Essential role for Stat5a/b in myeloproliferative neoplasms induced by BCR-ABL1 and JAK2(V617F) in mice

Abstract

STAT5 proteins are constitutively activated in malignant cells from many patients with leukemia, including the myeloproliferative neoplasms (MPNs) chronic myeloid leukemia (CML) and polycythemia vera (PV), but whether STAT5 is essential for the pathogenesis of these diseases is not known. In the present study, we used mice with a conditional null mutation in the Stat5a/b gene locus to determine the requirement for STAT5 in MPNs induced by BCR-ABL1 and JAK2(V617F) in retroviral transplantation models of CML and PV. Loss of one Stat5a/b allele resulted in a decrease in BCR-ABL1-induced CML-like MPN and the appearance of B-cell acute lymphoblastic leukemia, whereas complete deletion of Stat5a/b prevented the development of leukemia in primary recipients. However, BCR-ABL1 was expressed and active in Stat5-null leukemic stem cells, and Stat5 deletion did not prevent progression to lymphoid blast crisis or abolish established B-cell acute lymphoblastic leukemia. JAK2(V617F) failed to induce polycythemia in recipients after deletion of Stat5a/b, although the loss of STAT5 did not prevent the development of myelofibrosis. These results demonstrate that STAT5a/b is essential for the induction of CML-like leukemia by BCR-ABL1 and of polycythemia by JAK2(V617F), and validate STAT5a/b and the genes they regulate as targets for therapy in these MPNs.

Figure 1

Reduction in Stat5 gene dosage attenuates BCR-ABL1–induced CML-like MPN. Kaplan-Meier survival curve for recipients of Stat5a/b^fl/+ (n = 2) or Stat5a/b^fl/− (n = 4) BM transduced with p210MIGFP retrovirus, and recipients of Stat5a/b^fl/+ (n = 8) or Stat5a/b^fl/− (n = 11) BM transduced with p210MIGFPCre retrovirus is shown. The symbols indicate individual recipient mice, with the disease phenotype of each designated by the shading: black, CML-like MPN; white, B-ALL; black and white, mixed CML/B-ALL; gray, histiocytic sarcoma (HS). Relative to the p210MIGFP→Stat5a/b^fl/+ cohort, the survival of the p210MIGFP→Stat5a/b^fl/− and p210MIGFPCre→Stat5a/b^fl/+cohorts was significantly prolonged (P = .017 and P = .0009, respectively, by Mantel-Cox test). Relative to the p210MIGFPCre→Stat5a/b^fl/+ cohort, the survival of the p210MIGFPCre→Stat5a/b^fl/− cohort was also significantly prolonged (P = .008).

Figure 2

Mx-Cre–mediated deletion of Stat5a/b abolishes CML-like MPN induced by BCR-ABL1. (A) Kaplan-Meier survival curve for recipients of p210MIGFP-transduced BM from Mx-Cre;Stat5a/b^fl/+ donors (solid line), and for recipients of p210MIGFP-transduced BM from Mx-Cre;Stat5a/b^fl/− donors either untreated (dotted line) or treated (dashed line) with pIpC after transplantation as described in the “Methods.” The symbols indicate individual recipient mice, with the disease phenotype of each designated by the shading. No recipients in the Mx-Cre;Stat5a/b^fl/− + pIpC cohort developed hematologic disease. (B) PB leukocyte counts at day 50 after transplantation in untreated or pIpC-treated recipients of p210MIGFP-transduced BM from Mx-Cre;Stat5a/b^fl/+ (left) or Mx-Cre;Stat5a/b^fl/− (right) donors. (C) Flow cytometric plot of GFP expression in PB myeloid cells from 2 representative recipients (mice #39 and #38) of p210MIGFP-transduced BM from Stat5a/b^fl/− donors. (D) Southern blot analysis of the extent of recombination of the floxed Stat5a/b allele in genomic DNA from PB leukocytes at day 50 after transplantation. Nomenclature is as in supplemental Figure 1C. The small amount of wild-type Stat5a/b allele in recipients of Stat5a/b^fl/− BM (mice #35-40) represents contribution from radioresistant host lymphocytes. (E) Spleen weights at autopsy of untreated or pIpC-treated recipients of p210MIGFP-transduced BM from Mx-Cre;Stat5a/b^fl/+ (left) or Mx-Cre;Stat5a/b^fl/− (right) donors. (F) Western blot analysis of primary myeloerythroid cell extracts from a representative untreated recipient of p210MIGFP-transduced BM from Mx-Cre;Stat5a/b^fl/− donors that developed MPN and from 4 healthy pIpC-treated recipients. As controls, extracts from parental and BCR-ABL1–expressing Ba/F3 cell lines and BM from a nontransplanted Stat5 wild-type mouse were included. Proteins were immunoblotted with the indicated Abs against total or phosphorylated BCR-ABL1, STAT5, and CrkL. An anti-eIF4e immunoblot demonstrating equivalent protein loading is shown at the bottom.

Figure 3

Retrovirally transduced normal HSCs sustain myelopoiesis after Stat5 deletion. Genomic DNA from BM of recipients of MIGFP-transduced Stat5a/b^fl/− and Mx-Cre;Stat5a/b^fl/− donor BM was isolated 2 months after transplantation and analyzed by Southern blot for efficiency of Stat5a/b deletion and number of engrafting proviral clones, as in supplemental Figure 1C. Top panel: Analysis of efficiency of recombination of the floxed Stat5a/b allele in recipients of MIGFP-transduced Stat5a/b^fl/− BM (lanes 3-7) and in untreated (lanes 8-9) and pIpC-treated (lanes 10-12) recipients of MIGFP-transduced Mx-Cre;Stat5a/b^fl/− BM using a Stat5 probe. Bottom panel: Analysis of the number of engrafted retrovirally transduced HSC clones in the 3 cohorts using a GFP probe.

Figure 4

STAT5 is not required for B-lymphoid transformation by BCR-ABL1 in vitro or for maintanance of established B-lymphoid leukemia in viro. (A-B) STAT5 is not required for transformation of primary BM B-lymphoid cells by BCR-ABL1 in vitro. (A) Primary BM from non–5-FU-treated Stat5a/b^fl/+ and Stat5a/b^fl/− donor mice was transduced with p210MIGFP or p210MIGFPCre retrovirus and plated directly in agarose. Transformed pre-B lymphoid colonies were counted on day 10. There was no significant difference in the number of colonies arising from Stat5a/b^fl/− donor BM transduced with either p210MIGFP or p210MIGFPCre retrovirus (P = .88) or between any of the values assessed pairwise (t tests). (B) Serial dilutions of transduced BM were plated in triplicate on syngeneic stromal layers derived from nontransduced wild-type BM and cultured for 3 weeks, as described in the “Methods.” The plating density is indicated by the line color, the number of cultures that reached confluence (defined as 10⁶ nonadherent cells) is indicated on the ordinate, and the time to confluence on the abscissa. (C) Growth of BCR-ABL1–transformed B-lymphoblasts derived from Mx-Cre;Stat5a/b^fl/+ (top panel) or Mx-Cre;Stat5a/b^fl/− (bottom panel) donors either untreated (blue squares) or treated with IFN-β (red circles). Each curve represents data from an independent population of transformed cells. (D) Southern blot analysis of Stat5a/b recombination status from the experiment in panel C. Lanes 1 and 2 contain tail DNA from Stat5 wild-type and Stat5^fl/− mice, respectively. Note the complete deletion of the floxed Stat5a/b allele in IFN-treated lymphoblasts from Mx-Cre;Stat5a/b^fl/+ donors (lanes 7-10) and from Mx-Cre;Stat5a/b^fl/− donors (lanes 15-18). (E) Kaplan-Meier survival curve for unirradiated Balb/c Rag2^−/− recipients injected intravenously (1 × 10⁷ cells each, n = 16) with BCR-ABL1–expressing lymphoblasts derived from the in vitro transformation experiments in panels B and C. After engraftment of leukemia, half of each cohort (dashed lines) were treated with pIpC (arrowheads) to induce Cre recombinase. (F) Southern blot analysis of Stat5a/b recombination status in tumor tissue from 3 representative recipients of lymphoblasts from Mx-Cre;Stat5a/b^fl/− donors from panel E that were treated with pIpC (lanes 4-6) or untreated (lanes 7-9). Lanes 1, 2, and 3 contain tail DNA from Stat5wt, Stat5^fl/+, and Stat5^fl/− mice, respectively. Note the complete deletion of the floxed Stat5a/b allele in lymphoblasts from pIpC-treated recipients.

Figure 5

Reduction of Stat5 gene dosage impairs JAK2^V617F-induced polycythemia in vivo. Hematocrit (A), reticulocyte counts (B), and leukocyte counts (C) from PB of recipients of BM from Stat5a/b^fl/+ and Stat5a/b^fl/− donors transduced with JAK2^V617FMIGFP or MIGFPCre retrovirus analyzed on day 100 after transplantation. Hematocrit was significantly lower in the Stat5a/b^fl/− JAK2^V617FMIGFPCre group (P = .0114 vs Stat5a/b^fl/− JAK2^V617FMIGFP by t test).

Figure 6

Efficient deletion of Stat5 by Mx-Cre abolishes polycythemia and reticulocytosis induced by JAK2^V617F. Hematocrit (A) and reticulocyte counts (B) of untreated (−) or pIpC-treated (+) recipients of JAK2^V617FMIGFP-transduced BM from Mx-Cre;Stat5a/b^fl/+ (left) or Mx-Cre;Stat5a/b^fl/− (right) donors on day 60 after transplantation. Both values were significantly lower in the Stat5a/b^fl/− (+pIpC) group (P < .0001 vs Stat5a/b^fl/− [−pIpC] by t test). (C) Flow cytometric plots of GFP expression in PB leukocytes from 3 representative pIpC-treated recipients of JAK2^V617FMIGFP-transduced BM from Stat5a/b^fl/− donors, with GFP⁺ populations ranging from 15%-90%. (D) Spleen weights from the mice in panel A at time of autopsy on day 100. (E) Western blot analysis of primary BM myeloerythroid cell extracts from representative untreated (−) or pIpC-treated (+) recipients of JAK2^V617FMIGFP-transduced BM from Mx-Cre;Stat5a/b^fl/+ or Mx-Cre;Stat5a/b^fl/− donors from panel A. As controls, protein extracts from parental and JAK2^V617F-expressing Ba/F3 cell lines and BM from a normal mouse were used. Cell extracts were immunoblotted with the indicated Abs against total or phosphorylated STAT5 and ERK1/2 and against total JAK2 and BCL-X. An anti-eIF4e immunoblot demonstrating equivalent protein loading is shown at the bottom. Note that some STAT5 protein is detectable in pIpC-treated recipients because of contamination with cells of host origin.

Figure 7

Subclinical myeloproliferative disease and myelofibrosis induced by JAK2^V617F in the absence of STAT5. H&E stain of spleen (A) and liver (B) from a representative pIpC-treated recipient of JAK2^V617F MIGFP-transduced Mx-Cre;Stat5a/b^fl/− BM that engrafted with provirus-positive donor-derived HSCs. In the spleen, a lymphoid follicle (f) and areas of erythroid (e) and myeloid (m) cell infiltration are designated. Magnification is 200×. (C-D) Reticulin stain of BM from a non-pIpC–treated recipient of JAK2^V617F MIGFP-transduced Mx-Cre;Stat5a/b^fl/− BM that developed polycythemia (C) and a pIpC-treated recipient with normal hematocrit (D). Magnification is 600×. All images were obtained from a BH-2 microscope using a Q-Color5 digital camera and QCapture Pro acquisition software (Olympus).