Increased Γ-globin Expression Research Articles

Transcriptional activation of the γ-globin gene in baboons treated with decitabine and in cultured erythroid progenitor cells involves different mechanisms

To investigate the mechanism(s) responsible for increased gamma-globin expression in vivo in decitabine-treated baboons and in vitro in cultured erythroid progenitor cells (EPC) from adult baboon bone marrow (BM). Fetal liver, adult BM erythroid cells pre- and post-decitabine, and cultured EPCs were analyzed for distribution of RNA polymerase II, histone acetylation, and histone H3 (lys4) trimethyl throughout the gamma-globin gene complex by chromatin immunoprecipitation. DNA methylation of the gamma-globin promoter was determined by bisulfite sequencing. Expression of the baboon Igamma- and Vgamma-globin chains was determined by high performance liquid chromatography (HPLC). Expression of BCL11A, a recently identified repressor of gamma-globin expression, was analyzed by Western blot. Increased gamma-globin expression in decitabine-treated baboons and cultured EPC correlated with increased levels of RNA polymerase II, histone acetylation, and histone H3 (lys4) trimethyl associated with the gamma-globin gene consistent with a transcriptional activation mechanism. Cultured EPC expressed the Igamma- and Vgamma-globin chains in a pattern characteristic of fetal development. The level of DNA methylation of the gamma-globin gene promoter in EPC cultures was similar to BM erythroid cells from normal adult baboons. Different BCL11A isoforms were observed in BM erythroid cells and cultured EPC. The mechanism responsible for increased gamma-globin expression in cultured EPC was unexpectedly not associated with increased DNA hypomethylation of the gamma-globin gene promoter compared to normal BM erythroid cells, in contrast to BM erythroid cells of decitabine-treated baboons. Rather, increased fetal hemoglobin in EPC cultures was associated with a fetal Igamma/Vgamma chain ratio and a difference in the size of the BCL11A protein compared to normal BM erythroid cells.

The Role of DNA Damage/Stress Response Pathways in the Mechanism of Action of Decitabine

The DNA methyltransferase (DNMT) inhibitor decitabine increases fetal hemoglobin (HbF) in experimental primates and to therapeutic levels in patients with sickle cell disease. Decitabine-DNMT covalent adducts formed following incorporation of decitabine into newly synthesized DNA can activate DNA damage/stress response signal transduction pathways. To test the hypothesis that DNA damage/stress response pathways are involved in the ability of decitabine to increase γ-globin expression, experiments were performed in a chemical inducer of dimerization (CID)-dependent mouse fetal bone marrow (BM)-derived cell line containing the human β-globin gene locus in the context of a yeast artificial chromosome and in primary baboon erythroid progenitors derived from CD34+ BM cells grown in Iscove's media containing 30% fetal bovine serum, 2U/ml epo, 200ng/ml SCF, and 1uM dexamethasone. Addition of decitabine (1μM) to both the CID-dependent mouse BM cell line and erythroid progenitors on day 7 of culture increased phospho-H2A.X and phospho-p38 in Western blot assays demonstrating activation of the ATM/ATR and p38 MAP kinase signal transduction pathways. Decitabine increased γ-globin 280 fold and ε-globin 38 fold in the mouse BM cell line as determined by real time PCR using the ΔΔCT method with mouse α-globin as the standard. Co-addition of the p38 MAP kinase inhibitor SB203580 reduced γ-globin induction 93.4% and ε-globin induction 88.4% in the mouse BM cell line. In primary erythroid progenitors, decitabine increased γ-globin 2.8 fold and ε-globin 44 fold. SB203580 reduced decitabine-mediated ε-globin induction 46% while no significant effect on the level of γ-globin induction was observed. Decitabine also increased phospho-CREB in the mouse BM cell line, a target of the p38 MAP kinase pathway required for γ-globin induction by histone deacetylase inhibitors. These results suggest that activation of the p38 MAP kinase pathway is required for induction of ε- and γ-globin expression by decitabine in the mouse fetal BM cell line and may contribute to increased ε-globin expression in primary erythroid progenitors. Additional Western blot analysis showed that decitabine treatment of erythroid progenitors increased p21WAF1 and p27KIP1 48–72 hours following addition of drug. Examination of Wright's stained cytospin preparations from decitabine-treated cultured erythroid progenitors showed that orthochromatic erythroblasts were increased >2 fold and basophilic erythroblasts were decreased 5 fold compared to untreated controls 96–120 hours post-drug addition at doses ranging from 0.125 to 1.0mM. These results demonstrate that decitabine increases terminal erythroid differentiation within this cell population. As RB protein is required for cell cycle exit and terminal erythroid differentiation, we suggest that induction of terminal differentiation by decitabine is mediated by the effect of increased p21WAF1 and p27KIP1 on RB phosphorylation. Because of the known asynchrony of γ- and β-globin expression during erythroid differentiation, the ability of decitabine to induce terminal erythroid differentiation could play a role in its ability to increase γ-globin expression. We conclude that activation of DNA damage/stress response pathways by decitabine may contribute to its ability to increase γ-globin expression and induce terminal erythroid differentiation.

RNAi Targeting MBD2 Increases γ-Globin Expression in a CID-Dependent Human β-YAC Murine Fetal Bone Marrow Cell Line.

Increased fetal hemoglobin (HbF) can alleviate the symptoms and increase the life span of patients with sickle cell disease. The development of new methods to increase HbF in patients has been hampered by the lack of cell line models. Recently a new murine fetal bone marrow (FBM) cell line containing the human β-globin gene locus in the context of a yeast artificial chromosome (βYAC) whose growth is dependent on the chemical inducer of dimerization (CID) AP20187 (Ariad Pharmaceuticals) was described (Blau et al, J Biol Chem 280: 36642, 2005). The DNA methyltransferase (DNMTase) inhibitor 5-azacytidine increased γ-globin gene expression in this cell line suggesting that it is a good model system for studying the mechanism of DNA methylation in γ-globin gene silencing. To investigate the role of methylated DNA binding proteins in γ-globin silencing, we transduced this cell line with retrovirus vectors expressing shRNA targeting the methylated DNA binding proteins MBD2 and MBD3. Transduced pools were selected for puromycin resistance and the effect on expression of the target gene and γ-globin determined by real time PCR using the ΔΔCT method. Single cell clones were isolated by limited dilution platings for further analysis. No difference in γ-globin expression was observed between clones derived from the parental cell line and cells transduced with a nonsilencing control vector. MBD3 expression was decreased 70–85% but γ-globin expression was not affected in the selected pool or in 12 clones following transduction wth a vector targeting MBD3. In 2 selected pools transduced with an shRNA vector targeting MBD2, MBD2 expression was decreased 60 and 82% and γ-globin expression increased 3.7 fold and 4.76 fold. Mean expression of γ-globin was 3.83 fold higher in 14 clones that showed >90% decrease in MBD2 mRNA compared to 11 clonal isolates derived from cells transduced with the non-silencing control vector (p<.05). In three clones that showed the greatest induction of γ-globin, the fold change (mean +SD) of ε-, γ-, and β-globin expression was 3.98±1.75, 11.29±4.27 and 1.90±1.34, respectively, demonstrating that MBD2 knockdown increased both ε- and γ-globin expression. The level of MBD2 protein in these clones and in two additional clones that showed >90% decrease in MBD2 mRNA but no induction of γ-globin expression were compared by Western blot analysis. MBD2 was decreased >90% in all clones compared to the parental line. Decreased MBD2 levels were therefore not associated with increased γ-globin expression in all clones. We then compared shRNA-mediated MBD2 knockdown with the DNMTase inhibitor decitabine for effects on ε- and γ-globin expression. Expression of ε-globin was increased 76.0±6.1 fold, γ-globin 131±10.2 fold, and β-globin 3.6±0.38 fold following treatment with decitabine (1 × 10−6 M; 48 hours; n=3). We conclude that decreased expression of MBD2 following transduction of the CID-dependent mouse FBM βYAC cell line with an shRNA vector targeting MBD2 is associated with increased γ-globin expression, confirming previous results of analysis of βYAC-MBD2 knockout mice (Rupon et al, PNAS 103: 6617, 2006). However, the increase in γ-globin expression was approximately 10 fold greater following decitabine treatment than was achieved with shRNA-mediated MBD2 knockdown.

Additive Effect of Butyrate and Hemin on Fetal Hemoglobin Induction in Erythroid Cells from Sickle Cell Disease and Beta-Thalassemia.

High levels of fetal hemoglobin (HbF) are known to ameliorate the pathophysiology of β-globin disorders. The objective of this study is twofold: the first is to evaluate the efficacy of hemin as an inducer of HbF in erythroid cells from patients with sickle cell disease (SCD) and β-thalassemia (β-thal); the second is to determine if the combination of butyrate and hemin can induce higher levels of expression of HbF than either agent alone. BFU-E derived cells from the peripheral blood of two patients with homozygous SCD, three patients with β-thal, one patient with sickle β-thalassemia (S/β-thal) and one normal individual (AA) were cultured in the absence (control) or presence of butyrate (B), hemin (H) or butyrate and hemin (B+H). As expected, the levels of γ-globin mRNA [expressed as % γ/(β+γ)] increased upon butyrate exposure in progenitor-derived erythroid cells from SS and S/β-thal patients, and to a lesser extent in patients with β-thal (P = 0.01). In contrast, butyrate did not increase γ-globin expression in BFU-E derived colonies from the AA individual. Moreover, hemin exposure increased the γ/(β+γ) ratio in all subjects (P = 0.02). These findings confirm that hemin can be an effective HbF inducing agent in SCD and β-thal. Although the mechanism of induction of HbF by hemin is not known, unlike butyrate, hemin is clearly not a direct inhibitor of histone deacetylases and is likely to induce HbF by a different mechanism of action. Thus, we investigated the effect of the combination of hemin and butyrate on γ-globin gene expression. Interestingly, the combination of butyrate and hemin resulted in additive increases in the γ/(β+γ) ratios in all patients compared to butyrate alone (P = 0.03) or hemin alone (P = 0.01) (Table I). Just as importantly, exposure to both drugs resulted in a decrease in the α/(β+γ) mRNA imbalance in β-thal, which is the predominant pathophysiological feature of this disorder. In conclusion, combination therapy consisting of butyrate and hemin, which are two agents with different mechanisms of action and different toxicity profiles, may provide a more effective way of inducing HbF in patients with SCD and β-thal.Table ImRNASCDβ-ThalS/β-ThalAAn2311%γ/(β+γ)Control3642267.1B4550416.9H55555215B+H60615913α/(β+γ)Control3.18.91.81.9B2.07.72.91.7H3.07.51.71.0B+H2.96.42.21.3

Human KU812 Erythroleukemia Cells: A Model for Competitive γ-Globin Induction by Fetal Hemoglobin Inducing Drugs.

The control of human hemoglobin gene switching from fetal γ-globin to adult β-globin is important in erythroid maturation and treatment approaches for sickle cell disease. Drugs that reverse the γ to β-globin ratio (γ/β) have been used effectively in clinical settings. Interaction between globin genes and the locus control region is a widely accepted mechanism for competitive γ-gene silencing. Studies in K562 cells are limited and often not correlated with in vivo response in part due to a lack of β-globin expression. Therefore, we tested a KU812 erythroleukemia cell line containing active γ- and β-globin genes, as an in vitro model for screening HbF inducers. Cell viability was monitored by trypan blue exclusion and globin mRNA measured in K562 cells (γ-globin) and KU812 cells (γ and β globin) by quantitative-PCR (q-PCR). Treatments with the histone deacetylase (HDAC) inhibitors: sodium butyrate (NaB, 2mM), trichostatin A (TSA, 0.2–0.5μM), suberoylanilide hydroxamic acid (SAHA, 2.5–5μM) or STI571 (0.5μM) were completed. In addition, hydroxyurea (HU, 100μM) was tested as an HbF inducer control.Results: KU812 cells grown in suspension in IMDM and 10% fetal bovine serum showed viability and proliferative capacity similar to K562 cells. Drug concentrations for SAHA and TSA were decreased to attain acceptable cell viability for 48hr inductions. Controls studies in K562 cells treated with NaB and TSA showed 60–80% viability and 3.2-fold increased γ-globin expression normalized by GAPD (Gγ/GAPD, See Table). For the new agents SAHA (0–5μM) and STI571 (0–2μM) dose response studies showed 62% and 66% viability in K562 cells at 50μM, and a concomitant increase in γ/GAPD mRNA from 1.7 to 40-fold respectively. For KU812 cells at steady-state we observed 40-fold higher γ vs. β globin mRNA. Inductions with NaB, TSA and SAHA produced increased γ/β ratios up to 4.5-fold along with β-gene repression; a scenario desirable in sickle cell patients. Although STI571 induced γ-globin 8-fold this effect was countered by 14-fold β-globin induction to produce a net reversal of the γ/β ratio (0.58). Control studies with HU showed on average an 1.4-fold increase in γ/β globin by 48 hrs suggesting that modest changes in this competitive ratio is sufficient to achieve significant clinical benefits. These data support KU812 cells as a good model for testing competitive γ-globin activation by drug inducers, which could not be ascertained in K562 cells. We plan to correlate:γ/β ratios,HbF protein (by ELISA; Bethyl Lab. INC., Montgomery, TX, and 3) histone H3 and H4 acetylation levels (by western blot) in K562 and KU812 cells. Preliminary ELISA data showed 2.4 to 6.4-fold increase in HbF protein by NaB and SAHA respectively in K562 cells. This data combined with the favorable γ/β globin ratio observed with SAHA suggests this agent might be an efficacious HbF inducer in vivo.Summary: The γ/β globin mRNA ratio was determined in KU812 cells to establish a better measure of drug-mediated HbF induction in vivo. We will screen novel agents for therapeutic potential. KU812 cells might also serve as a good model to study the elusive mechanisms involved in the γ to β globin switch during erythroid maturation.K562KU812HbF InducerDrug conc.γ/GAPD Fold Inc.γ/β Fold Inc.None01.01.0NaB2mM3.2*4.51*TSA0.2–0.5 μM2.11*1.612SAHA2.5–5 μM1.733.75†*STI571M μ 0.540.0*0.58*HUμ 100M1.3*1.35*1TSA 0.5μM; 2TSA 0.2μM; 3SAHA 5μM;†SAHA 2.5μM;* p<0.05

Increased γ-globin expression in a nondeletion HPFH mediated by an erythroid-specif ic DNA-binding factor

In man, a shift from gamma- to beta-globin gene expression in erythroblasts underlies a switch from fetal to adult haemoglobin during development. In hereditary persistence of fetal haemoglobin (HPFH), inappropriately high gamma-globin expression in adult life is associated with deletions in the beta-globin cluster or with single-base changes upstream of the gamma-globin genes. To account for enhanced gamma-gene expression in HPFH of the non-deletion type, we tested the nuclear proteins of human erythroleukaemia cells that bind gamma-promoter sequences in vitro by correlating specific mutations in their binding sites with promoter activity. An erythroid-specific factor (GF-1) binds as a single molecule to the -195 to -170 region and contacts two TATCT(AGATA) motifs, but not the conserved octamer (ATGCAAAT) that separates them. We observe that a single change (at -175, T----C) found in HPFH leads to increased promoter activity only in erythroid cells. This effect is mediated by GF-1, the human counterpart of the chicken erythroid factor Eryf 1. The form of HPFH we studied here is an inherited disorder which can be ascribed to the action of a cell-specific DNA-binding factor on a mutant promoter.