A Remote Secondary Binding Pocket Promotes Heteromultivalent Targeting of DC-SIGN

Abstract

Dendritic cells (DC) are antigen-presenting cells coordinating the interplay of the innate and the adaptive immune response. The endocytic C-type lectin receptors DC-SIGN and Langerin display expression profiles restricted to distinct DC subtypes and have emerged as prime targets for next-generation immunotherapies and anti-infectives. Using heteromultivalent liposomes copresenting mannosides bearing aromatic aglycones with natural glycan ligands, we serendipitously discovered striking cooperativity effects for DC-SIGN⁺ but not for Langerin⁺ cell lines. Mechanistic investigations combining NMR spectroscopy with molecular docking and molecular dynamics simulations led to the identification of a secondary binding pocket for the glycomimetics. This pocket, located remotely of DC-SIGN's carbohydrate bindings site, can be leveraged by heteromultivalent avidity enhancement. We further present preliminary evidence that the aglycone allosterically activates glycan recognition and thereby contributes to DC-SIGN-specific cell targeting. Our findings have important implications for both translational and basic glycoscience, showcasing heteromultivalent targeting of DCs to improve specificity and supporting potential allosteric regulation of DC-SIGN and CLRs in general.

Figure 1

Discovery of Man-based glycomimetics as Langerin ligands. (a) A focused library of 27 mannosides was screened against Langerin in a ¹⁹F NMR RDA. The library was previously synthesized as FimH and LecB inhibitors for the developments of anti-infectives against E. coli and P. aeruginosa, respectively.⁻ The screening yielded biphenyl aglycone-bearing mannoside 9 as a promising hit that was subsequently modified to 42 to enable conjugation to liposomes via a sulfonamide linker. (b) The Ca²⁺-dependent interaction of 9 with the Langerin CBS was validated by the addition of EDTA in direct ¹⁹F R₂-filtered NMR binding experiments using the trifluoro methyl group. (c) The K_I value determination for acetylated 42 (→ 43, K_I = 0.25 ± 0.07 mM) via the ¹⁹F NMR RDA revealed a 40-fold affinity increase over the Man reference 45 (Table S4, K_I = 10 ± 1 mM). Data shown for 45 were previously published and the mannoside was prepared as previously described. (d) The affinity of 43 (K_D = 0.46 ± 0.09 mM) was validated via ¹⁵N HSQC NMR.

Figure 2

Heteromultivalent liposomes show enhanced binding to DC-SIGN⁺ cells. (a) Schematic depiction of heteromultivalent liposomes. Exemplarily for a composition comprised of 42-Lip and Man-Lip ligands. (b) Binding of homo- and heteromultivalent Man-Lip/42-Lip liposomes to Raji cells: 42-Lip up to 1.26% total lipid concentration does not show any binding. Man-Lip on its own starts showing binding at 4.75%. The combination of both ligands facilitates strong binding with increasing amounts of 42-Lip and concurrently decreasing amounts of Man-Lip. (c) Binding of Man-Lip/42-Lip liposomes is inhibited by addition of EDTA or mannan suggesting involvements of Ca²⁺ mediated interaction at the primary binding site. (d) The cooperative binding effect saturates around a molar 42-Lip /Man-Lip ratio of 0.5. (e) Heteromultivalent liposomes using 42-Lip in conjunction with two other natural DC-SIGN ligands (Fuc-Lip and LeX-Lip): All plotted data were adjusted to the same natural ligand concentration (2.5%). Due to different coupling efficiencies, the molar ratios for 42-Lip vary, albeit remaining within the effect saturation regime of above 0.5 (see Figure 2d). (f) Raw data of measurements prior to processing shown in (e). (g) Homo- and heteromultivalent liposomes comprising two natural ligands do not significantly change in their ability to bind DC-SIGN (exemplarily shown for Man-Lip and LeX-Lip).

Figure 3

Ligand-observed binding mode analysis for DC-SIGN. (a) Titration experiments using the ¹⁹F NMR RDA indicated interaction of 48 with DC-SIGNs CBS with affinity in the low millimolar range (K_I = 1.15 ± 0.01 mM). (b) The K_D value determined in ¹⁵N HSQC NMR titrations indicated higher affinity than measured with the ¹⁹F NMR RDA (K_D,¹⁵N HSQC = 0.46 ± 0.16 mM). (c) Direct ¹⁹F R₂-filtered NMR binding experiments using the trifluoromethyl group of 48 show incomplete inhibition in the presence of 4 mM EDTA or 200 mM mannose, suggesting a secondary Ca²⁺-independent binding mode. (d) STD NMR experiments in the presence of Ca²⁺ served to determine the Ca²⁺-dependent binding mode of 48 with DC-SIGN ECD. Specific STD effects were only observed in the presence of DC-SIGN ECD, allowing for epitopes to be determined from normalized STD′₀ values calculated from STD build-up curves at t_sat of 0.5, 1, 2, and 6 s (Figure S18). The STD NMR spectrum is magnified 4-fold. STD′₀ values in H3 and H4 indicated a mannose-type interaction with DC-SIGN, while a high STD′₀ in H1′ suggested the −CF₃ on phenyl ring A to be involved in binding. (e) ¹H STD NMR experiments in the presence of EDTA-d₁₁ served to validate the Ca²⁺-independent binding mode with DC-SIGN ECD. Epitopes were obtained from normalized STD values calculated from STD spectra at t_sat of 2 s and displayed a shift from a mannose-dependent binding to an interaction dominated by the biphenyl-moiety in C1 in the absence of Ca²⁺. The STD spectrum is magnified 4-fold. CBS-inhibition using an excess of mannose revealed similar STD NMR epitopes (Figure S19).

Figure 4

Interaction of 48 with a secondary DC-SIGN binding pocket. (a) CSPs observed in ¹⁵N HSQC NMR in the presence of Ca²⁺ confirm involvement of CBS residues in binding of 48. (b) Mapping of CSPs on X-ray structure of DC-SIGN (PDB code: 1SL4) corroborates interaction with the CBS. (c) Examples of CBS residues showing fast exchanging resonances and reduced intensity upon titration (N365 and D366). (d and e) ΔCSPs were determined by subtracting CSPs observed under noninhibitory conditions from those observed under CBS-inhibitory conditions (Figure S25). The CSP map shows increases for residues of the secondary binding pocket as well as in the short and long loop regions in the absence of Ca²⁺. Mapping of ΔCSPs on the X-ray structure of DC-SIGN (PDB code: 1SL4) locates the secondary binding pocket between α helix 2 and β sheet 0 and 1. (f) Examples of residues in the secondary binding pocket showing increased CSPs upon 48 addition (M270, Y268).

Figure 5

Binding of natural glycan ligands to DC-SIGN is positively modulated by binding of 48. (a) The best-scored pose obtained from biased molecular docking simulations agrees with our experimental NMR-based binding mode analysis. The receptor surface is colored according to its hydrophobicity. (b and c) ¹⁵N HSQC NMR experiments with the M270F mutant and 48 demonstrate abrogation of 48-binding in the absence of Ca²⁺. ΔCSPs were determined by subtracting CSPs observed under noninhibitory conditions from those observed under CBS-inhibitory conditions (Figures S29 and S30). Compared to wild-type DC-SIGN, mapping of ΔCSPs on the X-ray structure of DC-SIGN (PDB code: 1SL4) revealed no increase in CSPs for the secondary binding pocket under inhibitory conditions. (d) Exemplary resonances of residues in the secondary binding pocket (Y268 and Q306) demonstrate abrogation of 48-binding. (e) The allosteric binding site predicted by AlloSite shares residues with the herein identified secondary binding pocket. A pseudoligand is depicted in black spheres. (f) DC-SIGN⁺ Raji cells were incubated with AF647-functionalized liposomes carrying either 2.5 mol % Fuc-Lip as a CBS ligand (left panel) or 1.26 mol % 48 (right panel) in conjunction with soluble 48 or Fuc, respectively. MFIs were normalized to samples containing no 48 or Fuc and then averaged over the results from four biological replicates, each conducted as two technical replicates. Soluble 48 significantly enhances binding of Fuc-Lip liposomes to Raji DC-SIGN⁺ cells (*p < 0.05; n = 4; two-tailed, unpaired Student’s t test). In contrast, soluble Fuc did not increase binding of 48 liposomes to DC-SIGN⁺ Raji cells (p > 0.05; n = 4; two-tailed, unpaired Student’s t test) (lipophilic, red; hydrophilic, blue).

Figure 6

Proposed mechanism of heteromultivalent avidity enhancement for DC-SIGN. Binding of a secondary binding pocket-ligand (i.e., 42/48) causes structural rearrangements in the DC-SIGN CRD affecting the carbohydrate binding site. This allosteric activation might increase Ca²⁺ complexation or directly natural glycan ligand binding, resulting in cooperative avidity enhancement for the heteromultivalent liposome beyond the targeting of two independent binding sites.