FRINK Linked Data Fragments server

Matches in SemOpenAlex for { <https://semopenalex.org/work/W1551764037> ?p ?o ?g. }

• W1551764037 endingPage "19650" @default.
• W1551764037 startingPage "19643" @default.
• W1551764037 abstract "Nuclear factor I (NFI) was suggested to be involved in the expression of the human α-globin gene. Two established cell lines, which express α-globin differentially, were therefore compared for differences in binding of NFI at the α-globin promoter in vivo. HeLa cells, in which α-globin is repressed, show a high density promoter occupation with several proteins associated with structurally distorted DNA. Cell line K562, which is inducible for α-globin, surprisingly was found to be heterogeneous consisting mainly of cells (∼95%) unable to express α-globin. However, the promoter of the nonexpressing K562 cells was clearly different from that of HeLa cells, being occupied only at basal transcriptional elements. Therefore, the α-globin gene in these K562 cells may not be truly repressed, but in an intermediate state between repression and active transcription. The NFI site of the α-globin promoter appeared occupied in HeLa but free of proteins in K562 cells. All cells of both cell lines produce NFI, but the composition and DNA binding affinity of NFI species differ significantly between the two cell lines. Therefore, distinct forms of NFI may repress α-globin transcription in HeLa cells. However, NFI is apparently not involved in establishing the latent transcriptional state of the majority of K562 cells. Nuclear factor I (NFI) was suggested to be involved in the expression of the human α-globin gene. Two established cell lines, which express α-globin differentially, were therefore compared for differences in binding of NFI at the α-globin promoter in vivo. HeLa cells, in which α-globin is repressed, show a high density promoter occupation with several proteins associated with structurally distorted DNA. Cell line K562, which is inducible for α-globin, surprisingly was found to be heterogeneous consisting mainly of cells (∼95%) unable to express α-globin. However, the promoter of the nonexpressing K562 cells was clearly different from that of HeLa cells, being occupied only at basal transcriptional elements. Therefore, the α-globin gene in these K562 cells may not be truly repressed, but in an intermediate state between repression and active transcription. The NFI site of the α-globin promoter appeared occupied in HeLa but free of proteins in K562 cells. All cells of both cell lines produce NFI, but the composition and DNA binding affinity of NFI species differ significantly between the two cell lines. Therefore, distinct forms of NFI may repress α-globin transcription in HeLa cells. However, NFI is apparently not involved in establishing the latent transcriptional state of the majority of K562 cells. Expression of α-globin is regulated in vivo by the interplay of the locus control region at −40 kilobases and diverse promoter elements(1Higgs D.R. Wood W.G. Jarman A.P. Sharpe J.A. Lida J. Pretorius I.-M. Ayyub H. Genes & Dev. 1990; 4: 1588-1601Google Scholar, 2Jarman A.P. Wood W.G. Sharpe J.A. Gourdon G. Ayyub H. Higgs D.R. Mol. Cell. Biol. 1991; 11: 4679-4689Google Scholar). Activation of a particular gene in the α-globin cluster is supposed to be achieved by interaction of factors binding to the locus control region and factors binding to promoter and enhancer elements, thereby keeping the chromatin free of histones(3Felsenfeld G. Nature. 1992; 355: 219-224Google Scholar). However, this histone-free state which also correlates with DNase I hypersensitivity (4Gross D.S. Garrard W.T. Annu. Rev. Biochem. 1988; 57: 159-197Google Scholar) bestows upon the globin genes only transcriptional competence. Additional events or factors that bind at the regulatory elements are required for a particular gene to be actively transcribed. In the erythroid lineage, the major specific transcription factor for globin gene expression is GATA-1(5Evans T. Reitman M. Felsenfeld G. Proc. Natl. Acad. Sci. U. S. A. 1988; 85: 5976-5980Google Scholar, 6Martin D.I.K. Tsai S.-F. Orkin S.H. Nature. 1989; 338: 435-438Google Scholar, 7Wall L. deBoer E. Grosveld F. Genes & Dev. 1988; 2: 1089-1100Google Scholar). However, the promoter of the human α-globin gene contains no GATA-1 site but instead basal transcription elements, a possible SP1/α-IRP site(8Kim C.G. Swendman S.L. Barnhart K.M. Sheffery M. Mol. Cell. Biol. 1990; 12: 5966-5985Google Scholar), and a binding site for nuclear factor I (NFI; (9Zorbas H. Rein T. Krause A. Hoffmann K. Winnacker E.-L. J. Biol. Chem. 1992; 267: 8478-8484Google Scholar)). 1The abbreviations used are: NFInuclear factor IDMSdimethyl sulfatePCRpolymerase chain reactionFACSfluorescence-activated cell sorter. nuclear factor I dimethyl sulfate polymerase chain reaction fluorescence-activated cell sorter. NFI was originally isolated from HeLa cells as a host protein required for the efficient replication of adenovirus 2/5 DNA in vitro and in vivo(10Hay R.T. EMBO J. 1985; 4: 421-426Google Scholar, 11Nagata K. Guggenheimer R.A. Enomoto T. Lichy J.H. Hurwitz J. Proc. Natl. Acad. Sci. U. S. A. 1982; 79: 6438-6442Google Scholar). NFI specifically recognizes the DNA consensus sequence 5′-TGG(N6)GCCAA-3′(12Nagata K. Guggenheimer R.A. Hurwitz J. Proc. Natl. Acad. Sci. U. S. A. 1983; 80: 6177-6181Google Scholar, 13Gronostajski R.M. Nucleic Acids Res. 1986; 14: 9117-9132Google Scholar, 14De Vries E. van Driel W. van den Heuvel S.J.L. van der Vliet P.C. EMBO J. 1987; 6: 161-168Google Scholar). NFI binding sites are found in many viral and cellular promoters and enhancers (see (3Felsenfeld G. Nature. 1992; 355: 219-224Google Scholar, 4Gross D.S. Garrard W.T. Annu. Rev. Biochem. 1988; 57: 159-197Google Scholar, 5Evans T. Reitman M. Felsenfeld G. Proc. Natl. Acad. Sci. U. S. A. 1988; 85: 5976-5980Google Scholar, 6Martin D.I.K. Tsai S.-F. Orkin S.H. Nature. 1989; 338: 435-438Google Scholar) and 7-12 in (9Zorbas H. Rein T. Krause A. Hoffmann K. Winnacker E.-L. J. Biol. Chem. 1992; 267: 8478-8484Google Scholar)) suggesting a role of NFI as transcriptional regulator. Most of these genes display tissue specificity in their expression(15Borgmeyer U. Nowock J. Sippel A.E. Nucleic Acids Res. 1984; 12: 4295-4311Google Scholar, 16Chu H.-M. Fischer W.H. Osborne T.F. Comb M.J. Nucleic Acids Res. 1991; 19: 2721-2728Google Scholar, 17Courtois S.J. Lafontaine D.A. Lemaugre F.P. Durviaux S.M. Rousseau G. Nucleic Acids Res. 1990; 18: 57-64Google Scholar, 18deBoer E. Antoniou M. Mignotte V. Wall L. Grosveld F. EMBO J. 1988; 7: 4203-4212Google Scholar, 19Graves R.A. Tontonoz P. Ross S.R. Spiegelman B.M. Genes & Dev. 1991; 5: 428-437Google Scholar, 20Hennighausen L. Siebenlist U. Danner D. Leder P. Rawlins D. Rosenfeld P. Kelly Jr., T. Nature. 1985; 314: 289-292Google Scholar, 21Jose-Estanyol M. Danan J.-L. J. Biol. Chem. 1988; 263: 10865-10871Google Scholar, 22Knezetic J.A. Felsenfeld G. Mol. Cell. Biol. 1993; 13: 4632-4639Google Scholar). However, NFI is a ubiquitous factor, and it is not known whether it can influence transcription in a tissue-specific manner. NFI may be involved in transcription as a ubiquitous factor with specificity provided through association with other, cell-specific factors. Precedents for this mode include, for example, the association of the ubiquitous Jun and Fos with the lymphoid-specific NF-ATp factor(23Jain J. McCaffrey P.G. Miner Z. Kerppola T.K. Lambert J.N. Verdine G.L. Curran T. Rao A. Nature. 1993; 365: 352-355Google Scholar). Alternatively, NFI may act as a cell-specific transcriptional regulator in spite of its ubiquitous expression. The latter view is supported by the observation of the presence of different forms of NFI in different cell types (24Apt D. Chong T. Liu Y. Bernard H.-U. J. Virol. 1993; 67: 4455-4463Google Scholar, 25Goyal N. Knox J. Gronostajski R.M. Mol. Cell. Biol. 1990; 10: 1041-1048Google Scholar, 26Jackson D.A. Rowader K.E. Stevens K. Jiang C. Milos P. Zaret K.S. Mol. Cell. Biol. 1993; 13: 2401-2410Google Scholar). These forms can arise by expression of different NFI genes (24Apt D. Chong T. Liu Y. Bernard H.-U. J. Virol. 1993; 67: 4455-4463Google Scholar, 27Gil G. Osborne T.F. Goldstein J.L. Brown M.S. J. Biol. Chem. 1988; 263: 19009-19019Google Scholar), by differential splicing(28Santoro C. Mermod N. Andrews P.C. Tjian R. Nature. 1988; 334: 218-224Google Scholar), by diverse covalent post-translational modifications(29Yang B.-S. Gilbert J.D. Freytag S.O. Mol. Cell. Biol. 1993; 13: 3093-3102Google Scholar, 30Jackson S.P. Tjian R. Cell. 1988; 55: 125-133Google Scholar), or by heterodimerization (31Chodosh L.A. Baldwin A.S. Carthew R.W. Sharp P.A. Cell. 1988; 53: 11-24Google Scholar). In this context it is of interest whether NFI could contribute tissue specificity in α-globin gene expression. NFI has been, in fact, implicated in the multistep process of transcriptional activation of the human α-globin gene by in vitro(32Jones K.A. Kadonaga J.T. Rosenfeld P.J. Kelly T.J. Tjian R. Cell. 1987; 48: 79-89Google Scholar) and in vivo transient assays with reporter plasmids(28Santoro C. Mermod N. Andrews P.C. Tjian R. Nature. 1988; 334: 218-224Google Scholar, 33Mermod N. O'Neill E.A. Kelly T.J. Tjian R. Cell. 1989; 58: 741-753Google Scholar). These assays revealed a weak but clear stimulation of α-globin transcription following binding of NFI to the promoter sequence. Originally it was thought that stimulation occurs by binding to the general positive cis-acting CCAAT genetic element of this promoter(34Mellon P. Parker V. Gluzman Y. Maniatis T. Cell. 1981; 27: 279-288Google Scholar). It was this assumption which led to the definition of NFI as “CTF” (=CCAAT-box transcription factor) implying a role for NFI as a general transcription factor(32Jones K.A. Kadonaga J.T. Rosenfeld P.J. Kelly T.J. Tjian R. Cell. 1987; 48: 79-89Google Scholar). However, we demonstrated that specific and fairly strong binding of NFI actually occurs at an adjacent previously unrecognized NFI site within the α-globin promoter(9Zorbas H. Rein T. Krause A. Hoffmann K. Winnacker E.-L. J. Biol. Chem. 1992; 267: 8478-8484Google Scholar). Furthermore, in vivo analysis with reporter plasmids suffers from copy number effects and does not account for the influence of chromatin which is known to play an important role in gene transcription via the presence of specific histones, nucleosomes, and higher order structures, such as the 30-nm-diameter chromatin filament, locus boundary elements, and the nuclear matrix or scaffold(3Felsenfeld G. Nature. 1992; 355: 219-224Google Scholar). Chromatin structure is particularly important for NFI binding and function; for example, Lee and Archer (35Lee H.-L. Archer T.K. Mol. Cell. Biol. 1994; 14: 32-41Google Scholar) demonstrated recently that NFI can bind and activate the murine mammary tumor virus promoter from transiently transfected, “naked” plasmid templates, whereas the chromatin version of the same sequence in the same cell is refractory to NFI action. For these reasons, we wondered whether occupation of the NFI site of the α-globin promoter in the chromosomal context could be correlated with a particular transcriptional state of the α-globin gene. This would clarify a possible importance of this site in vivo and provide clues for an implication of NFI in α-globin gene expression and regulation in situ. To start approaching this question, we compared the in vivo footprints of an inducible (K562) and a noninducible (HeLa) cell line. Our present results suggest that the transcriptional state of K562 cells does not correlate with NFI binding to the α-globin promoter in vivo. Our data are also compatible with the hypothesis that NFI species found in HeLa cells could act as repressors of α-globin transcription. K562 cells were purchased from ATCC (CCL 243); HeLa cells were a laboratory stock. Logarithmically growing K562 or HeLa cells suspended in RPMI 1640 medium supplemented with 10% fetal calf serum were harvested by centrifugation (1000 rpm, at 4°C), washed in phosphate-buffered saline, and RNA was isolated according to a standard protocol(36Civelli O. Birnberg N. Herbert E. J. Biol. Chem. 1982; 257: 6783-6787Google Scholar, 37Chirgwin J.M. Przybyla A.E. MacDonald R.J. Rutter W.J. Biochemistry. 1979; 18: 5294-5299Google Scholar). Hemin induction of K562 cells was done with a final concentration of 50 μg/ml for the times indicated in Fig. 1A. Slot blots were performed on a Schleicher & Schuell apparatus (SRC 07210 Minifold II). 30 μg of total RNA in 100 μl of TE were added to 300 μl of 6.15 M formaldehyde, 10 × SSC, denatured for 15 min at 65°C, and transferred on a nitrocellulose filter. The filter was baked at 80°C for 2 h. Specific probes for the α-globin mRNA were oligonucleotides C and E of the first primer set also used in the ligation-mediated PCR. Glyceraldehyde-3-phosphate dehydrogenase specific probes were oligonucleotides 5′-CCAGTGAGCTTCCCGTTCAGCTC-3′ and 5′-CCACCACCCTGTTGCTGTAGCC-3′. They were radioactively labeled at the 5′ ends as described(9Zorbas H. Rein T. Krause A. Hoffmann K. Winnacker E.-L. J. Biol. Chem. 1992; 267: 8478-8484Google Scholar). The sarcosyl technique was used for hybridization(38Overbeek P.A. Merlino G.T. Peters N.K. Cohn V.H. Moore G.P. Kleinsmith L.J. Biochim. Biophys. Acta. 1981; 656: 195-205Google Scholar). Purification of baculovirus-expressed NFI (amino acids 1-257) from infected Spodoptera frugiperda (Sf9) cells is described in (9Zorbas H. Rein T. Krause A. Hoffmann K. Winnacker E.-L. J. Biol. Chem. 1992; 267: 8478-8484Google Scholar). HeLa and K562 whole nuclear extracts were prepared as described (39Dignam J.D. Martin P.L. Shastry B.S. Roeder R.G. Methods Enzymol. 1983; 101: 582-598Google Scholar) with the exception that 1 mM phenylmethylsulfonyl fluoride was added to all solutions. Protein concentration was determined by a standard method (40Bradford M.M. Anal. Biochem. 1976; 72: 248-254Google Scholar) and was between 2.5 and 4 μg/μl. Whole cell extracts from uninfected Sf9 cells were kindly provided by M. Stanglmaier. Oligonucleotides were synthesized by G. Arnold (Laboratory for Molecular Biology-Genzentrum, Martinsried). Oligonucleotide “α-G wt” contains the NFI site in the context of the α-globin promoter(9Zorbas H. Rein T. Krause A. Hoffmann K. Winnacker E.-L. J. Biol. Chem. 1992; 267: 8478-8484Google Scholar). Oligonucleotide L1/2 with a higher affinity NFI site has been described in (41Meisterernst M. Gander I. Rogge L. Winnacker E.-L. Nucleic Acids Res. 1988; 16: 4419-4435Google Scholar). Oligonucleotide “α-G mut.” in which the NFI site has been inactivated is the same as oligonucleotide “k” in (9Zorbas H. Rein T. Krause A. Hoffmann K. Winnacker E.-L. J. Biol. Chem. 1992; 267: 8478-8484Google Scholar). Purification and radioactive labeling of the oligonucleotides and the conditions for protein-DNA incubation are described(9Zorbas H. Rein T. Krause A. Hoffmann K. Winnacker E.-L. J. Biol. Chem. 1992; 267: 8478-8484Google Scholar). The amount of the labeled double-stranded oligonucleotide α-G wt was usually 5 fmol, whole nuclear extracts were 2-4 μl; cold competitors were added before the binding reaction in 100-fold molar excess. In supershift experiments, the binding reaction was on ice, then 1 μl of nonimmune or anti-NFI-antiserum (described in (42Krause A. Heterologous NFI Expression and Production of NFI-specific Antibodies. Ph.D. thesis, Institute for Biochemistry, Ludwig-Maximilians-University, Munich, Germany. 1993; Google Scholar)) was added, and incubation was continued for 15 min. Native polyacrylamide gel electrophoresis (acrylamide:bisacrylamide = 30:0.8) and autoradiography were as in (43Zorbas H. Rogge L. Meisterernst M. Winnacker E.-L. Nucleic Acids Res. 1989; 19: 7735-7748Google Scholar). Conditions for base-specific modification in vitro and piperidine cleavage were as described(44Church G.M. Gilbert W. Proc. Natl. Acad. Sci. U. S. A. 1984; 81: 1991-1995Google Scholar). For in vivo footprinting with dimethyl sulfate (DMS), K562 or HeLa cells were grown as described above. α-Globin induction of the K562 cells was with 50 μg/ml hemin for 24 h. Treatment with α-amanitin was at a final concentration of 10 μg/ml for 1 h. The cells were washed with an isotonic phosphate buffer and incubated at 3 × 107 cells per ml in RPMI containing 0.2% DMS (Merck). The reaction was at room temperature for 2 min and was stopped by adding to the cells 40 volumes of cold phosphate buffer with 2%β-mercaptoethanol and subsequently by removing the medium by centrifugation. Cells remain viable after this treatment as controlled by trypan blue exclusion. The DNA was extracted by a standard protocol and cleaved at modified residues with piperidine. To visualize the DNA sequence, the ligation-mediated PCR method was used essentially as described(45Garrity P.A. Wold B.J. Proc. Natl. Acad. Sci. U. S. A. 1992; 89: 1021-1025Google Scholar). For the extension step with the radioactive primer E, four PCR cycles were performed. The primer set used for analyzing the sense-strand was: primer C, 5′-CAGGAGACAGCACCATGGTGGGTTC-3′; primer D, 5′-GGTGGGTTCTCTCTGAGTCTGTGGG-3′; and primer E, 5′-AGTCTGTGGGGACCAGAAGAGTGCC-3′. The in vivo footprinting experiments were repeated several times; the results are reproducible in the sense that the patterns are exactly the same whenever the same in vivo methylated DNA batch was analyzed and similar, but never contradictory, in the variation of band intensities, when a different DNA batch or a different set of primers was used. Intracellular expression levels of α-globin and NFI were determined by the use of specific antibodies and flow cytometry (FACS) as described previously(46Förster R. Emrich T. Voss C. Lipp M. Biochem. Biophys. Res. Commun. 1993; 196: 1496-1503Google Scholar). Briefly, K562 cells and HeLa cells were washed twice in phosphate-buffered saline, fixed with 2.5% paraformaldehyde (10 min, 4°C), and permeabilized for 4 min at room temperature with 0.0025% digitonin (Sigma) in order to allow intracellular antibody binding. Cells were adjusted to 2 × 106 per ml in staining buffer (phosphate-buffered saline, 4% fetal calf serum, 5 mM EDTA, 0.1% NaN3) and were incubated for 20 min at 12°C with a rabbit anti-α-globin immune serum (Sigma; final dilution: 1:3000) or with protein A-purified rabbit anti-NFI polyclonal antibodies (100 μg/ml)(42Krause A. Heterologous NFI Expression and Production of NFI-specific Antibodies. Ph.D. thesis, Institute for Biochemistry, Ludwig-Maximilians-University, Munich, Germany. 1993; Google Scholar). After 2 washes in staining buffer, cells were incubated with F(ab')2 fluorescein isothiocyanate-conjugated anti-rabbit IgG and IgM antibodies (1:160, Dianova, Hamburg, FRG). Cells were washed twice, counterstained with propidium iodine (5 μg/ml in 4 mM sodium citrate, 0.1% Triton X-100, pH 7.0) and analyzed by flow cytometry (Becton Dickinson, Heidelberg, FRG). In order to define a system for the study of α-globin expression, we used K562 cells, a commonly used erythroleukemic cell line, for the investigation of expression of the globin genes (cf. for example, (47Lumelsky N.L. Forget B.G. Mol. Cell. Biol. 1991; 11: 3528-3536Google Scholar)). K562 cells can be stimulated to actively transcribe the α-globin gene after hemin induction for different periods of time (Fig. 1A). There is a clear difference in the amount of α-globin mRNA in K562 cells compared to the non-α-globin expressing HeLa control cells (at least 12.5-fold more α-globin mRNA at 72 h). However, in the K562 cell population, α-globin mRNA already displays a high uninduced level (0 h) and is induced by a comparatively very low factor of only about 2.5-fold after 72 h. We wondered whether this reflects a uniformly low level of induction of all or most of the K562 cells, or rather a heterogeneous composition of this cell line, with some cells expressing high levels of α-globin and others expressing low levels or no α-globin. We therefore determined α-globin expression in K562 cells by FACS analysis using specific anti-α-globin antibodies; we indeed detected different subpopulations (Fig. 1B). In the uninduced state, the great majority of cells essentially does not express α-globin, whereas a few cells show comparatively high expression; the latter may be the reason for α-globin mRNA being already detectable before induction (see Fig. 1A). After hemin stimulation, the fraction of cells expressing α-globin increases, but does not exceed ∼5% of the overall population. This means that a small number of cells are actually responsible for virtually all of the α-globin expression in this cell line. As expected, no HeLa cell expresses α-globin detectably (Fig. 1B), which parallels the RNA analysis (see above). Accepting that protein levels directly mirror ongoing α-globin mRNA synthesis, these results mean that most of the K562 cells do not actively transcribe the α-globin gene, i.e. their state of α-globin expression is equivalent to that of HeLa cells. Since HeLa cells could not be induced to express α-globin (data not shown), the condition, in which promoter elements of the α-globin gene in these cells are, may reflect the dormant, or fully repressed state. On the other hand, the few K562 cells, which are transcribing α-globin and can even be stimulated (e.g. by hemin; this work), define the active state of the α-globin gene. We wondered whether the α-globin promoter, in the great majority of K562 cells which are silent, is also in a repressed state as in HeLa cells. To compare the promoter structure of α-globin in the two cell lines, we performed in vivo dimethyl sulfate (DMS) footprinting analysis. Typical results are shown in Fig. 2, A and B; all data are summarized in Fig. 3. 2The displayed in vivo DMS protection results with K562 cells were obtained after 24 h of hemin induction. However, equivalent results were obtained with uninduced cells (data not shown). This is not surprising, since the footprints of the K562 genomic DNA are representative for the bulk of the cells (about 95%) which do not express α-globin in either case. To properly interpret the results, we consider band intensities to be altered in vivo only when they are flanked by any two guanosines, the intensity of which is not altered compared to the in vitro signals.Figure 3:Summary of the in vivo footprinting data of the α-globin promoter in K562 and HeLa cells. Symbols are as in Fig. 2. For comparison, a summary of NFI footprints is also displayed, which was obtained by methylation interference analysis of this region in vitro(9Zorbas H. Rein T. Krause A. Hoffmann K. Winnacker E.-L. J. Biol. Chem. 1992; 267: 8478-8484Google Scholar).View Large Image Figure ViewerDownload (PPT) The in vivo footprinting patterns obtained with both cell lines is clearly different from that of the protein-free DNA methylated in vitro. This indicates that, in vivo, several proteins occupy the α-globin promoter in both cell lines (Fig. 2, lanes 1 and 4 or 5 versus lanes 2 and 3). However, the in vivo footprinting pattern is distinct for each cell line (Fig. 2, lane 1 versus lane 4 or 5): HeLa cells generally display a high density occupation of the promoter with proteins, revealed by the DMS-protected guanosines (denoted by lines with open dots in Fig. 2, lane 1, and in Fig. 3), whereas K562 cells show only a minimal occupation of the promoter (Fig. 2, lane 4 or 5, and Fig. 3). Additionally, the promoter in HeLa cells shows a particular region of about 25 nucleotides with many DMS-hypersensitive purine residues (denoted by arrows in Fig. 2, lane 1, and in Fig. 3). Also, two neighboring cytosine residues at two sites (within the α-IRP site and 3′ of it; compare Fig. 3) become methylated to some extent in this region, presumably at the N-3 positions which can occur only after strong distortion of the double strand state of the DNA (cf., for example, “Discussion” in (48Clark L. Matthews J.R. Hay R.T. J. Virol. 1990; 64: 1335-1344Google Scholar), and references therein). For these reasons, we suggest that the DMS hypersensitivity unequivocally indicates a dramatic alteration of the secondary structure at this region of the DNA in HeLa cells. In contrast, unusual reactivity of the DNA bases was not observed in K562 cells. Interestingly, protein binding in K562 cells seems to happen exclusively next to basal transcription elements, such as CCAAT-box, ATA-box and cap-site, as opposed to HeLa cells, in which at least the last two sites appear essentially protein-free (Fig. 2, lane 4 or 5 versus lane 1; see summary in Fig. 3). In summary, despite the equivalence in the expression pattern, the in vivo footprinting data indicate that K562 cells possess an α-globin promoter structure clearly different from HeLa cells. In the latter, the promoter is packed tightly with proteins and the DNA structure is pronouncedly distorted. In contrast, in K562 cells, the promoter shows an “open” chromatin configuration with proteins bound only at distinct sequence elements. Therefore, we suggest that the α-globin gene in the analyzed K562 cells may not be truly repressed, as in HeLa cells, but in an intermediate state between repression and active transcription. The performed in vivo footprint analysis provides the opportunity to examine protein interactions at the NFI site of the α-globin promoter in the chromosomal context of the two cell lines (cf. introduction). In vivo footprints with K562 cells revealed no stable protein occupation of the NFI site (Fig. 2, lane 5), in spite of the fact that the DNA region does not seem to be particularly inaccessible due to tight protein packaging (cf. instead the corresponding region in HeLa cells; previous section). Since clear protection footprints are obtained only if a sufficiently high portion of the DNA site in question is stably occupied, one reason for the lack of NFI footprints in K562 cells could be a low rate or a merely transient protein binding to its site. α-Amanitin has been used to visualize binding of RNA polymerase II in vivo by trapping the enzyme at the promoter(49Wang W. Carey M. Gralla J.D. Science. 1992; 255: 450-453Google Scholar). In an attempt to enhance a hypothetical insufficient factor binding at the α-globin promoter by the same rationale, we therefore performed in vivo footprints after treatment of induced K562 cells with α-amanitin. Again, no occupation of the NFI site was detectable (data not shown). Thus, by the methods used, the NFI site within the α-globin promoter does not become bound in K562 cells. In contrast, binding of the NFI site is clearly evident in HeLa cells, where the first two guanosines of the first half of the NFI site are consistently found to be protected from in vivo methylation (underlined in GGG(N6)GCCAG; see Fig. 2, lane 1, and summary in Fig. 3). However, this protection pattern deviates from DMS footprints of NFI made in vitro ((9Zorbas H. Rein T. Krause A. Hoffmann K. Winnacker E.-L. J. Biol. Chem. 1992; 267: 8478-8484Google Scholar) and (14De Vries E. van Driel W. van den Heuvel S.J.L. van der Vliet P.C. EMBO J. 1987; 6: 161-168Google Scholar); see “Discussion”). 3T. Rein, R. Förster, A. Krause, E.-L. Winnacker, and H. Zorbas, unpublished results. Therefore, it is not possible to diagnose unambiguously whether protection of the NFI site of the α-globin promoter in HeLa cells is due to NFI binding. Nevertheless, it is clear that the NFI site present at the α-globin promoter is utilized differentially in each cell line, and this may point toward a distinct, hitherto unrecognized genetic function of this site in the chromosomal context. All conclusions in the last two sections are based on the analysis of the sense strand. In spite of the application of various experimental conditions for the PCR (use of formamide, deoxynucleotide analogues, dimethyl sulfoxide, different temperatures, and concentrations of compounds), we never obtained interpretable signals from the antisense strand. This is most probably due to the even higher GC content of the DNA upstream of the NFI site, which we believe is responsible for the attenuation of the signals beyond the displayed region also of the sense strand (not shown). Ambiguous annealing of the PCR primers to this region may impede the analysis of the antisense strand. Binding of several eukaryotic proteins to DNA is known to depend on the methylation state of cytosines. Most proteins are inhibited from binding when cytosines are methylated (at so called CpG or HTF islands, (50Bird A. Cell. 1992; 70: 5-8Google Scholar)), but some require methylated cytosines in order to bind(51Lewis J.D. Meehan R.R. Henzel W.J. Maurer-Fogy I. Jeppesen P. Klein F. Bird A. Cell. 1992; 69: 905-914Google Scholar, 52Meehan R.R. Lewis J.D. McKay S. Kleiner E.L. Bird A.P. Cell. 1989; 58: 499-507Google Scholar). Therefore, the distinct footprinting pattern of the α-globin promoter could be, at least in part, due to cell line specific methylation of this region of the genomic DNA. In particular, differential CpG methylation might have been a cause for lack of binding of NFI to the NFI site in K562 cells. The HTF islands of the α1- and α2-globin gene loci have been investigated in several cell types and tissues(53Antequera F. Macleod D. Bird A.P. Cell. 1989; 58: 509-517Google Scholar, 54Antequera F. Boyes J. Bird A. Cell. 1990; 62: 503-514Google Scholar, 55Bird A.P. Taggert M.H. Nicholls R.D. Higgs D.R. EMBO J. 1987; 6: 999-1004Google Scholar). In fact, they have been shown to be unmethylated in K562 cells(55Bird A.P. Taggert M.H. Nicholls R.D. Higgs D.R. EMBO J. 1987; 6: 999-1004Google Scholar), whereas the same loci appear heavily methylated in HeLa cells(53Antequera F. Macleod D. Bird A.P. Cell. 1989; 58: 509-517Google Scholar, 54Antequera F. Boyes J. Bird A. Cell. 1990; 62: 503-514Google Scholar). However, in HeLa cells also, unmethylated sites are apparent at the promoter region, just in front of the α-globin genes(53Antequera F. Mac" @default.
• W1551764037 created "2016-06-24" @default.
• W1551764037 creator A5011028978 @default.
• W1551764037 creator A5041221280 @default.
• W1551764037 creator A5042379499 @default.
• W1551764037 creator A5048232167 @default.
• W1551764037 creator A5084238530 @default.
• W1551764037 date "1995-08-01" @default.
• W1551764037 modified "2023-10-16" @default.
• W1551764037 title "Organization of the α-Globin Promoter and Possible Role of Nuclear Factor I in an α-Globin-inducible and in a Noninducible Cell Line" @default.
• W1551764037 cites W12952247 @default.
• W1551764037 cites W137984322 @default.
• W1551764037 cites W1501771233 @default.
• W1551764037 cites W1534194237 @default.
• W1551764037 cites W1548936220 @default.
• W1551764037 cites W1559431187 @default.
• W1551764037 cites W1602290343 @default.
• W1551764037 cites W1605657905 @default.
• W1551764037 cites W1608277376 @default.
• W1551764037 cites W1620698641 @default.
• W1551764037 cites W1897425192 @default.
• W1551764037 cites W1942467333 @default.
• W1551764037 cites W1947358227 @default.
• W1551764037 cites W1963318377 @default.
• W1551764037 cites W1968013290 @default.
• W1551764037 cites W1969373767 @default.
• W1551764037 cites W1970224035 @default.
• W1551764037 cites W1970234384 @default.
• W1551764037 cites W1975791071 @default.
• W1551764037 cites W1983512144 @default.
• W1551764037 cites W1984441851 @default.
• W1551764037 cites W1992614891 @default.
• W1551764037 cites W1995608753 @default.
• W1551764037 cites W1997086574 @default.
• W1551764037 cites W1997208313 @default.
• W1551764037 cites W2003893812 @default.
• W1551764037 cites W2008982775 @default.
• W1551764037 cites W2010193151 @default.
• W1551764037 cites W2015599928 @default.
• W1551764037 cites W2018458653 @default.
• W1551764037 cites W2021825044 @default.
• W1551764037 cites W2026181894 @default.
• W1551764037 cites W2027703149 @default.
• W1551764037 cites W2035416523 @default.
• W1551764037 cites W2037824747 @default.
• W1551764037 cites W2039826008 @default.
• W1551764037 cites W2041185099 @default.
• W1551764037 cites W2045821588 @default.
• W1551764037 cites W2046655724 @default.
• W1551764037 cites W2060333964 @default.
• W1551764037 cites W2065367918 @default.
• W1551764037 cites W2066156515 @default.
• W1551764037 cites W2068065077 @default.
• W1551764037 cites W2070159701 @default.
• W1551764037 cites W2072393635 @default.
• W1551764037 cites W2072718264 @default.
• W1551764037 cites W2074378056 @default.
• W1551764037 cites W2077903723 @default.
• W1551764037 cites W2080766637 @default.
• W1551764037 cites W2084327277 @default.
• W1551764037 cites W2085206941 @default.
• W1551764037 cites W2085492596 @default.
• W1551764037 cites W2086693806 @default.
• W1551764037 cites W2089019735 @default.
• W1551764037 cites W2090554343 @default.
• W1551764037 cites W2100475900 @default.
• W1551764037 cites W2100726754 @default.
• W1551764037 cites W2101448549 @default.
• W1551764037 cites W2101735825 @default.
• W1551764037 cites W2109030204 @default.
• W1551764037 cites W2111885793 @default.
• W1551764037 cites W2119956700 @default.
• W1551764037 cites W2137705880 @default.
• W1551764037 cites W2156510371 @default.
• W1551764037 cites W2157569342 @default.
• W1551764037 cites W2177510297 @default.
• W1551764037 cites W2240323633 @default.
• W1551764037 cites W2576734941 @default.
• W1551764037 cites W3164715 @default.
• W1551764037 cites W4238941978 @default.
• W1551764037 cites W4239139991 @default.
• W1551764037 cites W42654338 @default.
• W1551764037 cites W4293247451 @default.
• W1551764037 cites W73481865 @default.
• W1551764037 doi "https://doi.org/10.1074/jbc.270.33.19643" @default.
• W1551764037 hasPubMedId "https://pubmed.ncbi.nlm.nih.gov/7642653" @default.
• W1551764037 hasPublicationYear "1995" @default.
• W1551764037 type Work @default.
• W1551764037 sameAs 1551764037 @default.
• W1551764037 citedByCount "26" @default.
• W1551764037 countsByYear W15517640372012 @default.
• W1551764037 countsByYear W15517640372013 @default.
• W1551764037 countsByYear W15517640372014 @default.
• W1551764037 countsByYear W15517640372016 @default.
• W1551764037 countsByYear W15517640372019 @default.
• W1551764037 countsByYear W15517640372021 @default.
• W1551764037 crossrefType "journal-article" @default.
• W1551764037 hasAuthorship W1551764037A5011028978 @default.

• next