Pulmonary Matrix Metalloproteinase Excess in Hospital-acquired Pneumonia

Medizinische Klinik III and Klinik für Anaesthesiologie, Medical University of Lübeck, Lübeck, Germany

	ABSTRACT

TOP ABSTRACT METHODS RESULTS DISCUSSION REFERENCES

 
In hospital-acquired pneumonia, extracellular matrix destruction is common and may be caused by excessive activity of matrix metalloproteinases (MMPs). Thirty patients with hospital-acquired pneumonia and 16 control subjects were studied. We evaluated the concentrations of MMP-8, MMP-9, and tissue inhibitor of metalloproteinase-1 in mini-bronchoalveolar lavage fluid (mini-BALF) and blood using zymography and specific immunoassays. In patients with hospital-acquired pneumonia concentrations of MMP-8 and MMP-9 in mini-BALF were increased 10-fold, whereas their specific inhibitor tissue inhibitor of metalloproteinase-1 was not concomitantly increased. In 80% of patients with pneumonia, but in none of the control subjects, the active form of MMP-9 was detected by zymography. Zymography furthermore showed the banding pattern of neutrophil-derived MMP-9, indicating that neutrophils were the main source of MMP-9. Comparison of neutrophils from blood and mini-BALF showed higher basal release of MMPs by pulmonary neutrophils. Stimulation analysis indicated that pulmonary neutrophils were already maximally activated. In patients with detection of potentially pathogenic microorganisms, concentrations of MMPs were fivefold increased compared with patients with negative cultures. Furthermore, MMP-levels were related to clinical severity. These are the first data suggesting that neutrophil-derived MMPs are increased in hospital-acquired pneumonia in association to the detection of causative microorganisms and clinical severity.

　

Key Words: matrix metalloproteinases • TIMP-1 hospital-acquired pneumonia • polymorphonuclear neutrophils

Pneumonia is the leading cause of death among patients with hospital-acquired infections, with an overall mortality of at least 30%. Despite the advent of modern antibiotics mortality has still not declined (1, 2). A probable explanation is the multimorbidity of this patient group with impaired defense mechanisms. Colonization of the respiratory tract is common especially among severely ill patients (1) and might be due to increased bacterial adherence in the upper airways (3). In these patients, clearance of invading microorganisms is often ineffective leading to persistent inflammation and tissue destruction (2). Tissue destruction on the other hand facilitates persistence of microorganisms and per se initiates a proinflammatory response.

As a crucial feature of host defense against bacteria polymorphonuclear neutrophils (PMN) migrate into sites of inflammation (4). Physiologically, PMN are found in the vascular bed of the lung. In pneumonia, PMN invade the alveolar compartment (5). They are known to secrete different preformed enzymes stored in granules and vesicles including matrix metalloproteinases (MMPs) (6). MMPs are a family of zinc-endopeptidases capable of degrading components of the extracellular matrix (7). They play an important role in physiologic and pathologic processes concerning the extracellular matrix turnover as well as tissue degradation, repair mechanisms, or cell migration (8, 9). Two isoforms are secreted mainly from neutrophils: MMP-9 (92-kD gelatinase) and MMP-8 (neutrophil collagenase) (7).

MMPs are highly regulated at different levels: (1) gene expression and protein secretion (10, 11), (2) activation of latent MMP-proforms by different nonproteolytic agents and proteinases including pseudomonal elastase (10, 12), and (3) inhibition of activated MMPs. The most important nonspecific inhibitor is 2-macroglobulin (13). Specific inhibitors of MMPs are the tissue inhibitors of matrix metalloproteinases (TIMPs) (14). They are produced by different cells like macrophages, fibroblasts, and PMN (15). TIMP-1 and TIMP-2 are capable of inhibiting MMPs by forming stoichiometric 1:1 complexes with active MMPs.

An imbalance between the expression of metalloproteinases and their inhibitors is believed to generate tissue destruction or abnormal tissue repair (14). Previous studies have shown that MMPs are implicated in the pathogenesis of various pulmonary inflammatory diseases like acute respiratory distress syndrome (16), bronchiectasis (17), cystic fibrosis (18), interstitial lung disease (19), chronic obstructive pulmonary disease (20), and emphysema (21). All of these are characterized by influx of neutrophils and parenchymal destruction. Pneumonia is characterized by similar features, but the role of MMPs and TIMPs has not yet been studied.

We therefore investigated MMP-8, MMP-9, and TIMP-1 in mini-bronchoalveolar lavage fluid (mini-BALF) and plasma of patients with hospital-acquired pneumonia to determine if there is an increased concentration and activation of MMPs within the pulmonary compartment and if this increase is counterbalanced by a concomitant increase of TIMP-1. In addition, we evaluated whether the levels and activity of MMPs are related to clinical severity and the detection of causative microorganisms. As we hypothesized that PMN are the main source of MMPs in pneumonia, we furthermore investigated the exocytosis of MMP-8 and MMP-9 by pulmonary and peripheral PMN.

	METHODS

TOP ABSTRACT METHODS RESULTS DISCUSSION REFERENCES

 
Study Group
Thirty patients with hospital-acquired pneumonia were studied  . Hospital-acquired pneumonia was defined by the presence of clinical symptoms at 48 hours or more after hospital admission, new or progressive infiltrates on the chest radiograph, and at least two of the following criteria: (1) fever ( 38.5°C), (2) leukopenia or leukocytosis (white blood cells 4,000 or  12,000/mm³), or (3) purulent tracheal secretions (22). The modified clinical pulmonary infection score (CPIS) was used to estimate the severity of pneumonia (23). For controls, 16 persons who underwent elective cardiac surgery were studied .

fig.ommitted TABLE 1. Demographic data of 30 patients with hospital-acquired pneumonia and 16 respiratory healthy control subjects (mean ± sem).

 
Bronchoscopy and Mini-bronchoalveolar Lavage
Bronchoscopy was performed according to standard conditions as described previously (24). Mini-bronchoalveolar lavage (mini-BAL) was performed during bronchoscopy in nonventilated patients and via suction catheter in ventilated patients (25). One sample of mini-BAL was used for microbiologic cultures. The other part of minilavage was diluted with an equal volume of phosphate-buffered saline and vortexed until homogenized. To separate supernatants (mini-BALF) from the cell pellet, the sample was centrifuged at 400 x g for 10 minutes. Mini-BALF was stored at -70°C. Cell pellet was then resuspended in Roswell Park Memorial Institute medium. Total cell count was performed on a hemocytometer. Neutrophils were determined by differential cell count of a Wright-Giemsa–stained cytocentrifuge smear. The sample was diluted to a concentration of 10⁶ cells/ml.

Blood
To determine the concentrations of MMPs and TIMP-1 in plasma, ethylenediaminetetraacetic acid blood samples were centrifuged at 400 x g for 10 minutes, and the supernatant was stored at -70°C. For isolation of neutrophils venous heparinized blood was drawn, and Ficoll-Hypaque density centrifugation was performed as described (26).

Degranulation
To quantify neutrophil release of MMPs, 100,000 neutrophils of blood and mini-BAL were incubated in phosphate-buffered saline at 37°C (27). For stimulation, PMA (phorbol-myristate-acetate) was added. After 1 hour, incubation was stopped by centrifugation (27). Supernatants were carefully removed and stored at -70°C. Results were expressed in femtograms of protein released per PMN (fg/PMN).

Measurements of MMPs and TIMP-1 by Enzyme-linked Immunosorbent Assay
The concentrations of MMP-8, MMP-9, and TIMP-1 in mini-BALF, plasma, and cell supernatants were determined by specific enzyme-linked immunosorbant assays (BIOTRAK ELISA; Amersham Biosciences, Freiburg, Germany) according to the manufacturer's instructions. Each sample was assayed in duplicate, and the values were within the linear portion of the standard curve.

Gelatin Zymography
Gelatin zymography was performed as described (28): sodium dodecyl sulfate gels containing 0.1% gelatin were used to identify gelatinolytic proteases from mini-BALF and plasma. To identify the enzymes, MMP standard was loaded on each gel. Gelatinolyic activity was quantified using E.A.S.Y.Win32 imaging software (Herolab, Wiesloch, Germany).

Measurement of 2 Macroglobulin and Lactoferrin
2 Macroglobulin and lactoferrin in mini-BALF were determined by highly sensitive chemiluminescence immunoassays as described before (29).

Statistical Analysis
All data are presented as mean ± SEM. Statistics were performed with nonparametric tests. For unpaired samples the Mann–Whitney U test was used. Correction for multiple comparisons was done by Scheffe's method. Correlations were analyzed with Spearman's rank correlation. A value of p less than 0.05 was accepted as significant. Calculations were performed with Statistica for Windows, version 5 (StatSoft GmbH, Hamburg, Germany).

	RESULTS

TOP ABSTRACT METHODS RESULTS DISCUSSION REFERENCES

 
MMP-8, MMP-9, and TIMP-1 Levels in Plasma and Mini-BALF
MMP-8 and MMP-9 levels in mini-BALF were 13-fold higher in patients with pneumonia compared with control subjects with healthy respiratory systems. In plasma, MMP-8 was 12-fold and MMP-9 was fourfold increased. The level of TIMP-1 was significantly higher in plasma of patients with pneumonia compared with control subjects. In contrast, there was no increase of TIMP-1 in mini-BALF of patients with pneumonia  .

fig.ommitted Figure 1. (A) Concentrations of MMP-8, MMP-9 and TIMP-1 in mini-BALF of 30 patients with hospital-acquired pneumonia (dark gray bars) and 16 control subjects with healthy respiratory systems (light gray bars) (mean ± SEM). MMP-8 and MMP-9 are significantly increased compared with control subjects, whereas TIMP-1 levels show no significant difference between the two groups. (B) Concentrations of MMP-8, MMP-9, and TIMP-1 in plasma of patients with hospital-acquired pneumonia (dark gray bars) and control subjects with healthy respiratory systems (light gray bars) (mean ± SEM). Patients with pneumonia show significantly increased MMP-8, MMP-9, and TIMP-1 levels.

 
In patients with pneumonia, the levels of MMP-8 and MMP-9 in mini-BALF correlated with the percentage of neutrophils in mini-BALF (MMP-8: r = 0.41, p = 0.03; MMP-9: r = 0.43, p = 0.02). TIMP-1 levels in mini-BALF were not associated with the percentage of neutrophils.

Neutrophil Release of MMP-8, MMP-9, and TIMP-1
In patients with pneumonia, the basal release of MMP-8 and MMP-9 by pulmonary neutrophils was significantly higher than basal MMP release of neutrophils isolated from blood (MMP-8: 9.3 ± 1.8 fg/PMN versus 2.2 ± 0.5 fg/PMN, p < 0.001; MMP-9: 17.3 ± 5.2 fg/PMN versus 5.7 ± 1.8 fg/PMN, p < 0.05). After stimulation with PMA, peripheral PMN were able to release more MMPs. In contrast, there was no significant difference between basal and PMA-stimulated release of MMPs in pulmonary PMN  .

fig.ommitted Figure 2. Release of MMP-8 and MMP-9 by peripheral and pulmonary PMN of patients with pneumonia under basal conditions (light gray bars) and after stimulation with PMA (dark gray bars). Basal release of MMP-8 and MMP-9 was significantly higher in pulmonary than in peripheral PMN. After stimulation with PMA, release of MMP-8 and MMP-9 by peripheral PMN was significantly increased. In pulmonary PMN, stimulation with PMA did not significantly increase the MMP release.

 
Comparison of patients with pneumonia and control subjects showed no difference in basal or PMA-stimulated release of MMP-8 and MMP-9 by peripheral PMN (MMP-8 basal: 2.2 ± 0.5 versus 2.8 ± 0.8 fg/PMN, p = not significant; MMP-8 stimulated: 5.5 ± 1.0 versus 5.6 ± 1.0 fg/PMN, p = not significant; MMP-9 basal: 5.7 ± 1.8 versus 4.8 ± 1.3 fg/PMN, p = not significant; MMP-9 stimulated: 10.2. ± 2.5 versus 9.0 ± 1.8 fg/PMN, p = not significant). Comparison of MMP release by pulmonary neutrophils of patients and control subjects was not possible as the proportion of PMN in mini-BAL of control subjects was too low (PMN < 3%). Release of TIMP-1 by neutrophils of patients and control subjects was not detectable.

Determination of Gelatinolytic Activity
Gelatin zymography was performed in all patients and control subjects. In mini-BALF of patients with pneumonia, gelatinolytic bands of 220, 130, and 92 kD were detected in all samples. These bands correspond to the typical banding pattern of neutrophil-derived MMP-9 consisting of pro–MMP-9 (92 kD), human neutrophil lipocalin–pro-MMP-9 complex (130 kD) and the homodimeric form of MMP-9 (220 kD). The 92-kD band was the most prominent of these three bands in all patients. Furthermore, an 85-kD band corresponding to the activated form of MMP-9 was detected in 24 of the 30 patients with pneumonia (80%). All patients with potentially pathogenic microorganisms (PPM) showed an 85-kD band, and in 14 out of the 16 patients (87.5%) with a CPIS of 7 or more, an 85-kD band was detected. In mini-BALF of control subjects barely visible bands of 220, 130, and 92 kD, but no 85-kD band, were identified. Zymography of representative patients is shown in  . Analysis of plasma samples showed the three bands of neutrophil-derived MMP-9. Neither patients nor control subjects expressed bands of activated MMP-9 in plasma. Gelatinolytic activity of the bands corresponding to pro–MMP-9 (92 kD) and the activated form (85 kD) was further quantified by densitometric scanning . Densitometric results correlated significantly with the determination of MMP-9 by enzyme-linked immunosorbent assay (r = 0.86, p < 0.001).

fig.ommitted Figure 3. (A) Gelatin zymography of mini-BALF from representative patients with hospital-acquired pneumonia and a control subject with a healthy respiratory system. Lane 1: mini-BALF of a control subject, lanes 2 to 5: mini-BALF of four different patients with hospital-acquired pneumonia, lane 6: MMP standard. Gelatinolytic bands of pro–MMP-9 (92 kD), pro-MMP-9–lipocalin-complex (130 kD) and homodimeric MMP-9 (220 kD) are clearly visible in patients with pneumonia but only barely detectable in the control subject. Furthermore, the patients with pneumonia show an 85-kD band corresponding to the active form of MMP-9, which is not present in the control subject. (B) Densitometric analysis of gelatinolytic MMP-9 activity in mini-BALF and plasma. Results are expressed as percent activity versus MMP-9 standard. Gelatinolytic activity in mini-BALF and plasma was significantly higher in patients with pneumonia (dark gray bars) than in control subjects (light gray bars). In mini-BALF of patients with pneumonia, the shown results reflect the total activity of both active- and proform of MMP-9. The percentage of zymographic signal due to 85 kD was 10% of the total MMP-9 signal in zymography. In mini-BALF of control subjects and plasma of both groups, results reflect the activity of pro–MMP-9, as the activated form was not detectable.

 
In the presence of ethylenediaminetetraacetic acid, an inhibitor of MMPs, all of the bands of plasma and mini-BALF samples disappeared, whereas they were not affected by addition of phenylmethane sulfonyl fluoride, a serine protease inhibitor (data not shown). This indicates that gelatinolytic activity was due to the presence of MMPs.

Microbiological Findings and Levels of MMP and TIMP
Patients with hospital-acquired pneumonia were divided into two groups depending on culture results of mini-BAL. The PPM-group included 12 patients (number of patients in parentheses): Pseudomonas aeruginosa (3), P. aeruginosa and Enterobacter spp. (2), Staphyloccocus aureus (1), S. aureus and Streptococcus pneumoniae (1), Aspergillus fumigatus (2), Branhamella catarrhalis (1), Stenotrophomonas maltophilia and Citrobacter freundii (1), Klebsiella pneumoniae (1). The non-PPM group included 18 patients with residential flora of the oropharynx, Candida spp., or sterile cultures. The concentrations of MMP-8 and MMP-9 in mini-BALF were significantly higher in the PPM group compared with the non-PPM group (p < 0.005). In contrast, there was no significant difference between the concentrations of TIMP-1 in mini-BALF in the two subgroups  .

fig.ommitted Figure 4. Concentrations of MMP-8, MMP-9 and TIMP-1 in mini-BALF of the PPM-group (patients with PPM in microbiologic findings, including P. aeruginosa, S. aureus, A. fumigatus, and different gramnegative bacteria) and non-PPM group (normal flora of the oropharynx, Candida spp., or sterile culture). Levels of MMP-8 and MMP-9 were significantly increased in the PPM-group, whereas no significant difference between TIMP-1 levels of the two groups was detectable (mean ± SEM).

 
Clinical Severity and Levels of MMP and TIMP
Assessment of severity by the CPIS showed a mean score of 6.92 ± 0.26 points in patients with hospital-acquired pneumonia. The PPM-group showed a significantly higher CPIS than the non-PPM group (7.7 ± 0.12 versus 6.4 ± 0.15; p < 0.01). In patients with a CPIS of 7 or more, the concentrations of MMP-8 and MMP-9 were significantly higher than in patients with a CPIS of less than 7 (MMP-8: 513.62 ± 150.33 ng/ml versus 179.41 ± 75.01 ng/ml, p < 0.05; MMP-9: 1,123.84 ± 348.99 ng/ml versus 350.06 ± 129.38 ng/ml, p < 0.05). In contrast, TIMP-1 concentrations were not significantly different between the two groups (TIMP1: 725.08 ± 228.04 ng/ml versus 370.50 ± 97.24 ng/ml).

2-Macroglobulin and Lactoferrin in Mini-BALF
To estimate the alveolocapillary leakage, 2-macroglobulin was determined in mini-BALF. The concentration of 2-macroglobulin in patients with pneumonia was significantly higher than in control subjects, in whom 2-macroglobulin was detected only in trace amounts (1.57 ± 0.3 mg/L versus 0.17 ± 0.07 mg/L). No correlation was found between the concentrations of MMPs and 2-macroglobulin in mini-BALF.

To compare the amount of MMPs in mini-BALF with other contents of neutrophil granules released into mini-BALF, the concentration of lactoferrin was determined. The mean ± SEM concentration of lactoferrin in mini-BALF of patients with pneumonia was 646.91 ± 167.43 ng/ml. There was a positive correlation between MMP-8 but not MMP-9 levels and the concentration of lactoferrin in mini-BALF (MMP-8: p < 0.05, r = 0.43).

	DISCUSSION

TOP ABSTRACT METHODS RESULTS DISCUSSION REFERENCES

 
Pneumonia is characterized by a massive influx of neutrophils into sites of inflammation and frequently leads to severe pulmonary tissue injury (5). Matrix degrading proteases released by PMN may play an important role in this process. We found a significant increase of MMP-8 and MMP-9 in mini-BALF of patients with hospital-acquired pneumonia compared with control subjects. Furthermore, we showed that the excess of MMPs in the pulmonary compartment of patients with pneumonia was not counterbalanced by a concomitant increase of the specific inhibitor TIMP-1. These results implicate a local imbalance between MMPs and their inhibitor TIMP-1 with an excess of MMPs in the inflamed lung compared with control subjects with healthy respiratory systems. However, as MMPs are released as inactive proforms that need to be activated extracellularly, the detection of activated MMPs is crucial. Therefore we also performed zymographic analysis to distinguish between latent and activated MMP-9. In mini-BALF samples zymography demonstrated not only the 92-kD latent form of MMP-9 but also the 85-kD active form of MMP-9 in the vast majority of patients with pneumonia. This finding reveals that antiprotease defense was overwhelmed, leading to free MMP-9 activity in the alveolar compartment. In control subjects, the active form of MMP-9 was not detectable. In contrast to TIMP-1, 2-macroglobulin, which is an unspecific inhibitor of MMPs, was significantly increased in mini-BALF of patients with pneumonia. However, as 2-macroglobulin–bound MMPs are not detectable by enzyme-linked immunosorbent assay, the highly increased MMP levels in mini-BALF indicate that 2-macroglobulin levels were not sufficient for neutralizing MMPs. Furthermore, the detection of activated MMP-9 by zymography in most of the patients with pneumonia shows that MMPs were not sufficiently inhibited.

Increased, free MMP activity and an imbalance between MMP and TIMP have been described as a crucial feature in different diseases. Delacourt and coworkers demonstrated that an imbalance between MMP-9 and TIMP in sputum with high levels of free gelatinase activity plays an important role in tissue destruction in cystic fibrosis (18). Sepper and coworkers found that the degree of neutrophil collagenase (MMP-8) activity was correlated to the severity of bronchiectasis (17). In contrast, in diseases like asthma, which are characterized by increased deposition of collagen, a decreased molar MMP-9/TIMP-1 ratio correlates with more severe disease and an increased ratio is associated with steroid responsiveness (30, 31).

Our results suggest that an MMP–TIMP imbalance with unchecked MMP activity goes together with more severe pneumonia. First, in patients with severe pneumonia, as assessed by the CPIS, the MMP excess was higher than in patients with less severe disease, and the active form of MMP-9 was found in most cases of this group. Second, patients with detection of PPM showed higher concentrations of MMP-8 and MMP-9 in mini-BALF, and the active form of MMP-9 was present in all subjects. High-risk pathogens of the PPM-group such as P. aeruginosa, S. aureus, and A. fumigatus are known to be associated with an increased mortality (2). However, we did not find significant differences in mortality rates between the PPM and non-PPM group, which may be due to the limited size of our study group. Direct contribution to the production of MMPs by the pathogens themselves seems rather unlikely. Until now it has not been shown that MMP-8 or MMP-9 are produced by any of the detected pathogens. However, it is known that pseudomonal elastase that is secreted by P. aeruginosa proteolytically activates pro–MMP-9 (12). Thus, bacterial pathogens might induce activation of MMPs, which goes together with our data that in all patients of the PPM-group the activated form of MMP-9 was detectable.

The results of our study support our assumption that neutrophils are the main source of MMPs in pneumonia. First, zymographic analysis of mini-BALF showed the characteristic banding pattern of neutrophil-derived MMP-9 consisting of pro–MMP-9 (92 kD), human neutrophil lipocalin–pro-MMP-9-complex (130 kD), and a homodimeric form of MMP-9 (220 kD). Human neutrophil lipocalin is a 25-kD lipocalin that is stored in the secondary granules of PMN (32). It is associated with neutrophil gelatinases and is thus a marker for PMN-released MMP-9. It may play a role in the defense against microorganisms (32), but its function has not been fully characterized until now.

Second, the correlations between the percentage of neutrophils and the concentrations of MMP-8 and MMP-9 in mini-BALF suggest a possible association. Suga and coworkers found similar results in idiopathic interstitial pneumonia (33) as did Betsuyaku and coworkers in patients with emphysema (21). Third, comparison of MMP levels with lactoferrin, a glycoprotein secreted by PMN, showed a significant correlation between MMP-8 and lactoferrin in mini-BALF of patients with pneumonia but not between MMP-9 and lactoferrin. MMP-8 and lactoferrin are both exclusively stored in secondary neutrophil granules, whereas MMP-9 is also stored in specific gelatinase granules, which might be the reason for the lack of correlation. These data support the assumption that MMP release in pneumonia is mainly neutrophil-derived.

Besides, we demonstrated a difference between the MMP release of peripheral and pulmonary PMN in patients with pneumonia. Basal MMP-8 and MMP-9 release of pulmonary PMN was significantly higher than of peripheral PMN. After stimulation, pulmonary PMN were not able to release more MMP. In contrast, peripheral PMN released more MMP-8 and MMP-9. These data show that the pulmonary neutrophils of patients with pneumonia were already highly activated. In accordance with a previous study about MMP release by peripheral PMN in patients with asthma and chronic obstructive pulmonary disease (34), we showed no differences in release of MMPs by peripheral neutrophils from patients with pneumonia and control subjects, neither basally nor after stimulation.

The excess of MMPs in hospital-acquired pneumonia may have various consequences. It is well known that in case of pulmonary inflammation PMN are recruited to the alveolar compartment (5). It has been shown that the gelatinases MMP-9 and MMP-2 play a key role in migration of different leukocyte populations like T cells and neutrophils (6, 35). It is assumed that the release of MMP-8 and MMP-9 leads to a disruption of basement membrane components and thus facilitates neutrophil migration into extravascular tissue (6). Delclaux and coworkers demonstrated in vitro that MMP-9 is a major factor in PMN migration through basement membranes and that TIMP-1 inhibits PMN migration (36). Keck and coworkers showed that inhibition of MMPs by batimastat (BB-94)—a selective synthetic inhibitor of MMPs—inhibits PMN transmigration and reduces the alveolocapillary leakage (37). However, other authors question the importance of MMPs in basement membrane migration (38, 39). We did not directly investigate the role of MMPs in PMN migration and membrane degradation, but our data on 2-macroglobulin levels in mini-BALF indicate significant alveolocapillary leakage in patients with pneumonia. According to our results and previous studies we speculate that PMN are attracted to pulmonary capillaries by inflammatory mediators that also lead to PMN degranulation with exocytosis of MMPs (13). The released MMPs might then facilitate PMN migration into the alveolar compartment where other contents of PMN granules and reactive oxygen species can express their antimicrobial function. Whether MMPs themselves also have direct bactericidal properties is still not clear (15). However, in addition to the role of MMPs in the degradation of the extracellular matrix, MMPs have important regulatory functions. Opdenakker and coworkers showed that MMP-9 truncates chemokines such as interleukin-8 into a more active form, leading to increased MMP-9 levels that again lead to a more potent interleukin-8 (13).

Although MMPs might thus be essential for PMN migration, regulatory and effector functions, continuing inflammation and the positive feedback loop between interleukin-8 and MMP-9 may maintain excessive MMP activity. This finally can lead to severe extracellular matrix destruction and alveolocapillary leakage with increased risk of acute respiratory distress syndrome, the most serious complication of pneumonia. Torii and coworkers described high MMP-2 and MMP-9 levels in BALF of patients with acute respiratory distress syndrome. The MMP-9 levels were correlated with the number of neutrophils as well as with 7S collagen, a marker of basement membrane disruption. The authors strongly suggested that the increased concentration of collagenolytic enzymes in the lung is implicated in the pathogenesis of acute respiratory distress syndrome through the disruption of basement membrane structures (16). Keck and coworkers showed that MMP-9 is involved in pancreatitis-associated lung injury, a condition that is clinically and pathologically similar to acute respiratory distress syndrome (37). Finally, in a recent study attenuation of ventilator-induced lung injury by the synthetic MMP inhibitor Prinomastat was demonstrated in rats (40).

In summary we found excessive concentrations of MMP-8 and MMP-9 without a counterbalancing increase of TIMP-1 within the pulmonary compartment in hospital-acquired pneumonia. This was related to the detection of PPM and clinical severity. Extracellular matrix destruction may initiate a vicious cycle of release of proinflammatory cytokines, which by itself stimulates PMN invasion into the alveolar compartment and the expression of MMPs with consecutive aggravation of the inflammatory response. These findings may contribute toward understanding the role of MMPs in the pathogenesis of lung damage in hospital-acquired pneumonia. Further studies with larger patient groups are necessary to assess the prognostic impact of MMPs in hospital-acquired pneumonia.

	Acknowledgments

 
The authors thank Dr. T. Kohlmann for statistical advice.

	REFERENCES

TOP ABSTRACT METHODS RESULTS DISCUSSION REFERENCES

 

1. Lynch JP. Hospital-acquired pneumonia: risk factors, microbiology, and treatment. Chest 2001;119:373S–384S.

2. American Thoracic Society. Hospital-acquired pneumonia in adults: diagnosis, assessment of severity, initial antimicrobial therapy, and preventive strategies. Am J Respir Crit Care Med 1996;153:1711–1725.

3. Weinmeister KD, Dal Nogare AR. Buccal cell carbohydrates are altered during critical illness. Am J Respir Crit Care Med 1994;150:131–134.

4. Lehrer RI, Ganz T, Selsted ME, Babior BM, Curnutte JT. Neutrophils and host defense. Ann Intern Med 1988;109:127–142.

5. Brown GM, Brown DM, Donaldson K, Drost E, MacNee W. Neutrophil sequestration in rat lungs. Thorax 1995;50:661–667.

6. Borregaard N, Cowland JB. Granules of the human neutrophilic polymorphonuclear leukocyte. Blood 1997;89:3503–3521.

7. Braun J, O'Connor C. Measurement of proteases and antiproteases in bronchoalveolar lavage fluid. Eur Respir Rev 1999;9:76–85.

8. Ray JM, Stetler-Stevenson WG. The role of matrix metalloproteases and their inhibitors in tumour invasion, metastasis and angiogenesis. Eur Respir J 1994;7:2062–2072.

9. Shapiro SD, Senior RM. Matrix metalloproteinases: matrix degradation and more. Am J Respir Cell Mol Biol 1999;20:1100–1102.

10. Nagase H. Activation mechanisms of matrix metalloproteinases. Biol Chem 1997;378:151–160.

11. Pugin J, Widmer MC, Kossodo S, Liang CM, Preas HL II, Suffredini AF. Human neutrophils secrete gelatinase B in vitro and in vivo in response to endotoxin and proinflammatory mediators. Am J Respir Cell Mol Biol 1999;20:458–464.

12. Miyajima S, Akaike T, Matsumoto K, Okamoto T, Yoshitake J, Hayashida K, Negi A, Maeda H. Matrix metalloproteinases induction by pseudomonal virulence factors and inflammatory cytokines in vitro. Microb Pathog 2001;31:271–281.

13. Opdenakker G, Van den Steen PE, Dubois B, Nelissen I, Van Coillie E, Masure S, Proost P, Van Damme J. Gelatinase B functions as regulator and effector in leukocyte biology. J Leukoc Biol 2001;69:851–859.

14. Wojtowicz-Praga SM, Dickson RB, Hawkins MJ. Matrix metalloproteinase inhibitors. Invest New Drugs 1997;15:61–75.

15. Price B, Dennison C, Tschesche H, Elliott E. Neutrophil tissue inhibitor of matrix metalloproteinases-1 occurs in novel vesicles that do not fuse with the phagosome. J Biol Chem 2000;275:28308–28315.

16. Torii K, Iida K, Miyazaki Y, Saga S, Kondoh Y, Taniguchi H, Taki F, Takagi K, Matsuyama M, Suzuki R. Higher concentrations of matrix metalloproteinases in bronchoalveolar lavage fluid of patients with adult respiratory distress syndrome. Am J Respir Crit Care Med 1997;155:43–46.

17. Sepper R, Konttinen YT, Ding Y, Takagi M, Sorsa T. Human neutrophil collagenase (MMP-8), identified in bronchiectasis BAL fluid, correlates with severity of disease. Chest 1995;107:1641–1647.

18. Delacourt C, Le Bourgeois M, D'Ortho MP, Doit C, Scheinmann P, Navarro J, Harf A, Hartmann DJ, Lafuma C. Imbalance between 95 kDa type IV collagenase and tissue inhibitor of metalloproteinases in sputum of patients with cystic fibrosis. Am J Respir Crit Care Med 1995;152:765–774.

19. Choi KH, Lee HB, Jeong MY, Rhee YK, Chung MJ, Kwak YG, Lee YC. The role of matrix metalloproteinase-9 and tissue inhibitor of metalloproteinase-1 in cryptogenic organizing pneumonia. Chest 2002;121:1478–1485.

20. Segura-Valdez L, Pardo A, Gaxiola M, Uhal BD, Becerril C, Selman M. Upregulation of gelatinases A and B, collagenases 1 and 2, and increased parenchymal cell death in COPD. Chest 2000;117:684–694.

21. Betsuyaku T, Nishimura M, Takeyabu K, Tanino M, Venge P, Xu S, Kawakami Y. Neutrophil granule proteins in bronchoalveolar lavage fluid from subjects with subclinical emphysema. Am J Respir Crit Care Med 1999;159:1985–1991.

22. Torres A, Bauer TT, Leon-Gil C, Castillo F, Alvarez-Lerma F, Martinez-Pellus A, Leal-Noval SR, Nadal P, Palomar M, Blanquer J, et al. Treatment of severe nosocomial pneumonia: a prospective randomised comparison of intravenous ciprofloxacin with imipenem/cilastatin. Thorax 2000;55:1033–1039.

23. Pugin J, Auckenthaler R, Mili N, Janssens JP, Lew PD, Suter PM. Diagnosis of ventilator-associated pneumonia by bacteriologic analysis of bronchoscopic and nonbronchoscopic "blind" bronchoalveolar lavage fluid. Am Rev Respir Dis 1991;143:1121–1129.

24. Dalhoff K, Braun J, Kothe H, Korber M, Pein M, Wiessmann KJ. Oxidative metabolism of pulmonary phagocytes in acute pneumonia. Respiration 1994;61:144–149.

25. Papazian L, Thomas P, Garbe L, Guignon I, Thirion X, Charrel J, Bollet C, Fuentes P, Gouin F. Bronchoscopic or blind sampling techniques for the diagnosis of ventilator-associated pneumonia. Am J Respir Crit Care Med 1995;152:1982–1991.

26. English D, Andersen BR. Single-step separation of red blood cells: granulocytes and mononuclear leukocytes on discontinuous density gradients of Ficoll-Hypaque. J Immunol Methods 1974;5:249–252.

27. Carlson M, Hakansson L, Peterson C, Stalenheim G, Venge P. Secretion of granule proteins from eosinophils and neutrophils is increased in asthma. J Allergy Clin Immunol 1991;87:27–33.

28. Leber TM, Balkwill FR. Zymography: a single-step staining method for quantitation of proteolytic activity on substrate gels. Anal Biochem 1997;249:24–28.

29. Braun J, Schultek T, Tegtmeier KF, Florenz A, Rohde C, Wood WG. Luminometric assays of seven acute-phase proteins in minimal volumes of serum, plasma, sputum, and bronchioalveolar lavage. Clin Chem 1986;32:743–747.

30. Bosse M, Chakir J, Rouabhia M, Boulet LP, Audette M, Laviolette M. Serum matrix metalloproteinase-9:tissue inhibitor of metalloproteinase-1 ratio correlates with steroid responsiveness in moderate to severe asthma. Am J Respir Crit Care Med 1999;159:596–602.

31. Vignola AM, Riccobono L, Mirabella A, Profita M, Chanez P, Bellia V, Mautino G, D'Accardi P, Bousquet J, Bonsignore G. Sputum metalloproteinase-9/tissue inhibitor of metalloproteinase-1 ratio correlates with airflow obstruction in asthma and chronic bronchitis. Am J Respir Crit Care Med 1998;158:1945–1950.

32. Cowland JB, Borregaard N. Molecular characterization and pattern of tissue expression of the gene for neutrophil gelatinase-associated lipocalin from humans. Genomics 1997;45:17–23.

33. Suga M, Iyonaga K, Okamoto T, Gushima Y, Miyakawa H, Akaike T, Ando M. Characteristic elevation of matrix metalloproteinase activity in idiopathic interstitial pneumonias. Am J Respir Crit Care Med 2000;162:1949–1956.

34. Cataldo D, Munaut C, Noel A, Frankenne F, Bartsch P, Foidart JM, Louis R. Matrix metalloproteinases and TIMP-1 production by peripheral blood granulocytes from COPD patients and asthmatics. Allergy 2001;56:145–151.

35. Leppert D, Waubant E, Galardy R, Bunnett NW, Hauser SL. T cell gelatinases mediate basement membrane transmigration in vitro. J Immunol 1995;154:4379–4389.

36. Delclaux C, Delacourt C, D'Ortho MP, Boyer V, Lafuma C, Harf A. Role of gelatinase B and elastase in human polymorphonuclear neutrophil migration across basement membrane. Am J Respir Cell Mol Biol 1996;14:288–295.

37. Keck T, Balcom JH, Fernandez-del Castillo C, Antoniu BA, Warshaw AL. Matrix metalloproteinase-9 promotes neutrophil migration and alveolar capillary leakage in pancreatitisassociated lung injury in the rat. Gastroenterology 2002;122:188–201.

38. Betsuyaku T, Shipley JM, Liu Z, Senior RM. Neutrophil emigration in the lungs, peritoneum, and skin does not require gelatinase B. Am J Respir Cell Mol Biol 1999;20:1303–1309.

39. Mackarel AJ, Cottell DC, Russell KJ, FitzGerald MX, O'Connor CM. Migration of neutrophils across human pulmonary endothelial cells is not blocked by matrix metalloproteinase or serine protease inhibitors. Am J Respir Cell Mol Biol 1999;20:1209–1219.

40. Foda HD, Rollo EE, Drews M, Conner C, Appelt K, Shalinsky DR, Zucker S. Ventilator-induced lung injury upregulates and activates gelatinases and EMMPRIN: attenuation by the synthetic matrix metalloproteinase inhibitor, Prinomastat (AG3340). Am J Respir Cell Mol Biol 2001;25:717–724.