Fig 2: Analysis of secretion in Vti1b-/- platelets. A, expression levels of other known SNAREs and SNARE-associated proteins were measured in Stx8-/- platelets to ensure the defect results from Stx8 deletion. Levels of STX11, VAMP8, SNAP23, and MUNC13d and 18b were comparable between the genotypes (image representative of at least 3 independent experiments). B, a significant decrease in expression of Vti1b in Stx8-/- platelets was observed (31.13 ± 5.85% of WT expression, p < 0.01). C, similarly, Stx8 was also down-regulated in Vti1b-/- platelets. D, to confirm that only complete ablation of Stx8 leads to defective dense granule secretion, we analyzed ATP secretion from Stx8+/- platelets and found no defect, unlike in Stx8-/- platelets (n > 3, difference not significant but consistent for WT versus Stx8-/-, two-way ANOVA). E–G, to further confirm that deletion of Stx8 and not the associated down-regulation of Vti1b was responsible for defect in dense granule secretion observed, we measured secretion from dense (E) and a-granules (F) and lysosomes (G) in Vti1b-/- platelets. No difference was observed in any of the secretion events in Vti1b-/- platelets (n = 6). Error bars represent S.E.

Fig 3: Gp91phox colocalizes with SNAREs on the phagosomal membrane. (A) Confocal images of dendritic cells stimulated with zymosan particles for 1 h and immunostained for gp91phox (green in merge) together with Vti1b, stx7, VAMP8, stx8, SNAP23, stx12 or stx5 (magenta). Yellow arrowheads, phagosomes positive for both gp91phox and the indicated SNARE. BF, brightfield. Scale bar: 5 µm. (B) Pearson correlation coefficients for colocalization of gp91phox with the indicated SNAREs for zymosan-containing phagosomes (from panel A) and for unstimulated cells (w/o zymosan; Fig S2A). Colocalization of gp91phox immunostained with both Alexa-Fluor-488 and -568-labeled secondary antibodies is shown as a positive control (individual donors shown). (C) Significance levels (one-way ANOVA with Tukey's test) from panel B. Black text, zymosan-pulsed samples; red text, unstimulated samples. ns, not significant. (D) Normalized staining intensity values of phagosomal gp91phox as a function of normalized intensity values of the indicated SNAREs for individual phagosomes (log2 scale; phagosomes pooled from >3 donors are shown; solid lines, linear regression with 95% confidence intervals; ß, regression coefficients; R2, R-squared values). (E) Quantification of the phagosomal enrichment for gp91phox and/or the indicated SNAREs (from panel D, with normalized intensity values >1). (F) Phagosomal enrichment of gp91phox and the indicated SNAREs following stimulation with zymosan (n=6 donors for gp91phox; n=3 donors for the SNAREs). Results show means±s.e.m.

Fig 4: Analysis of aggregation and ADP rescue experiments in Stx8-/- platelets. A, aggregation was measured using lumi-aggregometer and amplitude was expressed as percentage decrease in optical density of the sample at t = 180 s (n = 4). The concentration-response curve for Stx8-/- was shifted to the right, but [thrombin] EC50 was not significantly different (95% CI of WT EC50: 0.042 to 0.090, 95% CI of Stx8-/- EC50: 0.069 to 0.101, p > 0.05). However, the maximum aggregation amplitude was significantly reduced in Stx8-/- at 0.075 units/ml of thrombin (26.1 ± 13.6% decrease in optical density in Stx8-/- versus 60.8 ± 18.2% in WT, p = 0.021, two-way ANOVA with Bonferroni post-test), with aggregation at 0.05 and 0.06 units/ml also consistently reduced, but not significantly. B, integrin aIIbß3 activation was measured by FACS, no difference in PE-JON/A binding was observed between genotypes suggesting an alternative mechanism for the observed aggregation defect. C, representative aggregation traces showing the effect of co-stimulation with 10 µm exogenous ADP response to EC50 concentration of thrombin (i) and CRP (ii). D, aggregation in response to co-stimulation with ~EC50 (0.05–0.075 units/ml) thrombin and 10 µm ADP was compared with EC50 thrombin alone (n = 5). Stx8-/- aggregation was significantly reduced in response to thrombin alone (37.8 ± 8.9 versus 63.4 ± 6.1% decrease in optical density, respectively, p < 0.05) and could be fully rescued with ADP (p > 0.05). The maximum aggregation to 1 unit/ml of thrombin was the same between genotypes (data not shown). Similarly, the effect of co-stimulation with ADP and EC50 CRP (0.3–0.5 µg/ml) was measured (n = 4). There was a significant reduction in aggregation in response to ~EC50 CRP in Stx8-/- platelets (30.0 ± 7.3 versus 46.3 ± 8.6% in WT, p < 0.05) that was fully rescued with ADP (p > 0.05). Maximum response (to 5 µg/ml of CRP) was the same (data not shown). E, secretion of ATP was also measured. Again, secretion in response to EC50 concentrations of thrombin or CRP was significantly reduced (p < 0.01), whereas co-stimulation with ADP could only partially rescue secretion when thrombin was used as primary agonist (p < 0.05). In the case of co-stimulation with CRP, ADP enhanced ATP secretion to a much greater extent in both WT and Stx8-/- platelets (ADP enhancement of 185.75 ± 42.55% of CRP alone, compared with 111.30 ± 20.66% of thrombin alone). Error bars represent S.E.

Fig 5: Quantification of granule secretion in Stx8-/- platelets. All concentration-response data were analyzed by appropriate curve fitting in GraphPad Prism. A, mouse platelets were stimulated in the aggregometer as described. Dense granule secretion was measured as an increase in luminescence relative to ATP standard for each stimulation. Data were normalized to maximum ATP secretion of WT platelets on the day of experiment. There was a significant increase in [thrombin] EC50 in Stx8-/- (95% CI of EC50: 0.085 to 0.11 units/ml) compared with WT platelets (95% CI of EC50: 0.068 to 0.086 units/ml) (n = 5, variable slope sigmoidal dose-response curve, p = 0.0248). Stx8-/- and WT secretion in response to 0.06 units/ml was, respectively, 1.742 ± 0.618 and 10.379 ± 3.767% of the maximum WT secretion on the day (n = 4, p > 0.05, ns), and in response to 0.075 units/ml of thrombin: 6.927 ± 5.970 and 38.933 ± 20.969% of the maximum WT secretion, p = 0.0339). B, the total 5-HT content was measured using 3H-labeled 5-HT as described. There was no significant difference in 5-HT content between genotypes (n = 3). C, a-granule secretion was measured by FACS. Median fluorescent intensity (MFI) of FITC-CD62P was quantified in response to increasing concentrations of thrombin. MFI was normalized to the maximum MFI of WT platelets on the day of experiment (n = 4). D, lysosome secretion was measured by ß-hexosaminidase enzyme activity in supernatants following stimulation with increasing concentrations of thrombin. Absorbance at 405 nm was measured, and values were expressed as percentage of TOTAL control for each subject (n = 6). There was no difference in [thrombin] EC50 between genotypes in a-granule or lysosome secretion (p > 0.05, variable slope sigmoidal dose-response curve). Error bars represent S.E.