β1 integrin−mediated signals are required for platelet granule secretion and hemostasis in mouse by Tobias Petzold, Raphael Ruppert, Dharmendra Pandey, Verena Barocke, Hannelore Meyer, Michael Lorenz, Lin Zhang, Wolfgang Siess, Steffen Massberg, and Markus Moser Blood Volume 122(15):2723-2731 October 10, 2013 ©2013 by American Society of Hematology

Characterization of platelets from β1 integrin mouse mutants. Characterization of platelets from β1 integrin mouse mutants. Platelet β1 integrin surface expression of wild-type (WT; β1+/+), HT (β1+/−), KO (β1−/−), TTAA (β1TTAA), and Hpm (β1Hpm) β1 integrin mice were analyzed by flow cytometry. (B) Geometric mean values of β1 integrin expression shown in panel A were corrected for isotype control and expressed as % of WT β1-integrin surface levels (n = 6 per group). (C) Platelet lysates were subjected to immunoblotting for β1-integrin, talin-1, kindlin-3, and glyceraldehyde-3-phosphate dehydrogenase (GAPDH). (D) Surface expression of β3 integrin, GPVI, and GPIX on platelets of indicated mouse strains. (E) Platelet β1 integrin activation was determined with the conformation-specific 9EG7 antibody via the use of flow cytometry upon stimulation with 50 ng/mL CVX, 0.1 U/mL thrombin, or 5 mM U46619. Acquired mean fluorescence intensity values were normalized to total β1 integrin expression levels and are shown as arbitrary unit (n = 3 for TTAA and Hpm, n = 4 for all other groups; bars represent mean values ± SEM; significance levels are indicated; *P < .05; **P < .01; n.s., not significant). (F) Active β3 integrins were determined with JON/A antibody after platelet stimulation with indicated stimuli. No significant difference between all tested groups was determined (n = 5 for HT and WT, n = 4 for all other groups; bars represent geometric mean values ± SEM). Tobias Petzold et al. Blood 2013;122:2723-2731 ©2013 by American Society of Hematology

β1 integrin−mediated spreading and outside-in signaling is independent of kindlin-integrin interaction. β1 integrin−mediated spreading and outside-in signaling is independent of kindlin-integrin interaction. (A) Spreading of thrombin-stimulated platelets on soluble collagen in the absence or presence of 0.75 mM Mn2+ and on fibrinogen. Representative pictures are shown 45 minutes after platelet seeding (scale bar represents 5 μm). (B) Spreading area of mutant platelets on soluble collagen is shown relative to wild-type (WT) platelets at 45 minutes (bars represent mean values ± SEM; significance level are indicated; *P < .05; **P < .01). (C) Integrin outside-in signaling was evaluated by FAK autophosphorylation at Tyr379. Washed platelets from wild-type (WT), β1 knockout (KO), and TTAA mice as well as from kindlin-3−deficient mice (K3(−/−)) were stimulated with 5 μg/mL fibrillar collagen and seeded on soluble collagen-coated surfaces in the absence or presence of Mn2+. To analyze basal FAK phosphorylation, unstimulated platelets were seeded on BSA. Cells were lysed after 30 minutes and subjected to immunoblotting for FAK Y397 phosphorylation and total FAK. Tobias Petzold et al. Blood 2013;122:2723-2731 ©2013 by American Society of Hematology

Reduced platelet adhesion and aggregate formation onto injured vessel walls of β1-null, β1TTAA, and β1hpm mice, whereas tail bleeding times of β1hpm mice are normal. Reduced platelet adhesion and aggregate formation onto injured vessel walls of β1-null, β1TTAA, and β1hpmmice, whereas tail bleeding times of β1hpmmice are normal. (A) Number of adherent fluorescently labeled platelets to the injured vessel wall was quantified after 5, 10, 15, and 60 minutes. Values are expressed as percentage of adherent wild-type (WT) platelets (n/mm2) at 5 minutes (WT, n = 9; HT, n = 6; knockout [KO], n = 6; TTAA, n = 4; Hpm, n = 8; bars represent mean values ± SEM in percent, significance level are indicated; *P < .05; **P < .01 by one-way analysis of variance followed by a Tukey multiple comparison test). (B) Tail-bleeding times of chimeric mice carrying the bone marrow of the indicated genotype (WT, n = 19; fl/fl, n = 6; HT, n = 10; KO, n = 9; TTAA, n = 15; TTAA/fl, n = 6; Hpm, n = 13; Hpm/fl, n = 6; Kindlin-3−/−, n = 4; cross line represents mean bleeding time and bars represent SEM, significance level are indicated; **P < .01). Tobias Petzold et al. Blood 2013;122:2723-2731 ©2013 by American Society of Hematology

Collagen-induced platelet aggregation and granule secretion requires β1 integrins. Collagen-induced platelet aggregation and granule secretion requires β1 integrins. (A) Washed platelets were stimulated with 5 μg/mL fibrillar collagen (black curves) or 50 ng/mL CVX (gray curves) and aggregation was recorded for 600 seconds. (B) Aggregation of β1-null (KO) and β1TTAA (TTAA) platelets in the presence of 100 μg/mL fibrinogen (gray curves) or 0.75 mM Mn2+ (black curves). (C) ATP release from platelet dense granules after stimulation of platelets with collagen. ATP was measured by luminescence after addition of luciferase luciferin reagent using the Lumi-aggregometer (Chronolog, Havertown, PA). Platelets were stimulated with either 5 μg/mL fibrillar collagen or 50 ng/mL CVX for 10 minutes. The relative luminescence after 8 minutes (ratio of collagen to CVX stimulated cells) is shown (bars represent mean values ± SEM of n = 3 independent experiments). (D) Serotonin release after stimulation with collagen. Platelets were stimulated with either 5 μg/mL fibrillar collagen or 50 ng/mL CVX for 8 minutes. The relative serotonin level (ratio of collagen to CVX stimulation) is shown (bars represent mean values ± SEM of n ≥ 3 independent experiments). (E) Quantification of the cellular fibrinogen content in resting platelets or 8 minutes after stimulation with either fibrillar collagen (5 μg/mL) or CVX (50 ng/mL; bars represent mean values ± SEM; significance levels are indicated; *P < .05; wild-type (WT), n = 3; knockout (KO), n = 3; TTAA, n = 3; hypomorphic (Hpm), n = 3; representative blots are shown in supplemental Figure 4B). (F) Platelet content of the α-granule cargo proteins vWF and thrombospondin-1 (TSP-1) in wild-type (WT) and β1-null (KO) platelets upon collagen and stimulation with CVX. (G) Ratio of PF-4 release from platelets stimulated with either 5 μg/mL fibrillar collagen or 50 ng/mL CVX for 8 minutes (bars represent mean values ± SEM of n ≥ 3 independent experiments). Tobias Petzold et al. Blood 2013;122:2723-2731 ©2013 by American Society of Hematology

Normal actin polymerization, Rac activation, and PAK phosphorylation in β1 hypomorphic platelets. Normal actin polymerization, Rac activation, and PAK phosphorylation in β1 hypomorphic platelets. (A) Platelets from wild-type (WT), β1-null (KO), and hypomorphic (Hpm) mice were stimulated with 5 µg/mL fibrillar collagen for 60 seconds in the presence of 0.5 mM RGDS peptide, and the relative F-actin content was measured by flow cytometry. Curves represent mean ± SEM increase in FITC phalloidin MFI (n = 3, significance level are indicated). (B) Same platelet populations were stimulated with 5 µg/mL fibrillar collagen for 30 and 60 seconds lysed and subjected to Rac-1 GTP pulldown experiments. Total Rac-1 loading is shown below. (C) Quantification of GTP-bound active Rac-1 pulldown experiments at 60 seconds (bars represent mean values ± SEM; significance level are indicated; *P < .05; **P < .01; wild-type (WT), n = 4; β1-null (KO), n = 3; hypomorphic (Hpm), n = 4). (D) Platelets from WT, KO, and Hpm mice were stimulated with 5 µg/mL fibrillar collagen for the indicated time points, lysed and subjected to immunoblotting for p-Thr432/402 PAK1/2 and p-Ser19 myosin light chain (pMLC). Total MLC is shown as loading control. Tobias Petzold et al. Blood 2013;122:2723-2731 ©2013 by American Society of Hematology