LPS-Induced G-CSF Expression in Macrophages Is Mediated by ERK2, but Not ERK1

Granulocyte colony-stimulating factor (G-CSF) selectively stimulates proliferation and differentiation of neutrophil progenitors which play important roles in host defense against infectious agents. However, persistent G-CSF production often leads to neutrophilia and excessive inflammatory reactions. There is therefore a need to understand the mechanism regulating G-CSF expression. In this study, we showed that U0126, a MEK1/2 inhibitor, decreases lipopolysaccharide (LPS)-stimulated G-CSF promoter activity, mRNA expression and protein secretion. Using short hairpin RNA knockdown, we demonstrated that ERK2, and not ERK1, involves in LPS-induced G-CSF expression, but not LPS-regulated expression of TNF-α. Reporter assays showed that ERK2 and C/EBPβ synergistically activate G-CSF promoter activity. Further chromatin immunoprecipitation (ChIP) assays revealed that U0126 inhibits LPS-induced binding of NF-κB (p50/p65) and C/EBPβ to the G-CSF promoter, but not their nuclear protein levels. Knockdown of ERK2 inhibits LPS-induced accessibility of the G-CSF promoter region to DNase I, suggesting that chromatin remodeling may occur. These findings clarify that ERK2, rather than ERK1, mediates LPS-induced G-CSF expression in macrophages by remodeling chromatin, and stimulates C/EBPβ-dependent activation of the G-CSF promoter. This study provides a potential target for regulating G-CSF expression.


Introduction
Granulocyte colony-stimulating factor (G-CSF), a hematopoietic growth factor, regulates the proliferation of neutrophil progenitors, and the differentiation of granulocyte lineages, and the survival and maturation of neutrophil progenitors, and their mobilization from bone marrow to peripheral tissues [1]. For decades, recombinant G-CSF has been widely used in patients receiving chemotherapy to increase the number of circulating hematopoietic progenitor cells and in certain patients with neutropenia. Endogenous G-CSF is produced by various types of solution in PBS. MMLV reverse transcriptase was from Promega (Madison, WI, USA). Super-Fect Transfection reagent was purchased from Qiagen (Hilden, Germany). Antibodies against ERK, phospho-ERK, p50, p65, C/EBPβ, β-actin, or lamin B were obtained from Santa Cruz Biotechnology (Santa Cruz, CA, USA). U0126 and PD98059 (specific MEK inhibitors) were purchased from Calbiochem (San Diego, CA) and were dissolved as stock solutions in DMSO, and were added to the culture medium 30 min before other treatments or as indicated with 0.1% DMSO in culture medium as the control.

Cell culture and LPS treatment
The RAW264.7 murine macrophage cell line was cultured as described previously [6]. THP-1, a human acute monocytic leukemia cell line, was maintained and induced to differentiate into macrophages using 160 nM of phorbol 12-myristate-13-acetate (PMA) as described previously [24]. G-CSF mRNA and protein levels were compared in untreated cells and cells pretreated with the indicated concentration of inhibitor or DMSO for 30 min, then LPS was added and the cells incubated for the indicated time; unless otherwise stated, LPS treatment was at 100 ng/ ml for 6 h and the concentration of U0126 was 10 μM.

RNA isolation and RT-PCR analyses
Total cellular RNA was isolated from cells using TRIzol reagent (Invitrogen) according to the manufacturer's protocol and its concentration was determined from the absorbance at 260 nm, then it was subjected to reverse transcription (RT)-polymerase chain reaction (PCR) analysis as described previously [6]. G-CSF, TNF-α, and GAPDH mRNA levels were determined by RT-PCR using the primers listed in S1 Table. Preparation of cytosolic and nuclear extracts Cytosolic and nuclear extracts were prepared as described previously [25]. Briefly, cells were washed twice with ice-cold PBS, incubated on ice for 15 min with lysis buffer (10 mM HEPES, pH 7.4, 10 mM NaCl, 1.5 mM MgCl 2 , 0.5 mM DTT, 0.2% Nonidet P-40, and 0.2 mM PMSF), and collected by gentle pipetting. After centrifugation at 500 xg for 5 min at 4°C, the supernatant was collected as the cytosolic extract, and the pellet was extracted by incubation for 15 min at 4°C with intermittent vortexing in nuclear extract buffer (20 mM HEPES-KOH, pH 7.9, 25% glycerol, 420 mM NaCl, 1.5 mM MgCl 2 , 0.2 mM EDTA, 0.5 mM DTT, 0.2 mM PMSF, and 1x protease inhibitor cocktail), followed by centrifugation at 12,900 xg for 10 min at 4°C, after which the supernatant was collected as the nuclear extract. The Bradford method (DC Protein Assay; Bio-Rad) was used to measure the protein concentrations in the extracts, which were then stored in aliquots at-80°C. antibody was detected using an improved chemiluminescence detection system (NEN Life Science Products). Protein concentrations were determined by the Bradford method (DC Protein Assay, Bio-Rad).

Quantitation of G-CSF in culture medium
The concentrations of human (THP-1 cells) or mouse (RAW264.7 cells) G-CSF in the culture medium were measured by enzyme-linked immunosorbent assay using human or mouse G-CSF Quantikine ELISA kits (R&D Systems) according to the manufacturer's instructions.

Plasmids
Expression plasmids encoding mouse p50, Oct-2, or C/EBPβ were, respectively, generous gifts from Dr. Neil D. Perkins, University of Michigan, Dr. Winship Herr, Cold Spring Harbor Laboratory, and Dr. Gerard Elbery, Section of Nephrology, Department of Pediatrics, University of Oklahoma Health Sciences Center. Plasmids encoding the T188A or S64A mutant of C/EBPβ were generated from the wild-type C/EBPβ-expressing plasmid by PCR mutagenesis and confirmed by sequencing analysis. A plasmid expressing the constitutively active form of ERK2 (CA-ERK2) was obtained from Upstate Co. (Charlottesville, VA, USA). All plasmids were amplified in the E. coli DH5α host strain and purified by equilibrium centrifugation in a CsCl −EtBr gradient as described previously [24]. The purity and stability of plasmid preparations were confirmed by agarose gel electrophoresis followed by ethidium bromide staining, and the DNA concentration was measured by UV absorption at 260 nm.

Transient transfection and reporter gene activity assay
Transient transfection was carried out as described previously [25]. Briefly, RAW264.7 cells were plated and cultured in 24-well plates overnight before transfection. To determine the role of ERK in regulating the G-CSF promoter activity, 0.5 μg of the pG-CSF(-283/+35)-Luc plasmid and 0.1 μg of the phRLTK plasmid were mixed with 0.15-0.6 μg of the expression plasmid containing p50, Oct-2, CA-ERK, or C/EBPβ cDNA or mixtures of the plasmid encoding C/ EBPβ and that encoding p50, Oct-2, or CA-ERK, and the total amount of DNA was adjusted to 1.2 μg with pcDNA3.1. The DNA mixture was transfected into RAW264.7 cells using Super-Fect transfection reagent, then 24 h later, cell lysates were prepared using the lysis buffer in the kit and Photinus and Renilla luciferase activities were measured using the Dual-Luciferase reporter assay system as described previously [25]. The light intensity produced by Photinus luciferase (test plasmid) was normalized to that produced by Renilla luciferase (control plasmid).

Chromatin immunoprecipitation (ChIP) assay
The ChIP assay was performed according to the manufacturer's instructions (Upstate Biotechnology Inc.) as described previously [6]. Briefly, after treatment, the cells were fixed with 1% formaldehyde for 10 min at 37°C to cross-link DNA and protein, collected by scraping, and sonicated to fragment the chromatin. Immunoprecipitation was performed using control rabbit IgG or rabbit antibodies against ERK, C/EBPβ, p50, p65, or histone 3 (Santa Cruz), the cross-links were reversed at 65°C for 4 h and digested with proteinase K (Sigma-Aldrich) for 1 h at 45°C to remove proteins, then the immunoprecipitated DNA was recovered by phenol/ chloroform extraction and ethanol precipitation before being used as template for PCR with the following primers: G-CSF promoter (-248 to -70), forward (5'-TGGCTGGaagAGAGGAA GAG-3') and reverse (5'-CTGGGGCAACTCAGGCTTA-3') or TNF-α promoter (-270 to -4), forward (5'-CTGATTGGCCCCAGATTG-3') and reverse (5'-CTTCTGCTGGCTGGCTGT-3') (GenBank: Y00467.1). Ten percent of the chromatin DNA used for immunoprecipitation was also subjected to PCR analysis (indicated as "input"). The number of PCR cycles was 30 or 31 for all ChIP experiments and 24 for the input samples.

DNase I accessibility assay
After 1% formaldehyde treatment, the cells were gently lysed with 500 μl of ice-cold Nonidet P-40 lysis buffer (10 mM Tris-HCl, pH 7.4, 10 mM NaCl, 3 mM MgCl 2 , 0.5% Nonidet P-40, 0.15 mM spermine, 0.5 mM spermidine), and the lysates centrifuged at 500 xg for 5 min at 4°C to pellet the nuclei, which were washed with 200 μl of Nonidet P-40 lysis buffer and centrifuged again at 500 xg for 5 min at 4°C. After gentle rinsing with 200 μl of ice-cold buffer A (100 mM NaCl, 50 mM Tris-HCl, pH 8.0, 3 mM MgCl 2 , 0.15 mM spermine, and 0.5 mM spermidine), the pelleted nuclei were suspended in buffer A containing 1 μM CaCl 2 for DNase I (Promega) treatment, which was performed at 37°C for 5 min for THP-1 cells or for 2 min for Raw264.7 cells, and stopped with ice-cold 10 mM EDTA, pH 8.0. After reversal of the formaldehyde cross-linking at 65°C for 4 h, proteinase K and RNase A were added to the sample, which were then incubated at 37°C overnight, then DNA was extracted using the standard phenol-chloroform technique. Care was taken to use cut-off tips and very gentle pipetting to reduce non-specific DNA sheering. After precipitation, the DNA was dissolved in H 2 O.
Real-time PCR was used to identify the DNase-sensitive sites using the primers (final concentration 400 nM) listed in S1 Table. DNase I-treated and non-digested DNA were tested in triplicate. All PCR reactions were performed with SYBR Green PCR Master Mix in an ABI PRISM 7300 Sequence Detection System (Applied Biosystems 7300 System Sequence Detection System software version 1.3.) ΔCt values were determined by subtracting the Ct value of the control reaction. The conditions for the PCR were as follows: 50°C for 2 min, 95°C for 10 min, 40 cycles of 95°C for 15 sec, and 60°C for 1 min. At the end of the PCR cycles, amplification specificity was confirmed by dissociation curve analysis, and the products were separated on a 3% agarose gel and stained with ethidium bromide for visual confirmation of the PCR product.

Statistical analyses
Results are shown as the mean + SD. Differences between mean values were evaluated using Student's t test and were considered significant at p <0.05.

Inhibition of ERK activation prevents LPS-induced G-CSF expression in macrophages
An initial experiment was designed to confirm the involvement of ERK activation in LPS-stimulated G-CSF production in macrophages. Phosphorylation of ERK1/2 was detected in mouse Raw264.7 macrophages after 15 to 60 min of treatment with LPS (100 ng/ml) and was blocked by 30 min pretreatment with U0126 (10 μM) (Fig 1A). To examine whether U0126 inhibited LPS-induced G-CSF production, Raw264.7 cells were incubated with or without U0126, then with LPS for 0 to 24 h. Fig 1B shows that LPS-induced G-CSF secretion in the culture medium was time-dependent and was abolished by U0126 pretreatment. In addition, the inhibitory effect of U0126 on LPS-induced G-CSF production was dose-dependent ( Fig 1C). Since LPS is known to increase G-CSF mRNA expression in mouse bone marrow-derived macrophage (BMDMs), we then examined whether the ERK inhibitors U0126 or PD98059 could block this effect in primary cells. The results in S2A and S2B Fig show that G-CSF mRNA levels were significantly increased after LPS treatment for 6 h and this effect was blocked by pretreatment cells with 0.01 or 0.1 μM of U0126 or 1 or 10 μM of PD98059. ERK2 is critical for LPS-induced G-CSF expression ERK1 and ERK2 are isoforms of the 'classical' MAPKs and both are activated by MAP/ERK kinase 1 (MEK1) and MEK2, members of the MAPKK family. To examine whether U0126 inhibited LPS-induced phosphorylation of ERK1/2 in THP-1 macrophages, cells were induced to differentiate into macrophages by incubation with 160 nM PMA, and then were incubated with LPS for 0.5, 1 or 2 h. Phosphorylation of ERK1/2 was detected in THP-1 macrophages after treatment with LPS (100 ng/ml) and was blocked by 30 min pretreatment with U0126 (10 μM) (S3 Fig). To determine whether ERK1 and/or ERK2 involved in LPS-induced G-CSF expression, ERK expression in THP-1 human acute monocytic leukemia cells was knocked down using shRNA clones ERK1a and ERK1b carrying two different ERK1 sequences or shRNAs ERK2a and ERK2b carrying two different ERK2 sequences. The luciferase shRNA was used as a negative control. Knockdown efficiency and specificity were confirmed by Western blotting analysis (Fig 2A and S4 Fig). After shRNA-lentiviral transduction and puromycin selection, the cells were induced to differentiate into macrophages by incubation with 160 nM PMA, and then were incubated with LPS or PBS for 4 h. As shown in Fig 2B and 2C, a significant increase at LPS-induced G-CSF protein and mRNA levels was detected in control luciferase shRNA knockdown and ERK1 knockdown cells, but not in ERK2 knockdown cells. Whereas, neither ERK1 nor ERK2 knockdown had a significant effect on LPS-induced TNF-α mRNA expression (Fig 2D), though both basal and LPS-induced TNF-α mRNA levels were slightly lower in the ERK2 knockdown cells. These results show that ERK2, but not ERK1, is essential for LPS-induced G-CSF expression, but is not a common regulator for the expression of LPS-activated genes.

Cooperative regulation of G-CSF promoter activation by ERK2 and C/ EBPβ
The results above suggested that ERK2 might involve in LPS-activated G-CSF expression at the transcriptional level. To further clarify whether ERK2 activation affected the G-CSF promoter activity, reporter assays were performed. Raw264.7 cells were transfected with a reporter plasmid containing the luciferase gene driven by the G-CSF promoter (-283/+35) (pG-CSF(-283/+35)-Luc), then were incubated with or without U0126, followed by LPS treatment for 6 h. As shown in S5A Fig, U0126 pretreatment reduced LPS-activated G-CSF promoter activity by 60%, supporting the idea that LPS-activated G-CSF promoter activity requires activation of the MEK/ERK pathway. We therefore investigated the role of ERK2 in G-CSF promoter activation and possible regulatory factors involved. Raw264.7 cells were co-transfected with the pG-CSF(-283/+35)-Luc reporter plasmid and (i) a plasmid encoding constitutively active ERK2 (CA-ERK2) or p50, Oct-2, or C/EBPβ, all of which are involved in LPS-regulated G-CSF expression, or (ii) with pairs of plasmids consisting of the C/EBPβ-encoding plasmid and the CA-ERK2, p50, or Oct-2 plasmid. As shown in Fig 3A, co-transfection with pG-CSF(-283/+35)-Luc and the plasmid encoding p50, Oct-2, C/EBPβ, or CA-ERK2 resulted, respectively, in a 0.3-, 0.7-, 6.2-, or 2.7-fold increase in luciferase activity compared to transfection with pG-CSF(-283/+35)-Luc alone. In addition, cotransfection with the pG-CSF(-283/+35)-Luc reporter plasmid, the CA-ERK2-encoding plasmid, and the plasmid encoding C/EBPβ resulted in the highest luciferase activity, with a 28-fold increase compared to the control with the Luc reporter plasmid alone. In contrast, co-transfection with the pG-CSF(-283/+35)-Luc reporter, CA-ERK2-encoding plasmid, and either the p50 or Oct-2 expression plasmid resulted in no significant increase in luciferase activity compared to that detected in the cells co-transfected with CA-ERK2-encoding plasmid alone (Fig 3A). These findings suggest the importance of the interaction between ERK2 and C/EBPβ in regulating LPSstimulated G-CSF promoter activity. Since ERK1/2 is known to phosphorylate C/EBPβ and thereby regulate its transcriptional factor activity [26], we next examined whether LPS treatment led to ERK1/2-mediated phosphorylation of C/EBPβ. As shown in Fig 3B, LPS treatment of RAW264.7 cells for 4 h increased both the protein and phosphorylation levels of nuclear C/ EBPβ, and phosphorylation level was partially inhibited, but not the protein level, by U0126 pretreatment. Serine 64 in the transactivation domain and threonine 188 in the regulatory domain of C/EBPβ, both have been proposed as potential target sites for phosphorylation by ERK [22,27,28]. To clarify the phosphorylation sites on C/EBPβ by LPS-activated ERK, these two amino acids were separately mutated to alanine by point mutation to create the C/EBPβ-S64A and C/ EBPβ-T188A mutants. In RAW264.7 cells co-transfected with the pG-CSF(-283/+35)-Luc reporter, the CA-ERK2-encoding plasmid, and either the wild-type C/EBPβ or one of the mutant C/EBPβ constructs, a 45% decrease in luciferase activity was detected in cells with C/EBPβ T188A compared to that with wild type C/EBPβ, whereas the C/EBPβ-S64A mutant had no significant effect (Fig 3C). We then mutated the consensus C/EBPβ binding site in the G-CSF promoter to examine its role in LPS-ERK-induced G-CSF expression. Co-transfection of Raw264.7 cells with increasing amounts of the CA-ERK2-expressing plasmid (0-0.6 μg) and the pG-CSF (-283/+35)-Luc reporter carrying either the consensus C/EBPβ binding site (WT: ATTTCA-CAAA) or the mutant C/EBPβ binding site (GCGTCACGCA) resulted in a 2.8-to 16-fold increase in luciferase activity using reporter carrying the WT binding site, but only a 1.6-to 4.3-fold increase with the reporter construct carrying the mutant C/EBPβ binding site (Fig 3D). Taken together, these results demonstrate that the ERK2-mediated increase in G-CSF promoter activity, at least in part, involves phosphorylation of C/EBPβ at Thr188.

U0126 inhibits LPS-induced binding of nuclear NF-κB and C/EBPβ to the G-CSF promoter region and decreases its accessibility to DNase
To determine whether U0126 inhibited LPS-induced G-CSF expression by affecting transcription factors known to regulate G-CSF expression, we analyzed levels of the transcriptional regulators including p50, p65, and C/EBPβ in the nucleus of RAW264.7 cells. As shown in Fig 4A, nuclear levels of p50, p65, and C/EBPβ were increased after 6 h of LPS treatment and this effect was unaffected by pretreatment with U0126. These results show that nuclear levels of these transcriptional factors are irrelevant for the ERK-regulated G-CSF transcription. We then used ChIP assays to investigate the effect of U0126 on the bindings of p50, p65, and C/EBPβ to the promoters of G-CSF and TNF-α in cells treated with LPS. Fig 4B shows that binding of p50, p65, or C/EBPβ to the G-CSF promoter and the TNF-α promoter was induced by LPS and that U0126 pretreatment inhibited binding of all three to the G-CSF promoter, but not the TNF-α promoter (Fig 4B). To clarify whether U0126 pretreatment affected LPS-stimulated NF-κB transcriptional activity, Raw264.7 cells were transfected with an NF-κB-driven luciferase reporter (pNF-κB-Neo-Luc), then, at 24 h after transfection, were incubated with LPS with or without U0126 pretreatment. As expected, LPS treatment for 6 h significantly upregulated NF-κB-driven luciferase activity and this effect was not affected by U0126 pretreatment (S5B Fig). These results indicate that LPS-triggered NF-κB activation, including nuclear levels of NF-κB and its transactivation activity, was not affected by U0126.
Activation of the MEK-ERK signaling pathway induces histone phosphorylation, causing nucleosome remodeling at the promoters of immediate early genes [29]. DNase accessibility analysis was therefore employed to determine whether LPS activation of ERK resulted in modulation of chromatin remodeling at the promoter region of the G-CSF gene. Nuclei of RAW264.7 or THP-1 macrophages pretreated with DMSO or U0126, then treated with LPS for 4 h were subjected to DNase I digestion and the amount of undigested DNA in the G-CSF and TNF-α promoter regions containing C/EBPβ binding sites was determined by quantitative real-time PCR. In Raw264.7 cells, the amount of undigested DNA in region -167/+12 of the G-CSF promoter ( Fig 5A) and region -270/-4 of the TNF-α promoter (Fig 5B) was decreased after LPS treatment, while U0126 pretreatment restored the amount of undigested G-CSF promoter DNA, but not TNF-α promoter DNA, to the control level. Similar results were detected in human THP-1 macrophages that LPS treatment decreased the amounts of undigested DNA within the promoter regions of G-CSF (Fig 6A) and TNF-α (Fig 6B) that contain a C/EBPβ binding site. Again, pretreatment with U0126 restored the amount of undigested DNA within the G-CSF promoter, but not the TNF-α promoter (Fig 6A and 6B). Furthermore,  Knockdown of ERK2, but not ERK1, decreases LPS-induced DNase I accessibility of the G-CSF promoter. THP-1 cells in which either ERK1 or ERK2 was knocked down using shRNA were induced to differentiate into macrophages. After incubation with DMSO or U0126 (10 μM), the cells were incubated with LPS (100 ng/ml) or PBS for 4 h, then nuclei were purified and subjected to DNase I digestion for 5 min. Genomic DNA was then isolated and the amount of undigested DNA within region -624/-450 of the G-CSF promoter (A) or within region -228/-71 of the TNF-α promoter (B) was determined by quantitative real-time PCR and expressed as a percentage of that in cells not treated with LPS and U0126. The results are the mean ± SD for three independent experiments. *p<0.05 compared to the corresponding control. knockdown of ERK2 in THP-1 cells resulted in an increased amount of undigested G-CSF promoter DNA (Fig 6A), but had no significant effect on the TNF-α promoter (Fig 6B) in LPStreated cells compared to LPS-treated knockdown control cells. Whereas, ERK1 knockdown had no significant effect on the amount of undigested G-CSF and TNF-α promoter DNA in LPS-treated cells compared to LPS-treated control knockdown cells. These results reveal that LPS treatment increases the accessibility of both the G-CSF and TNF-α promoters to DNase I and that U0126 pretreatment or ERK2 depletion abrogates LPS effect on DNase accessibility of the G-CSF promoter, but not the TNF-α promoter region.

Discussion
Although recombinant G-CSF is widely used clinically, regulation of endogenous G-CSF expression has not been well characterized. In this study, we used a highly selective inhibitor of MAP/ERK kinase, U0126, and shRNA knockdown of ERK1 or ERK2 to demonstrate that LPSinduced activation of ERK2, but not ERK1, was required for LPS-induced G-CSF expression in macrophages. Moreover, our results showed that LPS-activated ERK2 interacts with C/EBPβ and increases the accessibility of the G-CSF promoter to DNase I and transcriptional factor binding to the G-CSF promoter, leading to up-regulation of G-CSF expression.
ERK signaling is known associated with various cellular processes, including proliferation, differentiation, and survival. The two classical ERKs, ERK1 and ERK2, co-express in most tissues, but vary in relative abundance. They involve in the transcriptional regulation of multiple cellular processes by phosphorylating many substrates, including Elk1, c-Myc, and ribosomal S6 kinase, and by activating c-fos. An animal study showed that ERK2 knockout is lethal and cannot be compensated for by ERK1, while ERK1 knockout has only a mild phenotypic effect [30]. It is therefore proposed that ERK1 and ERK2 have different biological functions [31]. It is also known that ERK2 expression is more critical than ERK1 expression for survival [31]. In this study, we evidenced that ERK2, rather than ERK1, involved in LPS-induced G-CSF expression in macrophages, further illustrating the different functions of these two kinases.
ERK2 has been identified as a transcription regulator but the consequence of ERK2 signaling might be determined by the interaction of ERK2 and other proteins in either the nucleus or cytoplasm, such as nuclear factors, scaffold proteins, and anchoring proteins. For example, interaction between activated ERK2 and the IκB-NF-κB complex in the cytoplasm has been implicated in protecting myeloid leukemia cells against apoptogenic stimuli [32]. Although NF-κB is known to be a major transcriptional regulator of the expressions of inflammatory genes, NF-κB alone is not sufficient to drive transcription of G-CSF [4]. We have previously shown that, following LPS stimulation, Oct-2 binds to the promoters of the iNOS and G-CSF genes and this effect is reduced by inhibitors of PI3K, AKT, or mTOR, thus demonstrating the essential role of Oct-2 in regulating the transcription of these genes [6]. All these results suggest that both Oct-2 and NF-κB are essential for LPS-induced expression of G-CSF. However, as shown in Fig 4 and S6 Fig in this study, U0126 did not affect LPS-induced nuclear levels of NF-κB (p50 and p65) and Oct-2, but reduced the LPS-induced binding of NF-κB p50 and p65 to the G-CSF promoter, but not to the TNF-α promoter. In addition, LPS-induced accessibility of the G-CSF promoter to DNase I was prevented by U0126 or ERK2 knockdown, but not by ERK1 knockdown, and, again, no such effect was seen with the TNF-α promoter region (Figs 5  and 6). These results further suggest that ERK2 activation may specifically alter chromatin conformation in the G-CSF promoter region, thus increasing its accessibility to DNase I digestion and facilitating the binding of transcription factors [33]. Carlson et al. [34] identified a potential ERK2 substrate, ETV3, a transcriptional regulator that is extensively phosphorylated by activated ERK2 and loses its ability to bind to the GGAA-containing E box motif at the promoters of many genes. ETV3 has been suggested to function by cooperating with SBNO2, a transcriptional corepressor, to regulate the anti-inflammatory effects of IL-10 [35]. In our study here, we provide evidences that ERK2 mediates LPS-upregulated G-CSF promoter activity, at least in part, by interacting with C/EBPβ and increasing its ability to activate transcription. However, it remains unclear whether factors similar to ETV3 or other chromatin remodeling complexes involve in LPS-ERK2-regulated G-CSF expression in macrophages.
NF-κB controls many genes involved in inflammation, such as G-CSF, IL-6, TNF-α and iNOS. Although we show that U0126 has no effect on LPS-increased NF-κB-dependent reporter gene expression and nuclear protein levels of p50 and p65 in RAW 264.7 macrophage. LPS-induced NF-κB activation has been reported to be inhibited by PD98059 in murine J774 macrophages [36]. The discrepancy could be due to that different cells or different methods were used. In addition, studies have shown that inhibition of ERK1/2 has no effect on the activation of a stably transfected NF-κB-dependent reporter gene expression [37], or expression of NF-κB down-stream genes, such as IL-6, TNF-α and iNOS [38][39][40]. Furthermore, evidence showing that phosphorylation level of p65 at S468 and S536 is not inhibited by PD98059 [41,42]. These results suggest that ERK2 may not a critical regulator of NF-κB activation, and ERK2 regulates G-CSF expression is not through inhibition of NF-κB activation.
G-CSF is essential for the protective inflammatory response and for maintaining the balance between anti-and pro-inflammatory reactions in the inflammatory condition. Higher levels of G-CSF may cause an excessive inflammatory response and have been associated with morbidity and mortality of acute lung injury or with chronic inflammatory diseases, such as rheumatoid arthritis [9,10]. Knockout of G-CSF protects mice against collagen-induced arthritis, mostly due to prevention of infiltration by activated neutrophils [43]. Thus, inhibition of G-CSF activity is now considered as a new therapeutic strategy for rheumatoid arthritis and other inflammatory diseases.
In this report, we provide evidences for a pivotal role of ERK2 in LPS-induced G-CSF expression in macrophages. We show that ERK2 involves in LPS-induced G-CSF expression at two distinct steps; activation of ERK2 triggers local chromatin remodeling, which may facilitate binding of transcription factors to the G-CSF promoter, and activation of ERK2 cooperates with C/EBPβ to activate G-CSF promoter activity. These effects of ERK2 on G-CSF expression are not shared by ERK1. The precise mechanism by which ERK2 causes chromatin remodeling is not known yet and further studies are needed.  Fig. U0126 inhibits LPS-induced G-CSF promoter but not NF-κB-driven promoter activity. RAW264.7 cells were co-transfected with 1 μg of pG-CSF(−283/+35)-Luc (A) or the pTransNF-κB-Neo plasmid (B) and 0.05 μg of phRLTK. At 24 h post-transfection, the cells were pretreated with DMSO or 10 μM U0126 for 30 min, followed by addition of LPS (100 ng/ ml) for 6 h, then luciferase activities were determined using the Dual-Luciferase reporter assay system, and firefly luciferase activity was normalized to renilla luciferase activity, then the results were expressed relative to those for untreated control cells (C). The values are the mean ± SD for three independent experiments. Ã p < 0.01 compared to the LPS-treated cells in (A) and Ã p < 0.05 compared to the LPS-treated cells in (B).