Definitive Erythropoiesis Research Articles

“Definitive” Erythropoiesis Has Distinct Developmental Origins and Globin Expression Patterns in the Mammalian Embryo.

Abstract 2539Poster Board II-516A transient wave of primitive erythropoiesis begins at embryonic day 7.5 (E7.5) in the mouse as yolk sac-derived primitive erythroid progenitors (EryP-CFC) generate precursors that mature in the circulation and expand in numbers until E12.5. A second wave of erythroid progenitors (BFU-E) originates in the yolk sac beginning at E8.25 that generate definitive erythroid cells in vitro. These BFU-E colonize the newly forming liver beginning at E10.5, prior to the initial appearance there of adult-repopulating hematopoietic stem cells (HSCs) between E11.5-12.5. This wave of definitive erythroid yolk sac progenitors is proposed to be the source of new blood cells required by the growing embryo after the expansion of primitive erythroid cells has ceased and before HSC-derived hematopoiesis can fulfill the erythropoietic needs of the embryo. We utilized multispectral imaging flow cytometry both to distinguish erythroid lineages and to define specific stages of erythroid precursor maturation in the mouse embryo. Consistent with this model, we found that small numbers of definitive erythrocytes first enter the embryonic circulation beginning at E11.5. All maturational stages of erythroid precursors were observed in the E11.5 liver, consistent with these first definitive erythrocytes having rapidly completed their maturation in the liver. The expression of βH1 and εy-beta globin genes is thought to be limited to primitive erythroid cells. Surprisingly, examination of globin gene expression by in situ hybridization revealed high levels of βH1-, but not εy-globin, transcripts in the parenchyma of E11.5-12.5 livers. RT-PCR analysis of globin mRNAs confirmed the expression of βH1- and adult β1-, but not εy-globin, in E11.5 liver-derived definitive (ckit+, Ter119lo) proerythroblasts sorted by flow cytometry to remove contaminating primitive (ckit-, Ter119+) erythroid cells. A similar pattern of globin gene expression was found in individual definitive erythroid colonies derived from E9.5 yolk sac and from early fetal liver. In vitro differentiation of definitive erythroid progenitors from E9.5 yolk sac revealed a maturational “switch” from βH1- and β1-globins to predominantly β1-globin. βH1-globin transcripts were not observed in proerythroblasts from bone marrow or E16.5 liver or in erythroid colonies from later fetal liver. ChIP analysis revealed that hyperacetylated domains encompass all beta globin genes in primitive erythroid cells but only the adult β1- and β2-globin genes in E16.5 liver proerythroblasts. Consistent with their unique gene expression, E11.5 liver proerythroblasts have hyperacetylated domains encompassing the βh1-, β1- and β2-, but not εy-globin genes. We also examined human globin transgene expression in mice carrying a single copy of the human beta globin locus. Because of the overlapping presence and changing proportion of primitive and definitive erythroid cells during development, we analyzed sorted cell populations whose identities were confirmed by murine globin gene expression. We confirmed that primitive erythroid cells express higher levels of γ- than ε-globin and little β-globin. E11.5 proerythroblasts and cultured E9.5 progenitors express γ- and β-, but not ε-globin. E16.5 liver proerythroblasts express β- and low levels of γ-globin, while adult marrow proerythroblasts express only β-globin transcripts. In summary, two forms of definitive erythropoiesis emerge in the murine embryo, each with distinct globin expression patterns and chromatin modifications of the β-globin locus. While both lineages predominantly express adult globins, the first, yolk sac-derived lineage uniquely expresses low levels of the embryonic βH1-globin gene as well as the human γ-globin transgene. The second definitive erythroid lineage, found in the later fetal liver and postnatal marrow, expresses only adult murine globins as well as low levels of the human γ-globin transgene only in the fetus. Our studies reveal a surprising complexity to the ontogeny of erythropoiesis. Disclosures:No relevant conflicts of interest to declare.

Functional Analysis of Transcription Factor c-Maf in Embryonic Hematopoiesis.

Abstract 3646Poster Board III-582c-Maf is one of the large Maf (musculoaponeurotic fibrosarcoma) transcription factors that belong to the AP1 super family of basic leucine zipper proteins. c-Maf has been shown to be important for the regulation of IL-4 and IL-21 gene expression in T-helper-2 cells and the formation of the lens structure. A recent report has demonstrated that the augmentation of c-Maf gene expression was observed in myeloma cell lines and bone marrow samples of multiple myeloma patients. It suggests that c-Maf plays important role in multiple myeloma pathogenesis. However, the physiological roles of c-Maf in normal hematopoiesis are largely unknown and have never been analyzed in vivo setting. In this study, we investigated that c-Maf is indispensable for hematopoesis via analyzing c-Maf deficient (c-Maf−/−) mice. In the C57BL/6J background, c-Maf−/− mice were embryonic lethal and it died between embryonic day (E) 15 and E18. c-Maf−/− embryo looked pale and had evidence of anemia. In c-Maf−/− peripheral blood at E14.5, the number of enucleated erythrocytes was decreased compared that in wild-type (WT) littermates. Flow cytometric analysis of fetal liver cells indicated that mature red blood cells were reduced in c-Maf−/− embryos. In addition, Annexin-V positive apoptotic cells were increased in c-Maf−/− fetal liver. Interestingly, there was no significant difference between WT and c-Maf−/− fetal liver cells in erythroid colony counts on the colony-forming unit assay. These results indicated that c-Maf−/− erythroid cells can develop to mature cells, and the impaired definitive erythropoiesis in the c-Maf−/− mice is due to a non-cell-autonomous effect. Immunohistochemical staining of WT fetal liver using anti-c-Maf antibody showed that c-Maf was highly expressed in macrophage but not in erythroid cells. For this reason, we hypothesized that c-Maf−/− macrophage was responsible for anemia of c-Maf−/− embryo. The erythropoietic microenvironment in fetal liver mainly consists of clusters of erythroblastic islands that made up of central macrophage and surrounding erythroblasts. The compromise of erythroblasitc islands through defect of macrophage causes anemia and embryonic lethality. To verify whether c-Maf−/− macrophage cause fail to form erythroblasitic island, we isolated the erythroblasitc island from WT and c-Maf−/− fetal liver. As a result, the number of erythroblast surrounding the macrophage in erythroblastic island was significantly reduced in c-Maf−/− mice. Moreover, in vitro reconstitution experiments showed that c-Maf−/− erythroblasts were able to form erythroblastic island with WT macrophages, while c-Maf−/− macrophages and WT erythroblasts could not form erythroblastic island. These results strongly suggest that the cause of anemia in c-Maf−/− mice is due to abnormal function of macrophages in forming erythroblastic island. Finally, to clarify the molecular mechanism in formation of erythroblastic island, we found the target gene of c-Maf in fetal liver macrophages. The DNA microarray and real time RT-PCR analyses showed that the decreased expression of vascular adhesion molecule 1 (VCAM-1), known as an adhesion molecule expressed on macrophage of erythroblastic island, was observed in c-Maf−/− fetal liver macrophage. We conclude that c-Maf is indispensable for definitive hematopoiesis and erythroblastic island formation through crucial roll in fetal liver macrophages. Disclosures:No relevant conflicts of interest to declare.

A JAK2 Interdomain Linker Relays Epo Receptor Engagement Signals to Kinase Activation

JAK2 (Janus kinase 2) is essential for cytokine receptor signaling, and several lines of evidence support a causal role of an activating JAK2 mutation in myeloproliferative disorders. JAK2 activity is autoinhibited by its pseudokinase domain in the basal state, and the inhibition is released by cytokine stimulation; how engagement of the cognate receptor triggers this release is unknown. From a functional screen for gain-of-function JAK2 mutations, we discovered 13 missense mutations, nine in the pseudokinase domain and four in the Src homology 2 (SH2)-pseudokinase domain linker. These mutations identified determinants for autoinhibition and inducible activation in JAK2. Two of the mutants, K539I and N622I, resulted in erythrocytosis in mice. Scanning mutagenesis of the SH2-pseudokinase domain linker indicated that its N-terminal part was essential for interaction of JAK2 with the Epo receptor, whereas certain mutations in the C-terminal region conferred constitutive activation. We further showed that substitutions for Glu(543)-Asp(544) in this linker or Leu(611), Arg(683), or Phe(694) in the hinge proximal region of the pseudokinase domain resulted in activated JAK2 mutants that could not be further stimulated by Epo. These results suggest that the SH2-pseudokinase domain linker acts as a switch that relays cytokine engagement to JAK2 activation by flexing the pseudokinase domain hinge.

Chicken α‐globin switching depends on autonomous silencing of the embryonic π globin gene by epigenetics mechanisms

Switching in hemoglobin gene expression is an informative paradigm for studying transcriptional regulation. Here we determined the patterns of chicken alpha-globin gene expression during development and erythroid differentiation. Previously published data suggested that the promoter regions of alpha-globin genes contain the complete information for proper developmental regulation. However, our data show a preferential trans-activation of the embryonic alpha-globin gene independent of the developmental or differentiation stage. We also found that DNA methylation and histone deacetylation play key roles in silencing the expression of the embryonic pi gene in definitive erythrocytes. However, drug-mediated reactivation of the embryonic gene during definitive erythropoiesis dramatically impaired the expression of the adult genes, suggesting gene competition or interference for enhancer elements. Our results also support a model in which the lack of open chromatin marks and localized recruitment of chicken MeCP2 contribute to autonomous gene silencing of the embryonic alpha-globin gene in a developmentally specific manner. We propose that epigenetic mechanisms are necessary for in vivo chicken alpha-globin gene switching through differential gene silencing of the embryonic alpha-globin gene in order to allow proper activation of adult alpha-globin genes.

EKLF/KLF1 Controls Cell Cycle Entry via Direct Regulation of E2f2

Differentiation of erythroid cells requires precise control over the cell cycle to regulate the balance between cell proliferation and differentiation. The zinc finger transcription factor, erythroid Krüppel-like factor (EKLF/KLF1), is essential for proper erythroid cell differentiation and regulates many erythroid genes. Here we show that loss of EKLF leads to aberrant entry into S-phase of the cell cycle during both primitive and definitive erythropoiesis. This cell cycle defect was associated with a significant reduction in the expression levels of E2f2 and E2f4, key factors necessary for the induction of S-phase gene expression and erythropoiesis. We found and validated novel intronic enhancers in both the E2f2 and E2f4 genes, which contain conserved CACC, GATA, and E-BOX elements. The E2f2 enhancer was occupied by EKLF in vivo. Furthermore, we were able to partially restore cell cycle dynamics in EKLF(-/-) fetal liver upon additional genetic depletion of Rb, establishing a genetic causal link between reduced E2f2 and the EKLF cell cycle defect. Finally, we propose direct regulation of the E2f2 enhancer is a generic mechanism by which many KLFs regulate proliferation and differentiation.

Short-chain fatty acid–mediated effects on erythropoiesis in primary definitive erythroid cells

AbstractShort-chain fatty acids (SCFAs; butyrate and propionate) up-regulate embryonic/fetal globin gene expression through unclear mechanisms. In a murine model of definitive erythropoiesis, SCFAs increased embryonic β-type globin gene expression in primary erythroid fetal liver cells (eFLCs) after 72 hours in culture, from 1.7% (± 1.2%) of total β-globin gene expression at day 0 to 4.9% (± 2.2%) in propionate and 5.4% (± 3.4%) in butyrate; this effect was greater in butyrate plus insulin/erythropoietin (BIE), at 19.5% (± 8.3%) compared with 0.1% (± 0.1%) in ins/EPO alone (P < .05). Fetal γ-globin gene expression was increased in human transgene-containing eFLCs, to 35.9% (± 7.0%) in BIE compared with 4.4% (± 4.2%) in ins/EPO only (P < .05). Embryonic globin gene expression was detectable in 11 of 15 single eFLCs treated with BIE, but in0 of 15 ins/EPO-only treated cells. Butyrate-treated [65.5% (± 9.9%)] and 77.5% (± 4.0%) propionate-treated eFLCs were highly differentiated in culture, compared with 21.5% (± 3.5%) in ins/EPO (P < .005). Importantly, signaling intermediaries, previously implicated in induced embryonic/fetal globin gene expression (STAT5, p42/44, and p38), were not differentially activated by SCFAs in eFLCs; but increased bulk histone (H3) acetylation was seen in SCFA-treated eFLCs. SCFAs induce embryonic globin gene expression in eFLCS, which are a useful short-term and physiologic primary cell model of embryonic/fetal globin gene induction during definitive erythropoiesis.

Targeted deletion of Nm23/nucleoside diphosphate kinase A and B reveals their requirement for definitive erythropoiesis in the mouse embryo

The ubiquitously expressed nucleoside diphosphate kinases (Nm23/NDPK/Awd) are a large family of multifunctional enzymes implicated in nucleic acid metabolism and in normal and abnormal development. Here, we describe the generation and characterization of NDPK A- and B-deficient (Nme1(-/-)/Nme2(-/-)) mice in which >95% of the enzyme activity is eliminated. These mice are undersized, die perinatally, and exhibit a spectrum of hematological phenotypes including severe anemia, impaired maturation of erythrocytes, and abnormal hematopoiesis in the liver and bone marrow. Flow cytometric analysis of developing Nme1(-/-)/Nme2(-/-) erythroid cells indicated that the major iron transport receptor molecule TfR1 is attenuated concomitant with a reduction of intracellular iron, suggesting that TfR1 is a downstream target of NDPKs and that reduced iron in Nme1(-/-)/Nme2(-/-) erythroblasts is inhibiting their development. We conclude that Nm23/NDPKs play critical roles in definitive erythroid development. Our novel mouse model also links erythropoiesis and nucleotide metabolism.

ＤＮＡの分解異常による関節炎

A large amount of chromosomal DNA is actively degraded during some physiological contexts such as programmed cell death and definitive erythropoiesis. DNase II is an acid DNase, and localized in lysosomes. It is required to degrade DNA of apoptotic cells after macrophages engulf dying cells, and also to degrade the nuclear DNA expelled from erythroid precursor cells. In the fetal liver and thymus of the DNase II-/- mice, there are many abnormal macrophages which carry numerous undigested DNA, and the macrophages constitutively produce IFN-β. DNase II-/- mice die in feto due to severe anemia caused by the overproduction of IFN-β.To analyze the physiological role of DNase II in adult stage, we established two mice lines : DNase II-/- IFN-IR(type-I interferon receptor)-/- mice and mice with an induced deletion of the DNase II gene. Here we show the two mice lines (referred to DNaseII deficient mice) develop a chronic polyarthritis resembling human rheumatoid arthritis (RA). Histology showed severe synovitis accompanied by pannus formation and bone erosion. A set of inflammatory cytokine such as IL-1, IL-6, and TNF-α genes was strongly activated in the affected joints of these mice. Macrophages in bone marrow and spleen accumulated undigested DNA. Both gene expression of TNF-α in bone marrow and concentration of TNF-α in serum were elevated in DNaseII deficient mice before the disease onset. The disease development was blocked by administration of neutralizing antibody against TNF-α or the loss of TNF-α gene indicating that TNF-α played an essential role for the arthritis caused by impaired DNA degradation. In contrast, Rag2 deficiency (thus lack of B cells and T cells) did not affect the onset of the disease. These results show the arthritis in DNaseII deficient mice is dependent of cytokines and innate immunity but independent of acquired immunity.These findings add an evidence to the unique concept we have proposed: "If DNA is not properly degraded, DNA causes adverse effects to organisms." We suggest an unexpected possibility that impaired DNA degradation can cause arthritis. Abnormal activation of macrophages or innate immunity may trigger the arthritis in some cases of human RA. DNase II deficient mice which show well resembling characteristics with human RA will contribute to reveal the pathology and develop beneficial therapies in RA.

Budding and direct division of adult human erythrocytes in serum-free cultures

This paper documents the budding and division process of human erythrocytes in vitro using time-lapse light microscopy of hanging-drop preparations. The erythrocytes were prepared from normal adult human peripheral blood. Red blood cells showed cytoplasmic budding, segmentation, and direct division with delayed addition of erythropoietin in serum-free Dulbecco's modified Eagle's medium/nutrient mixture F-12 Ham. These observations are thought to be useful for developing model of definitive erythropoiesis and simple expansion of human erythrocytes.

Inactivation of G-protein-coupled Receptor 48 (Gpr48/Lgr4) Impairs Definitive Erythropoiesis at Midgestation through Down-regulation of the ATF4 Signaling Pathway

G-protein-coupled receptors (GPCRs), one of the most versatile groups of cell surface receptors, can recognize specific ligands from neural, hormonal, and paracrine organs and regulate cell growth, proliferation, and differentiation. Gpr48/LGR4 is a recently identified orphan GPCR with unknown functions. To reveal the functions of Gpr48 in vivo, we generated Gpr48-/- mice and found that Gpr48-/- fetuses displayed transient anemia during midgestation and abnormal definitive erythropoiesis. The dramatic decrease of definitive erythroid precursors (Ter119pos population) in Gpr48-/- fetal liver at E13.5 was confirmed by histological analysis and blood smear assays. Real-time PCR analyses showed that in Gpr48-/- mice both adult hemoglobin alpha and beta chains were decreased while embryonic hemoglobin chains (zeta, betaH1, and epsilony) were increased, providing another evidence for the impairment of definitive erythropoiesis. Furthermore, proliferation was suppressed in Gpr48-/- fetal liver with decreased c-Myc and cyclin D1 expression, whereas apoptosis was unaffected. ATF4, a key transcription factor in erythropoiesis, was down-regulated in Gpr48-/- fetal livers during midgestation stage through the cAMP-PKA-CREB pathway, suggesting that Gpr48 regulated definitive erythropoiesis through ATF4-mediated definitive erythropoiesis.

Differential Glucose Transporter Expression during Perinatal and Postnatal Erythropoiesis

Glucose is a major source of energy for living organisms and its transport in vertebrates is a universally conserved property. Of all cell lineages, human erythrocytes express the highest level of the Glut1 glucose transporter with >200,000 molecules/cell. However, we recently reported that erythrocyte Glut1 expression is a specific trait of vitamin C-deficient mammalian species, comprising only higher primates, guinea pigs and fruit bats (Montel-Hagen et al., Cell, 2008). We now show that in all other tested mammals, including mice, rats, dogs and cows, Glut1 is in fact transiently expressed in erythrocytes during the neonatal period. This is in marked contrast with humans, where Glut1 is present at equivalently high levels on both neonatal and adult RBC. In mice, we found that Glut1 expression was not associated with primitive erythropoiesis but was highly expressed during definitive fetal erythropoiesis. Indeed, this transporter was present at significantly earlier stages of erythropoiesis in fetal spleen and liver than immediately following birth. It was therefore important to determine whether erythrocyte Glut1 expression in mice is specifically associated with fetal erythropoiesis or alternatively, is common to any physiological state where an extensive erythropoiesis is provoked. Induction of a hemolytic anemia in adult mice resulted in a massive erythropoiesis with significant increase in glucose uptake but notably, Glut1 was not detected. Rather, in these conditions as well as following birth, Glut4, an insulin-sensitive transporter previously thought to be responsible for glucose uptake in muscle and adipose tissue, was highly expressed. Following birth, the concomitant repression of Glut1 and induction of Glut4 was associated with a significantly augmented ratio of the Sp3 to Sp1 zinc-finger transcription factors. Thus, in contrast to humans, murine Glut1 is highly expressed during definitive erythropoiesis and is then downregulated at birth. Further erythroid development is characterized by the upregulation of a distinct glucose transporter, Glut4. Expression of distinct glucose transporters in nonhuman erythrocytes, regulated at the transcriptional level, therefore characterizes different states of erythroid development and differentiation.

Iron Does Not “Jiggle Free” in Mitochondria: Is Mitoferrin the Only Answer?.

The developing erythron requires tremendous amounts of iron (Fe) for the synthesis of heme for hemoproteins, such as hemoglobin, and iron-sulfur (Fe/S) clusters of proteins, which are required to catalyze redox reactions and regulate Fe uptake and storage. The uptake of Fe from transferrin (Tf) involves the binding of Tf to its cognate receptor (TfR), followed by the endocytosis of the Tf-TfR complex.1,2 In the late endosome, the release of Fe3+ from TfR is achieved by acidification of the vesicle by the v-ATPase H+-pump. Steap3 reduces the liberated Fe3+ prior to its transport out of the endosome by the DMT1 transporter. In contrast to previous conventional models for a cytosolic intermediate state, new data have emerged showing the direct interorganellar transfer of Fe from the endosome to the mitochondria.3 Although it is assumed that the exported Fe is targeted to the mitochondria for eventual incorporation into heme and Fe/S clusters, our understanding of the precise mechanism of how Fe traverses the outer and inner mitochondrial membranes remains poorly understood. Work in yeast mutants have implicated the role of solute carriers, Mrs3/4p, in mitochondrial iron homeostasis and revealed that it is the reduced form of iron, Fe2+, that is imported into the mitochondria. Subsequent studies of the zebrafish mutant, frascati, led to the discovery of mitoferrin 1 (Mfrn1, slc25a37), the vertebrate ortholog of Mrs3/4, as the major iron importer across the inner mitochondrial membrane in developing erythroblasts.4 A structurally related paralog, mitoferrin 2 (Mfrn2, slc25a28), plays the analogous role of Fe importer in non-erythroid cells. Loss-of-function studies of Mfrn1 in the mouse have confirmed its requirement in mammalian primitive and definitive erythropoiesis and its essential role in heme and Fe/S biosynthesis. Several questions remain unanswered in mitochondrial Fe metabolism: Do the two Mfrn importers account for all Fe imported into the mitochondria? How does Fe get across the outer mitochondrial membrane to reach the Mfrn importers? How does the translocated Fe in the matrix ultimately reach ferrochelatase to form heme? How is Fe translocated across the inner mitochondrial membrane?

Abcb10 Physically Interacts with Mitoferrin1 to Enhance Its Stability for Heme Synthesis in the Erythroid Mitochondria

We previously described a zebrafish mutant, frascati (frs), with hypochromic anemia and erythroid maturation arrest. Using positional cloning, we identified mitoferrin1 (mfrn1, slc25a37) as the gene disrupted in frs mutants (Shaw GC, et al. 2006 Nature 440:96–100). Mfrn1, the principle iron-importer in the mitochondria, is essential for heme and iron-sulphur (Fe/S) cluster syntheses in erythroblasts, and is required for primitive and definitive erythropoiesis. The biochemistry for Mfrn1-mediated iron acquisition into the mitochondria of developing erythroblasts, however, is poorly understood. In order to identify Mfrn1-associated proteins involved in mitochondrial iron homeostasis, we employed the strategy of in vivo epitope-tagged affinity purification and mass spectrometry (MS). A “bait protein,” Biotag-mouse Mfrn1 (BT-Mfrn1), was engineered to affinity-purify the associated proteins. A series of control experiments were first conducted to demonstrate that the BT-Mfrn1 protein properly targeted to the mitochondria and functionally complemented the anemia in frs embryos. A two-dimensional Blue Native gel followed by SDS-PAGE and western blot showed Mfrn1 forms higher-order protein complexes with interacting proteins in the mitochondrial fraction. A stable mouse erythroleukemia (MEL) clone expressing BT-Mfrn1 protein was derived as a source for protein purification. Affinity-purified BT-Mfrn1 protein complexes were analyzed by MS in triplicate runs in comparison to control MEL cells. Abcb10, a GATA-1 induced ATP-binding cassette transporter highly expressed in hematopoietic tissues and involved in heme biosynthesis (Shirihai OS, et al. 2000 EMBO J. 19:2492–2502), was found as one of the proteins which physically associated with Mfrn1 during MEL differentiation. The Abcb10:Mfrn1 interaction was confirmed by immunoprecipitation (IP) and western analysis with endogenously expressed proteins in MEL cells and proteins expressed by transient transfection in heterologous cells. Mfrn1 was previously shown to have longer protein half-life in differentiated MEL cells compared to undifferentiated cells (Paradkar PN, et al. submitted). To test our hypothesis that the Abcb10:Mfrn1 interaction may enhance Mfrn1 protein stability, a pulse-chase assay using 35S-methionine labeled cells was performed, followed by IP. Abcb10 was found to stabilize Mfrn1 protein in COS7 cells co-transfected with Abcb10 and Mfrn1 compared to control cells transfected with Mfrn1 alone. Similar results were obtained when total steady-state Mfrn1 protein was evaluated by western analysis comparing lysates from Abcb10:Mfrn1 cotransfected and Mfrn1 transfected cells. Our data suggest that Abcb10 physically interacts with Mfrn1 to enhance its stability and promote Mfrn1-dependent mitochondrial heme biogenesis.

A Novel Model of Short Chain Fatty Acid (SCFA)- Mediated up-Regulation of Embryonic/Fetal Globin Genes during Definitive Erythropoiesis.

Persistence of fetal hemoglobin can ameliorate adult beta (β)-globin gene disorders such as sickle cell disease and β-thalassemia. Short chain fatty acids (SCFAs) up-regulate embryonic/fetal globin gene expression in vitro and in vivo and have great therapeutic potential. We first studied the SCFA-responsiveness of embryonic and fetal globin gene expression during short-term culture in a murine primary cell model. Pooled erythroid fetal liver cells (eFLCs) from wild-type (wt) and human β-globin gene locus-containing mice were examined after treatment with SCFAs. Cells were cultured in basal media alone or in basal media plus propionate (5mm), butyrate (1mM), and/or insulin (2 U/mL) and erythropoietin (10 mg/mL, ‘ins/EPO‘). Murine embryonic β-type globin gene expression, ((βH1+εY)/( βH1+εY+βMAJ)x100) was markedly increased at 72 hours in SCFA-treated wt eFLCs, from 1.7±1.2% at baseline to, at 72 hours, 4.88±2.21% in propionate, 5.40±3.39% in butyrate, 19.48±8.30% in butyrate & ins/EPO, compared with ins/EPO-only-treated wt eFLCs, at 0.3±0.3% (n ≥ 3 for each, p<.05). Human fetal γ-globin gene expression was also up-regulated in human transgenic eFLCs, at 29.0±5.0% in butyrate & ins/EPO vs. 3.07±1.08 in ins/EPO only (n=2, p<.05). PCR analysis of FACS-sorted individual eFLCs after culture confirmed that both ins/EPO- and butyrate & ins/EPO-treated cells express detectable α-globin (n=15 for both, p= n.s.). However, only butyrate & ins/EPO-treated cells also express detectable embryonic βH1 globin (11/15, p<.005). Apoptosis was not increased in SCFA-treated eFLCs, with or without ins/EPO. Erythroid differentiation was analyzed by FACS quantitation of CD71 and TER119 co-immunostaining. At 72 hours, 21.5±3.5% of ins/EPO- vs. 65.5±9.9% of butyrate-only- and 77.5±4.0% of propionateonly-treated erythroid cells were highly differentiated (p<.005). Cytospins confirmed a marked increase in erythroid differentiation in SCFA-treated eFLCs.We also investigated SCFA-associated phosphorylation of signaling molecules, notably STAT5 and p38, which had been described in other models of erythropoiesis. Western blotting of protein extracts from cultured eFLCs showed phosphorylation of STAT5 in ins/EPO- and butyrate & ins/EPO-, but not in butyrate-only-, treated cells. However, p38 was constitutively phosphorylated in all of these conditions. Inhibitors of p38 kinase activity, SB 203580 (5–100 μM) or PD 169316 (5–300 μM), added to each condition throughout culture, prevented SCFA-mediated induction of embryonic globin gene expression in a concentration dependent manner. Erythroid differentiation was also blocked by p38 inhibition. Our data suggest that SCFAs augment both pancellular embryonic globin gene expression and erythroid differentiation. However, in our model, p38 phosphorylation and kinase activity are not unique to SCFA treatment in eFLCs; p38-associated erythroid differentiation, common to eFLCs in all culture conditions tested, may be an essential precursor for additional, as yet uncharacterized, molecular triggers for SCFA-mediated reactivation of embryonic/fetal globin gene expression in eFLCs.

Transcriptional Regulation of Thrombopoiesis.

Abstract Over the past two decades, a number of key transcription factors have been identified that play essential roles in megakaryocyte development. These include GATA-1, GATA-2, Friend of GATA-1 (FOG-1), Runx-1, Cbf-β, Fli-1, GABPα, TEL, NF-E2 p45, Gfi-1b, and SCL/TAL. Importantly, mutations in genes encoding several of these have been linked to human disorders of thrombopoieisis. Germline GATA-1 mutations that disrupt binding to FOG-1 cause X-linked macrothrombocytopenia and dyserythropoietic anemia. Acquired GATA-1 mutations that lead to exclusive production of a short isoform (GATA-1s) play initiating roles in Down Syndrome Transient Myeloproliferative Disorder (DS-TMD) and subsequent Acute Megakaryoblastic Leukemia (DS-AMKL). Haploinsufficiency of Runx-1 causes Familial Platelet Disorder with Propensity to Develop AML (FPD/ AML). Heterozygous loss of the Fli-1 gene leads to the macrothrombocytopenia seen in Jacobsen’s (Paris-Trousseau) syndrome. Important outstanding questions include: how these transcription factors act together to control megakaryocyte terminal maturation; how they differentially act as activators or repressors depending on gene context; how they intersect with cell signaling pathways; how they may coordinate terminal megakaryocyte maturation with spatial location within the bone marrow; how they may control cell fate decisions of bipotential erythroid/megakaryocytic progenitor cells; and whether additional key transcription factors exist. Application of proteomic approaches involving multi-protein complex purification has provided novel insights into some of these questions. We have isolated GATA-1 containing complexes from megakaryocytic cells and identified the Krüppel-type zinc finger transcription factor ZBP-89 as a novel regulator of megakaryocyte and erythroid development. Knockdown of ZBP-89 expression in zebrafish embryos and mice results in blocked early megakaryopoiesis and definitive erythropoiesis, phenocopying aspects of GATA-1- and FOG-1-deficient animals. We have also found that the focal adhesion component Kindlin-3 co-localizes to the nucleus and interacts with FOG-1, suggesting a possible link between integrin signaling and megakaryocyte transcriptional control. Runx-1 multi-protein complex purifications have led to the identification of Fli-1 as a direct binding partner. This interaction results in synergistic transcriptional activation of megakaryocyte-specific genes. Interestingly, the interaction between Runx-1 and Fli-1 occurs preferentially in cells that are differentiating, even though both proteins are expressed abundantly in undifferentiated megakaryoblastic cells. This binding event correlates with assembly of a large complex containing Runx-1/ Fli-1/GATA-1/FOG-1 based on gel filtration chromatography experiments. These factors may, therefore, act as a megakaryocyte-specific enhancesome. Key future directions are aimed at elucidating the molecular mechanisms that regulate these protein-protein interactions and how cell signaling pathways may modulate them.

Chromatin Structure and Transcription of the Human Alpha Globin Locus during Erythroid Differentation

The human alpha globin genes are controlled by DNase hypersensitive sites (HS) HS-4, HS-8, HS-10, HS-33 and HS-40 upstream of the ζ gene. Among these, HS40 functions as a strong enhancer of the alpha like genes. The alpha globin genes are situated amidst actively transcribing genes, but are transcriptionally silent in non-erythroid cells including hematopoietic progenitor cells We have undertaken an analysis of the chromatin structure of the alpha globin locus, recruitment of transcription factors, and the transcriptional activity of the locus in CD34+ hematopoietic progenitor cells and upon their differentiation into erythroid cells. Chromatin immunoprecipitation (ChIP) followed by PCR analysis of all the regulatory and structural segments of the α-globin locus were performed using antibodies against chemically modified tails of histone H3, the insulator binding factor CTCF, transcription factors such as GATA-1 and NF-E2, and Pol II. Both H3Me2K4 and H3AcK9 modifications were present at HS48 and HS33 in CD34+ cells and substantially increase when these cells are differentiated into erythroid lineage. At the HS40 region, these modifications were present at a low level in CD34+ cells and did not change during erythroid differentiation. Among the α-like gene promoters, we find these modifications at the Mu and theta gene promoters in CD34+ cells and they increase during erythropoiesis. These modifications were absent at the zeta gene promoter consistent with the inactivity of this gene during definitive erythropoiesis. Overall the dominant HS40 enhancer possesses moderate levels of H3Me2K4 and H3AcK9 modifications, and its cognate major a-globin promoter is devoid of these modifications in CD34+ cells even when these cells are differentiated into erythroid lineage. The entire α-globin locus including the HS enhancer regions and a-like gene promoters did not contain the unphosphorylated (initiation) form of Pol II recruitment in CD34+ cells. When these cells differentiated into the erythroid lineage, Pol II was recruited at the HS40 and HS48 regions and at the Mu and theta promoters. Rearrangement of the CTCF binding sites at the α-globin locus occurs during differentiation of CD34+ cells into the erythroid lineage. In CD34+ cells, as in HeLa cells, the α-globin genes are flanked by multiple CTCF binding events at the 5′ and 3′ ends of the locus. At the 5′ end of the locus, the HS40 and HS48 sequences were surrounded by four CTCF binding sites at HS33, HS46, HS55 and HS90. At the 3′ end of the locus CTCF was observed at the theta globin promoter and at the 3′ end of the theta globin gene. Upon differentiation of the CD34+ cells into the erythroid pathway, CTCF recruitment is significantly reduced at HS90 and HS46 sequences, while the sites at HS55 and HS33 show increased CTCF binding. Thus, in contrast to the CD34+ cells, the HS40 and HS48 sequences are y flanked by two CTCF recruitment sites in erythroid cells. Such a differential placement of CTCF binding sites suggests that differential interaction among CTCF sites may regulate the effects of the HS-40 enhancer. In erythroid cells, a strong HS40 enhancer formed by virtue of the recruitment of the enhancer factors can overcome blocking by the downstream flanking CTCF site and this might be mediated by specific interactions between the two flanking insulators. The CTCF binding at the 3′ end of the theta globin gene is abolished during erythropoiesis of CD34+ cells. However, the recruitment of CTCF at the theta globin promoter is unchanged suggesting that the theta globin may be insulated by the influence of the α-globin enhancer sequences. We have detected transcripts from parts of the theta and zeta genes and intergenic regions in HeLa, NB4 and 06990 lymphoblastoid cells and primary erythroid cells in culture. The transcription of the locus was localized to certain regions, suggesting that there may be unappreciated transcriptional regulatory elements within the locus.

Hereditary Persistence of Fetal Hemoglobin: Old, New and Future Mutations in the aγ-Globin Gene-Proximal Region

Hereditary persistence of fetal hemoglobin (HPFH) is a result of mutations that prevent the silencing of the g-globin genes during the adult stage of definitive erythropoiesis. Two types of HPFH are recognized, deletional HPFH and non-deletional HPFH. Mutations in the later class have been identified in the proximal promoters of the Ag- and Gg-globin genes. Individuals homozygous for sickle cell disease or certain b-thalassemia mutations, that have in addition a HPFH mutation, do not suffer the deleterious effects of these diseases. These subjects provide the natural evidence supporting the clinical effort to reactivate fetal hemoglobin as the major treatment for SCD and b-thalassemias. Thus, understanding the molecular mechanisms regulating the g-globin genes is essential for identification of points of therapeutic intervention. Although the number of point mutations causing HPFH has grown over the years, the biochemical mechanisms affected by these alterations remains elusive. In addition, it is unlikely that all potential mutations have been identified in humans. A complete catalog of all potential HPFH point mutations, coupled with knowledge of the transcriptional processes affected by them will be an invaluable step towards effectively treating these diseases. We recently identified a novel T>A HPFH mutation in a GATA site at position -566 of the Ag-globin promoter, the most distal in the promoter to date, that affects binding of a GATA-1-FOG-1-Mi2 repressor complex. Since this study utilized mutated human b-globin locus yeast artificial chromosome (b-YAC) transgenic mice, where a second copy of the Ag-globin gene was introduced near the locus control region, we produced b-YAC transgenic mice containing the -566 mutation at the normally located Ag-globin gene. These mice display a mild HPFH phenotype, an approximately 3% increase in g-globin gene expression, compared to wild-type b-YAC mice. Chromatin immunoprecipitation (ChIP) studies demonstrated that this mutation prevents GATA-1 binding when g-globin is repressed in post-conception day 18 (E18) fetal liver, whereas recruitment was observed in wild-type b-YAC transgenic samples from the same developmental stage. These data are consistent with the presence of a GATA-1-mediated repressor complex at this GATA site when g-globin is not expressed. GATA-1-mediated repression may be a general mechanism of g-globin silencing. To begin testing this hypothesis, we utilized previously generated Ag-globin -117 G>A Greek HPFH b-YAC transgenic mice, which show a 5–8% increase in g-globin synthesis in adult erythropoiesis. Published data suggested that this mutation affects nearby GATA-1 binding. Our ChIP data confirmed these results, however the GATA-1 multi-protein complex that is affected may differ from that recruited to the -566 GATA binding site. Finally, we have developed a cell-based selection that is being used to identify a comprehensive set of Ag-globin HPFH promoter mutations. Chemical inducer of dimerization (CID)-dependent Ag-globin promoter-eGFP b-YAC bone marrow cells were derived from transgenic mice and mutagenized with N-ethyl, N-nitrosourea (ENU). These cells are normally GFP−; treatment with g-globin-inducers or the presence of the -117 Greek HPFH mutation results in GFP+ cells. GFP+ cells were collected by FACS and individual cell clones expanded so that genomic DNA could be isolated. Promoter proximal regions were amplified by four PCR primer sets and subjected to heteroduplex analysis with the corresponding wild-type Ag-globin promoter PCR products as the control amplicons. Twenty three heteroduplexes have been detected among 158 mutant clones screened. Most are clustered in the proximal promoter. These data suggest that we have produced HPFH mutations, likely consisting of those known in human populations, as well as novel sites that affect repressor binding or enhance recruitment of transcriptional activators.

Altered Erythroid and Megakaryocytic Differentiation in Mice Expressing a Unique Chromatin Remodeling Domain of Erythroid Krüppel-Like Factor (EKLF)

The zinc finger-encoding transacting factor EKLF binds key regulatory elements of many erythroid-specific genes, and is essential for definitive erythropoiesis. Mice lacking this factor (EKLF−/−) die of anemia by E15.5 of gestation, failing to activate β-globin gene transcription, and demonstrating a block in the erythroid differentiation program at the primitive erythroblast stage. In contrast, megakaryocytic progenitors are amplified in EKLFnull embryos, with increased Fli-1 gene expression (a marker of early megakaryocytic differentiation), consistent with the idea that EKLF modulates the megakaryocyticerythroid (M-E) differentiation switch. We have demonstrated that an amino terminal mutant of EKLF (Δ221EKLF), is required to induce chromatin remodeling at the β-globin promoter in an EKLF-null erythroid cell line, but additional amino terminal sequences are required for initiation of β-globin gene transcription (Brown et al., 2002). To evaluate the role of this chromatin remodeling (CR) domain in erythroid and megakaryocytic differentiation in vivo, we have generated a knock-in allele of EKLFCR allele. Similar to EKLF-null embryos, mice homozygous for this mutant allele die of anemia by E15.5 of gestation. In contrast to erythroid cells lacking EKLF, EKLFCR/CR progenitors demonstrateappropriate binding of the CR encoding domain to all EKLF-regulatory sequences;a block in erythropoiesis at a more a mature stage in differentiationa chromatin architecture and histone modification pattern at erythroid-specific genes that recapitulates the events observed in EKLF+/+ erythroblasts at a similar stage of erythroid ontogeny;a failure of terminal erythroid gene transcription.Examining the role of EKLFCR in megakaryopoiesis, we observedinhibition of megakaryocytic progenitor amplification in EKLFCR/CR fetal hematopoietic cell populations when compared to EKLF-null embryos;loss of Fli-1 gene expression in EKLFCR expressing cells;binding of the EKLFCR mutant protein to the Fli-1 promoter with inhibition of gene transcription;a repressed chromatin architecture at megakaryocytic gene loci.In contrast to these results, mice homozygous for a knockin allele encoding the zinc finger DNA binding domain alone (Δ253EKLF), a region shown previously to be sufficient for chromatin remodeling in vitro, demonstrate erythroid and megakaryocytic phenotypes that resemble those observed in EKLF-null hematopoietic progenitors. Taken together, our results suggest strongly that the unique EKLFCR domain is necessary and sufficient to modulate the chromatin-specific roles of EKLF at erythroid- and megakaryocytic-specific loci in definitive hematopoietic cells in vivo.

Definitive erythropoiesis in chicken yolk sac

The first wave of erythropoiesis in amniotic animals generates all primitive erythrocytes and takes place exclusively in yolk sac mesoderm. It is less clear, however, to what extent and for how long the yolk sac contributes to the second wave of erythropoiesis which gives rise to definitive erythrocytes for later embryonic and adult use. Here, we examine the initiation, duration, and site of definitive erythrocyte formation in chicken yolk sac. We show that the earliest definitive erythrocytes are generated in yolk sac venous vessels surrounding major arteries at embryonic day (E) 4-4.5, and that mature definitive erythrocytes enter circulating at E4.5-E5. This takes place at a time when yolk sac vasculature remodels extensively to generate paired arterial/venous vessels. The yolk sac remains the predominant site for definitive erythropoiesis from E5 to E10, and continues to generate definitive erythrocytes at least until E15. Similar to primitive erythropoiesis, definitive erythropoiesis in the yolk sac is accompanied by the expression of transcriptional regulators gata1, scl, and lmo2. Furthermore, our data suggest that one main source of definitive erythropoietic cells is the pre-existing vascular endothelial cells. It remains unclear whether yolk sac derived hematopoietic progenitors that do not undergo erythropoiesis in the yolk sac may take up intraembryonic niches and contribute to erythropoietic stem cell population after hatching.

Haem‐regulated eIF2α kinase is necessary for adaptive gene expression in erythroid precursors under the stress of iron deficiency

Haem-regulated eIF2alpha kinase (HRI) is essential for the regulation of globin gene translation and the survival of erythroid precursors in iron/haem deficiency. This study found that that in iron deficiency, fetal definitive erythropoiesis is inhibited at the basophilic erythroblast stage with increased proliferation and elevated apoptosis. This hallmark of ineffective erythropoiesis is more severe in HRI deficiency. Microarray gene profiling analysis showed that HRI was required for adaptive gene expression in erythroid precursors during chronic iron deficiency. The number of genes with expression affected more than twofold increased, from 213 in iron deficiency and 73 in HRI deficiency, to 3135 in combined iron and HRI deficiencies. Many of these genes are regulated by Gata1 and Fog1. We demonstrate for the first time that Gata1 expression in developing erythroid precursors is decreased in iron deficiency, and is decreased further in combined iron and HRI deficiencies. Additionally, Fog1 expression is decreased in combined deficiencies, but not in iron or HRI deficiency alone. Our results indicate that HRI confers adaptive gene expression in developing erythroblasts during iron deficiency through maintaining Gata1/Fog1 expression.