Improved electrical conductance through self-assembly of bioinspired peptides into nanoscale fibers

lable at ScienceDirect

Materials Chemistry and Physics xxx (2015) 1e8

Contents lists avai

Materials Chemistry and Physics

journal homepage: www.elsevier .com/locate/matchemphys

Improved electrical conductance through self-assembly of bioinspiredpeptides into nanoscale fibers

Rhiannon C.G. Creasey*, Yoshitaka Shingaya, Tomonobu Nakayama**

International Center for Materials Nanoarchitectonics (MANA), National Institute for Materials Science (NIMS), Namiki 1-1, Tsukuba, Ibaraki 305-0044,Japan

h i g h l i g h t s

* Corresponding author. Present address: SchoolNottingham, Nottingham NG7 2RD, United Kingdom.** Corresponding author.

E-mail addresses: [email protected]@nims.go.jp (T. Nakayama).

http://dx.doi.org/10.1016/j.matchemphys.2015.03.0340254-0584/© 2015 Elsevier B.V. All rights reserved.

Please cite this article in press as: R.C.G. Crnanoscale fibers, Materials Chemistry and P

g r a p h i c a l a b s t r a c t

� We designed a novel peptidesequence from natural amino acidswhich forms nanofibers in water.

� The nanofibers were investigated byAFM, fluorescence and CDspectroscopy.

� A film of self-assembled peptideshows conductivity in air andvacuum.

� We propose that stacking of phenyl-alanine between peptides leads toconductivity.

a r t i c l e i n f o

Article history:Received 26 December 2014Received in revised form19 February 2015Accepted 23 March 2015Available online xxx

Keywords:BiomaterialsAtomic force microscopy (AFM)Electrical conductivityNanostructures

a b s t r a c t

We investigated the electrical conductance of films consisting of bio-inspired peptide molecules and oftheir extended form, self-assembled nanoscale fibers. Here, the entirely natural and novel peptidesequence, GFPRFAGFP, was designed based on naturally occurring fibrous proteins. To attain electricalconductance, we implemented phenylalanine residues in the sequence such that the aromatic rings arepresent along face of the molecule. We confirmed self-assembly of nanoscale fibers in pure water afterincubating the peptides at 37 �C by AFM. The morphology and conformation of the incubated peptidefibers were studied using AFM, fluorescence spectroscopy and circular dichroism spectroscopy. It wasshown that very thin fibers with a single-molecule-level diameter form. The helical feature of the peptidebackbone and enhanced stacking of aromatic residues were also investigated. This aromatic stacking isimportant to our electrical measurements as, even in vacuum environment, films of non-incubatedGFPRFAGFP sometimes show apparent conductance while those containing self-assembled nanoscalefibers show stable and improved conductance. We propose that this effect may be due to extendedstacking of aromatic residues providing p e p conjugation along the fiber.

© 2015 Elsevier B.V. All rights reserved.

of Pharmacy, University of

am.ac.uk (R.C.G. Creasey),

easey, et al., Improved electrhysics (2015), http://dx.doi.o

1. Introduction

In the last decade, increasing needs for electrically conductivenanomaterials have been emerging for use in electronic devices [1],medical [2,3], and sensing [4,5] applications, among others. Inparticular, one dimensional conductive nanostructures are essen-tial for the construction of nanodevices [5e7]. One method of

ical conductance through self-assembly of bioinspired peptides intorg/10.1016/j.matchemphys.2015.03.034

R.C.G. Creasey et al. / Materials Chemistry and Physics xxx (2015) 1e82

construction at the nanoscale is utilizing self-assembly. Much self-assembly research has focused on mimicking nature's buildingblocks [8e10]; however, without modification these organic ma-terials are intrinsically insulating, or conduct via relatively slowmechanisms [11]. Some notable exceptions exist, such as the bac-teria Geobacter Sulfurreducens, capable of producing nanowireswith metal-like conductivity [5,12,13]. Also, recent studies haveshown that electron conductivity through natural peptides istheoretically plausible [12,14]. The suggestedmanner of conductionis via delocalized electron states and enhanced molecular orbitalstates from self-assembly of aromatic and charged residues. How-ever, thus far, the potential of this emergent functionality has notbeen utilized.

In this work, we present a novel short peptide formed entirelyfrom naturally occurring amino acids which dynamically self-assembles into conductive fibers. This peptide, containing resi-dues of glycine (G), phenylalanine (F), proline (P), arginine (R), andalanine (A) in the sequence GFPRFAGFP, was shown to self-assemble under aqueous conditions, forming long nanofibers.These nanofibers form from left-handed helical backbone confor-mations of the monomer peptides, allowing aromatic groupstacking. It was shown that a film made from these nanofibers isconductive, highlighting the potential for electronic functionality toemerge from the assembly of bioinspired materials. This andsimilar peptides offer controlled functionality and interestingproperties suitable for a variety of nanoelectronic and bio-nanotechnological applications.

2. Materials and experimental methods

2.1. Peptide

De Novo design of peptide was inspired by the fibrous proteinsfound in the extracellular matrix of mammalian tissues, such aselastin and collagen [15,16]. Utilizing similar sequences, the peptidebackbone comprising of GFPRFAGFP was prepared for this study forthe following two reasons: First, GFPRFAGFP was expected to adopta helical structure [15,17], allowing for one-dimensional assembly.Second, periodically inserted F was expected to promote intermo-lecular coupling [18e22], and increase conductance of the assem-bled construct owing to p e p conjugation [23.].

The chemical structure and the ball-and-stick model of aGFPRFAGFP molecule are shown in Fig. 1(a) and (b), respectively.

Also, the calculated polar surface area of the molecule is shownin Fig. 1(b), where the blue and the red areas highlight polar andapolar residues, respectively. Amphiphiles such as this moleculehave been extensively exploited for controllable aqueous self-assembly [8,24e26]. Fig. 1(c) e 1(e) show the CPK model of a sin-gle peptide down different axes of view, expressing dimensions ofapproximately 15.8 � 8.9 � 28.6 Å. The peptide model with PPIIbackbone [27] was initially built using VegaZZ (v3) [28]. Afterinserting a water layer of 10 molecules thickness, molecularconformation of the model was further optimized through MM2force field method [29].

All water used in this study is 0.22 mm filtered ultrapure MilliQwater (>18.2 MU cm). The designed peptide molecule, NH2-GFPRFAGFP-COOH, was synthesized and purified by EurofinsOperon (Operon Bio-Technology Co., Ltd., Japan). 1 mg lyophilizedaliquots were prepared in water to a concentration of 20 mg/mL asstock solution. Prior to further dilution as described, the stock so-lution was sonicated in an ice bath for 30 s.

2.2. Atomic force microscope (AFM) observations

GFPRFAGFP solution of 10 mg/mL in water was prepared and

Please cite this article in press as: R.C.G. Creasey, et al., Improved electrnanoscale fibers, Materials Chemistry and Physics (2015), http://dx.doi.o

incubated at 37 �C either overnight or for 7 days. The incubatedpeptide solutionwas further suspended to a concentration of 1 mg/mL inwater, and 5 mL was placed onto a freshly cleavedmica surfaceand incubated for 15 min. Prior to AFM observations, the samplewas again rinsed in water and dried under pure Argon gas toremove excess peptide from the surface. For comparison, peptidesamples which were prepared immediately after dilution (non-incubated) were used as a control. Images were acquired in ACtapping mode on an Asylum MFP3D SPM (Asylum Research, USA)using Olympus OMCL-AC240TS probes (nominal values: k ¼ 1.8 N/m, f ¼ 75 kHz, radius > 10 nm) in air at RT. A free air amplitude of500 mV was used (20e50 nm) with minimum engage setpoint(<10% of free air amplitude). The phase was monitored to ensurestable imaging conditions. A phase increase was observed onengaging with the surface, indicating that imaging was in therepulsive regime.

2.3. Optical spectroscopy measurements

GFPRFAGFP solution of 10 mg/mL was used for the opticalspectroscopy measurements of both incubated and non-incubatedpeptide. Samples were diluted to 0.1e0.3 mg/mL (measured by UVabsorbance at 257 nm) in water immediately prior to the followingmeasurements.

2.3.1. Fluorescence spectroscopyFluorescence spectroscopy measurements were carried out at

room temperature on a Hitachi F-7000 (Japan) fluorescence spec-trophotometer with a small volume cuvette (pathlength 0.1 cm)with excitation wavelength of 265 nm, emission wavelength in therange of 275e500 nm, and a scan speed of 240 nm/min. Emissiondata was normalized against the fluorescence peak of phenylala-nine (~281 nm) in order to average results of repeated measure-ments on separately prepared samples.

2.3.2. Circular dichroism (CD) spectroscopyCD spectroscopy was performed on a J-725 spectrophotometer

(Jasco Co. Ltd., Japan) using a small-volume cuvette (pathlength0.1 cm). CD spectra were acquired in the range of 185e500 nmwitha step of 0.2 nm, at a scan rate 100 nm/min. Dichroic angle wasmeasured with a resolution of 100 mdeg under an N2 gas flow rateof ~25 L/min. To reduce noise, we averaged over 3 scans per sample.We acquired CD spectra at temperatures of 4, 21, and 60 �C.

2.4. Electrical measurements

Electrical conductances were investigated for the films ofincubated and non-incubated peptide, biotin, and insulin, using aKeithley 4200 semiconductor parameter analyzer equipped with a4200-preamplifier (Keithley Instruments Inc., USA) which enabledthe highest relative resolution of 100� 10�18 A in current mea-surement. To prepare the films, the method was similar to that forAFM observations: 5 mL droplets of incubated and non-incubatedpeptide, biotin and insulin (SigmaeAldrich) solutions (10 mg/mLin water) were cast onto freshly cleaved mica substrates and driedunder Argon gas. Then, two molybdenum plates with a size of3mm� 6mmwere placed with a spatial gap of 1 mm, directly ontothe sample films (Fig. S1): Effectively, conductances of the filmswith a width of 3 mm and a length of 1 mm were measured byflowing current in the direction of the length. The currentevoltagecharacteristics (IeV curves) were collected by sweeping the voltageof one electrode from �2 to þ2 V (and vice versa) with steps of0.01 V while the other electrode was grounded. Measurementswere performed in ambient conditions or under vacuumconditions.

ical conductance through self-assembly of bioinspired peptides intorg/10.1016/j.matchemphys.2015.03.034

Fig. 1. (a) Chemical structure and (b) ball-and-stick 3D model with polar surface area layer superimposed (blue ¼ polar, red ¼ non-polar) of peptide GFPRFAGFP. (cee) CPK modelsof peptide in PPII structure with mm2 minimization applied along the (c) z, (d) x, and (e) y axes; LR, dR and dF are approximately 7.7, 3.1 and 4.0 Å, respectively. (For interpretation ofthe references to color in this figure legend, the reader is referred to the web version of this article.)

R.C.G. Creasey et al. / Materials Chemistry and Physics xxx (2015) 1e8 3

3. Results and discussion

3.1. AFM of peptide nanofibers

GFPRFAGFP dissolved readily in pure water at 20 mg/mL, withno precipitates, indicating no issue with solubility in aqueous sol-vent due to hydrophobic residues. By incubating the peptide so-lution at the physiologically relevant 37 �C, peptide fiber formationwas promoted as confirmed by AFM observations as follows.

In Fig. 2 (a), an AFM image shows fibrous structures existing on asample prepared by dropping a small volume of incubated peptideonto mica. It reveals a large number of thin fibers along with somethicker fibers, an example of which is indicated by the arrow headin Fig. 2 (a). Another example of thin fibers is shown at higherresolution in Fig. 2 (b). They are fairly uniform in height except forsmall bright spots appearing along the fiber. Although these fiberssometimes have tangles or branches as indicated by the arrowheads in Fig. 2 (b), they were predominantly individual fibers, witha length often more than 10 mm. We note that shorter segments(<100 nm) were less frequently observed (Fig. S2). This suggests afiber formation pathway which involves an infrequent critical nu-cleus step, after which a fast lag phase enables peptides to easilyaggregate into long fibers [30].

Long-range height modulations are observed along thick fibers(for example, Fig. 2 (c)). Such a modulation of the fiber heightsuggests hierarchical assembly of the smaller fibers into larger fi-bers, as schematically shown in Fig. 2 (d). In addition, we haveobserved branching of thicker fibers into thinner fibers (Fig. S3)

Please cite this article in press as: R.C.G. Creasey, et al., Improved electrnanoscale fibers, Materials Chemistry and Physics (2015), http://dx.doi.o

indicating hierarchical assembly.The height measurements of randomly selected 171 fibers are

summarized in Fig. 2 (e). As we see in the figure, small fibers of0.4e0.6 nm in height are very common, while no particular heightis dominant in the larger bins, which is also consistent with theabove speculation that a hierarchical assembly of smaller fibersforms the larger fibers, as opposed to an alternative assemblypathway.

The smaller fibers as shown in Fig. 2 (b) are likely a 1D assemblyof the peptide, with the most commonly measured height of about0.5 nm. This value is about 0.4 nm smaller than the size of a singlepeptide in the z axis, 0.89 nm [see Fig. 1(c)]. The difference betweenmodel height and measured height here may be due to an imagingcondition in the repulsive regime, which can reduce the apparentheight of features by up to 0.5 nm [31,32]. It should be noted that foran accurate height measurement, the sample modulus and condi-tions of imaging (such as the imaging force) must be understoodand accounted for. For example, in order to quantify the imagingforce, the cantilever must be carefully calibrated [33e35]. For thepeptide fibers observed herein, the exact height remains unknown.However, no possible model of a peptide monomer can account forthe height which has been measured in the more stable repulsiveregime. Therefore, our measurements only indicate that the pep-tide assembly is 1D, and can be considered to be in the samedimensional range as the model in Fig 1 (c).

Regarding the bumps of 0.14 nm (±0.10 nm, n ¼ 42) in heightalong the small peptide fibers [Fig. 2 (b)], they are unlike featuresseen on control samples (Fig. S4). It is, therefore, natural to think

ical conductance through self-assembly of bioinspired peptides intorg/10.1016/j.matchemphys.2015.03.034

Fig. 2. AFM height images in air of GFPRFAGFP fibers on mica showing (a) many thin fibers and one thicker fiber, (b) the smaller fibers, and (c) larger fiber with a twisted structureshown by irregular asymmetrical height features along the fiber. (d) Frequency distribution histogram of fiber heights measured by AFM section profiles, and (e) schematic ofpotential mechanism of hierarchical assembly of peptide into different sizes and shapes of fiber.

Fig. 3. Fluorescence emission spectra before (blue, dashed line) and after (red, solidline) incubation at 37 �C. (For interpretation of the references to color in this figurelegend, the reader is referred to the web version of this article.)

R.C.G. Creasey et al. / Materials Chemistry and Physics xxx (2015) 1e84

that they are part of peptides and are not due to contamination.Presumably, the bumps correspond to residues, such as R or F, oc-casionally pointing away from the peptide fiber. As shown in Fig. 1(e), R (or F) residue may protrude with a height of dR (dF) to LR (LF)which are estimated to be 0.4e0.8 nm based on van derWaals radiiof the elements. If we take the effect of under-estimation in a heightprofile into account, the observed heights of the bumps agree withthe height of irregularly protruded residues. This also suggests thatthe peptide fibers have some structural disorder in the 1Dassembly.

3.2. Optical spectroscopy of peptide solutions

Further insights into the structures of GFPRFAGFP monomersand fibers in solutionwere obtained through optical spectroscopiesas described below.

As shown by the blue dotted line in Fig. 3, fluorescence spec-troscopy measurement on peptide samples in solution before in-cubation, with an excitation of 265 nm, exhibits an emission peakaround 285 nm corresponding to F fluorescence. After incubation,there appears an additional peak just below 300 nm and alengthening of the emission tail after 340 nm as shown by a redsolid line. These features agree well with a reported effect of aro-matic interactions such as p e p stacking for fiber formation[18,36]. However, the effect is not pronounced in this samplebecause amorphous peptide remains after incubation as confirmedby AFM; this most likely reflects a relatively small amount of fibergrowth after incubation relative to the non-assembled peptide insolution.

To determine the secondary structure of the peptide backbone,CD spectroscopy and analyses were carried out by comparingmeasured spectra to known samples [36e38]. The peptideGFPRFAGFP is shown to contain a primarily left-handed helicalstructure, although the aromatic residues can confound estimationof secondary structure by CD spectroscopy [38,39].

Please cite this article in press as: R.C.G. Creasey, et al., Improved electrnanoscale fibers, Materials Chemistry and Physics (2015), http://dx.doi.o

As presented by blue solid lines in Fig. 4 (a) e (c), the non-incubated peptide monomers in solution at different tempera-tures of 4, 21, and 60 �C showed unique CD spectra in the far-UVrange. At 4 �C, where solvent interactions are minimized, thenon-incubated peptide shows a maximum around 220 nm. Thispeak is due to a combination of PPII conformation of the amidebackbone and the p e p* transitions of the aromatic residues ac-cording to previous studies [37,38]. A local minimum around200 nm is similar to that observed for denatured collagen and otherPPII helices [27,37,40]: the minimum is followed by a localmaximum around 191 nm and a negative drop-off towards 185 nm.These features below 195 nm are considered to be consistent withPPII [38,41] or may indicate a small contribution from another

ical conductance through self-assembly of bioinspired peptides intorg/10.1016/j.matchemphys.2015.03.034

Fig. 4. CD spectra of non-incubated (blue) and incubated (red) peptide in watermeasured at (a) 4 �C, (b) 21 �C and (c) 60 �C. (For interpretation of the references tocolor in this figure legend, the reader is referred to the web version of this article.)

Table 1FTIR assignments of deconvoluted peaks.

Peak Assignments (reference)

1625 cm�1 PPII [44], R stretching [45], aggregation [46], short extendedchains [47]

1652 cm�1 R stretching [48], disordered [49,50], a-helix [46,47]1675 cm�1 (C]O without H-bonds) PPII [44,46,51], b -turns [45,49,50]1694 cm�1 b-sheet [45], b-turn [52], PPII [53,54]

R.C.G. Creasey et al. / Materials Chemistry and Physics xxx (2015) 1e8 5

structure, such as b-turns [37,38]. Due to the short sequence of thepeptide and the aromatic contribution, it was impossible to confirmwithout ambiguity the secondary structure compared to referencespectra via traditional analyses using LINCOMB or Pepfit [42,43]. Inthe near-UV/visible region of the spectra (Fig. S5), a weakminima isobserved at 280 nm, followed by a very weak local maxima at375 nm, characteristic of F sidechains [39].

In support of the CD results, deconvolution of ATR-FTIR per-formed on solid-state films (Fig. S6) shows only minor conforma-tional changes before and after incubation in the amide I and IIregions. Peaks are observed at 1625, 1652, 1675 and 1694 cm�1 inthe amide I region; however, these are not trivial to assign forconformational structure. Some literature assignments are listed inTable 1.

Please cite this article in press as: R.C.G. Creasey, et al., Improved electrnanoscale fibers, Materials Chemistry and Physics (2015), http://dx.doi.o

In combinationwith the CD spectra, these results are most likelyconsistent with a primarily PPII backbone with some amount ofturn or disorder; this uncertainty can be attributed to the shortlength of the amino acid sequence. The structure of the assemblydiscussed herein is based upon the spectroscopic data primarilyachieved in solution, while the fiber formation and conductivitymeasurements are observed in the solid state. The FTIR data sug-gests that the peptide structure is not significantly altered in thesolid state; indeed, many protein assemblies have been reported inthis manner. Nonetheless it is worth noting that the peptideconformation may be slightly altered in the conductivitymeasurements.

After incubation under conditions shown to create peptide fi-bers, some changes are observed in the CD spectra as shown by redsolid lines in Figs. 4(a) e (c). At 4 �C, all peaks appear to be lessintense; this suggests greater flexibility in the peptide chromo-phores after self-assembly [38,55]. This seems counterintuitive, butmay in fact reflect altered hydrogen bonding states. The FTIRspectra in the Amide II and A regions confirm that hydrogenbonding has been altered after incubation (Fig. S6).

At 21 �C, the CD spectra from monomers and fibers are almostidentical, and remain similar even at 60 �C. However, the spectra at60 �C no longer display any peaks other than that attributed pri-marily to the aromatic chromophore at 220 nm, and this peak ismore pronounced in the incubated sample. Also, the minima at200 nm due to F are less intense after incubation, reflecting analtered electronic transition in the F residues after self-assembly.

3.3. Consideration on the peptide fiber formation

Self-assembly experiments were attempted in PBS (1X, pH 7.4),MES (0.1 M, pH 6) and KOH (0.5 mM, pH 10.7) buffers, along withacetonitrile, chloroform and ethanol solvents, in order to confirman electrostatic role of R residue in the peptide fiber formation.Fibers did not form under those conditions as judged from AFMobservations, although some form of 2D self-assembly was recog-nized in acetonitrile.

The spectroscopic observations suggest that only a minorconformational change is required to form fibers as we see in AFMimages (see Fig. 2). A possible model of the fibrous structures areschematically shown in Fig. S7, but we need further clarification forthe details of the conformational information and the growthprocesses of the peptide fibers. Nonetheless, the sequence of aminoacids is crucial to the formation of conductive fibers. It is wellknown that the presence and location of phenylalanine in a peptidesequence has a great impact on self-assembly [56], from amyloid-based sequences [18,57] to very short peptide based molecules[22,58]. Furthermore, the use of a helical peptide backbone allowsdifferent ‘faces’ of the peptide to be available for different func-tionalities [59], therefore aligning the F residues for closer packingin a self-assembled structure.

It is reasonable to think that the limiting step for fiber growth isthe formation of the critical nucleus. This structuremay be unstableor conformationally not favored at room temperature in water, asonly a very small number of short fibers are observed (Fig. S2). Thus

ical conductance through self-assembly of bioinspired peptides intorg/10.1016/j.matchemphys.2015.03.034

R.C.G. Creasey et al. / Materials Chemistry and Physics xxx (2015) 1e86

raising the temperature was necessary for formation of fibers, andfurther investigation of conditions may lead to an even greateryield of fibers.

3.4. Conductivity measurements

Now, we discuss electrical conductance of the films of incubatedpeptide fibers and other reference samples based on the resultsobtained by 2 terminal IeV measurements.

Fig. 5 (a) summarizes the measured conductance over 139measurements in air with a relative humidity of 50e60 % at20e25 �C, and in a vacuum of 1 e 5 � 10�6 mbar at 19 �C.Conductance in the range of 10�16 to 10�10 S was observeddepending on the type of sample and on the environmental

Fig. 5. (a) Histogram graph of conductance across sample films (as shown on the inset sch(orange) conditions. The size of each data point represents the portion of results for that samsweep) measured by 2 probes with 1 mm gap for freshly cleaved mica (orange), and films(blue) and after (red) incubation. These are representative curves taken from the same sampconditions. (For interpretation of the references to color in this figure legend, the reader is

Please cite this article in press as: R.C.G. Creasey, et al., Improved electrnanoscale fibers, Materials Chemistry and Physics (2015), http://dx.doi.o

conditions. The size of the circle in the figure represents the ratio ofobservations at corresponding conductance and the color of thecircle represents the environment (see figure caption). Measuredconductance for incubated peptide films in air is in the range of10�11 to 10�10 S (mean 84.44 pS (SD 77.0, n ¼ 35)) whereas theyshow slightly lower conductance, mostly around 10�11 S (mean24.204 pS (SD 15.98, n ¼ 18)), in vacuum condition. The films ofnon-incubated peptide monomers show conductance of the orderof 10�11 S (mean 32.70 pS (SD 26.4, n ¼ 25)) in air which is onlyslightly smaller than the case of the fibers. However, in vacuum, theconductance of the peptide monomer films widely scatters across arange of 10�14 to 10�11 S (mean 6.778 pS (SD 9.75, n ¼ 6)), which isaround 100 times smaller than the case of the peptide fiber films.An independent T-test of the vacuum condition incubated and non-

ematic of experimental setup), showing data points from ambient (blue) and vacuumple set which fall in the corresponding conductance. Representative IeV curves (dual

prepared on mica from insulin (purple), biotin (green) and GFPRFAGFP peptide beforele under (b) ambient (20e25 �C, 50e60 % RH) and (c) vacuum (1e5 � 106 mbar, 19 �C)referred to the web version of this article.)

ical conductance through self-assembly of bioinspired peptides intorg/10.1016/j.matchemphys.2015.03.034

R.C.G. Creasey et al. / Materials Chemistry and Physics xxx (2015) 1e8 7

incubated peptide results shows a significant difference in themean conductance acquired, despite the large standard deviations(P (T � t) two-tail ¼ 0.006).

Interestingly, the control samples of biotin, insulin and the baremica all show similar conductance in air and in vacuum: In air,conductance distributes across a range of 10�13 to 10�11 S (mean6.10 pS (SD 9.5, n ¼ 12), mean 18.33 pS (SD 18.9, n ¼ 13), and mean10.04 pS (SD 11.6, n ¼ 15), respectively) which is an order ofmagnitude smaller than the case of the peptide fiber, and, in vac-uum, all the measured conductance drops to the order of 10�14 S(mean 0.024 pS (SD 0.01, n¼ 4), mean 0.027 pS (SD 0.03, n¼ 4), andmean 0.016 pS (SD 0.01, n ¼ 8), respectively). In the case of controlsamples measured in air, some overlap of results occurs with thepeptide samples. This is likely due to a water layer on the samplesurface in ambient conditions, which may allow a small amount ofconductivity; this conductivity overlap of the control samples isgone in the vacuum condition.

Fig. 5 (b) and (c) show typical IeV curves measured in air and invacuum, respectively. It is noted that the measured IeV curves lookalmost linear but there always exists non-linear feature presum-ably owing to an effect of capacitance in our electrical circuitincluding the sample itself. Therefore, to minimize the effect, it wasquite important to perform themeasurement quite slowly, i.e., witha voltage sweep at a rate of less than 0.1 V/min. Here, we can clearlyunderstand that only an incubated peptide fiber film shows similarconductance both in air and in vacuum.

As we see in Fig. 5 (a), the conductance of the GFPRFAGFPpeptide films are, on average, higher than those of biotin and in-sulin films and that of the bare mica surface. Furthermore, incu-bated peptide fiber films reproducibly exhibit larger conductancethan non-incubated peptide monomer films especially in vacuum.Thus, we conclude that the electrical conduction through theincubated peptide fiber films is mostly realized without protondiffusion [10]. The other films (except for the non-incubated pep-tide) do require a water layer to show a conductance around10�11 S.

The reason why the non-incubated peptide monomer filmsshow variety of conductance in vacuum is surmised as follows.Owing to the existence of aromatic residues and possibly to afragmental p e p conjugation formed evenwithout incubation, thefilm of non-incubated peptide molecules would be conductive in alocal vicinity. This speculation is consistent with our results fromoptical spectroscopy measurements (Figs. 3 and 4), indicating localmolecular structures of incubated and non-incubated peptides arenot very different in a local vicinity. In addition, the drop-dry filmdoes not guarantee reproducible uniformity over a film withmillimeter scale, although the weight of the molecules incorpo-rated in the film is controlled. Such uncontrolled fluctuation indensity over the film and the existence of local conductivity mayhave resulted in the wide distribution for the conductance of non-incubated peptide monomer films as shown in Fig. 5 (a).

The measured conductance of the peptide-based films in theorder of 10�11 S are higher than those measured for amyloidicpeptide fibers [14,60,61]. They are, of course, extremely poor con-ductors compared to other conductive carbon-basedmaterials suchas carbon nanotubes and graphene [62e64], which have p conju-gated electron systems. Our results indicate that there exists p e pconjugated system contributing the electrical conductance throughthe film of incubated GFPRFAGFP peptide fibers. However, thenumber of fibers directly bridging between two electrodes may notbe large enough because we frequently observed one end of a fiberin frames of AFM images. Also, each fiber has irregular bumps,which could disrupt the conjugated conduction path, as we see inFig. 2 (a). So, the measured conductance would be considerablydecreased from the case of perfectly stacked aromatic residues.

Please cite this article in press as: R.C.G. Creasey, et al., Improved electrnanoscale fibers, Materials Chemistry and Physics (2015), http://dx.doi.o

Other works on conducting nanostructures have utilizedadvanced techniques such as microfabricated electrode arrays [65],STM [66], CP-AFM [13,67] andMP-AFM [68]. Such experimentsmayreveal high-resolution conductivity information of peptide struc-tures. The anisotropic conductivity property of individual fibers isof particular interest. We have attempted preliminary experimentsin measuring the nanoscale conductivity of individual fibers;however, the contact resistance between the metal probe and theorganic molecules was a large issue. The use of a macroscopicelectrode to contact many fibers simultaneously was the moststable measurement condition within the scope of this work. Themacroscopic experiment utilized in this paper is a useful proof-of-concept for checking if an assembly may be conductive beforeembarking on the comprehensive work of testing via the advancedtechniques mentioned above.

4. Conclusions

The novel GFPRFAGFP peptide presented herein is unlike anyknown peptide sequence thus far, despite drawing on inspirationfrommammalian elastic tissues. In simple aqueous conditions, it iscapable of forming fibrous nanostructures. However, the conditionsare not yet optimized, as evidenced by the small changes in spec-troscopic data before and after incubation to induce self-assembly.Electrical measurements of the films of the GFPRFAGFP monomersand fibers, biotin, and insulin show apparent conductance for thepeptide. This conductance is improved significantly after self-assembly; therefore, the extended stacking of aromatic residuesin the peptide sequence may play a vital role in electron transport.

An important aspect of this work is as follows. Although themeasured conductance is still small, we have shown that the DeNovo designed peptides using entirely natural amino acids self-assemble into nanoscale fibers at 37 �C with apparent conduc-tance. This leads to interesting possibilities for self-assembling andself-repairing conductive network structures, which may haveapplication as an intelligent network system.

Acknowledgments

CD spectroscopy measurements were carried out at the Mole-cule & Material Synthesis Platform in “Nanotechnology PlatformProject” operated by the Ministry of Education, Culture, Sports,Science and Technology (MEXT), Japan. The authors are grateful toM. Takeuchi, C. Schultz, and K. Allan for scientific discussions.

Appendix A. Supplementary data

Supplementary data related to this article can be found at http://dx.doi.org/10.1016/j.matchemphys.2015.03.034.

References

[1] A.Z. Stieg, A.V. Avizienis, H.O. Sillin, C. Martin-Olmos, M. Aono, J.K. Gimzewski,Adv. Mater. 24 (2012) 286e293.

[2] R. Rai, M. Tallawi, N. Barbani, C. Frati, D. Madeddu, S. Cavalli, et al., Mater. Sci.Eng. C 33 (2013) 3677e3687.

[3] J.T. Seil, T.J. Webster, Wiley Interdiscip. Rev. Nanomed. Nanobiotechnol. 2(2010) 635e647.

[4] F. Shamsi, H. Coster, K.A. Jolliffe, T. Chilcott, Mater. Chem. Phys. 126 (2011)955e961.

[5] A. Noy, A.B. Artyukhin, N. Misra, Mater. Today 12 (2009) 22e31.[6] Y. Xia, P. Yang, Y. Sun, Y. Wu, B. Mayers, B. Gates, et al., Adv. Mater. 15 (2003)

353e389.[7] J. Zhu, W. Jiang, Mater. Chem. Phys. 101 (2007) 56e62.[8] B. Adhikari, A. Banerjee, J. Indian Inst. Sci. 91 (2011).[9] H.R. Amaral, S. Kogikoski, E.R. Silva, J.A. Souza, W.A. Alves, Mater. Chem. Phys.

137 (2012) 628e636.[10] M. Amit, S. Appel, R. Cohen, G. Cheng, I.W. Hamley, N. Ashkenasy, Adv. Funct.

Mater. 24 (2014) 5873e5880.

ical conductance through self-assembly of bioinspired peptides intorg/10.1016/j.matchemphys.2015.03.034

R.C.G. Creasey et al. / Materials Chemistry and Physics xxx (2015) 1e88

[11] D.E. Friesen, T.J.A. Craddock, A.P. Kalra, J.A. Tuszynski, Biosystems (2014).[12] M. Vargas, N.S. Malvankar, P. Tremblay, C. Leang, J.A. Smith, P. Patel, et al.,

MBio 4 (2013) e00105e13.[13] M.Y. El-Naggar, G. Wanger, K.M. Leung, T.D. Yuzvinsky, G. Southam, J. Yang, et

al., Proc. Natl. Acad. Sci. U. S. A. 107 (2010) 18127e18131.[14] M. Amit, G. Cheng, I.W. Hamley, N. Ashkenasy, Soft Matter 8 (2012)

8690e8696.[15] S. Rauscher, S. Baud, M. Miao, F.W. Keeley, R. Pom�es, Structure 14 (2006)

1667e1676.[16] a V Persikov, J.A. Ramshaw, A. Kirkpatrick, B. Brodsky, Biochemistry 39 (2000)

14960e14967.[17] A. Rath, A.R. Davidson, C.M. Deber, Biopolymers 80 (2005) 179e185.[18] K.E. Marshall, K.L. Morris, D. Charlton, N. O'Reilly, L. Lewis, H. Walden, et al.,

Biochemistry 50 (2011) 2061e2071.[19] C.G. Claessens, J.F. Stoddart, J. Phys. Org. Chem. 10 (1997) 254e272.[20] G.B. McGaughey, J. Biol. Chem. 273 (1998) 15458e15463.[21] J. Zhou, X. Du, Y. Gao, J. Shi, B. Xu, J. Am. Chem. Soc. 136 (2014) 2970e2973.[22] J. Nanda, A. Biswas, A. Banerjee, Soft Matter 9 (2013) 4198.[23] N. Ashkenasy, W.S. Horne, M.R. Ghadiri, Small 2 (2006) 99e102.[24] S. Zhang, Nat. Biotechnol. 21 (2003) 1171e1178.[25] M. Gupta, A. Bagaria, A. Mishra, P. Mathur, A. Basu, S. Ramakumar, et al., Adv.

Mater. 19 (2007) 858e861.[26] C. Val�ery, F. Artzner, M. Paternostre, Soft Matter 7 (2011) 9583e9594.[27] B. Bochicchio, A.M. Tamburro, Chirality 14 (2002) 782e792.[28] A. Pedretti, L. Villa, G. Vistoli, J. Comput. Aided. Mol. Des. 18 (2004) 167e173.[29] N.L. Allinger, J. Am. Chem. Soc. 99 (1977) 8127e8134.[30] A.M. Morris, M.A. Watzky, R.G. Finke, Biochim. Biophys. Acta 1794 (2009)

375e397.[31] A. San Paulo, R. García, Biophys. J. 78 (2000) 1599e1605.[32] A.N. Round, M.J. Miles, Nanotechnology 15 (2004) S176eS183.[33] A.D. Slattery, J.S. Quinton, C.T. Gibson, Nanotechnology 23 (2012) 285704.[34] A.D. Slattery, A.J. Blanch, J.S. Quinton, C.T. Gibson, Ultramicroscopy 131 (2013)

46e55.[35] J.E. Sader, J. Lu, P. Mulvaney, 113702 (2014) 6e11.[36] M.J. Krysmann, V. Castelletto, A. Kelarakis, I.W. Hamley, R.A. Hule, D.J. Pochan,

Biochemistry 47 (2008) 4597e4605.[37] N.J. Greenfield, Nat. Protoc. 1 (2006) 2876e2890.[38] G.D. Fasman, ed., 1 (reprint, Plenum Press, New York, 2010.[39] D.M. Rogers, J.D. Hirst, Biochemistry 43 (2004) 11092e11102.[40] P.L. Atreya, V.S. Ananthanarayanan, J. Biol. Chem. 266 (1991) 2852e2858.

Please cite this article in press as: R.C.G. Creasey, et al., Improved electrnanoscale fibers, Materials Chemistry and Physics (2015), http://dx.doi.o

[41] S. Brahms, J. Brahms, G. Spach, A. Brack, Proc. Natl. Acad. Sci. U. S. A. 74 (1977)3208e3212.

[42] M.A. Amon, M. Ali, V. Bender, K. Hall, M.-I. Aguilar, J. Aldrich-Wright, et al.,J. Pept. Sci. 14 (2008) 714e724.

[43] R.G. Creasey, N.H. Voelcker, C.J. Schultz, Biochim. Biophys. Acta 1824 (2012)711e722.

[44] M.A. Bryan, J.W. Brauner, G. Anderle, C.R. Flach, B. Brodsky, R. Mendelsohn,J. Am, Chem. Soc. 129 (2007) 7877e7884.

[45] J. Kong, S. Yu, Acta Biochim. Biophys. Sin. (Shanghai) 39 (2007) 549e559.[46] M. Jackson, H.H. Mantsch, Crit. Rev. Biochem. Mol. Biol. 30 (1995) 95e120.[47] D.M. Byler, H. Susi, Biopolymers 25 (1986) 469e487.[48] A. Barth, C. Zscherp, Q. Rev, Biophys 35 (2002) 369e430.[49] A. Dong, P. Huang, W.S. Caughey, Biochemistry 29 (1990) 3303e3308.[50] R.R. De Chimie, A. Adochitei, G. Drochioiu, 56 (2011) 783e791.[51] A. Pepe, B. Bochicchio, J. Soft Matter. 2013 (2013) 1e7.[52] C. Chadefaux, A. Le Ho, L. Bellot-gurlet, 2009, 129, 137[53] N.S. Myshakina, S.A. Asher, J. Phys. Chem. B 111 (2007) 4271e4279.[54] A.V. Mikhonin, S.A. Asher, J. Phys. Chem. B 109 (2005) 3047e3052.[55] M.L. Tiffany, S. Krimm, Biopolymers 11 (1972) 2309e2316.[56] G. Fichman, E. Gazit, Acta Biomater. 10 (2014) 1671e1682.[57] E. Gazit, FASEB J. 16 (2002) 77e83.[58] C. Tang, R.V. Ulijn, A. Saiani, Langmuir 27 (2011) 14438e14449.[59] D.N. Woolfson, M.G. Ryadnov, Curr. Opin. Chem. Biol. 10 (2006) 559e567.[60] L.L. Del Mercato, P.P. Pompa, G. Maruccio, A. Della Torre, S. Sabella,

A.M. Tamburro, et al., Proc. Natl. Acad. Sci. U. S. A. 104 (2007) 18019e18024.[61] J. Castillo, S. Tanzi, M. Dimaki, W. Svendsen, Electrophoresis 29 (2008)

5026e5032.[62] Z. Wu, Z. Chen, X. Du, J.M. Logan, J. Sippel, M. Nikolou, et al., Science 305

(2004) 1273e1276.[63] S. Higuchi, O. Kubo, H. Kuramochi, M. Aono, T. Nakayama, Nanotechnology 22

(2011) 285205.[64] X. Li, G. Zhang, X. Bai, X. Sun, X. Wang, E. Wang, et al., Nat. Nanotechnol. 3

(2008) 538e542.[65] A.V. Avizienis, H.O. Sillin, C. Martin-Olmos, H.H. Shieh, M. Aono, A.Z. Stieg, et

al., PLoS One 7 (2012) e42772.[66] J.P. Veazey, G. Reguera, S.H. Tessmer, Phys. Rev. E 84 (2011) 060901.[67] L. Cai, H. Tabata, T. Kawai, Science 80 (2001), 211.[68] T. Nakayama, O. Kubo, Y. Shingaya, S. Higuchi, T. Hasegawa, C.-S. Jiang, et al.,

Adv. Mater. 24 (2012) 1675e1692.

ical conductance through self-assembly of bioinspired peptides intorg/10.1016/j.matchemphys.2015.03.034