Ojective 1: Demonstrate that Eltrombopag can bypass the TPO signaling defect induced by IFNγ IFNγ INHIBITS TPO- BUT NOT ELTROMBOPAG-INDUCED HSPC PROLIFERATION G-CSF mobilized human CD34+ cells from 3 independent healthy donors were cultured for 7 days in the presence of SCF and Flt3-L, supplemented with TPO or Eltrombopag, with or without IFNγ. CD34+ cell maintenance and proliferation were measured by flow cytometry. For all donors, the total CD34+ cell counts decreased 7.5-fold when exposed to both TPO and IFN-γ compared to cultures containing TPO only (p=0.022). In contrast, CD34+ cells cultured with Eltrombopag and IFNγ showed only a minimal reduction (1.5-fold) in CD34+ cell numbers compared to cultures containing Eltrombopag but no IFNγ. These data suggest that IFNγ inhibits TPO- but not eltrombopag-induced HSPC proliferation. To assess the impact of IFN-γ on long-term repopulating hematopoietic stem cells, CD34+ cells cultured for 7 days in the presence of TPO or Eltrombopag with or without IFN-γ were transplanted in NSG mice. Data from these transplantation experiments are pending. IFNγ IMPAIRS TPO- BUT NOT ELTROMBOPAG-INDUCED SIGNALING We next hypothesized that IFNγ negatively impacts HSPC proliferation by perturbing the TPO signaling pathway and that Eltrombopag bypasses the TPO signaling defect. Upon binding of TPO to its receptor, c-mpl, several major signaling pathways are induced including MAPK/ERK, proteinase kinase C, AKT and JAK/STAT (e.g. STAT5). Human G-CSF mobilized CD34+ cells from 4 independent healthy donors were first cultured for 24 hours with or without IFN-γ 2000IU/mL and STAT5 signaling was subsequently evaluated by flow cytometry following a brief (15 minutes) exposure to TPO 5ng/mL. On average, a 4.4-fold decrease in STAT5 signaling was observed in CD34+ cells exposed to IFN-γ, indicating impaired TPO signaling in the presence of IFN-γ (p<0.001). In contrast, eltrombopag resulted in a lower phosphorylation of STAT5 in CD34+ cells compared to TPO, consistent with the lower proliferation observed in ex vivo cultures with eltrombopag compared to TPO, but the intensity of STAT5 phosphorylation was preserved in the presence of IFNγ. HOW DOES ELTROMBOPAG BYPASS THE TPO SIGNALING DEFECT INDUCED BY IFNγ? Bypass intracellular negative regulators of TPO signaling. Recent studies indicate that IFN-γ upregulates suppressor of cytokine signaling (SOCS) molecules, resulting in inhibition of TPO-induced STAT signaling. We have performed transcriptome analysis of highly purified human HSC (CD34+CD38-CD45RA-CD90+CD49f+Rholo) exposed to TPO or Eltrombopag, using a single cell RNA-Seq approach. Preliminary data indicate a 10-fold downregulation of SOCS1 and SOCS3 expression in HSCs exposed to Eltrombopag compared to cells exposed to TPO, suggesting a SOCS-independent regulation of Eltrombopag signaling. Additional single cell RNA Seq and Q-PCR experiments are underway to confirm these data using highly purified HSC exposed to TPO or Eltrombopag in the presence or absence of IFNγ. Additional negative regulators of TPO signaling, including SHP-1 phosphatase, p27 cell cycle inhibitors, and down-modulation of c-Mpl expression. Bypass c-Mpl binding defect of TPO. We next carried out chemical crosslinking and microscale thermophoresis assays to investigate whether the inhibitory effect of IFNγ on TPO signaling may result from TPO inactivation due to direct physical interaction with IFNγ. We found that TPO specifically formed heterodimers with IFNγ with a 1:1 stoichiometry. Furthermore, we determined an apparent TPO-INFγ heterodimer dissociation constant of 35nM (Figure B). In contrast, no heterodimer formation was observed between IFNγ and the early-acting cytokine stem cell factor (SCF). In FY16, further In conclusion, our data confirm the negative impact of INF-γ on TPO signaling and propose cytokine heterodimerization as a novel mechanism for the clinically observed HSPC exhaustion and bone marrow failure in disorders of chronic inflammation. Ojective 2: Demonstrate that Eltrombopag has DNA repair activity in human HSPCs G-CSF mobilized human CD34+ cells from 5 independent healthy donors were cultured in the presence of SCF and Flt3-L (SF), SF and TPO (SFT), or SF and Eltrombopag (SFE) for 24 hours before exposure to 2Gy γ-irradiation, and then cultured for an additional 5 to 24 hours. DNA damage was quantified by flow cytometric determination of γH2AX expression, a marker of irradiation-induced DNA double-strand breaks (DSB), and CD34+ cell survival was measured by flow cytometry using Annexin V and a viability dye. There were significantly fewer γH2AX+ cells 5 hours post-irradiation when the culture included TPO or Eltrombopag than with SF alone (n=5). Five hours post-irradiation, cultures containing TPO or Eltrombopag had significantly increased percentages of live cells (n=5), as well as decreased percentages of cells undergoing apoptosis compared to cultures with SF alone (SFT 12.6 0.5% p=0.003; SFE 12.4 2.1% p=0.012; SF 21.5 3.7%, n=5). RT-qPCR arrays performed at 5 hours after irradiation on CD34+ cells cultured as above with SFT or SFE showed a significant decrease (p<0.05) of at least two-fold in several pro-apoptotic or cell cycle arrest genes (BBC3, CCNO, GADD45G, PPM1D) compared to CD34+ cells cultured with SF alone. Twenty-four hours post-irradiation, cells cultured with TPO or Eltrombopag had significantly increased percentages of live cells (n=3), and decreased percentages of dead cells compared to cells cultured with SF alone (SFT 9.75 1.0% p=0.013; SFE 16.3 0.6% p=0.032; SF 36.5 6.2%, n=3). Progenitor cell survival was assessed using the CFU assay. The number of colony-forming cells was 5.9 ( 0.4) and 3.6 ( 0.2) fold higher when cultured with TPO or Eltrombopag, respectively, before γ-irradiation than when cultured with SF alone (p=0.005 and 0.006, respectively, n=2). Survival of long-term repopulating HSCs was assessed by quantifying human CD45+ cell engraftment at least 2 months after intravenous injection of NSG mice with irradiated human CD34+CD38- cells pre-cultured for 24 hours with SF, SFT or SFE. Engraftment of cells cultured with TPO or Eltrombopag was significantly higher than engraftment obtained after injection of cells cultured with SF alone before γ-irradiation. We conclude that, analogous to TPO, Eltrombopag favors DNA DSB repair and, consequently, survival of both hematopoietic stem and progenitor cells after γ-irradiation. These pre-clinical data suggest that Eltrombopag may be of benefit in the treatment of patients with Fanconi Anemia (FA). This work will be presented at the American Society of Hematology (ASH) in December 2015. In FY16, additional studies will be conducted to pinpoint the DNA repair pathway (e.g. NHEJ) employed by Eltrombopag. We will also confirm the utility of Eltrombopag in FA by directly testing DNA repair activity in FA HSPCs. The latter, inaccessible in FA patients due to bone marrow failure, will be obtained via hematopoietic differentiation of iPSCs derived from FA patients skin fibroblasts. The laboratory of Dr. Juan Carlos Izpisua-Belmonte (Salk Institute, CA) has provided diseased and genetically corrected iPSC lines from patients with FA. A clinical trial testing Eltrombopag in patients with FA will also be initiated in FY16.

Project Start
Project End
Budget Start
Budget End
Support Year
Fiscal Year
Total Cost
Indirect Cost
U.S. National Heart Lung and Blood Inst
Zip Code
Townsley, Danielle M; Scheinberg, Phillip; Winkler, Thomas et al. (2017) Eltrombopag Added to Standard Immunosuppression for Aplastic Anemia. N Engl J Med 376:1540-1550
Ito, Sawa; Barrett, A John; Dutra, Amalia et al. (2015) Long term maintenance of myeloid leukemic stem cells cultured with unrelated human mesenchymal stromal cells. Stem Cell Res 14:95-104
Wu, Chuanfeng; Hong, So Gun; Winkler, Thomas et al. (2014) Development of an inducible caspase-9 safety switch for pluripotent stem cell-based therapies. Mol Ther Methods Clin Dev 1:14053
Sellers, Stephanie E; Dumitriu, Bogdan; Morgan, Mary J et al. (2014) No impact of lentiviral transduction on hematopoietic stem/progenitor cell telomere length or gene expression in the rhesus macaque model. Mol Ther 22:52-8