PEDS Advance Access originally published online on June 6, 2008
Protein Engineering Design and Selection 2008 21(9):561-566; doi:10.1093/protein/gzn033
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Structural simulation and protein engineering to convert an endo-chitosanase to an exo-chitosanase
1Department of Applied Chemistry 2Department of Biological Science and Technology, Institute of Bioinformatics, National Chiao Tung University, 1001 Ta-Hseh Road, Hsinchu, Taiwan 3Department of Agricultural Chemistry, Faculty of Horticulture, Chiba University, Matsudo 271, Japan 4Department of Marine Biotechnology, National Kaohsiung Marine University, 142 Hai-Chuan Road, Nan-Tzu Dist, Kaohsiung, Taiwan
5 To whom correspondence should be addressed. E-mail: ykl{at}cc.nctu.edu.tw (Y.-K. Li)/cycheng{at}mail.nkmu.edu.tw (C.-Y. C)
 |
Abstract |
|---|
 Top Abstract IntroductionMaterials and methodsResults and discussionFundingReferences |
|---|
To obtain an enzyme for the production of chito-disaccharides (GlcN2) by converting endo-chitosanase to exo-chitosanase, we chose an endo-chitosanase from Bacillus circulans MH-K1 (Csn) as the candidate for protein engineering. Using molecular modeling, two peptides with five amino acids (PCLGG) and six amino acids (SRTCKP) were designed and inserted after the positions of D115 and T222 of Csn, respectively. The inserted fragments are expected to form loops that might protrude from opposite walls of the substrate-binding cleft, thus forming a ‘roof’ over the catalytic site that might alter the product specificity. The chimeric chitosanase (Chim-Csn) and wild-type chitosanase (WT-Csn) were both over-expressed in Escherichia coli and purified nearly to homogeneity. The products formed from chitosan were analyzed by ESI-MS (electrospray ionization-mass spectrometry). A mixture of GlcN2, GlcN3 and GlcN4 was obtained with WT-Csn, whereas Chim-Csn formed, with a smaller catalytic rate (3% of WT-Csn activity), GlcN2 as the dominant product. Measurements of viscosity showed that, with similar amounts of enzyme activity, Chim-Csn catalyzed the hydrolysis of chitosan with a smaller rate of viscosity decrease than WT-Csn. The results indicate that, on inserting two surface loops, the endo-type chitosanase was converted into an exo-type chitosanase, which to our knowledge is the first chitosanase that releases GlcN2 from chitosan as the dominant product.
Keywords: chitosan/exo-chitosanase/mass spectrometry/structural modeling
|
Introduction |
|---|
|
TopAbstractIntroductionMaterials and methodsResults and discussionFundingReferences |
|---|
Chitosanases (EC 3.2.1.132 [EC] ), catalyzing the hydrolysis of the β-1,4 glycosidic linkage of chitosan, are classified into glycoside hydrolases (GH) in five main families – GH-5, 8, 46, 75 and 80 (http://afmb.cnrs-mrs.fr/CAZY/), based on the homology of the amino-acid sequences (Henrissat and Bairoch, 1996
). All have been shown to be the endo-type enzyme. In general, glycoside hydrolases can hydrolyze their substrates in an endo-type or exo-type manner: the exo-type glycoside hydrolases hydrolyze their substrates from one end of the substrate in an orderly manner to produce the corresponding monosaccharide and/or disaccharide, whereas the endo-type glycoside hydrolases randomly degrade polysaccharides at any accessible glycosidic bond to give a mixture of oligosaccharides. Many microbial chitosanase have been found to function as the endo-type enzyme, which exhibit powerful catalysis on chitosan to release chito-oligosaccharides with various degrees of polymerization (DP) (Cheng and Li, 2000; Cheng et al., 2006), but few chitosan-degrading enzymes possess the exo-type catalytic feature. Perhaps, at present, only exo-β-D-glucosaminidase (GH-2) (Cote et al., 2006), which is not classified as a member of the chitosanase families listed above, can cleave chitosan in an exo-type manner to generate glucosamine (GlcN). There is, so far, no successful method to produce a pure chito-oligosaccharide, such as GlcN2, directly with an enzyme; laborious separation is commonly employed, but allows an isolation of GlcNn (n = 2–6) in pure form on only a small scale.
Although exo-β-D-glucosaminidase, which is considered an exo-chitosanase, was first discovered and characterized in 1990 (Nanjo et al., 1990), the structure of this enzyme is not yet available, so it is difficult structurally to reconstruct a mutant with a desired function. One can, however, learn from nature about the structural designs of glycoside hydrolases in other GH families. In many glycoside hydrolases, such as GH-6, 8, 9, 17, 18, 28 and 55, the exo- and endo-type enzymes are present in the same family. As enzymes in the same family not only possess similarity of their amino-acid sequence, they are also commonly classified into the same clan, indicating that the endo- and exo-type enzymes might be endowed with similar amino-acid sequences and protein folding but with minor structural variation at the catalytic site. For example, both endo-1, 4-glucanase and cellobiohydrolase belong to GH-6 and are endo- and exo-type glucanases, respectively. Of their amino-acid sequences, one quarter are identical, and their protein foldings are similar with only few and minor differences (Fig. 1A and B). As shown in the exo-type glucanase (Fig. 1B), two loops were observed to protrude from opposite walls of the substrate-binding cleft to form a roof over the catalytic center. This feature probably results in an exo-type catalytic function of this glucanase. Further work has demonstrated the conversion of the exo-1,4-glucanase into an endo-1,4-glucanase on removing the two loops (Meinke et al., 1995).
|
A chitosanase from Bacillus circulans MH-K1 (Csn) (GH-46) hydrolyzed GlcN6 in an endo-splitting manner, and released GlcN2, GlcN3 and GlcN4 in almost equivalent proportions (Fukamizo et al., 2005
). The enzyme, composed of 259 amino acids (Ando et al., 1992), consists of two globular domains that generate the substrate-binding cleft (Fukamizo et al., 2005). Inspection of protein structures of Csn (PDB ID: 1QGI) and the endo-glucanase (PDB ID: 2BOD) in GH-6, Fig. 1C and A, respectively, reveals that both enzymes possess a similar structure that encloses a deep cleft for substrate binding and for catalysis. Our idea to convert an endo-chitosanase to an exo-chitosanase was derived from the above-mentioned information. We designed, and proposed to add, two loops at the appropriate positions of Csn by structure modeling, as shown in Fig. 1D, to form a prospective exo-type chitosanase. |
Materials and methods |
|---|
|
TopAbstractIntroductionMaterials and methodsResults and discussionFundingReferences |
|---|
Chitosan and chito-oligosaccharides
Chitosan with 85% degree of deacetylation was obtained from local suppliers in Taiwan. Chito-oligosaccharides (GlcN2, GlcN3, GlcN4, GlcN5 and GlcN6) were purchased (Seikagaku, Tokyo, Japan). All other chemicals (Sigma-Aldrich) were molecular biology grade.
Bacterial strains and plasmids
The vector pNCMO2 containing the B. circulans MH-K1 chitosanase gene was provided by Professor Ando in Japan. Two primers, N-term primer 5'-GCTCCCATGGCT TTCCATATGGCTTCTCCT-3' containing an NdeI restriction site (in bold) and C-term primer 5'-GACAATGTAATT GTTCCCTAC-3', were used for PCR amplification. The amplified DNA fragment (
1.0 kb) was digested with NdeI and XhoI restriction enzyme, and then ligated into pET22b (+). This new clone, pET-csn, containing the gene for mature Csn with an added methionine at the N-terminus, was transformed into Escherichia coli (E. coli) strain BL21 (DE3) for further protein expression.
The site-directed mutagenesis was performed according to the Quik-changeTM method (Stratagene). The basic procedure involved PCR amplification with pET-csn as the template and two synthetic oligonucleotides containing the desired mutation as primers. Insertion of peptides PCLGG and SRTCKP was achieved by double mutations of pET-csn. These two primer pairs were designed as follows: Csn-115-F: 5'-CA GGC GGC TCT AGA GAT CCG TGC CTG GGC GGT ACC CAT CCC GAT GGC C-3', Csn-115-B: 5'-G GCC ATC GGG ATG GGT ACC GCC CAG GCA CGG ATC TCT AGA GCC GCC TG-3', Csn-222-F': 5'-CTG AAT CAA GGC GCT ACT AGT CGT ACC TGC AAA CCG GGC TCA GAT ACG CTT C-3', and Csn-222-B': 5'-G AAG CGT ATC TGA GCC CGG TTT GCA GGT ACG ACT AGT AGC GCC TTG ATT CAG-3'.
Each PCR product was treated with DpnI restriction enzyme to cleave the methylated template DNA, transformed into E. coli and confirmed on DNA sequencing of the full gene.
Enzyme production and purification
All purification steps were performed at ambient temperature (25°C). A single colony of E. coli strain BL21 (DE3) harboring plasmid pET-csn was inoculated into LB medium (5 ml) containing ampicillin (0.1 mg/ml) and cultivated at 37°C for 12 h. The overnight culture was then transferred into a conical flask (1 l) containing LB medium (500 ml) with ampicillin (0.1 mg/ml) and IPTG (1 mM), at 37°C for 19–20 h. The cell pellet was obtained on centrifugation (6000g, 10 min, 4°C), resuspended in sodium phosphate buffer (10 ml, 20 mM, pH 7.0) and sonicated for 10 min. The lysate was centrifuged at 10 000g for 20 min at 4°C to obtain the cell-free extract. The extract was then applied onto a cation-exchanged chromatographic column (HiTrapTM in 5 ml SP-Sepharose; Amersham Biosciences). The column was pre-equilibrated with phosphate buffer (pH 7.0, 20 mM) and eluted with a linear gradient of NaCl (200–400 mM) in the same buffer at a flow rate 2 ml/min. The fractions with chitosanase activity were pooled and stored at 4°C for further experiments.
Chitosanase activity was analyzed by estimating the amount of the reducing ends of sugars according to the dinitrosalicylic-acid (DNS) method (Miller, 1995; Cheng et al., 2006). This standard assay, namely DNS assay, was prepared on mixing chitosan (0.3 ml, 1%, pH 6.0) and enzyme (0.3 ml) with suitable dilution in phosphate buffer (20 mM, pH 7.0) and incubated for 1 h at 37°C to allow for enzymatic hydrolysis; DNS reagent (0.6 ml) was then added, and the resulting mixture was boiled for 15 min, chilled and centrifuged to isolate the insoluble chitosan. The resulting adducts of reducing sugars were analyzed and measured spectrophotometrically at 540 nm. The absorption coefficient of the resulting adducts was determined to be 788 M–1 cm–1 when D-glucosamine served as the control sample. One unit of chitosanase activity is defined as the amount of enzyme required to release 1 µmol of detectable reducing sugars at 37°C in 1 min (Ando et al., 1992; Cheng and Li, 2000).
Two other methods used to monitor enzyme activity were viscometry (Ohtakara, 1988) and electrospray ionization-mass spectrometric (ESI-MS) analysis. Although the former method is cumbersome and insensitive, it has the potential to detect early variations in the reaction solution to determine the splitting pattern of exo- or endo-type. It was performed with a Brookfield DV-II viscometer (Model LVDV II+, USA). The enzymatic reactions (8 ml) were performed with Chim-Csn (15 µg) [or 0.3 µg wild-type chitosanase (WT-Csn)] at 25°C using chitosan (0.5%, w/v) as substrate and analyzed at appropriate intervals. The ESI-MS method is described later.
Protein concentration and purity determination
Protein concentration was determined according to the method of Bradford (1976) using bovine serum albumin as a standard. The purity of the enzyme was verified with sodium dodecyl sulfate–polyacrylamide gel electrophoresis with a gel system (12.5%, Laemmli, 1970) and enzyme (5 µg). The protein bands were visualized on staining with Coomassie brilliant blue R-250.
Identification of the disulfide bond in Chim-Csn
The disulfide bond was investigated on labeling via chemical treatment and further analyzed colorimetrically or with ESI-MS. Purified protein was treated with or without dithiothreitol (DTT, 10 mM) for 30 min at <25°C. It was then exposed to iodoacetamide (IAA, 50 mM, 50 mM Tris, pH 8.0) in the dark at room temperature for 2 h to alkylate any cysteine that existed or had been released from previous disulfide bonding. After chemical treatment, proteins were analyzed with ESI-MS as described in the following section. Alternatively, the purified protein was treated with 5,5-Dithiobis (2-nitrobenzoic acid, 10–50 mM) (DTNB; Ellman's reagent) (50 mM Tris, pH 8.0), and the absorbance of the sample supernatant was characterized at 412 nm against the control.
Electrospray ionization-mass spectrometry analysis
Mass spectra of enzyme were recorded with a quadrupole time-of-flight mass filter (Micromass, Manchester, UK) scanning with a ratio of mass to charge in the range 500–2500 units (m/z), with a scan 2 s/step and an interscan interval 0.1 s/step. In all ESI-MS experiments, the quadrupole scan mode was used under a capillary needle at 3 kV and the source block temperature 80°C. For the MS measurements, the desalted proteins were obtained from a column (Zip-Tip, Amersham Biosciences) or ultrafiltration spin column (Vivaspin, Vivascience). Protein samples (2–5 µg) were infused in the mass spectrometer with acetonitrile solution (10%, containing 0.1% formic acid as eluent).
A triple-quadrupole MS system (Quattro Micro; Micromass) was used for the analysis of the enzymatic product. Chitosan (0.05%, w/v) was mixed with appropriately diluted chitosanase (9.4 µg Chim-Csn, or 0.3 µg WT-Csn; total volume 5 ml). The enzymatic reaction was performed at 25°C; an aliquot (10 µL) withdrawn from the reaction mixture at a constant interval was injected into the quadrupole MS with an auto-sampler.
A structural modeling was employed to generate a model of chimeric chitosanase (Chim-Csn) on inserting peptides to form two loops protruding from opposite walls of the substrate-binding cleft. Each peptide consists of one cysteine that is near enough in space to form a disulfide bond. The procedures of structure modeling are listed as follows: First, we selected a high-resolution (1.6 Å) crystal structure of chitosanase [PDB accession code 1QGI (Saito et al., 1999)] from B. circulans MH-K1 as the protein template. Secondly, we selected two insertion points located at regions Arg114 to Thr116 and Thr222 to Ser225 because they are at the edge of the active site cleft. The insertion of the peptides into these two regions is expected to maintain the conformation of the active site and the core structure of chitosanase. Thirdly, to estimate the length of the inserted peptide, we calculated the space around both insertion points. Peptides containing 4–8 amino-acid residues were selected to form the protruding loop. Fourthly, we designed the inserting peptides based on the conformation and composition of β-turn. Proline and glycine are commonly observed in the β-turn because of the steric hindrance imposed by the turn geometry (Creighton, 1993). Each peptide required one cysteine for disulfide bond formation. Based on these criteria, peptides composed of a cysteine, several prolines and glycines, and others can be suggested and further examined. Fifthly, the modified protein sequence was modeled for a theoretical structure using SWISS-MODEL (Guex and Peitsch, 1997; Schwede et al., 2000, 2003). Sixth, SYBYL version 6.9 (Tripos, Inc.) was used for energy minimization. Parameters used in SYBYL were set as follows: initial optimization = simplex, termination = gradient, 0.05 kcal/(mol*Å), NB (non-bonded) cutoff = 8.0 Å, dielectric constant = 1.0 and maximum iterations = 500. Finally, we evaluated the structural models based on two criteria: the inserted peptide pair can form thermodynamically stable loops to form a roof over the catalytic site, and two cysteines on the peptide pair are near enough in space to form a disulfide bond.
|
Results and discussion |
|---|
|
TopAbstractIntroductionMaterials and methodsResults and discussionFundingReferences |
|---|
Protein engineering and structure modeling
More than 25 peptide pairs, such as [GLCG, AGLCGA], [PCLGG, ARCTKL], [PCLGG, SRCTKP] and [PCLGG, SRTCKP], were modeled based on the structure-modeling procedure. On observing these structural models, we found that proline directly influenced these two peptides to bend toward each other to form a disulfide bond. Finally, a Chim-Csn with two peptides of PCLGG and SRTCKP was designed and modeled as shown in Fig. 2.
|
The distances between Cβ atoms and S atoms of the two cysteines in the two inserted peptides were 5.06 and 5.99 Å, respectively, in this structure. Because of the dynamic flexibility of these two inserted peptides in an aqueous solution and the rotation of Cβ–SH bonds, these two cysteines can form a disulfide bond between the inserted peptide pair. The root-mean-square deviation (RMSD) between WT-Csn and the Chim-Csn was 0.6 Å (excluding the two inserted peptides). This RMSD value indicates that the conformation of the active site is preserved after insertion of the peptide pair. These results indicate Chim-Csn to be a satisfactory candidate for our work.
Plasmid construction, over-expression and purification of recombinant chitosanase
The Bacillus chitosanase gene was first reported to be expressed in the Bacillus brevis system (Saito et al., 1995), but we successfully employed the E. coli expression system in this work. A DNA fragment containing csn was translocated from pNCMO2 into the pET22b (+) vector for protein expression.
To construct the gene of the mature Csn, we removed the DNA fragment encoding the signal peptide while adding a codon of methionine (ATG) at the N-terminus of the mature protein. The Chim-Csn was constructed by a double mutation to insert peptide 1 and peptide 2 after the positions of D115 and T222 of Csn, respectively. The amino-acid sequences of Csn and the mutant are shown in Fig. 2. Both enzymes were over-expressed in E. coli and further purified with cation-exchange chromatography. The deduced molecular masses of WT-Csn and Chim-Csn are 29079 and 30122 u, respectively. ESI-MS analysis showed the molecular masses of WT-Csn and Chim-Csn to be 28948 and 29991 u, respectively (Fig. 3A and B). The difference 1043 u between the two recombinant proteins confirmed the insertion of the two peptides, whereas the lack of 131 u, relative to the deduced molecular mass, on WT-Csn (and on Chim-Csn) is explicable by the removal of the first methionine, which is common for proteins produced in E. coli.
|
Identification of the disulfide bond in Chim-Csn
As shown in Fig. 1D, two peptide loops, each containing a cysteine residue, were inserted in opposite walls of the substrate-binding cleft. These two cysteine residues were designed to form a disulfide bond crossing the protruding loop. The existence of a disulfide bond was examined by chemical labeling and further analyzed by colorimetry or ESI-MS. As expected, when Chim-Csn was treated with DTNB, no UV change at OD412 nm was detected. Further treatment with IAA showed no alteration of the molecular mass (Fig. 3C). These results indicate that there was no accessible free –SH group in Chim-Csn or WT-Csn. In contrast, when Chim-Csn was pre-treated with DTT and further with IAA, a 116-dalton increment in the molecular mass of Chim-Csn was observed (Fig. 3D), indicating that the disulfide bond on the roof was destroyed and the two free –SH groups became available for IAA labeling reaction. One more disulfide bond between Cys92 and Cys166 (Saito et al., 1999), which is buried inside the protein, cannot be labeled, like the case that modification with IAA WT-Csn failed under the same conditions. These results show that the two added cysteines formed a disulfide bond, which might alter the binding topology of the substrate and consequently convert the endo-chitosanase into an exo-type enzyme.
To distinguish the conversion of endo- to exo-type mechanisms, we monitored the variation of viscosity of chitosan (0.5% w/v), monitored in the presence of WT-Csn and Chim-Csn, separately. It is important to add a similar amount of enzyme activity in the viscometric assay to ensure a comparable rate of production of the reducing sugar. The slow initial decrease of viscosity indicates an exo-type of enzymatic catalysis. As Chim-Csn was found to exhibit activity 3% of that of WT-Csn by DNS assay, 50-fold more concentrated Chim-Csn, relative to the employed WT-Csn, was applied in this work. As shown in Fig. 4, WT-Csn decreased the viscosity of chitosan solution greatly at an early stage of the reaction with a rate 5.48 cp/min, whereas Chim-Csn decreased the viscosity at a much smaller rate, 1.48 cp/min. This result indicates that the splitting pattern of substrate-catalyzed Chim-Csn differs from the endo-type of WT-Csn. The viscosity measurements provided a positive clue showing that Chim-Csn promoted the hydrolysis of chitosan in an exo-type manner, but the results are also explicable with an assumption that the enzyme became processive as discussed in the cases of GH-18 chitinase (Horn et al., 2006; Sikorski et. al., 2006), GH-6 cellobiohydrolase II and GH-7 cellobiohydrolase I (Boisset et. al., 2000).
|
ESI-MS analysis of the enzymatic product
In general, an exo-type chitosanase can hydrolyze chitosan into a specific product, and an endo-type chitosanase hydrolyzes chitosan to form a mixture of chito-oligosaccharides. To confirm Chim-Csn is an exo-hydrolase and to analyze its hydrolytic product, we performed a temporal course experiment, monitoring the initial products with ESI-MS. As expected, WT-Csn acting as an endo-chitosanase simultaneously releases various oligosaccharides, such as GlcN, GlcN2 and GlcN3, from chitosan (Fig. 5), consistent with published results (Fukamizo et al., 2005). In contrast, during the entire catalytic reaction of Chim-Csn, GlcN2 was the dominant product. This result not only confirms that Chim-Csn is an exo-hydrolase, but also identifies the enzymatic product as GlcN2.
|
|
Funding |
|---|
|
TopAbstractIntroductionMaterials and methodsResults and discussionFundingReferences |
|---|
National Science Council and the MOE-ATU Program in Taiwan provided financial support.
|
Footnotes |
|---|
Abbreviations: GlcN, 2-amino-2-deoxy-D-glucopyranose; GlcNn, β-1,4-linked oligomer of GlcN with a polymerization degree n.
|
References |
|---|
|
TopAbstractIntroductionMaterials and methodsResults and discussionFundingReferences |
|---|
Ando A., Noguchi K., Yanagi M., Shinomiya H., Kagawa W., Hirata H., Yabuki M., Fujii T. J. Gen. Appl. Microbiol. (1992) 38:135–144.[CrossRef]
Boisset C., Fraschini C., Schülein M., Henrissat B., Chanzy H. Appl. Environ. Microbiol. (2000) 66:1444–1452.
Bradford M.M. Anal. Biochem. (1976) 72:248–254.[CrossRef][Web of Science][Medline]
Cheng C.Y., Li Y.K. Biotechnol. Appl. Biochem. (2000) 32:197–203.[CrossRef][Web of Science]
Cheng C.Y., Chang C.H., Wu Y.J., Li Y.K. J. Biol. Chem. (2006) 281:3137–3144.
Cote N., Fleury A., Dumont-Blanchette E., Fukamizo T., Mitsutomi M., Brzezinski R. Biochem. J. (2006) 394:675–686.[CrossRef][Web of Science][Medline]
Creighton T.E. Proteins: Structures and Molecular Properties (1993) 2nd edn. New York: W.H. Freeman and Company. 225–226.
Fukamizo T., Amano S., Yamaguchi K., Yoshikawa T., Katsumi T., Saito J., Suzuki M., Miki K., Nagata Y., Ando A. J. Biochem. (Tokyo) (2005) 138:563–569.
Guex N., Peitsch M.C. Electrophoresis (1997) 18:2714–2723.[CrossRef][Web of Science][Medline]
Henrissat B., Bairoch A. Biochem. J. (1996) 316:695–696.[Web of Science][Medline]
Horn S.J., Sørbotten A., Synstad B., Sikorski P., Sørlie M., Vårum K.M., Eijsink V.G. FEBS J. (2006) 273:491–503.[CrossRef][Medline]
Laemmli U.K. Nature (1970) 227:680–685.[CrossRef][Web of Science][Medline]
Meinke A., Damude H.G., Tomme P., Kwan E., Kilburn D.G., Miller R.C. Jr, Warren R.A.J., Gilkes N.R. J. Biol. Chem. (1995) 270:4383–4386.
Miller G.L. Anal. Chem. (1995) 3:426–429.
Nanjo F., Katsumi R., Sakai K. J. Biol. Chem. (1990) 265:10088–10094.
Ohtakara A. Methods Enzymol. (1988) 161:426–430.[CrossRef][Web of Science]
Saito J., Kita A., Nagata Y., Ando A., Miki K. Acta Crystallogr. D. (1995) 51:856–857.[CrossRef][Medline]
Saito J., Kita A., Higuchi Y., Nagata Y., Ando A., Miki K. J. Biol. Chem. (1999) 274:30817–30825.
Schwede T., Diemand A., Guex N., Peitsch M.C. Res. Microbiol. (2000) 151:107–112.[Medline]
Schwede T., Kopp J., Guex N., Peitsch M.C. Nucleic Acids Res. (2003) 31:3381–3385.
Sikorski P., Sørbotten A., Horn S.J., Eijsink V.G., Vårum K.M. Biochemistry (2006) 45:9566–9574.[CrossRef][Web of Science][Medline]
Received September 24, 2007; revised April 23, 2008; accepted May 7, 2008.
CiteULike 
Connotea 
Del.icio.us What's this?
This article has been cited by other articles:
|
|
 |
|
  A. Ochiai, T. Itoh, B. Mikami, W. Hashimoto, and K. Murata Structural Determinants Responsible for Substrate Recognition and Mode of Action in Family 11 Polysaccharide Lyases J. Biol. Chem., April 10, 2009; 284(15): 10181 - 10189. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||