Development of SM in mice transplanted with TRKB- and BDNF-modified hematopoietic stem/progenitor cells. (A-F) Representative histopathology (hematoxylin and eosin) and cytology (May-Grünwald-Giemsa) from 2 animals (A-E: mouse 1007, F: mouse 746). The World Health Organization criteria for human SM were fulfilled in diseased animals.^6,7  Multifocal, dense infiltrates of mast cells (≥15 mast cells in aggregates) were observed in different organs. (A-B) Sections of liver showing infiltration of mast cells in liver (×100, ×400). (C-D) Accumulation of mast cells in the red pulp of spleen (×100, ×400). (E) Cytospin of spleen showing mature, round (typical) mast cells with abundant cytoplasm filled with granules (×1000). Note phagocytosis of red cells in some mast cells (B,E). Spindle-shaped slightly hypogranular mast cells were also observed in other animals (data not shown). (F) Low-magnification views of bone marrow section showing extensive infiltration of mast cells in sternum and osteosclerosis (×100). However, cytospin of bone marrow cells from tibia showed <20% of mast cells. Moreover, kidney infiltration of mast cells was observed in some animals with SM (data not shown). It is important to note that spleen and liver were affected in all animals with SM, whereas some animals did not have infiltration of mast cells in bone marrow. Unfortunately, skin and gut were not collected for histological analyses, because we do not routinely analyze these for leukemogenesis studies. (G-K) Representative flow cytometric diagrams from animal 714. Flow cytometric analysis showing expression of transgenes (BDNF and TRKB, measured by enhanced green fluorescent protein (EGFP) and antibody against hemagglutinin-tag, respectively) in spleen (G, negative control shown as inset), expression of c-Kit and FcεRI in transgene positive (H) and transgene negative population (neither TRKB nor BDNF) (I), and expression of FcεRI and CD25 (marker for neoplastic mast cells) in transgene positive (J) and transgene negative populations (K, negative control for panels H-K shown as inset). These demonstrated that abnormal mast cells were derived from gene-modified cells. It is interesting to note that in SM mice, both autocrine and paracrine loops may have contributed to mastocytosis (Figure 1G), whereas we observed a clear selection for both lymphoid leukemia (supplemental Figure 4) and myeloid leukemia transformation by autocrine activation of TRKB.⁴  Activation of TRKB in hematopoietic stem/progenitor cells induced 2 different phenotypes in vivo, ie, SM and lymphoblastic leukemia. This might be due to different cell populations targeted,¹⁰  although other factors might also be involved.