Published online before print January 3, 2008
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
* Institutes for Surgical Research and
Cardiovascular Diseases, University of Munich, Munich, Germany
1 Correspondence: Institute for Surgical Research, University of Munich, Marchioninistr. 27, D-81377 Munich, Germany. E-mail: christoph.reichel{at}med.uni-muenchen.de
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
or platelet-activating factor (PAF). Using zymography, we detected a significant elevation of MMP-9 activity in response to the stimuli applied, and MMP-2 expression was not altered. However, treatment with a specific MMP-2/-9 inhibitor significantly abrogated elevated MMP-9 activity. As observed by intravital microscopy, all inflammatory conditions induced a significant increase in numbers of adherent and transmigrated leukocytes (>80% Ly-6G+ neutrophils). Blockade of gelatinases significantly diminished I/R- and MIP-1-induced leukocyte adherence and subsequent transmigration, and upon stimulation with PAF, gelatinase inhibition had no effect on leukocyte adherence but selectively reduced leukocyte transmigration. Concomitantly, we observed an increase in microvascular permeability after I/R and upon stimulation with MIP-1 or PAF, which was almost completely abolished in the inhibitor-treated groups. Using immunofluorescence staining and confocal microscopy, discontinuous expression of collagen IV, a major substrate of gelatinases within the perivascular basement membrane (BM), was detected in postcapillary venules. Analysis of intensity profiles demonstrated regions of low fluorescence intensity, whose size was enlarged significantly after I/R and upon stimulation with MIP-1 or PAF as compared with unstimulated controls. However, this enlargement was abolished significantly after inhibition of gelatinases, respectively. In conclusion, these data demonstrate that gelatinases strictly regulate microvascular permeability and BM remodeling during the early inflammatory response, whereas concomitant leukocyte recruitment is mediated by these proteases in a stimulus-specific manner.
Key Words: leukocyte • transmigration • permeability • basement membrane • gelatinases
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
6β1-integrin, ICAM-2, endothelial cell-selective adhesion molecule (ESAM), CD99, cluster designation 99/L2 (CD99L2), junctional adhesion molecule-A (JAM-A), and JAM-C, are critically involved in leukocyte migration through endothelial cells in vivo [4
]. However, less is known about the mechanisms promoting the passage of leukocytes through the perivascular basement membrane (BM). In this context, PECAM-1 and the integrin 6β1, as well as leukocyte proteases including neutrophil elastase (NE) and matrix metalloproteinases (MMPs), are suggested to facilitate this final step [5
]. MMPs represent a family of zinc-dependent endopeptidases, which are characterized by their ability to degrade components of the extracellular matrix [6 ]. Previous reports have implicated MMPs in numerous physiological and pathophysiological processes including tissue remodeling, angiogenesis, as well as cell differentiation and migration. Thereof, MMP-2 (gelatinase A) and -9 (gelatinase B) constitute the subgroup of gelatinases. These proteases are expressed by different cell types, such as endothelial cells and leukocytes, and serve as cofactors for the degradation of collagen IV, the integral structural component of venular BMs [7 ].
A variety of studies has demonstrated that inhibition of gelatinase activity effectively attenuates leukocyte infiltration of inflamed tissue [8 9 10 11 12 ]. Recently, we were able to show that blockade of MMP-9 reduces neutrophil as well as T cell recruitment during ischemia-reperfusion (I/R) and ameliorates subsequent liver damage. In addition, we found that MMP-9 is required for the motility of interstitially migrating leukocytes [13 ]. However, it is noteworthy that in contrast to these studies, there are also some contentious findings. In different in vitro models, blockade of gelatinases did not exert any effects on leukocyte migration [14 , 15 ]. Moreover, in mouse models of glomerulonephritis and bacterial pneumonia as well as in experimental peritonitis, leukocyte recruitment to the sites of inflammation was not affected by inhibition of MMP-2 and/or -9, respectively [16 17 18 ].
Interestingly, there is a growing body of evidence that the mechanisms underlying leukocyte extravasation are governed by the nature of the inflammatory model used. It has been shown that adhesion molecules, such as PECAM-1, ICAM-2, 6β1-integrin, and JAM-A, mediate transmigration of leukocytes in a stimulus-specific manner ([19
20
21
22
]; listed in Table 1
). Whether gelatinases control leukocyte extravasation in dependency of the inflammatory stimulus has not yet been studied.
|
View this table: [in a new window] |
Table 1. Stimulus-Specific Regulation of Leukocyte Transmigration in vivo
|
Therefore, the objective of the present study was to evaluate the functional relevance of gelatinases for each single step of the leukocyte recruitment process under different inflammatory conditions in vivo as well as to directly investigate the role of gelatinases for collagen IV remodeling within the perivascular BM.
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Surgical procedure
The surgical preparation was performed as originally described by Baez [24
] with minor modifications. Mice were anesthetized using a ketamine/xylazine mixture (100 mg/kg ketamine and 10 mg/kg xylazine), administrated by i.p. injection. The left femoral artery was cannulated in a retrograde manner for continuous blood pressure monitoring and the administration of microspheres to the cremaster vasculature (see below). The right cremaster muscle was exposed through a ventral incision of the scrotum. The muscle was opened ventrally in a relatively avascular zone, using careful electrocautery to stop any bleeding, and spread over the transparent pedestal of a custom-made microscopic stage. Epididymis and testicle were detached from the cremaster muscle and placed into the abdominal cavity. Throughout the procedure as well as after surgical preparation during intravital microscopy, the muscle was superfused with warm-buffered saline.
Leukocyte recruitment to the cremaster muscle was induced using three different protocols. In a first set of experiments, ischemia of the cremaster muscle was induced by clamping all supplying vessels at the basis of the cremaster muscle using a vascular clamp (Martin, Tuttlingen, Germany) as described before [25
]. Stagnancy of blood flow was then verified by intravital microscopy. After 30 min of ischemia, the vascular clamp was removed, and reperfusion was restored for 130 min. In a second set of experiments, inflammation was induced by intrascrotal injection of MIP-1 (600 ng in 0.3 ml PBS, R&D Systems Europe Ltd., Wiesbaden, Germany). In a third set of experiments, PAF (Sigma-Aldrich, Deisenhofen, Germany) was added to the superfusion buffer at a final concentration of 100 nM.
Intravital microscopy
The setup for intravital microscopy was centered around an Olympus BX 50 upright microscope (Olympus Microscopy, Hamburg, Germany), equipped for stroboscopic fluorescence epi-illumination microscopy. Light from a 75-W xenon source was narrowed to a near-monochromatic beam of a wavelength of 700 nm by a galvanometric scanner (Polychrome II, TILL Photonics, Graefelfing, Germany) and directed onto the specimen via a FITC filter cube equipped with dichroic and emission filters (DCLP 500, LP515, Olympus). Microscopic images were obtained with Olympus water immersion lenses [20x/numerical aperture (NA) 0.5 and 10x/NA 0.3] and recorded with an analog black-and-white charge-coupled device (CCD) video camera (Cohu 4920, Cohu, San Diego, CA, USA) and an analog video recorder (AG-7350-E, Panasonic, Tokyo, Japan). Oblique illumination was obtained by positioning a mirroring surface (reflector) directly below the specimen and tilting its angle relative to the horizontal plane. The reflector consisted of a round cover glass (thickness, 0.19–0.22 mm; diameter, 11.8 mm), which was coated with aluminum vapor (Freichel, Kaufbeuren, Germany) and brought into direct contact with the overlying specimen as described previously [26
]. For measurement of centerline blood flow velocity, green fluorescent microspheres (0.96 µm diameter, Molecular Probes, Leiden, The Netherlands) were injected via an arterial catheter, and their passage through the vessels of interest was recorded using the FITC filter cube under appropriate stroboscopic illumination (exposure, 1 ms; cycle time, 10 ms; 
=488 nm), integrating video images for sufficient time (>80 ms) to allow for the recording of several images of the same bead on one frame. Beads that were flowing freely along the centerline of the vessels were used to determine blood flow velocity (see below).
Quantification of leukocyte kinetics and microhemodynamic parameters
For off-line analysis of parameters describing the sequential steps of leukocyte extravasation, we used the Cap-Image image analysis software (Dr. Zeintl, Heidelberg, Germany). Rolling leukocytes were defined as those moving slower than the associated blood flow and quantified as described previously [27
]. Briefly, the rolling leukocyte flux fraction was determined from video recordings by counting all visible cells passing through a plane perpendicular to the vessel axis and dividing this number by the total leukocyte flux through the vessel, which can be estimated by the product of the systemic leukocyte count, mean blood flow velocity, and estimated vessel cross-sectional area. Firmly adherent cells were determined as those resting in the associated blood flow for more than 30 s and related to the luminal surface per 100 µm vessel length. Emigrated cells were counted in regions of interest (ROI), covering 75 µm on both sides of a vessel over 100 µm vessel length. By measuring the distance between several images of one fluorescent bead under stroboscopic illumination, centerline blood flow velocity was determined. From measured vessel diameters and centerline blood flow velocity, apparent wall shear stress was calculated, assuming a parabolic flow velocity profile over the vessel cross-section.
Experimental groups
Animals were assigned randomly to the following groups: sham-operated mice as well as mice receiving the MMP-2/-9 inhibitor III (Calbiochem, Darmstadt, Germany; 0.5 mg in 200 µl PBS, 5 min prior to onset of reperfusion/stimulation with MIP-1 or PAF) or vehicle (n=6 each group) undergoing I/R as well as stimulation with MIP-1 or PAF.
MMP-2/-9 inhibitor
The MMP-2/-9 inhibitor III is a synthetic cyclic decapeptide H-Cys-Thr-Thr-His-Trp-Gly-Phe-Thr-Leu-Cys-OH (also called CTTHWGFTLC or CTT peptide), which selectively inhibits MMP-2 (IC50, 10 µM) and MMP-9 activity (IC50, 10 µM) but does not affect activity of other MMPs including membrane-type 1-MMP, MMP-8, or MMP-13. Furthermore, the peptide has not been found to exhibit any apparent toxicity [28
]. The MMP-2/-9 inhibitor has been demonstrated in vitro to inhibit migration of tumor and endothelial cells. In vivo, this peptide binds to tumors, suppresses tumor formation, targets tumor vasculature, and prolongs survival of tumor-bearing mice, as well as reduces leukocyte recruitment under different inflammatory conditions [13
, 28
29
30
]. In addition, this inhibitor has been used to determine gelatinase activity in tissue samples using in situ zymography and to analyze the contribution of gelatinases in various biological processes including vasoconstriction and epithelial-mesenchymal transition [31
32
33
34
]. Although the mechanism for how the peptide inhibits gelatinase activity is not completely understood, it is suggested that the Trp residue in the histidine-tryptophane-glycine-phenylalanine (HWGF) motif may bind to the hydrophobic pocket of the substrate cleft in the gelatinases and that the His residue may act as a ligand for the catalytic Zn2+ ion [28
]. The dose used in our experiments has previously been shown to be effective in mice [13
, 28
].
Experimental protocols
I/R
Three postcapillary vessel segments in a central area of the spread-out cremaster muscle were randomly chosen among those that were at least 150 µm away from neighboring postcapillary venules and did not branch over a distance of at least 150 µm. After having obtained baseline recordings of leukocyte rolling, firm adhesion, and emigration in all three vessel segments, the MMP-2/-9 inhibitor was applied. Five minutes later, ischemia was induced for 30 min. Measurements, which took 
5 min, respectively, were repeated at 5, 30, 60, 90, and 120 min after onset of reperfusion, and blood flow velocity was determined at 125 min after reperfusion as described above.
Intrascrotal injection of MIP-1
Five minutes prior to intrascrotal injection of MIP-1, the MMP-2/-9 inhibitor was administered. After 3 h of stimulation with MIP-1, five vessel segments were randomly chosen as described above, and recordings of migration parameters were obtained.
Superfusion of PAF
Three postcapillary vessel segments were randomly chosen as described above. After having obtained baseline recordings, the MMP-2/-9 inhibitor was administrated. After 5 min, inflammation was induced by adding PAF to the superfusion buffer. Measurements were repeated every 30 min for up to 120 min.
After intravital microscopy, tissue samples of the cremaster muscle were taken for immunohistochemistry, confocal laser-scanning microscopy, as well as for zymography (see below). Blood samples were collected by cardiac puncture for the determination of systemic leukocyte counts using a Coulter ACT Counter (Coulter Corp., Miami, FL, USA). Animals were then killed by bleeding to death.
Microvascular permeability
Analysis of microvascular permeability was performed according to previous protocols with minor modifications [35
, 36
]. Briefly, the macromolecule FITC-dextran (5 mg in 0.1 ml saline, Mr150,000, Sigma-Aldrich) was infused intra-arterially after 30 min of ischemia and 90 min of reperfusion, after 150 min of stimulation with MIP-1, as well as after 90 min of stimulation with PAF, respectively. Five postcapillary vessel segments as well as the surrounding perivascular tissue were excited at 488 nm, and emission >515 nm was recorded by a CCD camera (Sensicam, PCO, Kelheim, Germany) 30 min after injection of FITC-dextran using an appropriate emission filter (LP 515). Mean gray values of fluorescence intensity were measured by digital image analysis (TILLvisION 4.0, TILL Photonics) in six randomly selected ROIs (50x50 µm2), localized 50 µm distant from the postcapillary venule under investigation. The average of mean gray values was calculated.
Immunohistochemistry
To determine the phenotype of transmigrated leukocytes, immunostaining of paraffin-embedded serial tissue sections of the cremaster muscle was performed. Sections were incubated with primary rat anti-mouse anti-Ly-6G, anti-CD45 (BD Biosciences, San Jose, CA, USA), or anti-F4/80 (Serotec, Oxford, UK) IgG antibodies. Afterwards, the paraffin sections were stained with commercially available immunohistochemistry kits (Ly-6G, CD45, Super Sensitive Link-Label IHC detection system, BioGenex, San Ramon, CA, USA; F4/80, Vectastain ABC kit, Vector Laboratories, Burlingame, CA, USA), obtaining an easily detectable reddish or brownish endproduct, respectively. Finally, the sections were counterstained with Mayer’s hemalaun. The number of extravascularly localized Ly-6G-, CD45-, or F4/80-positive cells was quantified by light microscopy (magnification 400x) on three sections (10 observation fields per section) from three individual animals per experimental group in a blinded manner, respectively. The number of transmigrated Ly-6G-positive cells (neutrophils) and F4/80-positive cells (monocytes/macrophages) is expressed as the percentage of total CD45-positive leukocytes.
Immunostaining and analysis of tissues by confocal microscopy
For the analysis of collagen IV expression, cremaster muscles were fixed in 4% paraformaldehyde. Tissues were then blocked and permeabilized in PBS, supplemented with 10% goat serum (Sigma-Aldrich) and 0.5% Triton X-100 (Sigma-Aldrich). After incubation with the primary rabbit anti-mouse collagen IV polyclonal antibody (Abcam, Cambridge, UK) at room temperature for 12 h, tissues were incubated with the secondary Alexa Fluor 488-linked goat anti-rabbit (Invitrogen, Carlsbad, CA, USA) antibody for 3 h at room temperature. Immunostained tissues were mounted in PermaFluor (Beckman Coulter, Fullerton, CA, USA) on glass slides and observed using a Leica SP5 confocal laser-scanning microscope (Leica Microsystems, Wetzlar, Germany) with an oil immersion lens (Leica; 40x/NA 1.25–0.75). Optical sections of tissue samples through the whole depth of the tissue were obtained using, as far as possible, the same settings for all samples analyzed. Z-stack digital images were collected optically at every 0.5 µm depth and applied to three-dimensional (3D) reconstruction analysis using Leica Application Suite software. To analyze the expression profile of collagen IV, 3D images of vessels were split in the middle along the longitudinal axis. Images of these "semi-vessels" were then analyzed for fluorescence intensity, as described elsewhere, using Leica Application Suite software [23
]. Briefly, ROIs within 3D images of semi-vessels were identified manually, and their intensity profile was compared with the average intensity of the entire vessel within the same field of view. Collagen IV LE sites were defined as those regions in which the average fluorescence intensity/unit area was less than 60% of the average fluorescence intensity in the whole vessel segment under investigation. LE sites from five vessel segments/tissue (n=4 mice per group) were analyzed. LE site size was determined using Leica Application Suite software, and LE site density was calculated for the total surface area of the semi-vessels.
Zymography
Metalloproteinase activity in the cremaster muscle was determined by enzyme zymography as described previously. Briefly, isolated cremaster tissue was immersed in liquid nitrogen for 1 min and rethawed at 4°C. Samples were homogenized in a cooled mortar, taken up in 100 µl Laemmli sample buffer, and incubated for 1 h at 4°C with mixing. Samples were electrophoresed in a 7.5% polyacrylamide gel containing 0.1% gelatin. To remove SDS, the gel was washed four times for 30 min each in washing buffer (50 mM Tris-HCl, pH 7.5; 10 mM CaCl2; 2.5% Triton X-100). The gel was incubated in incubation buffer (50 mM Tris-HCl, pH 7.5; 150 mM NaCl; 10 mM CaCl2) at 37°C overnight with shaking and subsequently stained with Coomassie blue. Murine MMP-2 and -9 (active enzyme and proenzyme, Calbiochem) were used to identify the gelatinolytic bands on the gels. The presence of gelatinolytic activity was identified as clear bands on a blue background after destaining. The results are presented as OD units. On our gels, primarily, the 105-kDa latent form of MMP-9 was detected. Only very little of the truncated form of MMP-9 associated with proteolytic activity in vitro was observed. These in vivo findings are frequently reported and are not inconsistent with functionality of MMP-9 as a protease in vivo [37
38
39
40
]. In this context, enzymatic activation of MMP-9 proenzyme is suggested to occur by binding at the cell surface, and activated MMP-9 is then degraded rapidly to prevent excess activity [41
]. Therefore, only low levels of active MMP-9 may be present in the cremasteric tissue.
Statistics
Data analysis was performed with a statistical software package (SigmaStat for Windows, Jandel Scientific, Erkrath, Germany). The ANOVA On Ranks test followed by the Student-Newman-Keuls test was used for the estimation of stochastic probability in intergroup comparisons. Mean values and SEM are given. P values <0.05 were considered significant.
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
. In contrast, systemic leukocyte counts were increased significantly in animals stimulated with PAF as compared with control. In addition, stimulation with PAF induced a significant decrease in blood flow velocity and wall shear rate as compared with unstimulated controls, and inner vessel diameters did not differ among groups. However, treatment with the MMP-2/-9 inhibitor did not significantly alter elevated systemic leukocyte counts as well as decreased blood flow velocities and shear rates in animals stimulated with MIP-1 or PAF. |
View this table: [in a new window] |
Table 2. Systemic Leukocyte Counts and Microhemodynamic Parameters
|
or stimulation with PAF as compared with controls. However, this increase in MMP-9 activity was diminished significantly in the inhibitor-treated groups, respectively.
 View larger version (13K): [in a new window] |
Figure 1. Activity of MMP-2 and MMP-9. Representative zymogram showing activity of MMP-2 and MMP-9 in sham-operated mice as well as mice treated with a MMP-2/-9 inhibitor or vehicle after I/R (30 min/120 min; A). Panels show results for MMP-2 (B) and MMP-9 activity (C) in sham-operated mice as well as in mice treated with a MMP-2/-9 inhibitor or vehicle after I/R (30/120 min) and upon stimulation with MIP-1 or PAF (mean±SEM for n=4 per group; #, P<0.05, vs. sham; *, P<0.05, vs. vehicle).
|
or PAF. In a first set of experiments, leukocyte responses were analyzed after I/R. The surgical procedure induced leukocyte rolling in postcapillary venules of the cremaster muscle (Fig. 2A ), which remained on this level throughout the entire reperfusion phase. No statistically significant differences were detected among groups.
 View larger version (33K): [in a new window] |
Figure 2. Leukocyte recruitment during I/R and upon stimulation with MIP-1. Leukocyte rolling, firm adherence, and transmigration were quantified in postcapillary venules of the cremaster muscle at baseline conditions prior to 30 min of ischemia as well as at 5, 30, 60, 90, and 120 min of reperfusion (A, C, and E) or after 180 min of stimulation with the CC chemokine MIP-1 (B, D, and F). Panels show results for sham-operated mice as well as for mice treated with a MMP-2/-9 inhibitor or vehicle at representative time-points (mean±SEM for n=6 per group; #, P<0.05, vs. sham; *, P<0.05, vs. vehicle).
|
After surgical preparation, the number of emigrated leukocytes was low and did not differ among groups. In contrast, after 120 min of reperfusion, there was a marked increase in numbers of transmigrated leukocytes in mice treated with vehicle as compared with sham-operated, control mice (29.4±1.5/104 µm2 vs. 14.9±1.4/104 µm2; Fig. 2E ). However, treatment with the inhibitor significantly diminished this increase to the level of sham-operated mice (17.2±3.2/104 µm2). Similar results for leukocyte firm adherence and transmigration were obtained after 60 min of reperfusion (Fig. 2C and E) .
Leukocyte recruitment upon stimulation with MIP-1
In a second set of experiments, leukocyte migration parameters were analyzed 3 h after intrascrotal injection of the CC chemokine MIP-1. No significant differences were detected in numbers of rolling leukocytes among experimental groups (Fig. 2B)
. In contrast, after stimulation with MIP-1, the number of firmly adherent (15.9±1.1/104 µm2 vs. 5.0±0.3/104 µm2; Fig. 2D
) and transmigrated (16.0±2.5/104 µm2 vs. 4.4±0.5/104 µm2; Fig. 2F
) leukocytes was elevated significantly in mice treated with vehicle as compared with controls. In the treated group, however, leukocyte adherence (9.4±0.4/104 µm2) and transmigration (7.7±0.6/104 µm2) were significantly attenuated.
Leukocyte recruitment upon stimulation with PAF
Finally, leukocyte responses were evaluated after stimulation with the chemoattractant PAF. Prior to the induction of inflammation, no significant differences were observed in numbers of rolling, adherent, and transmigrated leukocytes among groups (Fig. 3A
3B
3C
). However, in mice treated with vehicle (3.7±0.8%), the leukocyte rolling flux fraction was significantly lower after 120 min of stimulation with PAF than at baseline conditions (11.3±1.4%). Furthermore, stimulation with PAF significantly increased leukocyte firm adherence (10.2±1.5/104 µm2 vs. 3.7±0.7/104 µm2) and transmigration (22.8±2.5/104 µm2 vs. 9.5±2.4/104 µm2) in mice treated with vehicle as compared with controls. Interestingly, inhibition of gelatinases had no effect on leukocyte rolling flux fraction (3.4±0.7%) as well as on leukocyte firm adherence (12.9±1.4/104 µm2) after stimulation with PAF but significantly reduced the number of transmigrated leukocytes (17.2±1.1/104 µm2). Comparable results for leukocyte firm adherence and transmigration were obtained after 60 min of stimulation with PAF (Fig. 3B
and C)
.
 View larger version (17K): [in a new window] |
Figure 3. Leukocyte recruitment upon stimulation with PAF. Leukocyte rolling (A), firm adherence (B), and transmigration (C) were quantified in postcapillary venules of the cremaster muscle at baseline conditions as well as after 30, 60, 90, and 120 min of stimulation with the chemoattractant PAF. Panels show results for sham-operated mice as well as for mice treated with a MMP-2/-9 inhibitor or vehicle at representative time-points (mean±SEM for n=6 per group; #, P<0.05, vs. sham; *, P<0.05, vs. vehicle).
|
or PAF, over 80% of emigrated leukocytes were Ly-6G-positive neutrophils, and 10–20% were F4/80-positive monocytes/macrophages. No statistically significant differences were detected among experimental groups.
Microvascular permeability
As a measure of microvascular permeability, leakage of the macromolecule FITC-dextran into the perivascular space was determined. In mice treated with vehicle, there was a significant increase in the leakage of FITC-dextran observed after I/R (twofold; Fig. 4A
) as well as upon 180 min of stimulation with MIP-1 (2.5-fold; Fig. 4B
) or 120 min of stimulation with PAF (fivefold; Fig. 4C
) as compared with controls. In contrast, treatment with the MMP-2/-9 inhibitor almost completely abolished FITC-dextran leakage to the level of control animals, respectively.
 View larger version (15K): [in a new window] |
Figure 4. Microvascular permeability. As a measure of microvascular permeability, leakage of FITC-dextran into the perivascular space was quantified. Panels show results for sham-operated mice as well as for mice treated with a MMP-2/-9 inhibitor or vehicle after I/R (A; 30/120 min) and upon stimulation with MIP-1 (B) or PAF (C; mean±SEM for n=6 per group; #, P<0.05, vs. sham; *, P<0.05, vs. vehicle).
|
(14.0±0.8 µm2) or PAF (14.9±1.4 µm2) induced a significant increase in the average size of collagen IV LE sites as compared with unstimulated controls. This increase was significantly abolished to the level of unstimulated, control animals after inhibition of gelatinases, respectively. However, in unstimulated, control animals, 26.5 ± 5.2% of collagen IV LE sites were larger than 10 µm2 (Fig. 5C)
. In response to I/R (58.8±6.0%) as well as upon stimulation with MIP-1 (54.0±2.1%) or PAF (56.5±4.2%), the proportion of collagen IV LE sites larger than 10 µm2 to all collagen IV LE sites detected was increased significantly. In contrast, in inhibitor-treated animals, this elevation was diminished significantly to the level of unstimulated, control animals. Finally, no statistically significant differences were found in the density of collagen IV LE sites among experimental groups.
 View larger version (22K): [in a new window] |
Figure 5. Expression of collagen IV. Representative image of a postcapillary venule in the postischemic cremaster muscle immunostained for collagen IV. White rings show the position of selected LE sites (A). Using immunofluorescence staining and confocal microscopy, expression profiles in terms of size (B), size distribution (C), and density (D) of collagen IV LE sites were determined. Panels show results for unstimulated, control mice as well as for mice treated with a MMP-2/-9 inhibitor or vehicle after I/R (30/120 min) and upon stimulation with MIP-1 or PAF (mean±SEM for n=4 per group; #, P<0.05, vs. sham; *, P<0.05, vs. vehicle).
|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
6β1-integrin, and JAM-A have been reported to mediate leukocyte transmigration in a stimulus-specific manner in vivo (Table 1)
, the objective of our study was to explore whether gelatinases exhibit a similar profile of stimulus specificity in this context.
Therefore, the role of gelatinases in the leukocyte recruitment process was investigated in response to I/R as well as upon stimulation with the CC chemokine MIP-1 or the chemoattractant PAF using a specific MMP-2/-9 inhibitor [28
]. In the cremaster muscle of control mice, we detected LE of both gelatinases. However, in response to I/R as well as upon stimulation with MIP-1 or PAF, MMP-9 activity was enhanced significantly, whereas expression of MMP-2 was not altered. Our data are in line with recent findings describing constitutive expression of MMP-2 as well as rapid up-regulation of MMP-9 in the initial inflammatory response [7
, 42
, 43
]. Treatment with the MMP-2/-9 inhibitor significantly abolished the elevation of MMP-9 activity under the inflammatory conditions applied and did not significantly alter low constitutive activity of MMP-2.
Using RLOT intravital microscopy, we found that blockade of gelatinases did not affect leukocyte rolling in response to I/R as well as to stimulation with MIP-1. In contrast, in the postischemic liver, MMP-9 blockade attenuated leukocyte rolling by diminishing P-selectin translocation [13
]. The reason for these inconsistent findings is unclear and may be related to tissue-specific differences.
As the second step in the sequential process of leukocyte recruitment, I/R- as well as MIP-1-elicited leukocyte firm adherence was diminished significantly after inhibition of gelatinases. These results confirm our previous findings in the postischemic liver, as leukocyte accumulation in hepatic sinusoids was significantly attenuated after inhibition of MMP-9 [13
]. The observed effects can be explained by the ability of gelatinases to convert big endothelin-1 to endothelin-1, which in turn, up-regulates the expression of adhesion molecules such as ICAM-1 on endothelial cells and the β2-integrin membrane antigen-1 on the surface of neutrophils, finally leading to increased firm adherence of neutrophils [44
].
Consequently, I/R- as well as MIP-1-induced transmigration of neutrophils and monocytes was nearly abolished after blockade of increased gelatinase activity. These data are in agreement with several studies demonstrating a crucial role of gelatinases for the regulation of neutrophil extravasation [8
9
10
11
12
]. In this context, the ability of gelatinases to degrade collagen IV, the integral structural component of BM, had been suggested to be the key mechanism promoting gelatinase-dependent migration of leukocytes through venular walls. Recently, Nourshargh’s group [23
] identified regions within the perivascular BM, where the expression of collagen IV and other BM constituents is lower than the average vascular level. These LE areas are thought to be preferentially used by transmigrating leukocytes. Moreover, it has been shown that stimulation with IL-1β induces a transient enlargement of collagen IV LE sites, which is dependent on NE [23
]. In the present study, we demonstrate that in the early postischemic inflammatory response, as well as upon stimulation with the CC chemokine MIP-1, the vast majority of collagen IV LE sites within the perivascular BM is enlarged significantly. In addition, we found that inhibition of enhanced gelatinase activity significantly reduces the average size of these collagen IV LE sites, suggesting a crucial role of gelatinases in postischemic as well as MIP-1-induced remodeling of the venular BM. However, to which extent remodeling of the BM and the consequences of diminished leukocyte firm adherence contribute to gelatinase-dependent extravasation of neutrophils and monocytes after I/R or stimulation with MIP-1 cannot clearly be answered.
In a further set of experiments, leukocyte responses were analyzed upon stimulation with the chemoattractant PAF. Similar to the observations in response to I/R as well as to stimulation with MIP-1, inhibition of gelatinase activity had no effect on leukocyte rolling in PAF-stimulated cremasteric venules. In contrast, leukocyte firm adherence was not affected by blockade of gelatinases. Moreover, whereas transmigration of neutrophils in response to I/R (>85%) as well as stimulation with MIP-1 (>70%) were almost completely abolished, neutrophil extravasation upon stimulation with PAF was only attenuated by 30–40%. As we observed similar effects over the entire time course during I/R- as well as PAF-induced leukocyte extravasation, respectively, we conclude that our intravital microscopic findings depend on the nature of the respective inflammatory condition but not on the duration of the inflammatory response. In addition, the average size of collagen IV LE sites was elevated significantly upon stimulation with PAF as compared with unstimulated controls. In inhibitor-treated mice, however, this increase was diminished significantly, demonstrating again a crucial role of gelatinases in collagen IV remodeling during acute inflammatory processes. It is noteworthy that in unstimulated, control animals as well as in inhibitor-treated animals, a small proportion of collagen IV LE sites was still larger than the average vascular level. These LE sites might be constitutively "enlarged" and might thereby serve as exit points for leukocytes transmigrating independently of gelatinases. Taken together, these data suggest that gelatinases play a pivotal role for BM remodeling in the early inflammatory response and regulate the recruitment process of neutrophils and monocytes in a stimulus-specific manner in vivo.
Interestingly enough, our intravital microscopic observations are similar to recent findings describing a stimulus-specific control of leukocyte transendothelial migration by the adhesion molecules PECAM-1, ICAM-2, 6β1-integrin, and JAM-A (Table 1)
. The mechanisms underlying these observations are still unclear.
However, there is a growing body of evidence that inflammatory stimuli differentially induce extravasation of leukocytes. Thereby, chemoattractants, including PAF, LTB4, and TNF-, are thought to directly activate leukocytes [20
, 45
, 46
], and stimuli, such as IL-1β and I/R, might directly activate endothelial cells [47
48
49
]. Moreover, MMPs have been reported to modify nonmatrix substrates, such as cytokines and chemokines. In this context, MMP-9 has been shown to increase the chemotactic potency of certain chemoattractants (e.g., CXCL-8), whereas others become inactivated by this protease [50
]. Consequently, the existence of stimulus-dependent expression patterns of adhesion molecules as well as modulation of cytokine activity by gelatinases might provide the molecular basis for our intravital-microscopic observations.
In addition, a rather simple but potentially conceivable explanation for the stimulus-specific regulation of leukocyte transmigration would be the ability of distinct inflammatory stimuli to differentially disintegrate endothelial junctions. Thereby, certain cytokines are suggested to induce a rather low enlargement of endothelial junctions and with respect to earlier observations, engage adhesion molecules, including PECAM-1, ICAM-2, 6β1-integrin, and JAM-A, to actively facilitate leukocyte transmigration. In contrast, other chemoattractants might cause a profound disrupture of endothelial junctions, allowing a passive efflux of leukocytes independently of the control by those adhesion molecules.
As a measure for junctional integrity, changes in microvascular permeability were analyzed under the inflammatory conditions studied before. We could demonstrate that in response to I/R as well as MIP-1, microvascular permeability was just slightly enhanced, reaching approximately twice the level of control conditions. In contrast, stimulation with PAF caused a massive increase in microvascular permeability, which was more than twofold higher than the increase induced by I/R or MIP-1. These data support the concept of a stimulus-dependent disintegration of endothelial junctions. However, it is not clear whether leukocytes just passively extravasate through these opened junctions. Interestingly, gelatinases have been reported to degrade basal lamina proteins such as collagen IV, fibronectin, laminin, and heparane sulfate as well as to cleave junctional proteins including occluding, claudin-5, and vascular endothelial-cadherin [51
, 52
]. Therefore, we tested the effect of the MMP-2/-9 inhibitor on the regulation of microvascular permeability. We found that inhibition of gelatinases completely abolished the I/R- as well as MIP-1-induced elevation in microvascular permeability, suggesting a crucial role of this protease for the modulation of junctional integrity during the early inflammatory response. These results are in line with previous observations, as MMP-9 was critically involved in the regulation of microvascular permeability in different organs including lung, kidney, and brain [12
, 53
, 54
]. In contrast, inhibition of MMP-9 rather increased vasopermeability in a model of acute peritonitis [55
]. The reason for these inconsistent findings is not clear.
Interestingly, the massive enhancement of microvascular permeability upon stimulation with PAF was also abrogated almost completely after blockade of MMP-9. Despite inhibition of increased gelatinase activity as well as the subsequent reconstitution of junctional integrity, however, still a huge number of leukocytes emigrated into the perivascular tissue. Therefore, it seems unlikely that leukocytes just passively efflux through disruptured endothelial junctions, suggesting mechanisms facilitating leukocyte diapedesis independently from gelatinases. Although it is possible that PECAM-1, ICAM-2, the integrin 6β1 or JAM-A control PAF-induced transmigration of leukocytes after inhibition of gelatinases, the adhesion molecules ESAM, CD99, CD99L2, and JAM-C, as well as the protease NE are the more probable candidates. Moreover, it has been reported that leukocytes use different routes to overcome the endothelial barrier. Following morphological studies, the vast majority of emigrating leukocytes seems to migrate between intercellular junctions of endothelial cells into the perivascular tissue using the paracellular/junctional route, and a significant, smaller proportion of emigrating leukocytes has been reported to use the transcellular route [56
, 57
]. With respect to our intravital microscopy findings, it also seems possible that certain inflammatory stimuli, such as PAF, LTB4, and TNF-, might allow transmigrating leukocytes to use the transcellular route more intensively than others do.
In summary, our in vivo findings demonstrate that gelatinases strictly regulate microvascular permeability as well as remodeling of the perivascular BM during the initial inflammatory response, whereas concomitant, firm adherence, and (subsequent) transmigration of neutrophils and monocytes are mediated by these proteases in an inflammation-specific manner. These data highlight the stimulus dependency of the leukocyte recruitment process in vivo and might thereby help explain inconsistent findings in literature.
Received October 2, 2007; revised November 21, 2007; accepted November 29, 2007.
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
6 integrins and neutrophil elastase cooperate in mediating neutrophil transmigration J. Cell Sci. 118,2067-20766 integrin inhibits IL-1β- but not TNF--induced neutrophil transmigration in vivo J. Leukoc. Biol. 77,159-165
: tissue distribution, biochemistry, and function of a natural adherence molecule (ICAM-1) J. Immunol. 137,245-254[Abstract]This article has been cited by other articles:
 |
|
 |
|
  B. Sarraj, S. Massberg, Y. Li, A. Kasorn, K. Subramanian, F. Loison, L. E. Silberstein, U. von Andrian, and H. R. Luo Myeloid-Specific Deletion of Tumor Suppressor PTEN Augments Neutrophil Transendothelial Migration during Inflammation J. Immunol., June 1, 2009; 182(11): 7190 - 7200. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
