Alzheimer Amyloid-P Peptide Aggregation: Altemate Products of Fibril Formation

Tze-Hsien Jackson Huang

A thesis submitted in codormity with the requirements for the degree of Doctor of Philosophy

Graduate Department of Medicd Biophysics University of Toronto

O TH. Jackson Huang 200 1

uisitions and Acquisitions et Bib ~ographic Services senrices bibliographiques "9-

The author has graated a non- L'auteur a accordé une licence non excIPsive licence allowing the exchisive permettant à la National Library of Canada to BibIiotheque nationale du Canada de reprodace, Ioan, distri'bute or sell reproduire, prêter, distribuer ou copies of this thesis in microform, vendre des copies de cette thése sous paper or electronic fomats. la forme de microfiche/nlm, de

reproduction sur papier ou sur format électroniqpe .

The author retains ownership of the L'auteur conserve la propriété du copyright in this thesis. Neither the droit d'auteur qui protège cette thèse. thesis nor substantial extracts fkom it Ni la thèse ni des extraits substantiels may be printed or otherwise de ceae-ci ne doivent être imprimés reproduced without the author's ou autrement reproduits sans son permission. autorisation.

Alzheimer Amyloid-B Peptide Aggregation: Alternate Products of Fibril Formation

Tze-Hsien Jackson Huang Doctor of Philosophy 200 1

Graduate Department of Medical Biophysics University of Toronto

A m c f

In Alzheimer disease (AD), polymerization of the amyloid B peptide (AB) to form

fibrillar deposits in the brain is associated with neurodegeneration. Because of this,

researchers have worked to understand fibrillogenesis and the toxicity of fibrils.

However, recent discoveries suggest intermediate structures in AB fbril formation are

also toxic. We sought to gain a better understanding of intermediate structures in the

fibrillogenesis pathway, describe conditions where AB foms biologically inert

amorphous arnyioid rather than neurotoxic fibrillar amyloid, and elucidate some of the

extrinsic factors controlling conversion of Ab monomers to fibrils.

At Iow micromolar concentrations, Ap0, the 40 residue fom of AB, assembled

iato two types of soluble oligomea depending upon pH. Dimerftetramers with irreguiar

secondary structure formed at neutral pH while sphericd particles with a mass of 0.94

megadaitons and fbsheet secondary structure f o d at pH 3. Both structures were stable

for at least 4 weeks; this stability will ailow for m e r investigation using hi&-

resolution techniques.

In the Iaboratory, A$ often precipitates to fom non-specific aggregates. We

sought to characterize these aggregates for use as an in vitro mode1 for amorphous

amyloid. These aggregates were fhructured at peptide concentrations >10 pM and

unfolded at lower concentrations. The structured aggregates were tightiy packed

containing peptides inaccessible to water. Peptides in the unstnictured aggregates were

loosely packed, mobile and accessible to water. Structured aggregates appeared

protofibrillar and developed iato mature fibrils after several weeks whereas the

unstnrchwd aggregates were invisible by EM and did not generate fibriis. These findings

suggest the mstructured aggregates share mauy properties with the amorphous amyloid

of AD and may aid in studying morphous amyloid in vitro.

We used the organic osmolytes trimethylamine IV-oxide ( M O ) and glycerol as

mirnics of natuaIly occurring chaperone molecules to investigate their effects upon

fibrillogenesis. TMAO and glycerol accelerated the Ab randorn coi1 to B-sheet

conformational change. 'Ihis tninsition occurred with the immediate conversion of

amorphous uastructured aggregates to unifonn globular structures. TMAO and glycerol

aiso mediated the transformation of protofibrils to mature fibrils. Thus, the effects of

extrinsic factors such as chaperone molecules must be considered when studying AB

fibrillogenesis.

iii

Ackno wledgments

When a thesis is finally complete, the world knows a Little more about the subject

of investigation. Unfortunately, though the knowledge is then recorded the people who

helped make the thesis a reality remain unrecognized. This is my attempt at remedying

that situation.

My supervisor was Avi Chakrabartty. When 1 started with Avi in September

1994, the !ab was empty Save for him, a computer and some AB(9-25) peptide stored in a

mini-fkidge that actually belonged to another lab. Since then, he bas built a lab with state-

of-the-art instrumentation and a talented staff of post-docs, students and technicians. It is

a testament to his abilities as a scientist and a leader. 1 had the good fortune to be a part

of. this dynamic environment of gmwth and evolution. Through the experience t have

learned much and 1 am better for it. Th& you Avi.

Paul Fraser and Mike Rauth were members of my supervisory cornmittee. Of

course, Paul's knowledge of AD was invaluabIe, but what 1'11 remember moa is that his

sense of humour and no-nonsense attitude helped keep things in perspective. Mike was

generous, encouraging, unrelentingly thorough and made me work my ass off to get this

thesis ready .

My t h e at the Department wouid have been far less rich if it were not for my

feiiow lab members: Sandy Go, Paul Gorrnan, Cynthia Quan, Chandra Boon and Xiao-

Fei Qi. Because of you dl, the lab was more than a work place; it was a place of laughter

and h,

1 am grateful to my coIIaborators Dun-Sheng Yang, Nick Plaskos and Chris Yip.

Their expertise made this thesis much more complete.

JoAtme McLaurin joined the lab as a post-doc shoaly after 1 began. She has h c e

left to lead her own research in AD. We saw much together over three years that she was

here and 1 feel priviteged to have worked with her. 1 wiil always admire J o h e for her

integnty and dedication.

Jake Tyson and Stan Lidon are my fiïends. Through t h e s of doubt they were

there for me and they convinced me that the work was worthwhile. 1 would not have

completed this endeavour if it were not for them.

My parents, Suzan and Louis, deserve my moa heartfelt thanks. They taught me

the value of Ieaming and the power of knowledge. It has been their patience,

understanding and support in altowing me the t h e and fieedorn to pume my interests

that has contributed the most to making this work possible. For dl that you have done

and for your love, thank you.

CONTENTS

. CaAPTER 1: ALWIEIMER DISEASE ...-.. ..~~o..o..o.o..o..o...~...... I

Ag .................................................................................................................................................................... 5 A@ and ABfibriifirmation ................................................................... .. ........................................... 8

AB AND NEUROTOX~Y: THE Ag HYFOTHESIS OF AD ................................................................................. 12 EARLY-ONSET AD (EOAD) ....................................................................................................................... 13

APP mutations .............................................................................................................................................. 13 Mutations in PSI and PS2 ................... ................. .................................................................................. 15 Dmvn Synàiome ........................................................................................................................................... 16

LATE-ONSET AD (LOAD): GENETIC POLYMORPHISMS M APOLIPOPR~EIN E4 AND ~ - 2 - M ~ ~ R f f i t O B v t M ..................................................................................................................................... 17

ApoE ............................................................................................................................................................. 17 A2M .............................................................................................................................................................. 18

.................................................. .................................. MOLECULAR MECHAMSMS OF AB NEUROTOXIC~ .. 18 .................................................................................... CONTROVERSES AND ISSUES FACMG RESEARCH 19

........................................................................................................ What iS the newotoxic species in AD? 19 THESIS O m M E ................... ,rr ....... r ............................................................................................................ 2 1

Proprties ofsuIztbIe A@ .............................................................................................................................. 21 Properties of d m e amyioid ................. .. ................................................................................................... II

...................................... Extrinsic factors ~vhich injluence aggregational properties of AB 22 7 7 ......................... Chapter 2: Smcîural sîudies of soluble oligomers of the Akheimer j3arnyfoid peptide ,,

Chapter 3: Alternate aggregation pthways of the Akheimer &myloid peptide: an in vitro model of .......................................................................................................................... . d@ue arnyioid .........,,,.. .... 23

Chapter 4: lbfanipulating the amyloid-8 aggregation pathway with chemical chaperones ..................... 23 UEERmcm ....... .............................. ............................................................................................................ 25

CHAPTER 2: STRUCTüEùU !VüDIES OF SOLUBLE OLIGOMERS OF THE ALZHEZMER PAMYLOID PEPTIDE.....-.. ~ M . w ~ o o . m ~ H . . H ~ - e e n o m m i w t..mao.oo.mmr .-mm 45

ABSTRACT ....................................................... ,.,. ............................................................................................. 46 ........................................................................................ ............ ~ O O U ~ O N ~ ........................................... 47 ........................................................................................ .............................. MA- AND ~ O D S ....... 49

.......................................................................................................................................... Pepci& synthesh 49 Fluorescent labefing ................. ,., ................................................................................................................ 50 Prepation of stock peptide solutions that are f i e offibriI seecis ........................................................... 50 Meclmrement of peptik concentration ...................................................................................................... 5 1 Electron microscopy .................................................................................................................................... 52

.......................................................................................................................... Fluorescence spectroscopy 52 ................................................................................................................ Circular dichrokm spectroscopy 53 ................................................................................................................. Analyical ultracentrifugation ,,.. 54

............................................................................................................................. Atomic force micrmcopy 54 n ...... . ....... . ......................... "* .......................................................................................... 55

A scenanihg test for i&naing sotubze A&IO ofigomers ........................................................................... 55 Associrilian reactiom ofAm manitomci by emkonment-se~tsirive~uorescent proks and fluorescence monance energy rnmsfer ( F m ........................................................................................ 61 Molecular weighrs OfsufubIe AfiO oligomers .........................................,.................. . 66

List of Figures

AB nbrils APP processing and the amino acid sequence of AB42

Absorbance and fluorescence of AB40 aggregates at varying pH Solubility of AfM0 at pH 3,s and 7 Concentration dependence of AB40 association reactions Fluorescence spectra of Trp-Ag40 and AEDANS-AB40 at pH 5 Ultracentrifugation anaiysis of AB40 fibril intermediates

Sedhentation coefficient distniution from sedimentation velocity experiments at pH 3 Electron micrographs of platinum-carbon shadowed or negatively stained AB40 preparations (0.05 mg1m.L) Atomic force microscopy of APO at pH 3 Volume distribution of AM0 particles

2.10 CD spectroscopy of 20 pM AB40 at pH 3,s and 7 77

The NBD absorbante of AM0 samples CD spectra of AB40 aggregates The NBD fluorescence of AB40 samples Fluorescence polarization of AB40 Platinun-carbon shadow electron microscopy

The eEects of glycerol on AWO Effects of TMAO Morphologicd changes in peptide aggregates induced by chernicd c haperones Rotary platindcarbon shadowing electron microscopy Insolubility of the urismictured and fibrillar aggregates monitored by a fluorescent Iabeled AB peptide tracer

List of Abbreviafions

AFM ApoE APP BACE CD DS EM EOAD FRET LOAD NBD NFTs NMR PS W C S TMAFM T'MA0

a-2-macroglo bulin Amyloid fJ Alzheimer disease Afbderived diffusible ligands 5-(((amino)ethy l)amino)nap hthalene-l- sulfonic acid Atomic force microscopy Apolipoprotein E A m y loid precursor protein Beta-site APP-cleaving enzyme Circular dichroism Down syndrome EIectron microscopy Earlysaset Alzheimer disease Fluorescent resonance energy transfer Late-onset Alzheimer disease 7-nitrobenz-2-oxa- l,3-diazole Neuro fibtillary tangles Nuclear magnetic resonance Presenilin Platinum/car bon Svedberg unit (1 o - ' ~ seconds) Tapping mode atomic force microscopy Trimethy lamine oxide

Chapter 1 : Alzheimer Disease

Introduction

Alzheimer disease (AD) is a progressive neurodegenerative disorder which results

in dementia and death. In Canada, 3 16 500 people over 65 have dementia and of those,

64% have AD. It is estMated that by 2031, over 750 000 Canadians will have AD and

related dementias (Canadian Study of Heaith and Aging Working Group, 1994)

The disorder is initially manifested in memory deficit, then progresses to affect

language skills, judgment, reasoning, personality and behaviour. AD gradually

compromises a persoa's resistance to infections and other illnesses which are often the

ultimate cause of death. The time course for the progression of AD is typically eight

years but can range from three to twenty-five years. AD can occur at any age, but the

majority of AD patients are over age 60.

AD can be divided into the latesnset form (>60 yean) and the less common early

onset fonn ( ~ 6 0 years). Early onset AD (EOAD) is genetically inherited and thus far

three genes have been linked to this form. A number of risk factors have been associated

with late-onset AD (LOAD) some genetic, some environmental; however, the cause of

the disease is stiU unknown and at present there is no cure. Present treatment is Limited to

miniminng the consequences of the symptoms. As people born during the population

boom of the late 1940s and early 1950s reach 60 years of age, the ernotional, social and

economic impact of AD wil1 grow and hding a prevention or cure wiU gain even greater

urgency .

It is beIieved that a peptide, origïnaily termed A4 but now commonly refened to

as amyioid-B (AB), pIays an important role in the neurodegeneration of the disorder. This

chapter will describe some properties of AB and explain why this peptide is suspected to

be the prllnary causal agent of AD. Our knowledge of AB is Uicomplete and significant

inconsistencies rernain bringing doubt to the AB hypothesis. These probiems wil1 also be

discussed and the h a l portion of the chapter w i U outiine how this thesis addresses these

issues and endeavours to m e r some of the questions.

Amyloid and Neurofibrfllary Tangles

Pathologically, AD is characterized by the accumulation of protein deposits in the

brain. Alois Alzheimer (1907) fiat reported the presence of abnormal fiben within

neuronal ceIl bodies called neurofibrillary tangles (NFT) and extracellular deposits of

amyloidogenic proteins in plaques and cerebral blood vessels.

The main component of NFTs is paired helical filaments of the

hyperphosphorylated isoforms of the microtubule-associated protein tau. Extracellular

amyloid is composed prirnarily of the peptide amyloid-B (AB). Work in this field has

focussed on understanding the roles of these two molecules in the progression of AD.

Despite many advances, these roles are still unclear and whether NFTs and AB play

seminal causal roles in the initiai stages of AD or are only end resuits of other pathogenic

events rernains highly controversial. However, a significant amount of evidence is

accumulating to implicate AB as the primary cause responsible for the pathogenesis of

AD.

Figure 1.1. AB fibrils. Negative stain EM of AB fibnls isolated from AD autopsy tissue. x -160 000. (Men et al., 1983)

AmyIoid Plaques

Amyloid coliects to f o m three general types of deposits in AD brains: senile

plaques, diffuse plaques and cerebrovascular amyloid. Senile plaques are lesions 10-200

prn in diameter with a dense core of fibdlar amyloid (Figure 1.1). These plaques are

surrounded by degenerating and swollen nerve terminais (Müller-Hill and Beyreuther,

1989).

Diffuse plaques, unlike seniie plaques, contain Iittle or no fibrillar Ab and are

instead composed of morphous aggregates of AB (Yamaguchi et al., 1989). Diffuse

plaques are not surrounded by dystrophic neurites and are not associated with

neurodegeneration. Cerebrovascular amyloid is found in the walls of midl cerebral blood

vessels and contains fibrillar AB.

In addition to AB, other proteins associated with plaques have been identified

including amyloid-P component (Coria et ai., 1988); basement membrane components,

the serine protease inhibitor, cr-1-antichymotrypsin (ACT) (Abraham et al., 1988);

apolipoprotein E (ApoE)(Wisniewski and Frangione, 1992); cornpiement factors, C lq,

C3d and C4d (McGeer et ai., 1989); and heparan suïlitte proteogIycans (HSPG) (Snow et

al., 1988).

AB was fint identified by Glemer and Wong (1984) when they sequenced the

principal protein component of vascular amyloid. This peptide was 24/28 residues in

Iength. Subsequent work has shown that the length of AB is variable with the greatest

heterogeneity at the C-terminus. AB usualiy begins with Aspl and is 39-43 residues long

with the 40 and 42 amino acid peptides being the most common (Figure 1.2).

There is sorne specificity in regard to the length of Af3 and its distribution in the

brain. AB was found to extend primarily to residues 42/43 in senile plaques (Masters et

al., 1985). In vitro, AB42 is more fibrillogenic than the other f o m (Hilbich et al., 1991;

Jarrett et al., 1993; Iwatsubo et al., 1994; Younkin, 1995). AB40 is also present in

significant quantity in senile plaques and apparently its deposition precedes that of Ap2.

AB28, AB39, AB40 and AB42 are present in vascular amyloid (Joachim et al., 1988;

PreK et al., 1988); however, the AB40:Ap42 ratio is higher in vascular amyloid than in

senile plaques. AB40 is also found in soluble form at low concentrations in the biological

fluids of normal individuals (Haass et al., 1992; Seubert et al., 1992; Shoji et al., 1992;

Busciglio et al., 1993), as are AB28 and AB42 (Vigo-Pelfiey et al., 1993). In AD brains,

the levels of soluble A$ are elevated (Teller et al., 1996).

membrane

i - .

Amylotd Precursor Protein (APP) . - 4 - --, - --. . .- , Nt Ct

1 10 20 30 40 DAEFRHDSGYEVHHQKLVFFAEOVGSNKGAI I G L M V G G V V 1 A 1 t I I t 8 t

p-sectetase a-secte tase psecretase

Figure 1.2. APP processing and the amino acid sequence of AB42. AB42 is proteolytically cleaved fiom the amyloid precursor protein (APP). APP is a normal transmembraue protein with an extracellular N-terminus and a cytosoüc C-terminus. The sites of the enzymatic activities producing AB are indicated (adapted fiom Fraser et al., 1993).

AB is a proteolytic product of the B-amyloid precursor protein (APP), a large

type4 transmembrane protein encoded on chromosome 21 which is constitutively

expressed in many celI types including neuronal, glial, endothelial, epithelial, kidney and

spleen (Kang et al., 1987; Robakis et al., 1987; Tanzi et al., 1987; Weidemann et al.,

1989; Shoji et al., 1992). AB is encoded in the putative traasmembrane and extracellular

domains of APP. Several isofonns of APP with varying length have been identified

(APP563, APP695, APP751 and APP770) (Kang et al., 1987; Kitaguchi et cil., 1988;

Ponte et al., 1988; Tanzi et al., 1988; de Sauvage and Octave, 1989) al1 generated by

alternative splicing of mRNA Born a single 19 exon gene located on the long ami of

chromosome 21 (St George-Hyslop et al., 1987).

The function of APP is not known. Some possible functions include roles as an

autocrine factor to stimulate ce11 proliferation (Saitoh et al., 1989), as a mediator of

neurite outgrowth Ui response to nerve growth factor (Milward et al., 1992) and as a

modulator of ce11 adhesion (Schubert et al., 1989).

AB and AB fibril formation

The ultrastructure of AB fibrils was revealed when senile plaque amyloid from

AD brains was examined ushg electron microscopy (Teny et al., 1964; Kidd, 1964). AB

Ebrils are 70-90 A Ui diameter and, when viewed in cross-section, appear to be composed

of 5-6 sub-fibdar structures called filaments (Nafang, 1980). Staining of senile plaques

with Congo red produces green birefhgence under polarized light (Puchtler, et al., 1962)

indicating that fibds are composed of polymers with 8-sheet structure (Glenner et ai.,

1972). X-ray difhction adysis of unonented plaque cores suggests that peptides within

the nbril are arraoged as cross-6 pIeated sheets with a distaoce between poiypeptide

c h a h of 4.76 A and a distance between sheets of 10.6 A (Eschner et al., 1986). H o w

AB assembles into this complex, highly organized structure is unknown. Jarrea and

Lansbury (1993) have proposed that kineticaiiy, AB Brillogenesis can be described as a

two step process: nucleation and growth. Nucleation is the slow thermodynamically

unfavourabte self-association of monorneric AB in a random coi1 conformation to form an

oligomeric nucleus with 8-sheet secondary structure. Once the nucleus has formed,

continued addition of AB monomers becomes thermodynamically

fibril grows.

Though the overall kinetics may be simple the mechanism

€avouable and the

leading to the fmal

ultrastructure is likely rather complex, involving a host of conformationally varied

transient intermediates, The effort to determine fibril ultrastructure and understand the

process of fibril formation has been hindered because A$ is highly hydrophobie and

prone to precipitation into an amorphous aggregate. These characteristics prevent

crystallization thus precluding study by x-ray cry stallography . The inso lubility of AB also

limits the usefilness of examination by nuclear magnetic resonance (NMR).

Unpredictable aggregation makes kinetics experirnents dificult.

A breakthrough came for researchers when it was learned that synthetic AB

hgments behave much üke full length naturally occurring forms. Fibrils of AB([-28)

possess morphologies comparable to in vivo fibrils; the fibrils have cross-6 structure and

green buefikgence with Congo red staining is present (Kirschner et al., 1987).

Furthemore, fragments as short as Ap(12-28) and AB(l4-28) still Iead to amyloid fibril

formation (Gorevic et al., 1987). Thus the difficulties encouatered eariy on in puriQing

AB or synthesizing MI-length forms cotdd be circumvented by using short mode1

peptides. In addition, this development made it possible to characterize individual

sequence regions and to evaluate their contributions to fibril formation.

[n an aggregate, AB appears to associate into a double stranded B-sheet with a

tum located at positions 26-29 of the sequence (Hiibich et al., 1991). The hydrophobie

carboxyl-temiinal sequence, Ap(34-42) derived entirely fiom the trammembrane dornain

of APP, possesses unusually stable fbstnicture and exclusively adopts oligomeric,

intermolecular psheet conformation in aqueous solution; consequently, this region may

be responsible for directing the fotciing of complete AB (Haiverson et al., 1990).

An early study suggested that the sequence region encoded by residues 11-24

contains the cntical intrinsic information specifjhg fibril formation (Kirscher et al.,

1987) and subsequent work using substituted peptides has shown that fibrillogenesis is

hydrophobically driven by this domain (Hilbich et al., 1992). However, at least one other

force is also at play because B-sheet formation with AB is also pHodependent.

Specifically, fibnllar structure, for a nurnber of mode1 peptides which included d l or part

of this domain, is maximal within a pH rauge of 3-8 (Fraser et ai., 1991). Also, Barrow

and Zagoaki (1 99 1) made similar observations hding that Ap(1-3 9) and AB(1-42) had

maximum rates of aggregation at pH 5.5. This tendency for &sheets to form at acidic pH

suggests the involvement of the imidazole groups of His 6, 13, and 14 as well as the

carboxylic acid side chahs of Asp and Glu (residues 1,3,7, 11, 22 and 23).

Work with short peptide rnodels had proven f i t f u l for gaining knowledge of

some properties of AB, but extension to the in vivo behavior of full length AB must be

made with caution. Thus work continues with AB40 and Ap2, guided by lessons learned

nom the models. The factors which accounted for the unpredictable aggregation

behavior, such as batch to batch variations in synthesizing peptide, inconsistent

solubilization procedures and fdure to establish uniform startiug conditions that are fiee

of nucleating non-specifc aggregates or seeds, have been addressed and significant

advances have been made in describing the process of fibrillogenesis.

The overall kinetics do appear to follow a nucleation-dependent mechanism. Two

defuiing characteristics of this model are: 1) the presence of a lag tirne during which the

nucleus forrns and 2) the ability of a preformed nucleus to seed a soluble supersaturated

solution and initiate the polymerization phase. Using Congo red binding as a measure of

fibrilization, Wood et al. (1996) have shown that A$ polymerization is sigrnoidal with an

initial lag period and, fiilrtiermore, they show that this lag penod can be elirninated with

addition of pre-formed seeds.

Lending further suppoa to the nucleation-growth model of Ab fibrillogenesis,

soluble oligomers and protofibrillar structures have been isolated and their properties

suggest they have a precursor-product relationship with mature fibrils. Walsh et al.

(1997) have observed by size exclusion chromatography that AB40 associates to form

dimea. Wiîh incubation, an additional species with a rnolecular weight of >IO0 000

wouid appear. Electron micmscopy showed this species consisted of curved fibrils 6-8

nm in diameter and eOO nm in length. They termed these structures protofibrils and in

subsequent work found tbat protofibrils are in equilibrium with low rnolecular weight AB

(monomers and dimers), have B-sheet secondary structure, and Iead to the formation of

mature fibriIs (WaIsh et ai., 1999). In another in vitro investigation, Lambert et al. (1998)

reported that inhibition of fibril formation by clusterin (dso cdled ApoJ) or by

incubation at reduced temperature (48°C) produces oligomerïc structures they termed

A B-derived diffusMe ligands ( ADDLs). Whether ADDLs are phy siologically relevant

precursoa to fibril formation or side products of the inhibition of fibrillogenesis has yet

to be determined.

Other evidence indicates oligomeric Ab plays a role in vivo. Roher et al. (1996)

have purified stable dimeric and trimeric AB components fiom senile and vascular

amyloid deposits of AD brains while, Kuo et al. have isolated water soluble AB, dimeric

to oligomeric, fiom AD brain and cerebrospinal fluid at leveis 6 times higher than control

samples (Ku0 et al., 1996).

Since the sequencing of AB, our understanding of its physicai charactenstics,

biochemical properties and aggregational behaviour has grown signifcantly. However,

work with AB has been dnven by an assumption that AB is the cause of AD pathology.

The reasoning behind this conclusion is the subject of the following section.

A/3 and Neumtoxicity: the AB Hypothesis of AD

Researchers have hypothesized that AB is the causal agent in AD because AB is

the main component of senile plaques and because these plaques are associated with

dystrophie neetes. There is, however, no direct evidence establishing that Ab causes the

disease. The reasons for the hypothesis are: in both early- and late-onset forms of AD,

gene mutations appear to exert effects on AB by altering APP processing to AB andlor by

promoting AB deposition and amyloid accumulation; also, the neurotoxicity of AB fibrils

has been demonstrated in numemus ceiI culture studies (Püce et al., 1993; Simmons et al.,

1993; Lorenzo and Yankner, 1994).

Early-Onset AD (EOAD)

It is estimated that early-onset AD (also knom as familial AD) occurs in 10% of

all AD cases. It is geneticdy inherited and arises fiom mutations in at least three genes:

APP, presenilin 1 (PSI) and presenilin 2 (PS2). ALmost ail of these mutations are

missense substitutions resuiting in a single amino acid change.

APP mutations

APP generates an array of peptide products of which AB is ody one. Three

enzyme activities are involved in the proteolytic processing of APP (Figure 1.2): a-

secretase, p-secretase and y-secretase. The B-secretase cleaves APP to form two

fragments APPsf3, a -100 kD soluble NHz-terminal fragment, and C99 (12 kDa,

membrane-bound). a-secretase cleaves within the AB sequence at Lys 16 producing a

soluble NH&agment, APPsa, and the 10 kDa membrane-bound C83. Finally, the y-

secretase cleaves either C99 or C83 to produce the COOH-terminus of AB or the non-

pathogenic peptide p3, respectively.

APP mutations account for 0.1% of AD cases (Tanzi et al., 1996). The fust

identified mutation, V717I or "London", was found in APP adjacent to the region

encodkig AB. Other mutations and their proximity to the Ab motif suggest they have an

effect on APP processing by selectively promoting the cleavage of AB. Missense

mutations at the COOH-terminus, the London and "Indiana" (Vïl7F) mutations, produce

an increase in the proportion of AB42 relative to AB40 (Suniki et al., 1994). Under in

vitro conditions, AB42 is the more fibdiogenic and cytotoxic of the Ab peptides and

appears to be important for initiation of AB deposition (Hïlbich et al., 199 1 ; Jarrett et al.,

1993; Iwatsubo et al., 1994; Yolmkin, 1995). The 670/671 double mutation (Lys-Met to

Asn-Leu or "Swedish") located immediately adjacent to the NHt-terminus of AB,

enhances AB production 6-fold (Cai et al., 1993).

How these mutations cause the increased production of AB is d l in question, but

the proximity of the mutations to the termini of the A$ sequence within N P suggests

that alteration of APP processing may be the explanation. Because the B- and y-secretase

activities are required for the release of AB fiom APP, a great ded of effort has been

dïrected at identiQing these associated enzymes. Only recently have these proteases been

identified. Vassar et al. (1 999) cloned an aspartic protease with ail known characteristics

of 8-secretase which they termed BACE (beta-site APP-cleaving enzyme). BACE

overexpression încreased p-secretase cleavage products, anti-sense inhibition of BACE

mRNA decreased B-secretase products and purified BACE cIeaved APP-derived

substrates as expected of fl-semetase.

The y-secretase is respoosible for determinhg the COOH-terminal -cation

point for AB. The peptide length has a signifïcant impact on the fibrillogenic and

cytotoxic properties of AB. Thus, modulation of the y-secretase activity wouid be an

important therapeutic approach. Recently, Li and CO-workers made the fust important

step in identwg the proteins involved in this process when they reported that the two

EOAD-linked proteins PSI and PS2 contain the active site of the y-secretase activity (Li

et ai., 2000).

Mutations in Psi and PS2

In 1995, it was found that mutations which cosegregate with a substantial number

of EOAD pedigrees are contained in two genes, PSI and PS2. It is estimated that PSI and

PS2 mutations are nspoasible for as much as 50% of all EOAD cases. The PS1 gene is

located on chromosome 14 and encodes an amino acid sequence of 467 residues

(Sherrington et ai., 1995). PS2, encoded on chromosome 1, is 448 amino acids long

(Levy-Lahad et al., 1995). PS t and PS2 have 67% amino acid identity. More than 50

mutations in PSI have been described while 2 have been reported in PS2. It is predicted

that PSI has 7-9 transmembrane domains with an extended hydrophilic "loop" region and

both the NHt-terminus and COOH-termirius located in the cytosol (Lendon et al., 1997)

Mutant PSI alten APP processing by selectively increasing the production of

AB42 (Scheuner et al., 1996; Duff et al., 1996; Borchelt et al., 1996) and accelerates

amyloid deposition (Lemere et al., 1996). When PSI genes were deleted in mice neurons,

y-secretase activity was abolished. Also, Wolfe et al. (1999) reported that mutation of

either of two aspartate residues in transmembrane domains 6 and 7 nullified y-secretase

activity. However, PS 1 and PS2 bear Little homology to known aspartyl proteases. It was

uncertain whether the presenilins were cofactoa acting in conjunction with other proteins

in the proteolysis of APP or acted directly on APP. Li et al. (2000) found that transition-

state specinc y-secretase inhibitors diiected at the active site of aspartyl proteases bind

specincaily to the presenilins. Thus, the presenilins contain the catdytic site of the y-

semetase activity and act directly on APP. Subsecpentiy, Yu et al. (2000) reported that a

newly identified trausmembme protein they termed nicastrin complexes with PS IPS2

and the COOH-terminal denvatives to modulate the production of AB.

In PSlIPS2 mediated EOAD, the biological impact is severe; each of the PS 1

mutations, except for one, is niacient to cause AD. That PSI and PS2 are responsible for

the production of AB and that altered APP processing and increased AB deposition are

the primary outcornes of the PS lIPS2 mutations, provide strong support for the idea that

AB plays a seminal role in AD.

Down Syndrome

In Down syndrome (DS), extensive Afl deposition in the form of senile plaques

and accumulation of NFTs also occurs. Examination of DS patients dying at various ages

reveals a time course for progression fiom amorphous, non-fibrillar AB deposits to the

development of amyloid fibrils and senile plaques (Giaccone et al., 1989; Mann, 1989;

Motte and Williams, 1989). For this reason, amorphous amyloid has also been termed

"preamyloid". Amorphous amyloid plaques are found in DS subjects as early as the mid-

teens and are observed regularly in those in their twenties and thirties. These diffise

plaques are more prevalent in younger patients and are not associated with dystrophie

neurites. These observations suggest the possibility there is an age-dependent evolution in

structure and fiuiction, from an inert, non-pathogenic form to an active, neurotoxic form.

DS patients have k e copies of chromosome 21 (Wisniewski et al., 1985). It is

presumably the relative overexpression of the APP gene Located on chromosome 21 and

the consequent increase in AB b d e n which is responsible for the APlike pathology.

Late-Onset AD (LOAD): Genetic Polymorphisms in Apolipoprofein €4 and

a-2-Macrogfo bulin

Late-onset AD accounts for the majority of AD cases and is associated with

genetic and environmentai nsk factors. Unlike early onset AD where gene mutations are

the cause, the genetics in late onset AD Uivolve common population polymorphisms

within the genorne. On their own, each of these variants are largely innocuous; however,

when present with other variants in a cornplement of polymorphisms, these variants

appear to predispose an individuai to AD. Some 21 genes have been associated with

increased risk for AD including apolipoprotein E (ApoE), a-2-macroglobulin (A2M), a-

I-antichymotrypsin (ACT), the very low density lipoprotein receptor (VLDLR), the

LDL-receptor related protein (LW), bleomycin hydrolase (BH), estrogen receptor a

(ER-@, oeurotrophin-3 (NT3) and transferrin (TF). Of these, the strongest associations

have been established for ApoE and A2M.

ApoE

ApoE is a 34 kDa product of a 4 exon gene on chromosome 19 which is involved

in cholesterol metabolism (Mahley, 1988). Of the three major isofoms of ApoE, the ~4

allele is the one linked to increased risk for Iate-onset AD (Corder et al., 1993;

Strittmatter et al., 1993). Inheritance of the ~4 allele correlates with increased AB

deposition and higher seaile plaque density in the cerebrd cortex (Schmechel et al.,

1993). In vitro, ApoE promotes assembly of AB into EIarnents (Ma et al., 1994); but, the

precise role of ApoE allele ~4 in AD is unknown. Thomas Wisniewski has proposed that

ApoE may either act as a ''pathological chaperone" which induces pathological folding

and polymerization of AB, or that its hgments, which are themselves fibrilIogenic, may

initiate AB nbd formation (Wisniewski and Frangione, 1992; Wiiewski et al., 1995).

A2M

MM, a s e m protehase inhibitor, binds AB with high affinity. A2M can degrade

AB (Qiu et al., 1996) and in vitro, A2M has been observed to inhibit fibril formation and

toxicity (Du et al., 1998). For those reasons A2M polymorphisms became a candidate for

genetic risk in AD. One polymorphism, A2M-2 was found to confer increased risk for

AD (Blacker et al., 1998). A second, CPP V10001, was also identified and associated

with signifcantly increased A$ deposition (Liao et al., 1 998; My lly kangas et al., 1 999).

It is important to emphasize that the DNA variations in these genes are neither

causative to, nor necessary for, AD. It is the association of increased accumulation of AB

in carriers veMs non-carriers which suggests a critical rote for AB in the disorder.

Molecular Mechanisms of AB Neurotoxicity

As part of the effofort to establish a causal relationship between Ab and AD,

researchers have identified some potential mechanisms of AB toxicity . Some whic h have

been descn'bed are: induction of apoptosis (Loo et al., 1993), generation of toxic levels of

hydrogen peroxide (Behl et ai., 1994), abenant activation of ce11 d a c e receptors (Zhang

et al., 1994), porekhanne1 formation in the ce11 membrane causing altered ion

homeostasis (Arispe et al., 1993), binding to C lq and complement activation (Rogers et

al., 1992) and acceleration of tissue plasmuiogen activator ( P A ) mediated activation of

plasminogen to plasmin (Kingston et al., 1995).

Confroversies and lssues hcing AD Research

The existing genetic and neurotoxic evidence attests to an important

interrelationship between AB and AD, but it rernains to be proven that AB causes

pathogenesis of AD. Much of the uncertainty surrounds the fact that a poor correlation

exists between senile plaques and the extent of dementia (Terry et al., 1991). Also the

results fiom AB toxicity experiments have been inconsistent, as some studies show its

ability to kill cells while othen indicate AB does not cause significant ce11 death

(Stephenson and Clemens, 1992; Podlisny et ai., 1993). Finally, animal models of AD

exhibit A$ deposition but fail to develop other AD associated pathologies. AB

fibrillogenesis is a complex process. Perhaps, in the effort to dissect this process into its

component pathways a disproportionate emphasis has been placed on the fibnllar species.

The idea of fibrillar AB as the sole neurotoxic agent in the disease may be an

ovenimplification.

What is the neurotoxic species in AD?

Recentiy, more attention has been directed towards the soluble fonns of AB.

Monomerddimea have been fond to be toxic in cultures of rat hippocampal neuron glia

cells (Roher et al., 1996). Also, mal1 soluble oligomers (17 and 27 D a ) called A b

derived diffusibIe ligands (ADDLs) have been identifîed (Lambert et ai., 1998). ADDLs

disnipt long-term potentiation and are toxic at nanomolar concentrations when added to

mouse brah siice cultures. Walsh et al. (1999) have found that protofibrillat AB a6ects

normd metabolism in cultured neurons.

These findings indicate that soluble forms of AB are potent neurotoxhs. Possibly,

soluble AB rnay be the prhciple toxic species in AD. If so, this could help explain the

confIicting evidence in Linking AB deposition and fibrillar amyloid with the disorder.

Levels of soluble A$ may correlate better than senile plaque number with the extent of

dementia. Though soluble AB is found in both nomal and AD brains, soluble AB is

elevated in AD and is elevated prior to plaque formation in DS (Teller et al., 1996).

Inconsistencies in demonstrating AB toxicity rnay be due not only to the failure of

enniring the formation of WUar AB but also failure to consider the effects of soluble

AB forms. Animal models may not show AD pathologies other than amyloid deposition

because the models do not generate pathologically sigificant levels of soluble AB.

Clearly, the pathogenesis of AD involves more than the association of the AB

peptides to form fibrils. The importance of the formation of aiternate products, such as

soluble AB forms and diffuse amyloid, is now coming under closer scrutiny. The

discovenes of the B-secretase, BACE, and the y-secretase roles of PS 1, PS2 and nicastrin

give researchers promising targets for treating AD by controlling the metabolism of APP

to form AB. However, the development of effective therapies specifically targeting these

secretases remains a signiscant challenge. In the meantime, altemate means of treatment

m u t be explored. Among the most prornising strategies is the inhibition of AB

aggregation into toxic foms. The incompleteness of our knowledge of the pathways of

AB aggregation and of the factors which control the AB assembly remains a signifcant

hindrance in this effort.

The main objective of this present work is to untangle some of the confused

hterrelationships surroundhg alternate aggregation pathways of AB fibrillogenesis and

explain some of the factors which control the aggregationai behavior of AB. Two

products of aiternate aggregation will be examined, soluble forms of AB and diffuse

amy Ioid.

Properties of soluble AB

With respect to soluble forms of AB, some of the questions we asked were as

follows. What are the physicai properties of soluble AB? What is its conformation? What

are the sizes of these oligomers and what conditions detennine their stability? Answen to

these questions would help place where soluble A$ is located in the fibrillogenesis

pathway. Furthemore, quantitative analysis would provide information critical to

developing therapeutic strategies such as inhibition or reversai of AB association W o r

polymerization.

Properties of diffuse amyloid

The properties of diaise amyloid are also inadequately understood Is diffuse

amyloid the progenitor of i%riUar amyloid, and if so, what uitluences the switch fiom

diaise to fibrillar? Does diffuse amyloid decrease amyloid burden in the brain by acting

as an inert reservoir for AB?

In order to carry out a methodical investigation into the properties of d i f h e

amyloid, it

diffuse AB

wouid be helpful to establish in vitro conditions where diffuse AB forms. If

can be formed in vitro, then the application of fluorescence, CD and AFM

methods can be applied to gain new insights into the aggregational and secondary

structural characteristics of diffuse arnyloid.

Extrinsic factors which influence aggregational properties of AB

The evolution of $-sheet secondary structure accompanies the polyrnerization of

AB into amy loidogenic fibriis. What factors influence the conformational and consequent

aggregational characteristics of AB? Does folding determine whether the peptide

associates into oligomers, or protofibrils, or any other aggregated product? M a t are the

effects of arnyloid-associated proteins?

Chapter 2: Structural studies of soluble oligomers of the Alzheimer &amyloid

peptide

Structural Studies of Soluble Oligomers of the AIzheimer B-A myloid Peptide

describes how we were able to stabilize Ab oligomers. Maintenance of these structures

for extended periods of t h e allowed the application of fluorescence and circular

dichroism spectroscopy, analyticd ultracentrifugation, and electron and atomic force

microscopy techniques. These studies should aid in identifying the mechanism(s) by

which these oligomers play a role in the pathogenesis of Alzheimer disease, and aid in the

development of novel approaches to inhibit or abolish the neurotoxicity of these

AB oiigomea.

In this work 1 performed al1 the spectroscopie experiments. Atomic force

microscopy experiments were done by N. Plaskos and C. Yip at the Department of

Chernical Engineering and Applied Chemistry, University of Toronto and electron

microsopy experiments by D.S. Yang and P.E. Fraser at the Center for Research in

Neurodegenerative Disease, University of Toronto. The sedimentation experiments were

performed by S. Go at the analytical ultracentrifugation facility at the University of

Toronto,

Chapter 3: Altemate aggregation pathways of the Alzheimer bamyloid peptide:

an in vitro mode1 of diffuse amyloid

In the course of the work in the laboratory, non-specific AB aggregates often

formed. Codd these in vitro aggregates be the amorphous amyloid in diffuse plaques? In

Alternate Aggregation Pathways of the Alzheimer PAmyloid Peptide: an in vitro mode1

of dzBse mgdoid, we propose methods by which amorphous arnyloid can be formed in

vitro. These methods will allow for investigation into the aggregational and secondary

structural characteristics of the altemate aggregation pathways of AB. These fmdings will

provide a better understandmg of the fibril formation process and may have implications

in understandhg the reIationship between diffuse AB and amyloid fibrils.

1 completed al1 the spectroscopic experiments, while electron microscopy was

carried out by D.S. Yang and P.E. Fraser.

Chapter 4: Manipulating the arnyloid-f3 aggregation pathway with chemical

chaperones

The ability of AB to form fibrils and amorphous amyloid is intrinsic to the peptide

itself. However extrinsic factors wiiI dso play roles in the aggregational behaviour of AB.

Because the folding of AB takes place w i t . htraceliular compartments it likely involves

protein chaperones or other chaperone elements which are responsible for controllhg

protein foldùig. The chemicd agents trimethylamine N-oxide (TMAO) and gIycerol are

organic osmolytes which can influence protein folding by altering the hydration forces

around the peptide backbone. In this study, TMAO and giycerol were used to investigate

how folding and aggregation of AB can be iduenced by other molecules.

1 perfomed the fluorescence experiments and half the CD experiments. Electron

microscopy and the remainder of the CD experiments were completed by D.S. Yang and

P.E. Fraser. TMAFM was done by C. Yip.

Abraham, CB., Sekoe, DJ., and Potter, H. (1988). Immunochemical identification of the

serine protease inhibitor a-1-antichymotrypsin in the brain amyloid deposits of

Alzheimer's disease. Cell52,487-50 1.

Alzheimer, A. (1907) Uber eine eigematige Erkrankung der Hirnrinde. Allg. 2. Psychiatr.

Psychisch-Gericht. Med. 64,146- 148.

Arispe, N., Pollard, H.B., and Rojas, E. (1993) Giant multilevel cation channels formed

by Alzheimer disease amyloid b-protein [A$-(1-40)] in bilayer membrane. Proc. Natl.

Acad. Sci. USA 90,10573-10577.

Bmow, C.J., and Zagorski, M.G (1991). Solution structures of p-peptide and its

constituent fragments: relation to amy loid deposition. Science 253, 179- 1 82.

Blacker, D., et al. (1998) Alpha.2-macroglobulin is genetically associated with

Alzheimer's disease. Nat. Genet. 19,357-360.

Behi, C., Davis, J.B., Lesley, R, and Schubert, D. (1994) Hydrogen peroxide mediates

amyloid B protein toxicity. CeU 77,8 17-82?.

Blacker, D., Wilcox, M.A., Laird, N.M., Rodes, L., Horvath, S.M., GO, R.C., Perry, EL,

Watson, B.Jr., Bassett, S.S., Mcuinis, M.G., Albert, M.S., Hyman, B.T., and Tanzi, E U .

(1998) Alpha-2 macroglobulin is geneticaily associated with Alzheimer disease. Nat.

Genet. 19,357-360.

Borchelt, D.R., niinakaran, G., Eckman, C.B., Lee, M.K., Davenport, F., Ratovitsky, T.,

Prada, CM., Km, G., Seekins, S., Yager, D., Slunt, H.H., Wang, R, Seeger, M., Levey,

AL, Gandy, S.E., Copelmd, N.G., Jenkins, NA., Pnce, D.L., and Younkia, S.G. (1 996)

Familial Alzheimer's disease-linked presenilin 1 variants elevate AB 1 -42/ 1-40 ratio in

vitro and in vivo. Neuron 17,1005-101 3,

Busciglio, J., Gabuzda, D.H., Matsudaira, P., and Yankner, B.A. (1993) Generation of 8-

arnyloid in the secretory pathway in neuronal and non-neuronal cells. Roc. Nad. Acad.

Sci. USA 90,2092-2096.

Cai, X.D., Golde, T.E., and Younkin, S.G. (1993) Release of excess arnyloid protein

fiorn a mutant amyloid B protein precursor. Science 259,s 14-5 1 6.

Canadian Study of HeaIth and Aghg Workiag Group. (1994) Canadian study of heaith

and aging: study methods and prevalence of dementia. Can. Med. Assoc. 1.1 50,899-91 3.

Coria, F., Castaiio, E.M., Relli, F., Larrondo-Lillo, M., van Duinen, S., Shelanski, ML.,

and FrangÏone, B. (1988). Isolation and characterization of amyloid P component from

Alzheimer's disease and other types of cerebral amyloidosis. Lab. h v e d 58,454-458.

Corder, E.H., Saunders, A.M., Strittmatter, W.J., Schmechel, D.E., Gaskell, P.C., Small,

G.W., Roses, A.D., Haines, J.L., and Pericak-Vance, M.A. (1993) Gene dose of

apolipoprotein E type 4 allele and the risk of Alzheimer's disease in late onset f i l i e s .

Science 26 1,921-923.

de Sauvage, F., and Octave, LN. (1989). A novel mRNA of the A4 amyloid precursor

gene coding for a possibly secreted protein. Science 245,65 1-653.

Du, Y., Bales, K.R, Dodel, R.C., Liu, X., G h , MA., Homy LW., Little, S.P., and Paul,

S.M. (1 998) Mpha2-macroglobuh attenuates beta-amyloid peptide 1-40 fibril formation

and associated neurotoxicity of cuitured fetal rat cortical neurons. J. Neurochem. 70,

11 82-1 188.

Duff, K., Eckman, C., Zehr, C., Yu, X., h d a , C.M., Perez-TUT, L, Hutton, M., Buee, L.,

Harigaya, Y., Yager, D., Morgan, D., Gordon, M.N., Holcomb, L., Refolo, L., Zenk, B.,

Hardy, J., and YounkUi, S.G. (1996) Increased amyloid-b42(43) in brains of mice

expressing mutant presenilllil . Nature 383,710-71 3.

Fraser, P.E., Nguyen, J.T., Surewicz, W.K., and Kirscher, DA. (1991). pH-dependent

structurai transitions of Alzheimer amyloid BlA4 peptides. Biophysical J. 60, 1 190-120 1.

Fraser, P.E., Lévesque, L., and McLachlan, D.R. (1993) Biochemistry of Alzheimer's

disease amyloid plaques. Clin. Biochem. 26, 1-1 1.

Giaccone, G., Tagliavini, F., Linoli, G., Bouras, C., Frigerior, L., Frangione, B., and

Bugiaai, 0. (1 989). Down patients: extraceIlular prearnyloid deposits precede neuntic

degeneration and senile plaques. Neurosci. Lett. 97,232-238.

Glemer, G.G., Eanes, E.D., and Page, D.L. (1972) The relation of the properties of

Congo red-stained amyloid fibrils to the fl-conformation. J. Histochem. Cytochem. 20,

82 1-826.

Glenner, G.G., and Wong, C.W. (1984) Alzheimer's disease: initial report of the

purification and characterization of a novel cerebrovascuiar amyloid protein. Biochem.

Biophys. Res. Commua. 120,885-890.

Gorevic, P.D., Castaiio, E.M., Sarnia, R-, and Fmgione, B. (1987). Ten to fourteen

residue peptides of Alzheimer's disease protein are sufncient for amyloid f i bd formation

and its characteristic X-ray ciBiaction pattern. Biochem. Biophys. Res. Commun. 147,

854-862.

Haass, C, Schlossmacher, M.G., Hung, A.Y., Vigo-Pelney, C., MeLlon, A., Ostaszewski,

B-L., Lieberburg, L, KOO, E.H., Schenk, D., TepIow, D.B. and Sekoe, D.J. (1992)

Amyloid beta-peptide is produced by cuitured cells d d g normal metabolism. Nature

359,322-325.

Hdverson, K., Fraser, P.E., Kùschner, D.A., and Lansbury, P.T. (1990). Molecular

determinants of amyloid deposition in Alzheimer's disease: conformational studies of

synthetic beta-protein fragments. Biochemistry 29,2639-2644.

Hilbich, C., Kisters-Woike, B., Reed, J., Masters, C.L., and Beyreuther, K. (1991)

Aggregation and secondary structure of synthetic amyloid beta A4 peptides of

Alzheimer's disease. J. Mol. Biol. 218, 149-163.

Hilbich, C., Kister-WoUte, B., Reed, J., Masters, CL., and Beyreuther, K. (1992).

Substitutions of hydrophobie amino acids reduce the amyIoidogenicity of Alzheimer's

disease BA4 peptides. J. Mol. Biol. 228,460-473.

Iwatsubo, T., Odaka, A., Suniki, N., Mizusawa, K., Nukina, N., and &ara, Y. (1994)

Visualization of A842(43) and AB40 in senile plaques with end-specific AB

monoclonais: Evidence that an initially deposited species is AP2(43). Neuron 13,4543.

Janett, J.T., Berger, E.P., L m s b ~ y ~ P.T. Jr. (1993) The carboxy terminus of the beta

amy1oid protein is critical for the seeding of arnyloid formation: Implications for the

pathogenesis of Alzheimer's disease. Biochemistry 32,4693-4697.

Jarrett, J.T., and Lansbury, P.T.Jr. (1 993). Seeding "one-dimensionai crystdlization" of

arnyloid: a pathogenic mechanism in Alzheimer's disease and scrapie? Ce11 73, 1055-

1058.

Joachim, C.L., Duffy, L.K., Moms, J.H., and Selkoe, D.J. (1988) Protein chernical and

Unmunocytochemicd studies of meningovascular beta-amyloid protein in Alzheimer's

disease and nonal aging. Brain Res. 474,100-1 1 1.

Kang, J., Lemaire, H.-G., Unterbeck, A., Salbaum, J.M., Masters, C.L., Gneschik, K.H.,

Multhaup, G., Beyreuther, K., and MWer-Hill, B. (1987) The precursor of Alzheimer's

disease arnyloid A4 protein resembles a cell-dace receptor. Nature 325,733-736.

Kidd, M. (1964) Alzheimer's disease- An electron microscopical study. Brain 87,307-

320.

Kingston, LB., Castro, M.J., and Anderson, S. (1995) In vitro stimulation of tissue-type

plasminogen activator by Alzheimer amy loid B-peptide analogues. Nat. Med. 2, 1 3 8- 142.

Kirscher, D., Abraham, C., and SeIkoe, D. (1986) X-ray ciifTiaction fkom intraneuronal

paired helical filaments and extraneuronal amyloid fibers in Aizheimer disease indicates

cross-8 conformation. hoc. Natl, Acad. Sci. USA 83,503-507.

Kirscher, DA., Inouye, H., Duw, LX., Sinclair, A., Lind, M., and Selkoe, D.J. (1987).

Synthetic peptide homologous to beta protein fiom Alzheimer disease foms arnyloid-like

nbrils in viîro. Proc. Natl. Acad. Sci. USA 84,6953-6957.

Kitaguchi, N., Takahashi, Y., Tokushima, Y ., Shiojiri, S., and Ito, H. (1 988). Novel

precursor of Alzheimer's disease amyloid protein shows protease inhibitory activity.

Nature 33 1,530-532.

Kuo, Y.-M., Emmerling, M.R., Vigo-Pelfiey, C., Kasunic, T.C., Kirkpatrick, J.B.,

Murdoch, G.H., BaIl, MJ., and Roher, A.E. (1996) Water-soluble AB(N-40, N-42)

oligomers in normal and Alzheimer disease braim. J. Biol. Chem. 271,4077408 1.

Lambert, M.P., Barlow, A.K., Chrorny, B.A., Edwards, C., Freed, EL, Liosatos, M.,

Morgan, T.E., Rozovsky, L, Trommer, B., Viola, K.L., Wals, P., Zhang, C., Finch, C.E.,

Krant, GA., and ECIein, W.L. (1 998) Diffiisrile nonfibrillar iigands derived fiom AB1-42

are potent central nervous system neurotoxins. Proc. Nad. Acad. Sci. USA 95, 6448-

6453.

Lemere, Ch., Lopera, F., Kosik, K.S., Lendon, CL., Ossa, J., Saido, T.C., Yamaguchi,

H., Ruiz, A., Martinez, A.O., Madrigal, L., Hincapie, L., Arango, J.C., Anthony, D.C.,

Koo, CH., Goate, A., and Selkoe, D.J., (1996) The E280A presenilin 1 Alzheimer

mutation produced A842 deposition and severe cerebellar pathology. Nature Med. 2,

1146-1 150.

Lendon, CL., Ashall, F., and Goate, A.M. (1997) E x p l o ~ g the etiology of Alzheimer

disease using molecular genetics. I. Am. Med. Assoc. 277,825-83 1.

Levy-Lahad, E., Wasco, W., Poorkaj, P., Romano, D.M., Oshima, J., Pattingell, W.H.,

Chang-en, Y., Jondro, PD., Schmidt, S.D., Wang, K., Crowley, AC., Fu, Y.H.,

Guenette, S.Y., Galas, D., Nemens, E., Wijman, E.M., Bird, T.D., Schellenberg, G.D.,

and Tanzi, R.E. (1995) Candidate gene for the chromosome I familial Alzheimer's

disease locus. Science 269,973-977.

Li, Y.-M., XII, M, Lai, M.-T., Huang, Q., Castro, J.L., DiMuzÎo-Mower, J., Hamson, T.,

Lellis, C., Nadin, A., Neduveiii, J.G., Register, R.B., Sardana, M.K., Shearman, M.S.,

Smith, A.L., Shi, X.-P., Yin, K L , Shafer, Jh., and Gardell, S.J. (2000) Photoactivated

y-secretase inhibitors directed to the active site covdentiy label presenilin 1. Nature 405,

689-694.

Liao, A., Nitsch, RM., Greenberg, S.M., Finckh, U., Blacker, D., Albert, M., Rebeck,

G.W., Gomez-Ida, T., Clatworthy, A., Binetti, G., Hock, C., Mueller-Thomsen, T.,

Mann, U., Zuchowski, K., Beisiegel, U., Staehelin, H., Growdon, JH., Tanzi, RE., and

Hyman, B.T. (1998) Genetic association of an aipha2-rnacrogIobulin (Vd1000Ile)

polymorphism and Alzheimer's disease. Hum. Mol. Genet. 7, 1953-1956.

Loo, D.T., Copani, A.' Pike, C.J., Whittemore, E.R., Walencewicz, A.J., and Cotman,

C.W. (1993) Apoptosis is induced by fbamyloid in cultured central nervow system

neurons. Proc. Nad. Acad. Sci. USA. 90,795-7955.

Lorenzo, A. and Yankner, B. (1994) Beta-amyloid neurotoxicity requires fibril formation

and is inhibited by congo red. Proc. Nati. Acad. Sci. USA 91, 12243-12247.

Ma, J., Yee, A., Brewer, H.BJr., Das, S., and Potter, H. (1994) Amyloid-associated

proteins alpha 1-antichymotrypsin and apolipoprotein E promote assembly of Alzheimer

beta-protein into filaments. Nature 372,92-94.

Mahfey, R.W. (1988) Apolipoprotein E: Cholesterol transport protein with expanding

role in ceiI biology. Science 240,622-630.

Mann, D.M.A. (1989). Cerebrai amyloidosis, aging and Alzheimer's disease: a

contribution h m studies on Down's syndrome. NeurobioI. Agbg 10,397-399.

Masters, C.L., Simms, G., Weinman, N.A., Multlraup, G*, McDonaid, B.L., and

Beyreuther, K. (1985) Amyloid plaque core protein in Alzheimer disease and Down

syndrome. Proc. Natl. Acad. Sci. USA 82,4245-4249.

McGeer, P.L., Akiyama, H., and McGeer, E.G. (1989). Immune system response in

Alzheimer's dx'sease. Can. J. NeuroI. Sci. 16,5 16-527.

Merz, P.A., Wisniewski, H.M., Somerville, R.A., Bobin, S.A., Masters, C.L., and Iqbal,

K. (1983) Ultrastructural morphology of amyloid fibrils from neuritic and amyloid

plaques. Acta Neuropathol. 60,113-124.

Milward, E.A., Papadopoulos, R., Fuller, S.J., Moir, R.D., SrnaIl, D., Beyreuther, K., and

Masters, C.L. (1992). The amyloid protein precursor of Alrheimer's disease is a mediator

of the effects of nerve growth factor on neurite outgrowth. Neuron 9,129-137.

Motte, J., and Williams, R.S. (1 989). Age-related changes in the density and morphology

of plaques and neurofibrîilary tangles in Down syndrome brain. Acta Neuropathol. 77,

535-546.

Müiier-Hi& B., aad Beyreuther, K. (1989) MolecuIar biology of Alzheimer's disease.

Ann. Rev. Biochem. 58,287-307.

MyUykangas, L., Polvikoski, T., Suikava, R., Verkkoniemi, A., Crook, R., Tienari, P.J.,

Pusa, AX., Niinisto, L., O'Brien, P., Kontula, K., Hardy, J., Haltia, M. and Perez-Tur, 5.

(1999) Genetic association of alpha-2-macroglobulin with Alzheimer's disease in a

Finnish elderly population. AM. Neurol. 46,382-390.

Narang, H.K. (1980) High-resolution electron rnicroscopic analysis of the amyloid fibril

in Alzheimer's disease. J. Neuropath. Exp. Neurol. 39,621-63 1.

Pike C.J., Burdick, D., Waiencewicz, A.J.? Glabe, CG., and Conan, C.W. (1993).

Neurodegeneration induced by beta-amyloid peptides in vitro: the role of peptide

assembly state. J. Neurosci. 1 3, 1 676- 168%

Podlisny, M.B., Stephenson, D.T., Frosch, M.P., Tolan, D.R., Lieberburg, I., Clemens,

I.A., and Sekoe, D.J. (1993) Microinjection of synthetic amyloid B-protein in monkey

cere brai cortex fails to produce acute neurotolticity. Am. J. Pathol. l42,17-24.

Ponte, P., Godez-Dewhitt, P., Schilling, J., Milier, J., Hsu, D., Greenberg, B., Davis,

K., Wdlace, W., Lieberburg, I., Fuller, F., and Cordell, B. (1988). A new A4 amyIoid

mRNA contaios a domain homologous to serine proteinase inhibitors. Nature 33 1, 525-

527.

Preiii, F., Castao, E.M., GIenner, G.G., and Frangione, B. (1988) Differences between

vascular and plaque core amy 10 id in Alzheimer's disease. J. Neurochern. 5 1,648065 1.

Puchtler, H., Sweat, F., and Levine, M. (1962) On the binding of Congo red by amyloid.

J. Histochem. Cytochem. 10,355-364.

Qiu, W.Q., Borth, W.. Ye, Z., Haass, C., Teplow, D.B., Selkoe, D.J. (1996) Degradation

of amyloid beta-protein by a serine protease-dpha2-macroglobdin cornplex. J. Biol.

Chem. 27 1,8443-845 1.

Robakis, NX., Ramakrishna, N., Wolfe, G., and Wisniewski, KM. (1987). Molecular

cloning and characterization of a cDNA encoding the cerebrovascular and the neuritic

plaque amyloid peptides. hoc. Natl. Acad. Sci. USA 84,4 190-41 94.

Rogers, J., Cooper, N.R., Webster, S., Schultz, J., McGeer, P.L., Styren, S.D., Civin,

W.H., Brachova, L., Bradt, B., Ward, P., and Lieberburg, 1. (1992) Complement

activation by 8-amyloid in Alzheimer's disease. hoc. Natl. Acad. Sci. USA 89, 100 16-

10020.

Roher, A.E., Chaney, M.O., Kuo Y.-M., Webster, S.D., Stine, W.B., Haverkamp, L.J.,

Woods, A.S., Cotter, R.J., Tuohy, LM., Krafft, G.A., Bonneil, B.S., and Emmerling,

M.R. (1996) Morphology and toxicity of AB-(1-42) dimer denved from neuritic and

vascular amyloid deposits of Alzheimer's disease. J. Biol. Chem. 271,2063 1-20635.

S t George-Hyslop, P.H., Tami, R.E., Polinsky, R.J., Haines, J.L., Nee, L., Watkins, P.C.,

Myers, RH., Feldman, R.G., Pollen, D., Drachman, D., Growdon, J., Bruni, A., Foncin,

J.-F., Salmon, D., Frommelt, P., haducci , L., Sorbi, S., Piacentini, S., Stewart, G.C.,

Hobbs, W.J., C O M ~ ~ Y , P., and Gusella, J.F. (1987). The genetic deficit causing familial

Alzheimer's disease maps to chromosome 21. Science 23 5,885-889.

Saitoh, T., Smdsmo, M., Roch, I.-M., Kimma, N., Cole, G., Schubert, D., Olteadorf, T.,

and Schenk, DB. (1989). Secreted form of amyloid B protein precursor is involved in the

growth regdation of fibroblasts. CeD 58,615-622.

Scheuner, D., Eckman, C., Jensen, M., Song, X., Citron, M., Suzuki, N., Bird, T.D.,

Hardy, J*, Hutton, M., Kukull, W., Lanon, E., Levy-Lahad, E., Vütanen, M., Peskuid, E*,

Poorkaj, P., ScheIlenberg, G., Tanzi, R., Wasco, W., Lannfelt, L., Sekoe, D., and

Younkin, S.G. (1996) Secreted amyloid beta-protein similar to that in the s e d e plaques

of Alzheimer's disease is increased in vivo by the presenilin 1 and 2 and APP mutations

Linked to f d i a l Alzheimer's disease. Nat. Med. 2,864-870.

Schmechel, D.E., Saunders, A.M., Stritfmatter, W.J., Crain, B.J., Hdette, C.M., Joo,

S.H., Pericak-Vance, M.A., Goldgaber, DeT and Roses, A.D. (1993) fncreased amyloid

beta-peptide deposition in cerebrai cortex as a consequence of apolipoprotein E genotype

in late-onset Alzheimer disease. hoc. Natl. Acad. Sci. USA 90,9649-9653.

Schubert, D., Jin, L W . , Saitoh, T., and Cole, G. (1989). The regulation of amyloid

protein precursor secretion and its modulatory role in celi adhesion. Neuron 3,689-694.

Seubert, P., Vigo-Pelfiey, C., Esch, F., Lee, M., Dovey, H., Sinha, S., Schlossmacher,

M.G., Whdey, J., Swindlehurst, C., McCormack, R, Wolfert, R, Sekoe, DJ.,

Lieberburg, I., and Schenk, D. (1992) Isolation and cpantificatiori of soluble Alzheimer's

fbpeptide from bioiogical fluids. Nature 359,325-327.

Shehgton, R, Rogaev, EL, Liang, Y., Rogaeva, EA., Levesques, G., Ikeda, M., Chi,

HoT Lin, C., Li, G., Hohan, K., Tsuda, T., Mar, L., Foncia, J.F., Bruni, A.C., Montesi,

M.P., Sorbi, S., Rainero, I., Pinessi, L., Nee, L., Chumakov, I., Pollen, D., Brookes, A.,

Sanseau, P., Polinsky, R.J., Wasco, W., DaSilva, H.A.R., Haines, J.L., Pencak-Vance,

MA., Tanzi, RE., Roses, A.D., Fraser, PB., Rommens, J.M., and St. George-Hyslop, P.

(1995) Cloning of a gene bearing missense mutations in early omet familial Alzheimer's

disease. Nature 375,754760.

Shoji, M., Golde, T., Ghiso, J., Cheung, T., Estus, S., Shaffier, L., Cai, X., McKay, D.,

Tinter, R., Fraagione, B., and Younkh, S. (1 992). Production of the Alzheimer arnyloid fl

protein by normal proteolytic processing. Science 258, 1 26- 129.

Simmons, L.K., May, P.C., Tomaselli, K.J., Rydel, R.E., Fuson, K.S., Brigham, E.F.,

Wright, S., Lieberburg, L, Becker, G.W., Brems, D.N., and Li, W.Y. Secondary structure

of amyloid beta peptide correlates with neurotoxic activity in vitra(1993) Molec. P h m .

45,373-379.

Snow, AD., Mar, H., Nochlin, D., Kirnata, K., Kato, M., Suniki, S., Hassell, J., and

Wight, TN. (1988). The pmence of hepanui sulfate proteoglycans in the neuritic plaques

and congophilic angiopathy in Alzheimer's disease. Am. J. Pathol. 133,456-463.

Stephenson, D.T., and Clemens, J.A. (1992) In vivo effects of 8-amyloid implants in

rodents: lack of potentiation of damage associated with transient global forebrain

ischemia. Brain Res. 585,2350246.

Strittmatter, W.J., Saunders, A.M., Schmechel, D., Pericak-Vance, M., Enghild, J.,

Salvesen, G.S. and Roses, A.D. (1993) Apolipoprotein E: High-avidity binding to beta-

amyloid and increased fiequency of type 4 de le in Iate onset familial AIzheher disease.

Proc. Naa. Acad. Sci. USA 90,1977- 198 1.

Suzuki, N., Cheung, T.T., Cai, T.T., Cai X.D., Odaka, A., Otvos, L., Eckman, C., Golde,

T.E., and Younkin, S.G. (1994) An Uicrease percentage of long amyloid p protein

secreted by familial amyloid protein precursor (BPP717) mutants. Science 264, 1336-

1340.

TanP, R.E., Gusella, J.F., Wakins, P.C., Bruns, G.A.P., St-George-Hyslop, P.H., van

Keuren, M.L., Patterson, D., Pagan, S., Kurnit, DM., and Neve, R.L. (1 987). Amyloid p

protein gene: cDNA, mRNA distribution, and genetic Iinkage near the Alzheimer locus.

Science 235,880-884.

Tanzi, R.E., McClatchey, AL, Lamper& E.D., Villa-Komaroff, LL., Gusella, J.F., and

Neve, RL. (1988). Protease inhibitor domain encoded by an amyloid protein precursor

mRNA associated with AIzheimer's disease. Nature 33 1,528-530-

Tanzi, R E , Kovacs, DM., Kim, T.W., Moir, R.D., Guenette, S.Y., and Wasco, W.

(1996) The gene defects responsible for familial Alzheimer's disease. Neurobiol. Dis. 3,

159-168.

Teiler, J.K., Russo, C., DeBusk, L.M., Angelini, G., Zaccheo, D., Dagna-Bricarelli, F.,

Scarteuini, P., Bertolini, S., Mann, D.M.A., Tabaton, M., and Gambetti, P. (1996)

Presence of soluble amyloid p-peptide precedes amyloid plaque formation in Down's

syndrome. Nature Med. 2,93-95.

Teny, R.D., Gonatas, N.K., and Wiss, M. (1964) Ultrastructural studies in Alzheimer's

presenile dementia Am. I. Pathol. 44,269-297.

Terry, RD., Masliah, E., Salmon, D.P., Butters, N, DeTeresa, R., Hill, R., and Katmian,

R. (1991) Physical basis of cognitive aiterations Ui Alzheimer's disease: synapse loss is

the major conelate of cognitive impairment. AM. Neurol. 30,572-580.

Vassar, R., Bennett, B.D., Babu-Khan, S., Kahn, S., Mendiaz, E.A., Denis, P., Teplow,

DB., Ross, S., Amarante, P., Loeloff, R., Luo, Y., Fisher, S., Fuller, J., Edenson, S., Lile,

J., Jarosinski, M.A., Biere, A.L., Cunan, E., Burgess, T., Louis, J - C , Collins, F.,

Treanor, J., Rogers, G., and Citron, M. (1999) B-secretase cleavage of Alzheimer's

amyloid precursor protein by the transmembrane aspartic protease BACE. Science 286,

735-74L.

Vigo-PeEey, C., Lee, D., Keim, P., Lieberburg, I., and Schenk, D.B. (1993)

Characterization of beta-amyloid peptide from human cerebrospinal fiuid. JNeurochern.

61, 1965-1968.

Walsh, D.M., Lomakln, A., Benedek, G.B., Condron, MM., and Teplow, D.B. (1997)

Amyloid B-protein fibrillogenesis: detection of a protofibrillar intemediate. J. Biol.

Chem. 272,236422372.

Walsh, D.M., Hartley, D.M., KWWUO~O, Y., Fezoui, Y., Condron, M.M., Lomakh, A.,

Benedek, G.B., Sekoe, D.J., and Teplow, D.B. (1999) Amyloid 8-protein fibrillogenesis:

structure and biologicai activity of protofibdlar intermediates. I. Biol. Chem. 274,

25945-25952.

Weidemann, A., Konig, G., Bunke, D., Fischer, P., Saibaum, LM., Masters, CL., and

Beyreuther, K. (1989). Identification, biogenesis, and localization of precursors of

Alzheimer's disease A4 amyloid proteiu. Cell57, 1 1 5- 126.

Wisniewski, K.E., DaIton, A.I., McLachian, C., Wen, G. Y., and Wisniewski, H.M.

(1 985) Alzheimer's disease in Down's syndrome: clinicopathologic studies. Neurology

35,957-961.

Wisniewski, T., and Franpione, B. (1992) Apolipoprotein E: a pathological chaperone

protein in patients with cerebral and systemic amyloid. Neurosci. Lett. 135,235-238.

Wisniewski, T., Lalowski, M., Golabek, A.A., Vogel, T., and Frangione, B. (1995) Is

Alzheimer's disease an apolipoproteh E amyloidosis? Lancet 345,956-958.

Wolfe, M.S., Xia, W., Ostaszewski, BL., Diehl, T.S., Kimberly, W.T., and Seikoe, D.J.

(1999) Two trammembrane aspartates in presenilin-1 required for presenilin

endoproteolysis and gamma-secretase activity . Nature 3 98,s 1 3-5 1 7.

Wood, S.J., Maleeff, B., Hart, T., am 1 Wetzel, R. (1996) Physicai, morphologicd and

functional differences between pH 5.8 and 7.4 aggregates of the Alzheimer's amyloid

peptide AB. I. Mol. Biol. 256,870-877.

Yamaguchi, H., Nakazato, Y., Hirai, S., Shoji, M., and Harigaya, Y. (1989) ) Electron

rnicrograph of diffuse plaques. Initiai stage of s e d e plaque formation in the Alzheimer

brain. Am. J. Pathul. 135,593497-

YoudÜn, S.G. (1995) Evidence that AB42 is the real culprit in Alzheimer's disease. Am.

Neurol. 37,287-288.

Yu, G., Nishimura, M., Arawaka, S., Levitan, D., Zhang, L., Tandon, A., Song, Y.-Q.,

Rogaeva, E., Chen, F., Kawarai, T., Supala, A., Levesque, L., Yu, H., Yang, DA.,

Hoimes, E., Milman, P., Liang, Y., Zhang, DM., Xu, D.H., Sato, C., Rogaev, E., Smith,

M., Janus, C., Zhang, Y., Aebersold, R, Faner, L., Sorbi, S., Bruni, A., Fraser, P., and St.

George-Hyslop, P. (2000) N i c d modulates presenilin-mediated ootch/glp- l signal

transduction and &APP processing. Nature 407,4844.

Zhang, C., Lambert, M.P., Bunch, C., Barber, K., Wade, W.S., Krafft, G.A., and Klein

W.L. (1994) Focal adhesion kinase expressed by nerve cell lines shows increased

tyrosine phosphorylation in response to Alzheimer's AB peptide. J. Biol. Chem. 269,

25247-25250.

Chapter 2: Structural Studies of Soluble Oligomers of the

Alzheimer PAmyloid Peptide

Huang, T.H.J., Yang, D.-S, Plaskos, N-P., Go, S., Yip, CM., Fraser, P.E., and Chakrabartty, A. (2000) J. Mol. Biol. 297,73-87.

Recent studies have suggested that nonfibrillar soluble forms of AB peptides

possess neurotoxic properties and may therefore play a role in the molecular pathogenesis

of Alzheimer disease. We have identified solution conditions under which two types of

sohble oligomers of AB40 could be trapped and stabilized for an extended period of

t h e . The Erst type of oligomer comprises a mhme of dimedtetramers which are stabIe

at neutral pH and low micromolar concentration, for a period of at least 4 weeks. The

second type of oligomer comprises a nanow distribution of particles that are sphericai

when examined by electron microscopy and atomic force microscopy. The number

average molecular weight of this distribution of particles is 0.94 MDa, and they are stable

at pH 3 for at least 4 weeks. Circular dichroism midies indicate that the dhershetramers

possess irregdar secondary structure that is not a-helix or $-structure, while the 0.94

MDa particles contain 8-structure. Fluorescence resonance energy transfer experiments

indicate that AB40 moieties in arnyloid fibriis or protofibds are more similar in structure

to those in the 0.94 MDa particles than those in the dimersltetramers. These fmdings

indicate that soluble oligomeric forms of Ab peptides can be trapped for extended periods

of tirne, enabling their study by high resolution techniques that wouid not othenMse be

possible.

The AB f d y of peptides is derived by the enzymatic breakdown of the amyloid

precursor protein (APP), a 563-770-residue membrane protein expressed in neuronal and

non-neuronal tissue. The two most abundant forms of A$ are, respectively, the 40 and 42

residue peptides AB40 and AB42 Both forms are capable of assembling into 60-100 A

diameter 8-sheet fibrils that exhibit the characteristic cross-6 X-ray fiber diffraction

pattern, and yield a red-green biref'rlngence when stained with Congo Red.

Pathologically, a key hallmark of Alzheimer disease (AD) is the formation of

insoluble deposits of A$ in the brah (Selkoe, 1991; Terry, 1994), both as diffuse and

senile plaques. While senile plaques comprise clusters of AB fibrils (and amyloid

associated proteins) m u n d e d by dead and dying neurons, and reactive astrocytes and

microglial cells (Brion, 1992), difiùse plaques are composed of predominantly

amorphous Ab but may contain a small population of AB fibrils (Yamaguchi et al.,

1989). The most important difference between plaque-types is that neurodegeneration

surrounds s e d e plaques but not diffuse plaques.

A signifcant body of evidence exists implicating the involvement of sede plaque

formation in the pathology of AD. Neurons in close prowimity to AB tibrils are

dystrophie. The neumtoxicity of AB fibrils has been demonstrated in numerous ce11

culture studies (Pike et ai., 1993; Simmons et al., 1993; Lorenzo and Yaakner, 1994).

Mutations in the APP gene have been implicated in some instances of familial AD (Goate

et ai., 199 1; MurreIi et ai., 1991), a finding corroborated by the observation that

transgenic mice overexpressing a mutant form of APP exhibit some of the

neuropathological features of AD (Gmes et al., 1995). Remarkably, patients with Down

syndrome bear three copies of the APP gene and invariably display neuropathological

changes consistent with the development of AD. SUnilarly, mutations in the presenilin

genes have been Linked strongly with f d a l l, and appear to result in increased leveIs

of AB42 (Suzuki et al., 1994; Borchelt, 1996; Citron et al., 1997). These observations

have led to the so-cailed amyloid hypothesis that the formation of senile A8 plaques is a

major contributor but not necessarily the sole causal element in AD.

However, there remaius a poor correlation between the existence of senile plaques

and the extent of dementia (Terry et al., 1991). This argues that the treatment of AD with

therapies designed to retard, diminish, or even abolish senile plaque formation may have

little or no effect on dementia, the most prevalent symptorn of AD. A plausible

explmation for this poor correlation can be found in the recent reports that monomeric

AB peptides can assemble into a solubIe oligomeric state that is intermediate to the AB

monomer and fibril foms (Kuo et al., 1996; Roher et al., 1996; Teller et al., 1996;

Lambert et ai., 1998). Some of these oligomeric AB intemediates are potent neurotoxllis

that kill neurons in cultured hippocampai brain slices fiom mice, even at concentrations

as low as 5 nanomolar (Lambert et ai.,1998). Notably, the neurotoxicity is neuron-

speciflc as astrocytes in these brain slices are not eected. Furthemore, prior to Uutiating

neuronal death, these oligomeric AB intemediates abolish long-term potentiation, a

potentially cntical component of memory and leaming (Lambert et al., 1998). These

observations suggest the following scenario. Io the typicd AD patient, a mixed

popdation of AB oügomers and mature senile pIaques rnay be present. Upon release nom

s e d e plaques, or as fieely soluble separate molecuiar entities, these AB oligomers may

then interact destnictively with neurons. Through this association process, disruption of

neuronal processes such as long-term potentiation may resdt, leadmg to cognitive

changes and memory dysfiinction and ultimately to progressive dementia.

In order to characterue the structure of these potentidy toxic AB oligomers, we

devised strategies for trapping these species for extended periods of tixne, which then

enabled us to apply fluorescence and circular dichroism spectroscopy, analytical

ultracenhifugation, and electron and atomic force microscopy techniques. These studies

should aid in identifjhg the rnechanism(s) by which these oligomers play a role in the

pathogenesis of Alzheimer disease, and aid in the developrnent of novel approaches to

inhibit or abolish the neurotoxicity of these AB oligomers.

Materials and Methods

Peptide synthesis

Peptides were prepared by solid-phase synthesis on a PerSeptive Biosystems 9050

Plus peptide synthesizer, as peptide-amides using PAL-PEG-PS resin (PerSeptive

B iosy stems). An active ester coupling procedure, emplo ying O-(7-azaberuotriazo 1- 1 -YI)-

1,1,3,3-tetramethyluronium hexafluorophosphate of 9-fluorenylmethoxycarbonyl amino

a d , was used. The peptides were cleaved fiom the resin with 8 1 : 13: 1 :5 trifluoroacetic

acid: thioanisole: m-cresol: ethanedithiol mixture. M e r incubation for 1 hour at 25 OC,

the resin was removed by filtration and bromotrimethylsilane was added to a fiai

concentration of 12.5% (vfv). After incubation for 5 hours at O OC, the peptides were

precipitated and washed in cold ether. Peptide purit. and identity were connmied by

electrospray mass spectrometry. The expected molecufar weight of AM0 is 4330 ghol,

and the observed molecular weight was 4332.1 glmol.

Fluorescent labeling

Pnor to fluorescent labeling, a glycine residue was added to the N-terminus of

AB40 to act as a flexible linker between the fluorophore and peptide. A tryptophan

residue, 5-((2-(t-B0C)-y-glutamylarninoethy1)amino)nap hthalene- -suKonic acid (BOC-

glutamyl-AEDANS) or 6 - ( N - ( 7 - n i t r o b e a z - 2 - 0 x a 4 , 3 - d i ~ t ~ o l - 4 - y l ) ~ acid

(NBD hexanoic acid, Molecular Probes) was then coupled to this extended AB40

sequence during peptide synthesis to create TrpAB40, AEDANS-A$40, or NBD-AB40

respectively. Peptide purity and identity were confimed by electrospray mass

spectrornetry .

Preparation of stock peptide solutions that are free of fibril seeds

To obtain preparations of AB40 that are free of fibds, aggregates, and fibril

seeds, we used a modification of the method proposed by Walsh et al. (1999). Ab40

peptide was dissolved in 6.0 M guanidine hydrochloride. 30 m M N h O H aqueous

solution (pH 9.98) to obtain a final peptide concentration of 200-500 FM. The peptide

solution (-7 ml) was appiied to a G-75 gel Nmtion column (1.5 x 50 cm) using 0.15%

(vlv) W H solution (-pH 10) as the nmning buffer. The basic Mining b a e r was used

because AB40 aggregation is reduced at pH values above 7 (Levine, 1993). Peptide-

containing fiactions eluting at positions expected for monomeric peptide were pooled.

The concentration of the pooled preparations were between 150-1 80 PM.

During the course of the experiments, stock peptide solutions were stored for

several weeks at 4 OC; no changes in the behavior of these peptides were observed over

this time. We have found that this method of chromatographic separation (at -pH IO) of

AB40 monomen fÎom aggregates is an efficient way of removing aggregation seeds

(uapublished results).

Measurement of peptide concentration

The concentration of stock solutions of peptide stored in 0.15% (vlv) NH4OH

solution (-pH 10) were measured by W absorbance and cdculating the concentration

using Beer's law. Since the peptide stock solutions do not contain aggregates, light

scattering amfacts do not occur. In addition, the high W transparency of the 0.15% (vh)

N&OH buffer ensures that interference fiom bufFer constituents does not occur. Peptide

concentrations for individual experiments were calculated by multiplying the

concentration of the stock solution by the appropriate dilution factor.

Determination of peptide concentration by aromatic W absorbance is highly

accurate (error 1%), and offers a considerable advantage over other peptide assays such

as the ninhydrin assay or quantitative amino acid analpis which produce errors of about

k 5% (Marqusee et al., 1989; Chakrabartty and Hew, 199 1; Chakrabartty et al., 19%).

The concentration of stock peptide solutions that were fiee of f i bd seeds was

determined by tyrosine absorbance at 275 MI ( ~ 1 3 9 0 CK'M*'; Edelhoch, 1967) for

AB40, by tryptophan absorbance at 281 nm (~=5690 cm-'~- ' ; Edelhoch, 1967) for

Trp-Ap40, by AEDANS absorbance at 338 nm (~=6500 c ~ ' M - ~ ; Hudson and Weber,

1973) for AEDANS-Ap40, and by NBD absorbance at 480 nm (~=23,000 cm%';

Chantier and Szent-Gyorgyi, 1978) for NBD-Ap40. Absorbante measurements were

made on a Perkin Elmer Lambda 3B spectrophotometer.

Electron microscopy

Negatively stained fibrils were prepared by floating charged pioloform, carbon-

coated gids on peptide solutions (0.05 mg/mL AP40, pH 2.0-8.0). These solutions were

incubated for 1 day and 7 days. Tu control pH, the peptide solutions were made using a

buffer of 1 m M borate, 1 mM citrate and 1 mM phosphate. M e r the grids were blotted

and air-dried, the samples were stained with 1% (wh) phosphotungstic acid.

Platindcarbon (Pt/C) shadowing was performed using the glycerot spray

method of Tyler and Branton (1980). Solutions containing 0.025 mg/mL AB40 were

brought to 30-50% (v/v) glycerol, sprayed ont0 fieshly cleaved mica and fieeze-dried.

The preparations were shadowed on an Edwards E12E4 coater with WC (15 to 2 0 4 at

an angle of 5 to 7O, foUowed by an additional 30 to 40 A of carbon-cooting applied at an

angle of 90°. Film deposition was measured with a Baizers QSE2Ol quartz crystal

monitor. The WC replicas were floated off in distilled water and placed onto 300-mesh

copper grids.

Representative electron microscopy images of the peptide assemblies were

acquked on a Hitachi H-7000 TEM operated with an accelerating voltage of 75 kV.

Fluorescence spectroscopy

Steady-state fluorescence was measured at room temperature using a Photon

Technology International QM-1 fluorescence spectrophotometer equipped with excitation

intensity correction and a magnetic stirrer.

For measurernents of NBD fluorescence, emission spectra fiom 500 nm to 600

am were colIected (k,=470 m, 0.1 to 1 sednm, bandpass4 nm for excitation and

emission). For FRET measurements, emission spectra from 300 ~i to 550 nm were

collected (Aa=28 1 nm, 0.1 to 1 sec/nm, bandpas== nm for excitation and emission).

Excitation spectra were also acquired for TpAP40 (Â,=200 to 320 am, he356 nm,

0.1 sec/m, bandpasr4 nm for excitation and emission) and AEDANS-AB40 (L-250

to 400 nm, am=481 nm, 0.1 secfnm, bandpas& nm for excitation and emission). A

quartz cuvette with 1 cm path Iength and 0.5 mL volume was used. Spectra of samples

containhg only uniabeled AB40 were used to correct for light sca t te~g.

Samples were prepared so that Trp-AB40 and AEDANS-AB40 would be present

in a 1:l molar ratio. Stock peptides were miued, then diluted to the desired final

concentrations and pH using buffer (25 mM borate, 25 mM citrate and 25 m M

phosphate). M e r dilution, the pH was again measured and readjusted, if necessary, then

ali samples were incubated 18-24 hours prior to spectral acquisition

Circular dichroism spectroscopy

Circular dichroism (CD) spectra were recorded on an Aviv Circular Dichroism

Spectrometer mode1 62DS at 2S°C. Spectra were obtained fiom 200 to 300 nm (0.5 cm

path length, 0.5 nm steps and 1 nm bmdwidth) on a sample containhg 20 pM AB40 in

b&er (7.5 mM borate, 7.5 mM citrate and 7.5 mM phosphate).

Analytical ultracentrifugation

Sedimentation experiments were performed at 20°C on a Beckman XLI

Analytical Ultracentnfuge ushg an ANSO-Ti rotor. The sedimentation equilibrium nuis

using six-channel charcoal-Epon celis were performed for 24 hours before equilibrium

absorbance measurernents were taken. Molecdar weight determinations involved global

d y s i s of three different sarnple concentrations centrifuged at at Ieast three different

speeds. Molecuiar weights were caiculated using Beckman XLI data analysis software in

which absorbance vernis radiai position data were fitted to the sedimentation equilibriurn

equation using nonlinear least-squares techniques (Johnson et ai., 1981). The partial

specific volume and density of the sample was calculated using the program SEDNTERP

(Laue et ai., 1992) fiom the amino acid sequence and bufTer composition, respectively.

Sedimentation velocity measurements using double sector charcoalBpon cells

equipped with sapphire windows were performed on samples (20 PM) centrifuged at

12 300 x g and concentration distributions were detemhed using absorbance optics.

Time derivative analysis (Stafford, 1992) of the sedimentation velocity data was

performed using the Beckman XLI data andysis software.

Atomic force microscopy

Al1 solution tapping atomic force microscopy images were acquired using a

combination contactltapping mode liquid ce11 fîtted to a Nanoscope IIIA MuMMode

scanning probe microscope (Digital Instruments). AU images were acquired using 120

pm oxide-sharpened silicon nitride V-shaped cantilevers (Type DNP-S, Digital

Instruments). Prior to use, the AFM tips were exposed to W irradiation to remove

adventitious orgaaic contamhants fkom the tip d a c e . The AFM images were acquired

using the E scanning head, which has a maximum lateral scan area of 14.6 pm x 14.6 p.

In siru imaging was perfomed on samples prepared by transferring -5 pl of the peptide

solution onto fieshly cleaved mica, seding the sample in the Liquid cell, and fiIling the

ceii with b a e r solution. Al1 imaging was performed at tip scan rates from 1.5-2.5 Hz,

using cantilever drive frequencies of -8.9 kHz. Al1 images were captured as 512 x 512

scans and were low-pass filtered. Feature size and volumes were calculated using the

Digital Instruments Nanoscope software (version 4.21), and shareware image analysis

program, NM-Image (version 1.62).

A screening test for identifying soluble AB40 oligomers

To identi@ the presence of soluble oligomers of AB40, we examined both the

solubility and structural properties of the peptide under conditions known to initiate

fibrillogenesis. The key to this paraIlel approach involves the following: detection of both

structure formation and AB40 insolubility suggests the presence of fibrils or aggregates;

detection of structure formation under conditions where the peptide rernains soluble

suggests the presence of either stnictured monomers or oligomen; detection of neither

structure formation nor insolubility suggests the presence of unstnictured monomers.

Structural changes in AB40 were detected by a fluorescence-based assay which

mooitored changes in the local environment (e.g. polarity, solvent accessibility) of

individual AB40 molecules. AB40 solubility was asseued by absorbante spectroscopy.

The experiments were conducted at a peptide concentration of 20 pM which is in the

middle of the critical concentration range (10 and 40 IiM) for fibrillogenesis of AB40

(Harper and Lansbury, 1997). At this concentration, amyloid fibrils will be marginally

stable, and thus the likeIihood ofpopulating soluble oligomers may be increased.

Since AM0 aggregation is hown to be initiated at pH values less than 6 (Fraser

et al., 1992; Levine, 1993; Wood et al., 1996), the pH-dependence of the fluorescence

and solubility assays were compared. If soluble AB40 oligomers are significantly

populated at a particultu pH, then the relative levels of fluorescence should be greater

than that of insolubility.

Absorbante measurements were taken at 320 nm and performed as a function of

pH on samples incubated for 18 hrs at 2S°C. The absorption measurements produced a

bell-shaped c w e with maximum absorption at pH 5 (Figure 2.1). As AB40 does not

contain any chromophores that absorb at 320 nm, any absorption at this wavelength is

tikely the resdt of hcreased turbidity caused by AB40 insolubility. Since there was no

detectable absorption at pH values lower than 3.5 or above 7, this suggests that insoluble

AfM0 species formed between pH 3.5 and 7.

Figure 2.1. Absorbance and fluorescence of AB40 at varying pH. Absorbance and fluorescence techniques were used to investigate the aggregationai response of AM0 (20 CiM) to pH. Absorbance (hb,=320 MI) and NBD-fluorescence (A-470 nm, &=500- 600 nm) values were nomalized to facilitate cornparison (Nonnalized Value = (&k~Amin)/(A-Amh)y where A is absorbance or fluorescence). Turbidity measured via absorbmce (open circles) was undetectable below pH 3.5, increased up to pH 5 and again decreased to an undetectable level by pH 7. The shape of the fluorescence curve (solid circles) was broader and did not reach a minimum level until a higher pH of 8. Also, the fluorescence was still sensitized at the lowest tested pH.

The fluorescence assay involved labehg AB40 with the environment-

sensitive fluorophore NBD. Since NBD fluorescence is strongly quenched by water, any

increase in NBD fluorescence is likely a consequence of sequestering of the NBD

fluorophore through either conformationai changes in the AB40 peptide, or AB40 ses

association. NBD fi uorescence rnemrnents perforrned as a function of pH on mixtures

of 20 pM AB40 and 0.1 FM NBD-Iabeled AB40 hcubated for 18 hrs (25°C) also

produced a bell-shaped curve, with a maximum in NBD fluorescence at pH 5 (Figure

2.1). The NBD fluorescence curve was, however, s ignif icdy different fiom the

absorbance c w e . Specificaily, the pH-transitions were broader in the fluorescence curve

than the absorbance curve, and unlike the absorbance curve, the fluorescence cuve did

not retum to basal levels at low pH.

In summary, the results of the screening test for soluble AB40 oligomers indicate

that at pH values below 3.5 or higher than 7, there is structure formation in AB40 with

maintenance of solubility. The results provide evidence for the formation of either

soluble oligomers or stnictured monorneric AB40. Experiments described later in this

paper distinguish between these possibiiities.

Previous studies have show, however, that AB40 can form fibnls at pH 7.4

which exhibit rninimd turbidity (Wood et al., 1996). To address this possibility we

compared the W absorbance spectnmi of 20 pM AB40 solutions with varying turbidity

both before and d e r microfige centrifugation (30 min. at 13 300 x g). Based on tyrosine

absorbance at 275 nm ( ~ 1 3 9 0 cm-'&; Edeihoch, 1967), 20 pM AM0 solution should

have an absorbance of 0.028 at 275 nm. After overnight incubation (25 OC) at pH 3 and 7,

the absurbances at 275 nm of 20 ph4 A840 prior to centrifugation were 0.037 and 0.034,

respectively; after centcifbgation they were 0.036 and 0.030, respectively (Figure 2.2). in

addition, the W absorbance spectra of these peptide solutions contained features typical

of a peptide containing Tyr and Phe; namely, Tyr absorption bands at 280 nm and 275

nm, and fine structure of Phe absorption between 250 nm and 270 m. Shce the

absorbantes are close to the expected values and do not change much upon

centrifugation, it appears that rnajority of the AB40 molecules at pH 3 and 7 are soluble.

At pH 5, on the other hand, the fine structure of the aromatic absorption bands was

missing and the absorbance at 275 nm was 0.30 pior to centrifugation, which is ten-fold

higher than that expected for tyrosine absorbance at this concentration. This is indicative

of the presence of large aggregates that scatter Light. After centrifugation, absorbance was

almost zero (Figure 2.2) indicating the sedimentation of insoluble foms of AMO.

240 250 260 270 280 290 300 310 320 Wavelength (nm)

Figure 2.2. Solubility of AB40 at pH 3,5, and 7. AB40 samples (20 pM) were infubated for 18 h at 25OC at pH 3, 5, and 7 (buffer: 7.5 mM borate, 7.5 mM citrate and 7.5 mM phosphate), and the UV absorbance spectnim was measured before and after centrifugation in a microfuge (30 min. at 13 300 x g). The UV absorbance specûa before and ;ifter cenfigation are depicted as soiid and dashed h e s , respectively.

Taken together the above data do not exclude the possibility that a s m d fiaction

of AB40 molecules fom insoluble structures at pH 3 and 7. However, at these pH values

a majonty of the AB40 molecdes adopt structural states that reduce the exposure of the

NBD fluorophore and enable the AM0 rnoieties to remain in a soluble state. These f o m

could be either structured monomeric AB40 or structured soluble oiigorners.

Association reactions of AM0 rnonitored by environment-sensitive fluorescent

probes and fiuorescence resonance energy transfer (FRET).

Two different experiments were perfomed to distinguish whether the soluble

foms populated at pH 7 and pH 3 were monomenc or oligomenc Ap40. Fint, the efTect

of changes in AB40 concentration on NBD-fluorescence was examined at pH 7.0. The

concentration of NBD-labeled AB40 was held constant at 0.1 pM and the amount of

uniabeled AB40 was varied fiom O to 48 FM. The non-liaear relationship between the

NBD fluorescence and unlabeled AB40 concentration (Figure 2.3A) argues that as AB40

level is raised, the NBD reporter groups are increasingly more sequestered, and suggests

the formation of higher-order oligomeric AB40 species that preferentiaily sequester the

NBD group.

Figure 2.3. Concentration dependence of AB40 association reactions. Panel A: AB40 samptes of concentrations varying fiom O to 48 @A (pH 7) were incubated with 0.1 pM NBD-labeled AB40. Fluorescence increased non-Iinearly and saturated with increasing peptide concentration.

FRET between the energy donor TrpAM0 and the energy acceptor AEDANS- AB40 was measured and expressed as the acceptor/donor ratio to report on the degree of AB40 association. Panel B: At pH 7, the acceptorldonor ratio increased with concentration up to a maximum value of 1.6. Panel C: At pH 3 (solid circies) and pH 5 (open ckcles), AM0 association a h increased with concentration and reached maximum values of 17 and 18 at 50 pM total peptide concentration. These were more than ten-fold higher than the ratio for the corresponding AB40 concentration at pH 7.

Further evidence of soluble oligomers was obtained through fluorescence

resonance energy transfer (FRET) rneasurements. Previousfy we have reported how

FRET rneasurements employing Trp as the energy donor and AEDANS as the energy

acceptor can be used to monitor the association of AB peptides without interferhg with

the fibrillogenesis process (Huang et al., 1997). For FRET measurernents, two

N-terminus labeled versions of AB40 were synthesized: one with a Trp residue (denoted

Trp-AWO), and a second (denoted AEDANS-Ap40) with an AEDANS group (attached

via the side chah of glutamic acid).

The fluorescence emission and excitation spectra of mixed Trp-AM0 / AEDANS-

AB40 preparations at pH 5 clearly illustrate a significant quenching of the Trp

fluorophore emission signal and a concomitant enhancement of the AEDANS emission,

relative to the Trp-AB40, and AEDANS-AB40 alone (Figure 2.4A). Since donor

quenching of Trp and sensitized AEDANS emission are characteristic of an energy

transfer process, and the intensity of the fluorescence energy transfer between Trp and

AEDANS would increase with increasing proximity of the two fluorophores, these renilts

support the conclusion that oligomenzation or aggregation of the mixed peptide

preparation is occurring. Excitation spectra of the acceptor AEDANS in mixed

preparations were collected to con£Ïrm that the observed changes in fluorescence were the

result of fluorescence energy transfer and not changes in the intrinsic fluorescent

characteristics of AEDANS moiety induced by conformational changes in the peptide

(Figure 2-43).

We can obtah a quantitative measure of the degree of association by integrating

the AEDANS signal fiom 425-525 nm, and expressing a s value as a ratio relative to the

integrated Trp signal fiom 340-375 ML Furthemore, sùice the acceptor/donor ratio is

related to the efficiency of energy tramfer, the acceptoddonor ratio is inversely

proportional to the separation distance between donor and acceptor (Huang et al., 1997).

The acceptoddonor ratios of AB40 preparations at different total peptide concentration

were monitored at pH 3,5, and 7 (Figure 2.3B, C). The acceptoddonor ratios of ai l of the

sarnples increased nonlinearly with the concentration of AB40 present, and displayed a

saturation effect (Figure 2.3B, C). The increases in acceptor/donor ratios provide proof

that the AB40 molecules are associathg into oligomeric structures at pH 3,5, and 7. An

interesting Merence between the association reactions is that the acceptor/donor ratios at

pH 3 and 5 are similar to each other and approximately ten-fold higher than the ratios at

pH 7. Since these ratios are inversely proportional to the separation distance of donor and

acceptor fluorophores, these results suggest that the average fluorophore separation

distance of AB40 oligomers at pH 7 are fat greater than the separation distance of AB40

oligomerdaggregates at pH 3 and 5.

These results demonstrate that the AB40 monomer is capable of forming an

insoluble aggregate at pH 5, and of seKassociathg into soluble oligomenc species at pH

3 and 7. Moreover, our fluorescence energy W e r data argue that the soluble pH 3 and

7 species are stnictuxdly distinct.

200 300 400 500 600 Wavelength (nm)

Figure 2.4. Fluorescence spectra of Trp-AB40 and AEDANS-AM0 at pH 5. Panel A, &=281 MI): the Trp fluorescence in the mixed labeled peptide sample (dashed h e , 3 pM Trp-AWO and 3 pM AEDANS-AB40) was quenched relative to the fluorescence when only Trp-labeled peptide was present (dotted h e , 3pM Trp-AM0 and 3 p M AVO). AEDANS fluorescence in the mixed labeled peptide sample was sensitized relative to the sample contaùiing only AEDANS labeled peptide (solid h e , 3 pM AEDANS-AB40 and 3 pM AB40). Donor quenching and acceptor sensitization are indicative features of energy traosfer. Panel B: Excitation spectra of donor ody (dotted iine, 3 phd TrpAp0 and 3 p M AP0, L = 3 5 6 nm) and acceptor only (solid he, 3 p M AEDANS-AM0 and 3 pM Aw0, L m 4 8 1 nm) showed no changes in either shape or maximum excitation wavelength.

Molecular weig hts of soluble AM0 oligomers

The molecular weights of oligomeric AB40 species formed at pH 3,5, and 7 were

deterrnined through a global anaiysis of analytical ultracentrifugation data taken at three

dBerent centrifugation speeds on samples of varying concentration (5, 10, and 20 pM)

(Figure 2.5).

The number average molecular weight of AB40 at pH 3.0 calculated from

sedimentation equilibrium data was 938 000 Da (95% confidence interval:

880 000-994 000 Da) (Figure 2.5B). The sedimentation equilibrium experiments were

repeated on the same samples after 4 weeks incubation at 4 OC and no significant change

in molecular weight was observed (data not shown). Plots of h(absorbance) vs. radius2

were nonlinear and showed upward curvature (data not shown), indicating that the

presence of multiple species with different molecular weight. To investigate the

molecular mass heterogeneity of this AB40 preparation, sedimentation velocity

experiments were conducted and the data anaiyzed ushg the t h e denvative (dcldt)

technique (Stafford, 1992). The velocity data indicate that at pH 3.0, AB40 is not

monodisperse but rather exists as a unimodal distribution of aggregate sizes and

molecuiar weights (Figure 2.6). We wiiI refer to this population of AB40 particles, which

has an average sedimentation coefficient of 15.6 S, as 1 megadalton (MDa) AB40

particles.

g o.,

1 <.«/

0.04

Figure 2.5. Ultracentrifugation analysis of AM0 aggregates. Panel A: Data f?om sedimentation equili%nm d y s i s of AM0 at pH 7 (5, 10 and 20 PM; 4 660,s 900, 7 300,57 080,65 520, 74 550 x g) indicated that the number average molecdar weight was 12 100 Da (95% confidence interval4 1 17542 990 Da). Panel B: For AM0 at pH 3 (5,IO and 20 pM; 1 8992 370,2 710 x g) the number average molecular weight was 938 000 Da (confidence mtervai = 880 000-994 000).

Figure 2.6. Sedimentation coefficient distribution fiom sedirnentation veIocity experiments at pH 3. The average sedimentath coefficient of the particles in solution was 15.6 S and the disti'liution was unimodal.

In the case of the pH 5 preparation, no soluble species were detected by analytical

ultnicentrifbgation.

The number average moleculat weight of AB40 at pH 7.0 was 12 100 Da (95%

confidence interval = 11 175-12 990 Da). which is 2.8-fold greater than the weight of

monomenc AB40. This indicates that AB40 associates into oligomers that are, on the

average, larger than d h e a at pH 7. The sedimentation equilibrium experiments were

repeated on the same samples &et 4 weeks incubation at 4 O C and no significant change

in molecular weight was observed (data not show). Global fitting of the pH 7 data sets

to multi-state association models was perfomed to elucidate a plausible oligomerization

mechanism. The goodness of the fits were assessed by the distribution pattern of the

residtiais and the magnitudes of the variances. The pH 7 data sets could be fitted to

monomer-dirner (variance = 1 .O3 x IO-'), monomer-trimer (variance = 8.00 x 104), and

monomer-tetramer (variance = 7.91 x W6) models. While dl three of the two-state

models produced residuals of similar magnitudes and random distribution patterns, the

estimates of variance indicate the monomer-tetramer model gives the best fit. The pH 7

data could dso be fitted by the three state monomer-dher-tetramer model (variance =

7.70 x 103. However, performance of the F-test indicated the monomeraimer-tetramer

mode1 was not a better fit than the monomer-tetramer model, even at the 10% level of

significance. In smary, AB40 at pH 7 forms an equilibrium mixture of monomers and

tetramers, however, the presence of dimers cannot be ruied out

Efectron microscopy of AB40 aggregates and soluble oligomers

The AM0 preparations at pH 3.5, and 7 were examined by electron microscopy,

employiag both platinun-carbon (WC) shadowing and negative stainiag techniques.

Long slender nbrils (70-80 A in diameter) that were several microns long were reveded

by both negative stain d y s i s and WC shadowing of A840 sarnples prepared at pH 3

(Figure 2.7A, B). WC shadowing additionally reveaied the presence of abundant s m d

sphericd structures.

PIatinum/carbon shadowing and negative stain electron microscopy of AM0

samples prepared at pH 5 (Figure 2.7C, D) and pH 7 (Figure 2.E, F) revealed short

fibds with Uregular surfaces; the fibrils were approximately 10 nm in diameter and 100-

400 nm in length. The small spherical particles found in the pH 3 preparation were

noticeably absent in the pH 5 and 7 samples. While the firils formed at pH 3 Look more

like typical amyloid fibrils, those formed at pH 5 and 7 were shorter and had a more

irregular d a c e than typicai amyloid fibrils. The morphology and dimensions of these

structures may perhaps represent protofibrils, short flexible fibrîls that are 4-10 nm in

diameter and up to 200 nm in length (Harper et al., 1999; Walsh et al., 1999), rather than

mature amyloid tibrils.

Figure 2.7. Electron micrographs of plathum-carbon shadowed or negatively stained AB40 preparations (0.05 m m ) . WC shadowing (panel A) and negative staining @anel B) of AB40 prepared at pH 3 showed long slender fibrils. At pH 5 and pH 7, the morphology of the fibrils changed and appeared short and thick with a rough d a c e (panels C and E: WC shadowing, pH 5 and 7, respectiveIy; paneIs D and F: negative staining, pH 5 and 7, respectively). The pH 7 preparations contained fewer of these short thick fibrils compared to pH 5. W C shadowing at pH 3 revealed the presence of s m d sphencal particles which were not visualized by negative staining at the same pH. These spherical structures were dso noticeably absent at the other pH values. The scale bar represents 100 m.

Whiie EM is a powemil techniqye for investigating ultrastructure, it is ciifficuit to

quantify the re1ative abundance of different structures by EM. Comparison of the EM

results with solubility results cm help in this regard. Solubility analysis (Figure 2.2)

indicates both pH 3 and pH 7 preparations contain predominantly soluble foms of A$40,

while the pH 5 preparation is comprised of insoluble foms of AMO. Thus, the thùi fibrils

observed in the electron micrographs of pH 3 sample and the short thick atypical fibrils

(or protofibrils) found in the pH 7 sample must be composed of a relatively small fraction

of the total amount of AB40 present Conversely, the majority of the AB40 molecules in

the pH 5 preparation appear to have aggregated into short thick atypical fibrils (or

protofibrils). With regard to the mal1 spherical particles observed in the pH 3 sample,

these particles may represent the 1 megadalton particles detected by analytical

ultracentrifugation (Figure 2.5) and atomic force rnicroscopy was used to further

investigate this possibility.

Atomic force microscopy of AM0 fibrils and 1 rnegadalton AM0 particles

In situ AFM was performed to quantKy the dimensions of AB40 stnictures formed

in the pH 3 solution. Consistent with the results obtained by WC shadow electron

microscopy (Figure 2.7A), long slender fibrils and an abundance of small sphencal

particles were detected by AFM on the pH 3 AB40 sample (Figure 2.8). The size

homogeneity of the spherical particles was examined by meanrring the d i s t n i o n s of

particle voIumes found in the sarnple (Figure 20). While the shape and dimensions of the

AFM tip can contrziute to an overestimation of lateral f e a m size, the observed vertical

dimensions generally are Unmune to such eEects. The mean diameter of the spherical

particles was 15 nm (range = 8-18 nm). The volumes of the spherical partîcles had a

unimodal' distribution with a mode of 1 750 nm3 (range = 250-3 000 nm3). This

distniution was simiiar in shape to the distniution of sedimentation coefficients (Figure

2.6) obtained fiom time derivative analysis of sedimentation velocity data on the pH 3

sample.

The mode of the volume measurement of the particles was used to estimate the

molecular weighfusing two methods that are likely to produce an underesthate and an

overestimate of the true molecular weight. in the first method, it was assumed that the

density of the particle is equal to the reciprocal of the partial specific volume of AB40

(0.7362 ~ m ~ / ~ ) , and the molecular weight will be the product of the volume, density, and

Avogadro's number. This caiculation estimated the mean molecular weight of the

spherical particles to be 1.4 MDa. In the second method, the density of the particle was

assumed to be sirnilar to that of protein crynals; the molecular weight can be calculated

by dividing the mean particle density by V,, the ratio of the volume of the asymmetric

unit of a protein crystal and the molecular weight of the protein (Matthews, 1968). Values

of V,,, range between 1.75 A3/Tla and 3.5 A3Da for various protein crystais (Matthews,

1968). Based on these values of V,, the molecular weight range of the spherical particles

was cdculated to be between 0.5 MDa and 1 MDa. These molecuiar weight estimates are

in close agreement with the molecdar weight of 0.94 MDa determined by sedirnentatioo

equilibrium (Figure 2.5B); therefore, the srnd sphezical particles observed by AFM and

electron microscopy are identicai to the 1 MDa AB40 particles detected by andyticd

ultracentcifbgation.

Figure 2.8. Atomic force microscopy of Ap40at pH 3. AFM images showed the existence of the malI sphericd particles dong with long slender nbrils at pH 3.

O 5000 1 OOOg 15000 Volume (nm )

Figure 2.9. Volume distribution of AB40 particles. From the AFM images, the volume distribution of the spherical particles was unimodal with a mode of 1750 nm3. Two methods of molecdar weight caicdation gave results of 1.4 MDa and 0.5-1 MDa. These sizes a p e with the enimate of 0.94 MDa nom sedimentation equilibrium. Based upon these r d t s , the s m d spherical particles were named 1 MDa AM0 particles.

Secondary structures of ABQO at pH 3.5. and 7

At pH 7 and at an AB40 concentration of 20 pn, AB40 forms predorninantiy

tetramen and possibly dimers, as seen by analyticd ultracentrifugation. The

conesponding CD spectnnn of these oligomers (Figure 2.10) shows a strong negative

band below 200 om with a shoulder around 220 m. This spectnun indicates the presence

of little, if any, a-heüx or &structure; it may indicate the presence of irregular secondary

structure.

At pH 5, the peptide foms protofibrilsffibrils and the CD spectnun shows a

negative band at 225 nm and a positive band at 209 nm (Figure 2.10). Both of these

bands are atypical. However, evaluation of the secondary structure content from the CD

data at pH 5 is not possible because the sample is comprised exclusiveiy of insoluble

material. This matera scatters the incident light in a wavelength-dependent manner and

the scattered light is polarized (Tanford, 1961). Both effects alter the CD spectrum and

prevent evaluation of the secondary structure content

At pH 3, where the peptide foms predominantly 1 MDa particles, the CD

spectnmi is typical of fktmcture, with a single minimum at 21 8 nm and a maximum at

202 nm (Figure 2.10).

Figure 2.10. Circular dichroism spectroscopy of 20 jM AB40 at pH 3, 5 and 7. Upper panel shows the CD spectnun of AM0 at pH 3. At this pH, the 1 MDa particles were the predominant structwe. Characteristic of Fstructure, the spectnmi showed a minimum at 21 8 nm and a maximum at 202 m. At pH 5, where atypicai fibrils (or protonbrils) were abundant, the spectnmi showed atypical bands at 225 nm and 209 nm (middle panel). Under conditions which favoured the formation of dimerdtetramers (pH 7, 20 FM AfMO), the CD spectrum had a negative band below 200 mn with a shoulder around 220 MI suggesting iittle or no a-heIk or pstructure.

Discussion

The discovery that soluble oligomeric foms of AB possess neurotoxic activity

and can inhibit neurobiological processes such as long term potentiation (Lambert et al.,

1998) has impiicated these species in the molecdar pathogenesis of Alzheimer disease.

We have found conditions in which two different types of soluble oligomeric foms of

AB40 can be formed and stabilized for extended time periods. Using these conditions we

have investigated structural properties of these oligomers. A cornparison of these results

with other published data is discussed below.

Amy loid fibril formation under conditions where soluble AB40 oligomers are

stable

We fmd that soluble AB40 oligomers are significantly populated at peptide

concentrations S 20 pM and at pH 3 and 7. EM examination and solubility studies of

these samples show that a smali hction of AM0 molecules present in the samples form

short thick atypical fibrils at pH 7 and typical amyloid fibriIs at pH 3. While the fibrils

present in the pH 3 sample resemble mature amyloid fibrils, those in the pH 7 sample

more closely resemble protofibrils than mature amyloid fibrils. The reduced Ievel of

fTbrillogenesis and the absence of mature amyloid nbrils in the pH 7 sample may seem

surprisin& given that many laboratories have observed mature amyloid fibril formation of

AM0 at pH 7-8 (examples: Wood et al., 1996; Waish et al., 1999). One ükely reason for

this apparent inconsistency is the Iow peptide concentration (120 pM) employed in our

study. The critical concentration for nbril formation at neutral pH is between 10 and

40 pM (Harper and Lansbrrry, 1997), and most reports of mature amyIoid fibril formation

at neutral pH use signincantly higher concentrations. For example, Wood et al. (1996)

report the efficient growth of mature amyloid fibrils at 230 PM. Since the peptide

concentrations examined here are just at or below the critical concentration for fibril

formation, efficient growth of mature amyloid fibrils is not observed at neutral pH. The

presence of mature amyloid fibrils at pH 3 suggests that the critical concentration is lower

andtor the rate of fibnllogenesis is higher at pH 3 compared to pH 7. Rapid formation of

long thui fibrils at pH 1.0 bas been recently reported (Harper et al., 1999).

Atypical fibril formation at pH 5

Our observation that AB40 forms short atypical fibrils that may be similar to

protofibrils at 20 pM peptide concentration, pH 5, is in line with previously reported

observations. Using negative stain EM, Wood et al. (1996) have reported that AB40

(230 PM) at pH 5.8 fonns large amorphous aggregates rather than amyloid fibrils. A

recent AFM study of AB40 at pH 4.5 and 5.8 has s h o w that the peptide foms

protofibds exclusively at these pH values (Harper et al., 1999).

Small AM0 oligomers

There has been extensive published work characterizhg the size of soluble AB

oligomers. Size exclusion chromatography studies have niggested that A$ peptides at

neutrai pH form dimers predominantiy (Hilbich et al., 1992; Soreghan et al., 1994;

Garzon-Rodriguez et ai., 1997), aithough Afhimers and tetraruers also have been

detected (Barrow et al., 1992; Roher et ai., 1996). Results fiom FRET and fluorescence

decay experiments support the existence of a stable dimer form of AB40 (Garzon-

Rodriguez et al., 1997). The presence of low mo1ecula.r weight oligomers of AB40 has

aiso been investigated by Light scattering analysis; however, the analysis could not

distinguish between monorner and dimer (Walsh et al., 1999). These techniques,

however, have certain shortcomings that may introduce enors in the molecular weight

calculations. For example, analyses of chromatographie data are performed assuming that

the observed species are spherical in shape and aay interaction between the peptide and

the gel filtration column will introduce systematic errors in the molecular weight

estimates: similady, any exchange between various oligomeric forms which occurs on the

tirne scale of the experiment will also introduce errors in molecdar weight estimates. The

fluorescence experiments are biased by those fluorophore pairs that are closest in space

and thus may miss solubk species in which the fluorophore pairs are spatially distant.

Analytical ultracentrifugation is not prone to these types of problems. Our

sediientation equilibrium data of AB40 indicate that tetramen and possibly diiers are

present at neutrai pH. Interestingly, a symmetrical tetrameric structure would also be

compatible with the FRET and fluorescence decay data of Glabe and coworkers (Garzon-

Rodriguez et al., 1997). The presence of these multiple equiiibria will compiicate NMR

d y s i s of AB40 oligomers at neutral pH. It should be noted that sedimentation anaiysis

on AB40 has been performed in the past (Snyder et ai., 1994; Seilheimer et al., 1997);

however, those studies focused on the formation of large aggregates and full

chanicterization of the oiigomerization state of s m d oligomers was not attempted.

Our CD spectroscopy redts show that AM0 dimers/tetramers do not possess a-

helical or Bsheet structure. The observed lack of regular secondary structure in the mal1

AB40 oligomers is consistent with studies on the temperature dependence of ABfibril

formation, which suggested that the bindmg of monomeric AB to nascent nbds requirrd

large conformational changes and correspondingiy hi& activation energies (Kusumoto et

al., 1998). Structure prediction s~idies of AB42 have assumed that the dimer possesses

structure (Chaney et al., 1998) However, given our CD spectroscopy results on small

AB40 oligomers, it is conceivable the dimeric AB42 is dso devoid of a- and &structure.

If iodeed amyloid dbds are the toxic agent responsible for neurodegeneration,

then a plausible therapeutic tactic wouid be to design srna11 molecules that stabilize these

nonfibrillogenic dimersltetramers. Alternatively, the small Ab40 oligomen may

themselves represent the toxic species, a intriguing possibility given the recent report that

small AB oligornea stabilized by 80% formic acid are indirectly toxic to cultured primary

neurons (Roher et al., 1996).

One megadalton particles of AB40

At low pH (-3.0), AM0 f o m a nanow distribution of spherical particles with an

average m a s of approximately 1 MDa. The mass and morphology of these particles do

not change over the period of several weeks. Cornparison with pubüshed data suggests

that spherîcal particles similar to the observed 1 MDa particles are formed transiently at

neutrd pH during the fibrillogenesis process. Kinetic in situ AFM studies have

demonstntted that fibrils appear after the formation of 1 000 to 10 000 nm3 globular

aggregates (Kowalewski and HoItmian, 1999); the average volume of the 1 MDa

particles (1 750 m3, Figtue 2.9) is within this range. Another kioetic study of AB40

nbdlogenesis, employing AFM, has demonstrated that mail spherical particles (4 nm in

diameter) coexist with protofibrils during ealy phases of nbriilogenesis (Harper et al.,

1999); the sUe of these particles is within the size range of the I MDa particles. Fuially,

sedimentation andysis of relatively high concentrations of both AB40 and AB42 at

neutral pH has revealed the formation of soluble aggregates with molecdar weights

(0.85-2 MDa) (Snyder et al., 1994) that are close to the average mass of the 1 MDa

particles. These soluble aggregates appear pnor to formation of amyloid fibrils (Snyder et

ai., 1994).

The 1 MDa particles also have certain similarities with soluble oiigomenc f o m

of Aw2, termed AB-derived dinusible ligands (ADDLs), that have neurotoxic properties

(Lambert et al., 1998). ADDLs are spherical particles with diameters between 4.8-5.7 nrn.

Preparations containing ADDLs possess neurotoxic activity against cultured neurons and

can also inhibit long term potentiation in cultured brain slices (Lambert et al., 1998). The

1 MDa particles of AB40 (8-18 nm in diameter) are slightiy larger than the ADDLs,

suggesting that the additional two C-terminal residues present in AB42 reduces the size of

the s p h e b l particle forrned. Another major difference between the 1 MDa particle and

ADDLs is that gel filtration chromatography under basic conditions was used to obtain

aggregate-free starting material for the production of 1 MDa particles, while

dimethylmlfoxide treatment was used for the same purpose during production of ADDLs.

TepIow and coworkers (Walsh et al., 1999) have stressed the importance of using starting

peptide preparations that are fkee of f ibd seeds, and we have opted to use their

chrornatographic method for preparation of "seed-fiee" starting matenal.

Conclusions

We have demonstrated that soluble digomers of AB40 cm be trapped and

stabfied, allowing them to be studied by high resolution techniques that are inherentiy

slow. This approach may have generai applicability to the study of other fibrillogenic

peptides, such as islet amy loid polypeptide. Through a combination of absorbance,

fluorescence, and CD spectroscopy, analytical ultracentrifiigation, and electron and

atomic force microscopies, we have characterized two oligorneiic forms of AB40. The

nrst type of oligomer, AB40 tetramers and possibIy dimers, possess irregular secondary

structure and are clearly distinct h m those AB40 molecules that make up the fibril or

protofibril. The second oligomer, the 1 MDa particles, appear to be similar to kinetic

intermediates of the amyloid fibril that have been detected previously. It remains to be

seen whether these oligomers are toxic to neumns and studies are under way to estabiish

the stability of the intermediates under ce11 culture conditions. Furthemore, the

possibility rernains bat these oligomen may be intemediates on the fibril formation

pathway, and may have relevance under physiologie pH. Our detailed biophysical studies

of these oligomers have identified a new target for inhibitor strategies, and the

biophysical procedures we developed could serve as assays for identification and

evduation of potential inhibitors.

Barrow, C.J., Yasuda, A., Kenny, P.T.M., and Zagorski, M.G (1992). Solution

conformations and aggregational properties of synthetic p-peptides of Alzheimer's

disease. Analysis of cucular dichroism spectm. J. Mol. Biol. 225, 1075-1093.

Borchelt, D.R. (1 996) Familial Alzheimer's disease-linked presenili 1 variants elevate

Abetal-42/1-40 ratio in vitro and in vivo. Neuron 17,10054 0 13.

Brion, J.P. (1992) The pathology of the neuronal cytoskeleton in Alzheimer's disease.

Biochim. Biophys. Acta 1 160, 134-142.

Chaney, M.O., Webster, S.D., Kuo, Y.M., and Roher, A.E. (1998) Molecular modeling of

the Abetal-42 peptide fiom Alzheimer's disease. Protein Engineering 11,76 1-767.

Chantier, PD., and Szent-Gyorgyi, A.G. (1978) Spectroscopie studies on invertebrate

myosiils aad üght chains. Biochemistry 17,5440 - 5447 .

Chakrabartty, A., and Hew, C.L.(1991) The efEect of enhanced alpha-helicity on the

activity of a winter flounder antifkeze polypeptide. Eut. J. Biochem. 202,1057 - 1063.

Chakrabartty, A., Kortemme, T., Padmanabhan, S., and Baldwin, R.L. (1992)

Contribution of aromatic sidechains to far ultraviolet circular dichroism of helical

peptides and its effect on meamernent of helix propensities. Biochemistry 32, 5560-

5565.

Citron, M., Westaway. D., Xia, W., Carlson, G., Diehl, T., Levesque, G., Johnson-Wood,

K., Lee, M., Seubert, P., Davis, A., Kholodenko, D. Motter, R., Sherrington, R., Perry,

B., Yao, H., Strome, R., Lieberburg, L, Rommens, J., Kim, S., Schenk, D., Fraser, P., St

George Hyslop, P., and Selkoe, DJ. (1997) Mutant presenilins of Alzheimer's disease

increase production of 42-residue amyloid beta-protein in both transfected cells and

transgenic mice. Nature Medicine. 3,67-72.

Edelhoch, H, (1967). Spectroscopie determination of tryptophan and tyrosine in proteins.

Biochemistry 6,1948-1954.

Fraser, PE., Nguyen, J.T., Inouye, H., Surewicz, WX, Selkoe, D.J., Podlisny, MB., and

Kirscher, DA. (1992) Fibril formation by primate, rodent, and Dutch-hemonahagic

analogues of Alzheimer amyloid bprotein. Biochemistry 3 1,107 16- 1 0723.

Games, D., Adams, D., Alessandrini, R., Barbour, R., Berthelette, P., et al. (1995).

Alzheber-type neuropathology in transgenic mice overexpressing V717F beta-amyloid

precursor protein. Nature 373,523-527.

Ganon-Rodreguez, W., Sepulveda-Becerra, M., Milton, S., and Glabe, C.G. (1997)

Soluble amyloid Abeta(1-40) exists as a stable dimer at low concentrations. J. Biol.

Chem. 272,2 1037-21 044.

Goate, A., Chartier-Harh, M. C., MuIIan, M., Brown, I., Crawford, F., et al. (1991).

Segregation of a missense mutation in the amyloid precunor protein gene with familial

Alzheimer's disease. Nature 349,704706.

Harper, J.D., and Lansbury, P.T. (1997) Models of amyloid seeding in Alzheimer's

disease and scrapie: mechanistic tniths and physiological consequences of the t h e -

dependent soIubiIity of amyloid prote&. AM. Rev. Biochem. 66,385-407

Harper, J.D., Wong, S.S., Lieber, CM., and Lansbury, P.T. (1 999) Assembly of Abeta

amyloid protofibrils: an in vitro mode1 for a possible eady event in Alzheimer's disease.

Biochemistry 38,8972 - 8980.

Hilbich, C., Ester-Woke, B., Reed, J., Masters, CL., and Beyreuther, K. (1992).

Substitutions of hydrophobie amino acids reduce the amyloidogenicity of Alzheimer's

disease bA4 peptides. J. Mol. Biol. 228,460473-

Huang,T.H.J., Fraser, P.E., and Chakrabartty, A. (1997) Fibrillogenesis of Alzheimer

beta-amyloid peptides studied by fluorescence energy transfer. J. Mol. Biol. 269, 214-

224.

Hudson, E.N. and Weber, G. (1973). Synthesis and characterization of two fluorescent

suifhydryl reagents. Biochemistry 12,4 154-416 1.

Johnson, M.L., Corriea, J.J., Yphantis, DA., and Halvorson, H.R. (1981) Analysis of

&ta fiom the d y t i c a l ultracentrifuge by nodhear Ieast squares techniques. Biophys. J.

36,575-588.

Kowalewski, TZ., and HoItPnan, D. M. (1999) in situ atomic force microscopy study of

Alzheimer's beta-amyIoid peptide on different substrates: new insights into mechanism of

beta-sheet formation. h c . Natl. Acad. Sci. USA 96,3688-3693.

Kuo, Y.M., Emrnerling, M.R-, Vigo-Pelfiey, C., Kasunic, T.C., Kirkpatrick, J.B.,

Murdoch, G.H., Ball, M.J., and Rober, A.E. (1996) Water-soluble Abeta (N-40, N-42)

oligomers in normal and Alzheimer disease brains. J. Biol. Chem. 27 1,4077-408 1.

Kusumoto Y. Lomakin A. Teplow DB. Benedek GB. (1998) Temperature dependence

of amyIoid beta-protein fibrillization. Proc. Natl. Acad. Sci. USA 95, 12277.12282.

Lambert, M.P., Barlow, A.K., Chromy, B.A., Edwards, C., Freed, R., Liosatos, M.,

Morgan, T.E., Rozovsky, L, Trommer, B., Viola, K.L., Wais, P., Zhang, C., Finch, C.E.,

Krafft, GA., and Klein, W-L. (1998) Diffusible nonfibdar ligands derived fiom Abetal-

42 are potent centrai nervous system neurotoxins. Proc. Natl. Acad. Sci. USA 95,6448-

6453.

Laue, T.M., Shah, B.D., Ridgeway, T.M., and Pelletier, S.L., (1992) Andytical

UItracentrifugation in Biochemistry and Polymer Science (Harding, S., and Rowe, A.

Eds), pp. 90-125, Royal Society of Chemistry, Cambridge UX.

Levine, R (1993) Thioflavine T interaction with synthetic Alzheimer's disease b-amyloid

peptides: Detection of amyloid aggregation in solution. Protein Science 2,40é4 10.

Lorenzo, A. and Y-er, B. (1994) Betaamyloid neurotoxicity requires £ibn1 formation

and is inhibited by congo red. Proc. Natl. Acad. Sci. USA 9 1, 12243012247.

Matthews, B.W. (1 968) Solvent content of protein crystals. J. Mol. Biol. 33,49 1-497.

Marqusee, S., Robbias, V.H., and Baldwin, EL. (1989). Unusually stable helk formation

in short alanine-based peptides. Roc. Nad. Acad. Sci. USA 86,5286 - 5290.

MuneU, J., Farlow, M., Ghetti, B., and Benson, MD. (1991). A mutation in the amyloid

precursor protein associated with hereditary Alzheimer's disease. Science 254,97099.

Pike C.J., Burdick, D., Waiencewicz, A.J., GIabe, C.G., and Cotman, C.W. (1993).

Neurodegeneration induced by beta-amyloid peptides in vitro: the role of peptide

assembly state. I. Neurosci. 13,16764687,

Roher, A.E., Chaney, M.O., Kuo, Y.M.,Webster, S.D., Stine, W.B., Haverkamp, L.,

Woods, A.S., Cotter, RJ., Tuohy, LM., Kra& GA., Bomel, B.S., and Emmerling, M.,

R. (1996) Morphology and Toxicity of Abeta(l-42) dimer derived fiom neuritic and

vascular amy loid deposits of Alzheimer's disease. I. Biol. Chem. 27 1,2063 1 -2063 5.

Seilheimer, B., Bohnnann, B., Bondolfi, L., Muller, F., Stuber, F., and Dobeli, H. (1997)

The toxicity of the Alzheimer's beta-amyloid peptide correlates with a distinct f ibd

morphobgy. J. Struct. Biot. 1 19,59 - 71.

Sekoe, D.J. (199 1). The molecular pathology of Alzheimer's disease. Neuron 6, 487-

498.

Simmons, L.K., May, P.C., Tomaselfi, K.J., Rydel, R.E., Fuson, K.S., Brigham, E.F.,

Wright, S., Lieberburg, L, Becker, G.W., Brems, D.N., and Li, W.Y. Secondary structure

of amyloid beta peptide cornlates with neurotoxic activity in vino.(1993) Molec. Pharm.

45,373-379.

Snyder, S., Ladror, US., Wade, W.S., Wang, G.T., Barret, L.W., Matayoshi, E.D.,

Huffaker, H.J., Krafft, G. A., Holzman, T.F. (1994) Amyloid-beta aggregation: Selective

inhibition of aggregation in mixtures with different chah lengths. Biophys. L 67, 1216-

1228.

Soreghan, B., Kosmoski, J., and Glabe, C.G. (1994) Surfactant properties of Alzheimer

Abeta peptides and the mechanism of amyloid aggregation. J. Biol. Chem. 269,28551-

28554.

Stafford, W. (1992) Boundary analysis in sedimentation transport experiments: A

procedure for obtaining sedimentation coefficient distributions using the time denvative

of the concentration profile. And. Biochem. 203,295-30 1.

Suzuki, N., Iwatsubo, T., Odaka, A., Ishibashi, Y., Kitada, C. and Sara, Y. (1994) High

tissue content of soluble beta 1-40 is linked to cerebral amyloid angiopathy. Am. J.

Pathol. 145,452460

Taaford, C, Physical Chemistry of Macromolecules, p. 288, John Wiey and Sons Inc.,

New York, USA, 196 2 .

Teiier, JX., Russo, C., DeBusk, L.M., Angelini, G., Zaccheo, D. (1996) Presence of

soluble amyloid-beta peptide precedes amyloid plaque formation in Down's syndrome.

Nature Medicine 2,93-95.

Terry, RD., Masliah, E., Salmon, D.P., Butters, N, DeTeresa, R., Hill, R., and Katmian,

R (1991) Physical basis of cognitive alterations in Alzheimer's disease: synapse loss is

the major correlate of cognitive impairment. Am. Neurol. 30,572-580.

Terry, R.D. (1994) Neuropathological changes in Alzheimer disease. Prog. Brain Res.

101,383-390.

Tyler, J.M., and Branton, D. Rotary shadowing of extended molecules dned fiom

glycerol. (1980) J. Ultrastnict. Res. 7 1,95402.

Walsh, D., Hartiey, D.M., Kusumoto, Y., Fezoui, Y., Condron, MeMo, Lomakin, A.,

Benedek, G.B., Sekoe, D.J., and Teplow, D.B. (1999) Amyloid beta-protein

fibrillogenesis. Structure and bioIogicaI activity of protofibrillar intemediates. J. B iol.

Chem. 274,25945 - 25952.

Wood, S.J., Maleeff, B., Hart, T., and Wetzd, R (1996). Physical, morphological, and

fiinctional differences between pH 5.8 and 7.4 aggregates of the Alzheimer's amyloid

peptide AB. J. Mol. Biol. 256,870-877.

Yamaguchi H, Nakazato Y, Hirai S, et al. (1989) Electron micrograph of diffuse plaques.

Initial stage of senile plaque formation in the Alzheimer brain. Am. J. Pathol. 135, 593-

597.

Chapter 3: Aiternate Aggregation Pathways of the Alzheimer B-Amyloid Peptide: an in vitro Mode1 of Diffuse Amyloid

Adapted fiorn:

Huang, T.H. J., Yang, D.S., Fraser, PB., and Chakrabartty, A. (2000) J Biol Chem. 275, 36436-36440.

Absfmct

Deposition of Amyloid-8 (AB) aggregates in the brain is a defuiing characteristic

of Alzheimer Disease (AD). Fibrillar amyloid, found in the cores of senile plaques, is

surrounded by dystrophie netrites. In contrast, the amorphous AB (also called

preamyloid) in diffuse plaques is not associated with neurodegeneration. Depending on

the conditions, AB will also form fibrillar or amorphous aggregates in vitro. In this

present study, we sought to characterize the properties of the amorphous aggregate and

determine if we could establish an in vitro mode1 for amorphous AB. Circular dichroism

(CD) data indicated AB40 assembled to form either a B-stmctured aggregate or an

dolded aggregate, with the stnictrtred aggregate forming at high peptide concentrations

and the UIlStrtlctured aggregate fomiing at Iow A840 levels. The criticai concentration

separating these two pathways was 10 pM. Fluorescence ernission and polarization

showed the smictured aggregate was tightly packed, containing peptides that were not

accessible to water. Peptides in the unstnictured aggregate were loosely packed, mobile

and accessible to water. When examined by electron microscopy (EM), the structured

aggregate appeared as protofibrillar structures and formed classic amytoid fibrils over a

period of several weeks. The unstnictured aggregate was not visible by EM and did not

generate fibrils. These fmdings suggest the unstnicturrd aggregate shares many

properties with the amorphous AB of AD and conditions can be estabLished to form

amorphous AB in vitro. This would d o w for investigations to better understand the

relationship between fibrillar and amorphous AB and could have sign5ca.t impact upon

efforts to h d therapies for AD.

Introduction

Amyloid deposits, composed primarily of the amyIoid-fl (AB) peptide, are

associated with the neuropathology of Alzheimer Disease (AD). Though the precise

nature of the relationship between this 39-43 residue peptide and neurodegeneration is

still under investigation, a growing body of evidence supports the idea that AB plays a

primary causal role in the progression of the disease (Pike et al., 1993; Simmons et al.,

1993; Lorenzo and Yankner, 1994).

AB is a proteolytic product of the fLamyloid precursor protein (APP). APP, a

large type4 transmembme protein, is encoded on chromosome 21 and is constitutively

expressed in many cell types (Kmg et ai., 1987; Selkoe et al., 1988; Dyrks et al., 1988).

AB assembles in a nucleation-dependent process to form intemediate protofibrils and

uitimately fibrils 60-90 A diameter (Jarreît and Lansbury, 1992; Jarrett et al., 1993;

Walsh et al., 1997). Staining with Congo red and examination under polarized light

produces green birefihgence (PuchtIer et ai., 1962) indicating that amyloid deposits are

composed of peptide polymers with &sheet structure. This confornational characteristic

of AB fibrils is confumed by x-ray diffiraction andysis which shows cross-B fiber

diff'raction pattern with a distance between polypeptide ch* of 4.76 A and a distance

between sheets of 10.6 A (Kirschner et ai., 1986).

FibriIs are the main component of amyloid deposits knom as s e d e plaques.

Senile plaques are lesions 10-200 pm in diameter with a dense core of fibflar amyloid.

These plaques are mrrounded by degenerating and swollen m e tenninaIs (MüiIer-HiIi

and Beyreuther, 1989).

A second type of AB deposit found in AD brains is known as diffuse plaques.

UdÏke senile plaques, diaise plaques contain Little or no fibdlar AB and are instead

cornposed of amorphous AB (Yamaguchi et al., 1989). Amorphous A8 is non-

congophilic and is ody visible by immunocytochemicai methods, suggesting that these

deposits lack the ordered p-sheet secondary structure responsible for tinctorial and optical

properties of fibrillar amyloid. DiBise plaques are not surrounded by dystrophie neurites

and are not associated with neurodegeneration. It has been postulated that diaise plaques

are precursors to senile plaques and hence amorphous A$ has also been called

"preamyloid". Support for this idea cornes fiom studies with Down Syndrome (DS).

Patients with this disorder have three copies of chromosome 2 1 and exhibit extensive AB

deposition in thek brains (Wisniewski et al., 1985). There appears to be an age-dependent

progression tkom the amorphous AB deposits found in the brauis of young DS subjects to

the formation of senile plaques by adulthood (Giaccone et al., 1989; Mann, 1989).

If amorphous AB is the progenitor of AB fibrils then it would be of particular

usefulness to know what factors drive the conversion h m the benign amorphous AB of

diffuse plaques to the toxic fibrillar amyloid of senile plaques. An important early step in

galliing this information wodd be the establishment of in vitro conditions to form

amorphous AB aggregates. AB fibrillogenesis is a nucleation-dependent process which

can be signiscantly af3ected by the presence of s m d peptide aggregates acting as 'seeds'

(Walsh et al., 1999). In our work we have observed that AB preparations in which no

speciai precautions have been taken to eliminate these seeds would form non-specifk

aggregates visible by atomic force microscopy (Yang et al., 1999). We ask the question:

couid these in vitro aggregates be the amorphous AB in diffuse plaques?

To answer this question it was important to identify the means by which

amorphous AB could be recognized. What characteristics distinguish amorphous AB

fiom fibrillar amyloid?Although the conformation of AB in diffuse plaques is not known,

it is likely unstnictured. This provides a basis for devising an experimental plan to study

amorphous AB in vitm.

In the present shidy, we have developed a systematic methodology for identifjktg

diffuse amyloid. Sedimentation was used to determine the aggregation state of Ab.

Subsequently, CD spectroscopy, fluorescence ernission and fluorescence polarization

were employed to elucidate the intramoIecular and intennolecular structural

charactenstics of peptides within the aggregates. Finally, the gross morphology of the

aggregates was vinialized using electron microscopy.

Aggregation is a mdti-mo1ecuIar association reaction and is therefore

concentration dependent. In these experiments, concentration was varied to see how

peptide concentration influences AB self-association.

If we are able to establish Ni vitro conditions for amorphous AB formation, then

the methods used here WU allow for investigation into the aggregational and secondary

structural characteristics of the alternate aggregation pathways of AB. These fmdings will

provide a better understanding of the %riI formation process and may have implications

in understanding the relationship between diffuse amyloid and amyloid fibrils.

Peptide synthesis

Peptides were prepared by soiid-phase synthesis on a PerSeptive Biosystems 9050

Plus peptide synthesizer, as peptide-amides using PAL-PEG-PS resin (PerSeptive

Biosystems). An active ester couphg procedure, employing O-(7-azabenzotnazo1-l-y1)-

1,1,3,3-tetramethyluronium hexafluorophosphate of 9-fluorenylmethoxycarbony l amino

acids, was used. The peptides were cleaved from the resin with 8 1 : 13: 1 :5 trifluoroacetic

acid: thioanisole: m-cresol: ethanedithi01 mixture. M e r incubation for 1 hour at 25 OC,

the resin was removed by filtration and bromotrimethylsilane was added to a final

concentration of 12.5% (v/v). M e r incubation for 5 hours at O OC, the peptides were

precipitated and washed in cold ether. Peptide purity and identity were cont7rmed by

electrospray mass spectrometry.

Fluorescent labeling

Pnor to fluorescent labeling, a glycine residue was added to the N-terminus of

AB40 to act as a flexible Iinker between the fluorophore and peptide. 6-(N-(7-nitrobenz-

2sxa-1 J-diazol4yl)amino)hexanoic acid (NBD hexanoic acid) (Molecular Probes) was

then coupled to this extended AB40 sequence during peptide syntheas to create NBD-

AP0. Peptide pur@ and identity was confirmed by eIectrospray m a s spectrometry.

Preparation of stock peptide solutions and concentration determination

AB40 and NBD-AB40 preparations were lyophilized then dissolved in 10%

hexafluoro-2-propanol (HFIP). Peptide concentration was determined by tyrosine

absorbance at 275 nrn ( ~ 1 3 9 0 c c n f ' ~ ~ ' ) (Edelhoch, 1967) for A840 and by NBD

absorbance at 480 MI (~=23,000 cm- '~- ' ) (Chantler and Szent-Gyorgyi, 1978).

Absorbante measurements were made on a Perki. Elmer Lambda 3B spectrophotometer.

The stock solutions were stored at -20 O C untii required.

During the course of the experimeats, stock peptide solutions were stored for

several weeks; no changes in the behavior of these peptides were observed over this the.

Electron rnicroscopy

Negatively stained fibrils were prepared by fioating charged piolofom, carbon-

coated grids on peptide solutions. These solutions were incubated for 8 weeks before EM

preparation. To control pH, the peptide solutions were made using a 25 mM phosphate

buffet. M e r the gids were blotted and air-dried, the samples were stained with 1% (wlv)

phosphotungstic acid.

Platinum/carbon (WC) shadowing was performed using the glycerol spray

method (Huang et al., 2000). Solutions containing AM0 were brought to 30 to 50% (dv)

glycerol, sprayed onto freshly cleaved mica and fieeze-dried. The preparations were

shadowed on an Edwards E12E4 coater with WC (15 to 20A) at an angle of 5" to 7'

foiIowed by an additional 30 to 40 A of carbon-coating applied at an angle of 90'. Film

deposition was measured with a Balzea QSE2Ol quartz crystal monitor. The WC

replicants were floated off in distilled water and placed onto 300-mesh copper grids.

Representative electron rnicroscopy images of the peptide assemblies were

acquired on a Hitachi H-7000 TEM operated with an accelerating voltage of 75 kV.

Fluorescence emission spectroscopy

Steady-state fluorescence was measured at room temperature using a Photon

Technology Intemational QM-1 fhorescence spectrophotometer equipped with excitation

intensity correction and a magnetic stirrer.

For measurements of NBD fluorescence, emission spectra from 500 ~i fo 600

nm were collected (-70 m, averaging time=O.l or 1 sec, 1 ~1 steps, bandpass4 nxn

for excitation and emission). A quartz cuvette with 1 cm path length and 0.5 mi, volume

was w d . Spectra of samples containing only unlabeled AB40 were used to correct for

light scattering.

Samples were prepared with a fvted concentration of 0.1 pM NBD-AB40 as a

tracer. Stock peptides were &ed then diluted to the desired final concentrations and pH

using buffer (25 mM borate, 25 mM citrate and 25 mM phosphate). Mer dilution, the pH

was again measured and readjusted, if necessary, then d l samples were incubated 18-24

hours pnor to spectral acquisition.

Fluorescence depolarization spectroscopy

For steady-state polarkation experïrnents, the fluorescence spectrophotometer

was conflgured in the L-format and samples of NBD-AB40 and free NBD were excited at

478 m. The intensities (0 of the vertical (v) and horizontai (h) components of emission

at 530 nm were measured. Fluorescence polarization @) was calculated according to the

ept ion, ~C(I&GI'~)/(I~+G~,~) where nrst and second mbscripts refer to the positions

of the excitation and emission polarizers, respectively, and G=Ihv/Ihh. G is the factor used

to correct for detection ciifferences between vaticdy and horizontally polarized light.

A concentration series of AB40 was prepared (0.75, 1.5,3,6, 12,24 and 48 m M

AB40 with 0.1 pM NBD-Ap40; 25 m M phosphate, pH 7.0) and the fluorescence

polarization of the samples was measured.

Circular dichroism spectroscopy

Circuiar dichroism (CD) spectra were recorded on an Aviv Circuiar Dichroism

Spectrometer mode1 62DS at 2S°C. Spectra were obtained from 200 to 300 nm (0.5 cm

path length, 0.5 nm steps and 1 nm bandwidth) on a sample containhg 20 pM AB40 in

bufTer (7.5 m M borate, 7.5 mM citrate and 7.5 mM phosphate).

For the experiments we chose to use the 40 residue fom of AB (AWO). This

synthetic peptide was HFLC purified, but no attempt was made to remove aggregates or

fibril seeds which may have been present. A840 contains no intrinsic fluorophores thus it

was necessary to intmduce an extruisic fluorescent molecuie. This was accomplished by

attachhg the environment-sensitive fluorophore, NBD, to the N-terminus of AB40 via a

glycine residue. This glycine acts as a flexible linker to prevent any adverse steric effects

which may have been exerted by NBD. This labeled molecuie, NBD-AB40, was theo

added as a tracer to samples made with AP0. This approach has the distinct advantage

that most of the peptide wodd be wiId type while only a tiny component wodd be the

engineered molecde.

Concentration effect on AM0 aggregation

Samples with fnced tracer amowts of labeled peptide (0.1 pM MD-Ap40) and

ùicreasing concentrations of unlabeled peptide were prepared at pH 7 and dowed to

incubate ovemight. Sedimentation by centrifugation was used to test for the formation of

AM0 aggregates.

The absorbance of NBD at 490 nm was measured before centrifugation (Figure

3.1). The samples were then centrihiged and the NBD absorbance measured again. After

centrifugation littie if any NBD absorbance remained at al1 concentrations. nius,

sedimentable aggregates formed at al1 peptide concentrations. It was possible to pellet the

aggregate in a table top centrifuge set at 15 600 x g a relatively low g-force. This

suggested that the aggregates were large. Furthemore, since nearly al1 absorbance was

lost, most of the AM0 had aggregated into an insoluble form and only a midl fraction of

the peptide remained in solution.

Figure 3.1. The NBD absorbance of AB40 samples (O to 50 pM Ap40, 0.1 NBD- Ap0,25 mM phosphate, pH 7.0) were measured before (a) and after (O) centrifugation for 30 min. at 15 600 x g. Prior to centrifbgation absorbance at 490 nm rernained reIatively constant around 0.08 for concentrations up to 15 pM. By 20 p M A p 0 , absorbance increased with increasing concentration. AAer centrifugation, nearly ail of the absorbance at every concentration was Lost indicating that most of the AB40 was sedimentabLe.

Secondary structure of AM0 aggregates

The tinctond properties of amyloid suggest that fibrilla amyloid bas ordered

structure whiie amorphous AB does not. To see ifthere were any Merences in secondary

structure among the amyloid preparations CD spectroscopy was used (Figure 3.2). The

spectnim at lowest peptide concentration (8.7 pM) was representative of an do lded

conformation. Since the sedimentation data showed that most of the A B 0 is aggregated

at this concentration, this suggested that it is the aggregates which were unstnictured. At

17.1 pM the spectnim showed characterktics which were representative of fbstnictwe-

a positive band around 200 nm and a negative band at 218 nm. By 25.2 pM these bands

were quite pronounced The amount of fhructure content of AB40 aggregates reached a

maximum by 40.8 W. Thus, p-structure evolved as the total peptide concentration

increased. Light scatterhg due to the presence of aggregated material cm ofien confound

the interpretation of CD spectra. Because most of the peptide in these samples was

aggregated, this could have been a signincant problem. However, the specba here closely

resembled CD spectra of unfolded and B-sheet conformations. In this case, we are

confident light scattering effects have not advenely iduenced the results.

Thus, AB40 formed aggregates at aiI concentrations tested and there were two

distinct types: at low concentration, AB40 associated to form mstructured aggregates

while at high concentrations, AB40 associated to form &structure aggregates.

b -

- - -

L- -

-

O0 200 210 220 230 240 250 Wavelength (nm)

Figure 3.2. CD spectra of AB40 aggregates. SampIes of AB40 were prepared and incubated overnight (8.7, 17.1,25.2,33.1,40.8, 482, 55.4 @l Ap40, 5 m M phosphate, pH 7.0). The CD spectra dso showed a concentration-dependent change. The shape of the spectnun at 8.7 @l suggested that the AB40 in the sample contained little, if any, secondary structure. In con- the spectnim of the 55.4 pM AB40 sample, had a negative band at 218 nm and a positive band at around 200 nm, features which indicate structure. The extent of fbsheet adopted by AM0 evolved as totai peptide concentration increased,

In diaise plaques, amyloid appears wispy and non-compacted whereas s e d e

plaques appear dense and fibdar. These morphological characteristics suggest that

constituent peptides may have different solvent accessibility depending on whether they

are present in diffuse amyloid or fibrilla amyloid. To see if a similar situation existed

with our in vitro amyloid we looked at solvent accessibility using NBD as our

representative side chah. NBD, having a similar size and shape to tryptophan, could be

accepted into the aggregatefibril like other aromatic amino acids. Once introduced into

AB40, any effects of solvent accessibility will act on NBD as with other residues in the

sequence, with the advantage that any changes would be reported by the fluorescence

characteristics of the fluorophore.

Side-chain environment of AB40 aggregates

The fluorescence of NBD is quenched by water. When NBD-AB40 is not

assembled, NBD is accessible to water. NBD fluorescence would be quenched and

fluorescence emission in the sample wodd be low. When AB40 assembles, water wodd

be excluded and quenching reduced so fluorescence increases. If there were a fùrther

conformational change which results in a more compact structure that M e r sequesters

NBD fiorn water, then fluorescence wodd be further enhanced.

Again a fixed tracer amount of NBDAP40 was used in conjunction with an

increasing amount of imlabeled Ap0. With more AB40 there was increased fluorescence

emission, mggesting that solvent accessibility decreased with increasing concentration

(Figure 3.3). This change indicated that the aggregate at fow concentration was different

from the aggregate which formed at high concentration; specifically, that AB40

aggregates condensed with increasing concentration to f o m more compact structures

which effectively sequestered NBD fiom quenching by water. Ifthere were no change in

structure then the fluorescence would have remained flat throughout the increasing

concentrations.

Centrifugation of the samples and memernent of fluorescence produced redts

identical to those fiom the absorbance meanirements. Nearly ail the fluorescence in each

sarnple was lost after cenhihigation. This confumed that large sedimentable aggregates

formed at all concentrations. Furthemore, that most fluorescence was lost indicated the

majority of peptide aggregated and ody a small fiaction of AB40 comprised the soluble

component.

The evidence accumulated thus far supported the idea that we were producing in

vitro versions of diffuse and fibrillar amyloid, as these aggregates shared properties with

in vivo AD amyloid. The increase in fluorescence was consistent with formation of non-

compact structures at low peptide concentration and evolution of tightly packed water

excluding structures as protofibrils or fibrils formed with increasing peptide amounts.

The rnobility of sidechains wouid provide fuaher idormation as to how tightly packed

the peptides are in one aggregate type versus another. Another property of fluorescence,

fluorescence polarization, c m be used to investigate this.

Figure 3.3. The NBD fluorescence of AB40 samples (O to 50 pM AB40,O.l pM NBD- AP40,25 mM phosphate, pH 7.0) was measured before (0) and after (@) centrifugation for 30 min. at 15 600 x g. The fluorescence measurements showed a concentration dependence with the intensity of NBD fluorescence emission increasing with higher total peptide concentrations. Af'ter centrifugation, M e fluorescence remained in the sample supernatant, confimiing the conclusion fiom the absorbance experiments that most of the AB40 is in the forrn of large sedimentable aggregates.

Side-chah mobility of Aw aggregates

If a fluorophore is excited by polarized light then the emitted light will also be

polarized. The fluorophore wiil move during the time delay between excitation and

emission and this movement will result in the depolarization of the emitted light relative

to the excitation light. The degree of depolarization is dependent on the rotational

relaxation tirne; the greater the rate of rotation the lower its polarkation. In our system,

two motions will affect the polarization of the fluorescence emission: global motions and

local motions. Any changes in global motions wouid give Uiformation on the aggregation

state of NBD-AB40 while changes in local motions would provide information on the

packing state of ou. representative side-chain NBD.

The polarization of free NBD was first measured (Figure 3.4). This value was

reIatively low, 0.32, because it is a smail molecuie and has a short rotational relaxation

time. However, once attached to Ap0, a much larger molecule, its movement was more

reshicted and the measured polarization was higher. Polarization increased up to 10 @l

then plateaued. The 10 pM AB40 concentration may mark a transition between two

States. We knew that large aggregates formed at al1 concentrations, it was udikely then

that the change in polarization was due to global motion changes. Rather, it was more

likely the local motions of NBD were the cause. Thus, above 10 pM, M3D was contained

within the p-structure aggregate and was tightly packed and not free to move. 10

seems to be criticai concentration dividing the conversion between two aggregate types.

I - b

L. - b

* - L , . ...... -...... poieritatiori of fm NBD fiuomhore P

r . 1 . . - - I . . . . 1 - . . . l . . . . t . . . .

Figure 3.4. Fluorescence polarization of AB40. A concentration series of AB40 was prepared (0.75, 1.5,3,6, 12,24 and 48 pM AB40 with 0.1,0.2,0.4,0.8, 1 -6.3 1 and 6.4 pM NBD-AB40; 25 mM phosphate, pH 7.0) and the fluorescence polarkation of the samples were meamred. Fluorescence polarization was 0.49 at the Iowest peptide concentration tested. Polarization increased to a maximum value of 0.54 at 13.6 pM and declined slightiy up to the highest concentration of 54.4 W. The polarkation of the fiee NBD fluorophore is shown as a =ference (dotted line).

A$ formed two distinct aggregate types: an unstructured aggregate and the

structured aggregate. Not ody is the UnStNctured aggregate lacking regular secondary

structure but also its conformation is looser, solvent exposed and its constituent peptides

are not as tightly associated. On the other hanci, the peptides in the B-structured aggregate

showed B-sheet secondary stnicture, were tightly packed and excluded water fiom their

side-chah environments. These in vitro aggregates appeared to share many properties

with in vivo amyloid forms. As a final step we attempted to visudize the morphology of

the aggregates by employing EM

Ultrastructure of AM0 aggregates

AB40 preparations were examined by either platinum-carbon shadowing electron

microscopy or negative-stain etectron microscopy. At 0.1 pM NBD-AB40 and 5 pM

AB40 (Figure 3.5A), a concentration which exhibited aggregate formation but no ordered

secondary structure, ody a few maIl spherical structures were pment Though we know

that the unsûuctured aggregate was fomiing at this concentration, they were not visible

by this technique. As the peptide concentration was increased (0.1 pM NBD-AB40 and

50 @A Ap40, Figure 3.5B). these structures became abuadant; moreover, the sphericd

aggregates seemed to assemble into clusters and elongated aggregates shilar to

protofibnls. Because CD showed that p-structure evolved with concentration we

concluded that these protofibriIs were the p-structure aggregate. Afier 8 weeks

incubation, the high concentration AB40 sample (0.1 p M NBD-AB40 and 48 pM AP0,

pane1 C) had taken on the morphology of classic amyloid nbds.

Figure 3.5. Platinum-carbon shadow electron microscopy (EM) was penormed on samples of AfM0 at varying concentrations after ovemight incubation. Negative-stain EM was performed on a 48 pM sampIe which had been incubahg for 8 weeks. Al1 samples contained 25 mM phosphate and had a pH of 7.0. Bars indicate 100 m. Sorne spherical structures were visuaiized at 0.1 pM NBD-AB40 (panel A). As the amount of peptide was increased (panel B: 0.1 p M NBD-AB40,50 pM AB40) the number of structures increased. The structures at the higher concentration appeared to be aggregates of the sphericd structures which predomuiated at the Iower concentrations and resembled protofibrils. The micrograph of the 8 week old 0.1 @l NBBAw0,48 ph4 AM0 sample (panel C) showed that our synthetic peptides fonned classic amyloid fibrils under the in vitro conditions employed in the experiments.

Discussion

In the search to i d e n e the nelnotoxic species of A$ in AD, amorphous AB has

been ignored as a subject of investigation because diffuse plaques are not associated with

neurodegeneration. However, it is clear that amorphous AB is an important component of

the total amyloid load in the AD brain and, therefore, its properties should be better

understood. Previous studies have emphasized the importance of ensuring that

preparations of AB are seed-free in generating AB fibrils rather than non-specific

aggregates (Walsh et al., 1999). We reported previously that peptide stocks which have

been carefùily pre-treated to remove seeds generated soluble oligomenc structures which

formed pnor to fibrillar amyloid (Huang et al., 1999). We also observed that AB40

preparations where seeds have not been removed formed amorphous aggregates (Yang et

al., 1999). In the present study, we have found that these samples may provide a usehl in

vitro system for studying in vivo amorphous Ag.

Many of the experiments make use of AB40 labeled with the fluorescent probe

MD. The behavior of MD-AB40 appears to be comparable to that of AB40 itself. We

have reported previously that NBD-AB40 has M a r sohbility characteristics (Yang et

al., 1999). Furthemore, under defined conditions, AB40 labeled with NBD as well as

other fluorescent probes, Trp or AEDANS, self-associates to form oligomeric structures

B e tmiabeled AB40 (Huang et al., 1999). As a final precaution, to reduce the possibility

of the fluorophore afEecting the behavior of AWO, Iabeled peptide has been used in tracer

amounts*

There are at least two different aggregation pathways for Ap0 , one leading to a

fbstnichire aggregate and the other to an unstmctured aggregate. The f%stnictured

aggregate is cornposed of tightly packed AB40 peptides which have fbsheet secondary

structure and may be the protofibril precursor to AB fibrils. The unstructured aggregate

lacks repuiar secon* structure, its constituent peptides are not tightly associated and its

conformation is looser with the side-chains solvent exposed. The unstnictured aggregate

shares mmy properties with the diffuse amyloid in AD. Whether AB40 associates to form

the unstnictured aggregate or fibrils is determined by the total peptide concentration. The

critical concentration separatirtg these two pathways is 10 W. A possible explanation for

why the more compact and ordered fibrils form at higher concentration is that the nucleus

for fibril formation is stable only at peptide concentrations above 10 pM.

It has been suggested that there is a tirne-dependent relationship between

amorphous AB and fîbrillar amyloid where diffuse amyloid is the precursor to fibril

formation. Our results show this relationship may not only be one of time but rather one

which is dso based on concentration. Some mutations associated with AD increase A$

concentration. This would have the effect of shifting the aggregation pathway to favor

formation of Bstructured aggregate and hence AB fibrils. If nbrils are toxic it would be

cnticd to prevent this step nom t a h g place. A treatment m t e g y is to reduce AB

concentration below the critical concentration for conversion to fibrils and thereby

maintain amyloid in the non-neurotolac amorphous form making it unnecessary to

completely elimuiate AB.

It would be usehl to examine whether amorphous and fibrillar AB directly

interconvert, that is, once amorphous Ab is formed does it convert to fibrÏiIar amyloid?

Solid to soiid transitions are rare and require extreme conditio~, such as high pressure in

the case of coal-graphite-diamond conversion. It is unlikely that a such a transition is

taking place in vivo. A more plausible mechanism is that A $ monomers are the

intermediate between these two aggregates. An understanding of this interconversion

would be necessary to evaiuate whether reducing peptide concentration to control the

aggregation pathway would be sufficient to prevent AD pathogenesis.

Alternatively, the aggregation of AB into amorphous forms may generate, by

chance, nucleation sites fiom which structured fibrils can grow with the addition of

monometic AB. The present study indicates that concentration is an important factor

influencing amyloid morphology; however, the precise mechanisrns driving conversion

will have to be d e t e d e d in fiiture experiments.

The methodology used in this study can be deployed to test dmgs. Fluorescence

techniques cari be utilized as a preliminary screen for potential drugs which prevent

formation of fhtructured aggregates. Subsequently, CD spectroscopy can be applied for

confirmation.

Chantier, P.D., and Szent-Gyorgyi, A.G. (1978) Spectroscopic studies on invertebrate

myosins and light chains. Biochemistry 17,5440-5447.

Dyrks, T., Weidemann, A., Mdthaup, G., Salbaum, J.M., Lemaire, H.G., Kang, JOy

Muller-Hill, B., Masters, CL., and Beyreuther, K. (1988) Identification, traiisrnernbrane

orientation and biogenesis of the amyloid A4 precursor of Alzheimer's disease. EMBO.

J. 7,949-957.

Edelhoch, K. (1967). Spectroscopic determination of tryptophan and tyrosine in proteins.

Biochemistry 6,1948- 1954.

Giaccone, G., Tagliavini, F., Linoli, G., Bouras, C., Frigerior, L., Frangione, B., and

Bugiani, 0. (1989). Doms patients: extracellular preamyloid deposits precede neuritic

degeneration and s e d e plaques. Neurosci. Lett. 97,232-238.

Huang, T.H.J., Yang, D.-S, Plaskos, NP., GO, S., Yip, C.M., Fraser, P.E., and

Chakrabartty, A. (2000) Structural snidies of soIub1e oligomers of the Alzheimer &

amyloid peptide. J. Mol. Biol. 297,73987.

Jmett, J.T., and Lansbury, P.J. (1992) Amyloid fibril formation requires a chemicdy

discriminating nucleation event: studies of an amyloidogenic sequence fiom the bacterial

protein OsmB. Biochemistry 3 1,1234542352.

Janett, J.T., Berger, E.P., Lansbury, P.T. Jr. (1993) The carboxy terminus of the beta

amyloid protein is critical for the seeding of amyloid formation: Implications for the

pathogenesis of Alzheimer's disease. Biochemistry 332,46934697.

Kang, J., Lemaire, H.-G., Unterbeck, A., Salbaum, J.M., Masters, CL., Grzeschik, K.H.,

Muithaup, G., Beyreuther, K., and Müller-Hill, B. (1 987) The precursor of Alzheimer's

disease amyloid A4 protein resembles a cel1-dace receptor. Nature 325,733-736.

Kirscher, D., Abraham, C., and Seikoe, D. (1986) X-ray diffraction Crom intraneuronal

paired helical filaments and extraneuronal amyloid fibers in Alzheimer disease indicates

cross-6 confonnatioa. Proc. Natl, Acad. Sci. USA 83,503-507.

Lorenzo, A. and Yankner, B. (1994) Beta-amyloid neurotoxicity requires fibnl formation

and is inhibited by congo red. Roc. Natl, Acad. Sci. USA 91,12243-12247.

Mann, D.M.A. (1989). Cerebrai amyloidosis, aging and Alzheimer's disease: a

contnaution h m studies on Dom's syndrome. Neurobiol. Aghg 10,397-399.

Müller-Hill, B., and Beyreuther, K. (1989) Molecular biology of Alzheimer's disease.

AM. Rev. Biochem. 58,287-307.

Pike C.J., Burdick, D., Waiencewicz, A.J., Glabe, CG., and Cotman, C.W. (1993).

Neurodegeneration induced by beta-amyloid peptides Ni vitro: the role of peptide

assembly state. J. Neurosci. 13, 1676-1 687.

Puchtler, H., Sweat, F., and Levine, M. (1962) On the binding of Congo red by amyloid.

J. Histochem. Cytochem. 1 O, 355-364.

Selkoe, D.J., Podlimy, M.B., Joachim, C.L., Vickers, E.A., Lee, G., Fritz, L.C., and

Oltersdorf, T. (1988) Beta-amyloid precursor protein of Alzheimer disease occurs as 1 10-

to 135-kilocialton membrane-associated proteins in neural and nonneural tissues. Proc.

Nati. Acad. Sci. U S A 85,734197345.

Simmons, L.K., May, P.C., Tomaselli, K-J., Rydel, R.E., Fuson, K.S., Brigham, E.F.,

Wright, S., Lieberburg, L, Becker, G.W., Brems, DN., and Li, W.Y. Secondary structure

of amyloid beta peptide correlates with neurotoxic activity in vino.(1993) Molec. Pharm.

45,373-379.

Walsh, D.M., Lomakin, A., Benedek, G.B., Condron, MM., and Teplow, DB. (1997)

Amyloid beta-protein fibrillogenesis: detection of a protofibrillar intermediate. I. Biol.

Chem. 272,22364-223 72.

Walsh, D.M., Hartley, DM., Kusumoto, Y., Fezoui, Y., Condron, M.M., Lomakin, A.,

Benedek, GoB., Sekoe, D.J., and Teplow, D.B. (1 999) Amyloid &protein fibrillogenesis:

structure and biological activity of protofibriliar intermediates. J. Biol. Chem. 274,

25945-25952.

Wisniewski, K.E., Dalton, A.J., McLachlan, C., Wen, G. Y., and Wisniewski, H.M.

(1985) Alzheimer's disease in Dom's syndrome: clinicopathologic studies. Neuroiogy

35,957-961.

Yamaguchi, H., Nakazato, Y., Hirai, S., Shoji, M., and Harigaya, Y. (1989) ) Electron

micrograph of diffuse plaques. Initial stage of senile plaque formation in the Alzheimer

brain. Am. J. Pathol, 135,593-597.

Yang, D.-S., Yip, CM., Huang, T.H.J., Chakrabartty, A., and Fraser, P.E. (1999)

Manipulahg the amyloid-B aggregation pathway with chernical chaperones. 5. Biol.

Chm. 274,32970932974.

Chapter 4: Manipulating the Amyloid-$ Aggregation Pathway

with Chernical Chaperones

Adapted fiom:

Yang, D.-S., Yip, C.M., Huang, T.H.J., Chakrabartîy, A., and Fraser, P.E. (1 999) J. Biol. Chem. 274,32970-32974.

Amyloid-8 assembly into fibriliar structures is a defhing characteristic of

Alzheimer disease that is initiated by a conformationai transition from mndom coil to B-

sheet and a nucleation-dependent aggregation process. We have investigated the role of

organic osmolytes as chemicai chaperones in the amyloid pathway using glycerol to

mimic the effects of naturally occurrhg molecules. Osmolytes, such as the natwaily

occtmïng trimethylamine oxide (TMAO) and glycerol, correct folding defects by

preferentially hydrating partially denatured proteins and entropically stabilize native

conformations and polymeric states. TMAO and glycerol were found to rapidly

accelerate the AB random-to-p conformational change necessary for fiber formation.

This was accompanied by an immediate conversion of amorphous, unstnictwed

aggregates into uniform gio bdar and possibly nucleating structures. O mol yte- facilitated

changes in AB hydration also affected the k a l stages of amyloid formation and mediated

transition fiom the protofibd to mature fibers that are observed in vivo. These fidings

suggest that hydration forces can be used to control fibril assembly and may have

implications for the accumulation of AB within intracelldar compartments such as the

endoplasmic reticuium and for in vitro modeling of the amyloid pathway.

Amyloid plaques are a centrai featurr of Alzheimer disease (AD) pathology and

are considered to be a major factor in neuronal ce11 loss (Selkoe, 1997; Kisilevsky and

Fraser, 1997). Ultrastmcturally, plaques are fibrous masses composed primariiy of the

40-42 residue amyloid-B (AB) peptide. AB is derived intracellulady by endoproteolysis

of the integral membrane amyloid precursor protein (APP) that results in secretion of the

peptide by normal cellular pathways as well as intracellukir accumulation. Under

pathogenic conditions, AB self-associates into a wetl-defmed supramolecular fibril with

high p-sheet content (Inouye et al., 1993). AB polymerization is considered to be a two-

stage process initiated by the association of individual Af3 monomers into small

nucleating 'seeds' that is accompanied by a transition from a predorninately random coi1

to an arnyloidogenic 8-sheet conformation (Jamtt et al., 1993; Lomakin et al., 1997).

Subsequent to nucleation, the AB 'seeds' assemble in a chah-like mamer to yield an

intermediate protofiril structure (Stine et ai., 1996; Harper et al., 1997a; Walsh et al.,

1997) which may represent a common elernent of aU amyloid fibnls (Sunde et al., 1997;

Blake and Serpell, 1996). ProtofibriI conversion into the ramified fibrils observed in vivo

can be affécted by factors such as the amyloid-binding apoiipoprotein E (ApoE) which is

an AD rïsk factor (Schmechel et al., 1993), the relative quantity of the more

amyioidogenic AB species (Jarrett et al., 1993) and other chaperone elements that may

control AB ~e~association.

The observation that A$ is generated within intracellular compartments, including

the endoplasmic reticdum (Cook et ai., 1997; Hartman et al., 1997; Wild-Bode et al.,

1997; Skovronsky et al., 1998). which is the quality control site for protein folding, has

prompted us to investigate the role of chemicai chaperones in the AB folding pathway.

Under stress conditions or exposure to denaturants, heat shock proteins Oisp) and

chemical chaperones assist in stabiiizing correctly folded proteins. Through a process

known as osmotic remediation (Burg, 1995), chemical chaperones or organic osmolytes

including carbohydrates, fiee amino acids or methylamines (e.g., trimethylamine N-

oxide (TMAO)), effectively control protein foldhg through a preferential hydration of

exposed polypeptide backbone and sidechains of partidly dolded structures (Wang and

Bolen, 1997). Chemical agents such as glycerol and polyethylene glycol mimic these

hydration effects which creates a thermodynamically unstable state due to the

unfavourable entropic changes associated with the increased ordering of bound water

molecules (Timasheff et al., 1993). This can be rectified by folding of the protein into its

native conformation, which sequesters the exposed groups and excludes the osmolytes

fiom the protein domain. As a result the fiee energy of the native conformation is

substantially lower than the do lded state which is demonstrated by the ability of

glycerol to correct misfolded proteins within the endoplasmic reticdum (Brown et al.,

1997). Hydration eRects are equally important in protein polymerization where

osmolytes are excluded through increased protein-protein interactions and can, for

example, enhance the assembly and stability of microtubules (Sackett, 1997). The

prefedal assembly of protein polymer subunits, such as AB and tubulin, is a product of

theu unique structure, which results in the fibrous aggregate being the lowest energy

conformer. ûur Ourdings indicated that s i d a r forces contribute to the initiation of the

Ab random coi1 to Psheet transition and stabilization of the resuiting aggregates. These

observations suggest that chernical chaperones may be useful in modeling amyloid

plaque formation and rnay have some bearing on the cellular events hvolved Ui fibd

formation.

Ekperimental Procedures

Peptide synthesis and labeling

ArnyIoid-B peptides residues 1-40 (AB40) was synthesized and purified as

previously described (McLaurin et al., 1998). Peptides were exarnined irnmediately

following dissolution in aqueous b a e r and following prolonged incubation at high

concentration to promote srnail protofibrillar B-sheet aggregates. AB40 was Iabeled for

fluorescence studies using N,N'-dimethyl-N-(iodoacetyl)-N'-(7-nih~benz-2-oxa-1,3-

diazol4yl)ethylenediamitle (IANBD amide, Molecular Probes) as descnbed previously

(Huang et al., 1997). NBD-labeled peptide was purified by gel filtration

chromatography, disaggregated in 10% hexafluoro-2-propanol (HFIP) and stock

solutions stored at -20" C.

Secondary structure analysis

The effects of trùnethylamine oxide (TMAO) (Sigma) and glycerol (BDH

Chemicals) on AB conformation were determined by circdiv dichroism (CD). Peptides

were dissolved to a finai concentration of 50 pM in distiUed water or 10 mM phosphate

b a e r (pH 7). Peptide solutions were combined with the TMAO at concentrations

ranghg from 50-150 pM and with glycerol ranging &orn 1.2 to 6 M (10-50 % by

volume). Spectra were collected foliowing a 10 minute equilibration period and afier 48-

72 hrs incubation at room temperature. Spectra were acquired on a Jasco

spectropolarimeter Mode1 5-715 in a 0.1 cm path length celi over the wavelength range

190-250 MI with a 1 .O m band width, 0.1 nm resolution, 1 sec response time and a scan

rate of 2 0 - d m h . AU spectra were corrected by subtraction of any contributions from

btûfer, giycerol or po lyethylene glycol (PEG).

Electron and atomic force microscopy

AB aggregates were examined by phosphotungstic acid negative staining and

platinum/carbon (WC) shadowing techniques as previously described (Wang and Bolen,

1997). For shadowing studies, the samples were atomïzed ont0 freshly cleaved mica,

immediately plunged into liquid nitrogen and lyophiiized to eliminate drying artifacts that

codd be caused by changes in peptide concentration. Dned preparations were platinum

coated in an Edwards E12E4 coater and viewed on an Hitachi H-7000 operated with an

accelerating voltage of 75 kV.

For tapping-mode atomic force microscopy (TMAFM) studies, the peptides were

dissohed in 25 mM phosphate buEer (pH 7) and then adjusted to the desired TMAO or

giycerol content to a bai peptide concentration of 2.5 @M. Approlcimately 10 pL, of the

peptide solution was W e r r e d to a piece of freshly cleaved mica glued to a steel AFM

sample molmt The sample was then immediately sealed in the TMAFM liquid ce11 and

the cell nIled 6th b s e r solution. TMAFM imaging was conducted at room temperature

using a combination contactftapping mode Liquid cell fitted to a Digital Instruments

Nanoscope IIIA MultiMode scanning probe microscope. All images were acquired using

120 pm oxide-sharpened silicon nitride V-shaped cantilevers (Type DNP-S, Digital

Instruments) at a scan rate of -2 Hz at a samphg rate of 256 or 512 points per scan he.

Prior to use, the AFM tips were exposed to UV irradiation to remove adventitious organic

con taminants fkom the tip d a c e . W e a priori detefmination of the appropriate drive

frequency is dificuit owing to viscous couphg between the cantilever and the fluid

medium, which gives rise to multiple broad resonance peaks, optimal irnaging was

achieved at a cantilever drive fiequency of -8.9 kHz

Solubility measurements

The aggregation state of AB under the various solvent conditions was examined

by high speed centrifugation and assay of the soluble materiai using the fluorescent

labeled peptide as indicator. Solutions containing 0.1 pM NBD-labeled AB40 and 10 pM

unlabeled peptide were combined in 25 mM phosphate buffer (pH 7) containing kom O to

6.0 M glycerol. NBD-labeled AB40 fluorescence spectra were acquired; the samples

cenhifuged in a Beckman Airfuge at maximum velocity (135 000 x g) for 30 minutes;

and the fluorescence spectra of the supernatant was measured Spectra of control samples

containing only AB40 were collected and subtracted From the NBD-labeled AB40

fluorescence to correct for the effects of light scattering. Steady-state ffuorescence was

measured at room temperature using a Photon Technology International QM-1

fluorescence spectrophotometer equipped with excitation intensity correction. Emission

spectra fkom 500 rn to 600 nm were coUected (L478 nm, 0.1 secfnm, 8 nm bandpass

for excitation and emission) using a 0.2 x 1 cm path length, 0.5 mL cuvette. To

determine the effects of glycerol on the fluorescent probe, emission spectra of

unconjugated NBD at 0.1 p M was measured under comparable conditions.

ResuIts and Discussion

Chcular dichroism (CD) was used to evaluate the effects of the naturally

occurring osmolyte, TMAO, and glycerol on AB conformation. When dissolved in

aqueous buffers, AB residues 1-40 (AB40) initialiy exhibit a randorn coi1 conformation

indicative of an unordered structure (Figure. 4.1A). Consistent with previous reports, the

conformational changes of AB fiom random to 8-sheet (McLaurin et al., 1998) and

subsequent fibril formation (Harper et al., 1997b) can take from h o m to days depending

on the particular peptide batch and incubation conditions. In contrast, adjusting the

solution to 1.2 M glycerol(lO% vlv) resulted in an immediate folding of the peptide into

the amyloid-associated fbconformer. Increasing the glycerol level to 3 or 6 M (25-50%

VIV) produced a linear increase in the 8-sheet content as measured by the minima

htensity at 21 8 nm (Figure 4.1 A).

Figure 4.1. Panel A: The effects of glycerol on Ap40. CD of AB40 demonstrathg the immediate conversion from random coi1 to a B-sheet conformation with increasing giycerol concentrations. Panel B: Conformationai transitions of an incubated or "aged" AB40 comparable to the protofibril state iodicating a nmilar effect of increasing the & sheet content induced by glycetol-mediated solvation effects. Spectra are of the AB peptide in buffer (soiid h e ) and in the presence of 1.2 M (dash-dotted line), 3.0 M (dashed line) or 6.0 M (dotted he) giycerol.

These effects were not due to increased peptide concentration caused by

immiscibiiity in glycerol since CD studies conducted where AB concentrations were

doubled (50 pM to 100 p.M) to replicate the 50% glycerol conditions reveaied no change

in the random coil structure of AB40 (data not shown). A similar change was observed

for AM0 which had been preincubated to form B-sheet aggregates which represent the

early stages of fibril formation. In this case, elevating glycerol concentrations

proportionally increased the pre-existing fl-sheet content but to a slightly lesser

quantitative degree as compared to the random coi1 to B-conformationai change (Figure

4.1B). Identical resuits were obtained under al1 conditions in the presence of a low

molecular weight (400 Da) PEG at varylng concentrations (data not shown). These

results indicate that changes in protein hydration by chernical chaperones rapidly

accelerate the conformational transition required for amyloid formation.

GIycerol is a non-physiological mode1 of osmolyte activity and we therefore

investigated the effect of trimethylamine oxide (TMAO) which is found Ni vivo and acts

to maintai. correctly folded proteins in several species (Wang and Bolen, 1997). TMAO

produced a nmilar random-to-$ transition as that seen with glycerol but at significantly

lower concentrations of 50-150 pM (corresponding to molar ratios of 1:l and L:3

peptidelTMA0) (Figure 4.2A).

Figure 4.2. Panel A: Effects of TMAO. Circdar dichroism (CD) of AB40 demonstrating the imrnediate conversion fiom random coiI to a 8-sheet conformation with increasing concentrations of trimethylamine oxide (TMAO). Control AB in b a e r only (solid line), AfkTMAO at molar ratios of 1 : 1 (dashed-dotted Iine), 1 : 1.5 (dashed line), and 1 :3 (dotted he) are indicated. Panel B: Roportional increases in the p-sheet conformation as measured by the absorption at 218 nm in the presence of TMAO (closed circles) and giycerol (open squares) with AB initidly in the random conformation and following preincubation pnor to the addition of glycerol (closed squares) are dso indicated.

Similar to glycerol, TMAO increased the quantity of the fkonformation with the

pre-incubated peptide which Uiitially displayed a folded and aggregated structure (data

not shown). The increases in &sheet conformation by both TMAO and glycerol were

found to be roughly linear as determined by the absorption at 218 nm (Figure 4.2B).

However, at the higher concentrations of TMAO, the proportion of the B-conformation

appeared to be approaching a plateau. These fïndings suggest that TMAO is a more

active compound in controllhg the folcihg and aggregation state of AB which may reflect

a more potent effect of this osmolyte in an in vivo setting.

The morphological changes associated with the p-sheet conformation were

assessed by tapping mode atomic force microscopy (TMAFM) performed directly in

aqueous buffer. TMAFM of AB in the absence of the chemicai chaperones revealed

irregular aggregate rafts with an approximate thickness of 10 nm (Figure 4.3A). The

presence of aggregates was unexpected since it is generally assumed that the random coi1

conformation corresponds to a Mly solvated and soluble AB monomer. Since the

sarnples were not pretreated to remove srnaII peptide aggregates (e.g., submicron

filtration), these minor components could be present and senre as nuclei for non-specific

precipitation. Aiternatively, the amorphous deposits may be due to adsorption of A$ to

the mica substrate used for TMAFM. Addition of either TMAO or glycerol at

concentrations that produced the random-to-8 transition, as c o b e d by CD, redted in

a complete conversion of the amorphous AB aggregates to a mixture of protofibrüs and

smd, ellipsoïdal particles (Figure 4.3B,C).

Figure 4.3. Panel A: Morphological changes in peptide aggregates induced by chemicai chaperones. Atomic force microscopy of AB40 (2.5 PM) in phosphate buffer demonstrating the arnorphous aggregates. Panel B: A wide field scan of an identicai AB40 peptide in 6.0 M glycerol reveaiing the formation of protofibrils and smder non- fibrillar aggregates. Identical structures were observed in the presence of 150 pM TMAO. Panel C: Enlarged view of the protofibrils indicating the globular morphology and the discrete ellipsoidal aggregates (arrow) representing the tetmmenclpentameric aggregate. Scale bars are 100 nm in (panel A), 2500 nm (panel B) and 50 rn in (pane1 Ch

The protofibrils varied in length and displayed a globular, axial penodicity of

approximately 100 A that was comparable to their average diameter. These were

morphologicdy similar to previously reported protofibrils (Harper et al., 1997a; Walsh

et al., 1997). However, judging from wider field scans, the ellipsoid aggregates

predominated, with dimensions of 50 x 60 x 15 A (Figure 4.3C, mow). We note that

these dimensions are slightly overestimated due to convolution of the AFM tip shape

"th the aggregate shape.

Given an experimentally detemiined volume of -45 500 A3, these ellipsoid

particles represent AB tetramers or pentarners based on a calculated volume of

approximateiy 10 000 A3 for an AB (residues 1-42) molecule folded into a 2-stranded fL

sheet. These rnay represent an aggregate of the AB dimer which has been shown to be

stable under sirniIar aqueous conditions (Gazon-Rodriguez et ai., 1997). Comparable

structures have been observed by AFM (Roher et al., 1996; Kowaiewski and Holman,

1999) that have been termed AB-detived diffisible ligands (ADDLs) which may

represent the most neurotolric species (Lambert et al., 1998). The position of these smaii

aggregates on the amyloid pathway is presently unclear. A straightforward explanation

would be that there is a linear relationship where these are the progenitors of the

protofibrils which are generated by the direct polymerization of the ellipsoid aggregates.

Mtematively they rnay represent a side reaction with AB monomers shutthg between

these aad the protofibrils.

Assembly of the nucleating aggregates to fom protofibrils is considered to be the

slow kinetic phase of the amyloid pathway (Evans et aI., 1 995). This is foiIowed by the

thermodynamic phase with the transition to compacted amyioid fibrils, a process which

may also be affected by osmolyte-induced hydration. Previous AFM and negative stain

eIectron microscopy studies have defhed protofibrils as truncated and highly flexible

fibrillar structures that are the precursors to plaque-related fibrils (Stine et al., 1996;

Harper et al., 1997a; Walsh et ai., 1997). Following preincubation of a concentrated

AB40 solution, we have observed similar aggregates of similar morphology using

platindcarbon shadowing techniques (Figure 4.4A). Such protofibrils ranged in length

fkom 1-100 nm and were poorly contrasteci, suggesthg that they have a low profile. The

addition of glycerol resdted in a rapid conversion to straight, compacted fibrils that

extended over several hundred nanometers (Figure 4.4B) and occasionaIly displayed

helicai twisting (arrowheads). Exposure to chernical chaperones such as glycerol also

significantly improved the image contrast of the fibrils, presumably as a result of their

compaction into more defmed tubular assemblies. While some c w e d protofibril-like

structures remahed following glycerol treatment, there was a complete conversion to the

longer fibrils following several hours incubation (data not shown). Elongation occuned

at a significantly greater pace as compared to the several days required to convert AB

protofibrils in the absence of the glycerol chaperone.

Figure 4.4. Panel A: Rotary platinumlcarbon shadowing electron microscopy of AB40 demonstrating the presence of protofibrils in a preincubated or "aged" preparation. Transition to amyloid fibril Ui 6.0 M glycerol showing the formation of elongated and rigid fibers (panel B) that fiequently exhiited a helicai twisting (anowheads). Scale bar is 100 ML

To obtain a quantitative measure of aggregation under our experimentai

conditions, cenhifugation was employed using AB IabeIed with the fluorescent probe N7-

(7-nitro benz-2-oxa-I,3-diazol-4-yl)ethyleedie (NBD) . Our previous tlw rescence

resonance energy W e r (FRET) studies have indicated that the low concentrations of

the fluorophore-labeled A$ do not i n t e r e with the kinetics of fibril formation or the

morphology of the redting aggregates (Huang et al., 1997). In the current study, similar

peptide solutions containhg excess uniabeled (10 IiM) and NBD-labeled AB (0.1 @A)

were used. Prior to centrifugation, the fiuorescence intwsity increased Linearly with

increasing concentrations of glycerol, with the fluorescence at 6 M glycerol being

approximately twice that of the buf5er only sample (Figure 4.5). This increase was due to

glycerol-induced changes in the quantum yield of NBD as shown by the response of the

unconjugated label under these conditions. Following centrifugation, a 97% decrease in

fluorescence was observed indicating that the peptides are almost completely aggregated.

This was the case for the glycerol-containing samples as well as the initial aqueous

solution which is consistent with the amorphous aggregates obsenred by AFM. This is

likely due to the fact that the samples were not pretreated, for example, by filtration to

remove non-specific aggregates which would have examined the more direct conversion

of solubIe monomer to Ab aggngate. However, the resuits presented here with untreated

samples suggest that the conversion to the Psheet particles, that are induced by solvation

changes, rnay involve a shuttüng of rnonomers from the unordered structure to the

nucleating aggregates. An analogous situation may occur with the diffuse-to-senile

plaque AB interconversion that is seen in vivo and the use of chemicd chaperones may be

useftd in modehg this aspect of the amyloid pathway.

Glyceroi concentration (molar)

Figure 4.5. Insolubility of the unstnictured and fibrillar aggregates monitored by a fluorescent labeled AB peptide tracer. Peptides hcubated in aqueous phosphate b a e r or with increasing concentrations of glycerol were subjected to high speed centrifugation (135 000 x g). Recipitation of peptide was observed in aii cases as shown by the loss of fluorescence observed in the initial solutions (open cucles) as compared to the supernatant fluorescence following centrifugation (closed circles). The increase in fluorescence at higher gIyceroI concentrations is due to changes in the probe as shown by similar changes in unconjugated label (open sqgms).

Cumulatively these resuits demonstrate that changes in the solvation state of the

Af! peptide affects severai aspects of the amyloidogenic pathway. In general terms,

amyloid nbrils are initiated by the destabïhtion of a n o d cellular protein that leads to

a partially unfolded intermediate. This can be accelerated by point mutations as has been

shown for lysozyme (Booth et al., 1997) and transthyretin (Lai et al., 1996) which are

deposited as systemic plaques. If uncorrected, the do lded intermediate will dtimately

assemble into the 8-sheet aggregates that initiate fibril formation. It has been proposed

that chernical chaperones may, in part, regdate amyloid formation by stabilizuig the

lower energy native conformer to reduce the levels of unfolded proteins that are required

for the amyloidogenic pathway. This is consistent with the observation that naturaily

occturing organic osmolytes inhibit the conversion of the cellular prion protein (P~P') to

the protease resistant and amyloid forming prpSC associated with transmissible

encephalopathies (DebBunnan et al., 1997; Tatzelt et al., 1996). Similar stabilization by

giycerol of misfolded mutant transmembrane proteins within the endoplasmic reticulum

has been demonstrated for the chloride transporter associated with cystic fibrosis (Brown

et al., 1996; Sato et al., 1996) and aquaporins related to nephrogenic diabetes

(Tamarappoo and Verkman, 1998).

Although stabilizing native conformations of larger proteins c m affect amyloid -

formation, it is likely that AB does not have a conventional conformation and may

therefore exist in an unfolded and possibly amorphous state. Therefore, ifunchecked by

cellular control elements, the peptide wiU gradually deposit as diffuse and eventually

senile plaques. Our results indicate that solvation changes induced by chemicd

chaperones such as the naturaiIy occirning TMAO and the in vitro mode1 provided by

glycerol, rapidly accelerate the major steps in the amyloidogenic pathway. These include

both the early nucleation and conformational events as weil as the protofibril-to-fiber

conversion. This may be unique to proteins that normaily exist in polymerk foms as

show by chernical chaperone-mediated assembly of tubulin into microtubules and their

subsequent resistance to urea denaturation (Sackett, 1997). The significance of our

findings is that they reveai additional contributors to Ab amyloid formation and therefore

provide an additional tool to manipulate this pathway. Such information codd

potentially be used to develop or accelerate cellular models of AB aggregation and the

assesment of agents that modulate fibril formation.

Blake, C.C.F. and Serpeil, L. (1996) Synchrotron x-ray studies suggest that the core of

the transthyretin amyloid fibd is a continuous beta-sheet heiix. Structure 4,989-998.

Booth, D.R, Sunde, M., Bellotti, V., Robinson, C.V., Hutchinson, W.L., Fraser, P.E.,

Hawkins, PN., Dobson, CM., Radford, S.E., Blake, C.C.F. and Pepys, M.B. (1997)

Mability, unfolding and aggregation of human lysozyme variants underlying amyloid

fibnllogenesis. Nature 385,787093.

Brown, C.R., Hong-Brown, L.Q., Biwersi, J., Verkman, A.S. and Welch, W.J. (1996)

Chemicd chaperones correct the mutant phenotype of the delta F508 cystic Eibrosis

transmembrane conductance regdator protein. Celi Stress Chaperones 1, 1 17-25.

Brown, C.R., Hong-Brown, L.Q. and Welch, WJ. (1997) Correctkg temperature-

sensitive protein folding defects. I. Clin. Invest. 99, 1432- 1444.

Burg, M.B. (1995) MolecuIar basis of osmotic regdation. Am. I. Physiol. 268,983-996.

Cook, DG., Forman, M.S., Sung, K., Leight, S., KO~SOQ~ DL., Iwatmbo, T, Lee, V.M.

and Doms, R.W. (1997) Alzheimer's A beta(l-42) is generated in the endoplasmic

reticuldmtermediate cornpartment of NT2N ce&. Nature Med. 3,102 1 - 1023.

DebBurman, S.K., Raymond, G.J., Caughey, B. and Lindquist, S. (1997) Chaperone-

s u p e ~ s e d conversion of prion protein to its protease-tesistant form. Proc. Nad. Acad.

Sci. USA 94, 13938-43.

Evans, K.C., Berger, ER, Cho, C.G., Weisgraber, KH. and Lansbury, P.T., Jr. (1995)

ApoIipoprotein E is a kinetic but not a thermodynamic inhibitor of amyloid formation:

implications for the pathogenesis and treatment of Alzheimer disease. Proc. Natl. Acad.

Sci. USA 92,763-767.

Garzon-Rodriguez, W., Sepulveda-Becerra, M., Milton, S., and Glabe, C.G. (1997)

Soluble amyloid Abeta(1-40) exists as a stable dimer at low concentrations. J. Biol.

Chem. 272,21037-21044.

Harper, J.D., Wong, S.S., Lieber, C.M. and Lansbury, P.T. (1997a) Observation of

metastable Abeta amyloid protofibnls by atomic force microscopy. Chem. Biol. 4, 1 19-

125.

Harper, J.D., Lieber, C.M. and Lansbury, P.T., Jr. (199%) Atomic force rnicroscopic

imaging of seeded fibril formation and fibril branching by the Alzheimer's disease

amyloid-beta protein. Chem. Biol. 4,95 1-959.

Hartmann, T., Bieger, S.C., Bruhi, B., Tienari, PJ., Ida, N., Allsop, D., Roberts, G.W.,

Masters, CL., Dotti, CG., Unsicker, K. and Beyreuther, K. (1997) Distinct sites of

intraceiIular production for Alzheimer's disease Abeta40/42 amyloid peptides. Nature

Med. 3,1016-1020.

Huang,T.H.J., Fraser, P.E., and Chakrabartty, A. (1997) Fibriflogeaesis of Alzheimer

beta-amyloid peptides studied by tluorescence energy transfer. J. Mol. Biol. 269, 214-

224.

Inouye, H., Fraser, P.E. and Kirschner, D.A. (1993) Structure of beta-crystallite

assemblies fomed by Alzheimer beta-amyloid protein analogues: analysis by x-ray

difkction. Biophys. 1. 64,502-5 19.

Janett, J.T., Berger, E.P. and Lansbury, P.T., Jr. (1993) The carboxy terminus of the beta

amyIoid protein is criticd for the seedmg of amyloid formation: implications for the

pathogenesis of Alzheimer's disease. Biochemistry 32,469304697.

Kidevsky, R and Fraser, P.E. (1997) A beta arnyloidogenesis: unique, or variation on a

systemic theme? Cd. Rev. Biochem. Mol. Biol. 32, 361-404.

Kowaiewski, T.Z., and HoItzman, D. M. (1999) In situ atomic force microscopy study of

Alzheimer's betaamyloid peptide on dif5erent substrates: new insights mto mechanism of

beta-sheet formation. Proc. Nad. Acad Sci. USA 96,3688-3693.

Lai, Z., Colon, W. and Kelly, J.W. (1996) The acid-mediated denaturation pathway of

tmsthyretin yields a confionnational intemediate that can self-assemble into amyloid.

Biochemistry 35,6470-6482.

Lambert, M.P., Barlow, A.K., Chromy, B.A., Edwards, C., Freed, R., Liosatos, M.,

Morgan, T.E., Rozovsky, I., Trommer, B., Viola, K.L., Wals, P., Zhang, C., Finch, C.E.,

Krafft, G.A., and Klein, W.L. (1998) Difhible nonfibdlar ligands derived fiom Abetal-

42 are potent central nervous systern neurotoxins. Proc. Nat/. Acad Sci USA 95, 6448-

6453.

Lomakin, A., Teplow, DB., Kirschner, D.A. and Benedek, G.B. (1997) Kinetic theory

of fibrillogenesis of amyloid beta-protein. Proc. Nad. Acad. Sci. USA 94,7942-7947.

M c L a u ~ , J., Franklin, T., Chakrabartty, A. and Fraser, P.E. (1998) Phosphatidylhositol

and inositol hvolvement in Alzheimer amyloid-beta fibril growth and a r r e a J. Mol. Biol.

278,183-1 94.

Roher, A.E., Chaney, M.O., Kuo, Y.M., Webster, S.D., Stine, W.B., Haverkamp, L.J.,

Woods, A.S., Cotter, RJ., Tuohy, J.M., Kmfft, G.A., Bonnell, B.S. and Emmerhg, M.R

(1 996) Morphology and toxicity of AB(1-42) dimer denved fiom neuritic and vascular

amyloid deposits of Alzheimer's disease. J. Biol. Chem. 271,2063 1-20635.

Sackett DL. (1997) Nahrral osmolyte trimethylamine N-oxide stimulates tubulin

polymerkation and reverses urea inhibition. Am. J. Physiol. 273, R669-R676.

Sato, S., Ward, CL., Krouse, M E , Wine, J.J. and Kopito, ILR. (1996) Glycerol reverses

the misfolclhg phenotype of the most cornmon cystic fibrosis mutation. J. Biol. Chem.

271,635-638.

Schmechel, D.E., Saunders, A.M., Strittmatter, W.J., Crain, B.J., Hulette, C.M., Joo,

S.H., Pericak-Vance, M.A., Goldgaber, D. and Roses, A.D. (1993) Increased amyloid

beta-peptide deposition in cerebral cortex as a coasequence of apolipoprotein E genotype

in Iate-onset Alzheimer disease. Proc. Natl. Acad. Sci. USA 90,9649-9653

Seikoe, D.J. (1997) Alzheimer's disease: genotypes, phenotypes, and treatments. Science

275,630-63 1.

Skovronslq, D.M., Doms, R.W. and Lee VM. (1998) Detection of a novel intraneuronai

pool of insoluble amyloid beta protein that accumulates with time in culture. J. Cell Biol.

I41,IO3I-1039.

Stine, W.B., Jr., Snyder, S. W., Laciror, U.S., Wade, W.S., Miller, MF., Perm, T.J.,

Hoban, T.F. and Krafft, GA. (1996) The nonometer-scale structure of amyloid beta

vimalized by atomic force microscopy. J. Protein Chem. lS,l93-203.

Sunde, M., Serpell, L.C., Bartlam, M., Fraser, PX., Pepys, M., and Blake, C I . (1997)

Common core structure of amyloid fibrils by Synchrotron X-ray ciBiaction. S. MOI. BioI.

273,7290739,

Tamarappoo, B.K. and Verkman, A.S. (1998) Defective aquaporin-2 traficking in

nephrogenic diabetes insipidus and correction by chernical chaperones. J. Clin. Invest.

10 1,2257-2267

Tatzelt, J., Pmi.net, S.B. and Welch, W.J. (1996) Chernical chaperones interfere with the

formation of scrapie prion protein. EMBO J. 1 5,6363-63 73.

Timasheff, S.N. (1993) The control of protein stability and association by weak

interactions with water: how do solvents affect these processes? Annu. Rev. Biophys.

Biomol. Stmct. 22,67-97.

Walsh, D.M., Lomakin, A., Benedek, G.B., Condron, M.M. and Teplow, Da. (1 997)

Amyioid B-protein fibrillogenesis: detection of a protofibrillar intermediate. I. Biol.

Chem. 272,22364-22372

Wang, A. and Bolen, D.W. (1997) A nahtrally o c c ~ g protective system in urea-rich

cells: mechanism of osmolyte protection of proteins against urea denaturation.

Biochemistry 36,9101-9108.

Wdd-Bode, C., Yamazaki, T., Capell, A., Leimer, U., Steiner, H., Ihara, Y. and Haass, C.

(1997) IntraceUular generation and accumulation of amyloid beta-peptide terminating at

amino acid 42. J. Biol. Chem. 272,16085-16088.

Chapter 5: Discussion

Summaty

Within the process of fibrilIogenesis, A$ foms a wide variety of intermediate

structures and aitemate products depending on extemal conditions. Factors Like pH, total

AB peptide concentration, and the effects of extrinsic molecules have significant impact

on the structure of AB and its aggregates.

In Chapter 2, we saw that AB cm adopt two soluble oligomeric states depending

on pH. At pH 3, AB associates to fonn a 0.94 Mda particle, while at pH 7 AB forms a

dimerketramer. At pH 5, AB forms large insoluble aggregates. The 0.94 Mda particle and

the dimedtetramer are stable for at least 4 weeks. Cellular cornpartments have distinct pH

levels. Lysosomes are pH 3-5 and endosomes are pH 5-6. Extracellular fluid is pH 7.

Intracellular processing of APP and trafficking of AB through the various compartments

may have profound effects on their structure and biological activity. That soluble A$

forms can be isolated, and that they are stable for extended periods of tirne, permits the

use of high resolution biophysical methods and wiIl aid in the elucidating the

mechanisms of fibrillogenesis.

Chapter 3 described the establishment of an in vifro system which will ailow for

detailed characterization of amorphous AB. The findings Erom that work indicate peptide

concentration is one factor controlhg whether AB aggregates into an amorphous or

fibrillar form. Deposits of amorphous amyloid appear before deposits of fibrillar amyloid.

The possibility exists that in AD and DS it is the inabzty to clear AB and the consequent

build-up in AB to a point in excess of the criticai concentration for fibrillogenesis, which

is responsible for nbrillar amyloid a c c d a t i o n and senile plaque formation in the brain.

Beyond the intrinsic propeaies of AB itseE, extrinsic factors such as proteins

associateci with amyloid plaques may exert important effects on the structural aad/or

aggregational state of A$. For example, ApoJ has been reported to alter AB association to

favor the formation of soluble oligomers called ADDLs (Lambert et al., 1998). Also, AB

rnetabolism may take place within intracellular compartments containing protein

chaperones or chaperone elements which control protein folding. Thus, it would be

important to learn the effects of plaque-associated proteins and proteins that control

protein folding inside the relevant intracellular compartments. TMAO and glycerol

mimic naturally occuning chaperone proteins in that they stabilize native conformations

and polymerk States. These two osmolytes could be used to study their effects upon the

structure and stability of A$. The results of this work were described in Chapter 4.

TMAO and glycerol accelerated AB conformational change Eom randorn coi1 to B-sheet.

This transition accompanied the conversion of morphous aggregates to uniform globular

structures. These molecules also mediated the transformation of protofibrils to mature

fibrils. In vitro techniques used here can be applied to investigation of physiologically

relevant amyloidassociated protek.

Discussion

It is not difiicdt to appreciate the complexity of the fibril formation process, one

chmcterized by a multitude of intermediate steps and intncate interrelationships among

products. For a researcher in this field, the task of dissecting this process is a damting

one. This complexity, compounded by the challenges of working with a problematic

peptide, has driven researchers to focus the investigation on the starting point, AB, and on

the end-point, the AB nbril. Though much is yet to be done, a great deal has been

accomplished. We now have appreciable knowledge of the physicochemical properties of

the AB peptide and the s t r u c ~ and hctional characteristics of the AB fibril.

Tt is now necessary to extend the research to explore the areas in between the end-

points. As this happens, our understanding of fibrillogenesis will evolve f?om a simplined

description of discrete steps to a more sophisticated insight into the constellation of

dynamically inter-related products and side-products. Presently, soluble intermediates,

such as AB dime~s and ADDLs, are being characterized and their positions and roles in

the fibriI formation pathway are becoming better understood. Building upon this

knowledge was the pizmary goal of the present work. The approach taken was to first

identiQ conditions where intermediate and alternate products of fibrillogenesis could be

stabilized and then to develop methods to study and characterize these products. It would

then be possible to better defme the relationships among these states. To reach this level

of understanding, to know how each step in the process affects the others, would be

essential in developing treatments for AD especially with those strategies which attempt

to modulate AB directIy.

Some of the approaches to tackling the problem of AB in AD are: disaggregation

of AB deposits with antibodies against the AB peptide, modulation of APP processing and

inhibition of AB fibril formation.

Solomon et al. (1997) fomd that site-directed monoclonal antibodies agaiost AB

were able to disaggregate £%ds in vifru. Schenk et al. (1999) immunized PDAPP mice, a

mouse mode1 for AD which overexpresses mutant human APP and develops key

neuropathological featrnes of AD, with AB. They fomd that immunkation prevented

plaque formation, neuntic dystrophy and astrogliosis. These reports indicate an

immunological approach to AD therapy is one which holds promise.

With the identification of the eaymes involved in the & and y-secretase activities

of APP processing, another viable therapeutic stnitegy is to moddate APP proteolysis to

reduce or arrest AB production. [nhibitors were used in linkuig y-secretase activities to

PSI and PS2, demonstrating this type of approach is tractable. However, many

challenges, such as the delivery of such dnigs to the brain, have yet to be overcome.

Thus, other methods must also be examined.

If the toxic f o m of AB c m be clearly identifed, then an inhibitor which prevents

the formation of such an aggregate can be developed. Tjemberg et al. (1996) reported a

pentapeptide fiagrnent of AB was capable of binding full length AB and preventing fibril

formation. This approach is a sensible one, as it would be logical that a sequence region

of AB would have auto-affiity which binds AB but would lack the regions responsible

for continued polymerization. An alternative method is to find compounds that are

unrelated to A$ but exert inhibitory effects on fibrillogenesis. To this end, Wood et al.

(1 996) showed that hexadecyl-N-methylpipendinium (HMP) bromide selectively binds

AB aud prevents its assernbly into fibrils.

The successful deplopent of inhibitors as therapeutic agents depends a great

deai upon a refined understanding of AB polymerization. Such knowledge wili permit

targeting of the most relevant pathway intermediates in terms of toxicity. It would also

prevent incorrect targeting which may a d y d t in the production of higher Ievels of

toxic products. Because nbfiogenesis is a multi-step process, many intermediates exist.

Some of these intemediates, soluble oligomers for example, may be the more relevant

toxic species. Inhilition at one point in the pathway may disntpt the equilibrium arnong

species and actually inmase the population of the toxic intermediates. These are some of

the considerations which must be understood if an effective inhibitor-based therapy is to

be attained,

Future Work

An important step in the future work arising from this project is to expand the

studies to include AB42 because, of al1 the AB f o m , this peptide is the more toxic and

fibrillogenic. AB42 is difficult to synthesize and an effective protocol for making this

peptide has not yet been established in our Iab. Addressing this problem will necessarily

be the subject of future work.

Now that methods are in place to preferentidy stabilize oligomenc AB40, the

toxicity of these structures can be tested in neurotoxicity assays. A rat

pheochromocytoma cell line, PC-12, c m be used for this purpose. PC-12 cells cm be

induced to differentiate into b'neun>n-Iike" cens with the addition of nerve-growth factor.

AB40 oligomers can be added to these cultures and membrane disruption can be detected

using the lactate dehydrogenase release assay, while cell death can be foliowed using the

sulfhydryirhodamine B assay. After the initiai screening for toxicity with PC-12 ceus,

hippocampd neurons isolated from rat embryos can be used for confirmation of the

toxicity redts within a more biologically relevant context

It may also be possible to apply muitidimensionai NMR spectroscopy to elucidate

the 3D structures of AB40 oligomers. The large size of the 0.94 MDa particles would

likely make them UIISUitabIe for this type of d y s i s . The dimedtetramrs present at pH

7, on the other hand, are srna11 enough for NMR spectroscopy. TraditionaIly, the

minimum peptide concentration limit for NMR (1 mM) has Iunited its usefulness in

studying AB, as aggregates quickiy form at these concentrations. However, advances in

this field have made it possible to collect spectra fiom samples with peptide

concentrations as low as LOO pM. At this concentration, dimersltetnuners may be stable

for the duration of the NMR spectral acquisition. Finding conditions where high quality

NMR data c m be collected would be the key challenge in this endeavour. This task

wouid reap great rewards, because a detailed structural description wouid provide iasight

into the structure-function relationship of toxic AB40 aggregates and aid significantly in

the design of therapeutic antagonists.

The total concentration of AB is a factor which detennines whether AB associates

to fom diffuse AB or fibrillar AB. It would be useful to leam how diffuse AB converts to

fibnllar amyloid. A fmt step wouid be to elucidate the t h e dependence of this process.

The fluorescence of NBD-labeled AB40 can be followed as a hc t ion of time. As the

fluorophore is increasingly sequestered fkom water with the conversion to fibrils,

fluorescence will increase. Additionally, fluorescence polarization cm be monitored as a

firnction of time with polarization increasing as fibrils form. Kinetic models can be fitted

to the data and rate constants can be determined,

Molecules such as TMAO have signincant effects on fÏbril formation. The same

holds true for plaque-associated proteins. For example, ApoE promotes fibril formation

whiIe A2M inhibits nbril formation. Such proteins may play roles in converthg diffuse

AB to fibds. In studying their effects on AB, the presence of such proteins in the sampies

would interfere with techniques suc6 as analyticai ultracentrifugation and CD

spectroscopy. However, fluorescence would iikely remain unaffected. The change in

fluorescence concomitant with the evolution of fïbrillar structure nom diffuse arnyloid

can be foIIowed to study the effects of the proteins. These effects cm also be quantitated

with rate constants by fitting kinetic models to the fluorescence data.

It is aiso possible that plaque-associated proteins could affect the formation of

soluble AB oligomers. To follow the formation of difise aggregates versus soluble

oligomers versus fibnls the acceptoddonor ratio fiom FRET' experiments can be utiiized.

Due to the ciifferences in structure and peptide packhg of these aggregate types, the

separation distances between fluorophores could be distinct and would be reflected in the

acceptor/donor ratio. For example, the oligomers formed at pH 3 and pH 5 have an

acceptor/donor ratio of 18 and 17, respectively, at 48 @d, while the dimerltetramer at pH

7 has a ratio of 1.36 at 50 pbl (Figure. 2.3B and C). The formation of each aggregate

could be distinguished based on the acceptor/donor ratio and the effects of plaque-

associated proteins on the formation of these aggregates can be monitored.

The continuation of the work described in this thesis will assist in answerïng some

of the many questions still remaining regarding AB and will aid in the search for AD

therapies.

Lambert, M.P., Barlow, AX., Chromy, B.A., Edwards, C., Freed, R., Liosatos, M.,

Morgan, T.E., Rozovsiq, I., Trommer, B.. Viola, K.L., Wals, P., Zhang, C., Finch, CE.,

Krafft, G.A., and Klein, W.L. (1998) D i h i b l e nonfibdlar Ligands derived fiom AB142

are potent central nervous system neurotoxhs. Proc. Nad. Acad. Sci. USA 95, 6448-

6453 .

Schenk, D., Barbour, R, Dum, W., Gordon, G., Grajeda, H., Guido, T., Hu, K., Huang,

J., Johnson-Wood, K., Khan, K., Kholodenko, D., Lee, M., Liao, Z., Lieberburg, I.,

Motter, R., Mutter, L., Soriano, F., Shopp, G., Vasquez, N., Vandevert, C., Waker, S.,

Wogulis, M., Yednock, T., Games, D e T and Seubert, P. (1 999) Nature 400,173-177.

Solomon, B., Koppel, R., Ftankel, D., and Hanan-Aharon, E. (1997) Disaggregation of

Alzheimer B-amyloid by site-directed mAb. Proc. Natl., Acad. Sci. U.S.A. 94, 4109-

41 12.

Tjernberg, L.O., Niblund, J., Lindpvist, F., Johansscn, J., KarIstr6m, A.R., Thyberg, J.,

Terenius, L., and Nordstedt, C. (1996) Arrest of fhmyloid ftbril formation by a

pentapeptide Ligand. J. Biol. Chem. 271,8545-8548.

Wood, S.J., MaclCemie, L., Mdeeff, B., Hurle, M.R., and Wetzel, R. (1996) Selective

inhibition of A$ fîbd formation. J. Biol. Chem. 271.40864092.