Activating transcription factor 4 mediates hyperglycaemia-induced endothelial inflammation and retinal vascular leakage through activation of STAT3 in a mouse model of type 1 diabetes.

AIMS/HYPOTHESIS: There is convincing evidence that endoplasmic reticulum (ER) stress is implicated in the pathogenesis of diabetes and its complications; however, the mechanisms are not fully understood. This study aimed to dissect the role and signalling pathways of activating transcription factor 4 (ATF4) in ER-stress-associated endothelial inflammation and diabetic retinopathy. METHODS: ER stress and ATF4 activity were manipulated by complementary pharmacological and genetic approaches in cultured retinal endothelial (TR-iBRB) cells. Diabetes was induced by streptozotocin in heterozygous Atf4 knockout and wild-type mice. ER stress markers, inflammatory cytokines and adhesion molecules, activation of the signal transducer and activator of transcription 3 (STAT3) pathway, and retinal vascular permeability were measured. RESULTS: High-glucose treatment resulted in rapid induction of ER stress, activation of ATF4, and increased production of inflammatory factors in TR-iBRB cells. Suppressing ER stress or inhibiting ATF4 activity markedly attenuated high-glucose-induced production of intercellular adhesion molecule 1, TNF-α and vascular endothelial growth factor. Conversely, enhancing ER stress or overexpressing Atf4 was sufficient to induce endothelial inflammation, which was, at least in part, through activation of the STAT3 pathway. Furthermore, knockdown of the Stat3 gene or inhibiting STAT3 activity restored ER homeostasis in cells exposed to high glucose and prevented ATF4 activation, suggesting that STAT3 is required for high-glucose-induced ER stress. Finally, we showed that downregulation of Atf4 significantly ameliorated retinal inflammation, STAT3 activation and vascular leakage in a mouse model of type 1 diabetes. CONCLUSIONS/INTERPRETATION: Taken together, our data reveal a pivotal role of ER stress and the ATF4/STAT3 pathway in retinal endothelial inflammation in diabetic retinopathy.

Oxygen radical inhibition of nitric oxide-dependent vascular function in sickle cell disease.

Plasma xanthine oxidase (XO) activity was defined as a source of enhanced vascular superoxide (O(2)( *-)) and hydrogen peroxide (H(2)O(2)) production in both sickle cell disease (SCD) patients and knockout-transgenic SCD mice. There was a significant increase in the plasma XO activity of SCD patients that was similarly reflected in the SCD mouse model. Western blot and enzymatic analysis of liver tissue from SCD mice revealed decreased XO content. Hematoxylin and eosin staining of liver tissue of knockout-transgenic SCD mice indicated extensive hepatocellular injury that was accompanied by increased plasma content of the liver enzyme alanine aminotransferase. Immunocytochemical and enzymatic analysis of XO in thoracic aorta and liver tissue of SCD mice showed increased vessel wall and decreased liver XO, with XO concentrated on and in vascular luminal cells. Steady-state rates of vascular O(2)( *-) production, as indicated by coelenterazine chemiluminescence, were significantly increased, and nitric oxide (( *)NO)-dependent vasorelaxation of aortic ring segments was severely impaired in SCD mice, implying oxidative inactivation of ( *)NO. Pretreatment of aortic vessels with the superoxide dismutase mimetic manganese 5,10,15,20-tetrakis(N-ethylpyridinium-2-yl)porphyrin markedly decreased O(2)( small middle dot-) levels and significantly restored acetylcholine-dependent relaxation, whereas catalase had no effect. These data reveal that episodes of intrahepatic hypoxia-reoxygenation associated with SCD can induce the release of XO into the circulation from the liver. This circulating XO can then bind avidly to vessel luminal cells and impair vascular function by creating an oxidative milieu and catalytically consuming (*)NO via O(2)( small middle dot-)-dependent mechanisms.

Mouse model of congenital polycythemia: Homologous replacement of murine gene by mutant human erythropoietin receptor gene.

Mutations causing truncations of the cytoplasmic domain of the human erythropoietin receptor (EPOR) result in a dominantly inherited disorder-primary familial congenital polycythemia. This disorder is characterized by increased numbers of erythrocytes (polycythemia) and by in vitro hypersensitivity of erythroid precursors to erythropoietin. The consequences of EPOR truncation in nonerythroid tissues are unknown. We replaced the murine EPOR gene with a wild-type human EPOR gene and a mutant human EPOR gene that we initially identified in a patient with polycythemia. This mutation leads to an EPOR truncated after the first tyrosine residue of the intracellular domain. Mice heterozygous for this mutant allele and a wild-type human EPOR allele mimicked the human disorder. Interestingly, mice that were homozygous for the mutant human allele were severely polycythemic but viable. Our results provide a model for functional studies of EPOR-triggered signaling pathways in erythropoiesis. These animals can now be used to investigate the molecular pathophysiology of this gain-of-function EPOR mutation in erythroid tissue and in those nonerythroid tissues that express EPOR.

Novel transactivation domain in erythroid Kruppel-like factor (EKLF).

Erythroid Kruppel-like Factor (EKLF) is an erythroid-specific transcription factor that plays a critical role in gamma- to beta-globin gene switching during development. To identify essential domains required for EKLF transactivation function, we cotransfected a human erythroleukemia cell line (K562) with a locus control region gamma/Luc-beta/Cat reporter and an EKLF expression vector. In this assay EKLF mediates a 500-fold induction of beta/CAT expression compared with controls. To map essential transactivation domains, progressive NH(2)-terminal and internal deletion mutants of EKLF were constructed. All EKLF mutants were expressed at wild-type levels, localized to the nucleus, and bound DNA. When mutant EKLF proteins were tested for beta/CAT activation, a novel transactivation domain was identified. This novel domain, encompassing amino acids (aa) 140-358, is sufficient for maximal beta/CAT activation. An 85-amino acid subdomain within this region (aa 140-225) is essential for its activity. Interestingly, this central transactivation subdomain is functionally redundant with the amino-terminal domain (aa 1-139). Thus, EKLF possesses at least two potent transactivation domains that appear to function in a redundant manner.

Lentiviral vector transduction of hematopoietic stem cells that mediate long-term reconstitution of lethally irradiated mice.

Lentiviral vectors efficiently transduce human CD34(+) cells that mediate long-term engraftment of nonobese diabetic/severe combined immunodeficient mice. However, hematopoiesis in these animals is abnormal. Typically, 95% of the human cells in peripheral blood are B lymphocytes. To determine whether lentiviral vectors efficiently transduce stem cells that maintain normal hematopoiesis in vivo, we isolated Sca-1(+)c-Kit(+)Lin(-) bone marrow cells from mice without 5-fluorouracil treatment, and transduced these cells in the absence of cytokine stimulation with a novel lentiviral vector containing a GFP (green flourescent protein) reporter gene. These cells were transplanted into lethally irradiated C57Bl/6 mice. In fully reconstituted animals, GFP expression was observed in 8.0% of peripheral blood mononuclear cells for 20 weeks posttransplantation. Lineage analysis demonstrated that a similar percentage (approximately 8.0%) of GFP-positive cells was detected in peripheral blood B cells, T cells, granulocytes and monocytes, bone marrow erythroid precursor cells, splenic B cells, and thymic T cells. In secondary transplant recipients, up to 20% of some lineages expressed GFP. Our results suggest that quiescent, hematopoietic stem cells are efficiently transduced by lentiviral vectors without impairing self-renewal and normal lineage specification in vivo. Efficient gene delivery into murine stem cells with lentiviral vectors will allow direct tests of genetic therapies in mouse models of hematopoietic diseases such as sickle cell anemia and thalassemia, in which corrected cells may have a selective survival advantage.

Human gamma-globin gene promoter element regulates human beta-globin gene developmental specificity.

The persistence of fetal hemoglobin in many patients with deletion type beta thalassemias and the expression patterns of human globin genes in transgenic mice suggest that gamma- to beta-globin gene switching results primarily from competition of gamma- and beta-globin genes for interaction with the beta-globin locus control region (LCR). To define regulatory sequences that are essential for the competitive advantage of the gamma gene at early developmental stages, stable transgenic mouse lines were produced with LCR gamma-beta constructs containing deletions of gamma 5'-flanking DNA. All constructs contained the full 22 kb LCR, a 4.1 kb beta-globin gene and a gamma-globin gene with 1348, 383, 202, 130, 72 or 52 bp of 5'-flanking sequence. Primer extension analysis of yolk sac, fetal liver and blood RNA from these lines demonstrated that a region between -202 and -130 of the human gamma-globin gene promoter was required to suppress beta-globin gene expression at early developmental stages. Four transcription factor binding sites within this region [GATA(p), Oct1, GATA(d) and CACCC] were mutated independently in LCR gamma-beta constructs and transgenic mouse lines were produced. Only the gamma CACCC box mutation resulted in high levels of beta-globin gene expression in early embryos. These results demonstrate that the CACCC box of the human gamma-globin gene plays a critical role in human beta-globin gene developmental specificity. The data also suggest that gamma CACCC box binding factors mediate LCR-gamma interactions which normally enhance gamma-globin and suppress beta-globin gene expression in fetal erythroid cells.

Molecular mechanism for silencing virally transduced genes involves histone deacetylation and chromatin condensation.

Virally transduced genes are often silenced after integration into the host genome. Chromatin immunoprecipitation and nuclease sensitivity experiments now demonstrate that silencing of the transgene is characterized by deacetylation of histone H4 lysines and chromatin condensation. Trichostatin A treatment results in dramatic reactivation of gene expression that is preceded by histone acetylation and chromatin decondensation. Analysis of individual histone H4 lysines demonstrate that chromatin domain opening is coincident with rapid acetylation of histone H4 K5, K12, and K16 and that maintenance of the open domain is correlated with acetylation of histone H4 K8. Removal of trichostatin A results in rapid deacetylation of histone H4 K8, chromatin condensation, and transcription silencing. The results suggest that deacetylation of histone H4 lysines and coincident chromatin condensation are critically involved in the silencing of virally transduced genes.

Trichostatin A up-regulates human papillomavirus type 11 upstream regulatory region-E6 promoter activity in undifferentiated primary human keratinocytes.

Human papillomavirus (HPV) gene expression in squamous epithelia is differentiation dependent in benign patient lesions and in organotypic raft cultures of primary human keratinocytes (PHKs). Using the lacZ reporter in raft cultures, we previously showed that this transcriptional regulation of the HPV type 11 (HPV-11) enhancer-promoter located in the upstream regulatory region (URR) appears to have resulted from coordination between the transcription transactivators AP1, Oct1, and Sp1 in differentiated upper strata and the repressor C/EBP in proliferating basal cells. We report here that trichostatin A, a specific inhibitor of histone deacetylase, dramatically stimulated reporter gene activity from the wild-type HPV-11 URR or the C/EBP mutation in PHKs grown in undifferentiated submerged cultures. In epithelial raft cultures, up-regulation occurred predominantly in basal and parabasal strata; this effect was promoter specific, as expression of the lacZ reporter gene driven by the murine leukemia virus long terminal repeat (LTR), the keratin 14 promoter, or the involucrin promoter was not altered, nor was expression of endogenous keratin 10 and profilaggrin affected. However, the responses were not cell type or species specific, as identical results were observed for both HPV-11 URR-lacZ and LTR-lacZ in murine retrovirus producer cell lines of fibroblast origin.

Stochastic, stage-specific mechanisms account for the variegation of a human globin transgene.

The random insertion of transgenes into the genomic DNA of mice usually leads to widely variable levels of expression in individual founder lines. To study the mechanisms that cause variegation, we designed a transgene that we expected to variegate, which consisted of a beta-globin locus control region 5' HS-2 linked in tandem to a tagged human beta-globin gene (into which a Lac-Z cassette had been inserted). All tested founder lines exhibited red blood cell-specific expression, but levels of expression varied >1000-fold from the lowest to the highest expressing line. Most of the variation in levels of expression appeared to reflect differences in the percentage of cells in the peripheral blood that expressed the transgene, which ranged from 0.3% in the lowest expressing line to 88% in the highest; the level of transgene expression per cell varied no more than 10-fold from the lowest to the highest expressing line. These differences in expression levels could not be explained by the location of transgene integration, by an effect of beta-galactosidase on red blood cell survival, by the half life of the beta-galactosidase enzyme or by the age of the animals. The progeny of all early erythroid progenitors (BFU-E colony-forming cells) exhibited the same propensity to variegate in methylcellulose-based cultures, suggesting that the decision to variegate occurs after the BFU-E stage of erythroid differentiation. Collectively, these data suggest that variegation in levels of transgene expression are due to local, integration site-dependent phenomena that alter the probability that a transgene will be expressed in an appropriate cell; however, these local effects have a minimal impact on the transgene's activity in the cells that initiate transcription.

Rapid genotyping of mice with hemoglobinopathies and globin transgenes.

The hematology of the laboratory mouse has been well characterized. Normal genetic differences at the alpha- and beta-globin gene loci serve as useful markers for a wide variety of types of experimental studies. There are a number of naturally occurring or induced mutations that disrupt globin expression and produce thalassemic phenotypes. In addition, much has been learned of the workings of the globin locus control region from studies of transgenic mice, including those with mutations induced by targeted site-specific modifications. After a new mutation or transgene has been created, it must be maintained in living mice, and the genotypes of the offspring must be ascertained. While it is possible to determine genotypes by DNA analyses, such assays are time consuming and relatively expensive. An osmotic challenge test--originally developed for the genotyping of large-deletion alpha-thalassemia mutations in mice--has proven useful in detecting both severe and milder alpha- and beta-thalassemias, as well as some transgenic genotypes in mice carrying human globin genes. Reliable genotyping can, in some cases, be completed within a few minutes with minimal expense. Quantification of red cell fragility for a variety of thalassemic and transgenic mice is described here, along with a simplified test suitable for rapid, routine genotyping. The osmotic challenge test is perfectly reliable for distinguishing genotypes that cause significantly decreased release of hemoglobin from the red cells, but it is also useful for some of the conditions in which overall erythrocyte osmotic fragility is essentially normal.

Knockout-transgenic mouse model of sickle cell disease.

When transgenic mice that expressed human sickle hemoglobin were mated with mice having knockout mutations of the mouse alpha- and beta-globin genes, animals were produced that synthesized only human hemoglobin in adult red blood cells. Similar to many human patients with sickle cell disease, the mice developed a severe hemolytic anemia and extensive organ pathology. Numerous sickled erythrocytes were observed in peripheral blood. Although chronically anemic, most animals survived for 2 to 9 months and were fertile. Drug and genetic therapies can now be tested in this mouse model of sickle cell disease.

Reactivation of silenced, virally transduced genes by inhibitors of histone deacetylase.

Retroviral and adeno-associated viral sequences can dramatically silence transgene expression in mice. We now report that this repression also occurs in stably infected HeLa cells when the cells are grown without selection. Expression of a transduced lacZ gene (rAAV/CMVlacZ) is silenced in greater than 90% of cells after 60 days in culture. Surprisingly, high-level expression can be reactivated by treating the cells with sodium butyrate or trichostatin A but not with 5-azacytidine. When cell clones with integrated copies of rAAV/CMVlacZ were isolated, lacZ expression was silenced in 80% of the clones; however, lacZ expression was reactivated in all of the silenced clones by treatment with butyrate or trichostatin A. The two drugs also reactivated a silenced globin gene construct (rAAV/HS2alphabetaAS3) in stably infected K562 cells. Trichostatin A is a specific inhibitor of histone deacetylase; therefore, we propose that hyperacetylation of histones after drug treatment changes the structure of chromatin on integrated viral sequences and relieves repression of transduced genes. The reactivation of silenced, transduced genes has implications for gene therapy. Efficient viral gene transfer followed by drug treatment to relieve suppression may provide a powerful combination for treatment of various genetic and infectious diseases.

The bZIP transcription factor LCR-F1 is essential for mesoderm formation in mouse development.

LCR-F1 is a mammalian bZIP transcription factor containing a basic amino acid domain highly homologous to a domain in the Drosophila Cap 'N' Collar and Caenorhabditis elegans SKN-1 proteins. LCR-F1 binds to AP1-like sequences in the human beta-globin locus control region and activates high-level expression of beta-globin genes. To assess the role of LCR-F1 in mammalian development, the mouse Lcrf1 gene was deleted in embryonic stem (ES) cells, and mice derived from these cells were mated to produce Lcrf1 null animals. Homozygous mutant embryos progressed normally to the late egg cylinder stage at approximately 6.5 days post coitus (dpc), but development was arrested before 7.5 dpc. Lcrf1 mutant embryos failed to form a primitive streak and lacked detectable mesoderm. These results demonstrate that LCR-F1 is essential for gastrulation in the mouse and suggest that this transcription factor controls expression of genes critical for the earliest events in mesoderm formation. Interestingly, Lcrf1 null ES cells injected into wild-type blastocysts contributed to all mesodermally derived tissues examined, including erythroid cells producing hemoglobin. These results demonstrate that the Lcrf1 mutation is not cell autonomous and suggest that LCR-F1 regulates expression of signaling molecules essential for gastrulation. The synthesis of normal hemoglobin levels in erythroid cells of chimeras derived from Lcrf1 null cells suggests that LCR-F1 is not essential for globin gene expression. LCR-F1 and the related bZIP transcription factors NF-E2 p45 and NRF2 must compensate for each other in globin gene regulation.

Activation of delta-globin gene expression by erythroid Krupple-like factor: a potential approach for gene therapy of sickle cell disease.

Hemoglobin A2 (HbA2; alpha 2 delta 2) is a powerful inhibitor of HbS (alpha 2 beta 2(3)) polymerization. However, HbA2 levels are normally low in sickle cell patients. We show that a major reason for low delta-globin gene expression is the defective CACCC box at -90 in the delta-globin promoter. When the CACCC box defect in delta is corrected, expression of an HS2 delta /Luciferase reporter is equivalent to HS2 beta /Luciferase. Erythroid Krupple-like factor (EKLF), which binds to the CACCC box of the beta-globin gene and activates high-level expression, does not bind to the normal delta-globin promoter. Our goal is to design a modified EKLF that binds to the defective delta-globin promoter and enhances delta-globin gene expression. To test the feasibility of this strategy, we inserted the beta-globin CACCC box at -90 of the delta-globin gene promoter to produce an HS2 delta CAC-beta construct and quantitated human delta- and beta-globin mRNA in stably transformed murine erythroleukemia (MEL) cells. delta- Globin mRNA in these cells was 22.0% +/- 9.0% of total human globin mRNA (delta/delta + beta) as compared with 3.0% +/- 1.3% in the HS2 delta-beta control. In a second set of experiments a GAL4 DNA-binding site was inserted at -90 of the delta-globin gene to produce an HS2 delta GAL4-beta construct. This construct and a GAL4(1-147)/EKLF expression vector were stably transfected into MEL cells. delta-Globin mRNA in these cells was 27.8% +/- 7.1% of total human globin mRNA as compared with 9.9% +/- 2.5% in the HS2 delta GAL4-beta plus GAL4(1-147) control. These results show that delta-globin gene expression can be significantly increased by a modified EKLF. Based on these results, we suggest that modified EKLFs, which contain zinc fingers designed to bind specifically to the defective delta-globin CACCC box, may be useful in gene therapy approaches to increase HbA2 levels and inhibit HbS polymerization.

Stable transduction of recombinant adeno-associated virus into hematopoietic stem cells from normal and sickle cell patients.

Stable introduction of genes into human hematopoietic stem cells with self-renewing potential is a necessary requirement for gene therapy strategies. We have developed an adeno-associated virus (AAV) vector and a partial packaging cell line that produces recombinant AAV at a titer of 10(8) transducing particles per milliliter. A high-titer viral stock containing the CMV/lacZ gene was used to transfer lacZ sequences into CD34+ Lin-Thy+ hematopoietic stem cells purified from normal and homozygous sickle cell patients. After infection, the cells were cultured in two ways. In the first set of experiments, the cell were expanded 300-fold in liquid culture for 21 days and plated in methylcellulose. Burst-forming units-erythroid (BFU-E) and colony-forming units-granulocyte/macrophage (CFU-GM) were then analyzed for lacZ sequences. In the second set of experiments, infected cells were cultured for 6 weeks under conditions that maintain long-term culture-initiating cells (LTC-IC). Progenitors were plated in methylcellulose, and BFU-E were analyzed for lacZ DNA. Stable transduction of lacZ sequences was observed in 25% of the colonies in both sets of experiments. These results demonstrate for the first time that LTC-IC can be transduced stably with a recombinant AAV vector. The results suggest that AAV may be a useful vector for genetic therapy of sickle cell disease and other hematopoietic disorders.

End joining of genomic DNA and transgene DNA in fertilized mouse eggs.

A linear 5.2-kb HS2/beta-globin construct with an upstream KpnI terminus (4-nucleotide (nt) 3' protruding single strand, PSS) and a downstream SalI terminus (4-nt 5' PSS) was microinjected into fertilized mouse eggs. The injected DNA fragments integrated into the mouse genome primarily as a head-to-tail tandem array. Chromosome/transgene junctions were obtained from seven of eight transgenic animals. All of the junctions occurred in the proximity of a transgene KpnI end; a maximum loss of 8 nt from the transgene terminus was observed. Two of these junctions completely preserved the 4-nt KpnI 3' PSS. Transgene/transgene junctions from two animals were analyzed. SalI/KpnI junctions that completely preserved both the SalI 5' PSS and the KpnI 3' PSS were found in each animal. These are the first examples of complete nt preservation at junctions formed between a 5' PSS terminus and a 3' PSS terminus in transgenic mice. The data are consistent with the fill-in model of Thode et al. [Cell 60 (1990) 921-928] in which alignment proteins juxtapose 5' PSS and 3' PSS termini; DNA polymerase then utilizes the recessed 3'-OH of the 5' PSS terminus as a primer to synthesize DNA across the gap. This mechanism results in the formation of junctions with no loss of sequence. The results described in the present paper suggest that this mechanism may be involved in the formation of junctions in transgenic mice.

Mouse model of human beta zero thalassemia: targeted deletion of the mouse beta maj- and beta min-globin genes in embryonic stem cells.

beta zero-Thalassemia is an inherited disorder characterized by the absence of beta-globin polypeptides derived from the affected allele. The molecular basis for this deficiency is a mutation of the adult beta-globin structural gene or cis regulatory elements that control beta-globin gene expression. A mouse model of this disease would enable the testing of therapeutic regimens designed to correct the defect. Here we report a 16-kb deletion that includes both adult beta-like globin genes, beta maj and beta min, in mouse embryonic stem cells. Heterozygous animals derived from the targeted cells are severely anemic with dramatically reduced hemoglobin levels, abnormal red cell morphology, splenomegaly, and markedly increased reticulocyte counts. Homozygous animals die in utero; however, heterozygous mice are fertile and transmit the deleted allele to progeny. The anemic phenotype is completely rescued in progeny derived from mating beta zero-thalassemic animals with transgenic mice expressing high levels of human hemoglobin A. The beta zero-thalassemic mice can be used to test genetic therapies for beta zero-thalassemia and can be bred with transgenic mice expressing high levels of human hemoglobin HbS to produce an improved mouse model of sickle cell disease.

Autonomous, erythroid-specific DNase I hypersensitive site formed by human beta-globin locus control region (LCR) 5' HS 2 in transgenic mice.

The human beta-globin locus control region (LCR) is composed of four erythroid-specific, DNase I hypersensitive (HS) sites that are located 6 to 18 kb upstream of the epsilon-globin gene. The beta-globin LCR appears to have two major functions. First, the sequences "open" a chromosomal domain that includes the epsilon-, gamma-, and beta-globin genes and, second, the LCR directs high-level, erythroid-specific expression of each globin gene family member. An LCR subfragment containing only 5' HS 2 can confer these properties on a linked beta-globin gene. To determine whether 5' HS 2 can form an erythroid-specific, DNase I hypersensitive site in the absence of a linked globin gene, a 1.9-kb DNA fragment containing this site was injected into fertilized mouse eggs and DNase I hypersensitivity was analyzed in the animals that developed. In 9 of 10 transgenic mouse lines, the human 5' HS 2 fragment formed a DNase I hypersensitive site in fetal liver but not in fetal brain. These results suggest that human 5' HS 2 can function autonomously to organize an open chromatin domain specifically in erythroid cells.

Role of erythroid Kruppel-like factor in human gamma- to beta-globin gene switching.

Erythroid Kruppel-like factor (EKLF) is an erythroid-specific transcription factor that contains zinc finger domains similar to the Kruppel protein of Drosophila melanogaster. Previous studies demonstrated that EKLF binds to the CACCC box in the human beta-globin gene promoter and activates transcription. CACCC box mutations that cause severe beta-thalassemias in humans inhibit EKLF binding. Results described in this paper suggest that EKLF functions predominately in adult erythroid tissue. The EKLF gene is expressed at a 3-fold higher level in adult erythroid tissue than in fetal erythroid tissue, and the EKLF protein binds to the human beta-globin promoter 8-fold more efficiently than to the human gamma-globin promoter. Co-transfection experiments in the human fetal-like erythroleukemia cell line K562 demonstrate that over-expression of EKLF activates a beta-globin reporter construct 1000-fold; a linked gamma-globin reporter is activated only 3-fold. Mutation of the beta-globin CACCC box severely inhibits activation. These results demonstrate that EKLF is a developmental stage-enriched protein that preferentially activates human beta-globin gene expression. The data strongly suggest that EKLF is an important factor involved in human gamma- to beta-globin gene switching.

Retroviral vector sequences inhibit human beta-globin gene expression in transgenic mice.

The DNase I hypersensitive site 5' HS2 of the human beta-globin locus control region confers position-independent, high-level expression on the human beta-globin gene in transgenic mice. When a 5' HS2 beta-globin construct is flanked by retroviral vector sequences derived from Moloney Murine Leukemia Virus (MoMLV), expression of the beta-globin gene is severely inhibited. Apparently, one or more elements within the MoMLV genome is capable of repressing transcription of the human beta-globin gene in transgenic mice. A construct lacking the retroviral enhancer also fails to express the beta-globin gene, indicating that this region of the virus is not essential for repression. Further analysis may permit the identification of specific viral sequences that inhibit gene expression; these sequences could then be deleted or mutated to produce improved viral vectors.