Layer Thickness of Hydrophobin Films Leads to ... - ACS Publications

Article pubs.acs.org/Langmuir

Layer Thickness of Hydrophobin Films Leads to Oscillation in Wettability Leopold J. Gruner,* Kai Ostermann, and Gerhard Rödel Institute of Genetics, Technische Universität Dresden, 01217 Dresden, Germany S Supporting Information *

ABSTRACT: In nanobiotechnology, the properties of surfaces are often key to sensor applications. If analytes possess a low tolerance or affinity regarding the sensory substrate (surface), then the setup of mediators may be indicated. Hydrophobins enable biocompatible surface functionalization without significant restrictions of the physicochemical substrate properties. Because of the imperfect formation of hydrophobin films, a high variation in surface properties is observed. In this study, we report on the relation between the film thickness of hydrophobin-coated solid surfaces and their wettability. We found that the wettability of protein-coated surfaces strictly depends on the amount of adsorbed protein, as reflected in an oscillation of the contact angles of hydrophobin-coated silicon wafers. Fusion proteins of Ccg2 and HFBI, representatives of class I and II hydrophobins, document the influence of fused peptide tags on the wettability. The orientation of the first crystal nuclei plays a decisive role in the formation of the growing hydrophobin layers. Here, a simple method of deducing the film thickness of hydrophobin assemblies on solid surfaces is presented. The determination of the static contact angle allows the prediction of which part of the protein is exposed to possible analytes.

■

INTRODUCTION Biosensors typically encompass substrates such as biomolecules or whole cells that are immobilized on surfaces, allowing electric or physicochemical read-out. Usually, surface modifications are required for the respective applications. Changes in topology, altered wettability, and especially surface functionalization allow a more efficient accessibility of the immobilized substrates.1 During the last few years, a variety of coordinative methods have been developed in addition to covalent modification. Silanized surfaces are covalently modified substrates generated by the chemical immobilization of a bonding agent (mostly trialkoxysilanes) on silicon surfaces in the gas phase,2 in organic solvents,3 or in a combination of gas-phase deposition and water treatment at low temperatures (≤150 °C).4 Such composite materials are successfully used for DNA immobilization.5 In contrast, polylysine as a macromolecule is coordinatively immobilized on solid surfaces, allowing the adsorption of DNA and proteins.6 However, polylysine is often inhomogeneous in its surface coating, which leads to local differences or defects in electrostatic substrate binding. Besides the construction of irreversible covalently bound polyelectrolyte brushes,7,8 electrostatic immobilization can be achieved via layer-by-layer methods.9 For example, polyelectrolytes serve as biocompatible surfaces for the immobilization of biopolymers or whole cells, leading to a wide application spectrum in biosensors.10,11 © 2012 American Chemical Society

Proteins and polypeptides allow for reversible surface modification as well as the introduction of functional domains. In this regard, self-assembling proteins such as bacterial surface layer proteins or fungal hydrophobins that form ordered 2D structures at interfaces are of special interest.12,13 As in the case of surface coating with polylysines, the assembly of proteins is driven by electrostatic interactions. Ordered layer-by-layer deposition allows the controllable inversion of surface properties. Unlike polyelectrolytes, which show a polarity reversal from one layer to another, hydrophobins are expected to show an inversion in hydrophilicity because of their surfactant-like character. For application purposes, a high regularity and homogeneity of the assembled protein structures is required. This study focuses on hydrophobins, which are small globular proteins with amphiphilic properties. With regard to their amino acid pattern and physicochemical properties, hydrophobins are divided into two classes.14 The presence of both a hydrophobic and a hydrophilic patch gives them a high tendency to migrate to hydrophobic−hydrophilic interfaces. Unfortunately, this effect is also a major driving force for selfaggregation.14 Hydrophobin HFBI from Trichoderma reesei was chosen as a representative of class II hydrophobins. Depending on the protein concentration, HFBI is monomeric in solution (at concentrations below 50 ng/μL) or mainly tetrameric (at Received: October 30, 2011 Revised: January 5, 2012 Published: March 29, 2012 6942

dx.doi.org/10.1021/la204252y | Langmuir 2012, 28, 6942−6949

Langmuir

Article

concentrations higher than 150 ng/μL).15 In contrast, class I hydrophobins have a higher surface activity.16 In the case of Ccg2 from Neurospora crassa, the monomeric form was also observed at high protein concentrations in solution.17 In general, hydrophobins form symmetric structures, and in the case of class I hydrophobins, multiple rodlet structures can be observed after assembly at interfaces.18

■

(Multiscope, Optrel Berlin, Germany) equipped with a He−Ne laser (632.8 nm). Two different methods were used for hydrophobin immobilization on solid surfaces: the sessile drop and the drop-surface transfer methods. In the first case, clean and smooth silicon wafers (1 × 2 cm2) as well as Parafilm M (American National Can) and a Teflon foil (HBM) (both 1 × 2 cm2) were loaded with protein solutions in Tris buffer (50 mM at pH 8.5) in four different concentrations (10, 50, 100, and 200 ng/μL) for 15 to 60 min. With the drop shape method, a drop of hydrophobin solution in Tris buffer was positioned on Parafilm M and incubated for 30 min under ambient conditions. The film at the liquid/air interface was then transferred to solid substrates by touching the drop surfaces with the substrate. After several contacts, the substrate was removed. In both cases, surplus hydrophobin solutions were removed, and substrates were rinsed in a flow of ddH2O for 1 min to wash away weakly adsorbed proteins before protein-coated substrates were stored in dialysis buffer. Surface Characterization. After sample preparation, an exemplary investigation of layer morphology was done by liquid AFM to evaluate the homogeneity of chosen surface areas. Hydrophobin film thicknesses were measured in situ by nulling ellipsometry. The surface hydrophilicity of treated silicon wafers was determined by water contact angle measurements. In a final analysis, samples were treated with absolute ethanol to remove adsorbed protein, washed with ddH2O, and air dried. Afterward, the layer thickness was again determined by nulling ellipsometry. Atomic Force Microscopy. A JPK Nanowizard AFM (JPK Instruments, Germany) was used to image the protein films in liquid under ambient conditions at 21 ± 1 °C. Films were analyzed in buffer (50 mM Tris at pH 8.5 or 100 mM sodium chloride) via tapping mode. The DNP-S10 silicon nitride cantilever (Vecco Instruments Inc.) has a spring constant of 0.32 N·m−1, a resonance frequency in aqueous solution of ∼12 kHz, and a radius of curvature of ∼10 nm (for objects below ∼5 nm). Applied scanning parameters are the integral gain at 0.1−0.5 V, the proportional gain at 0−5 V, and the scan rate along the slow axis at 0.5−1 Hz. The samples were first aligned to the same orientation, and then a 10 μm × 10 μm section was examined to evaluate the homogeneity of the crystalline areas. Ellipsometry. Nulling ellipsometry was used to measure the spectral variation of the ellipsometric parameters Ψ (relative amplitude ratio) and Δ (relative phase shift) at a fixed angle of incidence of 68°. To calculate the layer thickness via a multilayer box model consisting of silicon, silicon dioxide, protein, and water, the refractive index of the protein layer was set to n = 1.375.21,22 The optical constants of silicon and silicon dioxide were taken from the literature.23 Determined values were averaged from six independent measurements recorded from different locations on a sample surface. All measurements were performed at 21 ± 1 °C and constant humidity in a uniformly airconditioned room. The relative standard deviation was in the range of 1−5%. Contact Angle Measurement. Static contact angles were measured in air by the sessile-drop method using a DSA10 system (Kruss GmbH, Germany) with an automatic dispenser (dispension rate 0.1 μL/s). Single values refer to six independent measurements at different substrate positions of settled 2 μL deionized water drops after a 5 s equilibration interval on a sample surface with a standard deviation of 1−3°.

MATERIALS AND METHODS

Cloning Strategy. Positional cloning of hydrophobin-encoding alleles was performed in two steps. Hf bI, encoding the mature version of HFBI(aa23−97) without an amino-terminal signal peptide, was amplified from the pET-17b-HFBIHA plasmid (Institute of Genetics, TU Dresden) using the primers listed in Table S1. The forward primer introduced restriction site NdeI at the 5′ end of the coding sequence, whereas the reverse primer allows the generation of chimeric hydrophobin fusion constructs. The fused sequences encode biomineralization tags that are derived from the R5 peptide, a repetitive subunit of the silaffine from Cylindrotheca f usiformis.19 The authentic R5 peptide (R5P) and a truncated version (XSR5P) were encoded by the oligonucleotides shown in Table S1. To avoid interference with proper protein folding and to ensure free tag accessibility, both domains were separated by a flexible (GGGGS)2 linker with a total length of approx 0.66 nm.20 Primers at the 3′ ends encode a stop codon and restriction site XhoI. After the restriction of PCR products with the respective endonucleases, the fragments were ligated in the pET28b(+) vector (Novagen, Germany), which had been digested with the same endonucleases. Ligation into pET28b(+) allows translational fusion with an N-terminal 6xHis tag that was used for purification. The resulting plasmids were transformed into E. coli TOP10. The inserted recombinant plasmids were analyzed by DNA sequencing. The pET23b-Ccg2HA plasmid encoding the mature form of the Ccg2 hydrophobin fused to a 3xHA tag was supplied from the strain collection, Institute of Genetics, TU Dresden (Figure S1). Expression and Purification of Hydrophobins. Plasmids encoding 6xHis-HFBI(aa23−97) -(R5P)2 and 6xHis-HFBI (aa23−97) (XSR5P)3 were transformed into E. coli SHuffle T7 Express cells (NEB, Germany) for expression. Overnight cultures were diluted in LB-Miller medium supplemented with kanamycin (30 ng/mL) to an optical density (OD600) of ∼0.4. Bacteria were grown at 37 °C, and gene expression was induced by adding IPTG to a final concentration of 0.4 mM. Cells were harvested 4 h after induction by centrifugation (4000g, 10 min, 4 °C). The cell pellet was resuspended in 50 mM Tris buffer (pH 7.0). Nucleic acids were removed by the addition of DNase A and RNase, and a protease inhibitor cocktail (Roche Diagnostics, Germany) was added to inhibit protein degradation. Cells were disrupted by three pulses of ultrasonification (9 cyc, 42 W, 5 min) with a Bandelin Sonopuls (Bandelin Electronic, Germany) and subsequent treatment with a French press. Proteins that were present in the insoluble fraction were isolated by extraction with urea buffer (pH 8.0). Proteins were purified via the His tag according to the manufacturer's protocol (Novagen, Germany) and analyzed by SDSPAGE. As an example, the results for HFBI-(XSR5P)3 are shown in Figure S2. Sample Preparation and Characterization of Self-Assembled Films. Isolated and purified hydrophobins were dialyzed against 50 mM Tris buffer (1 mM glutathion red., 0.2 mM glutathion ox., pH 8.5) at 4 °C for 24 h. For the separation of protein aggregates, the suspension was centrifuged (18 000g, 15 min, 4 °C). Silicon wafers were cleaned by ultrasonification in absolute ethanol for 10 min, and the excess solvent was removed by a high-pressure dry air treatment. To modify the wettability of the solid Si surface, Si wafers were etched with an oxidizing piranha solution (30 vol % H2O2/H2SO4 30:70) for 2 h at room temperature, rinsed with doubly distilled water (ddH2O) for 1 min, and dried with high-pressure air. This pretreatment minimizes the surface roughness and densifies the silanol groups (Si−OH) on the SiO2 surface. The thickness of the oxide layer on the silicon surface was determined using a nulling ellipsometer

■

RESULTS Fusion proteins consisting of HFBI and a truncated R5 peptide sequence (Figure S1) as well as Ccg2 were successfully expressed in E. coli, purified, and immediately used to coat prepared silicon surfaces, Parafilm M, and Teflon. The thickness of the adsorbed protein films was determined using precision nulling ellipsometry. Because hydrophobins possess distinct hydrophobic and hydrophilic patches, contact angle measurements were also chosen to characterize the hydrophobin-mediated wettability of the surface. In Figures 1 6943

dx.doi.org/10.1021/la204252y | Langmuir 2012, 28, 6942−6949

Langmuir

Article

Figure 1. Correlation between the protein film thickness and surface wettability. HFBI-(R5P)2-treated (upper graph) and HFBI-(XSR5P)3 (lower graph)-treated silicon wafers. Hydrophobin on pure and paraffin-waxed silicon (■) and on piranha-treated silicon (•). Tangents connect extreme points.

Figure 2. Correlation between surface wettability and a protein layer. Ccg2-HA-treated silicon wafer. Hydrophobin on pure silicon (■) and on paraffin-waxed silicon (•). Tangents connect extreme points. 6944

dx.doi.org/10.1021/la204252y | Langmuir 2012, 28, 6942−6949

Langmuir

Article

Table 1. Calculated Data from Figures 1 and 2a first layer

second layer

Si/SiO2/paraffin-waxed Si wafer a piranha-treated Si wafer

14.0 Å (55.7°) 13.8 Å (86.1°)

28.2 Å (50.9°)

a Si/SiO2/paraffin-waxed Si wafer a piranha-treated Si wafer

13.5 Å (58.4°) 15.0 Å (86.0°)

27.5 Å (89.7°) 26.2 Å (60.2°)

a Si/SiO2/piranha-treated Si wafer a paraffin-waxed Si wafer/Teflon

28.0 Å (45.1°) 22.8 Å (97.0°)

55.6 Å (82.3°) 55.5 Å (58.2°)

protein

third layer

HFBI-(R5P)2 on on HFBI-(XSR5P)3 on on Ccg2-HA on on

40.1 Å (65.6°)

81.6 Å (79.2°)

a

Extreme values of contact angle measurements (in parentheses) and the corresponding layer thicknesses (in bold) for protein-coated silicon wafers and piranha-etched silicon wafers are given.

and 2, the determined protein film thickness was related to the corresponding contact angle θ. The thickness of the SiO2 layer on the silicon substrates was 17 ± 1 Å. The protein films on the Si/SiO2 substrates had layer thicknesses of between 7 and 45 Å for HFBI-treated silicon wafers and 60 Å for Ccg2-treated silicon wafers. Overall, the measured contact angles θ of protein-coated Parafilm M, Teflon, and Si/SiO2 substrates are in the range from 44 to 95°. Contact angles of Teflon, Parafilm M, and ethanol-cleaned silicon substrates were 121 ± 0.9, 110.2 ± 1.5, and 45 ± 2°, respectively. The contact angle of the piranha-etched silicon wafers was too small to be determined accurately (θ ≤ 5°). Local inhomogeneity in the layers with submicrometer resolution could not be assessed with this method. The data showed no linear correlation between the layer thickness and wettability. The measured values were fitted with regression functions as shown in Figures 1 and 2. This approach allowed the determination of local extreme values shown in Table 1. By arithmetically averaging the determined extreme values, the protein film thickness for an HFBI(aa23−97)(R5P)2 layer was calculated to be dfilm = 14.0 ± 0.2 Å and that for the HFBI(aa23−97)-(XSR5P)3 layer was calculated to be dfilm = 13.6 ± 0.7 Å. For class I hydrophobin Ccg2, we determined a layer thickness of 27.4 ± 0.4 Å, a contact angle maximum of θ = 97 ± 0.6°, and a contact angle minimum of θ = 45 ± 1.2°. The influence of layer thickness on the surface wettability is documented by the alignment of tangents to the resulting curves by connecting the extreme values and the determination of the slope (Figure 1). As shown in Figure 1, minima are characterized on average by a higher number of values reflecting low contact angles, which may indicate a preferential tag direction in the hydrophilic medium. The resulting reduction in hydrophobicity leads to the formation of ordered, more homogeneous structures. The small number of structures, which expose the hydrophobic protein patch, seems to reflect an energetically unfavorable state in the aqueous media. Compared to the applied macroscopic measurements, liquid AFM images of hydrophobin layers provide microscopic resolution of the assembled protein layers. Images of silicon wafers coated with hydrophobins via the drop-shape method illustrate the formation of crystalline areas. For Ccg2-treated surfaces, small longitudinal structures with an average height of around 3 nm are detectable after a contact time between the solution and substrate of 1 min (Figure 3A). In contrast, a contact time of 10 min leads to the formation of unstructured areas with an average height of 6 nm (Figure 3B), where tubular structures are difficult to identify. AFM images of

immobilized HFBI show the formation of small regular structures with an average height of around 3 nm after a contact time of 1 min (Figure 4A). These small substructures are also visible after a contact time of 10 min or longer; however, they seem to be arranged on a complete layer (Figure 4B).

■

DISCUSSION The small globular hydrophobins have the ability to change the physicochemical nature of a surface after immobilization. Hydrophobins can be genetically modified and are conceivable as carriers for functional peptides and enzymes. Fusions between hydrophobins and other protein domains are possible at both the C and N termini of hydrophobins without affecting their structural integrity and high surface activity.24,25 Both termini are located within the hydrophilic protein patch.26,27 As seen in Figures 3 and 4, AFM images as well as other highresolution methods do not provide direct information about the orientation of hydrophobin monomers. Especially for sensor applications, the challenge is to guarantee a directed hydrophobin immobilization. As described in the literature, high protein concentrations and long incubation times lead to self-aggregation and the formation of thicker protein layers at interfaces.28 Interestingly, the formation of multilayers takes place in an ordered manner as evidenced by the sinusoidal correlation between the layer thickness and contact angle (Figures 1 and 2). Preferred orientations of protein units indicate active behavior in diffusion and assembly. In contrast to unspecific adsorption, it might be interpreted as a result of directed assembly. Furthermore, the results indicate that homogeneous layers are formed over a large area of the substrate by directed protein− protein interactions. The comparison of extrema in wettability and their corresponding film thicknesses in Figures 1 and 2 shows that the adsorbed hydrophobins form well-defined layers with mean thicknesses of 14.0 ± 0.2 and 13.6 ± 0.7 Å for HFBI constructs and 27.4 ± 0.4 Å for Ccg2 derivatives. The thicknesses of the adsorbed layers are in good agreement with those at the air− water interface with mean thicknesses of 13 ± 2 and 25 Å.26,29 The difference in wettability of the determined monolayers implies a different orientation of the proteins on the two substrates. Because of their amphiphilic character, hydrophobins are able to interact with substrates by hydrophobic or hydrophilic interactions depending on the surface wettability. The protein bilayers show an inversion in wettability compared to the monolayer, which implies that 6945

dx.doi.org/10.1021/la204252y | Langmuir 2012, 28, 6942−6949

Langmuir

Article

Figure 3. AFM topography images of class I hydrophobin Ccg2-HA. Hydrophobin films of Ccg2-HA created by the drop shape method after equilibration times of (A) 1 and (B) 10 min.

Figure 4. AFM topography images of class II hydrophobin HFBI-(R5P)2. Hydrophobin films of HFBI-(R5P)2 created by the drop shape method after equilibration times of (A) 1 and (B) 10 min.

Figure 1 from the first minimum thickness of coated silicon wafers indicate that the hydrophilic protein patch is exposed. However, the deviation in contact angle of approximately 3° may result from the different behavior of the exposed protein parts. This assumption is reinforced by the position of the first minimum in the inverted layer formation. Minima of different tagged HFBI proteins differ in Δθ by about 9.5°. It can be concluded that the fused tag is oriented into solution and directly influences the hydrophilic character of the fusion construct. The extreme values that were tangentially connected in Figure 1 indicate a protein-dependent effect on the surface properties. Depending on the hydrophobin fusion protein, the increase in layer thickness leads to different tendencies in wettability. In contrast to HFBI-(XSR5P)3 (Δθ = +3.9°), HFBI-(R5P)2 (Δθ = −4.7°) shows a decreasing trend in contact angles with increasing layer thickness. R5 peptides possess the ability to form self-assembled structures, which may lead to an interaction of the embedded fusion domains.31 As a result, the protein subunits harbor a slightly tilted position, leading to a more surface-orientated tag direction. Depending on the size and hydrophilicity of the fused tag, it leads to a decrease or increase in wettability. Comparing the values determined by macroscopic analytical methods, which imply a regular oscillation, with AFM images (Figure 4), a layer-by-layer formation of class II hydrophobin subunits becomes apparent.

HFBI bilayers are probably arranged in the same basic dimer orientation as observed for crystal structures.27 HFBI Adsorption onto Hydrophilic and Hydrophobic Solid Surfaces. Two HFBI fusion proteins were deposited on surfaces of varying wettability. Both constructs differ in the amino acid composition of their fused tags. By the calculation of the Kyte−Doolittle hydrophobicity index (KD), it was determined that the XSR5P domain (KD = −0.72) represents a slightly more hydrophobic polypeptide chain than does the R5P (KD = −1.80).30 Hakanpää et al. described the formation of a planar hydrophobic patch by solely hydrophobic aliphatic side chains (residues Leu12, Val23, Leu24, Leu26, Ile27, Leu29, Val59, Ala60, Val62, Ala63, Ala66, Leu67, and Leu68), which enables HFBI to interact with hydrophobic substrates.29 The calculated KD index is 2.97. For HFBI-coated piranha-etched silicon wafers, we identified a tag-independent contact angle of 86.0 ± 0.1° for the first minimum. This contact angle suggests that the hydrophilic protein patches may serve as surface mediators whereas the hydrophobic patch is exposed to the medium. This effect may be enhanced by the preferred affinity of the R5 peptide for hydroxyl-group-densified SiO2 surfaces.19 From X-ray measurements, it is known that the hydrophobic patch and the C and N termini flank the protein in the longitudinal direction.27 The maximum distance between both protein parts hinders a direct interaction. In contrast, pure silicon and paraffin-waxed wafers represent a more hydrophobic structure. The contact angles displayed in 6946

dx.doi.org/10.1021/la204252y | Langmuir 2012, 28, 6942−6949

Langmuir

Article

Ccg2 Adsorption onto Hydrophilic and Hydrophobic Solid Surfaces. The layer thickness (27.4 ± 0.4 Å) of Ccg2 as a representative of class I hydrophobins is in good agreement with the film thickness of 25 Å determined by AFM.26 Similarly, the overall structure is in line with a high content of β barrels.26 The small deviation in layer thickness may be explained by the type of layer formation. The protein layer at the water−air interface represents a more flexible system than a water−solid interface. Furthermore, the protein transfer to a solid interface via a dot-blotting method could lead to local defects. The determination of layer thicknesses by AFM also depends on the properties of the cantilever. Using the X-ray fiber diffraction of Ccg2 rodlets, a monomer size of 27 Å was determined.26 The contact angle of the minimum in Figure 2 is in good agreement with the literature value given for the exposed hydrophilic side of class I hydrophobin SC3.12 For the hydrophobic side of surface-assembled Ccg2, a wettability of θ = 97.0 ± 0.6° was determined. This is approximately 83% in comparison with the most surface-active protein SC3 (θ = 117 ± 8°).32 From the results of contact angle measurements, we can deduce that, in comparison to class II hydrophobins, immobilized Ccg2 does not form inverted layer structures on pure silicon substrates. Although untreated silicon seems to be sufficient to serve as a hydrophobic substrate for the immobilization of inverted class II hydrophobin structures, the formation of adequate class I assemblies requires much more hydrophobic surfaces. The absence of tubular structures in AFM images of multilayered Ccg2 assemblies (Figure 3B) implies a nondirected interaction of hydrophobin monomers. Furthermore, the minimal formation of supermolecular protein structures on the first completed layer may lead to an attenuation in the sinusoidal behavior of the wettability in relation to the thickness of the assembled protein as shown in Figure 2. Model. The layer-by-layer deposition of oppositely charged polyelectrolytes on surfaces allows a variation in layer thickness and its physical properties, which is affected by the number of charges.9 By analogy, hydrophobins also affect an alternation in surface properties from layer to layer. In contrast to polyelectrolytes, however, the alternating positioning of subunits is based on the domino-like protein structure (Figure 5). Considering the measurement deviation that results from the laser beam size (2 × 1 nm2) and the drop size (2 μL), the

applied methods produce only relative values. Because of a 6fold repetition of selective measurements, macroscopic results imply highly ordered areas. In agreement with the theory of protein crystallization,33,34 single-crystalline domains take up a preferred orientation and transfer it to further growing particles. The formation of nucleation points strictly depends on the nature of the substrate, as shown by Lopez et al.35 for surface layer proteins. Therefore, we created a model for the formation of partially homogeneous layers as shown in Figure 5. Depending on the orientation of the proteins, layers are formed with exposed or embedded fusion peptides. For the functionalization of surfaces in sensor technology, it is essential to know the orientation of the proteins. The oscillation in wettability results in an alternating influence of hydrophobic hydrophobin patches and the tagged domains, which implies an inverted positioning of the subsequent protein layer. Kyte−Doolittle-based calculations show the existence of an additive effect, which finally leads to an increase or decrease in the wettability of two adjacent layers.30 Depending on the protein concentration and incubation time, the formation of a multilayer is initiated. The growing inverted layer results in the ordered deposition of additional monomers or whole protein aggregates, which again serve as nucleation points. In contrast to the data of Kisko et al., organized single-crystalline domains seem to act as nucleation centers that lead to the formation of homogeneous layers.36 According to the literature, hydrophobin-coated surfaces show a wide variation in contact angle values. For example, contact angles of HFBI-modified hydrophobic surfaces (PCL, PDMS, or gold) range between 45.3 and 64°.37−40 For hydrophilic surfaces (glas), Askolin et al.37 published a contact angle of about 72 ± 15°. Kwan et al.41 reported a contact angle for Ccg2-coated Teflon of 80.4 ± 10.1. In none of these publications was the determination of the protein layer thickness the focus. Contact angle measurements just served as a simple test to prove the successful assembly of hydrophobins. Our model may provide an explanation for the reported variations of hydrophobin-treated surfaces. Our results provide a simple possibility to determine the orientation of hydrophobin monomers depending on the layer thickness. In comparison to other available (microscopic) surface-investigation methods such as AFM, scanning electron microscopy (SEM), or transmission electron microscopy (TEM), our macroscopic method offers several advantages. It is simple and fast in preparation and handling. For example, protein samples observed by SEM and TEM have to be coated with a thin layer of heavy metal to increase the contrast, but effects on the morphology of the protein layer cannot be excluded. Furthermore, the combination of two cheap surface observation methods allows a sufficient evaluation of the layer thickness of immobilized hydrophobins over large-scale areas. We were able to show by AFM in liquid that ellipsometric and contact angle measurements of dried and reswollen hydrophobin-coated samples yield results that are identical to those under native conditions. To ensure the maximal reactivity of modified and functionalized hydrophobins, directed monolayered protein deposition is necessary. Depending on the properties of the fused tag, its accessibility is crucial to applications in enzymatic and chemical processes. Our method allows not only a determination of the layer thickness and a definition of the exposed protein side but also qualitative statements about the content of similarly orientated protein units.

Figure 5. Model for protein-assembled monolayers. (A) Hydrophobic substrate−protein interaction allows the exposure of the C-terminal fused protein domain, which defines the surface wettability. (B) Protein-coated piranha-etched silicon wafer resulting in the inverted assembly of protein monomers. The coupled protein domain is embedded in the protein layer. 6947

dx.doi.org/10.1021/la204252y | Langmuir 2012, 28, 6942−6949

Langmuir

■

Article

CONCLUSIONS Experiments were performed for representatives of class I and class II hydrophobins. Genetically engineered hydrophobins differing in the hydrophilicity of the fused tag were constructed and used as model proteins. Our results show that the exposure of the hydrophobic patch of Ccg2 or HFBI and its C-terminal fused peptide alternates from layer to layer. The hydrophilic character of the fused peptide strongly influences the surface properties, which must be taken into account in the generation of protein layers. Furthermore, it plays a decisive role in the interaction between the protein and the substrate. The contact angles of monolayers formed by the different HFBI fusion proteins on silicon wafers that expose their hydrophobic patches are identical. This indicates that the fused domain has no direct influence on the wettability on this part of the protein. Derived minima and maxima in the hydropathy of hydrophobin-treated substrates represent multiple similarly structured hydrophobin layers. In contrast, both AFM images and the attenuation in the oscillation behavior of class I hydrophobin Ccg2 imply unstructured protein deposition on the first monolayer. We assume that these results can be transferred to other surface-active proteins. In Figure 5, a simple model of the layer formation of adsorbed protein is proposed, considering two preferred protein orientations. The model is in line with the experimental data obtained with ellipsometry, contact angle measurements, and AFM. The results indicate that the direction of monomer adsorbance plays an important role in the accessibility of the fused tag. For the design and construction of functionalized hydrophobin monolayers, it is essential to calculate the hydrophobic character of the fused tag to ensure that it will be exposed to the medium. Our study demonstrates that a higher amount of adsorbed protein must not be correlated with an increase in the accessibility of the fused domain. In conclusion, we established a simple method to determine the film thickness and orientation of surface-active proteins without using complex and expensive techniques. It is based on a comparison of film thicknesses and corresponding changes in protein-specific properties and allows quantitative and qualitative statements to be made about the homogeneity of hydrophobin-coated substrates.

■

of Physical Chemistry of Polymers, TU Dresden, Germany, for supporting our contact angle measurements and the Institute for Polymer Research, Dresden, Germany, for providing ellipsometer measurement time.

■

(1) Kasemo, B. Surf. Sci. 2002, 500, 656−677. (2) Evans, B.; White, T. E. J. Catal. 1968, 11, 336−341. (3) Boerio, F. J.; Cheng, S. Y. J. Colloid Interface Sci. 1979, 68, 252− 259. (4) Ek, S.; Iiskola, E. I.; Niinistö, L. J. Phys. Chem. B 2004, 108, 9650−9655. (5) Guo, Z.; Guilfoyle, R. A.; Thiel, A. J.; Wang, R.; Smith, L. M. Nucleic Acids Res. 1994, 22, 5456−5465. (6) Williams, R. C. Proc. Natl. Acad. Sci. U.S.A. 1977, 74, 2310−2315. (7) Miklavic, S. J.; Marčelja, S. J. Phys. Chem. 1988, 92, 6718−6722. (8) Chen, H.; Zajac, R.; Chakrabarti, A. J. Chem. Phys. 1996, 104, 10. (9) Decher, G. Science 1997, 277, 1232. (10) Blacklock, J.; Handa, H.; Manickam, D. S.; Mao, G.; Mukhopadhyay, A.; Oupický, D. Biomaterials 2007, 28, 117−124. (11) Borghol, N.; Mora, L.; Sakly, N.; Lejeune, P.; Jouenne, T.; Jaffrézic-Renault, N.; Othmane, A. J. Biotechnol. 2011, 151, 114−121. (12) Scholtmeijer, K.; Janssen, M. I.; Gerssen, B.; de Vocht, M. L.; van Leeuwen, B. M.; van Kooten, T. G.; Wösten, H. A. B.; Wessels, G. H. Appl. Environ. Microb. 2002, 68, 1367−1373. (13) Sleytr, U. B.; Györvary, E.; Pum, D. Prog. Org. Coat. 2003, 47, 279−287. (14) Torkkeli, M.; Serimaa, R.; Ikkala, O.; Linder, M. Biophys. J. 2002, 83, 2240−2247. (15) Szilvay, G. R.; Nakari-Setälä, T.; Linder, M. B. Biochemistry 2006, 45, 8590−8598. (16) Van der Vegt, W.; van der Mei, H. C.; Wösten, H. A. B.; Wessels, J. G. H.; Busscher, H. Biophys. Chem. 1996, 57, 253−260. (17) Mackay, J. P.; Matthews, J. M.; Winefield, R. D.; Mackay, L. G.; Haverkamp, R. G.; Templeton, M. D. Structure 2006, 9, 83−91. (18) Linder, M. B.; Szilvay, G. R.; Nakari-Setälä, T.; Penttilä, M. E. FEMS Microbiol. Rev. 2005, 29, 877−896. (19) Kröger, N.; Deutzmann, R.; Sumper, M. Science 1999, 286, 1129−1132. (20) Huston, J. S.; Levinson, D.; Mudgett-Hunter, M.; Tai, M.-S.; Novotný, J.; Margolies, M. N.; Ridge, R. J.; Bruccoleri, R. E.; Heber, E.; Crea, R.; Oppermann, H. Proc. Natl. Acad. Sci. U.S.A. 1988, 85, 5879− 5883. (21) Werner, C.; Eichhorn, K.-J.; Grundke, K.; Simon, F.; Grahlert, W.; Jacobasch, H. J. Colloids Surf., A 1999, 156, 3−17. (22) Reichelt, S.; Eichhorn, K.-J.; Aulich, D.; Hinrichs, K.; Jain, N.; Appelhans, D.; Voit, B. Colloids Surf., B 2009, 69, 169−177. (23) Synytska, A.; Appelhans, D.; Wang, Z. G.; Simon, F.; Lehmann, F.; Stamm, M.; Grundke, K. Macromolecules 2007, 40, 297−305. (24) Linder, M.; Szilvay, G. R.; Nakari-Setälä, T.; Söderlund, H.; Penttilä, M. Protein Sci. 2002, 11, 2257−2266. (25) Nakari-Setälä, T.; Aro, N.; Kalkkinen, N.; Alatalo, E.; Penttilä, M. Eur. J. Biochem. 1996, 235, 248−255. (26) Kwan, A. H. Y.; Winefield, R. D.; Sunde, M.; Matthews, J. M.; Haverkamp, R. G.; Templeton, M. D.; Mackay, J. P. Proc. Natl. Acad. Sci. U.S.A. 2006, 103, 3621−3626. (27) Hakanpäa,̈ J.; Szilvay, G. R.; Kaljunen, H.; Maksimainen, M.; Linder, M.; Rouvinen, J. Protein Sci. 2006, 15, 2129−2140. (28) Wang, X.; Graveland-Bikker, J. F.; de Kruif, C. G.; Robillard, G. T. Protein Sci. 2004, 13, 810−821. (29) Paananen, A.; Vuorimaa, E.; Torkkeli, M.; Penttilä, M.; Kauranen, M.; Ikkala, O.; Lemmetyinen, H.; Serimaa, R.; Linder, M. B. Biochemistry 2003, 42, 5253−5258. (30) Kyte, J.; Doolittle, R. F. J. Mol. Biol. 1982, 157, 105−132. (31) Knecht, M. R.; Wright, D. W. Chem. Commun. 2003, 24, 3038− 3039. (32) Wösten, H. A. B.; de Vocht, M. L. Biochim. Biophys. Acta 2000, 1469, 79−86.

ASSOCIATED CONTENT

S Supporting Information *

Oligodesoxynucleotides (ODN) used for DNA amplification. Mature hydrophobins HFBI from Trichoderma reesei fused to the R5 peptide (R5P) from Cylindrotheca f usiformis or a truncated version separated by a (GGGGS)2 linker (L). This material is available free of charge via the Internet at http:// pubs.acs.org.

■

REFERENCES

AUTHOR INFORMATION

Corresponding Author

**E-mail: [email protected]. Notes

The authors declare no competing financial interest.

■

ACKNOWLEDGMENTS This work was supported by the DFG Research Training Group “Nano- and Biotechniques for the Packaging of Electronic Systems” (DFG 1401/1). We thank the Institute 6948

dx.doi.org/10.1021/la204252y | Langmuir 2012, 28, 6942−6949

Langmuir

Article

(33) Rosenberger, F. J. Cryst. Growth 1986, 76, 618−636. (34) McPherson, A. Eur. J. Biochem. 1990, 189, 1−23. (35) Lopez, A. E.; Moreno-Flores, S.; Pum, D.; Slytr, U. B.; TocaHerrera, J. L. Small 2010, 6, 396−403. (36) Kisko, K.; Torkkeli, M.; Vuorimaa, E.; Lemmetyinen, H.; Seeck, O. H.; Linder, M.; Serimaa, R. Surf. Sci. 2005, 584, 35−40. (37) Askolin, S.; Linder, M.; Scholtmeijer, K.; Tenkanen, M.; Penttilä, M.; de Vocht, M. L.; Wösten, H. A. B. Biomacromolecules 2006, 7, 1295−1301. (38) Zhao, Z.-X.; Wang, H.-C.; Qin, X.; Wang, X.-S.; Qiao, M.-Q.; Anzai, J.; Chen, Q. Colloids Surf., B 2009, 71, 102−106. (39) Ahlroos, T.; Hakala, T. J.; Helle, A.; Linder, M. B.; Holmberg, K.; Mahlberg, R.; Laaksonen, P.; Varjus, S. P. I. Mech. Eng. J-J. Eng. 2011, 225, 1013−1022. (40) Zhang, M.; Wang, Z.; Wang, Z.; Feng, S.; Xu, H.; Zhao, Q.; Wang, S.; Fang, J.; Qiao, M.; Kong, D. Colloids Surf., B 2011, 85, 32− 39. (41) Kwan, A. H.; Macindoe, I.; Vukašin, Morris, V. K.; Kass, I.; Gupte, R.; Mark, A. E.; Templeton, M. D.; Mackay, J. P.; Sunde, M. J. Mol. Biol. 2008, 382, 708−720.

6949

dx.doi.org/10.1021/la204252y | Langmuir 2012, 28, 6942−6949