NMR studies on carbohydrate interactions with DC-SIGN towards a quantitative STD analysis

1771 Pure Appl. Chem. , Vol. 85, No. 9, pp. 1771–1787, 2013. http://dx.doi.org/10.1351/PAC-CON-12-11-14 © 2013 IUPAC, Publication date (Web): 22 July 2013 NMR studies on carbohydrate interactions with DC-SIGN towards a quantitative STD analysis* Cinzia Guzzi, Juan Carlos Muñoz-García, Pedro M. Enriquez-Navas, Javier Rojo, Jesús Angulo, and Pedro M. Nieto‡ Glycosystems Laboratory, Department of Bioorganic Chemistry, Instituto de Investigaciones Químicas (CSIC), Avda. Américo Vespucio 49, Sevilla 41092, Spain Abstract: The recent introduction of saturation transfer difference (STD) NMR has increased the tools for the study of protein–carbohydrate complexes. This is useful when it is combined with transfer nuclear Overhauser enhancement spectroscopy (NOESY) measurement, or when it is interpreted using the expected calculated values of transference, yielding additional, very valuable information for the study of this type of complex. The objective of this work is to cover the advances of the STD technique as exemplified by the investigations of DC-SIGN (dendritic cell-specific ICAM-3 grabbing non-inte- grin) recognition by simple carbohydrates or mimics of them, based on structures containing a terminal mannose or fucose. We also will discuss the methods for quantification of the STD values based on the initial growing rates with the saturation time. Keywords: carbohydrates; lectin; molecular interactions; molecular recognition; NMR. INTRODUCTION Dendritic cells (DCs) are key players in the initial response to pathogens as they are the first participants in the long series of events in host–pathogen interaction leading to activation of specific T cells [1,2]. They are found in epidermal and mucosal tissues and are thus able to quickly recognize new invading pathogens through the identification of pathogen-associated molecular patterns (PAMPs). Apart from the Toll-like receptor family of pattern recognition receptors (PRRs), they also possess C-type lectin receptors (CLRs) dedicated to the specific recognition of pathogen carbohydrate patterns [3]. Among these CLRs, DC-SIGN (dendritic cell-specific ICAM-3 grabbing non-integrin) has attracted a great deal of attention during the past decade. Initially highlighted for its role in HIV transmission to T cells [4], it has then been identified as a PRR hijacked by many other pathogens (e.g., some viruses, bacteria, fungi, and parasites) to escape immune response in their infectious processes [5,6]. More recently, DC-SIGN has also been found to be involved in the modulation of the immune response [7]. DC-SIGN is a type II transmembrane protein with a short cytosolic region, a transmembrane segment, and an extended extracellular domain (ECD) ending in a carbohydrate recognition domain (CRD) located at 320 Å above the cell surface to interact with potential antigens [8]. This ECD is divided into two regions: a neck region, involved in the tetramerization of the receptor constituted by seven and a *Pure Appl. Chem. 85, 1759–1900 (2013). A collection of invited papers based on presentations at the 26th International Carbohydrate Symposium (ICS 2012), Madrid, Spain, 22–27 July 2012. ‡Corresponding author: E-mail: [email protected] half repeating units, and a calcium-dependent CRD, which is responsible for the molecular recognition processes mediated by DC-SIGN [8]. In recent years, many strategies have arisen to block DC-SIGN CRD avoiding its use by pathogen glycoproteins [9–12]. Natural ligands are frequently recognized by DC-SIGN in a multimodal fashion by using not only its tetrameric state but also taking advantage of its organization into clustered patches at the cell membrane [13,14]. Several efforts have been made to apply a multiple ligand presentation platform (polymers, dendrimers, or nanoparticles) [15–18]. We initially demonstrated that dendrimers loaded with multiple copies of mannose were able to inhibit DC-SIGN/gp120 interaction [9]. However, mannose, as a natural ligand, is not specific enough for in vivo practical applications. The design of a ligand with good selectivity and basal affinity is of essential importance. We have been designing new synthetic molecules mimicking natural sugar properties, and we have studied the natural ligands in order to understand the recognition process to improve the characteristics of the ligand [19–22]. In this sense, we have been studying by NMR and saturation transfer difference (STD) the interaction between the monomeric natural ligands and/or the mimetics, and these studies are going to be reviewed here. BASIS OF STD Transference of saturation in transient complexes has been studied since the early 1960s [23,24], but in its present form, STD NMR, was proposed in 1999 by Mayer as a screening technique allowing the detection of the saturation acquired by the free ligand from the complex in the averaged signals in fast exchange dominated by the free ligand [25]. Since then, STD NMR has become a very versatile and robust ligand-observed NMR technique, with growing applications extending its applicability [26–29]. The STD NMR experiment relays in the difference between a 1D NMR experiment in which the receptor is selectively saturated by a low-power radio frequency (on-resonance experiment). In case of binding, the saturation acquired by the ligand signal would be detected by an increase of the corresponding signal in the difference experiment where a reference experiment (off-resonance) is subtracted by the on-resonance. The saturation, which is responsible for the difference between the signals, is accumulated in solution and its magnitude, firstly proportional to the saturation time, is carrying information about the binding constant and kinetics, geometry, and stoichiometry of the complex. Although the experiment is basically the same, the conditions, parameters, and modes of processing can be adjusted according to the application. In addition, a number of new implementations have been described, once the magnetization is transferred to the ligand it can be manipulated by hyphenated experiments [30,31] or by heteronuclear editing [32,33], extending its applications. The primary magnitude obtained from the STD experiment is the absolute STD value numeri - cally expressed as a percentage of the NMR signal relative to the non-saturating spectrum [100 × (Ioff-resonance – Ion-resonance)/Ioff-resonance]. The value can be reprocessed to obtain the relative STD to perform an epitope-mapping analysis by scaling to 100 % relative to the most intense absolute STD signal [30]. If the first one can be interpreted in terms of the existence of transient binding and is useful in the case of library screening [27], the second one has been used to map the interaction epitope, which in principle is dependent on the number and distances of saturated spins of the protein close to each ligand proton in the complex in a dipolar interaction fashion inversely proportional to the sixth power of the distance. Another STD-derived magnitude is the STD amplification factor (STD-AF) corresponding to the percentage of relative STD multiplied by the ligand excess over the receptor [30]. Additionally, the initial growth rate of the STD (STD0) can be measured by fitting the curve to a monoexponential equation (see below) [34]. C. GUZZI et al. © 2013, IUPAC Pure Appl. Chem. , Vol. 85, No. 9, pp. 1771–1787, 2013 1772 STD for epitope mapping Mannose di- and trisaccharides from gp-120 gp-120 is a heavily glycosylated protein that is found on the surface of the HIV envelope and is responsible for the first events that lead to viral entry into the cells and to further infection [35]. This process is mediated by the interaction of the glycan part of the gp120, from the virus envelope, with DC-SIGN. The process is mediated by the interaction of some sugars with a Ca2+ atom, which is present in the binding site of this lectin. This occurs via the coordination with two adjacent hydroxyl groups from mannose- or fucose-type sugar residues [36]. We start the studies by comparing two branched natural trisaccharides inspired by those enclosed in some mannose-containing natural structures (Manα(1-2)[Manα(1-6)]Man, 1, and Manα(1- 3)[Manα(1-6)]Man, 2). Basically, the differences are that the first one is included in the high-mannose (Man9GlcNAc2) structures found in the HIV envelope, while the second one constitutes the repeating unit of the cell-wall lipoarabinomannans, which is also recognized by DC-SIGN. The NMR study was performed recording NOESY experiments in the absence and presence of DC-SIGN EDC (ensuring that DC-SIGN is a tetrameric aggregate), as well as STD NMR experiments for epitope mapping (Fig. 1) [37]. The transfer NOESY experiments [38–40] (Fig. 1) indicate that in those conditions there is binding to CD-SIGN EDC from 1and 2as deduced from the change in sign of the NOE cross-peaks, relative to the free experiments. In spite of the apparent differences between the NOESY experiments of the trimannosides alone and in the presence of DC-SIGN, those are due to the changes in the correlation time, which allows for larger spin diffusion, and are consistent with the same conformation in the bound and free state. The STD at 2.5 s of mixing time data indicated that the larger contact with the receptor is with ring C for trisaccharide 1and for ring D for 2[37]. The second ring is the central one in both cases, while the weaker contacts are observed for the 1-6 linked mannose. The results are consistent with a model of binding in which the terminal non-1-6 linked mannose is the ring that is interacting with the Ca2+ atom, and the central sugar is following, while the terminal residue bound between the 1 and 6 positions is far from the main binding site. All these results allowed us to propose a model of interaction in which the interaction with the Ca2+ atom, which is known to be the key point for complex formation, occurs between two hydroxyl groups in the trans diequatorial 3 and 4 positions of the binding mannose, see Fig. 2. Mannose-containing mimetics We have also shown the potential multimodal binding of mannose derivatives to DC-SIGN (see below) [16,42]. A series of glycomimetics have been prepared by replacing in the mannose series a non-termi- nal residue by a cyclohexyl moiety with two amides with additional substituents capable of additional interactions [19]. After the α1-2 dimannoside scaffold” (or “trimannoside 1scaffold”), a further series of glycomimetics was prepared by replacing the “reducing end mannose” (or “internal mannose”) by a cyclohexyl moiety with two carboxymethyl esters as additional substituents capable of additional interactions [43]. Interestingly, in the case of 3b, the introduction of a triazol moiety had an additional spectroscopic benefit of favoring the dispersion of the cyclohexyl signals in respect to 3a without modifying the binding mode nor the affinity, thus facilitating a more accurate analysis of the STD data, see Fig. 3. © 2013, IUPAC Pure Appl. Chem. , Vol. 85, No. 9, pp. 1771–1787, 2013 Carbohydrate interactions with DC-SIGN 1773 C. GUZZI et al. © 2013, IUPAC Pure Appl. Chem. , Vol. 85, No. 9, pp. 1771–1787, 2013 1774 Fig. 1 Trisaccharide 1(A) and 2(B) with the relative STD values (2 s), (C and D, top) NOESY of the free trisaccharide registered at 500 MHz in 150 mM NaCl, 20 mM TRIS-d6, 4 mM CaCl2at 278 K using a mixing time of 600 ms, (C and D, bottom) transfer NOESY in the same conditions in the presence of 1:50 molar ratio of CD-SIGN EDC, (E and F) STD experiments at 2.5 s of saturation time in the same buffer as (C) and (D) for 1and 2, respectively. Comparing the NOESY of 3b and the transfer NOESY in the presence of DC-SIGN (Fig. 4), both show essentially the same NOE cross-peaks pattern but with the opposite sign, indicating the presence of binding (see interglycosidic NOE sort distances labeled in Fig. 4B and corresponding cross-peaks in Figs 4C and 4D). As the differences in the peak intensities can be attributed to the larger correlation time of the complex, this result is compatible with comparable conformations in both, the free and the bound ligand. © 2013, IUPAC Pure Appl. Chem. , Vol. 85, No. 9, pp. 1771–1787, 2013 Carbohydrate interactions with DC-SIGN 1775 Fig. 2 CRD DC-SIGN complex with an oligomannoside (pdb code: 2it5) showing the interaction of the terminal mannose with the Ca2+ atom [41]. Fig. 3 1H NMR spectra of 3a and 3b, notice the dispersion induced by the triazole moiety attached to the amino group of the linker. The crystallographic structure shown in Fig. 5A was used to calculate the theoretical STD values (C) that were compared with the experimental ones (B). The agreement with the experimental ones (B) is remarkable (see Fig. 5), explaining the high value of the initial growing rate of proton 6ax due to its proximity to Val 351. An additional mimic, a pseudo-trisaccharide, was prepared by introducing an extra mannose residue in the reducing end of the carbohydrate, 4(Fig. 6). The effects on the chemical shift dispersion induced by the triazole were smaller than those found for 4b. As both compounds interact with DC-SIGN with additional strength, they were subject to NMR analysis, and transfer NOE and STD NMR experiments were performed. Most of the intense STD signals were justified by the model structure (Fig. 6C) that resembles the interaction mode of the pseudo-disaccharide 4b. However, in this case some STD signals for the reducing end mannose were observed in spite of the longer distance from the recognition site. Attempts to fit to multiple conformations based on variations of the 1-6 linkage geometry were unsuccessful. Finally, the introduction of a new complex that interacts with the Ca2+ via the opposite mannose, at the reducing end (see Fig. 6D), justifies the results shown in Fig. 6E for a mixture 80–20 %. C. GUZZI et al. © 2013, IUPAC Pure Appl. Chem. , Vol. 85, No. 9, pp. 1771–1787, 2013 1776 Fig. 4 (A) Pseudo-disaccharides 3a and 3b, (B) crystallographic structure of the complex between 3a and DC-SIGN CRD (pdb:2xr5), in dotted lines the NOE distances observed in the (C) NOESY of the free 3b and (D) transfer NOESY of 3b in the presence of DC-SIGN using a mixing time of 200 ms, in the presence of 150 mM NaCl, 20 mM TRIS-d6, 4 mM CaCl2at 278 K and both registered at 500 MHz. © 2013, IUPAC Pure Appl. Chem. , Vol. 85, No. 9, pp. 1771–1787, 2013 Carbohydrate interactions with DC-SIGN 1777 Fig. 5 (A) Crystallographic structure of the complex between 3a and DC-SIGN (pdb: 2xr5) showing the interactions with the calcium atom and between H6 from the carbohydrate and Val 351, (B) experimental, and (C) calculated using CORCEMA-ST, STD growing curves for the complex DC-SIGN EDC and 3a. Experimental conditions were 500 MHz, 150 mM NaCl, 20 mM TRIS-d6, 4 mM CaCl2at 278 K, and 1:50 molar ratio of DC-SIGN. Fig. 6 (A) compound 4b, (B) experimental STD growing curves for 4b in the presence of DC-SIGN ECD, 1:50 molar ratio, (C) X-ray model for trisaccharide 4b (pdb:2xr6), (D) theoretical model corresponding to the inversion and rotation of the X-ray structure (notice that the mannose interacting with the calcium atom is the opposite from C), (E) theoretical STD growing curves corresponding to a mixture of conformations in proportions 80 to 20 (C to D). Fucose-containing derivatives Besides mannose-containing oligosaccharides, many of the natural ligands of DC-SIGN are also oligosaccharides that contain fucose. Binding of fucose derivatives by DC-SIGN, mainly Lewis-type antigen determinants, has been reported by several groups [44–46]. Fucose itself was reported to bind DC-SIGN with dissociation constant about 6 mM [47]. In order to get insight into their binding properties and modes, we have studied some fucose-containing compounds. The study started using a synthetic analogue of LeX, LeXOMe (5). The STD experiment reveals the interaction occurs mainly by the fucose residue (Fig. 7) [48]. A mimetic of LeX was synthesized, preserving the fucose moiety but replacing the glucosamine ring that has been substituted by a cyclohexyl ring and the galactose by a mimetic based on a cyclohexane with one ester and one amide, 6[49]. When the same experiment was performed on 6, we obtained STD signals only for the fucose signals with a relative intensity very related to that obtained for the LeXOMe [49]. Thus, we can assign the same binding mode to both compounds. The structures available from X-ray studies with fucose derivatives display a mode of inter - action that implies the interaction of the fucose ring with the Ca2+ atom, which in this case occurs mainly via hydroxyls in positions 3 and 4 that are in a cis-axial-equatorial disposition and not trans-di- equatorial as in the case of mannose (Fig. 8) [50]. Using this information, a new series of fucose mimics with better affinity was designed and synthesized [22], among them 7and 8were selected for NMR studies. The structure was motivated by the structure of 6, replacing the terminal cyclohexyl ring by an aromatic moiety in order to improve the stability of the complexes by establishing new interactions (CH-πor π–π) with either Val 351 or Phe 313, which are in the proximities. After a conformational study of 7(see Fig. 9 for the minimum energy conformer) the structural analysis of the respective complexes with DC-SIGN (EDC) was performed [48]. One of the main conclusions from this analysis is that, although the fucose ring still has the largest relative STD for the anomeric proton of the fucose, in both compounds the saturation is also transferred to C. GUZZI et al. © 2013, IUPAC Pure Appl. Chem. , Vol. 85, No. 9, pp. 1771–1787, 2013 1778 Fig. 7 Structures of LeXOMe and mimetic 6and STD experiments registered at 500 and 600 MHz, respectively, in the presence of 150 mM NaCl, 20 mM TRIS-d6, 4 mM CaCl2at 278 K using a saturation time of 2 s and 1:50 molar ratio of DC-SIGN. the aromatic ring, indicating the participation of those in the binding to DC-SIGN (Fig. 6) [48]. Nevertheless, the influence of the electronic density on the ring or the possibility of acting as hydrogen bond donor or acceptor seems to not have a particular impact on the structure of the complex as both compounds display similar epitope maps and very similar relative STD values. STD in competition experiments The ability of NMR to detect different signals from each individual nucleus makes it a powerful tool to perform competition experiments. We have performed STD NMR competition experiments in our system using the fucosyl amide 8, LeX5, and α-methyl mannoside, which differ on the chemical shift of the anomeric signal (Fig. 10). In one case, we recorded the STD growing curves for 8, while monitoring the anomeric proton. After we added LeXor mannoside, the same amount (1 mM), recording the © 2013, IUPAC Pure Appl. Chem. , Vol. 85, No. 9, pp. 1771–1787, 2013 Carbohydrate interactions with DC-SIGN 1779 Fig. 8 1sl5 crystallographic structure [50] showing the interaction between the Ca2+ atom and the fucose ring via O3 and O4. Fig. 9 Minimum energy conformation of compound 7showing the diagnostic NOE observed. 27. C. A. Lepre, J. M. Moore, J. W. Peng. Chem. Rev. 104, 3641 (2004). 28. M. Pellecchia, I. Bertini, D. Cowburn, C. Dalvit, E. Giralt, W. Jahnke, T. L. James, S. W. Homans, H. Kessler, C. Luchinat, B. Meyer, H. Oschkinat, J. Peng, H. Schwalbe, G. Siegal. Nat. Rev. Drug Discovery 7, 738 (2008). 29. M. Pellecchia, D. S. Sem, K. Wuthrich. Nat. Rev. Drug Discovery 1, 211 (2002). 30. M. Mayer, B. Meyer. J. Am. Chem. Soc. 123, 6108 (2001). 31. M. Vogtherr, T. Peters. J. Am. Chem. Soc. 122, 6093 (2000). 32. T. Diercks, J. P. Ribeiro, F. J. Cañada, S. André, J. Jiménez-Barbero, H.-J. Gabius. Chem.—Eur. J. 15, 5666 (2009). 33. C. Rauber, S. Berger. Magn. Reson. Chem. 48, 91 (2010). 34. M. Mayer, T. L. James. J. Am. Chem. Soc. 126, 4453 (2004). 35. D. D. Ho, J. C. Kaplan, I. E. Rackauskas, M. E. Gurney. Science 239, 1021 (1988). 36. H. Feinberg, D. A. Mitchell, K. Drickamer, W. I. Weis. Science 294, 2163 (2001). 37. J. J. Reina, I. Diaz, P. M. Nieto, N. E. Campillo, J. A. Paez, G. Tabarani, F. Fieschi, J. Rojo. Org. Biomol. Chem. 6, 2743 (2008). 38. A. A. Bothner-By, J. H. Noggle. J. Am. Chem. Soc. 101, 5152 (1979). 39. J. P. Albrand, B. Birdsall, J. Feeney, G. C. K. Roberts, A. S. V. Burgen. Int. J. Biol. Macromol. 1, 37 (1979). 40. F. Ni. Prog. Nucl. Magn. Reson. Spectrosc. 26, 517 (1994). 41. H. Feinberg, R. Castelli, K. Drickamer, P. H. Seeberger, W. I. Weis. J. Biol. Chem. 282, 4202 (2007). 42. M. Sanchez-Navarro, J. Rojo. Drug News Perspect. 23, 557 (2010). 43. M. Thépaut, C. Guzzi, I. Sutkeviciute, S. Sattin, R. Ribeiro-Viana, N. Varga, E. Chabrol, J. Rojo, A. Bernardi, J. Angulo, P. M. Nieto, F. Fieschi. J. Am. Chem. Soc. 135, 2518 (2013). 44. B. J. Appelmelk, I. van Die, S. J. van Vliet, C. Vandenbroucke-Grauls, T. B. H. Geijtenbeek, Y. van Kooyk. J. Immunol. 170, 1635 (2003). 45. D. A. Mitchell, A. J. Fadden, K. Drickamer. J. Biol. Chem. 276, 28939 (2001). 46. S. Meyer, E. van Liempt, A. Imberty, Y. van Kooyk, H. Geyer, R. Geyer, I. van Die. J. Biol. Chem. 280, 37349 (2005). 47. E. P. Mitchell, C. Sabin, L. Snajdrova, M. Pokorna, S. Perret, C. Gautier, C. Hofr, N. Gilboa- Garber, J. Koca, M. Wimmerova, A. Imberty. Proteins: Struct. Funct. Bioinf. 58, 735 (2005). 48. C. Guzzi, J. Angulo, F. Doro, J. J. Reina, M. Thepaut, F. Fieschi, A. Bernardi, J. Rojo, P. M. Nieto. Org. Biomol. Chem. 9, 7705 (2011). 49. G. Timpano, G. Tabarani, M. Anderluh, D. Invernizzi, F. Vasile, D. Potenza, P. M. Nieto, J. Rojo, F. Fieschi, A. Bernardi. ChemBioChem 9, 1921 (2008). 50. Y. Guo, H. Feinberg, E. Conroy, D. A. Mitchell, R. Alvarez, O. Blixt, M. E. Taylor, W. I. Weis, K. Drickamer. Nat. Struct. Mol. Biol. 11, 591 (2004). 51. A. T. Neffe, M. Bilang, I. Gruneberg, B. Meyer. J. Med. Chem. 50, 3482 (2007). 52. M. Pickhardt, G. Larbig, I. Khlistunova, A. Coksezen, B. Meyer, E. M. Mandelkow, B. Schmidt, E. Mandelkow. Biochemistry 46, 10016 (2007). 53. Y. Cheng, W. H. Prusoff. Biochem. Pharmacol. 22, 3099 (1973). 54. J. Angulo, P. M. Enriquez-Navas, P. M. Nieto. Chem.—Eur. J. 16, 7803 (2010). 55. P. M. Enríquez-Navas, M. Marradi, D. Padro, J. Angulo, S. Penadés. Chem.—Eur. J. 17, 1547 (2011). 56. V. Jayalakshmi, N. R. Krishna. J. Magn. Reson. 155, 106 (2002). 57. N. Rama Krishna, V. Jayalakshmi. Prog. Nucl. Magn. Reson. Spectrosc. 49, 1 (2006). 58. A. Bhunia, V. Jayalakshmi, A. J. Benie, O. Schuster, S. Kelm, N. R. Krishna, T. Peters. Carbohydr. Res. 339, 259 (2004). 59. J. Angulo, B. Langpap, A. Blume, T. Biet, B. Meyer, N. R. Krishna, H. Peters, M. M. Palcic, T. Peters. J. Am. Chem. Soc. 128, 13529 (2006). C. GUZZI et al. © 2013, IUPAC Pure Appl. Chem. , Vol. 85, No. 9, pp. 1771–1787, 2013 1786 60. V. Jayalakshmi, T. Biet, T. Peters, N. R. Krishna. J. Am. Chem. Soc. 126, 8610 (2004). 61. V. Jayalakshmi, N. R. Krishna. J. Magn. Reson. 168, 36 (2004). 62. V. Jayalakshmi, T. Biet, T. Peters, N. R. Krishna. J. Am. Chem. Soc. 127, 7261 (2005). 63. V. Jayalakshmi, N. R. Krishna. J. Am. Chem. Soc. 127, 14080 (2005). 64. A. Canales, J. Rodriguez-Salarichs, C. Trigili, L. Nieto, C. Coderch, J. M. Andreu, I. Paterson, J. Jimenez-Barbero, J. F. Diaz. ACS Chem. Biol. 6, 789 (2011). 65. M. G. Szczepina, D. W. Bleile, B. M. Pinto. Chem.—Eur. J. 17, 11446 (2011). 66. Y. Yuan, D. W. Bleile, X. Wen, D. A. R. Sanders, K. Itoh, H. W. Liu, B. M. Pinto. J. Am. Chem. Soc. 130, 3157 (2008). 67. J. Angulo, I. Diaz, J. J. Reina, G. Tabarani, F. Fieschi, J. Rojo, P. M. Nieto. ChemBioChem 9, 2225 (2008). 68. Y. Xu, I. P. Sugar, N. R. Krishna. J. Biomol. NMR 5, 37 (1995). 69. N. R. Krishna, D. G. Agresti, J. D. Glickson, R. Walter. Biophys. J. 24, 791 (1978). © 2013, IUPAC Pure Appl. Chem. , Vol. 85, No. 9, pp. 1771–1787, 2013 Carbohydrate interactions with DC-SIGN 1787