Crystal Structure Analysis of a (B/a)8-TIM Barrel Enzyme and Its Mutants : Insights into the Role of Interactions Between Termini in Influencing Protein Stability. Experimental and Computational Study of Protein-Surface-Pockets Occluded by Tryptophan Side-Chains

Abstract

Xylanases (EC 3.2.1.8) are glycosyl hydrolases that catalyze the hydrolysis of internal β-1,4 glycosidic bonds of xylan backbones, and have potential economical and environment friendly applications in the paper pulp, food, animal feed, detergent industries, bio-ethanol and bio-energy production systems. A xylanase from Bacillus sp. NG-27 (BSX), which is an extracellular endoxylanase, belonging to glycosyl hydrolase family 10 (GH10), shows optimum activity at a temperature of 70 °C and at a pH 8.5. It has a (β/α)8-triosephosphate isomerase (TIM) barrel fold, which has been studied concerning its function, structural properties, design and evolution. BSX, apart from thermo-alkalophilic features, shows resistance to SDS denaturation and protease K degradation. Hence, BSX serves as an important model system for fundamental understanding of the structure-stability-evolution relations of the ubiquitous TIM barrel fold. While the factors responsible for the thermal stability of GH10 xylanases have been analyzed, the improvement of thermostability of already thermostable enzymes is an important challenge. In general, there are large differences in optimal temperature (Tm) between hyperthermostable proteins with respect to their mesophilic homologs, indicating considerable scope available for introducing novel protein engineering approaches to improve protein stability. Thermostability and thermotolerance are of particular importance for industrial enzymes, because higher operating temperatures allow higher reactivity, higher bioavailability, higher process yield, lower viscosity, and reducing the risk of contamination. Thus, finding enzymes that can function at high temperatures has immense industrial importance and constitutes an active area of research. Earlier studies on enzymatic activity and thermostability of a recombinant BSX (RBSX) with different extreme N-terminus mutants by biochemical/biophysical methods showed that a single amino acid substitution (Val1→Leu) markedly enhanced the thermostability of recombinant xylanase from 70 °C to 75 °C without compromising its catalytic activity and showed higher cooperativity in the thermal unfolding transition. Conversely, substitution of Val1→Ala (V1A) at the same position decreased the stability of the protein from 70 °C to 68 °C. Furthermore, it was observed that substitution of Phe4 by Ala decreased the stability by ~4 °C whereas substitution of Trp6→Ala and Tyr343→Ala decreased the stability by ~10 °C with respect to RBSX. On the other hand, substitution of Phe4 by another aromatic residue Trp (F4W) did not change the stability and activity of RBSX. However, structural details were not available at that time, precluding any structure-based rationalization of stability changes resulting from a single amino acid substitution. The thesis reports the crystal structures of a recombinant xylanase from Bacillus sp. NG-27 (RBSX) and its various N-terminal and C-terminal mutants namely V1A, V1L, F4A, F4W, W6A, and Y343A. The crystal structure of RBSX (PDB ID: 4QCE) was solved at a resolution of 2.35 Å whereas those of V1A mutant (PDB ID: 4QCF) and V1L mutant (PDB ID: 4QDM) were solved at a resolution of 2.26 Å and 1.99 Å respectively. On the other hand, the crystal structure of F4A was solved at a resolution of 2.23 Å whereas F4W, W6A, and Y343A mutants were solved at a resolution of 2.22 Å, 1.67 Å, and 2.30 Å respectively. The availability of experimentally determined RBSX structure and its various mutant structures has enabled a critical examination including from a network perspective, of factors influencing thermal stability. The crystal structures in combination with computational analysis have provided valuable insights into the structural features that govern protein thermostability. The thesis candidate established a link between N-terminal to C-terminal contacts and RBSX thermostability. The study reveals that augmenting N-terminal to C-terminal noncovalent interactions is associated with enhancement of the stability of the enzyme. Perhaps, for the first time, the study provides a network perspective of N-terminal to C-terminal interactions and shows that the stabilizing interactions are not restricted to terminal regions but propagate to different parts of the protein structure. Furthermore, analysis of structures of different aromatic mutants of RBSX and structural bioinformatics studies were combined to understand the role of long-range aromatic cluster in the form of 'aromatic-clique' in the thermal stabilization of proteins. The results highlight an additional source of stability in thermophilic proteins, which could arise due to the prevalence of aromatic-cliques. In addition, the work exemplifies the experimental evidence specifically through long-range aromatic clique, in reiterating the role of interactions between N- and C-termini in protein stabilization. The thesis candidate demonstrated the experimental evidence depicting the role of partially solvent exposed tryptophan residues in shielding a surface pocket, which influenced the solvation of backbone atoms and stability of the RBSX enzyme. The candidate carried out a comprehensive database analysis of available crystal structures to look into the possible role of partially exposed tryptophan in hyperthermophilic proteins. The study provides strong evidence that partially exposed tryptophan side-chain is recruited in hyperthermophilic proteins for occluding potential surface pockets, to provide backbone solvent shielding and local stabilization. The overall structure of this thesis is further explained through a chapter wise description below: Chapter 1 | An introduction and outline of the thesis This chapter starts with a general introduction about the diversity of microorganisms and their ability to thrive in extreme environments such as high temperature. The research on these enterprising organisms offers not just the insights into the resilience of life on earth or possibilities of life elsewhere in the universe but also can provide exciting opportunities for a variety of industrial, environmental, biomedical, and pharmaceutical applications. While the adaptation of the cell inventory is important, it is a challenge for proteins to overcome high temperature in order to remain folded in the correct three-dimensional structure while maintaining adequate flexibility for their desired function. Hence, elucidation of the molecular basis of protein stability at extreme temperature continues to attract researcher over a board range of disciplines. The various structural features responsible for protein stability are outlined and the basic structural and molecular strategies for the adaptation to high temperatures revealed by structure analysis are delineated. Of all potentially deactivating factors of protein stability, temperature is the best studied. A brief outline of the strategies and approaches for the design of proteins to meet the desirable properties such as increased thermal stability are presented whereas the structural features responsible for stability of triosephosphate isomerase (TIM)-barrel fold is outlined under a separate section. Subsequently a short introduction of family 10 (GH10) xylanases, which has the ubiquitous TIM-barrel fold and their classifications are presented. A section is dedicated to describe various thermostable GH10 xylanases, their structural features responsible for stability, and current and potential biotechnological applications. At the end, the scope of the present work is detailed. Chapter 2| Crystallization, Data Collection and Data processing of recombinant BSX (RBSX) and its different variants: Chapter 2 presents the purification of recombinant xylanase from Bacillus sp. NG-27 (RBSX), its N-terminal variants (V1A, V1L), and aromatic variants (F4W, F4A, W6A, and Y343A). The expression and purification of RBSX and its variants were carried out at the laboratory of our collaborator Prof. V. S. Reddy, Plant Transformation Group, International Centre for Genetic Engineering and Biotechnology (ICGEB), New Delhi, India. Initial crystallization trials were screened by hanging drop vapor diffusion method and micro-batch diffusion method using crystallization-screening kits (Crystal Screen and Crystal Screen 2) from Hampton Research, USA and a laboratory made screen, which was based on reported crystallization condition of native BSX. After a few rounds of trials and optimization of the crystallization condition, diffraction quality rod shaped crystals of recombinant BSX (RBSX) were obtained within ten days, when 2-µl protein solution (10 g/ml) was mixed with 2-µl reservoir solution composed of 0.12 M MgCl2, 0.1 M NaCl, 0.1M Tris-HCl pH 8.5 and 15% PEG 8000. Subsequently, crystal was used for X-ray data collection and it diffracted X-rays to better than 2.2 Å at the home source at cryo-temperature (100 K). RBSX crystals belong to orthorhombic space group P212121 with unit cell parameters a = 54.77 Å, b = 75.65 Å, c = 179.91 Å and α = β = γ =90°. A three-dimensional screening grid was prepared based on crystallization condition of RBSX by carefully varying salt concentration (NaCl and MgCl2 from 10mM to 300mM in the interval of 10mM), different PEG variants (PEG 1000, PEG 3350, PEG 4000, PEG 8000, and PEG 10000) in the range of 5% to 20%. Tris-HCl buffer of pH 8.0 and of pH 8.5 was used in the concentration range of 0.05M and 0.1M respectively. Rod shaped crystals were obtained using hanging drop vapor diffusion method from the condition of 0.1M NaCl, 80mM MgCl2, 0.05M Tris-HCl pH 8.5 and 18 % PEG 8000 and 0.1M NaCl, 60mM MgCl2, 0.1M Tris-HCl pH 8.5 and 16 % PEG 8000 for V1L mutant and V1A mutant respectively. The diffracting crystals of F4A mutant were obtained from the condition of 0.1M NaCl, 140mM MgCl2, 0.05M Tris-HCl pH 8.5 and 15% PEG 8000 by using hanging drop vapor diffusion method. On the other hand, F4W and Y343A, crystals were grown by micro-batch diffusion method containing 1.1-μ1 ratio of protein and crystallization solution of 0.1M NaCl, 120mM MgCl2, 0.1M Tris-HCl pH 8.5 and 18% PEG 8000 and 0.1M NaCl, 150mM MgCl2, 0.1M Tris-HCl pH 8.5 and 15 % PEG 6000 respectively . W6A mutant crystals were grown by hanging drop vapor diffusion method of 0.1M NaCl, 160mM MgCl2, 0.05M Tris-HCl pH 8.5 and 20% PEG 8000. All the crystals were obtained at 20 °C-22 °C in 5-10 days, and were used for diffraction experiments (details in the table below). Table 1

Protein Space a b c α β γ X-ray source PDB group (Å) (Å) (Å) (°) (°) (°) ID RBSX P212121 54.77 75.65 176.91 90 90 90 Home-source 4QCE

V1A C2 73.57 80.12 69.90 90 110.81 90 Home-source 4QCF

V1L P212121 54.88 76.58 176.73 90 90 90 Synchrotron 4QDM

F4W P212121 55.27 77.32 176.75 90 90 90 Home- source 5EB8

F4A P212121 52.62 67.71 181.54 90 90 90 Home- source 5EFF

W6A P212121 54.99 76.60 181.54 90 90 90 Synchrotron 5EFD

Y343A C2 73.86 80.11 69.21 90 111.19 90 Home- source 5EBA

The quality of all dataset was assessed by SFCHECK. The data sets were found

appropriate and useful for structure determination as discussed in Chapter 3. Chapter 3 | Molecular Replacement, Model Building, Refinement, validation of recombinant xylanase (RBSX), and different mutant structures: Chapter 3 details the application of molecular replacement method to the structure solution of RBSX structure, N-terminal and aromatic mutants of RBSX, the course of iterative model building and the refinement carried out and the quality of the final protein structure models. The structure solution for all the structures was obtained by the molecular replacement (MR) method with the program PHASER-MR in the PHENIX package using a search model of native-enzyme (2F8Q). The asymmetric unit of RBSX, V1L, F4A, F4W, W6A crystals was expected to contain two molecules whereas V1A and Y343A crystal was expected to contain one molecule as indicated by Matthews’s coefficient calculation. The final round of refinement was carried out with restrained refinement with TLS parameters for all the structures. The most essential refinement statistics of the final model of RBSX, V1A, and V1L mutant structures are given in Table 2 whereas the same for aromatic mutant structures, F4A, F4W, W6A, and Y343A are given in Table 3. Table 2

Refinement Statistics RBSX V1A V1L Resolution (Å) 27.7-2.32 26.8-2.26 40.2-1.96 Rwork / Rfree (%) 17.9/22.7 17.4/22.5 15.2/19.0 Average B-factors (Å2) Protein 21.6 26.3 13.9 Ligand/ion 15.6 26.4 18.74 Water 20.6 27.2 23.2 RMSD Bond distance (Å) 0.007 0.005 0.019 Bond angles (◦) 1.123 0.955 1.802 Luzzati coordinate 0.279 0.269 0.175 error (Å) Working set Table 3

Refinement Statistics F4A F4W W6A Y343A Resolution (Å) 18.15-2.23 18.97-2.22 32.1-1.67 34.03-2.30 Rwork / Rfree (%) 17.8/24.0 16.8/21.0 15.68/18.58 17.8/23.0 Average B-factors (Å2) Protein 13.1 19.5 16.7 26.8 Ligand/ion 14.2 20.0 21.4 26.1 Water 11.5 28.9 25.9 30.7 RMSD Bond distance (Å) 0.0144 0.0088 0.0139 0.0063 Bond angles (◦) 1.593 1.228 1.5995 1.0951 Luzzati coordinate 0.261 0.252 0.176 0.293 error (Å) Working set Chapter 4 | Mutations at the extreme N-terminus modulate thermostability of RBSX: Implications of interactions between termini for stability This chapter details the structural analysis of RBSX and its various extreme N-terminus mutations in relation to their different thermostability scale. Although several factors have been attributed to thermostability, the stabilization strategies used by proteins are still enigmatic. Studies on a RBSX, which has the ubiquitous (β/α)8-TIM (Triosephosphate isomerase) barrel fold showed that just a single mutation, Valine1→Leucine (V1L), though not part of any secondary structural element, markedly enhanced the stability from 70 °C to 75 °C without loss of catalytic activity. Conversely, substitution of Valine1→Alanine (V1A) at the same position decreased the stability of the enzyme from 70 °C to 68 °C. To gain structural insights as to how a single extreme N-terminus mutation can markedly influence the thermostability of the enzyme, the candidate has determined the crystal structure of RBSX and two mutants. Based on computational analysis of their crystal structures including residue interaction network, a link was established between N- to C-terminal contacts and RBSX thermostability. The study reveals that augmenting N- to C-terminal non-covalent interactions is associated with the enhancement of the stability of the enzyme. Perhaps, for the first time, the study provides a network perspective of N-terminal to C-terminal interactions and shows that the stabilizing interactions are not restricted to terminal regions but propagate to different parts of the protein structure. In addition, several lines of evidence were discussed that point to support the structural coupling between the chain termini and implications of stability changes in different proteins. It is proposed that the strategy of mutations at the termini could be exploited with a view to modulate stability without compromising on enzymatic activity, or in general, protein function, in diverse folds where N- and C-termini are in close proximity. Chapter 5 | Role of long-range aromatic cluster in the structural stability of RBSX Chapter 5 describes the different aromatic mutant crystal structures of RBSX namely F4W, F4A, W6A, and Y343A and the structural comparison with the RBSX crystal structure. Systematic studies of different alanine mutations (F4A, W6A, and Y343A) to disrupt this aromatic cluster showed that substitution of Phe4, Trp6, and Y343 by alanine drastically decreased the stability of recombinant BSX (RBSX). It was observed that substitution of Phe4 by Ala (F4A) decreased the RBSX stability by ~5 °C whereas substitutions of Trp6 by Ala (W6A) and Tyr343 by Ala (Y343A) markedly decreased the stability of the enzyme by ~10 °C. On the other hand, substitution of Phe4 by Trp (F4W) did not result any change in its thermal unfolding pattern of the enzyme. We observed that the mutated amino acid residues (Phe4, Trp6, and Tyr343) in the RBSX structure are part of an ‘aromatic-clique’. An aromatic-clique is defined as a cluster of aromatic residues in which each residue interacts with all other residues within the cluster through aromatic interactions. The study reveals that the decreased stability shown by F4A, W6A, and Y343A mutants resulted from cumulative effects in the loss of aromatic interactions and disruption of aromatic-clique, and reduced van der Waal interactions. In addition, the work exemplifies the importance of interactions between N-terminal and C-terminal through aromatic contacts or packing in folding and stability of the TIM-barrel fold protein. The structure based multiple sequence alignment of RBSX with other GH10 xylanase from Bacillus organisms revealed that aromatic-clique of interest is fully conserved in B. halodurans (BHX) and Bacillus firmus (BFX) xylanases, which are thermostable in nature, like RBSX. On the other hand, this aromatic-clique is not conserved in the GH10 xylanases from Bacillus N137, Bacillus alcalophilus, which are reported as thermo-labile in nature. Furthermore, analysis of available crystal structures of different thermostable xylanases from GH10 family showed the prevalence of aromatic-clique that may be playing a critical role in their structure-stability and folding. Lastly, a comprehensive analysis of homologous pairs of proteins from (hyper)thermophilic and mesophilic organisms was carried out and observed the high occurrence of aromatic-cliques in the thermophilic proteins in comparison to their mesophilic homologs. These results highlight an additional source of stability in thermophilic proteins, which can arise due to the prevalence of aromatic-cliques. The findings reported in the thesis provide important lessons for engineering xylanases for industrial applications. The strategy of mutations based on clustering of aromatic pairs in the form of ‘aromatic-clique’ may be effectively applied to other enzymes and provides new insights for engineers to design proteins for biotechnological applications. Chapter 6 | Tryptophan occludes surface pocket: Implications for protein stability Chapter 6 describes the structural feature of a partially exposed tryptophan residue, which effectively occludes a surface pocket and plays a critical role in RBSX thermo-stabilization. As a part our long-standing interest in the structural analysis of thermostable proteins, it was observed that just a single mutation, W6A of a recombinant xylanase (RBSX) from Bacillus sp. NG-27 decreased the stability from 70 °C to 60 °C. To gain structural insights into how a single mutation W6A can remarkably influence the thermostability of the enzyme, we determined the crystal structure of W6A mutant and compared the same with the crystal structure of RBSX. We serendipitously observed that substitution of Trp6 by alanine (W6A) in the protein results a small surface pocket, which was shielded by the bulky side-chain of Trp6 in the native structure. Inspection of the molecular structure of native protein structure revealed that side chain of Trp6 occludes the surface pocket, sterically impeding entry of solvent molecules including water. We demonstrated the experimental evidence depicting how a partially exposed tryptophan, which was shielding a surface pocket (tryptophan-shield), can directly influence the backbone solvation, and modulate the stability of the enzyme. Furthermore, computational analysis of high-resolution structures of hyperthermophilic proteins reveals that bulky and aromatic indole side-chain of tryptophan effectively occludes surface pockets in several hyperthermophilic proteins. The study provides a strong evidence that partially exposed tryptophan side-chain is recruited in hyperthermophilic proteins for occluding potential surface pockets to provide backbone solvent shielding and local stabilization. Chapter 7 | Summary and future direction Chapter 7 summaries the important findings of the present study from the crystal structure and computational analysis of a recombinant xylanase (RBSX) and its various N-terminal and C-terminal mutants and also outlines the future direction of the work. Appendix A details SFCHECK output for the processed data for all the structures reported in the thesis. Appendix B Reprints of the publications