characterization of the dnase i hypersensitive site 3' of the human

Characterization of the DNase I Hypersensitive Site 3‘ of the Human @ Globin Gene Domain

By Donald E. Fleenor and Russel E. Kaufman

The members of the human globin gene family are flanked by strong DNase I hypersensitive sites. The collection of sites 5‘ to the 6 globin gene is able to confer high levels of expression of linked globin genes, but a function has not been assigned to the site 3’ to the @ globin gene (3’HSI). Our analysis of this DNase I super hypersensitive site shows that the region is composed of multiple DNase I sites. By examination of the DNA sequence, we have determined that the region is very A/T-rich and contains topoisomerase II recognition sequences, as well as several consensus

HE HUMAN @ GLOBIN gene domain contains five T known transcriptionally active genes that are regulated in a tissue and developmental stage-specific manner. The genes are distributed over 50 kb of DNA, all of which has been cloned and sequenced. The analysis of globin gene mu- tants and the study of @ globin transgenes have identified regulatory elements within and immediately surrounding globin genes. Antoniou et all have identified at least three separate regulatory elements in the @ globin gene required for the appropriate expression of human @ globin transgenes: a positively acting globin-specific promoter element, an enhancer located within the 3’ half of the gene, and an enhancer located approximately 800 bp downstream ofthe gene. How- ever, f l globin transgenes containing these regulatory elements are expressed in transgenic mice at a much lower level than the endogenous @ globin gene, with no apparent correlation to gene copy number.’ Grosveld et a13 showed that the inclusion of sequences flanking the @ globin gene domain con- fers completely regulated, high-level expression of @ globin transgenes, independent of their site of integration. These flanking sequences were originally described as a series of erythroid-specific, DNase I super-hypersensitive site^.^-^ Tal- bot et a1’ have shown that sequences necessary for this regulated expression reside within a 6.5-kb locus control region (LCR) containing four upstream hypersensitive sites, whereas Forrester et a1* further delimited the required regulatory sequences into a 2.5-kb cassette. Within the LCR, SHS2 exhibits potent enhancer activity in transient expression studies.’ Analyses of this region have shown that the tandem recognition sequence for AP-l/NF-E2 is sufficient to confer the enhancer activity. “J

Whereas transgenic studies have shown the activating properties of the DNase I hypersensitive sites within the 5’ LCR, no function has been determined for the 3’ hypersensitive site (3‘HS 1). Although initially described as an erythroid- specific site, Dhar et all2 have shown that 3‘HS 1, along with SHS5 and 5’HS2, is present in several nonerythroid, as well as nonhematopoietic cell lines. A possible function of the chromatin in these DNase I super-hypersensitive sites is to organize and maintain the @ globin gene domain in an overall DNase I-sensitive, transcriptionally poised state. In this state, the individual globin genes could be further modulated by factors affecting the transcription of each specific globin gene. In their studies of Ig genes, Cockerill and Garrard13 have proposed that such an organizing function may be performed by scaffold-associated regions (SARs). SARs are proposed to

binding motifs for GATA-1 and AP-IINF-EP. Gel mobility shift assays indicate that the region can interact in vitro with GATA-1 and AP-1 /NF-E2, and functional studies show that the region serves as a scaffold attachment region in both erythroid and nonerythroid cell lines. Whereas many of the physical features of 3’HSI are shared by 5’HS2 (a component of the 5’ locus control region), transient expression studies show that 3’ HSI does not share the erythroid- specific enhancer activity exhibited by 5’HS2. 0 7 993 by The American Society of Hematology.

allow the chromatin to be organized into topologic domains or loops, which are constrained by the nuclear framework. These SARs may act as fixed boundaries, isolating the @ globin domain from the effects, both positive and negative, of neighboring gene complexes. Similar functions have been reported for the A elements surrounding the chicken lysozyme geneI4 and the specialized chromatin structures (SCS) flanking the Drosophila hsp 70 gene.I5 In transgenic studies, these elements appear to protect their respective genes from the effects of adjacent chromatin as they confer position-independent, regulated expression. Jarman and Higgs16 have shown the presence of several SAR sites within the /3 globin gene domain, including a site that maps to 3’HSl.

We have undertaken an examination of 3’HS 1 to determine the potential role of this region in the regulation of the @ globin gene complex. We report here the cloning and initial characterization of the region that has shown potentially important structural features, including SARs, topoisomerase I1 recognition sites, several GATA- 1 binding sites, and an AP- 1 /NF-E2 recognition sequence clustered around the DNase I site. Whereas many of these features are shared by 5’HS2, the major activator of the 5’ LCR, 3’HSl does not possess enhancer activity when linked to a y globin reporter gene. We hypothesize that the DNase I sensitivity of this region may be the result of multiple protein interactions, many of which are found within 5’HS2, yet which function differently in the context of 3’HS 1, perhaps to limit the activation of the 5’ LCR.

MATERIALS AND METHODS

Cloning and sequencing. An EMBL3 library containing size- fractionated BamHI human genomic inserts was prepared according

From the Departments of Medicine and Biochemistry, Duke Uni-

Submitted March 18, 1992; accepted December 18, 1992. Supported by Sickle Cell Center Grant No. P60-HL28391 and by

National Institutes ofHealth Grant No. 5-ROI-DK38699. D.E.F. was the recipient of a Fellowship from the Cooleys Anemia Foundation. R.E.K. is a scholar of the Leukemia Society ofAmerica.

Address reprint requests to Donald E. Fleenor, PhD, Box 3250, Duke University Medical Center, Durham, NC 27710.

The publication costs of this article were defrayed in part by page charge payment. This article must therefore be hereby marked “advertisement” in accordance with 18 U.S.C. section 1734 solely to indicate this fact.

versity Medical Center, Durham, NC.

0 1993 by The American Society of Hematology. 0006-4971/93/81 I O-OOI2$3.00/0

Blood, Vol 81, No 10 (May 15), 1993: pp 2781-2790 2781

For personal use only.on April 12, 2018. by guest www.bloodjournal.orgFrom

http://www.bloodjournal.org/

http://www.bloodjournal.org/site/subscriptions/ToS.xhtml

2782 FLEENOR AND KAUFMAN

to the Stratagene protocol. The recombinant 4 was identified by screening with RK29, a 1.25-kb EcoRI fragment located approximately 20 kb 3' of the p globin gene.I7 Various restriction fragments were subcloned into MI3 vectors using standard procedures. Se- quencing template deletions were generated according to the method described by Dale et a1.I' Oligonucleotide primers were purified by high performance liquid chromatography (HPLC) using a Gen Pak Fax column (Waters, Milford, MA). Sequence reactions used the Sequenase protocol supplied by US Biochemical (Cleveland, OH).

K562 cells were maintained in RPMI containing 10% fetal calf serum (FCS). Cells were pelleted, washed with phosphate-buffered saline (PBS), and resuspended in ice-cold I O mmol/L Tris, pH 7.8,4 mmol/L MgCI2, I mmol/L EDTA, 0.3% NP-40. The cells were dounced with a type B pestle and the nuclei collected by centrifugation for I O minutes at 3.300g. The nuclei were resuspended in the aforementioned buffer, centrifuged, and resuspended in 33 mmol/L HEPES, pH 7.8,50 mmol/L NaCI, 6 mmol/ L MgClz at a concentration of 3 X IO7 nuclei/mL. DNase I was added to 200-pL aliquots of nuclei, for a final concentration ranging from 0 to 12.5 U/mL, and incubated at 37°C for I O minutes. The reaction was stopped by the addition of 8 pL 0.5 mol/L EDTA, 5 pL 20% sodium dodecyl sulfate (SDS), and 2 p L I O mg/mL proteinase K and incubating at 65°C for 30 minutes. The DNA was phenol/chloroform- extracted, ethanol-precipatated, restricted with various enzymes, and examined by Southern analysis.

Scaffold attachment assay. K562 nuclear scaffolds were prepared according to the methods of Mirkovitch et all9 and Gasser and Laemmli," with modifications. A suspension culture of 300 mL at an approximate concentration of 5 X IO5 cellsjmL was collected by centrifugation ( 5 minutes at 900g) and washed 3X with isolation buffer at room temperature. Isolation buffer contains 3.75 mmol/L Tris, pH 7.4,0.05 mmol/L spermine, 0.125 mmol/L spermidine, 0.5 mmol/L EDTA/KOH, pH 7.4, I % (vol/vol) thiodiglycol, 20 mmol/ L KC1,O. 1 mmol/L phenylmethysulfonyl fluoride (PMSF), and 0. I % aprotinin (Sigma, St Louis, MO). The cell pellet was resuspended in 10 mL ice-cold isolation buffer containing 0.1% digitonin and dounced with 20 strokes using a type B pestle. The nuclei were collected by centrifugation ( I O minutes at 9008) and washed twice with ice-cold isolation buffer containing 0.1% digitonin, 0.1 mmol/L CuS04, without EDTA. The nuclear pellet was resuspended in 100 pL ofthe buffer and divided into aliquots of 2 A260 units. The nuclei were warmed at 37°C for 20 minutes. After heating, 1.5 mL of LIS extraction buffer was added at room temperature. LIS extraction buffer contains 25 mmol/L 3,5-diiodosalicylic acid, lithium salt (Sigma), 5 mmol/L HEPES/NaOH, pH 7.4,0.25 mmol/L spermidine, 2 mmol/ L KC1, 2 mmol/L EDTA/KOH, pH 7.4, 0.1 mmol/L CuS04, 0.1% digitonin, 0.1 mmol/L PMSF, and 0.1% aprotinin. After I O minutes of incubation at room temperature, the histone-depleted nuclei were collected by centrifugation of 2,400g for 20 minutes at 4°C. Extracted nuclei were washed five times in digestion buffer containing 20 mmol/ L Tris, pH 7.4, 70 mmol/L NaCI, 20 mmol/L KCI, I O mmol/L MgClz, 0.125 mmol/L spermidine, 0.05 mmol/L spermine, 0.1% digitonin, 0. I mmol/L PMSF, and 0.1% aprotinin. The nuclei were digested with restriction enzymes (1 U/pL) in a total volume of 400 pL at 37°C with occasional agitation for 12 to 16 hours. Pellet and soluble fractions were separated by centrifugation in a microfuge for 2 minutes. The pellet was washed with digestion buffer to remove any residual solubilized fragments. Both fractions were subjected to proteinase K (50 pg/mL) and SDS (1%) treatment at 37°C for 1 hour. The DNA was extracted with phenol/chloroform and ethanol precipitated. Five micrograms of each fraction was electrophoresed on agarose gels, transferred to nitrocellulose, and probed with random- labeled restriction fragments. Hybridization was at 42°C in a solution

DNase I hypersensitivity assay.

containing SO% formamide, 4X SSC, 0.1% SDS, 0.1% NaPPi, I O mmol/L HEPES, pH 7.4, S X Denhardts, 50 pg/mL heparin, and 50 pg/mL salmon sperm DNA.

Gel mobility shift assay. Nuclear extract was prepared from tissue culture cells as described by Dignam et al." Gel shift assays were performed at room temperature in a 20-pL reaction containing I O mmol/L Tris, pH 7.5, 50 mmol/L NaC1, 1 mmol/L dithiothreitol (DTT), 1 mmol/L EDTA, 5% glycerol, 5 pg of nuclear extract, 1 pg poly (dI-dC), and 2 X IO4 cpm end-labeled double-stranded oligonucleotides (0.02 to 0.1 ng). The reactions were electrophoresed on a 0.5X Tris-borate-EDTA/6% polyacrylamide gel.

K562 cells were maintained in RPMI supplemented with 10% FCS. Cells were pelleted and resuspended in RPMI at IO' cells/mL, and 250-pL aliquots were electroporated using the BioRad Gene Pulser (BioRad, Richmond, CA) at a setting of 220 mV, 960 pF, and infinite resistance. Cells were harvested 20 to 24 hours after electroporation and luciferase assays were performed using an LKB 125 I Luminometer. Secreted alkaline phosphatase (SEAP) assays were performed as described.22

Transient luciferase assays.

RESULTS

DNase I hypersensitivity analysis. The region containing 3'HS 1 was cloned as a 16-kb BamHI fragment from an EMBL 3 genomic library (Fig 1). Previous reports differ as to the exact location of the hypersensitive (HS) site within the 16- kb genomic fragment. Grosveld et a13 place the site within the 944-bp Bgl II/HindIII fragment (DFl3) located approximately 20 kb 3' of the p globin gene, whereas Dhar et a123 have localized the site within the adjoining 1,335-bp EcoRI/ Bgl I1 fragment (DF14). Epner et a124 show that the 5.6-kb EcoRI proband is shortened to 4.3 kb, thereby placing the HS site near the Bgl I1 restriction site. We detect the 4.3-kb product, as well as a 5.2-kb product generated by a higher concentration of DNase I (Fig 2A). This would suggest the presence of an additional HS site approximately 400 bp from the 5' EcoRI site. This additional site was confirmed by di- gesting DNase I-treated nuclei with Bgl I1 and probing with

R R B H H R

X x x X

~ I RK29 DF14 DF13 DF26 DF28

DF31

DF22

Fig 1. Map of the genomic clone containing 3'HS1, located approximately 20 kb 3' of the 6 globin gene. Shown are the various subclones used in this study.




CHARACTERIZATION OF 3HS1

& & R B H H R - DF28 probe

5.6 kb

5.2 kb

4.3 kb

t 1.6

B B R I I

6

RK20 probe

2.5 kb

1.6 kb

4 R

4 4 B H

I I DF31 probe

2.27 kb

1.3 kb

900 bp Fig 2. DNase I hypersensitivity analysis of 3’HSl. K562 nuclei were treated with increasing concentrations of DNase 1, the DNA digested with (A) EcoRI, (0) 8g/ II. and (C) €coRI/Hindlll. Shown are the probands and the resulting DNase I digestion products. Arrows indicate the locations of the HS sites.

950 bp

850 bp





RK29 (Fig 2B). In this experiment, the 2.5-kb proband is cleaved to produce a 1.6-kb product, thereby confirming the location of the more 5' site. Our attempts to more accurately map the HS site near the Bgl I1 site suggest that there are at least two HS sites in this area. Figure 2C shows that the 2.27- kb EcoRIIHindIII proband is cleaved by DNase I to produce a broad band of 1.3 kb and two sharper bands of 950 and 850 bp. These data are consistent with there being a HS site on either side of the Bgl I1 restriction site. It is unclear whether we detect an additional 900-bp fragment, which is the result of cleaving at two HS sites. Therefore, these data indicate that super HS site 3'HS 1 is actually the summation of at least three sites.

To examine this region in detail, the entire 2.27-kb EcoRI/ Hind111 fragment (DF31) was subcloned into M13 vectors and sequenced in its entirety. Restriction fragments that flank this region (RK29 and DF22) were also subcloned into M 13 vectors and at least partially sequenced. The genomic sequence of this 3,750-bp region is presented in Fig 3. The previously described DNase I HS sites map to coordinates 1650, 2570, and 2670, An analysis of the sequence data showed several interesting features, perhaps the most notice- able being the high A/T-content. The 3,750-bp genomic sequence contains several localized regions of high A/T content, one of which is approximately 88% A/T over a span of 360 bp. Because genomic sequences of similar composition (A/ T-rich) have been shown to be enriched for SARs and topoisomerase I1 s i t e ~ , ' ~ ' ' ~ , * ~ our analysis of the region began with a search for these elements.

Our assay for scaffold association involved a series of functional studies similar to those described by Mirkovitch et Nuclei isolated from the erythroid cell line K562 were gently extracted with a diiodosalicylic acid (LIS) buffer to remove all histones and many nonhistone proteins, thereby causing the DNA fiber to decondense into loops that remain attached to the proteinaceous scaffold at specific points. After the removal of the LIS, restriction enzymes are added that cleave the majority of the DNA (60% to 70%) from the scaffold (soluble fraction), whereas those fragments bound to the scaffold are collected by centrifugation (pellet fraction). Equal amounts of DNA (by weight) from both fractions are then examined by Southern analysis for their differential enrich- ment of particular restriction fragments. Whereas most restriction fragments are liberated into the soluble fraction, fragments containing an SAR are predominantly found in the pellet fraction. Because the pellet fraction comprises approximately 30% to 40% of the total DNA, equal amounts of DNA from each fraction would exaggerate the relative abundance of a particular fragment within the pellet fraction. Therefore, we operationally define a restriction fragment containing an SAR as one in which at least 80% of the fragment is found within the pellet fraction.

Figure 4A is a Southern analysis of K562 nuclear scaffolds digested with either Bgl 11, EcoRI, or Xba and probed with RK29 (1.25-kb EcoRI fragment). The probe identified a 2.5- kb Bgl I1 fragment within the pellet fraction, as well as a I .25- kb EcoRI fragment and a 1.26-kb Xba fragment, both of which were liberated into their respective soluble fractions.

SARs.

These data suggest that the 2.5-kb Bgl I1 fragment is bound to the nuclear scaffold due to an SAR within the 1.3-kb Xba/ Bgl I1 sequence. To extend the SAR analysis 3' of RK29, an identical Southern blot was probed with the 1,334-bp EcoRI/ Bgl I1 fragment (DF14). The probe again identified the 2.5- kb Bgl I1 fragment, as well as the 5.6-kb EcoRI fragment, and the 1.3- and 0.88-kb Xba fragments within their respective pellet fractions (Fig 4B).

The 3' boundary of the SAR was determined by HindIII digestions of K562 nuclear scaffolds probed with DF26 or DF28 (Fig 4C). The Southern analyses indicate that the 2.4- kb HindIII fragment is found within the pellet fraction, whereas the adjacent 4.2-kb HindIII fragment is not signifi- cantly enriched within the pellet fraction, and is therefore not considered to contain an SAR. These data would therefore indicate that scaffold attachment points reside within a 4.5- kb region of genomic DNA defined as being flanked by the subclones RK29 and DF28. Our examination of scaffold association in a nonerythroid cell line (HeLa) shown not to contain 3'HSI shows a similarly bound region. These data suggest that attachment to the nuclear scaffold is not sufficient for the formation of the DNase I hypersensitive site and that its formation is a consequence of factors not present in HeLa cells.

Previously characterized SARs have been shown to be A/T-rich sequences associated with various structural and regulatory elements. For example, many of the attachment regions are enriched for sequences related to the consensus topoisomerase I1 cleavage se- q u e n ~ e . ' ~ , * ~ , ~ ~ This is potentially significant as topoisomerase I1 is a major structural component of the nuclear scaffold and may constitute part of the recognition requirements of SARs. As well as a structural role in DNA attachment, topoisomerase I1 may have an enzymatic role in manipulating the supercoiled state of chromatin loops, which may be important in both replication and gene expression. Therefore, the sequence surrounding 3'HS 1 was examined for sequences related to the vertebrate topoisomerase I1 consensus sequence.26 Two sequences that are 88.9% homologous (16 of 18) to the topoisomerase I1 consensus sequence were identified within the 3,750-bp region. These sequences are indicated by bold underlining in Fig 3.

In addition to topoisomerase sites, tissue-specific transcriptional enhancers have also been found in close association with SARs. 16*27 A comparison of previously identified erythroid-specific enhancers located 5' of the t globin gene' and 3' of the *-y globin2' and p globin shows no extensive sequence homology with which to compare with our sequence. However, each of the globin gene enhancer regions contains at least two copies of the GATA-I binding site m~t i f .~ ' .~ ' Transcriptional activation by GATA- 1 has been implicated in studies involving enhancers located 3' of the chicken and human /3 globin gene^.^',^^ An examination of our sequence data showed that the GATA- I recognition sequence is represented nine times within the 3,750-bp region. The GATA sites are indicated by bold type in Fig 3.

To determine whether these sites can function as GATA- I binding sites in vitro, we performed a series of mobility

Elements associated with SARs.




CHARACTERIZATION OF 3HS1

TCTAGAGAATTCTAAAAATCCACAGATTTAGAAATTCTAGGAATCCAAGAAATTACAATTTTAGATCTAC TACAATTTCACAATATAACATAACCAATAAACACGGACATTTTATTTACAT~TAAAAGATGGGTTTG TCACACAATTTAAAAATCTAGGGTCGAAAGCCACAGCAAAGGCAACCTCCTGGCACTATATCACAATTAGA TGACACCTGCATAGAGAAAGCAAAGTGGTCATGTTAGTCACTGA~GCCTC~T~CCCTGTGGGGATG I TCTACTCTTTTGTATTCCATCAGCTTTTTTCCCAGTAACAACACATCCTAGCAATAATATCCTGACGAGT TTCTAT~CTCCCCTCCTAAGATACAGTATCTTTGTGGTTGCATCCCCAGCCATCCCTCA TTTGGTTGAGGTTGGAATTTCAATGGGTGGGTTGAAT~TTCCCTAAACAAACTTGCCATGGTAAAATCA AGGCTAGCATTACTGTTCACACAAGGCACTGTGCACTGTGCTAAGATCCCCATCCTTATAGCAATGGGAGGGCTCAT AGGCAAGTCATGTCATGCAGAGGCCATTTCATCATCCCTTCTTGTCTTTGGGGACTTGGGTCTCTCCGGT TGTATAGGAGATTTCGCCAGGCATTTGAATCCTTCACTGCTAGTCATCCTGGACTCTGTGGTGTACAATA TCCATCTCTCCTATGTTCTGAGGTCAGAATGACCTCTCTCTTAACCAGTTGTCTGGGCCCAGGCTCCATT ATTGATATAGTCATGATCTCCTCTGTTGGGGATGAAGTAGGCAAAGTAAGCAATTTGAGGCACTAATTTACTTCTCAC ATTCTTTTCTTGAACAGAAAGATAGAACTGGAAA~AATACTAGTATAT~TTCAAAATTTTAGTTTTA ATAACATTTAATCAGACATAAAGGATTAATTATGGTAATGTGAATTTCAATAAAT~~~AG~CTAATATA AGTGTAACTGTGTAATATTCATACTTTTTCTGAAGGCTGAAGGCTTTACTAATTTGATATGGCATTACTTTTTATTG C T G C C A A A A C T A T T C T T A T T C C A C T G T G T G G T G A T G A G A A T AGATAGCT~CCCTGAAGCCATAGTAACCCCCTGGAGAAAAATTGGACCTGGAGTCTAGCAGCCTAGGTAT GGGTACTCGATTTCTTAGAAAGGATGCCTTTACAATTTCCTTTATCTT~TAAGGGTATTGAATGTAGAATT CTAGAATTTTCAGAGGACAACTTAAAATATGTGTAATAGTT~AATTATTTATCCTCATAAATTTAACTG TTCATTTTAATATATTTAAGGATGAATTTTTTAAAAGTTGATTTCATAAAAATGGGAATAGAAAGATGGT TCCATAGGCTGACTGAGAGTGTAGAGGAGGGATGGGAAGGGAAAGAAGTTGATCTTCAGTTAGACTAGAG GAATAAGTTTTAGTGATCTCTCACACTGCATAGTGAACACAGTTAATAATTATATTATGTATTTAAATTAA A A A T T G C T A A A A A A T A A A T A T T T T A T G T T C T C A C C A C A A A T T AGCTTGATAGACTCTCTCTACAATGTATATATAGATCAAACATCACATTGTATCCCATAACATATTATA~ ATATTATATATTTATATTATATATTATTATTGTATCCATTAATATATCACTTAT~TTTCCAGGCAAATA AAAAATGTTTTTAAAATATAAATTTATTTGTAACCTCCTTTTACTTTTCTGCTTGGTTTTCTTCTTTCAT TCAGTGTTTACCAGTTTCTTATAGTTAATTTATTTTTTAAGCTGTCTCACATTTTCTGAAGAAAAGGGAAC ATATTAAAGCCAACAAAACAAATACACTATCTTGCATGAGATGAT~TATGTCATGGTACAATCAAAGTAAGCAATGCT ATAAATCTTATAAAAACTTCTCAAATGGTTAGATGGCTACAGTTGAACAGATGGACCATGTCATATATTT TTTATAATGCTTCTAAGGTATGGCTAATTTTTT~TATTTTAGTAATGATGGGAATTATTATTTATAG AAATCTTATAAAATATATAATG~TATGTAAT~GTCTAGAT~TGTGTATATACATAATATATATT TATTACATAATATATAATATATAATGTATATTTATATATTACATGCATTATATATTAAAGGATTATAATACATT TTATATATTATATATTAATGTAATATGTTATTAAAGGATTATATACAATAATCTA~ACATTTTATGCTT ATATAATATATAATAAATATATAGTATATAATAAATATACACTATATATTTGTATCTATATATGTTTATA AAGTCATTCCTCTAATTAGGTCATAACCAGTTCAGGT~CTGG~TTTAAGCCTACTTCAGG~TGTG GTAAATAGATTCTCTCTGAACTAGCATATTCAGAATCATTAAAGGATCAGTCAGTTC~TGGACAAGTCTTATA GAATGTTCTTACCTCTTCAGCCATCCCAAGACTCTTGAGGGCCTGACCTCGCTTACACT~GCAGATCT GCCTTATGCATCACTGAAGTAGGGAGGGAAGAAAG~TGATGAA~ACTT~GACCCCTAGTGGTGTCCA GAAAAGACCATTAAAGGAATGACCTTTAAAGGATGGATGGACATACAATTTTTTGTCCAAGGCAGGACATGTGT GGGTGTCTTTCAGTAATTATGTTCTAAGAACAGC~CTCCACTGCCTTGGC~TAGGAATGTTTTA GTTCTATAGAATTATAAAGGATGAAGCTGTCTTTTAAACACACAATATACTTTCTCTATGTCTTTGGAAC~TGAC TATTGGTCATTACCCTATTTTAAAGTAAGCAAGTAAGCAAGTAATCACACAGGGAATTA~CTGAAAAGACAG~ AAAAAAAACCAAGAGATTTCTGCATATGTAGGTCAGTTTTAATCAGAGGGCATCAG~GACTCCTAAA AGAATGACCTGGTTATTATAATCACAGATTTGCGTTTTCCAAGTCAACATTCCAGACAGTGCTCAGAGGGG A T A C G A A A A C C C T T T T A T T T C T C C A G A C T C A A A T T C A C T G GGCATTGTTCTGGTTGCTGGGAACTCAGACTCAGACTGAGATACCATACACTGA~CTCAGATAGCATAAGACAAC ATGATGTCTTGGAAAACTGTAAAT~TTTTGT~TT~TACAGGTGGAGCATCTGGCACACCTGACA TATTGATCTTGTTTTTCTTTAAATCTTCATTTATTTATTTACCTTATC~CTATGCTCTTTCATCCTACCTT TCAAAACATATTTTAAAAAATCCTCCAACATGTATTTTGCTCTGGTAATCCCAAAAGGCTGATA~TCTCT ATGGTGGCAACATGGATAATACTGTTCCCCCATCTAGATGGTCTCA~CTTCTGTAT~AGT~G~GAA G C C T G A A T G A A A G T A G A T T T T T A A G C T T T G T A G C T A G T C T C C T GCATGAAAATAGATTTTTTTTTTCCTTTGGGGACAGAGT~GCTCTGTCGCCCAGACTGGAGTGCAATGG CGCGATCTCGGCTCACTGCAACTTCCACCTCCCAGGATC~GC~TT~CCTGCCTCAGTCTCCCAAGTA ACTGGGATTACAGGAGCACACTGCCATGCCCAGCTAATTA

2785

70 140 2 10 280 350 420 490 560 630 700 770 840 910 980 1050 1120 1190 1260 1330 1400 1470 1540 1610 1680 1750 1820 1890 1960 2030 2 100 2170 2240 2310 2380 2450 2520 2590 2660 2730 2800 2870 2940 3010 3080 3 150 3220 3290 3360 3430 3500 3570 3640 3710 3750

Fig 3. Sequence of genomic DNA containing 3’HSI . The AP-l/NF-EZ consensus sequence is underlined. Bold underlining indicates the topoisomerase II consensus sequences. The GATA-1 consensus sites are in bold type.





A Bgl Eco Xba D C D C D C

B Bgl ECO Xba D C D S P S

2.5

1.25

2.5 '

1.3

0.88

Hind P S

2.4

I

RK29 DF14 DF26 DF28

Eigl It

Eo0 RI

Xba I

RK29

Eco RI Hind 111

m m DF14 DF26 DF28

Fig 4. Scaffold association analysis of the region surrounding 3'HSl. Nuclear scaffolds were digested with various restriction enzymes and the DNA examined by Southern analysis as described in Materials and Methods. The Southern blots were probed with RK29 (A), DF14 (B). and DF26 or DF28 (C) . Solid boxes indicate restriction fragments enriched within the pellet fraction, whereas open boxes represent fragments liberated into the soluble fraction.

shift assays in which we compared one of our potential sites selected at random with known GATA-I binding sites located within the " y globin promoter. Double-stranded oligonucleotides prepared from the potential GATA- I site at position 2 135-2 140 and from the double GATA-I sites flanking an octamer binding site located at - 195 to - I69 of the " y globin promoter were end-labeled and incubated with either K562 or HeLa nuclear extract in the presence of various cold competitors.

Martin et al" have shown that the incubation of labeled double-stranded oligonucleotides containing the GATA- 1 binding motif from the " y globin promoter with K562 nuclear extract produces the formation of two specific complexes. The upper complex results from the binding of the ubiquitous octamer-binding protein, whereas the more rapidly migrating complex arises from the binding of GATA-I (Fig 5, lanes 2 and 3). Incubation of the putative GATA-I binding site from 3'HSI (hereafter designated 3'-AGA) with K562 nuclear extract results in the production of a single. specific complex that comigrates with the "y/GATA-1 complex (Fig 5, lanes 6 and 7). The inclusion of the unlabeled " y GATA-I motif oligonucleotides reduced the formation of the complex, suggesting that the complex was formed by the binding ofGATA- 1 (Fig 5, lane 8). Further evidence that 3'-AGA does indeed

bind GATA- I comes from the reciprocal experiment in which unlabeled 3'-AGA competed away the formation of the " y / GATA-I complex. while leaving the complex produced by the "yloctamer intact (Fig 5. lane 4). Therefore. it appears that GATA-I can interact in vitro with at least one of the GATA binding sites clustered within 3'HSl.

In addition to the multiple GATA-I binding sites. an AP- I/NF-E2 DNA sequence motif was detected at nucleotide position 1410. Previously, Ney et all0 have identified tandem AP-I/NF-E2 binding sites within 5'HS2 that function as an inducible enhancer in erythroid cells. The NF-E2 binding site is an AP-I site preceded by a GC dinucleotide. We performed gel shift experiments to determine whether the AP- I/NF-E3 motif within 3'HSI does indeed participate in a similar DNA/protein interaction. Double-stranded oligonucleotides representing the site were incubated with K562 or HeLa cell nuclear extract in the presence of various cold competitors (Fig 6). Incubation of the AP-I/NF-E2 oligos (3'AP-I/NF-E2) with K562 nuclear extract results in the formation of a major complex that is specifically competed away by the addition of the identical cold competitor (Fig 6. lanes 2 and 3). The inclusion of a cold competitor representing the tandem AP-I/NF-E2 sites within 5'HS2 inhibits the formation of the labeled complex. suggesting that binding by similar




CHARACTERIZATION OF 3HS1 2787

octamei

GATA 1

Fig 5. Gel mobility shift analysis of a represen- tative GATA-1 binding site motif within 3’HSl. The sequences of the double-stranded oligonucleotides representing the binding sites are: 3’-AGA, 5’TA- CACATTTATCTAGACTTG3’; and pro, B’GACAGATATTTGCATTGAGATAGTGY. lanes (probe, extract, cold comp): 1 : pro, -. -; 2: A~

pro, K562, -; 3: pro, K562, pro; 4: pro, K562.3’AGA; 5: 3’-AGA, -, -; 6: 3’-AGA, K562, -; 7 : 3’-AGA, K562, 3’AGA; 8: 3’-AGA, K562, pro.42

factors accounts for its formation (Fig 6. lane 4). The addition of an unlabeled double-stranded oligonucleotide representing a consensus CRE element, an element whose recognition sequence is similar to that of AP-I,33 did not inhibit the formation of the complex (Fig 6. lane 5). If the shifted complex is a consequence of AP- I binding. one would expect to detect such binding using extracts prepared from other cell types. This was shown to be the case as an identical result was ob- tained using HeLa cell nuclear extract (data not shown). Therefore, we cannot determine if this site binds NF-E2 or only AP-I in vitro.

Our analysis of 3’HSI also showed the presence of four GGTGG motifs (coordinates 446, 3266. 3433, and 3665). two ofwhich are interspersed within a cluster of three GATA motifs. GGTGG motifs within SHS2 and SHS3 have been reported to footprint using in vitrd4..” and in vivo tech- niques.36 Although we have no evidence that these sites within 3’HSI do indeed interact with nuclear proteins, it is interesting to note that the collection of GATA-I. AP-I/NF-E2, and GGTGG motifs are shared by many of the HS sites surrounding the @ globin gene domain.

Because many of the physical features of 3‘HS I are shared by 5’HS2, the component of the 5’ LCR shown to possess enhancer activity, we sought to determine whether 3‘HS I also exhibits enhancer activity. The 2.27-kb EcoRI/HindIII fragment (DF3 1) was cloned upstream of a y globin-luciferase reporter gene and its activity was assayed transiently in K562 cells (Fig 7). Transformation efficiencies were normalized by cotransfecting a plasmid that expresses secreted alkaline phosphatase.” Plasmids containing the 3’HSI region in either orientation showed no increase in luciferase activity relative to the y globin-luciferase parent plasmid. whereas the positive

control containing SHS2 showed a 50-fold increase in luciferase activity. Therefore, even though the two hypersensitive sites share common features. 3’HSI does not share the enhancer activity exhibited by 5’HS2.

DISCUSSION

Our examination of the DNase I super hypersensitive site located approximately 20 kb 3’ of the human @ globin gene (3’HSI) shows that the site is actually a composite of at least three HS sites surrounding a very A/T-rich region of DNA. The unusually high A/T content may allow for DNA con- formations that are more susceptible to DNase I digestion. Alternating purine-pyrimidine sequences have been shown to form cruciform structures”’ and Z-DNA,-’* and to exclude nucleosome formation,39 potentially resulting in a more open chromatin configuration. It is possible that the high A/T composition, and not any particular sequence. allows the DNA to form a more open configuration, and thereby become more susceptible to DNase digestion. However, the absence of a DNase I hypersensitive site at this position in HeLa cells would suggest that any structural conformation achieved solely on the basis of the DNA sequence is not sufficient to generate the site. Our data do not establish whether there is an erythroid-specific component to this DNase I site because recent studies indicate that it is not seen exclusively in erythroid cells.12 We hypothesize that the formation of the site requires the interaction of several trans-acting factors, and that some of them may be unique to erythroid lineages.

Within the genomic sequence surrounding 3‘HSl. we have identified sites of potential protein interaction. both on the basis of consensus recognition sequences for DNA binding factors and on experimental studies that show DNA/protein





I 2 3 4 5 Y luc 1 .o

Fig 6. Gel mobility shift analysis of the AP-lINF-EP binding site motif within 3'HSl. The sequences of the double-stranded oligonucleotides, with the binding motifs underlined, are: 3'AP-l/NF-E2, S'TAGGCTGACTGAGAGTGTAGAGGAGGGATG3'; 5'AP-l/NF-E2. 5'TGCTGAGTCATGATGAGTCATGC3'; CRE, S'CTAGACCCClTACGTCAGAGGCG3'. Lanes (probe, nuclear extract, cold comp): 1 : 3'AP-1 /NF-E2, -, -; 2: 3'AP-1 /NF-E2, K562, -; 3: 3'AP-l/NF-E2. K562, 3'AP-l/NF-E2; 4: 3'AP-l/NF-E2, K562, 5'AP-l/NF-E2; 5: 3'AP-1 /NF-E2, K562, CRE.'*

interactions (Fig 8). The genomic sequence was shown to contain the consensus recognition sequences for topoisomerase 11, GATA-I. and AP-I/NF-EZ, as well as the GGTGG motif. Functional studies demonstrate that the region serves as a scaffold attachment point, that at least one of the GATA sites can bind GATA-I in vitro. and that factor(s) binds to an AP-l/NF-E2 site within the region. Although the results from the gel mobility shift experiments d o not necessarily reflect the in vivo situation. these data are consistent with previous studies that have shown close associations between SARs, topoisomerase recognition sites. and enhancer ele-

The initial characterization of 3'HSI suggested that it was erythroid specific. A more extensive analysis indicates that it is present in cell lines of nonerythroid origin, but its sus- ceptibility to DNase I is variable. Because scaffold association is not erythroid specific. it may play a significant role in gen- erating the DNase I hypersensitive site. Likewise. the interactions of topoisomerase 11. or the binding of AP-I, both ubiquitous'proteins. may contribute to the formation of the site. Studies on the erythroid promoter of the porphobilinogen deaminase (PBGD) gene demonstrate the binding of a distinct erythroid-specific protein. NF-EZ. to a recognition sequence

ments. 13. I h.Z(l.2S.27

2.3 RHYluc 0.98 2 0.28

HS2 Y luc 49 2 18 Fig 7. Transient luciferase activity of various 7-luciferase plas-

mids in K562 cells. Transfection efficiencies were normalized by cotransfecting a CMV-SEAP plasmid and assaying for secreted alkaline phosphatase. Corrected values are presented as a comparison to the ?-luciferase plasmid.

shared by AP-I. Methylation interference and point muta- genesis experiments show that AP-I and NF-E2 have slight differences in their binding sequence specificities. Within the NF-E2 recognition sequence GCTGAGTCA. the 5' G residue appears critical for NF-E2 binding.4".4' This consensus NF- E2 recognition sequence. along with the internal AP-I binding motifare present in 3'HSI and may therefore serve as a binding site for NF-E2. Although our gel shift analysis of the region suggests the binding of the ubiquitous AP- I , this does not preclude the interaction of NF-E2 with the overlapping site. In addition to possible contributions by NF-EZ, we pro- pose that a clustering of GATA-I binding sites within this genomic region, although perhaps not sufficient by them- selves, may contribute to some erythroid-specific component ofthe hypersensitive site. Such a clustering of GATA-I binding sites may serve as a repository of transcription factors for the p globin gene domain.

The region may function to organize the domain by serving as a torsion absorber through its scaffold attachment points and topoisomerase sites to relieve supercoiling created during replication or transcription. Such a structure flanking the domain may also function by damping the effects of adjacent regulatory elements. A similar situation has been reported for the chicken lysozyme gene in that an SAR placed between

DNase I HS Sites

1 11 - An-rich

B H I I I

0 . . 0 . . 0.0 A A A A <

Fig 8. Features of 3'HSl. Shown are the locations of the DNase I HS sites, the A/T-rich region, and sequence motifs for (t) AP-l/NF- €2. (0) GATA-1, (A) GGTGG, and (e) topoisomerase II.




CHARACTERIZATION OF 3HS1 2789

the gene and its enhancer greatly diminished the effect of the enhancer.14 Within the 0 globin gene domain there is mount- ing evidence that regulatory elements exert their influence over many kilobases of DNA; therefore, it may be important to eliminate regulatory influences from adjacent genes. In addition to protecting the globin domain from outside influences, it may be equally important to limit the activation potential of the 5‘ LCR to members of the gene domain. In particular, as attempts to introduce globin genes linked to the LCR into human cells proceed, it may be important to limit the influence of the strong enhancer element to prevent activation of genes close to the DNA insertion site.

ACKNOWLEDGMENT

The authors thank Mary Watson, Sara Owens, and Katharina Ur- ban for excellent technical assistance and Bobbie Reeves for manu- script preparation.

REFERENCES

1. Antoniou M, deBoer E, Habets G, Grosveld F The human p- globin gene contains multiple regulatory regions: Identification of one promoter and two downstream enhancers. EMBO J 7:377, 1988

2. Townes TM, Lingrel JB, Chen HC, Brinster RL, Palmiter RD: Erythroid-specific expression of human beta globin globin genes in transgenic mice. EMBO J 4: 17 15, 1985

3. Grosveld F, van Assendelft GB, Greaves DR, Kollias G: Position independent, high level expression of the human beta globin gene in transgenic mice. Cell 51:975, 1987

4. Tuan D, Solomon WB, Li QL, London IM: The beta like- globin gene domain in human erythroid cells. Proc Natl Acad Sci USA 82:6384, 1985

5. Forrester WC, Thompson C, Elder JT, Groudine M: A developmentally stable chromatin structure in the human beta gene cluster. Proc Natl Acad Sci USA 83:1359, 1986

6. Forrester WC, Takegawa S, Papayannopoulou T, Stamatoy- annopoulos G, Groudine M: Evidence for a locus activation region: The formation of developmentally stable hypersensitive sites in globin- expressing hybrids. Nucleic Acids Res 15: 10159, 1987

7. Talbot D, Collis P, Antoniou M, Vidal M, Grosveld F, Greaves DR: A dominant control region from the human beta globin locus confemng integration site-independent gene expression. Nature 338: 352, 1989

8. Forrester WC, Novak U, Gelinas R, Groudine M: Molecular analysis of the human beta globin locus activation region. Proc Natl Acad Sci USA 865439, 1989

9. Tuan DYH, Solomon WB, London IM, Lee DP: An erythroid- specific, developmental-stage-independent enhancer far upstream of the human “beta-like globin” genes. Proc Natl Acad Sci USA 86: 2554, 1989

10. Ney PA, Sorrentino BP, McDonagh KT, Nienhius AW: Tan- dem AP-1 binding sites within the human beta globin dominant control region function as an inducible enhancer in erythroid cells. Genes Dev 4:992, 1990

1 1. Sorrentino B, Ney P, Bodine D, Nienhius A W A 46 base pair enhancer sequence within the locus activating region is required for inducted expression of the gamma-globin gene during erythroid dif- ferentiation. Nucleic Acids Res 18:2721, 1990

12. Dhar V, Nadi A, Schildkraut CL, Skoultchi AI: Erythroid- specific nuclease-hypersensitive sites flanking the human beta-globin domain. Mol Cell Biol 10:4324, 1990

13. Cockerill PN, Garrard WT: Chromosomal loop anchorage of the kappa immunoglobulin gene occurs next to the enhancer in a region containing topoisomerase I1 sites. Cell 44:273, I986

14. Stief A, Winter DM, Stratling WH, Sippel A E A nuclear DNA attachment element mediates elevated and position-independent gene activity. Nature 341:343, 1989

15. Kellum R, Schedl P A position-effect assay for boundaries of higher order chromosomal domains. Cell 64:941, 1991

16. Jarman AP, Higgs DR: Nuclear scaffold attachment sites in the human globin gene complexes. EMBO J 7:3337, 1988

17. Kaufman RE, Kretschmer PJ, Adams JW, Coon HC, Ander- son WF, Nienhuis AW: Cloning and characterization of DNA sequences surrounding the human gamma, delta, and beta globin genes. Proc Natl Acad Sci USA 77:4229, 1980

18. Dale RMK, McClure BA, Houchins J P A rapid single-stranded cloning strategy for producing a sequential series of overlapping clones for use in DNA sequencing: Application to sequencing the corn mi- tochondrial 18s rDNA. Plasmid 13:31, 1985

19. Mirkovitch J, Mirault M, Laemmli U K Organization of the higher order chromatin loop: Specific DNA attachment sites on nuclear scaffold. Cell 39:223, 1984

20. Gasser SM, Laemmli U K Cohabitation of scaffold binding regions with upstream/enhancer elements of three developmentally regulated genes of D. melanogaster. Cell 4652 1, 1986

2 I . Dignam J, Lebovitz R, Roeder R: Accurate transcription ini- tiation by RNA pol I1 in a soluble extract from isolated mammalian nuclei. Nucleic Acids Res 2:1476, 1983

22. Berger J, Hauber J, Hauber R, Geiger R, Cullen BR: Secreted placental alkaline phosphatase: A powerful new quantitative indicator of gene expression in eukaryotic cells. Gene 66:1, 1988

23. Dhar V, Mager D, Iqbal A, Schildkraut C L The coordinate replication of the human beta globin gene domain reflects its transcriptional activity and nuclease hypersensitivity. Mol Cell Biol 8: 4958, 1988

24. Epner E, Forrester WC, Kim C, Telling A, Envir T, Brice M, Papayannopoulou T, Groudine M: Chromatin structure and DNA replication patterns in normal and mutant beta globin gene loci, in Stamatoyannopoulos G, Nienhuis AW (eds): Regulation of Hemo- globin Switching, Proceedings of the Seventh Conference on He- moglobin Switching. Baltimore, MD, Johns Hopkins University, 1991, p 153

25. Gasser SM, Laemmli U K The organization of the chromatin loops: Characterization of a scaffold attachment site. EMBO J 5:5 1 1, 1986

26. Spitzner JR, Muller MT: A consensus sequence for cleavage by vertebrate DNA topoisomerase 11. Nucleic Acids Res 165533, 1988

27. Kollias G, Wrighton N, Hurst J, Grosveld F: Regulated expression of human gamma, beta, and hybrid globin genes in transgenic mice: Manipulation of the developmental expression patterns. Cell 4639, 1986

28. Bodine DM, Ley TJ: An enhancer element lies 3’to the human A gamma globin gene. EMBO J 6:2997, 1987

29. Trudel M, Constantini F: A 3’ enhancer contributes to the stage specific expression of the human beta globin gene. Genes Dev 1:954, 1987

30. Wall L, deBoer E, Grosveld F: The human beta globin gene 3’ enhancer contains multiple binding sites for an erythroid-specific protein. Genes Dev 2:1089, 1988

31. Martin DIK, Tsai S, Orkin SH: Increased gamma-globin expression in nondeletional HPFH mediated by an erythroid-specific DNA binding factor. Nature 338:435, 1989





32. Evans T, Reitman M, Felsenfeld G An erythroid-specific DNA binding factor recognizes a regulatory sequence common to all chicken globin genes. Proc Natl Acad Sci USA 85:5976, 1988

33. Boker JA, Roesler WJ, Vandenbark GR, Kaetzel DM, Hanson RW, Nilson JH: Characterization of the CAMP for the alpha-subunit of glycoprotein hormones and phosphoenolpyruvate carboxylase (GTP). J Biol Chem 263:19740, 1988

34. Talbot D, Philipsen S , Fraser P, Grosveld F: Detailed analysis of the site 3 region of the human beta globin dominant control region. EMBO J 9:2169, 1990

35. Philipsen S, Talbot D, Fraser P, Grosveld F The beta globin dominant control region: Hypersensitive site 2. EMBO J 92159, 1990

36. Strauss EC, Orkin SH: In vivo protein-DNA interactions at hypersensitive site 3 of the human beta globin locus control region. Proc Natl Acad Sci USA 89:5809, 1992

37. Haniford DB, Pulleyhlank DE: Transition of a cloned d(ATjn- d(ATjn tract to a cruciform in vivo. Nucleic Acids Res 13:4343, 1985

38. McLean MJ, Blaho JA, Kilpatrick MW, Wells RD: Consec- utive A-T pairs can adopt a left-handed DNA structure. Proc Natl Acad Sci USA 83:5884, 1986

39. Kunkel GR, Martinson HG: Nucleosomes will not form on double stranded RNA or over poly (dA)-poly(dT) tracts in recombinant DNA. Nucleic Acids Res 9:6869, 1981

40. Mignotte V, Wall L, deBoer E, Grosveld F, Romeo PH: Two tissue specific factors bind the erythroid promoter of the human porphobilinogen deaminase gene. Nucleic Acids Res 17:37, 1989

41. Mignotte V, Eleouet JF, Raich N, Romeo PH: Cis- and trans- acting elements involved in the regulation of the erythroid promoter of the human porphobilinogen deaminase gene. Proc Natl Acad Sci USA 86:6548, 1989

42. Fleenor DE, Kaufman RE: Chromosomal organization of the beta globin gene domain, in Stamatoyannopoulos G, Nienhuis AW (edsj: Regulation of Hemoglobin Switching. Proceedings of the Sev- enth Conference on Hemoglobin Switching. Baltimore, MD, Johns Hopkins University, 199 1, p I78




1993 81: 2781-2790

DE Fleenor and RE Kaufman beta globin gene domainCharacterization of the DNase I hypersensitive site 3' of the human

http://www.bloodjournal.org/content/81/10/2781.full.htmlUpdated information and services can be found at:

Articles on similar topics can be found in the following Blood collections

http://www.bloodjournal.org/site/misc/rights.xhtml#repub_requestsInformation about reproducing this article in parts or in its entirety may be found online at:

http://www.bloodjournal.org/site/misc/rights.xhtml#reprintsInformation about ordering reprints may be found online at:

http://www.bloodjournal.org/site/subscriptions/index.xhtmlInformation about subscriptions and ASH membership may be found online at:

Copyright 2011 by The American Society of Hematology; all rights reserved.Society of Hematology, 2021 L St, NW, Suite 900, Washington DC 20036.Blood (print ISSN 0006-4971, online ISSN 1528-0020), is published weekly by the American


http://www.bloodjournal.org/content/81/10/2781.full.html

http://www.bloodjournal.org/site/misc/rights.xhtml#repub_requests

http://www.bloodjournal.org/site/misc/rights.xhtml#reprints

http://www.bloodjournal.org/site/subscriptions/index.xhtml





characterization of the dnase i hypersensitive site 3' of the human

Documents