Tick saliva protein Evasin-3 modulates chemotaxis by disrupting CXCL8 interactions with glycosaminoglycans and CXCR2

Abstract

Chemokines are a group of chemotaxis proteins that regulate cell trafficking and play important roles in immune responses and inflammation. Ticks are blood-sucking parasites that secrete numerous immune-modulatory agents in their saliva to evade host immune responses. Evasin-3 is a small salivary protein that belongs to a class of chemokine-binding proteins isolated from the brown dog tick, Rhipicephalus sanguineus Evasin-3 has been shown to have a high affinity for chemokines CXCL1 and CXCL8 and to diminish inflammation in mice. In the present study, solution NMR spectroscopy was used to investigate the structure of Evasin-3 and its CXCL8-Evasin-3 complex. Evasin-3 is found to disrupt the glycosaminoglycan-binding site of CXCL8 and inhibit the interaction of CXCL8 with CXCR2. Structural data were used to design two novel CXCL8-binding peptides. The linear tEv3 17-56 and cyclic tcEv3 16-56 dPG Evasin-3 variants were chemically synthesized by solid-phase peptide synthesis. The affinity of these newly synthesized variants to CXCL8 was measured by surface plasmon resonance biosensor analysis. The K_d values of tEv3 17-56 and tcEv3 16-56 dPG were 27 and 13 nm, respectively. Both compounds effectively inhibited CXCL8-induced migration of polymorphonuclear neutrophils. The present results suggest utility of synthetic Evasin-3 variants as scaffolds for designing and fine-tuning new chemokine-binding agents that suppress immune responses and inflammation.

Keywords: C-X-C motif chemokine ligand (CXCL); chemokine; nuclear magnetic resonance (NMR); peptide chemical synthesis; protein structure; protein-protein interaction.

Figure 1.

NMR analysis of binding CXCL8 to met-Evasin-3. A, the chemical shift perturbation plot of 200 μm [¹⁵N,¹³C]CXCL8 amide peaks upon binding of 200 μm met-Evasin-3 at 37 °C, pH 4.5. B, schematic representation of CXCL8 secondary structure according to the dimer crystal structure (PDB code 1IL8). C, the chemical shift perturbation plot of 200 μm [¹⁵N,¹³C] met-Evasin-3 amide peaks upon binding of 200 μm CXCL8 at 44 °C, pH 4.5. Predicted O-glycosylation sites are marked by circles, and N-glycosylation is shown by triangles. D, schematic representation of met-Evasin-3 secondary structure predicted by CSI 3.0 in the unbound form (top) and in the complex with CXCL8 (bottom). Edge β-strands are shown by dark blue arrows, interior β-strands are shown by light blue arrows, and turns are shown by square brackets. Δppm values are expressed as the sum of square roots (Δ¹⁵N_{free–complex})²/6.51 + (Δ¹H_{free–complex})².

Figure 2.

¹⁵N heteronuclear NOE relaxation values of [¹⁵N,¹³C]met-Evasin-3 in free form (light blue) and in the CXCL8–[¹⁵N,¹³C]met-Evasin-3 complex (dark blue).

Figure 3.

A, overlay of CXCL8 structures in the CXCL8 dimer (dark blue; PDB code 1IL8) and in the [¹⁵N,¹³C]CXCL8–met-Evasin-3 complex (gray). B, ribbon representation of the ensemble of 10 lowest energy structures of tEv3 17–56 (PDB code 6QJB); disulfides are shown by yellow sticks. C, cartoon representation of the HADDOCK structure of the CXCL8–tEv3 17–56 complex. tEv3 17–56 is shown in light blue, and CXCL8 is in gray.

Figure 4.

Chromatographic elution profiles of 0.1 mg/ml CXCL8 in absence (light blue) and presence (dark blue) of 1 mg/ml of low-molecular-mass heparin (Fragmin). SEC chromatograms taken immediately after the addition and after a 1-h incubation at 37 °C with 0.1 mg/ml met-Evasin-3 are depicted in gray and black, respectively.

Figure 5.

A, overview of tcEv3 16–56 dPG folding and purification. The chromatograms of crude peptide mixture after HF cleavage and after folding are shown in black and dark blue, respectively. The results of the LC-MS analysis of purified tcEv3 16–56 dPG is shown in light blue; the calculated mass of [MH]⁺ is 4552.02 Da, and the observed mass is 4552.06 Da. B, SPR biosensor analysis of tcEv3 16–56 dPG upon binding to immobilized human CXCL8. The binding curve is plotted using maximal response signal for each injection. The apparent K_d value is calculated by fitting the data to a steady-state affinity model using a linear component. The fitted K_d is 13 nm, and R_max = 16.4, χ² = 0.202. C, HPLC analysis of stability of tcEv3 16–56 dPG in human plasma at 37 °C.

Figure 6.

Effect of Evasin-3 variants on CXCL8-induced PMN migration. Shown is PMN migration as a result of the addition of 1 nm CXCL8 (n = 10) compared with the control without chemoattractant (n = 10) and compared with the effect of 10 nm Evasin-3 (n = 10), tEv3 17–56 (n = 9), or tcEv3 16–56 dPG (n = 5) on CXCL8-induced PMN migration.