Aegyptin displays high-affinity for the von Willebrand factor binding site (RGQOGVMGF) in collagen and inhibits carotid thrombus formation in vivo

Abstract

Aegyptin is a 30 kDa mosquito salivary gland protein that binds to collagen and inhibits platelet aggregation. We have studied the biophysical properties of aegyptin and its mechanism of action. Light-scattering plot showed that aegyptin has an elongated monomeric form, which explains the apparent molecular mass of 110 kDa estimated by gel-filtration chromatography. Surface plasmon resonance identified the sequence RGQOGVMGF (where O is hydroxyproline) that mediates collagen interaction with von Willebrand factor (vWF) as a high-affinity binding site for aegyptin, with a K(D) of approximately 5 nM. Additionally, aegyptin interacts with the linear peptide RGQPGVMGF and heat-denatured collagen, indicating that the triple helix and hydroxyproline are not a prerequisite for binding. However, aegyptin does not interact with scrambled RGQPGVMGF peptide. Aegyptin also recognizes the peptides (GPO)(10) and GFOGER with low affinity (microM range), which respectively represent glycoprotein VI and integrin alpha2beta1 binding sites in collagen. Truncated forms of aegyptin were engineered, and the C-terminus fragment was shown to interact with collagen and to attenuate platelet aggregation. In addition, aegyptin prevents laser-induced carotid thrombus formation in the presence of Rose Bengal in vivo, without significant bleeding in rats. In conclusion, aegyptin interacts with distinct binding sites in collagen, and is useful tool to inhibit platelet-collagen interaction in vitro and in vivo.

Fig. 1. Biophysical properties of aegyptin

(A) Chromatographic analysis of aegyptin by size exclusion TSK gel G3000PW (arrow, apparent molecular mass 110 kDa) is superimposed to the elution pattern of molecular mass markers (in blue). Inset shows SDS-PAGE of purified recombinant aegyptin (arrowhead). Molecular mass standards used: thyroglobulin (670 kDa), immunoglobulin (158 kDa), ovalbumin (44 kDa), myoglobin (17 kDa) and vitamin B12 (1.4 kDa). (B) Inline multi-angle light scatter. The solid and blue lines represent absorbance 280 nm and MALS results, respectively. Inset shows the results between 10–20 mls. (C) CD spectra of aegyptin. Inset shows the proportion of α-helix, β-sheet, β-turn and unordered structures.

Fig. 2. Aegyptin interaction with collagen

(A) Surface plasmon resonance. Sensorgrams in black lines show aegyptin (in nM: a, 20; b, 10; c, 5; d; 2.5; e, 1.25) binding to immobilized soluble collagen type I. Data fitting using global two-state binding model is displayed as red line. (B) Sensograms show collagen (in nM: a, 5; b, 2.5, c, 1.25; d, 0.625; e, 0.3; f, 0.15 and g, 0.075) binding to immobilized aegyptin. (C) Solid phase binding assay. Aegyptin (0–1 µM) was incubated with immobilized collagen and binding estimated with anti-his mouse monoclonal antibody as described in Methods. (D) Fluorescence microscopy. Cover slips coated with fibrillar collagen were incubated with aegyptin-FITC for 20 min at room temperature and analyzed under fluorescence microscope (right upper panel), as described in Methods. Collagen incubated with PBS (negative control) did not display auto fluorescence under the same conditions (right lower panel). DIC for each condition is shown in the left lower and upper panels.

Fig. 3. Aegyptin displays high affinity for vWF binding site of collagen

Sensorgrams show aegyptin binding to immobilized (A) cross-linked RGQOGVMGF, (B) linear RGQOGVMGF, (C) cross-linked hydroxyproline-less RGQPGVMGF, (D) linear hydroxyproline-less RGQPGVMGF, and (G) heat-denatured collagen (90 min at 98°C). In (E) aegyptin was injected in different flow cells of the same sensor chip containing immobilized scrambled RGQPGVMGF, or collagen type III or RGQPGVMGF. Concentrations of recombinant aegyptin for (A)–(D) were (in nM: a, 50; b, 25; c, 12.5; d, 6.75; e, 3.1), for (E) was 1 µM; and for (G) was (in nM: a, 150; b, 75; c, 37.5; d, 18; e, 9; f, 4.5). Dissociation of aegyptin-ligand complex was monitored for 1800 seconds, and a global two-state reaction model was used to calculate kinetic parameters. (F) Inhibition of vWF binding to cross-linked RGQOGVMGF was estimated by ELISA in the presence of indicated concentrations of aegyptin.

Fig. 4. Aegyptin displays low affinity for GPVI or integrin α2β1 binding sites of collagen

Sensorgrams shows aegyptin binding to immobilized (A) cross-linked (GPO)₁₀ or (B) cross-linked GFOGER. Aegyptin concentration for (A) was (in µM: a, 2; b, 1.5; c, 1; d, 0.75; e, 0.5; and f, 0.25) and for (B) was (in µM: a, 3; b, 2; c, 1; d, 0.5; e, 0.3; and f, 0.15). Dissociation of aegyptin-ligand complex was monitored for 1800 seconds, and a global two-state reaction model was used to calculate kinetic parameters. (C) Functional assay using human platelet-rich plasma shows that aegyptin is ineffective to inhibit platelet responses to (GPO)₁₀ (2.5 µg/ml) but prevents collagen (2 µg/ml)-induced platelet aggregation. (D) Aegyptin failed to prevent washed human platelet adhesion to GFOGER under static conditions, but effectively inhibits platelet adhesion to collagen. No adhesion was detectable in the presence of EDTA.

Fig. 5. C-terminal-2 fragment of aegyptin binds to collagen

(A) Diagram of the constructs used for cloning and expression. (B) SPR experiments show C-terminus 2 fragment binding to aegyptin. (C) Sensorgrams of C-terminus-2 fragment (in nM: a, 250; b, 120; c, 60; d, 30; e, 15 and f, 5) binding to immobilized soluble collagen type I. Dissociation of aegyptin-collagen complex was monitored for 1800 seconds, and a global two-state binding model was used to calculate kinetic parameters. (D) Human platelet-rich plasma (2 × 10⁵/µl) was incubated with C-terminus-2 fragment (in µM: a, 0; b, 3; and c, 10) for 1 minute followed by addition of fibrillar Horm collagen (2 µg/ml, final concentration). Platelet aggregation was estimated by turbidimetry under test-tube stirring conditions. The tracings represent a typical experiment.

Fig. 6. Aegyptin prevents thrombus formation in vivo

(A) aegyptin (50 or 100 µg/kg) or PBS (control) was injected in the cava vein of rats and thrombosis was induced by slow injection (over 2 min) of 90 mg/kg body weight of Rose Bengal dye into the cava vein at a concentration of 60 mg/mL. Before injection, green light laser was applied to the desired site of injury from a distance of 3 cm and remained on for 80 minutes or until stable occlusion occurred. The number of animals tested for each condition is shown in the Figure. (B) Determination of the bleeding. Aegyptin at the indicated doses was administered i.v.; after 15 min of administration, the rat tail was cut 2 mm from the tip. The tail was carefully immersed in 40 ml of distilled water at room temperature, and blood loss (hemoglobin content) was estimated at 540 nm, after 60 min, as compared to a standard curve. Animals that received PBS were taken as control. In some experiments animals received heparin (1 mg/kg). Data represent the mean ± SEM of results obtained with 7 to 10 animals. *, P < 0.05.