Research Article | Open Access
Ramzan Ali, Muhammad Imtiaz Shafiq, "Sequence, Structure, and Binding Analysis of Cyclodextrinase (TK1770) from T. kodakarensis (KOD1) Using an In Silico Approach", Archaea, vol. 2015, Article ID 179196, 9 pages, 2015. https://doi.org/10.1155/2015/179196
Sequence, Structure, and Binding Analysis of Cyclodextrinase (TK1770) from T. kodakarensis (KOD1) Using an In Silico Approach
Thermostable cyclodextrinase (Tk1770 CDase) from hyperthermophilic archaeon Thermococcus kodakarensis (KOD1) hydrolyzes cyclodextrins into linear dextrins. The sequence of Tk1770 CDase retrieved from UniProt was aligned with sequences of sixteen CD hydrolyzing enzymes and a phylogenetic tree was constructed using Bayesian inference. The homology model of Tk1770 CDase was constructed and optimized with Modeller v9.14 program. The model was validated with ProSA server and PROCHECK analysis. Four conserved regions and the catalytic triad consisting of Asp411, Glu437, and Asp502 of GH13 family were identified in catalytic site. Also an additional fifth conserved region downstream to the fourth region was also identified. The structure of Tk1770 CDase consists of an additional N′-domain and a helix-loop-helix motif that is conserved in all archaeal CD hydrolyzing enzymes. The N′-domain contains an extended loop region that forms a part of catalytic domain and plays an important role in stability and substrate binding. The docking of substrate into catalytic site revealed the interactions with different conserved residues involved in substrate binding and formation of enzyme-substrate complex.
Enzymatic hydrolysis of polysaccharides is a method of choice in many industrial processes due to its high efficiency and better yields of the products as compared to acid hydrolysis. Glycoside hydrolases have been used for processing of starch, cellulose, hemicellulose, and cyclodextrins . Cyclodextrins (CDs) are cyclic oligosaccharides with six or more glucopyranosyl units linked through α-1,4 glycosidic bonds. Cyclodextrins with six, seven, and eight glucopyranosyl moieties are termed as α-, β-, and γ-cyclodextrins, respectively. In water CDs adapt a structure with all hydrophilic groups directed towards exterior surface and hydrophobic groups towards an internal cavity. The internal hydrophobic cavity of CDs allows them to be resistant to hydrolysis by common amylases and also form inclusion complexes with different organic molecules [2, 3]. CDs have many applications in food, agriculture, cosmetics, and pharmaceutical industry. The vast applications of CDs and their hydrolytic products create a need for efficient and specific enzymes for CDs hydrolysis [2, 4, 5]. The glycoside hydrolases have been classified into 14 clans and 133 families. Each clan consists of at least two families that share catalytic fold and mechanism according to database of carbohydrate active enzymes . The family GH13 (also called α-amylase superfamily), the largest family of glycoside hydrolases, is a member of clan GH-H along with families GH-70 and GH-77. The α-amylase family (GH13) is the most important family for industrial applications [7–9]. Although there is low sequence similarity among the enzymes of different families within this clan, they all exhibit certain structural features that have been conserved during evolution [10, 11].
The family GH13 is further classified into 35 subfamilies with at least 26 different specificities , including α-amylase (EC 188.8.131.52), pullulanase (EC 184.108.40.206), glucanotransferase (EC 220.127.116.11), and cyclodextrinase or cyclomaltodextrinase (EC 18.104.22.168). All of these subfamilies share a common catalytic domain comprising a TIM barrel or (β/α)8 barrel with a catalytic triad [13, 14] and a C-terminal domain consisting of β-strands only . Many enzymes also possess N- and/or C-terminal carbohydrate binding modules like CBM34, CBM20, CBM41, and CBM48 [12, 15]. The CD hydrolyzing enzymes include cyclodextrinase, maltogenic amylase, and neopullulanase that hydrolyze CDs into linear maltodextrins or maltose [16, 17]. Recently thermostable pullulan hydrolase III from Thermococcus kodakarensis (KOD1) has also been reported to hydrolyze CDs into maltose or glucose . The thermostable enzymes from hyperthermophiles have many advantages including higher rates of reaction, increased product yields, and decreased risks of contamination as compared to their mesophilic homologs [19–21]. Due to the advanced sequencing technologies and rapidly increasing numbers of genomes being sequenced, the number of sequences being classified as glycoside hydrolases is far exceeding the number of enzymes being structurally or biochemically characterized [22, 23]. Currently, GH13 family contains 26287 sequences but only 99 structures have been resolved . Till the date of writing this work, Protein Data Bank contains only six CD hydrolyzing enzymes (PDB IDs: 4EAF, 1EA9, 2XIE, 1J0H, 1H3G, and 1BVZ) . There is a need for better understanding of sequence and structural components of these proteins and their mechanism of catalysis as CDases. A bioinformatics approach can be used as a valuable predictive tool to provide information about structure and function of these enzymes.
In this work we have used an in silico approach to provide insight into the sequence, structural components, domain arrangement, catalytic machinery, and enzyme-substrate interactions of thermophilic cyclodextrinase (Tk1770) from Thermococcus kodakarensis (KOD1), an enzyme of potential industrial applications. This study provides the first attempt to use in silico approach to provide insight into the structure and key components of catalytic machinery of cyclodextrinase (CDase) from T. kodakarensis.
2. Materials and Methods
2.1. Sequence Retrieval, Alignment, and Phylogenetic Analysis
The amino acid sequence (UniProt ID Q5JJ59) of CDase from T. kodakarensis KOD1 (CDase-Tk; Tk1770) was retrieved from UniProtKB. A blast sequence similarity search was carried out against UniProtKB to find homologs of Tk1770. From the blast results sixteen different sequences of CD hydrolyzing enzymes from bacterial and archeal sources were selected for further studies (Table 1). The alignment of sequences was carried out with Clustal Omega and a rooted tree was generated using Bayesian inference method with default parameters [25, 26].
|A.A means amino acids. |
2.2. Homology Modeling
The Tk1770 CDase was subjected to NCBI BLAST against RCSB PDB (Protein Data Bank) to search suitable template(s) for comparative modeling. Multiple X-ray crystallographic structures (PDB ID: 4AEF, 1J0H, 4AEE, 1EA9, 1SMA, and 1WZL) with sequence identity from 56% to 29%, respectively, were selected as templates (Table 2). The sequences of target (Tk1770) and templates were aligned with Clustal Omega using UGENE program . The alignment and the PDB structures were used as inputs for homology modeling with Modeller v.9.14 . The model optimization was carried out by variable target function method (VTFM) with conjugate gradients (CG) and molecular dynamics (MD) with simulated annealing (SA) methods [27, 28]. The models generated by Modeller were scored on the basis of their DOPE (Discrete Optimized Protein Energy) values and the model with lowest DOPE score was selected for further studies. The homology model was further validated by ProSA-web server and PROCHECK [29, 30]. The model was refined by Modeller loop refinement functions and again validated for confidence. Thus, a reliable model was constructed and visualized using PyMOL .
2.3. Molecular Docking Studies
In order to investigate the enzyme-substrate interactions, the docking of substrates (α-, β-, and γ-cyclodextrins) into the active pocket of Tk1770 was carried out using AutoDock and MGL Tools v1.5.6 . The substrates were prepared by adding polar hydrogen atoms and partial charges. The protein model was prepared by adding polar hydrogens and Gasteiger charges. The grid map dimensions were set around the active site with all other parameters set to default and rigid docking was performed. The candidates poses of the substrates were scored on the basis of their binding energy in kcal/mol and the best poses with lowest binding energy (kcal/mol) were selected.
3. Results and Discussion
3.1. Sequence Alignment and Phylogenetic Tree
The sequence of Tk1770 consisting of 656 amino acids was aligned with sixteen CD hydrolyzing enzymes from the GH13 family (Figure 1). These sequences included eleven archeal enzymes and five bacterial enzymes having sequence identities from 28% to 60% with Tk1770 CDase (Table 1). All enzymes possess three major domains (i) an N-domain, (ii) a catalytic TIM barrel, and (iii) a C-domain [10, 34]. The sequence analysis showed that archeal enzymes contain two N-terminal domains (i.e., N′- and N-domain) in addition to the catalytic and C-domains, whereas the N′-domain is absent in all the bacterial CD hydrolyzing enzymes (Figure 1). A linker region from residues 190 to 203 in Tk1770 connects two N-terminal domains with two C-terminal domains. Four conserved regions of GH13 family in TIM barrel structure were identified from residues 299 to 310, 405 to 414, 433 to 441, and 496 to 502 with catalytic triad being Asp411, Glu437, and Asp502. An additional conserved region of amino acids 533–539 was also identified downstream to the conserved regions I–IV.
A rooted phylogenetic tree was constructed from alignment using MrBayes with rate matrix wag (fixed) to find evolutionary relationship. The tree was divided into three clades with all bacterial enzymes forming one clade and archeal enzymes divided into two clades (Figure 2). The tree showed that Tk1770 CDase is more closely related to THEGJ MAse and THES4 CDases with a sequence identity of 59% and 60%, respectively (Figure 2). The STAMF α-amylase shows 28% sequence identity with Tk1770 CDase and acts as outgroup in the phylogenetic tree. The α-amylases usually do not exhibit CD hydrolyzing activity and they also lack N′-domain. The α-amylase (STAMF α-amylase) from S. marinus is quite unique in this regard as it exhibits both CD hydrolyzing activity and additional N′-domain . It suggests that during the course of evolution the presence of N′-domain might be linked to CD hydrolyzing activity in archaea.
3.2. Homology Modeling
The homology modeling program Modeller v9.14  was used to construct 3D structure of Tk1770 with multiple templates as described in Materials and Methods. Out of five models generated the best model with lowest DOPE value was selected.
In homology modeling, sometimes the model might contain certain high-energy loops or residues with unusual geometry. Thus, the model selected was refined using Modeller built-in loop-refinement function on loops ranging from 3 to 7 amino acids in length and then validated with ProSA-web server and PROCHECK analysis . The overall quality of the model was estimated by ProSA server in terms of -score by comparing it with -score values of experimentally resolved protein structures in Protein Data Bank . Ramachandran plot validated all the nonglycine, nonproline residues to be in allowed regions and 87.9% of residues in most favorable regions. This verifies that all the residues exhibited accurate stereochemical positions.
Homology model of Tk1770 CDase was aligned with P. furiosus neopullulanase (PYRFU NPase) (PDB ID: 4AEF) for an analysis and comparison of the active site and other structural features. The overall structure of Tk1770 CDase folds into four major domains with two β-strands only N-terminal domains (i.e., N′- and conventional N-domain), connected to TIM barrel (A-domain) and a C-terminal domain, also consisting of β-strands. The structure of N′-domain of Tk1770 typically represents CBM48 with eight β-strands [15, 36]. The structural alignment of N′- or CBM48 domain of Tk1770 and PYRFU NPase revealed that both contain a loop that extends into the catalytic site. However, the extended loop of N′-domain of PYRFU NPase forms a more flexible helical turn as compared to the loop of Tk1770 (Figure 3). The substitution of P91 and S92 in extended loop region of Tk1770 in place of K89 and G90 in loop of PYRFU NPase might be responsible for this apparent decreased flexibility of loop in N′-domain of Tk1770 (Figure 3). Furthermore, K89 and G90 in extended loop of N′-domain in PYRFU NPase form strong hydrogen bonding with D460 and E470 of catalytic site. In Tk1770 S92 makes only one hydrogen bond with D460, thus reducing the interactions between the N′-domain and catalytic domain. Moreover, S92 in Tk1770 rotates backward to form hydrogen bond with Y93 that makes the loop more rigid. All of these factors may contribute to the decreased stability of the enzyme domains in Tk1770. Recently it was reported that the optimum temperature for Tk1770 CDase is 65°C, which is much lower as compared to optimal growth temperature for T. kodakarensis (85°C) and the optimum temperature for other archeal CD hydrolyzing enzymes .
The (β/α)8 barrel (A-domain) also contains a much larger B-domain between β-strand 3 and alpha-helix 3 from residues 306 to 403. The B-domain of all the archeal enzymes possesses a helix-loop-helix (HLH) motif that extends at the entrance of active site (Figure 3), but this HLH motif is absent in all five bacterial enzymes as shown in Figure 1. It has been reported that in order to maintain activity at high temperatures archaea might have adapted additional structural features. These features include N′-domain with a loop extension into the catalytic site and HLH motif that provides all necessary components for substrate binding and catalysis in a monomer [35, 38].
3.3. Docking of Substrates into the Catalytic Site
The docking of substrates into the catalytic site provided information about the interactions in enzyme-substrate complex. For this purpose, AutoDock was used to dock cyclodextrins, that is, α-, β-, and γ-cyclodextrins, into the active site of the Tk1770 CDase model. All of the conformations of ligands generated by the AutoDock were scored on the basis of their binding affinities in kcal/mol. The best poses of α-, β-, and γ-cyclodextrins were selected with binding energies of −8.8, −6.1, and −7.8 kcal/mol, respectively.
The docking results showed that, apart from interacting residues, a number of residues come in close proximity with substrates, especially hydrophobic residues like Y93, F95, F373, F374, and V376. In case of docked α-cyclodextrin, residues D502, R550, and S375 formed hydrogen bonds with hydroxyl groups of substrate (Figure 4). In our homology model, K94 in the loop extension of N′-domain forms a salt bridge with E504 from the active site and might contribute to the stability of two domains, in the same manner as observed by Park et al. in amylase/neopullulanase (4AEF) from P. furiosus . However, docking of β-CD showed strong interactions of K94 with hydroxyl groups of substrate and with E504 (Figure 4). Similarly, docking of γ-CD revealed interactions of K94, R97, and K364 with substrate (Figure 4). The amino acid K364 in helical region of HLH motif extends into the entrance of active pocket right above the F373 and F374 and might have a role in guiding the substrate into the active site. The aromatic amino acids Y88, Y93, and F95 that seem to be forming boundary wall of the active site and K94 protruding into the entrance of catalytic site are conserved in archeal homologs, except STAMF α-amylase.
Cyclodextrinase from hyperthermophilic archaea T. kodakarensis hydrolyzes cyclodextrins into linear maltodextrins. The sequence alignment of CD hydrolyzing enzymes confirmed that archaea have developed an additional N′-domain and a helix-loop-helix (HLH) motif in the B-domain that is absent in all bacterial homologs. The homology model constructed revealed that loop connecting β-strand 7 and β-strand 8 of N′-domain extends into the catalytic site (A-domain) and plays an important role in substrate binding. Residues Y88, Y93, K94, F95, and R97 in extended loop of N′-domain of Tk1770 CDase are conserved in CD hydrolyzing enzymes of archaea. Structural alignment between model and template (4AEF) indicated that P91 and S92 in loop extension of N′-domain of Tk1770 might decrease its flexibility and interactions with A-domain. This might contribute to the decreased stability of two domains in Tk1770.
The docking studies indicated that residues K94, R97, K364, S375, D502, E504, and R550 form hydrogen bonds with substrates. Residue K364 in the helix of HLH motif extending at the entrance of the catalytic site interacts with substrate and might be involved in guiding the substrate into the catalytic site. From these results it can be inferred that archeal CD hydrolyzing enzymes have developed catalytic machinery in which an extension of N′-domain not only constitutes a part of active pocket, but also plays an important role in substrate binding.
Conflict of Interests
The authors declare no conflict of interests exists.
Ramzan Ali and Muhammad Imtiaz Shafiq contributed equally to the paper.
- R. M. Kelly, L. Dijkhuizen, and H. Leemhuis, “Starch and α-glucan acting enzymes, modulating their properties by directed evolution,” Journal of Biotechnology, vol. 140, no. 3-4, pp. 184–193, 2009.
- E. M. M. Del Valle, “Cyclodextrins and their uses: a review,” Process Biochemistry, vol. 39, no. 9, pp. 1033–1046, 2004.
- W. J. Shieh and A. R. Hedges, “Properties and applications of cyclodextrins,” Journal of Macromolecular Science, Part A: Pure and Applied Chemistry, vol. 33, no. 5, pp. 673–683, 1996.
- Y. Nakagawa, W. Saburi, M. Takada, Y. Hatada, and K. Horikoshi, “Gene cloning and enzymatic characteristics of a novel γ-cyclodextrin-specific cyclodextrinase from alkalophilic Bacillus clarkii 7364,” Biochimica et Biophysica Acta—Proteins and Proteomics, vol. 1784, no. 12, pp. 2004–2011, 2008.
- K. Uekama, F. Hirayama, and H. Arima, “Pharmaceutical applications of cyclodextrins and their derivatives,” in Cyclodextrins and Their Complexes: Chemistry, Analytical Methods, Applications, chapter 14, pp. 381–422, Wiley-VCH, 2006.
- 2015, http://www.cazy.org.
- P. V. Aiyer, “Amylases and their applications,” African Journal of Biotechnology, vol. 4, no. 13, pp. 1525–1529, 2005.
- W. D. Crabb and C. Mitchinson, “Enzymes involved in the processing of starch to sugars,” Trends in Biotechnology, vol. 15, no. 9, pp. 349–352, 1997.
- T. Han, F. Zeng, Z. Li et al., “Biochemical characterization of a recombinant pullulanase from Thermococcus kodakarensis KOD1,” Letters in Applied Microbiology, vol. 57, no. 4, pp. 336–343, 2013.
- E. A. MacGregor, “An overview of clan GH-H and distantly related families,” Biologia, vol. 60, supplement 16, pp. 5–12, 2005.
- P. M. de Souza and P. D. O. e Magalhães, “Application of microbial α-amylase in industry—a review,” Brazilian Journal of Microbiology, vol. 41, no. 4, pp. 850–861, 2010.
- M. R. Stam, E. G. J. Danchin, C. Rancurel, P. M. Coutinho, and B. Henrissat, “Dividing the large glycoside hydrolase family 13 into subfamilies: towards improved functional annotations of α-amylase-related proteins,” Protein Engineering, Design and Selection, vol. 19, no. 12, pp. 555–562, 2006.
- M. Machovič and Š. Janeček, “The invariant residues in the α-amylase family: just the catalytic triad,” Biologia, vol. 58, no. 6, pp. 1127–1132, 2003.
- Š. Janeček, “How many conserved sequence regions are there in the α-amylase family?” Biologia, vol. 57, supplement 11, pp. 29–41, 2002.
- D. Guillén, S. Sánchez, and R. Rodríguez-Sanoja, “Carbohydrate-binding domains: multiplicity of biological roles,” Applied Microbiology and Biotechnology, vol. 85, no. 5, pp. 1241–1249, 2010.
- J. Matzke, A. Herrmann, E. Schneider, and E. P. Bakker, “Gene cloning, nucleotide sequence and biochemical properties of a cytoplasmic cyclomaltodextrinase (neopullulanase) from Alicyclobacillus acidocaldarius, reclassification of a group of enzymes,” FEMS Microbiology Letters, vol. 183, no. 1, pp. 55–61, 2000.
- K.-H. Park, T.-J. Kim, T.-K. Cheong, J.-W. Kim, B.-H. Oh, and B. Svensson, “Structure, specificity and function of cyclomaltodextrinase, a multispecific enzyme of the α-amylase family,” Biochimica et Biophysica Acta—Protein Structure and Molecular Enzymology, vol. 1478, no. 2, pp. 165–185, 2000.
- N. Ahmad, N. Rashid, M. S. Haider, M. Akram, and M. Akhtar, “Novel maltotriose-hydrolyzing thermoacidophilic type III pullulan hydrolase from Thermococcus kodakarensisxs,” Applied and Environmental Microbiology, vol. 80, no. 3, pp. 1108–1115, 2014.
- G. D. Haki and S. K. Rakshit, “Developments in industrially important thermostable enzymes: a review,” Bioresource Technology, vol. 89, no. 1, pp. 17–34, 2003.
- C. Bertoldo and G. Antranikian, “Starch-hydrolyzing enzymes from thermophilic archaea and bacteria,” Current Opinion in Chemical Biology, vol. 6, no. 2, pp. 151–160, 2002.
- M. W. Bauer, L. E. Driskill, and R. M. Kelly, “Glycosyl hydrolases from hyperthermophilic microorganisms,” Current Opinion in Biotechnology, vol. 9, no. 2, pp. 141–145, 1998.
- C. Vieille and G. J. Zeikus, “Hyperthermophilic enzymes: sources, uses, and molecular mechanisms for thermostability,” Microbiology and Molecular Biology Reviews, vol. 65, no. 1, pp. 1–43, 2001.
- S. K. Khare, A. Pandey, and C. Larroche, “Current perspectives in enzymatic saccharification of lignocellulosic biomass,” Biochemical Engineering Journal, vol. 102, pp. 38–44, 2015.
- 2015, http://www.cazy.org/GH13.html.
- K. Okonechnikov, O. Golosova, M. Fursov et al., “Unipro UGENE: a unified bioinformatics toolkit,” Bioinformatics, vol. 28, no. 8, pp. 1166–1167, 2012.
- F. Ronquist and J. P. Huelsenbeck, “MrBayes 3: bayesian phylogenetic inference under mixed models,” Bioinformatics, vol. 19, no. 12, pp. 1572–1574, 2003.
- B. Webb and A. Sali, Protein Structure Modeling with MODELLER. Protein Structure Prediction, Springer, 2014.
- N. Eswar, B. Webb, M. A. Marti-Renom et al., “Comparative protein structure modeling using MODELLER,” in Current Protocols in Protein Science, John Wiley & Sons, 2007.
- M. Wiederstein and M. J. Sippl, “ProSA-web: interactive web service for the recognition of errors in three-dimensional structures of proteins,” Nucleic Acids Research, vol. 35, no. 2, pp. W407–W410, 2007.
- R. A. Laskowski, M. W. MacArthur, D. S. Moss, and J. M. Thornton, “PROCHECK: a program to check the stereochemical quality of protein structures,” Journal of Applied Crystallography, vol. 26, no. 2, pp. 283–291, 1993.
- W. L. DeLano, The PyMOL Molecular Graphics System, DeLano Scientific, Palo Alto, Calif, USA, 2002.
- J. Ren, L. Wen, X. Gao, C. Jin, Y. Xue, and X. Yao, “DOG 1.0: illustrator of protein domain structures,” Cell Research, vol. 19, no. 2, pp. 271–273, 2009.
- G. M. Morris, H. Ruth, W. Lindstrom et al., “AutoDock4 and AutoDockTools4: automated docking with selective receptor flexibility,” Journal of Computational Chemistry, vol. 30, no. 16, pp. 2785–2791, 2009.
- M. Machovič, B. Svensson, E. Ann MacGregor, and Š. Janeček, “A new clan of CBM families based on bioinformatics of starch-binding domains from families CBM20 and CBM21,” FEBS Journal, vol. 272, no. 21, pp. 5497–5513, 2005.
- T.-Y. Jung, D. Li, J.-T. Park et al., “Association of novel domain in active site of archaic hyperthermophilic maltogenic amylase from Staphylothermus marinus,” The Journal of Biological Chemistry, vol. 287, no. 11, pp. 7979–7989, 2012.
- M. Machovič and Š. Janeček, “Domain evolution in the GH13 pullulanase subfamily with focus on the carbohydrate-binding module family 48,” Biologia, vol. 63, no. 6, pp. 1057–1068, 2008.
- Y. Sun, X. Lv, Z. Li, J. Wang, B. Jia, and J. Liu, “Recombinant cyclodextrinase from Thermococcus kodakarensis KOD1: expression, purification, and enzymatic characterization,” Archaea, vol. 2015, Article ID 397924, 8 pages, 2015.
- J.-T. Park, H.-N. Song, T.-Y. Jung et al., “A novel domain arrangement in a monomeric cyclodextrin-hydrolyzing enzyme from the hyperthermophile,” Biochimica et Biophysica Acta—Proteins and Proteomics, vol. 1834, no. 1, pp. 380–386, 2013.
Copyright © 2015 Ramzan Ali and Muhammad Imtiaz Shafiq. This is an open access article distributed under the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.