Jiwang Zhang - US grants

DESCRIPTION (provided by applicant): Our goal is to define the roles of Pten phosphorylation remodeling in hematopoietic stem cell (HSC) regulation. Pten, a tumor suppressor, has both lipid and protein phosphatase activities that inhibit PI3K/Akt and Fak-MAPK signaling respectively. It regulates many aspects of cell behavior, including proliferation, survival, adhesion and migration.1-4 The phosphatase activities of Pten are regulated by its c-terminal tail phosphorylation.5, 6 In addition, Pten many also have some phosphatase-independent functions. Many of Pten's biological functions are dependent upon protein-protein interactions which are mediated by its PDZ-motif.7 Knockout of Pten in mouse hematopoietic tissues results in abnormal activation of PI3K/Akt and Src signaling, which leads to uncontrolled HSC activation (G0 to G1 transition) and mobilization, followed by HSC decline. These mice develop myeloproliferative disorder (MPD) followed by acute myeloid/T lymphoid leukemia. Although Pten mutations are not commonly found in hematopoietic malignancies, including leukemia, p-Pten (the phosphorylated form of Pten) levels are increased in the abnormal blasts of most leukemic patients' bone marrow samples. Phosphorylation of Pten's c-terminal tail (ser380, thr382, and thr383) leads to a conformation change which may result in the blocking of its ability to bind to other partner proteins, the reduction of Pten phosphatase activity, and/or the alteration of the lifespan of the Pten protein. Our recent studies have suggested that the phosphorylation of Pten's c-terminal tail may not affect its lipid phosphatase activity but significantly compromises its protein phosphatase activity. The non-phosphorylated form of Pten (non-p-Pten) inhibits Src/Fak/p38 activity, thus repressing cell migration/invasiveness and inducing cell:cell contact inhibition of growth. p-Pten might have a dominant-negative function which induces cell-contact-related Src/Fak/p38 activation. We found that non-p-Pten is expressed in HSCs, while p- Pten levels are increased when HSCs enter the cell cycle; both of these events correspond to increased p- Src, p-Fak and p-p38 levels. Transduced over-expression of non-p-Pten preserves HSCs in a bone marrow niche-dependent manner, whereas transduced over-expression of p-Pten induces HSC/progenitors (HSC/Ps, from wild-type mice which have endogenous Pten expression) to differentiate to myeloid precursors. We propose that non-p-Pten maintains HSC quiescence and self-renewal ability through inducing cell:cell (HSCs and niche cells) contact-induced inhibition of growth by inhibiting Src/Fak/p38 signaling activities, whereas Pten's c-terminal phosphorylation alters its ability to bind to its partners and compromises its protein phosphatase activity. p-Pten promotes opposite functions to these through inducing cell:cell contact-related Src/Fak/p38 signaling. These studies will provide insights into how quiescent HSCs become activated and expand in number, and how we might be able to induce activated HSCs to revert back into quiescence in order to enhance their engraftment ability. This should greatly help our ability to expand HSCs in vitro and hence improve the outcome of clinical stem cell transplantation. It might be also help us to understand the nature of Pten c-terminal phosphorylation in leukemogenesis.

DESCRIPTION (provided by applicant): In addition to NF-?B-mediated survival/proliferation and caspase 8-mediated apoptosis, recent studies demonstrated that tumor necrosis factor-? (TNF hereafter) also induces a RIP1/RIP3-mediated necroptosis (a type of programmed necrosis) in most types of cells. We found that acute monocytic leukemia cells (AMoL hereafter), including subtypes M4 and 5 of acute myeloid leukemia, produce TNF. Compared to normal hematopoietic stem/progenitor cells (HSPCs), AMoL cells are resistant to TNF-induced apoptosis/necroptosis. TNF stimulates the clonogenic growth of AMoL cells at least partially by inducing RIP1/RIP3-mediated differentiation blockage. This pathway also mediates hematopoietic repression of TNF by inducing necroptosis/apoptosis. We propose that TNF represses AMoL cell differentiation by stimulating RIP1/RIP3 signal-dependent SOCS1 (suppressor of cytokine signaling-1) expression. We found that inactivation of TNF-RIP1/RIP3 signaling by a specific inhibitor or genetic deletion induces partial differentiation of AMoL cells which is correlated to the down-regulation of SOCS1 protein. The differentiation of TNF-RIP1/RIP3-inactivated AMoL cells can be further enhanced by interferon (IFN)-? or IFN-?, known inducers of AMoL cell differentiation. Our studies suggest that inhibition of TNF-RIP1/RIP3 might benefit AMoL patients by inducing leukemic cell differentiation while affording protection to normal HSPCs. Inhibiting TNF-RIP1/RIP3 signaling combined with IFN-? or IFN-? might to be a novel treatment approach for AMoL. We want to evaluate the efficacy of such treatment in vivo using murine leukemic models and in vitro using primary AMoL cells. We also intend to study the molecular mechanisms by which the TNF-RIP1/RIP3 signal regulates the level of SOCS1 in AMoL cells.

? DESCRIPTION (provided by applicant): Leukemia fusion genes, the fusion products of chromosomal translocations, are detected in >55% leukemia patients. Leukemic cells develop addiction to and reliance upon leukemia fusion proteins, the protein products of fusion genes, for their proliferation and survival. There is considerable speculation that targeted inactivation f leukemia fusion proteins is a most effective strategy for anti-leukemia therapy. We conjecture that inhibiting the production of fusion proteins by repressing the transcription and/or translatio of fusion genes represents a novel and perhaps more effective treatment for leukemia. We want to use leukemia cells that are RUNX-ETO+ (a fusion gene in which the ETO gene has been fused to the RUNX1 gene due to t(8;21)) as a model system to test such a hypothesis because the promoters for RUNX1 and the 5'UTRs for RUNX1 transcripts have been well documented. RUNX1 has two promoters, P1 and P2, which produce transcripts which have two distinct 5'UTRs. P1 transcripts have a <450bp 5'UTR the translation of which is regulated by a cap-dependent mechanism, while P2 transcripts have a ~1.6kb 5'UTR the translation of which is regulated by a cap-independent IRES-mediated mechanism. Interestingly, the expression of RUNX1 in hematopoietic stem/ progenitor cells (HSPC) is primarily controlled by P1, while the expression of RUNX1-ETO in leukemia cells is controlled by P2. Such use of alternative promoter and translational entities by RUNX1-fusion genes in leukemia cells provide us with an opportunity to treat RUNX1-fusion gene+ leukemia by specifically inhibiting the transcription/translation of the fusion genes with fewer effects on normal HSPC. Many inhibitors of translation have been used clinically to treat leukemia, but with limited success. All such inhibitors were designed to repress cap-dependent translation; inhibitors for cap-independent translation have not yet been identified. We want to: 1) develop morpholino antisense oligonucleotides (MPOs) to specifically inhibit RUNX- ETO by targeting the translation of P2-transcripts; 2) identify chemicals that can repress P2-controlled RUNX1-ETO expression and IRES-mediated RUNX1-ETO translation in leukemia cells by a large-scale screening of a small molecule library. To do so, we generated an RE-GFP reporter cell line by fusing a GFP to the last exon of RUNX1-ETO fusion gene in Kasumi-1 cells. The intensity of GFP in such reporter cells reliably reflects the levels of RUNX1-ETO protein. Using this reporter cell line, we will evaluate the efficiency of MPOs in order to identify the most efficient MPOs for potential clinical translation and screen a small molecule library in order to identify small molecules that can specifically repress the production of RUNX1-ETO protein in leukemia cells. The candidates identified in such screening will further be evaluated to determine whether they can repress the growth of RUNX1-fusion gene+ leukemia cells without affecting the growth of normal HSPC. We also want to determine whether the candidates inhibit the production of RUNX1-ETO by repressing transcription or translation of fusion gene.

Hematopoietic stem cell (HSC) transplantation is the most effective therapy for life-threatening hematopoietic diseases such as leukemia and many solid tumors. However such therapy is significantly limited by the shortage of HSC resources. In vitro HSC expansion is believed to be the most effective and applicable strategy to address this problem. Despite significant efforts toward this end, our ability to expand HSCs remains clinically unsuccessful because the key factors that promote HSC self-renewal have not been identified. Two reasons might explain why we failed to do so: 1) the levels of the factor(s) in bone marrow (BM) are very low which cannot be detected by previous used techniques; 2) a proper combination of factors might be required which needs a reliable assay to determine. HSC behaviors are tightly controlled by specialized BM niche cells and their secreted factors. We found that in mice, HSCs in the BM expand by approximately 100 times during the 1-2 weeks of postnatal development. We also detected an approximately 4-fold increase in functional HSC numbers on day 4 following a single treatment with 5FU. By taking the advantage of these two HSC expansion models, we examined the dynamic changes of niche cells by flow cytometry and their gene expression profiles by RNA-sequence. Such assays allow us to sensitively detect the changes in lower frequency of niche cells at 0.01% level and lower expressing genes. We specifically searched for the niche cells and their secreted factors that were increased prior to HSC expansion in both models. We speculate that such niche cells and factors might be potential candidates of HSC stimuli. In addition, we also developed a very simple and reliable ex vivo assay to evaluate the expansion of functional HSCs. We found that the dynamic changes of the CD31high endothelial cells and CD140a+CD51+ mesenchymal stem cells (MSCs), as well as a group of 78 cytokine-encoding genes expressed by these two types of niche cells are closely correlated with HSC expansion in both models. Among these cytokines, by functional analysis, we have already shown that R-spondin 2 (Rspo2) and Rspo3, Wnt agonists, can promote 15-20-fold expansion of murine HSCs in a Wnt-dependent manner. We intend to study the molecular mechanism by which Rspo2 stimulates HSC expansion (Aim 1). We will further characterize the phenotype of CD140a+CD51+ MSCs and CD31hi cells and determine the molecular mechanism by which these two types of niche cells synergistically promote HSC expansion (Aim 2). We want to determine how long the Rspo2 can sustain HSC expansion in vitro and whether Rspo2 induces genomic mutations in HSCs. We also want to evaluate whether Rspo2, CD140a+CD51+ MSCs and CD31hi cells can promote human cord blood HSCs and mobilized peripheral blood HSCs (Aim 3). We expect to identify the key stimuli of HSC self-renewal and develop an improved in vitro HSC expansion system which can be used for clinical HSC transplantation therapy. The results expected from this study will also help us to understand the molecular mechanism by which HSC activity is regulated in both normal and pathological hematopoiesis.

Both NF-?B and Stat3 are abnormally activated in leukemic blasts and are implicated in drug-resistance and poor prognosis, suggesting they could be potential targets for therapy. We found that inactivation of both NF-?B and Stat3 signaling pathways synergistically represses self-renewal and drug-resistance in leukemia stem cells (LSCs), suggesting a compensatory role for these two pathways in the pathogenesis of leukemia. Stat3? and Stat3? are two major splicing isoforms. Active Stat3? promotes tumor growth by regulating target gene expression (functions as a transcription factor) and controlling mitochondrial production of ATP and ROS (functions as a regulator of the electron transport chain), while Stat3? lacks a transactivation domain and functions as a dominant-negative to Stat3?. All currently used inhibitors of Stat3 only repress its transcriptional activity without taking consideration of its mitochondrial activity, which might explain why these inhibitors failed to repress leukemia in patients. It was reported that induction of the switch from Stat3? to Stat3? provides a better tumor repressive effect than inhibition of both isoforms. We found that we can induce such a switch by inhibiting the serine/threonine-protein kinases receptor-interacting protein kinase 1 (Rip1) and Rip3. Rip3 and NF-?B are parallel downstream signaling pathways of Rip1, mediating cytokine-induced kinase- dependent and -independent activities of Rip1. We found that a moderate level of activation of Rip1-Rip3 kinase signaling exists in acute myeloid leukemia (AML) cells with MLL1-rearrangement (MLL-r) or NPM1 mutation (NPM1c+). Rip1-Rip3 signaling plays distinct roles in normal hematopoietic stem/progenitor cells?HSPCs?and AML cells. In HSPCs, Rip1-Rip3 signaling mediates TNF? and IL1?-induced necroptosis, while in AML cells, the moderate activation of such signaling is required for maintaining the levels of Stat3? by inhibition of calpain (CAPN), a family of proteolytic enzymes. CAPN reduces Stat3? and enhances Stat3? by specifically cleaving Stat3? protein and also SFRS5, a splice regulator for alternative splicing for Stat3?. Inhibition of Rip1-Rip3 kinase signaling results in depletion of Stat3? and an increase of Stat3?. Our study suggested that, as with co-inhibition of Stat3 and NF-?B, co-inhibition of Rip1-Rip3 signaling and NF-?B also compromises self-renewal of LSCs and sensitizes AML to standard chemotherapy. We want to test our novel combination treatment regimen in primary human AML cells using xenograft models. We also intend to elucidate the molecular mechanisms by which Stat3 and NF-?B regulate self-renewal and drug-resistance in LSCs as well as the molecular mechanism by which Rip3 signaling regulates CAPN-dependent Stat3 isoform switch. The expected results of this study will allow us to determine whether combinations of currently known inhibitors of Rip1/Rip3 and NF-kB signaling could improve treatment for MLL-r and NPM1c+ AML when combined with standard chemotherapy. The mechanistic studies will provide detailed information allowing us to more effectively target the Rip3-CAPN-Stat3 pathway to treat AML.