Morphology of self-assembled structures formed by short peptides from the amyloidogenic protein tau depends on the solvent in which the peptides are dissolved

67

5

Research ArticleReceived: 7 April 2009 Revised: 14 July 2009 Accepted: 16 July 2009 Published online in Wiley Interscience: 27 August 2009

(www.interscience.com) DOI 10.1002/psc.1172

Morphology of self-assembled structuresformed by short peptides from theamyloidogenic protein tau depends on thesolvent in which the peptides are dissolvedNitin Chaudhary, Shashi Singh and Ramakrishnan Nagaraj∗

The aggregation behavior of peptides Ac-VQIVYK-amide (AcPHF6) and Ac-QIVYK-amide (AcPHF5) from the amyloidogenicprotein tau was examined by atomic force microscopy (AFM) and fluorescence microscopy. Although AcPHF5 did not showenhancement of thioflavin T (ThT) fluorescence in aqueous buffer, distinct aggregates were discernible when peptide wasdissolved in organic solvents such as methanol (MeOH), trifluoroethanol (TFE), and hexafluoroisopropanol (HFIP) dried onmica and examined by AFM. Self-association was evident even though the peptide did not have the propensity to formsecondary structures in the organic solvents. In dried films, the peptide adopts predominantly β-conformation which results inthe formation of distinct aggregates. ThT fluorescence spectra and fluorescence images indicate the formation of fibrils whenAcPHF6 solutions in organic solvents were diluted into buffer. AcPHF6 had the ability to organize into fibrillar structures whenAFM samples were prepared from peptide dissolved in MeOH, TFE, HFIP, and also when diluted into buffer. AcPHF6 showedpropensity for β-structure in aqueous buffer. In MeOH and TFE, AcPHF6 showed helical and β-structure. Morphology of thefibrils was dependent on peptide conformation in the organic solvents. The structures observed for AcPHF6 are formed rapidlyand long incubation periods in the solvents are not necessary. The structures with varying morphologies observed for AcPHF5and AcPHF6 appear to be mediated by surfaces such as mica and the organic solvents used for dissolution of the peptides.Copyright c© 2009 European Peptide Society and John Wiley & Sons, Ltd.

Keywords: amyloid fibrils; beta structure; peptide self-association; synthetic peptides

Introduction

There are several examples of short peptides that have the abilityto form amyloid fibrils depending on factors such as solvent fordissolution, pH of buffer, temperature, and incubation period[1–18]. It has been demonstrated that short peptides can alsobe designed de novo that have the ability to form amyloidfibrils [19,20]. In fact, tetrapeptides have been shown to formfibrils with morphology almost identical to fibrils associatedwith disease conditions [21]. Fibrils have also been obtainedfrom peptides that do not adopt β-structure and fold intohelix-turn-helix conformation [22]. The morphology of fibrils hasbeen examined by electron microscopy, atomic force microscopy(AFM), and fluorescence imaging techniques. Several studies haveindicated that the morphology of self-assembled peptides andproteins appears to depend on surfaces such as mica and graphite[23–31]. Few reports indicate that peptides derived from amyloid-forming proteins form nanotubes and other nanostructures withdifferent morphologies [32–34]. Formation of these structuresis mediated by organic solvents such as ethanol, acetone, andmethanol (MeOH). It appears that amyloid-forming peptides canform structures with varying morphologies on surfaces such asmica when deposited from organic solvents.

Short peptides that occur in the third repeat region of tau, amicrotubule associated protein that plays an important role ingrowth and maintenance of neuritic processes, have the ability toform fibrils [9,35]. In humans, there are six isoforms of tau. All thesix isoforms can self-assemble to form paired helical filaments, are

hyperphosphorylated, and give rise to neurofibrillary tangles inAlzheimer’s disease [36,37]. Fibril formation in a family of peptidesderived from the sequence VQIVYK indicated that Ac-VQIVYK-am (AcPHF6) is one of the most rapidly aggregating peptides.Fibril formation is complete in less than 20 min [9]. We haveinitiated investigations on the self-assembly of amyloid-formingpeptides in organic solvents [38]. In this paper, we describethe self-association of two tau-related peptides Ac-VQIVYK-am(AcPHF6) and Ac-QIVYK-am (AcPHF5) when dissolved in water andthree alcohols: MeOH and structure-inducing fluorinated alcohols,trifluoroethanol (TFE), and hexafluoroisopropanol (HFIP).

Materials and Methods

Materials

9-Fmoc amino acids were purchased from NovabiochemAG (Switzerland) and Advanced ChemTech (Louisville, KY).Peptide synthesis resin, PAL resin (5-(4-Aminomethyl-3,5-dimethoxyphenoxy)valeric acid resin), was purchased from

∗ Correspondence to: Ramakrishnan Nagaraj, Centre for Cellular and MolecularBiology, Council of Scientific and Industrial Research, Uppal Road, Hyderabad500 007, Andhra Pradesh, India. E-mail: [email protected]

Centre for Cellular and Molecular Biology, Council of Scientific and IndustrialResearch, Uppal Road, Hyderabad 500 007, Andhra Pradesh, India

J. Pept. Sci. 2009; 15: 675–684 Copyright c© 2009 European Peptide Society and John Wiley & Sons, Ltd.

67

6

N. CHAUDHARY, S. SINGH AND R. NAGARAJ

Advanced ChemTech (Louisville, KY). All other reagents wereof highest grade available.

Peptide Synthesis and Purification

Peptides were synthesized using standard Fmoc chemistry [39]and acetylated on resin [40]. The synthesized peptides werecleaved from the resin and deprotected using a mixture containing82.5% TFA, 5% phenol, 5% H2O, 5% thioanisole, and 2.5%ethanedithiol for 12–15 h at room temperature [41]. Peptideswere precipitated in ice-cold diethyl ether and purified on HewlettPackard 1100 series HPLC instrument on a reversed phase C18Bio-Rad column using a linear gradient of H2O and acetonitrile(0–100% acetonitrile) containing 0.1%TFA. Purified peptides werecharacterized using matrix-assisted laser desorption/ionizationtime-of-flight mass spectrometry on a Voyager DE STR massspectrometer (PerSeptive Biosystems, Foster City, CA) at theProteomics Facility in Centre for Cellular and Molecular Biology,India. After purification, the solvent (H2O–acetonitrile mixturecontaining 0.1%TFA) was evaporated and the peptides were storedas dry solids. Peptide stock solutions were prepared from the driedsolids in four different solvents: deionized water, MeOH, TFE, andHFIP. The concentrations of the peptides were calculated using amolar absorption coefficient of 1280 M−1 cm−1 at λ = 280 nm.

Thioflavin T Fluorescence Spectroscopy

Thioflavin T (ThT) fluorescence spectra were recorded onFluorolog-3 Model FL3-22 spectrofluorometer (Horiba Jobin Yvon,Park Avenue Edison, NJ). ThT fluorescence assay for AcPHF6 wasperformed using a modification of the method described by Naikiet al. [42]. Aliquots from 1.5 mM peptide stock solutions in H2O,MeOH, TFE, and HFIP were added to ThT (10 µM) in 50 mM phos-phate buffer, pH 7.0. The excitation wavelength was set at 450 nm,slit width at 2 nm, and emission slit width at 5 nm.

ThT fluorescence assay for AcPHF5 was performed using themethod described by Goux et al. [9]. The peptide stock solutionin water (1.5 mM) was diluted to a concentration of 10 µM into10 µM ThT in 20 mM MOPS buffer, pH 7.2 containing 150 mM NaCland fluorescence emission was recorded immediately and after20 min. The excitation wavelength was set at 440 nm, slit width at1 nm, and emission slit width at 5 nm.

ThT Fluorescence Microscopy

AcPHF6 (1.5 mM) from TFE and HFIP was diluted in ThT (10 µM) in50 mM phosphate buffer, pH 7.0 to get a concentration of 75 µM.These samples were incubated at room temperature. Microscopicslides were prepared by taking out 5 µl aliquots after different timepoints: immediately after diluting the peptide in ThT solution, afterday 1, and after day 5. Fluorescence images were recorded on ZeissAxiovert 200 microscope. Filter set with 450/50 nm excitation and510/50 nm emission was used. The images were pseudocolored.

Circular Dichroism

Circular dichroism (CD) spectra were recorded on Jasco J-715spectropolarimeter. Stock solutions of AcPHF5 (1.5 mM) andAcPHF6 (1.5 mM) were prepared in H2O, MeOH, TFE, and HFIP.Far-UV (195–250 nm) spectra of 100 µM peptides were recordedin H2O, MeOH, TFE, HFIP, and in 50 mM phosphate buffer, pH7.0 immediately after diluting from the H2O, MeOH, TFE, and HFIP

stock solutions. All the spectra were recorded in 0.1 cm path lengthcell using a step size of 0.2 nm, band width of 1 nm, and scan rateof 100 nm/min. The spectra were recorded by averaging 10 scansand corrected by subtracting the solvent/buffer spectra.

Secondary structural components were estimated using CON-TINLL program available in CDPro software package [43,44].Spectra were deconvoluted for data range of 195–240 nm usingSMP56 protein reference set [44].

Atomic Force Microscopy

For AFM studies, the peptide samples in solvents were depositedonto the freshly peeled mica surfaces and allowed to dry. Forpeptide samples in buffers, 10 µl volumes were deposited onfreshly peeled mica surface. After 10 min, the mica pieces werewashed gently with 40 µl deionized water and excess solutionremoved through the edge of mica using tissue paper. All thesamples were air dried before imaging. Samples in organic solventsand unbuffered water were not washed with any solvent. Theimages were acquired using tapping mode AFM (Multimode,Digital Instruments, Santa Barbara, CA). A silicon nitride probewas oscillated at 275–310 KHz and images were collected at anoptimized scan rate. Analysis was done using Nanoscope (R) III5.30 r1.

Fourier Transform Infrared Spectroscopy

Attenuated total reflection–Fourier transform infrared spec-troscopy (ATR–FTIR) spectra of dried peptide films were recordedon Bruker Alpha-E spectrometer with Eco ATR single reflectionATR sampling module equipped with ZnSe ATR crystal. AcPHF5(1.5 mM) and AcPHF6 (1.5 mM) solutions in MeOH, TFE, and HFIPwere spread out and dried as films on ZnSe crystal from eachsolvent. Each spectrum is the average of 24 FTIR spectra at aresolution of 4 cm−1.

Table 1. Results from deconvolution of CD spectra of AcPHF5 andAcPHF6 using CONTINLL software available in CDPro CD deconvolutionsoftware package

PeptideFraction

helixFraction

sheetFraction

turnFraction

unordered

AcPHF5 in H2O 0.052 0.225 0.231 0.492

AcPHF5 in MeOH 0.069 0.078 0.272 0.58

AcPHF5 in TFE 0.11 0.13 0.414 0.346

AcPHF5 in HFIP 0.062 0.195 0.294 0.448

AcPHF5 in Buffer from H2O 0.048 0.245 0.223 0.484

AcPHF5 in Buffer from MeOH 0.063 0.105 0.232 0.6

AcPHF5 in Buffer from TFE 0.055 0.142 0.24 0.563

AcPHF5 in Buffer from HFIP 0.047 0.109 0.201 0.643

AcPHF6 in H2O 0.197 0.243 0.256 0.304

AcPHF6 in MeOH 0.407 0.168 0.213 0.211

AcPHF6 in TFE 0.239 0.182 0.239 0.34

AcPHF6 in HFIP 0.074 0.229 0.31 0.388

AcPHF6 in Buffer from H2O 0.187 0.234 0.296 0.285

AcPHF6 in Buffer from MeOH 0.19 0.367 0.226 0.217

AcPHF6 in Buffer from TFE 0.038 0.323 0.228 0.41

AcPHF6 in Buffer from HFIP 0.06 0.298 0.201 0.442

Deconvolution was done for data range of 195–240 nm using SMP56protein reference set.

www.interscience.com/journal/psc Copyright c© 2009 European Peptide Society and John Wiley & Sons, Ltd. J. Pept. Sci. 2009; 15: 675–684

67

7

SELF-ASSEMBLY OF SHORT TAU PEPTIDES

Figure 1. Fluorescence of ThT (10 µM) in 50 mM phosphate buffer, pH 7.0with increasing concentration of AcPHF6. Stock solutions (1.5 mM) wereprepared in H2O, MeOH, TFE, and HFIP. Fluorescence emission at 480 nmwas plotted against peptide concentration. Inset shows the zoomed regionat lower peptide concentrations.

Results and Discussion

ThT Fluorescence Spectroscopy

AcPHF5 and AcPHF6 were dissolved in H2O, MeOH, TFE, and HFIPat a concentration of 1.5 mM. When AcPHF6 solutions were dilutedinto phosphate buffer, pH 7.0, concentration dependent increasein ThT fluorescence was observed (Figure 1). The increase influorescence, indicative of amyloid fibril formation, was dependenton the solvent the peptide was dissolved in. Maximum increasewas observed on dilution from TFE and HFIP. The data in the insetshow increase in fluorescence even at 6 µM peptide concentration.The increase is much more beyond a concentration of 40 µM.The data shown in Figure 1 indicate that the extent of fibril

formation depends on the organic solvent in which the peptide isdissolved and is initiated even at low concentration of the peptide.Although the concentration of the organic solvents is <7.5% inthe aqueous solution (<2.5% at 40 µM peptide concentration),solvent dependent fibril formation is observed. It appears that‘solvent history’ plays a role in fibril formation. No increase in ThTfluorescence was observed when similar experiments were carriedout with AcPHF5 (data not shown). The interaction of AcPHF6 fibrilswith ThT was visualized by fluorescence microscopy. Fluorescenceimages were examined from the peptide deposited on glass slidesfrom buffer solutions into which the peptide was diluted from TFEand HFIP.

ThT Fluorescence Microscopy

Fluorescence images of AcPHF6 at 75 µM concentration in 50 mM

phosphate buffer, pH 7.0 from HFIP and TFE solutions are shown inFigure 2. Images were recorded at different time points. Top panels(Panels A–C) and bottom panels (Panels D–F) show imaging ofthe peptide diluted from HFIP and TFE solutions, respectively.Panel A shows images immediately after diluting the peptide inbuffer. Most fibrils are short, <1 µm in lengths, but longer fibrils∼5–8 µm are also observed. The data clearly show that very shortfibrils also bind ThT and cause increase in its fluorescence. Panel Bshows image recorded after 24 h. There is a distinct increase in thelengths of fibrils. Most fibrils are more than 1 µm in length. After120 h of incubation, there is further increase in length of the fibrils(Panel C). Fibrils between 2 µm and 5 µm in length are present.This data clearly show an increase in the lengths of fibrils withincubation time.

Long fibrils are seen immediately after diluting the peptidefrom TFE stock solution (Panel D). The lengths of the fibrils varyconsiderably. There are very short (<200 nm) to very large fibrils(>6 µm). After 24 h of incubation, shorter fibrils were observed.Fibrils >6 µm are rarely seen (Panel E). When fluorescence imageswere recorded after 120 h of total incubation, decrease in length

Figure 2. ThT fluorescence images of AcPHF6. Peptide was diluted to 75 µM in 50 mM phosphate buffer, pH 7.0 from 1.5 mM stock solutions in HFIP andTFE. Samples were incubated at room temperature. Microscopic slides were prepared by taking out 5 µl aliquots after different time points. Panels A–Cand D–F show images obtained on dilution from HFIP and TFE, respectively: (A, D) immediately after dilution, (B, E) 24 h, and (C, F) 120 h of incubation atroom temperature. Scale bars represent 10 µm.

J. Pept. Sci. 2009; 15: 675–684 Copyright c© 2009 European Peptide Society and John Wiley & Sons, Ltd. www.interscience.com/journal/psc

67

8


Figure 3. Far-UV CD spectra of AcPHF5 in (A) H2O, (B) MeOH, (C) TFE, and (D) HFIP. Peptide concentration was 100 µM.

Figure 4. Far-UV CD spectra of AcPHF5 in 50 mM phosphate buffer, pH 7.0 when peptide was diluted from different stock solutions: (A) H2O, (B) MeOH,(C) TFE, and (D) HFIP. The concentration of the peptide stock solutions was 1.5 mM and CD spectra were recorded at a concentration of 100 µM.

of the fibrils was observed. Most fibrils range 1–4 µm in lengths(Panel F).

ThT fluorescence images indicate that apart from time ofincubation in aqueous buffer, the solvent in which the stocksolution was prepared is also a determinant of fibrillar morphology.When stock solution was prepared in TFE, long fibrils were formedalmost immediately in buffer, whereas from HFIP, fibril formationwas slower. The stock solutions (1.5 mM) were diluted 20-foldto get a concentration of 75 µM. The amount of organic solventpresent in the buffer (∼5%), though small, could also conceivablyinfluence fibril formation. The conformation of the peptides inorganic solvents and on transfer from organic solvents to aqueousmedium was examined by CD spectroscopy.

Circular Dichroism

Far-UV CD spectra of AcPHF5 were recorded in unbuffered H2O,MeOH, and structure-inducing fluorinated alcohols: TFE and HFIP(Figure 3). In H2O, MeOH, and HFIP (Panels A, B, D), the peptide islargely unordered. In TFE (Panel C), two minima∼203 nm and∼217nm are observed. Table 1 shows the secondary structures obtainedby deconvolution of the CD spectra by CONTINLL software usingSMP56 protein reference data set [44]. In TFE, AcPHF5 shows11% helical content and ∼41% turn content with a decreasein unordered conformation. When diluted from H2O and organicsolvents into phosphate buffer (50 mM, pH 7.0), AcPHF5 is in largelyunordered conformation (Figure 4, Table 1).


67

9


Figure 5. Far-UV CD spectra of AcPHF6 in (A) H2O, (B) MeOH, (C) TFE, and (D) HFIP. Peptide concentration was 100 µM.

Figure 6. Far-UV CD spectra of AcPHF6 in 50 mM phosphate buffer, pH 7.0 when peptide was diluted from different stock solutions: (A) H2O, (B) MeOH,(C) TFE, and (D) HFIP. The concentration of the peptide stock solutions was 1.5 mM and CD spectra were recorded at a concentration of 100 µM.

CD spectra of AcPHF6 in different solvents are shown in Fig-ure 5. In H2O, a broad minimum centered ∼212 nm is observed.Deconvolution of spectra suggests an ensemble of structurespopulating helix, turn, and β-structure. In MeOH and TFE (Fig-ure 5(B, C)), two minima ∼205 nm and ∼215 nm are observed.Deconvolution of the spectra indicate that in MeOH and TFE,helical and β-conformation are populated (Table 1). The spec-trum in HFIP (Figure 5(D)) indicates that the peptide is largelyunstructured. In 50 mM phosphate buffer, pH 7.0, spectra char-acteristic of β-structure is observed when diluted from H2Oand MeOH (Figure 6(A, B)). Table 1 shows that when AcPHF6is diluted from H2O, the β-structure content is not changedsignificantly. However, when diluted from MeOH and TFE,β-structure content is almost doubled with a loss in helical struc-

ture. The spectrum when transferred from HFIP is characteristicof largely unordered conformation (Figure 6(D)). However, CDdeconvolution shows that there is an increase in β-structure con-formation with concomitant decrease in turn conformation. Thestructures adopted by AcPHF6 in buffer appear to depend on thesolvent in which the peptide is dissolved before dilution. Whena large fraction of the peptide populates secondary structures insolvents, as in H2O, MeOH, and TFE, dilution into buffer results inlargely β-structure.

Atomic Force Microscopy

AcPHF5 was dissolved in organic solvents and incubated at roomtemperature for 48 h. The samples were then deposited on mica


68

0


Figure 7. AFM images of AcPHF5 from stock solutions prepared in organic solvents in (A–C) MeOH, (D–F) TFE, and (G–I), HFIP. Concentration of peptidewas 1.5 mM. Scale bars represent 1 µm. Arrows in panels E and F indicate lateral assembly of thin fibers giving rise to thicker fibrils.

Figure 8. AFM images of AcPHF6 from stock solutions prepared in H2O and organic solvents in (A, B) H2O, (C, D) MeOH, (E, F) TFE, and (G, H) HFIP.Concentration of peptide was 1.5 mM. Scale bar in panel H represents 500 nm. In all other panels, scale bars represent 1 µm.

and AFM images were recorded. From MeOH, the peptide formsring-like structures, ranging in diameter from ∼200 to ∼800 nmand in height from ∼2 to 5 nm (Figure 7(A–C)). Small aggregatesalong with some structures arranged in circular fashion areobserved as shown in Figure 7(C). These small aggregates couldbe intermediates in ring formation. From TFE, the peptide formslong fibrous aggregates (up to ∼3 µm in length) having taperingends along with very few, small amorphous-like aggregates

(Figure 7(D–F)). Thicker fibers are clearly seen to be composedof 2 or more thinner fibers as depicted with arrows. Thickestfibers were ∼40 nm in diameter while thinnest fibers were <4 nmsuggesting a hierarchy in the lateral assembly of fibers where fibersof varying thickness ranging from <4 to ∼40 nm were observed.From HFIP, relatively short curved fibrous aggregates are observed(Figure 7(G–I)) along with very small spherical aggregates. Thefibers appear to bend, arrange both laterally and near their ends,


68

1


Figure 9. AFM images of AcPHF6 diluted into 50 mM phosphate buffer, pH 7.0 from 1.5 mM stock solutions in (A, B) H2O, (C, D) MeOH, (E, F) TFE, and(G–H) HFIP. Peptide was diluted to 100 µM from 1.5 mM stock solutions and panels represent images obtained from diluted solutions. AFM slides wereprepared immediately after diluting in the buffer. Scale bars represent 1 µm.

Figure 10. AFM images of AcPHF6 diluted into 5 mM phosphate buffer, pH 7.0 from 1.5 mM stock solutions in (A) H2O, (B) MeOH, (C) TFE, and (D) HFIP.Peptide was diluted to 100 µM from 1.5 mM stock solutions and panels represent images obtained from diluted solutions. AFM slides were preparedimmediately after diluting in the buffer. Scale bars represent 1 µm.

giving rise to ring-like structures. The time of incubation in organicsolvents is not important as similar images were obtained fromfreshly dissolved peptide solutions. AcPHF5 was diluted into 50 mM

phosphate buffer to a final concentration of 100 µM and depositedon freshly peeled mica surface within 5 min. AFM images indicatedthe presence of small amorphous aggregates (data not shown).Thus, AcPHF5 has the propensity to form ordered self-assembledstructures when samples are prepared from organic solvents withthe morphology dependent on the solvent. Morphology of thestructures also appears to depend on the conformation adoptedby the peptide in the organic solvent. Fibrillar morphology isdiscernible when sample was prepared from TFE in which afraction of AcPHF5 populates helical and turn conformations.When samples were prepared from MeOH and HFIP in which thepeptide was largely unordered, ring-like circular structures areobserved. The results indicate that organic solvents favor self-association of AcPHF5 on the mica surface although the peptidedoes not have the ability to form ordered aggregates in aqueousenvironment either in solution or in contact with mica surface.

The AFM images of AcPHF6 prepared from H2O, MeOH, TFE,and HFIP solutions are shown in Figure 8. From H2O, MeOH,and TFE solutions, AcPHF6 forms amyloid-like fibrils, majorityof them ranging from 4 to 12 nm in height, diameter rangecharacteristic of mature amyloid fibrils (Figure 8(A–F)). From HFIP,clumps of short and long fibrous aggregates were seen distributed

over the mica surface along with small spherical aggregates(Figure 8(G, H)), the thickest fibrils being ∼12 nm thick. The shortfibrils resemble amyloid protofibrillar aggregates in dimensions.Formation of fibrils correlates with the ability of AcPHF6 to foldinto ordered conformation in water, MeOH, and TFE. Fibrils areobserved irrespective of the conformation the peptide adopts indifferent solvents. Even when helical conformation is populated(in MeOH and TFE), fibrils are observed. Fibrillar morphology isless distinctive from HFIP solutions in which the peptide is largelyunordered. Panels in Figure 9 show images of fibrils obtainedfrom aqueous buffer into which the peptide was transferredfrom different solvents. The peptides were diluted into 50 mM

phosphate buffer, pH 7.0, to a concentration of 100 µM, i.e. theconcentration at which intense ThT fluorescence was observed.These samples were deposited on freshly peeled mica pieces inless than 5 min after dilution. Fibrils are observed when transferredfrom H2O to buffer (Figure 9(A, B)). Panels C and D in Figure 9 showaggregates when peptide was transferred from MeOH to buffer.Panel C shows large aggregates which are formed by the clumpingof fibrillar aggregates. Majority of the fibrils range between 5 nmand 15 nm in height and fibrils formed by lateral association oftengive rise to thicker structures (Panel D). Fibrils formed by AcPHF6when transferred from TFE to buffer range from 5 to 10 nm in heightand several micrometers in length (Figure 9(E, F)). Figure 9(G, H)shows the aggregates formed by AcPHF6 when transferred from


68

2


Figure 11. FTIR spectra in the amide I region of peptides when dried from organic solvents. Left panels (Panels A–C) represent spectra of AcPHF5while right panels (Panels D–F) show spectra of AcPHF6 when dried from MeOH (Top panels), TFE (Middle panels) and HFIP (Bottom panels). Peptideconcentration in solvents was 1.5 mM.

HFIP to buffer. There are thin fibrils where the thickness of thefibrils lies between 4 nm and 8 nm. Their dimensions and curvedmorphology are suggestive of protofibril-like structures. Fibrilsthicker than 8 nm are rarely seen. These fibrils do not appear asrigid as the fibrils obtained in other cases, and it is very likely thatamorphous looking aggregates are formed by their clumping asfibrillar structures appear to come out of these aggregates.

Aggregation was also examined in buffer at low ionic strength(5 mM). Peptide stock solutions were diluted into 5 mM phosphatebuffer, pH 7.0 to a final peptide concentration of 100 µM, anddeposited on mica in less than 5 min. Figure 10 shows theaggregates formed from H2O to buffer (Panel A), MeOH tobuffer (Panel B), TFE to buffer (Panel C), and HFIP to buffer(Panel D). Samples from H2O to buffer and TFE to buffer show

distinct fibrillar aggregates (Panels A and C). When diluted from

MeOH, amorphous aggregates are observed (Panel B). Short

fibrils with large aggregates are observed when diluted from

HFIP to buffer (Panel D). In the case of H2O to buffer sample,

the fibrils are ∼8–10 nm in height but lateral assembly gives

rise to thicker (up to 25 nm) fibrils. The fibrils formed from

TFE to buffer and HFIP to buffer samples range from thin

protofibrillar like (2 nm) to thicker mature fibril like (10–14 nm).

In high ionic strength buffer (50 mM), the fibrils are thicker and

longer than those obtained in 5 mM buffer. The formation of

fibrils from MeOH, TFE, and HFIP is severely compromised in

5 mM buffer when compared to those in high ionic strength

buffer.


68

3


The AFM images show the ability of AcPHF6 to form fibrillarstructures when deposited on mica from organic solvents aswell as from aqueous solutions. The morphology and dimensionsof the fibrils however depend on the solvent the peptide wasdissolved in.

Fourier Transform Infrared Spectroscopy

Amide I region in FTIR spectra of peptides and proteins originatespredominantly from C O stretching vibration making it sensitiveto secondary structures [45–47]. ATR–FTIR spectra were recordedon dry peptide films to correlate the peptide secondary structureswith self-assembled structures on mica (Figure 11). AcPHF5 adoptslargely β-structure in the dried films (Panels A–C). However, theabsorption band is at 1626 cm−1 when peptide is dried fromMeOH (Panel A) as compared to 1632 cm−1 when dried fromTFE or HFIP (Panels B and C). Panels D–F represent the spectraof dried AcPHF6 films. The peptide adopts predominantly β-structure when dried from MeOH (Panel D), TFE (Panel E), andHFIP (Panel F). When dried from MeOH, the major absorptionband is at 1628 cm−1 as compared to those dried from TFEand HFIP where major bands are at ∼1631 cm−1. A smallband at 1664 cm−1 from MeOH sample and at 1661 cm−1

from TFE and HFIP could arise due to the presence of turnconformations.

The drying process causes structural changes in the peptides.The structural changes seem to depend on the initial structuresadopted by the peptides in the solvents. Although, bothpeptides adopt β-structure as their major component, minordifferences clearly evident in the spectra can be attributed todifferences in the strength of hydrogen bonding in the peptides[45–47].

The AcPHF6 sequence is an important determinant of fibril for-mation in tau [9,35]. The peptide sequence has been investigatedin an effort to delineate hydrophobic and length requirementsfor fibril formation [9]. Our results indicate that although removalof valine from AcPHF6 results in loss of fibril-forming property inaqueous solution, organized structures are formed on the micasurface from organic solvents. AcPHF5 has low propensity forordered conformation in solution. Ordered aggregated structuresformed on mica with β-structure could arise as a result of largeincrease in peptide concentration during the drying process. Ag-gregation behavior of AcPHF5 and AcPHF6 indicates that organicsolvents and surfaces such as mica play an important role in mod-ulating the morphology of the aggregated forms of the peptides.We have shown the self-assembly of an amyloidogenic peptidefrom human β2-microglobulin on mica surface when dried fromorganic solvents [38]. Unlike AcPHF5 and AcPHF6, self-assembledstructures were largely helical when dried from TFE and HFIPwhile β-structure was populated when dried from MeOH. Themorphology of the aggregates was also different from those ob-tained from AcPHF5 and AcPHF6. Low propensity for amyloidtype fibril formation does not preclude the formation of orga-nized structures from organic solvents especially when dried onsurfaces. In the case of AcPHF6, although fibrillar morphologypredominates, the morphology depends on the solvent in whichthe peptide is dissolved. The structures are formed rapidly andlong incubation periods for the formation of amyloid fibrils are notnecessary.

Fibril-forming peptides could be attractive candidates forforming self-assembled structures with varying morphologies onsurfaces mediated by organic solvents.

Acknowledgements

We thank Kiran Kumar in Dr. Tushar Chakraborty’s laboratory inthe Indian Institute of Chemical Technology, Hyderabad, for helpin recording FTIR spectra. Funding from CSIR Network ProjectNWP0035 is gratefully acknowledged.

References

1 Tjernberg LO, Callaway DJ, Tjernberg A, Hahne S, Lilliehook C,Terenius L, Thyberg J, Nordstedt C. A molecular model of Alzheimeramyloid beta-peptide fibril formation. J. Biol. Chem. 1999; 274:12619–12625.

2 Balbach JJ, Ishii Y, Antzutkin ON, Leapman RD, Rizzo NW, Dyda F,Reed J, Tycko R. Amyloid fibril formation by A beta 16–22, a seven-residue fragment of the Alzheimer’s beta-amyloid peptide, andstructural characterization by solid state NMR. Biochemistry 2000;39: 13748–13759.

3 Balbirnie M, Grothe R, Eisenberg DS. An amyloid-formingpeptide from the yeast prion Sup35 reveals a dehydrated beta-sheet structure for amyloid. Proc. Natl. Acad. Sci. USA 2001; 98:2375–2380.

4 Cottingham MG, Hollinshead MS, Vaux DJ. Amyloid fibril formationby a synthetic peptide from a region of human acetylcholinesterasethat is homologous to the Alzheimer’s amyloid-beta peptide.Biochemistry 2002; 41: 13539–13547.

5 Reches M, Porat Y, Gazit E. Amyloid fibril formation by pentapeptideand tetrapeptide fragments of human calcitonin. J. Biol. Chem. 2002;277: 35475–35480.

6 Jones S, Manning J, Kad NM, Radford SE. Amyloid-forming peptidesfrom beta2-microglobulin-Insights into the mechanism of fibrilformation in vitro. J. Mol. Biol. 2003; 325: 249–257.

7 Nilsson MR, Dobson CM. In vitro characterization of lactoferrinaggregation and amyloid formation. Biochemistry 2003; 42: 375–382.

8 Zanuy D, Ma B, Nussinov R. Short peptide amyloid organization:stabilities and conformations of the islet amyloid peptide NFGAIL.Biophys. J. 2003; 84: 1884–1894.

9 Goux WJ, Kopplin L, Nguyen AD, Leak K, Rutkofsky M, Shanmu-ganandam VD, Sharma D, Inouye H, Kirschner DA. The formationof straight and twisted filaments from short tau peptides. J. Biol.Chem. 2004; 279: 26868–26875.

10 Frare E, Polverino De Laureto P, Zurdo J, Dobson CM, Fontana A. Ahighly amyloidogenic region of hen lysozyme. J. Mol. Biol. 2004; 340:1153–1165.

11 Petty SA, Adalsteinsson T, Decatur SM. Correlations amongmorphology, beta-sheet stability, and molecular structure in prionpeptide aggregates. Biochemistry 2005; 44: 4720–4726.

12 Wadai H, Yamaguchi K, Takahashi S, Kanno T, Kawai T, Naiki H,Goto Y. Stereospecific amyloid-like fibril formation by a peptidefragment of beta2-microglobulin. Biochemistry 2005; 44: 157–164.

13 Nelson R, Sawaya MR, Balbirnie M, Madsen AO, Riekel C, Grothe R,Eisenberg D. Structure of the cross-beta spine of amyloid-like fibrils.Nature 2005; 435: 773–778.

14 Rojas Quijano FA, Morrow D, Wise BM, Brancia FL, Goux WJ.Prediction of nucleating sequences from amyloidogenicpropensities of tau-related peptides. Biochemistry 2006; 45:4638–4652.

15 Yamaguchi K, Naiki H, Goto Y. Mechanism by which the amyloid-likefibrils of a beta 2-microglobulin fragment are induced by fluorine-substituted alcohols. J. Mol. Biol. 2006; 363: 279–288.

16 Ivanova MI, Thompson MJ, Eisenberg D. A systematic screen ofbeta(2)-microglobulin and insulin for amyloid-like segments. Proc.Natl. Acad. Sci. USA 2006; 103: 4079–4082.

17 Sawaya MR, Sambashivan S, Nelson R, Ivanova MI, Sievers SA,Apostol MI, Thompson MJ, Balbirnie M, Wiltzius JJ, McFarlane HT,Madsen AO, Riekel C, Eisenberg D. Atomic structures of amyloidcross-beta spines reveal varied steric zippers. Nature 2007; 447:453–457.

18 Hamley IW. Peptide fibrillization. Angew. Chem. Int. Ed. Engl. 2007;46: 8128–8147.

19 Lopez De La Paz M, Goldie K, Zurdo J, Lacroix E, Dobson CM,Hoenger A, Serrano L. De novo designed peptide-based amyloidfibrils. Proc. Natl. Acad. Sci. U.S.A. 2002; 99: 16052–16057.


68

4


20 West MW, Wang W, Patterson J, Mancias JD, Beasley JR, Hecht MH.De novo amyloid proteins from designed combinatorial libraries.Proc. Natl. Acad. Sci. USA 1999; 96: 11211–11216.

21 Tjernberg L, Hosia W, Bark N, Thyberg J, Johansson J. Chargeattraction and beta propensity are necessary for amyloid fibrilformation from tetrapeptides. J. Biol. Chem. 2002; 277: 43243–43246.

22 Fezoui Y, Hartley DM, Walsh DM, Selkoe DJ, Osterhout JJ, Teplow DB.A de novo designed helix-turn-helix peptide forms nontoxic amyloidfibrils. Nat. Struct. Biol. 2000; 7: 1095–1099.

23 Kowalewski T, Holtzman DM. In situ atomic force microscopy studyof Alzheimer’s beta-amyloid peptide on different substrates: newinsights into mechanism of beta-sheet formation. Proc. Natl. Acad.Sci. USA 1999; 96: 3688–3693.

24 Wang Z, Zhou C, Wang C, Wan L, Fang X, Bai C. AFM and STM studyof beta-amyloid aggregation on graphite. Ultramicroscopy 2003; 97:73–79.

25 Losic D, Martin LL, Aguilar MI, Small DH. Beta-amyloid fibrilformation is promoted by step edges of highly oriented pyrolyticgraphite. Biopolymers 2006; 84: 519–526.

26 Zhu M, Souillac PO, Ionescu-Zanetti C, Carter SA, Fink AL. Surface-catalyzed amyloid fibril formation. J. Biol. Chem. 2002; 277:50914–50922.

27 Blackley HK, Sanders GH, Davies MC, Roberts CJ, Tendler SJ,Wilkinson MJ. In-situ atomic force microscopy study of beta-amyloidfibrillization. J. Mol. Biol. 2000; 298: 833–840.

28 Ban T, Morigaki K, Yagi H, Kawasaki T, Kobayashi A, Yuba S, Naiki H,Goto Y. Real-time and single fibril observation of the formationof amyloid beta spherulitic structures. J. Biol. Chem. 2006; 281:33677–33683.

29 Green JD, Goldsbury C, Kistler J, Cooper GJ, Aebi U. Human amylinoligomer growth and fibril elongation define two distinct phases inamyloid formation. J. Biol. Chem. 2004; 279: 12206–12212.

30 Goldsbury C, Kistler J, Aebi U, Arvinte T, Cooper GJ. Watchingamyloid fibrils grow by time-lapse atomic force microscopy. J. Mol.Biol. 1999; 285: 33–39.

31 Nayak A, Dutta AK, Belfort G. Surface-enhanced nucleation of insulinamyloid fibrillation. Biochem. Biophys. Res. Commun. 2008; 369:303–307.

32 Li H, Zhang F, Zhang Y, He J, Hu J. Organic solvents mediate self-assembly of GAV-9 peptide on mica surface. Acta Biochim. Biophys.Sin. (Shanghai) 2007; 39: 285–289.

33 Krysmann MJ, Castelletto V, McKendrick JE, Clifton LA, WH I,Harris PJ, King SM. Self-assembly of peptide nanotubes in an organicsolvent. Langmuir 2008; 24: 8158–8162.

34 Krysmann MJ, Castelletto V, Hamley IW. Fibrillisation of hydrophobi-cally modified amyloid peptide fragments in an organic solvent. SoftMatter 2007; 3: 1401–1406.

35 von Bergen M, Friedhoff P, Biernat J, Heberle J, Mandelkow EM,Mandelkow E. Assembly of tau protein into Alzheimer paired helicalfilaments depends on a local sequence motif ((306)VQIVYK(311))forming beta structure. Proc.Natl.Acad.Sci.USA 2000; 97: 5129–5134.

36 Goedert M, Spillantini MG, Cairns NJ, Crowther RA. Tau proteins ofAlzheimer paired helical filaments: abnormal phosphorylation of allsix brain isoforms. Neuron 1992; 8: 159–168.

37 Goedert M, Spillantini MG, Jakes R, Crowther RA, Vanmechelen E,Probst A, Gotz J, Burki K, Cohen P. Molecular dissection of the pairedhelical filament. Neurobiol. Aging 1995; 16: 325–334.

38 Chaudhary N, Singh S, Nagaraj R. Organic solvent mediated self-association of an amyloid forming peptide from beta(2)-microglobulin: an atomic force microscopy study. Biopolymers 2008;90: 783–791.

39 Atherton E, Sheppard RC. Solid Phase Synthesis: A Practical Approach.IRL Press: Oxford, 1989.

40 Stewart JM, Young JD. Solid Phase Peptide Synthesis. Pierce ChemicalCompany: Illinois, 1984.

41 King DS, Fields CG, Fields GB. A cleavage method which minimizesside reactions following Fmoc solid phase peptide synthesis. Int. J.Pept. Protein Res. 1990; 36: 255–266.

42 Naiki H, Higuchi K, Hosokawa M, Takeda T. Fluorometricdetermination of amyloid fibrils in vitro using the fluorescent dye,thioflavin T1. Anal. Biochem. 1989; 177: 244–249.

43 Sreerama N, Woody RW. Estimation of protein secondary structurefrom circular dichroism spectra: comparison of CONTIN, SELCON, andCDSSTR methods with an expanded reference set. Anal. Biochem.2000; 287: 252–260.

44 Sreerama N, Woody RW. On the analysis of membrane proteincircular dichroism spectra. Protein Sci. 2004; 13: 100–112.

45 Haris PI, Chapman D. The conformational analysis of peptides usingFourier transform IR spectroscopy. Biopolymers 1995; 37: 251–263.

46 Pelton JT, McLean LR. Spectroscopic methods for analysis of proteinsecondary structure. Anal. Biochem. 2000; 277: 167–176.

47 Surewicz WK, Mantsch HH, Chapman D. Determination of proteinsecondary structure by Fourier transform infrared spectroscopy: acritical assessment. Biochemistry 1993; 32: 389–394.


Documents

Morphology of self-assembled structures formed by short peptides from the amyloidogenic protein tau depends on the solvent in which the peptides are dissolved