Structure and Mechanical Properties of Spider Silk Films at the Air

Article pubs.acs.org/Langmuir

Structure and Mechanical Properties of Spider Silk Films at the Air− Water Interface Anne Renault,† Jean-François Rioux-Dubé,‡ Thierry Lefèvre,‡ Sylvie Beaufils,† Véronique Vié,† François Paquet-Mercier,‡ and Michel Pézolet*,‡ †

Institut de Physique de Rennes, UMR CNRS 6251, Université de Rennes, Bat 11A Campus de Beaulieu, 35042 Rennes Cedex, France ‡ Centre de Recherche sur les Matériaux Avancés, Département de Chimie, Université Laval, Québec, Canada G1V 0A6 ABSTRACT: The kinetics of adsorption of solubilized spider major ampullate (MA) silk fibers at the air−water interface and the molecular structure and mechanical properties of the interfacial films formed have been studied using various physical techniques. The data show that Nephila clavipes MA proteins progressively adsorb at the interface and ultimately form a highly cohesive thin film. In situ infrared spectroscopy shows that as soon as they reach the interface the proteins predominantly form β sheets. The protein secondary structure does not change significantly as the film grows, and the amount of β sheet is the same as that of the natural fiber. This suggests that the final β-sheet content is mainly dictated by the primary structure and not by the underlying formation process. The measure of the shear elastic constant at low strain reveals a very strong, viscous, cohesive assembly. The β sheets seem to form cross-links dispersed within an intermolecular network, thus probably playing a major role in the film strength. More importantly, the molecular weight seems to be a crucial factor because interfacial films made from the natural proteins are ∼7 times stronger and ∼3 times more viscous than those obtained previously with shorter recombinant proteins. Brewster angle microscopy at the air−water interface and transmission electron microscopy of transferred films have revealed a homogeneous organization on the micrometer scale. The images suggest that the structural assembly at the air−water interface leads to the formation of macroscopically solid and highly cohesive networks. Overall, the results suggest that natural spider silk proteins, although sharing similarities with recombinant proteins, have the particular ability to self-assemble into ordered materials with exceptional mechanical properties.

■

INTRODUCTION

In the case of spider silk, recombinant major ampullate (MA) spidroins of Nephila clavipes (rMaSpI and rMaSpII) having a molecular weight of approximately 65 kDa form elastic films at the air−water interface as a result of a network of aggregated proteins characterized by intermolecular β sheets.7 rMaSpII is found to have a higher affinity for the interface than rMaSpI.7 The recombinant MA spidroin of Araneus diadematus (ADF-4) adsorbed at the oil−water interface also form mechanically stable elastic films.8 Lower ADF-4 concentrations than for globular protein β-lactoglobulin are required to obtain highshear-modulus films.8 To our knowledge, natural MA silk proteins have not yet been studied at the air−water interface. The MA silk is commonly recognized as a hallmark of tough natural fiber that exhibits high strength and extensibility.9−12 It is used by orb-weaving spiders as lifeline, radii, frame, and mooring thread for the web. It is composed of two highmolecular-weight (260−350 kDa) spidroins,13 MaSpI and MaSpII, in a 80:20 ratio.14 The primary structures of these two proteins are very similar, being mainly composed of

The study of the behavior of proteins at interfaces, especially at the air−water interface, is an insightful strategy for obtaining information about protein amphipatic and self-assembly properties. For silk proteins, this avenue may help to understand better how this high-molecular-weight biopolymer self-organizes in the natural fiber and may help in developing innovative and useful environmentally friendly and biocompatible materials. Bombyx mori (B. mori) fibroins form stable films1,2 with a gellike3 or crystalline4,5 structure when adsorbed at the air−water interface. It has been suggested that protein adsorption is controlled by a mixed diffusion-kinetic model at low concentration (≤1 mg/mL) whereas it is limited by a diffusion-controlled model at high concentration (≥10 mg/ mL). Films transferred by Langmuir−Blodgett techniques are characterized by intermolecular β sheets.1 B. mori fibroins can also form stable films at the oil−water interface.2,6 The surface elastic modulus of interfacial films generally appears to be very large with respect to those formed by proteins such as lysozyme, insulin, and β-casein.2,3 © XXXX American Chemical Society

Received: December 19, 2012

A

dx.doi.org/10.1021/la401104m | Langmuir XXXX, XXX, XXX−XXX

Langmuir

Article

hundreds of repeat units flanked by N- and C-terminal nonrepetitive parts.15,16 The repetitive unit is made of polyalanine blocks of 4−10 amino acids with GGA tripeptides and a 20−30 amino acid-long Gly-rich sequence. As opposed to MaSpI, MaSpII contains proline residues. Polyalanine blocks adopt a β-sheet conformation in the fiber, whereas the Gly-rich segments adopt other, generally disordered, conformational structures.17−19 In this work, MA silk fibers of the spider Nephila clavipes have been dissolved in water, and the kinetics of adsorption of the natural proteins at the air−water interface has been characterized using different physical techniques, including surface pressure, ellipsometric angle, Brewster angle microscopy imaging, shear elastic constant, and infrared spectroscopy. The microstructure of the films formed at the interface has been characterized by transmission electron microscopy (TEM). The results obtained help us to understand the mechanism and mechanical properties of silk protein assembly better and may be useful in producing silk-inspired materials.

■

molecules. The gray level is correlated to the amount of matter. Our device does not allow the exact quantification of the number of molecules present at the interface, but the comparison between gray levels gives an indication of the relative number of molecules. The experiment was performed while simultaneously measuring the surface pressure. The lateral resolution was 3 μm. Shear Elastic Constant Measurements. The rheometer used21 is based on the action of a very light float applying a rotational strain to the surface through a magnetic couple generated by a pair of Helmholtz coils and a small magnetic pin deposited on the float. The float is a 10-mm-diameter paraffin-coated aluminum disc located at the air−water interface in the center of a 48-mm-diameter Teflon trough surrounded by a surface whose rigidity is measured. The subphase was 5 mm deep. Sensitive angular detection of the float rotation was achieved using a mirror fixed on the magnet that reflects a laser beam onto a differential photodiode. A small sinusoidal torque excitation (in the 0.01−100 Hz frequency range) was applied to the float by an oscillating field perpendicular to that of the solenoid. This field acts as a restoring torque equivalent to a surface having a rigidity of 0.16 mN/ m. The amplitude and phase of the float angular motion were then recorded. The measured amplitude of rotation is dependent on the restoring forces applied by the monolayer under shear deformation, whereas the phase provides information relative to the elasticity/ viscosity of the monolayer. The elastic constant, μ (or G′), and the surface viscosity, ηs, were calculated using the viscoelastic model.22 The amplitude and phase of the mechanical response of the pure subphase were first analyzed in the 0.01−100 Hz frequency range to verify the absence of any rigidity. This measurement takes approximately 1 h. The protein solution was then poured into the trough, and the mechanical response of the layer formed at the interface was recorded at a fixed frequency of 5 Hz. At the end of the kinetics measurements, when the shear elastic constant, μ (expressed in mN/m), reached a nearly constant value, a new measurement between 0.01 and 100 Hz was recorded. Rigidity measurements were carried out at 18 °C. Values of the elastic constant and viscosity are the means of two measurements each. In Situ Infrared Spectroscopy. Spectra were recorded using a Nicolet Magna 850 Fourier transform infrared spectrometer (Thermo Scientific, Madison, WI) equipped with a photovoltaic MCT detector (Kolmar Technologies, Newburyport, MA, USA). Polarization Modulation Infrared Reflection Adsorption Spectroscopy (PM-IRRAS). The PM-IRRAS method at the air− water interface is described in detail elsewhere.23 The solution was deposited in a 5-mm-deep rectangular Teflon trough (total volume of 10 mL). The polarization of the infrared beam was modulated with a PEM-90 photoelastic modulator (Hinds Instruments, Hillsboro, OR, USA) that was set for optimum efficiency at 1400 cm−1. The spectra were recorded at 8 cm−1 resolution with a scanning mirror velocity of 0.47 cm/s by coadding 1024 interferograms. The total acquisition time for each spectrum was about 16 min. Normalized PM-IRRAS spectra were obtained using eq 1

EXPERIMENTAL SECTION

MA Protein Solutions. N. clavipes spiders were collected in Florida (USA). Spiders were raised in the laboratory in 20 × 50 × 60 cm3 cages at 58 ± 5% relative humidity and 24 ± 2 °C. They were fed four times a week with small crickets and with 3 drops of 10% w/v glucose solution per week. The MA proteins were obtained by forced reeling the spiders at a speed of 1 cm/s. The silk fibers were dissolved in 6 M guanidine thiocyanate (GdnHSCN) at a concentration of approximately 1−3 mg/mL and held at 40 °C for 30 min. To remove GdnHSCN, the proteins were eluted through PD-10 Sephadex G-25 medium desalting columns (GE Healthcare Bio-Sciences AB, Uppsala, Sweden) equilibrated with 50 mL of ultrapure water (resistivity of 18.2 MΩ·cm and surface tension in the range of 72.8−73.0 mN/m). The final protein concentration was verified by UV absorption at 280 nm with a Cary 500 (Varian, Palo Alto, CA, USA) UV−vis−NIR spectrophotometer using an absorption coefficient of 0.594 mL·mg−1·cm−1. The solution was then diluted with ultrapure water to the desired bulk protein concentration, hereafter called Cb. Protein solutions were poured directly into the trough immediately after preparation (maximum elapsed time of 10 min), and the kinetics of protein adsorption was measured. The same protein solutions were used for PM-IRRAS and for ellipsometric/surface tension experiments, with the measurements being recorded simultaneously. Ellipsometry and Surface Tension Measurements. The ellipsometric angles (Δ and ψ) and surface pressure (π) were recorded simultaneously in the same trough. The ellipsometric measurements were carried out with a conventional null ellipsometer using a He−Ne laser operating at 632.8 nm.20 The surface pressure was measured with a Nima PS4 film balance (Coventry, England) using filter paper (0.5 × 1.0 cm2 Whatman paper, no. 1) as a Wilhelmy plate. The volume of the Teflon sample trough was 10 mL. All experiments were performed at room temperature between 19 and 21 °C. Protein concentrations in the subphase, Cb, of between 1 and 10 μg·mL−1 were investigated. Imaging Ellipsometer. For microscopic observation of the protein interfacial layer, a commercial null imaging ellipsometer (Accurion, nanofilm-EP3, NTF, Göttingen, Germany) was used. The ellipsometer is equipped with a motorized goniometer for an accurate selection of the angle of incidence, and it was used in a Brewster angle microscopy (BAM) configuration. A laser beam (532 nm) was reflected by the interface with an incident angle close to that of the Brewster angle (approximately 53°). The light passed thought a linear polarizer. Upon reflection from the interface, the beam was imaged onto a CCD camera through a long-working-distance 10× objective. The contrast on the images is due to a change in the reflection coefficient of the polarized incident light on the surface as a result of a change in the refractive index of the interface upon adsorption of

S(d) − S(0) ΔS = S S(0)

(1)

where S(d) and S(0) are the PM-IRRAS signals of the covered and uncovered subphases, respectively. The spectra were recorded as a function of time for at least 4 h after the beginning of the adsorption. They were corrected for baseline deviations using a cubic polynomial function and corrected for the water vapor contributions by subtracting a reference water vapor spectrum weighted such that the curvimetric length of the corrected spectra is minimized (Goormaghtigh, E., personal communication). Attenuated Total Reflection (ATR) Infrared Spectroscopy. For ATR experiments, the films formed from a 5 μg/mL solution after 5 h of adsorption (π ≈ 10−12 mN/m) were transferred onto one side of a multireflection germanium crystal (50 × 20 × 2 mm3) using the Langmuir−Schaeffer technique. Polarized spectra were recorded in order to investigate the molecular orientation at the interface. Each B


Langmuir

Article

spectrum is the result of the average of 1024 scans at 4 cm−1 resolution apodized with a Happ-Genzel function. To determine the secondary structure content, the isotropic spectrum was calculated from the polarized ones using the following equation A iso =

2A s + A p 3

where Ap and As are the absorbance polarized parallel and perpendicular to the plane of incidence, respectively. Then, the amide I region was decomposed using a method previously described24 in order to make an accurate estimation of the level of orientation of the β sheets. The minimum number of components (three) was used for the curve-fitting: two components at 1627 and 1697 cm−1 are due to β sheets, and a component at 1655 cm−1 is assigned to disordered structures. Other components at 1720 and 1593 cm−1 originate from carbonyl stretching and carboxyl stretching vibrations, respectively. Constraints on the position (±2 cm−1) and bandwidth (±4 cm−1) were applied during the curve-fitting calculation to avoid physically nonrelevant results. The content of secondary structures corresponds to the area of the amide I band component divided by the total area of the amide I band. The molecular orientation of the β sheets was then determined using the method developed by Marsh for the nonaxial symmetry of βsheet structures25 and Harrick’s thin film approximation.26 The calculation is based on the measurement of the dichroic ratios of the amide I band at 1626 cm−1 and amide II band at 1520 cm−1 as deduced from the peak heights of the polarized spectra. The dichroic ratios allow the determination of two order parameters ⟨P2(cosθ)⟩ and ⟨P2(cosβ)⟩, where θ represents the angle between the β-sheet plane and the normal to the interface and β is the angle of the β-strand axis within the plane of the β sheet.25 All spectral manipulations were performed using GRAMS/AI 8.0 (Thermo-Galactic, Salem, NH). Transmission Electron Microscopy (TEM). The films formed at the interface for a bulk concentration of 10 μg/mL after ∼5 h (π ≈ 10−12 mN/m) were transferred by gentle deposition of the TEM grid (coated with a Formvar film) with tweezers onto the interfacial films. The film was then pushed down into the solution to ensure the adhesion of the film on the grid. Then, the film−grid assembly was carefully pulled out of the solution. Finally, the samples were allowed to dry on a filter paper. The instrument used was a JEOL JEM-1230 electron microscope operating at 80 kV (JEOL, Tokyo, JPN) equipped with a Gatan DualVision (Gatan, Pleasantown, CA, USA) CCD camera as a detector. Analyses were carried out using bright-field mode and negative staining. Ammonium molybdate solution (2 wt %/v) was added to the grid as a contrast agent for a few seconds and then absorbed. Micrographs were visualized using Gatan Digital Micrograph software, version 3.4 (Gatan, Pleasantown, CA).

Figure 1. Surface pressure as a function time for a solution of MA proteins at different concentrations. The uncertainty in the surface pressure is ±0.2 mN/m.

natural MA proteins such as natural B. mori and recombinant MA silk proteins2,3,7,8 are able to form stable interfacial films. The ellipsometric angle (Figure 2), which is principally affected by the interfacial concentration, also increases as both

■

Figure 2. Ellipsometric angle as a function time for a solution of MA proteins at different concentrations. The uncertainty in the ellipsometric angle is ±0.5°.

RESULTS AND DISCUSSION Kinetics of Adsorption (Surface Pressure and Ellipsometric Angle). The kinetics of adsorption of silk proteins was first investigated with two complementary variables (i.e., the surface pressure and ellipsometric angle). The former technique is related to the surface activity of the molecules whereas the latter one depends on the protein density at the interface. The changes in the surface pressure (π) as a function of time are shown in Figure 1. A progressive adsorption of the spidroins at the interface is observed as monitored by the increase in π with time. As expected, the kinetics of adsorption during the first 2 to 3 h is faster as the concentration increases. (The decrease in π occurring at higher concentrations after 3 to 4 h of aging is addressed below.) This behavior is similar to that obtained previously with other silk proteins.2,3,7,8 The present data overall show that, because of their amphipathic properties,

the time and bulk concentration increase. These results suggest that spidroins continuously accumulate at the interface and that the film thickness increases with protein concentration. This observation is fully consistent with a progressive increase in the intensity of the infrared amide I band with time and concentration (see below). Therefore, no saturation of the interface was observed in the concentration range studied, and the film is able to accommodate all molecules that reach the interface. Using the relation proposed by Feijter,27 we can roughly estimate the interfacial concentration from the ellipsometric angle to be 2 to 3 mg/m2 after ∼4 h (at Cb ≥ 2 mg/mL). This value and the general pattern of the kinetics of adsorption are C


Langmuir

Article

After 15 h of adsorption, the elastic constant reached values of about 50−60, 90−100, and 150−170 mN/m at 3, 5, and 10 mg/mL, respectively. These values are much higher (up to 1 order of magnitude) than values obtained for rMaSpI and rMaSpII and other proteins such as ovalbumin. Indeed, the elastic constants measured under similar conditions are in the 10−20 mN/m range for rMaSpI and rMaSpII (at 50 μg/mL)7 and the 5−15 mN/m range for ovalbumin (at 1−10 μg/mL).28 Thus, natural MA protein seems to form very strong films at the air−water interface. It is noteworthy that although these values are comparable for these various proteins caution has to be taken in the interpretation of the absolute values that are recorded at a fixed frequency because this calculated value of μ is based on the approximation that the material is purely elastic.29 As described in the Experimental Section, measurements of the elastic constant μ as a function of the strain frequency and concentration were also recorded at the end of the kinetics measurements to investigate the linear viscoelastic behavior of the systems (Figure 4). The real and imaginary parts of μ were

very close to those obtained previously for recombinant protein rMaSpI,7 suggesting equivalent amphiphilic properties and affinity for the interface. This observation is consistent with the fact that the fiber is made from 80% MaSpI. Because the primary structure of natural proteins is very similar to that of rMaSpI, except for the molecular weight difference (∼65 kDa for rMaSpI and 260−350 kDa for the natural protein) and the absence of the C-terminal region for rMaSpI, these structural differences do not seem to impact the kinetics of adsorption and the propensity to form films. It is noteworthy that a decrease in the surface pressure for long adsorption times (more than 3 to 4 h) is observed at 5 and 10 μg/mL. This effect, which is more or less reproducible, has often been observed in the course of the experiments. A decrease in the number of protein molecules from the interphase to the subphase is unlikely because the ellipsometric angle increases for the same period of time, which indicates that the spidroins still accumulate at the interface. It rather suggests that the film formed at the interface is different from a typical model of isolated molecules or a fluid system such as liquid expanded phases. It is more consistent with the formation of a cohesive and solid film so that, instead of measuring only the surface tension at the interface, the Wilhemy plate is mainly constrained by mechanical forces imposed by a tridimensional network. Because the surface pressure is defined as the change in the water surface tension by (interacting) adsorbed molecules, it does not seem to be adapted to such rigid and cohesive films. Mechanical Interfacial Properties (Shear Elastic Constant). Figure 3 shows the shear elastic constant as a function

Figure 4. Mechanical response of the interfacial film formed by a solution of MA protein at 3, 5, and 10 μg/mL, as measured by the real and imaginary parts of the rotational amplitude of the float.

fitted with a viscoelastic model that allows a quantitative and accurate determination of the surface elastic constant μ and viscosity ηs of the films.22 The quality of the curve-fitting calculations of Figure 4 shows that the interfacial films are very elastic at low strain. The values of the elastic constant and viscosity are given in Table 1 for different protein concentrations. As can be seen, the elasticity and viscosity increase with protein concentration in

Figure 3. Shear elastic constant of the interfacial film as a function of time for a solution of MA protein at 3, 5, and 10 μg/mL, obtained at a rotational pulsation of the float of 5 Hz.

time for a solution of MA proteins at 3, 5, and 10 μg/mL. As can be seen, the shear elastic constant progressively increases with time and with concentration. These curves reflect the kinetics of adsorption, the progressive organization of the proteins at the interface, and the strengthening of the films as more proteins are involved. Similarly to ellipsometric data, the elastic constant progressively increases with time without a plateau, suggesting that the films can continuously incorporate proteins. These changes are concomitant with the increases in the surface pressure and ellipsometric angle, suggesting that intermolecular forces progressively develop within the interfacial film.

Table 1. Values of the Surface Viscosity (ηs) and Elastic Constant (μ as a Function of the Protein Concentration

D

bulk concentration Cb (μg/mL)

ηs (mN s/m)

μ (mN/m)

3 5 10

0.13 ± 0.02 0.17 ± 0.01 0.21 ± 0.05

60 ± 12 105 ± 10 131 ± 17


Langmuir

Article

Molecular Structure at the Interface (Infrared Spectroscopy). The PM-IRRAS spectra recorded in situ at the air− water interface are shown in Figure 5A as a function of time.

the bulk. Because the above results have shown that more proteins adsorb at the interface as Cb increases, it appears that the films are stronger and more efficiently dissipate energy when more molecules are present at the interface. The films are overall very resistant to mechanical constraints and concomitantly exhibit viscous properties. The comparison between the present data and those obtained previously for recombinant silk proteins rMaSpI and rMaSpII is very informative. For example, the elastic constant μ calculated using the same methodology from the results obtained previously for rMaSpI and rMaSpII at a bulk concentration of 5 μg/mL is 18 and 14 mN/m, respectively,7 comparable to the 10−12 mN/m values obtained at 4−60 μg/ mL for 48 kDa recombinant proteins of Araneus diadematus at the oil−water interface.8 These values are, however, much smaller than the 105 mN/m value obtained here for the natural MA proteins. Similarly, the calculated surface viscosity of the natural proteins at 5 μg/mL is 0.16 mN·s/m compared to 0.05 and 0.06 mN·s/m for rMaSpI and rMaSpII, respectively. Thus, although the behavior at the molecular level of the native proteins is similar to that of rMaSpI, the elastic modulus and the viscosity of the films formed by the native proteins are much higher than for rMaSpI, which clearly emphasizes the effect of molecular weight on the mechanical properties of the proteins films. The results obtained with the native proteins are characteristic of a strong, highly cohesive, dissipative system. The high elastic constant formed by interfacial silk protein films has already been mentioned even at low concentrations, in particular, with respect to globular or unordered proteins such as β-casein.2,3,8 This is also true for the natural MA proteins investigated in this study. The elastic modulus is indeed much higher than that of the 2D multilayer gel network formed by ovalbumin (∼ 3−30 mN/m at 1000 μg/mL depending on pH), a globular protein of egg white that also adopts a β-sheet conformation at the air−water interface.20,28 It is even an order of magnitude higher than that of the 2D network of polymerized filaments formed by G-actin (∼20 mN/m at 25 μg/mL).30 It is, however, much lower than the 450 mN/m (G′) value obtained for regenerated solutions of Bombyx mori silk at 10 μg/mL.3 Therefore, rheological measurements at the air− water interface generally show that proteins of small molecular weight compared to natural proteins form weaker films and that the native molecular weight of the silk proteins is a determinant in reaching strong mechanical resistance. The same conclusion obviously arises for the viscosity. This finding is consistent with a similar general property of viscoelastic polymers rationalized by the Rouse model that predicts an increase in the polymer melt viscosity as a function of molecular weight due to entanglement interactions.31−33 Interestingly, it has been shown that the tensile properties of a synthetic fiber produced with a full length recombinant MA protein of Nephila clavipes are much better than those obtained with truncated proteins.34 These results emphasize the benefits provided by a high-molecular-weight protein on the mechanical properties of silk materials, although the natural proteins are polydisperse.13 In addition, the C- and N-terminal regions were not included in the sequence of the full-length recombinant proteins so that their effect was unknown, but they did not seem to be determinant.34 This conclusion supports the hypothesis that the C-terminal region does not influence the mechanical properties of the films although it could affect the kinetics of adsorption and assembly.

Figure 5. PM-IRRAS spectra at the air−water interface recorded as a function of time for a 5 μg/mL solution. (A) Air−H2O interface. (B) Air−D2O interface.

The intensity increases with time showing that the proteins progressively adsorb at the interface in agreement with surface pressure and ellipsometry measurements. The first PM-IRRAS spectra recorded are dominated in the amide I region by bands at 1626 and 1697 cm−1 that are well known to be associated with two vibrational modes of the intermolecular β sheets resulting from protein aggregation.35−38 A third band near 1655 cm−1 is assigned to residual disordered secondary structures.39 The band at 1516−1537 cm−1 is due to the amide II vibration and is consistent with the β-sheet structure. These spectra are very close to those recorded for rMaSpI,7 which suggests that the differences in the mechanical properties of natural and recombinant protein films are not related to the secondary structure. The amide I band shows that the first molecules that adsorb at the interface are predominantly in the β-sheet conformation. However, circular dichroism (data not shown) has revealed that the secondary structure of MA proteins in the bulk solution is totally devoid of the β-sheet conformation. The absence of β sheets for solubilized MA fibers during at least 6 h has further been confirmed from more recent experiments using Raman optical activity and VCD at a protein concentration of 5 mg/ mL (data not shown). Thus, it seems that a rapid adsorptioninduced conformational change occurs when the silk proteins E


Langmuir

Article

reach the air−water interface. The same behavior has been observed for rMaSpI whereas rMaSpII exhibited upon adsorption a conformational rearrangement from a structure dominated by the α helix to a structure characterized by β sheets. Thus, the behavior of natural MA proteins reflects the predominance of MaSpI−MaSpI interactions. MaSpI−MaSpII interactions may be unobservable, but it is also possible that the rapid formation of β sheets by MaSpI accelerates the formation of β sheets by MaSpII. No further significant change of the shape of the amide I band can be noted with time, which suggests that spidroins gather at the interface without a further conformational change. A similar observation was made for rMaSpI,7 which seems consistent with the fact that the MA silk is made of 80% MaSpI. One can, however, notice an increase in the amide II/amide I intensity ratio upon adsorption, suggesting a reorientation of the proteins at the interface. To confirm these results, PM-IRRAS spectra were also recorded at the air−D2O interface (Figure 5B). In this case, the spectra are free from water vapor absorption, allowing a better investigation of the beginning of the adsorption kinetics. As can be seen, the first spectrum is characterized by two components located at 1618 and 1690 cm−1, which is again typical of intermolecular β sheets resulting from protein aggregation, which supports the above conclusion. Bands are shifted only with respect to the air−H2O spectra as a result of the deuteration of the amide groups. Again, apart from a shift of the high-frequency component (from 1618 to 1623 cm−1) as a function of time, no significant spectral modifications are observed, suggesting no further conformational change once the proteins have reached the interface. The intensity of the amide I band, however, increases with time, showing that more proteins adsorb at the interface in very good agreement with the ellipsometry results. The films formed at the interface have also been transferred onto germanium ATR crystals by the Langmuir−Schaeffer technique. Polarized spectra have been recorded to assess whether the proteins exhibit a preferential orientation within the film. The s- and p-polarized spectra are shown in Figure 6 as well as the calculated isotropic spectrum and its spectral decomposition. The amide I region is basically similar to that recorded in situ at the air−water interface with the two components at 1627 and 1697 cm−1 due to intermolecular β sheets and a broad component near 1655 cm−1 representative of disordered (or amorphous) structures. The amide II band arises at 1520 cm−1. The molecular orientation of the β sheets calculated from the polarized spectra of the transferred film shows that, taking into account the experimental uncertainty, the β sheets have essentially no preferential orientation with respect to the plane of the interface (⟨P2(cos θ)⟩ = 0.14 ± 0.09). Moreover, the value of ⟨P2(cos β)⟩ is 0.19 ± 0.16, suggesting a very small orientation of the strands, if any, in the plane of the β sheets. Thus, polarized IR data do not show any significant orientation of the proteins in the interfacial films. The spectral decomposition of the amide I band of the isotropic spectrum has also been used to estimate the β-sheet content. The results obtained show that the overall secondary structure is composed of β sheets (35 ± 3%) and other structural elements (65 ± 3%). Interestingly, this quantity of β sheets closely corresponds to that found in the natural MA fiber of Nephila clavipes.40,41 This suggests that the proportion of β sheets formed at the interface is optimal and that it represents

Figure 6. (A) Polarized ATR and (B) isotropic spectra of the interfacial film obtained from a 5 μg/mL solution after 5 h of adsorption (π ≈ 10−12 mN/m) before being transferred onto a germanium crystal.

the maximal number of β sheets that the protein can form. For the natural MA silk fiber, it has been shown that the β-sheet content corresponds to the percentage of amino acids involved in polyalanine blocks and those located next to them.40,41 The present data suggest that the same amino acid residues adopt a β-sheet conformation at the air−water interface, which further indicates that the number of β sheets formed in MA silk protein systems can be principally driven by the primary structure and only slightly by the underlying formation process. Finally, it is likely that, as in the natural fiber, the intermolecular β sheets are at the origin of the interactions that stabilize these interfacial films and give them their mechanical properties. In particular, aggregates formed at the interface through β sheets are likely to play the role of cross-links connecting amorphous regions, as in the natural fiber. Such junctions are likely important for the elastic response of the film network. Microstructural Characterization of the Interfacial Films. The morphology of the film resulting from 2 h of adsorption of a solution at a bulk concentration of 10 μg/mL (π ≈ 9 to 10 mN/m) has been studied in situ at the air−water interface by BAM imaging (Figure 7). The film appears to be very homogeneous on the micrometer scale. To highlight the contrast, the film has been cut with tweezers. Interestingly, the film (white area in Figure 7) exhibits a clear, rather sharp edge, suggesting that it is cohesive and behaves like a viscoelastic solid on large scales. This observation is consistent with rheological measurements obtained at low strain. Protein aggregates located beneath the film can be observed in the black area of the image. The shape of these structures suggests a F


Langmuir

Article

leading to a secondary structure characterized by the presence of β sheets. As proteins accumulate at the interface, adsorbed proteins adopt the same conformation dominated by β sheets during the whole adsorption process. The β-sheet content of the film is very close to that of the natural fiber, suggesting that the number of β sheets formed in MA silk protein systems is principally driven by the primary structure. It also appears that β sheets play a similar role as in the natural fiber by forming cross-links in an interfacial network, thus assuring the cohesion and the resistance of the film. BAM and TEM techniques suggest that the resulting films are rather homogeneous. Rheological experiments show that the films exhibit viscoelastic properties. They are viscous while being very elastic and cohesive at low strain compared to other proteins under similar conditions. The high strength and cohesion of the films also seem to be true at large scales as deduced from BAM and TEM images. It is probable that these remarkable mechanical properties mainly originate from the high molecular weight of natural proteins, providing, as in the natural fiber, a highly elastic modulus with an efficient capacity to dissipate energy with strain. This study thus underlines the crucial role that molecular weight could play in future synthetic systems made of silk proteins. It is anticipated that native-sized silk proteins would form in general more resistant materials than shorter or incomplete recombinant ones that lead to weaker systems. From a biomimetism prospective, although the mechanism of molecular assembly cannot be neglected, being able to obtain native-size proteins appears to be essential to producing materials with optimal mechanical properties. This study exemplifies the usefulness of investigations at the air−water interface to understand the behavior and self-assembly properties of silk protein systems better.

Figure 7. BAM images of the interfacial film obtained after 130 min (10 μg/mL) at the air−water interface after being cut with tweezers. The white area corresponds to the film, and the black area, to water. The image size is 392 μm × 502 μm, and the scale bar is 100 μm.

fractal microstructural organization, at least in the first step of the assembly process. The films obtained for a bulk concentration of 10 μg/mL after ∼5 h have also been collected directly at the air−water interface and transferred to TEM grids (Figure 8). The fact that the structural assembly formed at a water surface can be transferred to a substrate supports the formation of an autosupported and cohesive film. TEM images reveal that the films formed are rather homogeneous, although some aggregates are sparingly observed, especially at high (10 μg·mL−1) concentration (Figure 8, left). High magnifications of zones devoid of aggregates indeed seem to be relatively homogeneous (Figure 8, middle). The silk protein selfassembly at the interface thus seems to be relatively smooth and ordered. This observation contrasts with the presence of oriented fibers as proposed for B. mori films formed at the air− water interface and transferred by the Langmuir−Blodgett technique.2 The films often exhibit defects such as cracks and holes as a result of the transfer operation (Figure 8, right), which provides insights into its mechanical properties: the sharp, clean edges are characteristic of a broken solid film, consistent with the results obtained by BAM. These observations clearly indicate that the film is not fluid and possesses a certain degree of rigidity. This is reminiscent of the mechanical properties of the fiber that is stiff and extensible.

■

AUTHOR INFORMATION

Notes

The authors declare no competing financial interest.

■

ACKNOWLEDGMENTS

We thank Richard Janvier (Institut de Biologie Intégrative et des Systèmes (IBIS), Université Laval) for his technical assistance and for his insightful help in the interpretation of the TEM experiments. Part of this work was supported by the Natural Sciences and Engineering Research Council (NSERC) of Canada and the Fonds de Recherche du Québec − Nature et Technologies (FRQNT). Part of this work was supported by the Centre National de la Recherche Scientifique (PICS CNRS Franco-Canadien).

■

CONCLUSIONS This study shows that natural MA proteins have a propensity to form films at the air−water interface and that their amphiphilic properties are very similar to those of their shorter MaSpI recombinant counterparts. As protein molecules reach the interface, a rapid, spontaneous conformational change occurs,

Figure 8. TEM images of the interfacial film (10 μg/mL) transferred onto a grill after 5 h of adsorption (π ≈ 10−12 mN/m). (Left) 5 keV (scale bar = 2 μm). (Middle) 300 keV (scale bar = 50 nm). (Right) Other region, 5 keV (scale bar = 2 μm). G


Langmuir

■

Article

(24) Paquet-Mercier, F.; Lefèvre, T.; Auger, M.; Pézolet, M., Evidence by infrared spectroscopy of the presence of two types of β-sheets in major ampullate spider silk and silkworm silk. Soft Matter 2012. (25) Marsh, D. Dichroic ratios in polarized Fourier transform infrared for nonaxial symmetry of β-sheet structures. Biophys. J. 1997, 72, 2710−2718. (26) Harrick, N. J. Internal Reflection Spectroscopy; Harrick Scientific Corporation: Ossining, NY, 1987; p 327. (27) De Feijter, J. A.; Benjamins, J.; Veer, F. A. Ellipsometry as a tool to study the adsorption behavior of synthetic and biopolymers at the air-water inerface. Biopolymers 1978, 17, 1759−1772. (28) Pezennec, S.; Gauthier, F.; Alonso, C.; Graner, F.; Croguennec, T.; Brulé, G.; Renault, A. The protein net electric charge determines the surface rheological properties of ovalbumin at the air-water interface. Food Hydrocolloids 2000, 14, 463−472. (29) Vénien-Bryan, C.; Lenne, P.-F.; Zakri, C.; Renault, A.; Brisson, A.; Legrand, J.-F.; Berge, B. Characterization of the growth of 2D protein crystals on a lipid monolayer by ellipsometry and rigidity measurements coupled to electron microscopy. Biophys. J. 1998, 74, 2649−2657. (30) Renault, A.; Lenne, P.-F.; Zakri, C.; Aradian, A.; Vénien-Bryan, C.; Amblard, F. Surface-induced polymerization of actin. Biophys. J. 1999, 76, 1580−1590. (31) Ferry, J. D. Molecular Theory for Undiluted Amorphous Polymers and Concentrated Solutions; Network and Entanglements. Viscoelastic Properties of Polymers, 3rd ed.; John Wiley and Sons: New York, 1980; pp 224−263. (32) Mark, J.; Mgai, K.; Graessley, W.; Mandelkern, L.; Samulski, E.; Koenig, J.; Wignall, G., Physical Properties of Polymers, 3rd ed.; Cambridge University Press: Cambridge, U.K., 2003; p 519. (33) Riande, E.; Díaz-Calleja, R.; Prolongo, M. G.; Masegosa, R. M.; Salom, C., Viscoelastic Behavior of Polymers above Tg. In Polymer Viscoelasticity: Stress and Strain in Practice; Hudgin, D. E., Ed.; Marcel Dekker: New York, 2000. (34) Xiaa, X.-X.; Qiana, Z.-G.; Kib, C. S.; Park, Y. H.; Kaplan, D.; Lee, S. Y. Native-sized recombinant spider silk protein produced in metabolically engineered Escherichia coli results in a strong fiber. Proc. Natj. Acad. Sci. U.S.A. 2010, 107, 10459−14063. (35) Clark, A. H.; Judge, F. J.; Richards, J. B.; Stubbs, J. M.; Suggett, A. Electron microscopy of network structures in the thermally-induced globular protein gels. Int. J. Pept. Protein Res. 1981, 17, 380−392. (36) Jackson, M.; Haris, P. I.; Chapman, D. Fourier transform infrared spectroscopic studies of Ca2+-binding proteins. Biochemistry 1991, 30, 9681−9686. (37) Krimm, S.; Bandekar, J. Vibrational spectroscopy and conformation of peptides, polypeptides, and proteins. Adv. Protein Chem. 1986, 38, 181−364. (38) Susi, H.; Timasheff, S. N.; Stevens, L. Infrared spectra and protein conformations in aqueous solutions. I. The amide I band in H2O and D2O solutions. J. Biol. Chem. 1967, 242, 5460−5466. (39) Goormaghtigh, E.; Cabiaux, V.; Ruysschaert, J.-M., Determination of soluble and membrane protein structure by Fourier transform spectroscopy. II. Experimental aspects, side-chain structure and H/D exchange. In Physico-Chemical Methods in the Study of Biomembranes; Hilderson, H. J., Ralston, G. B., Eds.; Plenum Press: New York, 1994; Vol. 23, pp 363−403. (40) Jenkins, J. E.; Creager, M. S.; Lewis, R. V.; Holland, G. P.; arger, J. L. Y. Quantitative correlation between the protein primary sequences and secondary structures in spider dragline silks. Biomacromolecules 2010, 11, 192−200. (41) Lefèvre, T.; Rousseau, M.-E.; Pézolet, M. Protein secondary structure and orientation in silk as revealed by Raman spectromicroscopy. Biophys. J. 2007, 92, 2885−2895.

REFERENCES

(1) Muller, W. S.; Samuelson, L. A.; Fossey, S. A.; Kaplan, D. L. Formation and characterization of Langmuir silk films. Langmuir 1993, 9, 1857−1861. (2) Wang, L.; Xie, H.; Qiao, X.; Goffin, A.; Hodgkinson, T.; Yuan, X.; Sun, K.; Fuller, G. G. Interfacial rheology of natural silk fibroin at air/ water and oil/water interfaces. Langmuir 2012, 28, 459−467. (3) Yang, Y.; Dicko, C.; Bain, C. D.; Gong, Z.; Jacobs, R. M. J.; Shao, Z.; Terry, A. E.; Vollrath, F. Behavior of silk protein at the air−water interface. Soft Matter 2012, 8, 9705−9712. (4) Valluzzi, R.; Gido, S. P. The crystal structure of Bombyx mori fibroin at the air-water interface. Biopolymers 1997, 42, 705−717. (5) Valluzzi, R.; Gido, S. P.; Zhang, W.; Muller, W. S.; Kaplan, D. L. Trigonal crystal structure of Bombyx mori silk incorporating a threefold helical chain conformation found at the air−water interface. Macromolecules 1996, 29, 8606−8614. (6) Valluzzi, R.; Gido, S. P.; Muller, S. W.; Kaplan, D. L. Orientation of silk III at the air-water interface. Int. J. Biol. Macromol. 1999, 24, 237−242. (7) Renault, A.; Rioux-Dubé, J.-F.; Lefèvre, T.; Pezennec, S.; Beaufils, S.; Vié, V.; Tremblay, M.; Pézolet, M. Surface properties and conformation of Nephila clavipes spider recombinant silk proteins at the air−water interface. Langmuir 2009, 25, 8170−8180. (8) Vézy, C.; Hermanson, K. D.; Scheibel, T.; Baush, A. R. Interfacial rheological properties of recombinant spider-silk proteins. Biointerphases 2009, 4, 43−46. (9) Denny, M. The physical properties of spider’s silk and their role in the design of orb-webs. J. Exp. Biol. 1976, 65, 483−506. (10) Gosline, J.; Denny, M.; Demont, M. E. Spider silk as rubber. Nature 1984, 309, 551−552. (11) Gosline, J. M.; DeMont, M. E.; Denny, M. W. The structure and properties of spider silk. Endeavour 1986, 10, 37−43. (12) Vollrath, F.; Porter, D.; Holland, C. There are many more lessons still to be learned from spider silks. Soft Matter 2011, 7, 9595− 9600. (13) Sponner, A.; Schlott, B.; Vollrath, F.; Unger, E.; Grosse, F.; Weisshart, K. Characterization of the protein components of Nephila clavipes dragline silk. Biochemistry 2005, 44, 4727−4736. (14) Brooks, A. E.; Steinkraus, H. B.; Nelson, S. R. An investigation of the divergence og major ampullate silk fibers from Nephila clavipes and argiope aurantia. Biomacromolecules 2005, 6, 3095−3099. (15) Xu, M.; Lewis, R. V. Structure of a protein superfiber - spider dragline silk. Proc. Natl. Acad. Sci. U.S.A. 1990, 87, 7120−7124. (16) Hinnan, M. B.; Lewis, R. V. Isolation of a clone encoding a second dragline silk fibroin - Nephila clavipes dragline silk is a twoprotein fiber. J. Biol. Chem. 1992, 267, 19320−19324. (17) Hayashi, C. Y.; Shipley, N. H.; Lewis, R. V. Hypotheses that correlate the sequence, structure, and mechanical properties of spider silk proteins. Int. J. Biol. Macromol. 1999, 24, 271−275. (18) Hinman, M. B.; Jones, J. A.; Lewis, R. V. Synthetic spider silk: a modular fiber. Trends Biotechnol. 2000, 18, 374379. (19) Lewis, R. V. Spider silk: ancient ideas for new biomaterials. Chem. Rev. 2006, 106, 3762−3774. (20) Renault, A.; Pézennec, S.; Gauthier, F.; Vié, V.; Desbat, B. Surface rheological properties of native and S-ovalbumin are correlated with the development of an intermolecular β-sheet network at the airwater interface. Langmuir 2002, 18, 6887−6895. (21) Zakri, C.; Renault, A.; Berge, B. Comparison between macroscopic and microscopic measurements of the shear elastic constant of alcohol monolayers at the air-water interface. Physica B 1998, 248, 208−210. (22) Sarkis, J.; Vié, V.; Winder, S. J.; Renault, A.; Le Rumeur, E.; Hubert, J.-F. Resisting sarcolemmal rupture: dystrophin repeats increase membrane-actin stiffness. FASEB J. 2013, 27, 359−367. (23) Bourque, H.; Laurin, I.; Pézolet, M. Investigation of the poly(Llactide)/poly(D-lactide) stereocomplex at the air−water interface by polarization modulation infrared reflection absorption spectroscopy. Langmuir 2001, 17, 5842−5849. H


Structure and Mechanical Properties of Spider Silk Films at the Air

Recommend Documents