Rac1-mediated Mitochondrial H2O2 Generation Regulates MMP-9 Gene Expression in Macrophages via Inhibition of SP-1 and AP-1*

Aberrant matrix deposition is a hallmark of pulmonary fibrosis and is characterized by an imbalance between matrix deposition and degradation. We have previously shown that mice harboring a conditional deletion of the GTP-binding protein, Rac1, in macrophages are protected from asbestos-induced pulmonary fibrosis. To investigate the contribution of aberrant matrix degradation, we addressed the role of Rac1 in regulating expression of macrophage-specific MMP-9 (matrix metalloproteinase-9). We found that MMP-9 gene transcription was significantly greater in Rac1 null macrophages. Deletion and mutational analysis of the MMP-9 promoter revealed that both SP-1 and AP-1 are essential for MMP-9 transcription. Overexpression of constitutive active Rac1 (V12) revealed that H2O2 was derived from the mitochondria. Rac1-induced H2O2 generation down-regulated MMP-9 gene transcription, whereas catalase overexpression in WT cells enhanced MMP-9 expression. SP-1 interacted directly with both c-Jun and c-Fos, and H2O2 decreased this binding, suggesting that SP-1 and AP-1 function cooperatively to regulate MMP-9 transcription. Rac1-mediated H2O2 inhibited the ERK MAPK, which was essential for activation of SP-1 and AP-1. ERK activation and MMP-9 expression were recovered by overexpressing catalase or transfecting siRNA for the mitochondrial iron-sulfur protein, Rieske. These observations were recapitulated in vivo. MMP-9 mRNA was higher in alveolar macrophages isolated from Rac1 null mice and wild type mice given catalase. Rac1 regulates MMP-9 transcription via mitochondrial H2O2 generation, providing a potential mechanism by which Rac1 null mice fail to develop pulmonary fibrosis.

Matrix metalloproteinases (MMPs) 2 are a family of zinc-dependent endopeptidases that degrade extracellular matrix (ECM) proteins and are required for ECM remodeling. They play an important role in a wide variety of physiological conditions and disease states, some of which include normal development, inflammation, wound healing and repair, vascular diseases, and cancer growth and metastasis (1)(2)(3). Under normal conditions, the maintenance of the ECM is a dynamic process in which the production of ECM proteins is balanced by MMPinduced protein degradation. In contrast, in states of excessive tissue remodeling when production of ECM proteins exceeds proteolysis, tissue fibrosis can develop due to expansion of the ECM (4).
The regulation of MMP expression and activity occurs at many levels, including gene transcription, post-transcriptional processing, and proenzyme activation (5,6). Additional control occurs by the family of tissue inhibitors of metalloproteinases (TIMPs), which bind in a 1:1 ratio to the active site of MMPs (7). The generation of reactive oxygen species (ROS) has been shown to modulate both the expression and activity of MMPs (8,9). ROS are known to convert pro-MMPs to their active catalytic state by oxidizing the bond between a highly conserved cysteine residue and zinc in the catalytic domain (10). In addition, ROS increase the expression of MMPs via cell signaling pathways, such as MAPKs, that are regulated by redoxsensitive phosphatases (11)(12)(13). ROS oxidize cysteine residues in the catalytic domains of these phosphatases, rendering them inactive and, thereby, potentiate the signal of the respective kinases, resulting in increased MMP expression. In addition to other MAPKs, the ERK MAPK has been shown to modulate MMP expression (12). Inhibition of ROS by flavoenzyme inhibitors or antioxidants has been shown to decrease MMP expression (8,14,15).
The Rho family of GTP-binding proteins is composed of 20 members, one of which is Rac1 (16). These proteins play an important role in host defense. In particular, Rac1 is known to regulate assembly of the actin cytoskeleton, activation of the NADPH oxidase in non-phagocytic cells, the cellular transformation initiated by Ras oncogenes, and the migration, adhesion, and differentiation of cells (17)(18)(19). Its role in mediating ROS generation has been implicated in many disease processes, including fibrosis (17)(18)20). Studies in which Rac1 has been shown to regulate MMP production via increased ROS were performed in non-phagocytic cells, such as fibroblasts, epithelial cells, and adenocarcinoma cells, and reduction of ROS decreased MMP expression and activity (14,15,21,22). However, the molecular mechanism(s) by which Rac1 modulates MMP-9 gene transcription, and the basis of these actions in phagocytic cells has not been demonstrated.
We recently demonstrated that Rac1 activity in alveolar macrophages, via H 2 O 2 production, is required for asbestosinduced pulmonary fibrosis (23). We now investigated the mechanisms for Rac1-mediated ROS signaling focusing on the transcriptional regulation of MMP-9. Our results demonstrate that MMP-9 transcription was redox-sensitive and dramatically increased in Rac1 null cells due to increased MMP-9 promoter activity and mRNA synthesis. The source of Rac1induced H 2 O 2 was from the mitochondria. Constitutive active Rac1 (V12) and Rac1-mediated H 2 O 2 inhibited the ERK MAPK, which was essential for MMP-9 transcription. More importantly, MMP-9 expression was recovered by expressing siRNA for the mitochondrial iron-sulfur protein, Rieske. These in vitro observations were reproduced in vivo because alveolar macrophages obtained from Rac1 null mice exposed to asbestos expressed significantly more MMP-9 mRNA than cells from WT mice. The inhibitory effect of Rac1-induced H 2 O 2 was demonstrated by finding enhanced MMP-9 expression in macrophages obtained from asbestos-exposed WT mice given catalase. Moreover, MMP-9 Ϫ/Ϫ mice developed pulmonary fibrosis to the same degree as WT mice. These data indicate that Rac1-mediated mitochondrial H 2 O 2 generation suppresses MMP-9 gene transcription and provides a mechanism by which Rac1 null mice are protected from developing pulmonary fibrosis.
Mice-WT and Rac1 null C57BL/6 mice were used in these studies, and all protocols were approved by the University of Iowa Institutional Animal Care and Use Committee. Rac1 null mice were generated by selectively disrupting the Rac1 gene in cells of the myeloid lineage, as described previously (24). MMP-9 Ϫ/Ϫ mice were obtained from Jackson Laboratory (Bar Harbor, ME). Mice were administered 100 g of chrysotile asbestos in 50 l of normal saline intratracheally after being anesthetized with 3% isoflurane using a precision Fortec vaporizer (Cyprane, Keighley, UK). 21 days later, mice were euthanized with an overdose of isoflurane, and bronchoalveolar lavage (BAL) was performed. BAL cells were collected by centrifugation of the BAL for 10 min at 220 g and used for the determina-tion of MMP-9 mRNA. Over 90% of the BAL cells were macrophages.
To evaluate the effects of catalase, WT mice were exposed to asbestos as described above. The mice were then administered either catalase (2000 units/mouse) or water intratracheally every day for 20 days. BAL was performed on euthanized mice, and BAL cells were collected.
Cell Culture-WT and Rac1 null macrophages were obtained from WT and Rac1 null mice as described previously (23). Cells were maintained in RPMI 1640 medium containing 10% newborn calf serum, 0.5 mM ␤-mercaptoethanol, and penicillin/streptomycin. THP-1 and HFL-1 cells were obtained from the American Type Culture Collection and maintained in RPMI 1640 or Dulbecco's modified Eagle's medium, respectively, supplemented with 10% newborn calf serum and penicillin/streptomycin. All experimental conditions were performed in phenol red-free RPMI 1640 medium supplemented with 0.5% serum and penicillin/streptomycin. Where indicated, chrysotile asbestos was used at a final concentration of 10 g/cm 2 .
H 2 O 2 Generation-Release of H 2 O 2 from WT and Rac1 null macrophages was determined as described previously (25). The assay takes advantage of H 2 O 2 -mediated oxidation of HP to Complex I, which, in turn, oxidizes pHPA to a stable fluorescent (pHPA) 2 dimer. Mitochondria were isolated by lysing the cells in a mitochondrial buffer containing 10 mM Tris, pH 7.8, 320 mM sucrose, 0.2 mM EDTA, and protease inhibitors. Lysates were homogenized using a Kontes pellet pestle motor and centrifuged at 1,000 g for 10 min at 4°C. The supernatant was removed and kept at 4°C, and the pellet was lysed, homogenized, and centrifuged again. The two supernatants were pooled and centrifuged at 12,000 ϫ g for 15 min at 4°C. After the supernatant was discarded, the pellet was resuspended in mitochondrial buffer without sucrose. For membrane isolation, cells were lysed in a buffer containing 50 mM Tris⅐HCl, pH 8.0, 10 mM EDTA, and protease inhibitors. Lysates were homogenized using a Kontes pellet pestle motor and centrifuged at 3,000 rpm for 3 min at 4°C. Supernatants were centrifuged at 100,000 ϫ g for 1 h. After removal of the supernatant, the membrane pellet was resuspended in lysis buffer and incubated on ice for 30 min. Cells, mitochondrial, or membrane fractions were incubated in phenol red-, serum-free HBSS containing 1.6 mM pHPA, 95 g/ml HP, 6.5 mM glucose, 1 mM HEPES, 6 mM sodium bicarbonate. The formation of the (pHPA) 2 dimer in the medium was monitored by measuring fluorescence at excitation and emission wavelengths of 323 and 400 nm, respectively. Adenoviral Vectors-Replication-deficient recombinant adenovirus type 5 with the E1 region replaced with DNA containing the cytomegalovirus (CMV) promoter region alone (Ad.CMV), catalase cDNA (26), V12-Rac1 cDNA, or N17-Rac1 cDNA (27) downstream of the CMV promoter was obtained from the Gene Transfer Vector Core at the University of Iowa (Iowa City, IA). Cells were infected for 48 h with the vectors at a multiplicity of infection of 500 in RPMI 1640 medium containing 0.5% newborn calf serum.
Quantitative Real-time PCR-Total RNA from BAL cells and cells grown in culture was isolated by TRIzol and, following DNase treatment, subjected to reverse transcription using the reverse transcriptase kit Iscript (Bio-Rad). MMP-9, HPRT, and ␤-actin mRNA transcripts were determined by quantitative real time PCR using SYBR Green (Bio-Rad) and the respective primers on an IQ5 real-time PCR machine (Bio-Rad). The following primer sets were used: MMP-9 (5Ј-CCA CAT CTC CCT CCA GAA A-3Ј and 5Ј-CAC TTG GTG GTT TGC TAC GA-3Ј), MMP-12 (5Ј-AGA GCA GTG CCC CAG AGG TCA-3Ј and 5Ј-GGG GGT TTC ACT GGG GCT CCA TA-3Ј), HPRT (5Ј-CCT CAT GGA CTG ATT ATG GAC-3Ј and 5Ј-CAG ATT CAA CTT GCG CTC ATC-3Ј), and ␤-actin (5Ј-AGA GGG AAA TCG TGC GTG AC-3Ј and 5Ј-CAA TAG TGA TGA TGA CCT GGC CGT-3Ј). Data were calculated by the ⌬⌬Ct method. MMP-9 mRNA in cultured cells was normalized to HPRT mRNA, which amplified at approximately the same Ct as MMP-9. MMP-9 mRNA in BAL cells was normalized to ␤-actin. Results are expressed as arbitrary units of MMP-9 mRNA relative to the respective housekeeping gene.
Plasmids and Luciferase Assays-The 5Ј-flanking sequence of the human MMP-9 promoter (Ϫ1284/ϩ21) inserted between KpnI and HindIII restriction sites in the pGL3 basic expression vector was a generous gift from Dr. Jianming Xu (Baylor College of Medicine, Houston, TX) (28). Deletion constructs were generated by PCR using the full-length MMP-9 promoter (Ϫ1284/ϩ21) as a template with forward and reverse primers containing KpnI and HindIII restriction sites, respectively. Amplified products were resolved on agarose gels, purified, and directionally cloned into the KpnI and HindIII restriction sites of pGL3 basic vector.
SP-1-dependent gene expression was evaluated using a luciferase reporter plasmid (pSP-1-luc) driven by thee tandem copies of the SP-1 enhancer (GGGGCGGGGCG) linked to a herpes simplex virus thymidine kinase promoter. AP-1-dependent gene expression was evaluated by the previously described plasmid (29). Using PCR techniques, murine SP-1 cDNA (NM_ 013672) was directionally cloned into the pcDNA3.1D/V5-His-TOPO vector (Invitrogen). The pCMV-HA-ERK2 (K/A) plasmid has been described and was a generous gift from Dr. Roger Davis (University of Massachusetts) (30). The correct reading frame and sequence of all plasmids used in the study were verified by fluorescent automated DNA sequencing performed by the University of Iowa DNA Facility.
Cells were transfected using Effectene Transfection Reagent (Qiagen, Valencia, CA) according to the manufacturer's directions. To correct for transfection efficiency, cells were co-transfected with phL-TK vector encoding Renilla luciferase (Promega). 4 h after transfection, medium was replaced with fresh serum-containing medium, and cells were allowed to recover for 24 -48 h. Firefly and Renilla luciferase activities were determined in cell lysates using the Dual Luciferase reporter assay kit (Promega), according to the manufacturer's instructions, and are expressed in real light units.
Small Interfering RNA-Rac1 null cells and THP-1 cells were transfected with 100 nM control or ERK siRNA duplex (Dharmacon Research, Lafayette, CO) or human Rieske siRNA duplex (IDT, Iowa City, IA) utilizing DharmaFect 4 or 2 (Dharmacon Research), respectively, according to the manufacturer's instructions. 6 h after transfection, media were replaced, and cells were allowed to recover for 48 -72 h.
Chromatin Immunoprecipitation (ChIP) Assay-The ChIP assay was conducted using the SimpleChIP TM enzymatic chromatin IP kit (Cell Signaling) according to the manufacturer's instructions. Briefly, Rac1 null cells (ϳ40 ϫ 10 6 cells) were grown to confluence on 150-mm tissue culture plates and then treated with either vehicle or 25 M H 2 O 2 for 2 h. Cells were fixed using 1% formaldehyde for 15 min at room temperature, followed by termination of fixation using excess glycine. Cells were harvested, the cell nuclei were isolated, and the resulting nuclear pellets were treated with micrococcal nuclease (150 units for 20 min at 37°C) to digest DNA to fragments of about 0.3 kb. The resulting cross-linked chromatin preparations were used for input controls (2% of total) or for immunoprecipitation using 2 g of anti-SP-1 or anti-c-Jun antibodies (Santa Cruz Biotechnology, Inc.) or normal rabbit IgG antibody (Cell Signaling) as a negative control. Immunocomplexes were eluted, and the chromatin was subjected to reversal of cross-links followed by DNA purification as described in the protocol. PCR was performed using purified DNA and the following primers: 5Ј-GCT CCC ACA TGT GTG TGT C-3Ј and 5Ј-CCT AGC TCC AGC AGG CTG for the murine MMP-9 SP-1 binding site (76-bp product); 5Ј-GTA GTG TAA ACA CAC ACA CAC A-3Ј and 5Ј-AGT AAA ACG GAA TCA GTG ACC C-3Ј for the distal AP-1 site (116-bp product); and 5Ј-CCC CAC ACT GTA GGT TCT ATC C-3Ј and 5Ј-ATC CTG CCT CAA AGA GCC T-3Ј for the proximal AP-1 site (101-bp product). RPL30 primers (Cell Signaling) were used for used for PCR detection of the murine ribosomal protein gene locus (159-bp product). PCR products were resolved on a 2% agarose gel.
Zymogram Analysis-WT and Rac1 null cells were incubated for 24 h in RPMI containing 0.5% newborn calf serum in the absence of ␤-mercaptoethanol and phenol red. Conditioned media equivalent to 30 g of protein were separated on a 5% polyacrylamide gel containing 1 mg/ml gelatin under denaturing and non-reducing conditions. Gels were washed twice for 10 min each in 2.5% Triton X-100 and twice for 30 min each in reaction buffer (50 mM Tris, 5 mM calcium chloride, and 2 M zinc chloride, pH 8) containing 2.7% Triton X-100. After removal of the Triton X-100 with washing, the gels were incubated for 72 h at 37°C in reaction buffer. They were then stained with Coomassie Blue until clear zones representing gelatinolytic activity were observed. The bands co-migrated with purified MMP-9 protein.
Hydroxyproline Determination-BAL fluid was digested for 24 h at 112°C with 6 N hydrochloric acid. Hydroxyproline concentration was determined as described previously (23).
Immunoblot Analysis-Extraction of nuclear proteins and whole cell lysates were performed as described previously (31). Nuclear extracts, whole cell lysates, and conditioned media were separated by SDS-PAGE. Immunoblot analyses were performed with the designated antibodies, followed by the appropriate secondary antibody cross-linked to horseradish peroxidase.
Purification of SP-1-His-tagged Protein-Rac1 null cells were transfected with empty pcDNA3.1 vector or pcDNA3.1-SP-1-V5-His. After 24 h, cells were exposed to vehicle or H 2 O 2 (25 M) for 2 h. After treatments, cells were harvested in Buffer B (phosphate-buffered saline; 0.5 M NaCl, 1% Triton X-100, protease, and phosphatase inhibitors), the lysates were briefly sonicated on ice, and cellular debris was pelleted at 12,000 ϫ g for 10 min at 4°C. Talon metal (cobalt) affinity resin (Clontech) was added to each lysate, and samples were rotated overnight at 4°C. The SP-1-His proteins were eluted by adding protein sample buffer and heating at 95°C for 5 min.
Statistical Analyses-Statistical comparisons were performed using either an unpaired, one-tailed t test or one-way analysis of variance followed by Tukey's t test. Values in the figures are expressed as means with S.E., and p Ͻ 0.05 was considered to be significant.

MMP-9 Expression Is Increased in Rac1
Null Cells-Our previous observations demonstrated that deletion of Rac1 from macrophages protects mice from asbestos-induced pulmo-nary fibrosis (23). To determine whether increased matrix degradation in Rac1 null mice accounted, in part, for this effect, we evaluated MMP-9 expression in WT and Rac1 null macrophages. WT and Rac1 null cells were cultured for 24 h. MMP-9 protein and activity in conditioned medium were determined by immunoblotting and zymogram analysis, respectively. MMP-9 protein and activity were dramatically increased in conditioned medium collected from Rac1 null macrophages compared with WT cells (Fig. 1A). This difference was also observed at the mRNA level (Fig.  1B). To determine whether these changes were regulated at the transcriptional level, WT and Rac1 null cells were transiently transfected with a luciferase vector driven by the MMP-9 promoter (Ϫ1284/ϩ21 or MMP-9 wt ), and luciferase activity was determined 48 h after transfection. Consistent with the dramatically increased protein, activity, and mRNA levels in Rac1 null cells, MMP-9 promoter activity was 7-fold greater in Rac1 null cells compared with WT cells (Fig. 1C).
Rac1-mediated Inhibition Was Specific for MMP-9-To address the specificity of the effects of Rac1 on MMP-9, we measured relative mRNA expression of MMP-2 and MMP-12, which are other macrophage-specific MMPs. MMP-2 was not detected in WT or Rac1 null cells (data not shown), and there was no difference in mRNA expression of MMP-12 ( Fig. 2A).
The negative regulatory role of Rac1 was confirmed by overexpressing constitutively active Rac1 (V12) in Rac1 null cells. Cells expressing V12-Rac1 had a significantly lower abundance of MMP-9 mRNA than cells infected with the empty vector (Fig. 2B). Likewise, overexpression of a dominant negative Rac1 (N17) in WT macrophages increased MMP-9 mRNA levels in these cells (Fig. 2C). In aggregate, these results demonstrate that Rac1-mediated inhibition of macrophage-specific MMPs is exclusive for MMP-9, and Rac1 suppresses MMP-9 expression in macrophages at the level of transcription.
SP-1 and AP-1 Are Necessary for MMP-9 Gene Expression-Because MMP-9 appeared to be modulated at the level of transcription, we next addressed which regulatory elements in the MMP-9 promoter accounted for greater promoter activity in Rac1 null cells. MMP-9 gene expression is controlled by multiple transcription factors, including NF-B and AP-1, which are known to be involved in monocyte and macrophage host defense (29,31). Two AP-1 binding sites are present in the MMP-9 promoter and are separated by 447 bp. Directly upstream of the distal AP-1 site is a GC-rich region that contains an SP-1 binding site. Therefore, we investigated the role of these factors in the regulation of MMP-9 promoter activity in Rac1 null cells. Truncated constructs with the NF-B (Ϫ584/ ϩ21), SP-1 (Ϫ544/ϩ21), and AP-1 (Ϫ102/ϩ21 and Ϫ69/ϩ21)

Rac1 Inhibits MMP-9
AUGUST 6, 2010 • VOLUME 285 • NUMBER 32 consensus sites deleted from the MMP-9 promoter were cloned into the pGL3 basic expression vector (Fig. 3A). WT and Rac1 null cells were transiently transfected with each vector, and luciferase activity was determined. MMP-9 promoter activity was significantly greater in Rac1 null cells compared with WT cells (Fig. 3B). All of the truncation mutants lacked a significant region of the promoter spanning about 616 bp upstream of the NF-B site. We determined whether this region (Ϫ1285/Ϫ668) contributed to MMP-9 transcriptional regulatory activity. Promoter activity of Rac1 null cells transfected with the construct (Ϫ668/ϩ21) was not significantly different from that observed in cells transfected with the full-length plasmid (data not shown). Deletion of the NF-B binding site from the full-length MMP-9 promoter did not alter the activity of the promoter in Rac1 null cells. On the other hand, successive deletions of the SP-1 and the distal and proximal AP-1 binding sites progressively attenuated the promoter activity in Rac1 null cells (Fig.  3B). In contrast, none of the truncated mutants affected luciferase activity in WT cells. Deletion of all four regulatory elements, however, decreased MMP-9 promoter activity in Rac1 null cells to a level similar to that observed in WT cells. These results strongly suggest that the three cis-acting elements that bind SP-1 and AP-1 are required for driving MMP-9 gene expression in Rac1 null cells.
Based on the above results coupled with the fact that the truncated NF-B mutant had no effect on MMP-9 promoter activity, we further evaluated the relative contributions of SP-1 and AP-1. MMP-9 luciferase constructs with mutations in one or all of the three binding sites were generated (Fig. 3C). Results obtained with Rac1 null cells transfected with these mutants were compared with luciferase activity observed in WT cells transfected with MMP-9 wt . Single mutations in SP-1 (SP-1 mut ), distal AP-1 (AP-1 mut1 ), or proximal AP-1 (AP-1 mut2 ) binding sites significantly attenuated MMP-9 promoter activity in Rac1 null cells (Fig. 3D). The decreases in luciferase activities observed with SP-1 mut and AP-1 mut1 were similar, suggesting that both of these regions were of equal importance in regulating the MMP-9 promoter. Cells expressing the AP-1 mut2 resulted in a relatively greater decrease in MMP-9 promoter activity. None of these single mutations were sufficient to completely inhibit MMP-9 promoter activity in Rac1 null cells to the level observed in WT cells. In contrast, dual mutations of SP-1 and either the distal or proximal AP-1 binding sites, as well as triple mutations of all thee regulatory regions, were found to decrease the activity of the MMP-9 promoter in Rac1 null cells to WT levels. Taken together, these results suggest that Rac1 inhibits MMP-9 expression by regulating SP-1 and AP-1 transcriptional activity. Furthermore, none of the three cis-acting elements alone was sufficient to regulate MMP-9 transcription in macrophages.
SP-1 and AP-1 DNA Binding and Transcriptional Activity Are Increased in Rac1 Null Cells-To ascertain the role of SP-1 and AP-1 in regulating MMP-9 promoter activity, we determined whether there were differences in SP-1 and AP-1 DNA binding in WT and Rac1 null cells. Nuclear protein from WT and Rac1 null cells were incubated with labeled consensus oligonucleotides of SP-1 and AP-1 and separated on a polyacrylamide gel. There was significantly greater DNA binding of SP-1 and AP-1 in Rac1 null cells compared with WT cells (Fig. 4A). In fact, WT cells had essentially no SP-1 or AP-1 DNA binding, but NF-B DNA binding was equally present in both cells types (data not shown).
Because DNA binding does not necessarily correlate with transcriptional activity, we next determined the role of Rac1 in regulating SP-1-and AP-1-dependent transcription using a luciferase reporter plasmid. WT and Rac1 null cells were transiently transfected with either pSP-1-luc or pAP-1-luc. After 48 h, cells were harvested, and luciferase activity was measured. Luciferase activity in Rac1 null macrophages was at least 3-fold higher than in WT macrophages in both SP-1-driven (Fig. 4B) and AP-1-driven luciferase expression (Fig. 4C).
Total and phosphorylated SP-1 has been observed to regulate SP-1-mediated transcriptional activation (32)(33)(34)(35). Based on the observation that the increase in MMP-9 promoter activity in Rac1 null cells was dependent, in part, on SP-1 and there was greater DNA binding and transcriptional activity of SP-1 in Rac1 null cells, we next examined SP-1 nuclear expression and

Rac1 Inhibits MMP-9
its level of phosphorylation at Thr 453 , which has been associated with transcriptional activity (33)(34)(35). Rac1 null cells had significantly greater amounts of phospho-SP-1 (Thr 453 ) and total SP-1 in the nucleus (Fig. 4D). Similar levels of SP-1 were present in the cytoplasm of both cells (data not shown).
We also determined the relative expression of c-Fos and c-Jun, which form the AP-1 heterodimer in macrophages (36). We first determined levels of the phosphorylated form of c-Jun because it is necessary for transcriptional activity of AP-1. Rac1 null cells had significantly greater phospho-c-Jun compared with WT cells (Fig. 4E). Because induction of transcriptional activity also requires the translocation of c-Fos to the nucleus, we determined the levels of c-Fos expression in these cells. Rac1 null cells had significantly more nuclear c-Fos than WT cells (Fig. 4E). Taken together, these data demonstrate that Rac1 negatively regulates SP-1 and AP-1 activation and suggest that, in macrophages, optimal MMP-9 expression is modulated by SP-1 and AP-1.
H 2 O 2 Inhibits MMP-9 Expression-We have previously demonstrated that asbestos-induced pulmonary fibrosis in WT mice is accompanied by increased ROS generation in alveolar macrophages and that intratracheal administration of catalase significantly attenuates pulmonary fibrosis in these mice (23). Because Rac1 has been shown to increase H 2 O 2 generation (21, 37), we evaluated the effect of H 2 O 2 on MMP-9 expression. As expected, the generation of H 2 O 2 in WT cells spontaneously increased in a timedependent manner (Fig. 5A). In contrast, H 2 O 2 was barely detectable in Rac1 null cells over a prolonged period of time. This difference was associated with a higher rate of H 2 O 2 production in WT cells compared with Rac1 null cells (Fig. 5B).
We next determined the source of Rac1-mediated H 2 O 2 generation. THP-1 cells were infected with an adenoviral vector expressing either an empty construct or the constitutively active Rac1 (V12). 48 h later, mitochondrial and membrane fractions were isolated, and H 2 O 2 generation was determined. We found that overexpression of Rac1 had only a slight effect on H 2 O 2 generation from the membrane fraction (Fig. 5C), whereas the mitochondrial H 2 O 2 generation was increased greater than 6.5fold in cells expressing constitutive active Rac1 (Fig. 5D). Taken together, these data demonstrate that Rac1 increases H 2 O 2 generation and that the primary source is from the mitochondria.
To determine if H 2 O 2 generation contributed to MMP-9 transcription, WT cells were infected with adenoviral vectors containing either an empty construct or a catalase construct. After 48 h, total RNA was isolated. WT cells overexpressing catalase had significantly greater MMP-9 mRNA than cells expressing the empty vector (Fig. 6A). To confirm that H 2 O 2 inhibited MMP-9 expression transcriptionally, WT cells were transiently co-transfected with the MMP-9 wt luciferase vector and either an empty vector or the catalase expression vector. Luciferase activity was determined 48 h later. Similar to the mRNA levels, MMP-9 promoter activity was about 5-fold greater in WT cells overexpressing catalase (Fig. 6B).
In order to demonstrate the role of mitochondrial H 2 O 2 in MMP-9 expression, THP-1 cells were transfected with siRNA for the mitochondrial iron-sulfur protein, Rieske, and MMP-9 expression was determined in conditioned medium. We found that MMP-9 secretion was dramatically increased in cells expressing the Rieske siRNA (Fig. 6C).
To further confirm that H 2 O 2 inhibits MMP-9 expression, we examined the effects of H 2 O 2 addition on MMP-9 expression. Rac1 null cells were transiently transfected with the MMP-9 wt . After 48 h, cells were exposed to H 2 O 2 over a time course up to 2 h. Cells stimulated with H 2 O 2 had significantly less luciferase activity (Fig. 6D). The decrease in MMP-9 promoter activity by H 2 O 2 was similar at all times. We also examined the effect of H 2 O 2 on MMP-9 in vivo utilizing a ChIP assay. Rac1 null cells were cultured with or without exposure to H 2 O 2 for 2 h. SP-1 and c-Jun were immunoprecipitated from the DNAprotein complex, and PCR amplification was performed using primers to detect the SP-1, the distal AP-1, and the proximal AP-1 sites. We found that H 2 O 2 reduced SP-1 binding and abolished AP-1 DNA binding to the MMP-9 promoter (Fig. 6E). Taken together, these data demonstrate that Rac1induced H 2 O 2 generation inhibits MMP-9 expression at the level of transcription and that overexpression of catalase or knockdown of the mitochondrial iron-sulfur protein, Rieske, increases MMP-9 expression.
ERK MAPK Is Increased in Rac1 Null Cells and Is Negatively Regulated by H 2 O 2 -The signaling pathway linking H 2 O 2 generation to inhibition of MMP-9 was evaluated by investigating the role of the ERK MAPK because it has been shown to be involved in regulating MMP expression (12). We found that ERK was constitutively activated in Rac1 null cells and was barely detectable in WT cells (Fig. 7A). To evaluate ERK in a different manner, THP-1 cells were infected with adenoviral vectors containing an empty vector, a constitutive active Rac1 (V12), or a dominant negative Rac1 (N17) construct. After 48 h, cells were exposed to asbestos, and an immunoblot analysis was performed for ERK. Asbestos did not activate ERK in cells expressing the empty vector (Fig. 7B). ERK was suppressed below control levels in the cells expressing V12, whereas ERK activation was significantly increased in cells expressing the dominant negative (N17) Rac1, and this activation was augmented by asbestos exposure (Fig. 7B).
Because H 2 O 2 generation is significantly greater in WT macrophages, we questioned whether H 2 O 2 had a role in inhibiting ERK activation. THP-1 cells were infected with adenoviral vectors containing either an empty construct or a catalase construct, and after 48 h, cells were exposed to asbestos for 2 h. Asbestos decreased phospho-ERK in cells expressing the empty vector, and this was increased significantly in cells expressing the catalase vector alone. ERK activation was further enhanced by asbestos exposure in cells overexpressing catalase (Fig. 7C). To address the source of H 2 O 2 , we transfected THP-1 cells with either scrambled or Rieske siRNA and determined ERK activation. Knockdown of the mitochondrial iron-sulfur protein, Rieske, resulted in a dramatic increase in ERK activation with no change in the expression of total ERK (Fig. 7D). In aggregate, these data demonstrate that ERK is negatively regulated by Rac1 and mitochondrial H 2 O 2 .
ERK Modulates MMP-9 Transcription by Regulating SP-1 and AP-1 Nuclear Localization and Activation-To determine the effect of ERK on MMP-9 expression, Rac1 null cells were transfected with a scrambled or ERK siRNA. Nuclear extracts were isolated after 48 h and separated by SDS-PAGE. Knock- down of ERK decreased SP-1 nuclear levels and SP-1 phosphorylation (Fig. 8A). Likewise, c-Jun phosphorylation and c-Fos nuclear translocation were dramatically reduced in cells transfected with the ERK siRNA (Fig. 8B).
To confirm that these changes in SP-1 and AP-1 proteins regulated MMP-9 transcription, Rac1 null cells were transfected with the MMP-9 wt luciferase vector and either an empty or dominant negative ERK expression vector. After 48 h, cells were harvested, and luciferase activity was measured. MMP-9 promoter activity in Rac1 null macrophages was significantly inhibited in cells expressing the dominant negative ERK (Fig.  8C). In aggregate, these data demonstrate that ERK modulates MMP-9 transcription by regulating nuclear localization and phosphorylation of SP-1 and AP-1.
The above data coupled with our promoter mutation analysis suggest that both SP-1 and AP-1 are necessary for MMP-9 transcription in macrophages. Due to the fact that both SP-1 and AP-1 are required, we determined whether there was functional interaction between these transcription factors that regulated optimal MMP-9 expression. Rac1 null macrophages were transfected with either an empty vector or SP-1-V5-His. After 24 h, cells were stimulated with H 2 O 2 for 2 h. SP-1 expression was significantly enhanced in cells transfected with the SP-1-V5-His vector (Fig. 8D). Whole cell lysates were then subjected to His pull-down. Immunoblot analysis demonstrated that both phospho-c-Jun and c-Fos directly interacted with SP-1-V5-His. In contrast, stimulation with H 2 O 2 significantly inhibited this interaction (Fig. 8D).
Taken together, these data are the first to demonstrate that MMP-9 transcription is regulated by Rac1mediated H 2 O 2 generation by modulating the activation and direct interaction of SP-1 and AP-1 transcription factors.
Absence of Rac1 Increases MMP-9 Expression in Vivo-To address whether the above in vitro observations had biological significance, we determined MMP-9 gene expression in macrophages isolated from the lungs of WT and Rac1 null mice exposed to asbestos. Asbestos induces the generation of ROS, especially H 2 O 2 , in alveolar macrophages and, thus, recapitulates the in vitro conditions (23). We first confirmed that the alveolar macrophages from Rac1 null mice did not express Rac1 by performing an immunoblot analysis of whole cell lysates of BAL cells obtained from mice exposed to chrysotile asbestos. More importantly, alveolar macrophages obtained from asbestos-exposed Rac1 null mice had an active ERK MAPK in contrast to WT mice (Fig. 9A). To determine the relative expression of MMP-9 in mice after asbestos exposure, total RNA was isolated from alveolar macrophages, and MMP-9 mRNA was measured by quantitative real-time PCR. The in vivo data corroborated the in vitro data by demonstrating that MMP-9 mRNA was significantly greater in cells isolated from Rac1 null mice compared with WT mice (Fig. 9B).
In order to confirm that these differences were due to H 2 O 2 generation as our in vitro data suggest, we exposed WT mice to asbestos followed by daily administrations of either catalase or vehicle for 20 days. MMP-9 mRNA in macrophages obtained from WT mice administered catalase was more than 7-fold greater when compared with mice given vehicle alone, which also agrees with our in vitro results (Fig. 9C).
To verify the importance of MMP-9 in regulating the development of fibrosis, WT and MMP-9 Ϫ/Ϫ mice were exposed to asbestos. After 21 days, the mice were euthanized, and BAL fluid was obtained. WT and MMP-9 Ϫ/Ϫ mice developed a similar extent of fibrosis as determined by measuring hydroxyproline concentrations in the BAL fluid (Fig. 9D). We also performed Masson's trichome staining on lung sections obtained from WT, MMP-9 Ϫ/Ϫ , and Rac1 null mice 21 days after asbestos exposure. Both WT and MMP-9 Ϫ/Ϫ mice had extensive deposition of collagen in peribronchial and parenchymal lung sections, whereas Rac1 null mice has no collagen deposition after asbestos exposure (Fig. 9E).
To provide a direct link between regulation of MMP-9 expression by Rac1 and fibrosis development, human lung fibroblasts (HFL-1) were cultured for 24 h in conditioned

Rac1 Inhibits MMP-9
AUGUST 6, 2010 • VOLUME 285 • NUMBER 32 medium obtained from WT and Rac1 null macrophages in the presence of asbestos. MMP-9 activity in conditioned medium was determined by zymography. As shown above, MMP-9 activity was dramatically increased in conditioned medium collected from Rac1 null macrophages compared with WT cells (Fig. 9F). To determine the effect of MMP-9 on collagen deposition, we measured the amount of procollagen and collagen I secreted by HFL-1 cells incubated with WT and Rac1 null macrophage conditioned medium. Fibroblasts exposed to Rac1 null conditioned medium secreted significantly less procollagen and collagen I compared with cells incubated with WT medium (Fig. 9F). In aggregate, these results support the hypothesis that MMP-9 secreted by macrophages modulates collagen deposition by fibroblasts and provide direct evidence that lack of MMP-9 is sufficient for fibrosis development. Furthermore, these results indicate that in macrophages, MMP-9 expression is regulated by Rac1-mediated mitochondrial H 2 O 2 generation and that MMP-9 plays an important role in attenuating the development of pulmonary fibrosis.

DISCUSSION
There are limited data regarding the molecular mechanisms regulating MMP-9 expression and activity in pulmonary fibrosis. MMP-9 is a critical molecular target that regulates the fibrotic phenotype. In this study, we demonstrate for the first time that Rac1-mediated mitochondrial H 2 O 2 generation inhibits MMP-9 transcriptional activity via inhibiting SP-1 and AP-1 DNA binding, transcriptional activity, and direct interaction in macrophages. Evidence in support of this pathway includes the following: (i) Rac1 null cells exhibit constitutively high MMP-9 mRNA expression; (ii) promoter deletional and mutational analysis uncovered a regulatory region (Ϫ561/Ϫ77) that confers Rac1 inhibitory response elements; (iii) Rac1-induced H 2 O 2 generation was derived from the mitochondria; (iv) MMP-9 transcriptional activity was suppressed by Rac1induced H 2 O 2 ; (v) Rac1 and H 2 O 2 inhibited ERK MAPK, which is essential for SP-1 and AP-1 transcriptional activation; (vi) collagen deposition by human lung fibroblasts is abolished when cultured in conditioned medium from Rac1 null cells; and (vii) targeted deletion of Rac1, knockdown of the mitochondrial iron-sulfur complex III subunit, Rieske, and modulation of H 2 O 2 in mice differentially regulates expression of MMP-9. In aggregate, these observations provide a new mechanistic model linking Rac1 and mitochondrion-generated H 2 O 2 to the pathobiology of asbestosis-induced pulmonary fibrosis.
Rac1 is a member of the family of Rho GTPases, and it regulates several cellular functions, such as actin polymerization, ChIP assays were performed as described under "Experimental Procedures." Cross-linked chromatin from Rac1 null cells was subjected to immunoprecipitation using anti-SP-1 or anti-c-Jun or normal rabbit IgG. Input DNA (2% of total) and immunoprecipitated DNA were then analyzed by PCR using primers to the SP-1 binding site (IP1; 76-bp product), the distal AP-1 binding site (IP2; 116 bp), the proximal AP-1 binding site (IP3; 101 bp), or ribosomal protein L30 (IP4; 159 bp) as a negative control. PCR products were separated on a 2% agarose gel. FIGURE 7. Rac1 and H 2 O 2 modulate ERK activation. A, whole cell lysates of WT and Rac1 null cells were isolated. B, THP-1 cells were infected with adenovirus vector containing an empty construct, a constitutive active Rac1 (V12), or dominant negative Rac1 (N17). After 48 h, cells were exposed to asbestos, and whole cell lysates were isolated. C, THP-1 cells were infected with adenovirus containing either an empty or a catalase construct. After 48 h, the cells were exposed to asbestos, and whole cell lysates were isolated. D, THP-1 cells were transfected with 100 nM scrambled or Rieske siRNA, and 48 h later, whole cell lysates were isolated. Immunoblot analyses were performed for phospho-ERK (p-ERK) and ERK for activation and equal loading, respectively. the assembly of NADPH oxidase in non-phagocytic cells, cell adhesion, and cell differentiation (16). Compared with the other isoforms of Rac, Rac1 is ubiquitously expressed and is abundant in macrophages (16). Rac1 activation increases the generation of H 2 O 2 in cells (21,38). Pulmonary fibrosis is characterized by aberrant ECM remodeling, and this process can be influenced by H 2 O 2 . Previous studies have linked Rac1 to MMP activity through the activation of ROS. In fibroblasts, Rac1 increases ROS via activation of the NADPH oxidase and subsequent NF-B-dependent genes, including collagenase-1 (21). Here, we found that WT and Rac1 null cells had similar levels of NF-B DNA binding and that NF-B did not regulate MMP-9 expression. Rac1 has also been shown to up-regulate TIMP-1 via ROS-dependent activation of AP-1 (39). It is not clear from this study if MMPs were inhibited by TIMP-1. We found that there was no significant difference in TIMP gene expression between WT and Rac1 null cells (data not shown); thus, we focused on the transcriptional regulation of MMP-9. Our novel results unequivocally demonstrate that Rac1 plays a crucial role in negatively regulating the expression of the macrophage-specific matrix-degrading enzyme, MMP-9. This regulation is specific for MMP-9 and not other macrophage-specific MMPs.
In the current study, Rac1 regulated MMP-9 expression via H 2 O 2 production. In non-phagocytic cells, the primary supply of Rac1-mediated H 2 O 2 is the NADPH oxidase (21,37,40). In contrast, Rac2 regulates NADPH oxidase activity in macrophages (41)(42)(43). Another important source of H 2 O 2 is the mitochondria. Rac1 modulation of mitochondrial ROS generation has only been described in fibroblasts via engagement of integrins (44,45). Our study is the first to demonstrate that Rac1 induces H 2 O 2 generation from the mitochondria in macrophages. Moreover, our data using the Rieske siRNA suggest that complex III is important for H 2 O 2 generation. These data are novel in that Rac1-mediated H 2 O 2 abrogates MMP-9 transcription.
H 2 O 2 is known to cause extracellular post-translational activation of MMPs via oxidative cleavage of a conserved cysteine residue (10). In addition, H 2 O 2 also increases MMP-9 expression by activating cell signaling pathways, such as the MAPK pathway (11)(12)(13). Both the ERK and p38 MAPK pathways have been shown to be involved in MMP expression. The mechanism by which ROS, especially H 2 O 2 , increase MAPK activation is primarily by inhibition of redox-sensitive protein tyrosine and dual specificity MAPK phosphatases (46,47). The novel aspect of our study is that we found that ERK is inhibited by H 2 O 2 and is recovered by overexpression of catalase. The relationship of Rac1 to ERK activation is uncertain. One study suggests that Rac2 and RhoA cooperate with Raf to activate ERK2 (48). In contrast, our results demonstrate that Rac1 inhibits ERK activation by increasing the generation of H 2 O 2 . We have previously demonstrated that oxidative inactivation of mitogen-activated protein kinase phosphatases by ROS in macrophages results in p38 MAPK activation and inhibition of ERK (49,50). Because Rac1 is an upstream activator of the p38 MAPK, it is plausible that Rac1-mediated H 2 O 2 negatively regulates MMP-9 expression via this differential MAPK activation.
ROS, including H 2 O 2 , regulate transcription factors, such as AP-1. The proximal AP-1 site is common to all MMPs with the exception of MMP-2 and MMP-11 and has been reported to be essential for basal promoter activity of MMP-1, MMP-3, and MMP-9 (12,51). The proximal AP-1 site has long been thought to play a dominant role in the transcriptional activation of the MMP promoters, particularly in response to stimulation with phorbol esters (51). We have found that the macrophage AP-1 is a heterodimer primarily composed of c-Fos and c-Jun (52), and both c-Fos expression and c-Jun phosphorylation have been linked to MMP gene expression (12,53). The distal AP-1 site has not been investigated in most studies, but we found that both the distal and the proximal AP-1 sites contribute to optimal promoter activity. More importantly, our data demonstrate that although both AP-1 sites contribute to MMP-9 expression, the MMP-9 gene also requires the activity of another cis-acting element, SP-1, for optimal promoter activity. we emphasize the role of in vivo H 2 O 2 production in regulating MMP-9 gene expression in lung macrophages exposed to asbestos. In aggregate, these results define the molecular mechanisms by which Rac1 regulates MMP-9 expression and may provide important clues to prevent the development of pulmonary fibrosis.