The Morphology and Cellular Characterization of Thrombocytes in Adult Xenopus laevis.

In primates and rodents, platelets originate from the bone marrow megakaryocytes through a unique differentiation process with nuclear polyploidization, cytoplasmic maturation and proplatelet formation. In contrast, circulating thrombocytes of most non-mammalian vertebrates are particularly distinctive; the cells are large and nucleated. Adult Xenopus laevis may be an useful non-mammalian model for analyzing dynamic hematopoiesis because they are individually tolerable for time lapse analysis in vivo with sequential blood sampling, whereas classification of cell types has not been established yet. Microstructures of Xenopus thrombocytes observed with electron microscope exhibited structural characteristics largely resembling zebrafish thrombocytes with nucleated spindle cellular features (Thattaliyath et al., Blood 2005), and they had lobulated nuclear chromatin, granules, microparticles and open canalicular system-like-structures as in mammalian megakaryocytes. Since thrombocyte identification based on the morphological aspect was not sufficient, chemical staining with acetylecholinesterase and thiazole orange were performed. Additionally, mice were immunized by Xenopus peripheral blood cells to generate monoclonal antibodies, and two hybridomas producing IgG, respectively T12 and T5, were screened. T12⁺ (T12 positive) cells were morphologically typical thrombocytes. Flow cytometric analysis revealed that T12⁺ cells were also positive to anti-human GpIIb/IIIa polyclonal antibodies, and approximately 2-3% of whole peripheral blood cells were T12⁺/GpIIb/IIIa⁺ that distributed in FSC^low/SSC^low fraction. When T12 was injected into Xenopus to deplete T12⁺ cells in vivo, the detectable level of T12 in the circulation lasted for more than several weeks. Peripheral thrombocyte counts predominantly began to decrease immediately and reached their nadir at day 3, but white blood cell counts were not changed. RNA-rich blood cells considered as younger cells were then increasingly appeared, and finally the cell counts recovered to normal levels at day 10–15, indicating that in vivo depletion of T12⁺ cells induced thrombopoiesis and/or release of mature thrombocytes from the pool. T5 recognizing cells were classified into two populations by immunostaining and flow cytometry; T5⁺/GpIIb/IIIa⁺ cells were morphologically thrombocytic as the cells recognized by T12, while T5⁺/GpIIb/IIIa⁻ cells were spherical and similar appearance to lymphocytic cells. These observations raised some possibilities e.g.; antigen of T5 was a membrane protein common to both lymphocytes and thrombocytes, or T5⁺/GpIIb/IIIa⁻ cells were thrombocyte progenitors at earlier development stage than T12⁺/GpIIb/IIIa⁺ cells. Nevertheless only a few percent of T12⁺ and T5⁺ cells resided in peripheral blood, immunostaining revealed that the proportions of T12⁺/T5⁺ and T5⁺ cells in spleen were 10% and 70%, and T12⁺/T5⁺ and T5⁺ cells in liver were 5% and 20%, respectively. These suggest that spleen is predominantly involved in thrombopoiesis and/or thrombocyte storage in adult Xenopus. As T12 and T5 can be used successfully in flow cytometry and magnetic cell sorting, they should contribute us directly to elucidate the origin of circulating Xenopus thrombocytes and their cellular development process.