FN-β3 integrin interaction promotes soft matrix–mediated proplatelet formation. (A) Quantification of MK forming proplatelets depending on the matrix stiffness and genotype; mean statistics: 1-way ANOVA and Tukey multiple comparisons, n = 3 independent experiments, each dot is the mean of at least 9 replicates ±SEM; ^§§P < .01; ^§§§P < .001; ^§§§§P < .0001 comparing the control with substrates of same stiffness. (B) Proplatelet formation by MK grown in liquid culture; (i) bright field observation of MK forming proplatelets; arrows indicate proplatelets; (ii) quantification of the proportion of Itgb3^–/– MKs extending proplatelets vs control MKs; mean ± SEM, no significant difference (t test), n = 5 independent cultures; (iii) quantification of the proportion of Itgb1b3^–/– MKs extending proplatelets vs control MKs; mean ± SEM, no significant difference; n = 3 independent cultures. (C) BM explant experiment. (Left) Phase contrast observation of MK forming proplatelets at the periphery of the tissue explant; arrows indicate proplatelets. (Right) Quantification of proplatelet formation. Mean ± SEM, statistics, 1-way ANOVA and Tukey multiple comparisons, n = 10 to 16 independent marrows. (D) Mouse platelet count. One-way ANOVA and Tukey multiple comparisons, mean ± SEM, n = 10 to 20 mice. (E) (i), BM sections of control and Itgb1b3^–/– mice immunolabeled for MKs von Willebrand Factor vWF, red and for vessels (FABP4, green); representative of at least 3 marrow sections. (ii) quantification of MKs per mm²; mean ± SEM, n = 5 mice, 3 sections per marrow and 4 fields of observation per section; (iii), distance between MKs and sinusoid vessels; mean ± SEM; n = 5 mice; 3 sections per marrow and 3 to 4 fields of observation per marrow section. No significant difference (t test).