ARTICLE pubs.acs.org/Biomac
Ultrafine Cellulose Fibers Produced by Asaia bogorensis, an Acetic Acid Bacterium Akio Kumagai,† Masahiro Mizuno,† Naoto Kato,† Kouichi Nozaki,† Eiji Togawa,‡ Shigeru Yamanaka,§ Kazuo Okuda,|| Inder M. Saxena,^ and Yoshihiko Amano*,† †
)
Department of Bioscience and Textile Technology, Interdisciplinary Graduate School of Science and Technology, Shinshu University, Nagano 380-8553, Japan ‡ Forestry and Forest Products Research Institute, Tsukuba 305-8687, Japan § Faculty of Textile Science and Technology, Shinshu University, Ueda 386-8567, Japan Research and Education Faculty, Multidisciplinary Science Cluster, Kuroshio Science Unit, Kochi University, Kochi 780-8520, Japan ^ Section of Molecular Genetics and Microbiology, The University of Texas at Austin, Austin, Texas 78712, United States ABSTRACT: The ability to synthesize cellulose by Asaia bogorensis, a member of the acetic acid bacteria, was studied in two substrains, AJ and JCM. Although both strains have identical 16S rDNA sequence, only the AJ strain formed a solid pellicle at the air liquid interface in static culture medium, and we analyzed this pellicle using a variety of techniques. In the presence of cellulase, glucose and cellobiose were released from the pellicle suggesting that it is made of cellulose. Field emission electron microscopy allowed the visualization of a 3D knitted structure with ultrafine microfibrils (approximately 5 20 nm in width) in cellulose from A. bogorensis compared with the 40 100 nm wide microfibrils observed in cellulose isolated from Gluconacetobacter xylinus, suggesting differences in the mechanism of cellulose biosynthesis or organization of cellulose synthesizing sites in these two related bacterial species. Identifying these differences will lead to a better understanding of cellulose biosynthesis in bacteria.
’ INTRODUCTION Cellulose is the most abundant organic biopolymer in the biosphere. It is an indispensable material in many industries mainly for its chemical and structural properties that allow for a wide range of applications. Cellulose is synthesized by a variety of organisms including plants, bacteria, fungi, and a few animal species.1 However, most cellulose in nature is produced by plants, where it is a component of the primary and secondary cell walls. In addition, a large number of bacterial species have the potential to synthesize cellulose based on the presence of genes for cellulose biosynthesis that have been identified in these bacteria by genome sequencing. Gluconacetobacter xylinus, a member of the acetic acid bacteria (AAB) is the most studied cellulose-producing bacterium and it is a model organism for understanding cellulose biosynthesis. Macroscopic cellulose fibers obtained from both plants and G. xylinus consist of bundles of β-1,4-glucan chains that form a hierarchical structure. Each β-1,4-glucan chain is synthesized by a cellulose synthase (CesA), which is the catalytic subunit in a multisubunit complex that possibly contains other proteins. This multiprotein complex is viewed as a cellulose-synthesizing site, and these sites are organized as terminal complexes that may be linear, as in the case of G. xylinus,2 or form a rosette, as in the case of plants.3,4 The β-1,4-glucan chain extruded from a catalytic subunit associates with other glucan chains to form a subelementary fibril. These subelementary fibrils are bundled to form r 2011 American Chemical Society
microfibrils and finally the cellulose fibers.1 Current understanding is that the arrangement of cellulose synthesizing sites on the cell surface is characteristic of the organism and this arrangement determines the morphology of the cellulose product. Bacterial cellulose (BC) is an extracellular polysaccharide that is implicated in different biological functions depending on the bacteria that produce cellulose.1 Suggested functions include a floating device in G. xylinus,5 involvement in adhesion to plant roots in Agrobacterium tumefaciens,6 and biofilm formation in Escherichia coli.7 The amount of cellulose produced by bacteria, except for G. xylinus, is extremely low, and it has been a challenge to isolate and study the cellulose produced by most bacteria. For instance, in Rhizobium trifolii,8 Salmonella enterica serovar Typhimurium,9 and Chromobacterium violaceum,10 the extracellularly released microfibrils were isolated and identified as cellulose. The ability to analyze cellulose produced by different bacterial species is important because the features of the cellulose structure provide an insight into the mechanism of cellulose biosynthesis. To screen for new cellulose-producing bacteria among AAB, strains belonging to the genus Acetobacter, Gluconacetobacter, Asaia, and Kozakia, were grown in static cultures. Among the Received: April 22, 2011 Revised: May 30, 2011 Published: June 08, 2011 2815
dx.doi.org/10.1021/bm2005615 | Biomacromolecules 2011, 12, 2815–2821
Biomacromolecules
ARTICLE
Figure 1. Morphology of pellicles produced by G. xylinus (a), A. bogorensis AJ strain (b), and A. bogorensis JCM strain (c). Bacteria were cultured under static conditions in SH medium in culture tubes (upper) and Petri dishes (bottom) for 7 days at 25 C.
different strains, it was found that Asaia bogorensis JCM10569 produced a thin film on the surface of the culture medium. A. bogorensis, isolated from the flowers of tropical plants, is a Gram negative, aerobic, and rod-shaped bacterium that is peritrichously flagellated11 and that has some unique characteristics compared with other AAB. For instance, bacteria belonging to the genus Asaia show none or only a scanty production of acetic acid from ethanol because of the lack of ethanol oxidation activities and growth inhibition in the presence of 0.35% acetic acid, despite growing well even at pH 3.0.11,12 So far, there has been no report of cellulose production by A. bogorensis.13 In the present study, we have characterized the biofilm produced by A. bogorensis and demonstrated for the first time that it is composed of cellulose. Furthermore, we have isolated the cellulose produced by this bacterium and characterized it using a variety of techniques.
’ MATERIALS AND METHODS Bacterial Strains, Media, Growth Conditions, and Plasmids. Two substrains of A. bogorensis JCM10569, designated as AJ and JCM, were identified and used in the present study. G. xylinus ATCC53582 was used as a control strain for BC production. A. bogorensis and G. xylinus strains were grown in Schramm Hestrin (SH) medium5 (20 g/L glucose, 5.0 g/L yeast extract, 5.0 g/L peptone, 0.27 g/L Na2HPO4 3 12H2O, 0.115 g/L citric acid) at 25 C. A. bogorensis and G. xylinus were cultured under static conditions to obtain cellulose as a pellicle. E. coli DH5R, purchased from TaKaRa (Tokyo), was grown in LB medium at 37 C on a rotary shaker. Plasmid pBluescript II SK (+) (Stratagene) was used as the cloning vector. 16S rDNA Sequencing. Total genomic DNA from both strains of A. bogorensis was extracted using the ISOPLANT II kit (Nippon Gene) following the manufacturer’s instructions. DNA fragments specific for the 16S rDNA-coding regions of JCM and AJ strains were amplified using the two universal primers 20F (GAGTTTGATCCTGGCTCAG) and 1500R (GTTACCTTGTTACGACTT), designed by Yamada et al.11 and using the isolated genomic DNA as the template. Amplified
16S rDNA fragments were cloned in plasmid pBluescript II SK (+) for custom DNA sequencing at Hokkaido System Science. Isolation and Purification of Cellulose Pellicles. Cellulose pellicles produced by strains of A. bogorensis and G. xylinus were purified as previously reported.14 Pellicles obtained from static cultures were harvested and treated with 2% NaOH at 100 C for 1 h to remove bacterial cells and other ingredients, such as medium components. They were then washed thoroughly several times with deionized water to remove NaOH. The pellicles treated with NaOH were continuously purified by soaking in 1% NaClO solution for 15 h, followed by a complete wash with running water. Finally, the pellicles were suspended in a solution containing 1 mg/mL Proteinase K (TaKaRa), 0.1% SDS, and 0.01 M Tris-HCl buffer (pH 7.5) at 50 C for 24 h and washed several times with deionized water.
Enzymatic Digestion and Thin-Layer Chromatography (TLC). Pellicles treated with NaOH were homogenized and suspended at a concentration of 2 mg/mL in deionized water. The reaction mixture for digestion contained 0.2 mL of the suspended cellulose solution, 0.2 mL of 0.02 M sodium acetate buffer (SAB, pH 5.0), and 10 μL of 5% cellulase preparation (Cellusoft, Novozyme). Incubation was conducted at 30 C for 0, 1, 3, 6, and 12 h. For the analysis of the hydrolysate, 30 μL of the reaction mixture was spotted on a silica gel plate (Silica Gel 60 A, Whatman), dried, and then developed with a solvent system composed of chloroform, methanol, and water (90:65:15 v/v/v). After the solvent had evaporated completely, the spots of sugars on the TLC plate were visualized by spraying with a 30% sulfuric acid solution and heating the plate at 120 C for 15 min.
Attenuated Total Reflectance-Fourier Transform Infrared Spectroscopy (ATR-FTIR). Purified cellulose pellicles were freezedried and used for FT-IR. IR spectra were obtained at ATR mode by using the FT/IR-4200 (JASCO). The spectra were obtained with an accumulation of 100 scans and with a resolution of 4 cm 1 in the range of 4500 to 700 cm 1. CP/MAS 13C NMR Spectroscopy. The purified cellulose samples were freeze-dried and powdered. CP/MAS 13C NMR measurements were performed at 25 C on a Chemagnetics CMX300 spectrometer operating at 75.4 MHz. The samples, placed in a zirconia rotor for CMX300, were spun at 4.5 kHz, and the contact time for the CP process 2816
dx.doi.org/10.1021/bm2005615 |Biomacromolecules 2011, 12, 2815–2821
Biomacromolecules was 1.2 ms throughout this analysis. The spectra were calibrated using adamantane as a standard.
Field Emission Scanning Electron Microscopy (FE-SEM). The morphology of cellulose pellicles produced by A. bogorensis and G. xylinus was observed by field emission scanning electron microscope S-4100 (Hitachi) at high magnification. Cellulose pellicles purified with
Figure 2. TLC analysis of the products released from pellicles of G. xylinus and A. bogorensis AJ strain following treatment with cellulase for varying period of time (0, 1, 3, 6, and 12 h). Lane M shows the standardsG1, glucose; G2, cellobiose; G3, cellotriose; and G4, cellotetraose.
ARTICLE
NaOH were freeze-dried and then sputter-coated with gold using a twin coater JEC-550 (JEOL). The spattering of gold particles was carried out at 1.5 KVA for 2.5 min. Transmission Electron Microscopy (TEM). A. bogorensis cells, treated with 0.2% cellulase preparation (Cellusoft, Novozyme) to remove any attached cellulose, were suspended in 50 mM PBS containing 137 mM NaCl, 8 mM Na2HPO4 3 12H2O, 2.7 mM KCl, and 1.5 mM KH2PO4, and washed with the same buffer twice. A drop of the bacterial suspension in PBS was placed on a Formvar-coated grid (200 mesh, Nisshin EM), followed by a drop of 2% glucose solution. The grid was incubated for 30 min at room temperature to allow for cellulose synthesis. After removing excess solution with filter paper, a drop of solution containing cellobiohydrolase I conjugated with gold (CBH I-gold) was immediately placed on the grid and incubated for 15 min at room temperature. The grid was washed with a few drops of water and next washed with 0.25% Nonidet P-40 (Wako). The sample was negatively stained with 1.5% aqueous uranyl acetate solution for imaging. Negatively stained specimens were examined with JEM-2100 electron microscope (JEOL) at 200 kV. The CBH I-gold complex was prepared as described by Chanzy et al.15 We mixed 5 mL of a colloidal gold suspension (5 nm, Sigma) suspended in dilute acetic acid adjusted to pH 4.5 and 1 mL of 0.2 mg/mL CBH I derived from Trichoderma reesei in 0.05 M SAB with 0.5 mL of 5% NaCl and 0.5 mL of 1% PEG-20000 (JUNSEI). The solution was centrifuged at 4 C, 10 000 rpm for 60 min, and the supernatant was used as CBH I-gold.
Figure 3. FT-IR spectra in the region from 3800 to 700 cm 1 of pellicles produced by A. bogorensis AJ strain (black line) and G. xylinus (gray line) (a). The magnification of the region from 3400 to 3100 cm 1 (b) and 800 to 700 cm 1 (c) are shown with the peak heights indicating cellulose IR and cellulose Iβ and the ratio of cellulose IR and cellulose Iβ. 2817
dx.doi.org/10.1021/bm2005615 |Biomacromolecules 2011, 12, 2815–2821
Biomacromolecules
ARTICLE
’ RESULTS AND DISCUSSION Morphology of Pellicles Produced by Two Substrains of A. bogorensis. In this study, two substrains of A. bogorensis JCM10569,
one continuously subcultured (AJ) in our laboratory and the other (JCM) newly obtained from Riken Bioresource Center (Japan), were used. Interestingly, these two strains produced pellicles with different morphology in SH medium. (See Figure 1.) Both strains, AJ and JCM, produced pellicles on the surface of culture medium like G. xylinus. However, the color and hardness of pellicles were very different for the two A. bogorensis strains. The pellicle formed by the AJ strain was cream-white, whereas that from the JCM strain was pink. Furthermore, the pellicle obtained after 7 days cultivation of the JCM strain in Petri dish containing SH medium was easily broken by gentle agitation. In contrast, the pellicle from G. xylinus was solid and remained intact after agitation. The pellicle formed by the AJ strain was more stable compared with the pellicle produced by JCM strain but not as strong as the pellicle formed by G. xylinus. According to previous reports, colonies of A. bogorensis11,13 were observed to be pink to yellowish-white and production of cellulose was negative. These characteristics were similar to what we observed for the JCM strain. Because the two strains (AJ and JCM) showed different phenotypes, we determined the phylogenetic relationship between them using the 16S rDNA sequence. The 1338 bp of 16S rDNA sequence obtained from the two strains was completely identical, and it was also identical to 16S rDNA sequence registered as A. bogorensis JCM10569 by Yamada et al.11 This result suggests that these strains are phylogenetically identical, although they exhibit different phenotypes in pellicle production. In G. xylinus, it is well known that cellulose-deficient mutants arise spontaneously during continuous subculture.5,16 It is quite likely that a spontaneous mutation may have occurred in the original A. bogorensis JCM10569 during subculture giving rise to the present JCM strain. However, the genetic differences between A. bogorensis AJ and JCM strains remain to be identified at present. Characterization of Pellicle Produced by the AJ Strain of A. bogorensis. After purification using NaOH, NaClO, Proteinase K, and SDS, the pellicles obtained from static cultures of A. bogorensis AJ strain and G. xylinus were digested by cellulase, and the products released were analyzed by TLC. (See Figure 2.) The pellicle from G. xylinus, which is composed of cellulose, was digested to release glucose and cellobiose as final products after 12 h. The digestion products from the pellicle prepared from A. bogorensis AJ strain also showed the same pattern as G. xylinus, and no spots except for glucose and cellobiose were detected. To elucidate further the components of the pellicle from A. bogorensis, it was analyzed using FT-IR. The spectrum of the pellicle obtained from A. bogorensis showed a typical cellulose I pattern. (See Figure 3a.) The absorption spectra obtained from the pellicle from A. bogorensis was very similar to the spectra obtained from the pellicle from G. xylinus that is composed of crystalline cellulose. Both samples elicited signals in the area around 3700 to 3000 cm 1 and 3000 to 2850 cm 1 corresponding to the absorption spectra of the O H and C H bonds, respectively, and both samples also showed a strong absorption spectra around 1800 to 900 cm 1, which is a fingerprint region of typical cellulose I. These results confirm that the pellicle produced by A. bogorensis is composed of cellulose. The CP/MAS 13C NMR spectra of cellulose prepared from A. bogorensis AJ strain and G. xylinus were determined, and these are shown in Figure 4. All chemical shifts detected in the analyzed
Figure 4. CP/MAS 13C NMR spectra of cellulose produced by A. bogorensis (a) and G. xylinus (b).
Figure 5. FE-SEM images of the cellulose pellicle produced by A. bogorensis (a) and G. xylinus (b).
samples were assigned to C1 to C6, as suggested by Atalla and VanderHart.17 Native cellulose (cellulose I) is shown to be a composite of two distinct crystalline phases, namely, cellulose IR and Iβ, and the ratio of IR and Iβ differs in cellulose obtained from different species.18,19 In cellulose produced by G. xylinus, the C1 and C4 resonance lines of the spectra appear to be a triplet that is composed of an enhanced central line and a considerably smaller doublet besides the central line.20 Although the spectral patterns of both samples were almost the same, the signal intensity originating from cellulose IR in the C1 (106 ppm), the C4 (90 89 ppm), and 2818
dx.doi.org/10.1021/bm2005615 |Biomacromolecules 2011, 12, 2815–2821
Biomacromolecules
ARTICLE
Figure 6. TEM images of negatively stained A. bogorensis cells associated with fibers (a,b), and fibers only (c f) after labeling with CBH I-gold. The presence of cellulose microfibrils associated with a cell is observed by CBH I-gold labeling for A. bogorensis AJ strain (a) but not for the A. bogorensis JCM strain (b). CBH I-gold labeling is, however, observed in isolated fibers from both strains (c,d). Multiple flagella are also observed in both strains. Scale bar corresponds to 1 μm (a,b), 500 nm (c,d), and 50 nm (e,f).
the C6 (66 ppm) regions18 decreased slightly in the cellulose prepared from A. bogorensis AJ strain. Cellulose produced by higher plants, such as cotton and ramie, is rich in cellulose Iβ, whereas cellulose from the alga Valonia and G. xylinus is rich in cellulose IR.19 Cellulose produced by A. bogorensis AJ strain is rich in cellulose IR, but the amount of cellulose Iβ is higher in this bacterium than that observed in cellulose from G. xylinus. This is indicated by the lower intensities of C1 and C4 observed in the NMR spectra in cellulose from A. bogorensis as compared with cellulose from G. xylinus. Analysis of the FT-IR spectra also allows determination the ratio of IR and Iβ, bands near 3240 and 750 cm 1 are attributed to the IR phase and 3270 and 710 cm 1 to the Iβ phase (Figure 3b,c),21 and these observations confirm those obtained by NMR. In addition, crystallinity of cellulose produced by A. bogorensis AJ strain is lower than that of cellulose from G. xylinus based on the broad resonance lines for C4 (∼84 ppm) and C6 (∼62 ppm), which are assigned to the noncrystalline component of cellulose.22 These results confirm that the pellicle produced by A. bogorensis AJ strain contains mainly cellulose and that the structure of cellulose produced by this bacterium is slightly different from the cellulose produced by G. xylinus. Microscopic Analyses of Cellulose Produced by A. bogorensis. FE-SEM images of the cellulose product after purification
and freeze-drying are presented in Figure 5. Although cellulose from both A. bogorensis AJ strain and G. xylinus forms a 3D knitted structure, it is noteworthy that cellulose fibrils produced by A. bogorensis are much thinner than the cellulose ribbons from G. xylinus. As a result, the 3D network composed of cellulose fibrils from A. bogorensis is much finer compared with the G. xylinus network. The range of width of cellulose fibrils in A. bogorensis AJ strain and G. xylinus is about 5 20 nm and 40 100 nm, respectively, calculated from 50 individual cellulose fibers observed using FE-SEM. A. bogorensis cells producing cellulose were observed by TEM, as shown in Figure 6, and the cellulose fibers produced by AJ and JCM strains were compared. Many filamentous structures are observed around cells of both the A. bogorensis AJ and JCM strains, complicating the identification of cellulose fibers. A. bogorensis is a peritrichously flagellated bacterium,11 and these filamentous structures are believed to be flagella. To distinguish cellulose fibers in these filamentous structures, they were labeled with CBH I-gold, which binds specifically to crystalline cellulose through a cellulose binding module contained in CBH I. CBH I-gold associated specifically with the finest fibers in the AJ strain (Figure 6a,c). Although no cellulose fibers were observed around the cell in the JCM strain (Figure 6b), fragments of cellulose fiber 2819
dx.doi.org/10.1021/bm2005615 |Biomacromolecules 2011, 12, 2815–2821
Biomacromolecules separate from the bacterial cell were observed (Figure 6d). This observation suggests that the JCM strain also has the ability to produce cellulose like the AJ strain, but the amount of cellulose produced by this strain is much lower than that produced by the AJ strain. Moreover, the width of cellulose fibrils produced by A. bogorensis is
