Journal of Leukocyte Biology Myeloid cells, immune suppression, tumor immunology
HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
QUICK SEARCH:   [advanced]


     

Originally published online as doi:10.1189/jlb.0605302 on January 24, 2006

Published online before print January 24, 2006
This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Right arrow All Versions of this Article:
jlb.0605302v1
79/4/706    most recent
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow reprints & permissions
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Rahat, M. A.
Right arrow Articles by Lahat, N.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Rahat, M. A.
Right arrow Articles by Lahat, N.
(Journal of Leukocyte Biology. 2006;79:706-718.)
© 2006 by Society for Leukocyte Biology

Hypoxia reduces the output of matrix metalloproteinase-9 (MMP-9) in monocytes by inhibiting its secretion and elevating membranal association

Michal A. Rahat*,1, Barak Marom*,{dagger}, Haim Bitterman{dagger}, Lea Weiss-Cerem{dagger}, Amalia Kinarty* and Nitza Lahat*

* Immunology Research Unit and
{dagger} Ischemia-Shock Research Laboratory, Carmel Medical Center, Rappaport Family Institute for Research in the Medical Sciences, and Bruce Rappaport Faculty of Medicine, Technion, Haifa, Israel

1Correspondence: Immunology Research Unit, Carmel Medical Center, 7 Michal St., Haifa, 34362, Israel. E-mail: rahat_miki{at}clalit.org.il


   ABSTRACT
TOP
 ABSTRACT
INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 REFERENCES
 
Cellular hypoxia, characterizing tumors, ischemia, and inflammation induce recruitment of monocytes/macrophages, immobilize them at the hypoxic site, and alter their function. To migrate across the extracellular matrix and as part of their inflammatory functions, monocytes and macrophages secrete proteases, including matrix metalloproteinase-9 (MMP-9), whose expression is induced by proinflammatory cytokines [e.g., tumor necrosis factor {alpha} (TNF-{alpha})]. We show that hypoxia (<0.3% O2 for 48 h) reduced the output of TNF-{alpha}-induced proMMP-9 by threefold (P<0.01) in the U937 monocytic cell line and in primary human monocytes. TNF-{alpha} induced MMP-9 transcription by threefold, but no significant difference was observed in MMP-9 mRNA steady-state between normoxia and hypoxia, which inhibited the trafficking of proMMP-9 via secretory vesicles and increased the intracellular accumulation of proMMP-9 in the cells by 47% and 62% compared with normoxia (P<0.05), as evaluated by zymography of cellular extracts and confocal microscopy, respectively. Secretion of proMMP-9 was reduced by the addition of cytochalazin B or nocodazole, which inhibits the polymerization of actin and tubulin fibers, or by the addition of the Rho kinase inhibitor Y27632, suggesting the involvement of the cytoskeleton and the Rho GTPases in the process of enzyme secretion. Furthermore, attachment of proMMP-9 to the cell membrane increased after hypoxia via its interactions with surface molecules such as CD44. In addition, the reduced migration of monocytes in hypoxia was shown to be mediated, at least partially, by secreted MMP-9. Thus, hypoxia post-translationally reduced the secreted amounts of proMMP-9 by using two mutually nonexclusive mechanisms: mostly, inhibition of cellular trafficking and to a lesser extent, attachment to the membrane.

Key Words: human • macrophages • inflammation


   INTRODUCTION
TOP
 ABSTRACT
 INTRODUCTION
MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 REFERENCES
 
Monocytes and macrophages play a central role in the different stages of the inflammatory response, including recognition and elimination of foreign stimuli, initiation of the adaptive immune response, and regulation of the healing process. This enormous scope of activities requires their migration to inflicted tissues, their presence in these areas, and their fast adaptation to the changing microenvironment.

Tissue hypoperfusion and the disruption of blood supply to injured or inflamed tissues often cause the reduction of local oxygen tension (hypoxia). Tissue hypoxia is the common denominator in a variety of pathological conditions (e.g., malignant tumors, ischemia, trauma, infection, dermal wounds, and atherosclerosis) and a major trigger of local and systemic inflammatory response [1 ]. Hypoxia has been shown to induce monocyte and macrophage recruitment, to cause their accumulation in hypoxic regions [2 3 4 ], and to alter their morphology, expression of surface molecules, viability, phagocytosis, and release of cytokines [5 ]. However, only limited information is currently available about the mechanisms by which hypoxia exerts these effects.

Extravasation of monocytes from the blood vessel and their migration in the tissue require the degradation of the cellular basement membrane (BM) and the extracellular matrix (ECM). To this end, monocytes and macrophages, as well as other types of migrating cells, secrete several classes of proteins, including the matrix metalloproteinases (MMPs), which are a family of proteolytic enzymes that can collectively cleave all the components of the BM and ECM. In addition MMPs can cleave active or nonactive enzymes and membrane and ECM-bound molecules. Therefore, MMPs, which were mainly associated with ECM remodeling and cellular migration, are now considered to be involved in numerous cell functions in health and disease conditions [6 7 8 9 ]. Two of the MMPs, MMP-9 and MMP-2, degrade collagen IV, a main component of BM, and its denatured form gelatin and are called gelatinases. Similar to other MMPs, MMP-9 and MMP-2 are secreted in their latent zymogenic form (92 kDa and 72 kDa, respectively) and are cleaved by other MMPs or other proteases (e.g., plasmin) to yield the activated forms (84 kDa for MMP-9 and 68 kDa, 58 kDa, or 54 kDa for MMP-2) [8 , 9 ]. As MMP-2 is expressed constitutively in many cell types, and the expression of MMP-9 is mostly induced by proinflammatory stimuli, the ratio between the two gelatinases is used as a marker for the severity of the inflammatory process in various diseases [6 , 10 , 11 ]. Many inflammatory mediators induce the expression of MMP-9 in monocytes and macrophages [e.g., tumor necrosis factor {alpha} (TNF-{alpha}), endotoxins, interleukin (IL)-1, IL-8], probably through the activation of the transcription factors nuclear factor (NF)-{kappa}B and activated protein-1 [12 13 14 ], whereas anti-inflammatory mediators, such as IL-4, IL-10, and transforming growth factor-ß, suppress it [14 , 15 ]. To avoid their excessive activity, which may lead to tissue damage, MMPs are tightly regulated at several levels, including their transcription, secretion, and activation [9 , 16 , 17 ]. In addition, MMPs are secreted as latent zymogens, which need to be activated locally [9 , 18 ], and are inhibited by their secreted endogenous inhibitors [tissue inhibitor of MMPs (TIMPs)] as well as other protease inhibitors (e.g., {alpha}2-macroglobulin) [9 , 19 , 20 ]. Pro- and active MMPs have also been found to bind surface proteins, which affect their activity and function [21 ].

The effects of hypoxia on the expression and activity of MMP-9 have scarcely been studied, and conflicting data were reported in different cells and different oxygen concentrations. For example, hypoxia increased MMP-9 in brain endothelial cells [22 ], whereas human umbilical vein endothelial cells did not express MMP-9 in hypoxia [23 ]. In the human breast cancer cells, hypoxia increased MMP-9 expression in the MDA-MB-231 cell line but decreased it in the MCF-7 cell line [24 ]. The effect of hypoxia on MMP-9 in monocytes and macrophages has not yet been investigated. In this study, we show that prolonged hypoxia post-translationally leads to the reduction in the amounts of soluble proMMP-9 in the supernatants, by causing the accumulation of the intracellular enzyme via cytoskeletal-mediated inhibition of its trafficking. An elevation of MMP-9 associated with cell-surface molecules may support the existence of an additional mechanism, in which hypoxia suppresses monocyte MMP-9 secretion.


   MATERIALS AND METHODS
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
RESULTS
 DISCUSSION
 REFERENCES
 
Cells
Human monocytes were isolated from peripheral blood donated by healthy volunteers. Blood (20 ml) was taken in the presence of 0.2% EDTA, and peripheral blood mononuclear cells (PBMC) were separated by density centrifugation on Ficoll-Hypaque (Uni-Sep NovaMed, Israel). After three washes in phosphate-buffered saline (PBS), the PBMC were counted and plated in 60 mm dishes with Dulbecco’s modified Eagle’s medium (DMEM) and 20% fetal calf serum (FCS) at a concentration of 107 cells/plate, and after 2 h of incubation, the nonadherent cells were washed out extensively. Purity of the monocytes was routinely ≥80 ± 0.5%, as was evaluated by labeling the cells with anti-CD14 and anti-CD11b and analyzing the surface expression by flow cytometry. This part of the study was approved by the Carmel Medical Center Helsinki Committee (Israel). In addition, the human monocytic cell line U937 was cultured in DMEM with 10% FCS and antibiotics. To avoid possible masking of signals initiated by the exogenous stimuli, cells were incubated with DMEM without FCS before their exposure to the experimental conditions. Cells were subjected to normoxia or hypoxia for 48 h, with or without the addition of TNF-{alpha} (20 ng/ml, Peprotech-Cytolab, Rehovot, Israel). In all experiments, cell viability was determined using the XTT kit (Biological Industries, Kibbutz Beit-haemek, Israel).

Normoxic and hypoxic conditions
For normoxic conditions, cells were incubated in a regular incubator (21% O2, 5% CO2, 74% N2). Hypoxic incubation was performed in a sealed, anaerobic work station (Concept 400, Ruskin Technologies, Leeds, UK; Jouan, Saint Herblain, France), where the hypoxic environment (O2 <0.3%, 5% CO2, 95% N2), the temperature (37°C), and humidity (90%) are kept constant. Samples from the culture medium were taken at the end of the exposure to hypoxia to determine the partial pressures of O2 and CO2, as well as pH, using a blood gas analyzer ABL510 (Radiometer, Denmark). Hypoxic PO2 values were 23 ± 0.8 mmHg, PCO2 values were 39.5 ± 0.6 mmHg, and pH values were 7.4 ± 0.025.

Zymography
Expression and secretion of MMP-9 were analyzed by zymography, which enables detection of the zymogen and the activated forms of MMP-9. Primary monocytes (2x106) or U937 cells were incubated in the experimental condition, supernatants were collected, and cellular extracts were obtained from the cell line by harvesting the cells in the presence of radio immunoprecipitation assay (RIPA) buffer [PBS, 1% Nonidet P-40, 0.5% sodium deoxycholate, 0.1% sodium dodecyl sulfate (SDS), 10 µg/ml phenylmethylsulfonyl fluoride, 30 µg/ml aprotinin, 5 µg/ml leupeptin]. Protein concentrations were determined by the Bradford reagent, and equal amounts of nonreduced protein (30 µg) or equal volumes of the nonreduced conditioned media were loaded onto an 8% SDS-polyacrylamide gel electrophoresis (PAGE) containing 1% gelatin. After electrophoretic separation, gels were renatured by incubation in 2.5% Triton X-100 for 30 min, and gelatinase activity was revealed by overnight incubation in a developing buffer (50 mM Tris-HCl, pH 7.5, 200 mM NaCl, 5 mM CaCl2, 0.02% Brij35) and staining with 0.5% Coomassie blue G-250. The clear bands that were obtained against the blue background indicated the presence of MMP-9 in comparison with the molecular weight markers and standard positive control (Chemicon, Temecula, CA), which were run on each gel. The optical density (OD) of the bands was quantified using the bio-imaging system (Dinco and Renium, Jerusalem, Israel) and TINA software (Raytest, Straubenhardt, Germany).

Enzyme-linked immunosorbent assay (ELISA)
A commercial ELISA kit (R&D Systems, Minneapolis, MN) was carried out according to the manufacturer’s instructions. This ELISA enabled measurement of active and latent forms of MMP-9.

Quantitative real-time polymerase chain reaction (PCR) analyses
Total RNA was extracted from 2 x 106 U937 cells exposed to the experimental conditions using TriReagentTM (Molecular Research Center, Cincinnati, OH), according to the manufacturer’s instructions. Total RNA (5 µg) was transcribed to cDNA at 37°C for 1 h using 200 µM deoxynucleotides (Sigma Chemical Co., St. Louis, MO), 5 µM random hexamers (Amersham Pharmacia Biotech, Piscataway, NJ), 20 U RNAguard (Amersham Pharmacia Biotech), and 200 U/µl Moloney murine leukemia virus-reverse transcriptase (US Biochemicals, Cleveland, OH). MMP-9 mRNA expression levels were determined by quantitative real-time PCR on the cDNA samples using the TaqMan assay-on-demand kit with the ABI-PRISM 7000 (Applied Biosystems, Foster City, CA). Analysis was carried out in triplicates in a volume of 20 µl (2 min at 50°C, 10 min at 95°C, and a total of 40 cycles, each of 15 s at 95°C and 1 min at 60°C) for MMP-9 and the endogenous reference gene RPLP0, which does not change in hypoxia, and the comparative threshold (CT) cycle method was used. In each experiment, the normoxic, nonstimulated RNA sample was used as a calibrator to allow comparison of relative quantity between the samples.

Western blots analyses
After exposure to the experimental conditions, supernatants from U937 cells were collected and concentrated 50-fold by VivaSpin2 (Vivascience, Lincoln, UK), and equal volumes were loaded on a 10% SDS-PAGE. After electrophoretic separation, the proteins were transferred and fixed onto cellulose nitrate membranes (Schleicher and Schuell, Dassel, Germany) in transfer buffer (25 mM Tris, 180 mM glycine, 20% methanol, pH 8.3). The membranes were incubated for 1 h in blocking buffer (20% skimmed milk, 1% bovine serum albumin, 0.01% Tween-20, 10 mM Tris, pH 8.0, 150 mM NaCl) at room temperature and then probed for 1 h at room temperature with the diluted (1:100) mouse monoclonal anti-MMP-9 (Oncogene, Boston, MA). After washing three times in 1x Tris-buffered saline/Tween-20 (10 mM Tris, pH 8.0, 150 mM NaCl, 0.5% Tween-20), the membranes were incubated with horseradish peroxidase-conjugated donkey anti-mouse immunoglobulin G (IgG; Jackson ImmunoReasearch Laboratories, West Grove, PA), diluted 1:2000 in blocking buffer for an additional 1 h at room temperature, and then washed again. The enhanced chemiluminescence system (Amersham Pharmacia Biotech) was used for detection. The OD of the bands was quantified using the bio-imaging system and TINA software.

Immunoprecipitation
U937 cells were cultured and exposed to normoxia or hypoxia, with or without the addition of TNF-{alpha} for 48 h. Cells were harvested in RIPA buffer, and protein concentrations of the cellular extracts were measured by the Bradford reagent. To avoid nonspecific binding, 400 µg protein was incubated with 2 µg normal goat serum and precipitated using protein G plus agarose beads. The remaining proteins were incubated with 1 µg goat polyclonal anti-reversion-inducing cysteine-rich protein with Kazal motifs (RECK; R&D Systems) and rotated overnight at 4°C with protein G plus agarose. After centrifugation, the pellet was washed four times in PBS, resuspended in loading buffer, and boiled for 15 min and loaded onto a 10% SDS-PAGE. Similarly, immunoprecipitation of CD44 was performed using 2 µg normal mouse serum for nonspecific binding, 1 µg mouse monoclonal anti-CD44 (R&D Systems), and protein A-agarose. After electrophoretic separation, proteins were transferred and fixed onto a cellulose-nitrate membrane, and the membranes were then probed with anti-MMP-9, as described previously in the Western blot analysis.

Flow cytometry analysis
U937 cells were cultured in normoxia or hypoxia for 48 h, with or without the addition of TNF-{alpha}, and then, were labeled with monoclonal anti-MMP-9 and phycoerythrin (PE)-conjugated goat anti-mouse (Chemicon). Similarly, cells were labeled with the polyclonal antibody directed against RECK (Chemicon) and with fluorescein isothiocyanate-conjugated rabbit anti-goat or with the monoclonal antibody (mAb) directed against CD44 (Chemicon) and with the PE-conjugated goat anti-mouse. After washing, the cells were fixed in 0.1% formaldehyde and were analyzed using a Coulter-XL flow cytometer (Coulter Electronics, UK). Dead cells were excluded from the analysis by their forward- and side-way light-scattering properties.

Immunofluorescence
U937 cells were exposed to normoxia or hypoxia for 48 h, with or without TNF-{alpha}. The cells were collected into a tube and fixated with 3.7% formaldehyde for 10 min at room temperature. To avoid nonspecific binding, the cells were incubated with 4% donkey normal serum for 30 min at room temperature and then washed three times with PBS. Primary antibodies (goat polyclonal anti-MMP-9 or rabbit polyclonal anti-Rab8, Santa Cruz Biotechnology, CA) were incubated for 1 h at room temperature following three washes with PBS. Secondary antibodies (Rhodamine Red-X-conjugated donkey anti-goat IgG or Cy2-conjugated donkey anti-rabbit IgG, Jackson ImmunoResearch Laboratories) were incubated in the dark for 1 h at room temperature and then washed three times with PBS. Actin was stained with Alexa 488 phalloidin. Cells were then spread on coverslips, which were mounted on a slide with fluoromount G. Immunofluorescent images were acquired by confocal microscopy using the Bio-Rad MRC 1000 confocal system (Hercules, CA), and the images were analyzed for the integrated density of the fluorescence and for the area of the cells using the ImagePro Plus 4.5 software (Media Cybernetics, Silver Spring, MD).

Migration assay
Migration assays were performed in a modified Boyden chamber transwell (3 µm pores, Falcon). Briefly, a total of 5 x 105 U937 cells were added to the upper chamber, and the chemoattractant monocyte chemoattractant protein-1 (MCP-1; 100 ng/ml, R&D Systems) was added to the lower chamber. Basal migration and MCP-1-directed migration were determined after 48 h by adding a fixed volume of fluorescent beads (200 µl PKH26 reference beads, Sigma Chemical Co.), collecting the cells that migrated to the lower chamber, and determining their ratio in the sample relative to the total number of cells using flow cytometry.

Statistical analyses
All values are presented as means ± SE. The data were analyzed using repeated measures ANOVA. The Student Newman-Keuls multiple comparisons test was used to evaluate the significance between experimental groups, and P values exceeding 0.05 were not considered significant.


   RESULTS
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
DISCUSSION
 REFERENCES
 
Hypoxia and TNF-{alpha} do not change cell viability
Preliminary experiments indicated that differences in the amounts of MMP-9 in the supernatants were not visible before exposure of 48 h to hypoxia, and therefore, all experiments were conducted at that time-point. To verify that these conditions did not affect cell viability, we performed an XTT test at the end of the incubation period. Figure 1A and 1B , shows that exposure to hypoxia for 48 h, addition of the cytokine TNF-{alpha}, or their combination did not increase cell death in primary peripheral blood monocytes or in the monocytic U937 cell line.


Figure 1
View larger version (10K):
[in this window]
[in a new window]
  		Figure 1. Effect of hypoxia and TNF-{alpha} on cell viability. Cells (4x105) were plated in 96-well plates in a serum-free medium and incubated, with or without the addition of TNF-{alpha} (20 ng/ml) for 48 h in normoxia and hypoxia. Determination of cell viability was assessed by addition of XTT at the end of the incubation period and comparison with the experimental control (cells without TNF-{alpha} incubated in normoxia). (A) Cell viability of primary monocytes (n=4). (B) Cell viability of U937 monocytic cell line (n=11). Hypoxia or addition of TNF-{alpha} did not cause cell death.

 
Hypoxia reduces secretion of proMMP-9
MMP-9 is induced by inflammatory mediators such as TNF-{alpha}, and without such stimulation, only small amounts of the enzyme are produced. We analyzed the amounts of MMP-9 secreted by primary monocytes (Fig. 2A ) and the U937 monocytic cell line (Fig. 2B) by zymography. Using this method, which enables the detection of the zymogenic and the activated forms of MMP-9, only the latent form of the enzyme could be observed. Unlike the U937 cells, the primary monocytes exhibited a distinct band corresponding to proMMP-9, even without stimulation with TNF-{alpha}. In primary monocytes, hypoxia alone did not affect proMMP-9 levels. Addition of TNF-{alpha} in normoxia increased the amounts of secreted proMMP-9 by 2.3 ± 0.4-fold (P<0.01 relative to normoxia), and hypoxia attenuated the TNF-{alpha}-induced increase in proMMP-9 and maintained it at levels not significantly different from the normoxic control. Similarly, U937 cells, which were incubated in the absence of TNF-{alpha}, secreted small amounts of proMMP-9, and hypoxia alone did not cause a significant change (1.4±0.2-fold compared with the nontriggered cells in normoxia). Addition of TNF-{alpha} in normoxia increased proMMP-9 levels in the supernatants by 13.2 ± 1.8-fold compared with the control (P<0.001). However, hypoxia inhibited TNF-{alpha}-induced secretion of proMMP-9 (P<0.05). To further validate the zymography results, we evaluated the amounts of MMP-9 in the supernatants of U937 cells by ELISA. The TNF-{alpha}-induced amounts of MMP-9 in normoxia were 9.3 ± 2.6 ng/ml, and hypoxia reduced it to 3.3 ± 1.06 ng/ml (Fig. 2D) . Altogether, hypoxia reduced the secretion of TNF-{alpha}-induced proMMP-9 by threefold (P<0.001) and by 2.6-fold (P<0.01) in the monocytic cell line and in the primary monocytes, respectively.


Figure 2
View larger version (18K):
[in this window]
[in a new window]
  		Figure 2. Effect of hypoxia on secretion of proMMP-9 from monocytes. Cells (2x106) were plated in 24-well plates in a serum-free medium and were incubated in normoxia or hypoxia for 48 h, with or without the addition of TNF-{alpha} (20 ng/ml). At the end of incubation, supernatants were collected, and equal amounts were loaded onto a zymography gel. (A) Representative zymography gels. PC, positive control. (B) Densitometric analysis of proMMP-9 secretion from primary monocytes (n=4). (C) Densitometric analysis of proMMP-9 secretion from the U937 monocytic cell line (n=13). (D) MMP-9 amounts in supernatants of U937 cells (n=4). *, P < 0.05; **, P < 0.01; ***, P < 0.001, compared with nonstimulated cells in normoxia (control). §, P < 0.05; §§, P < 0.01; §§§, P < 0.001, compared with TNF-{alpha}-induced cells in normoxia and with nonstimulated hypoxia (C and D). Hypoxia reduced the secretion of TNF-{alpha}-induced MMP-9.

 
Hypoxia does not affect MMP-9 mRNA steady-state
To further study molecular mechanisms responsible for the hypoxia-induced reduction of proMMP-9 in the supernatants, we focused on the U937 cell line. To examine if hypoxia exerted its inhibitory effect at the transcriptional level, we used quantitative real-time PCR to measure the relative steady-state levels of MMP-9 mRNA in the different experimental groups. Each sample was normalized to the endogenous reference ribosomal gene RPLP0, which is similarly transcribed but does not change in hypoxia, as do other housekeeping genes [25 ]. As shown in Figure 3 , stimulation with TNF-{alpha} induced a 30-fold increase in the steady-state levels of proMMP-9 mRNA (P<0.01), but no significant differences were found between the induced levels in normoxia and hypoxia. Hence, addition of the inflammatory mediator TNF-{alpha} induced MMP-9 at the transcriptional level, but hypoxia had no transcriptional effect.


Figure 3
View larger version (11K):
[in this window]
[in a new window]
  		Figure 3. Effect of hypoxia on the steady-state MMP-9 mRNA. U937 cells (2x106) were plated in 24-well plates in a serum-free medium and were incubated in normoxia or hypoxia for 48 h, with or without the addition of TNF-{alpha} (20 ng/ml). At the end of incubation, total RNA was extracted and reverse-transcribed, and MMP-9 mRNA was amplified as described in Materials and Methods using the TaqMan primers and probes and ABI-PRISM 7000 sequence detection system. Samples were normalized to the endogenous reference gene RPLP0, and for each experiment, the calibrator was the nonstimulated normoxic cells (n=6). **, P < 0.01, compared with nonstimulated cells in normoxia or in hypoxia. TNF-{alpha} induced the transcription of MMP-9 mRNA, but hypoxia had no transcriptional or post-transcriptional effect on the steady-state MMP-9 mRNA in these cells. RQ, relative quantity.

 
Hypoxia increases the surface attachment of proMMP-9 via its binding to CD44
The possibility that hypoxia increases the degradation rate of secreted proMMP-9 was examined by Western blot analyses using the mAb directed against MMP-9 to detect MMP-9 fragments in the hypoxic or normoxic supernatants. Although the autoradiograms indicated that hypoxia reduced the levels of proMMP-9 in the supernatants, in accordance with the results observed by zymography, no additional, degraded bands of the protein were observed (data not shown), ruling out this possibility.

Cell-surface association of proMMP-9 was determined by flow cytometry, using the same mAb directed against MMP-9 (Fig. 4 ). In normoxia with the addition of TNF-{alpha}, only 15.7 ± 2.6% of the cells were stained positively for membranal MMP-9 with a mean fluorescence of 0.52 ± 0.06. However, exposure of the cells to hypoxia and TNF increased the percentage of cells exhibiting membranal attachment of MMP-9 to 49.3 ± 6.5%, with a mean fluorescence of 2.2 ± 0.33 (P<0.01 compared with TNF-stimulated, normoxic cells). Thus, hypoxia in the presence of TNF-{alpha} increased the membranal attachment of MMP-9.


Figure 4
View larger version (17K):
[in this window]
[in a new window]
  		Figure 4. Effect of hypoxia on surface attachment of proMMP-9. U937 cells (2x106) were plated in 24-well plates in a serum-free medium and were incubated in normoxia or hypoxia for 48 h with the addition of TNF-{alpha} (20 ng/ml). At the end of incubation, cells were labeled with anti-MMP-9 and PE-conjugated goat anti-mouse and subjected to flow cytometry analysis. A representative analysis in the presence of TNF-{alpha} (n=4) demonstrates that hypoxia increases the surface-associated binding of MMP-9. Light gray line, Isotype control; gray line, MMP-9 in normoxia; black line, MMP-9 in hypoxia.

 
To further understand this hypoxia-increased membranal attachment, we used flow cytometry using antibodies directed against two of the surface proteins that have been reported previously to bind MMP-9: CD44 and RECK [21 ]. In normoxia with TNF-{alpha}, 27.5 ± 13.3% of the cells were stained positively for surface expression of CD44 with a mean fluorescence of 1.8 ± 0.7 (Fig. 5A ). Hypoxia in the presence of TNF-{alpha} increased expression of surface CD44 to 39.4 ± 12.4% of the cells, with a mean fluorescence of 3.3 ± 0.9 (P<0.01). Similarly, 5.5 ± 3.1% of the TNF-{alpha}-stimulated, normoxic cells were positive for the surface expression of RECK with a mean fluorescence of 0.62 ± 0.2 (data not shown). Hypoxia and TNF-{alpha} increased the percentage of cells positively stained for RECK to 14.2 ± 1.5%, with a mean fluorescence of 0.91 ± 0.4 (P<0.05).


Figure 5
View larger version (14K):
[in this window]
[in a new window]
  		Figure 5. Effect of hypoxia on the binding of MMP-9 to CD44. U937 cells (2x106) were plated in 24-well plates in a serum-free medium and were incubated in normoxia or hypoxia for 48 h with the addition of TNF-{alpha} (20 ng/ml). (A) At the end of incubation, cells were labeled with anti-CD44 and PE-conjugated goat anti-mouse (n=5), and surface expression of CD44 was evaluated. Light gray line, Isotype control; gray line, CD44 in normoxia; black line, CD44 in hypoxia. (B) A representative experiment of immunoprecipitation (IP) with anti-CD44, followed by Western blot analysis (BL) with the monoclonal anti-MMP-9 (n=4), shows that in hypoxia, binding of MMP-9 to CD44 increased. N, Normoxia; H, hypoxia.

 
We next explored the effect of hypoxia on the binding of MMP-9 to CD44 and to RECK by immunoprecipitation of these proteins followed by Western blot analysis using the monoclonal anti-MMP-9 (Fig. 5B) . Immunoprecipitation of proMMP-9 using anti-CD44 resulted in binding, which was further increased by 18% (P<0.05) in hypoxia. A similar trend could be observed when anti-RECK was used for immunoprecipitation experiments; however, the weak signal produced could not be quantified by densitometry.

Hypoxia increases the intracellular levels of proMMP-9 by inhibiting its trafficking
To explore the possibility that hypoxia causes accumulation of intracellular proMMP-9 and therefore, its decreased secretion, cellular extracts from U937 cells were obtained after incubation for 48 h, with or without TNF-{alpha} in normoxia or hypoxia. Figure 6 shows that without stimulation by TNF-{alpha}, the cellular levels of proMMP-9, as measured by zymography, were low. However, addition of the cytokine increased intracellular proMMP-9 by 4.6 ± 0.7-fold in normoxia (P<0.001), and hypoxia increased it further by 47% (P<0.05). In search for the mechanisms mediating the retention of intracellular MMP-9, we further investigated the possibility that hypoxia interrupts its trafficking. As secretion of proteins containing a signal peptide such as proMMP-9 often occurs via secretory vesicles and as actin and tubulin cytoskeletal fibers are involved in trafficking of these vesicles, we took two separate approaches to investigate the involvement of these proteins in the trafficking of proMMP-9.


Figure 6
View larger version (14K):
[in this window]
[in a new window]
  		Figure 6. Effect of hypoxia on the accumulation of intracellular MMP-9. U937 cells (2x106) were plated in 24-well plates in a serum-free medium and were incubated in normoxia or hypoxia for 48 h, with or without the addition of TNF-{alpha} (20 ng/ml). At the end of incubation, cells were lysed, cellular proteins were extracted, and equal amounts of protein were loaded onto a zymography gel. (A) A representative zymography gel. (B) Densitometric analysis of intracellular proMMP-9 (n=7). *, P < 0.05, compared with TNF-{alpha}-stimulated cells in normoxia. Hypoxia increased the accumulation of intracellular TNF-{alpha}-induced MMP-9.

 
First, we disrupted the actin and tubulin microfibers by using cytochalazin B and nocodazole, which, respectively, inhibit their polymerization. As demonstrated in Figure 7A , cytochalazin B (1.25 and 2.5 µM) decreased proMMP-9 secretion in normoxia by 32% (P<0.05). In hypoxia, however, cytochalazin B had no additional effect beyond that already observed in hypoxia alone (Fig. 2B) . In normoxic cultures, the tubulin inhibitor nocodazole caused a significant, dose-dependent decrease in proMMP-9 secretion (Fig. 7B) , up to 76% (P<0.001) in the higher concentrations. Nocodazole, like cytochalazin B, did not further reduce the already-decreased amounts of proMMP-9 in hypoxia.


Figure 7
View larger version (21K):
[in this window]
[in a new window]
  		Figure 7. Effect of cytochalazin B and nocodazole on the secretion of MMP-9. U937 cells (2x106) were plated in 24-well plates in a serum-free medium and were incubated with TNF-{alpha} (20 ng/ml) in normoxia or hypoxia for 48 h, with the addition of increasing amounts of (A) the actin polymerization inhibitor cytochalazin B (n=5) or (B) the tubulin polymerization inhibitor nocodazole (n=5). At the end of incubation, supernatants were collected, and equal amounts were loaded onto a zymography gel. *, P < 0.05; **, P < 0.01; ***, P < 0.001, compared with TNF-{alpha}-stimulated cells in normoxia (control). Cytochalazin B, and to a greater extent nocodazole, inhibited the secretion of MMP-9 in normoxia in a dose-dependent manner, whereas in hypoxia, they exhibited no additional effect over that of hypoxia alone.

 
Second, we examined the effects of hypoxia on the colocalization of MMP-9, with actin or tubulin or with Rab8, a marker of secretory vesicles, by confocal microscopy. U937 cells were incubated in normoxia or hypoxia with TNF-{alpha} to induce proMMP-9 expression and stained for MMP-9, Rab8, actin, or tubulin. Isotype-matched control sera yielded no immunostaining or only a weak staining. Unstimulated cells did not show any staining with MMP-9, as the protein was not induced, but Rab8, actin, and tubulin were strongly stained in all cells (data not shown). Analysis of the images demonstrated that in hypoxia, the mean area of the cells was increased significantly by twofold (0.15±0.03 µm2 in normoxia and 0.26±0.03 µm2 in hypoxia; P<0.001). This expansion of the cells may have caused the reduced intensity of the fluorescent signal in hypoxia, as it was distributed over a larger area. To correct for this effect, the integrated OD (IOD), which takes into account the area of the cell, was measured. The IOD value for each fluorescently labeled protein was calculated separately and determined in three repetitions of the experiment, in at least five different fields. Consistent with the results obtained in Figure 6 , the IOD values of the intracellular MMP-9 staining were increased by 62% in hypoxia with TNF-{alpha} compared with normoxia with TNF-{alpha} (P<0.04). In normoxia, a strong colocalization was observed between MMP-9 and Rab8 in the merged image (Fig. 8A , panel C). The Rab8 fluorescent signal was reduced in hypoxia (Fig. 8A , panel A compared with panel D), and the MMP-9 signal was enhanced (Fig. 8A , panel B compared with panel E), and reduced colocalization of MMP-9 and Rab8 was observed (Fig. 8A , panel C compared with panel F). The density (IOD) of the actin and tubulin fibers was increased in hypoxia compared with normoxia by 95% (P<0.03) and 48% (P<0.04), respectively (Fig. 8B , panels A compared with D and panel B compared with E), suggesting changes in the cytoskeleton. MMP-9 was colocalized with the outer layer of actin fibers and with the inner layer of microtubules in normoxia (Fig. 8B , panels G and H), but exposure to hypoxia disrupted this colocalization (Fig. 8B , panels J and K).


Figure 8
View larger version (41K):
[in this window]
[in a new window]
  		Figure 8. Effect of hypoxia on the colocalization of MMP-9, Rab8, actin, and tubulin. U937 cells (2x106) were plated in 24-well plates in a serum-free medium and were incubated with TNF-{alpha} (20 ng/ml) in normoxia or hypoxia for 48 h. At the end of the incubation, cells were immunostained with (A) anti-MMP-9 (red, panels B and E) and anti-Rab8 (green, panels A and D) or with (B) anti-MMP-9 (red, panels C and F), anti-{alpha}-tubulin (blue, panels B and E), and phalloidin (green, panels A and D). The merged MMP-9 and Rab8 images from A + B and D + E are shown (panels C and F), and areas of colocalization are shown in yellow. The merged MMP-9 and tubulin images from B + C and E + F are shown (panels G and J), and areas of colocalization are shown in pink. The merged MMP-9 and actin images from A + B and D + E are shown (panels H and K), and areas of colocalization are shown in yellow. The merged MMP-9, tubulin, and actin images from A + B + C and D + E + F are shown (panels I and L), and areas of colocalization are shown in pink, pale blue, or white. Original magnification was x1800. Hypoxia diminished the colocalization of MMP-9 and Rab8 and clearly disrupted the colocalization of MMP-9 and tubulin.

 
Hypoxia signaling occurs partially via Rho kinase
We next looked at the possible signal transduction pathways by which hypoxia exerts its post-translational effects on proMMP-9 secretion. We incubated U937 cells in the presence or absence of TNF-{alpha} in normoxia and hypoxia, with several inhibitors that affect known signaling pathways involved in hypoxia and inflammation, and observed their effect on proMMP-9 levels in the supernatants. In normoxia with TNF-{alpha}, PD98050 [extracellular signal-regulated kinase (ERK)1/2 inhibitor], SB203580 (p38 inhibitor), Y27632 (Rho kinase inhibitor), chelerythrine [protein kinase C (PKC) inhibitor], and SN50 (NF-{kappa}B inhibitor) had no effect on the amounts of proMMP-9 in the supernatants (Fig. 9 ). The JNK inhibitor II, which inhibits the three JNK isoforms, dose-dependently decreased proMMP-9 amounts by up to 4.5-fold, relative to those measured in normoxia (P<0.01). A similar, dose-dependent reduction was observed in hypoxia with TNF-{alpha}, suggesting that TNF-{alpha} induction of proMMP-9 in the U937 cells was mediated through Jun MAPK and not through NF-{kappa}B. In addition, only in hypoxia, the Rho kinase inhibitor Y27632 dose-dependently reduced the amounts of secreted proMMP-9 (up to 46% inhibition, P<0.05), suggesting the involvement of the enzyme in the post-translational, hypoxia-mediated effects on proMMP-9.


Figure 9
View larger version (21K):
[in this window]
[in a new window]
  		Figure 9. Effect of inhibitors of different signaling pathways on the secretion of MMP-9. U937 cells (2x106) were plated in 24-well plates in a serum-free medium and were incubated, with or without the addition of TNF-{alpha} (20 ng/ml) in normoxia or hypoxia for 48 h. (A) In some of the wells, the following signaling inhibitors were also added: the ERK1/2 inhibitor PD98050 (50 µM), the p38 mitogen-activated protein kinase (MAPK) inhibitor SB203580 (100 µM), the Jun N-terminal kianse (JNK) inhibitor II (25 mM), the Rho kinase inhibitor Y27632 (10 µM), the PKC inhibitor chelerythrine (5 µM), and the NF-{kappa}B translocation inhibitor SN50 (50 µg/ml). At the end of incubation, supernatants were collected, equal amounts were loaded onto a zymography gel, and a densitometric analysis was performed (n=6). *, P < 0.05; **, P < 0.01 compared with the TNF{alpha}-induced cells in hypoxia. **, P < 0.01, compared with the TNF-{alpha}-induced cells in normoxia. §, P < 0.05, and §§, P < 0.01, compared with TNF-{alpha}-induced cells in hypoxia. (B) Addition of several amounts of the JNK inhibitor II, which inhibits all three JNK isoforms, resulted in a dose-dependent effect, demonstrating that TNF-{alpha} mediates its effects on MMP-9 induction via the JNK pathway in normoxia and hypoxia. *, P < 0.05 and **, P < 0.01 compared with TNF{alpha}-induced cells in normoxia or hypoxia. (C) Addition of several amounts of the Rho kinase inhibitor Y27632 demonstrated a dose-dependent response only in hypoxia, suggesting that Rho kinase is involved in the secretion of MMP-9 only in hypoxia.

 
Hypoxia-reduced MMP-9 affects migration of monocytes
To demonstrate that the hypoxia-reduced amounts of MMP-9 are physiologically relevant to the migration of monocytes, we performed a chemotaxis assay in normoxia and hypoxia with the addition of anti-MMP-9 (20 ng/ml) or the recombinant protein (20 ng/ml). Figure 10 shows that addition of TNF-{alpha} in normoxia enhanced U937 cell migration by 36.7 ± 3.9%, whereas only 4.1 ± 0.4% of the nonstimulated cells migrated. Hypoxia reduced the TNF-{alpha}-induced migration by twofold (17.4±3.4%, P<0.01), whereas nonstimulated cells in hypoxia did not change their migration rate. Addition of anti-MMP-9 inhibited the TNF-{alpha}-induced normoxic migration by 60% (P<0.01), and addition of the recombinant protein did not change the normoxic migration. However, in hypoxia, addition of the antibody had no further effect on migration, whereas the recombinant protein increased it by twofold (P<0.01) despite the hypoxic environment, restoring migration to normoxic values.


Figure 10
View larger version (14K):
[in this window]
[in a new window]
  		Figure 10. Effects of hypoxia on monocyte migration are mediated by proMMP-9. U937 cells (5x105) were added to the upper chamber of a Boyden chamber transwell and were subjected to normoxia or hypoxia for 48 h. Migration of the cells toward the chemoattractant MCP-1 (100 ng/ml), which was added to the lower chamber, was measured at that time, and in some of the wells, anti-MMP-9 (20 ng/ml) or recombinant MMP-9 (20 ng/ml) was added to the upper chamber. Results are presented as percentage of migrating cells above the basal level (n=4). Hypoxia reduced U937 migration, and this phenomenon was, at least partially, dependent on MMP-9.

 

   DISCUSSION
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
REFERENCES
 
In this study, we examined the hypoxia-induced control mechanisms of monocyte MMP-9, which is involved in their extravasation from blood vessels and migration in the damaged tissue. The results presented in this study show that hypoxia post-translationally reduces the amounts of TNF-{alpha}-induced proMMP-9 in supernatants derived from human primary and cell line monocytes by at least two mechanisms. The first is increased accumulation of intracellular proMMP-9 as a result of interference with its trafficking, which is regulated by hypoxia-induced alterations in the cytoskeleton. The second is increased surface attachment of proMMP-9 to the cell-surface CD44, the hyaluronan receptor, and to a much lesser extent, also to RECK, a membranal inhibitor of MMPs.

In the present study, we exposed macrophages to controlled hypoxia with ambient supernatant PO2 ranging between 22 and 24 mmHg. This level of oxygenation is relevant to clinical situations characterized by tissue hypoxia. Ambient tissue oxygen partial pressures below 30–40 mmHg diminish leukocyte killing and necrotic tissue proteolysis [5 , 26 ]. Furthermore, processes, which are of paramount importance for tissue repair and regeneration, such as fibroblast proliferation, collagen deposition, angiogenesis, and migration of epithelial cells, are also inhibited significantly under such hypoxia [26 ]. Significantly lower tissue PO2 is usually associated with irreversible damage with slim chances of recovery. We first confirmed (data not shown) that although cells were exposed to hypoxia (<0.3% O2), changes in the accumulation of proMMP-9 could be observed only after a prolonged culture period of 48 h. Such conditions may simulate the monocyte, which has already migrated into the tissue, arrived at the inflicted site where myriad inflammatory cytokines (e.g., TNF-{alpha}) are present, and remains subjected to prolonged or chronic hypoxia. Even in such stress, cell death did not occur, and macrophages remained resistant for an extended period of time. Thus, the reduction in the amounts of proMMP-9 observed in the supernatants was not a result of a decreased number of cells but an actual consequence of adaptation to the stress. Unlike the untriggered monocytic cell line, which exhibited low amounts of proMMP-9 in zymography of supernatants, the primary monocytes produced proMMP-9, which could be detected clearly in the absence of the strong inducer TNF-{alpha}, resulting in a smaller effect of the cytokine. Human primary monocytes were shown previously to secrete MMP-9 constitutively, without further stimulation, as a result of cell-matrix adhesive interactions mediated by engagement of integrins, probably caused by constant homeostatic activation [27 ], or reflecting activation caused by their purification procedure [28 ]. Nonetheless, although induction of proMMP-9 in the primary monocytes was smaller compared with that measured in the U937 cells, the inhibitory effect of hypoxia on proMMP-9 following stimulation with TNF-{alpha} was similar (threefold) in both cell types. Once this phenomenon was established, we have used only the monocytic cell line to study molecular mechanisms involved in its regulation, as those experiments demanded a larger number of cells.

A major checkpoint for regulation of MMP-9 expression is at the transcriptional level, and we have indeed demonstrated that TNF-{alpha} induction of the enzyme occurs at this level. However, hypoxia alone or hypoxia with TNF-{alpha} did not change the levels of the steady-state MMP-9 mRNA, as assessed by real-time PCR, and even increased the intracellular accumulation of the protein, indicating that it did not affect MMP-9 production but rather, inhibited its secretion or activation via a post-translational effect.

In our study, we could detect only the latent proMMP-9 form in zymography. This analysis relies on the SDS-PAGE separation of proMMP-9 from other proteins, which may bind it so that the possibility of an inhibitory protein (e.g., TIMP-1), binding to the soluble proMMP-9 and inhibiting its activity, is unlikely. Appearance of the proMMP-9 zymogen form is frequently observed in cultured cells, although the U937 cells and primary monocytes are known to contain potential proMMP-9 activators. Culture conditions do not favor proMMP-9 activation, as they promote rapid dissociation of membrane-bound proMMP-9 or allow the media to dilute soluble activators [21 ]. In addition, many studies described cellular activity of MMP-9, even in the absence of its 82 kDa-activated form or in the presence of specific inhibitors, thus presenting a dilemma in the understanding of proMMP-9 activation and function.

The post-translational reduction of proMMP-9 in the supernatants, without a change in its mRNA expression, could be explained by three possible nonmutually exclusive mechanisms: intracellular accumulation of the enzyme in the hypoxic cells, its increased attachment to the hypoxic cell membrane, and/or its increased proteolytic degradation in the hypoxic conditioned media. The third possibility was ruled out, as no degradation products of MMP-9 were found. However, we demonstrated that hypoxia caused proMMP-9 intracellular accumulation. In addition, a tendency toward elevated attachment of proMMP-9 to the membrane may have contributed, to some extent, to the end result.

Although not yet widely accepted, it has been shown in different nonmacrophage cell types that part of the secreted proMMP-9 can be associated with the cell surface. In human and murine malignant epithelial cells, binding of MMP-9 to the cell surface was mediated by several proteins, such as the {alpha}2(VI) chain of collagen IV [29 ], the adhesion molecule intercellular adhesion molecule-1 (ICAM-1) [30 ], the hyaluronan receptor CD44 [31 , 32 ], RECK, which functions as a membranal inhibitor of MMPs [33 34 35 ], and low-density lipoprotein receptor-related protein (LRP), which mediates endocytotic intake of proteins [36 ]. Binding of these proteins may localize the already secreted proMMP-9 to the front of the extending lamellipodia [{alpha}2(IV), CD44], inhibit its activity (RECK), induce its internalization (LRP), or result in cleavage of the binding molecule itself (ICAM-1). Generally, membrane-anchored MMP-9 would degrade membrane-associated proteins more efficiently than its secreted form, and secreted MMP-9 would process different substrates in a certain distance from the cell. Moreover, as a result of the inhibitory effects of several surface-binding ligands, bound and secreted MMP-9 would differentially modulate cell behavior and its surrounding. Our results show that the combination of hypoxia and TNF-{alpha} up-regulates the surface expression of CD44 and RECK, which seem to have opposite effects on MMP-9, and slightly increases the binding of MMP-9 to both. Therefore, the effects of hypoxia on the cell-associated functions of MMP-9 cannot be clarified by the present study.

Previous studies suggest that the cytoskeleton is a regulator of MMP-9 [37 , 38 ] and its family member MMP-2 [39 ]. In this study, four independent lines of evidences suggest indirectly that hypoxia affects proMMP-9 secretion by changing the cytoskeleton. First, zymography analysis of the cellular extracts reveals that intracellular accumulation of TNF-{alpha}-induced proMMP-9 occurs in hypoxia compared with normoxia. Of note, in these zymography gels, a lower band of the enzyme or a small smear could sometimes be observed, which did not correspond to the active form of the enzyme, and may have resulted from an under-glycosylated precursor, which needs to undergo full glycosylation during its trafficking through the secretory pathway [40 , 41 ]. Second, addition of increasing amounts of cytochalazin B and nocodazole, respective inhibitors of actin and tubulin polymerization, resulted in dose-dependent inhibition of proMMP-9 secretion to the medium. Inhibition occurred only in normoxia, and thus, we suggest that both types of cytoskeletal fibers, actin and tubulin, are depolarized in hypoxia and cause the observed inhibition of proMMP-9 secretion. The inhibitory effects of nocodazole are in accordance with a pervious report [42 ] in which nocodazole-induced depolymerization of microtubules in mouse mammary carcinoma cells inhibited MMP-9 secretion, and pretreatment with paclitaxel, which induces centralization of the microtubule system, enhanced secretion of the protease. Third, analysis of images obtained by confocal microscopy revealed that hypoxia caused intracellular accumulation of MMP-9, confirming the results of the zymography analysis. As the confocal image analysis was influenced by the increase of the area of cells in hypoxia, which is consistent with the enhanced attachment and spreading of macrophages in response to hypoxia [5 ], a correction was made, and the IOD values were measured. In addition, we looked for colocalization between MMP-9 and Rab8, a marker of secretory vesicles, in normoxia and hypoxia. Rab8 is a member of the small Rab GTPases family of proteins, which regulates different transport routes in the cell. Rab8 participates in directing secretory vesicles to the membrane through reorganization of actin and microtubules and induces the formation of new surface extensions [43 ]. Colocalization of MMP-9 and Rab8 was observed clearly in normoxia. However, in hypoxia, colocalization was diminished, as these two proteins were dissociated or as the hypoxia-induced impairment of the vesicles themselves resulted in reduced Rab8 fluorescence despite colocalization. The regulation of the cytoskeletal elements actin and tubulin and that of the secretory vesicles have been found to be coordinated in migrating cells [43 , 44 ]. Therefore, the hypoxia-induced alternations in these three cellular compartments are logical. In addition, a recent study strongly supports a role for cytoskeleton changes in hypoxia, as it showed that hypoxia increased tubulin stabilization and changed vesicle trafficking [45 ]. Fourth, by using the selective inhibitor Y27632, we demonstrated the involvement of Rho kinase (Rock) in the hypoxia-induced reduction of proMMP-9 in the supernatants. Rock and RhoA, which activates it, are members of the small Rho GTPases family, which is known as critical regulators of actin reorganization. By using the same Rock inhibitor Y27632, others have already shown that the RhoA/Rock pathway is activated in monocytes [46 , 47 ] and is associated with cytoskeletal alterations in hypoxia [48 ], affecting monocyte adhesion and migration. As we showed that the combination of Y27632 and hypoxia dose-dependently reduced proMMP-9 in the supernatants more than hypoxia alone, a modest, prosecretory role for the RhoA/Rock pathway can be suggested in the hypoxic, but not normoxic, monocyte. It is interesting that in osteoclasts, activation of Rock led to increased surface expression of CD44 [49 ], thus raising the possibility that the active RhoA/Rock pathway in hypoxia may be involved in the surface binding of MMP-9. However, the signal transduction pathways, which are responsible for the inhibition of proMMP-9 secretion in hypoxia, are still unclear and deserve further investigation.

The arrest of monocytes and macrophages in hypoxic areas of tumors and ischemic tissues is well-documented [2 3 4 ], although the mechanisms involved are only beginning to be elucidated. Hypoxia was shown to increase monocyte-directed migration and entrapment by differently modulating the expression of chemokines [e.g., CC chemokine ligand 2 (CCL2), CCL5, CCL3, CXC chemokine ligand 8] and their receptors (e.g., CC chemokine receptor 2, CXC chemokine receptor 4) near the hypoxic zone and inside this region [4 , 50 ]. Although many studies explored the various chemokine-related signaling pathways, our study focused on the hypoxia-induced regulation of the end-point protein MMP-9 in monocytes, which actually degrades the ECM and allows cell migration. We showed that the direct inhibitory effect of hypoxia on secretion of monocyte MMP-9 correlates with their reduced migration, and addition of the antibody or the recombinant enzyme demonstrated a direct role of MMP-9 in the process. Thus, reduced MMP-9 may be one of the mechanisms by which hypoxia sends a "stop signal" to macrophages and causes their accumulation in hypoxic sites. However, the binding to the membrane via CD44 may also account for additional, unexplored, physiological effects. In the inflammatory tissue or the cancerous lesion, monocytes may cross through areas of divergent, low-oxygen tensions, moving back and forth from normoxic to hypoxic regions. In this context, it is interesting to ask if hypoxia primes the monocyte to secrete more or less MMP-9. This question, which requires that monocytes be subjected to reoxygenation or to cycles of hypoxia/reoxygenation, was not addressed in this study. However, we and others have previously looked into the effects of hypoxia/reoxygenation on other macrophage inflammatory molecules. For example, although hypoxia inactivated the production of nitric oxide (NO) in mouse macrophages, reoxygenation restored it to normoxic values [51 , 52 ]. Similarly, hypoxia abolished the membranal expression of CD80 on human and mouse macrophages, and reoxygenation restored this expression [53 ]. Conversely, although hypoxia and lipopolysaccharide synergistically enhanced the secretion of the proinflammatory cytokine TNF, reoxygenation reduced it to its normoxic levels [53 ]. In all these examples, reoxygenation had no effect on macrophage viability. Thus, a general scheme emerges, whereby reoxygenation works to reverse the effects of hypoxia on macrophages. Moreover, as in this study, we have demonstrated that the post-translational effects of hypoxia on the secretion of these molecules are, at least in part, mediated by an alteration of the cytoskeleton. For example, hypoxia disrupts the protein–protein interaction of inducible NO synthase (iNOS) with the cytoskeletal protein {alpha}-actinin 4 and interferes with its intracellular cortical localization [51 ]. Thus, we could argue that the final outcome for each inflammatory molecule may depend on its intracellular localization or its secretory pathway: The cytoplasmic enzyme iNOS may be directed to the cortical region by cytoskeletal proteins, and disruption of this localization may inhibit its activity, the cytoskeletal-dependent internalization of the costimulatory membranal CD80 may be influenced by hypoxia, and the effects of hypoxia on secreted molecules such as cytokines or proteases may depend on their secretion pathway, which is facilitated by cytoskeletal fibers, as some (e.g., TNF) are secreted via the endosomal pathway, and others (e.g., MMP-9) are secreted by secretory vesicles. Hence, vesicle trafficking becomes a novel target for hypoxic regulation and may be of great importance in determining the outcome of inflammatory processes or in tumor invasion [45 ]. The direct inhibitory effect of hypoxia on secretion of monocyte MMP-9 and its enhancing influence on binding to the cell surface, as observed in this study, may enable divergent effects on migration, as well as other monocytic activities, depending on interacting molecules in the microenvironment. In addition, these results stress the importance of the cytoskeleton and of post-translational mechanisms in the hypoxia-mediated regulation of monocyte MMP-9.


   ACKNOWLEDGEMENTS
 
This work was supported in part by the Rappaport Family Institute for Research in the Medical Sciences and by Research Grant 5343 from the Chief Scientist Office of the Israeli Ministry of Health. M. A. R. and H. B. are equally contributing senior authors.

Received June 7, 2005; revised October 4, 2005; accepted November 9, 2005.


   REFERENCES
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 REFERENCES
 

  1. Nathan, C. (2003) Oxygen and the inflammatory cell Nature 422,675-676[CrossRef][Medline]
  2. Sica, A., Saccani, A., Mantovani, A. (2002) Tumor-associated macrophages: a molecular perspective Int. Immunopharmacol. 2,1045-1054[CrossRef][Medline]
  3. Grimshaw, M. J., Blackwill, F. R. (2001) Inhibition of monocytes and macrophages chemotaxis by hypoxia and inflammation—a potential mechanism Eur. J. Immunol. 31,480-489[CrossRef][Medline]
  4. Murdoch, C., Giannoudis, A., Lewis, C. E. (2004) Mechanisms regulating the recruitment of macrophages into hypoxic areas of tumors and other ischemic tissues Blood 104,2224-2234[Abstract/Free Full Text]
  5. Lewis, J. S., Lee, J. A., Underwood, J. C. E., Harris, A. L., Lewis, C. E. (1999) Macrophage responses to hypoxia: relevance to disease mechanisms J. Leukoc. Biol. 66,889-900[Abstract]
  6. Sellebjerg, F., Sorensen, T. L. (2003) Chemokines and matrix metalloproteinase-9 in leukocyte recruitment to the central nervous system Brain Res. Bull. 61,347-355[CrossRef][Medline]
  7. Seiki, M., Koshikawa, N., Yana, I. (2003) Role of pericellular proteolysis by membrane-type 1 matrix metalloproteinase in cancer invasion and angiogenesis Cancer Metastasis Rev. 22,129-143[CrossRef][Medline]
  8. Nagase, H. (1997) Activation mechanisms of matrix metalloproteinases Biol. Chem. 378,151-160[Medline]
  9. Visse, R., Nagase, H. (2003) Matrix metalloproteinases and tissue inhibitors of metalloproteinases: structure, function, and biochemistry Circ. Res. 92,827-839[Abstract/Free Full Text]
  10. Mawrin, C., Brunn, A., Rocken, C., Schroder, J. M. (2003) Peripheral neuropathy in systemic lupus erythematosus: pathomorphological features and distribution pattern of matrix metalloproteinases Acta Neuropathol. (Berl.) 105,365-372[Medline]
  11. Nygardas, P. T., Hinkkanen, A. E. (2002) Up-regulation of MMP-8 and MMP-9 activity in the Balb/c mouse spinal cord correlates with the severity of experimental autoimmune encephalomyelitis Clin. Exp. Immunol. 128,245-254[CrossRef][Medline]
  12. Moon, S. K., Cha, B. Y., Kim, C. H. (2004) ERK1/2 mediates TNF{alpha}-induced matrix metalloproteinase-9 expression in human vascular smooth muscle cells via the regulation of NF{kappa}B and AP-1: involvement of the ras dependent pathway J. Cell. Physiol. 198,417-427[CrossRef][Medline]
  13. Ma, Z., Shah, R. C., Chang, M. J., Benveniste, E. N. (2004) Coordination of cell signaling, chromatin remodeling, histone modifications, and regulator recruitment in human matrix metalloproteinase 9 gene transcription Mol. Cell. Biol. 24,5496-5509[Abstract/Free Full Text]
  14. Overall, C. M., Lopez-Otin, C. (2002) Strategies for MMP inhibition in cancer: innovations for the post-trial era Nat. Rev. Cancer 2,657-672[CrossRef][Medline]
  15. Ogawa, K., Chen, F., Kuang, C., Chen, Y. (2004) Suppression of matrix metalloproteinase-9 transcription by transforming growth factor-ß is mediated by a NF{kappa}B site Biochem. J. 381,413-422[CrossRef][Medline]
  16. Vincenti, M. P., Brinckerhoff, C. E. (2002) Transcriptional regulation of collagenase (MMP-1, MMP-13) genes in arthritis: integration of complex signaling pathways for the recruitment of gene-specific transcription factors Arthritis Res. 4,157-164[CrossRef][Medline]
  17. Van den Steen, P. E., Dubois, B., Nelissen, I., Rudd, P. M., Dwek, R. A., Opdenakker, G. (2002) Biochemistry and molecular biology of gelatinase B or matrix metalloproteinase-9 (MMP-9) Crit. Rev. Biochem. Mol. Biol. 37,375-536[CrossRef][Medline]
  18. Lazure, C. (2002) The peptidase zymogen proregions: nature’s way of preventing undesired activation and proteolysis Curr. Pharm. Des. 8,511-531[CrossRef][Medline]
  19. Gardner, J., Ghorpade, A. (2003) Tissue inhibitor of metalloproteinase (TIMP)-1: the TIMPed balance of matrix metalloproteinases in the central nervous system J. Neurosci. Res. 74,801-806[CrossRef][Medline]
  20. Opdenakker, G., Van Den Steen, P. E., Dubois, B., Nelissen, I., Van Coillie, E., Masure, S., Proost, P., Van Damme, J. (2001) Gelatinase B functions as a regulator and effector in leukocyte biology J. Leukoc. Biol. 69,851-859[Abstract/Free Full Text]
  21. Fridman, R., Toth, M., Chvyrkova, I., Meroueh, S. O., Mobashery, S. (2003) Cell surface association of matrix metalloproteinase 9 (gelatinase B) Cancer Metastasis Rev. 22,153-166[CrossRef][Medline]
  22. Kolev, K., Skopal, J., Simon, L., Csonka, E., Machovich, R., Nagy, Z. (2003) Matrix metalloproteinase-9 expression in post-hypoxic human brain capillary endothelial cells: H2O2 as a trigger and NFkB as a signal transducer Thromb. Haemost. 90,528-537[Medline]
  23. Ben-Yosef, Y., Lahat, N., Shapiro, S., Bitterman, H., Miller, A. (2002) Regulation of endothelial matrix metalloproteinase-2 by hypoxia/reoxygenation Circ. Res. 90,784-791[Abstract/Free Full Text]
  24. Canning, M. T., Postovit, L. M., Clarke, S. H., Graham, C. H. (2001) Oxygen-mediated regulation of gelatinase and tissue inhibitor of metalloproteinase-1 expression by invasive cells Exp. Cell Res. 267,88-94[CrossRef][Medline]
  25. Zhong, H., Simons, J. W. (1999) Direct comparison of GAPDH, ß-actin, cyclophilin, and 28S rRNA as internal standards for quantifying RNA levels under hypoxia Biochem. Biophys. Res. Commun. 259,523-526[CrossRef][Medline]
  26. Gordillo, G. M., Sen, C. K. (2003) Revisiting the essential role of oxygen in wound healing Am. J. Surg. 186,259-263[CrossRef][Medline]
  27. Chase, A. J., Bond, M., Crook, M. F., Newby, A. C. (2002) Role of nuclear factor-{kappa}B activation in metalloproteinase-1, -3, and -9 secretion in human macrophages in vitro and rabbit foam cells produced in vivo Arterioscler. Thromb. Vasc. Biol. 22,765-771[Abstract/Free Full Text]
  28. Rosengart, M. R., Arbabi, S., Garcia, I., Maier, R. V. (2000) Interactions of calcium/calmodulin-dependent protein kinase (CaMK) and extracellular-regulated kinase (ERK) in monocyte adherence and TNF{alpha} production Shock 13,183-189[Medline]
  29. Olson, M. W., Toth, M., Gervasi, D. C., Sado, Y., Ninomiya, Y., Fridman, R. (1998) High-affinity binding of latent matrix metalloproteinase-9 to the {alpha}(IV) chain of collagen IV J. Biol. Chem. 273,10672-10681[Abstract/Free Full Text]
  30. Sultan, S., Gosling, M., Nagase, H., Powell, J. T. (2004) Shear stress-induced shedding of soluble intercellular adhesion molecule from saphenous vein endothelium FEBS Lett. 564,161-165[CrossRef][Medline]
  31. Yu, Q., Stamenkovic, I. (2000) Cell surface-localized matrix metalloproteinase-9 proteolytically activates TGF-ß and promotes tumor invasion and angiogenesis Genes Dev. 14,163-176[Abstract/Free Full Text]
  32. Bourguignon, L. Y., Gunja-Smith, Z., Iida, N., Zhu, H. B., Young, L. J., Muller, W. J., Cardiff, R. D. (1998) CD44 v(3,8-10) is involved in cytoskeleton-mediated tumor cell migration and matrix metalloproteinase (MMP-9) association in metastatic breast cancer cells J. Cell. Physiol. 176,206-215[CrossRef][Medline]
  33. Takahashi, C., Sheng, Z., Horan, T. P., Kitayama, H., Maki, M., Hitomi, K., Kitaura, Y., Takai, S., Sasahara, R. M., Horimoto, A., Ikawa, Y., Ratzkin, B. J., Arakawa, T., Noda, M. (1998) Regulation of matrix metalloproteinase-9 and inhibition of tumor invasion by the membrane-anchored glycoprotein RECK Proc. Natl. Acad. Sci. USA 95,13221-13226[Abstract/Free Full Text]
  34. Noda, M., Oh, J., Takahashi, R., Kondo, S., Kitayama, H., Takahashi, C. (2003) RECK: a novel suppressor of malignancy linking oncogenic signaling to extracellular matrix remodeling Cancer Metastasis Rev. 22,167-175[CrossRef][Medline]
  35. Rhee, J. S., Coussens, L. M. (2002) RECKing MMP function: implications for cancer development Trends Cell Biol. 12,209-211[CrossRef][Medline]
  36. Hahn-Dantona, E., Ruiz, J. F., Bornstein, P., Strickland, D. K. (2001) The low density lipoprotein receptor-related protein modulates levels of matrix metalloproteinase 9 (MMP-9) by mediating its cellular catabolism J. Biol. Chem. 276,15498-15503[Abstract/Free Full Text]
  37. Chintala, S. K., Kyritsis, A. P., Mohan, P. M., Mohanam, S., Sawaya, R., Gokaslan, Z., Yung, W. K., Steck, P., Uhm, J. H., Aggarwal, B. B., Rao, J. (1999) Altered actin cytoskeleton and inhibition of matrix metalloproteinase expression by vanadate and phenylarsine oxide, inhibitors of phosphotyrosine phosphatases: modulation of migration and invasion of human malignant glioma cells Mol. Carcinog. 26,274-285[CrossRef][Medline]
  38. Chintala, S. K., Sawaya, R., Aggarwal, B. B., Majumder, S., Giri, D. K., Kyritsis, A. P., Gokaslan, Z., Rao, J. (1998) Induction of matrix metalloproteinase-9 requires a polymerized actin cytoskeleton in human malignant glioma cells J. Biol. Chem. 273,13545-13551[Abstract/Free Full Text]
  39. Tomasek, J. J., Halliday, N. L., Updike, D. L., Ahern-Moore, J. S., Vu, T. K., Liu, R. W., Howard, E. W. (1997) Gelatinase A activation is regulated by the organization of the polymerized actin cytoskeleton J. Biol. Chem. 272,7482-7487[Abstract/Free Full Text]
  40. Olson, M. W., Bernardo, M. M., Pietila, M., Gervasi, D. C., Toth, M., Kotra, L. P., Massova, I., Mobashery, S., Fridman, R. (2000) Characterization of the monomeric and dimeric forms of latent and active matrix metalloproteinase-9. Differential rates for activation by stromelysin 1 J. Biol. Chem. 275,2661-2668<