| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Molecular Biology Program, Memorial Sloan-Kettering Cancer Center, and Weill Graduate School of Medical Sciences, Cornell University, New York, NY 10021
 |
Abstract |
|---|
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionDisclosuresReferences |
|---|
/GABP
. Whereas GABP operates as an essential upstream activator, PU.1 assists the flanking "TTTAAA" element (–32/–27), a "weak" but essential TATA box, to bring TBP/TFIID to the transcription start site. PU.1 thus imparts a degree of cell specificity to the minimal promoter and provides a potential link between a number of signaling pathways and TFIID. However, a "strong" TATA box ("TATAAA") eliminates the need for the PU.1 binding site and for PU.1, but not for GABP. As GABP is widely expressed, a strong TATA box thus alleviates promyelocytic cell specificity of the def1 promoter. These findings suggest how the myeloid def1 promoter may have evolutionarily acquired its current properties. |
Introduction |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionDisclosuresReferences |
|---|
Consistent with these findings is the presence of def1 mRNA, albeit at relatively low levels, in the HL-60 human promyelocytic leukemia cell line (6, 8, 9, 10, 11, 12). HL-60 cells can be induced by retinoic acid (RA) to mature along the granulocytic pathway, with concomitant increase in def1 expression, thereby providing a model system for study of differentiation-specific gene regulation (13, 14, 15). It should be recognized, however, that some molecular events during drug-induced differentiation may perhaps not entirely reflect those occurring during normal granulopoiesis. In contrast, def1 transcripts have never been found in the related myeloblastic (KG-1), monoblastic (U-937), myeloblastic/erythroblastic (K-562), or lymphoid B and T cell lines, not even after extensive RA treatment (8, 12). Studies aimed at understanding this unique granulocytic expression of defensin genes must converge, eventually, at the identification of cis-regulatory control elements and the cognate transactivating factors (16, 17).
Instead of being strictly myeloid specific, many transcription factors involved in regulation of "myeloid" genes are more commonly expressed (18). Lineage specificity is controlled, in most of these cases, through unique factor combinations or related mechanisms (17, 19). A case in point is PU.1, a member of the ETS family of transcription factors, known to physically interact with other regulatory proteins or to otherwise function in activating, combinatorial arrangements (20, 21). Some factors implicated in the process also gain, or gain further, transactivating potential through posttranslational modifications (17, 22), which may either impart another layer of cell specificity and/or tie gene expression to a signaling pathway (22).
Previous studies have indicated that lineage- and stage-specific elements control myeloid def1 expression, a process modulated by serum factors, kinase activators and inhibitors, and mediators of inflammation (12, 23, 24), but the underlying mechanisms of basal and induced expression in promyelocytic cells are not yet fully understood. A minimal core promoter (–83 to +82), containing two functionally essential, ETS-like elements, can drive transcription in a quasi cell-specific manner (24, 25), perhaps in conjunction with an upstream C/EBP-like element of as yet unproven capacity (26, 27). The proximal GGAA element (–22/–19), downstream of a TTTAAA sequence, and the related distal one (–62/–59), each bind a different factor in vitro (24, 25). def1 induction is also directed through both elements and factors. Using a specific Ab in an in vitro DNA-binding assay (EMSA), it has previously been suggested that PU.1 was the leading candidate for specific interaction with the proximal GGAA site (25). PU.1 was subsequently purified and positively identified as the binding activity (24). As the next step to further dissect and reconstitute this system, we sought to purify and characterize the postulated ETS factor binding to the distal GGAA regulatory sequence ("D1 box"), termed D1-binding protein or D1BP.
In this study, we report identification of D1BP, biochemically and functionally, as the heterodimeric factor GA-binding protein (GABP). While GABP functions as an essential, typical upstream activator, PU.1 assists the flanking TTTAAA element, a weak yet essential TATA box, to bring TBP/TFIID to the def1 promoter near the transcriptional start site. As a result, PU.1 imparts a degree of cell specificity to the minimal promoter and provides a potential link between a number of signaling pathways and TFIID. However, PU.1 is incapable of functioning in such a capacity in a TATA-less (TGTAAA) def1 promoter. Conversely, creating a strong TATA box (TATAAA) in the def1 promoter eliminates the need for the PU.1 binding site and for PU.1, but not for GABP. Because GABP is more ubiquitously expressed, a strong TATA box mitigates promyelocytic cell specificity of the def1 promoter. Thus, depending on the switch, mutation of the same nucleotide can either eliminate or boost minimal promoter activity, thereby alleviating cell specificity in the latter case. This finding suggests a mechanism of how the myeloid def1 promoter may have evolutionarily acquired its current properties.
|
Materials and Methods |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionDisclosuresReferences |
|---|
Human promyelocytic leukemia HL-60 and myeloblastic leukemia KG-1 cells (American Type Culture Collection (ATCC)), and acute promyelocytic leukemia NB4 cells (28) were grown at 37°C in RPMI 1640 medium (Invitrogen Life Technologies) supplemented with 10% heat-inactivated FCS (HyClone Laboratories), nonessential amino acids, and penicillin and streptomycin at 5 µg/ml. HL-60 and NB4 cell cultures were passaged twice a week to maintain the cell density between 0.5–2 x 106 cells/ml. SL2 embryonic epithelial cells from Drosophila melanogaster Schneider line 2 (ATCC) were grown at 25°C in Schneider medium (Invitrogen Life Technologies) supplemented with 10% FCS. SL2 cell cultures were passaged every 3–4 days to maintain the cell density at 0.5–5 x 106 cells/ml. All cells were counted in hemocytometer chamber, and the viability was assessed by exclusion of 0.1% trypan blue.
Synthetic oligonucleotides
Complementary, single-stranded oligonucleotides were custom synthesized by Integrated DNA Technologies (IDT) and annealed before use as probes and/or as competitors in EMSAs. Only the sense sequences of each pair are listed here: D1, (5'-GACCCAACAGAAAGTAACCCCGGAAATTAG-3'); D1M2, (5'-GACCCAACAGAAAGTAACCCCAAGGATTAG-3'); D1M3, (5'-GACCCAACAGAAAGTAACCCCTGAAATTAG-3'); TA, (5'-CAAGACCTTTAAATAGGGGAAGTCCACTTG-3'); TAM2, (5'-CAAGACCTTTAAATAGGGCCCGTCCACTTG-3'); TA(T-31G), (5'-CAAGACCTGTAAATAGGGGAAGTCCACTTG-3'); TA(T-31A), (5'-CAAGACCTATAAATAGGGGAAGTCCACTTG-3'); TAM2(T-31A), (5'-CAAGACCTATAAATAGGGCCCGTCCACTTG-3'); PU.1(SV40), (5'-TGAAATAACCTCTGAAAGAGGAACTTGGTTAGGTA-3'); GABP (thrombopoietin), (5'-GTGAAGGCCCCCGGAAGT5ACGCCT-3'); Ets-1, (5'-CGGCCAAACCGGAAGTATGTGC-3'); Elk-1, (5'-TCCTGATCATCCACCGGAAGTGAG-3'); Elf-1, (5'-TAAACCCGGAAGTGTAGTACATC-3'); ER81, (5'-AACCCCCGCCGGAAGTACTGATCT-3'); Erg-2, (5'-CCCTGAGACCGGAAGTATTAGGCT-3').
Underlined nucleotides have been changed from the wild-type sequences. D1 and D1M3 (both sense and antisense) were also used for concatamerization by PCR to generate the surface ligand for affinity purification of the D1BP. For coupling to streptavidin-derivatized magnetic beads, the sense oligonucleotide of each pair was synthesized with biotin attached to a 6-carbon spacer at the 5' end; the complementary strand was without a biotin moiety.
EMSA
Prebinding of nuclear extract (5–10 µg) or respective column or other fractions to poly(dI:dC) was conducted at 25°C for 10 min in buffer containing 4% glycerol, 1 mM MgCl2, 0.5 mM EDTA, 0.5 mM DTT, 25 mM NaCl, 10 mM Tris-HCl (pH 7.5) and 0.05 mg/ml poly(dI:dC) (Amersham Biosciences). When indicated, unlabeled competitor oligonucleotides (200-fold molar excess) were included in the incubation mixture at this point. Radiolabeled probe (3.5 fM; 
2 x 104 cpm) was then added to the reaction above, mixed and incubated at 25°C for 20 min. One microliter of 10x gel-loading buffer, containing 250 mM Tris-HCl (pH 7.5), 0.2% of bromphenol blue, 0.2% xylene cyanol, and 40% glycerol, was added to the reaction and then loaded onto a 6% native gel (which was prerun for 90 min at 100 V) in 0.5x nondenaturing Tris-borate-EDTA (TBE) buffer. The electrophoresis was run at 25°C and 100 V for 3.5 h. The gel was then transferred onto Whatman paper, vacuum-dried, and exposed to Hyperfilm (Amersham Biosciences) for the desired period of time at –80°C and with an intensifier screen. For Ab supershift experiments, Abs (0.1 µg/µl; 1 µl volume) were added to the solutions after protein-DNA incubation for 20 min at 25°C, and the DNA-protein complexes were also resolved in 6% polyacrylamide gels. Abs recognizing PU.1, Ets-1, Ets-2, Elk-1, Elf-1, Egr-2, and Etv1 were purchased from Santa Cruz Biotechnology.
Preparation of DNA concatamers and attachment to magnetic beads
Multimers of the wild-type (D1) and the mutant (D1M3) DNA binding sites were generated by a PCR-based method (29) using complementary, single-stranded oligonucleotides of two direct repeats of D1 or D1M3, respectively. The forward single-stranded oligonucleotides were biotinylated at the 5' end. PCRs (50 µl vol) contained 460 ng of each primer, 8 µM dNTPs (Roche Molecular Biochemicals), and 2 U of Vent polymerase (New England Biolabs) in the following buffer: 10 mM KCl, 10 mM (NH4)2SO4, 3.5 mM MgSO4, 0.1% Triton X-100, and 20 mM Tris-HCl (pH 8.8; at 25°C). The cycling conditions were: 95°C for 2 min followed by 14 cycles of 95°C for 1 min, 55°C for 1 min, and 72°C for 3 min. PCR products were purified using QiaQwick kit (Qiagen), and analyzed by agarose gel electrophoresis in TBE buffer. Each PCR yielded 3–5 µg of DNA with sizes between 200 bp and 10 kb. The 5'-biotinylated, concatemerized DNA was attached to M280 streptavidin-coated magnetic beads (Dynal Biotech) using KilobaseBINDER kit (Dynal Biotech) according to the manufacturer’s instructions. Approximately 10–50% of DNA concatamers in the incubation mixture were attached to the beads (up to 8 µg/mg beads).
Protein purifications
D1 DNA-binding activity was purified from 1.4 x 109 exponentially growing NB4 cells, following established protocol (30). All procedures were performed at 4°C. Seventy-five milligrams of nuclear extract (31) was fractionated on a P11 phosphocellulose (Whatman) column, with 22-ml bed volume, equilibrated with buffer D (20 mM HEPES (pH 7.9), 0.2 mM EDTA, 0.5 mM DTT, 0.01% Nonidet P-40, 0.2 mM PMSF, 10% glycerol) containing 0.075 M KCl. Bound proteins were eluted using a 200-ml linear gradient of 0.075–0.85 M KCl in the same buffer. Fractions containing D1 DNA-binding activity (as monitored by EMSA) eluted at 0.1 M KCl, and were pooled and dialyzed overnight against 50 vol of buffer D containing 0.1 M KCl. All protein concentrations were determined using the Bradford assay (Bio-Rad) and BSA standards (Sigma-Aldrich). Affinity capture of D1-binding protein(s) on DNA-magnetic beads was then conducted (30). One-milligram beads with attached D1-concatamerized DNA (5 µg/mg; dsDNA/beads) were first washed in DNA-binding solution (20 mM HEPES (pH 7.9), 0.1M KCl, 0.2 mM EDTA, 0.5 mM DTT, 0.01% Nonidet P-40, 10% glycerol), mixed with 3.5 mg of the dialyzed P11 (0.1 M) protein fraction, and incubated for 3 h with rotation. Binding buffers always contained oligo(dI:dC) (30 bp in length; custom synthesized by IDT) and poly(dI:dC) (Amersham Biosciences) at 0.1 mg/ml each, except where noted. After binding, the beads were washed once with 1.2 ml of complete binding buffer, and three times with binding buffer in the presence of Escherichia coli double-stranded and ssDNA, at 0.1 mg/ml each, as nonspecific competitor. The beads were then eluted with 50 µl of binding buffer containing 0.5 M KCl, for 15 min on ice, and stored. The eluate was diluted in 1.2 ml of fresh, complete binding buffer and incubated for 3 h with 1-mg beads derivatized with attached mutant D1M3-concatamerized DNA to remove nonspecific nucleic acid-binding proteins by negative selection. The nonbound fraction was then taken for another round of positive selection on D1-concatamerized DNA beads, followed by salt elution of bound proteins, as described above. The final eluate was taken for EMSA analysis, SDS gel electrophoresis, Western blotting, and mass spectrometric identification. In addition, beads stored after each round of positive or negative selection were suspended in 50 µl of Laemmli sample buffer and also taken for gel analysis.
Mass spectrometry (MS)
Gel-resolved proteins were stained with Coomassie Blue R250 (Bio-Rad), bands excised and digested with trypsin, the mixtures fractionated on a Poros 50 R2 RP microtip, and resulting peptide pools analyzed by MALDI reflectron TOF (MALDI-reTOF) MS using a Bruker UltraFlex TOF/TOF instrument (Bruker Daltonics), as described (32, 33). Selected experimental masses (m/z) were taken to search the human segment of a nonredundant protein database (NR; 108,000 entries; National Center for Biotechnology Information, Bethesda, MD), using the MASCOT Peptide Mass Fingerprint (PMF) program, version 2.0.04 for Windows (Matrix Science), with a mass accuracy restriction better than 30 ppm, and maximum one missed cleavage site allowed per peptide. To confirm PMF results with scores 
40, mass spectrometric sequencing of selected peptides was done by MALDI-TOF/TOF (MS/MS) analysis on the same prepared samples, using the UltraFlex instrument in LIFT mode. Fragment ion spectra were taken to search NR using the MASCOT MS/MS Ion Search program (Matrix Science). Any tentative confirmation (Mascot score 30) of a PMF result thus obtained was verified by comparing the computer-generated fragment ion series of the predicted tryptic peptide with the experimental MS/MS data.
Plasmids
Luciferase reporter constructs were all derived from the pGL3' series of vectors (Promega). pGL3B-def1(–83/+82), (–34/+82) and (+11/+83), are basic (B) vectors containing human defensin 1 promoter inserts that span sequences from nucleotide –83 (Sau96I site) and, respectively, –34 and +11 (ExoIII truncations), to + 82 (ScaI site), as described (25). These three constructs will be further referred to as, respectively, a, b, and c. Mutants derived from pGL3B-def1(–83/+82) were constructed using the QuikChange (Stratagene) site-directed mutagenesis kit as per the manufacturer’s instructions. Appropriate sense and antisense mutation primers (35-nt long) were designed to introduce the following modifications (single, double, or triple sites): construct d, GAA to CCC (–21/–19); e, T to G (–31); f = d + e; g, T to A (–31); h = d + g; i, GGAA to AAGG (–62/–59); j = d +i; k = d + e + i; l = d + g + i. Mutant plasmid constructs were sequenced to confirm the desired alterations; sequence analysis was done at the DNA Service Laboratory, Biotechnology Center, Utah State University (Logan, UT). The expression plasmid pCMV-hGH (human growth hormone gene under control of a cytomegalovirus promoter) was obtained from Invitrogen Life Technologies and used throughout as an internal control for transfection efficiency (25). For the construction of expression plasmid pCDNA3.1/PU.1, mouse PU.1 cDNA (34) was excised with EcoRI from the PJ6 plasmid (gift from Dr. M. Klemsz, Indiana University School of Medicine, Indianapolis, IN), and inserted into the corresponding site of plasmid pCDNA3.1+ (Invitrogen Life Technologies). Human GABP and GABP cDNAs (35) were similarly excised from pBluescript plasmids (gift from Dr. T. Brown, Pzifer, Groton, CT) and inserted into pCDNA3.1+. Final constructs were confirmed by sequencing.
Recombinant proteins
Using PCR, BglII or BamH1 sites were introduced at the 5' and 3' ends of the open reading frames of GABP, GABP, and PU.1-containing plasmids (see above). The cDNAs were inserted into the BamH1 site of the His-tag expression vector pET-19b (Novagen), and transformed into E. coli strain BL21-Gold (DE3)p LysS (Stratagene). Cultures (250 ml) were grown at 27°C in Luria Broth containing 50 µg/ml ampicillin and 20 µg/ml chloramphenicol until the A600 reached 0.4. Expression was induced by the addition of isopropyl--D-thiogalactopyranoside to 1 mM. Cells were grown for a further 4 h at 27°C, then harvested by centrifugation and stored at –80°C until use. Cells were lysed at 4°C for 30 min in 10 mM sodium phosphate buffer (pH 7.4), containing 0.2 mg/ml lysozyme and 0.1% Nonidet P-40, and centrifuged at 12,000 x g for 20 min. His-tagged proteins were purified from cell lysates supernatants on 1-ml HisTrap columns (Amersham Biosciences) according to the manufacturer’s instructions. Proteins that eluted with a 300–500 mM gradient of imidazole were precipitated with 20% ammonium sulfate, and dissolved in 6 M urea, 50 mM K3PO4, 100 mM KCl, 5 mM DTT. They were then renatured by stepwise dialysis, for 4 h each at 4°C, against 50 vol of the same buffer with decreasing concentrations of 5 M, 3 M, 1 M, and no urea. Recombinant proteins were analyzed by EMSA, and also fractionated by SDS PAGE and identities confirmed by MS.
Production of anti-GABP ( and ) Abs
Peptides corresponding to residues 421–433 of GABP (CEQKKLAKMQLHG) and residues 365–381 of GABP (RQQLLKKEQEAEAYRQK; with an additional Cys residue at the N terminus) were synthesized by S. S. Yi in the Microchemistry and Proteomics Core Facility (Memorial Sloan-Kettering Cancer Center), and were conjugated to keyhole lympet hemocyanin using the Imject Maleimide-Activated Conjugation kit (Pierce). Polyclonal anti-GABP and anti-GABP Abs were individually produced by Pocono Rabbit Farm and Laboratory. Specific Abs were purified from rabbit sera after the first bleedings using prepacked Econo-Pac Protein A Cartridge (Bio-Rad) as per the manufacturer’s procedure. Purified Ig fractions (at 1 mg/ml) were stored at 4°C.
Western blotting
Protein solutions were separated by electrophoresis in SDS gels, and the proteins were transferred to a polyvinylidene difluoride Immobilon-P membrane (Millipore) in Tris-glycine-methanol buffer as described previously (30). The respective Abs were added to the membranes at 1 µg/ml in PBS/0.05% Tween 20 for 2 h at room temperature (RT). Anti-mouse or anti-rabbit-HRP-conjugated Abs (Santa Cruz Biotechnology) and ECL kit (Pierce) were used for visualization of the immune complexes.
Chromatin immunoprecipitation (ChIP)
ChIP was conducted using the Upstate Biotechnology Chromatin Immunoprecipitation Assay kit as per the manufacturer’s instructions. Briefly, 5 x 106 NB4 cells were treated with 1% formaldehyde for 15 min at RT. Cells were washed twice with PBS containing protease inhibitors (Roche), and lysed in 200 µl of SDS lysis buffer (1% SDS, 10 mM EDTA, 50 mM Tris-HCl (pH 8.1), and protease inhibitors). After incubation on ice for 10 min, the samples were sonicated on ice (14 x 10 s, at 15% amplitude; Fisher Sonic Dismembrator, Model 500) to an average DNA length of 200-3000 bp. Samples were the centrifuged for 10 min at maximum speed (Eppendorf centrifuge) at 4°C and the sonicated cell supernatant was diluted 10-fold with ChIP dilution buffer (0.01% SDS, 1.1% Triton X-100, 1.2 mM EDTA, 16.7 mM Tris-HCl (pH 8.1), 167 mM NaCl, and protease inhibitors). To reduce nonspecific background, each sample was precleared with 50 µl of protein A agarose (50% slurry) in the presence of 500 µg salmon sperm DNA for 30 min at 4°C with agitation. Incubations with specific Abs, and appropriate negative controls, were performed overnight at 4°C. The immune complexes were then collected with 60 µl of protein A agarose-50% slurry, again in the presence of 500 µg of salmon sperm DNA. After a 1-h rotation at 4°C, the beads were collected and washed once for 5 min with each of the following buffers: Low Salt Immune Complex Wash Buffer (0.1% SDS, 1% Triton X-100, 2 mM EDTA, 20 mM Tris-HCl (pH 8.1), 150 mM NaCl), high salt immune complex wash buffer (0.1% SDS, 1%Triton X-100, 2 mM EDTA, 20 mM Tris-HCl (pH 8.1), 500 mM NaCl), LiCl immune complex wash buffer (0.25 mM LiCl, 1% Nonidet P40, 1% deoxycholate, 1 mM EDTA, 10 mM Tris-HCl (pH 8.1)), and twice with TE buffer (10 mM Tris-HCl, 1 mM EDTA, pH 8). The washed resin was resuspended in 250 µl of freshly prepared elution buffer (1% SDS, 0.1 M NaHCO3). Cross-linking was reversed by incubation at 65°C overnight. DNA fragments were purified by phenol/chloroform extraction, amplified by PCR using primers that covered the –160/+11 region of def1 promoter (25), the –143/+8 region of the CD89 promoter (forward: 5'-GCTGTGAGGCAATAGATGTGGAA-3'; reverse: 5'-GGGTCCACTGTGCTGACAC-3'), or the –162/+137 region of the p40phox promoter (forward: 5'-GCGCCAAGGACTGACATC-3'; reverse: 5'-GAGCAGGTGGTGCGTCTC-3'), and analyzed by electrophoresis in 1% agarose gel/TBE.
Transient transfections
Transfections of HL-60 and KG-1 cells were performed as previously described (25). In brief, cells were diluted into the corresponding growth medium (see cell culture section), at densities of 4 x 105/ml, the day before transfection. After 18 h, cells (1 x 107 per transfection) were pelleted and washed with prewarmed (37°C) IMDM, centrifuged at 500 x g for 5 min at RT, resuspended at a density of 1 x 107 cells in 0.4 ml of warm IMDM containing 2.5 µg of pCMV-hGH plasmids. This suspension was added into the electroporation cuvette already containing the luciferase expression DNA constructs (<20 µl). Cells and plasmids were then mixed with a pipetter, incubated for 5 min at RT, followed by electroporation at 975 µF capacitance and 280 V using a Gene Pulser II (Bio-Rad), unless otherwise indicated. The cells were then transferred to 10 ml of warm IMDM with 10% FCS, the dishes swirled and incubated at 37°C for 5 h, and the cells harvested in 15-ml tubes by centrifugation at 500 x g for 5 min at RT. One milliliter of supernatant from each experiment was stored in an Eppendorf tube for human growth hormone (hGH) assay. Pellets were washed with 5 ml of PBS at RT, 300 µl of lysis buffer (1% Triton X-100, 25 mM Gly-Gly (pH 7.8), 15 mM MgSO4, 4 mM EGTA (pH 7.8), 1 mM DTT) was added, pellets were then resuspended, transferred to Eppendorf tubes, vortexed for 5 s and spun at full speed for 3 min at RT. Fifteen microliters of the above lysate was mixed with 300 µl of freshly made assay buffer, which contained 25 mM Gly-Gly (pH 7.8), 15 mM KPO4 (pH 7.8), 15 mM MgSO4, 4 mM EGTA (pH 7.8), 2 mM ATP (pH 7.8), 1 mM DTT. Relative light units were measured for 20 s in a model Monolight 2010 luminometer (Analytical Luminescence Laboratory) after addition of 100 µl of 1 mM D-luciferin potassium salt (Analytical Luminescence). The hGH was measured with the ELISA kit from the Nichols Institute as per the manufacturer’s instructions. Briefly, 100 µl of supernatant was mixed with an equal volume of Ab solution; the latter is a mixture of two mAbs, each one specific for a different and distinct epitope on the hGH molecule, to form a soluble sandwich complex in the presence of hGH. One of the Abs was 125I-labeled for detection while the other Ab is coupled to biotin. The reaction was then mixed with an avidin coated plastic bead and incubated for 90 min at RT while shaking (180 rpm). After two washes, the bead was counted in a gamma counter (LKB 1272 Clinigamma; Wallac) for 1 min.
Cotransfections of Drosophila SL2 cells
Exponentially growing SL2 cells were diluted to 5 x 105/ml and plated into 6-well tissue-culture dishes (1 ml/well). Cells were transfected at a ratio of 1:3 nucleic acid/DOSPER Liposomal Transfection Reagent (Roche) according to the manufacturer‘s instructions. One microgram of D-luciferase-containing pGL3B-def1 (–83/+82) wild-type or mutant reporter construct was cotransfected with equal amounts of the expression vector pcDNA3.1+ (Invitrogen Life Technologies) containing PU.1, GABP, and/or GABP and with 1 µg of plasmid expressing -galactosidase under control of a Drosophila actin promoter (gift from Dr. M. Baylies, Memorial Sloan-Kettering Cancer Center, New York, NY) as internal control for transfection efficiency (36). The DNA concentration for each transfection was equalized by addition of salmon sperm DNA. Twenty-four hours after transfection, cells were resuspended in 2 ml of fresh medium and incubated further for another 24 h. Transfected cells were harvested by centrifugation, washed once with PBS and resuspended in 100 µl of lysis buffer (25 mM Gly-Gly (pH 7.8), 15 mM MgSO4, 4 mM EGTA (pH 7.8), 1% Triton X-100, 1 mM DTT). Cells were then vortexed for 30 s, lysed by freezing and thawing and incubated on ice for 20 min. After centrifugation for 5 min at full speed in a table centrifuge, 50 µl of the supernatant was mixed with 300 µl of freshly made assay buffer (25 mM Gly-Gly (pH 7.8), 15 mM KPO4 (pH 7.8), 15 mM MgSO4, 4 mM EGTA, 2 mM ATP, 1 mM DTT). Luciferase activity was measured in a luminometer as described above. Five microliters of lysate was used to determine -galactosidase activity with the -Gal-Assay kit (Invitrogen Life Technologies) as per the manufacturer‘s instruction, and was used to normalize luciferase activity. The data shown are the mean values of at least three independent transfections, each in triplicate, and are expressed in percentage compared with the activity generated by the reporter construct alone.
|
Results |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionDisclosuresReferences |
|---|
def1 expression in promyelocytic leukemia cells is controlled by a combination of at least two nuclear factors, acting through two functional, ETS-like elements, TA and D1, in the minimal promoter (–83/+82). The proximal TA element binds PU.1 in vitro, and was successfully used for specific DNA affinity purification of that same factor from HL-60 cells (24). Next, identification of the D1BP was a key step to further resolve the mechanism of def1 minimal promoter activity. EMSA using a probe corresponding to the D1 box (Fig. 1) suggested highest levels of D1BP expression in promyelocytic NB4 cells (data not shown), which were therefore taken as the source for isolation of this factor. The same probe was used to assay for the D1BP during various steps of purification, as outlined in Fig. 2A. D1BP was retained on a P11 phosphocellulose column and eluted with 0.1 M KCl. TA-binding activity, containing PU.1 as assayed by Western blotting, eluted from the P11 column with 0.85 M KCl (data not shown). The subsequent procedure of positive/negative/positive affinity selection, involving wild-type (D1) and mutant (D1M3) fragments immobilized to magnetic particles, as previously developed in our laboratory (30), was applied for straightforward purification of the D1-binding activity. P11 fractions containing the desired activity were further enriched in D1BP by binding to wild-type (D1)n-affinity beads. Proteins eluted from the beads with 0.5 M KCl (Fig. 2B, lane 4; and 2C, lane 3) were then incubated with magnetic beads carrying mutant (D1M3)-DNA, differing in just a single base-pair from the wild-type sequence. The nonbound proteins (Fig. 2C, lane 4) were once more incubated with D1 beads for a final round of positive selection (Fig. 2B, lanes 5 and 6; and 2C, lanes 6 and 7). We consistently observed two shifted bands in EMSA throughout the purification procedure (Fig. 2B), as originally reported (25). Four bands observed after SDS gel electrophoresis of the eluted fraction (indicated with closed or open circles in Fig. 2C, lane 7) were excised and the proteins identified by a combination of PMF using MALDI-TOF MS, and MS-based sequencing using MALDI-TOF/TOF MS/MS.
|
|
, the DNA-binding subunit of the GABP/ heterodimer, a well-characterized member of the ETS-family of transcription factors (18, 37, 38). No data were obtained to identify any of the visualized bands as GABP. However, the protein was detected by Western blot analysis of an aliquot of the same fraction (data not shown). In addition, the DNA-binding activity was intact (Fig. 2B, lane 6), although some faster migrating bands, most likely indicating protein degradation, were also observed. In a comparable analysis of the prolactin gene promoter, the subunit of GABP was also identified by MS, while the GABP was not detected either (39). We do not have a ready explanation for these observations. The two other bands on the gel (
) were both identified as PARP1, an abundant nuclear protein that has frequently been observed to nonspecifically bind to nucleic acids (Ref. 30 ; M. Yaneva, H. Erdjument-Bromage, and P. Tempst, unpublished observations). In a parallel analysis, the P11 fractions (0.85 M salt elution) containing TA-binding activity were pooled and also purified on DNA-magnetic beads as previously described for HL-60 cell nuclear extracts (24). Mass spectrometric analysis confirmed that the TA-binding protein in NB4 cells was PU.1 as well (data not shown).
Binding specificities of recombinant and native (endogenous) GABP/GABP and PU.1 to the distal and proximal ETS elements
To demonstrate that GABP (+GABP) and PU.1 bind specifically to, respectively, D1 and TA sequences, His-tagged versions of these proteins were expressed in E. coli, and affinity purified on Ni-NTA columns, renatured and tested for DNA binding in EMSA. The results showed that rGABP but not bound to the D1 probe, forming the faster migrating band (Fig. 3A, lanes 2 and 3), consistent with earlier observations of GABP subunit binding to DNA (18). Presence of both subunits resulted in formation of an additional, slower migrating band, formed presumably by a complex of both rGABP and (Fig. 3A, lane 4). Neither protein bound to the TA probe (Fig. 3A, lane 8), whereas rPU.1 did so specifically (Fig. 3A, lane 9). Addition of both rGABP subunits did not affect this binding (Fig. 3A, lane 10). Binding of rPU.1 to the D1 sequence had already been ruled out in earlier studies (25), which was confirmed here (Fig. 3A, lanes 5 and 6). The results indicated that the rGABP/rGABP binds to the distal but not to the proximal ETS site, whereas rPU.1 clearly showed the opposite binding preference.
|
and (see Materials and Methods) in EMSA to specifically test for the presence of these proteins in the D1 DNA-binding complexes. Both anti-GABP and anti-GABP Abs, either alone or as an equimolar mixture, interfered with the formation of D1-protein complexes (Fig. 3B, lanes 3, 4, and 6), whereas purified preimmune Igs from the same animals did not have an effect (Fig. 3B, lane 5). Abs specific for other ETS-family proteins: PU.1, Elk-1, Ets-1, Ets-2, Elf-1, Etv-1, and Erg-1 did not have any effect on D1-binding activity (data not shown). Furthermore, dsDNA probes containing known binding sites for all these other factors did not effectively compete the D1 probe in an EMSA either, whereas a probe containing the GABP binding site of the thrombopoietin promoter (40) did so quite well (data not shown). Comparable EMSA results in all cases were observed using nuclear extract from NB4 cells as well (data not shown). These observations provide further proof that the endogenous GABP proteins are directly involved in binding to the distal ETS site in the def1 promoter in vitro.
To further analyze factor occupancy, we prepared biotinylated DNA fragments spanning different overlapping sequences of the core promoter region, either as monomers or concatamers, that were immobilized on streptavidin-coated magnetic beads as illustrated in Fig. 3C (upper panel). The D1 and D10 DNA fragments contained the distal ETS binding site; the D11 DNA fragment contains both the distal and the proximal ETS sites. These DNA beads were incubated with HL-60 nuclear extracts, washed and the eluted proteins analyzed for the presence of GABP/GABP and PU.1 by Western blotting. The results showed that PU.1 protein was only bound to D11 and not to D10 or D1 (Fig. 3C, lanes 4 and 5), i.e., this protein occupied the proximal but not to the distal ETS site. GABP was detected on the beads with D1 DNA, i.e., bound to the upstream, distal ETS site (Fig. 3C, lane 7). GABP binding to this site was significantly enhanced when a concatamerized D1 oligonucleotide was used for the pull-down (Fig. 3C, lane 8), reflecting its low level of expression in the nuclear extract. In separate experiments, GABP , as well as , could be detected as bound to multimer D10 and D11 fragments by Western blotting (data not shown). Thus, three lines of evidence clearly established that, in vitro, GABP specifically binds to the upstream GGAA site (–32/–25), whereas the PU.1 protein binds to the proximal GGAA site (–22/–19).
PU.1 and GABP bind to the def1 promoter in vivo
To verify in vivo occupancy of the def1 promoter by PU.1 and GABP/, in the context of chromatin in intact promyelocytic leukemia NB4 cells, we performed chromatin immuno-precipitation analyses. Abs against PU.1 and GABP did efficiently and specifically precipitate def1 promoter DNA (–160/+11), whereas preimmune sera could not, suggesting stable in vivo association of these two factors with the promoter (Fig. 4A, lanes 3 and 4, respectively). The proteins remained bound to the promoter in NB4 cells treated with 1 µM all-trans RA for 72 h (data not shown). Positive control ChIP analyses were conducted using the same cells as well as the anti-PU.1 and anti-GABP/GABP Abs, and different sets of primers (see Materials and Methods and Fig. 4 legend), corresponding to well-documented target genes of these factors, namely, the p40phox promoter (41) and the CD89 promoter (42). The results shown in Fig. 4, B and C, indicate that our analyses have been done properly and using adequate reagents.
|
To gain further mechanistic insights into defensin-1 core promoter function, it was necessary to dissect the proximal control region into its individual, constituent elements, and to explore the possibility of physical and functional interaction between their cognate binding proteins. The transcription start site (+1) was previously determined by primer extension and S1 nuclease protection (25), and the surrounding sequence is very similar to the initiator element consensus (43). Thus, the slightly upstream TTTAAATA sequence (–32/–25; underlined in Fig. 1) fulfills the requirements of a vertebrate TATA box by established criteria of weighted consensus sequence and location (44). Mutation of this sequence abrogates minimal promoter (–83/+82) activity in HL-60 cells, consistent with a functional role (25). The PU.1-binding, GGAA core sequence (–22/–19) just downstream from the TA-rich site is also functionally important, as its disruption resulted in reduced promoter efficacy in vivo.
We compared the sequence of the –33/–15 region of the def1 promoter with known eukaryotic TATA box and the PU.1 binding site base frequencies, either derived from 389 unrelated promoter sequences (44) or, respectively, established by PCR-mediated random site selection (45). Two observations can readily be made from inspecting the alignment in Table I. First, the two 15-nt long sites overlap by 11 bp. Second, the def1 promoter sequence is in full accordance with the conserved core elements of each site (i.e., TATAAA and GGAA), except that the second base of the TATA (the underlined nucleotide has been mutated) box is about 10 times more frequently an A in eukaryotic promoters than the T (91 A vs 9% T; see Table I) that occurs in this particular position (at –31; see Fig. 1) in the def1 promoter. We therefore investigated the role this nucleotide might play in def1 minimal promoter activity.
|
|
3% of wild type). Interestingly, a point mutation that created a bona fide TATA box (T-31 to A; construct g) increased minimal promoter activity to 160%. In combination with a PU.1 binding site knockout (construct h), the improved (or strong) TATA box increased activity still further, up to 250%. Similarly, a strong TATA box could also boost activity of a minimal promoter lacking both GABP- and PU.1 binding sites (construct l) by a factor of 10; from 5 to 50% of wild-type promoter. This, however, was still five times lower than the most active combination consisting of a strong TATA box, GABP+ and PU.1–. Introduction of a strong TATA box in a def1/Luc reporter construct analyzed in myeloblastic KG-1 cells also had a stimulatory effect on an otherwise much lower activity (5–10% of HL-60 cells), which was not further enhanced by disruption of the PU.1 binding site (Fig. 6B; constructs g and h). PU.1-site disruption had, by itself, little effect on reporter activity (construct d). These findings were, in fact, in perfect agreement with the above results obtained in HL-60 cells, because KG-1 cells do not express measurable levels of PU.1 protein (Ref. 25 ; data not shown), which makes the presence or absence of a functional binding site irrelevant. All the above observations suggest some intriguing mechanisms of activator and general transcription factor binding and cooperativity, which we will elaborate on in the discussion. Still, we first wanted to further establish in vivo function of each of the implicated factors, in combination with the effects of either a weak (TTTAAA) or strong (TATAAA) TATA box.
|
PU.1 expression is restricted to myeloid and B cells (46, 47), but GABP is rather ubiquitously expressed in human cells (18), which could potentially complicate coexpression analyses in HeLa or other type of heterologous cells. We therefore selected Drosophila SL2 cells for these experiments because they lack all the factors under study. SL2 cells were transfected with either wild-type or mutant (T-31 to A) def1 minimal promoter luciferase constructs along with expression vectors carrying cDNAs for either GABP , , PU.1, or equimolar mixtures of any of those. Expression of these proteins was under control of the CMV promoter which had previously been found to function in SL2 cells (36). Cotransfections with either GABP alone or GABP alone did not result in any measurable activity over background. However, cotransfection of two cDNAs expressing both and subunits resulted in a 2.5-fold enhanced activation of the wild-type def1 reporter construct (Fig. 7A). The effect of PU.1 cDNA expression alone was quite low. In contrast, cotransfections with all three cDNAs, expressing GABP , , and PU.1, resulted in a significant, close to 10-fold activation of the wild-type construct (Fig. 7A). Doubling the amount of cDNA for either GABP/GABP or PU.1 could not compensate for the absence of the other factor (data not shown). These results strongly suggest that GABP/GABP and PU.1 cooperate in def1 minimal promoter activation.
|
1.2x), whereas coexpression with both GABP subunits led to robust, 5-fold activation (Fig. 7B). This is also 4-fold over the normalized activity of a wild-type reporter in the presence of GABP/GABP. Again, these results revealed an intricate relationship between the ETS-factors and the protein(s) binding the TATA box, likely TBP/TFIID. Apparently, PU.1 operates in a positive or helper role when positioned on the wild-type promoter in close proximity to the weak TTTA box or to an inactive TGTA box but, conversely, has an appreciable negative effect when bound adjacent to a strong TATA box on an otherwise unaltered promoter. No such effects were observed for GABP binding to the distal control element, which consistently resulted in a 5- to 10-fold relative transcriptional activation in the presence of either a weak or a strong TATA box. |
Discussion |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionDisclosuresReferences |
|---|
through a proximal PU.1 binding site (24). Furthermore, PGE2 treatments greatly augment the RA response (23), whereas the inflammatory regulators do not (24), all suggesting a cAMP-stimulated, enhancer-like capacity for the D1 box and its binding protein, and supporting a model that D1BP and PU.1 are the endpoints of separate signaling pathways (24). As RA-dependent and -independent def1 regulation is strictly lineage specific, even within the class of myeloid blood cells (12, 23), transduction may be modulated by additional cis-acting regions outside the minimal promoter, or by the levels, modification states or interaction patterns of any of the binding proteins. As the next necessary step to further dissect and reconstitute this system, we have now identified D1BP biochemically and functionally as the heterodimeric factor GABP/GABP. GABP (18, 37) and particularly PU.1 (34, 48) have been frequently implicated in myeloid and myeloid-specific gene regulation, most often in combinations with various other factors (18, 49, 50, 51) but in a few cases in this specific pairing as well (38, 52). Through a series of in vitro and in vivo experiments, we have now conclusively established that, despite substantial sequence similarities between the two respective binding sites, there exists 1) a strict functional requirement for the presence of both factors and 2) mutually exclusive site-occupancies in the minimal promoter. This could be the result of critical flanking sequences influencing binding specificities of the respective factors, as well as requirements of specific protein-protein interactions with other nuclear factors and/or general transcription factors to initiate formation of an activating complex. In fact, we have obtained compelling evidence that, at least in the case of PU.1 binding to the proximal site, the second postulation may actually be the more important. Specifically, we propose here that the proximal ETS-site operates in conjunction with the adjacent TATA-like box to assemble a functional, primarily myeloid-specific, basal transcription complex.
The experimental observations supporting this model derive from a mutational analysis of the promoter region that contains both a TTTAAAT sequence (–32/–26) and a flanking PU.1 binding site (GGAA; –22/–19). While the TA rich element meets all criteria of a vertebrate TATA box (44), its precise sequence in the def1 promoter, in particular the T in the second position, only occurs in fewer than 10% of other genes (44). Selected mutations of this nucleotide, either alone or in combination with a disrupted PU.1-site, were evaluated for possible effects on PU.1-binding in vitro and on transcriptional activity in vivo. A summary of the results is presented in cartoon format in Fig. 8 to illustrate the protein-protein-DNA interactions that we believe are taking place at this locus in each configuration, and how those affect transcriptional activity (measured to increase from scenario A (lowest) to F (highest)). Panel D depicts the wild-type promoter. Functional disruption of the TTTA element, through a T-31 to G (as present in 0% of vertebrate TATA boxes) mutation, results in near compete loss of transcriptional activity (A and B). By contrast, changing T-31 to A (as present in 90% of vertebrate TATA boxes) results in a moderate increase (1.5x) of activity (E). The activity is further enhanced (2.5x) when the same mutation is paired with an inactive PU.1 binding site (F), whereas a PU.1-site knockout by itself results in considerable loss of activity (C), as anticipated.
|
R1b promoter (54). In contrast, putting a strong TATA box in the def1 promoter eliminates the need for a PU.1 binding site and by extension for PU.1, but not for GABP, a conclusion that was duly validated by functional analysis in fly cells. In fact, PU.1 is even less than optional in the presence of a strong TATA box as it appears it might interfere with optimal TFIID binding and function. As GABP is rather ubiquitously expressed, conversion to a bona fide TATA box could mitigate strict promyelocytic cell specificity of the def1 promoter, as shown in KG-1 myeloblastic cells. A unique feature about the TTTA element-plus-helper-site motif in the def1 promoter is that it could have originated by a single base pair change from a prototypical TATA box (A to T). Possibly, the existing wild-type promoter may have evolved in this way to become increasingly reliant on nearby positioning of PU.1 for optimal function, thereby resulting in a de facto myeloid-restricted activity. Furthermore, PU.1-binding affinity is regulated by phosphorylation in response to extracellular stimuli and cytokines, most notably by casein kinase II and p38 MAPK-dependent pathways (24, 55, 56, 57, 58). Such signals can now be readily, albeit indirectly, transduced to the general transcriptional machinery, thereby providing another layer of control in addition to activator-dependent mechanisms operating through GABP at the distal D1 box. Binding of active GABP, itself the endpoint of various signaling pathways and posttranslational modification events (18, 24, 59, 60), is also essential for transactivation, even in the presence of a strong TATA box.
Finally, it should be noted that our past observations on factor binding to specific GABP(D1)- and PU.1(TA)-probes following RA-treatment (25) correlate well with the reported changes in GABP and PU.1 expression (at the transcript and protein levels) during myeloid differentiation along the granulocytic pathway. Indeed, nuclear extracts from HL-60 cells showed increased PU.1-binding activity (by EMSA) and PU.1 protein levels (by immunoblot analysis) 2 days after induction (41). Reportedly, expression of GABP does not change during granulocytic differentiation but those of certain splicing forms of GABP, in contrast, are up-regulated (18, 61). In conjunction with differentiation-dependent phosphorylation of those same factors (24), these changes may fully account for the increased def1 transcription. The latter changes may very well be the more important as overexpression of cDNAs for either one factor, or both, did not affect def1 transcription significantly (data not shown).
In summary, the myeloid def1 promoter contains a weak TATA box plus one requisite and one helper site that specifically recruit GABP and PU.1 into an active transcription initiation complex. The need for a nearby PU.1 binding site for optimal TBP binding imparts a degree of cell specificity to the minimal promoter and provides a potential link between a number of signaling pathways and TFIID. The precise identities, roles, and regulation of other factors additionally controlling def1 expression during various stages of granulopoiesis through interactions outside the minimal promoter remain to be established.
|
Acknowledgments |
|---|
|
Disclosures |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionDisclosuresReferences |
|---|
|
Footnotes |
|---|
1 The Sloan-Kettering Microchemistry and Proteomics Core Facility is supported by National Institutes of Health Cancer Center Support Grant P30 CA08748. 
2 Address correspondence and reprint requests to Dr. Paul Tempst, Memorial Sloan-Kettering Cancer Center, 1275 York Avenue, New York, NY 10021. E-mail address: p-tempst{at}mskcc.org
3 Abbreviations used in this paper: def1, defensin-1; RA, retinoic acid; GABP, GA-binding protein; D1BP, D1-binding protein; MS, mass spectrometry; ChIP, chromatin immunoprecipitation; RT, room temperature; MALDI-reTOF, MALDI reflectron TOF; hGH, human growth hormone.
Received for publication August 17, 2005. Accepted for publication March 13, 2006.
|
References |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionDisclosuresReferences |
|---|
-defensin 1, 2, and 3 to the anti-HIV-1 activity of CD8 antiviral factor. Science 298: 995-1000. 
