The "beta-like-globin" gene domain in human erythroid cells

A fine-structure deletion map of human chromosome 11p: analysis of J1 series hybrids

Deletion analysis offers a powerful alternative to linkage and karyotypic approaches for human chromosome mapping. A panel of deletion hybrids has been derived by mutagenizing J1, a hamster cell line that stably retains chromosome 11 as its only human DNA, and selecting for loss of MIC1, a surface antigen encoded by a gene in band 11p13. A unique, self-consistent map was constructed by analyzing the pattern of marker segregation in 22 derivative cells lines; these carry overlapping deletions of 11p13, but selectively retain a segment near the 11p telomere. The map orders 35 breakpoints and 36 genetic markers, including 3 antigens, 2 isozymes, 12 cloned genes, and 19 anonymous DNA probes. The deletions span the entire short arm, dividing it into more than 20 segments and define a set of reagents that can be used to rapidly locate any newly identified marker on 11p, with greatest resolution in the region surrounding MIC1. The approach we demonstrate can be applied to map any mammalian chromosome. To test the gene order, we examined somatic cell hybrids from five patients, whose reciprocal translocations bisect band 11p13; these include two translocations associated with familial aniridia and two with acute T-cell leukemia. In each patient, the markers segregate in telomeric and centromeric groups as predicted by the deletion map. These data locate the aniridia gene (AN2) and a recurrent T-cell leukemia breakpoint (TCL2) in the marker sequence, on opposite sides of MIC1. To provide additional support, we have characterized the dosage of DNA markers in a patient with Beckwith-Wiedemann syndrome and an 11p15-11pter duplication. Our findings suggest the following gene order: TEL - (HRAS1, MER2, CTSD, TH/INS/IGF2, H19, D11S32) - (RRM1, D11S1, D11S25, D11S26) - D11S12 - (HBBC, D11S30) - D11S20 - (PTH, CALC) - (LDHA, SAA, TRPH, D11S18, D11S21) - D11S31 - D11S17 - HBVS1 - (FSHB, D11S16) - AN2 - MIC1 - TCL2 - delta J - CAT - MIC4 - D11S9 - D11S14 - ACP2 - (D11S33, 14L) - CEN. We have used the deletion map to show the distribution on 11p of two centromeric repetitive elements and the low-order interspersed repeat A36Fc. Finally, we provide evidence for an allelic segregation event in the hamster genome that underlies the stability of chromosome 11 in J1. The deletion map provides a basis to position hereditary disease loci on 11p, to distinguish the pattern of recessive mutations in different forms of cancer and, since many of these genes have been mapped in other mammalian species, to study the evolution of a conserved syntenic group.

Gamma delta beta-thalassemia due to a de novo mutation deleting the 5' beta-globin gene activation-region hypersensitive sites

gamma delta beta-Thalassemia is a rare disorder of hemoglobin biosynthesis, characterized molecularly by partial or complete deletions of the beta-globin gene complex of 100 kilobases (kb) or greater. Common to all mutants described has been the deletion of the most-5' sequences of the beta-globin complex. We have used the techniques of pulsed-field gel electrophoresis and polymerase chain reaction to study a patient with a clinical gamma delta beta-thalassemia phenotype. This subject developed a de novo deletion on a maternally inherited beta-globin gene chromosome involving approximately 30 kb of sequences 5' to the epsilon gene; the deletion extends from -9.5 kb to -39 kb 5' of epsilon and includes three of the four DNase I hypersensitive sites (at -10.9 kb, -14.7 kb, and -18 kb 5' of epsilon). The remaining sequences of the beta-globin complex, including the DNase I hypersensitive sites at -6.1 kb and all structural genes in cis to the deletion are physically intact, but presumably nonfunctional, as evidenced by the presence of a beta S-globin gene that is not expressed as a sickle hemoglobin. Deletion of DNase I hypersensitive sites on a previously functional beta-globin gene complex confirms the significance of these sites in regulating globin gene expression.

The human beta-globin locus activation region alters the developmental fate of a human fetal globin gene in transgenic mice

We linked a 3.3-kilobase fragment containing the entire A gamma-globin gene together with 1.3 kilobases of 5' flanking and 0.37 kilobase of 3' flanking DNA to a 2.5-kilobase fragment containing four of the developmentally stable hypersensitive sites normally located in the 5' region of the human beta-globin locus. This construct was injected into fertilized mouse eggs, and its expression was analyzed in the primitive and definitive erythroid cells, as well as the brain of 14-day embryos. All six transgenic individuals that contained intact copies of the construct expressed the transgene in an erythroid-specific fashion. Expression was observed in both primitive and definitive erythroid cells. This is in marked contrast to previous transgenic mice experiments using the same A gamma-globin gene fragment in isolation, where expression was restricted to primitive erythroid cells. Our results show that the region containing the developmentally stable globin locus hypersensitive sites changes the developmental stage specificity of a human fetal globin gene in transgenic mice. These observations imply that sequences additional to those used here are involved in the developmental control of fetal globin gene expression in vivo. The ability to express fetal globin in adult erythroid cells allows one to consider using fetal globin genes for gene therapy of sickle cell disease.

Human beta-globin gene expression in transgenic mice is enhanced by a distant DNase I hypersensitive site

Several lines of evidence suggest that erythroid-specific DNase I hypersensitive sites (HS) located far upstream of the human beta-globin gene are important in regulating beta-globin gene expression. We used the polymerase chain reaction technique to amplify and clone an 882-base-pair DNA fragment spanning one of these HS, designated HSII, which is located 54 kilobases upstream of the beta-globin gene. The cloned HSII fragment was linked to a human beta-globin gene in either the genomic (HSII-beta) or antigenomic (HSII-beta) orientation. These two constructs and a beta-globin gene alone (beta) were injected into fertilized mouse eggs, and expression was analyzed in liver and brain from day-16 transgenic fetuses. Five of 7 beta-transgenic fetuses expressed human beta-globin mRNA, but the level of expression per gene copy was low, ranging from 0.93 to 22.4% of mouse alpha-globin mRNA (average 9.9%). In contrast, 11 of 12 HSII-beta transgenic fetuses expressed beta-globin mRNA at levels per gene copy ranging from 31.3 to 336.6% of mouse alpha-globin mRNA (average 139.5%). Only three fetuses containing intact copies of the HSII-beta construct were produced. Two of three expressed human beta-globin mRNA at levels per gene copy of 179.2 and 387.1%. Expression of human beta-globin mRNA was tissue-specific in all three types of transgenic fetuses. These studies demonstrate that a small DNA fragment containing a single erythroid-specific HS can stimulate high-level human beta-globin gene expression in transgenic mice.

Synthesis of functional human hemoglobin in transgenic mice

Human alpha- and beta-globin genes were separately fused downstream of two erythroid-specific deoxyribonuclease (DNase) I super-hypersensitive sites that are normally located 50 kilobases upstream of the human beta-globin gene. These two constructs were coinjected into fertilized mouse eggs, and expression was analyzed in transgenic animals that developed. Mice that had intact copies of the transgenes expressed high levels of correctly initiated human alpha- and beta-globin messenger RNA specifically in erythroid tissue. An authentic human hemoglobin was formed in adult erythrocytes that when purified had an oxygen equilibrium curve identical to the curve of native human hemoglobin A (Hb A). Thus, functional human hemoglobin can be synthesized in transgenic mice. This provides a foundation for production of mouse models of human hemoglobinopathies such as sickle cell disease.

Molecular genetics of human blue cone monochromacy

Blue cone monochromacy is a rare X-linked disorder of color vision characterized by the absence of both red and green cone sensitivities. In 12 of 12 families carrying this trait, alterations are observed in the red and green visual pigment gene cluster. The alterations fall into two classes. One class arose from the wild type by a two-step pathway consisting of unequal homologous recombination and point mutation. The second class arose by nonhomologous deletion of genomic DNA adjacent to the red and green pigment gene cluster. These deletions define a 579-base pair region that is located 4 kilobases upstream of the red pigment gene and 43 kilobases upstream of the nearest green pigment gene; this 579-base pair region is essential for the activity of both pigment genes.

Activation and repression of a beta-globin gene in cell hybrids is accompanied by a shift in its temporal replication

To investigate whether a switch in the transcriptional activity of a gene is associated with a change in the timing of replication during the S phase, we examined the replication timing of the beta-globin genes in two different types of somatic cell hybrids. In mouse hepatoma (Hepa 1a) x mouse erythroleukemia (MEL) hybrid cells, the beta-globin gene from the MEL parent is transcriptionally inactivated and is later replicating than in the parental MEL cell line. In human fibroblast (GM3552) x MEL hybrid cells, the human beta-globin gene is transcriptionally activated, and all of the sequences within the human beta-globin domain (200 kilobases) we have examined appear to be earlier replicating than those in the parental fibroblast cell line. The chromatin configuration of the activated human beta-globin domain in the hybrids is relatively more sensitive to nucleases than that in the fibroblasts. Furthermore, major nuclease-hypersensitive sites that were absent in the chromatin flanking the distal 5' region of the human beta-globin gene cluster in the parental fibroblast cell line are present in the transcriptionally activated domain in the hybrid cell line. These results suggest that timing of replication of globin genes has been altered in these hybrid cells and thus is not fixed during the process of differentiation.

Delta zero-thalassemia in cis of beta Knossos-globin gene. Normal structure transient expression of the delta-globin gene

We have previously described the first homozygous cases of Hb Knossos in an Algerian family. The Hb A2 was completely absent, ascertaining the presence of a delta zero-thalassemia determinant in cis of the beta Knossos S gene. Here, we investigate the affected delta-globin gene. The complete DNA sequence of the gene and its 5' and 3' flanking regions was determined. Only two nucleotide changes were recorded: a C----T substitution at -199 and an AT insertion at -448 upstream from the cap site. To examine the involvement of these changes in gene function, the delta-gene was subcloned in an expression vector and introduced into COS cells. Analysis of RNA derived from these cells, using an S1 protection assay and dot-blot hybridization, revealed qualitatively and quantitatively normal transcription. The loss of delta-globin gene activity in vivo may be due to the alteration of a tissue-specific control.

Molecular analysis of the human beta-globin locus activation region

Recently, DNA sequences containing four erythroid-specific DNase I hypersensitive sites within 20 kilobases 5' of the human epsilon-globin gene have been identified as an important cis-acting regulatory element, the locus activation region (LAR). Subfragments of the LAR, containing either all or only the two 5' or two 3' hypersensitive sites were linked to the human beta-globin gene and analyzed for their effect on globin gene expression in stably transformed mouse erythroleukemia (MEL) cells. Constructs containing all four of the hypersensitive sites increase beta-globin mRNA levels 8- to 13-fold, while constructs with only the 5' or 3' sites increase globin expression to a lesser extent. No effect was seen when the constructs were assayed in 3T3 fibroblasts. All of the LAR derivatives form hypersensitive sites at the corresponding sequence position in MEL cells prior to and after induction of MEL cell differentiation. However, in 3T3 fibroblasts only the hypersensitive site corresponding to the previously described erythroid-specific -10.9 site was formed.

Cloning of cDNA for the major DNA-binding protein of the erythroid lineage through expression in mammalian cells

Genes expressed in erythroid cells contain binding sites for a cell-specific factor believed to be an important regulator for this haematopoietic lineage. Using high-level transient expression in mammalian cells, we have identified complementary DNA encoding the murine protein. The factor, a new member of the zinc-finger family of DNA-binding proteins, is restricted to erythroid cells at the level of RNA expression and is closely homologous between mouse and man.

Transcriptional promiscuity of the human alpha-globin gene

The human alpha-globin gene displays the unusual property of transcriptional promiscuity: that is, it functions in the absence of an enhancer when transfected into nonerythroid cell lines. It is also unusual in that its promoter region lies in a hypomethylated HpaII tiny fragment (HTF) island containing multiple copies of the consensus sequence for the SP1-binding site. We have investigated whether there is a relationship between these two observations. First, we investigated the mouse alpha-globin gene since it does not lie in an HTF island. We have demonstrated that it was not transcriptionally promiscuous. Second, we studied the transcriptional activity of the human alpha-globin gene in the absence of the GC-rich region containing putative SP1-binding sites and found a small (two- to threefold) but consistent positive effect of this region on transcriptional activity in both nonerythroid and erythroid cell lines. However, this effect did not account for the promiscuous nature of the human alpha-globin gene. We found that in a nonreplicating system, the human alpha-globin gene, like that of the mouse, required a simian virus 40 enhancer in order to be transcriptionally active in nonerythroid and erythroid cell lines. Since we only observed enhancer independence of the human alpha-globin gene in a high-copy-number replicating system, we suggest that competition for trans-acting factors could explain these results. Finally, our experiments with the erythroid cell line Putko suggest that there are no tissue-specific enhancers within 1 kilobase 5' of the human alpha-globin cap site or within the gene itself.

An erythroid-specific, developmental-stage-independent enhancer far upstream of the human "beta-like globin" genes

We have identified an erythroid-specific enhancer element far upstream of the human "beta-like globin" genes, at 10.2-11.0 kilobases 5' of the embryonic epsilon-globin gene, and thus at 53-54 kilobases 5' of the adult beta-globin gene. It is capable of enhancing the expression of a cis-linked test gene by up to 300-fold. This enhancer element is apparently developmental-stage-independent, as it is functional at the embryonic and the adult developmental stages in erythroid cells that are expressing the respective beta-like globin genes. The enhancer and globin promoter sequences work in synergy and are capable of conferring on a cis-linked gene the high transcriptional efficiency (enhancer function), erythroid specificity (enhancer and promoter functions), and developmental-stage specificity (promoter function) that are characteristic of the in vivo transcription of the beta-like globin genes in erythroid cells.

A dominant control region from the human beta-globin locus conferring integration site-independent gene expression

The regulatory elements that determine the expression pattern of a number of eukaryotic genes expressed specifically in certain tissues have been defined and studied in detail. In general, however, the expression conferred by these elements on genes reintroduced into the genomes of cell lines and transgenic animals has turned out to be at a low level relative to that of endogenous genes, and influenced by the chromosomal site of insertion of the exogenous construct. We have previously shown that if regions flanking the human beta-globin locus are introduced into the mouse genome along with the human beta-globin gene, a level of expression comparable to that of endogenous genes can be achieved that is also independent of integration site. We have now defined a dominant control region with these properties consisting of 6.5 kilobases of DNA encompassing erythroid cell-specific DNase I hypersensitive sites. The identification of such dominant control regions could have important applications in somatic gene therapy.

A single erythroid-specific DNase I super-hypersensitive site activates high levels of human beta-globin gene expression in transgenic mice

Erythroid-specific DNase I super-hypersensitive (HS) sites that are normally located far upstream of the human beta-globin locus were inserted immediately upstream of a 4.1-kb fragment containing the human beta-globin gene. These constructs (HS beta) and a construct containing the beta-globin gene alone (beta) were microinjected into fertilized mouse eggs, and expression was analyzed in erythroid fetal liver and brain of day-16 embryos that developed. Only 7 of 23 animals that contained the beta gene alone expressed human beta-globin mRNA in erythroid tissue, and the average level of expression per gene copy was 0.3% of endogenous mouse beta-globin mRNA. In contrast, 50 of 51 transgenic mice that contained various HS beta constructs expressed the transgene specifically in erythroid tissue. The average level of expression per gene copy for constructs containing all five upstream HS sites was 109% of endogenous mouse beta-globin mRNA. Constructs that contained a single super-hypersensitive site (HS II beta) expressed 40% as much human beta-globin as mouse beta-globin mRNA per gene copy. These results demonstrate that the HS VI site, normally located downstream of the human beta-globin locus, is not required for high-level expression. Furthermore, the results demonstrate that high levels of human beta-globin gene expression can be obtained in transgenic mice even when a relatively small fragment of DNA (1.9 kb) containing erythroid-specific super-hypersensitive site II (HS II) is inserted upstream of the human beta-globin gene.

Two regulatory domains flank the mouse H19 gene

The mouse H19 gene was identified by virtue of its coordinate regulation with the mouse alpha-fetoprotein gene. Both genes are expressed in the fetal liver, gut, and visceral endoderm of the yolk sac and are repressed shortly after birth in the liver and gut. They are both under the control of two trans-acting loci: raf, which affects the adult basal levels of the two mRNAs, and Rif, which affects their inducibility during liver regeneration. One crucial difference between the two genes is the activation of the H19 gene in mesoderm derivatives, skeletal and cardiac muscle. As a strategy for explaining both the similarities and differences in their modes of expression, the regulatory domains responsible for the expression of the H19 gene in liver were identified by transiently introducing the gene into a human hepatoma cell line. Two regions necessary for high-level expression of the gene could be identified, a promoter-proximal domain immediately preceding the start of transcription and an enhancer domain which lies between 5 and 6.5 kilobases 3' of the polyadenylation site. The 3' domain consists of two separable enhancer elements, each of which exhibits the properties of tissue-specific enhancers. Nucleotide sequence comparisons between the two H19 and three alpha-fetoprotein enhancers revealed limited similarities which are candidates for binding of common regulatory factors. Sequences which lie 3' of the gene are also required for the expression of the H19 gene following differentiation of teratocarcinoma cells into visceral endoderm.

The coordinate replication of the human beta-globin gene domain reflects its transcriptional activity and nuclease hypersensitivity

The temporal order of replication of DNA sequences in the chromosomal domain containing the human beta-globin gene cluster and its flanking sequences (140 kilobases) was measured and compared in two different human cell lines. In human erythroleukemia (K562) cells, in which embryonic and fetal globin genes are transcribed, all of the sequences we examined from the beta-globin domain replicated early during S phase, while in HeLa cells, in which globin genes are transcriptionally silent, these sequences replicated late during S. Potential sites of initiation of DNA replication within this domain were identified. The beta-globin gene domain was also found to differ with respect to the nuclease sensitivity of the chromatin in these two cell lines. In K562 cells, hypersensitive sites for endogenous nucleases and DNase I were present in the chromatin near the earliest-replicating segments in the beta-globin domain.

Developmental programs of human erythroleukemia cells: globin gene expression and methylation

We investigated the programs of globin gene expression in three known (K562, HEL, and KMOE) and three novel (OCI-M1, OCI-M2, and HEL-R) human erythroleukemic cell lines of adult origin. RNAs from induced and uninduced cells were analyzed for epsilon-, gamma-, delta-, and beta-, zeta-globin-specific transcripts. While high-level gamma-globin expression was common, the lines differed in their expression of embryonic (epsilon, zeta) and adult (delta, beta) globin mRNAs. The patterns of globin gene methylation were generally consistent with their observed expression profiles, with many of the same correlations being seen in normal cells. Although the programs of globin gene expression and methylation displayed by the lines appeared to be diverse, they were not random; rather, they made developmental sense, mimicking defined globin gene programs observed during normal human development. The characteristics exhibited by several of these lines suggest that they may have been derived from the transformation of multi- or oligopotent hematopoietic progenitor cells. We speculate that the expression of fetal or embryonic globins in these adult erythroleukemic cell lines is not an aberration of neoplastic transformation but is indicative of a fetal or embryonic potential in normal adult hematopoietic progenitors.

High-level erythroid expression of human alpha-globin genes in transgenic mice

The human alpha 1-globin gene was fused downstream of two erythroid-specific DNase I super-hypersensitive sites that are normally located upstream of the human beta-globin locus. This construct was injected into fertilized mouse eggs, and expression was analyzed in 16-day fetal livers and brains. All 11 fetuses that contained intact copies of the transgene expressed correctly initiated human alpha-globin mRNA in the erythroid fetal liver but not in brain. Levels of expression ranged from 4% to 337% of endogenous mouse beta-globin mRNA. A human alpha-globin construct that did not contain super-hypersensitive sites was not expressed. These results demonstrate that human beta-globin locus activation sequences can stimulate high levels of human alpha-globin gene expression in erythroid tissue of transgenic mice. The results also provide a foundation for experiments designed to coexpress human alpha- and beta-globin genes in transgenic mice and suggest a feasible approach for production of a mouse model for human sickle cell disease.

Transcriptional promiscuity of the human alpha-globin gene

The human alpha-globin gene displays the unusual property of transcriptional promiscuity: that is, it functions in the absence of an enhancer when transfected into nonerythroid cell lines. It is also unusual in that its promoter region lies in a hypomethylated HpaII tiny fragment (HTF) island containing multiple copies of the consensus sequence for the SP1-binding site. We have investigated whether there is a relationship between these two observations. First, we investigated the mouse alpha-globin gene since it does not lie in an HTF island. We have demonstrated that it was not transcriptionally promiscuous. Second, we studied the transcriptional activity of the human alpha-globin gene in the absence of the GC-rich region containing putative SP1-binding sites and found a small (two- to threefold) but consistent positive effect of this region on transcriptional activity in both nonerythroid and erythroid cell lines. However, this effect did not account for the promiscuous nature of the human alpha-globin gene. We found that in a nonreplicating system, the human alpha-globin gene, like that of the mouse, required a simian virus 40 enhancer in order to be transcriptionally active in nonerythroid and erythroid cell lines. Since we only observed enhancer independence of the human alpha-globin gene in a high-copy-number replicating system, we suggest that competition for trans-acting factors could explain these results. Finally, our experiments with the erythroid cell line Putko suggest that there are no tissue-specific enhancers within 1 kilobase 5' of the human alpha-globin cap site or within the gene itself.

Activation of developmentally mutated human globin genes by cell fusion

Human fetal globin genes are not expressed in hybrid cells produced by the fusion of normal human lymphocytes with mouse erythroleukemia cells. In contrast, when lymphocytes from persons with globin gene developmental mutations (hereditary persistence of fetal hemoglobin) are used for these fusions, fetal globin is expressed in the hybrid cells. Thus, mutations of developmental origin can be reconstituted in vitro by fusing mutant lymphoid cells with differentiated cell lines of the proper lineage. This system can readily be used for analyses, such as globin gene methylation, that normally require large numbers of pure nucleated erythroid cells, which are difficult to obtain.

The coordinate replication of the human beta-globin gene domain reflects its transcriptional activity and nuclease hypersensitivity

The temporal order of replication of DNA sequences in the chromosomal domain containing the human beta-globin gene cluster and its flanking sequences (140 kilobases) was measured and compared in two different human cell lines. In human erythroleukemia (K562) cells, in which embryonic and fetal globin genes are transcribed, all of the sequences we examined from the beta-globin domain replicated early during S phase, while in HeLa cells, in which globin genes are transcriptionally silent, these sequences replicated late during S. Potential sites of initiation of DNA replication within this domain were identified. The beta-globin gene domain was also found to differ with respect to the nuclease sensitivity of the chromatin in these two cell lines. In K562 cells, hypersensitive sites for endogenous nucleases and DNase I were present in the chromatin near the earliest-replicating segments in the beta-globin domain.

Two regulatory domains flank the mouse H19 gene

The mouse H19 gene was identified by virtue of its coordinate regulation with the mouse alpha-fetoprotein gene. Both genes are expressed in the fetal liver, gut, and visceral endoderm of the yolk sac and are repressed shortly after birth in the liver and gut. They are both under the control of two trans-acting loci: raf, which affects the adult basal levels of the two mRNAs, and Rif, which affects their inducibility during liver regeneration. One crucial difference between the two genes is the activation of the H19 gene in mesoderm derivatives, skeletal and cardiac muscle. As a strategy for explaining both the similarities and differences in their modes of expression, the regulatory domains responsible for the expression of the H19 gene in liver were identified by transiently introducing the gene into a human hepatoma cell line. Two regions necessary for high-level expression of the gene could be identified, a promoter-proximal domain immediately preceding the start of transcription and an enhancer domain which lies between 5 and 6.5 kilobases 3' of the polyadenylation site. The 3' domain consists of two separable enhancer elements, each of which exhibits the properties of tissue-specific enhancers. Nucleotide sequence comparisons between the two H19 and three alpha-fetoprotein enhancers revealed limited similarities which are candidates for binding of common regulatory factors. Sequences which lie 3' of the gene are also required for the expression of the H19 gene following differentiation of teratocarcinoma cells into visceral endoderm.

Developmental programs of human erythroleukemia cells: globin gene expression and methylation

We investigated the programs of globin gene expression in three known (K562, HEL, and KMOE) and three novel (OCI-M1, OCI-M2, and HEL-R) human erythroleukemic cell lines of adult origin. RNAs from induced and uninduced cells were analyzed for epsilon-, gamma-, delta-, and beta-, zeta-globin-specific transcripts. While high-level gamma-globin expression was common, the lines differed in their expression of embryonic (epsilon, zeta) and adult (delta, beta) globin mRNAs. The patterns of globin gene methylation were generally consistent with their observed expression profiles, with many of the same correlations being seen in normal cells. Although the programs of globin gene expression and methylation displayed by the lines appeared to be diverse, they were not random; rather, they made developmental sense, mimicking defined globin gene programs observed during normal human development. The characteristics exhibited by several of these lines suggest that they may have been derived from the transformation of multi- or oligopotent hematopoietic progenitor cells. We speculate that the expression of fetal or embryonic globins in these adult erythroleukemic cell lines is not an aberration of neoplastic transformation but is indicative of a fetal or embryonic potential in normal adult hematopoietic progenitors.

Asynchronous DNA replication within the human beta-globin gene locus

The timing of DNA replication of the human beta-globin gene locus has been studied by blot hybridization of newly synthesized BrdUrd-substituted DNA from cells in different stages of the S phase. Using probes that span greater than 120 kilobases across the human beta-globin gene locus, we show that the majority of this domain replicates in early S phase in the human erythroleukemia cell line K562 and in middle-to-late S phase in the lymphoid cell line Manca. However, in K562 cells three small regions display a strikingly different replication pattern than adjacent sequences. These islands, located in the inter-gamma-globin gene region and approximately 20 kilobases 5' to the epsilon-globin gene and 20 kilobases 3' to the beta-globin gene, replicate later and throughout S phase. A similar area is also present in the alpha-globin gene region in K562 cells. We suggest that these regions may represent sites of termination of replication forks.

The human beta-globin gene 3' enhancer contains multiple binding sites for an erythroid-specific protein

We have shown that the minimal enhancer fragment present in the 3'-flanking region of the human beta-globin gene contains four regions that bind nuclear proteins in vitro. By using gel mobility shift and DNase I footprinting assays, we were able to show that each of these regions binds an erythroid-cell-specific nuclear factor which we name NF-E1. This factor is present in erythroid cells at different developmental stages of globin gene expression. The recognition sequence of this protein (A/C Py T/A ATC A/T Py) is also present in the intragenic enhancer and the promoter of the beta-globin gene as well as in the promoter of other erythroid-cell specific genes. In addition to NF-E1, each of the four binding regions interacts with at least one other protein factor.

Regulated expression of genes inserted at the human chromosomal beta-globin locus by homologous recombination

We have examined the effect of the site of integration on the expression of cloned genes introduced into cultured erythroid cells. Smithies et al. [Smithies, O., Gregg, R.G., Boggs, S.S., Koralewski, M.A. & Kucherlapati, R.S. (1985) Nature (London) 317, 230-234] reported the targeted integration of DNA into the human beta-globin locus on chromosome 11 in a mouse erythroleukemia-human cell hybrid. These hybrid cells can undergo erythroid differentiation leading to greatly increased mouse and human beta-globin synthesis. By transfection of these hybrid cells with a plasmid carrying a modified human beta-globin gene and a foreign gene composed of the coding sequence of the bacterial neomycin-resistance gene linked to simian virus 40 transcription signals (SVneo), cells were obtained in which the two genes are integrated at the beta-globin locus on human chromosome 11 or at random sites. When we examined the response of the integrated genes to cell differentiation, we found that the genes inserted at the beta-globin locus were induced during differentiation, whereas randomly positioned copies were not induced. Even the foreign SVneo gene was inducible when it had been integrated at the beta-globin locus. The results show that genes introduced at the beta-globin locus acquire some of the regulatory properties of globin genes during erythroid differentiation.

Nuclease hypersensitivity in the beta-globin gene region of K562 cells

We have investigated chromatin structure in the beta-globin gene region of the K562 human erythroleukemic cell line by using S1 and DNase I nuclease sensitivity assays. Despite the lack of beta-globin gene expression in these cells, we find nuclease-hypersensitive sites to these enzymes in its 5' and 3' flanking regions in K562 chromatin. This result is in contrast to previous reports in which no hypersensitive sites were found in the immediate vicinity of this gene. In the 3' region, one major hypersensitive site at 0.9 kpb 3' and three minor hypersensitive sites at 0.7 kbp, 0.5 kbp 3' and 0.2 kbp 5' of the polyadenylation site were observed; these sites are very similar to those found in fetal liver and adult bone marrow cells in which the beta-globin gene is expressed. We find hypersensitive sites to both enzymes in the 5' region of the beta-globin gene: a major site 0.8 kbp 5' to the cap site, and two minor sites 1.2 and 1.5 kbp 5' to the cap site. The -0.8 kbp site is also present in plasmids containing the beta-globin gene. Our results suggest that the lack of beta-globin gene expression may be related to the lack of hypersensitivity sites in the immediate (150 bp) 5' flanking region of the beta-globin gene, as occurs in other active globin genes.

Evidence for a locus activation region: the formation of developmentally stable hypersensitive sites in globin-expressing hybrids

We have analyzed the chromatin structure of the human beta-globin locus in somatic cell hybrids resulting from the fusion of human non-erythroid cells and mouse erythroleukemia (MEL) cells. In these hybrids, the human adult beta-globin gene, but neither the embryonic nor fetal globin genes, is activated transcriptionally. In addition, the DNase I-resistant beta-like globin locus characteristic of the parental non-erythroid human cells (1,2) is reorganized over an approximately 80 kb region, including the formation of the developmentally stable hypersensitive sites 50 kb 5' and 20 kg 3' of the activated adult beta-globin gene (2,3). These results are consistent with the hypothesis that events occurring at the 5' and/or 3' developmentally stable hypersensitive sites are important, if not necessary, for the activation of the beta-globin locus.

Activation of the human beta-globin promoter in K562 cells by DNA sequences 5' to the fetal gamma- or embryonic zeta-globin genes

Regulatory sequences of the human fetal gamma-globin gene were studied by constructing composite gamma/beta globin promoters and comparing their function to that of intact beta promoters in human K562 cells. The beta-globin gene with either 1,600 or 127 basepairs of beta promoter sequence was not expressed after stable introduction into K562 cells, consistent with the known inactivity of the beta-globin gene in these cells. In contrast, a gamma/beta promoter composed of a gamma fragment spanning positions -408 to -137 joined to the 127-bp beta promoter was able to drive the beta-globin gene. The gene appeared to be inducible with hemin. A zeta-globin 5' flanking fragment also activated the beta promoter. The function of a series of composite gamma/beta promoters was then assessed by their ability to drive directly the neomycin resistance gene, again in stably transformed cells. The -408 to -137 gamma fragment activated the beta promoter in an orientation-specific manner in this assay. Deletion analysis showed that regulatory sequences were present between positions -259 and -137 of the fetal gamma-globin gene flanking region.

The beta-subunit of follicle-stimulating hormone is deleted in patients with aniridia and Wilms' tumour, allowing a further definition of the WAGR locus

One in 10,000 children develops Wilms' tumour, an embryonal malignancy of the kidney. Although most Wilms' tumours are sporadic, a genetic predisposition is associated with aniridia, genito-urinary malformations and mental retardation (the WAGR syndrome). Patients with this syndrome typically exhibit constitutional deletions involving band p13 of one chromosome 11 homologue. It is likely that these deletions overlap a cluster of separate but closely linked genes that control the development of the kidney, iris and urogenital tract (the WAGR complex). A discrete aniridia locus, in particular, has been defined within this chromosomal segment by a reciprocal translocation, transmitted through three generations, which interrupts 11p13. In addition, the specific loss of chromosome 11p alleles in sporadic Wilms' tumours has been demonstrated, suggesting that the WAGR complex includes a recessive oncogene, analogous to the retinoblastoma locus on chromosome 13. In WAGR patients, the inherited 11p deletion is thought to represent the first of two events required for the initiation of a Wilms' tumour, as suggested by Knudson from epidemiological data. We have now isolated the deleted chromosomes 11 from four WAGR patients in hamster-human somatic cell hybrids, and have tested genomic DNA from the hybrids with chromosome 11-specific probes. We show that 4 of 31 markers are deleted in at least one patient, but that of these markers, only the gene encoding the beta-subunit of follicle-stimulating hormone (FSHB) is deleted in all four patients. Our results demonstrate close physical linkage between FSHB and the WAGR locus, suggest a gene order for the four deleted markers and exclude other markers tested from this region. In hybrids prepared from a balanced translocation carrier with familial aniridia, the four markers segregate into proximal and distal groups. The translocation breakpoint, which identifies the position of the aniridia gene on 11p, is immediately proximal to FSHB, in the interval between FSHB and the catalase gene.

Differences in DNase I sensitivity and methylation within the human beta-globin gene domain and correlation with expression

We have analysed the chromatin features of DNA regions encompassing human epsilon, G gamma, A gamma, delta and beta globin structural genes in fetal and adult erythroid cells on the one hand and adult lymphocytes on the other. Highly purified nuclei from these cells were submitted to DNase I digestion and the kinetic data were obtained from the percentage of residual hybridization of defined regions in Southern blots. Our results, as others have shown by different approaches, indicate that the structural genes of the beta-globin cluster are generally more sensitive to DNase I in the erythroid cells than in non-erythroid cells. Thus a domain of DNase I sensitivity related to the committed state is defined. In addition we show that within this DNase-I-sensitive beta cluster domain, individual genes of the cluster are arranged in subdomains of differential DNase I sensitivity, which correlate with their expression status. Furthermore the differential expression of the two fetal genes in the fetal stage is shown to be directly proportional to the degree of hypomethylation of these genes.

A developmentally stable chromatin structure in the human beta-globin gene cluster

The DNase I-hypersensitive sites in the human embryonic beta-globin gene region have been mapped in erythroid-enriched fractions of disaggregated fetal livers, in adult nucleated red blood cells, and in fetal brain tissue. Our analysis of a region extending 11 kilobases (kb) 5' of the epsilon-globin gene reveals many minor nuclease-hypersensitive sites and one major site located 6.1 kb upstream of the epsilon-globin gene. All of these hypersensitive sites are erythroid-specific, and the major site is stable throughout erythroid development. As assayed by nuclear runoff transcription, little or no epsilon-globin gene expression is detectable in fetal or adult erythroid cells. Thus, the presence of the major hypersensitive site 5' of the epsilon-globin gene in both fetal and adult erythroid cells demonstrates that this site is not specifically correlated with transcription of the gene or with a particular stage of development. Rather, this site may reflect an early event in erythroid differentiation. In addition, DNase I has been used to probe the overall sensitivity of epsilon-globin chromatin in fetal erythroid cells. Our findings indicate that the epsilon-globin gene as well as the other genes in the beta-globin cluster reside within the chromatin domain that is more DNase I-sensitive than "bulk" chromatin.