Alpha-1 antitrypsin deficiency impairs lung antibacterial immunity in mice

Alpha-1 antitrypsin (AAT) is a major inhibitor of serine proteases in mammals. Therefore, its deficiency leads to protease–antiprotease imbalance and a risk for developing lung emphysema. Although therapy with human plasma-purified AAT attenuates AAT deficiency–related emphysema, its impact on lung antibacterial immunity is poorly defined. Here, we examined the effect of AAT therapy on lung protective immunity in AAT-deficient (KO) mice challenged with Streptococcus pneumoniae. AAT-KO mice were highly susceptible to S. pneumoniae, as determined by severe lobar pneumonia and early mortality. Mechanistically, we found that neutrophil-derived elastase (NE) degraded the opsonophagocytically important collectins, surfactant protein A (SP-A) and D (SP-D), which was accompanied by significantly impaired lung bacterial clearance in S. pneumoniae–infected AAT-KO mice. Treatment of S. pneumoniae–infected AAT-KO mice with human AAT protected SP-A and SP-D from NE-mediated degradation and corrected the pulmonary pathology observed in these mice. Likewise, treatment with Sivelestat, a specific inhibitor of NE, also protected collectins from degradation and significantly decreased bacterial loads in S. pneumoniae–infected AAT-KO mice. Our findings show that NE is responsible for the degradation of lung SP-A and SP-D in AAT-KO mice affecting lung protective immunity in AAT deficiency.


Introduction
Alpha-1 antitrypsin (AAT) belongs to the superfamily of serine protease inhibitors (SERPINs) and is an important regulator of neutrophil elastase (NE) and cathepsin G (1,2). Humans and mice have high serum levels of AAT, ranging at 1-2 mg/mL in humans and 3-4 mg/mL in mice, which increase upon injury or infection, thus underlining the (patho-) physiological need for sustained antiproteolytic activity in both species (1,3,4).
AAT is encoded by the SERPINA1 gene and mainly produced by hepatocytes but also by monocytes/macrophages and lung epithelial cells (5)(6)(7). Over 150 mutations have been identified in the SERPINA1 gene, some of which affect the concentration and/or functionality of AAT protein. The S (Glu264Val) and Z (Glu342Lys) mutations are most common and clinically significant and are both known to cause AAT deficiency (AATD) (2,5,(8)(9)(10)(11). Individuals homozygous for the Z allele exhibit roughly 90% reduced levels of circulating AAT and have an increased risk of developing chronic obstructive pulmonary disease (COPD) with emphysema (12)(13)(14)(15). Many patients with AATD are lifelong treated with AAT augmentation therapy (16). There is clinical evidence that therapy with weekly infusions of AAT decreases emphysema progression (17,18).
Current recommendations for AAT augmentation therapy are not based on randomized clinical trials but rather are oriented at serum AAT levels. Therefore, the standard augmentation therapy recommended for patients with AATD is defined as 60 mg AAT/kg body weight (b.w.) per week reaching approximately 50% of serum AAT levels found in healthy individuals (19). Many patients with AATD under standard augmentation therapy still experience emphysema progression. A recent prospective study revealed that double-dose augmentation therapy of 120 mg AAT/kg b.w. per week increased AAT levels to normal and reduced serine protease activities and hence markers of elastin degradation in bronchoalveolar lavage (BAL) fluids of patients with AATD (19). Results from such double-dose augmentation therapies suggest that current single-dose augmentation therapies may underestimate the need for AAT bioavailability in patients with AATD, thus calling for individualized AAT dosing considerations (19).
Alpha-1 antitrypsin (AAT) is a major inhibitor of serine proteases in mammals. Therefore, its deficiency leads to protease-antiprotease imbalance and a risk for developing lung emphysema. Although therapy with human plasma-purified AAT attenuates AAT deficiency-related emphysema, its impact on lung antibacterial immunity is poorly defined. Here, we examined the effect of AAT therapy on lung protective immunity in AAT-deficient (KO) mice challenged with Streptococcus pneumoniae. AAT-KO mice were highly susceptible to S. pneumoniae, as determined by severe lobar pneumonia and early mortality. Mechanistically, we found that neutrophil-derived elastase (NE) degraded the opsonophagocytically important collectins, surfactant protein A (SP-A) and D (SP-D), which was accompanied by significantly impaired lung bacterial clearance in S. pneumoniae-infected AAT-KO mice. Treatment of S. pneumoniae-infected AAT-KO mice with human AAT protected SP-A and SP-D from NE-mediated degradation and corrected the pulmonary pathology observed in these mice. Likewise, treatment with Sivelestat, a specific inhibitor of NE, also protected collectins from degradation and significantly decreased bacterial loads in S. pneumoniae-infected AAT-KO mice. Our findings show that NE is responsible for the degradation of lung SP-A and SP-D in AAT-KO mice affecting lung protective immunity in AAT deficiency.
We next examined the effect of hAAT therapy on total proteolytic and NE activity in BALF and BAL neutrophils of S. pneumoniae-infected mice of the 2 treatment groups. Using gelatin-based zymography ( Figure 3F), we observed substantial proteolytic (i.e., gelatinolytic) activity in BALF of S. pneumoniaeinfected AAT-KO mice (illustrated by white proteolysis bands in lanes 2-4), which was drastically blocked by the therapy with hAAT (lanes 5-7), and only partially blocked by oxidized AAT, which is known to be devoid of NE-inhibitory activity (27, 28) (lanes 8-10). In selected experiments, we also examined the effect of AAT therapy on BAL NE activity in mice. Baseline NE activity was observed in peripheral neutrophils from untreated WT and AAT-KO mice serving as controls ( Figure 3G). Of note, BAL neutrophils of S. pneumoniae-challenged AAT-KO mice treated with hAAT showed significantly lower NE activity compared with S. pneumoniae-infected, PBS-treated AAT-KO mice, thus illustrating a link between systemic therapy with hAAT and inhibition of NE activity in BAL neutrophils of S. pneumoniae-challenged AAT-KO mice ( Figure 3G). NE is a highly potent serine protease that is able to degrade lung elastin fibers and to activate proinflammatory cytokines (12,13,29). In line with this, we found significantly reduced levels of TNF-α, IL-1β, IL-6, CXCL1, and elastin-derived desmosine in BALF of S. pneumoniae-infected AAT-KO mice receiving therapy with hAAT ( Figure 3, H-L). Notably, reported differences in outcome between groups were not due to differences in leukocyte subset recruitment (data not shown).
Augmentation therapy improves lung antibacterial immunity in AAT-KO mice via inhibition of NE. To exclude potential immunomodulatory properties of hAAT therapy underlying the observed effects in S. pneumoniaeinfected AAT-KO mice, we treated AAT-KO mice with oxidized or native hAAT followed by infection with S. pneumoniae. Of note, oxidation of the methionine residues of AAT substantially decreases its ability to inhibit various proteases, including the serine protease NE, while preserving its immunomodulatory activity (27,28). AAT-KO mice augmented with hAAT demonstrated significantly improved bacterial clearance in BALF and lung tissue on day 2 postinfection, while no significant effect was observed in AAT-KO mice treated with oxidized hAAT (Figure 4, A and B). In additional experiments, S. pneumoniae-challenged NE-KO mice treated with vehicle or hAAT had similar CFU counts in their lungs (Figure 4, C and D). No significant differences in BALF cytokines were noted between groups (Figure 4, E-L). Thus, therapy with hAAT appears to improve lung protective immunity primarily by inhibition of proteolysis, specifically the activity of NE.
Effect of AAT deficiency on bacterial phagocytosis, killing, and burst induction in professional phagocytes. Based on the observation that hAAT-augmented AAT-KO mice coped better with pneumococcal challenge compared with vehicle-treated, S. pneumoniae-infected AAT-KO mice, we examined bacterial phagocytosis, killing, and burst induction in professional phagocytes of WT and AAT-KO mice in the absence or presence of hAAT. As shown in Figure 5, A-D, no differences in pneumococcal phagocytosis and killing were observed between rAMs or BM-PMNs of WT or AAT-KO mice, in either the absence or presence of hAAT. Moreover, BM-PMNs of WT and AAT-KO mice responded similarly to S. pneumoniae-or zymosan-induced respiratory burst ( Figure 5, E and F). These data illustrate that differences in antibacterial responses between WT and AAT-KO mice are not due to differences in pneumococcal phagocytosis and killing by the respective professional phagocyte subsets.
Human AAT remains biologically active in S. pneumoniae-infected AAT-KO mice. In the next set of experiments, we wished to determine whether hAAT employed in the current study would retain NE complex-forming Desmosine levels in BALF of untreated (CL) or S. pneumoniae-infected WT or AAT-KO mice, as indicated. *P ≤ 0.05, **P ≤ 0.01, ***P ≤ 0.001 compared with untreated or S. pneumoniae-infected WT mice, +++ P ≤ 0.001 compared with day 2 (Mann-Whitney U test, log-rank test). activity during pneumococcal infection. NE-AAT complex formation was examined by using BALF of S. pneumoniae-challenged AAT-KO mice receiving hAAT therapy or vehicle. As shown in Figure 6A, BALF collected from PBS-treated, S. pneumoniae-challenged AAT-KO mice incubated or not with NE in vitro did not show any AAT protein in Western blots ( Figure 6A, lanes 1, 3). However, AAT protein was easily detectable in BALF of S. pneumoniae-challenged AAT-KO mice treated with hAAT (lane 2), with 2 bands of different molecular size representing native AAT (lower band in lane 2) or NE-AAT complex (upper band in lane 2). Furthermore, the addition of purified NE to BALF of hAAT-augmented, S. pneumoniae-infected AAT-KO mice led to a nearly complete shift of the aforementioned lower band of native AAT to the protein band representing NE-AAT complexes (upper band in lane 4), clearly confirming the anti-NE activity of the hAAT preparation. As a control, blotting of hAAT protein gave a single band (lane 5), whereas hAAT preincubated with NE generated 2 AAT-specific bands of lower and higher molecular size representing native hAAT and NE-complexed hAAT, respectively (both lane 6). As expected, NE protein alone gave no signal in AAT-specific Western blotting (lane 7).
To confirm further that hAAT preserves anti-NE activity during pneumococcal pneumonia in AAT-KO mice, BALF of PBS-treated or hAAT-augmented, S. pneumoniae-challenged AAT-KO mice were subjected to kinetic NE activity assay. As shown in Figure 6B, BALF of PBS-treated and S. pneumoniaeinfected AAT-KO mice exhibited easily detectable NE activity similar to that of purified NE (serving as positive control). However, AAT purified from BALF of hAAT-augmented, S. pneumoniae-challenged AAT-KO mice showed substantially less NE activity ( Figure 6B). Hence, exogenously applied hAAT seems to be fully active during pneumococcal infection of AAT-KO mice and is not inactivated by any putative infection-evoked reactive intermediates.
Attenuated antibacterial immunity in AAT-KO mice is due to NE-dependent degradation of alveolar collectins surfactant protein A and D. Previous studies in patients with cystic fibrosis revealed that elastolytic activities in BAL fluids were inversely related to levels of respiratory opsonins (30). Therefore, we next questioned whether We next aimed to clarify whether NE was indeed responsible for the observed SP-A/SP-D degradation in S. pneumoniae-challenged AAT-KO mice. Therefore, we examined the effect of increasing concentrations of purified NE on SP-D protein levels in BALF of S. pneumoniae-challenged WT mice. As shown in Figure  7C, we found that NE even at a concentration of 0.5 μg was sufficient to degrade SP-D protein into even smaller SP-D fragments (lane 3), while addition of 2 μg NE per sample completely degraded SP-D protein in BAL fluids of S. pneumoniae-challenged WT mice ( Figure 7C, lane 5). Importantly, preincubation of purified NE with hAAT in vitro (in order to complex and thus inactivate the protease prior to addition to BALF) completely protected BALF-derived SP-D proteins from the degradation ( Figure 7C, lanes 6 to 9).
In the next set of experiments, we examined whether purified NE would also degrade recombinant SP-D (rSP-D) protein in vitro. Of note, addition of just 0.5 μg purified NE per sample was sufficient to nearly completely degrade rSP-D protein in vitro ( Figure 7D, lane 3 and 4), similar to the observation made in BAL fluids of WT mice ( Figure 7C).
We next hypothesized that if NE is the culprit to degrade collectins SP-A and SP-D both in vivo and in vitro, NE-deficient mice should be protected from NE-mediated depletion of alveolar collectins. Indeed, as shown in Figure 7E, SP-D protein was easily detectable in BALF of S. pneumoniae-challenged NE-KO mice, with similar band intensities observed in S. pneumoniae-challenged WT mice (compare lanes 1, 2 in Figure 7E with lanes 3, 4 in Figure 7B), and a similar pattern of SP-D was observed in AAT-treated, S. mean ± SD of n = 4-9 mice per time point and treatment group. (H-L) TNF-α, IL-1β, IL-6, CXCL1, and desmosine levels in BAL fluids of mice at days 2-3 postinfection. Data are shown as mean ± SD of n = 7-9 mice per time point and treatment group. *P ≤ 0.05, **P ≤ 0.01, ***P ≤ 0.001 compared with vehicle-treated mice, + P ≤ 0.05 compared with day 2 (Mann-Whitney U test, log-rank test). All data are representative of 2 independent experiments.  pneumoniae-challenged NE-KO mice. These data clearly support our hypothesis that neutrophil-derived NE is the major protease mediating degradation of lung collectins in S. pneumoniae-infected AAT-KO mice.

JCI
Using a sensitive ELISA for quantification of alveolar SP-D, we found that SP-D levels significantly increased in BALF of WT mice upon pneumococcal challenge compared with untreated WT mice, whereas upon S. pneumoniae challenge BALF SP-D levels of AAT-KO mice increased only weakly and nonsignificantly ( Figure 7F). Importantly, the infection-driven rise in SP-D levels could be significantly restored in BALF of AAT-KO mice upon therapy with hAAT ( Figure 7F).
Finally, we hypothesized that if NE-mediated degradation of lung collectins contributes to the disturbed antibacterial immunity in AAT-KO mice, blockade of NE with specific inhibitors should reverse this process. To this end, AAT-KO mice were treated with the highly specific NE inhibitor Sivelestat twice daily subsequent to S. pneumoniae infection. As shown in Figure 7G, therapeutic application of Sivelestat protected SP-D from NE-dependent degradation in BALF of S. pneumoniae-infected AAT-KO mice, relative to vehicle-treated, S. pneumoniae-challenged AAT-KO mice completely lacking alveolar SP-D protein and when compared with S. pneumoniae-challenged WT mice (compare lanes 3, 4 in Figure 7G with lanes 3, 4 in Figure 7B). Again, quantification of SP-D levels by ELISA confirmed significantly increased SP-D levels in BALF of S. pneumoniae-infected AAT-KO mice after treatment with Sivelestat, reaching a similar order of magnitude as observed in AAT-treated, S. pneumoniae-infected AAT-KO mice (compare Figure  7H with Figure 7F). Most importantly, we found that treatment with Sivelestat inhibited exaggerated NE activity, and thus loss of alveolar collectins led to significantly decreased bacterial loads both in BAL fluids and lungs of S. pneumoniae-infected AAT-KO mice (Figure 7, I and J).
Collectively, our data show that NE is the major culprit in degrading SP-A and SP-D in S. pneumoniaechallenged AAT-KO mice, thereby disturbing lung protective immunity in AAT-deficient mice. The latter can be restored by both hAAT augmentation therapy and using synthetic NE inhibitors.

Discussion
In the current study, we evaluated the impact of AATD on lung protective immunity against S. pneumoniae in mice in the absence or presence of AAT augmentation therapy. To reflect the clinical situation of patients with inherited AATD, we used recently generated AAT-KO mice lacking all 5 mouse serpin A1 paralogues, thus leading to complete AAT protein deficiency. AAT-KO mice were used at the age of 8-14 weeks while spontaneously developing lung emphysema only at advanced age of more than 35 weeks (25). Based on the data generated from this model, we show that AATD is detrimental to mice infected with S. pneumoniae but could be rescued upon treatment with plasma-purified hAAT.
Patients with AATD receive augmentation therapy primarily to slow down emphysema progression. To this end, patients with AATD typically receive standard doses of 60 mg/kg/wk intravenously to restore at least 50% of normal AAT serum levels (approximately 11 μM) (10). Although calculated standard doses may be sufficient to reverse the biochemical AAT defects in stable patients (10), little is known about serum AAT levels needed in case of microbial challenge where AAT turnover kinetics might be increased, as observed in the current study. Actually, previous studies showed that patients with AATD are more likely to suffer from pulmonary infections, which could be ameliorated by AAT therapy. A recent study by Balbi et al. reported that augmentation therapy in AATD patients with COPD significantly reduced the percentage of gram-positive and -negative bacteria (H. influenzae, S. pneumoniae) compared with controls (23). In addition, in a (subjective) questionnaire-based study by Lieberman, patients with ZZ AATD reported significantly fewer lung infections per year under augmentation therapy compared with those subjects with the same phenotype who were not receiving this therapy (31). Moreover, cystic fibrosis (CF) patients with deficient variants of AAT demonstrated an earlier carriage of P. aeruginosa, which was correlated with an earlier onset of P. aeruginosa lung infection (32). In contrast, CF patients receiving AAT therapy demonstrated significantly lower bacterial loads of P. aeruginosa (33). Together, these studies strongly support the view that AATD is associated with attenuated antibacterial immunity and that augmentation therapy contributes to the reduced incidence and severity of lung bacterial colonization and/or infection in AATD but also in non-AATD CF patients.
Lower elastolytic activities were reported in BALF of patients with AATD under standard augmentation therapy (19). In our experimental settings, bacterial infection significantly decreased plasma AAT levels in mice by approximately 25% (Figure 1), allowing us to predict that bacterial infection in patients with AATD would result in a transient drop of AAT plasma levels. Thus, the currently applied AAT therapy dosing regimen may be sufficient to diminish bacterial burden in patients with AATD but may be too low to adequately neutralize increasing NE activity due to the peripheral neutrophilia developing in response to acute bacterial infection. Based on a previous study (19) and our report, clinically relevant arguments to increase AAT augmentation therapy to more physiological levels may be (a) to facilitate an improved inhibition of ECM-degrading NE activity in the bronchoalveolar space of AATD patients (19) and (b) to better protect AATD patients against increased AAT turnover expectedly developing upon lung bacterial infections, which would also lower the risk of emphysema progression/exacerbation.
Various aspects may explain the observed infection-induced decrease of plasma AAT levels in mice. First, both elevated AAT consumption caused by peripheral neutrophilia and/or leakage of AAT into the alveolar space may have contributed to the observed drop in AAT protein levels. Moreover, an active trans-endo-/epithelial transport of AAT has also been described (34,35). Third, it is conceivable that septic liver injury developing in response to invasive pneumococcal disease may also result in decreased AAT serum levels, although this aspect is unlikely to be relevant in the current model of focal pneumonia.
AAT-KO mice were found to succumb to pneumococcal lung infection by day 2 postinfection. We found a markedly higher NE activity in BAL-derived neutrophils of S. pneumoniae-infected AAT-KO as compared with WT mice. This finding is in line with the lack of antiprotease activity of AAT in the KO but not in the WT mice, which actually responded to pneumococcal lung infection with significantly increased bronchoalveolar release of AAT. NE is a highly potent serine protease able to degrade lung ECM, such as elastin, and may facilitate proinflammatory cytokine activation in the inflammatory microenvironment (12, 13, 29). In the current study, we show that levels of desmosine, an amino acid released from elastin upon its degradation, increased substantially in BALF of S. pneumoniae-challenged AAT-KO relative to WT mice. These data demonstrate that NE actually triggers ECM degradation in the lungs of mice in the absence of AAT, and this process is profoundly inhibited upon AAT augmentation. Furthermore, analysis of the proinflammatory cytokines TNF-α, IL-1β, IL-6, and CXCL1 illustrated the effect of AAT augmentation therapy on lung inflammatory responses after bacterial challenge in nonaugmented as compared with augmented AAT-KO mice. Actually, these data show that augmentation therapy of S. pneumoniae-infected AAT-KO mice improved survival by effectively diminishing lung bacterial loads and the release of proinflammatory cytokines, which -in addition to the bacterial infection -may significantly contribute to lung damage. Together, these data clearly show an important role of alveolar AAT in blocking NE-mediated lung tissue damage in mice. Our findings are consistent with similar observations in AATD patients subjected to AAT augmentation therapy (19), thus supporting the clinical relevance of the current AAT-KO mouse model of AATD.
To explain the observed differences in antibacterial responses discriminating S. pneumoniae-challenged WT from AAT-KO mice, we examined the bacterial phagocytosis, killing, and burst capacity of alveolar macrophages and neutrophils from WT and AAT-KO mice in vitro. However, we did not note any significant differences between groups. Hence, it seems very unlikely that the observed differences in bacterial loads between WT and AAT-KO mice are due to principal differences in the phagocytosis/killing or respiratory burst capacity of professional phagocytes between these 2 experimental groups.
However, based on the observation that AAT-KO mice exhibited a drastically increased proteolytic burden in their lungs upon pneumococcal challenge, which was largely blocked by AAT therapy along with a significant reduction in lung bacterial loads, we hypothesized that NE was the major culprit to disturb lung antibacterial immunity in S. pneumoniae-challenged AAT KO. Previous studies in patients with CF showed that NE is capable of disturbing lung antibacterial immunity at different levels, including degradation of the opsonophagocytically important collectins SP-A and -D (36,37), as well as degradation of pathogen-specific immunoglobulins (30). According to our understanding, pathogen-evoked protective antibody responses were very unlikely to develop so quickly in our employed acute lung infection model. However, the hydrophilic lung surfactant proteins SP-A and SP-D are well-known to serve as opsonophagocytically relevant collectins to orchestrate lung host defense against various lung-tropic pathogens, including S. pneumoniae, based on in vitro studies and genetically engineered animal models (38)(39)(40)(41)(42)(43). In the current study, we provide evidence that NE is responsible for the observed degradation of SP-A and SP-D in S. pneumoniae-infected AAT-KO mice, which is accompanied by decreased antibacterial defense. Of note, in addition to the currently reported degradation of lung collectins by neutrophil-derived NE, peroxynitrite may also affect the binding activity of SP-A and SP-D, due to its strong nitrating and oxidizing effects (44). Other studies reported neutrophil-derived myeloperoxidase-dependent inactivation of SP-D in vivo (45). Actually, our findings provide a potentially novel and hitherto unrecognized molecular mechanism by which inherited AATD affects lung antibacterial immunity, besides its well-known effect on lung emphysema progression in patients with AATD, and how this defect can be corrected by augmentation therapy.
Chronic lung diseases like CF or COPD are characterized by a vicious circle of bronchoalveolar recruitment of neutrophils with subsequent excess release of NE culminating in end-stage emphysematous lung injury. At the same time, both CF and COPD patients have increased susceptibility to opportunistic bacterial infections, which at least in CF is causally responsible for excess lung neutrophil recruitment. A previous study suggested particularly NE to be responsible for mediating lung collectin depletion in BAL fluids of CF patients (46), although this study did not relate these findings to antibacterial immune defects. As mentioned before, a recent study by Balbi et al. (23) reported that AAT treatment significantly reduced total bacterial loads, including S. pneumoniae, in sputum samples of COPD patients when compared with untreated COPD patients. Therefore, when taking together previous studies and our work, it appears reasonable that the success of AAT augmentation therapy of (mostly neutrophil-dominated) chronic lung diseases should not only rely on long-term inhibition of emphysema development or progression but also should aim at inhibition of NE-dependent depletion of SP-A and SP-D in order to avoid defective antibacterial responses.
Collectively, we propose the following model: in WT mice, NE released into the alveolar airspace upon infection (at least partially) becomes complexed with AAT, thereby limiting both NE-driven degradation of lung ECM components as well as lung collectins SP-A and SP-D. In contrast, in AAT-KO mice, S. pneumoniae challenge triggers increased alveolar neutrophil accumulation and thus NE liberation that is not inhibited by AAT. This consequently triggers increased NE-dependent lung ECM degradation, as JCI Insight 2021;6(3):e140816 https://doi.org/10.1172/jci.insight.140816 determined by increased alveolar desmosine levels, and even further triggers a nearly complete functional depletion of SP-A and SP-D collectins within the alveolar compartment, thereby profoundly affecting lung bacterial pathogen elimination. Upon AAT augmentation therapy or application of NE-specific inhibitors, excessive neutrophil-derived NE activity is blocked, thus inhibiting ECM degradation and degradation of collectins, thereby facilitating a normalized bacterial pathogen elimination process similar to that observed in S. pneumoniae-challenged, AAT-sufficient WT mice.
As such, the current study expands our understanding of the impact of augmentation therapy on lung protective immunity in patients with inherited AATD and how this therapy might be combined in the future with more cost-efficient NE-specific inhibitor therapies. Importantly, however, we believe that these data may also be critically important to other neutrophil-dominated lung diseases where continuous recruitment of neutrophils is associated with defects in lung antibacterial immunity, including (but not limited to) CF, COPD, as well as bronchiectasis.

Methods
Animals. WT mice (C57BL/6) were purchased from Janvier Lab (Le Genest-Saint-Isle, France). AAT-KO mice lacking all of the 5 serpin A1 a-e paralogues expressed in mice were generated as described recently (25). NE-KO mice were purchased from The Jackson Laboratory (Bar Harbor, Maine, USA) (47). Age-and sex-matched mice were used for experiments at 8 to 14 weeks of age.
Culture and quantification of S. pneumoniae. For details, see Supplemental Methods. Growth kinetics of S. pneumoniae in the presence of AAT. For details, see Supplemental Methods. Infection of mice with S. pneumoniae. Infection of mice with capsular group 19F S. pneumoniae was done as recently described (49,50). For further information, see Supplemental Methods.
Determination of bacterial loads in BAL fluids and lung homogenates. For details, see Supplemental Methods. Oxidation of human purified AAT. Oxidization of native hAAT was achieved as described previously (27,28,51,52).
Treatment groups. Five treatment groups were established: WT mice were mock infected (PBS) or were infected with S. pneumoniae as described above, and endogenous AAT levels were determined in plasma and BALF at different time points postinfection.
The role of AAT deletion on evolving pneumococcal pneumonia was examined in AAT-KO mice and WT mice (serving as controls), which were infected orotracheally with S. pneumoniae, and disease progression was monitored during an observation period of 6 days.
The effect of AAT augmentation on antibacterial immunity in AAT-KO mice was studied by treating AAT-KO mice i.p. with either purified hAAT (2 mg/mouse twice daily) or PBS (vehicle control) at 6 hours prior to infection with S. pneumoniae and subsequent determination of CFU in BALF and lung tissue as well as assessment of survival and lung histopathology.
To exclude potential immunomodulatory effects of AAT on lung protective immunity against S. pneumoniae in mice, AAT-KO mice were treated i.p. with PBS, purified hAAT (exhibiting NE-inhibitory and putative immunomodulatory activity) (2 mg/mouse twice daily), or oxidized hAAT lacking NE-inhibitory activity (2 mg/mouse twice daily) (27,28,51,53) followed by infection with S. pneumoniae. In additional experiments, NE-deficient mice were treated with human purified AAT or PBS starting at 24 hours postinfection followed by determination of bacterial loads and cytokine responses in BALF and lung tissue.
To analyze the effect of selective NE inhibition on lung antibacterial immunity in AAT-KO mice, mice were treated i.p. with Sivelestat (2 mg/mouse twice daily) or saline (vehicle control), starting immediately