Search

Sox6 belongs to the Sry (sex-determining region Y)-related high-mobility-group-box family of transcription factors, which control cell-fate specification of many cell types. Here, we explored the role of Sox6 in human erythropoiesis by its overexpression both in the erythroleukemic K562 cell line and in primary erythroid cultures from human cord blood CD34+ cells. Sox6 induced significant erythroid differentiation in both models. K562 cells underwent hemoglobinization and, despite their leukemic origin, died within 9 days after transduction; primary erythroid cultures accelerated their kinetics of erythroid maturation and increased the number of cells that reached the final enucleation step. Searching for direct Sox6 targets, we found SOCS3 (suppressor of cytokine signaling-3), a known mediator of cytokine response. Sox6 was bound in vitro and in vivo to an evolutionarily conserved regulatory SOCS3 element, which induced transcriptional activation. SOCS3 overexpression in K562 cells and in primary erythroid cells recapitulated the growth inhibition induced by Sox6, which demonstrates that SOCS3 is a relevant Sox6 effector.

The Kit (White) gene encodes the transmembrane receptor of stem cell factor/Kit ligand (KL) and is essential for the normal development/maintenance of pluripotent primordial germ cells (PGCs), hematopoietic stem cells (HSCs), melanoblasts, and some of their descendants. The molecular basis for the transcriptional regulation of Kit during development of these important cell types is unknown. We investigated Kit regulation in hematopoietic cells and PGCs. We identified 6 DNase I hypersensitive sites (HS1-HS6) within the promoter and first intron of the mouse Kit gene and developed mouse lines expressing transgenic green fluorescent protein (GFP) under the control of these regulatory elements. A construct driven by the Kit promoter and including all 6 HS sites is highly expressed during mouse development in Kit+ cells including PGCs and hematopoietic progenitors (erythroid blast-forming units and mixed colony-forming units). In contrast, the Kit promoter alone (comprising HS1) is sufficient to drive low-level GFP expression in PGCs, but unable to function in hematopoietic cells. Hematopoietic expression further requires the addition of the intronproximal HS2 fragment; HS2 also greatly potentiates the activity in PGCs. Thus, HS2 acts as an enhancer integrating transcriptional signals common to 2 developmentally unrelated stem cell/progenitor lineages. Optimal hematopoietic expression further requires HS3-HS6.

Targeted expression to specific tissues or cell lineages is a necessary feature of a gene therapy vector for many clinical applications, such as correction of hemoglobinopathies or thalassemias by transplantation of genetically modified hematopoietic stem cells. We developed retroviral vectors in which the constitutive viral enhancer in the U3 region of the 3' LTR is replaced by an autoregulatory enhancer of the erythroid-specific GATA-1 transcription factor gene. The replaced enhancer is propagated to the 5' LTR upon integration into the target cell genome. The modified vectors were used to transduce human hematopoietic cell lines, cord blood-derived CD34(+) stem/progenitor cells, and murine bone marrow repopulating stem cells. The expression of appropriate reporter genes (triangle upLNGFR, EGFP) was analyzed in the differentiated progeny of transduced stem cells in vitro, in liquid culture as well as in clonogenic assay, and in vivo, after bone marrow transplantation in lethally irradiated mice. The GATA-1 autoregulatory enhancer effectively restricts the expression of the LTR-driven proviral transcription unit to the erythroblastic progeny of both human progenitors and mouse-repopulating stem cells. Packaging of viral particles, integration into the target genome, and stability of the integrated provirus are not affected by the LTR modification. Enhancer replacement is therefore an effective strategy to target expression of a retroviral transgene to a specific progeny of transduced hematopoietic stem cells.

NF-Y (CBF) is a CCAAT-binding trimer that activates 25 % of eukaryotic promoters. It contains putative histone fold motifs (HFMs) and distorts DNA. By using electrophoretic mobility shift assays with the twin CCAAT boxes of the human gamma-globin promoter and several combinations of subunit mutants, we dissected some of the structural features of CCAAT-box binding. NF-YA and NF-YC Q-rich domains significantly influence bending angles quantitatively, but not qualitatively, since they do not modify DNA orientation. They are both required for co-operative interactions among NF-Y molecules: for this, a precise alignement of two CCAAT boxes, 32 bp, three turns of the helix, is essential. Unlike the wild-type (wt) protein, steric hindrance does not impede simultaneous binding of the mutant composed of the short homology domains to CCAAT boxes closer than 22 bp: the addition of 11 amino acid residues to NF-YB and 13 to NF-YC flanking the HFM, restores wt behaviour. These stretches are predicted to form H2B-like alphaC and H2A-like alphaN fourth helices. A further support to this hypothesis comes from off-rates analysis of mutant combinations: the half-life of NF-Y, which is dependent on the type of NF-YB used, is extremely shortened, when the putative alphaC is present, nearly as much as in the wt NF-YB. These data (i) provide further evidence that NF-YB-NF-YC belong to the H2B-H2A subclasses, (ii) uncover new features of Q-rich domains, and (iii) define rules for NF-Y synergy that are potentially important for the regulation of many eukaryotic promoters., (Copyright 1999 Academic Press.)

GATA-1 is a transcription factor expressed both in the hematopoietic system and in the Sertoli cells of the testis, and is essential for correct erythropoiesis. Hematopoietic and Sertoli cells transcribe GATA-1 from two different promoters: the proximal (erythroid) is active in hematopoietic cells; the distal (testis) is active in Sertoli cells. We investigated by RT-PCR the possibility that GATA-1 might be transcribed from the testis promoter also in hematopoietic cells. Testis promoter-derived transcripts are present at low levels in vivo at all stages of hematopoietic development. Purified multipotent progenitors, fractionated into populations expressing low or high levels of GATA-1, do not contain any "testis" transcripts. However, when grown in vitro, they rapidly express GATA-1 from the testis promoter in the presence of Erythropoietin (Epo) but not in that of other growth factors. This result reflects an Epo-dependent differentiation event, rather than a direct effect of Epo. Indeed, immortalized progenitor cell lines which respond to both Epo and SCF, continue to express testis-derived transcripts when switched from Epo to SCF.

Anthracycline antitumor drugs such as aclacinomycin (ACM) and doxorubicin (DOX) used in subtoxic concentrations induce erythroid differentiation of the erythroleukemic cell line K562. To elucidate the possible role of erythroid genes of the erythropoietin receptor (EpoR) and the transcription factor GATA-1 in this effect, the regulatory regions of the above genes and human epsilon- and gamma-globin and porphobilinogen deaminase (PBGD) genes were fused to the firefly luciferase gene. The resulting reporter constructs were tested in a transfection assay of the erythroleukemic cell line K562 stimulated to differentiate by treatment with the anthracycline drugs ACM and DOX or hemin (HEM). The results showed activation of the tested promoters after cell treatment with ACM, but not with DOX or HEM. In contrast to the mouse EpoR gene promoter, the activity of the human EpoR gene promoter (-659/-60) in the reporter construct was not modified by addition of the first intron sequence. In ACM-treated K562 cells, EpoR gene promoter activity completely correlated with EpoR and GATA-1 mRNA levels and the degree of erythroid maturation. In addition, ACM strongly activated the erythroid gene promoters that contain GATA binding sites. Nevertheless, less activation was also observed for the GATA-1 gene promoter (-312/-31) lacking any known GATA binding sites. Insertion of the GATA-1 gene enhancer with two canonic GATA binding sites, stimulated the ACM activation effect for EpoR and GATA-1 promoter-containing constructs. Mutation of the enhancer GATA binding sites abolished this effect. All the regulatory regions tested (except gamma-globin promoter) were completely inactive in nonerythroid COS7 cells. These data indicate that (1) two structurally different anthracycline analogues, DOX and ACM, differ in their differentiation mechanisms, and (2) ACM switches on the erythroid program of K562 cells, at least in part because of interaction with a factor(s) that recognizes the GATA binding sites in the promoter region of erythroid genes leading to their activation.

Juvenile chronic myelocytic leukemia (JCML) is a rare disorder of early childhood. Characteristic of JCML are the progressive appearance of high levels of fetal hemoglobin (HbF), reflecting a true reversion to a fetal type of erythropoiesis, and the presence of colony-forming cells able to grow in vitro spontaneously in the absence of growth factors. To better understand the relationship between the erythroid abnormalities and the leukemic process, we analyzed the expression pattern of specific genes related to erythroid differentiation--GATA-1, EPOR, alpha-globin, beta-globin, and gamma-globin genes--in JCML peripheral blood (PB) cells and in vitro-derived colonies. Northern blot analysis of PB cells from five JCML patients indicated levels of GATA-1 transcripts much higher than those usually found in other types of leukemic cells, and S1 nuclease protection assay detected significantly increased expression of gamma-globin mRNA. Reverse transcription-polymerase chain reaction (RT-PCR) analysis of single granulocyte-macrophage colony-forming unit (CFU-GM) colonies, obtained in vitro in the absence of added growth factors from four JCML patients, detected GATA-1, EPOR, and globin (alpha and gamma) transcripts in most of the colonies tested, in contrast with control CFU-GM from normal bone marrow, which were positive only for GATA-1. Single JCML colonies were tested for the presence of two different transcripts; whereas alpha- and gamma-globin genes appeared mostly coexpressed, beta-globin mRNA was detected only in a minority of the gamma-globin-positive colonies, indicating that the leukemic pattern of hemoglobin synthesis is mainly fetal. In addition, the leukemic cells occurring during blast crisis of one of our patients displayed the typical features of a stem cell leukemia (CD34+, CD19-, CD2-, myeloperoxidase-). In this sorted CD34+ population, we detected the presence of a marker chromosome, der(12)t(3;12), previously identified in bone marrow cells at diagnosis and an expression pattern superimposable to that of the JCML colonies, consistently displaying a high gamma-globin:beta-globin mRNA ratio. The expression of erythroid markers within populations of leukemic cells, both in vivo and in vitro, supports the hypothesis that abnormal JCML erythroid cells may originate from the same mutated progenitor that sustains the growth of the leukemic cells.

With erythroid differentiation, committed progenitor cells acquire the ability to respond to erythropoietin (Epo). Epo interacts with target cells through the Epo receptor (Epo-R), whose expression is tightly regulated in a lineage-specific fashion. Epo-R expression is presumed to be progressively activated or repressed as cells progress along the erythroid or the myeloid pathway, respectively. Little is known of the mechanisms that underlie the erythroid-specific expression of the Epo-R gene. GATA-1, the major known transcription factor involved in Epo-R gene regulation, is not erythroid-specific. We have studied the regulation of the expression of the Epo-R gene in two related human Epo-responsive cell lines, UT-7 and UT-7 Epo. These lines express Epo-R at high levels because of amplification of the endogenous gene, which is apparently not rearranged. Treatment for 6 to 24 hours with the tumor promoter, phorbol myristate acetate (PMA), or 24 hours of growth factor starvation (Epo or granulocyte/macrophage colony-stimulating factor [GM-CSF]) decreased or increased the levels of Epo-R mRNA, respectively. In the case of growth factor starvation, the increase (approximately equal to threefold) in the level of Epo-R mRNA correlated directly with an increase in the rate of Epo-R gene transcription as measured by run-off assay. Both increases were observed as early as 3 hours after the growth factor was withdrawn and were reversible; levels of mRNA and transcription rates returned to baseline 3 hours after the cells were reexposed to growth factors. The changes in Epo-R expression after growth factor starvation were coordinated with changes in the level of expression of GATA-1 that were detected both at the mRNA and at the gene transcription level under these conditions (suggesting that GATA-1 was responsible for this upregulation). During PMA treatment, after a transient increase in Epo-R mRNA at 1 hour, a progressive decline in the level of Epo-R mRNA was observed; the level of Epo-R mRNA decreased by 50%, and fell below the level of detection by 6 and 24 hours, respectively. This decrement was explained in part by a fourfold reduction in the rate of gene transcription as well as a reduction (measured as levels of Epo-R mRNA in the presence of actinomycin D) in mRNA stability. The changes in transcription rate occurred in the absence of changes in the level of GATA-1 binding activity.(ABSTRACT TRUNCATED AT 400 WORDS)

In this report cis-acting sequences that direct transcription of the lecithin choleterol acyl transferase (LCAT) gene were identified. To assay the promoter activity, fragments from the 5' flanking region were fused upstream to the cloramphenicol acetyl transferase gene and transfected into Hep3B and HeLa cells. The gene sequences were active in both cell lines. A minimal promoter comprising only 71 bp is still fully active and contains a TATA box, a LFAI motif and two Sp1 binding sites. The activity of the promoter was entirely dependent on the Sp1 sites.

We have mapped the cap site of the human plasminogen mRNA by primer extension and PCR techniques and found that it is located at position -161 relative to the first ATG, 97 bases upstream to the 5' end of the previously isolated cDNA clone. Seven human hepatic and non hepatic cell lines and fresh liver cells were tested for human plasminogen mRNA expression: the liver and the liver derived HepG2 cell line represent the major site of plasminogen RNA synthesis while the other cell lines (Hep3B, HeLa, IMR, 293 CaCo and SW626) show much lower levels.

During a study of Sardinian families with hereditary persistence of fetal hemoglobin (HPFH), two unrelated subjects with unusually elevated Hb F levels were identified. By selective amplification of the A gamma gene promoter and hybridization to synthetic oligonucleotides, we demonstrate that these subjects are homozygous for the -117A gamma G----A substitution that is responsible for a form of nondeletional HPFH. The hemoglobin synthetic pattern of these patients is discussed.

Gamma-globin chain synthesis has been evaluated in individual bursts and subcolonies that were generated by normal adult blood BFU-Es in methylcellulose cultures containing semipurified erythropoietin (Ep) and then analyzed via either isoelectric focusing (IEF) of globin chains or immunofluorescence techniques. At variance with previously reported results, based on plasma clot culture and immunofluorescence, all bursts and subcolonies analyzed synthesize gamma-globin chains. Identification of gamma-chains has been confirmed by preparative IEF of HbF, followed by either carboxymethylcellulose chromatography or IEF analysis of the resulting globin chains. In all bursts analyzed, the relative synthesis of the two types of gamma-globin chains (G gamma and A gamma) shows an adult ratio (i.e., approximately 1:1). The results obtained via IEF have been confirmed by immunofluorescence studies, which apparently showed presence of at least some HbF-positive cells within all scrutinized bursts or subcolonies. The significance of these studies is discussed in the light of current hypotheses on mechanism(s) underlying HbF synthesis in normal adults.

Using in vitro assays, we show that nuclear proteins related to the Sp1 and GT-1 factors bind to a CACCC box sequence in the human beta-globin enhancer, adjacent to binding sites for the erythroid-specific factor NFE1 and the ubiquitous factor CP1. The same proteins are known to bind to the proximal, but not to the distal, CACCC, box in the human beta-globin promoter. A C G mutation in the promoter CACCC box, known to cause beta-thalassemia, greatly decreases protein binding to the CACCC box; the same effect is obtained when this mutation is introduced into the enhancer CACCC box.

A survey of hemoglobinopathies in Northern Sardinia allowed the identification of two subjects heterozygous for a new type of G gamma hereditary persistence of fetal hemoglobin (HPFH). The G gamma-globin gene from the HPFH chromosome shows the presence of a T----C substitution 175 nucleotides upstream of the CAP site, adding a new example of single-point mutations occurring in the promoter region of the gamma-globin genes and linked to HPFH phenotypes. In this case the mutation affects the 3' end nucleotide of a conserved octamer sequence known to be present in other regulatory elements of several genes.

We report that bromodeoxyuridine (BrdU) addition in semi-solid cultures of normal adult erythroid progenitors causes a sharp rise of gamma-globin gene expression in erythroid colonies. Control studies were carefully carried out to exclude the possibility of toxic effects exerted by the drug in these experimental conditions. In particular, BrdU addition induces a sharp increase in the level of relative gamma-globin synthesis and content in pooled BFU-E-derived colonies: this rise is clearly observed in single bursts of the mature type (largely composed of late erythroblasts) but not in immature ones (essentially comprising early erythroblasts). Furthermore, it is associated with an increase of the G gamma/G gamma + A gamma synthetic ratio from adult up to fetal like values. Reactivation of gamma-synthesis was observed even if BrdU was added to colonies composed essentially of early erythroblasts, ie, when BrdU was added to either bursts at day 10 of culture or late CFU-E-derived clones at day 1. These in vitro observations indicate modulation of gamma-synthesis at the stage of erythroblasts from normal adults. At the molecular level we suggest that BrdU, by replacing thymidine in DNA, may inhibit the switch from a fetal-like biosynthetic program expressed in early erythroblastic differentiation to the adult program expressed in later stages of maturation.

Restriction enzymes analysis of the DNA from two unrelated Italian families with Hb Hasharon, a variant Hb (alpha 47asp replaced by his) frequently occurring in the Polesine area in Italy, indicates that this variant is associated to an alpha globin gene deletion. The alpha Hash genotype most likely results from a mutation on an alpha thal2 genotype.

Spanish delta beta zero-thalassemia, a mild thalassemic condition characterized by increased level of hemoglobin (Hb) F production during adult life, is known to be due to a large deletion starting within the beta globin gene cluster and extending beyond the 3' breakpoint of any other similar deletional defects so far identified. By molecular cloning and by genomic mapping we now demonstrate that the deletion of Spanish delta beta zero-thalassemia ends at approximately 11 and 17 kilobases (kb) downstream to the 3' endpoints of black hereditary persistence of fetal hemoglobin (HPFH) type 1 and 2, respectively. As suggested by the complete characterization of this and other deletional defects involving the beta globin gene cluster, the 5' and 3' breakpoints of several deletions cluster in rather restricted DNA areas, further strengthening the idea that common molecular mechanisms may operate in causing these deletions.

Selective overexpression (50- to 100-fold) in adult erythroid cells of either G gamma or A gamma fetal globin gene is observed in hereditary conditions known as delta beta zero-thalassemia and hereditary persistence of fetal hemoglobin (HPFH). Recently, a C----T change at position -196 of an overexpressed A gamma globin gene from an Italian HPFH was hypothesized, on the basis of indirect evidence, to represent the cause of the functional defect. We now show that the same mutation is present in a different overexpressed A gamma-globin gene from a Sardinian patient with a different syndrome (delta beta zero-thalassemia). The Sardinian A gamma globin gene differs from both the HPFH and the normal A gamma globin gene at nucleotide 1,560 in the noncoding portion of the third exon, where an A is deleted. In addition, the mutant -196 A gamma-globin gene is linked to a normal beta globin gene in HPFH, and to a beta-thalassemic gene (beta 39CAG----TAG) in delta beta zero-thalassemia. These data strengthen the suggestion that -196 mutation is causally linked to the abnormal phenotype and raise the question of whether the same or multiple mutational events are responsible for the appearance of the -196 mutation in different syndromes.