Dynamic Adsorption Behavior of Poly(3-hydroxybutyrate

Abstract. Time-dependent adsorption behavior of poly(3-hydroxybutyrate) (PHB) depolymerase from Ralstonia pickettii T1 on a polyester surface was stud...
0 downloads 0 Views 491KB Size
Download PDF
Biomacromolecules 2005, 6, 2084-2090

2084

Dynamic Adsorption Behavior of Poly(3-hydroxybutyrate) Depolymerase onto Polyester Surface Investigated by QCM and AFM Yoshihiro Kikkawa,*,†,‡ Koichi Yamashita,*,‡,§ Tomohiro Hiraishi,‡ Masatoshi Kanesato,† and Yoshiharu Doi‡ Nanoarchitectonics Research Center, National Institute of Advanced Industrial Science and Technology (AIST), Tsukuba Central 4, 1-1-1 Higashi, Tsukuba, Ibaraki 305-8562 Japan, Polymer Chemistry Laboratory, RIKEN Institute, 2-1 Hirosawa, Wako-shi, Saitama 351-0198 Japan, and Advanced Development & Supporting Center, RIKEN Institute, 2-1 Hirosawa, Wako-shi, Saitama 351-0198 Japan Received February 1, 2005; Revised Manuscript Received March 16, 2005

Time-dependent adsorption behavior of poly(3-hydroxybutyrate) (PHB) depolymerase from Ralstonia pickettii T1 on a polyester surface was studied by complementary techniques of quarts crystal microbalance (QCM) and atomic force microscopy (AFM). Amorphous poly(L-lactide) (PLLA) thin films were used as adsorption substrates. Effects of enzyme concentration on adsorption onto the PLLA surface were determined timedependently by QCM. Adsorption of PHB depolymerase took place immediately after replacement of the buffer solutions with the enzyme solutions in the cell, followed by a gradual increase in the amount over 30 min. The amount of PHB depolymerase molecules adsorbed on the surface of amorphous PLLA thin films increased with an increase in the enzyme concentration. Time-dependent AFM observation of enzyme molecules was performed during the adsorption of PHB depolymerase. The phase response of the AFM signal revealed that the nature of the PLLA surface around the PHB depolymerase molecule was changed due to the adsorption function of the enzyme and that PHB depolymerase adsorbed onto the PLLA surface as a monolayer at a lower enzyme concentration. The number of PHB depolymerase molecules on the PLLA surface depended on the enzyme concentration and adsorption time. In addition, the height of the adsorbed enzyme was found to increase with time when the PLLA surface was crowded with the enzymes. In the case of higher enzyme concentrations, multilayered PHB depolymerases were observed on the PLLA thin film. These QCM and AFM results indicate that two-step adsorption of PHB depolymerase occurs on the amorphous PLLA thin film. First, adsorption of PHB depolymerase molecules takes place through the characteristic interaction between the binding domain of PHB depolymerase and the free surface of an amorphous PLLA thin film. As the adsorption proceeded, the surface region of the thin film was almost covered with the enzyme, which was accompanied by morphological changes. Second, the hydrophobic interactions among the enzymes in the adlayer and the solution become more dominant to stack as a second layer. Introduction Poly(hydroxyalkanoate)s (PHAs) are biodegradable and biocompatible thermoplastics, which are produced from renewable carbon sources by a wide variety of bacteria.1-5 A notable feature of PHAs is its biodegradability in the natural environment. PHAs can be degraded by extracellular poly(3-hydroybutyrate) (PHB) depolymerases excreted by a number of microorganisms.6-11 Several extracellular PHB depolymerases have been purified, and the structure and properties of the enzymes have been characterized.6-11 It has been reported that PHB depolymerase is composed of three * Corresponding authors. (Y.K.) Phone: +81-29-861-2955. Fax: +8129-861-3029. E-mail: [email protected]. (K.Y.) Phone: +81-48-4678000. Fax: +81-48-462-4631. E-mail: [email protected]. † National Institute of Advanced Industrial Science and Technology (AIST). ‡ Polymer Chemistry Laboratory, RIKEN Institute. § AD & S Center, RIKEN Institute.

domains such as catalytic, binding, and linker domains. The catalytic domain is capable of cleaving the ester bonds of PHAs. The binding domain has a function to adsorb on the solid PHAs surface. These two domains are flexibly connected by the linker domain. It has been proposed that the enzymatic degradation of PHA materials by PHB depolymerase takes place via two steps, namely, the adsorption of the enzyme onto the solid polyester surface by the function of the binding domain and subsequent hydrolysis of the polymer chains by the function of the catalytic domain. Thus, the adsorption process of PHB depolymerase is a key step in the hydrolysis of the PHAs materials. The binding characteristics of PHB depolymerases on polymer surfaces have been studied. Shinomiya et al.12 and Kasuya et al.13 have investigated adsorption of PHB depolymerases by using the fusions of binding domains with glutathione S-transferase. They suggested that recognition of water-insoluble PHAs by the binding domain included specific chemical interactions

10.1021/bm0500751 CCC: $30.25 © 2005 American Chemical Society Published on Web 04/27/2005

Behavior of Poly(3-hydroxybutyrate) Depolymerase

between several amino acids in the binding domain and the polyester surface as well as hydrophobic interactions. Visualization of an adsorbed PHB depolymerase on PHB single crystals has been performed through the formation of gold-conjugated antibody with the binding domain by transmission electron microscopy (TEM).12-15 The PHB depolymerase was shown to distribute on the crystal surface homogeneously, suggesting that the enzymes randomly adsorb on the PHB crystal surface. In our previous study,16 adsorption of PHB depolymerase from Ralstonia pickettii T1 onto thin films of poly(L-lactide) (PLLA), poly[(R)-3-hydroxyoctanoate] (PHO), polystyrene (PS), and polyethylene (PE) was investigated by using QCM. The PHB depolymerase was found to bind irreversibly onto all of the polymer films, and the amount of the adsorbed enzyme depended on the kind of polymer. However, no erosion of the films was observed. Further QCM experiments17 have revealed that the PHB depolymerase did not release from the melt-crystallized PHB surface spontaneously due to the strong binding interactions. On the other hand, the adsorbed PHB depolymerase was substituted by the hydrolytic-activity-disrupted PHB depolymerase molecules in a buffer solution. Murase et al.18 reported the morphological change of the solution-grown PHB single crystals during adsorption of mutated PHB depolymerase from R. pickettii T1, which lacks hydrolytic activity. They suggested that the mutant enzyme distorted the surface morphology around the less-ordered molecular chain-packing region despite lacking hydrolytic activity, resulting in the fragmentation of the single crystal along the crystallographic a axis. Direct AFM observation of PHB depolymrase molecules was successfully achieved on the PLLA thin film.19 It was found that PHB depolymerases strongly adsorb on the polyester surface and that small ridges were formed around them. In addition, the formation of hollows was recognized after the removal of the enzyme molecules. These results have demonstrated that the PHB depolymerase interacts with polyester chains through the binding domain enough to create a hollow on the surface of the polyester film. Although many studies have supplied useful information for understanding the adsorption function of PHB depolymerase as described above, dynamic behavior of the enzyme during the adsorption process is still unclear. In addition, the research on the adsorption of the PHB depolymerase has been limited to the initial stage of the enzymatic reaction despite the fact that the enzyme adsorption may continue during the reaction. In this paper, we studied the timedependent adsorption behavior of PHB depolymerase from R. pickettii T1 on PLLA amorphous thin films by both QCM and AFM to further clarify the characteristics of the interaction between the enzyme and the polyester surface. QCM is recognized as a versatile technique to follow adsorption processes at a solid/liquid interface in chemical and biological research.16,17,20-22 AFM is known as a powerful tool to observe the surface morphological changes during hydrolytic23-25 and enzymatic degradation.26-28 In addition, PHB depolymerase molecules can be directly observed by AFM without specific sample preparation.19 These comple-

Biomacromolecules, Vol. 6, No. 4, 2005 2085

mentary techniques of QCM and AFM provided progressive information on time-dependent dynamic behavior of PHB depolymerase during adsorption process, such as multistep adsorption processes and morphological changes of the enzyme. Experimental Section Material. PLLA was purchased from Polysciences Inc. (Warrington, PA). The PLLA was dissolved in chloroform and reprecipitated in methanol. Molecular weight data of the PLLA were evaluated by gel permeation chromatography (GPC) at 40 °C, using a Shimazu 10A GPC system and a 6A refractive index detector with two joint columns of Shodex K-80M and K-802. Chloroform was used as an eluent at a flow rate of 0.8 mL/min. A molecular weight calibration curve was obtained with polystyrene standards of low polydispersity. The number-average molecular weight (Mn) and molecular weight distribution (Mw/Mn) were measured to be 410 000 and 1.7, respectively. Extracellular PHB depolymerase from R. pickettii T1 was purified to electrophoretic homogeneity according to the methods by Tanio et al.6 Preparation of Amorphous PLLA Thin Films. The PLLA thin film of ca. 100 nm thickness was prepared on a cleaned silicon wafer or QCM oscillator by a spin-cast method. A chloroform solution [1.0%(w/v)] of the PLLA was cast on the supports at a rotation speed of 4000 rpm under dry air. The thickness of the thin film depended on the concentration of PLLA in chloroform. AFM measurement revealed that a 100 nm thick film was formed on the substrate when the concentration of the PLLA was 1.0%(w/v). The cast thin film was melted at 220 °C for 30 s and then rapidly quenched to 0 °C on an ice block, resulting in the formation of a completely amorphous PLLA thin film.19,22 QCM Analysis. A commercially available QCM setup of QCA-920 (SEIKO EG&G) was equipped with a cell with an inner volume of ca. 0.2 mL which was newly designed to reduce fluctuation of the signals due to injection of the sample solution. Its temperature was maintained at 25 ( 0.1 °C by circulating water. The oscillator was a polished 9 MHz AT-cut quartz crystal, on both sides of which Au electrodes were deposited (area size: 0.189 cm2 × 2). The mean roughness of the electrode surface was estimated to be 1.2 nm (30 × 30 µm) by AFM. According to the Sauerbrey equation,29 a frequency shift of 1.0 Hz corresponds to a mass change of 1.0 ng on the electrode (0.189 cm2). In our system, the resonant frequency and the resistance were monitored simultaneously to evaluate the influence of friction change. After stabilization of the QCM in 10 mM phosphate buffer solution (pH 7.4), the enzymatic reaction was initiated by replacing the buffer solution with the enzyme solution. The total volume used for replacement was about 1.5 mL. The frequency changes during the reactions were recorded by a microcomputer system. AFM Observation. The PLLA amorphous thin film was immersed into a 10 mM phosphate buffered solution containing PHB depolymerase at concentrations of 1, 2, and 4 µg/mL. The PLLA thin film was treated with the enzyme

2086

Biomacromolecules, Vol. 6, No. 4, 2005

Figure 1. Time courses of frequency changes (∆F) observed during enzyme adsorption on the amorphous PLLA surface by PHB depolymerase from R. pickettii T1 with different concentrations in the 10 mM phosphate buffer solution at 25 °C. The concentrations of the PHB depolymerase are 1 (A), 2 (B), and 4 µg/mL (C), respectively.

solution at 20 °C for various periods of time (2-120 min), and then the thin film was washed with Milli Q water. Morphologies of PHB depolymerases on amorphous PLLA thin film were observed by dynamic force mode (tapping mode) AFM (SII Nanotechnology Inc., Chiba, Japan, SPI3800/ SPA400) in air (25 °C). A 400 µm long silicon cantilever with a spring constant of 1.7 N/m was used for the AFM observation with a light tapping force (set-point value ) 0.80.9). The scan rate ranged from 0.5 to 1.0 Hz. The scan angle was set to 90°. Simultaneous registration was performed for height and phase images. Results QCM Analysis of PHB Depolymerase Adsorption. QCM was stabilized in 10 mM phosphate buffer solution (pH 7.4) without enzyme. Then, the buffer solution was replaced by enzyme solutions of different concentrations to start the enzymatic reaction. The PHB depolymerases have no ability to hydrolyze molecular chains of PLLA.15 Therefore, only the adsorption process of PHB depolymerase from Ralstonia pickettii T1 was directly observed by QCM. Figure 1 shows the typical frequency changes during adsorption of PHB depolymerase of various concentrations onto the amorphous PLLA thin film. Within the first 30 min after the replacement of the buffer solution with the enzyme solution, a sharp decrease in the QCM frequency was detected. This negative frequency shift corresponded to the weight increase of the PLLA thin films due to the rapid adsorption of PHB depolymerases onto the PLLA surface. At the later stage (after 30 min), the QCM frequency gently decreased to be almost constant after 60 min. In our previous paper,17 adsorption of mutant PHB depolymerase, which lost the hydrolytic activity of the catalytic domain, was studied on melt-crystallized PHB film by QCM. On replacing the buffer solution with the enzyme solution, fluctuation of QCM signals occurred for a few minutes so that initial adsorption of the mutated enzyme

Kikkawa et al.

Figure 2. Time-dependent frequency changes (∆F) during enzyme adsorption on the amorphous PLLA surface by PHB depolymerase from R. pickettii T1 in the 10 mM phosphate buffer solution at 25 °C, in which the enzyme solution in the QCM cell (4 µg/mL) was replaced with 10 mM phosphate buffer solution in the middle of the reaction. (A) The enzyme solution (4 µg/mL) was changed to the buffer solution during the reaction. (B) The reaction was monitored without replacing the enzyme solution.

could not be followed. However, in the present study, the suitable conditions and device for the QCM analysis provided the observation of the transient response of QCM signals to enzyme adsorption, as shown in Figure 1. The rate of initial adsorption increased with enzyme concentration. The total amount of the adsorbed enzyme was calculated from the frequency shift at an equilibrium state of the QCM signal using the Saurbrey equation.29 In the case of the enzyme at a concentration of 1 µg/mL, the adsorbed enzymes on the PLLA surface at equilibrium was calculated to be ca. 25 ng, whereas in the case of enzyme concentrations of 2 and 4 µg/mL, the total amount of adsorbed enzyme was ca. 40 ng on the electrode (0.189 cm2). To characterize the binding affinity of PHB depolymerase on the amorphous PLLA surface, the enzyme concentrations were changed in the middle of the enzymatic reaction by replacing the enzyme solution in the QCM cell. Figure 2 (line A) shows the time-course of frequency changes during enzymatic treatment of the amorphous PLLA surface, in which the enzyme solution of 4 µg/mL was replaced by a phosphate buffer solution without enzyme. This caused the frequency of the QCM signal to become constant. On the contrary, the frequency decrease was gently continued in the 4 µg/mL enzyme solution, as shown in Figure 2 (line B). AFM Observation of PHB Depolymerase Molecules. An amorphous PLLA film was prepared on a silicon wafer through thermal treatment at 220 °C and subsequent quenching to 0 °C. Morphologies of amorphous PLLA thin films and PHB depolymerase molecules were directly observed by AFM. Figure 3, parts A and B, shows the AFM height and phase images of the amorphous PLLA thin film, respectively. Smooth surface morphology was observed throughout the thin film. The average surface roughness (Ra) in 500 × 500 nm area was ca. 0.2 nm. Figure 3, parts C and D, shows the AFM height and phase images of PHB depolymerase molecules (1 µg/mL) dispersed on the PLLA

Behavior of Poly(3-hydroxybutyrate) Depolymerase

Figure 3. AFM images of PLLA thin film (A and B) and PHB depolymerase binding on the film surface (C and D). The frames A and C show the AFM height images, whereas the frames B and D are AFM phase images. The graph under frame C shows the crosssectional data over the distance represented by the white line in the image. Arrows in the line profile data indicate the small ridge around the adsorbed enzyme. The color contrast for the height images of frames A and C represents a total range of 10 nm, whereas that of frames B and D was -74 to -85° and -54 to -58°, respectively.

film for 2 min, respectively. In the height image of Figure 3C, white particles of 2.9 ( 0.6 nm in height and 22 ( 3 nm in width were observed on the PLLA surface. The dimensions of the particles were almost identical with those of the PHB depolymerases observed in our previous paper.19 In addition, it is obvious that phase contrast of the thin film before and after enzymatic treatment is different, and that the spherical object shows a different phase response against the PLLA surface. These results suggest that the particles on the PLLA film are PHB depolymerase molecules. Time-Dependent AFM Observation of Enzyme Adsorption. To follow the adsorption behavior of PHB depolymerase in the course of enzymatic treatment, timedependent AFM observations of enzyme molecules were performed for different concentrations of enzyme solutions. Figure 4 shows the time-dependent morphological changes on the surface of PLLA films when the enzyme solutions of 1, 2, and 4 µg/mL were dropped on the polyester surface. For the quantitative analysis of the AFM image, surface coverage of the polyester surface was estimated from Figure 4. Figure 5 shows the time-dependent surface coverage by PHB depolymerase molecules with different enzyme concentrations. As shown in Figures 4 and 5, fast adsorption of the enzymes occurred at an initial stage of adsorption, and the surface of the PLLA film was gradually covered with PHB depolymerase molecules as the enzyme adsorption proceeded. Then, the PHB depolymerase molecules seem to attach with each other to show the two-dimensional network

Biomacromolecules, Vol. 6, No. 4, 2005 2087

morphology. The density of the enzyme molecules increased with an increase in the enzyme concentration. For example, at an adsorption time of 120 min, the surface coverage in the enzyme concentration of 1 µg/mL was 60-70%, whereas in the case of 2 µg/mL, ca. 90% of the PLLA surface was covered with enzyme molecules. From the AFM height image in Figure 4, parts A and B, the PHB depolymerase molecules seemed to adsorb on the PLLA film as a monolayer. However, in Figure 4C at a concentration of 4 µg/mL, enzyme adsorption of the first layer was almost completed even at the adsorption time of 30-60 min, i.e., more than 90% of the PLLA surface was already covered with the enzymes. Figure 6 shows the cross-sectional data of Figure 4c-5 at 120 min. The second layer of the enzyme molecules was formed by the adsorption of PHB depolymerase molecules onto the adlayer of the enzymes. Similar observation can be also found in the AFM image of Figure 4B at an adsorption time of 120 min. Height Measurement of PHB Depolymerase by AFM. For the detailed study on the characteristic interaction of PHB depolymerase and the PLLA surface during adsorption, the height of the adsorbed enzymes was also calculated from the AFM images in Figure 4. Table 1 summarizes the height of PHB depolymerase on the PLLA surface measured in the course of adsorption. Interestingly, the height of the adsorbed enzyme has changed with time. In the case of enzyme concentrations of 1 and 2 µg/mL, the height of enzyme molecules which bound onto the PLLA surface at an early stage of 2-5 min was ca. 3-3.5 nm, whereas in the later stages of 30-120 min adsorption, the height of the enzyme was increased to be ca. 4-5 nm. For the enzyme adsorption in the 4.0 µg/mL solution, ca. 5 nm height enzyme molecules were found even at the early stage of reaction (2-5 min).

Discussion Multi-Step Adsorption of PHB Depolymerase. Binding of PHB depolymerase on the PLLA film was studied by QCM and AFM in terms of enzyme concentration and adsorption time. QCM analysis was carried out to follow the adsorption of PHB depolymerases on an amorphous PLLA surface. As shown in Figure 1, rapid negative frequency changes due to the enzyme adsorption were detected at an initial stage of enzymatic treatment until 30 min. The rapid frequency decrease was followed by a gradual negative frequency shift after 30 min. To determine the origin of the gentle frequency shift, the enzyme solution was replaced by buffer solution without enzyme in the middle of the enzymatic reaction, as shown in Figure 2 (line A). On replacing the enzyme solution with buffer, the decrease in the QCM signal was terminated, and the frequency showed a constant value. This result indicates that PHB depolymerase remained firmly bound to the PLLA surface. Therefore, the slow negative frequency changes after 30 min in line B of Figure 2 originates from further enzyme adsorption. In other words, the PHB depolymerase binds to the amorphous PLLA surface irreversibly via multistep adsorption, i.e., fast and slow adsorption regimes. These primary and subsidiary

2088

Biomacromolecules, Vol. 6, No. 4, 2005

Kikkawa et al.

Figure 4. Time-dependent AFM height images of PHB depolymerase from R. pickettii T1 adsorbed on PLLA amorphous thin film at 25 °C with various enzyme concentrations: (A) 1, (B) 2, and (C) 4 µg/mL. The vertical scale in the height images is 10 nm (dark to bright).

Figure 5. Surface coverage of the PLLA surface by PHB depolymerase measured from the AFM images in Figure 4: ([) 1, (b) 2, and (2) 4 µg/mL.

adsorption processes of PHB depolymerase can be supported by the AFM results in Figures 4 and 5 in terms of surface coverage on the PLLA film during adsorption of PHB depolymerase. At an initial stage of adsorption, the surface of the PLLA film was gradually covered with PHB depolymerase molecules as a monolayer. After the PLLA surface was completely covered with the monolayer, the second layer of the enzyme molecules started to form by additional

Figure 6. AFM height image of PHB depolymerase (4 µg/mL) adsorbed on PLLA surface for 120 min, and cross-sectional data at the white line region. This AFM image corresponds to the Figure 4c-5.

adsorption of PHB depolymerase molecules onto the adlayer of the enzymes. The multistep adsorption of PHB depolymerase depended on the enzyme concentration in the solution. QCM revealed

Biomacromolecules, Vol. 6, No. 4, 2005 2089

Behavior of Poly(3-hydroxybutyrate) Depolymerase Table 1. Height of PHB Depolymerase Molecules Adsorbed in the First Layer on Amorphous PLLA Surface Measured from AFM Images in Figure 4a height (nm)

conc. of enzyme (µg/mL)

2

5

30

60

120

0.5 1.0 2.0 4.0

2.4 ( 0.4 2.9 ( 0.6 3.4 ( 0.7 4.7 ( 0.5

3.3 ( 0.6 3.4 ( 0.7 3.5 ( 0.6 4.7 ( 0.6

3.2 ( 0.6 4.2 ( 0.6 4.6 ( 0.6 n.d.

3.2 ( 0.5 4.6 ( 0.6 4.9 ( 0.5 n.d.

3.2 ( 0.6 4.8 ( 0.5 n.d. n.d.

reaction time (min)

a n.d.: not detectable due to the high surface coverage by enzyme molecules.

that the total amount of bound enzyme on the PLLA surface and the initial rate of adsorption increased by increasing the enzyme concentration, as shown in Figure 1. Surface coverage in Figure 5 measured from time-dependent AFM images in Figure 4 also depended on the enzyme concentration. These results indicate that the enzyme concentration affects the total amount of adsorbed enzyme at an equilibrium state as well as initial adsorption rate. When the PHB film was used as an adsorption support for PHB depolymerase without hydrolytic activity, the amount of adsorption increased as the enzyme concentration increased.17 The present results on the PLLA film are in good agreement with those on the PHB film. Effects of PHB Depolymerase Adsorption on the Surface Properties of PLLA. The individual molecules of PHB depolymerase dispersed on the PLLA surface were observed by using AFM. The phase image in Figure 3D indicated that the PHB depolymerase had different phase signal against the PLLA surface. It is important to note that the phase signals around the PHB depolymerase molecules are also different from those of the enzyme and the PLLA surface. In a previous paper,19 we performed the direct observation of PHB depolymerase molecules from Pseudomonas stutzeri YM1006 and R. pickettii T1 (0.3 µg/mL) on PLLA film. Both enzymes strongly adsorbed onto the polyester surface, and a strong chemical interaction existed between the binding domain of PHB depolymerase and the ester bond of PLLA chains, resulting in movement of some polyester chains at the adsorption area to form the small ridges around the enzyme molecules. In addition, the small ridge showed a different phase response in comparison with the enzyme molecules and the PLLA surface. In the present study, small ridges were observed around the adsorbed enzymes as shown in Figure 3, parts C and D. The boundary of the ridges was blurry because of the relatively large diameter of the cantilever tip, which makes it difficult to detect the interface between the enzyme molecules and the small ridge. However, the region around the PHB depolymerase evidently showed different phase signals against the enzyme molecule and PLLA surface. This result indicates that the nature of the surrounding region of the enzyme is changed due to the binding interaction of PHB depolymerase to the PLLA chains. Morphological Changes of PHB Depolymerase on Amorphous PLLA Films. Quantitative AFM analysis revealed the detailed characteristics of PHB depolymerase

adsorption on the PLLA film. At the enzyme concentration of 0.5 µg/mL, the enzyme height was almost constant, whereas at 1-4 µg/mL, the height of PHB depolymerase on the surface of PLLA changed in the course of enzyme adsorption. This phenomenon may be related to the surface density of the enzyme adsorbed on the PLLA surface. Generally, proteins suffer multiple interactions with not only the adsorbed surfaces but also the neighboring proteins. In the case of the low concentration of enzyme solution or the early stage of enzyme adsorption, enzymes are randomly distributed on the PLLA surface, and individual enzymes feel less interaction with each other. On the contrary, when the PLLA surface was crowded with the bound enzymes, they are deformed and squeezed, resulting in the height increase of the enzyme. Thus, the deformation of PHB depolymerase molecules resulted in the increase in the height at a high density of the bound enzyme. In Figure 4, a two-dimensional network morphology was gradually established as the enzyme adsorption proceeded on the surface of PLLA. Many reports30-33 have suggested that proteins undergo conformational changes on hydrophobic supports and that hydrophobic residues in the proteins are allowed to contact with each other. Ta et al.34 studied the effects of adsorption supports on adsorption of bovine fibrinogen. They used highly oriented pyrolytic graphite (HOPG) and mica as adsorption supports. It was found that fibrinogen formed the branched network structure on HOPG, whereas on mica, it showed homogeneous adsorption as a cluster, suggesting the existence of strong interactions among proteins on HOPG. Taking into account these reports, a plausible mechanism can be proposed for the network formation observed in Figure 4. When the PLLA surface was covered with the enzymes at a high density, the distance between the enzyme molecules becomes closer. As a result, hydrophobic residues of the enzymes are allowed to interact with each other in addition to the binding interaction between the binding domain of the enzyme and the polyester surface. Thus, the two-dimensional network morphology is formed on the PLLA surface. Multilayer Adsorption of PHB Depolymerase. Timedependent QCM and AFM analyses have revealed that PHB depolymerase adsorbs on the amorphous PLLA surface by two steps. The first step was fast adsorption of the enzyme on the film surface to form a monolayer. Such a rapid adsorption was followed by a slow long-term adsorption process, in which the enzymes successively adsorbed as a second layer. Direct observation of the individual molecules of PHB depolymerase by AFM indicated the morphological changes of the enzyme adsorbed on the amorphous PLLA surface and network formation among the adsorbed enzymes. Kim et al.35 performed the direct AFM imaging of lysozyme adsorption on mica and found that two-stage adsorption occurs as fast monolayer formation and slow additional adsorption on the adlayer. They suggested that conformational changes of the proteins in the first layer caused exposure of the hydrophobic residue to interact with the proteins in solution. The present study suggests that adsorption of PHB depolymerase on the amorphous PLLA surface occurs by at

2090

Biomacromolecules, Vol. 6, No. 4, 2005

least two steps. First, enzyme molecules adsorb on the PLLA surface through chemical interactions between the binding domain of the enzyme and the ester bonds of PLLA chains, resulting in a gradual formation of the monolayer of PHB depolymerase molecules on the PLLA surface. This process is accompanied by the conformational change of the PHB depolymerase to form the network morphology due to the hydrophobic interaction between the enzymes. In the second step, after the PLLA surface is almost covered with the enzyme molecules, the hydrophobic interaction becomes dominant at an interface between the enzymes in the adlayer and those in the solution. Then, additional PHB depolymerase molecules stack on the adlayer as a second layer. The present results observed for amorphous PLLA film partly disagreed with our previous results for melt-crystallized PHB films, in which adsorption of PHB depolymerase occurred within 5 min to stack as a monolayer.17 The disagreement may be attributable to the different characters of the polyester surface, such as surface roughness, crystallinity, and type of polyester, etc. In the case of the meltcrystallized PHB film, the hydrolysis reaction by PHB depolymerase starts immediately after the adsorption of enzymes. As a result, the monolayer surface is in a mobile state, and it is difficult for additional enzymes to adsorb on an adlayer. Thus, the surface characters of the substrate would strongly affect adsorption behavior of PHB depolymerase because the adsorption function for PHB depolymerase is dependent on the crystallinity and/or the type of polyesters. Conclusions Adsorption behavior of PHB depolymerase from R. pickettii T1 was studied on the surface of amorphous PLLA thin films by QCM and AFM. The combination of the QCM and AFM results indicated that the binding process of PHB depolymerase can be divided into two regimes, i.e., rapid adsorption onto the polyester surface and slow coverage of the surface with morphological changes of the enzyme. From the QCM analysis, it was found that both the initial adsorption rate and the equilibrium amount of PHB depolymerase on the PLLA surface depended on the enzyme concentration in the solution. AFM observation revealed that binding of PHB depolymerase molecules causes a change in the surface character of the PLLA film around the adsorption area. In addition, morphological changes of the enzyme were detected as a height increase of PHB depolymerase molecules when the enzymes are adsorbed on the PLLA surface at a high density. Then, the enzymes followed to stack as a second layer through the hydrophobic interaction with the enzyme molecules in the first layer. These results indicate that surface properties under the binding enzyme strongly affects the adsorption behavior of PHB depolymerase. Acknowledgment. This work has been supported by the grants of Ecomolecular Science Research from RIKEN

Kikkawa et al.

Institute, SORST (Solution Oriented Research for Science and Technology) from Japan Science and Technology Agency (JST), and a Grant-in-Aid for Scientific research (No. 16550139) from the Ministry of Education, Culture, Sports, Science and Technology, Japan (to K.Y.). References and Notes (1) Holmes, P. A. In DeVelopments in Crystalline Polymers-2; Bassett, D. C., Ed.; Elsevier: London. 1988; p 1. (2) Anderson, A. J.; Dawes, E. A. Microbiol. ReV. 1990, 54, 450. (3) Doi, Y. Microbial Polyesters; VCH Publishers: New York, 1990. (4) Sudesh, K.; Abe, H.; Doi, Y. Prog. Polym. Sci. 2000, 25, 1503. (5) Lenz, R. W.; Marchessault, R. H. Biomacromolecules 2005, 6, 1. (6) Tanio, T.; Fukui, T.; Shirakura, Y.; Saito, T.; Tomita, K.; Kaiho, T.; Masamune, S. Eur. J. Biochem. 1982, 124, 71. (7) Mukai, K.; Yamada, K.; Doi, Y. Polym. Degrad. Stab. 1993, 41, 85. (8) Mergaert, J.; Webb, A.; Anderson, C.; Wouters, A.; Swings, J. Appl. EnViron. Microbiol. 1993, 59, 3233. (9) Jendrossek, D.; Schirmer, A.; Schlegel, H. G. Appl. Microbiol. Biotechnol. 1996, 46, 451. (10) Jendrossek, D. Polym. Degrad. Stab. 1998, 59, 317. (11) Jendrossek, D. In Biopolymers-3b; Steinbu¨chel, A., Doi, Y., Eds.; Wiley-VCH: Weinheim, Germany, 2001; p 41. (12) Shinomiya, M.; Iwata, T.; Kasuya, K.; Doi, Y. FEMS Microbiol. Lett. 1997, 154, 89. (13) Kasuya, K.; Inoue, Y.; Tanaka, T.; Akehata, T.; Iwata, T.; Fukui, T.; Doi, Y. Appl. EnViron. Microbiol. 1997, 63, 4844. (14) Iwata, T.; Doi, Y.; Kasuya, K.; Inoue, Y. Macromolecules 1997, 30, 833. (15) Iwata, T.; Doi, Y.; Tanaka, T.; Akehata, T.; Shiromo, M.; Teramachi, S. Macromolecules 1997, 30, 5290. (16) Yamashita, K.; Aoyagi, Y.; Abe, H.; Doi, Y. Biomacromolecules 2001, 2, 25. (17) Yamashita, K.; Funato, T.; Suzuki, Y.; Teramachi, S.; Doi, Y. Macromol. Biosci. 2003, 3, 694. (18) Murase, T.; Suzuki, Y.; Doi, Y.; Iwata, T. Biomacromolecules 2002, 3, 312. (19) Kikkawa, Y.; Fujita, M.; Hiraishi, T.; Yoshimoto, M.; Doi, Y. Biomacromolecules 2004, 5, 1642. (20) Janshoff, A.; Galla, H.-J.; Steinem, C. Angew. Chem., Int. Ed. Engl. 2000, 39, 4004. (21) Marx, K. A. Biomacromolecules 2003, 4, 1009. (22) Yamashita, K.; Kikkawa, Y.; Kurokawa, K.; Doi, Y. Biomacromolecules in press. (23) Shakesheff, K. M.; Davies, D. C.; Roberts, C. J.; Tendler, S. J. B.; Shard, A. G.; Domb, A. Langmuir 1994, 10, 4417. (24) Shakesheff, K. M.; Davies, D. C.; Domb, A.; Jackson, D. E.; Roberts, C. J.; Tendler, S. J. B.; Williams, P. M. Macromolecules 1995, 28, 1108. (25) Davies, D. C.; Shakesheff, K. M.; Shard, A. G.; Domb, A.; Roberts, C. J.; Tendler, S. J. B.; Williams, P. M. Macromolecules 1996, 26, 2205. (26) Kikkawa, Y.; Murase, T.; Abe, H.; Iwata, T.; Doi, Y. Macromol. Biosci. 2002, 2, 189. (27) Kikkawa, Y.; Hirota, T.; Numata, K.; Tsuge, T.; Iwata, T.; Abe, H.; Doi, Y. Macromol. Biosci. 2004, 4, 276. (28) Numata, K.; Hirota, T.; Kikkawa, Y.; Tsuge, T.; Iwata, T.; Abe, H.; Doi, Y. Biomacromolecules 2004, 5, 2186. (29) Sauerbrey, Z. Z. Phys. 1959, 155, 206. (30) Soderquist, M. E.; Wlton, G. J. Colloid Interface Sci. 1980, 75, 386. (31) Kondo, A.; Mihara, J. J. Colloid Interface Sci. 1996, 177, 214. (32) Buijs, J.; Hlady, V. J. Colloid Interface Sci. 1997, 190, 171. (33) Maste, M. C. L.; Norde, W.; Visser, A. J. W. G. J. Colloid Interface Sci. 1997, 196, 224. (34) Ta, T. C.; Sykes, M. T.; McDermott, M. T. Langmuir 1998, 14, 2435. (35) Kim, D. T.; Blanch, H. W.; Radke, C. J. Langmuir 2002, 18, 5841.

BM0500751