Letters pubs.acs.org/acschemicalbiology
Reduction of Lectin Valency Drastically Changes Glycolipid Dynamics in Membranes but Not Surface Avidity Julie Arnaud,† Julie Claudinon,‡,§ Kevin Tröndle,‡,§ Marie Trovaslet,∥ Göran Larson,⊥ Aline Thomas,† Annabelle Varrot,† Winfried Römer,*,‡,§ Anne Imberty,*,† and Aymeric Audfray† †
CERMAV-CNRS (affiliated to Grenoble Université and ICMG), BP53, 38041 Grenoble, France Institute of Biology II, Schänzlestraße 1, Albert-Ludwigs-University Freiburg, 79104 Freiburg, Germany § BIOSSCentre for Biological Signalling Studies, Schänzlestraβe 18, Albert-Ludwigs-University Freiburg, 79104 Freiburg, Germany ∥ Département de Toxicologie, IRBA, 24 av des Maquis du Grésivaudan, 38700 La Tronche, France ⊥ Department of Clinical Chemistry and Transfusion Medicine, Sahlgrenska University Hospital, 41345 Gothenburg, Sweden ‡
S Supporting Information *
ABSTRACT: Multivalency is proposed to play a role in the strong avidity of lectins for glycosylated cell surfaces and also in their ability to affect membrane dynamics by clustering glycosphingolipids. Lectins with modified valency were designed from the β-propeller fold of Ralstonia solanacearum lectin (RSL) that presents six fucose binding sites. After identification of key amino acids by molecular dynamics calculations, two mutants with reduced valency were produced. Isothermal titration calorimetry confirmed the loss of three high affinity binding sites for both mutants. Crystal structures indicated that residual low affinity binding occurred in W76A but not in R17A. The trivalent R17A mutant presented unchanged avidity toward fucosylated surfaces, when compared to hexavalent RSL. However, R17A is not able anymore to induce formation of membrane invaginations on giant unilamellar vesicules, indicating the crucial role of number of binding sites for clustering of glycolipids. In the human lung epithelial cell line H1299, wt-RSL is internalized within seconds whereas the kinetics of R17A uptake is largely delayed. Neolectins with tailored valency are promising tools to study membrane dynamics.
L
membrane bending.9−11 However, the role of glycolipid clustering in the internalization process is still unclear: the reduction of receptor binding sites did not inhibit the cellular entry of cholera toxin 12 and Shiga toxin 10 but reduced enormously the formation of membrane invaginations on cells and giant unilamellar vesicles for this latter protein. There is therefore a need for tailored lectin mutants to analyze the effect of multivalency upon cellular internalization of pathogens. The modular architecture of the β-propeller fold is well adapted for multivalent presentation and, therefore, for the study of the effect of valency. It consists of the repeat of small β-sheet domains referred to as “blade” arranged circularly. βPropellers of animal and fungal lectins are formed by the folding of one polypeptide chain with 5, 6, or 7 blades.13−15 A different β-propeller was identified for bacterial lectins from Ralstonia solanacearum (RSL) 16 and Burkholderia ambifaria (BambL).17 They present a pseudo-6-fold symmetry created by the trimerization of small peptides, each consisting of two blades (Figure 1A). This β-propeller has several advantages compared to the fungal or animal ones: it can be produced in
ectins are sugar-binding proteins that interact with carbohydrates noncovalently and specifically. They possess the ability to act as recognition molecules on cell surfaces, inside cells and in biological fluids. Thus, lectins are unique cellrecognition markers for carbohydrates. While sugars are able to carry biological information, lectins decipher the complex glycocode.1,2 Lectins display a large number of folds and a variety of binding site architectures. They are usually multivalent, either by oligomerization, tandem repetition, or surface presentation. 3 Multivalency is therefore a key component of lectin-carbohydrate interactions in many biological processes such as aggregation and cell surface attachment. Because receptors and ligands are displayed in several copies on the external side of viruses, bacteria, and cells, multivalency creates avidity and is able to enhance usually low unique site affinity by several orders of magnitude.4 Many pathogens use lectins for multivalent attachment on oligosaccharide epitopes on human epithelia.5 Cholera and Shiga toxins exploit plasma membrane glycosphingolipids as receptors to enter and infect cells via endocytic routes that do not involve the elaborate network of clathrin.6−8 These lectins can induce plasma membrane invaginations without the help of the cytosolic protein machinery via the dynamic construction of protein−lipid nanodomains whose intrinsic properties lead to © 2013 American Chemical Society
Received: February 8, 2013 Accepted: July 15, 2013 Published: July 15, 2013 1918
dx.doi.org/10.1021/cb400254b | ACS Chem. Biol. 2013, 8, 1918−1924
ACS Chemical Biology
Letters
Figure 1. (A) Overall structure of hexavalent trimeric RSL complexed with αMeFuc displayed as spheres (PDB code 2BT9). (B) Intramonomeric fucose binding site. (C) Energy contribution for the interaction between fucoside and binding site residues (dist 20) corresponds to 100% for wt-RSL and 70% for R17A. Scale bars: 5 μm.
membrane invaginations on cells and GUVs10 and on cellular internalization.10,12 Reduction of the number of binding sites did not inhibit, rather it attenuated, the internalization of cholera toxin or Shiga toxin in cells but inhibited the formation of membrane invaginations by Shiga toxin on glycolipidcontaining GUVs 10 pointing to different mechanisms of endocytosis in presence or absence of coat proteins. By combining structural with functional studies, we have discovered that both wt-RSL and W76A were able to induce membrane invaginations in giant unilamellar vesicles while the R17A mutant did not. The observation that the trimeric, trivalent R17A mutant, with binding sites that are 30 Å apart, is not able to form membrane invaginations complements nicely the data obtained for the pentameric Shiga toxin.10 Since the Bsubunit of Shiga toxin (StxB) can bind up to 15 receptor molecules and the StxB binding site mutant W37A still up to 10, it is yet extremely difficult to determine how many receptor molecules are indeed necessary for the formation of membrane invaginations and the cellular uptake. It gets more and more obvious that a certain degree of multivalency and/or topology of binding sites is requested in order to induce membrane invaginations. Over the years, many efforts have been made to understand the effect of ligand valency on receptor binding and pathogenicity,28,29 but not much has been invested on the lectin one.20 The hexavalent lectin RSL, in combination with its truly trivalent mutant R17A, represents unique tools for studying the effect of multivalency on synthetic and cellular systems. Using synthetic systems in combination with the trimeric RSL, we have gained new insights regarding the understanding of the role and the mode of action of lectin multivalency. It can be stressed that, on the one hand, the common pentameric protein scaffold of Shiga and cholera toxins, and the VP1 subunit of Simian Virus 40, is not an exclusive feature to promote membrane invaginations. On the other hand, the location of all carbohydrate binding sites on the same face of the protein promotes strong avidity for membrane but is not always sufficient for the formation of membrane invaginations as illustrated by the inactive trivalent mutant. This
There have been limited number of attempts to produce lectin valency mutants by altering the oligomer symmetry.12,20 A key point for tailoring valency is the identification of amino acids crucial for ligand interaction. Our MD calculations on RSL indicated that R17 and W76 are both strongly involved in ligand binding, confirming a previous docking study.21 Interestingly, the W76A mutant has already been characterized in order to quantify the role of aromatic amino acids.22 A decrease in affinity of 3 orders of magnitude (Kd = 1 mM) was observed, indicating that, as demonstrated in our work, some residual activity is conserved in absence of Trp. The influence of multivalency on interactions between endogenous or exogenous lectins and glycosylated cell surfaces is of particular interest for many research fields. Several modes of interactions have been described between lectins and (glyco)lipid rafts. First, multivalent lectins, such as galectins 23 or sialoadhesins,24 colocalize with glycosphingolipid-containing rafts, while norovirus capsids bind to the rim of galactosylceramide domains.25 Then, in vitro studies on model plasma membranes containing the ganglioside GM1 26 or on glycoengineered cells presenting lipid-anchored glycopolymers 27 demonstrated that multivalent lectins are able to redistribute the glycoconjugates by cross-linking and thus trigger the (re)organization of the membrane. More recently, it has been demonstrated that pentameric carbohydrate binding proteins such as the bacterial Shiga toxin and cholera toxin or the simian virus 40 (SV40) capsid bind to glycosphingolipids. They thereby induce plasma membrane curvature and thus directly promote their endocytic uptake into cells.9,10 Our results on the effect of wt-RSL on glycolipid-containing GUVs demonstrated that hexavalent objects, such as β-propeller lectins, are also able to induce membrane invaginations. This has been observed for natural and synthetic glycolipids. However, the clustering activity is somewhat dependent on the lipid backbone since glycosphingolipids resulted in a more effective sequestration of RSL in invagination than carbohydrate-bearing phospholipids. Mutants with reduced numbers of binding sites were previously obtained for Shiga and cholera toxins in order to determine the role of multivalency on the induction of 1922
dx.doi.org/10.1021/cb400254b | ACS Chem. Biol. 2013, 8, 1918−1924
ACS Chemical Biology
■
opens the route to further investigations about deciphering the interplay between the number of binding sites and their overall topologies. The results of the present work are also touching fundamental questions regarding evolution. Indeed, as pointed out previously,12 we need to understand if the evolutionary driving force for lectin multivalency is to enhance avidity for surface binding or to cluster glycosylated receptors for membrane deformation and cellular uptake. These processes might be highly regulated and dependent on cellular and microbial factors such as availability of different cell surface receptors and corresponding lectin affinities. With this first quantitative evaluation of the effect of valency on affinity/ avidity changes and on the formation of membrane invaginations, we can conclude that high avidity can be obtained with a relatively small number of binding sites while glycolipid clustering requires more binding sites and/or different topologies that have to be explored. For future research, engineered neoRSLs with tailored valency ranging from hexavalent to monovalent would be highly appreciated tools to elucidate the molecular mechanisms of binding, membrane invagination, trafficking, and other physiological processes induced by lectins.
■
Letters
ASSOCIATED CONTENT
S Supporting Information *
Details of methods, MD trajectory, DSC analysis, X-ray data collection and refinement statistics, ITC and SPR tables, GUV preparation, and movies of membrane invaginations. This material is available free of charge via the Internet at http:// pubs.acs.org. Accession Codes
Coordinates and structure factors of W76A and R17A mutants complexed with fucoside and fucose have been deposited in the Protein Data Bank, www.pdb.org (PDB ID code 3ZI8 and 4I6S).
■
AUTHOR INFORMATION
Corresponding Author
*E-mail:
[email protected];
[email protected]. Notes
The authors declare no competing financial interest.
■
ACKNOWLEDGMENTS This work has been funded by Agence Nationale de la Recherche Grant NeoLect (ANR-11-BSV5-002). A. Imberty acknowledges support from CNRS and J. Arnaud from Université Grenoble 1. W. Römer is supported by the Excellence Initiative of the German Research Foundation (EXC 294), by a grant from the Ministry of Science, Research and the Arts of Baden-Württemberg (Az: 33-7532.20) and by a starting grant of the European Research Council (Programme “Ideas” No. ERC-2011-StG 282105). The COST actions CM1102 and BM1003 and the Labex ARCANE (ANR-11LABX-003) are thanked for support. The authors are grateful to the SOLEIL synchrotron for provision of synchrotron radiation facilities and the CINES for calculation facilities. We thank S. Vidal for the generous gift of biotinylated fucoside. The technical help from the Proxima 1 beamline staff, E. Gillon, C. Bras, and L. Charlier is also acknowledged.
METHODS
Molecular Dynamics Calculations. A 32 ns trajectory of fully hydrated RSL complexed with αMeFuc was produced using the program NAMD 2.8 running with the Amber99SB and GLYCAM06 force fields 30 (See the SI for details). The sum of nonbonded interaction energy between the fucoside and the residues comprised within a sphere of 8 Å of the ligand was calculated throughout the trajectory, and averaged over the three symmetrically related intramonomeric binding sites. Production of RSL Mutants and Characterization. The pET25b-wt-RSL plasmid 16 was used as a template for site-directed mutagenesis using PCR (SI Table S1). Production wt-RSL and mutants were performed, as previously described.16 Details of production and affinity purification are available in the SI. Purified proteins were dialyzed against ultrapure water and lyophilized. The monodispersity and oligomerization state of the proteins were evaluated by dynamic light scattering at 1 mg mL−1 (Zetasizer Nano series, Malvern Instruments), and thermal stability was evaluated by DSC (see the SI). ITC experiments were performed with an ITC200 isothermal titration calorimeter (Microcal-GE Heathcare). SPR experiments were performed on a Biacore X100 instrument (GE Healthcare). Crystals of R17A/αMeFuc and W76A/fucose were obtained by the hanging drop vapor diffusion method. Details on ITC, SPR, crystallization conditions, data collection, and refinement procedure are given in the SI. Production of GUVs and Lectin Binding. GUVs were composed of DOPC (64.75 mol %), Texas Red DHPE (0.25 mol %), cholesterol (30 mol %), and various glycolipid species (5 mol %) (see the SI for preparation). GUVs were grown at RT using the electroformation technique on indium−tin oxide (ITO)-coated glass slides mostly as described.10,31 Wt-RSL or mutants (5 μg mL−1) were incubated with GUVs at RT and examined under an inverted confocal fluorescence microscope (NIKON A1R) equipped with a 60× oil immersion objective (Plan Apo, N.A. 1.40) using the resonant scanner. Incubation of Human Lung Epithelial Cells with Lectins. H1299 cells were incubated with either wt-RSL-Cy3 or R17A-Cy3 (2 μg mL−1) for 1 h at 37 °C. Cells were then fixed in 4 % paraformaldehyde for 15 min at room temperature (RT). The coverslips were mounted and analyzed by confocal fluorescence microscopy. Fluorescence intensity inside cells (n > 20) was evaluated after 1 h of lectin incubation using the “corrected total cell fluorescence” protocol of the ImageJ software.32
■
REFERENCES
(1) Ambrosi, M., Cameron, N. R., and Davis, B. G. (2005) Lectins: Tools for the molecular understanding of the glycocode. Org. Biomol. Chem. 3, 1593−1608. (2) Arnaud, J., Audfray, A., and Imberty, A. (2013) Binding sugars: From natural lectins to synthetic receptors and engineered neolectins. Chem. Soc. Rev. 42, 4798−4813. (3) Lis, H., and Sharon, N. (1998) Lectins: Carbohydrate-specific proteins that mediate cellular recognition. Chem. Rev. 98, 637−674. (4) Collins, B. E., and Paulson, J. C. (2004) Cell surface biology mediated by low affinity multivalent protein−glycan interactions. Curr. Opin. Chem. Biol 8, 617−625. (5) Imberty, A., and Varrot, A. (2008) Microbial recognition of human cell surface glycoconjugates. Curr. Opin. Struct. Biol 18, 567− 576. (6) Beddoe, T., Paton, A. W., Le Nours, J., Rossjohn, J., and Paton, J. C. (2010) Structure, biological functions, and applications of the AB5 toxins. Trends Biochem. Sci. 35, 411−418. (7) Eierhoff T., Stechmann B., Romer W. (2012) Pathogen and toxin entryHow pathogens and toxins induce and harness endocytotic mechanisms. In Molecular Regulation of Endocytosis (Ceresa B., Ed.) InTech, New York. (8) Johannes, L., and Romer, W. (2010) Shiga toxinsFrom cell biology to biomedical applications. Nat. Rev. Microbiol. 8, 105−116.
1923
dx.doi.org/10.1021/cb400254b | ACS Chem. Biol. 2013, 8, 1918−1924
ACS Chemical Biology
Letters
(9) Ewers, H., Romer, W., Smith, A. E., Bacia, K., Dmitrieff, S., et al. (2010) GM1 structure determines SV40-induced membrane invagination and infection. Nat. Cell Biol. 12, 11−12. (10) Romer, W., Berland, L., Chambon, V., Gaus, K., Windschiegl, B., et al. (2007) Shiga toxin induces tubular membrane invaginations for its uptake into cells. Nature 450, 670−675. (11) Romer, W., Pontani, L. L., Sorre, B., Rentero, C., Berland, L., et al. (2010) Actin dynamics drive membrane reorganization and scission in clathrin-independent endocytosis. Cell 140, 540−553. (12) Jobling, M. G., Yang, Z., Kam, W. R., Lencer, W. I., and Holmes, R. K. (2012) A single native ganglioside GM1-binding site is sufficient for cholera toxin to bind to cells and complete the intoxication pathway. mBio 3, e00401−00412. (13) Beisel, H. G., Kawabata, S., Iwanaga, S., Huber, R., and Bode, W. (1999) Tachylectin-2: Crystal structure of a specific GlcNAc/GalNAcbinding lectin involved in the innate immunity host defense of the Japanese horseshoe crab Tachypleus tridentatus. EMBO J. 18, 2313− 2322. (14) Cioci, G., Mitchell, E. P., Chazalet, V., Debray, H., Oscarson, S., et al. (2006) β-Propeller crystal structure of Psathgyrella velutina lectin: An integrin-like fungal protein interacting with monosaccharides and calcium. J. Mol. Biol. 357, 1575−1591. (15) Wimmerova, M., Mitchell, E., Sanchez, J. F., Gautier, C., and Imberty, A. (2003) Crystal structure of fungal lectin: Six-bladed βpropeller fold and novel recognition mode for Aleuria aurantia lectin. J. Biol. Chem. 278, 27059−27067. (16) Kostlanová, N., Mitchell, E. P., Lortat-Jacob, H., Oscarson, S., Lahmann, M., et al. (2005) The fucose-binding lectin from Ralstonia solanacearum: A new type of β-propeller architecture formed by oligomerisation and interacting with fucoside, fucosyllactose, and plant xyloglucan. J. Biol. Chem. 280, 27839−27849. (17) Audfray, A., Claudinon, J., Abounit, S., Ruvoën-Clouet, N., Larson, G., et al. (2012) The fucose-binding lectin from opportunistic pathogen Burkholderia ambifaria binds to both plant and human oligosaccharidic epitopes. J. Biol. Chem. 287, 4335−4347. (18) Chen, C. K. M., Chan, N.-L., and Wang, A. H. J. (2011) The many blades of the β-propeller proteins: Conserved but versatile. Trends Biochem. Sci. 36, 553−561. (19) Munoz, E. M., Correa, J., Fernandez-Megia, E., and Riguera, R. (2009) Probing the relevance of lectin clustering for the reliable evaluation of multivalent carbohydrate recognition. J. Am. Chem. Soc. 131, 17765−17767. (20) Olausson, J., Astrom, E., Jonsson, B. H., Tibell, L. A., and Pahlsson, P. (2011) Production and characterization of a monomeric form and a single-site form of Aleuria aurantia lectin. Glycobiology 21, 34−44. (21) Mishra, S. K., Adam, J., Wimmerova, M., and Koca, J. (2012) In silico mutagenesis and docking study of Ralstonia solanacearum RSL lectin: Performance of docking software to predict saccharide binding. J. Chem. Inf. Model. 52, 1250−1261. (22) Wimmerova, M., Kozmon, S., Necasova, I., Mishra, S. K., Komarek, J., et al. (2012) Stacking interactions between carbohydrate and protein quantified by combination of theoretical and experimental methods. PloS One 7, e46032. (23) Hsu, D. K., Chernyavsky, A. I., Chen, H. Y., Yu, L., Grando, S. A., et al. (2009) Endogenous galectin-3 is localized in membrane lipid rafts and regulates migration of dendritic cells. J. Invest. Dermatol. 129, 573−583. (24) Vinson, M., Rausch, O., Maycox, P. R., Prinjha, R. K., Chapman, D., et al. (2003) Lipid rafts mediate the interaction between myelinassociated glycoprotein (MAG) on myelin and MAG-receptors on neurons. Mol. Cell Neurosci. 22, 344−352. (25) Bally, M., Rydell, G. E., Zahn, R., Nasir, W., Eggeling, CB, M.E., et al. (2012) Norovirus GII.4 virus-like particles recognize galactosylceramides in domains of planar supported lipid bilayers. Angew Chem., Int. Ed. 51, 12020−12024. (26) Hammond, A. T., Heberle, F. A., Baumgart, T., Holowka, D., Baird, B., et al. (2005) Crosslinking a lipid raft component triggers
liquid ordered−liquid disordered phase separation in model plasma membranes. Proc. Natl. Acad. Sci. U.S.A. 102, 6320−6325. (27) Belardi, B., O’Donoghue, G. P., Smith, A. W., Groves, J. T., and Bertozzi, C. R. (2012) Investigating cell surface galectin-mediated cross-linking on glycoengineered cells. J. Am. Chem. Soc. 134, 9549− 9552. (28) Bernardi, A., Jiménez-Barbero, J., Casnati, A., Castro, C. D., Darbre, T., et al. (2013) Multivalent glycoconjugates as antipathogenic agents. Chem. Soc. Rev. 42, 4709−4727. (29) Gestwicki, J. E., Cairo, C. W., Strong, L. E., Oetjen, K. A., and Kiessling, L. L. (2002) Influencing receptor−ligand binding mechanisms with multivalent ligand architecture. J. Am. Chem. Soc. 124, 14922−14933. (30) Kirschner, K. N., Yongye, A. B., Tschampel, S. M., GonzalezOuteirino, J., Daniels, C. R., et al. (2008) GLYCAM06: A generalizable biomolecular force field. Carbohydrates. J. Comput. Chem. 29, 622− 655. (31) Mathivet, L., Cribier, S., and Devaux, P. F. (1996) Shape change and physical properties of giant phospholipid vesicles prepared in the presence of an AC electric field. Biophys. J. 70, 1112−1121. (32) Schneider, C. A., Rasband, W. S., and Eliceiri, K. W. (2012) NIH Image to ImageJ: 25 years of image analysis. Nat. Methods 9, 671−675.
1924
dx.doi.org/10.1021/cb400254b | ACS Chem. Biol. 2013, 8, 1918−1924
