Selective Inhibition of Integrin β3 Topology Provides a Safer Antithrombotic Strategy

Figures

• 

  FIGURE 1. Phage display screening for scFv clones against the β‐tail domain of integrin β3. (A) A hypothetical diagram illustrating a TMD topology‐specific blocker. (B) Positive scFv clones from four independent phage display screens were tested for binding to purified GST (control, ct) or GST‐βTD (βT). Binding levels were measured by ELISA and displayed as a white‐to‐black gradient, as indicated for each screen. Clone numbers are indicated to the left (04‐24) and right (28‐48) of the columns. (C) SDS‐PAGE analysis of representative clones purified as MBP fusions (arrow). Purification yields are indicated by a pink gradient. The bands indicated with the single (*) and double asterisk (**) are contaminants, presumably MBP‐VH and MBP, respectively. (D) Flow cytometry plots showing scFv (2 µm) binding to CHO and CHO/αIIbβ3 cells. Anti‐HA antibody detected scFv binding, and D57, an integrin αIIbβ3‐specific antibody, distinguished CHO/αIIbβ3 cells. Affinity (mean fluorescence of CHO/αIIbβ3 binding minus that of control clone S0‐00) and specificity (CHO/αIIbβ3‐to‐CHO binding ratio) are shown as brown and orange gradients, respectively. (E) Schematic representation of the orbital shaking‐induced…

• 

  FIGURE 2. scFv clone selection and antibody production. (A) Petal plots showing five properties of clones identified during screening. Clones with superior properties, S2‐07 and S3‐17, are highlighted in bold. (B) Schematic of the laminar flow‐induced fibrinogen (Fg) binding assay. (C) Mouse platelets attached to a collagen‐coated flow chamber were perfused with HBSS (10 mL/min, 40 dyne/cm2) containing biotinylated fibrinogen and 0.5 µm scFv for 5 min. Bound fibrinogen was analyzed as described in Figure 1F. (D) Fibrinogen binding was normalized to the S0‐00‐treated condition and shown as a bar graph (n = 4). Error bars indicate the mean ± SEM. ∗∗, p < 0.01; ∗∗∗, p < 0.001 (one‐way ANOVA using Tukey's multiple comparison test). (E) ELISA analysis of 2–07Ab and 3–17Ab binding to purified β3 βTD. Binding curves were fitted using a one‐site binding model. Control human Fc (hFc) binding is shown as squares. (F,G) Fibrinogen binding to mouse platelets under orbital shaking in the presence of 1 µm of the indicated antibody was analyzed by Western blotting. Quantified fibrinogen binding normalized to control hFc is shown (n = 3). Error bars indicate the mean ± SEM. ∗∗, p < 0.01; ∗∗∗, p < 0.001…

• 

  FIGURE 3. Effects of 3–17Ab on force‐induced exposure of the Leu694 residue in integrin β3. (A) A hypothetic illustration of force‐induced movement of integrin β3 βTD‐TMD. Three residues mutated into cysteines are shown. (B) CHO cells transfected with integrin αIIb and each β3 cysteine mutant were treated with a cell‐impermeable thiol‐reactive biotinylation reagent under static or orbital shaking conditions. After cell lysis, biotinylated integrin β3 was captured using streptavidin‐conjugated beads and analyzed by Western blotting. (C) Quantification of biotinylated integrin β3, normalized to the Glu683Cys mutant β3 (n = 4). Error bars indicate the mean ± SEM. ∗∗, p < 0.01 (one‐way ANOVA using Tukey's multiple comparison test). (D,E) Effects of 1 µm 2–07Ab on the force‐induced exposure of the Leu694Cys (n = 4). Error bars indicate the mean ± SEM. ∗, p < 0.05; ns, not significant (one‐way ANOVA using Tukey's multiple comparison test). (F,G) Effects of 1 µm 3–17Ab or abciximab on the force‐induced exposure of the Leu694Cys were analyzed as described in (D,E) (n = 7). Error bars indicate the mean ± SEM. ∗∗, p < 0.01; ∗∗∗, p < 0.001; ns, not significant (one‐way ANOVA using Tukey's multiple…

• 

  FIGURE 4. Effects of 3–17Ab on integrin β3 TMD embedding in the lipid bilayer. (A) Schematic diagram of fluorescence‐based assay using βTD‐TMD‐tail (Ile721Cys) embedded in nanodiscs. (B) Fluorescence emission spectra of bimane‐labeled βTD‐TMD‐tail (Ile721Cys) nanodiscs in the absence or presence of 10 µm THD with 1 µm control Fc (solid line) or 3–17Ab (dashed line). Mean fluorescence from three measurements was normalized to the peak intensity of the hFc control group. (C) Fluorescence emission spectra of mero60‐labeled βTD‐TMD‐tail (Ile721Cys) nanodiscs in the absence or presence of 100 µm EGCG with 1 µm control Fc (solid line) or 3–17Ab (dashed line). (D) CHO/αIIbβ3 cells were treated with 100 µm EGCG and 1 µm of each antibody for 30 min. PAC1 binding to cells was measured using flow cytometry, and the mean PAC1 fluorescence is shown as a bar graph (n = 6). Error bars indicate the mean ± SEM. ∗∗∗, p < 0.001 (one‐way ANOVA using Tukey's multiple comparison test). (E) Illustration of the interaction between integrin β3 βTD and scFv fused with MBP (left). Coomassie‐stained SDS‐PAGE of scFv before and after TEV cleavage is shown (right). The arrow and arrowhead indicate full‐length and…

• 

  FIGURE 5. Effects of 3–17Ab on ex vivo platelet functions. (A) Schematic diagram of stenotic flow chamber setup (left) and mouse whole blood perfusion assay (right). (B) Representative fluorescence images of the stenotic region (112 s) following perfusion in the presence of either hFc or 3–17Ab. Platelets were incubated with 1 µm antibody together with 1 µm DiOC6. Scale bar, 200 µm. (C,D) Quantification of the mean fluorescence intensity (C) and total area of platelet aggregates (D) in the stenotic region over time during perfusion (n = 5). Shaded areas represent mean ± SEM. ∗∗∗, p < 0.001 (repeated measures two‐way ANOVA). (E) Mouse platelets were incubated on a collagen‐coated surface in the presence of 1 µm antibody. Adherent platelets were fixed, stained with rhodamine‐phalloidin, and counted (n = 15 microscope fields from three independent experiments). Error bars indicate the mean ± SEM. ns, not significant (unpaired t‐test). (F) Mouse platelets were pretreated with 1 U/mL of α‐thrombin (or PBS), together with either 1 µm antibody or 100 µm eptifibatide, and incubated on a fibrinogen‐coated surface for 30 min. Adherent platelets were fixed and stained with rhodamine‐phalloidin, and…

• 

  FIGURE 6. Effects of 3–17Ab on in vivo platelet functions. (A) Schematic illustration of FeCl3‐induced carotid artery thrombosis model. (B,C) Representative blood flow traces showing the average flow rate (1.7 s intervals) in the carotid artery after injection of 10 mg/kg hFc (B) or 3–17Ab (C). Signal amplitude was normalized to the average blood flow from the initial 10 sec after FeCl3 application. (D) Kaplan–Meier plot showing the proportion of mice maintaining normal blood flow over time. Occlusion was defined as a sustained drop in normalized blood flow < 0.1 for >30 s (n = 8). Dotted lines indicate 95% confidence intervals. ∗∗p < 0.01 using the log‐rank (Mantel–Cox) test. (E) Representative histological images of carotid artery sections from each group, with platelets visualized by intravenous injection of 0.05 µmol/kg DiOC6 prior to surgery. Scale bar, 100 µm. (F) Mice were intravenously injected with 0.275 mg/kg collagen and 1.2 µg epinephrine, together with 10 mg/kg control hFc or 3–17Ab, as indicated. One hour later, ears were harvested for whole‐mount immunostaining. (G) Z‐stack (1 µm) images stained for CD31 (endothelium, red) and CD41 (platelets, green) were merged using ImageJ…