Fo

DO*†

R

hbc(ottpitaopiTsmflphtb

c

erlbn

Hr

8

t

Biochemical and Biophysical Research Communications 285, 58–63 (2001)

doi:10.1006/bbrc.2001.5123, available online at http://www.idealibrary.com on

0CA

luorescence Anisotropy: A Method for Early Detectionf Alzheimer b-Peptide (Ab) Aggregation

avid Allsop*,1 Linda Swanson,† Susan Moore,* Yvonne Davies,* Amber York,*,2

mar M.A. El-Agnaf,* and Ian Soutar†Department of Biological Sciences, Lancaster University, Lancaster LA1 4YQ, United Kingdom; andDepartment of Chemistry, Dainton Building, University of Sheffield, Sheffield S3 7HF, United Kingdom

eceived May 30, 2001

(Ab), a 39–43 residue peptide derived from proteolyticpmlvfnnotthcnptpsobfitgptplg(iseal

M

Ma

Time-resolved anisotropy measurements (TRAMS)ave been used to study the aggregation of the-amyloid (Ab) peptide which is suspected of playing aentral role in the pathogenesis of Alzheimer’s DiseaseAD). The experiments, which employ small quantitiesf fluorescently-labelled Ab, in addition to the un-agged peptide, have shown that the sensitive TRAMSechnique detects the presence of preformed “seed”articles in freshly prepared solutions of Ab. More

mportantly, as 100 mM solutions of Ab containingagged Ab at a concentration level of either 0.5 or 1 mMre incubated, the TRAMS prove capable of detectionf the peptide aggregation process through the ap-earance of a continuously increasing “residual an-

sotropy” within the time-resolved fluorescence data.he method detects Ab aggregation in its earliesttages, well before complexation becomes apparent inore conventional methods such as the thioflavin Tuorescence assay. The TRAMS approach promises torovide a most attractive route for establishment of aigh-throughput procedure for the early detection ofhe presence of amyloid aggregates in the screening ofiological samples. © 2001 Academic Press

Key Words: Alzheimer’s disease; amyloid; Ab; fluores-ein; time-resolved fluorescence anisotropy.

Alzheimer’s disease (AD), which leads to dementia andventual death, is characterized by the presence of neu-ofibrillary tangles and senile plaques in the brain. Theatter lesions are composed of amyloid fibrils surroundedy dystrophic and dying neurons. The principal compo-ent of the highly insoluble amyloid fibrils is b-amyloid

Abbreviations used: AD, Alzheimer’s disease; Ab, b-amyloid peptide;FIP, 1,1,1,3,3,3-hexafluoro-2-propanol; TRAMS, time-resolved anisot-

opy measurements; FRET, fluorescence resonance energy t.1 To whom correspondence should be addressed. Fax: 44 (0)1524

43854. E-mail: d.allsop@lancaster.ac.uk.2 Present address: University of Maryland, Baltimore County, Bal-

imore, MD 21250.

58006-291X/01 $35.00opyright © 2001 by Academic Pressll rights of reproduction in any form reserved.

rocessing of the larger amyloid precursor protein. Accu-ulating evidence supports the hypothesis that Ab fibril-

ogenesis is a seminal pathogenic event in AD (for re-iews see (1, 2)). One of the supporting pieces of evidenceor this “amyloid hypothesis” is the fact that Ab is toxic toeurons, both in vivo and in vitro (2). The development ofeurotoxicity appears to be connected with the formationf aggregates of Ab, but the precise molecular organiza-ion of the neurotoxic form of Ab and the mechanism ofoxicity have not been defined, although the generation ofydrogen peroxide may play a role in toxicity (3, 4). Re-ent attention has focussed on the suggestion that small,onfibrillar oligomers or “protofibrils” may represent aarticularly toxic species (5–7). Inhibiting or reversinghe early stages of Ab oligomerization is thus seen an aossible therapeutic strategy for AD, and developing sen-itive methods for the detection of early oligomeric formsf Ab is an important objective. Various techniques haveeen employed to monitor aggregation of Ab and otherbril-forming proteins and peptides (8–11), but most ofhese methods lack the sensitivity required to detect ag-regation in its very early stages. Fluorescence methodsrovide both sensitivity and selectivity, and have a dis-inct advantage over other techniques in that it shouldrove possible to detect the binding of a fluorescently-abelled protein to a preformed, nontagged protein aggre-ate, which could have important diagnostic implications12). In this paper, we elaborate upon our earlier prelim-nary reports (13, 14) that time-resolved anisotropy mea-urements (TRAMS) can detect Ab aggregation in itsarliest stages and represent, arguably, one of the mostttractive routes for establishing high-throughput amy-oid aggregation assays.

ATERIALS AND METHODS

Peptides. The nonlabelled Ab 1-40 peptide was synthesized on ailligen 9050 peptide synthesizer starting from Val-PEG-PS resins

nd using Fmoc N-protection, as described previously (15). Synthetic

Ab 1-40 fluo-peptide, bearing a fluorescein tag, was obtained fromN

flpstcamflmdmaTt3L(msaitagwws

orccgm

m3NPa4

R

tlfmvi

opmvitsi

flflwmotcr

wpolt

opaottsdmcfl3otauta

mttfl7a

Vol. 285, No. 1, 2001 BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS

ew England Nuclear Life Science Products (Boston, MA).

Fluorescence spectroscopy and anisotropy measurements. Theuo-peptide was initially dissolved in 1,1,1,3,3,3-hexafluoro-2-ropanol (HFIP) in order to remove any preformed aggregates, andtored frozen at 280°C in small working aliquots. These werehawed immediately before use, and the HFIP was evaporated byentrifugation under vacuum. The fluo-peptide was then dissolved invery small volume of 0.1 M acetic acid, before the addition of 100M Tris buffer, pH 7.4, to give a final concentration of 1–2 mMuo-peptide (nonaggregation control). In the aggregation experi-ents, an equal volume of nonlabelled Ab 1-40, freshly dissolved in

istilled water, was added to the nonaggregation control to give aixture which was either 0.5 or 1 mM with respect to fluo-peptide

nd 100 mM in non-labelled peptide, in 50 mM Tris buffer, pH 7.4.ime-resolved anisotropy measurements were taken after incuba-ion of the control or aggregating peptides for various time periods at7°C. Fluorescence spectra were obtained using a Perkin ElmerS50 spectrometer. Time-resolved anisotropy measurements

TRAMS) were carried out on a specially modified Edinburgh Instru-ents 199 time-correlated single photon counter. The excitation

ource used was an IBH nanoLED 05 light emitting diode (LED) withrepetition rate of 1 MHz and a FWHM of 1.1 ns. Vertically polar-

zed light was used to excite the sample and fluorescence intensitiesransmitted by a polarizer were analyzed in planes parallel, I \ (t),nd perpendicular, I'(t), to that of the polarized excitation. A “tog-ling procedure” was employed in the collection of I \ (t) and I'(t)herein the orientation of the analyzer was sequentially alteredhile memory quarters in the multichannel analyzer were switched

imultaneously.

Negative stain electron microscopy (EM). One microliter samplesf nonlabelled Ab 1-40 (100 mM in 50 mM Tris buffer, pH 7.4) wereemoved after incubation at 37°C, mixed with 4 ml of 2 mM octylglu-oside and 5 ml of 1% uranyl acetate, applied to formvar/carbonoated grids, and left for 10 min. Excess fluid was blotted off, and therid was air dried before examination in a Jeol JEM-1010 electronicroscope.

Thioflavin T assays. Samples (15 ml) of nonlabelled Ab 1-40 (100M in 50 mM Tris buffer, pH 7.4) were removed after incubation at7°C and added to 2 ml of 10 mM thioflavin T in 50 mM Glycine/aOH, pH 9.0. Duplicate samples were scanned three times in aerkin Elmer LS50 fluorescence spectrometer before and immedi-tely after the addition of peptide (16). Results are for excitation at50 nm and emission at 482 nm.

ESULTS AND DISCUSSION

Fluorescence anisotropy can be used to determinehe rate at which a fluorescent species tumbles in so-ution. Since the rate of reorientation of a particle is aunction of its size, fluorescence anisotropy measure-ents, particularly time-resolved experiments, offer a

ery sensitive and informative means of interrogatingnteractions between molecules.

A time-resolved anisotropy experiment uses pulsesf polarized light for excitation of the fluorescent sam-le. The probability of absorption of this radiation by aolecule depends upon the orientation of its transition

ector with respect to the plane of polarization of thencident excitation. Consequently, molecules are pho-oselected according to the extents to which their ab-orption axes are aligned parallel to the plane of polar-zation at the moment of incidence. The resultant

59

uorescence is polarized, its degree of anisotropy re-ecting, at a given time after excitation, the extent tohich molecular randomization has occurred throughicroBrownian motion. The rate of rotational motion

f the fluorescent species can be quantified, throughhe decay of emission anisotropy, r(t), in terms of aorrelation time tc. For a spherical rotor, the anisot-opy decay may be expressed as

r~t! 5 roexp~2t/tc!, [1]

here r o is the intrinsic anisotropy, a characteristicroperty of a given fluorophore. The larger the volumef the particle (and its associated solvent sheath), theonger it takes to reorient in solution, and the greaterhe value of its tc.

A commercially available, fluorescently labelled formf synthetic Ab(1–40) was used in our anisotropy ex-eriments. This Ab variant bears a fluorescein markerttached to a cysteine residue substituted at position 7f the amino acid sequence. The fluo-peptide wasreated with 1,1,1,3,3,3,-hexafluoro-2-propanol (HFIP)o disperse any pre-formed aggregates (17). Figure 1hows that the anisotropy of the fluorescence from ailute solution (2 mM) of fluo-peptide on its own, im-ediately after solubilization (Tris buffer, pH 7.4), de-

ayed rapidly to zero (tc ; 1 ns), indicating that theuo-peptide molecules tumble rapidly in solution at7°C. Incubation of this fluo-peptide solution for peri-ds of up to 6 days resulted in no significant change inhe anisotropy decay, indicating that no detectable self-ssociation or aggregation of the fluo-peptide occursnder these conditions. This is to be expected becausehe threshhold concentration for nucleation-dependentggregation of Ab 1-40 is in the range 10–40 mM (18),

FIG. 1. Anisotropy decay of Ab 1-40 (100 mM normal peptide 1 1M fluo-peptide) following incubation for various times at 37°C. Theraces in order of increased ‘r’ values after anisotropy decay are dueo 21 mM fluo-peptide only (r decays rapidly to zero) and then 1 mMuo-peptide plus 100 mM normal peptide after incubation for 0, 1, 24,2, 96, and 120 h. (Note: These are “raw” anisotropy decay curvesnd contain some “contamination” from scattered light).

and pretreatment of Ab with HFIP significantly re-twahuu

(gmhftc

sbttdthcactaswavtaccflsoatgtpottpmglnR

characterized by tc values greatly in excess of the flu-o(ft

laii2brtHwvsidtTifanrof

ss

Irmcttwtrpt

lccptoeTp

Vol. 285, No. 1, 2001 BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS

ards the aggregation process (17). Glabe and co-orkers (19) using analogous fluorescent Ab variantsnd fluorescence resonance energy transfer, FRET,ave concluded that under conditions similar to thosesed here, the peptide exists as stable dimers. Indeed,sing the equation

tc 5hVRT , [2]

where h is solvent viscosity, V is molar volume, R isas constant, and T is absolute temperature) we esti-ate that the rotor (if assumed to be a spherical entity)as a radius of the order of 1.5 nm, which is reasonableor a dimeric form of the peptide, and is consistent withhe hydrodynamic radius of 1.8 6 0.2 nm for dimeric Abalculated from quasielastic light-scattering data (20).

On addition of unlabelled Ab (1–40) to freshly dis-olved fluo-peptide (to give a solution of 100 mM unla-elled peptide and 1 mM fluo-peptide), we observedhat the anisotropy, r, no longer attained zero withinhe time range over which the fluorescence could beetected (Fig. 1). These changes in the appearance ofhe anisotropy decays are similar to those which weave observed upon mixing varying amounts of 2 inter-omplexing, synthetic macromolecules {poly(acryliccid) and poly(dimethylacrylamide)} one being fluores-ently tagged (21, 22). In the latter case, the changes inhe anisotropy decay profiles reflected the inhibition,s aggregation occurred, of the motion of the chainegments of the two water-soluble polymers, each ofhich exists in the form of a flexible coil, in dilutequeous solution. The complex formed, however has aery rigid structure, characterized by a very long rota-ional correlation time, in solution, which is presum-bly that for molecular reorientation of the entire spe-ies (21, 22). In the case of the peptide mixture, it islear that, as soon as mixing is effected, some of theuo-peptide becomes incorporated into a larger, morelowly relaxing species. It would appear that a fractionf the fluo-peptide immediately complexes with smallmounts of preformed aggregate (or seed) present inhe 100 mM solution of unlabelled peptide. These ag-regates must be large enough for the emission fromhe bound fraction of the fluo-peptide to suffer no de-olarization during the time (ca. 30 ns) for which flu-rescence is observable, their presence being charac-erized by the appearance of a “residual anisotropy” inhe time-resolved decay profiles. Samples of normaleptide taken at “time zero” for negative stain electronicroscopy (EM) (see Fig. 2a) revealed the presence of

lobular structures 8–24 nm in diameter, and slightlyarger ellipsoids with dimensions of 16–28 by 20–32m (similar to those described by Nybo et al. (23)).otation of particles of these dimensions would be

60

rescein’s excited state lifetime. Consequently, eitheror both) of these species could represent the pre-ormed seed present in the unlabelled peptide to whichhe fluo-peptide immediately becomes attached.

Upon incubation of the mixture of labelled plus un-abelled peptide, the contribution from the residualnisotropy, r`, to the overall anisotropy decay profilencreased progressively with time (Fig. 1). After 1–7 hncubation, the peptide was seen by EM (Figs. 2b andc) to assemble into larger chain-like groups, whichecame fibrillar in structure, and displayed elementsesembling so-called “protofibrils” (6, 7, 20) and even-ually (after .24 h) “mature” amyloid fibrils (Fig. 2d).owever, the increases in r` were apparent evenithin the first 30 min of incubation. In contrast, con-entional thioflavin T fluorescence assays (16) oneparate samples of 100 mM nonlabelled Ab 1-40,ncubated under similar buffer and temperature con-itions, proved incapable of detecting peptide aggrega-ion at times of incubation which were less than 7 h.hioflavin T is a reagent, the fluorescence from which

s enhanced when it adsorbs on to a hydrophobic sur-ace (16). It is commonly used to detect the presence ofggregates formed from amyloidogenic peptides. Asoted above, significant changes in thioflavin T fluo-escence above background levels (P , 0.05) occurrednly after 7 h incubation of the peptide (Fig. 3) in trialsor comparison with the TRAMS experiments.

The time-resolved anisotropy of the fluorescence ob-erved from the fluo-peptide was found to be well de-cribed by a function of the form,

r~t! 5 r1exp~2t/tc1! 1 r`. [3]

n this expression, tc1 is the correlation time for Abotational relaxation and r o 5 r 1 1 r`. Use of thisodel function resulted in the consistent recovery of a

omponent, tc1, of the order of 0.7–1.0 ns. In addition,he value of r` increased with the incubation time ofhe system until the latter approached ca. 2000 h, byhich time abundant fibrillar structures could be de-

ected by EM. Thereafter, the relative contribution of` to the overall anisotropy decay of r(t) decreased,resumably because, at this stage, the aggregated pep-ide rapidly fell out of suspension.

The lifetime, tf , of the excited state of the fluoresceinabel (ca. 4 ns) was unaffected by the formation ofomplexes with the unlabelled peptide throughout theourse of the incubation experiments. In addition, re-lacement of the fluo-peptide by an equivalent concen-ration of free fluorescein did not result in the devel-pment of a measurable anisotropy in the fluorophore’smission in an incubating mixture over a 48 h period.hus, the effects produced by addition of the unlabelledeptide to solutions of its fluo-peptide analogue are not

di

taltmal

Itrataohwrqa

vsaa

Vol. 285, No. 1, 2001 BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS

riven by any tendency of the fluorescein dye itself tonteract with the Ab peptide aggregates.

Under the incubation conditions adopted here, onlywo types of fluorescent species are detectable in thenisotropy data; the “uncomplexed” Ab dimer andarge aggregates. In general, for a system containingwo species of differing dimension but having a com-on fluorescence lifetime, tf , the resultant emission

nisotropy would exhibit a dual exponential decayaw, i.e.,

R~t! 5 r1exp~2t/tc1! 1 r2exp~2t/tc2!. [4]

FIG. 2. Ab 1-40 aggregates observed by negative stain electronarious time periods. (a) Time 0 h. Small globular or ellipsoid structtructures take on a more linear form (arrowheads). A few protofilamppear more fibrillar, and greater numbers of protofilaments (arrowspproximate diameter of 8 nm. Prefibrillar structures and protofilam

61

f, as in the current case, one of the species is so largehat tc2 @ tf , the second term appears to be constant, at2 (5 r`). For diagnostic purposes this is a bonus:ggregation is detectable, at the very earliest times,hrough the appearance and growth of this residualnisotropy. For studies of the kinetics and mechanismf peptide oligomerization and assembly, on the otherand, use of a much longer-lived luminescent markerould be advantageous, since this would allow the

otational relaxation of the large aggregates to beuantified. The present observations show that nucle-tion of a growth centre is followed by rapid peptide

croscopy. A 100 mM solution of peptide was incubated at 37°C fors are seen to from groups or chains (arrowheads). (b) Time 1 h. Thes are observed (arrows). (c) Time 3 h. Linear structures (arrowhead)e seen. (d) Time 24 h. Numerous amyloid fibrils are seen, having ants are less evident. Scale bar 100 nm.

miureent) aren

clt

oeonm1aaiQbgtectmwbmscapsfflpcmtcabaa

Elucidation of the molecular events that initiate oli-gossuc

A

efi

R

1

1

vaw

Vol. 285, No. 1, 2001 BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS

omplexation to form aggregates of significant size (ateast 25 nm across), which coexist with Ab dimers inhe incubating mixture.

Fluorescence anisotropy has clear advantages overther techniques for detecting and monitoring thearly stages of amyloid oligomerization. Past studiesn the kinetics of Ab aggregation have utilized tech-iques such as turbidity, thioflavin T binding, sedi-entation, Congo red binding, and HPLC (8–11, 16,

8). These methods provide information on the appear-nce of high molecular weight aggregates, or the dis-ppearance of soluble peptide, but early intermediatesn the oligomerization process cannot be detected.uasielastic light scattering spectroscopy (QLS) haseen used to monitor the various phases of Ab aggre-ation (20, 24), but cannot be used to study early in-ermediates of oligomerization of Ab when they are inquilibrium with high molecular weight forms. Re-ently, size-exclusion gel filtration has been also usedo study Ab oligomerization (6, 7, 20). However, thisethod detects only the formation of a high moleculareight gel-excluded fraction, shown to contain protofi-rils, but no early intermediates. The anisotropyethod can detect oligomerization in its very early

tages, and, furthermore, the specificity of the fluores-ent tagging should allow the detection of preformedggregates in biological samples. In this respect, theotential of the approach has already been demon-trated by analysis of samples of cerebrospinal fluidrom patients with Alzheimer’s disease by means ofuorescence correlation spectroscopy (12). The princi-al advantage of fluorescence anisotropy over fluores-ence correlation spectroscopy is that the formerethod should be much simpler to adapt for high-

hroughput analysis. This is also where the fluores-ence anisotropy approach scores over FRET, becausehigh-throughput drug screening or diagnostic assay

ased on fluorescence anisotropy would use smallermounts of fluorescently-tagged peptide so that thessay would be cheaper to run.

FIG. 3. Ab 1-40 aggregation as monitored by a standard thiofla-in T fluorometric assay. Graph shows the mean effect of peptideddition (from incubation of 100 mM peptide at 37°C) to thioflavin T,ith 95% confidence limits, at each time point.

62

omerization, is crucial to understanding the pathol-gy and developing potential therapeutics for AD. Thistudy provides a new strategy for detecting the earlytages of oligomerization, and for studying the molec-lar mechanisms, kinetics and inhibition of this pro-ess.

CKNOWLEDGMENTS

We thank The Alzheimer’s Disease Society, The Parkinson’s Dis-ase Society UK (Dr. Omar El-Agnaf), and Lancaster University fornancial support, and Peter Diggle for statistical advice.

EFERENCES

1. Hardy, J., and Allsop D. (1991) Amyloid deposition as the centralevent in the aetiology of Alzheimer’s disease. Trends Pharmacol.Sci. 12, 383–388.

2. Selkoe, D. J. (1999) Translating cell biology into therapeuticadvances in Alzheimer’s disease. Nature Suppl. 399, A23–A31.

3. Huang, X., Atwood, C. S., Hartshorn, M. A., Multhaup, G., Gold-stein, L. E., Scarpa, R. C., Cuajungco, M. P., Gray, D. N., Lim, J.,Moir, R. D., Tanzi, R. E., and Bush, A. I. (1999) The Ab peptideof Alzheimer’s disease directly produces hydrogen peroxidethrough metal ion reduction. Biochemistry 38, 7609–7616.

4. Turnbull, S., Tabner, B. J., Moore, S., Davies, Y., El-Agnaf,O. M. A., and Allsop, D. (2001) a-synuclein implicated in Parkin-son’s disease catalyses the formation of hydrogen peroxide invitro. Free Radical Biol. Med. 30, 1163–1170.

5. 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, J. M.,Krafft, G. A., Bonnell, B. S., and Emmerling, M. R. (1996) Mor-phology and toxicity of Ab-(1-42) dimer derived from neuritic andvascular amyloid deposits of Alzheimer’s disease. J. Biol. Chem.271, 20631–20635.

6. Walsh, D. M., Hartley, D. M., Kusumoto, Y., Fezouli, Y., Con-dron, M. M., Lomakin, A., Benedek, G. B., Selkoe, D. J., andTeplow, D. B. (1999) Amyloid b-protein fibrillogenesis; structureand biological function of protofibrillar intermediates. J. Biol.Chem. 274, 25945–25952.

7. Hartley, D. M., Walsh, D. M., Ye, C. P., Diehl, T., Vassilev, P. M.,Teplow, D. B., and Selkoe, D. J. (1999) Protofibrillar intermedi-ates of amyloid b-protein induce acute electrophysiologicalchanges and progressive neurotoxicity in cortical neurons.J. Neurosci. 19, 8876–8884.

8. El-Agnaf, O. M. A., Guthrie, D. J. S., Walsh, D. M., and Irvine,G. B. (1998) The influence of the central region containing resi-dues 19–25 on the aggregation properties and secondary struc-ture of Alzheimer’s b-amyloid peptide. Eur. J. Biochem. 256,560–569.

9. El-Agnaf, O. M. A., Bodles, A. M., Guthrie, D. J. S. Harriott, P.,and Irvine, G. B. (1998) The N-terminal region of non-Ab com-ponent of Alzheimer’s disease amyloid is responsible for its ten-dency to assume b-sheet and aggregate to form fibrils. Eur.J. Biochem. 256, 560–569.

0. Wood, S. J., Wetzel, R., Martin, J. D., and Hurlle, M. R. (1995)Prolines and amyloidogenicity in fragments of the Alzheimer’speptide Ab/A4. Biochemistry 34, 724–730.

1. Clements, A., Allsop, D., Walsh, D. M., and Williams, C. (1996)Aggregation and metal-binding properties of mutant forms of theamyloid Ab peptide of Alzheimer’s disease. J. Neurochem. 66,740–747.

12. Pitschke, M., Prior, R., Haupt, M., and Riesner, D. (1998) Detec-

1

1

1

1

1

1

seeding in Alzheimer’s disease and scrapie: Mechanistic truths

1

2

2

2

2

2

Vol. 285, No. 1, 2001 BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS

tion of single amyloid b-protein aggregates in the cerebrospinalfluid of Alzheimer’s patients by fluorescence correlation spectros-copy. Nature Med. 4, 832–834.

3. Allsop, D., Swanson, L., Moore, S., Davies, Y., El-Agnaf, O. M. A.,and Soutar, I. (2000) Sensitive detection of b-amyloid aggrega-tion by means of time-resolved fluorescence anisotropy. Neuro-biol. Aging 21, S20 (abstract).

4. Allsop, D., Twyman, L., Davies, Y., Moore, S., York, A., Swanson,L., and Soutar, I. (2001) Modulation of b-amyloid production andfibrillisation. Biochem. Soc. Symp. 67, 1–14.

5. El-Agnaf, O. M. A., Goodwin, H., Sheridan, J. M., Frears, E. R.,and Austen, B. M. (2000) Improved solid-phase syntheses ofamyloid proteins associated with neurodegenerative diseases.Pro. Pep. Lett. 7, 1–8.

6. Naiki, H., Higuchi, K., Hosokawa, M., and Takeda, T. (1989)Fluorometric determination of amyloid fibrils in vitro using thefluorescent dye, thioflavin T. Anal. Biochem. 177, 244–249.

7. Szabo, Z., Klement, E., Jost, K., Zarandi, M., Soos, K., andPenke, B. (1999) An FT-IR study of the b-amyloid conformation:Standardization of aggregation grade. Biochem. Biophys. Res.Commun. 265, 297–300.

8. Harper, J. D., and Lansbury, P. T. (1997) Models of amyloid

63

and physiological consequences of time-dependent solubility ofamyloid proteins. Ann. Rev. Biochem. 66, 385–407.

9. Garzon-Rodriguez, W., Sepulveda-Becerra, M., Milton, S., andGlabe, C. G. (1997) Soluble amyloid Ab-(1-40) exists as a stabledimer at low concentrations. J. Biol. Chem. 272, 21037–21044.

0. Walsh, D. M., Lomakin, A., Benedek, G. B., Condron, M. M., andTeplow, D. B. (1997) Amyloid b-protein fibrillogenesis: Detectionof a protofibrillar intermediate. J. Biol. Chem. 272, 22364–22372.

1. Soutar, I., and Swanson, L. (1990) Fluorescence anisotropy stud-ies of polyelectrolyte mobility and interpolyelectrolyte complex-ation in aqueous solution. Macromolecules 23, 5170.

2. Soutar, I., Swanson, L., Thorpe, F. G., and Zhu, C. (1996) Fluo-rescence studies of the dynamic behaviour of poly(dimethylac-rylamide) and its complex with poly(methacrylic acid) in dilutesolution. Macromolecules 29, 918.

3. Nybo, M., Svehag, S.–E., and Holm Nielsen, E. (1999) An ultra-structural study of amyloid intermediates in Ab 1-42 fibrillogen-esis. Scand. J. Immunol. 49, 219–223.

4. Lomakin, A., Teplow, D. B., Kirschner, D. A., and Benedek, G. B.(1997) Kinetic theory of fibrillogenesis of amyloid b-protein.Proc. Natl. Acad. Sci. USA 94, 7942–7947.