Sara C. Wagnera, Meike Roskampb, Helmut Cölfenc, Christoph Böttcherd, Sabine Schlecht*b and Beate Koksch*a
aInstitute of Chemistry and Biochemistry, Freie Universität Berlin, Takustraße 3, 14195 Berlin, Germany. E-mail: koksch@chemie.fu-berlin.de; Fax: +49 30 838 55644; Tel: +49 30 838 55344
bInstitute of Chemistry and Biochemistry, Freie Universität Berlin, Fabeckstraße 34–36, 14195 Berlin, Germany. E-mail: schlecht@chemie.fu-berlin.de; Fax: +49 30 838 53310; Tel: +49 30 838 52423
cDepartment of Colloid Chemistry, Max-Planck-Institute of Colloids and Interfaces, 14424 Potsdam, Germany
dResearch Centre for Electron Microscopy, Freie Universität Berlin, Fabeckstraße 36a, 14195 Berlin, Germany
First published on 17th November 2008
The nanoparticle–peptide interaction described here is based on electrostatic forces and the pH value can act as a trigger to direct the organization of functionalized nanoparticles in a reversible and repeatable manner. The ability of the peptide to interact with the charged gold nanoparticles is directly related to its helical structure and was not found for a random coil peptide with the same net charge. Interestingly, the interaction with nanoparticles seems to induce a fibrillation of the coiled coil peptide.
In previous studies, we reported the de novo design of a coiled coil peptide model system for electrostatic binding to charged membranes.9 Here, we apply the same design principles for the interaction with negatively charged Au nanoparticles containing a shell of mercaptoundecanoic acid (MUA) (Fig. 1a). We chose the coiled coil peptide here referred to as VW05, in which positions a and d are exclusively occupied by leucine and positions e and g contain glutamic acid and lysine, respectively (Fig. 1b). Furthermore, arginine, a highly basic amino acid, is present in position f yielding a solvent-exposed positively charged edge on the helical cylinder. The peptide was N-terminally labelled with anthranilic acid (Abz) to facilitate concentration determination and to enable visualization of the peptide after gel electrophoresis. An attractive electrostatic interaction with Au/MUA nanoparticles is only expected under conditions that provide for a net positive charge of peptide. In contrast, a net negative charge of the peptide should preclude this attractive interaction. Since the coiled coil structure provides a defined multivalent template, the positively charged peptide mediates an assembly of nanoparticles as shown for one dimension in Fig. 1a. Due to the size ratio between peptide helix and nanoparticle, this assembly has more than one dimension. In order to determine the importance of the coiled coil structure as a structural template, we also constructed a peptide VW05-ref that has the same amino acid composition as VW05 but a primary structure that does not follow a heptad repeat (Fig. 1b). In particular, leucine residues were placed such that the formation of a “knobs into holes” packing is not possible and, thus, the thermodynamic driving force for coiled coil formation is reduced. Therefore, this control peptide was designed to exist primarily as a random coil monomer in solution.
![]() | ||
Fig. 1 (a) General scheme of coiled coil peptide-induced assembly of Au/MUA nanoparticles, shown for one dimension. The length of the peptide helix and the size of nanoparticle can be estimated to be 3 nm and 6.2 nm, respectively. (b) Sequence of coiled coil peptide VW05, and non-coiled coil peptide VW05-ref. (c) CD spectra of 100 μM VW05 (pH 9 and 12) and VW05-ref (pH 9). |
Circular dichroism (CD) spectroscopy reveals that VW05 adopts an α-helical coiled coil structure at pH 9.0 and 12.0, as indicated by two minima at 208 and 222 nm (Fig. 1c).10 Between pH 9.0 and 12.0, the peptide's calculated net charge changes from +3.6 to −4.9, respectively (Fig. S2†). The CD spectrum of VW05-ref at pH 9.0 does not reveal a minimum at 200 nm which would be characteristic of a completely unfolded structure, but reveals an α-helical content of 21%, calculated from the molar ellipticity at 222 nm.10,11 The fact that there is a finite α-helical content may be a consequence of glutamic acid's and leucine's propensity for forming α-helices.12 Importantly, the oligomerization of peptide VW05-ref was ruled out by analytical ultracentrifugation measurements, whereas a trimeric complex was observed in the case of VW05 (Fig. S4†). A detailed discussion of ultracentrifugation measurements is given in the supporting information.
UV/Vis absorption spectroscopy was used to monitor the surface plasmon resonance band of Au/MUA nanoparticles. A distinct red-shift and broadening of the surface plasmon resonance maximum (λmax) are associated with a decrease in the interparticle distance for a given dielectric environment.13 In the following results, it turns out that several factors such as pH, peptide concentration (cp) and incubation time affect the value of λmax. At pH 9.0, an absorption maximum of Au/MUA nanoparticles at 522.0 nm is observed in the absence of peptide and for extended incubation times (Fig. 2a). Upon the addition of VW05, the λmax becomes red shifted (Fig. 2a). This spectroscopic change increases with cp and yields an initial λmax of 536.0 nm at 30 μM. With regard to the effect of incubation time, these values vary directly with one another; for example, at a cp of 10 μM, the λmax changes from 530.1 nm to 536.8 nm over a period of six hours. Fig. 2b shows the pH dependence of the absorption maximum at a fixed VW05 concentration of 10 μM and an incubation time of six hours. Interestingly, although the calculated pI value of VW05 is 10.4, a greater value of λmax of 524.7 nm compared to the λmax in the absence of peptide was determined at pH 10.5. This result implies that the net charge of the peptide is not the only factor that determines attraction between peptide and nanoparticle. The presentation of the solvent-exposed arginine surface upon coiled coil formation, considering a pKA value of arginine of 12.514, may still support electrostatic interactions at a pH of 10.5. Nevertheless, the addition of 10 μM VW05 at pH 12.0 results in the same value of λmax observed at pH 9.0 in the absence of peptide.
![]() | ||
Fig. 2 The position of the absorption maxima of 0.05 μM Au/MUA nanoparticles (a) in the presence of different concentrations of VW05 at pH 9.0 after 0 h and 6 h of incubation; at cp >10 μM, strong scattering effects hinder absorption measurements after 6 hours of incubation; (b) at 10 μM VW05 and various pH values; (c) at different concentrations of VW05-ref and pH 9.0; (d) four cycles of pH variation between 9.0 and 12.0 in the presence of 30 μM VW05. The sample was allowed to incubate for 5 min after each pH change. |
In contrast to the coiled coil peptide VW05, incubation of the nanoparticles with the control peptide VW05-ref does not lead to a noticeable change in the absorption maximum at pH 9.0. In this case, the slight increase in λmax may be due to a change in the ionic strength of the solution with increasing cp.
Gel electrophoresis was used to investigate whether an interaction between both peptides and the Au/MUA nanoparticles takes place at pH 9.0. Au/MUA nanoparticles showed a band of relatively high mobility that decreases with increasing cp of VW05 and is dramatically reduced at cp≥ 10 μM (Fig. 3a). To monitor the position of peptide next to the Au/MUA nanoparticles, the agarose gel was also visualized by fluorescence emission following excitation at 314 nm, corresponding to the absorption wavelength of the peptide's UV-label (Fig. 3b). Although, at cp of 50 μM, VW05 clearly shows a well-defined low-mobility band that corresponds to a net positive charge, no peptide can be visualized at the same cp in the presence of Au/MUA nanoparticles. We assume that the peptide is contained in the nanoparticle aggregates and that quenching effects caused by the absorption of nanoparticles interfere with the detection of peptide. An increase in cp up to 500 μM unveils an electrophoretic separation of the excess peptide from the nanoparticle–peptide aggregate. These results confirm that an interaction takes place between VW05 and the nanoparticle, one that cannot be observed under the same conditions for VW05-ref; in this case, electrophoretic separation of colloids and peptides occurs independently of cp (Fig. 3c–d). The greater electrophoretic mobility of VW05-ref compared to VW05 is a consequence of their different oligomerization states, monomeric in the case of VW05-ref and trimeric in the case of VW05.
![]() | ||
Fig. 3 Agarose gels of VW05 (a, b) and VW05-ref (c, d) at pH 9.0, as visualized by (a, c) visible light and (b, d) fluorescence emission following excitation at 314 nm. The numbers refer to the peptide concentration in μM in the presence (black) and in the absence (white) of 0.05 μM Au/MUA nanoparticles. |
The peptide-induced assembly of Au/MUA nanoparticles was examined more precisely by transmission electron microscopy (TEM). The average diameter of the Au core of these nanoparticles is 5.3 ± 0.5 nm. At pH 9.0, mainly isolated Au nanoparticles are detected that are evenly spread if prepared on the surface of hydrophilized carbon covered grids (Fig. S5a†). Infrequently, regions of more densely packed particle monolayers are observed, and we attribute this to the sample blotting procedure and subsequent drying of the grids (see ESI†). At pH 12.0, the spatial distribution of Au/MUA nanoparticles is comparable to that described for pH 9.0. However, in the presence of VW05 at pH 9.0, the morphology is dominated by particle assemblies that vary in diameter from 100 nm to few μm (Fig. S5c–d).
Neither the addition of VW05 at pH 12.0 nor the addition of VW05-ref at pH 9.0 appears to induce this type of nanoparticle organization (Fig. S6 and S7). Here, the observed structures resemble the features of the pure nanoparticles, which is consistent with the results of the UV/Vis and gel electrophoretic studies.
In order to obtain direct information on the assemblies’ native solution state, we additionally employed cryo-TEM measurements. We expected to learn about the structural aspects of the peptide nanoparticle interaction as the peptide could not be visualized in the dried samples. Cryo-TEM measurements of pure Au/MUA at pH 9.0 confirmed the above observation that the particles are isolated in solution in the absence of the peptide. No objects other than pure gold nanoparticles were detectable. The stabilizing layer of MUA is not visible in the images. The number of nanoparticles per area (per volume actually) was significantly lower in the cryo-specimen (see overview in Fig. S5†) if compared to the dried samples. This can be understood if one keeps in mind that freezing aliquots of the sample for a direct microscopical observation retains the particle concentration in solution which otherwise is dramatically increased if the sample is dried on a surface leading to a more dense arrangement of particles. Therefore the aggregation effect of nanoparticles in the presence of VW05 becomes even more impressive. The overview in Fig. 4c shows a large assembly of particles spanning the complete micrometre hole of the perforated carbon film. At higher magnification the participation of the peptide is clearly visible (Fig. 4d). Ultrathin fibers connect the nanoparticles in a network-like manner and multiple linear connections between the particles are visible. The diameter of the fibers ranges from ultrathin single helices of about 1 nm diameter up to slightly larger bundles of 3 nm diameter, which corresponds to a monomeric and trimeric helical organization, respectively. The fiber length in the assembly network is difficult to determine, however, peptide fibers not involved in the assemblies show dimensions in the micrometre range. In contrast, such fibers cannot be observed in the absence of nanoparticles. Due to the low concentration of the peptide the unambiguous identification of its prefiber organization in cryo-TEM images is nearly impossible. However, fibril formation of 10 μM VW05, in the absence of nanoparticles, can be ruled out by ultracentrifugation measurements. We therefore assume that the high local peptide concentration on the surface area of the nanoparticles decreases the lag time for nucleation and assists the α-helical fibrillation of peptide. Such a potential of nanoparticles to induce protein fibrillation has recently been reported for amyloid forming peptides.15
![]() | ||
Fig. 4 (a and b) Cryo-TEM micrographs of 0.05 μM Au/MUA nanoparticles at pH 9.0 and at different magnifications. The particles are mostly isolated and, as is obvious from (b), absent from any further visibly adherent material. Scale bars are 500 nm (a) and 50 nm (b), respectively. (c) Cryo-TEM image showing a large assembly (overview) of 0.05 μM Au/MUA nanoparticles in the presence of 10 μM of the peptide VW05 (scale bar is 200 nm). (d) The magnified detail from (c) indicates the multilayered organization of the VW05 mediated assembly as well as multiple fiber-like features (thought to correspond to α-helices) interconnecting the gold nanoparticles in a network-like manner (arrows). Scale bar is 25 nm. |
We were able to demonstrate the superiority of the α-helical coiled coil folding motif not only as a defined structural template but also as a self-organizing system for the assembly of Au/MUA nanoparticles. Since the nanoparticle–peptide interaction is based on electrostatic forces, we further demonstrated that pH control can be applied as a trigger to direct the organization of functionalized nanoparticles. Remarkably, a cyclic change in the pH of solution can be used to reversibly direct the assembly of colloids and peptides, as shown by UV/Vis spectroscopy and TEM (Fig. 2d and Fig. S8). A key feature of the system described here is that organization of the nanoparticles into defined networks can be directed by the external factor of pH, and that the organization is reversible and repeatable.
We believe that this study contributes valuable insight into the de novo design of small peptides for the development of bionanostructures. Our results offer a novel route to nanoparticle–peptide hybrid systems whose assembly can be controlled in a switchable manner by pH conditions. The system described here is currently being further developed to yield nanoparticle networks that have defined geometries and can be applied to the production of useful hybrid materials with tunable optical properties.
The vitrified samples were subsequently transferred under liquid nitrogen into a Tecnai F20 FEG transmission electron microscope (FEI Company, Oregon, USA) using the Gatan (Gatan Inc., California, USA) cryoholder and -stage (Model 626). Microscopy was carried out at 94 K sample temperature using the microscopes low dose protocol at a calibrated primary magnification of 62 k× and an accelerating voltage of 160 kV (FEG - illumination). Images were recorded using an EAGLE 2k-CCD device (FEI Company, Oregon, USA) at full 2048 by 2048 pixel size. The defocus was chosen in all cases to be 980 nm.
The equilibrium experiments have been evaluated with the model independent M* Program17 (Kristian Schilling, Nanolytics 2007) to determine the weight average molar mass for the different loading concentrations.
In addition, the sedimentation equilibrium concentration profiles have been fitted globally to a non-interacting ideal single species model over the entire investigated concentration range and three speeds (WINNONLIN v. 1.06.).18
Additionally the molar mass of the peptides was independently determined from their molar mass distribution, which was calculated from sedimentation velocity experiments using the program SEDFIT v. 10.09 beta.19 The molar mass average was calculated by integration of the distribution. The sedimentation velocity experiments were performed at 60,000 rpm and 25 °C using absorption optics at 320 nm and three different concentrations (300, 100 and 50 μM for VW05 and 400, 200 and 50 μM for VW05-ref).
Footnote |
† Electronic supplementary information (ESI) available: Experimental details and additional figures. See DOI: 10.1039/b813429d |
This journal is © The Royal Society of Chemistry 2009 |