Analysis of Platelet Alpha and Dense Granule Release Using Small Volumes of Blood

A significant limitation of the gold-standard platelet function test, light transmission aggregometry (LTA), is the relatively large volume of blood required, as well as need for a platelet count in the resulting platelet-rich plasma of not much less than 100,000/μL, thus potentially precluding this test in pediatric and thrombocytopenic patients. Flow cytometric assessment of platelet activation in diluted whole blood avoids these limitations, typically employing anti-CD62P (P-Selectin) binding reflecting alpha granule release and PAC-1 binding reflecting conversion of GPIIb/IIIa to its active conformation. We have recently reported the use of tandem mass spectrometry to assess platelet dense granule release of deuterated serotonin in comparably diluted whole blood samples (https://doi.org/10.1093/ajcp/aqz094). In the present study we examine the possibility of adapting and optimizing the luciferin-luciferase (L-L) approach often used together with LTA for the assay of platelet dense granule ATP release from 10-fold diluted whole blood in parallel with flow cytometric analysis.

We optimized the L-L reaction with commercially available reagents and used a conventional lumi-aggregometer to measure real-time secretion of platelet dense granules following agonist stimulation in 10-fold diluted whole blood obtained from healthy adult volunteers, with or without specimen stirring at 1000 RPM. Direct comparison of optimized reagents to a commercial L-L reagent kit showed an improved lower limit of detection for exogenously added ATP by at least one order of magnitude. Initial studies were performed under non-stirring conditions. Dose-dependency of agonist was observed, with representative results for released ATP of 243.1 ± 29.7 pmol/10⁷ platelets in response to 40 μM TRAP (thrombin receptor-activating peptide) and 18.9 ± 2.4 pmol/10⁷ platelets in response to 1 nM convulxin (GPVI receptor activator) (mean ± SEM, N = 15). In contrast, virtually no ATP secretion was observed in response to 10 μM ADP or to 10 μg/mL collagen. Despite the quite low platelet count (as low as 11,000/μL to date) following the 10-fold dilution of whole blood, introducing 1000 RPM stirring to these samples now permitted robust ATP release in response to collagen as low as 2 μg/mL. There continued to be little or no ATP release, however, in response to 10 μM ADP even with the addition of stirring.

Flow cytometry was used to interrogate agonist-stimulated alpha granule release and GPIIb/IIIa conformation change. Following incubations with agonist and fluorescently-labeled antibodies, the 10-fold diluted whole blood was further diluted an additional 10-fold, and gating based on light scatter and binding of anti-CD61 antibody used to identify individual platelets. Under non-stirring conditions, TRAP, ADP, and convulxin all produced strong P-selectin exposure, as well as conversion of GPIIb/IIIa to its active conformation. Analysis of samples that had undergone 1000 RPM stirring during incubation with agonist showed quite similar results in response to TRAP. While still strongly positive, the responses to ADP in the stirred specimens showed decreases typically of at least 50%. Without stirring, platelet responses to 2-10 μg/mL collagen were virtually absent, whereas in the stirred specimens significantly positive anti-CD62P and PAC-1 binding were now observed.

The combined flow cytometric and L-L studies required <1 mL blood. The results demonstrate the dramatic differences observed in the release of platelet alpha and dense granules, depending upon conditions and assay systems employed. We have now shown that with agonist addition to the same 10-fold diluted whole blood samples employed in flow cytometric platelet analysis, dense granule release may also be assessed by modifications of the L-L methodology already employed for lumi-aggregation studies in many laboratories. We have additionally shown that by employing stirring of the specimen during incubation with agonist, collagen itself may be employed as platelet stimulus, rather than having to rely upon convulxin. The difference in requirements for alpha versus dense granule release, however, remains exemplified by ADP: While a potent stimulus of alpha granule release, ADP employed in the 10-fold diluted whole blood setting appeared incapable of producing significant dense granule release either in the absence or presence of sample stirring.