Glycosaminoglycan Interactions with Chemokines Add Complexity to a Complex System

Abstract

Chemokines have two types of interactions that function cooperatively to control cell migration. Chemokine receptors on migrating cells integrate signals initiated upon chemokine binding to promote cell movement. Interactions with glycosaminoglycans (GAGs) localize chemokines on and near cell surfaces and the extracellular matrix to provide direction to the cell movement. The matrix of interacting chemokine-receptor partners has been known for some time, precise signaling and trafficking properties of many chemokine-receptor pairs have been characterized, and recent structural information has revealed atomic level detail on chemokine-receptor recognition and activation. However, precise knowledge of the interactions of chemokines with GAGs has lagged far behind such that a single paradigm of GAG presentation on surfaces is generally applied to all chemokines. This review summarizes accumulating evidence which suggests that there is a great deal of diversity and specificity in these interactions, that GAG interactions help fine-tune the function of chemokines, and that GAGs have other roles in chemokine biology beyond localization and surface presentation. This suggests that chemokine-GAG interactions add complexity to the already complex functions of the receptors and ligands.

Keywords: chemokine oligomerization; chemokine structure; chemokine therapeutics; chemokines; glycosaminoglycans/GAGs; heparan sulfate.

Figure 1

The role of GAGs in cell recruitment. Chemokines (yellow circles) produced by the underlying tissue are immobilized on cell surface GAGs. Circulating leukocytes first roll upon interaction with selectins (pale gray symbols), followed by firm adhesion to integrin ligands (dark gray symbols) following integrin activation on the leukocytes after which they transmigrate into the tissue. Although chemokines are depicted as a single species, many different chemokines may be involved.

Figure 2

Chemokine tertiary structure, chemokine–receptor interactions, and chemokine–glycosaminoglycan interactions. (A) Structure of a typical chemokine, illustrated by a monomeric subunit of CXCL8 (PDB ID 1IL8 [48]). (B) Structure of CCR5 bound to a variant of CCL5 with a modified N-terminus (PDB ID 5UIW [52]). Chemokine recognition sites 1, 1.5 and 2 (CRS1, CRS1.5 and CRS2) are highlighted in green, blue and salmon. In CRS1, residues that contribute to GAG binding are highlighted in dark blue (R17 and 40s-loop BBXB motif residues R44 (not visible), K45 and R47). Tyrosine sulfates are highlighted as orange and yellow spheres interacting with the 40s-loop cluster and R17. (C) Model of CCL5 in complex with a chondroitin sulfate (CS) hexasaccharide from paramagnetic relaxation enhancement and intra- and intermolecular nuclear Overhauser effect constraints (supplementary data to [56]). GAG binding residues R17, R44, K45 and R47 are highlighted in dark blue. Comparison with (B) shows that GAG and receptor utilize similar epitopes on the chemokine core domain (CRS1) and sulfation is involved in both cases. Consequently, binding of GAGs and receptors to chemokines are generally mutually exclusive. Panels (A–C) are not to scale.

Figure 3

Chemokine oligomers and their ability to bind chemokine receptors, or not. (A) Structure of a typical CXC dimer, illustrated by CXCL12 (PDB ID 3GV3) with individual subunits colored in blue and orange. (B) Structure of a typical CC dimer, illustrated by vMIP-II (PDB ID 2FHT) with individual subunits colored in blue and orange. (C) Docked model of the vMIP-II dimer to CXCR4 shows that CC dimers are sterically incompatible to bind receptor; the orange subunit completely overlaps and clashes with the receptor. (D) Docked model of the CXCL12 dimer with CXCR4 shows that CXC dimers are compatible with receptor binding. Models from (C,D) are reported in [50].

Figure 4

Chemokines adopt a wide range of oligomeric structures, sometimes in response to different GAGs. Structures are shown with the different chemokine subunits color-coded in sky blue and gray, and the main GAG binding residues highlighted in dark blue. (A) The CCL2 dimer, which binds to HS (PDB ID 1DOM [68]); residues highlighted include R18, K19, R24, K49. K58, H66 [55]. (B) The CCL2 tetramer, which binds to heparin (PDB ID 1DOL [86]); residues highlighted include R18, K19, R24, K49. K58, H66 [55]. (C) The CXCL4 tetramer, which binds to heparin (PDB ID 1RHP [85]); residues highlighted include R20, R22, K46, R49, K61, K62, K65, K66 [87]. (D) The CCL5 polymer (PDB ID 5DNF [84]); residues highlighted include K55, K56 and R59 [58,88].

Figure 5

Functional effects of chemokine–GAG interactions. (A) The GAG chains of proteoglycans form a hydrated adhesion-resistant surface on endothelial cells that inhibits leukocyte transmigration. (B,C) Binding of chemokines to the HS chains of proteoglycans may promote: clustering of the syndecans and chemokine–receptor independent signaling (B); or remodeling of the endothelial glycocalyx to enable leukocyte transmigration (C). Figure adapted from [31].

Figure 6

Proposed modes of action of antibodies 1B6 and 1F11. In the updated model of chemokine activity the directional signal is produced by a localized cloud of soluble chemokine. (A) The majority of 1B6 molecules are bound to GAG-displayed CXCL10 and free chemokines in the soluble gradient are not inhibited. (B) The binding of the antibody to soluble chemokines inhibits cell migration induced by the soluble chemokine cloud and may interfere with the formation of the chemokine gradient.