DaltonTransactions

PAPER

Cite this: DOI: 10.1039/c4dt01991a

Received 1st July 2014,Accepted 10th July 2014

DOI: 10.1039/c4dt01991a

www.rsc.org/dalton

Self-assembly of stable oligomeric and fibrillaraggregates of Aβ peptides relevant to Alzheimer’sdisease: morphology dependent Cu/heme toxicityand inhibition of PROS generation†

Kushal Sengupta,‡ Sudipta Chatterjee,‡ Debajyoti Pramanik, Somdatta Ghosh Dey*and Abhishek Dey*

Large and small aggregates of Aβ peptides, resembling the morphology and dimensions of fibrillar and oli-

gomeric forms of Aβ respectively, relevant to Alzheimer’s disease, are stabilized on electrodes using self-

assembly. Both of these forms were found to bind redox active Cu and heme, resulting in active sites

having distinctive biophysical properties. The reduced metal bound Aβ active sites of both the oligomeric

and fibrillar forms of Aβ produce detrimental partially reduced oxygen species (PROS). While the larger

aggregates of heme-Aβ produce more PROS in situ, the smaller aggregates of Cu-Aβ produce more

PROS. 8-Hydroxy quinoline and methylene blue are inhibitors of Cu and heme bound Aβ respectively,

and are shown to efficiently reduce PROS formation in the oligomeric forms. However, these inhibitors

are ineffective in reducing the toxicities of the Cu and heme bound Aβ peptides in the fibrils, making

them significantly more lethal than the smaller Aβ aggregates.

Introduction

Alzheimer’s disease (AD), a common cause of senile dementia,is one of the most recurrent progressive neurodegenerative dis-eases associated with memory loss, cognitive decline, moodswing, behavior change, physical disability, ultimately leadingto death.1–3 The aggregation of amyloid β (Aβ) peptides intoinsoluble, mature and ordered fibrillar forms through acascade of events, resulting from the proteolysis of theamyloid precursor protein (APP), is the key for the etiology ofthis disease.4,5 In AD affected brains, the soluble monomericAβ is gradually converted to fibrils and finally into plaquesthrough metastable and transient oligomeric and protofibrillarintermediates.5–7 Although fibrils were initially thought to bethe direct cause of pathogenesis and intracellular toxicity,8

later on the intermediates were found to be the major threatsin a number of amyloid diseases including AD, Parkinson’sdisease and Type II diabetes.6,9–12 Recently these fibril pre-cursors, oligomers and protofibrils, have been isolated andcharacterized by various groups using electron microscopy,

atomic force microscopy (AFM), circular dichroism, H/D exchangeexperiments, size exclusion chromatography, etc.8,13–16 Oligo-mers randomly coalesce into protofibrils, which subsequentlyevolve to form ordered aggregates that eventually leads to fibrilformation (Fig. 1). Reports suggest that oligomeric forms of Aβare also toxic by themselves as it attains a pseudo-micellarglobular shape15,17,18 with the hydrophilic residues projectingoutward, leading to easy binding of metals and production ofreactive oxygen species (ROS), thus leading to toxicity.19,20

Oligomers have been categorically classified into low mole-cular weight (assembly of 2–8 monomers) and high molecularweight (assembly of ∼20–40 monomers) oligomers havingdiameters ranging from 4 to 40 nm.13,15 These oligomersreadily fold to adopt cross β-sheet secondary structures thatlead to the formation of fibrillar forms on the surface of mem-brane bilayers, making it difficult to investigate their reactiv-ities in detail. Along with parallel/cross β-sheets, helicalstructures and anti-parallel β-sheets are also found in the finalstage of fibrillogenesis.21

Apart from the amyloid hypothesis, the physiologically rele-vant redox active metal, Cu, and cofactors like heme also bindto Aβ peptides and are shown to generate partially reducedoxygen species (PROS), which is further linked to oxidativedamage of healthy brain tissue.22–24 Biophysical studies indi-cate that Cu and heme bind well defined active sites in the Aβpeptides.24–26 It is conceivable that these active sites will bepresent in both oligomeric and protofibrillar forms of Aβ,

†Electronic supplementary information (ESI) available: AFM data, SERRS-RDEdata, PROS data. See DOI: 10.1039/c4dt01991a‡These authors contributed equally.

Department of Inorganic Chemistry, Indian Association for the Cultivation of

Science, Jadavpur, Kolkata, 700032, India. E-mail: icad@iacs.res.in,

icsgd@iacs.res.in

This journal is © The Royal Society of Chemistry 2014 Dalton Trans.

Publ

ishe

d on

14

July

201

4. D

ownl

oade

d by

Ind

ian

Ass

ocia

tion

for

the

Cul

tivat

ion

of S

cien

ce o

n 05

/08/

2014

15:

10:3

0.

View Article OnlineView Journal

i.e., small and large aggregates of Aβ, and their reactivities maydepend on the aggregate size as observed in the recent past byseveral groups where they have worked with Cu bound Aβspecies varying in aggregation size and form.27–30 Similarly,inhibition of their reactivity either by chelation or by suppres-sion of PROS generation may depend on the size of the aggre-gates as well; which remain unexplored.

In the recent past, our group has shown that human Aβ1–16, extended to accommodate a Cys (AβCys) residue at theC-terminal, spontaneously assembles on the Au electrode.31

These scaffolds can bind both heme and Cu2+ under physio-logical conditions and can be used as a platform for testingthe extent of PROS generated by Cu and heme bound Aβ pep-tides and their potential inhibitors. Here, controlled formationof large and small Aβ peptide aggregates bound to Cu andheme is reported, which resembles that of the protofibrills orfibrils and oligomers, respectively. The data show radicaldifferences in their relative PROS formation and response toinhibitors like 8-hydroxyquinone (HQ, Cu chelator) andmethylene blue (MB, heme inhibitor).

Experimental detailsMaterials

All reagents were of the highest grade commercially availableand were used without further purification. Aβ (1–16) peptidesappended with a terminal cysteine, AβCys (sequence: Asp-Ala-Glu-Phe-Arg-His-Asp-Ser-Gly-Tyr-Glu-Val-His-His-Gln-Lys-Cys),have been used for this study. The peptide was purchasedfrom GL Biochem (Shanghai) Ltd with >95% purity. Hemin,octanethiol (C8SH), potassium hexaflurophosphate (KPF6)and the buffers were purchased from Sigma-Aldrich.Di-sodium hydrogen phosphate dihydrate (Na2HPO4·2H2O),1-cysteine, copper sulfate, methylene blue, 8-hydroxy quino-line and imidazole were purchased from Merck. Au waferswere purchased from Platypus Technologies (1000 Å of Au on50 Å of Ti adhesion layer on the top of a Si(III) surface). Trans-parent Au wafers (100 Å of Au on 10 Å of Ti) were purchasedfrom Phasis, Switzerland. Au and Ag discs for the rotatingring disc electrochemistry (RRDE) and surface enhancedresonance Raman spectroscopy coupled to rotating discelectrochemistry (SERRS-RDE) experiments were purchasedfrom Pine Instruments, USA.

Instrumentation

A spectrophotometer (Agilent Technologies, model 8453) fittedwith a diode-array detector was used to obtain absorptiondata. All electrochemical experiments were performed using aCH Instruments (model CHI710D) Electrochemical Analyzer.Biopotentiostat, reference electrodes, and a Teflon® platematerial evaluating cell (ALS Japan) were purchased fromCH Instruments. The RRDE set-up from Pine Research Instru-mentation (E6 series ChangeDisk tips with AFE6M rotor) wasused to obtain the RRDE data. The AFM data were obtained atroom temperature in a Veeco dicp II (model no. AP-0100)instrument bearing a phosphate doped Si cantilever (1–10 Ω cm,thickness 3.5–4.5 μm, length 115–135 μm, width 30–40 μm,resonance frequency 245–287 kHz, elasticity 20–80 N m−1).Surface Enhanced Resonance Raman data were collected usinga Trivista 555 spectrograph (Princeton Instruments) and using413.1 nm excitation from a Kr+ laser (Coherent, Sabre InnovaSBRC-DBW-K).

Methods

Construction of the electrodesFormation of mixed self-assembled monolayer (SAM). Gold

wafers were cleaned electrochemically, first by electrolysiswhere it was held at a high positive potential (2.1 V) for a fewseconds and then by sweeping several times between 1.5 V and−0.3 V in 0.5 M H2SO4. 0.1 mM Aβ (1–17) in 100 mM phos-phate buffer (pH 7) was used as the SAM solution. Freshlycleaned Au wafers or discs were thoroughly rinsed with tripledistilled water, and purged with N2 gas and immersed in theSAM solution for 2 days. This resulted in large aggregatedAβ surfaces (AβWT). After 2 days the substrates were immersedin 0.1 mM C8SH (in EtOH) solution for 30 min in orderto obtain C8SH diluted surfaces (AβC8SH). For preparing amixed SAM of AβCys and 1-cysteine, a mixed SAM solutionof 0.1 mM Aβ (1–17) and 0.9 mM 1-cysteine was used toobtain Aβ1:9.

Attachment of heme and Cu on to SAM. Gold wafers or discsimmersed in the deposition solution were taken out andrinsed with triple distilled deionized water in order to removeany excess adsorbate and dried with N2 gas to remove the residualsolvent. The wafers were then inserted into a Plate Material Evalu-ating Cell (ALS Japan) and the discs were mounted on a platinumring disc assembly (Pine Instruments, USA). 1 mM Hemin

Fig. 1 2D schematic representation of sequential steps involved in amyloid peptide fibrillogenesis.

Paper Dalton Transactions

Dalton Trans. This journal is © The Royal Society of Chemistry 2014

Publ

ishe

d on

14

July

201

4. D

ownl

oade

d by

Ind

ian

Ass

ocia

tion

for

the

Cul

tivat

ion

of S

cien

ce o

n 05

/08/

2014

15:

10:3

0.

View Article Online

solution and 10 μM CuSO4 solution in DMSO were used as theincubating solutions. Heme-Aβ, Cu-Aβ surfaces were preparedby incubating the modified Au surfaces with the respectiveDMSO solutions for 10–15 min and 2 hours respectively. Thesurfaces, after incubation with the respective solution, wererinsed with DMSO for 3 min to remove excess physiadsorbedheme or Cu (if any).

Atomic force microscopy (AFM) experiments. Freshly cut Auwafers were taken for each AFM analysis where AyCys modifiedsurfaces were made as described in the previous section. Thesurfaces were thoroughly rinsed with triple deionised waterbefore analysis. AFM data were obtained at room temperaturein a Veeco dicp II instrument bearing a phosphate dopedSi cantilever (1–10 Ω cm, thickness 3.5–4.5 μm, length115–135 μm, width 30–40 μm, resonance frequency 245–287 kHz,elasticity 20–80 N m−1).

Absorption spectroscopy. For absorption experiments, co-valently attached heme on modified transparent Au electrodeswas used. Surface preparation was done as described in theprevious section.

Electrochemical experiments. All electrochemical (cyclic vol-tammetry and linear sweep voltammetry) experiments weredone in pH 7 buffer (until otherwise mentioned) containing100 mM Na2HPO4·2H2O and 100 mM KPF6 (supportingelectrolyte) using Pt wire as the counter electrode and Ag/AgClas the reference electrode. All the potentials in the text andfigures have been adjusted vs. NHE.

Cyclic voltammetry (CV) experiments. CV data of heme-AβWT

and heme-Aβ1:9 were collected in deoxygenated pH 7 bufferunder anaerobic conditions to obtain the redox potentials andin air saturated pH 7 buffer to obtain ORR current using modi-fied electrodes.

Partially reduced oxygen species (PROS) experiment. The plati-num ring and the gold disc were both polished with aluminapowder (size: 1 µ, 0.5 µ and 0.03 µ) and electrochemicallycleaned in 0.5 M H2SO4 (as mentioned in previous section forAu wafers) and inserted into the RRDE tip, which is thenmounted on the rotor and immersed into a cylindrical glasscell which is equipped with Ag/AgCl reference and Pt counterelectrodes. The collection efficiency of the RRDE set-up ismeasured in a 2 mM K3[Fe-(CN)6] and 0.1 M KNO3 solution at10 mV s−1 scan rate and 300 rpm rotation speed. The collec-tion efficiency (CE) generally recorded during these experi-ments was 20 ± 2%. The potential at which the ring was heldduring the collection experiments for detecting H2O2 wasobtained from the literature.

Coverage calculation. The coverage for a particular species isestimated by integrating the oxidation and reduction currentsof the respective species.

Surface enhanced resonance Raman spectroscopy coupled to rotat-ing disc electrode (SERRS-RDE) experiments. Ag discs are cleanedusing alumina powder (grit sizes 1, 0.3 and 0.05 microns) andthen roughened in 0.1 M KCl solution using reported pro-cedures and immersed in Ay (1–19) solutions to form SAM.The roughened modified Ag discs are then inserted into theRRDE set-up for collection of SERRS data. Heme is attached to

the SAM in a similar manner to that described in previous sec-tions. Experiments are done using an excitation wavelength of413.1 nm and the power used at the electrode surface wasaround 10–12 mW. While collecting the spectra at the resting/oxidized state the disc is held at 0 V and at −0.5 V (vs. Ag/AgCl)to obtain a reduced spectrum.

Inhibition studies. Inhibition by methylene blue and8-hydroxy quinoline were studied using various concentrationsof their solutions, respectively, in pH 7 buffer in each case.

Results and discussion

AβCys assemble on Au electrode forming large aggregates of8–12 nm height and ∼100–120 nm width which extend forfew micrometers resembling the fibrillar form of Aβ (AβWT)(Fig. 2a and 2b). Since the Aβ peptides are covalently tetheredto the electrode at the C-terminus, the formation of these walllike structures indicates the formation of parallel β-sheets.However, comparing the height and width observed here withthe literature reports of the protofibrils and fibrils of Aβ, weassume these large aggregates to be likely a mixture of boththese fibrillar forms (Fig. 3).13,15,32 In the presence of a hydro-phobic diluent such as C8SH (AβC8SH), isolated clusters areobserved (ESI Fig. S1†).31 Similarly, when an Au electrode isimmersed in a mixed SAM solution of AβCys and a hydrophilicdiluent 1-cysteine (1 : 9), a homogeneous distribution of smallclusters is observed (Aβ1:9) (Fig. 2c). Note that the height andwidth of the isolated clusters with both hydrophobic or hydro-philic diluents range between 4–5 nm and 16–22 nm, respect-ively (Fig. 2d). This height is consistent with the length of a 17

Fig. 2 AFM images of AβCys assembled on the Au electrode. (a) Wall-like structures of the aggregated AβCys i.e. without a diluent (AβWT), (c)isolated structures of the AβCys peptides in the presence of 90%1-cysteine as a diluent (Aβ1:9). (b) and (d) are their respective height dis-tribution profile diagrams.

Dalton Transactions Paper

This journal is © The Royal Society of Chemistry 2014 Dalton Trans.

Publ

ishe

d on

14

July

201

4. D

ownl

oade

d by

Ind

ian

Ass

ocia

tion

for

the

Cul

tivat

ion

of S

cien

ce o

n 05

/08/

2014

15:

10:3

0.

View Article Online

residue peptide. The width and 3D image of these clustersindicate the presence of ∼20–40 Aβ monomers aggregated intoglobular structures resembling the high molecular weight oli-gomers (Fig. 3). This assembly on the electrode surface wherethe hydrophilic Aβ peptide projects out to the electrolytic

solution can thus be compared to a cross-section of the pseudo-micellar oligomeric form, observed both in vivo and in vitro,where the hydrophilic residues project outward into the extra-cellular domain.15,17 Unlike in vivo, where the oligomers aretransient species involved in fibril formation and hence arenot stable entities, here these oligomers, bearing the sameorganization, are stable species offering an unprecedentedopportunity for investigation of their reactivities/toxicities.Thus the large aggregates and the small clusters formed onthe electrode surfaces under different depositing conditionsmimic the organization of fibrils and large oligomers of Aβformed in vivo.

These peptide assemblies readily bind both heme and Cuunder physiological conditions. The heme bound AβWT andAβ1:9 surfaces show intense soret transitions at 399 nm and401 nm respectively, similar to that reported previously (Fig. 4aand b).31 Upon binding an external ligand like imidazole thesoret band shifts from 399 nm to 422 nm in the case of heme-AβWT and from 401 nm to 428 nm for heme-Aβ1:9 (Fig. 4a andb). Note that a similar shift of soret from 400 nm to 426 nmwas observed for heme-AβC8SH.31 The differences in the valuesmay reflect the difference in the heme environment in theaggregated and isolated Aβ strands and thus SERRS-RDE33 wasused for a better understanding of the systems (vide infra).

CV experiments with heme-AβWT and heme-Aβ1:9, per-formed in the absence of oxygen, show a reversible FeIII/FeII

Fig. 3 Probable arrangements of AβWT (left) and Aβ1 : 9 (right) on theAu surface as observed in their respective AFM images.

Fig. 4 Characterization of the modified surfaces by absorption and CV experiments. Absorption data of heme bound AβWT (a) and Aβ1:9 (b) obtainedin pH 7 buffer under the resting state with (green) or without (red) the presence of the imidazole (Imd) ligand. (c) CV data of heme-AβWT (blue) andheme-Aβ1:9 (orange) modified electrodes in deoxygenated pH 7 buffer at a scan rate of 1 V s−1. (d) CV data of Cu-AβWT (green) and Cu-AβC8SH (red)modified electrodes in pH 7 buffer at a scan rate of 50 mV s−1.

Paper Dalton Transactions

Dalton Trans. This journal is © The Royal Society of Chemistry 2014

Publ

ishe

d on

14

July

201

4. D

ownl

oade

d by

Ind

ian

Ass

ocia

tion

for

the

Cul

tivat

ion

of S

cien

ce o

n 05

/08/

2014

15:

10:3

0.

View Article Online

redox couple at −130 mV and −160 mV vs. NHE, respectively(Fig. 4c), which appeared at −150 mV for heme-AβC8SHsamples.31 The surface coverage of these samples was obtainedto be 7 × 1012 molecules cm−2 and 4.4 × 1012 molecules cm−2

for the large and small aggregates, respectively. Thus almostsimilar numbers of heme sites are present on the electrodeirrespective of the aggregate size. Similar experiments with Cu-AβWT in pH 7 shows a reversible CuII/CuI couple at 370 mV vs.NHE (Fig. 4d). In our previous report it was observed thatwhen Cu is bound to isolated Aβ clusters (Cu-AβC8SH) thecouple appears at slightly more negative potentials at 300 mV(Fig. 4d).31 The coverages calculated for the large and smallCu-Aβ aggregates are 1.04 × 1013 and 7.92 × 1012 moleculescm−2, respectively.34

SERRS-RDE of heme bound to Aβ aggregates (i.e. heme-AβWT) under oxidizing conditions shows the oxidation state,coordination number, and spin state marker bands, υ4, υ3 andυ2 respectively, at 1371 cm−1, 1489 cm−1 and 1569 cm−1

(Fig. 5a, blue). For heme-AβC8SH and heme-Aβ1:9 (i.e., smallaggregates) the υ4, υ3 and υ2 bands appear at 1374 cm−1,1491 cm−1 and 1572 cm−1; and 1372 cm−1, 1491 cm−1 and

1570 cm−1, respectively (Fig. 5a and ESI Fig. S2†). These valuesindicate the presence of a six coordinate (6C) high-spin (HS)FeIII species as the major species in both the cases.35–37 Ashoulder can be observed in the υ2 band corresponding to 6Clow-spin (LS) FeIII species at around 1585 cm−1. The intensityof this shoulder is greater in the case of small aggregates(Fig. 5a, orange) relative to the large aggregates. Thus theSERRS data clearly indicate the presence of significant 6C LSFeIII species in small aggregates relative to large aggregates.The nature of the heme active site in the reduced state couldbe probed using the SERRS-RDE set-up, where SERRS data arecollected by applying a constant or variable potential at theelectrode.33 At a reducing potential of −200 mV (vs. NHE) theSERRS-RDE spectra of both heme-AβWT and heme-AβC8SH showthe presence of HS FeII species with the υ4, υ3 and υ2 bandsappearing at 1358 cm−1, 1491 cm−1 and 1580 cm−1; and at1358 cm−1, 1492 cm−1 and 1583 cm−1, respectively(Fig. 5b).35,37 Minor signal from some residual HS FeIII speciesis also observed (ESI Fig. S3†).

In the presence of O2, CV experiments with heme-AβWT andCu-AβWT show an irreversible cathodic catalytic current, corres-ponding to electrocatalytic O2 reduction, instead of the revers-ible redox couple (ESI Fig. S4†). Reaction of O2 with reducedCu or heme-Aβ complexes leads to the formation of H2O2

along with H2O because neither the Cu nor the heme activesite alone possesses 4e− required to reduce O2 to H2O. Whenimmobilized on an electrode the additional electrons neededcould be provided by the electrode at sufficiently reducedpotential. However at intermediate potentials PROS are pro-duced which can be detected in situ under steady state con-ditions using RRDE.38 While the amount of PROS released byheme-AβWT is 23.7 ± 0.5%, heme-Aβ1:9 produces 16.3 ± 0.3%PROS at 0 V vs. NHE (Table 1, Fig. 6). A similar PROS value wasalso observed with C8SH as a diluent. This reflects that hemecan potentially produce more stress when bound to Aβ aggre-gates relative to small isolated peptide strands. However anopposite trend was observed in the case of Cu-Aβ active sites.The amount of PROS released by the large aggregates ofCu-AβWT is 19.5 ± 1.5% (Fig. S5, ESI†), in contrast to 31 ± 1%PROS produced by the small isolated aggregates of Cu-AβC8SH(Table 1, Fig. 6) at 0 V vs. NHE. Thus the Cu bound smalleraggregates of Aβ can potentially result in more oxidative stressthan the larger clusters.

In isolated small clusters HQ (∼15 nM) acts as a chelator ofthe bound Cu and removes it from Aβ. MB (∼15 μM) leads tosignificant inhibition of O2 reduction by heme and also acts asan anti-oxidant.31 Unlike previous results, in Cu chelationexperiments performed with HQ using the same concentration

Table 1 Amount (%) of PROS produced by various species

Cofactors Heme Cu

AβWT (without diluents) 23.7 ± 0.5 19.5 ± 1.5Aβ1:9 (1-cysteine as diluent) 16.3 ± 0.3 NDAβC8SH (C8SH as diluent) 17 ± 2 31 ± 1

Fig. 5 Characterization of the heme modified surfaces by SERRS-RDEexperiment. SERRS-RDE data of heme-AβWT (blue) and heme-AβC8SH(orange) modified electrodes in pH 7 buffer under oxidizing potential(i.e. resting state) (a) and under reducing potential in anaerobicpH 7 buffer (b).

Dalton Transactions Paper

This journal is © The Royal Society of Chemistry 2014 Dalton Trans.

Publ

ishe

d on

14

July

201

4. D

ownl

oade

d by

Ind

ian

Ass

ocia

tion

for

the

Cul

tivat

ion

of S

cien

ce o

n 05

/08/

2014

15:

10:3

0.

View Article Online

(15 nM) and time (30 min) of incubation, large Cu-AβWT aggre-gates showed minimal removal of Cu from the aggregates.Varying both the concentration of HQ and the time of incu-bation with Cu-AβWT, it was observed that removal of Cu,bound to large Aβ aggregates, took greater concentration andmore time. It required >80 nM of HQ and 6 hours incubationrelative to 15 nM and 2 h in the small aggregates (Fig. 7a).Thus the chelation process was both thermodynamically andkinetically more inefficient for the large Cu-Aβ aggregates rela-tive to the smaller aggregates. The inhibition of O2 reductionvis-à-vis PROS production by MB was even more unsuccessfulin the larger aggregates. Even a high concentration of 500 μMMB was not sufficient enough for efficient inhibition of O2

reduction by heme-AβWT by even 50%, even after 6 h of incu-bation (Fig. 7b). In contrast, PROS production could be inhib-ited within 10 minutes and complete inhibition of O2

reduction within 6 h using only 15 μM concentration of MB inthe isolated small aggregates.

Conclusion

In summary, the UV-Vis, CV and SERRS data indicate that theactive site environments of heme-Aβ are different in large andsmall aggregates of Aβ peptides where the latter contain sig-nificant population of a LS ferric active site in addition to theHS ferric active site exclusively present in the larger aggregates.It was concluded that the larger aggregates of heme-Aβ andCu-Aβ, mimicking the Aβ fibrils, are capable of generatingneurotoxic PROS. Likewise, smaller aggregates, mimicking Aβoligomers, also produce significant PROS and thus both aredetrimental. The large heme-Aβ aggregates were found toproduce more PROS than the smaller heme-Aβ aggregates.Alternatively, the smaller Cu-Aβ aggregates were found toproduce more PROS than the larger ones. Given the fact thatthe E0 of Cu-Aβ is higher than the physiological reducingagents like ascorbic acid and glutathione, the smaller Cu-Aβaggregates, i.e. oligomers, are significantly neurotoxic com-

Fig. 7 3D graphical presentation of the inhibition studies. Variation of(a) Cu current (obtained from the CV peak intensities) and O2 reductioncurrent by heme (b) with the concentration of the respective inhibitors,HQ and MB, and time of incubation.

Fig. 6 PROS and RRDE data of the modified surfaces. (a) Percentage of PROS produced by heme and Cu bound various actives sites in pH 7 buffermeasured at physiological potentials (0 V vs. NHE). (b) RRDE plots of heme bound species showing the disc (ID) and ring (IR) currents.

Paper Dalton Transactions

Dalton Trans. This journal is © The Royal Society of Chemistry 2014

Publ

ishe

d on

14

July

201

4. D

ownl

oade

d by

Ind

ian

Ass

ocia

tion

for

the

Cul

tivat

ion

of S

cien

ce o

n 05

/08/

2014

15:

10:3

0.

View Article Online

pared with the larger fibrils. However, the oligomers werefound to be more susceptible to both removal of Cu fromCu-Aβ and inhibition of PROS production by heme-Aβ. Inhi-bition of the same was thermodynamically (took greater con-centration) and kinetically (took more time) difficult in the Aβfibrils likely due to sterically hindered access to these activesites. The fact that inhibition of their PROS production, eitherby chelation or by binding, is dramatically ineffective com-pared to the oligomers necessitates breaking of the fibrils tooligomers prior to any Cu/heme specific treatment.

Conflict of interest

The authors declare no competing financial interests.

Acknowledgements

We thank the SERC Fast Track Scheme SR/FT/CS-34/2010(SGD) and SR/S1/IC-35/2009 (AD), Department of Science andTechnology, Govt. of India for funding this research. K. S.,S. C. and D. P. thank the CSIR, India, for Senior ResearchFellowship.

References

1 P. G. Nelson, Curr. Alzheimer Res., 2005, 2, 497.2 A. Rauk, Chem. Soc. Rev., 2009, 38, 2698.3 K. P. Kepp, Chem. Rev., 2012, 112, 5193.4 J. A. Hardy and G. A. Higgins, Science, 1992, 256, 184.5 J. Hardy and D. J. Selkoe, Science, 2002, 297, 353.6 R. Kayed, E. Head, J. L. Thompson, T. M. McIntire,

S. C. Milton, C. W. Cotman and C. G. Glabe, Science, 2003,300, 486.

7 M. Faendrich, J. Mol. Biol., 2012, 421, 427.8 D. M. Walsh, D. M. Hartley, Y. Kusumoto, Y. Fezoui,

M. M. Condron, A. Lomakin, G. B. Benedek, D. J. Selkoeand D. B. Teplow, J. Biol. Chem., 1999, 274, 25945.

9 J. P. Cleary, D. M. Walsh, J. J. Hofmeister, G. M. Shankar,M. A. Kuskowski, D. J. Selkoe and K. H. Ashe, Nat. Neuro-sci., 2005, 8, 79.

10 A. Deshpande, E. Mina, C. Glabe and J. Busciglio, J. Neuro-sci., 2006, 26, 6011.

11 S. Lesne, M. T. Koh, L. Kotilinek, R. Kayed, C. G. Glabe,A. Yang, M. Gallagher and K. H. Ashe, Nature, 2006, 440,352.

12 C. Haass and D. J. Selkoe, Nat. Rev. Mol. Cell Biol., 2007, 8,101.

13 M. Arimon, I. Diez-Perez, M. J. Kogan, N. Durany, E. Giralt,F. Sanz and X. Fernandez-Busquets, FASEB J., 2005, 19,1344.

14 D. Losic, L. L. Martin, A. Mechler, M.-I. Aguilar andD. H. Small, J. Struct. Biol., 2006, 155, 104.

15 I. A. Mastrangelo, M. Ahmed, T. Sato, W. Liu, C. Wang,P. Hough and S. O. Smith, J. Mol. Biol., 2006, 358, 106.

16 Y. Zhang, D. L. Rempel, J. Zhang, A. K. Sharma,L. M. Mirica and M. L. Gross, Proc. Natl. Acad. Sci. U. S. A.,2013, 110, 14604.

17 W. Yong, A. Lomakin, M. D. Kirkitadze, D. B. Teplow,S.-H. Chen and G. B. Benedek, Proc. Natl. Acad. Sci. U. S. A.,2002, 99, 150.

18 R. Sabate and J. Estelrich, J. Phys. Chem. B, 2005, 109,11027.

19 L. L. Iversen, R. J. Mortishire-Smith, S. J. Pollack andM. S. Shearman, Biochem. J., 1995, 311, 1.

20 L. Hou and M. G. Zagorski, J. Am. Chem. Soc., 2006, 128,9260.

21 N. Carulla, M. Zhou, E. Giralt, C. V. Robinson andC. M. Dobson, Acc. Chem. Res., 2010, 43, 1072.

22 D. G. Smith, R. Cappai and K. J. Barnham, Biochim.Biophys. Acta, Biomembr., 2007, 1768, 1976.

23 A. Rauk, Dalton Trans., 2008, 1273.24 P. Faller, C. Hureau and O. Berthoumieu, Inorg. Chem.,

2013, 52, 12193.25 D. Pramanik and S. G. Dey, J. Am. Chem. Soc., 2011, 133, 81.26 D. Pramanik, C. Ghosh and S. G. Dey, J. Am. Chem. Soc.,

2012, 133, 15545.27 C. Opazo, X. Huang, R. A. Cherny, R. D. Moir, A. E. Roher,

A. R. White, R. Cappai, C. L. Masters, R. E. Tanzi,N. C. Inestrosa and A. I. Bush, J. Biol. Chem., 2002, 277,40302.

28 L. Guilloreau, S. Combalbert, A. Sournia-Saquet,H. Mazarguil and P. Faller, ChemBioChem, 2007, 8, 1317.

29 R. C. Nadal, S. E. J. Rigby and J. H. Viles, Biochemistry,2008, 47, 11653.

30 C.-L. Fang, W.-H. Wu, Q. Liu, X. Sun, Y. Ma, Y.-F. Zhao andY.-M. Li, Regul. Pept., 2010, 163, 1.

31 D. Pramanik, K. Sengupta, S. Mukherjee, S. G. Dey andA. Dey, J. Am. Chem. Soc., 2012, 134, 12180.

32 Z. Wang, C. Zhou, C. Wang, L. Wan, X. Fang and C. Bai,Ultramicroscopy, 2003, 97, 73.

33 K. Sengupta, S. Chatterjee, S. Samanta and A. Dey, Proc.Natl. Acad. Sci. U. S. A., 2013, 110, 8431.

34 Note that in the case of Aβ 1 : 9, Cu was found to bind theopen –COO-head groups of the diluents (i.e., from1-cysteine) and thus further experiments were not per-formed with Cu-Aβ 1 : 9 surfaces as the additional Cu couldnot be removed specifically.

35 S. Choi, J. J. Lee, Y. H. Wei and T. G. Spiro, J. Am. Chem.Soc., 1983, 105, 3692.

36 T. K. Das, M. Couture, H. C. Lee, J. Peisach, D. L. Rousseau,B. A. Wittenberg, J. B. Wittenberg and M. Guertin, Biochem-istry, 1999, 38, 15360.

37 A. Gomez de Gracia, L. Bordes and A. Desbois, J. Am. Chem.Soc., 2005, 127, 17634.

38 A. J. Bard and L. R. Faulkner, Electrochemical Methods: Fun-damentals and Applications, John Wiley & Sons, Inc.,New York, 2001.

Dalton Transactions Paper

This journal is © The Royal Society of Chemistry 2014 Dalton Trans.

Publ

ishe

d on

14

July

201

4. D

ownl

oade

d by

Ind

ian

Ass

ocia

tion

for

the

Cul

tivat

ion

of S

cien

ce o

n 05

/08/

2014

15:

10:3

0.

View Article Online