Development of an Analysis Toolkit, AnalysisFMO ... - ACS Publications

Subscriber access provided by University of Rhode Island | University Libraries

Application Note

Development of analysis toolkit to visualize interaction energies generated by fragment molecular orbital calculations Takaki Tokiwa, Shogo Nakano, yuta yamamoto, Takeshi Ishikawa, Sohei Ito, Vladimir Sladek, Kaori Fukuzawa, Yuji Mochizuki, Hiroaki Tokiwa, Fuminori Misaizu, and Yasuteru Shigeta J. Chem. Inf. Model., Just Accepted Manuscript • DOI: 10.1021/acs.jcim.8b00649 • Publication Date (Web): 05 Dec 2018 Downloaded from http://pubs.acs.org on December 6, 2018

Just Accepted “Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted online prior to technical editing, formatting for publication and author proofing. The American Chemical Society provides “Just Accepted” as a service to the research community to expedite the dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully peer reviewed, but should not be considered the official version of record. They are citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or consequences arising from the use of information contained in these “Just Accepted” manuscripts.

is published by the American Chemical Society. 1155 Sixteenth Street N.W., Washington, DC 20036 Published by American Chemical Society. Copyright © American Chemical Society. However, no copyright claim is made to original U.S. Government works, or works produced by employees of any Commonwealth realm Crown government in the course of their duties.

Page 1 of 19 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Journal of Chemical Information and Modeling

Development of an analysis toolkit to visualize interaction energies generated by fragment molecular orbital calculations Takaki Tokiwa1,2$, Shogo Nakano3$*, Yuta Yamamoto4, Takeshi Ishikawa5, Sohei Ito3, Vladimir Sladek6, Kaori Fukuzawa7,8, Yuji Mochizuki4,8, Hiroaki Tokiwa4*, Fuminori Misaizu1, and Yasuteru Shigeta2,9* 1Department

of Chemistry, Graduate School of Science, Tohoku University, 6-3, Aoba, Aramaki, Aoba-ku, Sendai, Miyagi 980-8578, Japan.

2Department

of Physics, Graduate School of Pure and Applied Sciences, University of Tsukuba, 1-1-1 Tennodai, Tsukuba, Ibaraki 305-8571, Japan.

3Graduate

Division of Nutritional and Environmental Sciences, University of Shizuoka, 52-1 Yada, Suruga-ku, Shizuoka422-8526, Japan.

4Department 5Department

of Chemistry, Rikkyo University, 3-34-1 Nishi-Ikebukuro, Toshima-ku, Tokyo 171-8501, Japan.

of Molecular Microbiology and Immunology, Graduate School of Biomedical Sciences, Nagasaki University, 1-12-4 Sakamoto, Nagasaki 852-8523, Japan.

6Institute 7School

of Chemistry - Centre for Glycomics, Dubravska cesta 9, 84538 Bratislava, Slovakia.

of Pharmacy and Pharmaceutical Sciences, Hoshi University, 2-4-41 Ebara, Shinagawa-Ku, Tokyo 1428501, Japan.

8Institute 9Center

$

of Industrial Science, The University of Tokyo, 4-6-1, Komaba, Meguro-ku, Tokyo 153-8505 Japan.

for Computational Sciences, University of Tsukuba, 1-1-1 Tennodai, Tsukuba, Ibaraki 305-8577, Japan.

These authors contributed equally to this work.

1 ACS Paragon Plus Environment

Journal of Chemical Information and Modeling 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Abstract In modern praxis, a knowledge-driven design of pharmaceutical compounds relies heavily on protein structure data. Nonetheless, quantification of the interaction between protein and ligand is of great importance in the theoretical evaluation of the ability of a pharmaceutical compound to comply with certain expectations. The FMO (fragment molecular orbital) method is handy in this regard. However, the physical complexity and the number of the interactions within a protein-ligand complex render analysis of the results somewhat complicated. This situation prompted us to develop the toolkit AnalysisFMO, which should enable a more efficient and convenient workflow with FMO data generated by quantum-chemical packages such as GAMESS, PAICS, and ABINIT-MP. AnalysisFMO consists of two separate units, RbAnalysisFMO, and the PyMOL plugins. The former can extract inter-fragment interaction energies (IFIEs) or pair interaction energies (PIEs) from the FMO output files generated by the aforementioned quantum-chemical packages. The PyMOL plugins enable visualization of IFIEs or PIEs in the protein structure in PyMOL. We demonstrate the use of this tool on a lectin protein from Burkholderia cenocepacia in which FMO analysis revealed the existence of a new interaction between Gly84 and fucose. Moreover, we found that second-shell interactions are crucial in forming the sugar binding site. In the case of bilirubin oxidase from Myrothecium verrucaria (MvBO), we predict that interactions between Asp105 and three His residues (His401, His403, and His136) are essential for optimally positioning the His residues to coordinate Cu atoms to form a Type 2 and two Type 3 Cu ions.

Availability and Implementation: RbAnalysisFMO and the PyMOL plugins use Ruby and Python (including NumPy), respectively. This software is freely available at http://dfns.ushizuoka-ken.ac.jp/labs/proeng/custom20.html/. Contact: [email protected] (Shogo Nakano); [email protected] (Hiroaki Tokiwa); [email protected] (Yasuteru Shigeta)


Page 2 of 19

Page 3 of 19 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


1. Introduction Protein structural data generated by X-ray crystallography, nuclear magnetic resonance (NMR), and cryo-electron microscopy (cryo-EM), which are represented as PDB records, are broadly utilized in both basic and applied studies, including drug design and protein engineering. More than 140,000 datasets have been deposited in PDB and are freely available in public databases like RCSB PDB. On the other hand, PDB data only contain geometrical information, and it is not straightforward to determine which residues contribute to features such as ligand binding, let alone to obtain any quantitative measure of these interactions. Estimation of pair interaction energies in a protein is generally performed by computational approaches. In this regard, molecular mechanics (MM) calculation is one of the broadly utilized approaches because of its low computational cost and ease of use. Commonly, force-field parameters, such as CHARMM (1) and AMBER (2), are utilized to estimate the energies. On the other hand, a drawback of MM is its low accuracy in recovering interaction energies stemming from electron correlation, such as π-π stacking and hydrophobic interactions. In this sense, ab initio quantum mechanical calculations greatly surpass the MM method in accuracy, but their high computational cost inhibits their straightforward application to larger molecular systems such as proteins. The fragment molecular orbital (FMO) method, which was specifically developed to circumvent these limitations, facilitates the use of ab initio quantum mechanical methods for proteins (3). PIEs, sometimes denoted as IFIEs, are calculated in a manner that accounts for the correct monomer polarization due to the external electrostatic field of all residues in the protein complex. Many examples of FMO applications have been reported, e.g., estimation and improvement of the interaction between an inhibitor and its target protein via partial geometry optimization (4, 5). Although this calculation can provide valuable insight into protein functions, its widespread application is limited, partly due to the complexity of the associated data analysis. Several quantum chemistry packages that include modules for performing FMO calculations [GAMESS (6), ABINIT-MP (7), and PAICS (8)] have individual GUI visualization tools for FMO: FU (FMO utility), BioStation Viewer, and PV (PaicsView), respectively. However, there is no universal visualization tool for these packages, and users must learn how to use them one 3 ACS Paragon Plus Environment


by one. This might deter some potential users from utilizing FMO calculations. To address this issue, we developed the toolkit AnalysisFMO. The use of our toolkit should lead to a more efficient and convenient workflow for FMO data generated by the aforementioned quantumchemical packages. AnalysisFMO includes RbAnalysisFMO, a program for extracting PIEs/IFIEs from an output file, and the PyMOL plugins PyGAMESS, PyPAICS, and PyABINIT-MP, which can visualize the IFIEs and/or PIEs in PyMOL, a popular molecular graphical user interface (GUI) visualization system (9). Several software packages for PDB visualization exist, e.g., Chimera (10), PyMOL (9), VMD (11) etc. PyMOL is a widely utilized viewer for analysis of biomolecules, and it has been referenced by more than 10,000 since 2002 (9). PyMOL utilizes the object-oriented language Python. In addition to its excellent visualization performance, PyMOL also empowers its application programming interface, the PyMOL API, which enables users to extend the functions of PyMOL through plugins written by Python. Currently, many plugins are being developed: GROMACS GUI, which can generate input data for the molecular dynamics simulation software GROMACS (12); CAVER, which searches for and visualizes tunnels and channels in biomolecules (13); and others. However, there is no plugin capable of visualizing protein-protein or protein-ligand interactions obtained by quantum mechanical calculations. Residue interaction analysis is useful in various applied studies such as rational protein design (14), in silico drug design (15), and elucidation of enzymatic mechanisms (16). Hence, this plugin has the potential to accelerate the usage of FMO in such studies.


Page 4 of 19

Page 5 of 19 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


2. Methods Computer environment to run an FMO analysis toolkit for visualization of FMO results In this study, we developed two software tools to visualize the results of FMO calculations: RbAnalysisFMO and a set of PyMOL plugins, written in the Ruby and Python programming languages, respectively. The tools will be described in the following sections. RbAnalysisFMO requires several Ruby libraries: Logger and Nokogiri [sparklemotion/nokogiri, GitHub. https://github.com/sparklemotion/nokogiri/blob/master/LICENSE.md

(accessed

August

3,

2018)], which are used for writing output in HTML and XML formats. RbAnalysisFMO can convert the output file of FMO calculations to a CSV file containing interaction energies represented as IFIEs or PIEs. The CSV file is one of the input files for the PyMOL plugins. The plugins utilize the Tkinter, Python megawidgets (Pmw), and NumPy libraries. Installation can be completed from the PyMOL Plugin installation tool (menu bar in external GUI of PyMOL → Plugin → Plugin Manager → Install new Plugin → Choose file …). After installation, the plugins can be confirmed in the Plugin submenu.

Plugin Flowchart A flowchart showing how to run both the Ruby script and PyMOL plugins is provided in Fig. 1. First, users must perform FMO calculations on a given protein or protein-ligand complex using a package such as GAMESS, PAICS, and ABINIT-MP. These packages generate output files, usually several Mb in size when the calculation is completed. RbAnalysisFMO can extract information about the pair interaction energies and write it to a new, smaller CSV file (Fig. 1). Details will be indicated in the next section. Then, the corresponding PDB and the resultant CSV files can be loaded through the PyMOL plugins: the CSV files prepared from GAMESS, PAICS, and ABINIT-MP data are handled by PyGAMESS, PyPAICS, and PyABINIT-MP, respectively. After loading the data, we can select two modes, “All-pairs” and “Selected-pairs”, depending on what we want to represent/display (Fig. 1). For the “All-pairs” mode, users have to prepare a CSV file containing 5 ACS Paragon Plus Environment


n × n pairs of IFIEs or PIEs, where n represents the total number of residues in the PDB data which were utilized in the FMO calculation. In the plugin GUI, users should provide an aminoacid residue number and a chain name, and then all residues interacting with the defined residue are automatically searched based on the information in the CSV file. For “Selected-pairs” mode, users have to prepare a CSV file containing only IFIEs or PIEs between a user-defined aminoacid residue and other residues. In Table S1, we provide a list of features and supported tools for the AnalysisFMO toolkit.

Ruby script (RbAnalysisFMO) to convert the output file of FMO calculations As indicated in the Methods section, the Ruby script can extract interaction energies from an output file of an FMO software, e.g., IFIEs and PIEs, either between two amino-acid residues in the protein and/or between an amino-acid residue and a ligand. RbAnalysisFMO can analyze two different modes as a fragment-target (ligand, peptide, etc.) interactions (Selected-pairs mode) or fragment-fragment interactions (All-pairs mode). In order to visualize, and hence facilitate the interaction analysis, the output is plotted as bar graphs or a heat map using the Gnuplot program [Gnuplot, http://www.gnuplot.info/ (accessed September 5th, 2018)]. For “Selected-pairs” and “All-pairs” modes, RbAnalysisFMO saves the extracted IFIEs or PIEs as CSV and text files in one- and two-dimensional tables, respectively. It is possible to run FMO calculations in such a way that all pair interactions are calculated or in such a manner that one chooses a particular fragment and only its pair interactions with the rest of the system are evaluated. In the latter case, the “All-pairs” mode cannot be run. When RbAnalysisFMO is executed in the “All-pairs” mode, a modified PDB file is generated. This file contains the Mulliken charges calculated at the FMO2 level of theory inserted into the temperature factor column of PDB. The “Selected-pairs” mode does not output these charges because only interaction energies between the selected fragment pairs are evaluated.


Page 6 of 19

Page 7 of 19 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


PyMOL plugin - the “All-pairs” mode After converting the output to the CSV file, we are ready to run a PyMOL plugin via its GUI. First, the CSV file generated by RbAnalysisFMO and the PDB file are loaded; see field “1” in Fig. S1. Then, users have to select either “All-pairs” or “Selected-pairs” mode; see field “2” in Fig. 2. The “All-pairs” mode enables visualization of IFIEs of residue-residue interactions. Users can define the chain identifier and the amino acid residue number for which interactions should be visualized; see “Input chain name …” and “Input amino acid residue numbers…” in “3” in Fig. S1. The plugin is executed by clicking the “Apply” button in “5” of Fig. S1. At this point, the program picks the interaction energies between the target residue and all remaining residues of the protein. These are divided into two groups: one contains negative (attractive) energies and the other positive (repulsive) energies. The energies for the two groups are individually normalized. Based on the normalized energies, the amino acid residues forming the top three most attractive/repulsive interactions are selected. Red or blue sticks are drawn between the target residue and the three most attractive and/or repulsive interaction partners: red represents a stable (attractive) interaction, whereas blue denotes an unstable (repulsive) interaction. The display of the red and blue sticks can be controlled in field “3” shown in Fig. S1. Stick width reflects the magnitude of the interaction energy.

PyMOL plugin - the “Selected-pairs” mode The “Selected-pairs” mode can be applied if, in the FMO calculation, the user requested only the pair interactions between one residue and all remaining residues. This can be useful for analysis of protein-ligand interactions. First, the PDB and CSV files have to be loaded. To execute the plugin in the “Selected pairs” mode, users have to select it in field “2” in Fig. S1. In contrast to the “All pairs” mode, users now have to set the parameters in “4” of Fig. S1. First, select the desired energy term in the top row (ct+mix) in “4” in Fig. S1. The availability of different energy terms depends on which FMO code was used, and hence, which plugin is appropriate. The PIEDA (Pair Interaction Energy Decomposition Analysis) (17) formalism separates the total interaction energy “Etotal” into the electrostatic energy “Ees”, the exchange repulsive energy, “Eex”, the energy for charge transfer + MIX, “Ect+mix”, and the dispersion 7 ACS Paragon Plus Environment


energy “Edisp”. Additionally, one can select the internal pair energy “Eij-Ei-Ej”, defined as the difference of the dimer energy Eij and the monomer energies Ei and Ej. Finally, the explicit embedded charge transfer energy “Tr(Dij+Vij)” can be selected. These options are available in PyGAMESS and PyABINIT-MP-PIEDA. For the IFIEs (PyPAICS and PyABINIT-MP), a total of three tabs will be available: the Hartree-Fock energy “HF”, the MP2 correlation energy “dMP2” (which roughly corresponds to the dispersion energy), and the MP2 total energy, which is the sum of the two energies (in kcal/mol). Next, users must define the “distance” value. Then, the plugin selects residues that are located within the specified distance from the ligand. The selected amino-acid residues are colored by reference to the normalized energies, generated using the same procedure as in the “All-pairs” mode. Here, color saturation denotes the magnitude of energies.


Page 8 of 19

Page 9 of 19 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


3. Results and Discussion In the following sections, we will demonstrate how the analysis of FMO interaction energies can be carried out using our code, and how it can contribute to obtaining deeper insight into ligand binding than a simple analysis of the crystal structure.

Interaction Energy Analysis of Fucose Specific Lectin, BC2L-C Specific recognition of glycoconjugates covering cell surfaces is the initial step in pathogenic infection (18). Burkholderia cenocepacia, an opportunistic bacterium that often causes rapid pulmonary disorder (19), recognizes sugars using three PA-IIL-like proteins (BC2LA, BC2L-B, and BC2L-C) that belong to a soluble fucose/mannose-binding lectin family (20). BC2L-C has a novel structure among lectins, having a trimeric assembly, and is very similar to tumor necrosis factor-1α (20). BC2L-C exhibits fucose-binding activity, and its crystal structure indicates that the side chains of Thr74 and Arg111 and the main-chain atoms of Thr83 and Arg85 contribute to the interaction with fucose (20). Quantitative assessment of these interactions is difficult using only the crystal structure. FMO calculation was performed on the structure of BC2L-C (PDB ID: 2WQ4) to quantitatively assess the interaction between BC2L-C and fucose. The FMO package PAICS was utilized in the calculation. The visualization of the interaction energies (IFIEs) between selenomethylated fucose (SFU) and other amino-acid residues in “Selected-pairs” mode is shown in Fig. 2A; here, the PyPAICS plugin was used. Among the amino-acid residues located within 3 Å of SFU, four amino-acid residues (Gly84 [chain A], Arg85 [chain A], Thr74 [chain C], and Arg111 [chain C]) exhibited attractive interactions. The magnitude of the interaction energies with these residues descends in the order Arg111 > Arg85 > Gly84 > Thr74. Notably, the mainchain atoms of Gly84 interact with SFU more strongly than those of Thr74; however, Gly84 was not indicated as an interacting residue of SFU in the PDB record. This indicates that FMO calculation can reveal additional interactions that are overlooked in analysis of the crystal structure alone.



Recent work showed that long-range interactions within the secondary shell could contribute to ligand recognition and modulation of enzymatic activity (21). Motivated by this, we decided to verify the existence of such interactions in BC2L-C. For this purpose, the “All-pairs” mode is a useful tool. The stabilizing/attractive interactions between Ser76 (chain B) and other amino-acid residues (Thr83, Gly84, and Arg85 in chain C) that form the sugar binding site are indicated in Fig. 2B. Color saturation and stick thickness are used to indicate the strengths of the interactions. Hence, we can easily predict that the amino-acid residues contribute to the stabilization in the order Gly84 > Arg85 > Thr83 (Fig. 2B). The results suggest that Ser76, which is located in the secondary shell, contributes to the formation of the sugar binding site of BC2L-C by interacting with binding site residues.

Prediction for metal coordination mechanism of bilirubin oxidase In the previous sections, we demonstrated how the toolkit could visualize IFIEs or PIEs on protein structures in PyMOL. Next, we elucidate the biological functions of certain amino acids using the structure of bilirubin oxidase (MvBO, PDB ID: 2XLL), a MvBO is a Cu-dependent enzyme that catalyzes the conversion from bilirubin to biliverdin (22). The enzyme contains three types of Cu atom (type [T] 1, 2, and 3), and the catalytic center is near a T2 Cu and a pair of T3 Cu atoms (Fig. 2C) (22-24). The residues His136, His401, and His403 coordinate these Cu atoms. Kataoka et al. indicated that mutation at Asp105, which is not coordinated directly to the Cu atoms, affects the ligation of Cu atoms in MvBO; in the Asp105Ala and Asp105Asn mutants, one and two Cu atoms, respectively, are missing relative to the native structure (23). However, the reason why these mutations reduce the number of Cu atoms remains unknown. To attempt an explanation, we performed FMO calculation of MvBO. The calculation was performed with GAMESS software, and all Cu atoms in MvBO were omitted because in the FMO scheme, the convergence of monomeric SCF, including the multicenter metal region and dimer-SCC iteration, is computationally hard. The top three amino-acid residues that strongly interacted with Asp105 are shown in Fig. 2D; Asp105 formed attractive interactions with His401, His403, and His136 (Figs. 2D and 2E). Residue-residue interaction is one of the main determinants of the physiological functions of proteins, including protein folding, substrate 10 ACS Paragon Plus Environment

Page 10 of 19

Page 11 of 19 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


binding, and cofactor stabilization in binding pockets. Hence we infer that the interactions of the copper-binding histidines with Asp105 are important for their stabilization in optimal positions for Cu binding and performance of their redox functions (24). In summary, as in other proteins, Asp105 in MvBO serves to localize the three His residues for coordination of Cu atoms in optimal positions for their functions, as shown in Fig. 2D.



4. Conclusion Here, we have outlined our toolkit and its application to a variety of proteins to illustrate its usefulness. These examples proved that intermolecular interactions within proteins or in proteinligand complexes could be quantitatively depicted using our toolkit. In particular, the role of the Asp105 residue in MvBO can be assessed on the basis of the interaction energy obtained by FMO. Our toolkit is highly general because it can visualize the results from different FMO software packages, including GAMESS, PAICS, and ABINIT-MP. In conclusion, we believe that the use of these plugins is a step towards more convenient and illustrative interpretation of FMO data, and could contribute to rational drug design and protein design.


Page 12 of 19

Page 13 of 19 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


Acknowledgements H.T. acknowledges the Rikkyo SFR project, 2014-2016, and the MEXT Supported Program for the Strategic Research Foundation at Private Universities, 2013-2018. V.S. thanks project VEGA 2/0035/16. The computations in this work were performed using the Research Center for Computational Science, Okazaki, Japan; the Center for Computational Sciences (CCS) at University of Tsukuba, Japan; and the facilities of the Supercomputer Center, the Institute for Solid State Physics, The University of Tokyo, Japan.

Funding This project was supported by the Japan Science and Technology Agency (JST) and the National Bioscience Database Center (NBDC).



Figure legends Figure 1. Flowchart of the AnalysisFMO toolkit.

Figure 2. Analysis of IFIEs for the TNF-like trimeric lectin domain from Burkholderia cenocepacia (BC2L-C, PDB ID: 2WQ4) using the AnalysisFMO toolkit. In “Selected-pair” mode, we searched for interactions between SFU, in which a Se atom replaces the S atom, and other amino-acid residues located within 3 Å of SFU (A). IFIEs were calculated utilizing secondorder Møller−Plesset (MP2) perturbation theory with the cc-pVDZ basis set. Attractive interactions are colored in red (A). In the “All-pairs” mode, we searched for interactions between S76 (chain B) and other residues (A). Attractive interactions are indicated by connecting the Cα atoms of the two residues with a red stick (B). Analysis of PIEs for bilirubin oxidase from Myrothecium verrucaria (MvBO, PDB ID: 2XLL) using PyGAMESS. The active-site structure of MvBO (C). Three His residues (H401, H403, and H136) coordinate with Type 2 (T2) and Type 3 (T3) Cu atoms. Representation of PIEs between D105 and other amino-acid residues (D). Top three amino-acid residues forming interactions with D105 are connected with red colored sticks. PIE values for the top five strongly interacting residues with D105 (E).


Page 14 of 19

Page 15 of 19 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


References 1.

Brooks, B. R., Bruccoleri, R. E., Olafson, B. D., States, D. J., Swaminathan, S., and Karplus, M. (1983) CHARMM: A program for macromolecular energy, minimization, and dynamics calculations, J. Comput. Chem. 4, 187-217.

2.

Case, D. A., Cheatham, T. E., 3rd, Darden, T., Gohlke, H., Luo, R., Merz, K. M., Jr., Onufriev, A., Simmerling, C., Wang, B., and Woods, R. J. (2005) The Amber biomolecular simulation programs, J. Comput. Chem. 26, 1668-1688.

3.

Fedorov, D. G., and Kitaura K. (2009) The Fragment Molecular Orbital Method: Practical Applications to Large Molecular Systems, CRC Press.

4.

Heifetz, A., Trani, G., Aldeghi, M., MacKinnon, C. H., McEwan, P. A., Brookfield, F. A., Chudyk, E. I., Bodkin, M., Pei, Z., Burch, J. D., and Ortwine, D. F. (2016) Fragment Molecular Orbital Method Applied to Lead Optimization of Novel Interleukin-2 Inducible T-Cell Kinase (ITK) Inhibitors, J. Med. Chem. 59, 4352-4363.

5.

Heifetz, A., Aldeghi, M., Chudyk, E. I., Fedorov, D. G., Bodkin, M. J., and Biggin, P. C. (2016) Using the fragment molecular orbital method to investigate agonist-orexin-2 receptor interactions, Biochem. Soc. Trans. 44, 574-581.

6.

Schmidt, M. W., Baldridge, K. K., Boatz, J. A., Elbert, S. T., Gordon, M. S., Jensen, J. H., Koseki, S., Matsunaga, N., Nguyen, K. A., Su, S. J., Windus, T. L., Dupuis, M., and Montgomery, J. A. (1993) General Atomic and Molecular Electronic-Structure System, J. Comput. Chem. 14, 13471363.

7.

Tanaka, S., Mochizuki, Y., Komeiji, Y., Okiyama, Y., and Fukuzawa, K. (2014) Electroncorrelated fragment-molecular-orbital calculations for biomolecular and nano systems, Phys. Chem. Chem. Phys. 16, 10310-10344.

8.

Ishikawa, T., and Kuwata, K. (2009) Fragment molecular orbital calculation using the RI-MP2 method, Chem. Phys. Lett. 474, 195-198.



9.

Delano, W., L. (2002) The PyMOL molecular graphics system, (Delano Scientific, S. C., CA., Ed.).

10.

Pettersen, E. F., Goddard, T. D., Huang, C. C., Couch, G. S., Greenblatt, D. M., Meng, E. C., and Ferrin, T. E. (2004) UCSF Chimera--a visualization system for exploratory research and analysis, J. Comput. Chem. 25, 1605-1612.

11.

Phillips, J. C., Braun, R., Wang, W., Gumbart, J., Tajkhorshid, E., Villa, E., Chipot, C., Skeel, R. D., Kale, L., and Schulten, K. (2005) Scalable molecular dynamics with NAMD, J. Comput. Chem. 26, 1781-1802.

12.

Makarewicz, T., and Kazmierkiewicz, R. (2013) Molecular dynamics simulation by GROMACS using GUI plugin for PyMOL, J. Chem. Inf. Model. 53, 1229-1234.

13.

Chovancova, E., Pavelka, A., Benes, P., Strnad, O., Brezovsky, J., Kozlikova, B., Gora, A., Sustr, V., Klvana, M., Medek, P., Biedermannova, L., Sochor, J., and Damborsky, J. (2012) CAVER 3.0: a tool for the analysis of transport pathways in dynamic protein structures, PLOS Comput. Biol. 8, e1002708.

14.

Kiss, G., Celebi-Olcum, N., Moretti, R., Baker, D., and Houk, K. N. (2013) Computational enzyme design, Angew. Chem. Int. Ed. Engl. 52, 5700-5725.

15.

Abe, T., Sando, A., Teraoka, F., Otsubo, T., Morita, K., Tokiwa, H., Ikeda, K., Suzuki, T., and Hidari, K. I. (2014) Computational design of a sulfoglucuronide derivative fitting into a hydrophobic pocket of dengue virus E protein, Biochem. Biophys. Res. Commun. 449, 32-37.

16.

Nakano, S., Okazaki, S., Ishitsubo, E., Kawahara, N., Komeda, H., Tokiwa, H., and Asano, Y. (2015) Structural and computational analysis of peptide recognition mechanism of class-C type penicillin binding protein, alkaline D-peptidase from Bacillus cereus DF4-B, Sci. Rep. 5, 13836.

17.

Fedorov, D. G., and Kitaura, K. (2007) Pair interaction energy decomposition analysis, J. Comput. Chem. 28, 222-237.

18.

Imberty, A., and Varrot, A. (2008) Microbial recognition of human cell surface glycoconjugates, Curr. Opin. Struct. Biol. 18, 567-576. 16 ACS Paragon Plus Environment

Page 16 of 19

Page 17 of 19 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


19.

Courtney, J. M., Dunbar, K. E., McDowell, A., Moore, J. E., Warke, T. J., Stevenson, M., and Elborn, J. S. (2004) Clinical outcome of Burkholderia cepacia complex infection in cystic fibrosis adults, J. Cyst. Fibros. 3, 93-98.

20.

Sulak, O., Cioci, G., Delia, M., Lahmann, M., Varrot, A., Imberty, A., and Wimmerova, M. (2010) A TNF-like trimeric lectin domain from Burkholderia cenocepacia with specificity for fucosylated human histo-blood group antigens, Structure 18, 59-72.

21.

Sommer, D. J., Vaughn, M. D., and Ghirlanda, G. (2014) Protein secondary-shell interactions enhance the photoinduced hydrogen production of cobalt protoporphyrin IX, Chem. Commun. (Camb) 50, 15852-15855.

22.

Koikeda, S., Ando, K., Kaji, H., Inoue, T., Murao, S., Takeuchi, K., and Samejima, T. (1993) Molecular cloning of the gene for bilirubin oxidase from Myrothecium verrucaria and its expression in yeast, J. Biol. Chem. 268, 18801-18809.

23.

Kataoka, K., Kitagawa, R., Inoue, M., Naruse, D., Sakurai, T., and Huang, H. W. (2005) Point mutations at the type I Cu ligands, Cys457 and Met467, and at the putative proton donor, Asp105, in Myrothecium verrucaria bilirubin oxidase and reactions with dioxygen, Biochemistry 44, 70047012.

24.

Frieden, C. (1971) Protein-protein interaction and enzymatic activity, Annu. Rev. Biochem. 40, 653-696.



Figure 1. Flowchart of analysisFMO toolkit. 190x275mm (96 x 96 DPI)

ACS Paragon Plus Environment

Page 18 of 19

Page 19 of 19 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


Analysis of IFIEs for the TNF-like trimeric lectin domain from Burkholderia cenocepacia (BC2L-C, PDB ID: 2WQ4) using the AnalysisFMO toolkit. In “Selected-pair” mode, we searched for interactions between SFU, in which a Se atom replaces the S atom, and other amino-acid residues located within 3 Å of SFU (A). IFIEs were calculated utilizing second-order Møller−Plesset (MP2) perturbation theory with the cc-pVDZ basis set. Attractive interactions are colored in red (A). In the “All-pairs” mode, we searched for interactions between S76 (chain B) and other residues (A). Attractive interactions are indicated by connecting the Cα atoms of the two residues with a red stick (B). Analysis of PIEs for bilirubin oxidase from Myrothecium verrucaria (MvBO, PDB ID: 2XLL) using PyGAMESS. The active-site structure of MvBO (C). Three His residues (H401, H403, and H136) coordinate with Type 2 (T2) and Type 3 (T3) Cu atoms. Representation of PIEs between D105 and other amino-acid residues (D). Top three amino-acid residues forming interactions with D105 are connected with red colored sticks. PIE values for the top five strongly interacting residues with D105 (E). 190x275mm (96 x 96 DPI)




Page 20 of 19

Development of an Analysis Toolkit, AnalysisFMO ... - ACS Publications

Recommend Documents