Atomic Force Microscopic Observation of in Vitro Polymerized Poly[(R

Shun Sato , Yusuke Ono , Yukiko Mochiyama , Easan Sivaniah , Yoshihiro Kikkawa , Kumar Sudesh , Tomohiro Hiraishi , Yoshiharu Doi , Hideki Abe and ...
0 downloads 0 Views 655KB Size
Download PDF
Biomacromolecules 2005, 6, 2671-2677

2671

Atomic Force Microscopic Observation of in Vitro Polymerized Poly[(R)-3-hydroxybutyrate]: Insight into Possible Mechanism of Granule Formation Tomohiro Hiraishi,*,† Yoshihiro Kikkawa,‡,§ Masahiro Fujita,‡ Yahaya Mohd Normi,‡,| Masatoshi Kanesato,§ Takeharu Tsuge,⊥ Kumar Sudesh,| Mizuo Maeda,† and Yoshiharu Doi‡,⊥ Bioengineering and Polymer Chemistry Laboratories, RIKEN Institute, 2-1 Hirosawa, Wako-shi, Saitama 351-0198, Japan, Nanoarchitectonics Research Center, National Institute of Advanced Industrial Science and Technology, Tsukuba Central 4, 1-1-1 Higashi, Tsukuba, Ibaraki 305-8562, Japan, School of Biological Science, Universiti Sains Malaysia, 11800 Minden, Penang, Malaysia, and Department of Innovative and Engineered Materials, Tokyo Institute of Technology, 4259 Nagatsuta, Midori-ku, Yokohama, Kanagawa 226-8502, Japan Received February 1, 2005; Revised Manuscript Received May 3, 2005

Atomic force microscopy (AFM) was used to study the formation and growth of poly[(R)-3-hydroxybutyrate] (PHB) structures formed in the enzymatic polymerization of (R)-3-hydroxybutyryl coenzyme A [(R)-3HBCoA] in vitro. Poly(3-hydroxyalkanoate) (PHA) synthase (PhaCRe) from Ralstonia eutropha, a class I synthase, was purified by one-step purification and then used for in vitro reactions. Before the reaction, PhaCRe molecules were deposited on highly oriented pyrolytic graphite (HOPG) and observed as spherical particles with an average height of 2.7 ( 0.6 nm and apparent width of 24 ( 3 nm. AFM analysis during the initial stage of the reaction, that is, after a small amount of (R)-3-HBCoA had been consumed, showed that the enzyme molecules polymerize (R)-3-HBCoA and form flexible 3HB polymer chains that extend from the enzyme particles, resulting in the formation of an enzyme-nascent PHB conjugate. When a sufficient amount of (R)-3-HBCoA was used as substrate, the reaction rapidly increased after the first minute followed by a slow increase in rate, and substrate was completely consumed after 4 min. After 4 min, spherical granules continued to grow in size to form clusters over 10 µm in width, and in later stages of cluster formation, the cluster developed small projections with a size of approximately 100-250 nm, suggesting qualitative changes of the PHB clusters. Moreover, the high-resolution AFM images suggested that globular structures of approximately 20-30 nm apparent width, which corresponds to the size of PhaCRe, were located on the surface of the small PHB granule particles. Introduction Poly[(R)-3-hydroxyalkanoate]s (PHAs) are biodegradable biopolymers produced by a wide variety of bacteria and accumulated as intracellular inclusions. Therefore, PHAs can be used as environmentally friendly materials in place of conventional petrochemical-based plastics.1 PHA synthases, key enzymes that catalyze PHA biosynthesis from (R)-3-hydroxyacyl-CoA [(R)-3-HACoA] substrates, have been classified into four classes on the basis of their substrate specificity and subunit composition. Class I (e.g., Ralstonia eutropha), class III (e.g., Allochromatium Vinosum) and class IV (e.g., Bacillus megaterium) enzymes preferentially incorporate short-chain substrates (C3-C5 monomers), while class II enzymes (e.g., Pseudomonas * Corresponding author: Phone +81-48(467)9312; fax +81-48(462)4658; e-mail [email protected]. † Bioengineering Laboratory, RIKEN Institute. ‡ Polymer Chemistry Laboratory, RIKEN Institute. § National Institute of Advanced Industrial Science and Technology. | Universiti Sains Malaysia. ⊥ Tokyo Institute of Technology.

oleoVorans)incorporatemedium-chain(C6-C14)monomers.1-7 Class I enzymes consist of a single subunit, while class III and IV enzymes are composed of two different subunits.6-12 PHA inclusions have been extensively studied and it is known that several proteins exist on the surface of the inclusions.1 For R. eutropha, a typical PHA-producing bacterium, these surface proteins include PHA synthase (PhaCRe), phasin (PhaP), repressor protein (PhaR), and PHA depolymerase (PhaZ). The morphologies of PHA granules from several PHA-producing bacteria have been examined mainly by electron microscopy.13,14 PHA granules in native strains have diameters ranging from 200 to 500 nm. Thus, R. eutropha cells contain about 10 PHB (poly[(R)-3hydroxybutyrate]) granules, with a diameter of 500 nm. PHB granules from recombinant bacteria, harboring all or only some of the PHA biosynthesis genes, vary in size, demonstrating that the expression of PHA-related proteins affects PHA inclusion size and perhaps morphology.15 Although electron microscopy is effective in evaluating the morphology of PHA inclusions, sample preparation may cause artifacts that could alter the PHA inclusions before

10.1021/bm0500749 CCC: $30.25 © 2005 American Chemical Society Published on Web 07/21/2005

2672

Biomacromolecules, Vol. 6, No. 5, 2005

observation. By contrast, atomic force microscopy (AFM) allows direct imaging of the surfaces of bio-based materials and therefore is an attractive method for this purpose.16 In our previous studies, the surface morphologies of PHA granules from Pseudomonas sp. 61-3,17 Comamonas acidoVorans,18 and R. eutropha19 were investigated by AFM. Studies on Pseudomonas sp. 61-3, which simultaneously produces PHB homopolymer and a copolymer with mediumchain 3-hydroxyalkanoate (3HAMCL) units of 6-12 carbon atoms, revealed the differences between the PHB homopolymer and PHA copolymer granules. Equally interesting, for granules isolated from C. acidoVorans and R. eutropha, the AFM experiments supported the existence of an envelope and showed 30-nm globular structures around the inclusions, reminiscent of the boundary layer and network of organized structures seen on the surface of PHA granules from R. eutropha reported by Dennis et al.20 In vitro PHA polymerization of (R)-3-HACoA has been investigated with PHA synthases (PhaC) from several bacteria and especially from R. eutropha, the enzyme often being produced in recombinant E. coli. The focus of these studies has been on the kinetics and the mechanism of PHB granules formation in vitro.21-28 As a recent example, Nobes et al.22 visualized the in vitro growth of PHB granules directly by transmission electron microscopy (TEM) and cryo-TEM. On the basis of their results and the kinetics of in vitro polymerization, they also simulated PHB granule formation in vitro.29,30 In the present study, we report the direct observations of PHB structures formed in vitro by AFM and suggest that the nascent PHB chain is elongated from a PhaCRe molecule to form a synthase-PHB conjugate. Several of these nascent PHB conjugates coalesce, which leads to formation of PHB clusters and morphological changes of the resulting granule surface. AFM high-resolution phase imaging of PHB granules allows observation of small particles 20-30 nm in apparent size, which is the identical size of PhaCRe that densely cover the surface of PHB granules. These AFM observations not only provide new information regarding surface morphology of PHB granules formed in vitro but also directly support a plausible mechanism of in vitro PHB granule formation proposed in previous studies.22,30,31 Materials and Methods Chemicals. (R)-3-Hydroxybutyryl coenzyme A [(R)-3HBCoA] was synthesized as described by Schubert et al.32 Other chemicals were purchased from Kanto Chemicals (Tokyo, Japan) or Wako Chemicals (Osaka, Japan) and used without further purification. Bacterial Strains and Growth Conditions. Escherichia coli JM109 and BL21(DE3) were used as cloning and expression hosts, respectively. E. coli was usually grown in Luria-Bertani (LB) broth (1% bacto-tryptone, 0.5% yeast extract, and 0.5% NaCl, pH 7.0) containing 50 µg/mL ampicillin. Plasmid vector pET-15b was used for expression of recombinant PhaCRe. Preparation of the plasmid DNA from E. coli and transformation of E. coli were carried out by standard procedures.33 All restriction enzymes and related

Hiraishi et al.

reagents for DNA manipulation were commercially available and used according to the suppliers’ recommendations. Construction, Expression, and Purification of PhaCRe. Recombinant PhaCRe was prepared as described previously.28 A DNA fragment encoding the whole PhaCRe gene was amplified by PCR from plasmid pGEM′CABRe as a template. The PCR product was digested by NdeI and BamHI and introduced into pET-15b, resulting in the formation of plasmid pET-15b::phaCRe. E. coli BL21(DE3) cells harboring pET-15b::phaCRe were grown at 37 °C to an OD600 ) 0.6-0.8. After isopropyl β-D(-)-thiogalactopyranoside (IPTG) addition to a final concentration at 0.1 mM, the cells were cultured at 30 °C for 4 h and harvested by centrifugation. The collected cells were disrupted with a French press. The suspension was centrifuged and the resultant supernatant was subjected to PhaCRe purification. The supernatant was applied to a Ni-NTA Superflow column, and PhaCRe was eluted with a stepwise gradient from 19.6 to 154 mM imidazole. The purity of the eluted enzyme solution was determined by sodium dodecyl sulfate-polyacrylamide gel electrophoretic (SDS-PAGE) analysis with a precision prestained molecular marker (Japan Bio-Rad Laboratories).34 Protein was stained with Coomassie brilliant blue R250 (Kanto, Tokyo). The protein concentration was determined by the Bradford method, and bovine serum albumin (BSA) was used as a standard.35 In Vitro Enzymatic Synthesis of PHB. PHB was initially synthesized in vitro in the presence of 100 µM (R)-3-HBCoA by a modified method based on the assay developed by Song et al.24 The polymerization was carried out at 30 °C in 2 mL of 100 mM Tris-HCl buffer solution at pH 8.0 containing 10 µg of BSA, 0.25% 6-O-(N-heptylcarbamoyl)methyl R-Dglucopyranoside (Hecameg), and suitable amounts of (R)3-HBCoA. The reaction was started by addition of the enzyme solution to a final concentration of 100 nM. The polymerization reaction was monitored at 236 nm to follow the consumption of (R)-3-HBCoA, and a 5 µL aliquot of the reaction mixture after termination of the reaction was diluted 50-fold to be used for the AFM observations of the nascent PHB chain. When a higher concentration of (R)-3HBCoA (3.0 mM) was used, aliquots were periodically taken and used directly for AFM observation or polymer formation assays. PhaCRe activity at a (R)-3-HBCoA concentration of 100 µM was assayed spectrophotometrically by measuring the hydrolysis of the thioester bond of the substrate according to the method of Fukui et al.36 Alternatively, at 3.0 mM (R)3-HBCoA, the reaction progress was followed by the 5,5dithiobis(2-nitrobenzoic acid) (DTNB) assay. A 5 µL aliquot of the reaction mixture was added to 20 µL of 5% trichloroacetic acid (TCA) solution. To this solution was added 1 mL of a 1 mM DTNB solution. The absorbance at 412 nm was recorded to determine the conversion of (R)3-HBCoA, by use of a molar coefficient at 412 nm of 13 700. PHB structure formation during the reaction was followed semiquantitatively by measuring the turbidity at 600 nm. AFM Observation of PHB Structures. A 5 µL aliquot taken from the reaction mixture was deposited on a freshly

AFM Observation of in Vitro Polymerized PHB

Biomacromolecules, Vol. 6, No. 5, 2005 2673

cleaved highly oriented pyrolytic graphite substrate (HOPG; ZYH quality, Veeco) for 1 min, washed with Milli-Q water, and allowed to dry in air. Morphologies of PHB structures on the HOPG substrate were observed by dynamic force (tapping) mode AFM (Seiko Instruments Inc., SPI3700/ SPA300) in air (25 °C). Long silicon cantilevers (400 µm) with a spring constant of 1.5 N/m were used for the AFM observation with a light tapping force. The set-point value (set-point amplitude/free oscillating amplitude) was 0.8-0.9. The scan rate ranged from 0.5 to 1.0 Hz. The scan angle was set to 90°. Height, amplitude, and phase images were simultaneously obtained. The average radius of the cantilever tips was determined to be 10.4 ( 1.3 nm by use of a silicon grating (TGT01, Micromasch, Estonia) as the tip characterization standard. After PHB clusters were observed on the HOPG substrate by AFM, they were treated with proteinase K overnight in the presence of 2 mM CaCl2, resulting in the removal of PhaCRe from the granules. The treated samples were washed with Milli-Q water, allowed to dry in air, and examined by AFM to follow morphological differences observed in the PHB structures caused by proteinase K treatment. Results and Discussion Initial Elongation of the Nascent PHB Chain. Figure 1A shows that the surface of the HOPG substrate, just after cleaving, was relatively smooth with an average roughness (Ra) ) 0.2-0.3 nm/500-nm square, and therefore it was a suitable substrate for AFM observation of PhaCRe and PHB molecular chains. To observe individual PhaCRe molecules, 5 µL of a dilute enzyme solution (2 nM) was placed on the HOPG substrate, followed by immersion in Milli-Q water to remove the excess enzyme molecules in solution. Figure 1B shows that the PhaCRe molecules are dispersed as spherical particles on the HOPG substrate. The average height and apparent width of the enzymes were measured to be 2.7 ( 0.3 and 24 ( 3 nm, respectively (the estimated radius of the PhaCRe enzyme is 3.6 ( 0.5 nm). To follow the elongation of nascent PHB chains by PhaCRe molecules, PHB was enzymatically synthesized in the presence of 100 µM (R)-3-HBCoA. Figure 2 shows the timedependent (R)-3-HBCoA consumption by PhaCRe at 30 °C. When the enzymatic solution (100 nM) was added to the reaction mixture, the absorbance at 236 nm decreased with a short lag phase, and after 5 min, the reaction was terminated. An aliquot from the resultant reaction mixture was diluted and subjected to AFM observation. Figure 1C,D shows the AFM height images for the PHB molecular chain elongated from PhaCRe. Figure 1C indicates that several particles are dispersed on the HOPG, and the fibrils elongated from the particles. As seen in Figure 1B, the particles recognized in Figure 1C are PhaCRe molecules. Thus, it is suggested that the fibrillar entities from the particles are PHB molecular chains. Figure 1D is the highly enlarged AFM image of the dotted squared region indicated in Figure 1C. This image reveals that the PHB molecular chain grows from the PhaCRe. The height of the fibrillar chain was measured to be 0.3 ( 0.1 nm as shown in Figure 1E. These results

Figure 1. AFM height images of (A) HOPG substrate, (B) PhaCRe dispersed on HOPG substrate, and (C) PhaCRe elongating the PHB molecular chain after 5 min reaction on HOPG substrate. (D) Enlarged AFM image of the dotted squared region in panel C; the area for the line profile data is indicated by the two allows. (E) Line profile data at the region pointed by white arrows in panel D.

Figure 2. Time dependence of (R)-3-HBCoA consumption during in vitro PHB synthesis at 30 °C. The reaction mixture was composed of 10 µg of BSA, 0.25% Hecameg, 100 nM enzyme, and 100 µM monomer in 2 mL of 100 mM Tris-HCl buffer solution at pH 8.0. The reaction was monitored at 236 nm.

suggest that an individual enzyme particle polymerizes (R)3-HBCoA to extend a 3HB polymer chain and forms an enzyme-PHB conjugate in the initial stage of the reaction, that is, after a small amount of (R)-3-HBCoA consumption. Growth of PHB Structures in in Vitro Polymerization. For evaluation of PHB structure formed by in vitro polymerization, the enzymatic synthesis of PHB was carried out in the presence of sufficient amounts of (R)-3-HBCoA (3.0 mM) and the reaction was started by addition of a PhaCRe solution (300 nM) to the reaction mixture at 30 °C. The scattering at 600 nm was measured to follow the PHB

2674

Biomacromolecules, Vol. 6, No. 5, 2005

Figure 3. Scattering at 600 nm of reaction mixture and conversion of (R)-3-HBCoA during in vitro PHB polymerization (A) at 30 °C. The reaction mixture was composed of 10 µg of BSA, 0.25% Hecameg, 300 nM enzyme, and 3 mM monomer in 2 mL of 100 mM Tris-HCl buffer solution at pH 8.0. AFM amplitude images of PHB granules synthesized in vitro at reaction times of (B) 1 min, (C) 3 min, and (D) 10 min.

structure formation semiquantitatively, and aliquots were periodically taken from the reaction mixture to measure the conversion of (R)-3-HBCoA to PHB. Figure 3A shows the scattering of the reaction mixture at 600 nm (solid line) and the conversion of (R)-3-HBCoA during in vitro PHB polymerization (dotted line) against reaction time. After addition of the enzyme solution to the reaction mixture, (R)3-HBCoA was completely consumed within 4 min. The scattering increased slowly during the first minute, followed by a rapid increase, indicating that PHB structures >600 nm in size began to develop after 1 min, at which time 1 mM (R)-3-HBCoA had been consumed by 300 nM PhaCRe enzyme. These results suggest that PhaCRe polymerizes (R)3-HBCoA and promptly generates an enzyme-PHB conjugate and that the progression from micelles to granules begins within 1 min, resulting in PHB structures >600 nm in diameter under these conditions. After 4 min of reaction time, the scattering at 600 nm continued to increase slowly despite the complete consumption of (R)-3-HBCoA, suggesting that

Hiraishi et al.

there is continued coalescence of PHB granules after the elongation of nascent polymer chains. AFM Observation of PHB Structures Formed in Vitro. To observe the growth of PHB structures formed by in vitro polymerization, aliquots from the reaction mixture taken after 1, 3, and 10 min were observed via AFM (Figure 3, panels B, C, and D, respectively). Figure 3B shows that PHB granules of different particle sizes coexist after reaction for 1 min. The smaller granules have a diameter of approximately 50 nm, while larger granules with a diameter of approximately 200-700 nm seemed to result from the coalescence of 2-5 granules with a diameters of approximately 100-250 nm. These results imply that the heterogeneous formation and coalescence of PHB granules proceeds in the initial stage of the reaction. Figure 3C shows an AFM image of PHB structures on HOPG after a 3 min reaction at 30 °C. Fewer but clearly larger PHB structures were seen at 3 min compared to the number seen at 1 min. The larger structures coexisted with small PHB granules, suggesting that the small spherical granules (diameters of approximately 150-400 nm) interacted to form larger clusters during the 1-3 min interval. These AFM observations correlate with an increase of the scattering at 600 nm (Figure 3A). Thus, PHB synthesis appears to proceed concomitantly with coalescence of synthase-PHB conjugates (PHB granules) to form larger clusters of PHB granules. After additional incubation of the reaction mixture (10 min reaction time), when enzymatic synthesis of PHB had already finished, PHB structures on HOPG were also observed by AFM. In this stage, huge PHB clusters were found on the HOPG substrate, with lengths (over 10 µm) in the direction of the long axis that were much longer than those observed at 3 min. Figures 3D and 4A show examples of these huge PHB clusters after 10 min of reaction time. Interestingly, these figures clearly revealed many fine particles extruding from the surface of PHB granule clusters. Nobes et al.22 demonstrated that PHB granules contain water to form relatively large structures in the early stage of the reaction. Subsequent slower growth of the granules is accompanied by coalescence and water ejection with reaction time.21 In the present study, the smooth surface of the PHB granule clusters formed in the early reaction stage has changed to a rougher surface that was composed of clear particles. Therefore, morphological surface changes of PHB granule clusters may be due to qualitative changes of PHB clusters by water ejection. Further AFM observation was performed for the individual small PHB granules exposed on the surface of PHB granule clusters after 10 min of reaction. Figure 4 panels B and C show highly enlarged AFM amplitude and phase images of the PHB granule clusters in Figure 4A. The phase image in Figure 4C revealed that a high density of smaller particles is located on the surface of 100-250 nm granules. Figure 4D shows the cross section of the height and phase shift at the white line region in Figure 4C. Each particle covering the surface of the 100-250 nm granules could be resolved as a periodical phase shift. From these data, the particle size was measured to be ca. 20-30 nm, which corresponds to

AFM Observation of in Vitro Polymerized PHB

Biomacromolecules, Vol. 6, No. 5, 2005 2675

Figure 5. (A) AFM amplitude image of PHB granule cluster synthesized in vitro at reaction times of 10 min after proteinase K treatment. (B, C) Highly enlarged AFM images from panel A: (B) amplitude image and (C) phase image. Arrows indicate (a) the flat region and (b) the globular region, respectively. Figure 4. (A) AFM amplitude image of PHB granule cluster synthesized in vitro at a reaction time of 10 min. (B, C) Highly enlarged AFM images from panel A: (B) amplitude image and (C) phase image. (D) Cross section along the white line indicated in panel C. Dotted and solid lines indicate the height and phase profiles, respectively.

the size of the PhaCRe enzyme molecule shown in Figure 1B. These results indicate that PHB granules are covered with PhaCRe enzyme molecules. If the hydrophilic PhaCRe molecules are located on the surface of each granule, the surface properties of PHB granule clusters should exhibit some hydrophilicity. This assumption is supported by the insolubility of the in vitro synthesized PHB clusters in chloroform (data not shown). Proteinase K is a serine-type endopeptidase that effectively and rather nonspecifically hydrolyzes peptide bonds. Assuming that the protrusions on the surface of PHB granules are PhaCRe, they might be hydrolyzed by proteinase K and removed from the surface. Figure 5A shows the AFM amplitude image of the PHB granules cluster after proteinase K treatment. In this figure, the overall size of PHB cluster remained unchanged, but there were morphological changes resulting in two different structures on the PHB cluster

surface. One is a flat surface, while the other is spherical structure. The size of the spherical structure was approximately 200-300 nm and relatively comparable with that of the PHB protrusions before the proteinase K treatment, as shown in Figure 4A. The highly enlarged amplitude and phase images of Figure 5A are shown in Figure 5 panels B and C, respectively. From Figure 5C, the flat region (arrow a) and the globular region (arrow b) on the PHB granules cluster showed different phase contrast, suggesting that surface properties of the flat region are different from those in the globular region. Such different surface characteristics may be attributed to the crystallinity of granules or remaining PhaCRe, despite the proteinase K treatment. Figure 6 shows high-resolution AFM images of Figure 5B. The AFM amplitude and phase images of Figure 6A,B present information on the regions composed of flat surfaces and spherical structures with diameters of 200-300 nm. To evaluate the surface roughness of the globular structures with bumpy surfaces, higher resolution analysis was carried out (Figure 6C,D). In these figures, the 200-300 nm spherical structures are shown to be assembled from smaller particles of approximately 50 nm in size. The size of the spherical

2676

Biomacromolecules, Vol. 6, No. 5, 2005

Hiraishi et al.

Figure 6. Enlarged AFM images from Figure 5B. Images of the region where the flat surface and the globular structures coexist: (A) amplitude image and (B) phase image. Images at higher magnification of the globular structure region: (C) amplitude image and (D) phase image. Panel E shows a cross section along the white line in panel D. Dotted and soild lines indicate the height and phase profiles, respectively.

Figure 7. Plausible mechanism of PHB granule growth during in vitro polymerization on the basis of AFM observations.

structures is slightly larger after proteinase K treatment as shown in Figures 4B and 6C, and the boundary between the particles is less distinguishable. Figure 6E shows the height and phase contrast profiles along the white line in Figure 6D. There was a periodical phase shift with ca. 50 nm in distance, which was larger than the size of PhaCRe molecules and corresponded with the size of the smaller particles in the initial stage of the reaction. Furthermore, the surface of each small particle (Figure 6C) was smoother after the treatment, suggesting that the PhaCRe molecules were removed from the surface of the PHB granules by proteinase K treatment. Gerngross and Martin21 demonstrated that proteinase K treatment of PHB granules removes the PhaCRe molecules from the granules without decreasing the polymer molecular weight. In this study, the proteinase K-treated PHB granule clusters exhibited smoother surfaces and became more soluble in chloroform. Therefore, removal of PhaCRe molecules may cause morphological changes of the PHB granule surface and enhance the solubility of PHB clusters in chloroform. On the basis of our AFM observations, combined with an earlier model of PHB granule growth,22,30,31 we propose a plausible formation process of PHB structures in vitro (Figure 7). First, PhaCRe reacts with (R)-3-HBCoA to synthesize and extrude a short PHB chain from the enzyme in solution, resulting in the formation of an enzyme-PHB conjugate.

Then, the enzyme further consumes (R)-3-HBCoA to extend the nascent PHB chain, which starts to form a growing granule. At the second stage, small PHB particles of approximately 50 nm in size begin to coalesce with each other. After reaching a PHB particle size of approximately 100 nm in diameter, the particles start to coalesce to form PHB granule clusters, with simultaneous consumption of substrate. The coalescence continues after synthesis of PHB, and the size of the resulting PHB granule cluster exceeds 10 µm. During this stage, there may be water ejection from the PHB granule clusters, resulting in changes of surface morphology, with the protrusions of small 100-250 nm PHB granules that are densely covered with proteins on the cluster surface. As a result, large PHB granule clusters with rough surfaces are produced in vitro compared with in vivo, because PHB granules can easily move and collide with other particles in vitro. Conclusions. This is the first report on the direct AFM observation of PHB structures formed by in vitro polymerization with PHA synthase (PhaCRe) from Ralstonia eutropha. PhaCRe completely converted 100 µM (R)-3-hydroxybutyryl CoA [(R)-3-HBCoA] to PHB within 5 min, and AFM observation of the reaction under these conditions showed fibrillar strands of 0.3 nm thickness that extended from a particle of approximately 25 nm in apparent size, suggesting that a PHB chain grew from PhaCRe in the initial stage of

AFM Observation of in Vitro Polymerized PHB

the polymerization, that is, after consumption of a small amount of (R)-3-HBCoA. AFM observations at higher (R)3-HBCoA concentration revealed that small PHB granules combined into globular structures of approximately 150400 nm in size, followed by coalescence of these globular structures to form large PHB granule clusters. Proteinase K treatment of the PHB granules demonstrated that the granule surface is covered with PhaCRe molecules, which form an insoluble surface against chloroform. These results provide valuable new information and strongly support the mechanism of in vitro PHB polymerization proposed by Nobes et al.22 and Gerngross and Martin.21 Especially, an AFM highresolution phase image of PHB granules shows direct information on surface morphology of the clusters, suggesting that small protrusions of 20-30 nm in apparent size, which correspond the apparent size of PhaCRe, densely cover the surface of PHB granules on the cluster. Acknowledgment. We especially thank Dr. Ken′ichiro Matsumoto for his technical assistance in (R)-3HB-CoA substrate synthesis, and we appreciate the assistance provided by Dr. C. T. Nomura for the English correction of our paper. This research was supported by grants for Ecomolecular Science Research from RIKEN Institute and for SORST (Solution Oriented Research for Science and Technology) from the Japan Science and Technology Agency (JST). Supporting Information Available. Size estimation of PhaCRe, AFM observation of PHB chains, and histograms of PHB particles after different reaction times. This material is available free of charge via the Internet at http:// pubs.acs.org. References and Notes (1) Sudesh, K.; Abe, H.; Doi, Y. Prog. Polym. Sci. 2000, 25, 1503. (2) Liebergesell, M.; Sonomoto, K.; Madkour, M.; Mayer, F.; Steinbu¨chel, A. Eur. J. Biochem. 1994, 226, 71. (3) Haywood, G. W.; Anderson, A. J.; Dawes, E. A. Biotechnol. Lett. 1989, 11, 471. (4) Ren, Q.; de Roo, G.; Kessler, B.; Witholt, B. Biochem. J. 2000, 349, 599. (5) Kraak, M. N.; Smits, T. H. M.; Kessler, B.; Witholt, B. J. Bacteriol. 1997, 179, 4985. (6) Huisman, G. W.; Wonink, E.; Meima, R.; Kazemier, B.; Terpstra, P.; Witholt, B. J. Biol. Chem. 1991, 266, 2191. (7) McCool, G. J.; Cannon, M. C. J. Bacteriol. 2001, 183, 4235.

Biomacromolecules, Vol. 6, No. 5, 2005 2677 (8) Slater, S. C.; Voige, W. H.; Dennis, D. E. J. Bacteriol. 1988, 170, 4431. (9) Liebergesell, M.; Steinbu¨chel, A. Eur. J. Biochem. 1992, 209, 135. (10) Peoples, O. P.; Sinskey, A. J. J. Biol. Chem. 1989, 264, 15298. (11) Gerngross, T. U.; Snell, K. D.; Peoples, O. P.; Sinskey, A. J.; Csuhai, E.; Masamune, S.; Stubbe, J. Biochemistry 1994, 33, 9311. (12) Liebergesell, M.; Steinbu¨chel, A. Appl. Microbiol. Biotechnol. 1993, 38, 493. (13) Ellar, D.; Lundgren, D. G.; Okamura, K.; Marchessault, R. H. J. Mol. Biol. 1968, 35, 489. (14) Fuller, R. C. Int. J. Biol. Macromol. 1992, 25, 21. (15) Wieczorek, R.; Pries, A.; Steinbu¨chel, A.; Mayer, F. J. Bacteriol. 1995, 177, 2425. (16) Kikkawa, Y.; Fujita, M.; Hiraishi, T.; Yoshimoto, M.; Doi, Y. Biomacromolecules 2004, 5, 1642. (17) Matsumoto, K.; Matsusaki, H.; Taguchi, K.; Seki, M.; Doi, Y. Biomacromolecules 2002, 3, 787. (18) Sudesh, K.; Gan, Z.; Maehara, A.; Doi, Y. Polym. Degrad. Stab. 2002, 77, 77. (19) Sudesh, K.; Gan, Z.; Matusmoto, K.; Doi, Y. Ultramicroscopy 2002, 91, 157. (20) Dennis, D.; Liebig, C.; Holley, T.; Thomas, K. S.; Khosla, A.; Wilson, D.; Augustin, B. FEMS Microbiol. Lett. 2003, 226, 113. (21) Gerngross, T. U.; Martin, D. P. Proc. Natl. Acad. Sci. U.S.A. 1995, 92, 6279. (22) Nobes, G. A. R.; Jurasek, L.; Marchessault, R. H.; Martin, D. P.; Mutaux, J. L.; Chanzy, H. Macromol. Rapid Commun. 2000, 21, 77. (23) Zhang, S.; Takagi, Y.; Lenz, R. W.; Goodwin, S. Biomacromolecules 2000, 1, 244. (24) Song, J. J.; Zhang, S.; Lenz, R. W.; Goodwin, S. Biomacromolecules 2000, 1, 433. (25) Zhang, S.; Kolvek, S.; Lenz, R. W.; Goodwin, S. Biomacromolecules 2003, 4, 504. (26) Jossek, R.; Reichelt, R.; Steinbu¨chel, A. Appl. Microbiol. Biotechnol. 1998, 49, 258. (27) Qi, Q.; Steinbu¨chel, A.; Rehm, B. H. A. Appl. Microbiol. Biotechnol. 2000, 54, 37. (28) Normi, Y. M.; Hiraishi, T.; Taguchi, S.; Abe, H.; Sudesh, K.; Najimudin, N.; Doi, Y. Macromol. Biosci. (in press). (29) Jurasek, L.; Nobes, G. A. R.; Marchessault, R. H. Macromol. Biosci. 2001, 1, 258. (30) Jurasek, L.; Marchessault, R. H. Biomacromolecules 2002, 3, 256. (31) Madison, L. L.; Huisman, G. W. Microbiol. Mol. Biol. ReV. 1999, 63, 21. (32) Schubert, P.; Steinbu¨chel, A.; Schlegel, H. G. J. Bacteriol. 1988,170, 5837. (33) Sambrook, J.; Fritsch, E. F.; Maniatis, T. Molecular Cloning: a Laboratory manual, 2nd ed.; Cold Spring Harbor Laboratory Press: Cold Spring Harbor, NY, 1989. (34) Laemmli, U. K. Nature 1970, 227, 680. (35) Bradford, M. M. Anal. Biochem. 1976, 72, 248. (36) Fukui, T.; Yoshimoto, A.; Matsumoto, M.; Hosokawa, S.; Saito, T.; Nishikawa, H.; Tomita, K. Arch. Microbiol. 1976, 110, 149.

BM0500749