In Vivo Platelet Production Occurs from an Expanding Pool of Non-Apoptotic Megakaryocytes in a Mouse Model of Passive Immune Thrombocytopenia

Abstract 1214

Immune thrombocytopenia (ITP) is the most common acquired thrombocytopenia in children. Typically, external triggers as infections or vaccinations cause the rise of antibodies that crossreact with antigens expressed on the platelet surface. These anti-platelet antibodies are mostly directed against glycoprotein complexes GPIIb/IIIa or GPIb/IX/V, resulting in an increased turnover of antibody-decorated platelets which are then sequestered by the reticuloendothelial system. Recently, it has been suggested that thrombocytopenia might also be due to an insufficient platelet production as serum of some patients with ITP can impair the maturation of CD34+ hematopoietic stem cells to bone marrow megakaryocytes (MKs) in vitro or abrogate the formation of proplatelets in an in vitro culture system. The accelerated platelet turnover demands the generation of platelets de novo. Bone marrow smears often reveal normal or slightly increased MKs, although they seem to be smaller and of altered morphology. However, very little is known about the consequences of anti-platelet antibodies on bone marrow MKs in vivo and in situ. Here, we took advantage of a simple animal model of passive ITP by single or multiple intraperitoneal injections of an anti-GPIb antibody into mice. MKs were evaluated by multi-color immunofluorescence histology on whole femur sections in a modified staining procedure that bypasses decalcification. MK numbers on day 3 were doubled in response to a single injection and tripled on day 8 when mice were injected additionally on day 3 and 7. In these mice platelet counts were up to 2000/nL on day 10, indicating the power to produce platelets. MK area per section was transiently upregulated on day 3 in single injected mice and quadrupled after multiple injections on day 8 before shrinking below norm on day 14. Staining with an anti-rat IgG antibody showed that the antibody was present on MKs within the bone marrow several hours to days after injection. The signal was present for 5 days and no antibody was detected on day 7. MKs had an overall normal morphology and showed no signs of apoptosis or DNA blebbing. All MKs analyzed were negative for TdT in a classical TUNEL assay, indicating that there were no single strand breaks. As platelet counts rose markedly while the antibody was still present on the MK surface, we sought to identify whether the pool of MKs is expanded or formed de novo. To address this, mice where fed with nucleotide analogue EdU for up to 12 days and femur sections stained with Click-It-647 reagent to stain for newly incorporated DNA while mice were treated with anti-platelet antibody or isotype control. We found EdU-positive MKs after 12 days in control isotype-injected mice indicating the de novo formation from hematopoietic stem cells. In antibody-injected mice, newly formed MKs were negative or stained weakly for EdU on day 12, suggesting that they arise partially from an existing pool of progenitors. Finally, we analyzed platelet formation in vivo by imaging of the cranial bone marrow of GPIIb-eYFP-heterozygous mice. The depletion antibody was labeled with Atto-590-fluorophore and injected hours before imaging. Vasculature was counterstained by Quantum dots. We found that MKs residing at the bone marrow were decorated with the antibody and released pre- and proplatelets into the vasculature, indicating that platelet biogenesis can occur in the presence of anti-platelet antibodies on MKs. Our data thus provide novel insight into the pathomechanism of platelet production in patients with ITP.