The human IL-15Tg mice developed a fatal lymphoproliferative disorder. (A) Survival plots of IL-15Tg mice on different backgrounds. IL-15Tg (B1) and IL-15Rα^−/−IL-15Tg mice were compared in terms of their life span with IL-15Tg mice showing a shortened survival (n = 12, P < .001). IL-15Rα–IL-15 double-Tg and IL-15Tg mice were compared as well (n = 15, P < .001). IL-15Rα^−/−-IL-15Tg mice lived as long as their littermates, whereas IL-15Rα–IL-15 double-Tg mice died at a very young age because of uncontrolled expansion of CD8 T cells. (B) Expansion of CD44^hiCD8 T cells in the IL-15Tg mice at the benign and terminal stages of leukemia. Splenocytes from a WT mouse (i), a young IL-15Tg (3-month-old) mouse (ii), a 13-month-old IL-15Tg mouse (iii), a 13-month-old IL-15Rα^−/−IL-15Tg mouse (iv), and a 3-month-old IL-15Rα–IL-15 double-Tg mouse (v) were stained with anti-CD3, anti-CD44, and anti-CD8 Abs and analyzed by flow cytometry. The majority (> 95%) of the splenocytes expressed CD3 and CD8 in the terminal-stage leukemic IL-15Tg mice (iii), whereas in the benign-phase IL-15Tg mice (ii) and IL-15Rα^−/−IL-15Tg mice (iv), CD8 T cells maintained an altered but stable homeostasis (with 80% and 44% of CD8⁺ cells in the CD3-expressing population, respectively). In the terminal-phase leukemic IL-15Tg mice (iii), T cells almost completely displaced other types of cells from the spleen. In the IL-15Rα–IL-15 double-Tg mice (v), even young mice (< 3 months old) displayed displacement of peripheral lymphocytes with CD8 T cells (> 92% of the T-cell compartment). Data represent results from 4-6 mice per group (see supplemental Figure 2 for a summary of all data). (C) Development of lymphoma/leukemia in WT mice transplanted with B1 (B1-001) IL-15Tg cells. Four weeks after the IV injection of 1 × 10⁶ B1 leukemic cell line cells, the WT recipient mice developed tumor masses in their spleens (left, indicated with an arrow). As a surrogate marker to monitor leukemic cell expansion, the concentrations of hIL-15 in the sera of transplanted mice were determined by a specific IL-15 ELISA (right, n = 8). (D) Constitutive phosphorylation and DNA binding of STAT5 in leukemic cells. In the left panel, constitutive tyrosine-phosphorylation of the STAT5a molecules in the ex vivo B1 leukemic cells (left lane) was demonstrated using an anti-pSTAT5a Ab by immunoblotting. An anti-STAT5a/b Ab was used to quantitate the amount of the STAT protein in each lane. In the right panel, STAT5 shows constitutive DNA binding in the ex vivo B1 leukemic cells as demonstrated using an EMSA assay. The target oligonucleotides were 5′-agttattagaaatTTCAAGGAAgtgacaacagag-3′ (STAT5 WT), 5′agttattagaaatTTCAACCTTgtgacaacaga-3′ (STAT5 mutant, mutation underlined), which were annealed, ³²P-labeled, and used as the probe. An anti-STAT5a/b Ab was used to supershift the DNA-protein complexes. The lanes indicate without (i) and with (ii) anti-STAT5a/b Ab. (E) Confirmation of the IL-15 dependency of the leukemic cells. Inclusion of the anti–hIL-15 Ab (MAB247; R&D Systems) abrogated the constitutive proliferation of IL-15Tg leukemic cells as assessed by [³H]-thymidine incorporation, suggesting that these cells depend on an autocrine supply of IL-15 for their growth and survival. The addition of anti-mouse IL-2/15Rβ Ab (TMβ1) led to a partial inhibition with B1/K2 clones, whereas the inhibition by TMβ1 was complete, with normal CD8 T cells stimulated with hIL-15. The data represent 3 independent experiments (n = 3 each).