Phosphorylation Of RUNX1 Research Articles

CDK9 phosphorylates RUNX1 to promote megakaryocytic fate in megakaryocytic-erythroid progenitors

AbstractThe specification of megakaryocytic (Mk) or erythroid (E) lineages from primary human megakaryocytic-erythroid progenitors (MEPs) is crucial for hematopoietic homeostasis, yet the underlying mechanisms regulating fate specification remain elusive. In this study, we identify RUNX1 as a key modulator of gene expression during MEP fate specification. Overexpression of RUNX1 in primary human MEPs promotes Mk specification, whereas pan-RUNX inhibition favors E specification. Although total RUNX1 levels do not differ between Mk progenitors (MkPs) and E progenitors (ErPs), there are higher levels of serine-phosphorylated RUNX1 in MkPs than ErPs, and mutant RUNX1 with phosphorylated-serine/threonine mimetic mutations (RUNX1-4D) significantly enhances the functional efficacy of RUNX1. To model the effects of RUNX1 variants, we use human erythroleukemia (HEL) cell lines expressing wild-type (WT), phosphomimetic (RUNX1-4D), and nonphosphorylatable (RUNX1-4A) mutants showing that the 3 forms of RUNX1 differentially regulate expression of 2625 genes. Both WT and RUNX1-4D variants increase expression in 40%, and decrease expression in another 40%, with lesser effects of RUNX1-4A. We find a significant overlap between the upregulated genes in WT and RUNX1-4D–expressing HEL cells and those upregulated in primary human MkPs vs MEPs. Although inhibition of known RUNX1 serine/threonine kinases does not affect phosphoserine RUNX1 levels in primary MEPs, specific inhibition of cyclin dependent kinase 9 (CDK9) in MEPs leads to both decreased RUNX1 phosphorylation and increased E commitment. Collectively, our findings show that serine/threonine phosphorylation of RUNX1 promotes Mk fate specification and introduce a novel kinase for RUNX1 linking the fundamental transcriptional machinery with activation of a cell type–specific transcription factor.

AMPK-induced novel phosphorylation of RUNX1 inhibits STAT3 activation and overcome imatinib resistance in chronic myelogenous leukemia (CML) subjects

Imatinib resistance remains an unresolved problem in CML disease. Activation of JAK2/STAT3 pathway and increased expression of RUNX1 have become one reason for development of imatinib resistance in CML subjects. Metformin has gained attention as an antileukemic drug in recent times. However, the molecular mechanism remains elusive. The present study shows that RUNX1 is a novel substrate of AMP-activated kinase (AMPK), where AMPK phosphorylates RUNX1 at Ser 94 position. Activation of AMPK by metformin could lead to increased cytoplasmic retention of RUNX1 due to Ser 94 phosphorylation. RUNX1 Ser 94 phosphorylation resulted in increased interaction with STAT3, which was reflected in reduced transcriptional activity of both RUNX1 and STAT3 due to their cytoplasmic retention. The reduced transcriptional activity of STAT3 and RUNX1 resulted in the down-regulation of their signaling targets involved in proliferation and anti-apoptosis. Our cell proliferation assays using in vitro resistant cell line models and PBMCs isolated from CML clinical patients and normal subjects demonstrate that metformin treatment resulted in reduced growth and improved imatinib sensitivity of resistant subjects.

Cell Cycle Regulates Phosphorylation of RUNX1 to Modulate Megakaryocyte-Erythroid Progenitor Fate Specification

Human Megakaryocytic-Erythroid Progenitors (MEP) produce megakaryocytic progenitors (MkP) and erythroid progenitors (ErP). Though some of the players have been identified, the molecular mechanisms underlying the MEP fate decision have not yet been determined. Using a functional single cell CFU-Mk/E assay, and single-cell RNA sequencing (scRNAseq), we have revealed that MEP cell cycling regulates MEP fate decisions with decreased cycling promoting Mk fate commitment and increased cycling promoting E fate commitment (Lu et al, Cell Reports, 2018). Our data point to RUNX1 (aka AML1), already known to be important for both Mk and E maturation, as playing a key role in MEP fate determination. RUNX1 target genes vary significantly between Common Myeloid Progenitors (CMP), MEP, MkP, and ErP. For example, the RUNX1 targets MPL, FLI1, and THBS1 are higher in MkP than in MEP and lowest in CMP and ErP. Analysis of scRNAseq data indicates that 11.8% and 9.3% of differentially (adj.p < 0.05, fold change > 2) expressed genes are predicted RUNX1 targets when comparing MEP to MkP and MEP to ErP, respectively (p=2.9e-8 and 5.6e-16). However, RUNX1 mRNA levels do not change significantly between CMP, MEP, MkP and ErP. We therefore assessed whether total RUNX1 protein levels and post-translational modifications change with Mk and E lineage commitment. Intracellular staining normalized to levels in CMP revealed that total RUNX1 protein is elevated 1.34-fold from CMP to MEP, is highest (1.65-fold) in MkP, and is intermediate in ErP (1.5-fold) consistent with RUNX1 perhaps promoting Mk over E fate. RUNX1 can be phosphorylated by cyclin dependent serine/threonine kinases (CDKs) as well as Src tyrosine kinases. Significant differences in RUNX1 phosphoprotein levels were revealed by intracellular FACS assays for differential phosphorylation of RUNX1. Phosphoserine modified RUNX1 (Ser21, Ser276, and Ser 397) levels, which likely reflect active RUNX1, are highest in MkP (1.28-fold over MEP), and diminished in ErP (-1.32-fold over MEP). To determine whether RUNX1 and phosphor-RUNX1 variants affect MEP commitment, we overexpressed wildtype (WT) RUNX1 or RUNX1-S4D (phospho-mimetic on S249, S277, T273, S276 residues). WT RUNX1 overexpression affected fate specification of primary MEP (p<0.05) with an increase in the Mk-only colony percentage from 24% to 43% at the expense of E-only colonies (from 26% to 16%) with little change in bipotent CFU-Mk/E. RUNX1-S4D more dramatically (p<0.04) increased Mk-only colonies (from 25% to 75%) and decreased E-only (from 25% to < 4%) and Mk/E (from 50% to 21%) colonies. Surprisingly, overexpression of RUNX1-S4D in FACSorted primary human ErP caused them to develop more (p<0.01) Mk/E-colonies (from 12% to 60%) with a dramatic decrease in E-only colonies (from 85% to 28%) suggesting that the E commitment of ErP is reversed (or reprogrammed) by activated RUNX1. In contrast, RUNX1 variants mimicking tyrosine phosphorylation (RUNX1-Y6D) promote E fate specification. Tyrosine phosphorylation on RUNX1 is mediated by Src family kinases (Huang te al, Genes Dev., 2012). We found that Src kinase inhibition results in increasing Mk lineage bias (from 20% to 71%) suggest differential phosphorylation of RUNX1 plays a pivotal role in MEP fate specification. Epigenetic modifications induced by phospho-serine vs phospho-tyrosine mimetics are currently under evaluation. Combining our data on cell cycle control of MEP with RUNX1 reveals that RUNX1 inhibition prevents the Mk bias induced by slowing the cell cycle. For example, the RUNX1 inhibitor Ro5-3335 causes MEP to have an Er bias from 27% to 61% while the mTOR inhibitor Rapamycin (slows MEP cell cycle) induces a Mk bias from 19% to 52%. The combination of Ro5-3335 plus Rapa results in increased Er from 27% to 45% suggest that RUNX1 influences MEP fate specification downstream of the cell cycle machinery. In summary, the phosphorylation status of RUNX1 in primary human hematopoietic progenitors regulates MEP fate commitment. Detailed mechanisms linking the cell cycle machinery to RUNX1 protein levels and function toward Mk vs E specification are being explored. Disclosures No relevant conflicts of interest to declare.

PAK4 phosphorylating RUNX1 promotes ERα-positive breast cancer-induced osteolytic bone destruction.

The biological function of nuclear PAK4 in ERα-positive breast cancer osteolytic bone destruction remains unclear. Here, we find that the nuclear PAK4 promotes osteoclastogenesis and tumor-induced osteolysis via phosphorylating RUNX1. We show that nuclear PAK4 interacts with and phosphorylates RUNX1 at Thr-207, which induces its localization from the nucleus to the cytoplasm and influences direct interaction with SIN3A/HDAC1 and PRMT1. Furthermore, we reveal that RUNX1 phosphorylation by PAK4 at Thr-207 promotes osteolytic bone destruction via targeting downstream genes related to osteoclast differentiation and maturation. Importantly, we verify changes in RUNX1 subcellular localization when nuclear PAK4 is positive in breast cancer bone metastasis tissues. Functionally, we demonstrate that RUNX1 phosphorylation promotes osteolytic bone maturation and ERα-positive breast cancer-induced osteolytic bone damage in the mouse model of orthotopic breast cancer bone metastasis. Our results suggest PAK4 can be a therapeutic target for ERα-positive breast cancer osteolytic bone destruction.

PS1326 IMPLEMENTATION OF FUNCTIONAL ASSAYS TO CHARACTERIZE AND DETERMINE THE CLINICAL IMPACT OF RUNX1 VARIANTS OF UNCERTAIN SIGNIFICANCE IN SPORADIC DISEASES AS WELL AS IN THE CONTEXT OF FPDMM

Background:Runt‐related transcription factor 1 (RUNX1) encodes for a transcription factor subunit. RUNX1 binds core binding factor β (CBFβ) and together they function as a transcriptional regulator. While dimerization with CBFb augments DNA‐binding affinity of RUNX1, its subsequent phosphorylation enhances its capacity to activate transcription. Regarding hematopoiesis, RUNX1 is known as a master regulator of transcription. In hematological malignancies, somatic genetic alterations of RUNX1 are frequently observed. Pathogenic germline variants of RUNX1 cause familial platelet disorder with associated myeloid malignancies (FPDMM). FPDMM is an autosomal dominant inherited disease associated with mild to moderate thrombocytopenia, platelet aggregation defects, and an increased risk to develop hematological malignancies, mainly MDS and AML. Functional consequences of RUNX1 missense variants detected in diagnostic analyses are often unclear.Aims:Develop a first set of assays, in order to functionally characterize RUNX1 variants of uncertain significance (VUS) and determine their clinical impact. As positive controls, known pathogenic RUNX1 variants were investigated in parallel.Methods:The capability of RUNX1 variants to bind to CBFβ was analyzed by a flow cytometry‐based fluorescence energy transfer (FRET) assay. Expression and phosphorylation of RUNX1 variants were investigated by means of western blotting. Finally, the ability to activate transcription was assessed using four independent luciferase reporter constructs. All assays were performed in HEK293T cells.Results:In comparison to RUNX1 wild type, we analyzed nine missense VUS and six known pathogenic variants of RUNX1. We observed significant impairment in all functional assays for known pathogenic variants. For six out of nine RUNX1 VUS, we gained evidence for their functional relevance. In the FRET assay, one RUNX1 VUS showed no dimerization and reduced dimerization ability was detected for two other VUS. By mobility shift in western blots, four VUS showed markedly decreased amounts of phosphorylation compared to wild type protein. Impaired phosphorylation was detectable for two additional VUS. Results were confirmed using a specific antibody against phosphorylated Ser249 of RUNX1. In contrast to RUNX1 wildtype, three VUS displayed no activation of all four luciferase reporters. Decreased activation capability of individual reporters was seen for three other VUS. In addition, variants lacking transcriptional activation ability, including the known pathogenic variants, tended to accumulate in HEK293T cells indicating their reduced degradation.Summary/Conclusion:In a proof of principle approach, we succeed to establish functional assays addressing dimerization, phosphorylation, and transcriptional activation of RUNX1. Functional data of individual variants are in line with the hierarchy of molecular events of RUNX1 (i.e., 1. dimerization, 2. phosphorylation, and 3. transcriptional activation). Our assays can be used to test both somatic and germline variants and support the classification of VUS as pathogenic if significant functional impairment is observed. This allows testing of healthy family members to exclude carriers as potential HSCT donors and implement rational screening measures. In the future, we will study additional VUS and establish further assays to characterize RUNX1 variants in more detail.

Runx1 Phosphorylation by Src Increases Trans-activation via Augmented Stability, Reduced Histone Deacetylase (HDAC) Binding, and Increased DNA Affinity, and Activated Runx1 Favors Granulopoiesis

Src phosphorylates Runx1 on one central and four C-terminal tyrosines. We find that activated Src synergizes with Runx1 to activate a Runx1 luciferase reporter. Mutation of the four Runx1 C-terminal tyrosines to aspartate or glutamate to mimic phosphorylation increases trans-activation of the reporter in 293T cells and allows induction of Cebpa or Pu.1 mRNAs in 32Dcl3 myeloid cells, whereas mutation of these residues to phenylalanine to prevent phosphorylation obviates these effects. Three mechanisms contribute to increased Runx1 activity upon tyrosine modification as follows: increased stability, reduced histone deacetylase (HDAC) interaction, and increased DNA binding. Mutation of the five modified Runx1 tyrosines to aspartate markedly reduced co-immunoprecipitation with HDAC1 and HDAC3, markedly increased stability in cycloheximide or in the presence of co-expressed Cdh1, an E3 ubiquitin ligase coactivator, with reduced ubiquitination, and allowed DNA-binding in gel shift assay similar to wild-type Runx1. In contrast, mutation of these residues to phenylalanine modestly increased HDAC interaction, modestly reduced stability, and markedly reduced DNA binding in gel shift assays and as assessed by chromatin immunoprecipitation with the -14-kb Pu.1 or +37-kb Cebpa enhancers after stable expression in 32Dcl3 cells. Affinity for CBFβ, the Runx1 DNA-binding partner, was not affected by these tyrosine modifications, and in vitro translated CBFβ markedly increased DNA affinity of both the translated phenylalanine and aspartate Runx1 variants. Finally, further supporting a positive role for Runx1 tyrosine phosphorylation during granulopoiesis, mutation of the five Src-modified residues to aspartate but not phenylalanine allows Runx1 to increase Cebpa and granulocyte colony formation by Runx1-deleted murine marrow.

Multiple phosphorylation sites are important for RUNX1 activity in early hematopoiesis and T‐cell differentiation

RUNX1 is essential for definitive hematopoiesis and T-cell differentiation. It has been shown that RUNX1 is phosphorylated at specific serine and threonine residues by several kinase families. However, it remains unclear whether RUNX1 phosphorylation is absolutely required for its biological functions. Here, we evaluated hematopoietic activities of RUNX1 mutants with serine (S)/threonine (T) to alanine (A), aspartic acid (D), or glutamic acid (E) mutations at phosphorylation sites using primary culture systems. Consistent with the results of knockin mice, RUNX1-2A, carrying two phospho-deficient mutations at S276 and S293, retained hematopoietic activity. RUNX1-4A, carrying four mutations at S276, S293, T300, and S303, showed impaired T-cell differentiation activity, but retained the ability to rescue the defective early hematopoiesis of Runx1-deficient cells. Notably, RUNX1-5A, carrying five mutations at S276, S293, T300, S303, and S462, completely lost its hematopoietic activity. In contrast, the phospho-mimic proteins RUNX1-4D/E and RUNX1-5D/E exhibited normal function. Our study identifies multiple phosphorylation sites that are indispensable for RUNX1 activity in hematopoiesis.

Differential effects of inhibition of bone morphogenic protein (BMP) signalling on T‐cell activation and differentiation

Bone morphogenetic proteins (BMPs) are involved in patterning and cellular fate in various organs including the thymus. However, the redundancy of BMPs and their receptors have made it difficult to analyse their physiological roles. Here, we investigated the role of BMP signalling in peripheral CD4(+) T cells by analysing the effects of an inhibitor of BMP signalling, dorsomorphin. Dorsomorphin suppressed phosphorylation of SMAD1/5/8, suggesting that BMP signalling naturally occurs in T cells. At high doses, dorsomorphin suppressed proliferation of T cells in a dose-dependent manner, inducing G1 arrest. Also, dorsomorphin suppressed Th17 and induced Treg-cell differentiation, while preserving Th2 differentiation. Dorsomorphin efficiently suppressed IL-2 production even at low doses in mouse CD4(+) T cells, suggesting that the BMP-Smad signalling physiologically regulates IL-2 transcription in these cells. In addition, recombinant BMP2 induced a dose-dependent multiphasic pattern of IL-2 production, while Noggin suppressed IL-2 production at higher doses in Jurkat cells. Notably, BMP signalling controlled the phosphorylation of RUNX1, revealing the molecular nature of its effect. Collectively, we describe multiple effects of dorsomorphin and Noggin on T-cell activation and differentiation, demonstrating a physiological role for BMP signalling in these processes.

Phosphorylation of RUNX1 by Cyclin-dependent Kinase Reduces Direct Interaction with HDAC1 and HDAC3

RUNX1 regulates formation of the definitive hematopoietic stem cell and its subsequent lineage maturation, and mutations of RUNX1 contribute to leukemic transformation. Phosphorylation of Ser-48, Ser-303, and Ser-424 by cyclin-dependent kinases (cdks) increases RUNX1 trans-activation activity without perturbing p300 interaction. We now find that endogenous RUNX1 interacts with endogenous HDAC1 or HDAC3. Mutation of the three RUNX1 serines to aspartic acid reduces co-immunoprecipitation with HDAC1 or HDAC3 when expressed in 293T cells; mutation of these three serines to alanine increases HDAC interaction, and mutation of each serine individually to aspartic acid also reduces these interactions. GST-RUNX1 isolated from bacterial extracts bound in vitro translated HDAC1 or HDAC3, and these interactions were weakened by mutation of Ser-48, Ser-303, and Ser-424 to aspartic acid. The ability of RUNX1 phosphorylation and not only serine to aspartic acid conversion to reduce HDAC1 binding was demonstrated using wild-type GST-RUNX1 phosphorylated in vitro using cdk1/cyclinB and by exposure of 293T cells transduced with RUNX1 and HDAC1 to roscovitine, a cdk inhibitor. Finally, RUNX1 or RUNX1(tripleD), in which Ser-48, Ser-303, and Ser-424 are mutated to aspartic acid, stimulated proliferation of transduced, lineage-negative murine marrow progenitors more potently than did RUNX1(tripleA), in which these serines are mutated to alanine, suggesting that stimulation of RUNX1 trans-activation by cdk-mediated reduction in HDAC interaction increases marrow progenitor cell proliferation.

Tyrosine Phosphorylation of Runx1 In Megakaryocytes by Src Family Kinases

AbstractAbstract 742Runx1 and its cofactor, CBF-beta, are the most frequent targets of chromosomal translocations in human leukemias. Point mutations in Runx-1 also occur in some cases of myelodysplastic syndrome and undifferentiated leukemia. During normal hematopoiesis, Runx1 is required for the ontogeny of all definitive hematopoietic stem cells and for the proper maturation of megakaryocytes (Mks) and lymphocytes. Despite these critical roles, the regulation of Runx1 activity via cell signaling pathways remains incompletely understood. Here, we report that Runx-1 is tyrosine phosphorylated in Mks. This occurs on multiple residues and is mediated by src-family tyrosine kinases (SFKs). Loss of Runx1 tyrosine phosphorylation correlates with phorbol ester induced differentiation of L8057 megakaryoblastic cells, suggesting a negative regulatory function. Consistent with this model, retroviral expression of a tyrosine non-phosphorylatable mutant Runx1 molecule increases primary murine fetal liver Mk maturation and Runx1 target gene expression to a greater extent than wild type Runx1. Moreover, treatment of wild type primary Mks with SFK inhibitors markedly enhances Mk maturation, as previously reported (Lannutti BJ, et al 2005 Blood;105:3875-3878). Treatment of L8057 cells with the pan-tyrosine phosphatase inhibitor Na3VO4, significantly increases Runx1 tyrosine phosphorylation levels, suggesting that tyrosine phosphorylation of Runx1 is dynamically regulated under steady-state conditions. Using a proteomic approach, we found that Runx1 physically interacts with the non-receptor tyrosine phosphatase SHP-2 (Ptpn11). We validated this interaction and showed that it occurs via direct interactions involving the Runx1 runt domain. ShRNA mediated knock down of SHP-2 in L8057 cells increases Runx1 tyrosine phosphorylation levels. Conditional knockout of SHP-2 in Mks using SHP-2fl/fl, PF4-Cre mice leads to reduced peripheral blood platelet counts and delayed platelet recovery following transient anti-GPIb antibody induced immune thrombocytopenia. Lastly, we show that treatment of TPA-induced L8057 cells with Na3VO4 markedly diminishes binding between Runx1 and the key Mk transcription factor GATA-1. Taken together, our data suggest that tyrosine phosphorylation of Runx1 via SFKs inhibits Runx1 function. Dephosphorylation, at least in part via SHP-2, relieves this inhibition and promotes Mk maturation. These effects are likely mediated through altered protein-protein interactions. Disclosures:No relevant conflicts of interest to declare.

Accelerated Leukemogenesis by Truncated CBFβ-SMMHC Defective in High-Affinity Binding with RUNX1

Dominant RUNX1 inhibition has been proposed as a common pathway for CBF leukemia. CBF beta-SMMHC, a fusion protein in human acute myeloid leukemia (AML), dominantly inhibits RUNX1 largely through its RUNX1 high-affinity binding domain (HABD). However, the type I CBF beta-SMMHC fusion in AML patients lacks HABD. Here, we report that the type I CBF beta-SMMHC protein binds RUNX1 inefficiently. Knockin mice expressing CBF beta-SMMHC with a HABD deletion developed leukemia quickly, even though hematopoietic defects associated with Runx1-inhibition were partially rescued. A larger pool of leukemia-initiating cells, increased MN1 expression, and retention of RUNX1 phosphorylation are potential mechanisms for accelerated leukemia development in these mice. Our data suggest that RUNX1 dominant inhibition may not be a critical step for leukemogenesis by CBF beta-SMMHC.

Phosphorylation of RUNX1 by Cyclin-Dependent Kinase Reduces Direct Interaction with HDAC1 and HDAC3 and Stimulates Marrow Progenitor Proliferation.

Abstract 2508Poster Board II-485RUNX1 helps direct the formation of definitive HSC during embryogenesis and the subsequent maturation of the lymphoid, myeloid, and megakaryocytic lineages in adults. RUNX1 regulates both lineage-specific genes and G1 to S cell cycle progression, the latter in part via induction of cyclin D3 and cdk4 transcription. Recent studies demonstrate that RUNX1 homologs in sea urchin blastocyts or in C. elegans seam cells, a stem cell subset, also stimulate cell proliferation. Of note, RUNX1 activities are commonly perturbed in human AML or ALL cases. RUNX1 residues S48, S303, and S424 fit the consensus, (S/T)PX(R/K), for phosphorylation by cyclin dependent kinases (cdk). Mutation of these residues to the phosphomimetic aspartic acid increases RUNX1 trans-activation, and mutation to alanine reduces trans-activation, without perturbing p300 interaction (Zhang et al 2008). We now demonstrate that cdk phosphorylation of RUNX1 reduces its interaction with HDAC1 and HDAC3. First, co-immunoprecipitation of FLAG-tagged HDAC1 or HDAC3 with myc-tagged RUNX1, RUNX1(tripleA), with S48, S303, and S424 mutated to alanine, or RUNX1(tripleD), with these three residues mutated to aspartic acid, was assessed using extracts from transiently transfected 293T cells. Both HDAC1 and HDAC3 bound RUNX1(tripleD) with greatly reduced affinity compared to RUNX1 or its tripleA variant. Consistent with previous mapping of HDAC1 and HDAC3 interaction to the C-terminal residues 380–432 of RUNX1, when this same assay was repeated using RUNX1(S48D), RUNX1(S303D), or RUNX1(S424D), carrying single residue mutations to aspartic acid, only S424D reduced HDAC1 binding, and S424D had a greater effect than S48D or S303D on HDAC3 binding. Second, when GST-RUNX1, GST-RUNX1(424A), or GST-RUNX1(424D) fusion proteins isolated from bacteria were incubated with in vitro translated, S35-methionine labeled HDAC1 or HDAC3, reduced interaction was again seen to the 424D variant. Finally, incubation of wild-type GST-RUNX1 with cdk1/cyclinB led to reduced interaction with radio-labeled HDAC1 or HDAC3, indicating that cdk phosphorylation and not just change of cdk target residues to aspartic acid reduces HDAC binding to RUNX1. In addition, endogenous HDAC1 in the Ba/F3 hematopoietic cell line co-immunoprecipitated with exogenous RUNX1, and we are attempting to optimize methods to detect interaction between the endogenous proteins. We previously provided data suggesting that RUNX1 cdk phosphorylation stimulates proliferation of the Ba/F3 hematopoietic cell line (Zhang et al 2008). We now find that RUNX1 or RUNX1(tripleD) reproducibly stimulate proliferation of transduced, lineage-negative murine marrow progenitors more potently than RUNX1(tripleA). Together our findings indicate that stimulation of RUNX1 trans-activation activity by cdk-mediated reduction in direct HDAC1 and HDAC3 interaction facilitates hematopoietic progenitor cell proliferation. Future studies will endeavor to determine whether activation of cdk in HSC or myeloid progenitors, for example by Notch or Wnt signaling, stimulates RUNX1 cdk phosphorylation to facilitate induction of cell cycle regulatory genes and cell proliferation. Conversely, reduced RUNX1 cdk phosphorylation may facilitate HSC quiescence. Disclosures:No relevant conflicts of interest to declare.

Phosphorylation of Runx1 by Cyclin-Dependent Kinases Regulates Its Interaction with HDAC1 and HDAC3.

Runx1/AML1 is a key regulator of hematopoiesis and leukemic transformation, as RUNX1(−/−) mice do not develop definitive hematopoietic stem cells, and sever alleukemic oncogenes, e.g. AML1-ETO, CBFβ-SMMHC, or TEL-AML1, inhibit Runx1activities. We have investigated regulation of cell cycle progression by Runx1. Runx1stimulates G1 to S cell cycle progression in hematopoietic cell lines and in transduced myeloid progenitors, and inhibition of Runx1 by CBFβ-SMMHC or AML1-ETO slows G1 progression. Runx1 induces cdk4 and cyclin D3 transcription, and exogenous cdk4, cyclin D2, or c-Myc overcomes inhibition of G1 progression by CBF oncoproteins. In addition to regulating cell cycle progression, Runx1 protein levels are themselves increased as hematopoietic cells progress from G1 to S to G2/M, though mRNA levels remain constant. Runx1 contains three consensus cdk sites, (S/T)PX(R/K), S48, S303, and S424, and using phospho-specific antisera we find that each of these is modified in hematopoietic cells. Mutation of these serines to aspartic acid, mimicking phosphorylation, increases trans-activation of a reporter containing four CBF sites or the TCRβ promoter, whereas mutation to alanine reduces trans-activation. p300 interacts similarly with Runx1(tripleA) and Runx1(tripleD). We have now evaluated interaction of HDACs1–8 with these variants and Runx1 and find that both HDAC1 and HDAC3 have reduced affinity for RUNX1(tripleD), as assessed by co-immunoprecipitation from transiently transfected 293T cells. Evaluation of single serine residue mutants (S48D, S303D, and S424D) demonstrates reduced affinity of HDAC1 or HDAC3 specifically for the Runx1(S424D) mutant, consistent with previous mapping of the Runx1:HDAC1 and Runx1:HDAC3 interactions to this region of Runx1. Thus, cdk phosphorylation of Runx1 S424 reduces affinity for HDAC1 and HDAC3, increasing Runx1 trans-activation potency. Regulation of Runx1 activity by cdks may control key developmental processes, including expansion of definitive HSC during development and regulation of the balance between adult HSC quiescence and proliferation.

PEBP2-β/CBF-β–dependent phosphorylation of RUNX1 and p300 by HIPK2: implications for leukemogenesis

The heterodimeric transcription factor RUNX1/PEBP2-beta (also known as AML1/CBF-beta) is essential for definitive hematopoiesis. Here, we show that interaction with PEBP2-beta leads to the phosphorylation of RUNX1, which in turn induces p300 phosphorylation. This is mediated by homeodomain interacting kinase 2 (HIPK2), targeting Ser(249), Ser(273), and Thr(276) in RUNX1, in a manner that is also dependent on the RUNX1 PY motif. Importantly, we observed the in vitro disruption of this phosphorylation cascade by multiple leukemogenic genetic defects targeting RUNX1/CBFB. In particular, the oncogenic protein PEBP2-beta-SMMHC prevents RUNX1/p300 phosphorylation by sequestering HIPK2 to mislocalized RUNX1/beta-SMMHC complexes. Therefore, phosphorylation of RUNX1 appears a critical step in its association with and phosphorylation of p300, and its disruption may be a common theme in RUNX1-associated leukemogenesis.

Phosphorylation of Runx1 at Ser249, Ser266, and Ser276 is dispensable for bone marrow hematopoiesis and thymocyte differentiation

Runx1, one of three mammalian runt-domain transcription factor family proteins, is essential for definitive hematopoiesis. Based on transfection assays, phosphorylation of Runx1 at the three serine residues, Ser249, Ser266, and Ser276, was thought to be important for trans-activation activity of Runx1. By using “knock-in” gene targeting, we generated mouse strains expressing mutant Runx1 protein that harbored a combined serine-to-alanine substitution at either of two residues, Ser249/Ser266 or Ser249/Ser276. Either mutation resulted in a lack of major phosphorylated form of Runx1. However, while loss of definitive hematopoiesis and impaired thymocyte differentiation was observed following the loss of Runx1, these phenotypes were rescued in those mice lacking the major phosphorylated form of Runx1. These results not only challenge the predicted regulation of Runx1 activity by phosphorylation at these serine residues, but also reaffirm the effectiveness of “knock-in” mutagenesis as a powerful tool for addressing the physiological relevance of post-translation modifications.

Cyclin-dependent kinase phosphorylation of RUNX1/AML1 on 3 sites increases transactivation potency and stimulates cell proliferation

AbstractRUNX1/AML1 regulates lineage-specific genes during hematopoiesis and stimulates G1 cell-cycle progression. Within RUNX1, S48, S303, and S424 fit the cyclin-dependent kinase (cdk) phosphorylation consensus, (S/T)PX(R/K). Phosphorylation of RUNX1 by cdks on serine 303 was shown to mediate destabilization of RUNX1 in G2/M. We now use an in vitro kinase assay, phosphopeptide-specific antiserum, and the cdk inhibitor roscovitine to demonstrate that S48 and S424 are also phosphorylated by cdk1 or cdk6 in hematopoietic cells. S48 phosphorylation of RUNX1 paralleled total RUNX1 levels during cell-cycle progression, S303 was more effectively phosphorylated in G2/M, and S424 in G1. Single, double, and triple mutation of the cdk sites to the partially phosphomimetic aspartic acid mildly reduced DNA affinity while progressively increasing transactivation of a model reporter. Mutation to alanine increased DNA affinity, suggesting that in other gene or cellular contexts phosphorylation of RUNX1 by cdks may reduce transactivation. The tripleD RUNX1 mutant rescued Ba/F3 cells from inhibition of proliferation by CBFβ-SMMHC more effectively than the tripleA mutant. Together these findings indicate that cdk phosphorylation of RUNX1 potentially couples stem/progenitor proliferation and lineage progression.

Cyclin-Dependent Kinase Phosphorylation of RUNX1/AML1 on Three Sites Increases Trans-Activation Potency and Stimulates Cell Proliferation.

RUNX1/AML1 regulates lineage-specific genes during hematopoiesis and also stimulates G1 cell cycle progression. CBFβ-SMMHC or AML1-ETO dominantly inhibit RUNX1 and slow G1 progression in hematopoietic cell lines or in murine or human marrow progenitors, cdk4, cyclin D2, or c-Myc overcome inhibition of proliferation by these CBF oncoproteins, exogenous RUNX1 stimulates G1 progression, and stimulation of G1 via deletion of p16INK4a or expression of E7 cooperates with CBFβ-SMMHC or TEL-AML1 to induce acute leukemia in mice. Induction of cdk4 or cyclin D3 transcription may underlie stimulation of G1 progression by RUNX1. Remarkably, the C. elegans ortholog of RUNX1, RNT-1, also stimulates G1 progression and couples stem cell proliferation with differentiation. Not only does RUNX1 regulate cell cycle progression, but in addition RUNX1 levels increase as hematopoietic cells progress from G1 to S and from S to G2/M. Within RUNX1, S48, S303, and S424 fit the cdk phosphorylation consensus, (S/T)PX(R/K). Phosphorylation of RUNX1 by cyclin dependent kinases on serine 303 was shown to mediate destabilization of RUNX1 in G2/M. We now find that S48 and S424 are also phosphorylated by cdk1 or cdk6. S48, S303, or S424 phosphopeptide antiserum that we developed specifically recognized kinased GST-RUNX1 and interacted with RUNX1 expressed in 293T cells or in the Ba/F3 hematopoietic cell line. S48 phosphorylation of RUNX1 paralleled total RUNX1 levels during cell cycle progression, S303 was more effectively phosphorylated in G2/M, and S424 in G1. Single, double, and triple mutation to alanine or to the partially phosphomimetic aspartic acid progressively diminished or increased trans-activation, such that the tripleA mutant activated a RUNX1 reporter 5-fold less potently than the tripleD mutant. Aspartic acid does not perfectly mimic serine phosphorylation, as illustrated by the much greater affinity of our antisera for wild-type RUNX1 versus RUNX1(tripleD), suggesting that the biologic effect of RUNX1 cdk phosphorylation is even more significant. The p300 co-activator retained interaction with the tripleA variant. The tripleD RUNX1 mutant rescued Ba/F3 cells from inhibition of proliferation by CBFβ-SMMHC more effectively than the tripleA mutant. Cdk phosphorylation of RUNX1 on three sites increases its ability to active transcription and to stimulate proliferation, potentially coupling entry of stem/progenitors into cycle with induction of genes required for hematopoietic lineage progression, such as those encoding myeloperoxidase, neutrophil elastase, the M-CSF receptor, and PU.1.

Involvement of P-TEFb/CDK9 in Cooperative Transcriptional Activation of Megakaryocytic Genes by GATA-1 and RUNX1 and in Programming of Megakaryocytic Development In Vivo.

Programming of megakaryocytic differentiation requires precise coordination of multiple signal transduction and transcription pathways. Previous in vivo and in vitro studies have implicated RUNX1 and GATA-1 as transcription factors that collaborate in the execution of this program. Analysis of the mechanism for the synergy of these two factors revealed induction of RUNX1 hyperphosphorylation by GATA-1 coexpression. A pharmacologic screen identified roscovitine as an inhibitor of the transcriptional cooperation, implicating a cyclin-dependent kinase (Cdk). A screen employing a panel of dominant-negative Cdk mutants identified Cdk9 as a critical component of the GATA-1-RUNX1 cooperation. In addition, HEXIM1, an endogenous Cdk9 inhibitor, similarly blocked transcriptional synergy. Furthermore, two kinase inhibitory compounds, DRB and flavopiridol, also blocked GATA-1-RUNX1 cooperation at concentrations specific for Cdk9 inhibition. Regarding the mechanism for GATA-1 induction of RUNX1 phosphorylation, coimmunoprecipitation experiments showed GATA-1 binding to both Cdk9 and cyclinT1. To examine the role of P-TEFb in primary megakaryocytic differentiation, human CD34+ cells in megakaryocytic cultures underwent treatment with 50 nM flavopiridol, a dose selective for Cdk9 inhibition. This treatment blocked megakaryocytic polyploidization while having no effect on the cell cycle properties of the non-megakaryocytic cells in the cultures. The treatment also impaired upregulation of CD41. Extending these findings to an in vivo model system, mice underwent treatment with daily low dose flavopiridol (5–7 mg/kg/day), a regimen previously shown to have no toxicity. Wild type C57BL/6 (wt BL/6) mice were compared with the ΔneoΔHS strain (GATA-1Lo) which has diminished GATA-1 expression in megakaryocytes. After only 1 week of treatment, the GATA-1Lo mice developed worsening thrombocytopenia associated with new-onset anemia, with several dying after 2 weeks of treatment. Flow cytometry on marrow from the treated GATA-1Lo mice revealed a marked expansion of abnormal megakaryocytes showing coexpression of CD71 plus CD41 and loss of polyploidization. Marrow and spleen histology showed extensive replacement by immature-appearing megakaryocytes with hypolobulated nuclei, as well as frequent pyknotic megakaryocytes. The control mice, flavopiridol treated wt BL/6 and saline treated GATA-1Lo, displayed none of these abnormalities. Additional experiments determined the flavopiridol effect on the GATA-1Lo mice to be completely reversible, with normalization of all parameters 2 weeks after ending treatment. In aggregate, these data implicate P-TEFb recruitment by GATA-1 in mediating cooperative activation of megakaryocytic promoters with RUNX1. This pathway may depend in part on the direct phosphorylation of RUNX1 by Cdk9. In mice, a synthetic lethal relationship between megakaryocytic GATA-1 deficiency and Cdk9 inhibition exists, manifesting as a fulminant but reversible megakaryocytic proliferative disorder reminiscent of the Down syndrome-associated megakaryocyte proliferations. A model is proposed in which P-TEFb, as a component of GATA-1-RUNX1 transcriptional complexes, plays an integral role in the specific programming of megakaryocytic differentiation, with particular importance in the unique cell cycle changes associated with this lineage.

Regulation of RUNX1 Transcriptional Function by GATA-1 Kinase Recruitment: A Novel Pathway for Megakaryocytic Gene Regulation.

RUNX1 and GATA-1 both play essential roles in the transcriptional programming of normal mammalian megakaryocytic development, deficiencies of either factor having similar phenotypic consequences. We have previously characterized physical and functional interactions between these two factors, and others have confirmed analogous cooperations in Drosophila and Danio homologs. We now present data on molecular mechanisms for the cooperation of these two factors in the transcriptional activation of the megakaryocytic aIIb integrin promoter. In these studies, GATA-2 also physically interacted with RUNX1 but failed to cooperate in transcriptional activation. In fact, increasing amounts of GATA-2 repressed the functional interplay between GATA-1 and RUNX1. Through generation of GATA-2/GATA-1 chimeras, we identified a conserved subdomain within the GATA-1 amino terminus that was both necessary and sufficient for transcriptional cooperation with RUNX1. Coexpression of wild type GATA-1 or of cooperating GATA-2/GATA-1 chimeras, but not of GATA-2 or of non-cooperating chimeras, induced a mobility shift in wild type RUNX1. Using immunoprecipitation followed by immunoblot with a panel of phosphospecific antibodies, we found GATA-1 to induce RUNX1 phosphorylation at recognition sites for cyclin-dependent kinases (cdks). Treatment of cells with roscovitine, a specific cdk inhibitor, blocked the transcriptional cooperation of GATA-1 with RUNX1 and eliminated the RUNX1 mobility shift caused by GATA-1 coexpression. Mutagenesis of RUNX1 identified a cluster of serine/threonine-proline (S/TP) sites collectively required for the transcriptional augmentation and mobility shift induced by GATA-1. In addition, intact DNA binding by RUNX1 was required for cooperation with GATA-1. These results provide a new paradigm for cooperation of interacting transcription factors, in which one partner recruits a kinase leading to phosphorylation and activation of the other partner. Furthermore, these results provide a biochemical basis for the previously inexplicable functional differences between GATA-1, which promotes megakaryocytic maturation, and GATA-2 which promotes proliferation without maturation.

Orientation-Dependent Regulation of RUNX1 and GATA-1 Transcriptional Function by the p300/CBP Coactivators.

The transcription factors RUNX1 and GATA-1, as well as the coactivators p300/CBP, have been implicated in the regulation of primary megakaryopoiesis through studies of knockout mice. In particular, p300/CBP has previously been shown to serve as a coactivator for both RUNX1 and GATA-1 in the transactivation of hematopoietic target genes. Coactivator orientation within transactivating complexes is generally not considered to influence the degree of transcriptional activity, reflecting an inherent flexibility in the spatial requirements for coactivator function. Experiments to further explore this issue showed coexpression of p300 to enhance cooperative transcriptional activation by wild type RUNX1 and GATA-1. Enforced recruitment of p300/CBP to GATA-1 through fusion of GATA-1 with the p300/CBP docking module of adenoviral E1A, E1A(1–89), enhanced GATA-1 activity alone, regardless of whether the fusion was to the amino or carboxy terminal of GATA-1. However, enforced recruitment of p300/CBP to the amino terminus of GATA-1 completely eliminated cooperation with RUNX1, and enforced recruitment of p300/CBP to the carboxy terminus of GATA-1 diminished cooperation with RUNX1. Both GATA-1 fusions retained physical interaction with RUNX1. Similarly, fusion of E1A(1–89) directly to the amino terminus of RUNX1 completely eliminated its transcriptional activity, while fusion to the carboxy terminus diminished RUNX1 transcriptional activity. For both the GATA-1 and the RUNX1 fusions, these repressive effects were attributable to the ectopic recruitment of p300/CBP because a mutation within E1A(1–89) known to specifically diminish p300/CBP recruitment, R2G, rescued the transcriptional activities. Addressing the mechanism of repression by ectopic p300/CBP, we found that E1A(1–89)-GATA-1 caused diminished serine phosphorylation within RUNX1, an effect opposite to that of wild type GATA-1 which enhanced RUNX1 phosphorylation. Similarly, E1A(1–89)-RUNX1 showed complete loss of phosphorylation on cdk target sites, as compared with wild type RUNX1. RUNX1-E1A(1–89) showed diminished phosphorylation on cdk target sites, as compared with wild type RUNX1. From these results, we conclude that p300/CBP may function as a coactivator for the RUNX1-GATA-1 complex when recruited to endogenous, wild type domains. By contrast, ectopic recruitment of p300/CBP to RUNX1, particularly to the amino terminus, targets RUNX1 for repression through inhibition or reversal of phosphorylation. Our results thus offer a novel paradigm in which the function of p300/CBP, coactivator versus repressor, may be determined by its mode of recruitment to certain transcriptional complexes. Notably, some transcription factors, such as GATA-1, have relaxed requirements for the topology of coactivator recruitment, whereas other transcription factors, such as RUNX1, have stringent requirements in this regard.