A talin mutant that impairs talin-integrin binding in platelets decelerates αIIbβ3 activation without pathological bleeding

Abstract

Tight regulation of integrin affinity is critical for hemostasis. A final step of integrin activation is talin binding to 2 sites within the integrin β cytoplasmic domain. Binding of talin to a membrane-distal NPxY sequence facilitates a second, weaker interaction of talin with an integrin membrane-proximal region (MPR) that is critical for integrin activation. To test the functional significance of these distinct interactions on platelet function in vivo, we generated knock-in mice expressing talin1 mutants with impaired capacity to interact with the β3 integrin MPR (L325R) or NPLY sequence (W359A). Both talin1(L325R) and talin1(W359A) mice were protected from experimental thrombosis. Talin1(L325R) mice, but not talin(W359A) mice, exhibited a severe bleeding phenotype. Activation of αIIbβ3 was completely blocked in talin1(L325R) platelets, whereas activation was reduced by approximately 50% in talin1(W359A) platelets. Quantitative biochemical measurements detected talin1(W359A) binding to β3 integrin, albeit with a 2.9-fold lower affinity than wild-type talin1. The rate of αIIbβ3 activation was slower in talin1(W359A) platelets, which consequently delayed aggregation under static conditions and reduced thrombus formation under physiological flow conditions. Together our data indicate that reduction of talin-β3 integrin binding affinity results in decelerated αIIbβ3 integrin activation and protection from arterial thrombosis without pathological bleeding.

Figure 1

Generation of mice expressing talin1 mutants in platelets. (A) Sequencing chromatograms of mutated regions of Tln1(L325R) and Tln1(W359A) ES cells. Genomic DNA isolated from targeted ES cells was used as a template for polymerase chain reaction (PCR) using primers that amplified the mutated sequences, and PCR amplicons were sequenced. (B) Mouse breeding strategy to obtain mice with Tln1(L325R) and Tln1(W359A) expressing platelets. (C) Western blot analysis showing similar levels of talin expression in control (Tln1^wt/flCre⁺) and mutant (Tln1^L325R/flCre⁺ and Tln1^W359A/flCre⁺) platelets. Par4-AP (D) or convulxin-induced (E) αIIbβ3-integrin activation (JON/A-PE binding). Reduced levels of talin1 protein in Tln1^WT/fl Cre⁺ platelets had minimal effect relative to Tln1^fl/fl Cre^– platelets. Bar graphs represent MFI ± SEM (n = 6, 3 independent experiments). *P < .05.

Figure 2

Analysis of thrombosis and hemostasis in talin1 mutant mice. (A) Tln1^W359A/fl Pf4-Cre⁺ (Tln1^WA/flCre⁺) and Tln1^L325R/fl Pf4-Cre⁺ (Tln1^LR/flCre⁺) mice are protected from FeCl₃-induced thrombosis of the carotid artery. Time to vessel occlusion was determined using a Doppler flow probe after 3 minutes of application of 10% FeCl_3. (B) The incidence of gastrointestinal bleeding in talin mutant and control mice was determined using a guaiac-based hemoccult test. Numbers of hemoccult-positive/total mice are shown for each group. (C) Bleeding times and (D) blood loss volume in the indicated mice after tail resection (n = 17-56 mice/group). *P < .05, **P < .01, ***P < .001.

Figure 3

Platelets expressing talin1(W359A) exhibit partially impaired αIIbβ3 activation. (A-B) αIIbβ3-integrin activation (JON/A-PE binding) was measured in washed platelets isolated from mice with the indicated genotypes. Platelets were stimulated for 10 minutes with increasing concentrations of Par4-AP (A) or convulxin (B), stained with JON/A-PE, and immediately analyzed by flow cytometry. Bar graphs represent MFI ± SEM (n = 6, 3 independent experiments). (C) Representative images of rhodamine-phalloidin–stained talin1 mutant platelets spread on fibrinogen-coated glass in the presence of 100 μM ADP for 45 minutes. (D) Quantitation of platelet area (μm²); n = 5 independent experiments; mean ± SEM. *P < .05, **P < .01, ***P < .001.

Figure 4

Talin1(W359A), but not talin1(L325R), impairs binding of talin to the β3 integrin tail. (A) Representation of the β3 integrin-talin complex structure (PDB 2H7E). Shown in red is the ribbon view of the β3 tail; light gray is the surface view of the talin F3 subdomain. Talin residues leucine 325 (L325) and tryptophan 359 (W359) are shown in green and blue, respectively. (B) Representative BiaCORE sensorgrams of THD binding to immobilized β3-integrin cytoplasmic domain (tail). Biotinylated β3 tail was immobilized to a neutravidin sensor chip. Indicated concentrations of wild-type or mutant talin head were injected, and response curves were measured as the difference between the experimental chamber and a reference chamber lacking immobilized β3 integrin. (C) Association rate constants (k_a), dissociation rate constants (k_d), and equilibrium dissociation constants (K_D) between β3-integrin tail and THD mean ± SEM (n = 3).

Figure 5

Expression of talin1(W359A) causes decelerated αIIbβ3 activation in platelets. The kinetics of αIIbβ3 activation were assessed in real time by flow cytometry. Jon/A-PE and Par4-AP were added simultaneously (arrow) to platelets of the indicated genotype. Jon/A-PE binding (integrin activation) was monitored continuously for 10 minutes. Tln1^W359A/flCre⁺ (Tln1^WA/flCre⁺) platelets were compared with (A) Tln1^wt/wtCre⁺(WT), Tln1^wt/flCre⁺, Tln1^fl/flCre⁺, and Tln1^L325R/flCre⁺ (Tln1^LR/flCre⁺) platelets, and CalDAG-GEFI^−/− (CDGI^−/−) platelets. Traces are representative of 3 independent experiments. (B) Real-time Jon/A-PE binding data are shown normalized for maximum binding within each group. Maximum values for the indicated groups were calculated as the average MFI over the final 10 seconds of the 10-minute assay. (C) Maximum velocity of αIIbβ3 activation, determined as the maximal rate of change of MFI over time (ΔMFI/minute). *P < .05, **P < .01, ***P < .001.

Figure 6

Delayed aggregation in platelets expressing talin1(W359A). (A) Aggregation response of washed Tln1^wt/flCre⁺ (black line), Tln1^W359A/flCre⁺ (gray line), Tln1^L325R/flCre⁺ (gray dashed line), and Tln1^fl/flCre⁺ platelets (black dashed line) stimulated with 5 μg/mL (LD) collagen, 25 μg/mL (HD) collagen, 1 μM U46619, 200 ng/mL convulxin, 200 μM (LD), or 600 μM (HD) Par4-AP. (B) Calcium mobilization in platelets labeled with the calcium-sensitive dye Fluo-4 and stimulated with Par4-AP in the presence of 1 mM Ca²⁺. (C) Time course of Rap1 activation in platelets stimulated with Par4-AP. The bottom panel shows total Rap1 as a loading control. Results are representative of 3 independent experiments.

Figure 7

The effects of fluid shear stress on talin1 mutant platelet thrombus formation ex vivo. Platelets in heparinized whole blood were labeled with anti–GPIX-Alexa488 and perfused over fibrillar collagen type I at low (100^–s) or high (1200^–s) shear rates. Adhesion of platelets was monitored continuously with a Nikon Eclipse TE300 inverted microscope (Nikon Instruments Inc., Melville, NY). (A-B) Representative images of platelet adhesion after 10 minutes of perfusion. (C-D) Image analysis. Platelet adhesion over collagen was quantified by measuring surface area coverage (percentage of total area) with Slidebook 5.0 software. Graphs show mean ± SEM (6 independent experiments). *P < .05, **P < .01, ***P < .001.