| Nucleic Acids Research | Pages |
Cloning and characterization of human oncostatin M promoter
Introduction
Materials And Methods
Materials
Cell lines and culture conditions
Isolation and characterization of genomic OSM clones
Primer extension assay
RNase protection assay (RPA)
Plasmids for transient transfections
Transient transfection assays
Mini-preparation of nuclear extract
Wild-type and mutant oligonucleotides used in EMSA
Electrophoretic mobility shift assay (EMSA)
Results
Cloning and sequencing of the OSM promoter
Determination of the 5[prime]-end of OSM mRNA
OM promoter elements
Sequences required for basal activity of the OSM promoter
Transcriptional activation of the OSM promoter in response to GM-CSF requires STAT sequences
STAT5b binds to the M-STAT element and drives activated transcription
Discussion
Acknowledgements
References
Cloning and characterization of human oncostatin M promoter
Received May 18, 1999; Revised and Accepted October 8, 1999
ABSTRACT Oncostatin M (OSM), an IL-6 subfamily cytokine, inhibits proliferation and causes morphological changes in many tumor cell lines. GM-CSF, phorbol-12-myristate-13-acetate (PMA), and lipopolysaccharide (LPS) induce OSM expression. To investigate the mechanisms governing OSM promoter activity, we have cloned and partially sequenced an 8.5 kb fragment of human genomic DNA immediately 5[prime] of the OSM coding region and mapped the transcription start site. Transient transfection assays with a series of 5[prime] deletion plasmids demonstrated maximal reporter activity in U937 cells with a minimum 304 bp construct. The 5[prime]-proximal region of the human OSM gene contains a C/EBP consensus element around -45 bp and several GC-rich regions around -60, each of which is responsible for basal promoter activity. Electrophoretic mobility shift assay coupled with supershift analysis confirmed the presence of a cis-acting binding site for activated STAT5 complexes following GM-CSF treatment. Furthermore, transient transfection studies demonstrated a loss of GM-CSF responsiveness in reporter constructs containing mutations within this STAT element. Our results establish that C/EBP and an as yet unidentified GC-rich binding transcription factor are responsible for basal OSM promoter activity, while GM-CSF-stimulated OSM expression is driven by activated STAT5 complexes binding to a cis-acting STAT element on the OSM promoter.
INTRODUCTION
Human oncostatin M, produced primarily by activated macrophages and T cells (1,2), is a multifunctional hematopoietic cytokine that exists in its mature form as a 28 kDa glycoprotein translated from an ~1.8 kb mRNA (3). Immature OSM is synthesized as a 252 amino acid precursor from which ~25 amino acids of signal sequence and 31 amino acid residues at the C-terminus of the pro-OSM are cleaved during post-translational modification prior to secretion (4). The human OSM gene is located on chromosome 22q12 and is immediately adjacent to the gene for leukemia inhibitory factor (LIF) (5,6). OSM and LIF are both members of the IL-6 subfamily of cytokines based on gene structure, amino acid homology and protein structure, and they are expressed in many of the same cell types (7-9).
OSM, first isolated from PMA-treated U937 lymphoma conditioned medium, was initially characterized based on its ability to inhibit the growth of the A375 melanoma cell line (1). It has shown a cytostatic growth inhibitory effect on a number of other tumor-derived cell lines including breast cancer (10-12). OSM has also been shown to induce differentiation of breast cancer and leukemia cell lines in culture (13,14). Conversely, OSM acts as an autocrine growth stimulator of AIDS-associated Kaposi sarcoma spindle cells (15) and as a mitogen for vascular smooth muscle cells (16). In addition to its role as a growth modulator, OSM seems to play a role in inflammatory disease, as evidenced by increased synovial fluid OSM mRNA in persons with rheumatoid arthritis (17,18), suggesting that OSM may contribute to joint inflammation in rheumatoid arthritis through perturbations in proteoglycan metabolism. Most strikingly, OSM expression appears to correlate with malignancy, as indicated by immunohistochemistry studies performed in our laboratory that show positive immunoreactivity in both tumor cells and invading inflammatory cells within the tumor sections (abstracts presented at the AACR Annual Meetings, 1998 and 1999).
Some other functions of OSM include up-regulation of hepatic LDL receptor in vitro and in vivo (19,20), inhibition of embryonic stem cell differentiation (21,22) and up-regulation of tissue inhibitor of metalloproteinases-1 (23). Studies by Wallace et al. (24) have shown that OSM augments platelet counts by acting as a megakaryocyte maturation factor without increasing other circulating blood cell counts. Wijelath et al. (25) suggest that OSM released at the site of vascular injury could stimulate angiogenesis by inducing bFGF synthesis, resulting in endothelial cell proliferation, migration and acquisition of spindle shape morphology. As demonstrated by Liu et al. (26), OSM, similar to the effects of VEGF, VRP and bFGF, enhances tyrosine phosphorylation of RAFTK, thereby coordinating signaling from the cell surface receptors of these growth factors to the cytoskeleton and providing a means to modulate cell growth and function. These combined results indicate that OSM, like many other growth regulators, is multifunctional.
OSM is known to exert its biological effects through two separate receptor complexes. The type I receptor consists of a heterodimer between the shared OSM/LIF receptor (LIFR) and the signal transducing molecule gp130 (27,28). The type II receptor complex consists of a heterodimer between an OSM-specific receptor (OSMR) and gp130 (29). Binding of OSM to either complex initiates receptor signaling through the JAK/STAT pathway, with preferential activation of STAT3 (30). The OSM-specific receptor is necessary, but not sufficient, for a growth response to OSM, as evidenced by the observation that cell lines lacking OSM receptors do not respond to OSM, while some but not all cells having OSM receptors show a response (10,31).
Progress in understanding the OSM expression mechanism in response to cytokine induction has come from studies of the murine homolog (mOSM) (32). However, no specific deletions or characterization of factors binding to this region were performed and verified. In addition, more and more reagents, such as cisplatin, phytohemagglutinin, lipopolysaccharide (LPS), and viral infections, are also strong activators of both murine and human OSM expression (3,33-35 and references therein). The exact mechanism determining OSM expression, however, remains unknown.
Accordingly, we have undertaken the cloning and characterization of the 5[prime]-flanking region of the human OSM gene in order to better understand transcriptional regulation of OSM. Using chimeric OSM promoter-luciferase reporter constructs, we demonstrate that this region functions as a constitutively active transcriptional promoter, which, like its murine counterpart (32,36), is capable of being stimulated by GM-CSF, but through a different element. We have localized the cis-acting element responsible for OSM induction to -193, and confirm its identity as a consensus binding site for activated STAT5 homodimers. These results provide a firm molecular basis for future studies investigating possible therapeutic manipulation of this clinically and physiologically important cytokine.
MATERIALS AND METHODS
The nucleotide sequence for the human oncostatin M proximal promoter has been deposited in the GenBank database under accession no. AF129855.
Materials
PMA was purchased from Calbiochem (catalog no. 524400-Q), poly(dI-dC)·poly(dI-dC) from Pharmacia Biotech (Piscataway, NJ). Oligonucleotides were synthesized by Gibco BRL (Life Technologies, Grand Island, NY). Rabbit polyclonal antibodies, specifically recognizing STAT1, STAT3, STAT5b and C/EBP[beta] were purchased from Santa Cruz Biotechnology (Santa Cruz, CA). GM-CSF, IFN[gamma] and TNF[alpha] were purchased from PeproTech (Rocky Hill, NJ). All other chemicals were from Sigma (St Louis, MO), unless otherwise indicated.
Cell lines and culture conditions
The human promonocytic cell line U937 and the human hepatocellular carcinoma cell line HepG2 were obtained from the American Type Culture Collection (ATCC, Rockville, MD). U937 cells were grown in RPMI 1640 medium supplemented with 10% heat-inactivated fetal calf serum (FCS) (HyClone Laboratories, Logan, UT), 100 U penicillin and 100 µg/ml streptomycin, and maintained at 37°C in a humidified atmosphere containing 5% CO2. HepG2 cells were maintained in minimal essential Eagle's medium (EMEM; Sigma, St Louis, MO) containing 10% fetal bovine serum (Summit Biotechnology, Fort Collins, CO), 100 U/ml penicillin G and 100 µg/ml streptomycin. Cell cultures were always passaged twice a week to maintain a cell density of between 2 × 105 and 1 × 106 cells/ml. Cells were counted in a hemocytometer chamber and viability assessed by 0.1% trypan blue exclusion. For transient transfection experiments, U937 cells were seeded at 2.5 × 105 cells/ml, electoporations were performed 16 h later, and inducers were included in the post-electroporation medium and then left in the culture medium for the desired time period. The concentrations of the inducers were as follows: 400 nM PMA, 2 ng/ml GM-CSF, 50 ng/ml OSM, 10 ng/ml IFN[gamma] and 5 ng/ml TNF[alpha]. The same concentrations were used to treat cells for purification of nuclear extracts for gel shift assays.
Isolation and characterization of genomic OSM clones
The OSM cDNA was used to screen a human genomic P1 library. A positive clone was isolated and subsequent sequence analysis confirmed its identity. A DNA fragment extending 8.5 kb upstream from the first ATG was subcloned into the pGL3Basic luciferase plasmid vector (Promega, Madison, WI). The OSM coding sequence was removed from this clone by PCR-coupled DNA subcloning. To accomplish this, a sense primer containing luciferase vector sequence with an appended MluI site, together with an antisense primer, designed from the OSM 5[prime]-untranslated region with an appended SalI restriction site, were used to carry out PCR using the aforementioned 8.5 kb subclone as template. This PCR product was then digested with MluI and SalI and cloned into the MluI and XhoI sites of the pGL3 basic luciferase vector. Then 8.3 kb of the 5[prime]-portion of the insert DNA derived from PCR was replaced by the corresponding genomic fragment by restriction enzyme EcoRI digestion and cloning. The 0.2 kb of the 3[prime]-portion derived from the PCR was sequenced and confirmed no mutations.
Primer extension assay
An antisense primer, OSM1A (5[prime]-AGCGTCCTCTGTGTGA-GCAGTAC-3[prime]), was designed from the first exon of the OSM gene. Two micrograms of mRNA from control or GM-CSF (2 ng/ml, 1 h)-treated U937 cells were annealed to 0.1 pmol of 32P-labeled OSM1A primer. Extension was carried out in AMV reverse transcriptase extension buffer in the presence of 2.8 mM sodium pyrophosphate and 1 U of AMV reverse transcriptase. The product of the extension reaction together with a dideoxy sequencing reaction as size marker were run on a 6% denaturing polyacrylamide gel and visualized by autoradiography.
RNase protection assay (RPA)
RPA was performed using a 32P-labeled RNA probe prepared by transcribing a 285 bp EcoRI-ScaI (-194 to +91) fragment of the OSM gene cloned into pBluescript KS vector. The probe was hybridized to 2 µg of mRNA, which was purified from control, PMA-treated (400 nM, 1 h), or GM-CSF-treated (2 ng/ml, 1 h) U937 cells and subsequently digested with 3.5 µg and 25 U of RNases A and T1, respectively (Boehringer Mannheim, Indianapolis, IN). The products together with a size marker were analyzed on an 8% denaturing polyacrylamide gel and visualized by autoradiography.
Plasmids for transient transfections
A promoterless luciferase reporter vector, pGL3-Basic (Promega), was used in the course of these studies. The expression plasmid pRL (sea pansy Renilla reniformis gene under control of an SV40 promoter) was also used throughout as an internal control for transfection efficiency (Promega). The 0.9 kb and 304, 189, 90 and 61 bp constructs were all derived from the 8.5 kb clone and generated respectively with SacI, BstNI, BstBI, StuI and BspHI restriction enzyme digestions. The 3STAT, -3STAT, 2STAT, -2STAT, 1STAT, -1STAT and -31 plasmid constructs were all generated by PCR by utilizing a series of synthetic sense oligonucleotide primers derived from the respective part of the promoter. All the above plasmids have 3[prime]-ends anchored 11 bp upstream of the first ATG (+53 to +55). All plasmids, including mutant constructs, were sequenced to confirm identity and desired alterations. Sequence analysis was performed at the DNA Service Laboratory, Bio-technology Center, Utah State University (Logan, UT).
Transient transfection assays
Transfection of tissue culture cells and luciferase assays were carried out as described by Ma et al. (37). The dual luciferase reporter assay system (Promega) was used for the transient transfection studies. Briefly, 18 µg plasmid DNA (weight amounts of insert-containing plasmids adjusted to be equimolar with the control) together with 10 ng of normalization pRL plasmid in serum-free RPMI 1640 medium, were transfected into U937 cells by electroporation at 2950 µF capacitance, 186 [Omega] resistance and 200 V charging voltage using a BTX600 electroporation unit (BTX Inc., San Diego, CA). The cells were then transferred to 10 ml warm RPMI 1640 containing 10% FBS with or without inducer and incubated at 37°C for 5-7 h. The cells were then harvested and the pellets were lysed in 250 µl of passive lysis buffer, vortexed for 5 s and spun at 2000 g for 5 min at 4°C. Ten microliters of the above lysate were used for luciferase activity measurement for 10 s in an Auto Lumat LB953 luminometer (E.G. & G. Berthold, Aliquippa, PA). Quenching of sample luciferase activity followed by normalization of transfection efficiency was performed by the addition of 100 µl of Stop & GloTM buffer (Promega, Madison, WI) to each sample followed by luciferase measurement as above. Beetle luciferase expression of samples was divided by that of the internal control plasmid. HepG2 cells were transfected by the method of calcium phosphate co-precipitation essentially as described previously (38). Briefly, cells were plated at 2.5 × 105 cells/well in 24-well tissue culture plates the day before transfection. One hour before transfection, fresh medium was added. Calcium phosphate precipitates containing (per well) 352 ng of luciferase reporter plasmid plus 352 pg of pRL normalization plasmid were prepared. The DNA/calcium phosphate precipitates were incubated with the cells at 37°C for 5 h, at which time the cells were washed once with PBS, incubated with 15% glycerol in 1× HEPES-buffered saline (HBS) for 1 min, washed twice more with PBS, and refed with EMEM containing 10% FBS. The following day, some wells received 2 ng/ml GM-CSF for 4 h. Lysates were then prepared and luciferase measurements were taken as described above.
Mini-preparation of nuclear extract
Nuclear extract was prepared as described (37) with modifications. Briefly, 108 cells were pelleted at 500 g for 5 min at room temperature. The pellet was resuspended in 1.5 ml ice-cold PBS and transferred to an Eppendorf tube and spun for 10 s at full speed. The pellet was then resuspended in 1× packed cell volume (p.c.v.) of cold buffer containing 10 mM HEPES-KOH pH 7.9, 1.5 mM MgCl2, 10 mM KCl, 0.5 mM DTT and 0.2 mM PMSF by flicking the tube and leaving it on ice for 15 min. The reaction mixture then was passed five times through a syringe with a 23G needle and spun for 20 s at full speed. The supernatant was discarded and the pellet resuspended in 2/3 p.c.v. of ice-cold buffer containing 20 mM HEPES-KOH pH 7.9 at 4°C, 25% glycerol, 1.5 mM MgCl2, 420 mM NaCl, 0.2 mM EDTA, 0.5 mM DTT and 0.2 mM PMSF. After incubating the suspension on ice for 30 min while stirring, it was centrifuged for 5 min with 13 000 g at 4°C. The clear supernatant was then collected, quickly frozen in liquid nitrogen and stored at -80°C until use. Protein concentrations were determined using the Bradford assay (Bio-Rad Laboratories, Hercules, CA) and bovine serum albumin standards (Sigma).
Wild-type and mutant oligonucleotides used in EMSA
The following double-stranded oligonucleotides were synthesized for use in the various EMSA experiments discussed in the text and figure legends. Four 5[prime]-protruding bases were added for Klenow fill-in labeling reactions. Underlined nucleotides have been changed from the wild-type sequences. TAAA (GCATAAAGTGGCTGCCAGCC); 1STAT (CATGTTCCCAGAAGGCTGCCCTCCC); 2STAT (GAATTCGAAGAAAACAGGA-GGAGGA); 3STAT (CAGTTCCCTGAAGATTGGCATGGAC); 1STATm (CGTGCTAGCCCATGGGCCCAGAAG-CCTGCC); 2STATm (CGCGTGCTAGAAGGCGAAGAAAA-CAGGAG); 3STATm (CGTGCTAGCCCTGGCCCAGGGC-CCTGAAGATGGCATG).
Electrophoretic mobility shift assay (EMSA)
EMSA was performed exactly as described by Ma et al. (37). In brief, pre-binding of 10-18 µg of the nuclear extract to poly(dI-dC) was carried out in a 10 µl reaction volume at 30°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)·poly(dI-dC). For competition experiments, the competing oligonucleotides (200-fold molar excess) were included in this pre-incubation mixture. Radiolabeled oligonucleotide probe (3.5 fmol, ~2 × 104 c.p.m.) was then added to the reaction mixture and incubated at 30°C for 20 min. One microliter of 10× gel loading buffer, containing 250 mM Tris-HCl pH 7.5, 0.2% bromophenol blue, 0.2% xylene cyanol and 40% glycerol, was then added to the reaction for loading onto a 4-6% native gel (which was pre-run for 90 min at 150 V in 0.5× non-denaturing TBE buffer), which was run at 250 V for ~3 h. The gel was transferred onto Whatman paper, vacuum dried and exposed to film (Fuji RX, Stamford, CT) for the desired period of time at -80°C with an intensifier screen. For antibody supershift experiments, antibodies (0.1 µg in 1 µl volume) were included in the pre-incubation mixture.
RESULTS
Cloning and sequencing of the OSM promoter
A 657 bp OSM cDNA probe (from +1 to +657) (5) was used for human genomic library screening at Genome Systems Inc. (St Louis, MO). A P1 clone was obtained and the identity of this clone was confirmed by partial sequencing analysis, which demonstrated that this clone contains the entire OSM coding region as reported previously (5) in addition to a >20 kb upstream DNA fragment. An 8.5 kb DNA fragment with the 3[prime]-end 11 bases upstream of the translation initiation codon was subcloned into a luciferase reporter vector, pGL3Basic (see Materials and Methods for cloning details). The restriction map and putative transcription factor binding sites of this clone have been characterized by our studies (Fig. 1).
Figure 1. Restriction map of the 8.5 kb DNA fragment flanking the human OSM gene in the 5[prime]->3[prime] direction. (A) The solid bar represents the genomic DNA fragment isolated from a human P1 clone and the hatched boxes represent three exons of the OSM gene. (B) The size in kilobases and identity of the restriction enzyme sites mapped within this region are indicated. The hatched box represents the 5[prime]-portion of the first exon of the OSM gene. (C) The diagram represents 304 bp of 5[prime]-flanking sequence containing the proximal promoter region, transcription initiation site and portion of OSM exon I. An arrow indicates the transcription initiation site. Putative transcription factor binding sites for C/EBP (circle), STAT (square), Ap1 (diamond) and Ets (open circle) and the GC-rich region (oval) are indicated.
Determination of the 5[prime]-end of OSM mRNA
To determine the transcription initiation site of the OSM mRNA, primer extension and RPA were carried out. In the primer extension assays, the 32P-end-labeled antisense primer, OSM1A (5[prime]-AGCGTCCTCTGTGTGAGCAGTAC-3[prime]), derived from the OSM first exon (Fig. 2A), was annealed to 2 µg of mRNA isolated from GM-CSF-treated U937 cells (1 h). The OSM1A primer generated a major 81 base product (Fig. 2B), indicating that the first exon is 86 bp. To support this result, RPA was performed using a riboprobe derived from a genomic DNA clone that covers the possible transcription initiation site, exon I and part of intron I (Fig. 2A). This probe protected a major 86 base fragment (Fig. 2C), a result consistent with that seen with primer extension. Collectively, these data indicate that exon I is 86 bp, with the adenosine nucleotide of the first ATG located at +53.
Figure 2. Primer extension and RPA were used to define the transcription start site. (A) The OSM gene with its first two exons is depicted as a solid line and filled boxes. The heavily shaded areas within the filled boxes represent the OSM coding region. The primer used for primer extension, OSM1A, is drawn above the OSM gene and its position relative to exon I is indicated by dotted lines. The relative size and position of the probe used for RPA is indicated underneath the OSM gene. Figures are not drawn to scale. (B) Primer extension analysis was performed using primer (OSM1A) annealed to 1 µg of mRNA from either normal or GM-CSF-induced U937 cells. The extended product (indicated by the arrow) is shown compared with the `A' lane from a dideoxy sequencing reaction which used 32P-labeled OSM1A as a primer and a plasmid containing OSM 5[prime]-flanking genomic sequence as template. (C) RPA. A single 86 bp protected product was detected in mRNA taken from GM-CSF-induced, but not from control, U937 cells.
OM promoter elements
An 835 bp genomic fragment surrounding the hOSM initiation site was sequenced from both strands, and the sequence was compared to that of its murine counterpart (Fig. 3). High homology between these two promoters extends 750 bp 5[prime] of the transcription start site, after which the level of homology decreases. An overall 87% homology between the human and murine sequences occurs within this 835 bp region. The DNA sequence immediately upstream of the defined transcription start site does not contain a consensus TATA box, but a consensus CCAAT element was localized at -48 to -44. A GenBank submotif search of this 835 bp region with the MatInspector program was performed with core similarity set at 0.8 and matrix similarity at 0.85, which revealed numerous putative transcription factor binding motifs. These include three readily recognizable STAT elements [named proximal STAT (P-STAT), middle STAT (M-STAT) and distal STAT (D-STAT)] and two AP1 sites. Most interestingly, a reversed Ets motif with the core sequence GGAA overlapped with the D-STAT element. All the above-mentioned elements are conserved in both human and mouse except the D-STAT element, which is unique to humans. No obvious repeat sequences were identified, though GC content is >65%. A condensed GC-rich region (25 out of 33 bp) was identified between -94 and -61 (Fig. 3).
Figure 3. DNA sequence comparison of human and murine OSM promoters. The first nucleotide of the OSM mRNA is double underlined and designated +1. The OSM ATG translation initiation codon is also underlined. The numbers on the right side of each line indicate the position of the last nucleotide of that line relative to the +1 transcription initiation nucleotide. Dashed lines within the sequence have been inserted to provide maximal sequence homology. Identity between human and murine DNA sequence is indicated by an asterisk. Based on sequence analysis, possible transcription factor binding sites and GC-rich regions are shaded and labeled.
Sequences required for basal activity of the OSM promoter
To identify functional cis-acting elements required for basal OSM gene expression, a series of 5[prime] deletion mutants of the OSM promoter were constructed upstream of the firefly luciferase gene in the pGL3 Basic vector. These constructs were tested for luciferase activity following transient transfection into the human promonocytic cell line U937 and HepG2 cells. Luciferase analysis demonstrated that the minimum promoter sequence required for full activity in U937 cells resides within 194 bp (Fig. 4). Further deletion to -109 resulted in a 35% decrease in activity, and deletion to -94 resulted in a 55% decrease. At -31 the reporter activity decreased to a level equal to the promoterless control (Fig. 4). In HepG2 cells, promoter activities of the -835, -304 and -194 constructs were not significantly different from that of the promoterless control plasmid. However, promoter activities of the remaining -109, -94, -61 and -31 constructs in HepG2 cells closeley mirrored those of the same constructs transfected into U937 cells. We then performed mutational analysis of putative cis-acting elements by nucleotide substitution. Mutation of the putative Ets (TTCC->GGCC), P-STAT (TTCCCAGAA->GGCCCAGAA), AP1 (TTGGGGTCATG->TTGGGCATATG), GC-rich (CCCT-CCCCCCTCCCCATCCCCACCC->CCCTACCCCCTACCC-ATCCCTACCC) and C/EBP (CCAAT->GATAT) motifs within the -304 bp region demonstrated that the putative C/EBP element, and to a lesser extent the GC-rich region, are important elements responsible for full basal promoter activity (Fig. 5).
Figure 4. A series of OSM promoter deletion mutant constructs were transfected into U937 (black bars) and HepG2 (light bars) cells to determine the minimum basal promoter. The left portion shows the plasmid constructs with size indicated in base pairs, along with the relative location of putative transcription factor binding sites. The right portion shows promoter activity measured in relative light units (RLU) obtained from each transfection. Promoter activity is normalized for transfection efficiency by dividing OSM promoter-driven luciferase expression by that of co-transfected SV40-driven renilla luciferase expression as measured using its own substrate. The results represent the mean of three experiments.
Figure 5. Two or three bases within putative transcription factor binding motifs or the GC-rich region were mutated (indicated by X) on the -304 bp OSM promoter-luciferase construct. These constructs were then transfected into U937 (black bars) and HepG2 (light bars) cells to determine the role of each element in OSM basal transcription. Normalization for transfection efficiency was done as stated for Figure 4.
Transcriptional activation of the OSM promoter in response to GM-CSF requires STAT sequences
Northern blot analysis showed that PMA and GM-CSF induced OSM mRNA expression identically in U937 cells (data not shown). A subsequent RPA was performed demonstrating that GM-CSF-induced OSM expression peaked at 1 h (Fig. 6). To precisely locate the control elements that regulate GM-CSF-induced OSM expression in U937 cells, transient transfection assays were performed in the presence or absence of GM-CSF. As shown in Figure 7, GM-CSF treatment resulted in a 1.2-fold increase in mutant M-STAT reporter gene expression, whereas luciferase activity of the same construct containing a wild-type M-STAT element was induced 3-fold by GM-CSF in U937 cells. Similar experiments were performed with mutant D-STAT and P-STAT constructs, but there was no significant effect on promoter activity after GM-CSF induction. When transfected into HepG2 cells, the identical constructs were uninducible by identical GM-CSF treatment. These results suggest that in U937 cells, M-STAT is the proximal OSM promoter main GM-CSF-responsive STAT element.
Figure 6. RPA detected OSM message from U937 cells. RPA was carried out with 2 µg poly(A)+ mRNA isolated from untreated U937 cells and cells which had been treated with GM-CSF for time periods ranging from 0.5 to 24 h. mRNA was hybridized to a 32P-labeled antisense riboprobe derived from a fragment of the OSM-containing genomic clone (see Fig. 2). The sizes of the molecular marker and the protected product are indicated.
Figure 7. Transient transfection assays were carried out using either a wild-type or mutant STAT reporter plasmid construct in U937 (black bars) and HepG2 (light bars) cells separately. Each plasmid construct consisted of six identical transfections. Three were treated with 2 ng/ml GM-CSF for 7 h post-transfection, and the remaining three were left untreated (empty bars). Transfection efficiency determination was performed as described above. Luciferase activity in untreated cells was arbitrarily assigned a value of 1 and luciferase activity from treated cells is expressed as a fold increase over control.
STAT5b binds to the M-STAT element and drives activated transcription
Nuclear protein extracts from control or GM-CSF-treated U937 cells were used along with the M-STAT consensus oligonucleotide probe (5[prime]-GAATTCGAAGAAAACAGGAGGAGGA-3[prime]) in EMSA (Fig. 8). Specific and strong protein-DNA complexes were formed in the reaction containing GM-CSF-treated but not in control U937 nuclear extracts. This complex could be competed by a 200-fold molar excess of unlabeled M-STAT probe. Antibody raised against STAT5b could supershift the complexes of GM-CSF-treated nuclear extracts, clearly showing that the protein involved is STAT5.
Figure 8. EMSA combined with a competition and supershift experiment demonstrates that GM-CSF-induced STAT5b binds to the OSM M-STAT probe. Nuclear extracts were prepared from U937 cells which were either treated (+) or not treated with 2 ng/ml GM-CSF for 1 h. These nuclear extracts were then used in the EMSA in the presence (+) or absence of 200-fold molar excess of cold probe or different types of anti-STAT polyclonal antibodies. The arrow indicates the location of the STAT5b:DNA complex, while the arrowhead indicates the location of the anti-STAT5b supershifted complex.
DISCUSSION
We isolated an 8.0 kb DNA fragment upstream of the OSM gene and sequenced 1 kb of the 3[prime]-portion. This is the first time that the human OSM proximal promoter sequence has been systematically scrutinized, although a similar study in mouse was undertaken previously (32). Using primer extension and RPA, we defined the transcriptional unit of the human OSM gene and demonstrated that the first exon is 86 bp in length and contains the first ATG codon. Yoshimura et al. (32) described two transcription start sites at nt 52 (major) and 81 (minor) upstream with respect to the ATG on the murine OSM promoter. Our study, however, detected only a single major start site corresponding to 52 bp upstream with respect to the ATG on the human OSM promoter.
Functional characterization of the 5[prime] upstream region demonstrated that the minimum sequence necessary for full activity resides within 304 bp upstream of the transcription initiation site. Within this 304 bp, sequence analysis identified many putative cis-acting transcriptional regulatory element binding sites including STAT, Ets, Ap1 and C/EBP, and a GC-rich region. Our preliminary gel shift and supershift assays have indicated that the STAT5b and C/EBP[beta] proteins specifically bind to the M-STAT and C/EBP elements, respectively, though binding of other putative transcription factors needs to be confirmed. We also determined by sequence analysis that the OSM promoter possesses a transcriptional initiator (Inr)-like sequence of -3 CCCAGCCG +5 (consensus YYCAYYYY) (39) located around the initiation nucleotide A. In conjunction with the GC-rich region and C/EBP, this element may be responsible for accurate initiation. By itself, however, this element does not appear to support transcription initiation, as we demonstrated by using the pGL3-31 construct in a transient transfection assay. pGL3-31 contains only the Inr-like sequence and a putative TATA-like sequence, but failed to elicit any promoter activity (Fig. 4) as compared with the promoterless control plasmid. Although no consensus TATA box was found, a rather high GC content region occurs around -75 bp. These facts taken together indicate that this promoter might belong to the TATA-less promoter family, members of which possess a typical GC-rich region. This is consistent with the knowledge that housekeeping proteins and most growth factors and receptors are encoded and regulated by TATA-less promoter genes (40 and references therein). In this context, the GC-rich region within the proximal region of the OSM promoter, when located in front of a TATA-less promoter, functions as the anchor for the basal transcription machinery (41). In support of this hypothesis, we demonstrated that factor(s) binding to this region is absolutely necessary for basal promoter activity. Conversely, mutations within this GC-rich region obliterated basal reporter gene expression (Fig. 5). In addition, the proximal promoter element CCAAT plays an equally important role in determining the basal transcriptional activity of the OSM gene, because mutations made within this element decreased the promoter activity to a level comparable to that of the promoterless control plasmids. Therefore, disruption of either element completely abolished OSM promoter activity. This requirement for intact C/EBP-binding and GC-rich regions seems to indicate that a cooperative interaction between C/EBP and a putative GC-rich-binding factor is necessary for baseline transcription to occur. In contrast, mutation of either the putative Ap1 or Ets element has no impact on OSM basal promoter activity.
A previous study found that murine OSM expression is restricted to bone marrow and spleen, and that normally it is not expressed in liver, lung, ovary, small intestine, kidney and brain (32). In vitro studies implicate mOSM as an immediate early gene expressed in murine myeloid and lymphoid cell lines in response to a subset of cytokines such as IL2, IL-3, EPO and GM-CSF (32,36). Alternatively, in human U937 promonocytes (3) and T lymphocytes (2), OSM expression could be induced, respectively, by PMA and T cell-specific mAb 9.3 or phytohemagglutinin. To understand the mechanisms that direct regulated OSM expression, we examined whether these inducers share a common regulatory pathway and/or a common cis-acting element on the OSM promoter. Preliminary RPA experiments conducted by our laboratory showed that both PMA and GM-CSF are capable of inducing OSM expression to a similar extent and with a similar time course (data not shown) in U937 cells. Unlike PMA, which activates numerous signal transduction pathways, GM-CSF binds to a single specific receptor on the cell surface and preferentially activates the JAK/STAT signal transduction pathway, culminating in transcription of target genes mediated by GAS or STAT elements. In fact, careful sequence analysis and GenBank database searching against 835 bp of the OSM proximal promoter region revealed many transcription factor binding motifs, among them three STAT elements.
To provide continuity with our studies involving basal OSM promoter activity, we chose to continue using U937 cells as a model system for dissecting the mechanisms responsible for GM-CSF-induced OSM expression. Figure 6 clearly indicates the kinetics of GM-CSF-induced OSM expression, which is evident by 0.5 h, reaches a maximum at 1 h, and remains at low levels for as long as 24 h following treatment. It had been suggested that OSM expression can also be induced by IFN[gamma] and TNF[alpha] (42). However, we failed to see detectable OSM expression by RT-PCR analysis after either IFN[gamma] or TNF[alpha] treatment of U937 cells. OSM has been shown to act as an autocrine growth activator in Kaposi's sarcoma spindle cells (43). To test whether OSM has an autocrine effect in our model system, we treated U937 cells with human recombinant OSM at a concentration of 50 ng/ml with sampling at time points from 1 to 16 h, and were unable to identify OSM expression in an RT-PCR assay. This is consistent with the observation that U937 cells do not have large numbers of type II high affinity OSM receptors (44).
In our further study of GM-CSF-induced OSM expression, we discovered that the human OSM promoter possesses a third STAT element located upstream of the two conserved downstream STAT elements that are present in the murine OSM promoter. Mutational inactivation of this STAT element, however, had no effect on inducibility of the OSM promoter construct by GM-CSF. The fact that this distal STAT element is not evolutionarily conserved between murine and human OSM also provides further evidence that this element does not seem to be involved in GM-CSF induction. The most proximal conserved STAT element on the murine OSM promoter, which lies within 100 bp of the transcription start site, is postulated to be the cis-acting sequence responsible for controlling mOSM expression by GM-CSF and other STAT activating factors (32). In contrast, mutational inactivation of the proximal conserved STAT element on the human OSM promoter had no effect on GM-CSF inducibility. Rather, inactivation of the M-STAT element significantly, but not completely, reduces GM-CSF inducibility. These assertions have been confirmed by both transient transfection analysis using wild-type and mutant M-STAT plasmid constructs (Fig. 7) and by EMSA for the identification of specific STAT5b DNA-binding proteins (Fig. 8). It is possible that cooperative interaction of STAT with certain members of the basal transcriptional machinery may be a requisite for full activation of OSM expression. In support of this hypothesis, recent studies have demonstrated a requirement for cooperative binding of STAT1/3 and Sp1, etc., for full STAT inducibility (45,46).
To investigate the tissue specificity of the OSM promoter, we chose HepG2 as a non-hematopoietic cell system, which we know does not express endogenous OSM. As expected, GM-CSF had no influence on the OSM proximal promoter, indicating that the STAT motifs do not respond to GM-CSF induction in these cells (Fig. 7). Since it is not known whether HepG2 cells express GM-CSF receptor, we cannot exclude the possibility that lack of the GM-CSF receptor may be responsible for the uninducibility of these cells by GM-CSF. In addition, analysis of basal luciferase activity showed that there may be a transcriptional silencer(s) between nt -194 and -109 that is utilized in HepG2 cells and not in U937 (Fig. 4). This strongly supports our current conclusion that the OSM promoter is under cell-specific control. Due to the presence of this putative silencer between -194 and -109, basal promoter activity drops to a very low level in HepG2 cells. Mutations made at the GC-rich region and the C/EBP motif reduce promoter activity in a manner similar to that observed in U937 cells, thus providing additional support for our finding that the GC-rich region and C/EBP element both contribute to basal promoter activity.
The OSM gene is clustered together with LIF on chromosome 22 and LIF has similar biological functions, but with very little similarity in the 5[prime]-promoter and 3[prime]-non-coding intron. Therefore, it is possible that these two genes are regulated by a common promoter locus that is far upstream of both genes. As emphasized by Agarwal and Rao (47), gene regulation really is an integrated function. The results from this study, however, will facilitate increased understanding of OSM expression regulation in vivo. Of interest to our laboratory is expression of OSM in the context of the tumor microenvironment and its implication in either tumor progression or the metastatic process. Although the biological function of OSM is still a matter of debate, elucidation of the mechanism governing its expression will allow purposeful control of its expression or perhaps give clinicians an important prognostic marker for tumor development.
ACKNOWLEDGEMENTS
This work was supported by the Department of Veterans Affairs (Office of Research and Development, Medical Research Service) and by the Mountain States Medical Research Institute, Boise, ID.
REFERENCES
*To whom correspondence should be addressed at present address: Research Service (151), VA Medical Center, 500 West Fort Street, Boise, ID 83702-4598, USA. Tel: +1 208 422 1156; Fax: +1 208 422 1155; Email: rvestal{at}micron.net Present address: Jingwen Liu, Palo Alto Veterans Affairs Health Care System, 3801 Miranda Avenue, Palo Alto, CA 94304, USA
This page is run by Oxford University Press, Great Clarendon Street, Oxford OX2 6DP, as part of the OUP Journals
Comments and feedback: jnl.info{at}oup.co.uk
Last modification:
Copyright© Oxford University Press, 1999.
CiteULike 
Connotea 
Del.icio.us What's this?
This article has been cited by other articles:
|
|
 |
|
  S. P. Kastl, W. S. Speidl, C. Kaun, K. M. Katsaros, G. Rega, T. Afonyushkin, V. N. Bochkov, P. Valent, A. Assadian, G. W. Hagmueller, et al. In Human Macrophages the Complement Component C5a Induces the Expression of Oncostatin M via AP-1 Activation Arterioscler. Thromb. Vasc. Biol., March 1, 2008; 28(3): 498 - 503. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 E. A. Nelson, S. R. Walker, W. Li, X. S. Liu, and D. A. Frank Identification of Human STAT5-dependent Gene Regulatory Elements Based on Interspecies Homology J. Biol. Chem., September 8, 2006; 281(36): 26216 - 26224. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 M. Buitenhuis, B. Baltus, J.-W. J. Lammers, P. J. Coffer, and L. Koenderman Signal transducer and activator of transcription 5a (STAT5a) is required for eosinophil differentiation of human cord blood-derived CD34+ cells Blood, January 1, 2003; 101(1): 134 - 142. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 P. Repovic and E. N. Benveniste Prostaglandin E2 Is a Novel Inducer of Oncostatin-M Expression in Macrophages and Microglia J. Neurosci., July 1, 2002; 22(13): 5334 - 5343. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 S. L. Wyszomierski and J. M. Rosen Cooperative Effects of STAT5 (Signal Transducer and Activator of Transcription 5) and C/EBP {beta} (CCAAT/Enhancer-Binding Protein-{beta}) on {beta}-Casein Gene Transcription Are Mediated by the Glucocorticoid Receptor Mol. Endocrinol., February 1, 2001; 15(2): 228 - 240. [Abstract] [Full Text]
|
|
|
|
|
|
G. B. Ehret, P. Reichenbach, U. Schindler, C. M. Horvath, S. Fritz, M. Nabholz, and P. Bucher DNA Binding Specificity of Different STAT Proteins. COMPARISON OF IN VITRO SPECIFICITY WITH NATURAL TARGET SITES J. Biol. Chem., February 23, 2001; 276(9): 6675 - 6688. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||