alzheimer's disease (AD) is the most common form of dementia, a disease that gradually destroys brain cells and leads to progressive decline in mental function. The presence of high densities of neuritic plaques composed of Abeta in the cerebral cortices is a criterion for the post-mortem diagnosis of AD. The view that Abeta deposition drives the pathogenesis of AD (amyloid hypothesis) has received support from a wide range of molecular, genetic, and animal studies. This hypothesis has been the focus of therapeutic intervention leading to the development of anti-Abeta immunotherapy as a potential treatment. There is a great deal of evidence that supports the capacity of immunization against Abeta to reduce amyloid pathology and restore memory function in transgenic mouse models of amyloidogenesis.However, as a result of anti-Abeta immunotherapy, many investigators have reported increased severity of cerebral amyloid angiopathy (CAA) and increased incidences of microhemorrhage. The mechanism/s responsible for the redistribution of Abeta to the vasculature is unclear. We examine two possible mechanisms that may influence the severity of CAA following immunization; the rate of Abeta clearance with deglycosylated antibodies via a dose response study and anti-Abeta antibody epitope specificity. Dose response results with a deglycosylated antibody showed that lower doses resulted in greater clearance of amyloid and significant improvements in cognition, suggesting that clearance mechanisms become saturated with high doses of antibody.Treatment with antibodies directed against different epitopes of Abeta implied that the degree of parenchymal Abeta clearance determines the extent of vascular Abeta accumulation; epitope specificity is not critical in directing the vascular accumulation. Passive anti-Abeta immunization can prevent Abeta deposition in APP transgenic mice. We investigated amyloid accumulation after immunization was terminated, and discovered that after treatment, amyloid began to accumulate as a factor of time and gradually built up but never reached the Abeta levels in control APP mice. These data suggest that delayed deposition of amyloid leads to long term delays in AD associated pathology. These data strongly support the use of prophylactic immunotherapy treatments, and it appears that existing amyloid deposits will require interventions that actively clear amyloid as the only means to efficiently reduce brain Abeta in AD.
502
Dissertation (Ph.D.)--University of South Florida, 2007.
504
Includes bibliographical references.
516
Text (Electronic dissertation) 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 142 pages.
Includes vita.
590
Advisor: David Morgan, Ph.D.
653
Alzheimer's disease.
Amyloid.
Angiopathy.
Microglia.
Transgenic mice.
690
Dissertations, Academic
z USF
x Molecular Pharmacology and Physiology
Doctoral.
773
t USF Electronic Theses and Dissertations.
4 856
u http://digital.lib.usf.edu/?e14.2145