Introduction

Top Abstract Introduction Materials and Methods Results Discussion References

The movement of polymorphonuclear neutrophils (PMN) from the circulation to extravascular sites is essential for host defense. This movement, called emigration, involves adhesion to and migration between endothelial cells and migration across subendothelial basement membranes. Great strides have been made in elucidating the mechanisms by which PMN adhere to and travel between endothelial cells (1). Much less well-defined are the mechanisms by which PMN traverse subendothelial structures, including basement membranes and other extracellular matrixes. This is the least-understood step in the process of PMN migration to extravascular sites of inflammation (6).

Among the unresolved issues about PMN emigration is whether focal proteolytic degradation of subendothelial basement membranes is required. Some evidence indicates a requirement for proteinases that may be PMN-derived or endothelial cell-derived (2). However, the necessity for proteinases is not certain, judging from studies showing that PMN pass through discontinuities in the basement membranes of alveolar capillaries in experimental bacterial pneumonia (7), and that PMN movement through basement membranes laid down by endothelial cells (8) and through reconstituted basement membranes (9) is not impaired by proteinase inhibitors. It is also possible that proteinase requirements differ between sites of emigration so that, for example, emigration in the pulmonary circulation, which occurs in alveolar capillaries (7), may involve different mechanisms than those in the systemic circulation, where emigration occurs in postcapillary venules (10).

PMN granules contain multiple proteinases that might be involved in PMN emigration (11). These include the matrix metalloproteinases (MMPs) gelatinase B (MMP-9; 92-kD gelatinase) and neutrophil collagenase (MMP-8), and the serine proteinases elastase, cathepsin G, proteinase 3, azurcidin, and urokinase-type plasminogen activator. The MMPs are stored in granules from which they are readily released when PMN are engaged by a variety of ligands, including most neutrophil chemoattractants.

Type IV collagen and entactin are substrates for gelatinase B (12). The capacity of gelatinase B to degrade these basement membrane components, along with the finding that inhibition of PMN metalloproteinases blunts PMN migration across basement membranes in vitro, has suggested an important role for gelatinase B in PMN emigration (13). Recently, we generated mice with a targeted mutation of the gelatinase B gene (14). In the present study we have used these mutant mice to investigate the consequences of gelatinase B deficiency upon PMN emigration in experimentally induced acute inflammation.

	Materials and Methods

Top Abstract Introduction Materials and Methods Results Discussion References

Materials

Lipopolysaccharide (LPS) (Escherichia coli 055:B5), myeloperoxidase (MPO) (EC 1.11.1.7), and zymosan were purchased from Sigma Chemical Co. (St. Louis, MO). Thioglycollate broth was from Difco Laboratories (Detroit, MI). Recombinant human interleukin (IL)-8 was from R&D Systems (Minneapolis, MN). Zymosan-activated serum (ZAS) was generated by incubating rat serum with zymosan (10 mg/ml) at 37°C for 1 h, followed by heat inactivation at 56°C for 30 min and removal of the zymosan by centrifugation. ZAS was diluted in Hanks' balanced salt solution (HBSS) with Ca²⁺ and Mg²⁺ containing 0.1% bovine serum albumin (BSA). Matrigel Invasion Chambers were from Becton-Dickinson Labware (Bedford, MA). Klebsiella pneumoniae and Staphylococcus aureus, provided by Abderrazzaq Belaaouaj (Washington University, St. Louis, MO), were grown in tryptic soy broth.

Gelatinase B-Deficient Mice

Gelatinase B-deficient (gelatinase B/) mice were generated as described (14). Briefly, most of exon 2 was deleted, resulting in loss of gelatinase B function. Gelatinase B/ mice of mixed genetic background (129/SvEv and CD1) were used for most studies. To confirm that the results were not strain-dependent, selected studies were done on gelatinase B/ mice having a pure 129/SvEv genetic background. Gelatinase B+/+ mice were from the same genetic background (mixed or 129/SvEv) as the gelatinase B/ animals. Gelatinase B/ mice are indistinguishable from their gelatinase B+/+ littermates by gross appearance, weight, fertility, longevity, and organ histology, although there is delayed osteogenesis at the end plates of long bones during development (14). Gelatinase B+/+ and gelatinase B/ mice, 8 to 10 wk old, were used for studies of PMN emigration in the lungs and peritoneum. All procedures were approved by the Washington University Animal Studies Committee.

Blood Leukocyte Counts

Blood was collected by retro-orbital venous plexus sampling in polypropylene tubes containing ethylenediaminetetraacetic acid. Complete blood counts were determined using a Baker-9000 automated cell counter (BioChem ImmunoSystems, Allentown, PA). Leukocyte differentials were performed on Wright-stained blood smears.

PMN Emigration in the Lungs

PMN emigration in the lungs was determined indirectly, using the entry of PMN into alveolar spaces as determined by the bronchoalveolar lavage (BAL) fluid (BALF) content of PMN. These data were supplemented by histologic analysis. After an intraperitoneal injection of ketamine and xylazine for sedation and anesthesia, the trachea was exposed and LPS, 200 µg, or saline only was instilled as a bolus (50 µl) into the trachea using a 28-gauge needle and U-100 insulin syringe (Becton Dickinson Co., Franklin Lakes, NJ). During the injection the animals were held head up at an approximately 45-degree angle and rocked gently to facilitate distribution of the instillate. The neck incision was sutured with 4-O silk (Ethicon Inc., Somerville, NJ). The cell count of the BALF was determined using a hemocytometer after lysing the red blood cells with Tris-buffered ammonium chloride buffer (pH 7.2). Leukocyte differential cell counts were performed upon Wright-stained preparations (LeukoStat; Fisher Scientific, Pittsburgh, PA). A total of 200 cells was counted per BALF.

PMN Emigration in the Peritoneal Cavity

Mice were injected with 1 ml of 4% thioglycollate broth intraperitoneally (5). They were killed 4 h later and the peritoneal cavity was lavaged twice with 10 ml of saline. The leukocyte count and differential of the peritoneal lavage fluid was determined using a hemocytometer and cytospins stained with LeukoStat (Fisher Scientific).

PMN Emigration in the Skin

Intradermal injection of IL-8 leads to an accumulation of PMN at the injection site (15). A single intradermal injection of IL-8 (50 ng in 50 µl of saline) was given to gelatinase B/  and B+/+ mice as previously described (15). Each mouse also received an injection of 50 µl of saline at a separate site. After 4 h, the skin at each injection site was excised and either fixed in formalin for histologic analysis or subjected to MPO activity assay as a measure of PMN accumulation.

MPO Activity

The MPO activity of skin was determined as previously described (15). Briefly, the tissue was homogenized in 0.5 ml of cold 0.1 M Tris-Cl, pH 7.6, and 0.15 M NaCl and freeze-thawed three times. After centrifugation (600 × g, 5 min), 50-µl supernatants were reacted with H₂O₂ (0.0005%) in the presence of o-dianisidine dihydrochloride (0.167 mg/ml) for 15 min. The reaction was stopped by the addition of 50 µl of 10% sodium azide and the optical density at 450 nm was measured. The assay was done in triplicate. Protein concentrations of the tissue extracts were determined by the Bio-Rad dye binding assay (Bio-Rad, Hercules, CA) using BSA as the standard.

BALF Protein and Gelatinase Activity

At 24 h after intratracheal injections of LPS or saline, animals were killed by CO₂ narcosis and the lungs were lavaged three times with 0.6 ml of saline through a tracheal cannula. The BALF was centrifuged (600 × g, 5 min) and the supernatant removed and frozen at 70°C. BALF protein concentrations were determined by the Bio-Rad dye binding assay (Bio-Rad), using BSA as the standard. The gelatinolytic activity in BALF was assessed by zymography (16). Four microliters of cell-free BALF was subjected to sodium dodecyl sulfate polyacrylamide gel electrophoresis (SDS-PAGE) under nonreducing conditions in gels containing 1 mg/ml gelatin. After electrophoresis, gels were soaked in 2.5% Triton X-100 for 30 min, rinsed briefly, and then incubated at 37°C for 24 h in 100 mM Tris HCl, pH 7.4, containing 10 mM CaCl₂. The gels were stained with Coomassie blue R-250 and destained in 5% acetic acid and 10% methanol.

PMN Migration through Matrigel

The capacity of PMN to migrate through Matrigel, a prototype basement membrane secreted by the Engelbreth- Holm-Swarm tumor, was assessed in a modified Boyden chamber assay (13). PMN were isolated from bone marrow (17) or from the peritoneal cavity 4 h after injection of thioglycollate. Bone marrow-derived PMN, used for most studies, were suspended at 1.5 × 10⁶/ml in HBSS with Ca²⁺ and Mg²⁺ containing 0.1% BSA and placed in the upper compartment. The lower compartment contained 0 to 2.0% ZAS in HBSS. The two compartments were separated by an 8-µm pore-size membrane coated with Matrigel. After a 3-h incubation at 37°C in humidified 5% CO₂- air, the filter was stained with LeukoStat (Fisher Scientific) and mounted on a glass slide. The number of cells that migrated to the underside of the filter in five random high-power fields (× 400) was determined for each of triplicate filters for each experimental condition. Each migration assay was done using PMN pooled from two or three gelatinase B/ or B+/+ mice.

Bacterial Killing by PMN

To assess the possible effects of gelatinase B deficiency upon a primary function of PMN (namely, host defense against bacteria), bacterial killing by gelatinase B-deficient PMN was compared with killing by gelatinase B+/+ PMN. Gelatinase B/ and B+/+ mice were injected with 4% thioglycollate (1 ml per mouse). After 3 h these animals were killed and their peritoneal exudates, which were almost exclusively PMN, were collected and washed with phosphate-buffered saline (PBS) and used for assays of bacterial killing, as previously described (18). In 96-well plates, 0.1 ml of PMN suspension (10⁶ cells/ml) were seeded. After 30 min the medium in each well was replaced, gently removing nonadherent PMN, and then 10⁷ K. pneumoniae or S. aureus was added to each well. After 15 min, the wells were washed to remove bacteria that were not cell-associated. PMN from some wells were lysed in PBS containing 0.1% Triton X-100. PMN in other wells were allowed to incubate for an additional 30 min, after which they were lysed. PMN lysates obtained at both time points were plated and the number of viable bacteria was determined. The number of viable bacteria present after 15 min of incubation was regarded as the baseline of PMN-associated bacteria. The number of viable bacteria still present in the PMN 30 min later, in comparison with the initial cell-associated number, was used as a measure of bacterial killing.

Statistical Analysis

All data are expressed as means ± standard error of the mean (SEM). Statistical differences between groups were determined using Student's unpaired t test. Significance was defined at the P < 0.05 level unless otherwise stated.

	Results

Top Abstract Introduction Materials and Methods Results Discussion References

PMN in Peripheral Blood

There was no difference in total circulating leukocyte counts and the percentages of neutrophils, lymphocytes, monocytes, and eosinophils between gelatinase B/ and B+/+ mice (Table 1).

BALF PMN after Intratracheal LPS

As described by others (19, 20), LPS induced alveolar septal thickening with intra-alveolar PMN and mononuclear cells (not shown). PMN accumulated in alveolar spaces after intratracheal instillation of LPS, comparably in B/ mice and B+/+ mice (Figure 1). Similar results occurred whether the gelatinase B/ mice were of pure 129/SvEv background or had a mixed (129/SvEv and CD1) genetic background.

View larger version (45K): [in this window] [in a new window]   Figure 1.   BALF content of PMN after intratracheal LPS. Gelatinase B+/+ and gelatinase B/ mice received 200 µg of LPS in saline by intratracheal injection. BAL was performed 3, 6, 12, or 24 h later. (A) The number of PMN in BALF; (B) the percentage of all cells in BALF that are PMN. There was no difference between gelatinase B+/+ and gelatinase B/ mice. Each data point represents four or more mice ± SEM.

BALF PMN Protein and Gelatinase Activity

The protein concentration in BALF after 24 h after LPS instillation was increased similarly in gelatinase B/ and B+/+ mice to 474 ± 118 and 388 ± 41 mg/liter, respectively (Figure 2A). These changes indicate that gelatinase B deficiency did not affect the increase in microvascular permeability elicited by LPS.

View larger version (41K): [in this window] [in a new window]   Figure 2.   Protein content and gelatinase activity of BALF 24 h after intratracheal LPS. (A) The total protein concentration of BALF was increased significantly (P < 0.05) compared with intratracheal saline in both gelatinase B+/+ and gelatinase B/ mice. Each group represents three or more mice ± SEM. (B) Gelatin zymography of BALF from gelatinase B+/+ and gelatinase B/ mice. Fluids from nontreated mice (lanes 1 and 4) and saline-treated mice (lanes 2 and 5) did not show gelatinolytic activity corresponding to gelatinase B. BALF from gelatinase B+/+ mice given 200 µg of LPS intratracheally (lane 3) showed a prominent band of gelatinolytic activity at ~ 105 kD corresponding to murine gelatinase B (29). The higher molecular-weight zones of gelatinolytic activity in this lane represent gelatinase B-lipocalin complexes at ~ 130 kD and multimers of gelatinase B at higher molecular weights. BALF from gelatinase B/ mice that received LPS (lane 6), as expected, showed no gelatinolytic activity corresponding to gelatinase B. All lanes showed gelatinolytic activity at ~ 67 kD, reflecting gelatinase A.

BALF from gelatinase B+/+ control animals given intratracheal LPS (Figure 2B, lane 3) contained prominent gelatinolytic activity at ~ 105 kD corresponding to gelatinase B, as reported by others (21). Activity was also present at ~ 130 kD consistent with the presence of gelatinase B and lipocalin complexes. The gelatinase B in these complexes would be derived from PMN because lipocalin is restricted to PMN (22). As expected, BALF from gelatinase B-deficient mice did not contain gelatinolytic activity at either 105 or 130 kD (Figure 2B, lane 6). Nontreated and saline-treated mice, both gelatinase B+/+ and B/ (Figure 2B, lanes 1, 2, 4, and 5) showed minimal gelatinase activity at 66 kD, probably representing gelatinase A. This activity was slightly increased after LPS treatment in both types of animals.

PMN Emigration in the Peritoneum

Thioglycollate medium was used to elicit acute intraperitoneal inflammation. At 4 h after intraperitoneal instillation of thioglycollate, there was no difference in PMN emigration between gelatinase B+/+ and gelatinase B/ mice. In 13 gelatinase B+/+ mice, peritoneal lavage yielded 15.3 ± 1.9 × 10⁶ leukocytes, of which 79 ± 5% were PMN; whereas lavage of six gelatinase B/ mice yielded 21.8 ± 3.5 × 10⁶ leukocytes, of which 82 ± 2% were PMN. Accordingly, PMN emigration in the peritoneal cavity was not impaired in gelatinase B/ mice.

PMN Emigration in the Skin

In both gelatinase B+/+ and gelatinase B/ mice, PMN were recruited into the skin by injection of the PMN chemoattractant IL-8 at 4 h (Figure 3B). The PMN were predominantly extravascular. Accordingly, measurements of whole-skin MPO activity could be regarded as indicative of PMN emigration. Skin MPO activity was increased approximately fourfold by IL-8 in both gelatinase B+/+ and gelatinase B/ mice (Figure 3A).

View larger version (72K): [in this window] [in a new window]   Figure 3.   Skin MPO activity and PMN localization in the skin after intradermal IL-8. IL-8 (50 ng in 50 µl of saline) or saline was injected intradermally into gelatinase B+/+ and gelatinase B/ mice. At 4 h later the injection sites were excised and assayed for MPO activity or fixed for histologic analysis. There was no difference in MPO activity between the two types of mice after IL-8, and the MPO activity was significantly increased (P < 0.05) after the injection of IL-8 compared with activity after the injection of saline only (n = 4 or 5 in each group; ± SEM) (A). After IL-8, there were numerous extravascular PMN in both gelatinase B+/+ and gelatinase B/ animals (original magnification: ×400) (B).

PMN Migration through Matrigel

In addition to the in vivo PMN recruitment experiments described previously, we also examined whether gelatinase B/ PMN can traverse a basement membrane-like substitute in vitro. Bone marrow-derived PMN from gelatinase B/ and B+/+ mice migrated through Matrigel-coated membranes with equal efficiency in response to ZAS (Figure 4). Similar results were obtained using PMN obtained from the peritoneal cavity after thioglycollate injection (data not shown).

View larger version (29K): [in this window] [in a new window]   Figure 4.   Migration of PMN through Matrigel in response to ZAS. Bone-marrow PMN were incubated on Matrigel-coated micropore filters to assess their ability to traverse this prototype basement membrane (see MATERIALS AND METHODS). The results shown are the number of PMN that migrated through the membrane per high-power field (original magnification: ×400). Gelatinase B/ and B+/+ PMN displayed comparable migration in response to several concentrations of ZAS. Shown are the means ± SEM of a representative experiment from three separate experiments, each done in triplicate.

Bacterial Killing by PMN

To assess whether gelatinase B deficiency impairs bacterial killing, PMN from gelatinase B+/+ and gelatinase B/ mice were incubated with bacteria and the residual bacteria were quantified. Both types of PMN reduced the number of viable S. aureus approximately 80% within 30 min (Figure 5A). In contrast, neither type of PMN significantly reduced the number of viable K. pneumoniae in 30 min (Figure 5B).

View larger version (39K): [in this window] [in a new window]   Figure 5.   Bacterial adhesion and killing by gelatinase B+/+ and gelatinase B/ PMN. PMN isolated from gelatinase B+/+ and gelatinase B/ mice were incubated with (A) S. aureus (SA) or (B) K. pneumoniae (KP) (see MATERIALS AND METHODS). After 15 min to establish a baseline of PMN-assocated bacteria, or after an additional 30 min to assess bacterial killing, the PMN were washed and then lysed with 0.1% Triton X-100. The lysates were plated in serial dilutions and the number of viable bacteria was determined as colony-forming units (CFU). Data shown are the means (± SEM) of two separate experiments done in triplicate.

	Discussion

Top Abstract Introduction Materials and Methods Results Discussion References

Gelatinase B deficiency results in a developmental phenotype of delayed ossification of endochondral bone (14). This defect correlates with impaired apoptosis of hypertrophic chondrocytes and retarded capillary invasion into the growth plate. However, there is no obvious disturbance in leukopoiesis. The bone marrow has normal precursors (T. Betsuyaku, unpublished observations) and the peripheral blood leukocyte number and differential are normal. Moreover, gelatinase B-deficient animals display normal growth, fertility, and life span, and are not prone to infections in a barrier animal facility. The fact that gelatinase B-deficient mice have normal numbers of circulating PMN simplified the comparison of PMN emigration between gelatinase B/ and gelatinase B+/+ mice. Direct comparisons would have been difficult if gelatinase B deficiency was associated with neutropenia, as occurs, for example, in granulocyte colony-stimulating factor receptor- deficient mice (23).

Gelatinase B is released from PMN when they are stimulated by chemokines (11). However, the physiologic function(s) of gelatinase B in PMN is not known. Because gelatinase B has proteolytic activity against some basement-membrane components (12), it is plausible that secretion of gelatinase B from PMN is involved in PMN emigration. Indeed, this speculation has been supported by studies of human PMN migration through Matrigel and through amnionic membranes (13, 24).

We designed experiments involving lung, peritoneum, and skin because the mechanisms of PMN emigration may be different at different sites (5). In studies of the role of CD11/CD18 in PMN emigration, CD18/ mutants had virtually no dermal emigration in response to an irritant, but had normal emigration in the lung in response to Streptococcus pneumoniae (5). We observed that, unlike CD18/ mutants, gelatinase B deficiency did not impair PMN emigration in any of the three sites. The present findings in the skin, using IL-8 as a chemoattractant, mirror recent data showing that PMN accumulate in the skin normally in gelatinase B/ mice in response to C5a generated by intracutaneous immune complexes of BP 180 and anti-BP 180 (25). It should be noted that the present studies are confined to PMN emigration. Normal PMN emigration in gelatinase B deficiency does not exclude the possibility of defective emigration of lymphocytes and eosinophils, which also contain gelatinase B and in which gelatinase B has been implicated in emigration (26, 27).

Because previous studies suggesting a role for gelatinase B in PMN emigration were done in vitro (13), we tested PMN migration through Matrigel-coated micropore membranes in response to the chemotactic activity of ZAS. Gelatinase B/ and B+/+ PMN were indistinguishable in their capacity to migrate through Matrigel. The difference between the present results and other in vitro data is not readily explained. More important, the present studies point to the need for caution in extrapolating from in vitro results. In vitro experimental systems for studying PMN emigration do not include the potential of proteolytic activities and inhibitors from resident cells such as microvascular endothelium (2, 9). In vitro systems may also have limited applicability to in vivo conditions if they use homogenous membranes, such as Matrigel, that lack pores like those in subendothelial basement membranes.

Several possibilities may explain normal PMN emigration despite gelatinase B deficiency. First, PMN emigration into the lung, peritoneal cavity, and skin may not depend upon proteolytic activity. The present results, together with the findings of discontinuities in pulmonary capillaries and postcapillary venules (7), make this an attractive possibility. Second, it is possible that proteolytic degradation of subendothelial basement membranes is required for PMN emigration at some sites, but that proteinases other than gelatinase B have this role. This possibility seems unlikely, however, because inhibitors of serine proteinases do not impair PMN emigration (8, 9) and because neutrophil elastase-deficient mice display normal PMN emigration in the lung and peritoneal cavity in response to bacterial infection (18). On the other hand, it is possible that proteinases expressed by endothelial cells or other resident cells adjacent to the basement membrane are involved in emigration (2). Another possibility is that the gelatinase B null mutation alters the profile of PMN proteinases. However, we have found that gelatinase B/ PMN have normal neutrophil elastase and cathepsin G activities (T. Betsuyaku, unpublished observations). We have not tested gelatinase B/ PMN for neutrophil collagenase (MMP-8), but this proteinase does not have the same substrate specificity as gelatinase B and therefore would not be expected to compensate for gelatinase B (28).

Because host defense against bacteria is a major function of PMN, we tested the capacity of gelatinase B-deficient PMN to kill bacteria in vitro. Against two organisms, no defect in killing was observed in comparison with gelatinase B+/+ PMN. These findings contrast with the recent discovery that PMN lacking neutrophil elastase have impaired killing of gram-negative bacteria in vitro and that all mice lacking neutrophil elastase died in response to a dose of intraperitoneal K. pneumoniae that resulted in a 50% mortality in gelatinase B+/+ mice (18). It would appear that gelatinase B does not have as critical a role in host defense against bacteria as does neutrophil elastase; however, studies in vivo are needed to further clarify the importance of gelatinase B to the bactericidal capacity of PMN.

In summary, the present results indicate that gelatinase B is not required for PMN emigration. The physiologic role of gelatinase B in PMN remains to be determined.