Chemokines selectively recruit and activate a variety of cells during inflammation. Interactions between cell surface glycosaminoglycans (GAGs) and chemokines drive the formation of haptotactic or immobilized gradients of chemokines at the site of inflammation, directing this recruitment. Chemokines bind to glycosaminoglycans on human umbilical vein endothelial cells (HUVECs) with affinities in the micromolar range: RANTES > MCP-1 > IL-8 > MIP-1alpha. This binding can be competed with by soluble glycosaminoglycans: heparin, heparin sulfate, chondroitin sulfate, and dermatan sulfate. RANTES binding showed the widest discrimination between glycosaminoglycans (700-fold), whereas MIP-1alpha was the least selective. Almost identical results were obtained in an assay using heparin sulfate beads as the source of immobilized glycosaminoglycan. The binding of chemokines to glycosaminoglycan fragments has a strong length dependence, and optimally requires both N- and O-sulfation. Isothermal titration calorimetry data confirm these results; IL-8 binds heparin fragments with a K(d) of 0.39-2.63 microM, and requires five saccharide units to bind each monomer of chemokine. In membranes from cells expressing the G-protein-coupled chemokine receptors CXCR1, CXCR2, and CCR1, soluble GAGs inhibit the binding of chemokine ligands to their receptors. Consistent with this, heparin and heparin sulfate could inhibit IL-8-induced neutrophil calcium flux. Chemokines can therefore form complexes with both cell surface and soluble GAGs; these interactions have different functions. Soluble GAG chemokines complexes are unable to bind the receptor, resulting in a block of the biological activity. Previously, we have shown that cell surface GAGs present chemokines to the G-protein-coupled receptors, by increasing the local concentration of protein. A model is presented which brings together all of these data. The selectivity in the chemokine-GAG interaction suggests selective disruption of the haptotactic gradient may be an achievable therapeutic approach in inflammatory disease.

Chemokines are 8-10 kDa proteins involved in the control of leukocyte trafficking and activation. In free solution, chemokines are monomers at physiologic concentrations, although many multimerize at higher concentrations. Cell surface heparan sulfate may sequester chemokines, increasing their local concentrations and facilitating their binding to receptors expressed on leukocytes. In competitive binding assays using immobilized heparin, a 2-3-fold increase in the bound radiolabeled chemokine was seen with increasing concentrations of unlabeled chemokine in the nanomolar range. Unlabeled chemokine concentrations between 0.25 and 50 microM were needed to compete the bound radioactivity. This biphasic competition curve was not seen for N-methyl-L25 IL-8, a variant of IL-8 which is unable to dimerize. In addition, complexes of chemokine and heparin eluted from gel filtration columns with apparent molecular masses of 33-60 kDa, suggesting that chemokine multimerization had occurred. The physiological relevance of this multimerization process was seen from studies using human endothelial cells. The endothelial cell binding sites for IL-8, RANTES, and MCP-1 were deduced to be glycosaminoglycans since competition assays showed the biphasic curves and micromolar IC50 values seen in studies with immobilized heparin, and mRNA for known chemokine receptors was not detected. Furthermore, digestion of endothelial cell monolayers with glycosaminidases decreased chemokine binding by up to 80%. Glycosaminoglycans can act as modulators of the ligand binding affinity of chemokine receptor-bearing cells. Removal of glycosaminoglycans from CHO cells expressing chemokine receptors CXCR1, CCR1, or CCR2 resulted in 40-70% decreases in the binding of RANTES, MCP-1, IL-8, and MIP-1alpha. Our data show that cell surface glycosaminoglycans induce polymerization of chemokines, increasing their local concentration and therefore enhancing their effects on high-affinity receptors within the local microenvironment.

Extension of recombinant human RANTES by a single residue at the amino terminus is sufficient to produce a potent and selective antagonist. RANTES is a proinflammatory cytokine that promotes cell accumulation and activation in chronic inflammatory diseases. When mature RANTES was expressed heterologously in Escherichia coli, the amino-terminal initiating methionine was not removed by the endogenous amino peptidases. This methionylated protein was fully folded but completely inactive in RANTES bioassays of calcium mobilization and chemotaxis of the promonocytic cell line THP-1. However, when assayed as an antagonist of both RANTES and macrophage inflammatory polypeptide-1␣ (MIP-1␣) in these assays, the methionylated RANTES (Met-RANTES) inhibited the actions of both chemokines. T cell chemotaxis was similarly inhibited. The antagonistic effect was selective since Met-RANTES had no effect on interleukin-8-or monocyte chemotractant protein-1-induced responses in these cells. Met-RANTES can compete with both [ RANTES is a member of a large family of cytokines, known as chemokines, which have the ability to recruit and activate a wide variety of proinflammatory cell types (1). They are small polypeptides of 8 -10 kDa and have been further classified into CXC or CC chemokines based on the spacings of the cysteine residues proximal to the amino terminus. CXC chemokines primarily activate neutrophils, whereas CC chemokines have effects on several leucocyte cell types. RANTES is a CC chemokine, and in vitro it can produce chemotaxis and activation of monocytes, eosinophils, and T cells, particularly CD4 ϩ CD45RO ϩ (memory) T cells (2), but not neutrophils. These results imply a role for RANTES in diseases such as allergen induced late phase skin reactions or in allergic asthma. This hypothesis is strengthened by the fact that large amounts of RANTES are found in nasal polyp tissues, which are rich in infiltrating eosinophils (3). In addition, injection of RANTES into dog skin has been shown to induce a large eosinophilic infiltrate in vivo (4), and migration of human T lymphocytes was observed on injection of human RANTES into a human/severe combined immune deficiency mouse model (5).MIP-1␣ shares an overlapping cell-type specificity with RANTES in vitro (6, 7) and has been shown to elicit an inflammatory response mediated through mast cell degranulation in vivo (8). A common receptor for these two CC chemokines has been cloned (9, 10) and is a member of the seven transmembrane G-protein linked receptor family. Recombinant expression of the receptor has shown that it can transduce a functional response on stimulation by both chemokines.We report the purification of human RANTES expressed heterologously in Escherichia coli. In this system, the protein retains its initiating methionine residue, which renders it inactive as an agonist, while enabling it to antagonize effects induced both by RANTES and MIP-1␣. It is able to compete for binding of both the radiolabeled ligands on THP-1 cells and to the recombinant RANTES/MIP-1...

CC chemokine receptor (CCR)4, a high affinity receptor for the CC chemokines thymus and activation-regulated chemokine (TARC) and macrophage-derived chemokine (MDC), is expressed in the thymus and spleen, and also by peripheral blood T cells, macrophages, platelets, and basophils. Recent studies have shown that CCR4 is the major chemokine receptor expressed by T helper type 2 (Th2) polarized cells. To study the in vivo role of CCR4, we have generated CCR4-deficient (CCR4−/−) mice by gene targeting. CCR4−/− mice developed normally. Splenocytes and thymocytes isolated from the CCR4−/− mice failed to respond to the CCR4 ligands TARC and MDC, as expected, but also surprisingly did not undergo chemotaxis in vitro in response to macrophage inflammatory protein (MIP)-1α. The CCR4 deletion had no effect on Th2 differentiation in vitro or in a Th2-dependent model of allergic airway inflammation. However, CCR4−/− mice exhibited significantly decreased mortality on administration of high or low dose bacterial lipopolysaccharide (LPS) compared with CCR4+/+ mice. After high dose LPS treatment, serum levels of tumor necrosis factor α, interleukin 1β, and MIP-1α were reduced in CCR4−/− mice, and decreased expression of MDC and MIP-2 mRNA was detected in peritoneal exudate cells. Analysis of peritoneal lavage cells from CCR4−/− mice by flow cytometry also revealed a significant decrease in the F4/80+ cell population. This may reflect a defect in the ability of the CCR4−/− macrophages to be retained in the peritoneal cavity. Taken together, our data reveal an unexpected role for CCR4 in the inflammatory response leading to LPS-induced lethality.

The activation of leukocytes by chemokines is believed to be mediated via binding of chemokines to glycosaminoglycan chains of the extracellular matrix. The binding site on the chemokine interleukin-8 (IL-8) for the glycosaminoglycan heparin has been characterized using a systematic series of site-directed mutants of IL-8 in which the basic residues of the protein have been replaced by alanine. Mutation of K64 and R68 caused the largest decrease in affinity for a heparin Sepharose matrix, with smaller effects seen with mutations of K20, R60, and K67. Heparin-derived disaccharides that could disrupt the IL-8-heparin Sepharose interaction were identified by a competitive binding assay. Heteronuclear NMR spectroscopic titration of 15N-labeled IL-8 with a trisulfated disaccharide revealed a cluster of residues on IL-8 which were perturbed by disaccharide binding. These data identify a heparin-binding surface on IL-8 that includes the C-terminal alpha-helix and the proximal loop around residues 18-23. The heparin-binding site is spatially distinct from the residues involved in receptor binding.