Detection of two tissue-specific DNA-binding proteins with affinity for sites in the mouse beta-globin intervening sequence 2

An intronic SNP in a RUNX1 binding site of SLC22A4, encoding an organic cation transporter, is associated with rheumatoid arthritis

Rheumatoid arthritis is a common inflammatory disease with complex genetic components. We investigated the genetic contribution of the cytokine gene cluster in chromosome 5q31 to susceptibility to rheumatoid arthritis in the Japanese population by case-control linkage disequilibrium (LD) mapping using single nucleotide polymorphisms (SNPs). Here we report that there is significant association between rheumatoid arthritis and the organic cation transporter gene SLC22A4 (P = 0.000034). We show that expression of SLC22A4 is specific to hematological and immunological tissues and that SLC22A4 is also highly expressed in the inflammatory joints of mice with collagen-induced arthritis. A SNP affects the transcriptional efficiency of SLC22A4 in vitro, owing to an allelic difference in affinity to Runt-related transcription factor 1 (RUNX1), a transcriptional regulator in the hematopoietic system. A SNP in RUNX1 is also strongly associated with rheumatoid arthritis (P = 0.00035). Our data indicate that the regulation of SLC22A4 expression by RUNX1 is associated with susceptibility to rheumatoid arthritis, which may represent an example of an epistatic effect of two genes on this disorder.

Dual regulatory role of human cytomegalovirus immediate-early protein in IL1B transcription is dependent upon Spi-1/PU.1

Activation of IL1B gene transcription has been shown to play a crucial role in human cytomegalovirus (HCMV) infection. We previously reported that HCMV immediate-early (IE) proteins vigorously transactivate IL1B expression without the need for a normally essential upstream enhancer. This activation appears to depend upon protein-protein tethering between IE2, which provides a transcription activation domain (TAD), and the DNA-binding domain of the transcription factor Spi-1. We now show a distinct mechanism by which IE1 and IE2 mediate both weak Spi-1-independent and vigorous Spi-1-dependent IL1B transcription from the -59 to +12 IL1B core promoter. These results demonstrate that in contrast to non-viral, enhancer-mediated, transactivation of IL1B, the IE mechanism is not absolutely dependent upon Spi-1. However, Spi-1 is required for vigorous transcription. Additionally, we have discovered that IE1, which cooperates with IE2 to transactivate IL1B, has minimal activity in the absence of IE2 and Spi-1. Furthermore, IE1 is a dual-acting factor, which can either activate or repress IL1B, depending on the presence of both IE2 and the Spi-1 TADs. Therefore, the relative expression of IE1 and IE2, which varies during HCMV infection, may provide a molecular mechanism by which IL1B can be repressed, thus, avoiding clearance by the host.

Autocrine induction of the human pro-IL-1beta gene promoter by IL-1beta in monocytes

IL-1beta is produced primarily by activated monocytes/macrophages. We report in this study that IL-1beta induces the human pro-IL-1beta (IL1B) gene promoter in human THP-1 monocytic cells. The -131 to +12 minimal IL1B promoter was induced by IL-1beta in a dose-dependent manner. The promoter possesses two important transcription factor binding motifs, one for an ETS family transcription factor Spi-1 (PU.1), and the other a binding site for NF-IL6 (CCAAT/enhancer binding protein beta). Autocrine promoter activity was completely inhibited by mutation of the Spi-1 site. Mutation of the NF-IL6 binding motif caused partial loss of activity. EMSAs using THP-1 cell nuclear extracts indicated that IL-1beta significantly induced Spi-1 binding to its target site within the IL1B promoter that was maximal at 1 h after stimulation, correlating with the kinetics of IL-1beta induction. The importance of Spi-1 was supported by our observation that Spi-1-deficient EL4 thymocytes exhibited IL-1beta-induced activity only after transfection with a Spi-1 expression vector. Moreover, TNFR-associated factor 6 also required Spi-1 to activate the promoter. Transfection studies using Spi-1 mutant constructs showed that the TATA-binding protein binding and glutamine-rich domains of Spi-1 were important for IL-1beta induction, whereas LPS induction required the proline, glutamic acid, serine, and threonine-rich domain containing serine 148 as well as the TATA-binding protein and glutamine-rich domains. We conclude that the IL1B promoter is an IL-1beta-responsive sequence as a result of its ability to bind Spi-1 in response to IL-1beta.

Human monocyte/neutrophil elastase inhibitor (MNEI) is regulated by PU.1/Spi-1, Sp1, and NF-kappaB

Human monocyte/neutrophil elastase inhibitor (MNEI) is a specific inhibitor of the neutrophil azurophil granule proteases including elastase. To understand the physiological mechanisms that regulate expression of MNEI, we dissected a 1.0 kb region upstream of exon 1. On transient transfection, promoter activity of MNEI-luciferase constructs was highest in U937 myeloid cells, followed by K562 hematopoietic cells, followed by HeLa cervical carcinoma cells, indicating that the MNEI promoter is most active in myeloid cells and is also active in non-myeloid cells. Three transcription factor binding elements, which confer the majority of activity, are located within the first 180 base pairs of the promoter, one of which, located at -128, was active in U937 and K562 cells but inactive in non-myeloid HeLa cells. The three proximal elements were identified by transient transfection, mutation, gel shift and competition assays as Sp1 at -170, PU.1/Spi-1 at -128, and Sp1 at -66. The trans-acting factors that bind and control these elements were detected, and their identity confirmed by antibody supershift assays. Further upstream at -821, an additional regulatory element was identified controlled by NF-kappaB, which supports the highest levels of MNEI transcriptional activity. In U937 cells, reporter gene expression by the MNEI-luciferase construct that included the NF-kappaB element was two- to three-fold greater than the construct without the element. In addition, treatment of myeloid cells with lipopolysaccharide, a complex glycolipid of gram-negative bacteria, activated NF-kappaB to bind the -821 element, together suggesting that enhancement of expression of the anti-inflammatory MNEI gene is linked to innate immune responses to bacterial infection.

Correction of the NMR structure of the ETS1/DNA complex

The ETS family of transcription factors consists of a group of proteins that share a highly conserved 85 amino acid DNA-binding domain (DBD). This family recognizes a consensus sequence rich in purine bases with a central GGAA motif. A comparison of the published three-dimensional structures of the DBD/DNA complexes of ETS1 by NMR [Werner et al. (1995) Cell, 83, 761-771] and the related Pu.1 by X-ray crystallography [Kodandapani et al. (1996) Nature, 380, 456-460] reveals an apparent discrepancy in which the protein domains bind with opposite polarity to their target sequences. This surprising and highly unlikely result prompted us to reexamine our NMR structure. Additional NMR experiments now reveal an error in the original interpretation of the spectra defining the orientation of the ETS1-DBD on DNA. It was originally reported that the ETS1-DBD bound to DNA with a bipartite motif involving major groove recognition via a helix-turn-helix element and minor groove recognition via protein side-chain intercalation. The presence of intercalation was deduced on the basis of numerous NOEs between several amino acids in the protein and a resonance at 12.33 ppm originally assigned to a DNA imino proton. New NMR experiments now conclusively demonstrate that this resonance, which is located within the DNA imino proton region of the spectrum, arises from the hydroxyl proton of Tyr86. Realization of this error necessitated reanalysis of the intermolecular NOEs. This revealed that the orientation of the ETS1-DBD in the complex is opposite to that originally reported and that a tryptophan residue does not intercalate into the DNA. The calculation of a new ensemble of structures based on the corrected data indicates that the structure of the ETS1-DBD/DNA complex is indeed similar to the X-ray structure of the Pu.1-DBD/DNA complex.

Granulocytic Differentiation of Normal Hematopoietic Precursor Cells Induced by Transcription Factor PU.1 Correlates With Negative Regulation of the c-myb Promoter

Numerous transcription factors allow hematopoietic cells to respond to lineage- and stage-specific cytokines and/or to act as their effectors. The transcription factors PU.1 and c-Myb are essential for hematopoiesis, most likely acting at distinct stages of differentiation, but sharing a common set of target genes. To determine whether PU.1 and c-Myb are functionally interrelated, murine bone marrow (BM) cells and 32Dcl3 murine myeloid precursor cells were infected with a retrovirus carrying a PU.1 cDNA and assessed for myeloid colony formation and for granulocytic differentiation, respectively. Compared with noninfected normal BM cells or to cells infected with an empty virus, hematopoietic precursor cells expressing PU.1 formed an increased number of interleukin-3 (IL-3) and granulocyte colony-stimulating factor (G-CSF )–stimulated colonies. Moreover, granulocytic differentiation of 32Dcl3 cells constitutively expressing PU.1 was accelerated, as indicated by morphology and by expression of differentiation markers. Downregulation of c-Myb protein levels by expression of an antisense c-myb construct was also associated with a faster kinetics of 32Dcl3 granulocytic differentiation. Sequence analysis of the 5′ flanking region of the c-myb gene revealed a consensus PU box at position +16 to +21 able to specifically interact in electrophoretic mobility shift assays with either bacterially synthesized PU.1 protein or whole cell extracts from differentiated 32Dcl3 cells. Transient expression of PU.1 in cotransfection assays in different cell lines resulted in inhibition of chloramphenicol acetyl transferase activity driven by different segments of the c-myb promoter. Moreover, such an effect was dependent on an intact PU box. Thus, the ability of PU.1 to potentiate terminal myeloid differentiation appears to involve downregulation of c-myb expression, an essential step during differentiation of hematopoietic precursor cells.

Characterization of the human and mouse Fli-1 promoter regions

Overexpression of the Fli-1 gene has been shown to be involved in retrovirus-induced mouse tumors. Cloning of the 5' flanking sequence of the mouse and human Fli-1 exon 1 was performed. At least two major transcription initiation sites were localized respectively at 143 and 114 nucleotides upstream of the previously defined mouse Fli-1 cDNA 5' end. The sequences flanking the CAP sites show good conservation between human and mouse (94%). The promoter region contains a potential TATA box lying 30 bp from one of the major identified CAP sites. Several conserved elements, such as GATA, EBS, GC rich, AP-2, AP-3 elements and a repetition of GA were observed next to the two major CAP sites. Furthermore, this latter was shown to form a H-DNA structure in vitro by S1 nuclease sensitivity experiments. The highly conserved 5' non-translated region of exon 1 is predicted to form a very stable hairpin structure which could regulate the Fli-1 expression at the post-transcriptional level. In Cas-Br-E-induced tumors, all the proviruses are found clustered within 35 nucleotides directly upstream the Fli-1 ATG start codon, thus deleting the hairpin structure from the transcript. Promoter activity was tested using the CAT reporter gene transfected in mouse and human erythroid cell lines. No promoter activity could be detected with various mouse Fli-1 promoter-CAT constructs containing 600 bp of the 5' flanking region, the complete exon 1, the 5' end of intron 1 and/or retroviral LTR sequence. Constructions of the human homologue containing nearly 1.5 kbp of Fli-1 5' flanking region was also inactive in transfected cells. These results suggest that multiple levels of regulation might control the Fli-1 expression.

Structural and functional characterization of the human CD36 gene promoter: identification of a proximal PEBP2/CBF site

CD36 is a cell surface glycoprotein composed of a single polypeptide chain, which interacts with thrombospondin, collagens type I and IV, oxidized low density lipoprotein, fatty acids, anionic phospholipids, and erythrocytes parasitized with Plasmodium falciparum. Its expression is restricted to a few cell types, including monocyte/macrophages. In these cells, CD36 is involved in phagocytosis of apoptotic cells, and foam cell formation by uptake of oxidized low density lipoprotein. To study the molecular mechanisms that control the transcription of the CD36 gene in monocytic cells we have isolated and analyzed the CD36 promoter. Transient expression experiments of 5'-deletion fragments of the CD36 promoter coupled to luciferase demonstrated that as few as 158 base pairs upstream from the transcription initiation site were sufficient to direct the monocyte-specific transcription of the reporter gene. Within the above region, the fragment spanning nucleotides -158 to -90 was required for optimal transcription in monocytic cells. Biochemical analysis of the region -158/-90 revealed a binding site for transcription factors of the polyomavirus enhancer-binding protein 2/core-binding factor (PEBP2/CBF) family at position -103. Disruption of the PEBP2/CBF site markedly diminished the role of the PEBP2/CBF factors in the constitutive transcription of the CD36 gene. The involvement of members of the PEBP2/CBF family in chromosome translocations associated with acute myeloid leukemia, and in the transcriptional regulation of the myeloid-specific genes encoding for myeloperoxidase, elastase, and the colony-stimulating factor receptor, highlights the relevance of the regulation of the CD36 gene promoter in monocytic cells by members of the PEBP2/CBF family.

Expression of the transcription factor, Spi-1 (PU.1), in differentiating murine erythroleukemia cells Is regulated post-transcriptionally. Evidence for differential stability of transcription factor mRNAs following inducer exposure

Increased expression of the transcription factor Spi-1 (PU.1) results from retroviral insertion in nearly all Friend spleen focus-forming virus-transformed murine erythroleukemia cell lines and exposure of these cells to Me2SO, induces their differentiation and decreases Spi-1 mRNA level by 4-5-fold. While these results suggest that alterations in Spi-1 expression have significant effects on erythroblast growth and differentiation, neither the cause nor the effect of the decrease in Spi-1 expression that follows Me2SO exposure has been established. The experiments described here demonstrate that the effect of inducers on Spi-1 expression is regulated post-transcriptionally. Nuclear run-off transcriptions demonstrated that Spi-1 transcription was not decreased following Me2SO exposure. Additionally, expression of a recombinant Spi-1 mRNA under transcriptional control of a constitutively active Rous sarcoma virus promoter was regulated identically to endogenous Spi-1 mRNA. The ability of Me2SO to destabilize Spi-1 mRNA was selective, as the stability of the erythroid transcription factors GATA-1 and NF-E2 were not similarly effected. The effect of Me2SO on the stability of Spi-1 mRNA provides a novel means of altering gene expression in these cells and is likely to have significance for the differentiation of these cells.

Nuclear factor binding sites in human beta globin IVS2

The second intron of the human beta globin gene (beta IVS2) has been previously identified as a region required for proper expression of beta globin. To further characterize this region, we have footprinted the entire beta IVS2 and have analyzed regions of interest by electrophoretic mobility shift assay. Through these studies we have identified four utilized binding sites for the erythroid regulatory factor GATA-1, two sites bound by general transcription factor Oct-1, two sites bound by the nuclear matrix attachment DNA binding protein special A-T-rich binding protein 1, and a site bound by a potential homeobox protein. Additionally, we have found several factors displaying temporal or tissue specificity by electrophoretic mobility shift assay, which may be potentially involved in the regulation of beta globin expression. These proteins are not supershifted by antibodies to factors important in erythroid regulation such as GATA-1, NFE-2, or YY1, or by antibodies against more general transcription factors.

Partial cloning of the rat choline acetyltransferase gene and in situ localization of its transcripts in the cell body of cholinergic neurons in the brain stem and spinal cord

We have isolated recombinant lambda (lambda) phages which contain a part of the rat choline acetyltransferase (ChAT) gene. Restriction and Southern blot analyses using synthetic oligonucleotides indicate that these clones overlap one another and contain at least four exons which reside in 16.4 kb of sequence encoding from the middle to the 3' end, but not the 5'-region, of the rat ChAT gene. Partial sequence analyses revealed that the clones contain an exon whose nucleotide sequence corresponds to a highly conserved region of ChAT during evolution. RNase protection mapping experiments show that sequences represented by this exon are expressed at high levels in the spinal cord of adult rats and at low but detectable levels in PC12 cells. By using the genomic sequences, including the exon, as a hybridization probe, we have detected ChAT mRNAs in situ in rat tissues. In situ hybridization experiments using radioactive and non-radioactive probes revealed that cholinergic motoneurons in the spinal cord, the laterodorsal tegmental nucleus as well as the hypoglossal nucleus in the brain stem were labeled, suggesting that the genomic sequence can be used as a probe to measure the ChAT mRNA levels in those cholinergic neurons. The results also indicate that the non-radioactive method gives a better resolution in localizing the expression of ChAT transcripts in the cytoplasm of cholinergic neurons.

Downstream activating sequence within the coding region of a yeast gene: specific binding in vitro of RAP1 protein

Using a gel retardation assay, a protein factor that specifically interacts with a 33 bp intragenic sequence of the highly expressed and glucose-inducible SRP1 gene of Saccharomyces cerevisiae has been detected. This binding site is located in a transcribed region and within the open reading frame (positions +710 to +743 relative to the first base of the initiation codon). A mutant strain carrying a deletion of this binding site showed a dramatic decrease in steady-state levels of SRP1 transcripts. This decline is not the result of a decrease in mRNA stability, since expression of hybrid genes in which the SRP1 promoter was replaced by the heterologous CYC1 promoter was not affected by the binding site deletion. These findings suggest that the 33 bp sequence contains a cis-acting downstream activating element which is involved in the transcriptional activation of the SRP1 promoter. Sequence comparisons showed similarities between a site located within the 33 bp sequence and the high-affinity consensus binding site of the RAP1/GRF1 (also named TUF) factor and methylation interference experiments confirmed that this site was involved in the protein-DNA interaction. Both the results of competition experiments with upstream activating sequences of ribosomal protein genes (UASrpg), which are targets for RAP1 binding, and determination of the apparent molecular weight of the affinity-purified DNA-binding protein indicated that RAP1 factor recognized the SRP1 33 bp element. The 33 bp sequence was found to be unable to provide UAS activity when placed upstream of the TATA box and transcription start site.

Inhibition of the mutant p53 gene in transformation assays

Mutation of the p53 gene is a key element in the development of several human cancers. Intron 4, a noncoding region of the p53 gene, is required for optimal expression of that gene. We have previously shown that nuclear protein binds intron 4 and have defined the protein-binding site. In this paper we address the question, "Does the mutant p53 gene's ability to transform cells to the malignant phenotype depend on protein binding to intron 4?" Using an in vitro assay in which the mutant p53 gene and Ha-ras oncogene cooperate in transformation of cells to the malignant phenotype, we determined the ability of mutant mouse p53 gene constructs, with and without two base pair substitutions at the intron 4 protein-binding site, to participate in malignant transformation. On Day 1, 5 x 10(5) rat embryo fibroblasts were transfected by the calcium phosphate procedure with 10 micrograms of both a mutant p53 gene construct and Ha-ras oncogene. Malignant transformation was evidenced by the formation of discrete foci of heaped-up cells. After 14 days of incubation at 37 degrees C in DMEM and 10% fetal calf serum (8% CO2), the cells were stained with cresyl violet and the foci counted. In three separate experiments, the presence of two base pair substitutions at the intron 4 protein-binding site caused a significant decrease in the number of foci formed (P less than 0.05).

Localization and DNA sequence of a replication origin in the rhodopsin gene locus of Chinese hamster cells

A chromosomal origin of DNA replication has been localized within the single-copy rhodopsin gene locus in Chinese hamster (line CHO) cells using two methods. In the first method, single-copy segments were identified at 3 to 15 kb intervals within approximately 75 kb (kb = 10(3) bases) of cloned genomic DNA containing the early-replicating rhodopsin gene near its middle. The cloned single-copy segments were then used as hybridization probes to quantify the replication of their corresponding genomic segments as synchronized cells progressed into S phase. In the second method, genomic DNA synthesized in vivo or in permeabilized early S phase cells was hybridized with slot-blots of the cloned single-copy DNA segments to identify the earliest replicating part of the 75 kb mapped region. The first method indicates that the earliest replicating DNA is located within a 10 kb region beginning 4 kb upstream from and extending 1 kb beyond the rhodopsin gene. The second method confirms the location in the vicinity of the rhodopsin gene and indicates that the earliest replicating region is located within or very near the 4.5 kb rhodopsin gene itself. An extended region of 12 kb that encompasses the entire early-replicating region has been sequenced for analysis and comparison with currently characterized origin regions associated with the CHO dihydrofolate reductase (dhfr) and human c-myc genes. There are several sequence similarities between the dhfr rhodopsin origin regions, including common transcription promoter consensus sequences, rodent Alu repeats with their 3'-A+T rich flanking sequences, A+T-rich yeast ARS and Drosophila SAR consensus sequences, and simple (GA)n repeats, but there are no extended regions of direct similarity. The rhodopsin gene locus is the second sequenced CHO origin region.

A differentiation stage-specific factor interacts with mouse carbonic anhydrase form I gene and a conserved sequence in mammalian beta-globin genes

We have identified in mouse erythroleukemic (MEL) cells a novel factor which binds to the 3' flanking region of beta-globin genes. Upon induction, this DNA binding factor disappears as beta-globin gene transcription increases. The factor protects 20-30 base pairs (bp) of a sequence which is tightly conserved in beta-globin genes including chicken, human, mouse and rabbit. A very similar sequence binds the factor in the mouse carbonic anhydrase form I gene, whose transcription diminishes upon induction of MEL cells. The factor, or a closely related form, is detected in nonerythroid cells and binding sites have been detected in other genes, including c-myb and immunoglobulin heavy chain-enhancer. We suggest that this factor could play a role in determining the timing of gene expression in several differentiating cell types.

Diverse soybean actin transcripts contain a large intron in the 5' untranslated leader: structural similarity to vertebrate muscle actin genes

Plant actins are encoded by complex and highly divergent multigene families. Despite the general lack of intron conservation in animal, fungal and protist actin genes, evidence is presented which indicates that higher plant actin genes have an untranslated leader exon with structural similarity to that found in vertebrate actin genes. All functional higher plant actin genes sequenced to date contain a potential intron acceptor site in the 5' untranslated region 10 to 13 nucleotides upstream of the initiator ATG. A leader specific cDNA probe hybridized to sequences over 1.0 kbp upstream from the coding region confirming the presence of an upstream exon. Primer extension of mRNA with gene-specific oligonucleotides was used to analyze the 5' untranslated exon and leader intron from four divergent soybean actin genes, SAc3, 4, 6 and 7. The 5' ends of all four mRNAs are heterogeneous. The consensus promoter elements of the SAc7 actin promoter were identified. Gene specific primer extension sequencing of actin mRNAs indicated that splicing of the 5' leader intron occurred at the predicted acceptor site in SAc6 and SAc7. The SAc6 and SAc7 5' untranslated exons are small (88-111 nt) and the leader introns are relatively large (844-1496 nt). The presence of an intron within the 5' RNA leader and an intron which splits a glycine codon at position 152 in all plant actin genes and all vertebrate muscle actin genes suggests that these structures may have been conserved due to a functional role in actin expression. The 5' regions of these two soybean actin genes contain many unusual features including (CT) repeats and long stretches of pyrimidine-rich DNA. The possible roles of the upstream exon/intron and the C + T-rich regions are discussed.

Sequence similarities among monkey ori-enriched (ors) fragments

Nucleotide sequences have been determined for eight ors (ori-enriched sequence) fragments isolated from monkey DNA by a method that was designed to enrich for origins of DNA replication [Kaufmann et al., Mol. Cell. Biol. 5 (1985) 721-727]. Evidence has been presented that some or possibly all of these sequences can serve, albeit inefficiently, as oris in vivo [Frappier and Zannis-Hadjopoulos, Proc. Natl. Acad. Sci. USA 84 (1987) 6668-6672]. Two of the fragments were found to contain the long terminal repeat-like elements of the 'O-family' of moderately repetitive sequences that are present in human DNA as a transposon-like element [Paulson et al., Nature 315 (1985) 359-361]. Extensive pair-wise comparisons of the sequences failed to detect any statistically significant common sequences, except for long asymmetrically distributed A + T-rich stretches. Nonetheless, when the ors fragments were examined for the presence of published consensus sequences, seven of eight were found to contain the control sequence described by Dierks et al. [Cell 32 (1983) 695-706], and the same seven of eight were found to contain both the scaffold attachment region T consensus [Gasser and Laemmli, Cell 46 (1986) 521-530] and the minimal Saccharomyces cerevisiae autonomously replicating sequence consensus [e.g., Palzkill and Newlon, Cell 53 (1988) 441-450].

Structure and evolution of a human erythroid transcription factor

Vertebrate erythroid cells contain a tissue-specific transcription factor referred to as Eryf 1 (ref. 1), GF-1 (ref. 2) or NF-E1 (ref. 3), for which binding sites are widely distributed in the promoters and enhancers of the globin gene family, and of other erythroid-specific genes. Aberrant binding of the human factor to a mutant site has been implicated in one form of hereditary persistence of fetal haemoglobin (HPFH; ref. 2). The complementary DNAs for both the chicken cEryf 1 (ref. 11) and mouse mEryf 1 (ref. 12) encoding genes have recently been cloned. We report here the cloning of the cDNA for the human Eryf 1 encoding gene. The central third of the hEryf 1 cDNA, containing two 'finger' motifs, is almost identical to that of chicken or mouse. The amino-and carboxy-terminal thirds of the human protein are similar to those of mouse, but are strikingly different from the corresponding domains in chicken. The evidence indicates that these erythroid regulatory factors evolved from a common precursor composed of two distinct kinds of repeated domains, which subsequently evolved at greatly different rates.

Analysis of possible repressor elements in the 5'-flanking region of the human beta-globin gene

Human beta-globin gene expression is confined predominantly to the adult with little or no expression of this gene occurring during embryonic or fetal life. The lack of expression of this gene in embryonic and fetal erythroid tissue could be due to the absence of required positive regulatory factors in these cells or the presence of negative regulatory factors which prevent expression of the adult globin gene. To test the repressor model, we have used a gel electrophoretic mobility shift assay to identify regions in the human beta-globin gene which bind proteins found in K562 cells, a cell line that expresses embryonic and fetal globins but not adult beta-globin. DNA fragments comprising the entire human beta-globin gene were assayed using nuclear proteins from K562 cells, and four regions were found that bind proteins. These are located within the 5'-flanking region, within the first and second introns, and at the 3'-flanking region of the gene. Previous studies have suggested the presence of potential repressor sites 5' of exon 2. For this reason, we examined whether the lack of the binding regions in the 5'-flanking sequence allow expression of the human beta-globin gene in transgenic mice during embryonic life. beta-globin gene expression was confined to adult life, indicating that if a transcriptional repressor is responsible for inactivating this gene in embryonic tissue, it is not regulated solely by sequences upstream from -122 bp in the 5'-flanking region of the human beta-globin gene.

The erythroid-specific transcription factor Eryf1: a new finger protein

The erythroid-specific transcription factor Eryf1 binds to DNA sites within regulatory regions of every member of both the alpha- and beta-globin families in chicken. The distribution of these sites suggests that Eryf1 may serve as a general "switch" factor for erythroid development. We have cloned the cDNA for Eryf1 and show that the corresponding mRNA is present in all erythroid lineages, but is absent from non-erythroid cells. We demonstrate that the cDNA encodes the specific Eryf1 binding activity found in erythrocytes. Eryf1 is a basic 38 kd protein containing a pair of highly similar "fingers" with the motif Cys-x-x-Cys-x17-Cys-x-x-Cys. The amino acid sequences of these regions bear no resemblance to those found in other regulatory proteins with a similar arrangement of cysteine residues. Our evidence suggests, furthermore, that transition metal ions are unusually tightly bound, or may not be necessary for the sequence-specific DNA binding of Eryf1.

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.

Evolution and transcription of old world monkey globin genes

Although apparently inactive in the whole animal, the delta globin genes from three species of Old World monkey (rhesus, baboon and green monkey) are all functional in an in vitro transcription assay. Their activities in vitro are similar to that of the functional human delta gene. A fourth monkey gene, from the colobus monkey, is transcribed approximately fivefold less efficiently than the others. This reduced in vitro activity results from a 20 base-pair deletion, which removes the normal site of mRNA initiation. When the deletion is repaired by site-directed mutagenesis, transcriptional activity increases to the level observed for the other delta genes. We also report the complete nucleotide sequences of the colobus beta, colobus delta, and rhesus delta genes. Sequence comparisons show that the delta and beta genes in the same species have not exchanged genetic information since the divergence of the human and monkey lineages. Phylogenetic analysis of these sequences affirms that the Old World monkey delta genes are evolving more rapidly than their functional counterparts. Moreover, the rate of replacement substitutions has risen to equal that of non-coding DNA, as expected for genes no longer under selective constraint.

In vitro characterization of tissue-specific nuclear proteins preferentially bound to the mouse beta-globin gene during MEL cell terminal differentiation

Using DNA restriction fragments of the mouse beta-globin gene and other promoter-containing DNA fragments (LTR-MMTV and pBR322) as controls, we have characterized by protein blotting, in extracts of mouse erythroleukemia (MEL) cells, specific nuclear DNA binding proteins with a preferential affinity for the beta-globin DNA. Some proteins (110 and 75 kDa) appear in differentiated MEL cells while others (100, 95, and 35 kDa) are present in immature MEL and normal erythroblast cells and bind selectively to the far-upstream region of the gene. These proteins could modulate either positively or negatively the expression of the beta-globin gene and maybe, of other genes, during the terminal differentiation of erythroid cells.

Regulation of erythroid cell-specific gene expression during erythropoiesis

The aim of our group's work over the past few years has been to investigate the molecular mechanisms regulating erythroid cell-specific gene expression during erythroid cell differentiation. In addition to the alpha-globin gene, we have focussed on two non-globin genes of interest encoding the rabbit red cell-specific lipoxygenase (LOX) and the mouse glutathione peroxidase (GSHPX), an important seleno-enzyme responsible for protection against peroxide-damage. Characterisation of the GSHPX gene showed that the seleno-cysteine residue in the active site of the enzyme is encoded by UGA, which usually functions as a translation-termination codon. This novel finding has important implications regarding mRNA sequence context effects affecting codon recognition. The regulation of the GSHPX and red cell LOX genes has been investigated by functional transfection experiments. The 700 bp upstream of the GSHPX promoter seems to function equally well when linked to the bacterial chloramphenicol acetyl transferase (CAT) gene and transfected into mouse erythroid or fibroblast cell lines. However, the presence of tissue-specific DNase I hypersensitive sites (DHSS) in the 3' flanking region of the GSHPX gene suggests that such sites may be important in its regulation in the various cell types in which it is highly expressed, i.e., erythroid cells, liver and kidney. The transcription unit of the RBC LOX gene has also been defined and 5' and 3' flanking regions are being investigated for erythroid-specific regulatory elements: a region upstream of the LOX gene gives increased expression of a linked CAT gene when transfected into mouse erythroid cell lines compared to non-erythroid cell lines.(ABSTRACT TRUNCATED AT 250 WORDS)

Human CCAAT-binding proteins have heterologous subunits

We have characterized three distinct proteins present in HeLa cell extracts that specifically recognize different subsets of transcriptional elements containing the pentanucleotide sequence CCAAT. One of these CCAAT-binding proteins, CP1, binds with high affinity to CCAAT elements present in the human alpha-globin promoter and the adenovirus major late promoter (MLP). A second protein, CP2, binds with high affinity to a CCAAT element present in the rat gamma-fibrinogen promoter. Finally, the third CCAAT-binding protein is nuclear factor I (NF-I), a cellular DNA-binding protein that binds to the adenovirus origin of replication and is required for the initiation of adenoviral replication. CP1, CP2, and NF-I are distinct activities in that each binds to its own recognition site with an affinity that is at least three orders of magnitude higher than that with which it binds to the recognition sites of the other two proteins. Surprisingly, CP1, CP2, and NF-I each appear to recognize their binding site with highest affinity as a multisubunit complex composed of heterologous subunits. In the case of CP1, two different types of subunits form a stable complex in the absence of a DNA-binding site. Moreover, both subunits are present in the CP1-DNA complex. We thus propose the existence of a family of related multisubunit CCAAT-binding proteins that are composed of heterologous subunits.