Distinct functional defects in platelet subpopulations after hematopoietic stem cell transplantation

Introduction Total body irradiation (TBI) prior to hematopoietic stem cell transplantation (HSCT) is often the last therapeutic option for hemato-oncological diseases. Complications include bleeding due to prolonged thrombocytopenia but also thrombosis, such as veno-occlusive disease. The role of platelets in HSCT-related complications is yet ill defined.

Method Studies were conducted using C57B6/J wildtype mice subjected to lethal TBI followed by HSCT with bone marrow from ubiquitously dsRed-expressing reporter mice. Platelet surface receptor expression and function was assessed by flow cytometry (FC) and aggregometry. In vitro thrombus formation was analyzed using a microfluidic collagen-coated flow chamber ([Fig. 1]).

Fig. 1 In vitro thrombus formation on collagen after HSCT; Whole blood from transplanted mice was perfused over a collagen-coated surface on d14, d21 and d28 after HSCT. While the total surface coverage was slightly increased, the thrombus volume was comparable to untreated controls. Donor-derived (dsRed+, magenta) platelets perferentially adhere to the thrombus edge. Platelets were counterstained with an anti-GPIX Dylight488 antibody (cyan).

Results Platelet receptor expression of GPIb/IX/V, CD9, GPIIb/IIIa, CLEC-2, GPVI, and integrin α2β1 was reduced between day d1 and d14 after HSCT, but returned to control values by d28. Platelet activation in response to high dose thrombin (Thr), collagen-related peptide (CRP-XL) or convulxin (Cvx) was severely impaired even on d28. Receptor expression correlated with platelet size of donor- (dsRed+) or recipient-derived (dsRed-) platelets. On d14, dsRed+ platelets were larger, while their receptor expression was comparable to control platelets, indicative of an overall decreased receptor density. dsRed- platelets were smaller with an even higher reduction of receptor expression. By d28, size and receptor levels of dsRed+ platelets normalized back to control values, while dsRed- platelets were still significantly smaller with markedly reduced expression of numerous receptors. Assessment of platelet function upon stimulation with Thr, CRP-XL or Cvx revealed hyporeactivity of dsRed- platelets, while dsRed+ platelets were unexpectedly hyperreactive compared to controls. In line with our FC data, platelet aggregation was blunted after Thr activation and virtually absent when activated with collagen or CRP-XL, even on d28. When we analyzed in vitro thrombus formation in a whole blood collagen flow chamber total surface coverage was mildly increased, although the thrombus volume remained comparable to control levels. This finding is in contrast to our FC and aggregometry results and suggests a yet unidentified involvement of plasmatic coagulation factors. Intriguingly, we found that dsRed+ platelets remained preferentially attached to the edge of the growing thrombi, implying that functionally distinct platelet subpopulations after HSCT could affect thrombus initiation and growth.

Conclusion Our platelet subpopulation analysis demonstrated that activity of recipient-derived platelets was massively and sustainably impaired after HSCT, while donor-derived platelets are hyperreactive, which is masked when the entire population is analyzed. Virtually unaltered in vitro thrombus formation implies an unrecognized involvement of the coagulation cascade after HSCT. The identification of a hyperreactive donor-derived platelet subpopulation provides a possible explanation for post-HSCT thrombosis.