Journal of Leukocyte Biology
HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
QUICK SEARCH:   [advanced]


     

Originally published online as doi:10.1189/jlb.0705390 on May 30, 2006

Published online before print May 30, 2006
This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Right arrow All Versions of this Article:
jlb.0705390v1
80/2/383    most recent
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow reprints & permissions
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Hale, J. S.
Right arrow Articles by Weis, J. H.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Hale, J. S.
Right arrow Articles by Weis, J. H.
(Journal of Leukocyte Biology. 2006;80:383-398.)
© 2006 by Society for Leukocyte Biology

Transcriptional control of Pactolus: evidence of a negative control region and comparison with its evolutionary paralogue, CD18 (ß2 integrin)

J. Scott Hale1, Timothy J. Dahlem1, Rebecca L. Margraf, Irina Debnath, Janis J. Weis and John H. Weis2

Division of Cell Biology and Immunology, Department of Pathology, University of Utah School of Medicine, Salt Lake City

2Correspondence: Department of Pathology, University of Utah School of Medicine, Salt Lake City, Utah 84132. john.weis{at}path.utah.edu


   ABSTRACT
TOP
 ABSTRACT
INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 REFERENCES
 
The mouse Pactolus and CD18 genes are highly conserved paralogues. The expression patterns of these genes are diverse in that most cells of hematopoietic lineage express CD18, but Pactolus is only expressed by maturing neutrophils. The minimal promoters of these two genes are homologous, including the conservation of two tandem PU.1-binding sites upstream of the transcriptional start site. To define the means by which these two structurally similar but functionally distinct promoters operate, a series of reporter assays, electrophoretic mobility shift assay (EMSA) and chromatin immunoprecipitation analyses, were performed. Transfection of Pactolus constructs into mouse macrophages, which do not express Pactolus, defined a negative control element within the first 100 base pairs. The presence of this negative regulatory site, distinct from the PU.1-binding site, was confirmed by EMSA oligonucleotide competition and gene reporter assays of Pactolus/CD18 chimeric constructs. Although PU.1 binding can be detected on Pactolus and CD18 minimal promoter segments with EMSA, only the CD18 promoter shows PU.1 binding in vivo, suggesting that the negative regulatory protein may block PU.1 from binding to the Pactolus promoter, thus inhibiting transcription of the gene. Sequence analysis of the negative control region in the Pactolus promoter suggested potential control by Snail and/or Smad families of transcription regulators. EMSA supershift analysis with antibodies against these proteins, using extracts from macrophages and mucosal mast cells, identified specific binding of Smuc to the promoter element, including a Smuc/PU.1/DNA trimeric complex. These data implicate Smuc as blocking Pactolus transcription in cells expressing PU.1 (and CD18) but not Pactolus.

Key Words: granulocytes • neutrophils • macrophages


   INTRODUCTION
TOP
 ABSTRACT
 INTRODUCTION
MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 REFERENCES
 
Pactolus is a mouse ß-integrin-like protein, which was discovered in a differential display analysis between different lineages of cells derived from bone marrow precursor cells [1 ]. Pactolus is expressed exclusively in the mouse by maturing neutrophils [2 ]. The vast bulk of the protein is retained in storage granules, where it undergoes extensive glycosyl modifications [3 ]. Upon terminal activation of the neutrophil, intracellular Pactolus is shuttled to the cell surface, presumably serving as a tag for the uptake of apoptotic/necrotic neutrophils by phagocytic macrophages. Although sequence conservation indicates Pactolus is similar to the ß-integrin subunits [4 ], it is expressed on the cell surface without any obvious {alpha}-chain partner [1 ].

Comparison of the genes and coding sequences of Pactolus and CD18 (ß-2 integrin subunit) suggests Pactolus is a duplication of CD18. The Pactolus gene has been altered, compared with CD18, by deletion in coding sequences within the metal ion-binding domain (which is critical for {alpha}-chain binding and ligand recognition) and altered sequences within the transmembrane/cytoplasmic protein domains [1 ]. Pactolus is not an activating or phagocytic receptor for neutrophils, and CD18 is [3 ], nor is Pactolus required for appropriate neutrophil migration into sites of inflammation [3 ].

One intriguing difference found between the Pactolus and the CD18 genes (based on their sequence conservation) is their differential expression profiles. The CD18 gene encodes the ß-integrin subunit, which pairs with at least four different {alpha}-chain subunits to form a variety of complexes such as lymphocyte function-associated antigen-1 and membrane-activated complex-1 [5 ]. The CD18 gene is expressed by T and B cells and by macrophages and granulocytes; analysis of CD18-deficient mice demonstrates loss of function of these cell types for activation, migration, and phagocytosis [6 , 7 ]. Pactolus, conversely, is only found in vivo to be expressed by maturing mouse neutrophils and is not expressed by macrophages or lymphocytes. The Pactolus gene has been found in the mouse and rat genomes but is not present in the human genomic sequence.

The promoter elements of the human CD18 gene have been examined by a number of groups; however, there are no similar reports detailing the transcriptional control elements of the murine CD18 gene. Reporter gene constructs possessing just the first 79 base pairs (bp) of the human CD18 promoter provide the same transcriptional induction as constructs containing 918 bp of the promoter, suggesting that the key transcriptional elements are just upstream of the transcriptional start site [8 ]. The human CD18 minimal promoter has been defined as possessing two distinct PU.1 sites within the first 74 bp of the promoter. Transfection analyses have shown that elimination of either of the two PU.1 sites diminishes reporter gene activity [8 , 9 ]. The primary ets family member, which binds to these sites, appears to be PU.1, although at least in the human, GA-binding protein (GABP) also plays a role, potentially competing for binding with PU.1 [10 ]. At least two SP1 sites also appear to be functional within this minimal promoter construct and can act cooperatively with GABP to help drive transcription [11 ]. Human CD18 expression can be induced by retinoic acid [12 ]. Such induction appears to be controlled directly by a retinoic acid response element as well as by synergistic interactions between GABP and p300/cyclic adenosine monophosphate-response element-binding protein-binding protein. One report suggested that the human CD18 gene required a site 3.5 kb upstream of the transcription initiation site for efficient promoter function in transgenic mice (a human CD18 promoter/CD4 reporter gene construct) [13 ]; however, the necessity of this region for CD18 transcription has not been demonstrated. PU.1 has long been recognized as a critical transcription factor required for the expression of a variety of myeloid promoters [14 15 16 17 ]. Indeed, PU.1 acts as a master switch, in that inhibition of its activity can block hematopoiesis [18 ], and most mice lacking the protein die at birth (or close thereafter) or live with severely compromised innate and acquired immune responses [19 ].

The sequence and structural analysis (size of exons and intron/exon junctions) of the Pactolus and CD18 mouse genes suggest that Pactolus is a duplication of the CD18 gene [4 ]; they are most alike to one another than to any other member of the ß-integrin subunit family. However, they have clearly evolved different mechanisms to control their transcription in different cell types. In this report, we investigate the mechanism(s) by which such control is asserted and how these two genes have diverged to provide their present tissue specificity of expression. We describe the presence of a novel, negative regulatory element within the Pactolus promoter, which binds one or more repressor protein(s) apparently capable of blocking PU.1-dependent transcription. When present, this protein may suppress Pactolus transcription by inhibiting the binding of PU.1 to its binding site(s) within the promoter element.


   MATERIALS AND METHODS
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
RESULTS
 DISCUSSION
 REFERENCES
 
Electrophoretic mobility shift assay (EMSA)
Nuclear extracts were prepared from RAW 264.7 cells and mucosal and connective mast cells [20 ], and all steps were conducted at 4°C following a standard protocol [21 ]. Cells (7x106) were washed in ice-cold phosphate-buffered saline (PBS) and lysed in 1 mL Buffer A [10 mM HEPES, pH 7.9, 0.05% Nonidet P-40 (NP-40), 10 mM KCl, 1 mM EDTA, 1 mM EGTA, 10 µg/mL aprotinin, 10 µg/mL leupeptin, 1 mM phenylmethylsulfonyl fluoride (PMSF), and 1 mM dithiothreitol (DTT)] for 10 min. Lysed cells were centrifuged for 5 min at 4000 g,and supernatant was removed by Pasteur pipette. Nuclear pellets were resuspended in 40 µL Buffer B (10 mM HEPES, pH 7.9, 400 mM NaCl, 25% glycerol, 50 mM MgCl2, 1 mM EDTA, 1 mM EGTA, 10 µg/mL aprotinin, 10 µg/mL leupeptin, 1 mM PMSF, and 1 mM DTT) and incubated for 60 min. Samples were centrifuged for 10 min at 14,000 g to pellet the nuclei, and 50 µL supernatant was recovered. Protein concentration was determined by the Bradford assay. DNA fragments were amplified from C57B6 genomic template DNA using 35 cycles, 55° annealing, and 5 s elongation. The Pactolus-108 fragment was amplified using primers 2167F (5'-AGATCTGCTCCATGAGTAGAGCAAC-3') and 2168R (5'-ATCGATTCTCTGGACGCTTGTGGCAG-3'). Pactolus prox was amplified using primers 1954F (5'-GGCTGCAGACAGAAGAGGAAGCACCAGA-3') and 2168R (5'-ATCGATTCTCTGGACGCTTGTGGCAG-3'). Pactolus dist was amplified using primers 2167F (5'-AGATCTGCTCCATGAGTAGAGCAAC-3') and 2169R (5'-GGCTGCAGCTGGGAGGAAGTGGAGTGA-3'). CD18-117 was amplified using primers 2212F (5'-GGAGATCTTACCCACATGTGGGCCAATC-3') and 2211R (5'-GGATCGATTGTGGTTGTGTCTAAAGAAC-3'). CD18 prox was amplified using primers 2269F (5'-GGCTGCAGACTTAGAACAGGAAGTGTC-3') and 2211R (5'-GGATCGATTGTGGTTGTGTCTAAAGAAC-3'). CD18 dist was amplified using primers 2212F (5'-GGAGATCTTACCCACATGTGGGCCAATC-3') and 2270R (5'-GGCTGCAGCCAGGAGGAAGTTGAGTG-3'). Amplified fragments were subcloned into the pCR 2.1-TOPO plasmid (Invitrogen, Carlsbad, CA). Probes were prepared by restriction digestion of the plasmids containing the probe insert using BglII and ClaI for pac-108, BglII and EcoRI for CD18-117, BglII and PstI for Pac dist and CD18 dist, and ClaI and PstI for Pac prox and CD18 prox. Probes were end-labeled with [{alpha}-32P]deoxy-cytidine 5'-triphosphate using the Klenow fragment of DNA polymerase I. Labeled probes were separated from the vector by electrophoresis on a 2% agarose gel and isolated using the Qiagen gel purification protocol. Binding reactions were performed in a total volume of 30 µL containing 50,000 counts per minute-labeled probe, 2.5 µg poly(dI:dC), and binding buffer (final 3 mM HEPES, pH 7.9, 0.7 mM MgCl2, 0.7 mM DTT, 3% glycerol, and 70 µM EDTA). Binding reactions contained 2–3 µg nuclear extract. Some reactions contained 1 µg double-stranded competitor oligo (added before nuclear extract) or 4 µg antibody (Santa Cruz Biotechnology, CA; added after nuclear extract). Reactions were incubated for 15 min at room temperature and electrophoresed for 2 h at 200 V on a 4.5% polyacrylamide gel (0.5xTris-boric acid-EDTA buffer). Gels were dried and exposed to X-ray film for 6–48 h.

Tissue culture, cell isolation, and reverse transcriptase-polymerase chain reaction (RT-PCR)
All cells were cultured at 37°C with 5% CO2. Mouse RAW 264.7 cells were cultured in Dulbecco’s modified Eagle’s medium (DMEM) with 10% fetal calf serum (FCS; Hyclone, Logan, UT), 1% penicillin-streptomycin (Pen-Strep, Life Technologies, Gaithersburg, MD), 10 mM HEPES (Gibco, Grand Island, NY), 1 mM minimum essential medium sodium pyruvate (Gibco), and 2 mM L-glutamine (Gibco). Mucosal-like mast cells (MMC) and connective tissue-like mast cells (CTMC) were derived as described [20 ]. Neutrophil-enriched and macrophage/lymphocyte-enriched samples were obtained from mouse bone marrow by step density centrifugation using 1.08 g/ml Percoll. The bottom, dense samples were highly enriched for maturing neutrophils, and the top phase was enriched for the less-dense, nongranulocyte-maturing lymphocytes, monocytes, and macrophages. Total splenocytes were obtained by manually shearing the splenic capsule, lysing the red blood cells, and washing the remaining cells in PBS by centrifugation. Total RNA was isolated from bone marrow and spleen cell samples using CsCl [22 ]. RNA was resuspended in water and quantified by A260 absorbance. cDNA was synthesized from 5 µg RNA, 10 µl 5x first-strand buffer, 5 µl 5 mM deoxy-unspecified nucleoside 5'-triphosphate (dNTP), 5 µl 0.1 M DTT, 2 µl 1.25 mM random primers (Life Technologies), 2 µl Moloney murine leukemia virus RT, and water to a final volume of 50 µl. The reaction was incubated at 37°C for 1 h. cDNA was purified using a PCR purification kit (Qiagen, Valencia, CA). Amplification was performed, and samples were analyzed by denaturing polyacrylamide electrophoresis in DNA-sequencing gels as described previously [23 ].

Luciferase reporter constructs
Luciferase reporter constructs were prepared by cloning fragments of the Pactolus and CD18 promoter by PCR and ligating into the pGL3 basic vector (Promega, Madison, WI), which contains the Firefly luciferase gene. The pac-79 fragment was amplified using primers 1820F (5'-AAGCTTCACTCCACTTCCTCCCAGG-3') and 1697R (5'-AAGCTTTCTCTGGACGCTTGTGGCAG-3'). CD18-88 was amplified using primers 1957F (5'-GGAGATCTTCACTCAACTTCCTCCTGGG-3') and 1956R (5'-GGAAGCTTTGTGGTTGTGTCTAAAGAAC-3'). Fragments pac-79 and CD18-88 were digested with BglII and HIII and ligated into their respective restriction sites in the pGL3 basic plasmid (Promega). Chimeric pac/CD18 promoter constructs were engineered by constructing a PstI restriction site between the proximal and distal PU.1 sites on the Pactolus and CD18 promoters. Chimeric constructs were made by combining PstI and HIII-digested pac-54 fragments or the CD18-63 fragment with the double-stranded oligo of pac distal BglII/PstI overhang or CD18 distal BglII/PstI overhang and double-ligated into the BglII and HIII restriction sites of the pGL3 basic plasmid. The pac-54 fragment was amplified using primers 1954F (5'-GGCTGCAGACAGAAGAGGAAGCACCAGA-3') and 1697R (5'-AAGCTTTCTCTGGACGCTTGTGGCAG-3'). CD18-63 fragment was amplified using primers 1955F (5'-GGCTGCAGACTTAGAACAGGAAGTGTCA-3') and 1956R (5'-GGAAGCTTTGTGGTTGTGTCTAAAGAAC-3'). Double-stranded oligo pac distal BglII/PstI overhang was made by annealing an equimolar amount of oligo 1950F (5'-GATCTTCACTCCACTTCCTCCCAGCTGCA-3') and complimentary oligo 1951R (5'-GCTGGGAGGAAGTGGAGTGAA-3'), heating to 100°C, and cooling. Double-stranded oligo CD18 distal BglII/PstI overhang was made by annealing equimolar amounts of oligo 1952F (5'-GATCTTCACTCAACTTCCTCCTGGCTGCA-3') and complimentary oligo 1953R (5'-GCCAGGAGGAAGTTGAGTGAA-3'), heating to 100°C, and cooling. Fragment pac(CD18) was made by first doing two separated PCR reactions, amplifying C57B6 genomic DNA using primers 2361F (5'-GGAGATCTTCACTCCACTTCCTCCCAGG-3') and 2358R (5'-AGCCGCTGAGGCCCCCCCGTCTGGTGCTTCCTCTTCTGTC-3') and primers 2356F (5'-GGGGGGGCCTCAGCGGCTACAAGCGTCCAGAGAAAGC-TTCC-3') and 2357R (5'-GGAAGCTTTCTCTGGACGCTTGT-3'). The purified products of these two reactions were combined and amplified by PCR using primers 2361F and 2357R to make the pac(CD18) fragment, and Fragment CD18(pac) was made by first doing two separate PCR reactions amplifying C57B6-genomic DNA using primers 1957F (5'-GGAGATCTTCACTCAACTTCCTCCTGGG-3') and 2360R (5'-CACGGCAGACACACCTTATGAGCCTGACACTTCCTGTTCTA-3') and primers 2359F (5'-TAAGGTGTGTCTGCCGTGTTCTTTAGACACAACCACAAAGCTTCC-3') and 1956R (5'-GGAAGCTTTGTGGTTGTGTCTAAAGAAC-3'). The purified products of these two reactions were combined and amplified by PCR using primers 1957F and 1956R to make the CD18(pac) fragment. Fragments pac(CD18) and CD18(pac) were digested with HIII and BglII and ligated into their respective restriction sites in the pGL3 basic plasmid. pac-1081 was amplified using primers 1699F (5'-CCCGGGCAATGTGTTAGCATAAGAAC-3') and 1697R (5'-AAGCTTTCTCTGGACGCTTGTGGCAG-3'). SmaI and HIII-digested pac-1081 fragment was ligated into the respective restriction sites of the pGL3 basic plasmid. The pac-538 pGL3 construct was made by digesting the pac-1081 pGL3 construct with NheI, separating the vector from the -1081/-539 fragment by agarose gel electrophoresis, and re-ligating the vector.

Transient transfection and luciferase reporter assay
RAW cells (5x106) were resuspended in 0.65 mL DMEM (10% FCS and 1% Pen-Strep) and pipetted into a 0.4-cm electroporation cuvette. Equimolar amounts of plasmid relative to 10 µg pGL3 plasmid DNA were added. pRL-cytomegalovirus (CMV; 1 µg) containing the Renilla luciferase gene (Promega) was included to serve as a control for transfection efficiency. Cells were incubated on ice for 5 min, electroporated at 280 V and 960 uF using a gene pulser (Bio-Rad, Hercules, CA), and incubated for 5 more min on ice. Transfected cells were transferred into 10 mL media and cultured for 46 h before harvesting. Cells were harvested by scraping, centrifugation at 1000 g, removal of the media, and resuspension in 5 mL cold PBS. Cells were centrifuged again and aspirated, and the cell pellet was resuspended in 200 µL passive lysis buffer. Lysis was done by freezing at –70°C for >10 min and a 2-min quick-thaw at 37°C. Samples were centrifuged for 4 min at 13,000 g, and 160 µl cell lysate was recovered. Each lysate (20 ul) was analyzed by dual luciferase assay protocol (Progmega) using a MLX microtiter plate luminometer (Dynex, Chantilly, VA). Background signal was subtracted from Firefly and Renilla readings, and data were normalized for transfection efficiency by the ratio of Firefly:Renilla light emission. All assays were analyzed as a percentage of activity relative to the CD18-88 construct.

Chromatin immunoprecipitation (ChIP)
All steps were performed at 4°C or on ice, except where indicated, following a modified protocol [24 25 26 ]. RAW 264.7 cells were harvested, washed two times with PBS with 0.1% bovine serum albumin (BSA). Cells (3x106) were pelleted for each ChIP reaction, and PBS was removed. DNA/protein cross-linking was done by resuspending cell pellets in 1% formaldehyde and incubating for 40 min at room temperature. Glycine was then added, incubated for 5 min, and then centrifuged to pellet cells. Pellet was washed 1x with PBS [0.1% BSA+protease inhibitor "Complete Minipill" (Roche, Indianapolis, IN); cells were resuspended in radioimmunoprecipitation assay (RIPA), 50 mM NaCl, 25 mM Tris, pH 7.5, 1 mM EDTA, 0.1% sodium dodecyl sulfate (SDS), 1% w/v deoxycholate, 1% NP-40, and 1% BSA and protease inhibitor Complete Minipill (Roche)], incubated for 20 min on ice, and then vortexed briefly. Lysed samples were sonicated five times at a Power 4 for 25 s each time. Sonicate was centrifuged for 10 min at 14 K. Supernate was taken and centrifuged again to pellet debri out. For the input sample, 50 µl supernate was mixed with 450 µL TE, and cross-linking was resolved by heating at 65°C overnight and then purified by Qiagen column protocol. The remaining supernate was used for the immunoprecipitation, and 4 ug isotype control, Oct 1, and PU.1, whole rabbit immunoglobulin G (IgG), rabbit polyclonal anti-Oct 1, and rabbit polyclonal anti-PU.1 (Santa Cruz) were added to each respective immunoprecipitation; the immunoprecipation reaction was done for 4 h, rotating at 4°C. Immunoprecipated reactions were added to sheep anti-rabbit IgG-conjugated Dynabeads, which were previously blocked and washed with PBS (1 mg/mL BSA and 0.1 mg/mL herring sperm DNA) and incubated overnight. Dynabeads were washed, using a magnetic rack, two times with RIPA, two times with RIPA + 0.1 mg/mL herring sperm DNA, two times with RIPA + 0.1 mg/mL herring sperm DNA and 500 mM NaCl, two times with RIPA with 250 mM LiCl, and two times with TE. Finally, TE was added, and cross-linking was resolved by incubating overnight at 65°C. Supernate was then separated from Dynabeads using a magnetic rack, and soluble DNA was purified using a Qiagen column protocol. PCR was done on each sample (input, isotype control, Oct 1, PU.1) and H20 using primers specific for the Pactolus promoter, CD18 promoter, CD21 promoter, and actin-intragenic region for 30 cycles with 6 s elongation and 55°C annealing temperature. PCR products were subjected to electrophoresis on a sequencing gel, and the gel was dried and exposed to X-ray film overnight.

Real-time PCR analysis of the preceding ChIP protocol was done using fluorescence detection in 10 µl reactions containing 3.0 ul input or ChIP DNA, 5 µM each primer, 0.5 U Amplitaq DNA polymerase, 110 ng TaqStart antibody (Clontech Laboratories, Palo Alto, CA), 0.8 mM dNTP, 1x light cycler buffer [3 mM MgCl2, 50 mM Tris-HCl (pH 8.3), 500 mg/ml BSA], and 1:30,000 dilution of SYBR Green I (Molecular Probes, Eugene, OR). Samples were loaded into capillary tubes and incubated in a fluorescence thermocycler (LightCycler LC24, Idaho Technology, Salt Lake City, UT) for denaturing at 94°C for 1 s, for annealing at 60°C for Pactolus and CD18 and 65°C for actin for 1 s, and for extension and fluorescence detection at 72°C for 1 s. This cycle was repeated 40 times, preceded by an initial incubation at 95°C for 60 s and followed by a melting curve acquired by cooling to 65°C and slowly heating at 0.2°C/s to 94°C, and fluorescence data were collected at 0.2°C intervals [27 ]. Relative quantification was done using a fourfold dilution series of input DNA to create a DNA concentration standard curve. The mean of three replicates for each sample was then used to determine the fold enrichment of target and reference genes. (ChIP target/input target)/(ChIP reference/input reference) = fold enrichment.

Western blot protein analysis
Cells (3.0–3.5x106 cell for Raw/2PK3 or 2.4x107 cell for thymus) were lysed in 0.5 ml RIPA buffer [50 mM NaCl, 25 mM Tris-HCl, pH 7.5, 1 mM EDTA, pH 7.5–8.0, 0.1% SDS (w/v), 1% NaDeoxycholate (w/v), 1% NP-40 (v/v), 1 mM PMSF, and one Complete Mini (Roche) protease inhibitor cocktail tablet] for 30 min at 4°C with rotation and vortexing. Lysates were cleared at 14,000 revolutions per minute for 10 min and the supernatant, transferred to a fresh microcentrifuge tube. Immunoblotting was preformed by standard methods with antibodies specific for the following proteins: actin (rabbit polyclonal, Catalog No. A-2066, Sigma Chemical Co., St. Louis, MO), Smuc (goat polyclonal, Catalog No. sc-10439X, Santa Cruz Biotechnology), anti-rabbit (donkey polyclonal, Catalog No. 711-035-152, Jackson Immunoresearch Laboratories, Inc., West Grove, PA), and anti-goat (donkey polyclonal, Catalog no. 705-035-147, Jackson Immunoresearch Laboratories, Inc.).


   RESULTS
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
DISCUSSION
 REFERENCES
 
Comparison of Pactolus and CD18 minimal promoters
A comparison of the sequences of the mouse CD18 and Pactolus promoters demonstrates a high level of conservation between the two (Fig. 1A ) and with that, of human CD18 and rat Pactolus (Fig. 1B) . All four of these promoter units possess two PU.1-like sites in conserved positions within the promoters, suggesting all of these genes are regulated positively by PU.1. The CD18 (mouse and human) and Pactolus promoters (mouse and rat) lack a canonical TATA box, thus requiring transcription factors such as PU.1 to help anchor the transcription initiation complex. Although these four promoter units are highly homologous, both of the Pactolus promoters possess an apparent deletion (compared with the CD18 promoters) just downstream of the proximal PU.1 site. The expression of Pactolus is restricted to neutrophils, and CD18 is expressed in T and B cells, macrophages, and granulocytes. However, if these two genes’ promoters are so structurally similar, then how is this differential transcription regulated?


Figure 1
View larger version (21K):
[in this window]
[in a new window]
  		Figure 1. Alignment of the murine Pactolus and CD18, rat Pactolus, and human CD18 minimal promoters. (A) Both mouse promoters contain proximal and distal putative PU.1 sites (as marked with bold type) and lack a TATA box. The shaded line denotes segments of each promoter representing 83% sequence homology. The dashed line denotes segments of each promoter, which show a high level of sequence divergence. The sequence shown for each promoter is equivalent to the pac-108 and CD18-117 probes used in EMSA experiments. The Pactolus and CD18 distal and proximal fragments are also indicated. (B) Clustal alignment of the mouse Pactolus (Mus Min Pac Pro) and CD18 (Mus Min CD18 Pr), rat Pactolus (Rat Min Pac Pro), and human CD18 (Hum Min CD18 Pr) minimal promoter elements. Sequences were obtained from genomic databases. *, Residues conserved in all four sequences.

 
Mapping control regions within the Pactolus promoter
We chose a luciferase reporter gene assay system to map transcriptional control elements of the Pactolus gene. One difficulty in following this procedure was the lack of a transfectable cell line that expresses Pactolus, and Pactolus expression is limited in vivo to maturing neutrophils. Two mouse cell lines, mouse promyelocytic and 32Dc13, possess characteristics of neutrophils after induction with all-trans retinoic acid or granulocyte-colony stimulating factor, respectively [28 29 30 31 ]; however, induction of such cells resulted in only a modest increase of Pactolus expression, and the transfection efficiency of such cells was very low (data not shown). Thus, at this time, there are no cell lines, which normally express Pactolus and can be used for reporter assays. Therefore, we decided to take the opposite approach: to use a transfectable cell line that expresses CD18 and PU.1 but not Pactolus, into which we could introduce mutant Pactolus promoter constructs. The immediate objective was to attempt to define the site(s) within the promoter element that inhibited expression of Pactolus in that cell. We chose the RAW mouse monocyte/macrophage cell line for these assays.

For this series of experiments, we chose to focus on control elements within the first 1081 bp of the Pactolus gene. The human CD18 gene analyses implicated the first 79 bp of the minimal promoter as having the most dramatic effect on transcription. The close homology of the mouse and human CD18 promoters and that of the mouse Pactolus promoter suggest that the minimal promoter in the two mouse genes may also have the greatest impact on transcriptional control. Although there may be additional control elements within these two mouse genes (and the human CD18 gene), which exert positive effects upon transcription, a negative control element-blocking transcription of Pactolus in, for example, macrophages would be assumed to be localized to the minimal promoter region. A series of promoter deletion constructs was prepared in which the Pactolus promoter was inserted upstream of the luciferase reporter gene (or no insertion: pGL3), and such constructs were then transfected into RAW cells. As a relevant, positive control, we used the mouse minimal CD18 promoter of 88 bp. As the level of expression of CD18 and Pactolus within mouse neutrophils is similar, both promoters should provide about equal reporter activity in a permissive cell. As shown in Figure 2A , all of the Pactolus constructs, whether they possessed 1081 bp, 538 bp, or 79 bp of the promoter, demonstrated minimal reporter gene activity in the RAW transfections, compared with the normalized activity of the CD18 promoter. There was no significant difference in reporter function between any of the Pactolus promoter-containing constructs. If the function of the minimal Pactolus promoter was suppressed by distal elements, then removing them by serial deletion should have allowed the reporter strength of the Pactolus promoter to approach that of CD18. As that was not observed, the presence of a negative control element within the Pactolus minimal promoter element (79 bp) is suggested.


Figure 2
View larger version (11K):
[in this window]
[in a new window]
  		Figure 2. Functional analysis of the Pactolus minimal promoter. (A) Luciferase assays of RAW 264.7 cells transiently transfected with reporter constructs containing Pactolus promoter fragments of varying lengths. The molecular equivalent of 10 µg pGL3 plasmid of all constructs was cotransfected with 1 µg pRL-CMV plasmid (to serve as a transfection efficiency control) into 5 x 106 cells, which were grown for 46 h after transfection and then harvested and assayed for Firefly and Renilla luciferase activity. Relative light units were calculated relative to readout from the reporter construct containing the CD18-88 promoter fragment that contains proximal and distal PU.1-binding sites. Data shown are the average of three experiments, each assay performed in duplicate. Error bars denote standard deviation. There is no statistical difference in reporter expression between the Pactolus-containing constructs (-1081, -538, or -79). (B) Chimeric promoters of Pactolus and CD18 are as described in Materials and Methods. Shown are luciferase assays of RAW 264.7 cells transiently transfected with the CD18-88 promoter in pGL3, Pactolus-79 promoter fragment in the pGL3 plasmid, pGL3-negative control, and chimeric promoters containing the distal PU.1 fragment of the Pactolus or CD18 promoter linked to the proximal fragment of the CD18 or Pactolus promoter, both in pGL3. Readout was normalized relative to the CD18-88 reporter construct. Data shown are the average of six independent experiments, each performed in duplicate. Error bars represent standard deviation.

 
Defining the transcriptional repressor function(s) associated with the Pactolus promoter
We sought to identify this putative, negative regulatory element in the Pactolus minimal promoter by creating a number of chimeric constructs, which swapped the proximal and distal segments of the CD18 and Pactolus promoters (see Materials and Methods). As shown in Figure 2B , reporter assays were carried out in RAW cells with constructs that possessed the following: pac/CD18, distal region of Pactolus promoter fused to the proximal region of CD18; CD18/pac, distal region of CD18 fused to the proximal region of the Pactolus promoter. The three control plasmids possessing the unaltered -79-bp minimal Pactolus promoter, the minimal CD18-88-bp promoter, and the empty vector cassette pGL3 were also assayed. The pac/CD18 construct demonstrated the same level of expression as the CD18-88 control, suggesting that the distal region of the Pactolus promoter did not have a negative impact on the proximal PU.1 site of the CD18 promoter. In contrast, the proximal Pactolus sequence did suppress the activity of the distal CD18 sequence to a level of expression comparable with the native Pactolus promoter sequences. As highlighted in Figure 1 , the proximal regions of the two promoters demonstrate a high level of sequence divergence, and the distal regions are highly conserved.

Contrasting nuclear protein-binding sites between the Pactolus and CD18 promoters
The preceding data suggested that the Pactolus minimal promoter possesses a negative control element that is capable of blocking PU.1-dependent transcription in RAW cells. We sought to determine if we could map DNA-binding proteins to such a site using EMSA. As shown in Figure 3 , the banding patterns between the two full-length minimal promoter fragments (pac-108 and CD18-117) produced similar patterns (Fig. 3A) . When the two sets of proximal and distal fragments were analyzed, it was apparent that the main binding sites for both promoters were within the proximal region. The previous analyses of the CD18 human promoter inferred PU.1 binding to both sites within the promoter region [8 , 10 ]. These EMSA data suggest that the proximal site in the CD18 promoter (as well as the Pactolus promoter) is much more active in protein complex formation than the distal site.


Figure 3
View larger version (28K):
[in this window]
[in a new window]
  		Figure 3. EMSA analysis of promoter fragments of the Pactolus and CD18 promoters with RAW cell nuclear extracts. EMSA analyses were performed as described in Materials and Methods. (A) Probes were incubated in the presence or absence of RAW cell nuclear extract. (B) EMSA analysis comparing the Pac proximal and CD18 proximal fragments. Double-stranded competitor oligo (1 µg) was added to each binding reaction where indicated before the addition of nuclear extract. Competitor oligo 55 contains the Pactolus proximal PU.1 site (see Fig. 5 ). Antibodies were added to the mix after the extract. Specific complexes denoted with the bars and lowercase letters are described in the text.

 
The specificity of PU.1 binding to these two minimal promoter elements was investigated using a competing site-specific oligonucleotide [oligo 55 (see Fig. 5 and Materials and Methods)] and supershift analysis with anti-PU.1 (Fig. 3B) . The Pactolus proximal fragment gave three major bands (Bands a–c). The presence of the competing oligo 55 greatly diminished the intensity of Bands a and c but had little effect on Band b. A nonsense oligo without the PU.1 ets-binding site had no effect on the EMSA pattern compared with control, as did an isotype control antibody (data not shown). The presence of the PU.1 antibody alone created the supershift bands marked as d, indicating PU.1 binding to that fragment. These data suggest that PU.1 and at least one additional protein bind to the Pactolus proximal region but do so at different sites.


Figure 5
View larger version (21K):
[in this window]
[in a new window]
  		Figure 5. Defining discrete protein-binding sites on the Pactolus proximal probe region. EMSA oligo-competition experiments were performed with the Pactolus proximal probe (pac proximal) and specific double-stranded oligonucleotides. RAW cell nuclear extract (2–3 µg) was added to each reaction except Lane 1 (free probe). (A and B) Unlabeled, double-stranded, competitor oligo (1 µg, as indicated) was added to evaluate protein/DNA complexes. (C) Sequence of Pactolus proximal probe (sense strand shown only)-delineating position and sequence of competitor and mutant competitor oligos. Mutated nucleotides are shown for each mutant competitor. 43m1 is the mutant oligo for oligo 43, in which the CCAG sequence was replaced with TGTC. The 38 oligo mutant series (38m1–5) are as shown.

 
Similar analysis of the CD18 proximal promoter fragment gave contrasting results (Fig. 3B) . Two primary bands (Bands e and f) disappear when the competing oligonucleotide is added, suggesting factors do bind to the proximal PU.1 site. However, only one of these bands (Band e) is extinguished with the anti-PU.1 antibody, suggesting that the protein-producing Band f is not PU.1 but is likely a member of the ets family (thus recognizing the core PU.1 site). The human CD18 promoter shows binding of PU.1 and GABP to such sites [10 , 11 ]; we have not determined if Band f is a result of GABP binding. It should be pointed out that such a complex is also faintly evident with the Pactolus proximal probe; this band is present with extract alone, is lost when oligo 55 is used as competition, and does not shift with the anti-PU.1 antisera. A lower band (Band g) is apparent; its quantity is unaffected with the various treatments. Thus, the CD18 and Pactolus proximal fragments demonstrate protein-binding constituents more complex than simply PU.1 binding.

Contrasting DNA-binding complexes formed on Pactolus promoter elements by Pactolus-expressing and -nonexpressing cells
We have previously shown that Pactolus expression is high in the in vitro-derived CTMC samples but low in the in vitro-derived MMC cells [1 ]; unfortunately, these primary cells do not transfect well for reporter gene assays (data not shown), but they are useful for the isolation of nuclear extracts. The three Pactolus promoter probes described above were used in EMSA analyses with nuclear extracts from RAW cells, MMC, and CTMC. EMSA analysis of the entire minimal promoter element (pac-108) demonstrated a number of specific shifts in the RAW, MMC, and CTMC extracts (Fig. 4A ). Antibodies specific for PU.1, Elk 1, and Oct 1 were used to define these binding partners. It is notable that only one shift was observed in the Pactolus-expressing CTMC extracts (Band c), and this shift was solely a result of PU.1 (Band d). A doublet was observed in the Pactolus-nonexpressing MMC sample (Band b), which did not supershift with the PU.1 antibody. MMC do express CD18 and PU.1 at levels comparable with the CTMC (data not shown).


Figure 4
View larger version (21K):
[in this window]
[in a new window]
  		Figure 4. EMSA analysis of Pactolus promoter fragments using Pactolus-expressing (CTMC) and Pactolus-nonexpressing (RAW and MMC) cells extracts. Labeled fragment was incubated without (Lane 1) or with (all other lanes) 3 ug nuclear extract. Antibdoy (4 µg; anti-PU.1, anti-Elk 1, and anti-Oct 1) was added to the binding reaction where indicated. Specific complexes denoted with the bars and lowercase letters are described in the text. (A) The entire minimal promoter region, the pac-108 probe, was incubated with nuclear extracts of RAW cells, MMC, and CTMC. (B) The Pactolus proximal probe (pac proximal) was incubated with nuclear extracts of RAW cells, MMC, and CTMC. (C) The Pactolus distal probe (pac distal) was incubated with nuclear extracts of the RAW, MMC, and CTMC cells.

 
The EMSA shift patterns between the proximal (Fig. 4B) and distal fragments (Fig. 4C) demonstrated that the bulk of PU.1 binding in the RAW and CTMC samples was to the proximal PU.1 site, not the distal site. Thus, the CTMC and RAW samples showed distinct PU.1 supershifts with the proximal fragment; no such shifts were readily evident with the distal fragment. There was little binding of CTMC extract proteins to the distal probe. All three of these cell types express CD18 and PU.1; the simplicity of the EMSA pattern with the Pactolus-expressing CTMC extract versus the more complex patterns seen with the RAW and MMC extracts (both of which do not express Pactolus) suggests additional factors are produced by these cells to block PU.1-dependent transcription of the Pactolus gene.

Mapping protein-binding sites in the Pactolus proximal promoter region
The diffuse nature of banding patterns in the RAW extracts with the entire promoter region was clarified into distinct bands with the proximal probe, including supershift species with the anti-PU.1 antibody. To map more definitively the protein-binding sites within the Pactolus proximal protein region, we performed EMSA experiments using competitive oligonucleotides from regions of the proximal PU.1 site (Fig. 5A ). The extract alone gave at least three distinct bands (Lane 2; Bands a–c), of which Bands a and c were extinguished when the PU.1 site oligo was added (Lane 3). Oligo 38, which possesses the 23 bp, just 3' of the PU.1 site, quenched Band b, as did mutant oligo 38m1. Mutant oligo 38m5 demonstrated some depression of Band b as well, and oligos 38m2, 38m3, and 38m4 had no effect on this band shift. The sequences used in this series of mutations were chosen randomly. These data suggest the binding site for this second protein complex is between the sequences –21 and –34.

This analysis was extended using a different set of oligo competitors (Fig. 5B) . Oligo 55 was created larger than just the minimal PU.1 site; thus, those nucleotides flanking the PU.1 site in the oligo could influence more than just PU.1 binding. Oligo 43 was created to overlap between oligo 55 and oligo 38 (Fig. 5C) , and a mutant form of 43 (43m1) was derived in which CCAG had been converted to TGTC. Oligo 20 was created to possess the first 20 bp of the minimal Pactolus promoter. As shown in Figure 5B , RAW extracts with the Pactolus proximal probe generated the bands observed previously (Lane 2). Addition of oligo 55 alone quenched Bands A and C (known to possess PU.1 from previous supershift assays; Fig. 3B and data not shown), and addition of oligo 43 and 43m1 did not block the formation of any of the bands. When oligo 55 was used in series with oligos 43, 43m1, 38, and 20, the sequence unique to oligo 38 blocked the appearance of Band b, providing additional confirmation that this second protein-binding site is within the –21 to –34 region.

We further investigated the DNA/protein complexes formed with the RAW cell nuclear extracts using a series of sense, nonsense, and mutant competitive oligonucleotides to try to narrow down the nucleotides required for this second protein-binding site. In Figure 6A , a sense and nonsense oligo set, 55 s and 55n, respectively, were used to determine that the band-shift prominently identified in the presence of oligo 55 (shown as Band a) is dependent on the sense sequence. Mutations within the competitive oligo 55 (55m3 and 55m4) blocked the enhanced formation of Band a (and 55m1 and 55m2 did not), indicating those sequence alterations altered a binding site. The mutant oligonucleotides 55m5, 55m6, and 55m7 also did not block the formation of Band a, generating patterns identical to that of 55 s, 55m1, and 55m2 (data not shown). In Figure 6B , the effect of another oligo, 33, was examined. The sequence of this oligo was described previously (see Fig. 5 ) as the second major protein-binding site within the Pactolus proximal sequence. Oligo 55 was used in these assays to promote the formation of Band a; the intensity of this band is depressed when oligo 38 is used, lost with a sense sequence of oligo 33 (33 s), and unaffected with a nonsense oligo 33 (33n). The specific residues within the oligo 33 sequence responsible for factor binding were elucidated with competitive mutant oligonucleotides (Fig. 6C) . Mutant oligo 33m2 had the most dramatic effect (compared with the 33 s oligo), indicating the principal binding site for this factor is centered in Residues 4–6 (TAA) of the oligo 33 sequence. In total, the EMSA data suggest that there are two primary protein-binding sites in the Pactolus proximal promoter region: the PU.1 site (–42 to –48) and a second site within the oligo 33 sequence, defined by oligonucleotide 33m2 as including nucleotides –27 to –29.


Figure 6
View larger version (45K):
[in this window]
[in a new window]
  		Figure 6. Defining protein-binding sites within the Pactolus proximal region. EMSA analysis with the Pactolus proximal probe and RAW cell extracts. (D) Competitive, double-strand oligonucleotides (only single-strand sequences shown) are delineated. (A) Competition of EMSA complexes with sense (55 s) and nonsense (55n) for the –55 to –34 oligonucleotide 35. The four 55 mutant oligos are noted. Nonsense oligonucleotides possess the same nucleotides as sense but in a randomized order. (B) Competition of EMSA complexes with a sense (33 s) and nonsense (33n). Oligo 38 is shown in Figure 5 . (C) Competition of EMSA complexes with mutant forms of oligonucleotide 33. (D) Positions of the series of mutant oligonucleotides are shown. Schematic of the Pactolus proximal probe with the position of native and mutant oligonucleotide sequences shown.

 
A negative regulatory element maps to the second, non-PU.1 protein-binding site
To determine if the –16 to –38 core sequence within the Pactolus promoter possessed a negative regulatory function, we created a series of constructs in which this sequence was inserted into the analogous region of the minimal CD18 promoter [CD18(pac)]. In addition, the analogous CD18 sequence (–23 to –43) was inserted in the minimal Pactolus promoter [pac(CD18); Fig. 7 ]. These constructs were then transfected into RAW cells, along with controls, and analyzed. The introduction of the CD18 sequence into the minimal Pactolus promoter demonstrated a level of expression roughly equivalent to that of the minimal CD18 promoter, suggesting the loss of negative regulation. In contrast, the insertion of the putative Pactolus negative regulatory sequence into the CD18 minimal promoter resulted in a suppression of reporter activity roughly equivalent to the native Pactolus promoter (pac-79). As a control, a random, nonsense sequence was also inserted within the same region of the CD18 promoter to ensure that those CD18 sequences were not positive regulatory sequences, and their loss would not repress reporter activity; this control construct gave a level of expression comparable with the CD18-88 minimal promoter control (data not shown). These data suggest that the second, non-PU.1, protein-binding site within the Pactolus promoter is capable of conferring transcriptional suppression on the Pactolus and CD18 promoters.


Figure 7
View larger version (15K):
[in this window]
[in a new window]
  		Figure 7. Reporter gene analysis of the putative negative regulatory region of the Pactolus minimal promoter. A series of constructs was prepared (see Materials and Methods) in which the internal 18 bp (denoted by bold italics) from the Pactolus and CD18 minimal promoters were substituted. Only a portion of the promoter region is shown; the proximal PU.1 site is denoted with the dark underline. RAW cells were transiently transfected as described previously. The Pactolus-79 and CD18-88 promoters were used as controls; luciferase values were based against the CD18-88 promoter as 100%. Data are averages of three different experiments, and each assay was done in duplicate. Error bars represent standard deviation.

 
Suppression of Pactolus expression in RAW cells is linked to blocked binding of PU.1 to the Pactolus promoter region
The preceding EMSA data suggest that RAW cell-derived PU.1 can bind to the Pactolus promoter in vitro, although the gene is transcriptionally silent in that cell. To determine if PU.1 actually binds the Pactolus promoter sequences in vivo, we used a modified ChIP using RAW cell chromatin. RAW cells were treated with formaldehyde to covalently link proteins to the chromatin, and the DNA was sheared by sonication and fragments immunoprecipitated with anti-PU.1, anti-Oct 1, and an isotype control antibody. Purified fragments were then amplified with PCR and analyzed by gel electrophoresis. As shown in Figure 8A , the DNA fragments immunoprecipitated with PU.1 did include the CD18 promoter but not the Pactolus promoter. As a negative control, the CD21 promoter, which contains Oct 1-binding sites but not PU.1 sites, was purified in the Oct 1 immunoprecipitation but not the PU.1 isolation. CD21 is not expressed in RAW cells. Finally, the actin gene, which does not possess PU.1- or Oct 1-binding sites, was negative for both immunoprecipitations (as well as was the isotype control).


Figure 8
View larger version (9K):
[in this window]
[in a new window]
  		Figure 8. Chromatin immunoprecipitation of PU.1 binding to Pactolus and CD18 promoters. (A) RAW cells were treated with formaldehyde, lysed, and sonicated. DNA/protein complexes were immunoprecipitated with whole rabbit IgG (Iso), Oct 1 antibody, and PU.1 antibody. Input is the starting material for the immunoprecipitation. Formaldehyde cross-linking was resolved, and PCR was preformed using primers specific for the Pactolus, CD18, and CD21 promoters. PCR, using primers for the actin gene, was used as a control for genomic DNA contamination. (B) ChIP analysis of RAW cells essentially as above, except the PCR amplication products were quantified by real-time PCR. Calculations to determine fold enrichment are described in Materials and Methods.

 
This experiment was repeated and analyzed by real-time PCR, in which the relative levels of precipitated gene sequences were compared directly with the internal actin control (see Materials and Methods). As shown in Figure 8B , the CD18 promoter was isolated by PU.1 precipitation (but not with the isotype control antibody), and the Pactolus promoter was not. The data from these two panels demonstrate that the Pactolus promoter within the RAW cell is not bound with PU.1 but that the CD18 promoter is. These assays cannot discriminate between PU.1 binding to the proximal- or distal-binding sites; thus, binding to both sites in the Pactolus promoter in RAW cells must be blocked. Whether or not the protein binding to the –22 to –34 site physically blocks PU.1 from binding in vivo to the Pactolus promoter site(s) is not known.

Defining the Snail family member Smuc as a protein binding within the Pactolus proximal region
Scanning the Pactolus proximal region for DNA-binding protein sites identified two distinct motifs for members of the Snail and Smad families of transcriptional regulators (Fig. 9A ) [32 33 34 35 36 ]. Both are present as tandem inverted repeats and are partially encompassed by the oligo 33-blocking site. The 33m2-altered sequences described in Figure 6 are the TAA residues within the 3' Snail family site. Members of the Smad and Snail families, many of which are expressed in bone marrow-derived cells, are known to function as negative transcriptional regulators. We examined the expression of various Smad and Snail family members by RT-PCR. Total bone marrow was separated by a Percoll step gradient, and the cells in the bottom pellet (enriched for maturing granulocytes) and cells in the top layer (enriched for less-dense cells, such as lymphocytes, monocytes, and macrophages) were isolated. As shown in Figure 9B , the granulocyte fraction possessed transcripts primarily for Smad 4, and the less-dense fraction possessed transcripts for Smad 3, 4, 5, and 7. Of the various Snail family members, Smuc transcripts were identified within the top but not the bottom phase of this gradient (Fig. 9C) . Our model, based on the data in this report, is that such an inhibitory factor would be transcribed preferentially in cells that express PU.1 but do not express Pactolus (cells within the top, not bottom, layer of the gradient); thus, Smad 3, 5, and 7 and Smuc were likely candidates. EMSA supershift analysis was carried out with antisera specific for PU.1, Smuc (Smc), and the three Smad proteins using RAW nuclear extracts. As shown in Figure 9D , PU.1 specifically bound to the complex marked as Band b, and the Smuc antibody resulted in the loss of Bands a and b. These data suggest Band b consists of a complex of PU.1, Smuc, and the DNA probe, and Band a is likely Smuc-binding alone [or in a complex with a protein(s) other than PU.1]. The three antibodies against the Smad proteins (Smad 3, 5, and 7) had no effect on any of the bands, indicating a lack of relevant binding sites for these proteins within the fragment. The assay was repeated in the presence of oligo 55 to enhance the production of the complex marked "c". The PU.1 and Smad antisera had no effect on this band, and the Smuc antisera not only supershifted Band a but also depressed the quantity of Band c. Supershift assays were also done with other members of the Snail family (SnaI and Slug); no specific complexes were detected (data not shown).


Figure 9
View larger version (56K):
[in this window]
[in a new window]
  		Figure 9. Smuc but not Smad protein complexes are evident within the Pactolus proximal region. (A) Localization of two sets of putative protein-binding motifs within the Pactolus proximal fragment. The proximal PU.1 site is marked within the box. Snail and Smad family motifs are tandem, inverted repeats and lie between the proximal PU.1 site and the transcription initiation site. (B) RT-PCR analysis of transcripts obtained from neutrophil-enriched (BOTTOM) and macrophage/lymphocyte-enriched (TOP) bone marrow samples and total splenic cells (SPLEEN). Gene-specific oligonucleotides for the noted gene products are within Table 1 ; all span at least one intron. dH20 is a water control with no oligonucleotides added. RT-PCR cycle numbers: SMAD 1–9, 25 cycles; ß-actin, 16 cycles; Pactolus (PAC 3'), 26 cycles; PU.1, 26 cycles. Products analyzed by DNA sequencing gel electrophoresis. (C) Same as B above except with Smuc oligonucleotides (25 cycles). (D) EMSA supershift analysis with RAW nuclear extracts, the Pactolus proximal probe, and antibodies against PU.1, Smuc (Smc), or Smad 3, 5, or 7 (SM3, SM5, SM7, respectively). Oligo 55 was added to the binding reactions as shown. Bands denoted as a–c are described in text. (E) Same as C, except nuclear extracts were obtained from cultured MMC. Antibody against EF1 was used as control. (F) Western blot analysis of total cell lysates from RAW macrophage and 2PK3 B cell lines and total thymus. Loading control of actin protein is shown. Molecular weight ladder is marked at right of panel.

 
MMC and RAW cells express PU.1 and CD18 but do not express Pactolus; thus, we extended these supershift EMSA assays using extracts from MMC to determine if they would also form Smuc complexes with the Pactolus proximal fragment (Fig. 9E) . The formation of complexes of this fragment with the MMC extract was enhanced in the presence of oligo 43 (see Fig. 5 ). Addition of anti-PU.1 and anti-EF1 had no effect on the MMC band shifts; however, incubation with the anti-Smuc antibody specifically removed Band a. Thus, Smuc complexes can also form with the Pactolus promoter fragment with the MMC extracts. To verify that the commercial anti-Smuc reagent was specific for the protein, we performed Western blot analysis using total extracts from RAW cells, 2PK3, and thymocytes (known to produce Smuc). As shown in Figure 9F , the anti-Smuc antibody is specific for a doublet of ~32,000mr, the size reported for Smuc.

To further determine if the Smuc protein binds directly to the negative regulator site in the proximal promoter of Pactolus, we created Probe 38, (Fig. 10A ) a double-stranded oligonucleotide corresponding to the sequence responsible for transfer of repression, –16/–38. An additional oligonucleotide set was prepared in which the three nucleotides, TAA, presumably required for Smuc binding, were mutated to a CGC sequence (Probe 38 MUTANT). These probes were used in supershift and competition EMSA gels with RAW and MMC nuclear extracts. As shown in Figure 10B , the RAW nuclear extracts with the two probes identified a doublet-shift (denoted A) with the correct Probe 38. This doublet was missing from the Mutant 38 probe shift. This doublet was supershifted in the presence of the Smuc antibody and was lost in the presence of cold, competing Probe 38 DNA but not in the presence of cold, competing Probe 38 Mutant (38M) DNA. Shown in Figure 10C , this doublet formed with the RAW extracts is unaffected in the presence of antisera against a variety of the Smad proteins. Similarly, the use of MMC extracts identified a single, specific band (denoted B), which was also lost in the presence of the Smuc antisera and the competing cold Probe 38 sequence but was unaffected by the Smad antisera and the cold Probe 38 Mutant sequence. Similar EMSA analysis was also performed with CTMC nuclear extract using the Probe 38 or Probe 38 mutant sequences (data not shown). As expected, no specific EMSA shifts were detected with these extracts.


Figure 10
View larger version (30K):
[in this window]
[in a new window]
  		Figure 10. Defined EMSA analysis of Smuc binding to the putative negative regulator site. EMSA supershift and competition assays were performed as described previously. (A) Sequence of sense (Probe 38) and mutant Probe 38 corresponding to the negative regulator site (normal and mutated nucleotides are enlarged). (B) EMSA supershift and competition analysis with RAW nuclear extract with labeled Probe 38 (left panel) and labeled Probe 38 Mutant (right panel). NE denotes nuclear extract alone, and the bracket A denotes the sequence-specific band. Smuc indicates the addition of that antisera. Unlabeled Probe 38 (COLD 38) and Probe 38 Mutant (COLD 38M) were used as competition oligonucleotides. The panel was cut off above the free probe. (C) EMSA supershift and competition analysis with RAW and MMC nuclear extracts with labeled Probe 38 performed as above. Brackets A and B denote Smuc-specific bands in RAW and MMC nuclear extracts, respectfully. The free probe bands were omitted from this panel to allow for increased size and clarity of the data. Antibodies and cold, competing oligonucleotides in each lane are noted.

 

   DISCUSSION
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
REFERENCES
 
In this report, we have analyzed the structure and function of the Pactolus gene minimal promoter. The Pactolus gene, found in mice and rats, is a direct duplication of the CD18 gene; the human genome does not contain a Pactolus gene. The murine CD18 and Pactolus gene promoters share a number of highly conserved characteristics, including two tandem PU.1-binding sites separated by 16–18 residues, and the absence of a TATA box. These similarities are in stark contrast to the dissimilar tissue specificity of expression; CD18 is expressed by lymphocytes, macrophages, and granulocytes, and Pactolus is restricted to cells of the neutrophil lineage. The key question addressed in this study is how this differential transcriptional regulation is achieved when the basic structures of the promoters are so identical.

The RAW macrophage cell line that expresses CD18 but does not express Pactolus was used in transfection reporter, EMSA, and ChIP assays to define functional elements of the Pactolus promoter. Reporter assays suggest that the negative control elements that block Pactolus transcription in RAW cells are contained within the minimal promoter region. EMSA clearly demonstrates PU.1 binding to the proximal PU.1 site of Pactolus in vitro, but no PU.1 binding was detected in vivo as determined by ChIP assay. Reporter assays further suggest that the Pactolus promoter possesses a site within the –16 to –38 region that inhibits transcription and also binds an unknown factor. This region of the promoter is noteworthy in that two sets of inverted repeat motifs are apparent: one set known to be recognized by members of the Snail family and the second set recognized by Smad proteins. Antibody supershift analysis of this region with reagents against members of the Snail and Smad families indicated complexes with Smuc, one of the Snail proteins.

The Snail-related gene family is highly conserved from Drosophila to mammals, the protein products of which have been described as involved with cell differentiation [32 , 37 ]. These proteins are basic helix-loop-helix transcription factors, which bind to a consensus E-box motif. There are four members of this family in mammals: SnaI, Slug, Smuc, and Scrt, which is also known as Scratch and appears to be involved primarily with neural development [38 ]. Smuc is expressed abundantly in skeletal muscle tissue, where it has been shown to regulate MyoD activity by competing for E box sequences and inhibiting the differentiation of myoblasts [35 , 39 ]. The Smuc protein consists of two distinct sets of domains: the zinc finger region for DNA binding at the C terminus of the protein and an N-terminal repressor domain. Our data suggest that Smuc directly inhibits PU.1-dependent transcription of the Pactolus gene, although further experimentation is required to fully document that conclusion. Ectopic expression of Smuc within the CTMC cells should be especially informative, as small interfering RNA suppression of Smuc transcription (presumably inducing Pactolus transcription) in macrophages may well prove lethal to the cell. The SnaI and Slug proteins offer functional parallels to our model for Smuc. SnaI and Slug have been shown to repress E-cadherin and Claudin-1 transcription by binding to proximal E-boxes within the gene’s promoter, presumably blocking access of the promoter to activating transcription factors [33 , 40 , 41 ]. In addition, SnaI has been shown to recruit Sin3A/histone deacetylase 1/2 (HDAC1/HDAC2) complexes to the E-cadherin gene to affect transcriptional repression [42 ].

A survey of the literature identified other genes that are also negatively regulated by factors binding to similarly arranged motifs in their promoters. For example, the terminal transferase gene promoter possesses an activating ets family-binding site, which requires occupancy for gene activation. However, Ikaros dimers can suppress transcription of the gene by occupying an adjacent site that blocks the ets-activating protein from binding and assisting in transcription [43 ]. A similar motif was observed in the promoter of the acute-phase reactant protein serum amyloid A1 gene, which similarly serves to suppress gene transcription [44 ].

We intentionally focused our analyses in this report on a Pactolus promoter element of 1081 bp (or less), comparing its activity to a minimal promoter element of the mouse CD18 promoter of 88 bp. Analysis of evolutionarily conserved regions (ECR; which can identify transcriptional control sites) between the mouse and human CD18 genes did not identify any such elements 5' of coding sequences (conservation between these two genes was limited to coding regions; http://ecrbrowser.dcode.org) [45 ], suggesting that if additional regulatory sites are present in these genes, they have not been conserved. Analyses of the human CD18 gene have not identified any transcriptional control sequences other than those within the minimal promoter element [8 9 10 11 , 46 ]. ECR analysis between the rat and mouse Pactolus genes showed a high level of sequence homology in all regions of the gene (and those genes flanking), indicating this region of the genome has seen little sequence diversification during rodent evolution. It is interesting that ECR analysis did not define homology between the mouse Pactolus (and mouse CD18) minimal promoter sequence compared with the human CD18 promoter element, although they do share significant sequence conservation (Fig. 1B) .

The Pactolus gene is clearly a direct duplication of CD18 [4 ], yet the Pactolus protein has a number of characteristics that set it aside from CD18. Pactolus does not require an {alpha}-chain for cell surface expression, Pactolus is held in different granules within the neutrophil than CD18, Pactolus is heavily modified by sialic acid residues, and CD18 is not, and the Pactolus gene is only expressed in neutrophils, and CD18 is expressed in virtually all cells of bone marrow derivation [1 2 3 ]. We have constructed a variety of Pactolus-containing expression constructs (colinear with the neo-resistance cassette) and retrovirus constructs to transfect/infect murine T and B cells and macrophages, as well as control fibroblasts such as Chinese hamster ovary (CHO) cells. These constructs do allow for constitutive, cell-surface expression of a partially glycosylated form of Pactolus on CHO cells; however, no neoresistant, Pactolus-expressing cells were obtained from transfections or infections into T, B, or macrophage cell lines (data not shown). Our conclusion from these experiments is that the expression of Pactolus in cells other than neutrophils (such as T and B cells and macrophages) is lethal. Therefore, it appears to be critical that the Pactolus promoter not be functional in an inappropriate cell.

PU.1 is presumed to be the key positive transcription factor controlling CD18 and Pactolus expression in neutrophils. PU.1 is also active and required for B cell and macrophage-lineage development. Thus, blocking Pactolus transcription in B cells and macrophages cannot be accomplished by down-modulating the expression of PU.1 but instead, must be accomplished by blocking PU.1 functional binding to the Pactolus promoter. If this model is correct for Pactolus regulation, then would other genes that also depend on PU.1 transcriptional activation but cannot be expressed in macrophages and lymphocytes use the same regulatory components? For example, human proteinase-3 expression requires PU.1 transactivation in maturing neutrophils, yet its expression is lost in the maturation of promyelocytic cells [47 ]. Expression of this proteinase, which is usually localized in azurophilic granules within the neutrophil, could easily be considered deleterious to the health of a maturing macrophage or B cell. The same type of transcriptional control scenario has been described for other neutrophil and eosinophil granule constituents that also demonstrate a dependence on PU.1 but whose expression is absent in macrophage and B cell lineages [14 ]. It will be of interest to determine if the promoters of these genes (and others like them that display the same transcriptional phenotype) possess the same control sequences that we have defined within the Pactolus minimal promoter.


View this table:
[in this window]
[in a new window]
  		Table 1. Oligonucleotides Used in PCR and RT-PCR Analyses: Listed 5'–3'

 

   ACKNOWLEDGEMENTS
 
This research was supported by National Institutes of Health (NIH) Grants AI-42032, AI-060618, and AI-24158 (J. H. W.) and NIH Grants AI-32223 and AR-43521 (J. J. W.). The authors particularly thank Sean Garrison and Andrias Hojaard for experimental assistance and all members of the Weis laboratories for insightful criticisms.


   FOOTNOTES
 
1 These authors contributed equally to this work. Back

Received July 14, 2005; revised April 10, 2006; accepted April 11, 2006.


   REFERENCES
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 REFERENCES
 

  1. Chen, Y., Garrison, S., Weis, J. J., Weis, J. H. (1998) Identification of Pactolus, an integrin ß subunit-like cell-surface protein preferentially expressed by cells of the bone marrow J. Biol. Chem. 273,8711-8718[Abstract/Free Full Text]
  2. Garrison, S., Hojgaard, A., Patillo, D., Weis, J. J., Weis, J. H. (2001) Functional characterization of Pactolus, a ß-integrin-like protein preferentially expressed by neutrophils J. Biol. Chem. 276,35500-35511[Abstract/Free Full Text]
  3. Garrison, S., Hojgaard, A., Margraf, R., Weis, J. J., Weis, J. H. (2003) Surface translocation of Pactolus is induced by cell activation and death, but is not required for neutrophil migration and function J. Immunol. 171,6795-6806[Abstract/Free Full Text]
  4. Margraf, R. L., Chen, Y., Garrison, S., Weis, J. J., Weis, J. H. (1999) Genomic organization, chromosomal localization, and transcriptional variants of the murine Pactolus gene Mamm. Genome 10,1075-1081[CrossRef][Medline]
  5. Gahmberg, C. G. (1997) Leukocyte adhesion: CD11/CD18 integrins and intercellular adhesion molecules Curr. Opin. Cell Biol. 9,643-650[CrossRef][Medline]
  6. Mizgerd, J. P., Kubo, H., Kutkoski, G. J., Bhagwan, S. D., Scharffetter-Kochanek, K., Beaudet, A. L., Doerschuk, C. M. (1997) Neutrophil emigration in the skin, lungs, and peritoneum: different requirements for CD11/CD18 revealed by CD18-deficient mice J. Exp. Med. 186,1357-1364[Abstract/Free Full Text]
  7. Walzog, B., Scharffetter-Kochanek, K., Gaehtgens, P. (1999) Impairment of neutrophil emigration in CD18-null mice Am. J. Physiol. 276,G1125-G1130
  8. Rosmarin, A. G., Caprio, D., Levy, R., Simkevich, C. (1995) CD18 (ß 2 leukocyte integrin) promoter requires PU.1 transcription factor for myeloid activity Proc. Natl. Acad. Sci. USA 92,801-805[Abstract/Free Full Text]
  9. Bottinger, E. P., Shelley, C. S., Farokhzad, O. C., Arnaout, M. A. (1994) The human ß 2 integrin CD18 promoter consists of two inverted Ets cis elements Mol. Cell. Biol. 14,2604-2615[Abstract/Free Full Text]
  10. Rosmarin, A. G., Caprio, D. G., Kirsch, D. G., Handa, H., Simkevich, C. P. (1995) GABP and PU.1 compete for binding, yet cooperate to increase CD18 (ß 2 leukocyte integrin) transcription J. Biol. Chem. 270,23627-23633[Abstract/Free Full Text]
  11. Rosmarin, A. G., Luo, M., Caprio, D. G., Shang, J., Simkevich, C. P. (1998) Sp1 cooperates with the ets transcription factor, GABP, to activate the CD18 (ß2 leukocyte integrin) promoter J. Biol. Chem. 273,13097-13103[Abstract/Free Full Text]
  12. Bush, T. S., St Coeur, M., Resendes, K. K., Rosmarin, A. G. (2003) GA-binding protein (GABP) and Sp1 are required, along with retinoid receptors, to mediate retinoic acid responsiveness of CD18 (ß 2 leukocyte integrin): a novel mechanism of transcriptional regulation in myeloid cells Blood 101,311-317[Abstract/Free Full Text]
  13. Kwiatkowski, B. A., Embree, L. J., Ritchie, K. A., Hickstein, D. D. (1996) Human leukocyte integrin CD18 promoter directs low levels of expression of a mutated human CD4 reporter gene in leukocytes of transgenic mice Biochem. Biophys. Res. Commun. 222,601-606[CrossRef][Medline]
  14. Gombart, A. F., Kwok, S. H., Anderson, K. L., Yamaguchi, Y., Torbett, B. E., Koeffler, H. P. (2003) Regulation of neutrophil and eosinophil secondary granule gene expression by transcription factors C/EBP {epsilon} and PU.1 Blood 101,3265-3273[Abstract/Free Full Text]
  15. Panopoulos, A. D., Bartos, D., Zhang, L., Watowich, S. S. (2002) Control of myeloid-specific integrin {alpha} Mß 2 (CD11b/CD18) expression by cytokines is regulated by Stat3-dependent activation of PU.1 J. Biol. Chem. 277,19001-19007[Abstract/Free Full Text]
  16. Li, S. L., Schlegel, W., Valente, A. J., Clark, R. A. (1999) Critical flanking sequences of PU.1 binding sites in myeloid-specific promoters J. Biol. Chem. 274,32453-32460[Abstract/Free Full Text]
  17. Li, S. L., Valente, A. J., Qiang, M., Schlegel, W., Gamez, M., Clark, R. A. (2002) Multiple PU.1 sites cooperate in the regulation of p40(phox) transcription during granulocytic differentiation of myeloid cells Blood 99,4578-4587[Abstract/Free Full Text]
  18. Voso, M. T., Burn, T. C., Wulf, G., Lim, B., Leone, G., Tenen, D. G. (1994) Inhibition of hematopoiesis by competitive binding of transcription factor PU.1 Proc. Natl. Acad. Sci. USA 91,7932-7936[Abstract/Free Full Text]
  19. Simon, M. C. (1998) PU.1 and hematopoiesis: lessons learned from gene targeting experiments Semin. Immunol. 10,111-118[CrossRef][Medline]
  20. Smith, T. J., Ducharme, L. A., Weis, J. H. (1994) Preferential expression of interleukin-12 or interleukin-4 by murine bone marrow mast cells derived in mast cell growth factor or interleukin-3 Eur. J. Immunol. 24,822-826[Medline]
  21. Dignam, J. D., Lebovitz, R. M., Roeder, R. G. (1983) Accurate transcription initiation by RNA polymerase II in a soluble extract from isolated mammalian nuclei Nucleic Acids Res. 11,1475-1489[Abstract/Free Full Text]
  22. Chirgwin, J. M., Przybyla, A. E., MacDonald, R. J., Rutter, W. J. (1979) Isolation of biologically active ribonucleic acid from sources enriched in ribonuclease Biochemistry 18,5294-5299[CrossRef][Medline]
  23. Tan, S. S., Weis, J. H. (1992) Development of a sensitive reverse transcriptase PCR assay, RT-RPCR, utilizing rapid cycle times PCR Methods Appl. 2,137-143[Medline]
  24. Johnson, T. A., Wilson, H. L., Roesler, W. J. (2001) Improvement of the chromatin immunoprecipitation (ChIP) assay by DNA fragment size fractionation Biotechniques 31,740-742[Medline]
  25. Dedon, P. C., Soults, J. A., Allis, C. D., Gorovsky, M. A. (1991) A simplified formaldehyde fixation and immunoprecipitation technique for studying protein-DNA interactions Anal. Biochem. 197,83-90[CrossRef][Medline]
  26. Weinmann, A. S., Farnham, P. J. (2002) Identification of unknown target genes of human transcription factors using chromatin immunoprecipitation Methods 26,37-47[CrossRef][Medline]