Chemokines orchestrate leukocyte trafficking and function in health and disease. Heterophilic interactions between chemokines in a given microenvironment may amplify, inhibit, or modulate their activity; however, a systematic evaluation of the chemokine interactome has not been performed. We used immunoligand blotting and surface plasmon resonance to obtain a comprehensive map of chemokine-chemokine interactions and to confirm their specificity. Structure-function analyses revealed that chemokine activity can be enhanced by CC-type heterodimers but inhibited by CXC-type heterodimers. Functional synergism was achieved through receptor heteromerization induced by CCL5-CCL17 or receptor retention at the cell surface via auxiliary proteoglycan binding of CCL5-CXCL4. In contrast, inhibitory activity relied on conformational changes (in CXCL12), affecting receptor signaling. Obligate CC-type heterodimers showed high efficacy and potency and drove acute lung injury and atherosclerosis, processes abrogated by specific CCL5-derived peptide inhibitors or knock-in of an interaction-deficient CXCL4 variant. Atheroprotective effects of CCL17 deficiency were phenocopied by a CCL5-derived peptide disrupting CCL5-CCL17 heterodimers, whereas a CCL5 α-helix peptide mimicked inhibitory effects on CXCL12-driven platelet aggregation. Thus, formation of specific chemokine heterodimers differentially dictates functional activity and can be exploited for therapeutic targeting.

Chemokines and galectins are simultaneously upregulated and mediate leukocyte recruitment during inflammation. Until now, these effector molecules have been considered to function independently. Here, we tested the hypothesis that they form molecular hybrids. By systematically screening chemokines for their ability to bind galectin‐1 and galectin‐3, we identified several interacting pairs, such as CXCL12 and galectin‐3. Based on NMR and MD studies of the CXCL12/galectin‐3 heterodimer, we identified contact sites between CXCL12 β‐strand 1 and Gal‐3 F‐face residues. Mutagenesis of galectin‐3 residues involved in heterodimer formation resulted in reduced binding to CXCL12, enabling testing of functional activity comparatively. Galectin‐3, but not its mutants, inhibited CXCL12‐induced chemotaxis of leukocytes and their recruitment into the mouse peritoneum. Moreover, galectin‐3 attenuated CXCL12‐stimulated signaling via its receptor CXCR4 in a ternary complex with the chemokine and receptor, consistent with our structural model. This first report of heterodimerization between chemokines and galectins reveals a new type of interaction between inflammatory mediators that can underlie a novel immunoregulatory mechanism in inflammation. Thus, further exploration of the chemokine/galectin interactome is warranted.

BackgroundGlycoprotein VI (GPVI) is the essential platelet collagen receptor in atherothrombosis, but its inhibition causes only a mild bleeding tendency. Thus, targeting this receptor has selective antithrombotic potential.ObjectivesThis study sought to compare compounds interfering with platelet GPVI–atherosclerotic plaque interaction to improve current antiatherothrombotic therapy.MethodsHuman atherosclerotic plaque–induced platelet aggregation was measured in anticoagulated blood under static and arterial flow conditions (550/s, 1,100/s, and 1,500/s). Inhibition by dimeric GPVI fragment crystallizable region of IgG (Fc) masking GPVI binding sites on collagen was compared with that of 3 anti-GPVI antibodies: BLO8-1, a human domain antibody; 5C4, a fragment antigen-binding (Fab fragment) of monoclonal rat immunoglobulin G; and m-Fab-F, a human recombinant sFab against GPVI dimers.ResultsGPVI-Fc reduced plaque-triggered platelet aggregation in static blood by 51%, BLO8-1 by 88%, and 5C4 by 93%. Under arterial flow conditions, BLO8-1 and 5C4 almost completely inhibited platelet aggregation while preserving platelet adhesion on plaque. Inhibition by GPVI-Fc, even at high concentrations, was less marked but increased with shear rate. Advanced optical imaging revealed rapid persistent GPVI-Fc binding to collagen under low and high shear flow, upstream and downstream of plaque fragments. At low shear particularly, platelets adhered in plaque flow niches to GPVI-Fc–free segments of collagen fibers and recruited other platelets onto aggregates via ADP and TxA2 release.ConclusionsAnti-GPVI antibodies inhibit atherosclerotic plaque-induced platelet aggregation under static and flow conditions more effectively than GPVI-Fc. However, potent platelet inhibition by GPVI-Fc at a higher shear rate (1,500/s) suggests localized antithrombotic efficacy at denuded or fissured stenotic high-risk lesions without systemic bleeding. The compound-specific differences have relevance for clinical trials targeting GPVI-collagen interaction combined with established antiplatelet therapies in patients with spontaneous plaque rupture or intervention-associated plaque injury.

To delineate ligand binding and functional characteristics of the human B1 kinin receptor, a stable clone of Chinese hamster ovary cells expressing a single class of binding sites for [3H]des-Arg10-lysylbradykinin with a Kd of 0.3 nM and a Bmax of 38 fmol/mg protein ( approximately 40,000 receptors/cell) was isolated. Studies with peptide analogs showed that a lysine residue at position 1 (based on the lysylbradykinin sequence) of ligands was essential for high affinity binding to the human B1 receptor. In marked contrast to cloned Chinese hamster ovary cells expressing the human kinin B2 receptor, which internalized approximately 80% of the ligand within 5 min upon exposure to 2 nM [3H]bradykinin, exposure of cells expressing the B1 receptor to 1 nM [3H]des-Arg10-lysylbradykinin resulted in minimal ligand internalization. Stimulation of the B1 receptor led to inositol phosphate generation and transient increases in intracellular calcium, confirming coupling to phospholipase C, while immunoprecipitation of photoaffinity-labeled G-proteins from membranes indicated specific coupling of the receptor to Galphaq/11 and Galphai1,2. The B1, unlike the B2, receptor does not desensitize (as demonstrated by continuous phosphoinositide hydrolysis), enhancing the potential role of this receptor during inflammatory events.