The dynamics of globin gene expression and gene therapy vectors.

The most important level of the regulation of the beta-globin genes is by activation of all of the genes by the locus control region (LCR). Part of the developmental regulation of the locus is achieved by competition of the genes for the interaction with the LCR. Although this level of gene regulation is quantitatively of less importance than the direct repression mechanism for the early genes, it has important implications and has provided an excellent assay to probe the regulation of transcription at the single cell level. The results of these studies indicate that the LCR interacts with individual globin genes and that LCR/gene interactions are dynamic with complexes forming and dissociating continually. We conclude that transcription only appears to take place while the LCR and gene interact and that the level of transcription is determined by the frequency and duration of such interaction rather than by changes in the rate of transcription of the promoters. This mechanism has clear implications for the design of vectors for the purpose of gene therapy.

The dynamics of globin gene expression and position effects.

We have used gene competition to study the regulation of the human beta-globin locus in transgenic mice as a model system of a multigene locus. The locus is regulated by the locus control region (LCR), which is required for the expression of all the genes. Analysis of the locus at the single-cell level shows that the LCR appears to interact directly with the genes via a looping mechanism. This interaction is monogenic, and the level of transcription is determined by the frequency and stability of LCR/gene complex formation. These parameters are dependent both on the distance between the LCR and gene(s), and the concentration of transcription factors in the nucleus. Disturbance of complex formation leads to position effects, particularly when the locus is integrated in a heterochromatic environment.

The dynamics of globin gene expression and gene therapy vectors.

The most important level of regulation of the beta-globin genes is by activation of all of the genes by the Locus Control Region (LCR) and repression of the early genes by an as yet unknown factor acting on sequences flanking the genes. Superimposed on this is a mechanism in which the early genes (epsilon and gamma) suppress the late genes (delta and beta) by competition for the interaction with the LCR. Although this extra level of gene regulation is quantitatively of less importance than the direct repression mechanism, it has important implications and has provided an excellent assay system to probe the regulation of transcription at the single cell level. These studies indicate that the LCR interacts with individual globin genes and that LCR/gene interactions are dynamic with complexes forming and dissociating continually. The levels of expression of each of the genes appear to depend on: 1) the frequency of interaction which is itself dependent on the distance of the gene to the LCR, 2) the affinity of the LCR for the gene and 3) the stability of the LCR/gene complex. The latter two are dependent on the balance of transcription factors. We conclude that transcription only appears to take place while the LCR and gene interact and that the level of transcription is determined by the frequency and duration of such interaction rather than by changes in the rate of transcription of promoters.

Two phases in the establishment of HOX expression domains.

Vertebrate HOX genes are thought to play a role in the transmission of anteroposterior (A-P) positional information to the cellular precursors of embryonic axial and paraxial structures. We present an in situ analysis of the expression of two adjacent HOX genes, Hox-2.3 and Hox-2.4. We discuss the data on early gene expression in the light of what is known about cell behavior, cellular fate maps, and A-P regionalization during gastrulation. After the genes are switched on in the most posterior part of the embryo at the late primitive streak stage, the expression domains spread, in a way that does not correlate with cell lineage, toward more cranial positions within and along the streak in mesoderm and ectoderm. A second phase of HOX gene expression takes place at the early somite stage after the expression domains have reached the region around and lateral to the anterior part of the streak (the equivalent of the chicken Hensen's node). HOX gene expression during this second phase is probably clonally transmitted and would be directly associated with the acquisition of A-P positional identity by the progenitors of paraxial mesoderm and neurectoderm which are located anterolaterally to the node.

Transcription complex stability and chromatin dynamics in vivo.

Distant regulatory sequences affect transcription through long-range chromatin interactions. Visualization of transcriptional activity of genes that compete for distant elements, using the globin locus as a model, has revealed the dynamics of chromatin interactions in vivo. Multiple genes appear to be transcribed alternately rather than at the same time to generate several messenger RNAs in one cell. The regulator may stably complex with one gene at a time and switch back and forth between genes in a flip-flop mechanism.

Comparative analysis of Igf-2/H19 imprinted domain: identification of a highly conserved intergenic DNase I hypersensitive region.

The closely linked, reciprocally imprinted mouse genes, insulin-like growth factor II (Igf-2) and H19, provide a significant paradigm for studies of the mechanism of parental imprinting. Most studies have focused on regions within and proximate to the genes, but an analysis of the whole region is essential to unravel how expression of these genes is controlled. A comparative long-range analysis of a 130-kb genomic region containing both genes was therefore carried out using material from normal and chromosome 7 maternal uniparental disomic embryos. We examined DNA methylation, chromatin structure assessed by hypersensitivity to DNase I, and regions that show strong conservation among mammalian species. The critical boundaries of differential DNA methylation and DNase I hypersensitivity are apparently confined within and proximate to the Igf-2 and H19 genes. However, we have identified a novel intergenic region that is conserved in mammals, GC-rich, and unmethylated in embryos and contains major DNase I hypersensitive sites.

The role of EKLF in human beta-globin gene competition.

We have investigated the role of erythroid Kruppel-like factor (EKLF) in expression of the human beta-globin genes in compound EKLF knockout/human beta-locus transgenic mice. EKLF affects only the adult mouse beta-globin genes in homozygous knockout mice; heterozygous mice are unaffected. Here we show that EKLF knockout mice express the human epsilon and gamma-globin genes normally in embryonic red cells. However, fetal liver erythropoiesis, which is marked by a period of gamma- and beta-gene competition in which the genes are alternately transcribed, exhibits an altered ratio of gamma- to beta-gene transcription. EKLF heterozygous fetal livers display a decrease in the number of transcriptionally active beta genes with a reciprocal increase in the number of transcriptionally active gamma genes. beta-Gene transcription is absent in homozygous knockout fetuses with coincident changes in chromatin structure at the beta promoter. There is a further increase in the number of transcriptionally active gamma genes and accompanying gamma gene promoter chromatin alterations. These results indicate that EKLF plays a major role in gamma- and beta-gene competition and suggest that EKLF is important in stabilizing the interaction between the Locus Control Region and the beta-globin gene. In addition, these findings provide further evidence that developmental modulation of globin gene expression within individual cells is accomplished by altering the frequency and/or duration of transcriptional periods of a gene rather than changing the rate of transcription.

Intergenic transcription and transinduction of the human beta-globin locus.

We have identified novel nuclear transcripts in the human beta-globin locus using nuclear run-on analysis in erythroid cell lines and in situ hybridization analysis of erythroid tissue. These transcripts extend across the LCR and intergenic regions but are undetectable in nonerythroid cells. Surprisingly, transient transfection of a beta-globin gene (epsilon, gamma, or beta) induces transcription of the LCR and intergenic regions from the chromosomal beta-globin locus in nonerythroid cell lines. The beta-globin genes themselves, however, remain transcriptionally silent. Induction is dependent on transcription of the globin gene in the transfected plasmid but does not require protein expression. Using in situ hybridization analysis, we show that the plasmid colocalizes with the endogenous beta-globin locus providing insight into the mechanism of transinduction.

The human beta-globin locus control region confers an early embryonic erythroid-specific expression pattern to a basic promoter driving the bacterial lacZ gene.

The beta-globin locus control region (LCR) is contained on a 20 kb DNA fragment and is characterized by the presence of five DNaseI hypersensitive sites in erythroid cells, termed 5'HS1-5. A fully active 6.5 kb version of the LCR, called the muLCR, has been described. Expression of the beta-like globin genes is absolutely dependent on the presence of the LCR. The developmental expression pattern of the genes in the cluster is achieved through competition of the promoters for the activating function of the LCR. Transgenic mice experiments suggest that subtle changes in the transcription factor environment lead to the successive silencing of the embryonic epsilon-globin and fetal gamma-globin promoters, resulting in the almost exclusive transcription of the beta-globin gene in adult erythropoiesis. In this paper, we have asked the question whether the LCR and its individual hypersensitive sites 5'HS1-4 can activate a basic promoter in the absence of any other globin sequences. We have employed a minimal promoter derived from the mouse Hsp68 gene driving the bacterial beta-galactosidase (lacZ) gene. The results show that the muLCR and 5'HS3 direct erythroid-specific, embryonic expression of this construct, while 5'HS1, 5'HS2 and 5'HS4 are inactive at any stage of development. Expression of the muLCR and 5'HS3 transgenes is repressed during fetal stages of development. The transgenes are in an inactive chromatin conformation and the lacZ gene is not transcribed, as shown by in situ hybridization. These data are compatible with the hypothesis that the LCR requires the presence of an active promoter to adopt an open chromatin conformation and with models proposing progressive heterochromatization during embryogenesis. The results suggest that the presence of a beta-globin gene is required for LCR function as conditions become more stringent during development.

Altered DNA-binding specificity mutants of EKLF and Sp1 show that EKLF is an activator of the beta-globin locus control region in vivo.

The locus control region of the beta-globin cluster contains five DNase I hypersensitive sites (5'HS1-5) required for locus activation. 5'HS3 contains six G-rich motifs that are essential for its activity. Members of a protein family, characterized by three zinc fingers highly homologous to those found in transcription factor Sp1, interact with these motifs. Because point mutagenesis cannot distinguish between family members, it is not known which protein activates 5'HS3. We show that the function of such closely related proteins can be distinguished in vivo by matching point mutations in 5'HS3 with amino acid changes in the zinc fingers of Sp1 and EKLF. Testing their activity in transgenic mice shows that EKLF is a direct activator of 5'HS3.

Chromatin interaction mechanism of transcriptional control in vivo.

We have used a kinetic analysis to distinguish possible mechanisms of activation of transcription of the different genes in the human beta globin locus. Based on in situ studies at the single-cell level we have previously suggested a dynamic mechanism of single genes alternately interacting with the locus control region (LCR) to activate transcription. However, those steady-state experiments did not allow a direct measurement of the dynamics of the mechanism and the presence of loci with in situ primary transcript signals from two beta-like genes in cis has left open the possibility that multiple genes in the locus could initiate transcription simultaneously. Kinetic assays involving removal of a block to transcription elongation in conjunction with RNA FISH show that multiple beta gene primary transcript signals in cis represent a transition between alternating transcriptional periods of single genes, supporting a dynamic interaction mechanism.

Erythroid Krüppel-like factor (EKLF) is active in primitive and definitive erythroid cells and is required for the function of 5'HS3 of the beta-globin locus control region.

Disruption of the gene for transcription factor EKLF (erythroid Krüppel-like factor) results in fatal anaemia caused by severely reduced expression of the adult beta-globin gene, while other erythroid-specific genes, including the embryonic epsilon- and fetal gamma-globin genes, are expressed normally. Thus, EKLF is thought to be a stage-specific factor acting through the CACC box in the beta-gene promoter, even though it is already present in embryonic red cells. Here, we show that a beta-globin gene linked directly to the locus control region (LCR) is expressed at embryonic stages, and that this is only modestly reduced in EKLF-/- embryos. Thus, embryonic beta-globin expression is not intrinsically dependent on EKLF. To investigate whether EKLF functions in the locus control region, we analysed the expression of LCR-driven lacZ reporters. This shows that EKLF is not required for reporter activation by the complete LCR. However, embryonic expression of reporters driven by 5'HS3 of the LCR requires EKLF. This suggests that EKLF interacts directly with the CACC motifs in 5'HS3 and demonstrates that EKLF is also a transcriptional activator in embryonic erythropoiesis. Finally, we show that overexpression of EKLF results in an earlier switch from gamma- to beta-globin expression. Adult mice with the EKLF transgene have reduced platelet counts, suggesting that EKLF levels affect the balance between the megakaryocytic and erythroid lineages. Interestingly, the EKLF transgene rescues the lethal phenotype of EKLF null mice, setting the stage for future studies aimed at the analysis of the EKLF protein and its role in beta-globin gene activation.

Heterochromatin effects on the frequency and duration of LCR-mediated gene transcription.

Locus control regions (LCRs) are responsible for initiating and maintaining a stable tissue-specific open chromatin structure of a locus. In transgenic mice, LCRs confer high level expression on linked genes independent of position in the mouse genome. Here we show that an incomplete LCR loses this property when integrated into heterochromatic regions. Two disruption mechanisms were observed. One is classical position-effect variegation, resulting in continuous transcription in a clonal subpopulation of cells. The other is a novel mechanism resulting in intermittent gene transcription in all cells. We conclude that only a complete LCR fully overcomes heterochromatin silencing and that it controls the level of transcription by ensuring activity in all cells at all times rather than directly controlling the rate of transcription.