Inflammation is a localized, protective response to trauma or microbial invasion that destroys the injurious agent and the injured tissue. Neutrophil elastase (NE), a serine protease stored in the azurophil granules of polymorphonuclear neutrophils, digests microbes after phagocytosis. NE can also digest microbes extracellularly but is associated with tissue damage and inflammatory disease. In this study, we show that polymorphonuclear neutrophils from mice deficient in serine protease inhibitor 6, a weak intracellular NE inhibitor, had increased susceptibility to self-inflicted lysis because of increased NE activity. The resulting transient increase in local extracellular NE activity was within a narrow range that resulted in the clearance of Pseudomonas aeruginosa but did not damage the lung. Therefore, deficiency in a weak intracellular inhibitor of NE results in an acute inflammatory response that protects from P. aeruginosa but does not cause lung disease.

Polymorphonuclear neutrophils (PMN)3 provide the first line of defense from infection through the elimination of bacteria and fungi. The microbicidal molecules used by PMNs are grouped into oxygen-dependent and oxygen-independent systems (1). The latter system comprises serine proteases such as neutrophil elastase (NE), cathepsin G (Cat G), and proteinase 3 (PR3) and other polypeptides that alter microbe structural integrity (1). In mice, NE is required for host defense against Gram-negative bacteria (2) and kills Escherichia coli by degrading its major outer membrane protein A after phagocytosis (3). Activated PMNs also release a variety of molecules, including NE, during their egress from the vasculature, when phagocytosis is impaired or the dose of microbe is overwhelming (4). The release of NE during inflammation causes a proteolytic environment, which can kill microbes and modulate the inflammatory response. However, a wide substrate repertoire has lead to the incrimination of NE in the tissue damage that gives rise to inflammatory diseases. Diseases associated with infection with Pseudomonas aeruginosa in the lung typify this situation. NE is thought to protect from P. aeruginosa through the degradation of flagellin (5) but is also associated with tissue destruction in pneumonia-mediated acute lung injury and cystic fibrosis (CF) (6).

The lifespan of a PMN is typically only ∼6 h after they leave the circulation, and so large numbers are rapidly recruited from the blood to the site of infection (7) then disposed of by apoptotic programmed cell death (PCD) (8, 9). It is assumed that the physiological reason for preventing the self-lysis of PMNs is 2-fold. First, to ensure sufficient viability so that they can move to the site of infection then phagocytose and eliminate microbes, Second, to safely remove damaged PMNs and avoid the release of NE and other noxious agents that may cause tissue damage and inflammatory disease. How PMN breakdown is prevented by protection from self-inflicted damage is not well understood.

One clue for how PMNs avoid self-inflicted damage may come from recent insights into how CTLs avoid being killed by their own toxic effector serine proteases such as granzyme B (GrB) (10). Regulation of serine proteases is mainly achieved through inhibitors of the serine protease inhibitor (serpin) superfamily (11). Several serpins with homology to OVA (ova-serpins) lack a traditional signal peptide sequence and are located in the cytoplasm of cells (12). The human ova-serpin proteinase inhibitor PI9 is a potent inhibitor of GrB (13). In mice, serine protease inhibitor 6 (Spi6), a homolog of PI9, also inhibits GrB in vitro (14) and is expressed in CTLs and NK cells (15). Using mice deficient in Spi6, we have shown that by inhibiting GrB, Spi6 protects CTLs from self-inflicted injury (10). The increased programmed cell death of Spi6 CTLs resulted in impaired survival and a decrease in CTL-effector function, most notably in impaired clearance of virus. Therefore, for CTLs, protection by Spi6 from self-inflicted damage is required for effector function.

PI9 is expressed in PMNs (16) and is also a weak inhibitor of NE (17). Therefore, suppression of NE by endogenous serpins may be a mechanism by which PMNs avoid self-inflicted damage. In the present study, experiments with Spi6 knockout (KO) mice revealed Spi6 as a weak intracellular inhibitor of NE in PMNs. Spi6 deficiency rendered PMNs susceptible to lysis. Direct instillation into the lung demonstrated a role for extracellular NE in the elimination of P. aeruginosa. Therefore, despite the susceptibility of Spi6 KO PMNs to disintegration, the subsequent release of NE protected from lethal pneumonia caused by P. aeruginosa by increased clearance from the lung. The acute inflammatory response of Spi6 KO mice to P. aeruginosa did not give disease because NE was increased to within a narrow protective range that did not damage tissue. These findings suggest that increased PMN immunity to P. aeruginosa can have a positive impact on disease.

Spi6 KO mice were generated in the C57BL/6 background by homologous recombination (10). Spi6 KO mice were crossed with NE KO mice (C57BL/6/129/Sv) (2) to generate NE−/−Spi6−/− (NE KO × Spi6 KO), NE−/−Spi6+/− (NE KO), NE+/−Spi6−/− (Spi6 KO), and NE+/−Spi6+/− wild-type (WT) mice (all C57BL/6/129/Sv F2 background). WT C57BL/6 mice (B6 mice) were purchased from Jackson Laboratory. All mice were maintained and bred under standard specific pathogen-free conditions. All experiments with mice were performed in compliance with the University of Chicago Institutional Animal Care and Use Committee regulations.

Recombinant Spi6 was generated in the pGEX-3X expression system in E. coli as a fusion protein with glutathione transferase (GST) and purified to homogeneity (>90% pure), using standard procedures recommended by the manufacturer (Amersham Bioscience). Rabbits were primed with 0.35 mg of recombinant GST-Spi6 in Freund’s complete adjuvant then boosted twice with 0.15 mg then finally with 0.35 mg in Freund’s incomplete adjuvant.

Thioglycollate (TG) (3%) (Difco Laboratories) was injected i.p (0.7 ml/20 g mouse) and PMNs were harvested after 4 h (107/mouse; 80% pure) (18). Total detergent extracts of PMNs were prepared in 10 mM Tris-HCl (pH 7.5), 150 mM NaCl, 10% glycerol, 1% Triton X-100, 2 mM EDTA, 1 mM PMSF, and 1× protease inhibitor cocktail (Roche), then nuclei removed by centrifugation at 3000 × g for 20 min. Protein extracts from PMNs (100 μg per lane) or human embryonic kidney epithelial 293 cells (19) (obtained from American Type Tissue Culture Collection) (15 μg) transfected with DNA encoding 3×FLAG-tagged Spi6 were resolved by SDS-PAGE (10%) then immunoblotted against anti-Spi6 rabbit serum (1/500 dilution) overnight in 5% skimmed milk. Spi6 was detected after probing with goat anti-rabbit IgG conjugated to HRP (2 μg/ml) and chemiluminescence (ECL kit; Amersham Biosciences). To control for equal loading, blots were stripped and re-probed for actin mAb clone ACTN05 (RDI Research Diagnostics) at 0.5 μg/ml and anti-mouse IgG HRP (Sigma-Aldrich) at 2 μg/ml. Endogenous Spi6 is detected as a 42-kDa protein and 3×FLAG-Spi6 as a 44-kDa protein.

PMNs were activated (2 × 106) with P. aeruginosa (2 × 107) at 37°C for 1–6 h then lysed by sonication in hypotonic buffer (50 mM PIPES, 50 mM KCL, 5 mM EGTA, 2 mM MgCl2, 5 mM DTT, pH 7.6). Lysates were centrifuged at 3000 × g for 20 min to remove nuclei then 15,000 × g for 30 min and supernatant (cytosol) removed (10). Protein concentration was determined by Lowry assay (Bio-Rad).

Recombinant Spi6 was generated in the pGEX-3X expression system in E. coli as a fusion protein with GST, using standard procedures recommended by the manufacturer (Amersham Bioscience). The GST tag was removed by factor X proteolysis and recombinant (r) Spi6 (44 kDa) purified to homogeneity. Purified human NE (HNE) activity (Athens Research and Technology) was determined by measuring the hydrolysis of MeOSuc-AAPV-AMC (200 μM) (Calbiochem) at 25°C in 0.1 M HEPES, 0.15 M NaCl, 0.1% PEG 8000 (pH 7.4) using a Hitachi F-200 fluorescence spectrophotometer (excitation 380 nm and emission 440 nm).

HNE activity (20 nM) decreased with time to an endpoint of zero activity in accordance with the exponential decay function (1) after incubation with recombinant Spi6.

\[\mathrm{percentage\ of\ activity}{=}100{\times}e^({-}k_{\mathrm{obs}}\ {\times}\ \mathrm{t})\]

(kobs = observed pseudo-first order rate constant, t = time).

HNE incubated alone showed insignificant losses in activity. Reactions performed at different fixed concentrations (400, 600, and 800 nM) of Spi6 gave pseudo-first order rate constants (kobs), which increased in proportion to the Spi6 concentration. An apparent second order inhibition rate constant (kapp) 1.0x ± 0.7 × 104 M−1s−1 was determined from the reaction (2)

\[k_{\mathrm{obs}}\ {=}\ k_{\mathrm{app}}\ {\times}\ {[}\mathrm{Spi6}{]}_{o}\]

(kapp = apparent second order rate constant for the inhibition reaction; [Spi6]o is the nominal concentration of Spi6 in the reaction).

To determine the stoichiometry of the inhibition (SI) reaction, HNE (20 nM) was incubated with varying molar ratios of Spi6 (1.25 to 10× HNE), insufficient to completely inhibit the enzyme, then enzyme activity followed until an endpoint activity was reached. The abscissa intercept of this plot indicated that complete inhibition of elastase activity required ∼15 mol of Spi6 per mol of enzyme.

NE activity in subcellular fractions, tissue culture supernatant, or bronchoalveolar lavage (BAL) was determined by measuring the hydrolysis of MeOSuc-AAPV-pNA (200 μM) (Calbiochem) at 25°C in 20 mM HEPES, 0.15 M NaCl, 0.01% BSA, 2% DMSO, 1% dimethylformamide (pH 7.5), using a Spectra MAX 250 spectrophotometer (Molecular Devices) at 405 nm. One unit of NE activity is defined as the amount of enzyme that hydrolyzes defined 1 pmol/min labeled substrate under these conditions.

Cat G activity in tissue culture supernatant was determined by measuring the hydrolysis of MeO-Suc-Ala-Ala-Pro-Val-pNA (2 μM) (Calbiochem) at 25°C in 20 mM HEPES, 0.15 M NaCl, 0.01% BSA, 2% DMSO, 1% dimethylformamide (pH 7.5) at 405 nm. One unit of Cat G activity is defined as the amount of enzyme that hydrolyzes defined 1 pmol/min labeled substrate under these conditions.

Lysozyme activity was determined by measuring the hydrolysis of Micrococcus lysodeikiticus cell wall (25 μg/ml) labeled with fluorescein (EnzCheck lysozyme assay kit; Molecular Probes) at 25°C in 0.1 M sodium phosphate, 0.1 M NaCl (pH 7.5). The activity of lysozyme in tissue culture supernatant was calculated from standard curves of purified lysozyme vs fluorescence at 518 nm according to the instructions of the manufacturer.

The percentage of dead PMNs was determined by staining with propidium iodide (PI-positive = dead), according to the manufacturer’s instructions (Molecular Probes). The lysis of PMNs was determined by assaying lactate dehydrogenase (LDH) activity in tissue culture supernatant according to the manufacturer’s instructions (Promega). PMNs were stained with annexin VFITC and PI to determine the percentage of apoptotic cells (PIAnnexin V+) according to the manufacturer’s instructions (BD Pharmingen). Assays for caspase 3 were performed on cytosol extracts from PMNs in reaction buffer (10 mM PIPES (pH 7.4), 8 mM DTT, 2 mM EDTA, 0.1% CHAPS) at 30°C using Ac-DEVD-pNA (Calbiochem) at 0.2 mM (10, 20). Specific activity was determined by normalizing for the amount of protein. Units of activity were as defined before (10). Activation specific death (percentage of PI+), lysis (LDH activity), or apoptosis (percentage of YOPRO-1+ or caspase 3 activity) were determined by subtracting the measurement from PMN incubated alone from that observed after incubation with P. aeruginosa for 6 h.

P. aeruginosa was grown overnight in 10 ml of LB at 37¡ C then 5 ml of overnight culture was added to 5 ml of fresh Luria-Bertani (LB) and grown for ∼3 h to OD 0.2–0.3. Bacteria were washed twice then resuspended in PBS at a concentration of 1.0 OD/ml. Mice (6–8 wk old; 16–18 g) were infected with a lethal dose of P. aeruginosa (intranasally (i.n.), 50 μl, 5 × 107 CFU/mouse) (5) and survival measured over time. P. aeruginosa titers were determined on LB plates after incubation at 37°C overnight. In brief, the whole right lung was dissected and put in PBS (5 ml), homogenized, then serially diluted, and plated on LB plates incubated at 37°C overnight. Plates with 10–300 colonies per plate were counted, and the titer was determined based on the dilution. BAL was recovered by lavaging twice with PBS (1 ml) after intratracheal injection (5). Leukocyte numbers were determined in BAL using standard cytometric analysis after lethal infection with P. aeruginosa. After lavage, lung histology was performed, mid-sagittal sections of the left lung fixed with PFA and stained with H&E. Mice were infected with P. aeruginosa, then after 4 h, instilled with HNE (40 μl in PBS) by i.n. injection and analyzed over time.

Survival curves were compared using the log rank test and significance of difference indicated by two-tailed p values (Prism version 4.0). The significance of difference for single means was measured using two-tailed Student’s t tests.

The ova-serpin PI9 is the human homologue of Spi6, which in addition to inhibiting GrB (13) can also inhibit NE (17). We determined whether Spi6 is a physiologically relevant inhibitor of NE, by examining NE activity in PMNs from C57BL/6 Spi6 KO mice (10). PMNs were elicited from the peritoneum by TG injection then harvested after 4 h (18). The specificity of anti-Spi6 anti-serum was first confirmed on Spi6 cDNA-transfected 293 cells by Western blot, and then used to show the expression of Spi6 in PMNs from B6 mice (42 kDa) but not Spi6 KO mice (Fig. 1,A). Spi6 lacks a signal peptide and like other OVA serpins is located in the cytoplasm (10). Therefore, we determined whether Spi6 could suppress the activity of cytoplasmic NE, which we have previously detected in activated mouse PMNs (2). TG-elicited peritoneal PMNs were stimulated by P. aeruginosa then at various time points cytosolic extracts prepared after detergent lysis (10). Enzyme assays revealed a 3-fold increase in NE specific activity in cytosolic fractions from Spi6 KO compared with B6 PMNs (Fig. 1 B). Therefore, we conclude that Spi6 is a physiological inhibitor of cytoplasmic NE in P. aeruginosa activated PMNs.

FIGURE 1.

Spi6 is an intracellular inhibitor of NE. A, Western blots of extracts from B6 or Spi6 KO PMNs and 293T cells (mock transfected or transfected with 3xFLAG-Spi6 DNA). Upper blot probed with anti-Spi6 and lower blot with anti-actin Abs. Arrows indicate the presence of endogenous Spi6 (42 kDa) and 3xFLAG-Spi6 (44 kDa). B, Mean specific activity (SA) of NE from the cytosol of PMNs ± SEM (n = 4 mice) after stimulation in vitro with P. aeruginosa. Kinetics (Spi6 400 nM) (C) and stoichiometry of inhibition of HNE activity (D) (20 nM) by recombinant Spi6.

FIGURE 1.

Spi6 is an intracellular inhibitor of NE. A, Western blots of extracts from B6 or Spi6 KO PMNs and 293T cells (mock transfected or transfected with 3xFLAG-Spi6 DNA). Upper blot probed with anti-Spi6 and lower blot with anti-actin Abs. Arrows indicate the presence of endogenous Spi6 (42 kDa) and 3xFLAG-Spi6 (44 kDa). B, Mean specific activity (SA) of NE from the cytosol of PMNs ± SEM (n = 4 mice) after stimulation in vitro with P. aeruginosa. Kinetics (Spi6 400 nM) (C) and stoichiometry of inhibition of HNE activity (D) (20 nM) by recombinant Spi6.

Close modal

To determine whether Spi6 can directly inhibit NE, the ability of recombinant Spi6 to inhibit purified HNE was examined under pseudo-first order kinetic conditions. The mean apparent second order inhibition rate constant (kapp) was determined as 1.0 x ± 0.7 × 104 M−1 s−1 (n = 3 determinations) (Fig. 1,C), at a SI of ∼15:1 (Fig. 1 D), giving a corrected rate of enzyme inhibition of 1.5 × 105 M−1 s−1. Compared with other NE-specific extracellular serpins such as α1-anti-trypsin (α1-AT) (kapp = 107 M−1 s−1, SI = 1:1) (21) and secretory leukoprotease inhibitor (22) (kapp = 2 × 106 M−1 s−1, SI = 1:1), Spi6 is a relatively weak inhibitor of NE.

Spi6 protects CTLs from self-inflicted damage by suppressing GrB activity in the cytoplasm (10). We determined whether suppression of NE by Spi6 protected PMNs from self-inflicted damage by examining the viability of Spi6 KO PMNs. TG-elicited PMNs were activated with P. aeruginosa then cell death measured by detecting the uptake of propidium iodide (PI) (23, 24). The percentage of Spi6 KO PMNs undergoing death (PI+) was consistently higher than B6 control PMNs from ∼2 h after P. aeruginosa activation (Fig. 2,A). For example, 4 h after P. aeruginosa activation the percentage of PI+ Spi6 KO PMNs was approximately three times higher (p = 0.002) than B6 PMNs. The increased onset of Spi6 KO PMNs death resulted in increased cellular lysis and disintegration, as evidenced by increased release of the cytosolic marker LDH (Fig. 2 B). Thus, after 6 h there was about a 4-fold increase in Spi6 KO PMN lysis over controls (p = 0.0004).

FIGURE 2.

Spi6 KO PMNs are susceptible to lysis. A, Death (mean percentage PI+) ± SEM (n = 4 mice); B, lysis (mean extracellular LDH activity) ± SEM (n = 4 mice); and C, activation-specific death (mean percentage PI+) ± SEM (n = 4 mice); D, lysis (mean extracellular LDH activity) ± SEM (n = 4 mice) and apoptosis measured by mean percentage of annexin V+ PIof PMNs ± SEM (n = 4 mice) (E) or specific activity (SA) of caspase 3, 6 h after stimulation in vitro with P. aeruginosa (F).

FIGURE 2.

Spi6 KO PMNs are susceptible to lysis. A, Death (mean percentage PI+) ± SEM (n = 4 mice); B, lysis (mean extracellular LDH activity) ± SEM (n = 4 mice); and C, activation-specific death (mean percentage PI+) ± SEM (n = 4 mice); D, lysis (mean extracellular LDH activity) ± SEM (n = 4 mice) and apoptosis measured by mean percentage of annexin V+ PIof PMNs ± SEM (n = 4 mice) (E) or specific activity (SA) of caspase 3, 6 h after stimulation in vitro with P. aeruginosa (F).

Close modal

We next determined whether the susceptibility to death was due to the increased NE activity in Spi6 KO PMNs. PMNs were susceptible to cell death upon in vitro culture; therefore, to measure PMN death specifically due to activation by P. aeruginosa we subtracted the percentage of PI+ PMNs cultured alone from the percentage cultured with P. aeruginosa after 6 h (referred to as activation-specific cell death). The effect of NE deficiency on the activation-specific cell death by P. aeruginosa was examined in NE KO x Spi6 KO double-deficient mice, which were generated by crossing NE KO mice (C57BL/6/129/Sv) (2) with Spi6 KO mice (C57BL/6 mice). In all experiments, double KOs, single KO and WT littermate controls were in the C57BL/6/129/Sv background. We observed a complete absence of activation-specific death observed in NE KO PMNs (C57BL/6/129/Sv F2 background), indicating a role for NE in inducing the death of activated WT PMNs (C57BL/6/129/Sv F2 background) (Fig. 2,D). In contrast, Spi6 KO PMNs (C57BL/6/129/Sv F2 background) underwent ∼3-fold (p = 0.03) more activation-induced cell death compared with WT PMNs. The susceptibility of Spi6 KO PMNs to activation-induced cell death could be corrected by NE deficiency, as evidenced by an absence of cell death in NE KO x Spi6 KO PMNs (Fig. 2 D). Therefore, we conclude that protection from NE is a physiological mechanism by which Spi6 protects PMNs from cell death induced by bacterial activation.

To measure PMN apoptosis we detected the expression of phosphotidylserine on the outer plasma membrane by annexin V staining (23) and the activation of caspase 3 in the cytoplasm (10). Although we observed an increase in the activation-specific death (Fig. 2,D) or lysis (Fig. 2,E) Spi6 KO PMNs, we did not observe any increase in activation-specific annexin V staining (Fig. 2,E) or caspase 3 activation (Fig. 2 F). We conclude that Spi6 protects PMNs from self-inflicted, lysis by suppressing NE-dependent necrotic death.

We wanted to determine whether the increased susceptibility of Spi6 KO PMNs to lysis increased the release of microbicidal components. Concomitant with increased cellular lysis, there was a 6-fold (p = 3 × 10−5) increase in NE activity in the extracellular environment of Spi6 KO PMNs as a result of increased lysis (Fig. 3,A). We also observed a similar increase in the release of another effector serine protease, Cat G (p = 1 × 10−5) (Fig. 3,B) and lysozyme (p = 2 × 10−5) (Fig. 3,C), which is also stored in azurophilic granules. The substrates of lysozyme are mucopolysaccarides and not serine proteases (1), and so it is very unlikely that Spi6, which is a serpin, cross-inhibits lysozyme. Therefore, we conclude that increased lysis of Spi6 KO PMNs leads to the nonspecific release of cellular components, which include cytoplasmic markers such as LDH (Fig. 2,B), and effector molecules such as NE, Cat G and lysozyme (Fig. 3), which are normally stored in azurophilic granules.

FIGURE 3.

Increased release of microbicidal enzymes by Spi6 KO PMNs. Mean activity of NE (A), Cat G (B), and lysozyme (C) 6 h after stimulation in vitro with P. aeruginosa. Mean activities are ± SEM (n = 4 mice).

FIGURE 3.

Increased release of microbicidal enzymes by Spi6 KO PMNs. Mean activity of NE (A), Cat G (B), and lysozyme (C) 6 h after stimulation in vitro with P. aeruginosa. Mean activities are ± SEM (n = 4 mice).

Close modal

To examine the physiological relevance of the increased release of azurophilic enzymes by Spi6 KO PMNs, we first determined whether such microbicidal enzymes can directly kill a bacterial pathogen in vivo. NE can degrade the essential flagellin protein of P. aeruginosa (5), and so it is possible that as with another Gram-negative bacterium E. coli, (2, 3), extracellular NE may directly kill P. aeruginosa at the site of infection. To test this, we determined the effect of direct intranasal instillation of control B6 mice with purified HNE on infection by P. aeruginosa. We found that the survival curves after P. aeruginosa infection at the doses of 1.8 (p = 0.0025) and 2.4 U/kg HNE (p = 0.0039) were significantly different from those of control-treated mice (Fig. 4,A). Whereas the survival curves of mice treated with the lowest dose of 1.2 (p = 0.251) or the highest dose of 5.9 U/kg HNE (p = 0.821) were not significantly different from those of undosed mice (Fig. 4,A). The increased survival of mice dosed over the range of 1.8–2.4 U/kg HNE from pneumonia was due to increased clearance of P. aeruginosa in the lung (Fig. 4,B). Although the high dose of HNE (5.9 U/kg) resulted in efficient clearance of P. aeruginosa (Fig. 4,B) it induced considerable lung damage, characterized by destruction of normal alveolar structure (Fig. 5,B), and so did not increase survival over that of undosed mice (Fig. 4,A). This lung damage was comparable to that seen in previous HNE instillation studies in animals (25) and in human chronic obstructive pulmonary disease (COPD) (6). The beneficial inflammatory response induced by instillation with HNE at a dose of 1.8 U/kg was characterized by an early transient increase in NE activity (1.4-fold higher than undosed mice, p = 0.02) and PMN numbers (1.7-fold higher than un-dosed mice, p = 0.02) as measured in BAL (Table I). Protective doses of HNE were different to pathological doses in another important regard. The lungs of B6 mice instilled with HNE at 1.8 U/kg that survived P. aeruginosa-induced pneumonia showed no gross signs of damage 50 days after infection (Fig. 5,C) and were indistinguishable from those of uninfected mice (Fig. 5 A). We conclude that an increase in the activity of extracellular NE can provide increased immunity to P. aeruginosa in the lung.

FIGURE 4.

Extracellular NE protects against P. aeruginosa infection. A, Percentage survival of B6 mice (n = 20–24 mice for each dose of HNE) after instillation with HNE (0–5.9 U/kg mouse) and infection with P. aeruginosa. ∗, Significantly different survival curves (B) mean titer ± SEM (n = 6) of P. aeruginosa 12 h after infection in the lung in B6 mice instilled with HNE (0–5.9 U/kg mouse). These data are representative of two independent experiments.

FIGURE 4.

Extracellular NE protects against P. aeruginosa infection. A, Percentage survival of B6 mice (n = 20–24 mice for each dose of HNE) after instillation with HNE (0–5.9 U/kg mouse) and infection with P. aeruginosa. ∗, Significantly different survival curves (B) mean titer ± SEM (n = 6) of P. aeruginosa 12 h after infection in the lung in B6 mice instilled with HNE (0–5.9 U/kg mouse). These data are representative of two independent experiments.

Close modal
FIGURE 5.

Effect of NE on lung histology after infection with P. aeruginosa. H&E stained midsagittal slices of lungs (original magnification, ×10) from uninfected B6 mice (A) or P. aeruginosa infected B6 mice, instilled with 5.9 U/kg HNE, 24 h after infection (B) or P. aeruginosa infected B6 mice instilled with 1.8 U/kg HNE, 27 days after infection (C). Uninfected Spi6 KO mouse (D) or Spi6 KO mice (E) 50 days after infection with P. aeruginosa.

FIGURE 5.

Effect of NE on lung histology after infection with P. aeruginosa. H&E stained midsagittal slices of lungs (original magnification, ×10) from uninfected B6 mice (A) or P. aeruginosa infected B6 mice, instilled with 5.9 U/kg HNE, 24 h after infection (B) or P. aeruginosa infected B6 mice instilled with 1.8 U/kg HNE, 27 days after infection (C). Uninfected Spi6 KO mouse (D) or Spi6 KO mice (E) 50 days after infection with P. aeruginosa.

Close modal
Table I.

Effect of HNE instillation in the lung after P. aeruginosa infection

Time (h)BAL NE (μU)BAL PMNs (/103)
 − − i.n. NE 
7.3 ± 0.23 11.9 ± 1.8 n.d. n.d. 
12 15.8 ± 2.0 21.8 ± 3.6 15.5 ± 1.7 26.6 ± 2.5 
24 13.8 ± 1.3 12.5 ± 0.22 n.d. n.d. 
648 died died 
Time (h)BAL NE (μU)BAL PMNs (/103)
 − − i.n. NE 
7.3 ± 0.23 11.9 ± 1.8 n.d. n.d. 
12 15.8 ± 2.0 21.8 ± 3.6 15.5 ± 1.7 26.6 ± 2.5 
24 13.8 ± 1.3 12.5 ± 0.22 n.d. n.d. 
648 died died 
a

NE activity (μU) and the number of segmented PMNs were determined for 1 ml of BAL from B6 mice and are ±SEM (n = 3–8 mice). Mice were infected with P. aeruginosa then after 4 h, either dosed i.n. with HNE (1.8 U/kg mouse) (+) or not (−). The dose was calculated based on the specific activity (U/mg) of HNE from the manufacturer (see Materials and Methods) and the weight of the mouse.At certain time points mice had died or data was not determined (n.d.). There was a significant difference (p = 0.02) in NE activity after 6 h and PMN number (p = 0.02) after 12 h between undosed and dosed mice.

The susceptibility of Spi6 KO PMNs to self-inflicted lysis results in an increase of NE from dead PMNs. Given the bactericidal and pathological effects of extracellular NE we examined the effect of Spi6 deficiency on PMN immunity to P. aeruginosa infection. Spi6 KO mice were significantly (p = 0.009) resistant to lethal infection with P. aeruginosa (Fig. 6,A), due to increased clearance from the lung (Fig. 6,B). Spi6 deficiency resulted in an increase in 2.6-fold increase in extracellular NE activity in the lung (Table II), which was within the protective range determined by HNE instillation experiments (Table I). Histological examination confirmed that the lungs of Spi6 KO mice that survived P. aeruginosa-induced pneumonia showed no gross signs of damage 50 days after infection (Fig. 4, D and E).

FIGURE 6.

A protective acute inflammatory response in the lungs of Spi6 KO mice. A, Percentage survival of B6 and Spi6 KO mice (n = 12 total mice for each experimental group). ∗, Significantly different survival curves and mean bacterial titer (± SEM) in the right lung over time after infection with P. aeruginosa (B). These data are representative of three independent experiments

FIGURE 6.

A protective acute inflammatory response in the lungs of Spi6 KO mice. A, Percentage survival of B6 and Spi6 KO mice (n = 12 total mice for each experimental group). ∗, Significantly different survival curves and mean bacterial titer (± SEM) in the right lung over time after infection with P. aeruginosa (B). These data are representative of three independent experiments

Close modal
Table II.

Effect of Spi6 deficiency on inflammation in the lung after P. aeruginosa infection

Time (h)BAL NE (μU)BAL PMNs (/103)BAL macrophage(/103)BAL lymphocyte(/103)
B6Spi6 KOB6Spi6 KOB6Spi6 KOB6Spi6 KO
0.3 ± 0.1 
5.4 ± 1.1 6.7 ± 0.8 n.d. n.d. n.d. n.d. n.d. n.d. 
12 3.9 ± 1.8 10.0 ± 1.7 9.4 ± 1.0 61.1 ± 8.4 1.1 ± 0.1 0.7 ± 0.04 0.5 ± 0.2 0.6 ± 0.2 
24 10.6 ± 1.7 9.0 ± 1.7 n.d. n.d. n.d. n.d. n.d. n.d. 
800 died n.d. died n.d. died n.d. died n.d. 
1200 died died died died 
Time (h)BAL NE (μU)BAL PMNs (/103)BAL macrophage(/103)BAL lymphocyte(/103)
B6Spi6 KOB6Spi6 KOB6Spi6 KOB6Spi6 KO
0.3 ± 0.1 
5.4 ± 1.1 6.7 ± 0.8 n.d. n.d. n.d. n.d. n.d. n.d. 
12 3.9 ± 1.8 10.0 ± 1.7 9.4 ± 1.0 61.1 ± 8.4 1.1 ± 0.1 0.7 ± 0.04 0.5 ± 0.2 0.6 ± 0.2 
24 10.6 ± 1.7 9.0 ± 1.7 n.d. n.d. n.d. n.d. n.d. n.d. 
800 died n.d. died n.d. died n.d. died n.d. 
1200 died died died died 
a

NE activity (μU) and the number of segmented PMNs were determined for 1 ml of BAL from B6 or Spi6 KO mice and are ± SEM (n = 3–7 mice). At certain time points mice had died or data was not determined (n.d.). There was a significant difference in the activity of NE (p = 0.03) and PMN number (p = 0.002) in BAL in B6 compared to Spi6 KO mice after 12 h.

The acute inflammatory response in Spi6 KO mice was also characterized by an increase in PMN influx into the lungs (7-fold higher than B6 controls) (Table II). The increased infiltration of inflammatory cells was limited to PMNs, as there was no increase in the levels of macrophages or lymphocytes in lung interstitium of Spi6 KO compared with B6 control mice (Table II). In addition to direct killing, the increase in extracellular NE may have also been responsible for the increase in infiltration of PMNs to the site of infection, further driving the inflammatory response in Spi6 KO mice (26, 27, 28). Taken together these data suggests a role for elevated NE is driving a protective acute inflammatory response in Spi6 KO mice.

It is widely accepted that apoptosis is important for the down-regulation of leukocyte responses. Thus, factors such as Spi6, which are required for CTL viability, are in turn required for effector function (10). Similarly, we show here that Spi6 is required for PMN viability by protecting them from NE. However, in contrast to CTLs, effector function was not diminished by the reduced viability of Spi6 KO PMNs. In fact, the necrotic disintegration of Spi6 KO PMNs resulted in an increased release of NE, which protected from lethal infection with P. aeruginosa. Therefore, self-inflicted lysis does not impair PMN function but rather results in a protective acute inflammatory response.

Death-inducing serine proteases are sequestered in cytotoxic granules to protect effector leukocytes from self-inflicted damage. Therefore, it was surprising that immunofluorescence studies on live intact cells revealed the presence of NE in the cytoplasm of PMNs (2) and GrB in the cytoplasm of CTLs (10). The present study reveals that Spi6 protects PMNs from death-induced from cytoplasmic NE, in an analogous fashion to how Spi6 protects CTLs from cytoplasmic GrB (10). However, we do not know how either NE or GrB reaches the cytoplasm. It is possible that it may be that NE and GrB are “misdirected” to the cytoplasm during the processes of phagolysosome formation in PMNs or granule polarization and release in CTLs.

Phagocytic neutrophils confer immunity by engulfing invading microbes (29). Experiments with protease deficient mice (2, 18, 30) support the long-held view that subsequent microbial killing is affected by the contents of cytoplasmic granules released into phagocytic vacuoles (1). This process requires a live cell and so, as for CTL killing, PMN viability is required for phagocytic immunity. Therefore, it was surprising that Spi6 deficiency, rather than impair PMN function, actually improved it. We believe that the release of NE from disintegrating Spi6 KO PMNs was responsible to the increased killing of P. aeruginosa. Our findings are consistent with PMNs being very short-lived effectors in which extracellular NE is directly microbicidal for Gram-negative bacteria (2, 3, 31). The susceptibility of Spi6 KO PMNs to break down resulted in the increased release of other microbicidal enzymes such as Cat G and lysozyme (Fig. 3). Therefore, it would seem probable that the disintegration of Spi6 KO PMNs leads to a release of a wide spectrum of microbicidal molecules into the extracellular milieu. In addition to the release of microbicidal molecules, Spi6 KO PMN breakdown may also result in the release of nuclear constituents that form neutrophil extracellular traps, which are thought to facilitate the killing of bacteria outside of the phagolysosome (31). Taken together our findings suggest that compared with WT PMNs, Spi6 KO PMNs may also be better able to not only kill P. aeruginosa more effectively through extracellular NE, but also be better able to kill Gram-positive bacteria and other microbes through the release of a wide spectrum of effector molecules. Evidence for a protective role for PMN lysis in host defense comes from clinical reports of the control of Candida albicans infection by extracellular microbicidal factors released from dying PMNs in healthy individuals (4). However, whether or not the release of microbicidal factors through PMN lysis is a physiologically important event in the defense of WT hosts from infection, is yet to be directly determined.

In humans (32, 33) and mice (34), α1-AT deficiency results in about a 10–20-fold increase in extra-cellular NE activity in the lung of COPD patients. More specifically in the case of inflammatory disease associated with P. aeruginosa infection, NE has been implicated in pneumonia-mediated acute lung injury and CF (6). However, Spi6 is a relatively weak inhibitor of NE and so in Spi6 KO mice, NE activity was increased to within the narrow permissible range (∼2–3 times higher than WT) that eliminated P. aeruginosa clearance without tissue damage. In addition, the protective, acute inflammatory response in Spi6 KO mice seemed to be associated with an exclusive increase in PMN infiltration. This, in combination with the susceptibility of Spi6 KO PMNs to death, biased the inflammatory response toward a short-lived increase in PMN-mediated killing without an increase in prolonged macrophage and lymphocyte-mediated inflammation. The acute inflammatory response in Spi6 KO mice was therefore critically different to clinically important inflammatory diseases, such as COPD (6) and sepsis (35), where the infiltration of macrophages and lymphocytes are thought to amplify responses through powerful proinflammatory cytokines. We suggest that the vigorous early clearance of P. aeruginosa by elevated NE is anti-inflammatory because it removes the potent inflammatory stimuli associated with flagellin and other P. aeruginosa components (5).

If deficiency in Spi6 improves immunity to P. aeruginosa then what selective advantage does Spi6 afford mice? In this article, we have studied acute bacterial infection and it is possible that Spi6 may be required to down-regulate PMN in chronic infections, which would otherwise lead to inflammatory diseases. An additional explanation for Spi6 selection during evolution, may be that it provides immunity to virus by protecting CTLs from self-inflicted damage by GrB (10). Therefore, although Spi6 may curtail PMN immunity to Gram-negative bacteria, it may be selected for in mice because it is required for CTL immunity to viruses.

It is often assumed that chronic inflammation occurs as a result of uncontrolled proinflammatory events. However, it is perhaps equally likely that failure, or inefficiency, of the normal resolution process is responsible for tipping the balance toward persistent inflammation, tissue injury and disease. Therefore, the failure of treatments for inflammatory diseases based on the neutralization of inflammatory cytokines (e.g., TNF-α, IL-1) or receptors (e.g., TNFR, TLR) may be because these therapies impair bacterial clearance (35). Targeting homologues of Spi6 in humans, may lead to alternative strategies to treat inflammatory diseases by inducing transient but protective PMN responses. Improved PMN function would increase microbial clearance and so would tip the balance of the inflammatory response toward resolution.

We thank G. Franzoso and S. Rutschmann for critical reading and D. Tharpe for help in preparing the manuscript.

The authors have no financial conflict of interest.

The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

1

This work was supported by National Institutes of Health Grant AI45108 (to P.G.A-R.).

3

Abbreviations used in this paper: PMN, polymorphonuclear neutrophils; Spi, serine protease inhibitor; NE, neutrophil elastase; SI, stoichiometry of the inhibition; HNE, human NE; KO, knockout; GrB, granzyme B; LDH, lactate dehydrogenase; PI, propidioum iodide; COPD, chronic pulmonary obstructive disease; Cat G, cathepsin G; TG, thioglycollate; WT, wild type; GST, glutathione transferase; BAL, bronchoalveolar lavage; i.n., intranasally; LB, Luria-Bertani.

1
Borregaard, N., J. B. Cowland.
1997
. Granules of the human neutrophilic polymorphonuclear leukocyte.
Blood
89
:
3503
-3521.
2
Belaaouaj, A., R. McCarthy, M. Baumann, Z. Gao, T. J. Ley, S. N. Abraham, S. D. Shapiro.
1998
. Mice lacking neutrophil elastase reveal impaired host defense against gram negative bacterial sepsis.
Nat. Med.
4
:
615
-618.
3
Belaaouaj, A., K. S. Kim, S. D. Shapiro.
2000
. Degradation of outer membrane protein A in Escherichia coli killing by neutrophil elastase.
Science
289
:
1185
-1188.
4
McNamara, M. P., J. H. Wiessner, C. Collins-Lech, B. L. Hahn, P. G. Sohnle.
1988
. Neutrophil death as a defence mechanism against Candida albicans infections.
Lancet
2
:
1163
-1165.
5
Lopez-Boado, Y. S., M. Espinola, S. Bahr, A. Belaaouaj.
2004
. Neutrophil serine proteinases cleave bacterial flagellin, abrogating its host response-inducing activity.
J. Immunol.
172
:
509
-515.
6
Dar, K. J., R. G. Crystal.
1999
. Inflammatory lung disease: molecular determinants of emphysema, bronchitis, and fibrosis. G. Gallin, and I. Goldstien, and R. Snyderman, eds.
Inflammation: Basic Principles and Clinical Correlates
3rd ed.
1061
-1082. Raven Press, New York.
7
Bainton, D..
1999
. Developmental biology of neutrophils and eosinophils. G. Gallin, and I. Goldstien, and R. Snyderman, eds.
Inflammation: Basic Principles and Clinical Correlates
3rd ed.
13
-34. Raven Press, New York.
8
Savill, J. S., A. H. Wyllie, J. E. Henson, M. J. Walport, P. M. Henson, C. Haslett.
1989
. Macrophage phagocytosis of aging neutrophils in inflammation: programmed cell death in the neutrophil leads to its recognition by macrophages.
J. Clin. Invest.
83
:
865
-875.
9
Savill, J., V. Fadok, P. Henson, C. Haslett.
1993
. Phagocyte recognition of cells undergoing apoptosis.
Immunol. Today
14
:
131
-136.
10
Zhang, M., S. M. Park, Y. Wang, R. Shah, N. Liu, A. E. Murmann, C. R. Wang, M. E. Peter, P. G. Ashton-Rickardt.
2006
. Serine protease inhibitor 6 protects cytotoxic T cells from self-inflicted injury by ensuring the integrity of cytotoxic granules.
Immunity
24
:
451
-461.
11
Silverman, G. A., P. I. Bird, R. W. Carrell, F. C. Church, P. B. Coughlin, P. G. W. Gettins, J. A. Irving, D. A. Lomas, C. J. Luke, R. W. Moyer, et al
2001
. The serpins are an expanding superfamily of structurally similar but functionally diverse proteins.
J. Biol. Chem.
276
:
33293
-33296.
12
Remold-O’Donnell, E..
1993
. The ovalbumin family of serpin proteins.
FEBS Lett.
315
:
105
-108.
13
Sun, J., C. H. Bird, V. Sutton, L. McDonald, P. B. Coughlin, T. A. De Jong, J. A. Trapani, P. I. Bird.
1996
. A cytosolic granzyme B inhibitor related to the viral apoptotic regulator cytokine response modifier A is present in cytotoxic lymphocytes.
J. Biol. Chem.
271
:
27802
-27809.
14
Sun, J., L. Ooms, C. H. Bird, V. R. Sutton, J. A. Trapani, P. I. Bird.
1997
. A new family of 10 murine ovalbumin serpins includes two homologs of proteinase inhibitor 8 and two homologs of the granzyme B inhibitor (proteinase inhibitor 9).
J. Biol. Chem.
272
:
15434
-15441.
15
Phillips, T., J. T. Opferman, R. Shah, N. Liu, C. J. Froelich, P. G. Ashton-Rickardt.
2004
. A role for the granzyme B inhibitor serine protease inhibitor 6 in CD8+ memory cell homeostasis.
J. Immunol.
173
:
3801
-3809.
16
Hirst, C. E., M. S. Buzza, C. H. Bird, H. S. Warren, P. U. Cameron, M. Zhang, P. G. Ashton-Rickardt, P. I. Bird.
2003
. The intracellular granzyme B inhibitor: proteinase inhibitor 9, is up-regulated during accessory cell maturation and effector cell degranulation, and its overexpression enhances CTL potency.
J. Immunol.
170
:
805
-815.
17
Dahlen, J. R., D. C. Foster, W. Kisiel.
1999
. Inhibition of neutrophil elastase by recombinant human proteinase inhibitor 9.
Biochim. Biophys. Acta
1451
:
233
-241.
18
Tkalcevic, J., M. Novelli, M. Phylactides, J. P. Iredale, A. W. Segal, J. Roes.
2000
. Impaired immunity and enhanced resistance to endotoxin in the absence of neutrophil elastase and cathepsin G.
Immunity
12
:
201
-210.
19
Graham, F. L., J. Smiley, W. C. Russell, R. Nairn.
1977
. Characteristics of a human cell line transformed by DNA from human adenovirus type 5.
J. Gen. Virol.
36
:
59
-74.
20
Liu, N., S. M. Raja, F. Zazzeroni, S. S. Metkar, R. Shah, M. Zhang, Y. Wang, D. Bromme, W. A. Russin, J. C. Lee, et al
2003
. NF-κB protects from the lysosomal pathway of cell death.
EMBO J.
22
:
5313
-5322.
21
Beatty, K., J. Bieth, J. Travis.
1980
. Kinetics of association of serine proteinases with native and oxidized α-1-proteinase inhibitor and α-1antichymotrypsin.
J. Biol. Chem.
255
:
3931
-3934.
22
Ying, Q. L., M. Kemme, S. R. Simon.
1994
. Functions of the N-terminal domain of secretory leukoprotease inhibitor.
Biochemistry
33
:
5445
-5450.
23
Vermes, I., C. Haanen, H. Steffens-Nakken, C. Reutelingsperger.
1995
. A novel assay for apoptosis: flow cytometric detection of phosphatidylserine expression on early apoptotic cells using fluorescein labelled Annexin V.
J. Immunol. Methods
184
:
39
-51.
24
Idziorek, T., J. Estaquier, F. De Bels, J.-C. Ameisen.
1995
. YOPRO-1 permits cytofluorometric analysis of programmed cell death (apoptosis) without interfering with cell viability.
J. Immunol. Methods
185
:
249
-258.
25
Janoff, A., B. Sloan, G. Weinbaum, V. Damiano, R. A. Sandhaus, J. Elias, P. Kimbel.
1977
. Experimental emphysema induced with purified human neutrophil elastase: tissue localization of the instilled protease.
Am. Rev. Respir. Dis.
115
:
461
-478.
26
Hubbard, R. C., G. Fells, J. Gadek, S. Pacholok, J. Humes, R. G. Crystal.
1991
. Neutrophil accumulation in the lung in α 1-antitrypsin deficiency: spontaneous release of leukotriene B4 by alveolar macrophages.
J. Clin. Invest.
88
:
891
-897.
27
Loike, J. D., B. Sodeik, L. Cao, S. Leucona, J. I. Weitz, P. A. Detmers, S. D. Wright, S. C. Silverstein.
1991
. CD11c/CD18 on neutrophils recognizes a domain at the N terminus of the A α chain of fibrinogen.
Proc. Natl. Acad. Sci. USA
88
:
1044
-1048.
28
Shapiro, S. D., N. M. Goldstein, A. M. Houghton, D. K. Kobayashi, D. Kelley, A. Belaaouaj.
2003
. Neutrophil elastase contributes to cigarette smoke-induced emphysema in mice.
Am. J. Pathol.
163
:
2329
-2335.
29
Metchnikoff, E..
1891
. Lecture VII.
Lectures on the Comparative Pathology of Inflammation: Delivered at the Pasteur Institute in 1891
Dover Publications, New York.
30
Reeves, E. P., H. Lu, H. L. Jacobs, C. G. Messina, S. Bolsover, G. Gabella, E. O. Potma, A. Warley, J. Roes, A. W. Segal.
2002
. Killing activity of neutrophils is mediated through activation of proteases by K+ flux.
Nature
416
:
291
-297.
31
Brinkmann, V., U. Reichard, C. Goosmann, B. Fauler, Y. Uhlemann, D. S. Weiss, Y. Weinrauch, A. Zychlinsky.
2004
. Neutrophil extracellular traps kill bacteria.
Science
303
:
1532
-1535.
32
Gadek, J. E., G. A. Fells, R. L. Zimmerman, S. I. Rennard, R. G. Crystal.
1981
. Antielastases of the human alveolar structures: implications for the protease-antiprotease theory of emphysema.
J. Clin. Invest.
68
:
889
-898.
33
Gadek, J. E., H. G. Klein, P. V. Holland, R. G. Crystal.
1981
. Replacement therapy of α 1-antitrypsin deficiency: reversal of protease-antiprotease imbalance within the alveolar structures of PiZ subjects.
J. Clin. Invest.
68
:
11581165
34
Martorana, P. A., T. Brand, C. Gardi, P. van Even, M. M. de Santi, P. Calzoni, P. Marcolongo, G. Lungarella.
1993
. The pallid mouse: a model of genetic α 1-antitrypsin deficiency.
Lab. Invest.
68
:
233
-241.
35
Cohen, J..
2002
. The immunopathogenesis of sepsis.
Nature
420
:
885
-891.