Published online before print October 28, 2004
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
-responsive region in the NCF2 promoter
Department of Veterinary Molecular Biology, Montana State University, Bozeman
1 Correspondence: Veterinary Molecular Biology, Montana State University, Bozeman, MT 59717. E-mail: mquinn{at}montana.edu
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
(TNF-), which activates AP-1, we hypothesized that TNF- might regulate NCF2transcription via AP-1. In support of this hypothesis, we show here that NCF2 promoter-reporter constructs are up-regulated by TNF- but only when AP-1 factors were coexpressed. Consistent with this observation, we also demonstrate that NCF2 mRNA and p67phox protein are up-regulated by TNF- in various myeloid cell lines as well as in human monocytes. It was surprising that mutagenesis of the AP-1 site in NCF2 promoter constructs did not eliminate TNF- induction, suggesting additional elements were involved in this response and that AP-1 might play a more indirect role. Indeed, we used NCF2 promoter-deletion constructs to map a novel TNF--responsive region (TRR) located between –56 and –16 bp upstream of the translational start site and demonstrated its importance in vivo using transcription factor decoy analysis. Furthermore, DNase footprinting verified specific binding of factor(s) to the TRR with AP-1 binding indirectly to this region. Thus, we have identified a novel NCF2 promoter/enhancer domain, which is essential for TNF--induced up-regulation of p67phox.
Key Words: neutrophil • promoter • transcriptional regulation • NADPH oxidase • monocyte
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Although flavocytochrome b contains the entire electron transport apparatus of the NADPH oxidase, it does not function independently in the cell and requires association of cytosolic oxidase proteins for normal enzymatic activity (reviewed in ref. [8 ]). Of the cytosolic components, p67phox is considered to play one of the most critical roles in NADPH oxidase activation and is the limiting cytosolic oxidase cofactor [9 , 10 ]. Indeed, NADPH oxidase activity can be reconstituted in vitro in the absence of p47phox by combining flavocytochrome b with high concentrations of p67phox and Rac [11 , 12 ] or chimeric p67phox-Rac fusion proteins [13 , 14 ], supporting the concept that p67phox plays a more distinct role in electron transport via direct interaction with flavocytochrome b [15 , 16 ] and Rac1/2 [17 ]. Although the specific role of p67phox in electron transfer is still not fully understood, it has been proposed that p67phox may regulate the NADPH oxidase by facilitating electron flow from NADPH to the flavin center of flavocytochrome b [18 ] and possibly by contributing to the binding of NADPH [19 ]. In support of this idea, it has been shown that p67phox binding can induce conformational changes in flavocytochrome b, resulting in electron flow and O2•– production [20 21 22 ]. Additionally, recent studies suggest that a temporal sequence of interactions involving flavocytochrome b, Rac1/2, and p67phox results in p67phox conformational changes that ultimately trigger electron flow within flavocytochrome b (reviewed in ref. [23 ]).
Based on the importance of p67phox in NADPH oxidase function, we hypothesized that transcriptional regulation of neutrophil cytosolic factor 2 (NCF2), the gene coding for p67phox, might play an important role in oxidase regulation [24 ]. In monocytes, cytokine treatment has been shown to regulate transcription of oxidase components [25 ]. Although phagocytosing neutrophils have been shown to undergo dramatic changes in gene expression [26 ], less is known about transcriptional regulation of oxidase genes in these cells. Nevertheless, recent studies in myeloid cell lines indicate NCF2 is regulated somewhat differently than the other oxidase proteins during cellular differentiation. For example, p67phox is generally expressed later than the other components, and its expression appears to correlate most closely with oxidase activity, supporting its role as a key factor limiting the respiratory burst during differentiation [27 28 29 ]. If p67phox is indeed the key limiting factor in the formation of an active NADPH oxidase, it could be argued that slight changes in gene expression might significantly affect the amount of O2•– generated or the duration of the burst.
Previous analysis of the NCF2 promoter region demonstrated the presence of a binding site for the activator protein 1 (AP-1) transcription factor, which was required for basal transcription of NCF2 promoter-reporter constructs [24
, 30
]. Using chromatin immunoprecipitation assays, Li et al. [30
] demonstrated binding of AP-1 to the AP-1 site of the endogenous p67phox promoter, verifying a key role for AP-1 in NCF2 regulation. Based on previous studies showing that endogenous p67phox is up-regulated by tumor necrosis factor (TNF-) [31
, 32
] and that TNF- activates AP-1 via Jun N-terminal kinase (JNK) [33
], we hypothesized that TNF--mediated induction of NCF2 involved the AP-1 binding site described above. Thus, we analyzed the role of this promoter element in TNF- regulation of NCF2. We show here that although AP-1 is necessary for TNF--induced up-regulation of NCF2 promoter-reporter constructs, the previously characterized AP-1 binding site was not required. Instead, we have identified a novel TNF--responsive region (TRR) within the promoter/enhancer of NCF2, which does not contain an AP-1 consensus binding site but is essential for TNF- responsiveness. Furthermore, we demonstrated an indirect interaction of AP-1 with the TRR.
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
was from Research Diagnostics, Inc. (Flanders, NJ). Phorbol myristate acetate (PMA) and luminol were from Sigma Chemical Co. (St. Louis, MO).
Cells and cell culture
HL-60 cells were grown in Dulbecco’s modified Eagles’s medium (DMEM) supplemented with 20% fetal bovine serum (FBS). K562, U937, MonoMac1, and MonoMac6 cells were grown in RPMI 1640 supplemented with 10% FBS and penicillin/streptomycin. For TNF- treatments, cell lines were resuspended in fresh media at 5 x 105 cells/mL and treated with 20 ng/mL TNF- for 20–24 h. K562 and MonoMac6 cells were maintained at 5 x 105 cells/mL and seeded at 2.5 x 105 cells/mL in fresh media 1 day prior to transfection.
Neutrophils and monocytes were purified from human blood. Neutrophils were purified using dextran sedimentation followed by Histopaque 1077 gradient separation and hypotonic lysis of red blood cells, as described previously [34 ]. Cell preparations were routinely >95% pure, as determined by light microscopy, and >98% viable, as determined by trypan blue exclusion. Monocytes were isolated by density gradient centrifugation using OptiPrep medium (Greiner BioOne, Longwood, FL) [35 ]. Monocyte preparations contained <10% lymphocytes and no granulocytes, as determined by flow cytometric scatter profiles and CD14 immunostaining (Becton Dickinson, Franklin Lakes, NJ).
Purified neutrophils and monocytes were suspended at 105 and 106 cells/mL, respectively, in RPMI 1640 supplemented with 10% FBS and penicillin/streptomycin and treated with TNF- (20 ng/mL) for 6–48 h (monocytes) and 0.5–24 h (neutrophils).
Quantitative real-time-polymerase chain reaction (RT-PCR)
Total RNA was isolated from cells using QIAshredder and RNeasy mini kits (Qiagen, Valencia, CA). RNA was DNase-treated on the RNeasy minicolumns using an RNase-free DNase set (Qiagen). First-strand synthesis was achieved by incubating 0.5–1.0 µg total RNA with 1 µL 50 ng/µL random primer (Invitrogen, Carlsbad, CA) and H2O (to 12 µL) for 5 min at 72°C, followed by 2 min on ice. Superscript II (Invitrogen) was added (1 µL), followed by 8 min at room temperature and 50 min at 42°C. cDNA (1 µL) was amplified, per the manufacturer’s suggestions, using SYBR Green PCR core reagents (Applied Biosystems, Foster City, CA). Reactions were run for one cycle at 50°C for 2 min, one cycle at 95°C for 10 min, and 40 cycles of 95°C for 15 s plus 60°C for 1 min using an ABI GeneAmp 5700 sequence detection system. NCF2 and TNF receptor-associated factor 1 (TRAF-1) mRNA levels were quantified relative to ribosomal 18S mRNA by measuring SYBR green incorporation during quantitative RT-PCR using the relative standard curve method. Calculations and statistical analyses were performed as described in the manufacturer’s protocol and in User Bulletin #2 for the ABI PRISM 5700 sequence detection system. Relative changes in NCF2 or TRAF-1 mRNA levels between treated and nontreated cells are expressed as fold increase compared with the basal level of expression in untreated cells (1.0).
Primer sequences were as follows: NCF2-specific primers, (forward) 5'-ATTACCTAGGCAAGGCGACG-3' and (reverse) 5'-TCTGGGTGGAGGCTCAGCT-3'; TRAF-1-specific primers, (forward) 5'-GCCTTTCCGGAACAAGGTC-3' and (reverse) 5'-CGTCAATGGCGTGCTCAC-3'; and 18S-specific primers, (forward) 5'-TCGAGGCCCTGTAATTGGAA-3' and (reverse) 5'-CCCAAGATCCAACTACGAGCTT-3'.
Luciferase reporter-promoter constructs
PCR was used to generate the NCF2 promoter-reporter deletion mutants using pGL3-p67(–2017) [i.e., 2017 bp 5' of the translational start site (TLS)] as template [24
]. For 5' deletion mutants, all forward primers contained a KpnI restriction site and the NCF2 promoter sequence necessary to generate the desired deletion. The reverse primer spanned the TLS of NCF2 and contained an XhoI restriction site. For 3' deletion mutants, the forward primer contained a KpnI site and NCF2 promoter sequence beginning at –500 bp upstream from the TLS. All reverse primers contained an XhoI site and NCF2 promoter sequence necessary to generate the desired deletion. The amplified DNA was digested with KpnI and XhoI and ligated into KpnI/XhoI-digested pGL3-basic.
The pGL3-p67(500)–(58–36) construct was generated by first removing a KpnI/PstI insert (–500 to –37 bp) from pGL3-p67(500). A PCR-amplified and KpnI/PstI-digested insert (–500 to – 59 bp) was ligated into this vector. The pGL3-3X(58–36) was generated by annealing a 5' phosphorylated oligonucleotide containing three copies of the p67phox promoter sequence (–58 to –36 bp) to its complementary strand. This double-stranded DNA was gel-purified and ligated into a SmaI-digested pGL3 vector. The AP-1 site-specific mutant [pGL3-p67(500)–AP1] was generated using the QuikChange site-directed mutagenesis kit (Stratagene, Cedar Creek, TX) and pGL3-p67(500) as template. All constructs were confirmed by cycle sequencing.
Cycle sequencing was performed on an MJ Research PTC-200 thermocycler using Big Dye terminators, and the samples were run on an ABI 310 genetic analyzer.
Transient transfections
The day prior to transfection, K562 cells were resuspended at 2.5 x 105 cells/mL in fresh DMEM. For transfections, 107 cells were resuspended in 800 µL RPMI containing 25 µg luciferase reporter construct and 5 µg pRL-TK vector to control for transfection efficiency. Cells were electroporated (BioRad Gene Pulser II) at 950 µF and 210 V in 0.4 cm cuvettes and resuspended in 8 mL complete media. TNF- (20 ng/mL) was added to media 4 h post-transfection. At 24 h post-transfection, cells were assayed for firefly and Renilla luciferase activity using the dual-luciferase reporter assay system (Promega, Madison, WI). Briefly, cells were washed in phosphate-buffered saline and resuspended in 50–100 µL lysis buffer (Promega), and 5–30 µL cell lysate was assayed for luciferase activity using a Lumat LB 9507 luminometer (EG&G Berthold, Germany). Subsequently, 50–100 µL Stop and Glo reagent (Promega) was added to the sample, which was then analyzed for Renilla luciferase activity.
DNase footprinting
Fluorescence-based DNase footprinting was performed as described [36
]. Briefly, a 327-bp NCF2 probe containing the TRR was fluorescently labeled on one strand using PCR and a 6-carboxyfluorescein (6-FAM)-labeled 5' primer (Applied Biosystems). The sequence of the 6-FAM-labeled forward PCR primer was 5'-AGCAGCTGATGTACTTCCTCTCTCCTC-3', and the nonlabeled reverse primer was 5'- GGGCTCGAGCGGTCCACCAGGGACATGATTAGGTAG-3'. PCR was performed using pGL3-p67(500) and pGL3-p67(500)–AP1 template DNA and Pfu DNA polymerase, and the labeled probes were gel-purified with QIAquick gel extraction kits (Qiagen) and used for DNase footprinting reactions, which were carried out as suggested using the core footprinting system (Promega). Briefly, 400 ng probe was incubated in binding buffer ± nuclear extract (3 µg) on ice for 10 min, followed by 45 s incubation with 0.25 U RQ1 RNase-free DNase. For competition reactions, 20x double-stranded competitor DNA was incubated with extract for 15 min at room temperature prior to addition of probe. The sequences of the sense strand for the TRR, AP-1, and random competitor DNAs were as follows: TRR: 5'-CCAACCTGTCTTCTCCCTGTCTCCTGCAGCTCTCTTGGCCTCCT-3';AP-1: 5'-CATGGTAGGGTTATGAGTCAGTTGCCAAAAG-3'; and random: 5'-AGTCAGTCAGTCAGTCAGTCAGTCAGTCAGTCAGTCAGTC-3'. Reactions were extracted with phenol:chloroform, precipitated, washed with 70% ethanol, and loaded on 6% polyacrylamide gels using ROX-350 size standards for reference (Applied Biosystems). Gels were run on an ABI 377 genetic analyzer and analyzed with Genescan software (Applied Biosystems).
Transcription factor decoy
MonoMac6 cells were seeded at 2.5 x 105 cells/mL the day prior to transfection and transfected at a density of 3–4 x 105 cells/mL. For transfections, 106 cells were centrifuged at 60 g for 10 min, resuspended in 100 µL nucleofector solution, and electroporated using a Nucleofector (Amaxa Biosystems, Germany) minus DNA or plus 1 µM TRR or random phosphorothioate double-stranded oligonucleotide (sequences shown above in DNase footprinting). Cells were transferred to a six-well tissue-culture plate with 500 µL complete media. At 4 h post-transfection, 1.4 mL complete media containing 20 ng/mL TNF- was added, and the cells were incubated for 20–24 h. RNA was isolated using QIAshredder columns and RNeasy mini kits (Qiagen). RNA was DNase-treated on the RNeasy mini columns using a RNase-free DNase set (Qiagen). After quantification, RNA was used for quantitative RT-PCR analysis.
Measurement of monocyte O2•– production
Human blood monocytes were cultured ±20 ng/mL TNF-. At the indicated times, cells were collected and washed with RPMI 1640 containing no phenol red or FBS and resuspended at 1–3 x 106/mL in the same buffer. For chemiluminescence assays, monocytes were aliquotted into white polystyrene microtiter plates (Corning, Corning, NY) at 1–3 x 105cells/well in RPMI 1640 containing 25 µM luminol. Cells were activated by the addition of 80 nM PMA, and the reactions were monitored for 60 min on a Fluoroskan Ascent FL microtiter plate reader (ThermoElectron Corp., Milford, MA) with a 1-min reading interval. Integrated chemiluminescence was determined over the full 60-min kinetic assay using Ascent software. All time-points and treatments were performed in triplicate.
Immunoblot analysis
Cell lysates were prepared by sonication in the presence of protease inhibitors (Sigma Chemical Co.), and protein content of the samples was determined using the bicinchoninic acid method (Pierce Chemical Co., St. Louis, MO). Samples (4 µg/lane) were separated by sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) on polyacrylamide gradient gels and transferred to nitrocellulose membrane as described [37
]. Prestained molecular weight standards (Amersham Biosciences, Piscataway, NJ) were included on all gels for reference. Previously characterized human p67phox monoclonal antibody 81.1 [38
] was used to probe Western blots, followed by an alkaline phosphatase-conjugated goat anti-mouse immunoglobulin G secondary antibody (BioRad, Richmond, CA). Developed blots were analyzed using an IS-1000 Alpha Imager digital imaging system (Alpha Innotech, San Leandro, CA).
Statistical analysis
One-way ANOVA was performed on the indicated sets of data, followed by Tukey’s pair-wise comparisons (GraphPad Prism Software, San Diego, CA). Differences at P < 0.05 were considered to be statistically significant.
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
-treated myeloid cells up-regulated NCF2 mRNA in HL-60 cells [32
] and p67phox protein in mouse bone marrow-derived macrophages [31
]. To evaluate this response in a wider range of cells commonly used to study NADPH oxidase function, quantitative RT-PCR and immunoblot analysis were used to compare mRNA and protein levels in a variety of untreated and TNF--treated myeloid cells and cell lines. Cell lines and primary cells chosen for these studies included: promyelocytic HL-60, promonocytic U937, monocytic MonoMac1 (CD14–) and MonoMac6 (CD14+) cell lines, as well as freshly isolated human blood neutrophils and monocytes.
In all the myeloid cell lines analyzed, transcription of endogenous NCF2 promoter was clearly induced by TNF- (Fig. 1A
). We observed a two- to threefold increase in NCF2 mRNA in all of the cell lines except for MonoMac1 cells, which reproducibly showed a five- to tenfold increase in NCF2 message. Similarly, TNF- induced a twofold increase in NCF2 mRNA in human monocytes. To further characterize the monocyte response, we analyzed NCF2 mRNA levels in monocytes treated for various times (0, 6, 12, 24, and 48 h) with or without TNF-. A small increase in message was observed in TNF--treated monocytes at 12 h; however, this difference was not statistically significant. In contrast, statistically significant (P<0.001) increases in NCF2 mRNA levels were observed in cells treated with TNF- for 24 h (average 3.1-fold) and 48 h (average 2.6-fold; Fig. 2
). Although increases in NCF2 mRNA levels were consistently observed at these times, it should also be noted that the relative increases were quite variable from donor to donor [e.g., the relative increase in message varied from 1.4- to eightfold at 24 h (N=7) and from 1.8- to 5.1-fold at 48 h (N=6)].
 View larger version (25K): [in a new window] |
Figure 1. TNF- up-regulates NCF2 message and p67phox protein levels in myeloid cells. (A) Quantitative RT-PCR was used to determine relative levels of NCF2 mRNA in total RNA isolated from cells cultured for 24 h without (untreated) or with 20 ng/mL TNF-. The results are presented as fold induction compared with basal levels of NCF2 mRNA and are representative of at least three experiments (inset indicates average induction). (B) Immunoblot analysis was used to evaluate p67phox expression in untreated and TNF--treated cells. Lysates were prepared from cells cultured for 24 h without or with 20 ng/mL TNF-, separated by SDS-PAGE (4 µg/lane), and blotted with monoclonal anti-p67phox. Representative of three experiments.
|
 View larger version (14K): [in a new window] |
Figure 2. Kinetics of TNF--induced up-regulation of NCF2 message. Quantitative RT-PCR was used to determine relative levels of NCF2 mRNA in total RNA isolated from human monocytes cultured for 0, 6, 12, 24, and 48 h without (Untreated) or with 20 ng/mL TNF- (+ TNF). The results are presented as relative levels of NCF2 mRNA (mean±SEM) from three (6 and 12 h) or six experiments. *, Statistically significant difference (P<0.001) between untreated and TNF--treated cells at each time-point.
|
-treated human neutrophils (Fig. 1A)
. To evaluate whether changes in NCF2 message might have occurred at earlier times, we also analyzed neutrophils treated with TNF- over a wide range of times (30 min to 24 h) but still found no detectable changes in NCF2 message (data not shown).
To evaluate if the observed changes in mRNA corresponded to changes in protein levels, we analyzed p67phox expression in untreated and TNF--treated cells by immunoblotting. As shown in Figure 1B
, we found that increases in NCF2 message levels correlated with increases in p67phox protein expression (Fig. 1B) , and densitometric analysis of the blots confirmed this correlation (not shown). Although there were no obvious differences in NCF2 mRNA levels between untreated and TNF--treated neutrophils to suggest a change in corresponding protein, we were concerned that subtle changes in p67phox protein might not be evident because of the high basal levels of p67phox already present in neutrophil samples. To address this issue, neutrophil samples were serially diluted and reanalyzed by immunoblotting. Again, no changes in p67phox levels were observed between untreated and treated cells (data not shown), as is consistent with the lack of change observed for NCF2 mRNA in these cells.
Evaluation of the role of the AP-1-binding site in TNF--induced up-regulation of NCF2 promoter-reporter constructs
In previous studies, we identified an AP-1 binding site in intron 1 of NCF2 and demonstrated that it was essential for basal promoter activity [24
]. As TNF- activates AP-1 via JNK, we hypothesized that TNF--dependent up-regulation of NCF2 might be a result of AP-1 activation via TNF- stimulation of JNK. To test this hypothesis, an NCF2 promoter-reporter construct, pGL3-p67(500), containing –500 to –1 bp upstream of the NCF2 ATG start site was transiently transfected into K562 cells to evaluate effects of TNF-. K562 cells were used for these experiments, as they express no detectable levels of endogenous c-Fos and c-Jun, as determined by immunoblot analysis, and there was relatively low basal activity of the NCF2 promoter-reporter constructs in this cell line (unpublished data).
As shown in Figure 3
, transcription of the NCF2 promoter-reporter construct pGL3-p67(500) was slightly up-regulated by cFos and cJun (approximately twofold) in the absence of TNF- treatment. Treatment of these cells with TNF- induced a much higher response (approximately fivefold); however, this response occurred only in cells cotransfected with c-Fos and c-Jun (Fig. 3)
. In comparison, control cells cotransfected with pGL3-basic vector and cFos/cJun exhibited little change relative to basal activity when treated with TNF-, indicating the response was specific for pGL3-p67(500). TNF--mediated induction was absent in cells transfected with only one of the AP-1 factor expression constructs (data not shown). These results implicated a role for AP-1 factors in the TNF--dependent up-regulation of NCF2 and supported our hypothesis. It was surprising that mutation of the AP-1 site [pGL3-p67(500)–AP1] did not affect the response to cFos/cJun/TNF-, whereas the induction by c-Fos/c-Jun alone was lost (Fig. 3)
. These data demonstrate that although the AP-1 site is necessary for c-Fos/c-Jun-dependent induction in the absence of TNF-, it does not seem to be required for the TNF--mediated response. Consequently, the essential AP-1 factors required for TNF- responsiveness do not seem to use the previously mapped AP-1 binding site.
 View larger version (20K): [in a new window] |
Figure 3. The AP-1 site is not required for TNF- up-regulation of the NCF2 promoter-reporter construct. Log-phase K562 cells were transiently transfected with empty pGL3-basic vector (control), NCF2 wild-type [pGL3-p67(500)], or mutant AP-1 site [(pGL3-p67(500)–AP1] promoter-reporter constructs along with empty pcDNA3.1 expression vector (open and shaded bars) and pcDNA3.1-cFos and pcDNA3.1-cJun expression constructs (cross-hatched and solid bars), as well as the pRL-TK construct (transfection control). TNF- (20 ng/mL) was added to the appropriate cultures 4 h post-transfection. At 24 h post-transfection, untreated (open and cross-hatched bars) and TNF--treated (shaded and solid bars) cells were lysed, and luciferase activity was determined. Luciferase activity was corrected for transfection efficiency and is reported relative to the activity of the promoterless vector (pGL3-basic). The data are presented as mean ± SEM, N = 3, where each construct was transfected in triplicate. Representative of five independent experiments.
|
-responsive region of the NCF2 promoter response, we initially used the NCF2 promoter-reporter deletion constructs pGL3-p67(500), pGL3-p67(300), pGL3-p67(150), pGL3-p67(75), and pGL3-p67(40), as described previously [24
]. K562 cells were cotransfected with the NCF2 promoter-reporter constructs and c-Fos and c-Jun expression constructs, followed by TNF- treatment. As shown in Figure 4
, cells transfected with pGL3-p67(75) were fully responsive to TNF-, and a complete loss of TNF- responsiveness occurred with pGL3-p67(40), implying essential regulatory sequences were present in this region (–75 to –40 bp). Comparison of this region with transcription factor databases indicated a putative DNA binding site at –70 to –67 bp (GGGG) for myeloid zinc finger protein MZF1 [39
]. As AP-1 can regulate promoters indirectly through interactions with other transcription factors [40
, 41
], we evaluated whether AP-1 could be signaling through MZF1 in the TNF- response. To evaluate this possibility, we generated promoter/reporter constructs with MZF1 and MZF1/AP-1 site-specific mutations. However, neither of these mutations blocked TNF--induced up-regulation of the NCF2 promoter, eliminating MZF1 as a regulatory factor in this response (data not shown).
 View larger version (22K): [in a new window] |
Figure 4. Mapping of the TNF--responsive region of the NCF2 promoter. K562 cells were transfected with wild-type pGL3-p67(500) and various deletion mutants of this NCF2 construct, along with pcDNA3.1-cFos, pcDNA3.1-cJun, and pRL-TK. Transfected cells were TNF--treated and analyzed as described in Figure 3
. Numbers in parentheses indicate the number of nucleotides upstream of the ATG site that is included in each construct, superscripts indicate nucleotides that have been removed from the pGL3-p67(500) construct, and 3X(56–36) denotes three tandem repeats of bp –56 to –36. TNF- responsiveness of constructs is presented as percentage relative to the pGL3-p67(500) response, which was set at 100%. The hatched bars indicate the region of the promoter/enhancer identified as the TRR (–56 to –16 bp upstream of the TLS), and the location of the previously identified AP-1 site is indicated. The data are representative of five experiments.
|
response between constructs pGL3-p67(56) and pGL3-p67(48). In confirmation of these data, deletion constructs missing the region between –56 and –36 [pGL3-p67(500)–(56–36)] were similarly unresponsive to TNF- induction. However, constructs consisting of three tandem repeats of the –56 to –36 sequence [pGL3-3X(56–36)] did not restore TNF- responsiveness. These results indicated that although the sequence between –56 and –36 was important for the TNF- response, it was not sufficient, and there was additional sequence downstream of this region that was required. To investigate this issue, 10 bp increments of sequence were added back to the 3' end of the wild-type sequence (–500 to –57 bp) to generate additional promoter-reporter constructs. It is not surprising that the addition of increments from –56 to –36 [pGL3-p67(500–36)] did not restore the TNF- response. Whereas inclusion of an additional 10 bp [pGL3-p67(500–26)] resulted in a partial response, the full TNF- response was achieved only when a fourth increment was added [pGL3-p67(500–16)]. Using the data generated from these promoter-reporter constructs, we were able to map the TRR in the NCF2 promoter to a 40-bp segment spanning the 3' end of intron 1 and the 5' end of exon 2 (–56 to –16 bp upstream of the TLS; Fig. 4
). This sequence is located downstream of the transcription initiation sites located in intron 1 (–167 bp, –121 bp, –104 bp, and –94 bp upstream of the TLS), identified in our initial characterization of the NCF2 promoter [24
]. As there were no other obvious consensus binding sites for known transcription factors within this 40-bp sequence, we concluded that it represents a unique binding site for a known factor or a binding site for a novel transcription factor.
Analysis of the interaction of AP-1 with the TRR
To establish whether nuclear factors bound to the TRR, fluorescence DNase footprinting [36
] was performed with a 327-bp probe containing the TRR and fluorescently labeled on the sense strand, as described. As shown in Figure 5A
, the region of the probe containing the TRR was protected from DNase digestion when MonoMac6 nuclear extract was present, providing direct evidence that a factor(s) binds this region of the promoter. As expected, the region of the probe containing the AP-1 site was also protected with MonoMac6 nuclear extract (Fig. 5A)
. Similar DNase footprinting results were obtained using nuclear extracts isolated from TNF--treated cells (data not shown). Furthermore, when TRR competitor was added, the region of the probe containing the TRR became unprotected (Fig. 5B)
, verifying that this protected region of the probe corresponded to the TRR sequence. It is interesting that in the presence of TRR competitor DNA (Fig. 5B)
, the region of the probe containing the AP-1 site also became unprotected, leading us to hypothesize that AP-1 bound directly to the TRR or that AP-1 interacted with factors that bind the TRR. If AP-1 bound directly to the TRR, then a region of the TRR should also be unprotected in the presence of a double-stranded oligonucleotide containing an AP-1 consensus site. However, as shown in Figure 5C , only the AP-1 site was unprotected when using AP-1 competitor DNA, as noted by an increase in peak height in the region of the probe containing the AP-1 site, demonstrating that AP-1 must be interacting indirectly with the TRR. A random double-stranded oligonucleotide did not affect the footprint, demonstrating the specificity of the TRR and AP-1 competitor DNA (Fig. 5D)
.
 View larger version (43K): [in a new window] |
Figure 5. AP-1 interacts indirectly with the TRR. A 327-bp fluorescent double-stranded NCF2 promoter probe (400 ng) containing the TRR was subjected to DNase treatment or preincubated with MonoMac6 nuclear extract (3 µg) followed by DNase treatment, as described. (A and E) Comparison of DNase footprints of wild-type (WT; A) or AP-1 mutant (E) NCF2 TRR probe alone (orange peaks) and probe incubated with nuclear extract (green peaks). *, Areas being protected, noted by decreased peak heights (green) in these regions. (B–D, F, G) Comparison of DNase footprints of wild-type (B–D) and AP-1 mutant (F and G) NCF2 probes incubated with nuclear extract (green peaks) and with nuclear extract plus competitor DNA (brown peaks). The specific competitor DNA is indicated in each panel. The locations of the TRR and the AP-1 site in the probe, relative to the DNase footprint, were determined by comparison with ROX-350 molecular weight markers and are shown schematically below the footprints. *, Areas being unprotected by the competitors, noted by increased peak heights (brown) in these regions. Representative of at least three experiments.
|
Investigation of the role of the TRR in TNF- up-regulation of NCF2 in vivo
To evaluate the physiological relevance of the TRR, we used transcription factor decoy analyses to determine if the TRR was important for TNF--induced up-regulation of NCF2 in vivo. MonoMac6 cells were transfected with double-stranded oligonucleotide DNA corresponding to the TRR sequence or random sequence. After 24 h ± TNF- treatment, RNA was isolated, and quantitative RT-PCR was used to compare relative amounts of NCF2 message. As shown in Figure 6
(upper panel), transfection of 1 µM TRR oligonucleotide significantly inhibited TNF--induced up-regulation of endogenous NCF2 when compared with random oligonucleotide. To confirm that the TNF- treatment was effective and that the inhibition by the TRR decoy oligonucleotide was specific, relative levels of mRNA for TRAF-1, a TNF--responsive gene, were also evaluated. As shown in Figure 6
(lower panel), TRAF-1 message increased with TNF- treatment, and this increase was not affected by the TRR or the random decoy double-stranded oligonucleotide. Thus, these data clearly demonstrate the importance of the TRR in TNF--mediated up-regulation of NCF2 in vivo.
 View larger version (17K): [in a new window] |
Figure 6. The TRR is required for TNF--induced up-regulation of NCF2 in vivo. MonoMac6 cells (106) were transfected ± 1 µM double-stranded TRR or random oligodeoxynucleotide (ODN) and treated for 20–24 h ± 20 ng/mL TNF-. RNA was isolated and analyzed by quantitative RT-PCR to determine relative amounts of NCF2 (upper panel) or TRAF-1 (lower panel) mRNA relative to ribosomal 18 s mRNA levels. The data are presented as mean ± SEM, N = 3, and are representative of three experiments. Statistically significant differences between TNF--treated and untreated cells (*, P<0.05; **, P<0.01; +, P<0.001) and between TNF--treated and TNF-/ODN-treated cells (#, P<0.01) are indicated.
|
-treated human blood monocytes-treated monocytes, we considered whether this effect correlated with the level of O2•– production in these cells. Human monocytes were incubated in the presence or absence of TNF- for 0, 6, 24, and 48 h, and O2•– production was determined. In untreated monocytes, there were no significant differences in PMA-stimulated O2•– production at 6 and 24 h compared with fresh monocytes, whereas the overall level of O2•– production was significantly lower in monocytes cultured for 48 h (Fig. 7
). In addition, monocytes cultured in the presence of TNF- for 6 and 24 h generated essentially the same levels of O2•– as untreated cells. In contrast, monocytes cultured for 48 h in the presence of TNF- produced significantly higher levels of O2•– relative to untreated cells after PMA activation (Fig. 7)
. These results are generally consistent with those previously reported for monocyte cultures [42
] and support the idea that the presence of TNF- seems to preserve O2•–-generating capacity in cultured monocytes.
 View larger version (15K): [in a new window] |
Figure 7. Changes in O2•– production by cultured human monocytes, which were cultured in the presence (solid bars) or absence (open bars) of 20 ng/mL TNF-. At the indicated time-points, cells were collected and activated with 80 nM PMA. The kinetics of O2•– production was monitored by a chemiluminescence assay, and luminescence was integrated over 60 min, as described. The data are presented as mean ± SEM, N = 3, and are representative of three experiments. Statistically significant differences between TNF--treated and untreated cells (*, P<0.05) are indicated.
|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
is a potent cytokine produced by many cell types in response to inflammation, infection, and injury (reviewed in ref. [43
]). Cellular responses to TNF- include activation, migration, fever, proliferation, differentiation, and apoptosis [44
]. Although TNF- exposure can activate a caspase cascade leading to apoptosis [45
], more commonly the result is activation of two transcription factors, AP-1 and NF-
B, which leads to the induction of genes involved in chronic and acute inflammatory responses [46
, 47
]. The existence of an AP-1 site in the NCF2 promoter and data suggesting p67phox is up-regulated by TNF- prompted us to investigate the role of TNF- in the regulation of NCF2 and to determine if the previously identified AP-1 site in intron 1 [24
, 30
] was required for this regulation. We demonstrate conclusively here the up-regulation of NCF2 mRNA and p67phox protein levels in TNF--treated promyelocytic and monocytic cell lines, as well as in primary monocytes but not in neutrophils.
In contrast to the data presented here, Dusi et al. [25
] previously reported p67phox expression was unchanged in monocytes treated for 48 h with TNF-, as compared with untreated cells. Although the reason for this difference may simply be a result of methodological differences and assay sensitivity, another plausible explanation could be donor variability. As mentioned in our results, the consistent increase in NCF2 mRNA levels observed at 48 h of TNF- treatment was quite variable from donor to donor, ranging from 1.8- to 5.1-fold. This variability might not have been detected in the two blots performed by this group, and we evaluated this issue using quantitative RT-PCR and immunoblotting of a number of samples.
Considering the differences in regulation and kinetics of the respiratory burst between monocytes and neutrophils (reviewed in ref. [48
]), the lack of TNF- up-regulation of NCF2 in neutrophils was not surprising. Activated neutrophils produce an immediate respiratory burst, when compared with monocytes, which peaks at 2–10 min, depending on the stimulus [49
, 50
]. Considering the large stores of NADPH oxidase components already present [51
, 52
] and the rapid response time observed for neutrophils [53
], it is unlikely that there would be sufficient time for TNF--mediated regulation of NCF2 to significantly affect the oxidative response. Conversely, it is well documented that neutrophils become primed by TNF- treatment, resulting in an enhanced production of O2•– when the cells encounter a subsequent physiological stimulus (reviewed in refs. [49
, 54
]). Neutrophil priming by TNF- has been shown to induce a number of effects, including degranulation, up-regulation of cell-surface receptors, and up-regulation of receptor affinity, thereby enhancing neutrophil responsiveness in acute inflammation (reviewed in ref. [54
]).
Compared to neutrophils, monocytes initially respond with a lower level of NADPH oxidase activity, which gradually increases over time and peaks at 
1 h but can still be detected out to several hours [50
, 55
]. Monocytes can also be stimulated transiently, a process that is not normally observed in neutrophils [56
]. Furthermore, neutrophils and monocytes respond differently to various activating agents [54
]. Overall, the kinetics of activation, the extended duration of O2•– production, and the ability to be transiently stimulated suggest that NADPH oxidase is regulated quite differently in phagocyte subsets and that gene regulation could contribute to the overall modulation of oxidant production in monocytes during chronic inflammation (reviewed in ref. [48
]).
We cannot rule out the possibility that the TNF- response observed in this study was a result of TNF--induced cell differentiation [57
], as most of the cells tested are relatively undifferentiated myeloid cell lines. However, the MonoMac6 cell line, which shows a surface phenotype resembling that of mature monocytes, and freshly isolated blood monocytes demonstrated at least a twofold increase in NCF2 mRNA and subsequent p67phox protein in response to TNF- treatment. If p67phox is the limiting factor in assembly and activation of the NADPH oxidase, then a twofold increase in p67phox protein levels could significantly influence reactive oxygen species (ROS) production. This idea is supported by previous studies of Uhlinger et al. [9
], who showed p67phox dose-dependently enhanced NADPH oxidase activity in a semi-recombinant, cell-free system containing fixed concentrations of p47phox. In cultured monocytes, however, the situation is far more complex with changes potentially occurring in the levels and/or function of many of the oxidase proteins [42
, 50
, 58
]. In culture, monocyte differentiation into macrophages is accompanied by a gradual loss of NADPH oxidase activity [59
]. In our monocyte preparation, this process seems to begin somewhere around 48 h, as indicated in our functional assays (see Fig. 7
). In the presence of TNF-, oxidase activity is significantly preserved at a higher level; however, it appears that the differentiation process will ultimately result in loss of activity whether TNF- is present or not. We demonstrate here that TNF- induces up-regulation of monocyte p67phox at 24 and 48 h; however, this effect did not always correlate with changes in oxidase activity. For example, similar oxidase activity was observed at 24 h in untreated and TNF--treated cells. One possible explanation for this finding is that small increases in p67phox do not make much of a difference in normal cells expressing the full complement of all oxidase components. This possibility is supported by kinetic studies in recombinant cell-free assays showing that the ability of additional p67phox to enhance oxidase activity at a given p47phox concentration was saturable [9
]. At 48 h, p67phox and oxidase activity were increased, suggesting a correlation. Whether this correlation is causal or unrelated to p67phox expression is still unclear. During monocyte differentiation, a number of extremely complex changes appear to contribute to loss of oxidase function, including loss of granule pools of flavocytochrome b [58
] and decreased expression of flavocytochrome b subunits and p47phox [42
]. In this context, small increases in p67phox expression could significantly impact NADPH oxidase activity. Nevertheless, further studies are clearly necessary to determine the functional significance of TNF--induced up-regulation of p67phox and its effects on generation of oxidants in vivo.
Activation of AP-1 by TNF- during an inflammatory response results in expression of various genes, such as proinflammatory cell adhesion molecules and tissue-remodeling proteases [33
, 47
]. The identification of an AP-1 site within intron 1 of NCF2 suggested that TNF--mediated up-regulation of p67phox might involve AP-1 activation via JNK and subsequent binding of AP-1 to its site in the NCF2 promoter. Indeed, promoter-reporter assays demonstrating the requirement for c-Fos and c-Jun in TNF--mediated up-regulation of NCF2 supported this line of reasoning. Thus, it was somewhat surprising to find that the previously characterized AP-1 site in the NCF2 promoter was not necessary for this response. To explain this observation, we hypothesized that AP-1 might be regulating the promoter indirectly via protein–protein interactions with another factor that specifically bound to the promoter at a different site, that AP-1 might be regulating a factor that was more directly involved in NCF2 regulation, or that AP-1 was binding a novel site in the promoter/enhancer of NCF2. In evaluation of these possibilities, we used NCF2 promoter-reporter constructs to map a 40-bp sequence of the promoter, which was necessary for the TNF- response. This region, located between –56 and –16 bp upstream of the translational start site, was designated as the TRR and represents a unique binding site for a known factor or a binding site for a novel transcription factor. It is interesting that we also observed a polymorphism in the TRR region during the initial cloning of the promoter/enhancer of NCF2, which was located at –47 to –43 bp upstream of the TLS. Three different sequences (5'-CCCTC-3', 5'-CTCCC-3', and 5'-CTCTC-3') were observed in this region after sequencing genomic DNA from five additional individuals. We did not find that this polymorphism noticeably affected the TNF- responsiveness of promoter/reporter constructs; however, further studies are required to determine if there is any physiological relevance of this polymorphism in vivo.
Although there was no suggestion of consensus binding sites for known transcription factors in the TRR, the TRR sequence is T-C (sense strand)- and A-G (antisense strand)-rich, which is also a feature of the core-binding site for PU.1 (5'-C/AGGAA/T-3') [60 , 61 ]. As PU.1 was previously shown to regulate NCF2 [24 , 30 , 62 ], we considered whether PU.1 might be involved in TRR regulation. We were able to demonstrate the binding of 35S-methionine-labeled PU.1 protein to the TRR sequence in electrophoretic mobility shift assays (EMSA). However, no supershift was observed when HL-60 nuclear extracts were incubated with PU.1 antibody and analyzed by EMSA (K. A. Gauss, unpublished data), suggesting that PU.1 from nuclear extracts does not bind the TRR probe. The variability in sequence of known PU.1 binding sites, as described by Li et al. [63 ], suggests that PU.1 is fairly promiscuous in binding to A-G-rich sequences. Thus, it is not clear whether the binding of labeled PU.1 protein to the TRR was physiologically relevant or whether it was simply a result of high amounts of added PU.1 protein binding nonspecifically (promiscuously) to the A-G-rich TRR. Further studies will be necessary to determine the relevance, if any, of PU.1 binding to the TRR.
DNase footprinting was used to show that nuclear factors do bind the TRR, as the entire region was protected from DNase, demonstrating the regulatory significance of this sequence. The footprint studies also demonstrated that the TRR was protected with nuclear extract from untreated and TNF--treated cells. It is not clear from these results if it is the same or different factors that bind to this sequence under both conditions. One possibility is that it is the same factor binding under both conditions, and it becomes activated only after TNF- treatment, possibly by the recruitment of a coactivator(s) such as the cyclic adenosine monophosphate response element-binding protein-binding protein (CBP) [64
]. Recruitment of CBP to a NCF2 cis element has been shown previously by Eklund and Kakar [62
], who demonstrated that this particular element was activated by PU.1, interferon (IFN) regulatory factor 1, and IFN consensus-binding protein and that recruitment of CBP to this site was the mechanism of activation by these factors. Another possible explanation is that completely different factors bind this region as a result of transcription factor competition for the same or an overlapping DNA binding site. An example of this type of competition between factors is demonstrated by the activating transcription factor Nkx2.5 and the transcriptional repressor 
EF1/ZEB1, which vie for an overlapping DNA binding site within the enhancer region of the pro-colla2 gene [65
]. In any case, we found that the TRR is constitutively bound by nuclear protein, clearly demonstrating its importance as a regulatory region in the promoter/enhancer of NCF2.
The use of double-stranded TRR and AP-1 competitor DNA in DNase footprinting studies demonstrated AP-1 was interacting indirectly with the TRR. This type of indirect interaction of AP-1 with DNA has been demonstrated previously [40
]. In particular, the monocyte-specific macrophage-colony stimulating factor (M-CSF) receptor gene, which is regulated by AP-1, does not contain an AP-1 binding site. In this case, c-Jun interacts with the ETS domain of PU.1 to act as a coactivator of PU.1, resulting in the expression of the M-CSF receptor gene. DNase footprinting studies shown here support promoter-reporter studies implicating AP-1 in the TNF- response and substantiate the hypothesis that AP-1 interacts with another factor that binds directly to the promoter. Together, these data suggest a model where AP-1 acts as a coactivator in the regulation of NCF2 by binding the TRR indirectly via another protein, which binds directly to the TRR. Furthermore, transcription factor decoy experiments confirmed the physiological relevance of the TRR in TNF--mediated regulation of NCF2 in vivo. Of great interest, of course, is the identification of the NF or factors that bind the TRR, and current studies using EMSA, DNA-affinity purification, and yeast one-hybrid screening are under way to address this issue.
Contact-mediated signaling of monocytes by stimulated T lymphocytes is a potent inflammatory mechanism that triggers massive up-regulation of the proinflammatory cytokines interleukin-1 and TNF- in monocytes [66
, 67
]. Activated macrophages play a pivotal part in the pathogenesis of chronic inflammatory diseases such as rheumatoid arthritis and atherosclerosis (reviewed in refs. [48
, 68
]). As an example, recent data suggest that the production of O2•– and other oxidants by activated macrophages functions at several different levels in contributing to the pathogenesis of atherosclerosis, including oxidation of lipids and induction of stress responses that alter cell function (reviewed in ref. [48
]). Furthermore, O2•– is an effective scavenger of nitric oxide and thus, could regulate endothelial relaxation and also generate highly reactive peroxynitrite [69
]. The studies provided here suggest an additional pathogenic mechanism whereby TNF- produced by activated macrophages could serve as an autocrine/paracrine regulator of macrophage NCF2 and consequently, p67phox expression. This could result in an increased and/or prolonged production of O2•–, further contributing to the pathogenesis of chronic inflammatory diseases. It is interesting that AP-1 has also been shown to be activated by H2O2 produced by the NADPH oxidase in the macrophage cell line NR8383 [70
], again implicating an autocrine/paracrine regulatory mechanism. Furthermore, Barbieri et al. [71
] recently showed that NADPH oxidase-generated ROS played an important role in monocyte differentiation into macrophages, suggesting an additional role of ROS in inflammatory responses within the atheroma. Overall, a more complete understanding of the role of transcriptional regulation in NADPH oxidase function will be essential in elucidating inflammatory disease pathogenesis.
Received June 8, 2004; revised October 5, 2004; accepted October 12, 2004.
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
but not colony stimulating factor 1 J. Leukoc. Biol. 55,530-535[Abstract]
-interferon Cancer Res. 52,2530-2537B: a pivotal transcription factor in chronic inflammatory diseases N. Engl. J. Med. 336,1066-1071 RI. Oxidase deactivation precedes internalization of cross-linked receptors J. Immunol. 143,2640-2649[Abstract]-induced p67phox and gp91phox expression J. Immunol. 163,6095-6105EF1/ZEB1 Mol. Cell. Biol. 24,6151-6161 and tumor necrosis factor receptors by human monocytic cell lines upon direct contact with stimulated T cells Eur. J. Immunol. 26,2404-2409[Medline]This article has been cited by other articles:
 |
|
 |
|
  K. A. Gauss, L. K. Nelson-Overton, D. W. Siemsen, Y. Gao, F. R. DeLeo, and M. T. Quinn Role of NF-{kappa}B in transcriptional regulation of the phagocyte NADPH oxidase by tumor necrosis factor-{alpha} J. Leukoc. Biol., September 1, 2007; 82(3): 729 - 741. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 M. C. B. Ammons, D. W. Siemsen, L. K. Nelson-Overton, M. T. Quinn, and K. A. Gauss Binding of Pleomorphic Adenoma Gene-like 2 to the Tumor Necrosis Factor (TNF)-{alpha}-responsive Region of the NCF2 Promoter Regulates p67phox Expression and NADPH Oxidase Activity J. Biol. Chem., June 15, 2007; 282(24): 17941 - 17952. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 I. A. Schepetkin, L. N. Kirpotina, A. I. Khlebnikov, and M. T. Quinn High-Throughput Screening for Small-Molecule Activators of Neutrophils: Identification of Novel N-Formyl Peptide Receptor Agonists Mol. Pharmacol., April 1, 2007; 71(4): 1061 - 1074. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
