Skip to main content
NCBI home page
iScience logoLink to iScience
. 2020 Jun 2;23(6):101231. doi: 10.1016/j.isci.2020.101231

Unveiling Molecular Recognition of Sialoglycans by Human Siglec-10

Rosa Ester Forgione 1, Cristina Di Carluccio 1, Juan Guzmán-Caldentey 2, Rosa Gaglione 1, Filomena Battista 1, Fabrizio Chiodo 3, Yoshiyuki Manabe 4, Angela Arciello 1, Pompea Del Vecchio 1, Koichi Fukase 4, Antonio Molinaro 1,4, Sonsoles Martín-Santamaría 2, Paul R Crocker 5, Roberta Marchetti 1,6,∗, Alba Silipo 1,∗∗
PMCID: PMC7306591  PMID: 32629603

Summary

Siglec-10 is an inhibitory I-type lectin selectively recognizing sialoglycans exposed on cell surfaces, involved in several patho-physiological processes. The key role Siglec-10 plays in the regulation of immune cell functions has made it a potential target for the development of immunotherapeutics against a broad range of diseases. However, the crystal structure of the protein has not been resolved for the time being and the atomic description of Siglec-10 interactions with complex glycans has not been previously unraveled. We present here the first insights of the molecular mechanisms regulating the interaction between Siglec-10 and naturally occurring sialoglycans. We used combined spectroscopic, computational and biophysical approaches to dissect glycans' epitope mapping and conformation upon binding in order to afford a description of the 3D complexes. Our outcomes provide a structural perspective for the rational design and development of high-affinity ligands to control the receptor functionality.

Subject Areas: Biochemistry, Biochemistry Methods, Structural Biology, Data Analysis in Structural Biology

Graphical Abstract

graphic file with name fx1.jpg

Open in a new tab

Highlights

  • •

    We unveiled the molecular basis of sialoglycans recognition by Siglec-10

  • •

    The conformation of sialoglycans drives the interaction with the protein

  • •

    Siglec-10 is able to recognize and bind complex N-glycans

  • •

    Our outcomes may open the venue for the design and development of novel glycomimetics

Biochemistry; Biochemistry Methods; Structural Biology; Data Analysis in Structural Biology

Introduction

Siglecs (sialic acid-binding immunoglobulin-like lectins) constitute a family of cell surface immunomodulatory receptors belonging to the immunoglobulin (Ig) superfamily that act as critical regulators of immune system (Crocker et al., 2007, Duan and Paulson, 2020, Macauley et al., 2014, Läubli and Varki, 2019) through the specific recognition of sialic acids, hallmarks of vertebrate self-glycans. Their ability to control the immune balance, triggering tolerogenic or immunogenic responses, explains why aberrant sialic acid-Siglec interactions are linked to a broad spectrum of pathologies including infection, autoimmunity, and cancer (Bochner and Zimmermann, 2015, Lübbers et al., 2018, Adams et al., 2018). Furthermore, feared human pathogens evolved the ability to shield their envelope glycans with sialic acid, mimicking Self Associated Molecular Patterns (Liu et al., 2017, Chang and Nizet, 2014) and exploiting the interaction with inhibitory Siglecs to escape host immune surveillance. In this context, the Siglec-sialoglycan axis is, therefore, an emerging attractive therapeutic target to prevent or affect the course of several immune (cancer, autoimmune, and infectious) diseases.

Siglecs are mainly expressed on innate and adaptive immune cells and display a broadly similar domain organization. These single-pass transmembrane proteins possess an extracellular N-terminal V-set domain, involved in the recognition of specific sialoglycans, and most of them encompass, in their cytoplasmic tails, one or more immunoreceptor tyrosine-based inhibitory motifs (ITIMs), involved in negative intracellular signaling pathways (Crocker et al., 2007, Duan and Paulson, 2020, Macauley et al., 2014).

Siglec-10 is an inhibitory receptor belonging to the CD33-related Siglecs family (Whitney et al., 2001) expressed on dendritic cells, human leukocytes, and B cells (Chen et al., 2009, Bandala-Sanchez et al., 2018, Bandala-Sanchez et al., 2020). It is characterized by five extracellular domains and three tyrosine-based motifs in its cytoplasmic tail, exhibits only 40%–48% homology to other members of CD33-related Siglecs group, and furthermore can bind with high affinity to α-2,6- or α-2,3-sialoglycans as well as to GT1b ganglioside (Rapoport et al., 2003). Siglec-10 is the human homolog of murine Siglec-G, which is known to regulate the tolerance of the B1 sub-class of B cells, dampening autoimmune responses upon sialoglycans recognition, although with yet unknown molecular mechanisms (Chen et al., 2014). Moreover, the leading cause of bacterial gastroenteritis worldwide and a major causative trigger of the Guillain-Barré syndrome, Campylobacter jejuni (Phongsisay, 2016), promotes an anti-inflammatory response mediated by its recognition by Siglec10 through the pseudaminic acid (sugar chemically related to sialic acid) decorating the bacterial flagellum (Stephenson et al., 2014). On the other hand, the interaction of Siglec-10 with the heavily glycosylated, mucin-like cell surface protein CD24, exposed on a variety of tumor cells, is an innate immune checkpoint essential for mediating anti-tumor immunity (Barkal et al., 2019, Zhang et al., 2015). Interestingly, expression of CD24 and Siglec-10 has also been observed at the fetal-maternal interface during the first 3 months of pregnancy, suggesting a possible role of Siglec-10 in establishing the immune tolerance in the placenta between the fetus and the mother (Sammar et al., 2017).

To decipher the complex mechanisms regulating host immune suppression following sialoglycan recognition by Siglec-10, the detailed knowledge of the molecular recognition and binding events represents a mandatory step. In turn, these may be translated into highly specific novel diagnostics and therapeutics, able to exploit Siglec-10 anti-inflammatory pathways for the modulation of immune cellular responses. In this context, and given the lack of any structural knowledge on Siglec-10, we used here a multidisciplinary strategy, combining spectroscopic, computational, and biophysical approaches, providing the first insights into the molecular mechanisms regulating Siglec-10 recognition of cognate (naturally occurring) sialyloglycans, demonstrating the critical role played by the glycans' conformation.

Results

Here, we tackled the important question of how Siglec-10 recognizes and binds, at a molecular level, different sialylated glycans, including 3′SLN and 6′SLN, 1 and 2 respectively, as well as naturally occurring complex-type N-glycans (Figure S1). By coupling spectroscopic approaches (such as NMR, fluorescence spectroscopy) to biochemical and computational methodologies (such as homology modeling, docking, molecular dynamics, and CORCEMA-ST calculations) we have obtained and validated 3D models of ligand-protein complexes. We therefore provide the dynamic characterization of the molecular interactions occurring in solution between Siglec-10 and cognate sialoglycans, mapping their binding epitopes and defining their bioactive conformation.

Siglec-10 Binding Specificities

The binding specificities of Siglec-10 were primarily evaluated by a solid-phase enzyme immunoassay, ELISA, and fluorescence analysis, using a recombinant soluble form of the protein in which the full extracellular region was fused to the Fc region of human IgG1. As depicted in Figure 1A, the ability of Siglec-10 to recognize terminal α2,6- and α2,3-linked sialic acid on various sialoglycans was demonstrated in ELISA experiments, indicating that Siglec-10 similarly recognized all the sialylated conjugates tested with a significant optical density (OD) compared with the negative control (PBS). Analogously, the fluorescence titration of increasing amounts of sialoglycans into a fixed concentration of Siglec-10 indicated that, regardless of the nature of the Neu5Ac-Gal linkage, both α2,3- or α2,6-linked sialoglycans could interact with the protein with binding constants in the micromolar range (Figure 1B).

Figure 1.

Figure 1

Open in a new tab

Siglec-10 Binding to Sialoglycans

(A) Siglec-10 binding to sialylated conjugates detected by ELISA. ELISA plates were coated with synthetic conjugates carrying terminal sialic acids with α2,3 or α2,6 linkages, followed by the addition of a human Fc chimera of Siglec-10. The detection of the interaction was evaluated with an anti-Human IgG-HRP (horseradish peroxidase) conjugate followed by substrate addition. The experiment has been performed three times in duplicate, showing similar results. Error bars indicate standard deviations of a representative experiment performed in duplicate. Data are normalized on BSA-treated wells used as blocking agent.

(B) Fluorescence titration. Fluorescence spectra of Siglec-10 (black lines) and in the presence (colored lines) of increasing amounts of the ligand sialoglycan 1 (upper panel) and 2 (lower panel), respectively. The insets report the binding isotherm and the values of the binding constants (Kb) obtained as described in the text.

Molecular Basis of Sialoglycans Recognition by Siglec-10

A combination of ligand-based NMR techniques and computational studies was employed to evaluate the epitope mapping (group epitope mapping) and the bioactive conformation of different sialoglycans (Figure S1) upon binding to Siglec-10.

Group Epitope Mapping and Bioactive Conformation of 1

The saturation transfer difference (STD) NMR analysis (Marchetti et al., 2016, Meyer and Peters, 2003, Mayer and Meyer, 1999) (Figure 2) conducted on 3′-Sialyl-N-acetyllactosamine 1 in the presence of Siglec-10 confirmed the binding and allowed a detailed description of the structural elements of 1 when accommodated into the protein binding pocket (Figure 2A). The comparison of the STD NMR spectrum with the corresponding reference (off-resonance) showed that only some protons of ligand 1 exhibited STD enhancements. STD effects were calculated by use of (I0 – Isat)/I0, where (I0 – Isat) is the intensity of the relative signal in the STD NMR spectrum and I0 is the peak intensity of an unsaturated reference spectrum (off-resonance). Siglec-10 gives the maximum saturation transfer to the acetyl group protons belonging to K residue of trisaccharide 1, so this STD signal was set at 100% and the other ligand 1 protons percentages were derived accordingly. The sialic acid residue (K) resulted essential for the interaction with Siglec-10, giving the most significant contribution to the molecular recognition process. Indeed, the N-acetyl group of K showed the highest STD signal, followed by the protons signals H-6 and H-7 affected by a rather high saturation transfer, close to 80%. Significant STD enhancements, around 60%, were also observed for protons H-5, H-8 and H-9, indicating their strong involvement in the interaction, whereas weaker STD signals were observed for the diastereotopic H-3 protons of K and for some protons of the galactose unit B. Conversely, the N-acetylglucosamine residue (A) almost did not contribute to the binding process, as confirmed by the absence of any corresponding STD signal. We have also constructed the STD build-up curves (Angulo and Nieto, 2011, Yan et al., 2003) (Figure 2C, Table S1) to overcome possible artifacts due to different protons T1 relaxation times and thus to precisely map the interacting epitope of ligand 1. The strong involvement of K in the binding was confirmed, as well as the orientation of B pointing toward the protein binding site.

Figure 2.

Figure 2

Open in a new tab

Analysis of the Group Epitope Mapping and Bioactive Conformation of Trisaccharide 1 upon Binding with Siglec-10

(A) STD-derived epitope mapping on the molecular envelope of ligand 1 in its bioactive minus gauche conformation.

(B) Reference 1H NMR spectrum (black) and 1D STD NMR spectrum (red) of the Siglec-10/ligand 1 mixture (molecular ratio 1:100, T = 298 K).

(C) STD build-up curves were derived by the monoexponential equation STD = STDmax[1 – exp(–ksatt)], constructed by fitting the STD factors at different saturation times, from 1 to 5 s. In the exponential curve, STD represents the observed signal intensity, STDmax is the asymptotic maximum of the curve, ksat is the observed saturation rate constant, and t is the saturation time.

(D) Selected NOE build-up curves of ligand 1 in the free and bound states by selective excitation of H-3 of Gal residue.

The determination of the epitope map of trisaccharide 1 (group epitope mapping) was complemented with the analysis of the conformational behavior of 1 in its free state and upon binding with Siglec-10. The accessible conformational space of 1 was accomplished by molecular mechanics (MM) and molecular dynamics (MD) simulations, together with nuclear overhauser effect (NOE)-derived experimental data, to define the preferred torsional angles ligand 1 adopted in the free and bound states. In the free state, a conformational equilibrium was observed between three main conformers, namely, g–/g+/t, differing by the value of the Φ (C1-C2-O-C3’) torsion angle between Sia and Gal residues (−60°/60°/180° data not shown) (Poppe et al., 1997, Forgione et al., 2020). It is worth noting that, in terms of topology, all these conformers adopt a rather extended conformation in which the sugar units fill a cone-like region of space (Xu et al., 2009) governed by a value of θ higher than 110°, where θ represents the angle between the carbon C-2 of Neu5Ac residue and C-1 atoms of the subsequent Gal and GlcNAc units (Chandrasekaran et al., 2008). During the MD simulation, ligand 1 mainly explored the g– (Φ = −60°) and t (Φ = 180°) conformations; conversely, the Ψ torsion angle (C2-O-C3’-H3’) remained stable along the dynamics around an average value of −11°. MD simulation results were supported by nuclear overhauser effect spectroscopy (NOESY) NMR analysis (Marchetti et al., 2016, Poveda and Jiménez-Barbero, 1998), confirming the existence of an equilibrium between different conformations of 1 in its free state (Table 1).

Table 1.

Conformational Behavior of 1—Key Intra-Molecular Neu5Ac-Gal Distances

Distance
1H Neu5Ac–1H Gal		 Exp. Free State Exp. Bound State Conformer g-
Φ = −60° Ψ = −11°		 Conformer g+
Φ = 60° Ψ = −11°		 Conformer t
Φ = 180° Ψ = −11°
H3ax Neu5Ac -H3 Gal 2.86 3.90 4.10 3.11 2.18
H3eq Neu5Ac -H3 Gal 3.65 – 4.40 2.03 3.40
H8 Neu5Ac -H3 Gal 3.41 3.19 3.36 6.17 4.31
Open in a new tab

The proton-proton distances were obtained by the sel-NOE experiments, acquired at different mixing times irradiating the B3 proton, using B1-B3 as reference.

As for the bound state, transferred-NOESY (tr-NOESY) NMR experiments were acquired to highlight changes in the conformation of 1 when interacting with Siglec-10. The analysis of tr-NOESY demonstrated a selection of the g– conformer (Φ/Ψ torsion angles of −60°/−11°) of 1 upon binding. In particular, in the tr-NOESY spectrum we observed a decrease of the cross-peak intensity of B3-K3ax, an increase of B3-K8, and the absence of B3-K3eq cross peak (Figure S2, left panel). The crucial inter-residual 1H–1H distances (Table 1) were extracted by construction of NOE build-up curves, clearly revealing a preference for the g– conformation of 1 when bound to Siglec-10 (Table 1, Figures 2C and 2D).

Group Epitope Mapping and Bioactive Conformation of 2

The binding profile of 6′-sialyl-N-acetyllactosamine 2 (Figure S1) by Siglec-10 was investigated by means of STD NMR (Figure 3); changes in the relative enhancements and differences in the multiplicity of STD signals were detected. The epitope mapping of ligand 2 highlighted the protons specifically involved in the interaction with Siglec-10. The strongest STD signal belonged to the N-acetyl group of non-reducing Neu5Ac, K residue, providing an STDepitope fit value of 100% (Table S2). As for the other proton signals belonging to sialic acid, only H-6 and H-7 showed high STD effects, thus pointing significantly close to the receptor-binding pocket, whereas the other protons exhibited STD values around or below 30%. Furthermore, the galactose moiety (B) of 2 received a significantly weaker magnetization transfer from the protein, corresponding to low STD values, and thus seemed less involved in the interaction with Siglec-10 if compared with 1. The N-acetylglucosamine (A) of 2 revealed no STD enhancements, suggesting that it was likely solvent exposed.

Figure 3.

Figure 3

Open in a new tab

Analysis of the Group Epitope Mapping and Bioactive Conformation of Trisaccharide 2 upon Binding with Siglec-10

(A) Reference 1H NMR spectrum (black) and 1D STD NMR spectrum (red) of the Siglec-10/ligand 2 mixture (molecular ratio 1:150, T = 298 K).

(B) STD build-up curves derived by fitting the experimental data to the monoexponential equation STD = STDmax[1 – exp(–ksatt)].

(C) STD-derived epitope mapping on the molecular envelope of ligand 2 in its bioactive bent gt conformation.

The conformational behavior of 2 in its free and bound states was determined by combining NMR and computational techniques. Here, the Neu5Ac-α-(2-6)-Gal glycosidic linkage was defined by an extra torsion angle (Poppe et al., 1992), namely, ω (O′6−C′6−C5’−O′5), that influences the entire three-dimensional structure of 2. In detail, three different rotamers differing in the ω value can coexist in solution, namely, gg/tg/gt (Patel et al., 2014) (ω −60°/180°/60°, respectively), the gt conformer being the most populated in the free state (Di Carluccio et al., 2019). Comparing the experimental NOE-derived and theoretical distances (Table 2, Figure S2 right panel), a preference was evident for the geometry with φ/ψ/ω torsion angles of −60°/180°/60° around Sia-Gal linkage (Family I, gt conformer) (Poppe et al., 1992) both in the free and bound states. The gt preference of 2 corresponded to a bent conformation that implicates the back-folding of the sialic acid residue and was characterized by an open umbrella-like topology (θ angle below 110° [Xu et al., 2009]) as supported by the small NOE contact between H-5 of sialic acid residue and the methyl group of N-acetylglucosamine (Figures 3C and S2 right panel).

Table 2.

Conformational Behavior of 2—Key Intra-Molecular Distances

Distance Exp Free State Exp Bound State Family I
Φ = −60° Ψ = 180° Ω = 60°		 Family II
Φ = 180° Ψ = 180° Ω = 60°
H1 Gal -H4 GlcNAc 2.52 2.50 2.40 2.40
H3eq Neu5Ac -H3 Gal / / 4.93 3.84
H8 Neu5Ac -H3 Gal / / 4.58 3.37
H3eq Neu5Ac -H3 Gal 4.20 4.30 4.43 2.53
H8 Neu5Ac -H3 Gal 4.10 4.30 4.25 2.35
H5 Neu5Ac -CH3GlcNAc 4.00 4.00 4.30 9.60
Open in a new tab

Intra- and inter-residual proton-proton distances extracted from the results of the MD simulations and translated into NOE contacts according to a full-matrix relaxation approach. The corresponding average distances obtained for the simulation from <r−6> values were compared with those estimated from the NOESY/tr-NOESY NMR experimental results.

Homology Modeling of Siglec-10

Since the crystal structure of Siglec-10 has not been solved yet, we opted to predict the protein 3D structure by means of computational methods, with the aim to describe a 3D model of Siglec-10/sialoglycans complexes. The homology model of the CRD (carbohydrate recognition domain) of Siglec-10 was derived by using Siglec-8 (PDB–ID: 2N7A), Siglec-7 (PDB-ID: 2HRL), Siglec-5 (PDB-ID: 2ZG2), and Siglec-3 (PDB-ID: 5IHB) as structural templates (see Supplemental Information for further details). The models were computed with different homology modeling servers and then refined by MD simulations as implemented within the Amber 14 package (Case et al., 2014) (Figures S3 and S4). The structural models of Siglec-10 displayed high similarity along with the corresponding templates. By comparing such models, the characteristic fold of the V-set domain of Siglecs CRD was observed, even though some differences occurred in the conformation of the variable region of Siglecs, namely, the inter-strand CC′ loop (Figure S4).

To select the most accurate model, a detailed analysis was performed, focusing the attention on the orientation of the binding-site residues and the flanking loops. Four promising models (Figure S4) were subsequently used for docking calculations with sialoglycans by means of Autodock 4.2 (Morris et al., 2009). Selected ligand-protein complexes were then submitted to MD simulations to monitor the stability in solution (Figures S4B and S4C). The most consistent protein/ligand complexes obtained from docking and MD were subjected to CORCEMA-ST (complete relaxation and conformational exchange matrix) calculations (Jayalakshmi and Krishna, 2002), a program that evaluates the predicted STD-NMR intensities for a proposed molecular model of a ligand-receptor complex. In order to validate the given complexes, the predicted fractional intensity changes for different ligand protons were compared with the experimental STD-NMR data through a normalized root-means square deviation (RMSD) value, also known as the NOE R factor. Thus, docking, MD simulations, and CORCEMA-ST analysis permitted the selection of MODEL IV as the optimal Siglec-10 model, predicted by using Siglec-7 structural template (Figure S4). It is worth to note that the selected sialoglycans/Siglec-10 complexes exhibited not only the best agreement between theoretical and experimental STD data but also the highest stability along the simulations.

3D-View of Siglec-10 in Interaction with 1

A three-dimensional view of the Siglec-10/trisaccharide 1 complex was achieved by computational studies. Docking calculations were performed using trisaccharide 1 in each of its three possible conformations (g-, g+, t). Selected docked poses were then submitted to MD simulations and subjected to CORCEMA-ST calculations. The best fit between predicted and experimental data, corresponding to the lowest NOE R factor value (0.29), was given by the MODEL IV/1 complex reported in Figure 4 (upper panel), carrying the trisaccharide 1 in the extended g- conformation, denoting good agreement between experimental and predicted STD data. Of note, the profiles of the theoretical STD data achieved for the other two models, bearing the trisaccharide 1 in g and t conformation, respectively, were far from the experimental STD effects, resulting in higher NOE R factor values and suggesting that these models should be excluded (Figure S6). The best selected Siglec-10/ligand 1 model (Figure 4, upper panel) displayed good stability along the MD simulations, further validating the preference of the ligand for the g- conformation in the bound state, as suggested by the stability of the Φ glycosidic torsion angle between the sialic acid and the galactose moiety during most of the simulation time (see Supplemental Information).

Figure 4.

Figure 4

Open in a new tab

Docking/MD and CORCEMA-ST Analysis of 1 (Upper Panel) and 2 (Lower Panel) in the Interaction with Siglec-10

(A) Best binding pose (STD, tr-NOESY, and MD based) for the docking of the ligand bound form into binding site of Siglec-10 model derived by using Siglec-7 as structural template. The carbohydrate and interacting amino acids are shown as sticks. Dashed black lines represents hydrogen bonds.

(B) Two-dimensional plots representing the interactions between the ligand and the residues of the binding site of Siglec-10 model. Dotted arrows indicate hydrogen bonds with functional groups from side chains and solid arrows those with functional groups of the backbone. The residues shown are close to the ligand and participate in hydrophobic and polar interactions.

(C) Comparison between experimental (dashed line) and theoretical (solid line) STD data for the best model of ligand bound to the structural model of Siglec-10 CRD derived by CORCEMA-ST analysis.

Examination of the binding mode of 1 by Siglec-10 (model IV) revealed the typical architecture of Siglecs binding site, characterized by a solvent-exposed, highly positively charged cavity constituted by the C, F, and G β strands and surrounded by the CC' and GG' loops (Figure 4 upper panel and Figure S4). The Neu5Ac residue was placed at the top side of the G-strand, making several polar interactions with the receptor amino acids. Such interactions established by the Neu5Ac moiety were previously reported to be crucial for recognition of sialylated glycans by other CD33-related Siglecs (Crocker et al., 2007, Attril et al., 2006, Attril et al., 2008, Pröpster et al., 2016). During the 100ns MD simulation of the model IV/1 complex, it was possible to observe high stability of the complex until 50 ns, without applying any restriction (Figure S5 upper panel). Alongside the simulation, ligand-protein interaction analysis identified relevant polar contacts between the trisaccharide 1 Neu5Ac unit and the residues located into the binding site of Siglec-10, namely, Arg-119, Arg-127, Asn-129, along with hydrophobic interactions involving Phe-21 and Tyr-128. In detail, the sialic acid carboxylate established a stable ionic interaction with the ammonium group from the highly conserved Arg-119. Furthermore, the hydroxyl groups of the glycerol chain of Neu5Ac formed hydrogen bonds with Asn-129 and Arg-20 residues of the protein. Also, a polar interaction between the acetamide moiety of Sia and Arg-127 residue was found. The H-bonds between the OH group at C-8 position of 1 and Asn-129 amide, and the OH group at C-9 and Arg-119 CO group were maintained along most of the simulation. The NH group of Neu5Ac acetamide moiety formed a hydrogen bond with the Arg-127 backbone, which was steadily held along the simulation. Additionally, Tyr-128 established key CH-π interactions with the C-7 and C-9 hydrogens of sialyl glycerol moiety. The only buried portion of 1 was the methyl group of Neu5Ac unit, which laid in a groove formed by Phe-21 and Tyr-128, thereby establishing CH-π interactions. In the complex, the Gal residue formed polar contacts with the binding-site residues of the protein, namely, Asn-129 and Glu-66, whereas the GlcNAc residue pointed far from the receptor surface, as confirmed by STD NMR analysis (Figure 4 upper panel and Figure S5 upper panel).

According to CORCEMA-ST analysis (Figure 4C upper panel), the protons belonging to the Neu5Ac units exhibited the highest STD effects, confirming that Neu5Ac was the major determinant of the interaction with Siglec-10. The strong STD value observed for the methyl group of the acetamide moiety of sialic acid corresponded to the close contacts between this moiety and Phe-21 and Tyr-128 side chains. Furthermore, the stable hydrogen bond established between acetamide moiety and Arg-127 backbone reinforced this interaction. Regarding STD data on the sialic acid glycerol side chain, the higher STD effect observed for the H-7 in comparison with H-8 and H-9 was ascribable to the CH-π interaction between H-7 and Tyr-128. Although the hydroxyl groups at positions 8 and 9 made direct hydrogen bonds with the backbone of Asn-129, the protons H-8 and H-9S pointed away with respect to the Tyr-128 group, consequently leading to an STD effect lower than the other Neu5Ac protons. Despite the contacts between the CC′ loop of the protein with the galactose ring in trisaccharide 1 bound conformation, weaker STD effects were predicted for the Gal protons. In addition, the STD value estimated for the N-acetylglucosamine moiety were negligibly small, thus confirming that the residue was not involved in the interaction with the protein (Figure 4 upper panel).

3D-View of Siglec10 in Interaction with 2

To get a three-dimensional perspective of the interaction of trisaccharide 2 with Siglec-10 CRD, docking calculations were performed by using the previously selected Siglec-10 structure and following a similar protocol to that for trisaccharide 1 (see above and Supplemental Information for further details). MD simulations of the most representative complexes were performed to assess the stability of the binding poses (Figure S4C); then, with the aim of selecting the best 3D model of interaction in line with STD data, CORCEMA-ST calculations were carried out. The lowest NOE R factor, with the value of 0.26, was obtained for the complex reported in the lower panel of Figure 4 in which the trisaccharide 2 adopted a gt conformation (see also Figure S6B).

The terminal sialic acid was again the major determinant of ligand binding, establishing several polar interactions with the protein, which were comparable with those observed for the trisaccharide MODEL IV/1 complex (Figure 4 upper panel). As in trisaccharide 1, the methyl group of Neu5Ac lies in the hydrophobic groove defined by Phe-21 and Tyr-128 (Figure 4 lower panel). Furthermore, the Neu5Ac residue of 2 formed the key salt bridge between its carboxylate and the guanidinium group of Arg-119; hydrogen bonds between the backbone of Asn-129 and the hydroxyl groups of the glycerol chain, namely, OH-8 and OH-9, were also observed. Notably, these interactions remained stable along the MD simulation. As observed for trisaccharide 1, a direct involvement of the Gal unit of 2 in the binding was evident with the hydrogen bond occurring between the Gal hydroxyl group in position 3 and the Glu-66 residue. Furthermore, the protein did not engage interactions with the GlcNAc unit of 2.

From the MD simulation performed on the selected MODEL IV/2 complex, it was possible to observe that, although 2 remained steadily anchored to the protein via the sialic acid moiety, the bound ligand explored both gt and tg conformations along the dynamic (Figure S5 lower panel). The ω angle, indeed, fluctuated between ω = 60° and ω = 180°. This suggested that ligand 2 displayed higher conformational flexibility, even upon binding, with respect to ligand 1. However, from the CORCEMA-ST analysis, the best fit between experimental and theoretical data was obtained for the gt conformer (Figure 4C lower panel). In detail, the highest predicted STD value was given by the methyl group of the N-acetamide moiety of Neu5Ac. This could be ascribable to the vicinity of the acetamide group toward the aromatic residues Phe-21 and Tyr-129. For the same reasons discussed for 1, the H-7 of the sialic acid glycerol side chain of trisaccharide 2 displayed higher STD effects with respect to the protons in position 8 and 9. Despite the proximity of the CC′ loop of the protein to the galactoside ring of the trisaccharide 2, lower STD effects were observed for the galactose moiety of 2. In complete agreement with the experimental STD data, no significant STD effects were predicted for the GlcNAc moiety (Figure 4 lower panel).

Molecular Interaction between Siglec-10 and Complex-type N-glycans

The interaction between naturally occurring glycans and Siglec-10 has also been investigated in order to unveil if and how a different presentation of the glycan epitope could influence its recognition by the receptor protein. Two biantennary complex-type undecasaccharides-Asn (ligands 3 and 4, Figure S1), bearing α2,3- and α2,6-linked sialic acid at glycan antennae termini, were used. Interestingly, NMR and computational methods suggested that only the outermost portion of the complex glycoconjugates was accommodated in the binding pocket of the protein with the inner residues not participating in the molecular recognition (Figure 5). Analysis of STD NMR spectra acquired in the presence of Siglec-10 (Figure S7) clearly revealed the key role of the sialic acid in the recognition and binding process, and the epitope mapping of glycans 3 and 4 was indeed comparable with those obtained with the smaller ligands 1 and 2; the acetyl group of the sialic acid displayed the strongest STD intensity and other STD enhancements corresponded to H-5/H-9 protons of the Neu5Ac moiety. Furthermore, STD effects for H-4 and H-6 of the galactose unit, in the case of the α2,6 undecasaccharide, and for H-3/H-6, in the case of the α2,3-linked sialoglycan, were observed. In addition, the protons belonging to the other sugar residues of the branches and the N-glycan pentasaccharide core did not appear in the STD spectra, clearly suggesting that the recognition was mainly driven by the Neu5Ac-Gal moieties. The other sugar residues were indeed solvent exposed and far from the protein surface.

Figure 5.

Figure 5

Open in a new tab

Molecular Recognition of Longer N-glycans by Siglec-10

(A) Full view of the 3-D complexes of ligand 3 (green) and ligand 4 (pink) using Siglec-10 (model IV) structural template obtained by docking calculations.

(B) Superimposition of ligand 1 (yellow) and ligand 3 (green) in the Siglec-10 (model IV) binding site.

(C) Superimposition of ligand 2 (blue) and ligand 4 (pink) in the Siglec-10 (model IV) binding site.

These results were further confirmed by modeling 3 and 4 into the Siglec-10 binding site (Figure 5). As shown in the resulting three-dimensional structures of the corresponding complexes, good predicted binding poses were obtained for both ligands, confirming that longer complex-type glycans can be accommodated in the protein binding pocket, displaying a similar network of interactions with respect to those established by sialoglycans 1 and 2 upon protein binding.

In detail, as expected, both complex glycoconjugates interacted with Siglec-10 binding-site residues mostly via the Neu5Ac moiety. Indeed, the ionized carboxylic group of sialic acid formed a polar interaction with the lateral chain of Arg-119 and Tyr-128 contributed significantly to the recognition establishing CH−π contacts with the Neu5Ac glycerol chain.

In the case of sialoglycan 3, the hydroxyl groups of Neu5Ac lateral chain were also involved in hydrogen bonds with Arg-127 and Arg-20. The Sia acetamide moiety formed a hydrogen bond with the Arg-127 backbone, whereas the Phe-21 residue on Siglec-10 was responsible for the hydrophobic interaction with Neu5Ac acetyl group. The Gal moiety was also placed inside the binding pocket, making polar contacts with Glu-66 and Asn-129 residues. A hydrogen bond between Asn-129 and the hydroxyl group at C-9 of Neu5Ac was instead observed for sialoglycan 4 in the interaction with Siglec-10. As in 3, the Gal moiety of 4 was in close proximity to the receptor surface, making polar contacts with Glu-66 residue. In agreement with NMR results, in both 3D complexes, the inner residues of the oligosaccharides were far from the receptor surface and did not directly contribute to the binding. Of note, the conformation of the sialylated lactosamine branches of 3 and 4, involved in the interaction with Siglec-10, was comparable with that adopted by the sialoglycans 1 and 2, respectively, upon binding. Moreover, independently from the sialic acid linkage to the adjacent galactose moiety, both complex glycans preferentially exhibited an extended conformation when placed in the binding site of the protein, although a certain flexibility around the Man-α-(1,6)-Man glycosidic linkage was observed (Figure 5).

Conclusions

In summary, the interaction of Siglec-10 with differently sialylated glycans was here investigated by an integrated approach based on NMR, docking, and molecular modeling. Although the crystal structure of Siglec-10 has not been solved yet, the orthogonal methods applied here allowed a molecular description of the recognition and binding processes driving Siglec-10-sialyloglycans complexes, providing unique information about dynamics and molecular interactions. The ligands' epitopes in their bound conformations were indeed profiled and consistent 3D models of the interactions were presented.

A comparison of the epitope maps obtained for α2,3- and α2,6-sialoglycans, which constitute the terminal end of the canonical sialylated complex-type N-glycans exposed on mammalian cells, highlighted that in each case the binding process to Siglec-10 was mainly driven by the sialic acid moiety, giving a 100% STDepitope fit, and galactose unit, whereas the N-acetylglucosamine pointed away from Siglec-10 binding pocket. On the other hand, some differences were evident from the STD-derived epitope maps, indicating that ligand 2 established less contacts within the protein binding pocket, as demonstrated by the lower involvement in the interaction of protons H-5, H-8, and H-9 of sialic acid and of the adjacent galactose unit.

The bound conformations of α2,3- and α2,6-sialoglycans were obtained by combining NOE-derived data with MM and MD simulations. The g- conformer of the α2,3-sialoglycans was selected upon binding to Siglec-10; instead, the α2,6-sialoglycans mainly assumed a bent gt conformation both in the free and bound states. Thus, our data suggested that, upon binding, the terminal portion of sialylated glycan chains preferentially adopted a cone-like or an umbrella-like topology depending on the sialic acid linkage to the underlying galactose residue.

Extensive homology modeling studies provided a 3-D model of human Siglec-10. Then, further computational approaches, namely docking and MD studies, together with CORCEMA-ST protocol, were used to either indirectly support the reliability of the 3-D model of Siglec-10 and to obtain structural insights into the molecular basis of N-glycan recognition by Siglec-10. Our three-dimensional models of the Siglec-10/sialoglycan complexes highlighted the architecture of the binding site essentially composed of hydrophobic (Phe-21), polar (Tyr-128, Asn-129), basic (Arg-119, Arg-127), and acid (Glu-66) residues. In detail, the Arg-119, Arg-128, Asn-129 residues were directly involved in the interaction with sialic acid, suggesting they shaped a region relevant for the molecular recognition, matching those previously reported for other studied members of the Siglec family (Crocker et al., 2007). Moreover, we observed that the CC' loop belonging to the Siglec-10 V-set domain established key interactions with the adjacent galactose moiety that appears to participate in the interaction process, whereas the GlcNAc residue showed no relevant contribution to the binding.

The ability of Siglec-10 to recognize both α-2,3 and α-2,6 linkages to a similar extent indicates a lower specificity (Macauley et al., 2014, Blixt et al., 2003) compared with other homologous Siglecs, likely due to a wider binding site. The CC′ loop region predicted for Siglec-10, indeed, points outward to the binding residues, in contrast to some other CD33-like Siglecs (for example, Siglec-5, Siglec-3, and Siglec-8) but as already reported for Sialoadhesin, Siglec-4, and Siglec-7 (Figure 6). This could account for the greater capacity of the protein to flexibly accommodate sialoglycans with different shapes and conformation, α2,3-sialoglycan 1 in its extended topology and the α2,6-sialoglycan 2 in a more compact bent conformation. Furthermore, also longer biantennary glycoconjugates,ligands 3 and 4 (Figure S1), could be accommodated in the protein binding pocket where the terminal ends preferentially adopted a cone-like or an umbrella-like topology respectively. Their outermost region was directly involved in polar and hydrophobic interactions with amino-acidic residues, whereas the inner moieties of the oligosaccharides pointed away from the protein surface. Of note, for both α2,3- and α2,6-linked complex type N-glycans 3 and 4, an extended conformation of the entire undecasaccharide seemed to be preferred (Figure 5) and, interestingly, the topology of the terminal sialylated portions oriented differently the long saccharide chains with respect to the binding site.

Figure 6.

Figure 6

Open in a new tab

Siglec-10 CC′ Loop Shape

Comparison of Siglec-10 (model IV) loop region with other CD33-like Siglec, namely, Siglec-5 (PDB ID: 2zg2) and Siglec-7 (PDB ID: 2hrl). According to the predicted model, Siglec-10 exhibits an “open” loop, similarly to Siglec-7.

Given the pivotal roles that Siglecs play in the regulation of immune cell functions, they have been widely recognized as targets for the development of immunotherapeutics against a broad range of diseases. The receptor-sialoglycan interactions can mediate events like cell adhesion, signaling and interaction, inhibition or regulation of the immune cell activation, and uptake of sialylated pathogens. The disturbance of the sialic acid-Siglec axis can contribute to the onset and development of diseases such as autoimmunity, infection, inflammation, aging, and cancer. Worthy of note is the role played by the inhibitory receptor Siglec-10 in several crucial patho-physiological processes. Siglec-10 is engaged by sialylated glycoproteins, as tumor-expressed CD24, to promote the tumor immune evasion, or by dangerous pathogens, including Campylobacter, to subvert the host immune system. Although our understanding of the biological roles of Siglecs has been expanding over the past decade, a detailed molecular knowledge of their interaction with glycan ligands is still an unmet challenge. For instance, despite Siglec-10 is an appealing target for the development of glyco-based immunotherapy (Angata et al., 2015), neither the three-dimensional structure of protein nor the molecular basis of its interactions with complex glycans had been previously unraveled. Thus, the outcomes shown here have improved our knowledge of the molecular mechanisms occurring between Siglec-10 and sialylated glycans, providing a structural point of view for the design and development of high-affinity ligands, in their bioactive conformations, able to control the receptor functionality.

Limitation of the Study

This study describes the interactions of the biologically relevant protein Siglec-10 with naturally occurring sialoglycans by using different biophysical techniques, including ELISA, steady-state fluorescence, and mainly NMR spectroscopy, combined with computational chemistry and CORCEMA-ST calculations. In the absence of high-resolution structural information on the receptor Siglec-10, we resorted to homology modeling following a convergent approach of combining different templates for the homology calculation and different modeling approaches to validate the generated models (docking, MD simulations, CORCEMA-ST). However, the structural conclusions we derived from homology modeling will require the definition of the three-dimensional structure of the protein Siglec-10 to avoid misinterpretation of the data.

Further development of this work may also include the search of natural ligands for Siglec-10. The functional significance of the Siglec-10/CD24 interaction has been previously shown; however, other heavily N-glycosylated and/or mucin-type glycoproteins could be further candidate ligands involved in the interaction with Siglec-10.

Resource Availability

Lead Contact

Roberta Marchetti (roberta.marchetti@unina.it).

Materials Availability

This study did not generate new unique reagents.

Data and Code Availability

This study did not generate datasets.

Methods

All methods can be found in the accompanying Transparent Methods supplemental file.

Acknowledgments

The authors sincerely thank GlyTech, Inc. for providing N-glycan ligand 3 and 4. This study was supported by the project “GLYTUNES” funded by MIUR Progetti di Ricerca di Rilevante Interesse Nazionale (PRIN 2017) (2017XZ2ZBK, 2019–2022) to A.S.; by progetto POR SATIN and Progetto POR Campania Oncoterapia to A.M.; by the European Commission (H2020-MSCA- 814102-SWEET CROSSTALK project) to A.M., R.M., and A.S.. This project has received funding from the European Research Council (ERC) under the European Union's Horizon 2020 research and innovation program under grant agreement No 851356 to R.M. FSE, PON Ricerca e Innovazione 2014–2020, Azione I.1 “Dottorati Innovativi con caratterizzazione Industriale” is acknowledged for funding the PhD grant to R.E.F. Grants by the Spanish Ministry of Science MICINN (CTQ2017-88353-R and fellowship BES 2015–071588 to J.G.-C.) and Wellcome Trust 103744/Z/14/Z to P.R.C. are acknowledged.

Author Contributions

A.S. and R.M. conceived and designed the project. R.M., R.E.F., C.D.C., A.S., and A.M. carried out NMR, docking, and CORCEMA experiments and analyzed the results. R.E.F., J.G.-C., and S.M.-S. performed homology modeling and MD simulations. P.R.C., A.A., and R.G. produced the protein. F.C. performed ELISA experiments. P.D.V. and F.B. carried out fluorescence analysis. K.F. and Y.M. provided the glycans. All the authors wrote, revised, and reviewed the manuscript.

Declaration of Interests

The authors declare no competing interests.

Published: June 26, 2020

Footnotes

Supplemental Information can be found online at https://doi.org/10.1016/j.isci.2020.101231.

Contributor Information

Roberta Marchetti, Email: roberta.marchetti@unina.it.

Alba Silipo, Email: silipo@unina.it.

Supplemental Information

Document S1. Transparent Methods, Figures S1–S7, and Tables S1 and S2
mmc1.pdf (2.3MB, pdf)

References

  1. Adams O.J., Stanczak M.A., von Gunten S., Läubli H. Targeting sialic acid-Siglec interactions to reverse immune suppression in cancer. Glycobiology. 2018;28:640–647. doi: 10.1093/glycob/cwx108. [DOI] [PubMed] [Google Scholar]
  2. Angata T., Nychola C.M., Macauley M.S. Therapeutic targeting of siglecs using antibody- and glycan-based approaches. Trends Pharmacol. Sci. 2015;36:645–660. doi: 10.1016/j.tips.2015.06.008. [DOI] [PMC free article] [PubMed] [Google Scholar]
  3. Angulo J., Nieto P.M. STD-NMR: application to transient interactions between biomolecules—a quantitative approach. Eur. Biophys. J. 2011;40:1357–1369. doi: 10.1007/s00249-011-0749-5. [DOI] [PubMed] [Google Scholar]
  4. Attril H., Takazawa H., Witt S., Kelm S., Isecke R., Brossmer R., Ando T., Ishida H., Kiso M., Crocker P.R., van Aalten D.M. The structure of siglec-7 in complex with sialosides: leads for rational structure-based inhibitor design. Biochem. J. 2006;397:271–278. doi: 10.1042/BJ20060103. [DOI] [PMC free article] [PubMed] [Google Scholar]
  5. Attril M.A., Trandem K., Sun P.D. Structural implications of siglec-5-mediated sialoglycan recognition. J. Mol. Biol. 2008;375:437–447. doi: 10.1016/j.jmb.2007.10.009. [DOI] [PMC free article] [PubMed] [Google Scholar]
  6. Bandala-Sanchez E., Bediaga G.N., Goddard-Borger E.D., Ngui K., Naselli G., Stone N.L., Neale A.M., Pearce L.A., Wardak A., Czabotar P. CD52 glycan binds the proinflammatory B box of HMGB1 to engage the Siglec-10 receptor and suppress human T cell function. Proc. Natl. Acad. Sci. U S A. 2018;115:7783–7788. doi: 10.1073/pnas.1722056115. [DOI] [PMC free article] [PubMed] [Google Scholar]
  7. Bandala-Sanchez E., Bediaga N.G., Naselli G., Neale A.M., Harrison L.C. Siglec-10 expression is up-regulated in activated human CD4+ T cells. Hum. Immunol. 2020 doi: 10.1016/j.humimm.2020.01.009. [DOI] [PubMed] [Google Scholar]
  8. Barkal A.A., Brewer R.E., Markovic M., Kowarsky M., Barkal S.A., Zaro B.W., Krishnan V., Hatakeyama J., Dorigo O., Barkal L.J. CD24 signalling through macrophage Siglec-10 is a target for cancer immunotherapy. Nature. 2019;572:392–396. doi: 10.1038/s41586-019-1456-0. [DOI] [PMC free article] [PubMed] [Google Scholar]
  9. Blixt O., Collins B.E., van den Nieuwenhof I.M., Crocker P.R., Paulson J.C. Sialoside specificity of the siglec family assessed using novel multivalent probes: identification of potent inhibitors of myelin-associated glycoprotein. J. Biol. Chem. 2003;278:31007–31019. doi: 10.1074/jbc.M304331200. [DOI] [PubMed] [Google Scholar]
  10. Bochner B.S., Zimmermann N. Role of siglecs and related glycan-binding proteins in immune responses and immunoregulation. J. Allergy Clin. Immunol. 2015;135:598–608. doi: 10.1016/j.jaci.2014.11.031. [DOI] [PMC free article] [PubMed] [Google Scholar]
  11. Di Carluccio C., Crisman E., Manabe Y., Forgione R.E., Lacetera A., Amato J., Pagano B., Randazzo A., Zampella A., Lanzetta R. Characterisation of the dynamic interactions between complex N-glycans and human CD22. ChemBioChem. 2019 doi: 10.1002/cbic.201900295. [DOI] [PubMed] [Google Scholar]
  12. Case D.A., Berryman J.T., Betz R.M., Cai Q., Cerutti D.S., Cheatham T.E., Darden T.A., Duke H.G., Goetz A.W., Gusarov S. University of California, San Francisco; 2014. AMBER 14. [Google Scholar]
  13. Chandrasekaran A., Srinivasan A., Raman R., Viswanathan K., Raguram S., Tumpey T.M., Sasisekharan V., Sasisekharan R. Glycan topology determines human adaptation of avian H5N1 virus hemagglutinin. Nat. Biotechnol. 2008;26:107–113. doi: 10.1038/nbt1375. [DOI] [PubMed] [Google Scholar]
  14. Chang Y.C., Nizet V. The interplay between Siglecs and sialylated pathogens. Glycobiology. 2014;24:818–825. doi: 10.1093/glycob/cwu067. [DOI] [PMC free article] [PubMed] [Google Scholar]
  15. Chen G.Y., Tang J., Zheng P., Liu Y. CD24 and Siglec-10 selectively repress tissue damage-induced immune responses. Science. 2009;323:1722–1725. doi: 10.1126/science.1168988. [DOI] [PMC free article] [PubMed] [Google Scholar]
  16. Chen G.-Y., Brown N.K., Zheng P., Liu Y. Siglec-G/10 in self–nonself discrimination of innate and adaptive immunity. Glycobiology. 2014;24:800–806. doi: 10.1093/glycob/cwu068. [DOI] [PMC free article] [PubMed] [Google Scholar]
  17. Crocker P.R., Paulson J.C., Varki A. Siglecs and their roles in the immune system. Nat. Rev. Immunol. 2007;7:255–256. doi: 10.1038/nri2056. [DOI] [PubMed] [Google Scholar]
  18. Duan S., Paulson J.C. Siglecs as immune cell checkpoints in disease. Annu. Rev. Immunol. 2020 doi: 10.1146/annurev-immunol-102419-035900. [DOI] [PubMed] [Google Scholar]
  19. Forgione R.E., Di Carluccio C., Kubota M., Manabe Y., Fukase K., Molinaro A., Hashigushi T., Marchetti R., Silipo A. Structural basis for Glycan-receptor binding by mumps virus hemagglutinin-neuraminidase. Sci. Rep. 2020;10:1589. doi: 10.1038/s41598-020-58559-6. [DOI] [PMC free article] [PubMed] [Google Scholar]
  20. Jayalakshmi V., Krishna N.R. Complete relaxation and conformational exchange matrix (CORCEMA) analysis of intermolecular saturation transfer effects in reversibly forming ligand-receptor complexes. J. Magn. Reson. 2002;155:106–118. doi: 10.1006/jmre.2001.2499. [DOI] [PubMed] [Google Scholar]
  21. Läubli H., Varki A. Sialic acid-binding immunoglobulin-like lectins (Siglecs) detect self-associated molecular patterns to regulate immune responses. Cell. Mol. Life Sci. 2019;77:593–605. doi: 10.1007/s00018-019-03288-x. [DOI] [PMC free article] [PubMed] [Google Scholar]
  22. Liu Y.C., Yu M.M., Chai Y.F., Shou S.T. Sialic acids in the immune response during sepsis. Front. Immunol. 2017;8:1601. doi: 10.3389/fimmu.2017.01601. [DOI] [PMC free article] [PubMed] [Google Scholar]
  23. Lübbers J., Rodríguez E., van Kooyk Y. Modulation of immune tolerance via siglec-sialic acid interactions. Front. Immunol. 2018;9:2807. doi: 10.3389/fimmu.2018.02807. [DOI] [PMC free article] [PubMed] [Google Scholar]
  24. Macauley M.S., Crocker P.R., Paulson J.C. Siglec-mediated regulation of immune cell function in disease. Nat. Rev. Immunol. 2014;14:653–666. doi: 10.1038/nri3737. [DOI] [PMC free article] [PubMed] [Google Scholar]
  25. Marchetti R., Perez S., Arda A., Imberty A., Jimenez-Barbero J., Silipo A., Molinaro A. “Rules of Engagement” of proteins-glycoconjugates interactions: a molecular view achievable by NMR spectroscopy and molecular modeling. Chemistryopen. 2016;5:274–296. doi: 10.1002/open.201600024. [DOI] [PMC free article] [PubMed] [Google Scholar]
  26. Mayer M., Meyer B. Characterization of ligand binding by saturation transfer difference NMR spectroscopy. Angew. Chem. Int. Ed. 1999;38:1784–1788. doi: 10.1002/(SICI)1521-3773(19990614)38:12<1784::AID-ANIE1784>3.0.CO;2-Q. [DOI] [PubMed] [Google Scholar]
  27. Meyer B., Peters T. NMR spectroscopy techniques for screening and identifying ligand binding to protein receptors. Angew. Chem. Int. Ed. 2003;42:864–890. doi: 10.1002/anie.200390233. [DOI] [PubMed] [Google Scholar]
  28. Morris G.M., Huey R., Lindstrom W., Sanner M.F., Belew R.K., Goodsell D.S., Olson A.J. AutoDock4 and AutoDockTools4: automated docking with selective receptor flexibility. J. Comput. Chem. 2009;30:2785–2791. doi: 10.1002/jcc.21256. [DOI] [PMC free article] [PubMed] [Google Scholar]
  29. Patel D.S., Pendrill R., Mallajosyula S.S., Widmalm G., MacKerell A.D., Jr. Conformational properties of α- or β-(1→6)-linked oligosaccharides: Hamiltonian replica exchange MD simulations and NMR experiments. J. Phys. Chem. B. 2014;118:2851–2871. doi: 10.1021/jp412051v. [DOI] [PMC free article] [PubMed] [Google Scholar]
  30. Phongsisay V. The immunobiology of Campylobacter jejuni: innate immunity and autoimmune diseases. Immunobiology. 2016;221:535–543. doi: 10.1016/j.imbio.2015.12.005. [DOI] [PubMed] [Google Scholar]
  31. Poppe L., Struike-Prill R., Meyer B., van Halbeek H. The solution conformation of sialyl-et(2 6)-lactose studied by modern NMR techniques and Monte Carlo simulations. J. Biomol. NMR. 1992;2:109–136. doi: 10.1007/BF01875524. [DOI] [PubMed] [Google Scholar]
  32. Poppe L., Brown G.S., Philo J.S., Nikrad P.V., Shah B.H. Conformation of sLex Tetrasaccharide, free in solution and bound to E-, P-, and L-selectin. J. Am. Chem. Soc. 1997;119:1727–1736. [Google Scholar]
  33. Poveda A., Jiménez-Barbero J. NMR studies of carbohydrate–protein interactions in solution. Chem. Soc. Rev. 1998;27:133–144. [Google Scholar]
  34. Pröpster J.M., Yang F., Rabbani S., Ernst B., Allain F.H.-T., Schubert M. Structural basis for Siglec-8-glycan recognition. Proc. Natl. Acad. Sci. U S A. 2016;113:E4170–E4179. doi: 10.1073/pnas.1602214113. [DOI] [PMC free article] [PubMed] [Google Scholar]
  35. Rapoport E., Mikhalyov I., Zhang J., Crocker P., Bovin N. Ganglioside binding pattern of CD33-related siglecs. Bioorg. Med. Chem. Lett. 2003;13:675–678. doi: 10.1016/s0960-894x(02)00998-8. [DOI] [PubMed] [Google Scholar]
  36. Sammar M., Siwetz M., Meiri H., Fleming V., Altevogt P., Huppertz B. Expression of CD24 and Siglec-10 in firsttrimesterplacenta: implications for immune tolerance at the fetal-maternal interface. Histochem. Cell Biol. 2017;147:565–574. doi: 10.1007/s00418-016-1531-7. [DOI] [PubMed] [Google Scholar]
  37. Stephenson H.N., Mills D.C., Jones H., Milioris E., Copland A., Dorrell N., Wren B.W., Crocker P.R., Escors D., Bajaj-Elliott M. Pseudaminic acid on Campylobacter jejuni flagella modulates dendritic cell IL-10 expression via Siglec-10 receptor: a novel flagellin-host interaction. J. Infect. Dis. 2014;210:1487–1498. doi: 10.1093/infdis/jiu287. [DOI] [PMC free article] [PubMed] [Google Scholar]
  38. Whitney G., Wang S., Chang H., Cheng K.Y., Lu P., Zhou X.D., Yang W.P., McKinnon M., Longphre M. A new siglec family member, siglec-10, is expressed in cells of the immune system and has signaling properties similar to CD33. Eur. J. Biochem. 2001;268:6083–6096. doi: 10.1046/j.0014-2956.2001.02543.x. [DOI] [PubMed] [Google Scholar]
  39. Xu D., Newhouse E.I., Amaro R.E., Pao H.C., Cheng L.S., Markwick P.R., McCammon J.A., Li W.W., Arzberger P.W., Distinct Glycan topology for avian and human sialopentasaccharide receptor analogues upon binding different hemagglutinins: a molecular dynamics perspective. J. Mol. Biol. 2009;387:465–491. doi: 10.1016/j.jmb.2009.01.040. [DOI] [PMC free article] [PubMed] [Google Scholar]
  40. Yan J., Kline A.D., Mo H., Shapiro M.J., Zartler E.R. The effect of relaxation on the epitope mapping by saturation transfer difference NMR. J. Magn. Reson. 2003;163:270–276. doi: 10.1016/s1090-7807(03)00106-x. [DOI] [PubMed] [Google Scholar]
  41. Zhang P., Lu X., Tao K., Shi L., Li W., Wang G., Wu K. Siglec-10 is associated with survival and natural killer cell dysfunction in hepatocellular carcinoma. J. Surg. Res. 2015;194:107–113. doi: 10.1016/j.jss.2014.09.035. [DOI] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

Document S1. Transparent Methods, Figures S1–S7, and Tables S1 and S2
mmc1.pdf (2.3MB, pdf)

Data Availability Statement

This study did not generate datasets.

Articles from iScience are provided here courtesy of Elsevier

RESOURCES