GPVI expression is linked to platelet size, age, and reactivity

Figures

• 

  Figure 1.. Thrombus formation with different platelet fractions. Platelet- and leukocyte-depleted whole blood was reconstituted with unsorted, large, small, or small (mass adjusted) platelets and perfused for 4 minutes at a wall shear rate of 1000 s−1 over a coated microspot containing collagen type I. Thrombus build-up was evaluated by image capturing every 30 seconds during the 4 minutes of blood perfusion and end point analysis after postlabeling and perfusion. (A) Representative brightfield images (upper row) and fluorescent images of DiOC6 (middle row) or P-selectin expression (lower row) of the different platelet fractions after 4 minutes of perfusion. (B-C) Quantification of thrombus build-up (kinetics, n = 6) with brightfield images of platelet deposition (B) and fluorescence images of DiOC6 deposition (C). Correction for platelet surface differences (∼30% between large and small platelets10) would increase surface area coverage (SAC) of small platelets to similar values as the determined SAC for mass-adjusted small platelets. (D-E) Quantification of endpoint analysis (n = 6) of brightfield images of multilayer deposition (D) and fluorescence images of P-selectin expression (E). (F)…

• 

  Figure 2.. Population analysis of multicolor flow cytometry data with platelet surface and activation markers on the size-separated platelet fractions. Isolated, size-separated platelets were stimulated with 5 µg/mL CRP-XL, 20 µM TRAP-6, or 5 µM 2-MeSADP and stained simultaneously for PAC1, GPVI, CD36, CD63, and GPIX for 15 minutes before fixation and measuring 10 000 events per sample. Clustering tools (FlowSOM) and a dimensionality reduction algorithm (tSNE) were used to analyze and visualize these multicolor flow cytometry data. (A) Representative tSNE plots with a FlowSOM population filter of the different platelet fractions (unsorted, large, and small) under different stimulations. A heatmap showing the individual profiles of platelet populations per sample is shown in supplemental Figure 2. (B) Characterization of the FlowSOM clusters (5 populations; 1-5). (C-E) Distribution profile of platelet populations upon stimulation with CRP-XL (C), TRAP-6 (D), or 2-MeSADP (E), respectively, resulting from FlowSOM analysis for the different platelet size fractions. Mean ± standard error of mean; n = 4. *P < .05.

• 

  Figure 3.. GPVI-rich platelets are more reactive than GPVI-poor platelets, regardless of size. PAC1 (integrin activation) (A), CD63 (dense granule secretion) (B), and P-selectin (α-granule release) (C) expression levels were determined for GPVI-rich (top 25% expressing) platelets and for GPVI-poor (rest of platelets; 75%) platelets after 15 minutes of stimulation with CRP-XL (5 µg/ml), 2-MeSADP (5 µM), or TRAP-6 (20 µM) for the different size fractions. Mean ± standard error of mean; n ≥ 4.

• 

  Figure 4.. Population analysis of platelet surface markers on the size-separated platelet fractions. Isolated, size-separated platelets were stained for CD36, CD42b, CD61, GPVI, and HLA for 15 minutes before fixation, and 10 000 events were recorded per sample. Clustering tool (FlowSOM) and dimensionality reduction algorithm were used to analyze and visualize these multicolor flow cytometry data. (A) Representative dimensionality reduction algorithm plots with a FlowSOM populations filter for the different platelet fractions. (B) Characterization of the FlowSOM clusters (3 populations; I-III). (C) Distribution profile of platelet populations resulting from FlowSOM analysis for the different platelet size fractions. (D) The proportion of RNA-positive platelets was determined by T-oligonucleotides in separated large and small platelets. (E) Correlation of RNA-content and HLA-I expression was determined by costaining using T-oligonucleotides and HLA-A,B,C-antibody in separated large and small platelets. (F) Relation between RNA content and GPVI expression. Mean ± standard error of mean; n ≥ 4. *P < .05, **P < .01. MFI, mean fluorescence intensity.

• 

  Figure 5.. Signal transduction is carried by RNA-positive platelets. Signal transduction was determined by the percentage of pAKT-AF488–positive cells after stimulation with 5 µg/mL (final) collagen in large and small platelets (A) and RNA-positive and RNA-negative platelets (B) as determined by costaining with T-oligonucleotides and pAKT-AF488 antibody. (C) Maximum pAKT-AF488 positive large and small as well as RNA-positive and RNA-negative platelets after stimulation with 5 µg/mL (final) collagen. Mean ± standard error of mean; n ≥ 4. **P < .01. pAKT, phosphorylated AKT; n.s., not significant.

• 

  Figure 6.. Platelet function and GPVI expression decrease during platelet storage. Quantification of GPVI (A), P-selectin expression of platelets stimulated with CRP-XL 5 µg/mL (B). (C) Immature platelet fraction. Platelet deposition (D), Thrombus multilayer formation (E) and P-selectin expression (F), each given as a percent surface area coverage (SAC). Platelets were reconstituted in platelet- and leukocyte-depleted whole blood perfused over collagen-coated surfaces. Platelet concentrates were prepared from buffy coats, split and stored at room temperature (RT, red lines) or 4°C (blue lines). n ≥ 4. *P < .05, **P < .01. Blue and red asterisks represent significant changes over time; black asterisks represent significant differences between storage conditions.

• 

  Figure 7.. Findings confirmed in patients with higher proportion of immature platelets (ITP patients). (A) Platelet indices per donor, including count, mean platelet volume (MPV), flow cytometric forward scatter median fluorescence intensity (MFI), proportion of RNA-rich platelets obtained as in supplemental Figure 5, immature platelet fraction (IPF), and whether donor received thrombopoietin (TPO) receptor agonist. (B-D) Flow cytometric measurements of expression levels of GPVI and HLA-I in resting platelets (B), and PAC1 (C) and CD63 (D) upon stimulation with CRP-XL (5 µg/mL), 2-MeSADP (5 µM), or TRAP-6 (20 µM) per donor. Details on color coding are given in panel A. PAC1 (E) and CD63 (F) expression levels were determined for GPVI-rich (top 25% expressing) platelets and for GPVI-poor (rest of platelets; 75%) platelets for each agonist. Mean ± standard error of mean; n = 2 per group.