Abstract
Numerous studies have shown how periplasmic binding proteins (PBPs) bind substrates with exquisite specificity, even distinguishing between sugar epimers and anomers, or structurally similar ions. Yet, marked substrate promiscuity is also a feature encoded in some PBPs. Except for three sub-Ångström crystal structures, there are no reports of hydrogen atom positions in the remaining (> 1000) PBP structures. The previous X-ray crystal structure of the maltodextrin periplasmic-binding protein from Thermotoga maritima (tmMBP) complexed with oligosaccharide showed a large network of interconnected water molecules stretching from one end of the substrate binding pocket to the other. These water molecules are positioned to form multiple hydrogen bonds, as well as forming interactions between the protein and substrate. Here we present the neutron crystal structure of tmMBP to a resolution of 2.1 Å. This is the first neutron crystal structure from the PBP superfamily and here we unambiguously identify the nature and orientation of the hydrogen bonding and water-mediated interactions involved in stabilizing a tetrasaccharide in the binding site. More broadly, these results demonstrate the conserved intricate mechanisms that underlie substrate-specificity and affinity in PBPs.
Similar content being viewed by others
Introduction
Periplasmic binding proteins (PBPs) are the ligand-binding components of ATP-binding cassette (ABC) transporters and are sometimes associated with chemotaxis systems. Members of the PBP superfamily mediate the uptake and transmembrane transport of a diversity of metabolically important solutes in bacteria, such as carbohydrates, ions, amino acids, and polyamines, to name a few.
Despite the wide variation in PBP ligand size and chemical functionality the two domain PBPs have a conserved αβ-fold, with each domain encompassing a central β-stranded core that is sandwiched between surrounding α-helices1. Ligand binding at the interdomain interface induces a hinge-bending motion in which the two solvent exposed polar interfaces of each domain close down on and envelop ligands in a largely desolvated binding site reminiscent of the protein interior2,3,4. However, in many instances water molecules remain within the binding site, and these solvent interactions with ligands have been shown to be important in broadening PBP substrate specificity or fine tuning highly specific binding5,6. For example, in PBPs that share the oligopeptide-binding protein fold, water molecules have been shown to act as adapters, matching the ligand hydrogen-bonding requirements and permitting a diverse array of substrates to be accommodated within a single binding site7,8. In some cases, rather than serving to increase promiscuity, water molecules are also used in PBPs to impart specificity beyond what can be encoded within the repertoire of protein backbone and side-chains6,9. In yet others, PBPs have been shown to have bipartite binding sites, where each sub-site differentially uses water to impart either specificity or promiscuity2,10. Taken together, the diversity of PBP binding site adaptation mechanisms suggests that determining water molecule orientation and hydrogen bonding patterns within the PBP substrate binding site is indispensable for understanding the fine tuning of ligand recognition within this class of proteins.
More than 1000 crystal structures of PBPs have been determined thus far. Of these, only three were determined at sub-Ångstrom resolution; glucose-binding protein (PDB ID:2FVY)11, glutamine-binding protein (PDB ID: 4KQP)12, and a phosphate-binding protein (PDB ID: 4F1V)13. In these three crystal structures, the resolution of the diffraction data permitted the explicit positioning of many protein hydrogen atoms, yet no hydrogen atoms could be modeled on solvent molecules. In all others, the position of key hydrogen atoms and patterns of hydrogen bonding interactions within the PBP binding site are inferred, rather than experimentally determined. Here we report the first neutron crystal structure of a PBP, namely a maltodextrin-binding protein (MBP) from a cluster of three related proteins found in Thermotoga maritima (tmMBP)14. This study reveals the explicit hydrogen bonding interactions and water network within the substrate binding site at an unprecedented level of detail, lending valuable insights into the intricate mechanisms that underpin the derivation of substrate-specificity and affinity in PBPs.
Results
Neutron structure of maltotetraose bound tmMBP
The genome of T. maritma encodes for three isoforms of maltose binding proteins; the tmMBP2 isoform was used in the present study and is called tmMBP going forwards. Large crystals of tmMBP that were suitable for neutron diffraction experiments were grown in sitting drops using nine-well siliconized glass plates. Each sitting drop was made by mixing 400 μl of protein with an equal volume of the mother liquor supplemented with 10% (vol./vol.) deuterated glycerol to slow the rate of crystallization and growth, and incubated at 20 °C. Large crystals (3–10 mm3) appeared within 30 days and were transferred weekly into sandwich boxes with fresh deuterated mother liquor to promote replacement of exchangeable hydrogen atoms for deuterium. A total of four mother liquor exchanges were performed. Crystals were transferred to quartz capillaries with a D2O mother liquor plug and sealed with wax for room temperature neutron and X-ray data collection.
The room temperature X-ray crystal structure of the tmMBP complexed with maltotetraose was determined at 1.70 Å with an Rwork/Rfree of 14.1/17.5%. The X-ray model contained a total of 3099 non-hydrogen atoms (C, N, O, S), of which 2893 were protein atoms, 45 ligand atoms, and 161 solvent atoms. Prior to joint refinement, hydrogen atoms were added to the structure, where exchangeable positions had mixed hydrogen/deuterium occupancy. This starting model was used for joint X-ray/neutron refinement in Phenix15,16 where H/D occupancy and position was experimentally modelled based on the neutron diffraction data. The final room temperature neutron structure was determined to a resolution of 2.10 Å and refined to an Rwork/Rfree of 24.3/27.3%. Data collection statistics and details on the final refined model are given in Table 1.
Overall, the room temperature 2.1 Å neutron structure of tmMBP (Fig. 1) is very similar (r.m.s.d. = 0.36 on backbone atoms) to the published X-ray structure determined at cryogenic temperatures (PDB ID: 6DTS)14. The neutron structure enabled the modelling of deuterium atoms at exchangeable sites on the protein and the bound ligand, as well as deuterium atoms on solvent water molecules. The nuclear density maps showed clearly resolved density for the maltotetraose substrate in the binding site (Fig. 1). Importantly, the neutron structure also resolves the position and orientation of D2O solvent molecules within the substrate binding cavity, which were previously inferred to provide hydrogen bonding interactions in the earlier cryogenic X-ray model (Fig. 1). Thus, the neutron structure shows directly how the network of water molecules that traverse the entire length of the binding cavity contribute to and mediate hydrogen bonding interactions between the protein and substrate.
Hydrogen bonding interactions in the tmMBP maltotetraose binding site
The tmMBP binding site uses a network of polar amino acids to satisfy the hydrogen bonding potential of many of the maltotetraose hydroxyl groups (Fig. 1b,c). Yet, a distinct pattern of hydroxyl group recognition is observed depending on where the carbohydrate ring is situated within the tmMBP binding site. A total of 8 specific hydrogen bonds, defined here as mediated either through direct protein/carbohydrate hydrogen bonds or through directly-bound water molecules, are observed for the first two sugar rings (ring 1/R1 and ring 2/R2) (Fig. 1b). The hydroxyl moieties of the first two sugar rings are all coordinated by specific hydrogen bonds, apart from the R1 O6 hydroxyl. The hydrogen bond formed between Glu111 and the R2 O2 hydroxyl is a short hydrogen bond17, with 2.7 Å between heavy atoms, yet the hydrogen atom is not found to be delocalized from a canonical position. A total of 4 specific hydrogen bonds are formed with the remaining two rings (ring 3/R3 and ring 4/R4) (Fig. 1c). The hydrogen bond formed between Asp66 and the ring 3 O2 hydroxyl is also a short hydrogen bond, with 2.7 Å between heavy atoms, yet it is also a canonical hydrogen bond.
The differential hydrogen bonding pattern of the R1/R2 and R3/R4 subsites suggests the tmMBP has a bipartite binding site. The bulk (8/12 hydrogen bonds) of the specific interactions occur in the “specificity subsite” surrounding the R1/R2 carbohydrate rings, whereas the remaining subsite leaves 4 of the hydroxyl groups with no specific hydrogen bonds (Fig. 1c). The water molecules bound in the specificity subsite are responsible for forming 3 of 8 hydrogen bonds and are thus important in binding of saccharides in tmMBP. The neutron diffraction data allows for directly orienting the water molecules rather than inferring their positions and hydrogen bonding pattern (Fig. 2). Each water molecule forms three hydrogen bonds, two with the binding pocket side chains and one with maltotetraose (Fig. 2). These water molecules all have lower than average b-factors (average of 20.7 Å2) from the remainder of the other solvent molecules (average of 40.5 Å2). Indeed, water three has the lowest b-factor of all solvent molecules. Interestingly, although the canonical MBP, Escherichia coli maltose-binding protein (ecMBP), binds maltodextrins in a shifted register (where tmMBP ring 2 corresponds to the ecMBP ring 1), the amino acids coordinating water 3, and the hydrogen bonds formed, are conserved as in tmMBP (Fig. 2d)18. Moreover, water 3 is also the solvent molecule with the lowest b-factor in ecMBP18. Taken together, these observations highlight the importance of the water network in the tuning the specificity of MBPs in general.
Discussion
Neutron crystallography is a powerful tool for determining hydrogen and water orientations in a protein structure19. Despite the wealth of knowledge underlying the fine-tuning of molecular recognition in the PBP superfamily, the 2.1 Å neutron structure of tmMBP is the first neutron crystal structure of any PBP. The bipartite tmMBP binding sight holds a large number of water molecules. Some of these have a direct role in mediating substrate specificity and are thus important for tight binding and specific recognition of saccharides. This contrasts with the paradigmatic role of the entropic contribution of water molecules to ligand binding affinity. It was recently shown that perturbing the PBP water network can significantly alter substrate-binding affinity, suggesting that binding pocket solvent molecules serve as an evolutionary constraint to maintain and modulate the affinity of substrate interactions6. Indeed the present work demonstrates the conservation of this concept in MBPs and thus reinforces the fundamental importance of water molecules in tuning PBP specificity6,20.
Methods
Expression, purification, and crystallization of tmMBP
For the neutron structure of tmMBP, unlabeled tmMBP was expressed and subjected to H/D exchange as described in the Results section. Unlabeled tmMBP was recombinantly expressed in E. coli and purified as previously described14. Briefly, Enfors minimal media expressed tmMBP was purified in a two-step chromatographic process, consisting of an initial nickel affinity chromatography column followed by size exclusion chromatography. The purified protein was extensively dialyzed in 20 mM Tris pH 7.5 and 40 mM NaCl, concentrated to 25 mg/ml, and crystallized using previously optimized conditions of 24–25% (wt./vol.) PEG 3350, 0.2 M sodium acetate, 0.1 M BisTris pH 5.5 in H2O solution14.
X-ray and neutron diffraction data collection and joint refinement
Neutron diffraction data were collected on the IMAGINE neutron diffractometer at the High Flux Isotope Reactor, Oak Ridge National Laboratory21,22,23. A total of 35 quasi-Laue diffraction images were collected (λ = 2.78–4.50 Å) at room temperature, with exposure time of 15.5 h per frame. Data were collected at two crystals settings, with the crystal rotated 10° on φ between successive images. Diffraction images were processed and integrated using the LAUEGEN package24,25. Wavelength normalization and scaling between Laue diffraction images was performed using LSCALE. The combination of low crystal symmetry and the cylindrical detector array limited overall neutron data completeness to ~ 73%, and to 51% in the highest 2.1 Å resolution shell (Table 1). The same crystal used for neutron diffraction was then used to collect room temperature X-ray diffraction dataset on a Rigaku 007HF diffractometer with an a EIGER R 4 M hybrid photon counting detector. Data were collected using the strategy mode in CryAlisPro (Rigaku. The Woodlands, Texas), with exposure times of 1 s per 0.25° phi-rotation and omega scans of 54–78° at 20 crystal settings. The data were integrated using CryAlisPro and reduced and scaled in AIMLESS26.The X-ray data extended to 1.70 Å and was 100% complete (Table 1). The previously determined structure of tmMBP (PDB code 6DTS) was used for phasing in Phaser15. Joint X-ray and neutron refinement was performed using PHENIX Refine16,27 and manual model building of the single chain found in the asymmetric unit of tmMBP was carried out in COOT28.
Data availability
The X-ray and neutron structure factors and the final refined atomic model for maltotetraose bound tmMBP have been deposited to the Protein Data Bank under accession code 8DHD.
References
Duan, X., Hall, J. A., Nikaido, H. & Quiocho, F. A. Crystal structures of the maltodextrin/maltose-binding protein complexed with reduced oligosaccharides: flexibility of tertiary structure and ligand binding. J. Mol. Biol. 306, 1115–1126 (2001).
Cuneo, M. J., Beese, L. S. & Hellinga, H. W. Ligand-induced conformational changes in a thermophilic ribose-binding protein. BMC Struct. Biol. 8, 50. https://doi.org/10.1186/1472-6807-8-50 (2008).
Magnusson, U. et al. Hinge-bending motion of D-allose-binding protein from Escherichia coli: Three open conformations. J. Biol. Chem. 277, 14077–14084 (2002).
Sharff, A. J., Rodseth, L. E., Spurlino, J. C. & Quiocho, F. A. Crystallographic evidence of a large ligand-induced hinge-twist motion between the two domains of the maltodextrin binding protein involved in active transport and chemotaxis. Biochemistry 31, 10657–10663 (1992).
Vergara, R. et al. The interplay of protein–ligand and water-mediated interactions shape affinity and selectivity in the LAO binding protein. FEBS J. 287, 763–782 (2020).
Darby, J. F. et al. Water networks can determine the affinity of ligand binding to proteins. J. Am. Chem. Soc. 141, 15818–15826 (2019).
Sleigh, S., Seavers, P., Wilkinson, A., Ladbury, J. & Tame, J. Crystallographic and calorimetric analysis of peptide binding to OppA protein. J. Mol. Biol. 291, 393–415 (1999).
Zainol, M. K., Linforth, R. J., Winzor, D. J. & Scott, D. J. Thermodynamics of semi-specific ligand recognition: The binding of dipeptides to the E. coli dipeptide binding protein DppA. Eur. Biophys. J. 50, 1103–1110 (2021).
Kim, M. & Cho, A. E. The role of water molecules in stereoselectivity of glucose/galactose-binding protein. Sci. Rep. 6, 1–10 (2016).
Munshi, P. et al. Molecular details of ligand selectivity determinants in a promiscuous beta-glucan periplasmic binding protein. BMC Struct. Biol. 13, 18. https://doi.org/10.1186/1472-6807-13-18 (2013).
Borrok, M. J., Kiessling, L. L. & Forest, K. T. Conformational changes of glucose/galactose-binding protein illuminated by open, unliganded, and ultra-high-resolution ligand-bound structures. Protein Sci. 16, 1032–1041 (2007).
Fulyani, F. et al. Functional diversity of tandem substrate-binding domains in ABC transporters from pathogenic bacteria. Structure 21, 1879–1888 (2013).
Elias, M. et al. The molecular basis of phosphate discrimination in arsenate-rich environments. Nature 491, 134–137 (2012).
Shukla, S. et al. Differential substrate recognition by maltose binding proteins influenced by structure and dynamics. Biochemistry 57, 5864–5876. https://doi.org/10.1021/acs.biochem.8b00783 (2018).
Adams, P. D. et al. PHENIX: A comprehensive python-based system for macromolecular structure solution. Acta Crystallogr. D Biol. Crystallogr. 66, 213–221 (2010).
Afonine, P. V. et al. Joint X-ray and neutron refinement with phenix. refine. Acta Crystallogr. Sect. D: Biol. Crystallogr. 66, 1153–1163 (2010).
Kumar, P. et al. Low-barrier and canonical hydrogen bonds modulate activity and specificity of a catalytic triad. Angew. Chem. 131, 16406–16412 (2019).
Quiocho, F. A., Spurlino, J. C. & Rodseth, L. E. Extensive features of tight oligosaccharide binding revealed in high-resolution structures of the maltodextrin transport/chemosensory receptor. Structure 5, 997–1015 (1997).
Helliwell, J. R. Fundamentals of neutron crystallography in structural biology. Methods Enzymol. 634, 1–19. https://doi.org/10.1016/bs.mie.2020.01.006 (2020).
Oh, B.-H., Ames, G. & Kim, S.-H. Structural basis for multiple ligand specificity of the periplasmic lysine-, arginine-, ornithine-binding protein. J. Biol. Chem. 269, 26323–26330 (1994).
Meilleur, F. et al. The IMAGINE instrument: First neutron protein structure and new capabilities for neutron macromolecular crystallography. Acta Crystallogr. D Biol. Crystallogr. 69, 2157–2160 (2013).
Meilleur, F., Kovalevsky, A. & Myles, D. A. in Methods in Enzymology Vol. 634 69–85 (Elsevier, 2020).
Meilleur, F., Coates, L., Cuneo, M. J., Kovalevsky, A. & Myles, D. A. The neutron macromolecular crystallography instruments at Oak Ridge National Laboratory: Advances, challenges, and opportunities. Crystals. 8, 388 (2018).
Helliwell, J. R. et al. The recording and analysis of synchrotron X-radiation Laue diffraction photographs. J. Appl. Crystallogr. 22(5), 483–497. https://doi.org/10.1107/S0021889889006564 (1989).
Campbell, J. W., Hao, Q., Harding, M. M., Nguti, N. D. & Wilkinson, C. LAUEGEN version 6.0 and INTLDM. J. Appl. Cryst. 31(3), 496-502. https://doi.org/10.1107/S0021889897016683.
Collaborative, C. P. The CCP4 suite: Programs for protein crystallography. Acta Crystallogr. Sect. D, Biol. Crystallogr. 50, 760 (1994).
Afonine, P. V. et al. Towards automated crystallographic structure refinement with phenix. refine. Acta Crystallogr. Sect. D: Biol. Crystallogr. 68, 352–367 (2012).
Emsley, P. & Cowtan, K. Coot: Model-building tools for molecular graphics. Acta Crystallogr. D Biol. Crystallogr. 60, 2126–2132 (2004).
Acknowledgements
We would like to thank Dr. Flora Meilleur and Dr. Andrey Kovalevsky for assistance with data collection on the IMAGINE beamline. SS would like to thank Dr. Gabriela Schröder for helping with joint neutron-X-ray refinement. We would also like to acknowledge Dr. Kevin Weiss for his assistance with bioreactor runs. The ORNL Center for Structural Molecular Biology (FWP ERKP291) was supported by the Office of Biological and Environmental Research of the U.S. Department of Energy. Research at the SNS and High Flux Isotope Reactor of ORNL was sponsored by the Scientific User Facilities Division, Office of Basic Energy Sciences, U.S. Department of Energy.
Author information
Authors and Affiliations
Contributions
Conceptualization and experimental design: S.S., D.A.M. and M.J.C.; investigation, data analysis and interpretation: S.S., D.A.M. and M.J.C.; manuscript writing, reviewing, and editing: S.S., D.A.M. and M.J.C.
Corresponding authors
Ethics declarations
Competing interests
The authors declare no competing interests.
Additional information
Publisher's note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Rights and permissions
Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article's Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by/4.0/.
About this article
Cite this article
Shukla, S., Myles, D.A. & Cuneo, M.J. Mapping periplasmic binding protein oligosaccharide recognition with neutron crystallography. Sci Rep 12, 17647 (2022). https://doi.org/10.1038/s41598-022-20542-8
Received:
Accepted:
Published:
DOI: https://doi.org/10.1038/s41598-022-20542-8
This article is cited by
-
Introduction to neutron scattering
ChemTexts (2023)
Comments
By submitting a comment you agree to abide by our Terms and Community Guidelines. If you find something abusive or that does not comply with our terms or guidelines please flag it as inappropriate.