Rapid Report pubs.acs.org/biochemistry
Neutron Crystallographic Studies Reveal Hydrogen Bond and WaterMediated Interactions between a Carbohydrate-Binding Module and Its Bound Carbohydrate Ligand S. Zoë Fisher,† Laura von Schantz,‡,⊥ Maria Håkansson,§ Derek T. Logan,§,∥ and Mats Ohlin*,‡ †
European Spallation Source, S-221 00 Lund, Sweden Department of Immunotechnology, Lund University, Medicon Village, S-223 81 Lund, Sweden § SARomics Biostructures AB, Medicon Village, S-223 81 Lund, Sweden ∥ Department of Biochemistry and Structural Biology, Lund University, S-221 00 Lund, Sweden ‡
S Supporting Information *
for substantially improving the determination of critical features of ligand binding, active site architecture, and the role of solvent and H-bonds in binding or catalysis. Despite the benefits of using neutron crystallography in structural biology, there are challenges remaining related to the preparation of large crystals (∼1 mm3) for neutron studies. We were recently able to prepare large (∼1.6 mm3) H/Dexchanged crystals of a genetically engineered type B CBM (X-2 with a Leu to Phe mutation at position 110, X-2 Leu110Phe) derived from CBM4-2 of the Xyn10 xylanase from Rhodothermus marinus in complex with a xyloglucan heptasaccharide ligand [XXXG shown in Figures S1 and S2; α-DXylp-(1−6)-β-D-Glcp-(1−4)[α-D-Xylp-(1−6)]-β-D-Glcp-(1− 4)[α-D-Xylp-(1−6)]-β-D-Glcp-(1−4)-β-D-Glcp].9,10 This module belongs to the family 4 CBM as defined by the Carbohydrate-Active enZYmes Database (CAZY; http:// www.cazy.org). CBM X-2 Leu110Phe, like CBM4-2, is a module with affinity for a range of carbohydrates (features of these modules are summarized in Table S1).11 It binds the heptasaccharide XXXG motif (nomenclature defined by Fry et al.12) but not the galactosylated nonasaccharide variant XLLG with high affinity.13 The structure of this complex has in the past been determined to ultrahigh resolution (1.0 Å) using X-ray crystallography,13 but even at this resolution, important Hbonding interactions had to be inferred as the H themselves were not readily visible.10 We now report on the first neutron structure of a CBM, determined to 1.6 Å resolution, and discuss the details of ligand binding interactions. Table 1 contains a summary of all interactions between XXXG and X-2 Leu110Phe. O2 and O3 of XYS1173 of the ligand are involved in a number of key interactions that help anchor XXXG in the ligand-binding cleft. Tyr149 was determined by affinity electrophoresis to be important for recognition of the branched xyloglucan structure but not for the recognition of linear xylan and β-glucan, an observation supported by X-ray crystallography data.13 In the neutron crystal structure, we indeed observe that Tyr149 is a H-bond donor to O3 of the branching
ABSTRACT: Carbohydrate-binding modules (CBMs) are key components of many carbohydrate-modifying enzymes. CBMs affect the activity of these enzymes by modulating bonding and catalysis. To further characterize and study CBM−ligand binding interactions, neutron crystallographic studies of an engineered family 4-type CBM in complex with a branched xyloglucan ligand were conducted. The first neutron crystal structure of a CBM− ligand complex reported here shows numerous atomic details of hydrogen bonding and water-mediated interactions and reveals the charged state of key binding cleft amino acid side chains.
C
arbohydrate-binding modules (CBMs) are integral components of many carbohydrate-modifying enzymes. Such enzymes are critical for many organisms’ supply of energy, but they also have substantial potential in bioengineering processes, including the biofuel industry and for the production of defined oligosaccharides.1−3 The properties of a CBM can profoundly affect the activity of the associated enzyme.4,5 A detailed study of the binding of modules to carbohydrate substrates will provide an understanding of their mode of action and also how their functionality in a given situation can be refined. Consequently, the structures of many CBMs have been determined by X-ray crystallography, revealing details of their mode of binding to their carbohydrate ligands.6,7 Binding interactions, such as through hydrogen bonds (H-bonds) and water, are made through hydrogen (H) atoms and are believed to be a key to these interactions, especially in type B and C CBMs.6 Unfortunately, most H atoms are not visible in structures determined using X-ray crystallography, not even in structures determined at very high resolution. This is due to the poor X-ray scattering power of H atoms. Instead, their presence and possible roles in ligand interactions have to be inferred. In contrast, H [as well as its isotope, deuterium (D)] readily scatters neutrons, similar to the “heavier” atom types found in proteins (C, N, O, and S). D, in contrast to H, does not contribute to a large incoherent background in neutron scattering, and it is common procedure to either fully deuterate or H/D exchange samples (D everywhere vs. D only in labile positions).8 Altogether, neutron diffraction is a powerful tool © XXXX American Chemical Society
Received: September 28, 2015 Revised: October 9, 2015
A
DOI: 10.1021/acs.biochem.5b01058 Biochemistry XXXX, XXX, XXX−XXX
Biochemistry
Rapid Report
The interactions observed in the neutron structure that involve XYS1173 support the inability of X-2 Leu110Phe to bind a further substituted sugar, XLLG (Figure S3). XLLG carries a galactose linked through O2 of XYS1173, and binding it will require a reorientation of XYS1173, disrupting the Hbonds described above. In contrast to the interactions with XYS1173, XYS1174 is highly flexible in the crystal structure and is not engaged directly with the CBM. XYS1174 can also be substituted at the O2 position, but this is not expected to affect binding to X-2 Leu110Phe. It is rather the collection of interactions surrounding XYS1173 that help explain the binding specificity for XXXG over XLLG. Residues Phe69 and Phe110 are important for ligand crossreactivity and provide a crucial π-stacking interaction with BGC1171.11,13 In addition to its π-stacking role, the backbone carbonyl group of Phe110 also forms a H-bond with BGC1170, further orienting the ligand within the binding site. BGC1171 is engaged in a water-mediated H-bond to the CBM through residues Glu112 and Thr104 (Figure 3).
Table 1. XXXG Binding Interactions to X-2 Leu110Phe sugar residue
interacting partner(s)
distance (Å)
BGC1170 (O3) BGC1171 (O2) BGC1171 (O3) BGC1171 (ring) XYS1173 (O2)
Phe110 (-CO) Glu112, Thr104 via S25 (O) XYS1175 (O4) Phe69, Phe110 π-stack His146 (-ND1) Arg115 (-NH1, -NH2) Tyr149 (-OD) Pro68 π-stack Thr74 (-OG1)
2.7 3.4 2.6 ∼4.0 3.0 2.6 2.7 ∼4.0 3.0
XYS1173 (O3) XYS1173 (ring) XYS1175 (O3)
XYS1173 of XXXG (Figure 1). This interaction explains the observation that Tyr149 is critical for binding of xyloglucan.13
Figure 1. Tyr149 is a H-bond donor to O3 of XYS1173. Residues are shown as yellow balls and sticks; exchanged D atoms are colored cyan with the 2Fo − Fc nuclear density shown as light blue mesh, contoured at 1.5σ.
In addition, O2 of XYS1173 is involved in two H-bonds, one as a H-bond donor to neutral His146 and the other as a Hbond acceptor from Arg115 (Figure 2). Arg115 is in turn coordinated on the opposite side through a H-bond with Glu112 (Figure 2). This coordination leads to an additional interaction with O6 of the carbohydrate backbone residue BGC1170, from which XYS1173 establishes its branch (Figure S1). Finally, there is an additional weak interaction in which the XYS1173 ring packs against the π-face of the Pro68 ring (Figure 2).14
Figure 3. Water-mediated H-bonding interactions between BGC1171 (green thicker sticks), water S25, and residues Glu112 and Thr104. Protein residues are shown as yellow balls and sticks; exchanged D atoms are colored cyan, and ligand carbon atoms are colored green. Maps have been omitted for the sake of clarity, and the rest of the sugar visible is shown as thin sticks for context.
Finally, the -OG1 group of Thr74 and O3 of XYS1175 engage in another H-bond between the CBM and the branched XXXG ligand (Figure S4). The apo form of the parent CBM X-2 and the derived X-2 Leu110Phe variant (Protein Data Bank entry 2y6h) contain a number of ordered water molecules within their binding site.11 At least 10 of these water molecules are displaced by carbohydrate ligands as they bind in the ligand-binding cleft.11,13 Similar to structural analyses in galectin and xylose isomerase, here too some water O atoms are displaced by carbohydrate O atoms in the complex structure.15,16 This is likely providing a beneficial entropic effect on the binding event and serves as a chemical template for ligand binding. From X-ray crystallographic studies, it is apparent that both xylan and xyloglucan each displace a similar number of water molecules. In contrast, the presence of branching XYS1173 and XYS1175 causes displacement of two additional water molecules and one additional water molecule, respectively. There are also compensatory movements in selected amino
Figure 2. D from O2 of XYS1173 is involved in ligand binding through two H-bonds, to neutral His146 and Arg115, and a weak πstacking interaction to Pro68. Residues are shown as yellow balls and sticks; exchanged D atoms are colored cyan with the 2Fo − Fc nuclear density shown as light blue mesh, contoured at 1.5σ. B
DOI: 10.1021/acs.biochem.5b01058 Biochemistry XXXX, XXX, XXX−XXX
Biochemistry
Rapid Report
possible to understand the importance of Glu112 through its H-bonding interactions (Figures 2 and 3). Glu112 is involved in the correct positioning of Arg115 so that it can directly interact with branched XYS1173, as well as water-mediated ligand binding to BGC1171. In this way, explicit knowledge and analysis of the binding site architecture through neutron crystallography can aid in the rational design of higher-affinity CBMs. We do point out that the module in this study is genetically engineered, but there are no specific reasons to believe that the chemical basis of its ligand binding would be significantly different from that of the native CBMs. It shares 96% sequence identity, the same fold, and a similar specificity with a native module (CBM4-2) (Table S1). In conclusion, neutron crystallographic studies of ligand binding to a CBM revealed several important interactions that modulate binding and ligand specificity: (1) proper positioning of key amino acid side chains through H-bonded networks, (2) participation of water in preparing the active site for ligand binding (both displaced waters and ordered water adjacent to the ligand-binding site), (3) water-mediated H-bonding to the protein, and (4) the contribution of two π-stacking interactions (Table 1). Importantly, although water molecules may be present in CBM−carbohydrate complexes (as exemplified by Szabo et al.,19 Charnock et al.,20 and von Schantz et al.11), their direct contribution to carbohydrate binding has been questioned.21,22 We here demonstrate the potential of neutron crystallography to visualize hydrogen bonding patterns involving water molecules in CBM−ligand interactions, an approach that directly lends support to or rejects the existence of these interactions.4 This investigation thus demonstrates how neutron diffraction crystallography can complement conventional X-ray crystal structure analyses and develop our understanding of CBM−ligand interactions.
acid side chains in the binding cleft upon carbohydrate binding. To accommodate XYS1173, Arg142 has moved, and for XYS1175, an adjacent Gln72 has been displaced. Despite major water displacement upon carbohydrate binding, three apparently conserved water molecules remain in the binding site: S25, S26, and S29 (Figure 4).
Figure 4. Conserved water network adjacent to the XXXG ligandbinding site in X-2 Leu110Phe. Residues and waters are shown as balls and sticks, as labeled, and exchanged D atoms are colored cyan. 2Fo − Fc nuclear density is shown as light blue mesh, contoured at 1.5σ.
The precise orientation of these waters and their ability to participate in H bonding are , however, not visible in structures determined by X-ray crystallography alone. This lack of atomic detailed information about light atoms limits our understanding of solvent roles in ligand binding and active site architecture. The neutron crystal structure reveals the orientation of the solvent D atoms, thus making it possible to directly observe Hbonds to the surrounding protein in the complex of X-2 Leu110Phe with XXXG. The three remaining water molecules in the complex are very well-ordered with clear 2Fo − Fc nuclear density (Figure 4). Inspection of their positions shows their participation in a H-bonded network that culminates in water S25 being perfectly positioned to accept a H-bond from BGC1171, while H-bonding with -OG1 of Thr104 and -OE2 of Glu112 (Figure 3). As these waters are situated adjacent to the ligand-binding site, we hypothesize that these waters are engaged in similar H-bonding patterns in the apo form. Indeed, inspection of the apo X-ray crystal structure shows that these water molecule’s O positions are conserved compared to the XXXG complex structure. Through their orientation and interactions, they appear to be involved in the establishment of binding site geometry that is prepared and ready to bind the ligand. Further neutron crystallographic studies of different complexes and of the apo form would be very revealing. Interestingly, another related CBM (XG-34), engineered to be specific for xyloglucan, carries a Glu112Asp mutation and has a substantially different orientation of Arg115 in its apo Xray crystal structure.17 Inspection of the X-2 Leu110Phe ligandbinding cleft shows that perturbation of Arg115 will negatively affect ligand binding, and clearly, the H-bond with Glu112 is very important for preparing the binding area (Figure 2). Unfortunately, there is no X-ray crystal structure of CBM XG-34 in complex with its ligand. However, it should be noted that a 2 order of magnitude improvement in its ligand binding affinity was measured upon reversing the mutation to the native Glu at position 112.18 From the current neutron work, it is
■
ASSOCIATED CONTENT
* Supporting Information S
The Supporting Information is available free of charge on the ACS Publications website at DOI: 10.1021/acs.biochem.5b01058. A detailed description of crystallization methods, data collection, and joint X-ray and neutron refinement, an additional table (Table S1), figures, and crystallographic statistics (Table S2) (PDF)
■
AUTHOR INFORMATION
Corresponding Author
*E-mail:
[email protected]. Phone: +46 46 222 4322. Present Address ⊥
L.v.S.: Alligator Bioscience AB, Medicon Village, S-223 81 Lund, Sweden.
Funding
We acknowledge financial support from the European Commission under the Seventh Framework Program by means of the grant agreement for the Integrated Infrastructure Initiative N. 262348 European Soft Matter Infrastructure (ESMI). Notes
The authors declare no competing financial interest. C
DOI: 10.1021/acs.biochem.5b01058 Biochemistry XXXX, XXX, XXX−XXX
Biochemistry
■
Rapid Report
(16) Saraboji, K., Hakansson, M., Genheden, S., Diehl, C., Qvist, J., Weininger, U., Nilsson, U. J., Leffler, H., Ryde, U., Akke, M., and Logan, D. T. (2012) The carbohydrate-binding site in galectin-3 is preorganized to recognize a sugarlike framework of oxygens: ultrahigh-resolution structures and water dynamics. Biochemistry 51, 296− 306. (17) Gullfot, F., Tan, T. C., von Schantz, L., Karlsson, E. N., Ohlin, M., Brumer, H., and Divne, C. (2010) The crystal structure of XG-34, an evolved xyloglucan-specific carbohydrate-binding module. Proteins: Struct., Funct., Genet. 78, 785−789. (18) von Schantz, L., Gullfot, F., Scheer, S., Filonova, L., Cicortas Gunnarsson, L., Flint, J. E., Daniel, G., Nordberg-Karlsson, E., Brumer, H., and Ohlin, M. (2009) Affinity maturation generates greatly improved xyloglucan-specific carbohydrate binding modules. BMC Biotechnol. 9, 92. (19) Szabo, L., Jamal, S., Xie, H., Charnock, S. J., Bolam, D. N., Gilbert, H. J., and Davies, G. J. (2001) Structure of a family 15 carbohydrate-binding module in complex with xylopentaose. Evidence that xylan binds in an approximate 3-fold helical conformation. J. Biol. Chem. 276, 49061−49065. (20) Charnock, S. J., Bolam, D. N., Nurizzo, D., Szabo, L., McKie, V. A., Gilbert, H. J., and Davies, G. J. (2002) Promiscuity in ligandbinding: The three-dimensional structure of a Piromyces carbohydrate-binding module, CBM29−2, in complex with cello- and mannohexaose. Proc. Natl. Acad. Sci. U. S. A. 99, 14077−14082. (21) Flint, J., Bolam, D. N., Nurizzo, D., Taylor, E. J., Williamson, M. P., Walters, C., Davies, G. J., and Gilbert, H. J. (2005) Probing the mechanism of ligand recognition in family 29 carbohydrate-binding modules. J. Biol. Chem. 280, 23718−23726. (22) Pell, G., Williamson, M. P., Walters, C., Du, H., Gilbert, H. J., and Bolam, D. N. (2003) Importance of hydrophobic and polar residues in ligand binding in the family 15 carbohydrate-binding module from Cellvibrio japonicus Xyn10C. Biochemistry 42, 9316− 9323.
ACKNOWLEDGMENTS We thank MAX-lab beamline staff Roberto Appio and Johan Unge and also Tobias Schrader and Andreas Ostermann at BIODIFF.
■
REFERENCES
(1) Dodd, D., and Cann, I. K. (2009) Enzymatic deconstruction of xylan for biofuel production. GCB Bioenergy 1, 2−17. (2) Harris, P. V., Xu, F., Kreel, N. E., Kang, C., and Fukuyama, S. (2014) New enzyme insights drive advances in commercial ethanol production. Curr. Opin. Chem. Biol. 19, 162−170. (3) Varnai, A., Makela, M. R., Djajadi, D. T., Rahikainen, J., Hatakka, A., and Viikari, L. (2014) Carbohydrate-binding modules of fungal cellulases: occurrence in nature, function, and relevance in industrial biomass conversion. Adv. Appl. Microbiol. 88, 103−165. (4) Hervé, C., Rogowski, A., Blake, A. W., Marcus, S. E., Gilbert, H. J., and Knox, J. P. (2010) Carbohydrate-binding modules promote the enzymatic deconstruction of intact plant cell walls by targeting and proximity effects. Proc. Natl. Acad. Sci. U. S. A. 107, 15293−15298. (5) Tomme, P., Van Tilbeurgh, H., Pettersson, G., Van Damme, J., Vandekerckhove, J., Knowles, J., Teeri, T., and Claeyssens, M. (1988) Studies of the cellulolytic system of Trichoderma reesei QM 9414. Analysis of domain function in two cellobiohydrolases by limited proteolysis. Eur. J. Biochem. 170, 575−581. (6) Boraston, A. B., Bolam, D. N., Gilbert, H. J., and Davies, G. J. (2004) Carbohydrate-binding modules: fine-tuning polysaccharide recognition. Biochem. J. 382, 769−781. (7) Gilbert, H. J., Knox, J. P., and Boraston, A. B. (2013) Advances in understanding the molecular basis of plant cell wall polysaccharide recognition by carbohydrate-binding modules. Curr. Opin. Struct. Biol. 23, 669−677. (8) Blakeley, M. P., Hasnain, S. S., and Antonyuk, S. V. (2015) Subatomic resolution X-ray crystallography and neutron crystallography: promise, challenges and potential. IUCrJ 2, 464−474. (9) Nordberg Karlsson, E., Bartonek-Roxå, E., and Holst, O. (1997) Cloning and sequence of a thermostable multidomain xylanase from the bacterium Rhodothermus marinus. Biochim. Biophys. Acta, Gene Struct. Expression 1353, 118−124. (10) Ohlin, M., von Schantz, L., Schrader, T. E., Ostermann, A., Logan, D. T., and Fisher, S. Z. (2015) Crystallization, neutron data collection, initial structure refinement and analysis of a xyloglucan heptamer bound to an engineered carbohydrate-binding module from xylanase. Acta Crystallogr., Sect. F: Struct. Biol. Commun. 71, 1072− 1077. (11) von Schantz, L., Håkansson, M., Logan, D. T., Walse, B., Ö sterlin, J., Nordberg Karlsson, E., and Ohlin, M. (2012) Structural basis for carbohydrate-binding specificity–a comparative assessment of two engineered carbohydrate-binding modules. Glycobiology 22, 948− 961. (12) Fry, S. C., York, W. S., Albersheim, P., Darvill, A., Hayashi, T., Joseleau, J.-P., Kato, Y., Lorences, E. P., Maclachlan, G. A., McNeil, M., Mort, A. J., Grant Reid, J. S., Seitz, H. U., Selvendran, R. R., Voragen, A. G. J., and White, A. R. (1993) An unambiguous nomenclature for xyloglucan-derived oligosaccharides. Physiol. Plant. 89, 1−3. (13) von Schantz, L., Håkansson, M., Logan, D. T., Nordberg Karlsson, E., and Ohlin, M. (2014) Carbohydrate binding module recognition of xyloglucan defined by polar contacts with branching xyloses and CH-Pi interactions. Proteins: Struct., Funct., Genet. 82, 3466−3475. (14) Zondlo, N. J. (2013) Aromatic-proline interactions: electronically tunable CH/pi interactions. Acc. Chem. Res. 46, 1039−1049. (15) Kovalevsky, A. Y., Hanson, L., Fisher, S. Z., Mustyakimov, M., Mason, S. A., Forsyth, V. T., Blakeley, M. P., Keen, D. A., Wagner, T., Carrell, H. L., Katz, A. K., Glusker, J. P., and Langan, P. (2010) Metal ion roles and the movement of hydrogen during reaction catalyzed by D-xylose isomerase: a joint x-ray and neutron diffraction study. Structure 18, 688−699. D
DOI: 10.1021/acs.biochem.5b01058 Biochemistry XXXX, XXX, XXX−XXX