Asian Journal of Scientific Research

Year: 2020 | Volume: 13 | Issue: 1 | Page No.: 92-110
DOI: 10.3923/ajsr.2020.92.110

In silico Structural Modelling and Comparative Analysis of β-mannanases from Psychrophilic and Mesophilic Arthrobacter sp.

Wan Nur Shuhaida Wan Mahadi, Clemente Michael Wong Vui Ling, Kenneth Francis Rodrigues and Cahyo Budiman

Abstract: Background and Objective: Endo-1,4-β-mannanase (β-mannanase, EC 3.2.1.78) is an industrially important enzyme which catalyses the hydrolysis of mannan-based polysaccharides. This enzyme is produced by psychrophilic and mesophilic groups of Arthrobacter sp., yet none of them were structurally characterized. This study aims to decipher the structural features of Arthrobacter β-mannanases that might responsible for their cold adaptation. Material and Methods: Thirty amino acid sequences encoding β-mannanases from Arthrobacter sp. were retrieved from GenBank and subjected to series of analysis of amino acid profiling and structural homology modelling using Phyre2 and SWISS-MODEL. Results: Structural alignment showed the catalytic residues (2 glutamic acids) were conserved among β-mannanase suggesting that they might shares catalytic mechanism. Psychrophilic β-mannanases showed remarkable differences from the mesophilic ones in the content of hydrophilic, particularly negatively charged, residues and proline, which were thought to be important for its cold adaptation. Three-dimensional model of all Arthrobacter β-mannanases forms a classic α/β barrel motif consisting of 8 helices and 9 β-sheets structures, except for psychrophilic ones, which having 8 helices and 8 β-sheets. Conclusion: Adaptation of Arthrobacter β-mannanases towards cold temperature involves structural adjustment particularly on structural flexibility and amino acid distribution.

Keywords: β-mannanase, Arthrobacter sp., cold-adaptation, structural modelling and mannan

INTRODUCTION

β-mannanase or specifically endo-1,4-β-mannanase (EC 3.2.1.78) is an enzyme belonging to the glycosyl hydrolase (GH) families that catalyses the hydrolysis of β-D-1,4-mannosidic linkages in mannan based polysaccharides, leaving shorter chains of oligosaccharides^1-3. These enzymes offer many applications, has many demands in various industries and is produced by many microorganisms in extracellular forms^3,4. Arthrobacter sp. is one of the bacterial species known to produce β-mannanases and is widely found in soil and extreme contaminated environments including the Antarctic region^5-8. This group has been found to have a growing temperature of 0-45°C which makes this bacteria to be either categorised as psychrophilic or mesophilic group^9,10. GenBank houses at least 30 amino acid sequences of β-mannanases from Arthrobacter sp. which have been isolated from psychrophilic and moderate mesophilic Arthobacter sp. Nevertheless, to our knowledge, none of β-mannanases of Arthrobacter sp. has been structurally studied extensively so far. While a number β-mannanases structures were deposited in protein data bank (PDB), none of them originated from Arthobacter sp. In addition, there are no inclusive reports on psychrophilic β-mannanases from any organism.

Structural studies of psychrophilic enzymes gain wide interest as the mechanisms by which the enzymes adapt to low temperature are diverse. Particularly, the cold adaptation mechanism of psychrophilic β-mannanases remains poorly understood. X-ray crystallography and nuclear magnetic resonance (NMR) spectroscopy are undoubtedly powerful tools to decipher the protein structures at atomic level. However, the technologies are costly and tedious^11,12. On the other hand, it is important to note that amino acid sequence of proteins dictate the structural properties of the protein¹³. Furthermore, the amino acid sequence stores the information required for the determination and characterization of a protein molecule’s functions, physical and chemical properties¹⁴. Accordingly, harnessing the amino acid sequence through a series of various platforms of computational biology (in silico) analysis is considerably a viable tool as an alternative over other costly technologies. In silico structural analysis, structural homology modelling is known to be getting more popular within this decade. This is due to the sigmoidal growth of deposited experimental data in the protein data bank (PDB) and a rapid progress of computation technology. The availability of experimental data at PDB provides various structural templates for the structural modelling with better accuracy¹⁵.

In this study, structural features of Arthobacter β-mannanases isolated from various environments were analysed using various in silico platforms. This involves amino acid composition, chemical properties and three-dimensional structural models of these β-mannanases. The comparative analysis between mesophilic and psychrophilic Arthobacter β-mannanases was then used as a basis to discuss the cold adaptation mechanism of this enzyme.

MATERIALS AND METHODS

Sequence and structural retrieval: The amino acid sequences of 30 mannan endo-1,4-β-mannosidase (β-mannanase) protein was retrieved from the GenBank (https://www. ncbi.nlm.nih.gov/genbank/). The study was conducted at Biotechnology Research Institute, Universiti Malaysia Sabah, Malaysia, for the period of October, 2018-March, 2019. For comparison, previously reported β-mannanases were included in this study including CfMan26A (Cellulomonas fimi), MeMan5A (Mytilus edulis) and TrMan5A (Trichoderma reesei). The complete list of organisms used in this study is shown in Table 1.

The POWER web interface (http://power.nhri.org.tw/) was connected for phylogenetic tree development and investigation in light of the Dayhoff-PAM method using the amino acid sequence of the proteins¹⁶. The Kitsch (http:// caps.ncbs.res.in/iws/phylip_files/kitsch.html) program (Fitch-Margoliash slightest squares strategy) was utilized for investigation of phylogenetic tree¹⁷. The unwavering quality of the assessed trees was assessed by using the Bootstrap strategy with 1000 replications.

Primary structural analysis and secondary structural analysis: ExPasy (Expert Protein Analysis System) ProtParam server¹⁸ has been 100 utilized for physiochemical characterization of the β-mannanases. These parameters include theoretical isoelectric points (pI), molecular weights, the total number of positive and negative residues, instability index, aliphatic index and the grand average hydropathy (GRAVY)^19-21. PSIPRED was used for secondary structure analysis of the proteins²². It is a server that uses the principle of 2 feeds which are the forward and neural networks.

Analysis of conserved catalytic residues: Analysis of the sequences was conducted using Clustal Omega-MSA for obtaining pairwise distance^23,24.

The catalytic residues of the β-mannanases were determined using sequence alignment from 3 reported and well-studied β-mannanases from Cellulomonas fimi, Mytilus edulis and Trichoderma reesei. The active sites of those β-mannanases (E175/E282, E177/E308 and E169/E276, for Cellulomonas fimi, Mytilus edulis and Trichoderma reesei, respectively) were used as references^25-27.

Homology protein modelling: The modelling of the 3D structure of the β-mannanases was performed using SWISS-MODEL and Phyre2 servers²⁸ and displayed using PyMoL software²⁹. The models were then evaluated by Verify 3D (http://toolkit.tuebingen.mpg.de/modeller/verity3d/), the Ramachandran plot using RAMPAGE (mordred.bioc.cam.ac. uk/~rapper/rampage.php), GMQE (Global Model Quality Estimation) and QMEAN (the Qualitative Model Energy Analysis) scores³⁰. The B-factor of each protein was calculated from the ResQ server (http://zhanglab.ccmb.med.umich. edu/)³¹.

RESULTS

Arthrobacter producing β-mannanases: Table 1 showed the 30 Arthrobacter strains producing 1,4-β-D mannanase from various environments including soil, water and plants. Based on their growth temperatures, 4 strains were identified as psychrophilic bacteria while 16 strains were classified as mesophilic bacteria. In addition, the optimum growth temperatures for the rest of 10 strains are unknown. Phylogenetic analysis of the thirty β-mannanases and 3 of the control was shown in Fig. 1. The distance between Arthrobacter β-mannanases were relatively close to each other, exceptionally with the three controls and A16 which is a slightly different species as mentioned previously.

Primary structure profile: Primary structure analysis showed that all β-mannanases from these bacteria were multi-domain proteins with a catalytic domain as a structural region responsible for the catalytic activity (Fig. 2). Amino acid sequences of these β-mannanases were found to have low similarities to the well-studied β-mannanases: TriMan5A, CFMan26, MeMAN5A (Table 2). Interestingly, the similarities among Arthrobacter β-mannanases also varied from 17.67-87.95%. Further analysis of the amino acid sequences shown in Fig. 3 revealed that the theoretical sizes and isoelectric point (pI) of the full-length were higher than the catalytic domain. Further, the instability index values, the aliphatic index and the grand average of hydrophobicity (GRAVY) of the full-length of Arthrobacter β-mannanases were also remarkably higher than that of the catalytic domain.

Amino acid profile: Figure 4 revealed that the full-length of Arthrobacter β-mannanases were dominated by hydrophobic residues followed by neutral, hydrophilic, charged and pro residues. Similar patterns were also found in the catalytic domain. Nevertheless, the hydrophilic residues were found to be higher than the neutral residues. Overall, the full-length had remarkable higher hydrophobic, neutral and proline residues than the catalytic domain.

Further, Table 3 revealed that β-mannanases from psychrophilic bacteria were found to have higher hydrophilic and charged residues than the mesophilic group. Meanwhile, neutral residues and proline of both groups were similar. Further, Table 3 also showed that hydrophilic residues of the full-length psychrophilic Arthrobacter β-mannanases were higher than that of the mesophilic group. By contrast, remarkable differences between the catalytic domain of β-mannanases from psychrophilic and mesophilic groups were observed in hydrophobic and hydrophilic groups (Table 3). In addition, Table 3 also showed that the instability index of mesophilic Arthrobacter β-mannanases was found to be higher than that of psychrophilic Arthrobacter β-mannanases. This was particularly observed in the catalytic domains of both β-mannanases. In addition, the aliphatic index and GRAVY of full-length and catalytic domains of mesophilic Arthrobacter β-mannanases were also higher than that of psychrophilic Arthrobacter β-mannanases.

Secondary structure profile: Figure 5 showed the full-length and catalytic domain of Arthrobacter β-mannanases were dominated by coil structures, followed by α-helix, β-sheet and β-turn secondary structures. Furthermore, Table 4 revealed that the full-length and catalytic domain of psychrophilic and mesophilic Arthrobacter β-mannanases were dominated by random coil structures, followed by α-helix, β-sheet and β-turn contents (Table 4).

Among these secondary structures, only α-helix contents of the full-length psychrophilic and mesophilic Arthrobacter β-mannanases were found to be considerably different. Meanwhile, the catalytic domain of the psychrophilic group has lower content of random coil secondary structures than that of the mesophilic group, while the other secondary structures were considerably comparable (about 1% difference only).

Three-dimensional model structures: The 30 selected models as shown in Fig. 6 were mostly built based on the template of endoglucanase H of Clostridium thermocellum (PDB ID:2V3G) under the SCOP domain of D2V3GA1.

All these models were selected as these models met the standards for structural quality parameters including structural geometry (Ramachandran plot), GCMQE, QMEAN and Verify-3D. The Ramachandran plot revealed that most of the residues of Arthrobacter β-mannanases were located in favoured regions (94-98%). In addition, all selected structures had GMQE scores ranging from 0.51-0.96 which were considered to be in the moderate to high score level. Meanwhile, the selected models have acceptable z-score (QMean), ranging from -4-0, which was considered a good structure model. Besides, 3D-1D score residues (Verify-3D) of all selected models were considered to be a good model as the scores were ranging from 96-100% (higher than the minimum standard of 65%). The best model selected were mostly built based on Phyre2 (19 models) and the SWISS-MODEL (11 models) platforms.

Figure 6 showed that all Arthrobacter β-mannanases folded into a classic (β/α)₈-barrel whereby the helices and strands form a solenoid that curved around to close on itself in a doughnut shape. The parallel β-strands formed the inner wall of the doughnut (hence, a β-barrel), whereas the α-helices formed the outer wall of the doughnut. Each β-strand connected to the next adjacent strand in the barrel through a long right-handed loop that included one of the helices, so that the ribbon N-to-C colouring at the top view proceeded in rainbow order around the barrel. In general, Table 5 indicated all structures showed high similarity among each other as indicated by low RMSD values of Cα-atoms (0.00-0.036 Å), with the exception of strain A16. Comparative analysis between the models of Arthrobacter β-mannanases from psychrophilic and mesophilic revealed the structures were highly similar (Table 5). Furthermore, 2 active sites of glutamic acids between the psychrophilic and mesophilic β-mannanases also indicated the residues were in highly-conserved placement (Fig. 7). In addition, B-factors for the 1st and 2nd catalytic residues of the psychrophilic group ranged from 20.00-24.10 (average: 21.93) and 25.57-32.65 (average: 29.76). Meanwhile, B-factors for the 1st and 2nd catalytic residues of the mesophilic group ranged from 20.00-24.10 (average: 21.93) and 25.01-32.51 (average:30.71) (Table 6).

DISCUSSION

Primary structural analysis revealed that the size and amino acid number of the Arthrobacter β-mannanases are varied.

This might be due to the presence of an additional domain (apart from its catalytic domain), which is the carbohydrate-binding domain (CBM) which functions to anchor substrates more effectively^32,33. Interestingly, Fig. 1 showed that most of Arthrobacter β-mannanases (except for Strain A16 and A13) were in the distinct branches or nodes of the well-known β-mannanases (MeMan5A, TriMan5A and CfMan26A). This might suggest that the historical evolution of Arthrobacter β-mannanases are different from MeMan5A, TriMan5A and CfMan26A and possibly exhibit unique properties. Meanwhile, pI value of the Arthrobacter β-mannanases also varied which suggests different pH adaptation of these proteins. It was previously confirmed that the pI value of an enzyme is related to the pH optimum for their catalytic activity and plays an important role in the solubility of the enzyme³⁴. The instability index values of the full-length and catalytic domain of Arthrobacter β-mannanases were found to be lower than 40 (Table 3). The index refers to the prediction of protein stability in the test tube, whereby the greater index reflects lower stability. In particular, a protein with the index of lower than 40 is considerably stable¹⁹. This suggests that Arthrobacter β-mannanases are generally stable in the test tube.

In addition, Table 3 indicated that the aliphatic index showed that the index of mesophilic Arthrobacter β-mannanase is considerably higher than that of psychrophilic one. This is plausible since psychrophilic proteins were characterized by lower thermal stability than its mesophilic counterparts³⁵. It is interesting that the GRAVY index value of the mesophilic Arthrobacter β-mannanases were found to be higher than that of the psychrophilic group. This suggested that the mesophilic Arthrobacter β-mannanases is more hydrophobic than the psychrophilic group. To note, hydrophobic interaction was known to play important role in the thermal stability of protein. Therefore, thermo-stable proteins are usually characterized by higher GRAVY index and more hydrophobic residues than the thermo-labile proteins^36,37. Nevertheless, Table 3 indicated that compositions of hydrophobic residues between mesophilic and psychrophilic Arthrobacter β-mannanase were similar. Note that the composition only refers to the number of hydrophobic residues, regardless of the magnitude of their hydrophobicity. Accordingly, similar number of hydrophobic residues between these 2 groups does not necessarily imply that both are similarly hydrophobic. As the GRAVY index is obviously different, it is assumed that the psychrophilic Arthrobacter β-mannanase is dominated by less bulky hydrophobic residues which then leads to less hydrophobic than the mesophilic ones.

Overall features of amino acid composition of Arthrobacter β-mannanase indicated the domination of the hydrophobic residues (Fig. 4). Similarly, β-mannanase from alkaliphilic Bacillus sp. N16-5 was also reported to be dominated by hydrophobic residues¹¹. Apart from the involvement of hydrophobic residues in structural stability, these residues might also be catalytically important for substrate binding and were often found in the substrate-binding pocket^38-40. The domination of hydrophobic residues in Arthrobacter β-mannanase was followed by polar uncharged and charged residues. The uncharged residues were reported to be heavily involved in the solvation of the protein as the residues were mostly located on the surface of the protein⁴¹. On the other hand, hydrophilic residues are believed to play a role in the formation of any ionic interaction, hydrogen bond and the Van Der Waals interaction stabilizing the proteins.

Interestingly, while generally cold-adapted enzymes are characterized by lower charged residues^42,43, Table 3 indicated that charged residues of psychrophilic Arthrobacter β-mannanase were found to be slightly higher than the mesophilic ones (Table 3). Nevertheless, Gianese et al.⁴⁴ indicated that rather than the composition (number) of charged residues, their spatial distribution is more important to cold-adaptation strategy. Thermophilic enzymes tend to have more charged residues in their flexible regions in order to increase the stability by compensating the structural entropy. In addition, the ratios of positively and negatively charged residues were also reported to be the factors for cold-adaptation⁴⁵. Nevertheless, Table 3 showed that both psychrophilic and mesophilic Arthrobacter β-mannanases showed lower positively charged residues compared to the negatively charged residues. This indicated that adjustments of charged residues might not be the main strategy of thermal adaptation for Arthrobacter β-mannanase. Alternatively, the structural distributions of charged residues not the number, are the main factors affecting the thermal adaptation of the enzyme. Nevertheless, further structural investigation on the distribution is needed to confirm this assumption.

Secondary structural analysis revealed that the full-length of psychrophilic Arthrobacter β-mannanase had a higher content of random coil as compared to the mesophilic one. This indicated that the full-length adopted a strategy of destabilizing some regions, particularly the coil, to adapt to the low temperature. The coil structure was known to be the flexible region in the protein structure and often determines protein stability. In some cases, the region is found to be disordered due to high flexibility⁴⁶. Accordingly, cold-adapted enzymes often have a more flexible region as compared to the mesophilic or thermophilic counterparts. Further, the average helical structure content of psychrophilic Arthrobacter β-mannanase was considerably lower than the mesophilic one (Table 4, Fig. 5). This is believed to be associated with the structural stability as the longer helical structure is reported to be more stable than the shorter one⁴⁷. It is interesting that the catalytic domain has comparable helical content (<1% differences) than the full-length, which suggested that the adaptation to low temperature by psychrophilic Arthrobacter β-mannanase is dictated by its full-length structure.

Furthermore, the fact that all 3D-models of psychrophilic and mesophilic Arthrobacter β-mannanase were similar (Fig. 6, Table 5) and indicated that the tertiary structure has no or little effect on the temperature adaptation of the enzyme. Rather, secondary and local structure arrangements as discussed above, play a more important role in the adaptation. Indeed, the B-factor of the second glutamic acid active site of the psychrophilic Arthrobacter β-mannanase is similar yet has a wider range compared to that of the mesophilic one (Fig. 7, Table 6). The B-factor reflects the fluctuation of atoms about their average positions which suggests that the protein with higher B-factor is considerably highly dynamic (flexible)⁴⁸. Accordingly, the second glutamic acid active site of the psychrophilic Arthrobacter β-mannanase exhibited higher flexibility than that of the mesophilic one. It is known that the flexibility of the area around the catalytic pocket is one of the strategies of cold-adapted enzymes. Nevertheless, the high structural similarity of psychrophilic and mesophilic Arthrobacter β-mannanase and conservation of the catalytic sites reflected that the enzymes shared similar mechanisms. This also implied that structural adjustment for thermal adaptation might be independent from the catalytic mechanism, however, it may indirectly have affected the catalysis properties.

CONCLUSION

β-mannanases from various Arthrobacter strains were shown to have high genetic relatedness and share the catalytic sites which indicated that the enzymes might employ similar catalytic mechanism. Nevertheless, comparative analysis on these β-mannanases displayed a variation of the physio-chemical and structural properties of the enzymes. The variation is believed to be associated with the adaptation of mechanisms to their respective environments. In particular, cold adaptation of psychrophilic β-mannanases was achieved by a series of adjustments on the secondary structure formation, flexibility (B-factor) around the active sites as well amino acid composition (hydrophilic, particularly negatively charged, residues and proline residues).

SIGNIFICANCE STATEMENT

This study discovers the structural differences of psychrophilic and mesophilic Arthrobacter β-mannanase that can be beneficial for further studies and industrial application. This study will help the researcher to uncover the critical areas of structural regulation of β-mannanases in thermal adaptation that many researchers were not able to explore. Thus, a new theory on the mechanism by which Arthrobacter β-mannanases adapt to low temperature may be arrived at.

ACKNOWLEDGMENT

This study was supported by SBK0253-SG-2016 and GUG0105-1/2017 research grants.

REFERENCES