, , , , , , ,

, , , ,

, , , , , , ,

, , , ,

*Department of Pathology, The University of Melbourne, Parkville, Victoria, Australia

�Centre for Neuroscience, The University of Melbourne, Parkville, Victoria, Australia

�Mental Health Research Institute, Parkville, Victoria, Australia

§Department of Clinical Neurosciences, St Vincent’s Health, Fitzroy, Victoria, Australia

¶Bio21 Molecular Science and Biotechnology Institute, The University of Melbourne, Parkville, Victoria, Australia

Accumulation of amyloid-b (Ab) is a critical factor in thedevelopment of Alzheimer’s disease (AD). Ab is a majorconstituent of the amyloid deposits that characterise the ADbrain (Glenner and Wong 1984a,b; Masters et al. 1985;Selkoe et al. 1986), and many reports describe the mecha-nisms by which Ab contributes directly to neuronal failure(see Crouch et al. 2008 for review). Targeting Ab accumu-lation for treating AD has strong therapeutic potential, andunderstanding the factors that determine Ab levels in thebrain is therefore an area of intense research focus.

Steady-state levels of Ab are determined by relative ratesof synthesis and degradation. In healthy subjects fractionalsynthesis and clearance rates for Ab in cerebrospinal fluid areestimated at 7.6% and 8.3% per hour (Bateman et al. 2006)indicating that although Ab is produced in healthy subjects,

its accumulation is prevented by the activity of functionalclearance mechanisms. Factors contributing to increased Abproduction have been widely reported, including cases of

Received August 15, 2008; revised manuscript received November 21,2008; accepted December 11, 2008.Address correspondence and reprint requests to Dr Peter J. Crouch,

PhD, Department of Pathology, The University of Melbourne, Parkville,Vic. 3010, Australia. E-mail: pjcrouch@unimelb.edu.auAbbreviations used: AD, Alzheimer’s disease; APP, amyloid precursor

protein; Ab, amyloid-b; CHO, Chinese Hamster Ovary; fAD, familialAD; IDE, insulin degrading enzyme; MMP, matrix metalloprotease;MPACs, metal-protein attenuating compounds; NEP, neprilysin; PBS,phosphate-buffered saline; PS1, presenilin 1; SDS, sodium dodecylsulfate; SRCD, synchrotron radiation circular dichroism; TCNB, NaCl,Tris, CaCl2 and Brij-35; ThT, thioflavin-T.

Abstract

Accumulation of neurotoxic amyloid-b (Ab) is central to

the pathology of Alzheimer’s disease (AD). Elucidating the

mechanisms of Ab accumulation will therefore expedite the

development of Ab-targeting AD therapeutics. We examined

activity of an Ab-degrading protease (matrix metalloprotease

2) to investigate whether biochemical factors consistent with

conditions in the AD brain contribute to Ab accumulation by

altering Ab sensitivity to proteolytic degradation. An Ab amino

acid mutation found in familial AD, Ab interactions with zinc

(Zn), and increased Ab hydrophobicity all strongly prevented

Ab degradation. Consistent to all of these factors is the pro-

motion of specific Ab aggregates where the protease cleavage

site, confirmed by mass spectrometry, is inaccessible within

an amyloid structure. These data indicate decreased degra-

dation due to amyloid formation initiates Ab accumulation by

preventing normal protease activity. Zn also prevented Ab

degradation by the proteases neprilysin and insulin degrading

enzyme. Treating Zn-induced Ab amyloid with the

metal-protein attenuating compound clioquinol reversed

amyloid formation and restored the peptide’s sensitivity to

degradation by matrix metalloprotease 2. This provides new

data indicating that therapeutic compounds designed to

modulate Ab-metal interactions can inhibit Ab accumulation

by restoring the catalytic potential of Ab-degrading proteases.

Keywords: Alzheimer’s disease, amyloid-b, oligomer, prote-

olysis, therapeutic, zinc.

J. Neurochem. (2009) 108, 1198–1207.

JOURNAL OF NEUROCHEMISTRY | 2009 | 108 | 1198–1207 doi: 10.1111/j.1471-4159.2009.05870.x

1198 Journal Compilation � 2009 International Society for Neurochemistry, J. Neurochem. (2009) 108, 1198–1207� 2009 The Authors

familial AD (fAD) that involve mutations to genes encodingthe amyloid precursor protein (APP) or the secretases thatcleave Ab from APP (Chartier-Harlin et al. 1991; Goateet al. 1991; Citron et al. 1992; Games et al. 1995; Sher-rington et al. 1995). However, relatively few reports describefactors that cause Ab accumulation via decreased rates ofdegradation. Several proteases have the potential to degradeAb (Carson and Turner 2002; Eckman and Eckman 2005),and some reports provide evidence that decreased proteaseexpression contributes to Ab accumulation in vivo. Forexample, neprilysin (NEP) mRNA levels in the post-mortemAD brain are lowest in regions of the brain most susceptibleto plaque formation (Yasojima et al. 2001a,b), and NEPexpression and activity correlates inversely with Ab depo-sition in the brain (Fukami et al. 2002; Miners et al. 2006;Farris et al. 2007; Hellstrom-Lindahl et al. 2008). In addi-tion, Ab levels are increased in the phosphate-buffered saline(PBS)-soluble fraction of brains from matrix metalloprotease(MMP) knockout mice (Yin et al. 2006), and other reportshave shown a correlation between decreased functionality ofinsulin degrading enzyme (IDE) and AD (Kim et al. 2007;Zhao et al. 2007). Importantly, elevating expression ofspecific proteases using a range of therapeutic strategieshas been shown to decrease Ab levels (Marr et al. 2003,2004; White et al. 2006; Hemming et al. 2007; Jiang et al.2008), and targeting Ab-degrading enzymes as a potentialtherapeutic strategy for AD has therefore been proposed(Nalivaeva et al. 2008).

In addition to decreased expression of Ab-degradingproteases, Ab accumulation in the AD brain may be the resultof decreased Ab sensitivity to protease activity, possibly dueto decreased access of the protease to its catalytic site on thesubstrate. The implication of this scenario is that Abaccumulation could proceed even in the presence of functionalproteases. This has considerable significance for the develop-ment of AD therapeutics because the efficacy of strategies thataim to increase protease expression will be determined bywhether the Ab substrate is amenable to protease-mediateddegradation. In the present study, we examined the hypothesisthat biochemical factors consistent with conditions in the ADbrain alter the Ab substrate’s sensitivity to degradation.

Experimental procedures

MaterialsRecombinant human MMP, NEP, and IDE were purchased from

R&D Systems (Minneapolis, MN, USA). Synthetic Ab1–40 (Ab40)and Ab1–42 (Ab42) were purchased from W.M. Keck Foundation

Biotechnology Resource Laboratory (Yale University, New Haven,

CT, USA). Synthetic Ab1–42 with glutamate-22 substituted with

glutamine (Ab42E22Q) was prepared using methods described

previously (Barnham et al. 2003). WO2 and G2–11 antibodies to

Ab were prepared as described previously (Ida et al. 1996).

Streptavidin-labeled europium was purchased from PerkinElmer

Life Sciences (Melbourne, Vic., Australia). All other chemicals were

from Sigma-Aldrich (Castle Hill, NSW, Australia).

Recombinant proteasesMatrix metalloprotease-2 at 100 lg/mL in TCNB buffer [150 mM

NaCl, 50 mM Tris, 10 mM CaCl2, 0.05% (v/v) Brij-35, pH 7.5] was

incubated at 37�C for 1 h with 4-aminophenylmercuric acetate

(prepared in dimethylsulfoxide) added to a final concentration of

1 mM.NEPwas prepared in 25 mMTris, 100 mMNaCl (pH 8.0) and

IDE in 25 mM Tris, 150 mM NaCl (pH 7.5), both at 100 lg/mL.

Before use, the protease preparations were diluted with TCNB buffer

as required. Control reaction mixtures contained equivalent volumes

and concentrations of all chemicals and reagents minus the protease.

Synthetic and biological Ab preparationsSynthetic Ab was dissolved in 1,1,1,3,3,3-hexafluoro-2-propanol

and its concentration determined by absorbance spectrometry and

reference to an extinction co-efficient (Ciccotosto et al. 2004).

Aliquots were dried by vacuum centrifugation and stored at )20�C.Dried, 1,1,1,3,3,3-hexafluoro-2-propanol-treated Ab samples were

reconstituted with 20 mM NaOH then diluted with PBS (145 mM

NaCl, 7.5 mM Na2HPO4, 0.24 mM NaH2PO4, pH 7.2) to 100 lM.

Freshly prepared 100 lM Ab solutions were either: (i) diluted

further with PBS to 10 lM and used immediately, (ii) aged at 37�Cfor 1, 5 or 24 h before diluting to 10 lM with PBS, (iii) diluted to

10 lM with PBS then aged at 37�C for 6 or 24 h. For some

experiments, 10 lM Ab42 was prepared using PBS containing

ZnCl2 to give a final Ab : Zn molar ratio of 1 : 1. In other

experiments, Ab : Zn solutions aged at 37�C were treated with

10 lM clioquinol before using. Conditioned media containing Abwas collected from Chinese Hamster Ovary epithelial cells

expressing human APP695 (CHO-APP cells) as described previ-

ously (White et al. 2006). Mouse brain extracts containing Abwere

prepared from 8-month old APP/presenilin 1 (PS1) mice (Jankow-

sky et al. 2001) as described previously (Crouch et al. 2005).

Protease and Ab reaction mixturesWhen using synthetic or brain-derived Ab, the Ab/protease reactionmixtures contained 6.7 lL TCNB buffer, 1.1 lL protease, and

2.2 lL Ab sample. When using CHO-APP conditioned media

samples the reaction mixtures contained 1.1 lL protease and 8.9 lLconditioned media. Zero–100 ng MMP was used as specified and

reaction mixtures were incubated at 37�C for 0–120 min. For all

experiments with NEP or IDE, the reaction mixtures contained

25 ng protease and were incubated at 37�C for 60 min. For some

experiments the reaction mixtures were supplemented with ZnCl2 or

CuCl2 at the concentrations shown. Reactions were terminated as

described below depending on the Ab detection methods used.

SDS–PAGE and western blot analysis of AbReaction mixtures were terminated by adding 4· denaturing sample

buffer [100 mM Tris, 10% (v/v) glycerol, 4% (w/v) sodium dodecyl

sulfate (SDS), 4% (v/v) b-mercaptoethanol, 0.01% (w/v) brom-

ophenol blue] as follows: for synthetic Ab assays 2 lL reaction

mixture was added to 35.5 lL MQ water and 12.5 lL 4· sample

buffer; for brain Ab assays 10 lL reaction mixture was added to

3.3 lL 4· sample buffer; for CHO-APP media assays 10 lL

� 2009 The AuthorsJournal Compilation � 2009 International Society for Neurochemistry, J. Neurochem. (2009) 108, 1198–1207

Therapeutic restoration of Ab degradation | 1199

reaction mixture was added to 5 lL MQ water and 5 lL 4· sample

buffer. Terminated reaction mixtures were incubated at 100�C for

5 min, proteins resolved on 10% Bis–Tris Novex gels (Invitrogen,

Carlsbad, CA, USA), then transferred to polyvinyldene fluoride

membranes. Membranes were probed with WO2 (Ida et al. 1996)and Ab bands visualised by enhanced chemiluminescence (Lumigen

TMA-6; GE Healthcare, Notting Hill, Victoria, Australia). Ab bands

were quantified by densitometry (IMAGEJ 1.38· software, NIH,

Bethesda, MA, USA) to determine protease-mediated Ab degrada-

tion (expressed as %Ab degraded compared to Ab levels in

reactions mixtures that contained no protease).

Surface enhanced laser desorption ionisation time of flight massspectrometryReaction mixtures (10 lL) were terminated with 10 lL 8 M urea

containing 10% (v/v) Triton X-100, then loaded onto PS10 activated

ProteinChips (Ciphergen Biosystems, Fremont, CA, USA) pre-

coupled with the WO2 antibody for Ab. ProteinChips were

incubated at 20�C for 3 h, then washed to remove unbound

proteins, and incubated twice with 1 lL sinapinic acid (50%

saturated solution in 1 : 1 acetonitrile : trifluoroacetic acid). WO2-

immunoreactive proteins bound to the ProteinChips were detected

using a PBSIIc surface enhanced laser desorption ionisation time of

flight mass spectrometer (Ciphergen Biosystems). Peaks were

analysed using CIPHERGEN PROTEINCHIP software 3.1.

Double antibody capture ELISA for AbReaction mixtures were terminated by adding the MMP inhibitor

1,10-phenanthroline to a final concentration of 10 mM and Ablevels determined using double antibody capture ELISA (White

et al. 2006). Protease-mediated Ab degradation (expressed as %Abdegraded) was determined relative to Ab levels in reactions mixtures

that contained no protease.

MMP activity assays using artificial MMP substrateMatrix metalloprotease (17 ng) was added to 200 lL reaction

mixtures containing 10 lM Mca-PLGL-Dpa-AR substrate (R&D

Systems) in TCNB buffer then incubated at 37�C for 2 h. ZnCl2 or

CuCl2 were included in the reaction mixture at 0–100 lM by

supplementing the TCNB buffer. MMP-mediated cleavage of the

substrate in the leu-gly domain removes the fluorescence-quenching

effect of the Dpa moiety and therefore enables Mca fluorescence to

be measured at 360 nm excitation and 405 nm emission. Ab-mediated substrate degradation is expressed relative to maximal

fluorescence obtained for reactions mixtures that contained no

additional metal ion.

Synchrotron radiation circular dichroism spectroscopySynchrotron radiation circular dichroism (SRCD) was carried out on

station 12.1 at the Daresbury Synchrotron Radiation Source. SRCD

data for Ab42 and Ab42E22Q were collected at 37�C in a 0.5 mm

fused silica cell with an Ab concentration of 20 lM. Spectra were

collected between 170 and 260 nm using 1 nm increments and a

dwell time of 1 s.

Electron microscopyAmyloid-b1–42 solutions aged at 37�C in the presence of ZnCl2 for

24 h then either used directly for electron microscopy or first treated

with 10 lM clioquinol. Alternatively, 10 lM Ab42 and Ab42E22Qsolutions were aged for 6 h at 37�C before using for electron

microscopy. Ab solutions were spotted onto carbon-coated copper

grids, allowed to incubate for 30 min, then excess solution removed.

Samples were stained with 0.5% (w/v) uranyl acetate and analysed

on a Siemens 102 transmission electron microscope operating with a

voltage of 60 kV.

Thioflavin-T assayThe amyloid content of Ab solutions was determined using the

thioflavin-T (ThT) assay described previously (Smith et al. 2007).The effects of clioquinol on Zn-induced amyloid formation were

determined by adding clioquinol to 10 lM before adding the ThT.

Statistical analysesAll experiments were performed with 3–8 replications. Abdegradation values (mean ± SEM) are expressed relative to control

reactions that contained no MMP. Statistical significance for the

amount of Ab degraded or the effect of a treatment on Abdegradation and/or amyloid formation was determined using the

Student’s T-test (hypothesised difference of zero). p-Values shown infigures indicate the level of significance.

Results

Ab is rapidly degraded by MMPTo establish appropriate assay conditions for MMP-mediatedAb degradation we determined time- and dose-dependentparameters for the reaction mixtures. Increasing the Ab/MMP reaction time up to 120 min facilitated an increase indegradation of monomeric Ab (Fig. 1a) and increasing theprotease concentration increased the amount of Ab degraded(Fig. 1b). ELISA analyses (Fig. 1c) showed good consis-tency with the western blot results (Fig. 1a and b), with bothdetection methods indicating 65–80% Ab degradation whenreaction mixtures containing 25 ng protease were incubatedfor 60 min. Consistent with previous reports (Roher et al.1994; Backstrom et al. 1996; Yan et al. 2006) surfaceenhanced laser desorption ionisation time of flight massspectrometry analysis showed MMP-mediated generation ofan Ab fragment with a mass equal to Ab1–34 (Ab34),indicating cleavage between Ab residues leu34 and met35(Fig. 1d).

Serum-free conditioned media from CHO-APP cells wasincubated with or without MMP for 60 min. WO2-immuno-reactive Ab species with an apparent mass equal tomonomeric Ab were detected by western blot (not shown)and the presence of MMP decreased abundance of theapparent Ab monomer by �55% (Fig. 1e). Similarly, theaddition of MMP to reaction mixtures containing APP/PS1mouse brain extract decreased abundance of the apparent Abmonomer by �50% (Fig. 1e). Notably, 100 ng MMP wasrequired to degrade Ab in brain extracts to the same extentthat 25 ng MMP degraded Ab in CHO-APP conditioned

Journal Compilation � 2009 International Society for Neurochemistry, J. Neurochem. (2009) 108, 1198–1207� 2009 The Authors

1200 | P. J. Crouch et al.

media. This may reflect the overall higher protein content ofthe mouse brain extracts and therefore the presence ofadditional protease substrates.

Aged Ab is resistant to degradationBefore adding to the Ab/MMP reaction mixture we aged100 lM Ab40 and Ab42 preparations at 37�C for 0, 1, 5 or24 h. Increasing the age of the Ab preparations decreased theamount of monomeric Ab degraded (Fig. 2a), indicating atime-dependent alteration to the Ab substrate that decreasedits sensitivity to degradation. Sensitivity of the shorter Ab40substrate to degradation also decreased with aging of thepeptide (Fig. 2a), but this effect was less evident compared tothe longer Ab42. As Ab is an amyloidogenic peptide thatoligomerises and forms higher order structures with time, weexamined whether these higher order structures were sensi-tive to degradation. To obtain Ab preparations expected tocontain some monomeric Ab as well as a range of higherorder structures we aged 100 lM Ab42 for 5 hrs at 37�Cbefore incubating for 60 min with or without MMP. The

aged Ab42 preparations contained Ab monomers, dimers,trimers, and tetramers, as well as less discrete species rangingfrom 35 to 100 kDa (collectively called high molecularweight aggregates) (Fig. 2b). MMP decreased abundance ofAb monomers, but did not affect Ab dimers, trimers,tetramers or high molecular weight aggregates (Fig. 2c).All of the higher order Ab structures shown in Fig. 2b areSDS-stable, indicating covalent cross-linking which may bethe basis of their observed resistance to MMP2. Alterna-tively, the SDS-stable dimers, trimers, and tetramers may besensitive to proteolytic degradation but partitioned withinlarger SDS-soluble Ab aggregates in the Ab solution, andthereby protected from protease activity. Establishing this,however, is beyond capacity of our current experimentalconditions.

Ab42E22Q is less sensitive to degradationDecreased degradation due to aging of the Ab peptide (Fig. 2)is consistent with time-dependent formation of proteaseresistance Ab amyloid. To test this directly, we compared

Fig. 1 Time- and dose-dependent degra-

dation of Ab. (a) Representative western

blot image and densitometry data showing

degradation of synthetic Ab42 after incu-

bating with 25 ng MMP2 for 0–120 min. (b)

Representative western blot image and

densitometry showing degradation of syn-

thetic Ab42 after incubating with 0–100 ng

MMP2 for 60 min. (c) ELISA assay for Ab42

after incubating with 25 ng MMP2 for

60 min in the presence or absence of the

MMP inhibitor 1,10-phenanthroline. (d)

Surface enhanced laser desorption ionisa-

tion time of flight mass spectrometry spec-

tra showing MMP2-dependent generation

of a product with a mass equal to Ab34.

Scale of Y-axis is equal for both spectra.

Average m/z value for peaks labeled

Ab42 = 4513 (n = 8 spectra, Ab42 mass =

4515), and peak labeled Ab34 = 3787

(n = 8 spectra, Ab34 mass = 3787). (e)

Densitometry analysis of western blot ima-

ges shows MMP2 degrades Ab in condi-

tioned media from CHO-APP cells when

incubated with 25 ng MMP2 for 60 min, as

well as Ab in APP/PS1 mouse brain ex-

tracts when incubated with 100 ng MMP for

60 min. Ab degradation values (mean ±

SEM, n = 3–6) are expressed relative to

control reactions that contained no MMP.

p-Values in (e) indicate Ab degradation

relative to respective controls.

� 2009 The AuthorsJournal Compilation � 2009 International Society for Neurochemistry, J. Neurochem. (2009) 108, 1198–1207

Therapeutic restoration of Ab degradation | 1201

wild-type Ab42 and mutant Ab42E22Q for rates of amylo-idogenesis and their subsequent sensitivity to degradation. TheE22Q substitution found in Dutch familial AD has been shownto accelerate the formation of Ab aggregates and amyloid(Murakami et al. 2003). We therefore aged wild-type Ab42and Ab42E22Q at 37�C for 0 or 6 h before determining Absecondary structure, amyloid content, and susceptibility todegradation. SRCD analysis of freshly prepared Ab solutionsrevealed Ab42 was present predominantly as random coil(strong negative signal at 197 nm) with small amounts of a-helix and b-sheet (smaller negative signal between 210 and220 nm) compared to the Ab42E22Q spectrumwhich showedpredominantly b-sheet conformation (positive signal at192 nm and a broad negative signal at 215 nm) (Fig. 3a).Despite the difference in secondary structure, ThT fluores-cence indicated the levels of amyloid content were notsignificantly different in freshly prepared peptide solutionswhen using 10 lM solutions (Fig. 3b, 0 h incubation) as perthe subsequent Ab degradation assays, or when using 20 lMsolutions as per the SRCD assays. At 20 lM, Ab42 ThTfluorescence = 29.1 ± 10.4 FU and Ab42E22Q fluores-cence = 32.0 ± 2.2 (p = 0.401). With aging, the amyloidcontent for both Ab42 and Ab42E22Q increased, but this was

more pronounced for Ab42E22Q (Fig. 3b, 6 h incubation).Electron microscopy analyses indicate the increased ThTsignal for Ab42E22Q after 6 h aging was due to acceleratedfibril formation (Fig. 3c), and a comparison of the peptides’sensitivity to MMP after aging for 0 or 6 h revealed theaccelerated rate of amyloid formation in AbE22Q coincidedwith decreased susceptibility to degradation (Fig. 3d).

Clioquinol reverses Zn-induced Ab resistance todegradationPrevious studies have shown Zn and Cu differentially affectAb aggregation and amyloidogenesis (Bush et al. 1994;Atwood et al. 1998; Smith et al. 2007). We found Zn, butnot Cu, induced a dose-dependent increase in Ab resistanceto MMP2 at low lM concentrations (Fig. 4a). The absenceof any inhibitory effects for Zn when measuring degradationof the Mca-PLGL-Dpa-AR substrate (Fig. 4a) indicate theeffects seen for Ab were substrate-specific, and not due todirect inhibition of MMP2. By contrast, comparable effectsfor high lM Cu towards degradation of both Ab and theMca-PLGL-Dpa-AR substrate indicate relatively weak,direct inhibition of MMP2 (Fig. 4a). Similar effects observedfor Zn on Ab degradation by NEP and IDE (Fig. 4b) indicate

Fig. 2 Aging Ab decreases the peptide’s

susceptibility to degradation by MMP2. (a)

Densitometry analysis of western blot re-

sults showing relative rates of Ab42 and

Ab40 degradation. Age of Ab preparation

(0, 1, 5, or 24 h) indicates the time 100 lM

Ab preparations were incubated at 37�Cprior to diluting to 10 lM then incubating

with MMP2 for 60 min. (b) Western blot

image showing MMP2 degrades mono-

meric Ab but not low molecular weight

oligomers or high molecular weight (HMW)

aggregates. Ab preparations (100 lM) were

aged for 5 h at 37�C before diluting to

10 lM and incubating with MMP2 for a

further 60 min. HMW aggregates are arbi-

trarily defined as non-discrete immunore-

activity between �35 and 100 kDa. The two

separate western blot panels shown are

from the same membrane exposed for dif-

ferent periods of time. (c) Densitometry

analysis of western blot results shown in

(b). Ab degradation values (mean ± SEM,

n = 3–4) are expressed relative to control

reactions that contained no MMP.

Journal Compilation � 2009 International Society for Neurochemistry, J. Neurochem. (2009) 108, 1198–1207� 2009 The Authors

1202 | P. J. Crouch et al.

the mechanism for increased protease resistance is notspecific for MMP2. Aging Ab42 in the presence ofequimolar Zn induced the formation of ThT-positive amyloidstructures, and consistent with the capacity for metal-proteinattenuating compounds (MPACs) to resolubilise Ab aggre-gates (Huang et al. 1997), we were able to show that Zn-induced Ab amyloid formation could be reversed using theMPAC clioquinol (Fig. 4c and d). Importantly, Zn-inducedAb amyloid was relatively resistant to degradation (16%degraded), but the addition of clioquinol increased Abdegradation to 58% (Fig. 4e), therefore partially restoringsensitivity to degradation.

Discussion

Central to AD pathology is cerebral accumulation of Ab,and the development of therapeutics to decrease brain Ablevels has therefore received intense research focus (Mastersand Beyreuther 2006). However, decreasing Ab levels in thebrain requires an understanding of the mechanisms that

contribute to its accumulation. In fAD, mutations to APP orthe secretases that cleave Ab from APP directly increase therates of Ab production (Chartier-Harlin et al. 1991; Goateet al. 1991; Citron et al. 1992; Games et al. 1995;Sherrington et al. 1995). Therefore, in the case of fAD,Ab accumulation appears to be the result of productionexceeding degradation. Supporting this, and indicating thatAb homeostasis may be restored by elevating proteaseexpression levels, some studies have shown that brain Ablevels in mice over-expressing mutant APP and/or presenilin1 can be decreased by elevating levels of Ab-degradingproteases (Marr et al. 2003; Hemming et al. 2007; El-Amouri et al. 2008). Alternatively, localised Ab accumula-tion in the AD brain appears to correlate with regions whereexpression of the Ab-degrading protease NEP is lowest(Yasojima et al. 2001a,b). Collectively, these studies indi-cate Ab accumulation may be the result of increased Abproduction and/or decreased expression of Ab-degradingproteases. However, very few studies have examined howfactors directly affecting the Ab substrate may contribute to

Fig. 3 Increased amyloidogenesis makes Ab42E22Q less susceptible

to degradation by MMP2. (a) Synchrotron radiation circular dichroism

spectra showing freshly prepared Ab42E22Q exhibits characteristic b-

sheet secondary structure compared to Ab42 which exhibits pre-

dominantly random coil with small amounts of a-helix and b-sheet. (b)

ThT fluorescence showing that although freshly prepared 10 lM Ab42

and Ab42E22Q solutions contain comparable levels of Ab amyloid, the

amyloid content of Ab42E22Q increases more rapidly than wild-type

Ab42 when aged at 37�C for 6 h. (c) Electron microscopy images

indicate the elevated ThT signal for Ab42E22Q compared to Ab42

after aging 10 lM solutions for 6 h (shown in b) is due to increased

abundance of amyloid fibril-like structures. (d) Densitometry analysis

of western blots shows decreased sensitivity of Ab42E22Q after aging

for 6 h. Ab degradation values (mean ± SEM, n = 3–4) are expressed

relative to control reactions that contained no MMP. p-Values in (b)

and (d) indicate significant difference between Ab42 and Ab42E22Q.

� 2009 The AuthorsJournal Compilation � 2009 International Society for Neurochemistry, J. Neurochem. (2009) 108, 1198–1207

Therapeutic restoration of Ab degradation | 1203

Ab accumulation by decreasing sensitivity to endogenousproteases.

Altered APP processing due to fAD mutations increasesrelative abundance of longer forms of the peptide (Suzukiet al. 1994) and the addition of two hydrophobic amino acidresidues makes Ab42 more susceptible to forming aggregatesthat bind the amyloid specific dye ThT (Soto and Castano1996). In addition to this, Ab42 is the predominant form ofAb within ThT-positive plaques in the AD brain (Shinkaiet al. 1997; Miller et al. 2006). Our data show that the moreamyloidogenic Ab42 is less sensitive to degradation thanAb40 (Fig. 2a), and that increased ThT fluorescence of theAb42E22Q mutant peptide associated with ‘Dutch’ fADcorrelates directly with decreased degradation (Fig. 3b andd). These data indicate that fAD-associated mutations thatpromote Ab amyloid formation may contribute to Abaccumulation by decreasing the peptide’s sensitivity toproteolytic degradation. Importantly, our SRCD, ThT and

electron microscopy data for Ab42 compared to Ab42E22Qshow that while a high initial b-sheet content in theAb42E22Q may contribute to subsequent accelerated for-mation of ThT-positive amyloid and fibril-like structures(Fig. 3a, b and c), the freshly prepared, b-sheet richAb42E22Q is still susceptible to degradation (Fig. 3d). Thisindicates that while factors conducive to Ab, b-sheetformation may be an important determinant of Ab accumu-lation, b-sheet rich Ab per se is still sensitive to proteolyticturnover. Supporting this, previous studies have shown Zn,but not Cu, induces the formation of b-sheet structure in Ab(Yang et al. 2000; Syme et al. 2004) and in a modelamyloidogenic peptide (Pagel et al. 2008). And as discussedbelow, Zn induces the formation of protease resistant, ThT-positive Ab amyloid, but Cu does not (Fig. 4a). Together,these data indicate Ab amyloid formation, possibly due tofactors that induce an initial increase in b-sheet content,contributes to Ab accumulation by preventing the activity of

Fig. 4 Zinc-induced Ab amyloid is resistant to degradation by MMP2.

(a) Relative MMP activity towards synthetic Ab42 or the artificial flu-

orogenic substrate Mca-PLGL-Dpa-AR (PLGL) in reaction mixtures

supplemented with 0–100 lM ZnCl2 or CuCl2. Ab42 degradation was

determined by densitometry analysis of western blot results. PLGL

degradation was determined by fluorescence of the cleaved substrate.

(b) Densitometry analysis of western blot results shows that the

presence of 100 lM ZnCl2 inhibits degradation of Ab by the alternate

proteases neprilysin (NEP) and insulin degrading enzyme (IDE). (c)

ThT fluorescence data showing that 10 lM Ab42 aged for 24 h in the

presence of 10 lM ZnCl2 forms ThT-positive Ab amyloid, and that

subsequent addition of 10 lM clioquinol (CQ) decreases the ThT

signal by 80%. (d) Electron microscopy showing that aging Ab42 in the

presence of Zn induced the formation of large Ab aggregates, and that

subsequent addition of 10 lM clioquinol caused amyloid disaggrega-

tion. (e) Densitometry analysis of western blot results shows that when

aged in the presence of ZnCl2 for 24 h Ab42 is relatively resistant to

degradation by MMP, but that the addition of clioquinol to the aged Ab-

Zn solution restores the peptide’s sensitivity to MMP. Ab degradation

values (mean ± SEM, n = 3–4) are expressed relative to control

reactions that contained no protease. Ab ThT fluorescence expressed

relative to Ab samples not treated with clioquinol. p-Values in (b), (c),

and (e) indicate significance of treating with clioquinol (CQ) or the

significance of adding ZnCl2 to the Ab/protease reaction mixtures.

Journal Compilation � 2009 International Society for Neurochemistry, J. Neurochem. (2009) 108, 1198–1207� 2009 The Authors

1204 | P. J. Crouch et al.

endogenous Ab-degrading enzymes. Consistent with this, theleu34-met35 cleavage site within the hydrophobic domain ofAb (Fig. 1d) would be inaccessible within an amyloidstructure.

While gene-specific mutations affecting proteins such asAPP or presenilin may explain Ab accumulation in fAD,there is a paucity of data to explain mechanisms of Abaccumulation in the much more prevalent sporadic AD(sAD). One potential mechanism is metal ion dyshomeosta-sis (see Crouch et al. 2007 for review), and strong data tosupport this come from a study where AD model mice werecrossed with mice deficient in the Zn transporter ZnT3, aprotein associated with releasing Zn into the synaptic cleft.ZnT3)/) mice exhibited a 50% decrease in amyloid plaqueburden compared to ZnT3+/+ littermates (Lee et al. 2002).As the ZnT3)/) and ZnT3+/+ mice expressed the samelevel of APP, these data indicate the role for synaptic Zn inAb accumulation and plaque formation (Bush 2003).Consistent with our observations above for Ab42E22Q,previous studies have already shown Zn induces b-sheet andsubsequent amyloid formation in a truncated form of Ab(Yang et al. 2000). Furthermore, Ab plaques in the AD braincontain b-sheet rich, ThT-positive amyloid Ab (Miller et al.2006), and Zn concentrations in Ab plaques are 30–70%higher than in plaque-free areas of the AD brain (Lovellet al. 1998). These data demonstrate the presence of Zn-enriched Ab amyloid deposits in the AD brain. Our data inFig. 4a and b indicate that Zn-induced formation of proteaseresistance may be an important early event in enabling Ab toaccumulate due to decreased efficacy of endogenous Ab-degrading enzymes. By contrast, the presence of Cu did notsubstantially alter the rate of Ab degradation (Fig. 4a). LikeZn, Cu is enriched in amyloid plaques of the AD brain(Lovell et al. 1998) and it can promote Ab oligomerisationand toxicity (Smith et al. 2006, 2007). However, Cu and Znpromote the formation of structurally different Ab aggre-gates. For example, under supra-molar Cu : Ab conditionsCu promotes Ab aggregation, but the aggregates formed arenot ThT-positive amyloid (Smith et al. 2007). Further to this,supra-molar Cu concentrations prevent the formation of ThT-positive Ab amyloid fibrils (Smith et al. 2007), indicatingthat while Zn accelerates Ab amyloid formation, Cupromotes aggregation down an alternate, non-amyloidogenicpathway. Ab accumulation in the AD brain may therefore bedetermined, in part, by whether the peptide interacts with Znor Cu. Data in Fig. 4 show that Zn-induced Ab amyloidformation correlates with decreased Ab degradation, butconditions consistent with Cu-induced formation of non-amyloid Ab aggregates do not prevent Ab degradation. Thisindicate that protease cleavage sites are inaccessible withinthe amyloid structure, but that they are still accessible on Cu-induced aggregates. The data in Fig. 4b indicate that thispossibility extends to the alternate Ab-degrading proteasesNEP and IDE.

The significance of showing Ab amyloid formation as adeterminant of the peptide’s sensitivity to degradation istwofold. First, it provides mechanistic insight for Abaccumulation even in the absence of increased Ab produc-tion or decreased protease expression, and may therefore bepertinent to Ab accumulation in sAD. Secondly, it is relevantto the development of AD therapeutics that target the Abpeptide. For example, the MPACs clioquinol and PBT2 bothdecrease the number of amyloid plaques in the brain whenadministered to transgenic mice expressing mutated humanAPP (Cherny et al. 2001; Adlard et al. 2008). BecauseMPACs disaggregate Ab in vitro (Huang et al. 1997; Chernyet al. 1999) the mechanism of action for decreased Ab levelsis believed to involve solubilisation of amyloid plaques byremoving Ab-bound metal ions. Plaque solubilisation alonehowever could not be expected to decrease overall brain Ablevels. In fact, liberating Ab from plaques could be expectedto elevate levels of soluble Ab. But in addition to decreasingplaque numbers, PBT2 and clioquinol also promotedecreased levels of soluble Ab in the brain (Cherny et al.2001; Adlard et al. 2008). Our study provides for the firsttime mechanistic data to indicate that in addition todecreasing the amyloid plaque burden, MPACs such asclioquinol can decrease overall brain Ab levels by reversingZn-induced amyloid formation and therefore restoring Absensitivity to proteolytic degradation. This provides newsupport for the development of therapeutics that aim toprevent Ab amyloid formation by showing the strategy canreinstate normal Ab-degrading activity of endogenousproteases. The efficacy of such an Ab-targeting strategycould be increased substantially if administered with atherapeutic that also elevates expression of the Ab-degradingproteases.

Acknowledgements

This work was supported by the National Health and Medical

Research Council (NHMRC) through Program Grant 400202 to

ARW, CLM, and KJB, and the University of Melbourne Early

Career Researcher Grant Scheme (Project #500144) to PJC. ARW is

a NHMRC RD Wright Fellow, KJB is a NHMRC Senior Research

Fellow. SRCD was carried out with the support of the Daresbury

Synchrotron Radiation Source and the assistance of David Clarke.

Travel for DJT to perform SRCD work was funded by the

Australian Nuclear Science and Technology Organisation through

the Access to Major Resource Facilities Program. Electron

microscopy was performed under the expert guidance of staff from

the University of Melbourne Bio21 EM Unit.

References

Adlard P. A., Cherny R. A., Finkelstein D. I. et al. (2008) Rapid res-toration of cognition in Alzheimer’s transgenic mice with 8-hy-droxy quinoline analogs is associated with decreased interstitialAb. Neuron 59, 43–55.

� 2009 The AuthorsJournal Compilation � 2009 International Society for Neurochemistry, J. Neurochem. (2009) 108, 1198–1207

Therapeutic restoration of Ab degradation | 1205

Atwood C. S., Moir R. D., Huang X., Scarpa R. C., Bacarra N. M. E.,Romano D. M., Hartshorn M. A., Tanzi R. E. and Bush A. I.(1998) Dramatic aggregation of Alzheimer Ab by Cu(II) is inducedby conditions representing physiological acidosis. J. Biol. Chem.273, 12817–12826.

Backstrom J. R., Lim G. P., Cullen M. J. and Tokes Z. A. (1996) Matrixmetalloproteinase-9 (MMP-9) is synthesized in neurons of thehuman hippocampus and is capable of degrading the amyloid-bpeptide (1–40). J. Neurosci. 16, 7910–7919.

Barnham K. J., Ciccotosto G. D., Tickler A. K. et al. (2003) Neurotoxic,redox-competent Alzheimer’s b-amyloid is released from lipidmembrane by methionine oxidation. J. Biol. Chem. 278, 42959–42965.

Bateman R. J., Munsell L. Y., Morris J. C., Swarm R., Yarasheski K. E.and Holtzman D. M. (2006) Human amyloid-b synthesis andclearance rates as measured in cerebrospinal fluid in vivo. Nat.Med. 12, 856–861.

Bush A. I. (2003) The metallobiology of Alzheimer’s disease. TrendsNeurosci. 26, 207–214.

Bush A. I., Pettingall W. H., Multhaup G., Paradis M., Vonsattel J.-P.,Gusella J. F., Beyreuther K., Masters C. L. and Tanzi R. E. (1994)Rapid induction of Alzheimer Ab amyloid formation by zinc.Science 265, 1464–1467.

Carson J. A. and Turner A. J. (2002) Beta-amyloid catabolism: roles forneprilysin (NEP) and other metallopeptidases? J. Neurochem. 81,1–8.

Chartier-Harlin M. C., Crawford F., Houlden H. et al. (1991) Early-onsetAlzheimer’s disease caused by mutations at codon 717 of the b-amyloid precursor protein gene. Nature 353, 844–846.

Cherny R. A., Legg J. T., McLean C. A. et al. (1999) Aqueous disso-lution of Alzheimer’s disease Ab amyloid deposits by biometaldepletion. J. Biol. Chem. 274, 23223–23228.

Cherny R. A., Atwood C. S., Xilinas M. E. et al. (2001) Treatment witha copper–zinc chelator markedly and rapidly inhibits b-amyloidaccumulation in Alzheimer’s disease transgenic mice. Neuron 30,665–676.

Ciccotosto G. D., Tew D., Curtain C. C. et al. (2004) Enhanced toxicityand cellular binding of a modified amyloid b peptide with a methi-onine to valine substitution. J. Biol. Chem. 279, 42528–42534.

Citron M., Oltersdorf T., Haass C., McConlogue L., Hung A. Y., SeubertP., Vigo-Pelfrey C., Lieberburg I. and Selkoe D. J. (1992) Mutationof the b-amyloid precursor protein in familial Alzheimer’s diseaseincreases b-protein production. Nature 360, 672–674.

Crouch P. J., Blake R., Duce J. A. et al. (2005) Copper-dependentinhibition of human cytochrome c oxidase by a dimeric conformerof amyloid-b1–42. J. Neurosci. 25, 672–679.

Crouch P. J., White A. R. and Bush A. I. (2007) The modulation of metalbio-availability as a therapeutic strategy for the treatment of Alz-heimer’s disease. FEBS J. 274, 3775–3783.

Crouch P. J., Harding S. M., White A. R., Camakaris J., Bush A. I. andMasters C. L. (2008) Mechanisms of Ab mediated neurodegener-ation in Alzheimer’s disease. Int. J. Biochem. Cell Biol. 40, 181–198.

Eckman E. A. and Eckman C. B. (2005) Ab-degrading enzymes:modulators of Alzheimer’s disease pathogenesis and targets fortherapeutic intervention. Biochem. Soc. Trans. 33, 1101–1105.

El-Amouri S. S., Zhu H., Yu J., Marr R., Verma I. M. and Kindy M. S.(2008) Neprilysin: an enzyme candidate to slow the progression ofAlzheimer’s disease. Am. J. Pathol. 172, 1342–1354.

Farris W., Schutz S. G., Cirrito J. R. et al. (2007) Loss of neprilysinfunction promotes amyloid plaque formation and causes cerebralamyloid angiopathy. Am. J. Pathol. 171, 241–251.

Fukami S., Watanabe K., Iwata N. et al. (2002) Ab-degrading endo-peptidase, neprilysin, in mouse brain: synaptic and axonal locali-

zation inversely correlating with Ab pathology. Neurosci. Res. 43,39–56.

Games D., Adams D., Alessandrini R. et al. (1995) Alzheimer-typeneuropathology in transgenic mice overexpressing V717F b-amy-loid precursor protein. Nature 373, 523–527.

Glenner G. G. and Wong C. W. (1984a) Alzheimer’s disease and Down’ssyndrome: sharing of a unique cerebrovascular amyloid fibrilprotein. Biochem. Biophys. Res. Commun. 122, 1131–1135.

Glenner G. G. and Wong C. W. (1984b) Alzheimer’s disease: initialreport of the purification and characterization of a novel cerebro-vascular amyloid protein. Biochem. Biophys. Res. Commun. 120,885–890.

Goate A., Chartier-Harlin M. C., Mullan M. et al. (1991) Segregation ofa missense mutation in the amyloid precursor protein gene withfamilial Alzheimer’s disease. Nature 349, 704–706.

Hellstrom-Lindahl E., Ravid R. and Nordberg A. (2008) Age-dependentdecline of neprilysin in Alzheimer’s disease and normal brain:inverse correlation with Ab levels. Neurobiol. Aging 29, 210–221.

Hemming M. L., Patterson M., Reske-Nielsen C., Lin L., Isacson O. andSelkoe D. J. (2007) Reducing amyloid plaque burden via ex vivogene delivery of an Ab-degrading protease: a novel therapeuticapproach to Alzheimer disease. PLoS Med. 4, e262.

Huang X., Atwood C. S., Moir R. D., Hartshorn M. A., Vonsattel J. P.,Tanzi R. E. and Bush A. I. (1997) Zinc-induced Alzheimer’s Ab1–40 aggregation is mediated by conformational factors. J. Biol.Chem. 272, 26464–26470.

Ida N., Hartmann T., Pantel J., Schroder J., Zerfass R., Forstl H.,Sandbrink R., Masters C. L. and Beyreuther K. (1996) Analysis ofheterogeneous bA4 peptides in human cerebrospinal fluid andblood by a newly developed sensitive western blot assay. J. Biol.Chem. 271, 22908–22914.

Jankowsky J. L., Slunt H. H., Ratovitski T., Jenkins N. A., CopelandN. G. and Borchelt D. R. (2001) Co-expression of multipletransgenes in mouse CNS: a comparison of strategies. Biomol. Eng.17, 157–165.

Jiang Q., Lee C. Y., Mandrekar S. et al. (2008) ApoE promotes theproteolytic degradation of Ab. Neuron 58, 681–693.

Kim M., Hersh L. B., Leissring M. A. et al. (2007) Decreased catalyticactivity of the insulin-degrading enzyme in chromosome 10-linkedAlzheimer disease families. J. Biol. Chem. 282, 7825–7832.

Lee J. Y., Cole T. B., Palmiter R. D., Suh S. W. and Koh J. Y. (2002)Contribution by synaptic zinc to the gender-disparate plaque for-mation in human Swedish mutant APP transgenic mice. Proc. NatlAcad. Sci. USA 99, 7705–7710.

Lovell M. A., Robertson J. D., Teesdale W. J., Campbell J. L. andMarkesbery W. R. (1998) Copper, iron and zinc in Alzheimer’sdisease senile plaques. J. Neurol. Sci. 158, 47–52.

Marr R. A., Rockenstein E., Mukherjee A., Kindy M. S., Hersh L. B.,Gage F. H., Verma I. M. and Masliah E. (2003) Neprilysin genetransfer reduces human amyloid pathology in transgenic mice.J. Neurosci. 23, 1992–1996.

Marr R. A., Guan H., Rockenstein E., Kindy M., Gage F. H., Verma I.,Masliah E. and Hersh L. B. (2004) Neprilysin regulates amyloid bpeptide levels. J. Mol. Neurosci. 22, 5–11.

Masters C. L. and Beyreuther K. (2006) Alzheimer’s centennial legacy:prospects for rational therapeutic intervention targeting the Abamyloid pathway. Brain 129, 2823–2839.

Masters C. L., Simms G., Weinman N. A., Multhaup G., McDonaldB. L. and Beyreuther K. (1985) Amyloid plaque core protein inAlzheimer disease and Down syndrome. Proc. Natl Acad. Sci. USA82, 4245–4249.

Miller L. M., Wang Q., Telivala T. P., Smith R. J., Lanzirotti A. andMiklossy J. (2006) Synchrotron-based infrared and X-ray imagingshows focalized accumulation of Cu and Zn co-localized with

Journal Compilation � 2009 International Society for Neurochemistry, J. Neurochem. (2009) 108, 1198–1207� 2009 The Authors

1206 | P. J. Crouch et al.

b-amyloid deposits in Alzheimer’s disease. J. Struct. Biol. 155, 30–37.

Miners J. S., Van Helmond Z., Chalmers K., Wilcock G., Love S. andKehoe P. G. (2006) Decreased expression and activity of neprilysinin Alzheimer disease are associated with cerebral amyloid angi-opathy. J. Neuropathol. Exp. Neurol. 65, 1012–1021.

Murakami K., Irie K., Morimoto A., Ohigashi H., Shindo M., Nagao M.,Shimizu T. and Shirasawa T. (2003) Neurotoxicity and physico-chemical properties of Ab mutant peptides from cerebral amyloidangiopathy: implication for the pathogenesis of cerebral amyloidangiopathy and Alzheimer’s disease. J. Biol. Chem. 278, 46179–46187.

Nalivaeva N. N., Fisk L. R., Belyaev N. D. and Turner A. J. (2008)Amyloid-degrading enzymes as therapeutic targets in Alzheimer’sdisease. Curr. Alzheimer Res. 5, 212–224.

Pagel K., Seri T., von Berlepsch H., Griebel J., Kirmse R., Bottcher C.and Koksch B. (2008) How metal ions affect amyloid formation:Cu2+- and Zn2+-sensitive peptides. Chembiochem 9, 531–536.

Roher A. E., Kasunic T. C., Woods A. S., Cotter R. J., Ball M. J. andFridman R. (1994) Proteolysis of Ab peptide from Alzheimerdisease brain by gelatinase A. Biochem. Biophys. Res. Commun.205, 1755–1761.

Selkoe D. J., Abraham C. R., Podlisny M. B. and Duffy L. K. (1986)Isolation of low-molecular-weight proteins from amyloid plaquefibers in Alzheimer’s disease. J. Neurochem. 46, 1820–1834.

Sherrington R., Rogaev E. I., Liang Y. et al. (1995) Cloning of a genebearing missense mutations in early-onset familial Alzheimer’sdisease. Nature 375, 754–760.

Shinkai Y., Yoshimura M., Morishima-Kawashima M., Ito Y., ShimadaH., Yanagisawa K. and Ihara Y. (1997) Amyloid beta-proteindeposition in the leptomeninges and cerebral cortex. Ann. Neurol.42, 899–908.

Smith D. P., Smith D. G., Curtain C. C. et al. (2006) Copper mediatedamyloid-b toxicity is associated with an intermolecular histidinebridge. J. Biol. Chem. 281, 15145–15154.

Smith D. P., Ciccotosto G. D., Tew D. J., Fodero-Tavoletti M. T.,Johanssen T., Masters C. L., Barnham K. J. and Cappai R. (2007)Concentration dependent Cu2+ induced aggregation and dityrosine

formation of the Alzheimer’s disease amyloid-b peptide. Bio-chemistry 46, 2881–2891.

Soto C. and Castano E. M. (1996) The conformation of Alzheimer’s bpeptide determines the rate of amyloid formation and its resistanceto proteolysis. Biochem. J. 314, 701–707.

Suzuki N., Cheung T. T., Cai X. D., Odaka A., Otvos L. Jr, Eckman C.,Golde T. E. and Younkin S. G. (1994) An increased percentage oflong amyloid b protein secreted by familial amyloid b proteinprecursor (b APP717) mutants. Science 264, 1336–1340.

SymeC.D.,Nadal R.C., RigbyS. E. andViles J.H. (2004)Copper bindingto the amyloid-b (Ab) peptide associated with Alzheimer’s disease.Folding, coordination geometry, pH dependence, stoichiometry, andaffinity of Ab-(1–28): Insights from a range of complementaryspectroscopic techniques. J. Biol. Chem. 279, 18169–18177.

White A. R., Du T., Laughton K. M. et al. (2006) Degradation of theAlzheimer disease amyloid b-peptide by metal-dependent up-reg-ulation of metalloprotease activity. J. Biol. Chem. 281, 17670–17680.

Yan P., Hu X., Song H. et al. (2006) Matrix metalloproteinase-9degrades amyloid-b fibrils in vitro and compact plaques in situ.J. Biol. Chem. 281, 24566–24574.

Yang D. S., McLaurin J., Qin K., Westaway D. and Fraser P. E. (2000)Examining the zinc binding site of the amyloid-b peptide. Eur. J.Biochem. 267, 6692–6698.

Yasojima K., Akiyama H., McGeer E. G. and McGeer P. L. (2001a)Reduced neprilysin in high plaque areas of Alzheimer brain: apossible relationship to deficient degradation of b-amyloid peptide.Neurosci. Lett. 297, 97–100.

Yasojima K., McGeer E. G. and McGeer P. L. (2001b) Relationshipbetween b amyloid peptide generating molecules and neprilysin inAlzheimer disease and normal brain. Brain Res. 919, 115–121.

Yin K. J., Cirrito J. R., Yan P. et al. (2006) Matrix metalloproteinasesexpressed by astrocytes mediate extracellular amyloid-b peptidecatabolism. J. Neurosci. 26, 10939–10948.

Zhao Z., Xiang Z., Haroutunian V., Buxbaum J. D., Stetka B. andPasinetti G. M. (2007) Insulin degrading enzyme activity selec-tively decreases in the hippocampal formation of cases at high riskto develop Alzheimer’s disease. Neurobiol. Aging 28, 824–830.

� 2009 The AuthorsJournal Compilation � 2009 International Society for Neurochemistry, J. Neurochem. (2009) 108, 1198–1207

Therapeutic restoration of Ab degradation | 1207