| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
* Interdisciplinary Center for Clinical Research BIOMAT, Aachen, Germany;
Department of Surgery, University Hospital, RWTH Aachen, Germany;
Department of Medicine, Williams Middleton Memorial, University of Wisconsin, Madison, Wisconsin, USA;
Department of Medicine II, University Hospital, Mannheim, Germany;
|| Department of Medicine, San Francisco Veterans Affairs Medical Center, University of California, San Francisco, California, USA; and
¶ Department of Nephrology and Clinical Immunology, University Hospital, Aachen, Germany
1Correspondence: Interdisciplinary Center for Clinical Research BIOMAT, University Hospital, Pauwelsstr. 30, 52074 Aachen, Germany. E-mail: plynen{at}ukaachen.de
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Key Words: polymers • wound healing • cell-specific gene regulation • transgene
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
. Meshes are constructed of nonabsorbable (polypropylene (PP), polyvinylidenfluoride (PVDF), polytetraafluorethylene (PTFE), polyethylenterephthalate), absorbable polymers (polyglycolide/polylactid (PG)) or combinations thereof. The implantation of nonabsorbable polymeric "biomaterials" induces activation of cytokine cascades and proteases that exhibit characteristics of a chronic inflammatory reaction. The inflammatory reaction and the accompanying fibrosis are prerequisites for mesh fixation within the tissue and the reinforcement of the abdominal wall. However, the mesh-induced inflammatory reaction frequently causes complications, including seroma formation, mesh shrinkage and migration, adhesion, infection, and pain (2)
. Earlier studies have revealed that the extent of a foreign body reaction largely depends on the type of alloplastic material introduced (3)
. Absorbable meshes have been designed, e.g., polyglactin 910, which lose 50% of their mechanical stability within 3 wk and are degraded within 3 mo. A compromise taking advantage of both material properties has been developed in a composite polypropylene/polyglactin mesh. These optimized meshes exhibit a diminished foreign body reaction with improved biocompatibility (4)
.
In addition to this approach to optimize mesh integration and enforce the abdominal wall, an open question is whether there are alternate means of beneficially influencing the foreign body reaction. To address this, an in-depth understanding of the molecular mechanisms that guide the extent of the foreign body reaction is required. Common cellular components of such a reaction are infiltrating macrophages that are dispersed in the developing granuloma (5)
andthe presence of foreign body giant cells (FBGC). These large cells are formed by macrophage fusion (5)
. A stimulus for fusion may be the mesh filaments that are too large for phagocytosis. FBGCs, as well as macrophages, have the capability to synthesize a plethora of proinflammatory cytokines, including transforming growth factor-ß (TGF-ß), platelet-derived growth factor (PDGF), and vascular endothelial growth factor (VEGF). These factors are regarded as key players that direct the extent of organ fibrosis, particularly by influencing the phenotypic behavior of surrounding fibroblasts (6)
.
In addition to proinflammatory cytokines, enzymes involved in matrix degradation have an impact on the extent of a foreign body reaction. The matrix metalloproteinases (MMPs) are particularly relevant as they are abundant, have distinct substrate specificities, and induce a proteolytic cellular phenotype (7)
. A key role for MMP-2 (72 kDa collagenase, gelatinase A) in tissue remodeling, embryogenesis, and angiogenesis has been demonstrated (8, 9). In diseases such as arthritis (10)
, cancer (11)
, atheroma (12)
, and tissue ulceration (13)
, MMP-2 enzymatic activities are up-regulated at the sites of inflammation or cellular invasion. Beyond its capability to hydrolyze components of the extracellular matrix (ECM), MMP-2 directly affects cellular phenotypes, proliferation rates, and the inflammatory reaction (14, 15). MMP-2 transforms mesenchymal cells and fibroblasts to actively proliferating and migratory phenotypes in vitro (14, 16). Recently, our group observed a transition of renal tubular epithelial cells to a myofibroblastic phenotype that is dependent on the MMP-2 expression level (17)
. In an animal wound-healing model, blockage of MMP-2 enzymatic activity by the MMP inhibitor Ilomastat dampens the inflammatory cell infiltration (18)
. After acute myocardial infarction in mice, pharmacological inhibition of MMP-2 activity improves the survival rate by preventing cardiac rupture and delaying postinfarct remodeling through a reduction in macrophage infiltration (19)
. In chronic ulcers and wound fluids of hernia patients MMP-2 enzymatic activity has been detected, whereas it is absent in healthy skin (13)
. Furthermore, Bellon et al. (20)
described increased MMP-2 expression levels in primary fibroblasts obtained from patients with direct inguinal hernia compared to healthy controls.
The extent of MMP-2 expression and enzymatic activity is regulated at the transcriptional, translational, and post-translational levels (21
, 22
23
24
25
26)
and is closely coordinated with tissue inhibitors of matrix metalloproteases (TIMPs) expression (27
, 28)
. Specific regulatory elements governing MMP-2 gene transcription that reside up to –1686 bps relative of the translational start site have been identified (24
, 29
30
31
32
33)
. A strong enhancer element denoted response element-1 (RE-1) is located at –1322 /–1282 bps of the rat MMP-2 gene (24)
, which is evolutionarily conserved and is similarly operative within the human gene at –1657/–1619 bps relative to the transcriptional start site (34)
. At least five distinct transcription factors have been shown to bind to the RE-1. These include activating protein (AP)-2 (AP 2) (33)
, Y-box protein-1 (YB-1) (34)
, nonmetastasizing protein 23ß (nm23ß) (35)
, signal transduction and activator of transcription factor 3 (Stat3) (36)
, and p53 (31)
. Transcriptional regulation of MMP-2 synthesis in vivo is similarly coordinated by distinct subsets of transcription factors, as shown by a recent study on hind-limb ischemia (37)
.
Given the finding that MMP-2 plays a pivotal role in wound healing and the inflammatory response, we set out to unravel the molecular mechanisms that govern MMP-2 gene transcription following biomaterial implantation in vivo. Mesh materials were implanted in transgenic reporter mice harboring ß-galactosidase (LacZ) reporter genes driven by defined regulatory sequences of the MMP-2 gene. The spatial and temporal transcriptional regulation of the MMP-2 gene were analyzed at the cellular level and compared with protein synthesis and enzymatic activity.
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
. In brief, mouse strain F8 harbors a ß-galactosidase reporter gene under control of MMP-2 regulatory sequences –1686/+423, which extend to the middle of the second exon. To determine the significance of the RE-1 that extends from –1282/–1322, F8del mice were created that harbor MMP-2 regulatory sequence –1241/+423, thereby excluding the RE-1.
Study design
A total of 72 male MMP-2/LacZ transgenic CD1-tg mice (body wt 25 to 50 g) were included in this study. All studies were performed in accordance with the guidelines of the "Deutsche Tierschutzgesetz" (50.203.2-AC 46, 38/02) and the National Institutes of Health (Bethesda, MD, USA) guidelines for the use of laboratory animals. To analyze wound healing, a dermal full-thickness incision was performed followed by subcutaneously (s.c.) preparation without mesh implantation or by additional mesh implantation. The observation period was 7, 21, and 90 d. Meshes (n=6 for each mesh) were randomly implanted bilateral of the abdominal midline in F8 and F8del mice.
The following polymeric meshes (0.5x0.5 cm in size) were implanted: 1) short-term absorbable polyglactin 910 meshes (Vicryl®, Ethicon, Norderstedt, Germany) with glycolide and lactide at a ratio of 9:1 constructed of multifilaments weighing 53.6 g/m2; 2) nonabsorbable, heavy weight polypropylene meshes (Prolene®, Ethicon, Norderstedt, Germany) constructed of monofilaments weighing 108.5 g/m2; and 3) combined polypropylene and polyglactin 910 meshes (Vypro II®, Ethicon, Norderstedt, Germany) low-weight and multifilamentous structure weighing 83 g/m2 (PP 52 g/m2, PG 31 g/m2).
Operation
The operative procedure was performed under sterile conditions and general anesthesia by intramuscular (i.m.) administration of ketamine (10%; Sanofi-Ceva, Düsseldorf, Germany) and xylazine (Rompun 2%; Bayer, Leverkusen, Germany). Full- thickness dermal incisions extending over 1.5 cm were performed 1 cm bilateral of the abdominal midline. Polymers were implanted s.c. 1 cm distal to the xiphoid process. Skin closure was achieved with Prolene® 3/0 sutures. At 7, 21, and 90 d after implantation, animals were euthanized by isoflurane asphyxiation and decapitation. Explants were split and either fixed in 10% formalin for morphological analysis or snap-frozen and stored at –80°C for activity assays.
Histological and immunohistochemical analyses
For histological studies 3 µm tissue sections were stained with hematoxylin/eosin. Immunohistochemistry was performed using the avidin-biotin-complex and diaminobenzidine as chromogen. To detect cells with MMP-2 promoter activity, sections were incubated with monoclonal anti-ß-galactosidase antibody (Europa, Cambridge, UK; 1:1700) followed by biotinylated rabbit anti-mouse IgG (Dako, Gostrup Denmark, 1:300). Immunohistochemical staining was performed for MMP-2 (polyclonal rabbit, 1:1000; Biomol, Hamburg, Germany), 
-smooth muscle actin (SMA, polyclonal rabbit, Abcam, Cambridge, UK, 1:300), CD68 (polyclonal rabbit, 1:100; Santa Cruz Biotechnology, Santa Cruz, CA, USA), YB-1 [polyclonal rabbit (34)
], followed by incubation with goat anti-rabbit IgG antibody (1:500, Dako). Coexpression of ß-galactosidase with MMP-2, SMA, CD68, and YB-1 was performed by immunofluorescence using Alexa Fluor 488-conjugate goat anti-mouse IgG, donkey anti-goat IgG and goat anti-rabbit IgG secondary antibodies (Molecular Probes, Eugene, OR, USA). Semiquantitative analyses of immunohistochemical staining at mesh/tissue interfaces (area 100 µmx100 µm, 400-fold magnification) were performed with digital imaging analyzing software (Image-Pro Plus, Media Cybernetics, Silver Spring, MD, USA). For each mesh 90 measurements were obtained from 6 animals in each group. Control stains were performed by omitting the primary antibodies.
Zymography
Serial frozen sections were placed on MMP in situ ZymoFilms composed of a 7-µm-thick layer of gelatin on a polyester base (Wako Chemicals, Neuss, Germany) or MMP-PT in situ ZymoFilm, which contains the MMP inhibitor 1,10-phenanthroline. Films were incubated at 37°C for 24 h and stained with Biebrich Scarlet. Gelatin digestion is visualized as transparency. Two observers evaluated the gelatinolytic activity independently according to a scoring system: (–: negative; +: weak; ++: moderate; +++: strong) in six independent tissue samples for each subgroup and time point.
Plasmid constructs
The luciferase reporter plasmid pGL2Promoter (pGL2P, Promega, Madison, WI, USA) was used to examine RE-1-mediated reporter gene expression in mouse dermal fibroblasts. The pGL2P-RE-1 promoter construct contains the isolated RE-1 sequence in the context of a heterologous SV40 promoter (32)
.
Cell culture, transient cell transfection, and luciferase assay
NIH 3T3 mouse fibroblasts were grown in Dulbecco’s Modified Eagle’s Medium supplemented with 10% heat-inactivated FBS and 2 mM glutamine. Transient cell transfections were performed with NHDF NucleofactorTM Kit (Amaxa Biosystems, Köln, Germany). Twenty-four hours after transfection, cells were cultured in the presence of polypropylene mesh and/or cocultured with 106 mouse RAW 264.7 macrophages for 24 h. Thereafter, cells were lysed and luciferase activity was measured as reported (31)
.
Statistics
Statistical analyses were performed using the Statistical Package for the Social Sciences (SPSS). Data were organized according to intervention (sham, mesh), duration of implantation, and type of transgenic mice. Differences between the groups were evaluated for statistical significance by ANOVA. In case of comparison between several groups -rate adjustment according to Bonferroni was performed. Differences of ß-galactosidase expression in F8 and F8del mice were tested with the Mann-Whitney U nonparametric rank test. P-values <0.05 were considered to be significant. All data are presented as mean values ± SD.
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
representative findings for HE staining are depicted. A prominent inflammatory cell infiltrate was present in all subgroups encompassing the mesh implantation at day 7. By day 90 this infiltrate completely resolved in PG mesh recipients. In contrast, the nonabsorbable PP mesh material induced persistent inflammatory cell infiltrates and granuloma formation.
|
As these findings indicate a time-dependent influx of inflammatory cells, primarily macrophages and leukocytes, we set out to quantify their number by means of CD68 staining. In accord with Fig. 1
, a profound increase was found in macrophage number at day 7 in all subgroups receiving alloplastic mesh material, as analyzed by double-blinded visual count and as summarized in Fig. 2
. In the PG, PP and composite PP/PG mesh subgroups macrophage influx was significantly higher than in sham-operated animals (Fig. 2
, upper panel). Remarkably, at day 7, no significant difference was found between the mesh subgroups. At day 21 the overall number of infiltrating macrophages was unchanged in all subgroups, whereas at day 90 a two-fold increased number of macrophages in the PP and PP/PG groups could be detected in comparison to the PG and sham subgroups. These differences were statistically significant.
|
In addition to infiltrating macrophages, the number of fibroblasts within the scar tissue was analyzed (Fig. 2
, lower panel). At day 90 the fibroblast number was significantly higher in the PP subgroup when compared to the sham group (P=0.03). Such an increase of fibroblast number has previously been reported by Klinge et al. (38)
and is in accord with granuloma formation and fibrosis in humans with mesh engraftments, as well as experimental models. A tendency toward higher fibroblast number was also detected for the multifilamenteous PP/PG mesh subgroup; however, this did not reach statistical significance (39)
.
Polymeric mesh implantation induces MMP-2 protein expression and gelatinolytic activity
Wound healing and scar tissue formation encompass elevated synthesis rates of diverse ECM components, as well as increased ECM turnover. Our focus was to assess the spatial and temporal expression of MMP-2 and gelatinolytic activity. Immunohistochemistry for MMP-2 (brown color) is provided in Fig. 3
A, the corresponding morphometric analysis is shown in Fig. 3B
. Staining specificity was assessed with an irrelevant primary antibody yielding no background stain with the secondary antibody (Fig. 3A
, upper left). The vicinity of the mesh filaments enhanced MMP-2 protein expression over time (shown for PP, Fig. 3A
, lower panel). For the PG group this subsided within 90 d, whereas it was still present at day 90 with PP mesh filaments (Fig. 3B
).
|
To evaluate the overall gelatinolytic activity within the tissue specimens, in situ zymography was performed with fresh frozen tissue sections. Representative in situ zymograms for time points 7, 21, and 90 d are depicted in Fig. 3C
, and semiquantitative analyses are summarized in Table 1
. At day 7, moderate-to-strong gelatinolytic activity was observed in all subgroups within the wound bed. For the mesh groups it was especially abundant in the vicinity of the mesh filaments (Fig. 3A
, Table 1
). In sham-operated animals gelatinolytic activity was low at day 21 and no longer detected at day 90, as was with the PG mesh, and is likely due to degradation of the mesh filaments. For groups with PP mesh implantation, a pronounced gelatinolytic "corona" was present surrounding the filaments throughout the entire observation period up to day 90. MMP-inhibitor 1,10-phenanthroline blocked substrate turnover, underscoring the specificity of the detected enzymatic activity.
|
Cell-specific activation of the MMP-2 promoter by mesh filaments
It is known that MMP-2 protein synthesis is regulated at the level of transcription (40)
. To assess transcriptional activity conferred by sequences –1686/+423 (F8 animals) and –1241/+423 (F8del animals, RE-1 deleted) immunohistochemistry for ß-galactosidase was performed and analyzed by morphometry in both mouse strains. For F8 transgenic mice harboring the extended promoter sequences –1686/+423, representative images are shown in Fig. 4
A. In sham-operated animals only weak staining for ß-galactosidase was present at all examined time points. In subgroups with mesh implantation ß-galactosidase staining was detected in cell clusters within the granuloma and in single cells infiltrating the mesh pores. At day 90 staining for the PG subgroup was indistinguishable from the sham group, whereas PP- and PP/PG-related ß-galactosidase expression was observed at all time points. Morphometric analysis revealed significantly enhanced ß-galactosidase expression in PP and PP/PG mesh subgroups, compared with sham-operated mice at days 7 and 21. Induction of ß-galactosidase expression was significant for both strains, F8 and F8del, after mesh implantation (Fig. 4B
). As immunofluorescence double staining indicated a distinct expression of ß-galactosidase in macrophages and fibroblasts (Fig. 5
A, B), we additionally performed morphometric analyses with distinction of macrophages vs. fibroblasts. PP and PP/PG mesh-implantation significantly induced ß-galactosidase expression in both animal strains (Fig. 4C
, left panel). In these mesh subgroups, F8 animals exhibited significantly higher levels of ß-galactosidase expression in macrophages than F8del animals. These results indicate that PP-induced MMP-2 expression in macrophages is primarily conferred by sequences –1686/ –1241 bps, including the RE-1, whereas the remainder of the proximal promoter sequence has only marginal transcriptional activity.
|
|
In F8del animals the induction of ß-galactosidase expression by mesh materials was predominantly detected in fibroblasts (Fig. 4C
, right panel). Again there was a significant difference between the expression levels of both mouse strains. In contrast to our findings with macrophages, F8del animals exhibited significantly higher ß-galactosidase expression in fibroblasts. These observations are in accord with a repressive function of the sequences –1686/–1241 bps on MMP-2 gene transcription in these cells.
The location of fibroblasts and macrophages in relation to the foreign body determines their MMP-2 expression level
Cells in direct contact to the mesh filaments were, by morphological criteria, mostly fibroblasts. To assess their MMP-2 protein expression levels and determine whether the cells exhibited features of myofibroblasts, immunofluorescence staining for -smooth muscle actin and MMP-2 was performed. A lining of MMP-2-positive cells that "encapsulate" the PP filament was observed in F8 mice (Fig. 5A
, upper panel). Concurrently, these cells exhibited strong -smooth muscle actin expression (Fig. 5A
, lower panel). Staining for ß-galactosidase indicated that the MMP-2 promoter/lacZ gene is not active in these cells. These results indicate that myofibroblasts residing in the immediate vicinity of mesh filaments express large amounts of MMP-2 protein; however, the transcription rate driven by MMP-2 promoter sequences extending up to –1686 bps (including the RE-1) was low. Deletion of sequences –1686/–1241 bps actually enhanced MMP-2 gene transcription (compare Fig. 4C
).
To assess the transcriptional activity of the MMP-2 promoter sequences in macrophages from F8 and F8del animals, staining for CD68 and ß-galactosidase was performed. Macrophages were present throughout the foreign body granuloma in both animal strains and the number of CD68 positive macrophages decreased with distance from the mesh filaments (Fig. 5B
). Whereas macrophages expressed ß-galactosidase at high levels in F8 animals (Fig. 5B
, upper panel), no staining of CD68 and ß-galactosidase was detected in F8 del animals (Fig. 5B
, lower panel). To correlate these results with gelatinolytic activity, images of in situ zymography were analyzed at 400-fold magnification (Fig. 5C
). In tissue surrounding mesh filaments, isolated spots of gelatinolysis were present. These most likely correspond to the spatial distribution of macrophages that are positive for ß-galactosidase as well as CD68 (compare Fig. 5B
). These findings suggest that a subpopulation of CD68-positive macrophages in the vicinity of the mesh filaments exhibits strong ß-galactosidase and MMP-2 expression and transcription rates, which are dependent on promoter sequences extending up to –1686 bps relative to the translation start site, including the RE-1.
Macrophages inhibit MMP-2 promoter activity in fibroblasts via response element-1
The in vivo findings indicated that there is an intimate cell-cell crosstalk between macrophages and fibroblasts, as well as cells and mesh filaments. To analyze this crosstalk in isolation, an in vitro model system for fibroblasts (NIH 3T3 cells) and macrophages (RAW 264.7 cell line) was utilized. A cell culture system was adopted, which allowed for the direct contact of cells with meshes, e.g., polypropylene filaments. As the in vivo data indicated a distinct role of the RE-1 in fibroblasts (suppression) and macrophages (induction), the corresponding 40 bp sequence element was cloned into the SV40 promoter carrying pGL2Promoter reporter construct (denoted pGL2P-RE-1). Mouse fibroblasts (NIH 3T3) harboring the pGL2P-RE-1 construct exhibited a 7-fold enhanced relative luciferase activity when compared with cells carrying the empty control plasmid pGL2Promoter (Fig. 6
). To evaluate a crosstalk between macrophages and fibroblasts on the RE-1-dependent promoter activity, RAW 264.7 macrophages (106 cells per well) were added and incubated with fibroblasts for 24 h. This cocultivation led to down-regulation of RE-1-dependent gene transcription by more than 50%. Contact of NIH 3T3 fibroblasts with polypropylene mesh material had no significant effect on RE-1-dependent gene transcription, whereas the addition of RAW 264.7 macrophages significantly down-regulated RE-1-dependent gene transcription by 
60%. Taken together, the trans-repression of the RE-1 transcriptional activity indicates an intimate crosstalk of RAW 264.7 macrophages with fibroblasts. Fibroblast contact with polypropylene had no affect on the RE-1-dependent transcription rate.
|
Model of cell-specific zonal regulation of MMP-2 gene transcription
To summarize the in vitro and in vivo findings of RE-1-dependent, as well as RE-1-independent, MMP-2 gene transcription a model was derived that is depicted in Fig. 7
. This model concentrates on the two main cell types characterized in our study, fibroblasts and macrophages. For fibroblasts, a strong up-regulation of MMP-2 expression was observed in the immediate vicinity of the mesh materials PP, as well as PG, in zone 1. This MMP-2 expression did not coincide with up-regulated MMP-2 promoter/LacZ gene transcription rates. Indeed, the RE-1 conferred a repressive function on gene transcription in fibroblasts when they had intimate contact with macrophages. For fibroblasts that are not immediately adjacent to mesh filaments, low levels of MMP-2 expression were observed (compare Fig. 5A
), which are located in zone 2. As macrophages may trans-repress the MMP-2 gene transcription in fibroblast in vitro via the RE-1 (compare Fig. 6
) and the in vivo data indicated a general repressive function of RE-1 in fibroblasts (compare Fig. 4C
), it is assumed that the RE-1 conferred a repressive function in these fibroblasts.
|
For fibroblasts that are more distant to the mesh filaments, such a trans-repressive effect may be diminished (zone 3). This was not further evaluated in the present study.
For macrophages the data indicate strong up-regulation of MMP-2 gene transcription and protein expression in cells adjacent to the mesh filaments in vivo (compare Fig. 5B
). The spatial distribution of these macrophages corresponded to the spot-like gelatinolytic activity detected in zone 1. In macrophages residing in zones 1 and 2 the RE-1 conferred most of the transcriptional activity (compare Fig. 4C
). CD68-positive macrophages more distant to mesh filaments no longer exhibited trans-activation of the RE-1 element and are assigned to zone 3.
Expression of transcription factor YB-1 in macrophages with MMP-2 transcription
Our findings demonstrate that sequences –1686/–1241 bps including the RE-1 conferred most of the transcriptional activity in macrophages in the presence of mesh materials. Previously, we have shown that Y-box protein-1 (YB-1) is a major, cell type-specific trans-activator of MMP-2 transcription in vitro (32)
. To evaluate whether cells with high MMP-2 transcription rates also express YB-1, we performed immunofluorescence staining for CD68/YB-1, CD68/ß-galactosidase, as well as YB-1/ß-galactosidase. As described above, macrophages that were adjacent to the mesh filaments were immunopositive for ß-galactosidase (Fig. 8
, upper panel). Additionally, a strong CD68/YB-1 staining was detected (Fig. 8
, middle panel) as was also true for YB-1/ß-galactosidase. These findings indicate that YB-1 may be a crucial regulator of MMP-2 expression in activated macrophages in vivo, which is in accord with previous studies that tested the influence of YB-1 on gene expression in vitro (31, 41).
|
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
and fibroblast-mediated synthesis and remodeling of ECM. At the cellular level we could assess MMP-2 gene transcription, protein synthesis, and gelatinolytic activity, resulting in a model with zonal transcriptional regulation through distinct regulatory elements of the MMP-2 gene promoter. This model is summarized in Fig. 7
and best illustrates our findings that are only accessible through comparative analyses of MMP-2/LacZ transgenic mice harboring sequences –1686/+423 bps or a truncated form of the MMP-2 promoter. For fibroblasts from both transgenic strains with immediate contact to the biomaterial, a myofibroblastic phenotype was demonstrated by staining for -smooth muscle actin. These cells exhibit high-level synthesis of MMP-2 (zone 1). It is unclear whether these fibroblasts have an elevated MMP-2 transcription rate or whether post-transcriptional regulatory events take place (21, 22), considering that sequences –1686/+423 were not transcriptionally active in these cells. CD68-positive macrophages that are adjacent to the biomaterial also synthesized MMP-2 at high levels and exhibited a proteolytic phenotype. In these cells sequences –1686/–1241, including the RE-1, were mainly involved in the trans-activation of gene transcription; similar to the macrophages that are not immediately adjacent to the biomaterial but rather interspersed throughout the granuloma.
For fibroblasts that do not exhibit a myofibroblastic phenotype and that reside more distant to the mesh filaments, only minute amounts of MMP-2 synthesis were detected (Fig. 5A
, zone 2). LacZ gene transcription was diminished in these fibroblasts when gene regulation was controlled by the MMP-2 regulatory sequences –1686/–1241 (including the RE-1). These findings suggest a silencing effect of the RE-1 in fibroblasts that are located in zone 2. Intriguingly, the RE-1 enhanced LacZ gene transcription of macrophages in zone 2. Such a bifunctional role of a regulatory element as silencer or enhancer has previously been observed for the RE-1 in the glomerulum, where a single transcription factor, YB-1, may exert stimulatory and repressive functions via the same element (32)
. The most likely explanation relates to differences in partnering of YB-1 with coactivators of transcription, e.g., p53 (43)
. In zone 3, CD68-positive macrophages no longer synthesized excessive amounts of MMP-2 and the sequences –1686/+423 did not activate transcription in these cells.
Our findings clearly indicate that MMP-2 protein synthesis is highly regulated at the transcriptional level. The chosen antibody for immunohistochemistry detects pro-MMP-2 as well as active MMP-2 enzyme, therefore a distinction between protein synthesis and proteolytically active enzyme cannot be made. However, the results obtained by in situ zymography are consistent with those obtained by immunohistochemistry.
By means of a cell coculture model system with mouse 3T3 skin fibroblasts and macrophages, we could demonstrate a cellular crosstalk resulting in trans-repression of RE-1-dependent gene transcription in fibroblasts. Zhu et al. (44)
provided evidence for such a crosstalk in vitro: CD147-positive monocytes/macrophages that were isolated from patients suffering from rheumatic arthritis induced MMP-2 secretion and gelatinolytic activity in human dermal fibroblasts; however, MMP-2 transcription rates were not determined (ND). Eerola et al. (45)
have studied porcine heterotopic bronchial allografts to assess MMPs during obliterative bronchiolitis development. The authors observed intense matrix metalloproteinase-2 expression during the onset of inflammation and fibroproliferation in endothelial cells, fibroblasts, and macrophages. The observation that recovery was rapid in immunosuppressed allografts hints at a crucial role of inflammatory cells. These data further support the hypothesis of a crosstalk between macrophages and fibroblasts via MMP-2 regulation in conferring fibrosis.
MMP-2 is known to modulate tumor cell-host interactions, angiogenesis, and tumor growth (46)
. Though in vitro studies have suggested that MMP-2 is expressed by cancer cells, in situ studies have indicated that MMP-2 is mainly expressed by surrounding stroma cells (47)
. Analysis of fibroblastic cells in human carcinomas indicated that the expression pattern of the MMP-2 gene was closely related to that of the MT-MMP gene (47)
. Such findings support the hypothesis that proteolytic activities originating from the stromal component of human carcinomas have a critical role in tumor progression. In our hands polymeric meshes induced persistent transcriptional activation of the MMP-2 gene in macrophages. Previous studies have investigated MMP-2 protein and gene expression in monocytes/macrophages only in vitro. In THP-1 cell cultures, chitosan-DNA nanoparticles induced MMP-2 protein expression, although the effect was transient with a peak after 48 h (48)
. Divergent time courses of macrophage activation by biomaterials in vivo and in vitro may be due to differences in ECM composition, cytokine signaling, and cell activation (46)
.
We compared the monofilamenteous heavy wt PP mesh with the multifilamenteous, low wt combined PP/PG mesh. Though polyglactin is completely resorbed by day 90, we could not detect differences in macrophages and fibroblast numbers or in cell-specific MMP-2 transcription and enzyme activation between both groups. From previous studies it is known that the combined PP/PG mesh elicits a diminished foreign body reaction with improved biocompatibility regarding the whole tissue reaction (4)
. Our data indicate that cell-specific MMP-2 regulation within the foreign body granuloma is independent of mesh structure and wt and is simply guided by the mesh-host interface. Thus, the beneficial effect of PP/PG meshes may be due to the lower amount of implanted PP material.
In the future modulation of MMP-2 expression might represent a novel approach for reorganizing cell-cell interactions with the aim of optimizing mesh integration into the tissue environment. A key regulator and therapeutic target might be the Y-box protein-1 (YB-1) that binds to the RE-1 and cell-specifically trans-activates, as well as trans-represses, the MMP-2 promoter (32)
. Our findings suggest that YB-1 mediates the transcriptional activation in macrophages that are adjacent to the mesh filaments. It has to be proven if selective blockage of the cell type-specific trans-activator YB-1 dampens chronic foreign body reactions by down-regulating MMP-2 transcription in macrophages. Forced overexpression of YB-1 in the liver inhibited organ fibrosis in a model of CCl4-induced liver damage (49)
.
In conclusion, our model analyzing MMP-2 gene expression in transgenic mice revealed distinct MMP-2 promoter activation dependent on localization and cell types but independent of mesh structure. These findings highlight the complexity of MMP-2 gene regulation, a key molecule for matrix remodeling. The established transgenic reporter mice may serve as useful models for further in vivo testing of "optimized" biomaterials with the aim of disrupting the perpetual MMP-2 gene activation in macrophages as a consequence of chronic foreign body reaction. As hernia disease with mesh implantation is one of the most common surgical procedures, an in-depth understanding of the underlying molecular mechanisms activated in the foreign body reaction may help to devise improved tools and medications for better outcomes.
|
ACKNOWLEDGMENTS |
|---|
Received for publication June 26, 2006. Accepted for publication October 31, 2006.
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE |
