Molecular Characteristics of Platelet Formation Studied by Genetic Profiling.

Abstract 4584

Megakaryocyte (Mk) development and proplatelet formation are regulated at multiple levels although terminal stage molecular signals are incompletely delineated. We applied a customized, platelet-restricted oligonucleotide gene chip comprised of 432 genes to characterize genetic changes that occur in CD34⁺ cells differentiated along the Mk lineage, using a cytokine cocktail containing IL-6 (25 ng/mL), IL-11 (25 ng/mL), FLT3 (50 ng/mL), IL-1β (10 ng/mL), thrombopoietin (TPO, 50 ng/mL), and stem cell factor (SCF, 25 ng/mL). After day (D) 5, >60% of cells remained CD41 (αIIB)-positive by flow cytometric analysis, with 90% CD41-positivity at D20. Positive-selection using a β3 (glycoprotein IIIa) anti-CD61 antibody and magnetic beads was completed daily, and allowed for physical separation and comparative genetic profiles of the CD61⁺ and CD61^- cell fractions. Gene-wise progressive correlation analysis of aggregate CD61⁺ to CD61^- cell fractions demonstrated a correlation distribution centered between -0.2 to 0, confirming the distinct genetic profiles of CD61⁺ and CD61^- cells. Unsupervised hierarchical clustering of the CD61⁺ cells – in conjunction with 41 normal platelet control profiles – confirmed the presence of two discriminatory dendograms that separated D1-15 profiles from D16-21 and normal platelet profiles, establishing that a genetic switch towards late Mk/proplatelet development occurred around D15/D16. Aggregate gene expression between D1-15 and D16-21 subsets identified 252 differentially-expressed genes (p < 0.05), only 12 of which were upregulated in D16-21 samples (11 of these 12 genes were differentially expressed between day-matched CD61^- cell fractions, confirming the specificity of these changes). The conjunction of this 12-gene subset with a common 5-member pathway cluster of progressively induced genes identified three transcripts [transforming growth factor β-2 (TGFB2), CD99 antigen (CD99), and thromboxane A2 receptor (TBXA2R)] that were specifically probed to elucidate transcript and protein expression patterns in circulating platelets. Dual-color quantitative flow cytometry for these three proteins (along with ITGB3, β3, as control for a non-induced gene) was completed in both permeabilized (intracytoplasmic) and non-permeabilized (cell-surface) platelets, in parallel with thiazole orange (TO) as a measure of platelet RNA content; for all analyses, RNA and protein content in individual platelets were normalized by size and volume to adjust for size/volume differences. Despite a wide distribution of platelet size, RNA content, and protein expression, both the RNA and protein concentrations remained relatively constant. Furthermore, the RNA concentration remained tightly restricted within a narrow window, establishing that this parameter is critically maintained for all platelets unrelated to age. For all platelets, there was poor correlation (r <0.15) between RNA and protein concentrations, although correlation coefficients increased considerably in the subset of platelets with the highest (top 3%) RNA content. This was evident for both intracytoplasmic (TGFB2, r=0.41; CD99, r=0.55; TBXA2R, r=0.64) and cell-surface (TGFB2; r=0.34; CD99, r=0.73; TBXA2R, r=0.58) determinations, with smaller correlation coefficients for ITGB3 (0.21 and 0.13, respectively). In summary, these data demonstrate that (1) a genetic switch leading to mature platelet profiles (and presumably proplatelet formation) occurs at D15/D16 of MK maturation; and (2) in late-stage highly expressed transcripts, RNA concentration predicts protein concentration (both intracytoplasmic and cell-surface) in the subset of platelets with highest RNA content. These data provide insights into molecular differences between newly-formed and older platelets, and may have implications for elucidating previously-described functional differences among platelet subsets.