All animal studies were approved by the Institutional Animal Care and Use Committee at the University of California, San Francisco (Protocol# AN152943). Male mice (12–15 weeks old) were either purchased (The Jackson Laboratory, Bar Harbor, ME, USA) or bred at the animal facility at University of California, San Francisco. Purchased mice were allowed to acclimatize to their new housing for at least 1 week before any experiments on them were conducted. ASC and NLRP3 KO mice were a kind gift from V. Dixit (Genentech) and IL-1R and Caspase 1 KO mice were a kind gift from A. Ma (UCSF). All mice were on a C57B/L6 background except for Caspase 1 KO mice which were on a 129 background and thus also Caspase 11 deficient. B6 Albino mice were provided by Kevin P. Francis (Perkin Elmer). tdTomato/MRP8 Cre mice were generously provided by Zena Werb (UCSF).

Only male mice were used in our experiments primarily to reflect the fact that trauma disproportionately affects human males. Based on our previous studies, we used group sizes of 6–10 for all experiments (25, 34). All mice for a given experiment were either littermates or purchased/bred such that they were age matched. Mice used in these experiments were randomly chosen either to undergo the various surgeries (sham vs. IR) or treatments (±specific treatments), therefore there was no attempt made to blind the individuals conducting the experiments, however, in situations where mice received a treatment or control before IR surgery or infection after surgery, the individual collecting the organs/plasma and generating the enzyme-linked immunosorbant assay (ELISA)/qPCR data was unaware of which mice received which specific treatment.

A murine model of unilateral left PA occlusion was used, as we have described previously (25). Briefly, anesthetized mice [using IP tribromoethanol (Avertin^®); Sigma-Aldrich] were orally intubated, given buprenorphine (IP; Harry Schein, Melville, NY, USA), and placed on a rodent ventilator, using tidal volumes of 225 µL (7.5 cm³/kg), and a respiratory rate of 180 breaths/min (assuming an average mouse weight of 30 g). A left thoracotomy via the interspace between the second and third ribs was performed and the left PA was identified and ligated using a slip knot suture with 7-0 or 8-0 prolene monofilament suture. The end of the suture was externalized through a narrow bore (27 g) needle to the anterior chest wall. Prior to closure of the thorax, the left lung was reinflated with positive end pressure. Local anesthetic (3–4 drops of 0.25% bupivacaine) was applied topically prior to skin closure. The total period of mechanical ventilation and surgery was approximately 20–25 min. After skin closure, mice were extubated and allowed to recover from anesthesia. After 30–60 min of ischemia, the ligature on the PA was released and left lung reperfusion started. At the experimental end point times, mice were euthanized and the blood and lungs were collected.

Blood was collected from anesthetized mice via cardiac puncture using a heparinized syringe, centrifuged (14,000g, 5 min) and the plasma separated, flash frozen in liquid nitrogen and stored at −80°C. Lower portions of the left lungs were excised and placed in either Trizol^® (Life Technologies, Carlsbad, CA, USA) at −80°C for RNA isolation. Levels of cytokines and chemokines were quantified in plasma. Lungs, spleen, liver, and kidneys were collected as well for colony-forming unit (CFU) measurements in infection experiments as described later.

Sham (control) mice underwent left thoracotomy and all other procedures at precisely the same time points as experimental mice, except the left PA was not isolated and the slip-knot was not tied or externalized.

All mice (sham and IR) received equivalent durations of mechanical ventilation (20–25 min) and were left spontaneously breathing during their recovery from anesthesia and the remainder of the ischemia period and subsequent reperfusion or equivalent periods in the sham mice.

While this lung IR procedure has high initial survival rates of 80–90% on average, some mice die from irreparable damage to the PA or left bronchus during the slip-knot placement. Mice that did not survive the surgery or the reperfusion period due to technical complications in the surgical procedure (predominantly, left bronchus or left PA injury) were excluded from the study.

Wild-type C57B/L6 mice were pretreated with either a caspase 1 inhibitor (ac-YVAD-cmk peptide, Bachem Americas, Inc., Torrence, CA, USA) given approximately 30 min before left lung ischemia or an IL-1β blocking antibody (AB-401-NA; R&D Systems, Minneapolis, MN, USA) given 18–22 h before left lung ischemia. The YVAD peptide was prepared per the manufacturer’s instructions. Briefly 5 mg of powder was dissolved in DMSO to create a 50 mg/mL stock solution, which was then diluted in PBS to 50 µg/mL and 1 mg/kg was given to each mouse. Vehicle control was prepared similarly but without adding the YVAD peptide.

For the IL-1β blocking antibody, each mouse received 100 µg IP. Control mice received a control IgG polyclonal goat antibody (100 µg IP). Lungs and plasma were collected 1 h after reperfusion as described earlier.

A murine model of E. coli peritonitis was introduced into C57BL/6 mice that received 30 min ischemia followed by 3 h reperfusion. The mice were then infected with E. coli intraperitoneally and 2 h later sacrificed and plasma and lungs collected for measurement of circulating levels and local production of cytokines and chemokines, respectively.

A murine model of E. coli pneumonia was generated as follows: E. coli O111 was recovered from frozen stocks by streaking on an LB agar plate. Following overnight growth, single colonies were used to inoculate 5 mL of LB broth and grown at 37 C overnight with shaking. The following day, 2.5 mL of the overnight culture was used to inoculate 50 mL of LB broth and grown for 4–6 h at 37°C with shaking. Once the optical density was between 0.7 and 1.0, the cultures were spun down at 3,000 rpm for 10 min, rinsed in sterile PBS, and then resuspended in LB broth containing glycerol (30%) and aliquots were made for infection experiments and flash frozen on dry ice, followed by storage at −80°C. Sample aliquots were thawed to confirm the stability of CFU counts before using them in animal experiments.

Escherichia coli was instilled in specific mice (wild type and various knockouts as described) via intratracheal (IT) instillation of 2 µL/g mouse weight (60 µL for a 30 g mouse) and the CFU corresponding to this was measured with each thawed aliquot (approximately 4–5 × 10^9–10 given IT to each mouse). Mice were first anesthetized with 3% isoflurane in an anesthetic chamber with oxygen as the carrier gas. After loss of pedal reflex, the mouse was quickly placed on an intubating slide. The vocal cords were visualized using an external fiberoptic light source and a curved forceps to lift the tongue up and to the left. E. coli was then administered using a 27 g syringe with a curved catheter (0.6 mm diameter) on the needle. Successful instillation was accompanied by a characteristic change in the spontaneous breathing pattern of the mouse. For mice that received the IR surgery, the E. coli IT instillation was performed as described above at 3 h after the start of the reperfusion period.

Following E. coli IT instillation, the mice were allowed to recover from anesthesia and returned to their cages and to a BSL2 room. These mice were observed thereafter at 1 h later, and then daily until the 48 h endpoint was reached.

Plasma and organs were then collected as described above and the organs were homogenized (Tissue-Tearor-Biospec Products, Bartlesville, OK, USA) in sterile PBS. Serial 10-fold dilutions were then performed to measure CFU after overnight incubation of LB agar plates at 37°C.

TaqMan-specific inventoried gene primers for glyceraldehyde 3-phosphate dehydrogenase (GAPDH), beta actin, interleukin (IL)-6, were used to measure the message levels of these genes in lung tissue (Life Technologies, Carlsbad, CA, USA).

Lung tissue was homogenized (Tissue-Tearor) and total RNA isolated using Trizol^®. We used the High Capacity RNA-to-cDNA reverse transcription Kit using 1 µg messenger RNA per reaction (Life Technologies). Quantitative real-time polymerase chain reaction was performed using the ABI Prism 7000 Sequence Detection System (Life Technologies). Run method: polymerase chain reaction activation at 95°C for 20 s was followed by 40 cycles of 1 s at 95°C and 20 s at 60°C.

The average threshold count (Ct) value of 2–3 technical replicates was used in all calculations. The average Ct values of the internal controls (GAPDH, beta actin) were used to calculate ΔCt values for the array samples. Data analysis was performed using the 2^−ΔΔCt method, and the data were corrected for statistical analysis using log transformation, mean centering, and autoscaling (35–37).

Concentrations of IL-6, IL-12p70, chemokine (C–C motif) ligand (CCL) 2/monocyte chemotactic protein (MCP) 1, chemokine (C–X–C motif) ligand (CXCL) 1 in mouse plasma were determined using the corresponding mouse Quantikine kits (R&D Systems, Minneapolis, MN, USA). All assays were performed according the manufacturer’s supplied protocol. Standard curves were generated and used to determine the concentrations of individual cytokines or chemokines in the sample.

Neutrophil superoxide production was assessed to determine whether neutrophils recruited following IR exhibited an activated phenotype. DHR 123 (Molecular probes) is a highly sensitive reactive oxygen species (ROS)-detection probe (38, 39) and thus, was chosen to assess intracellular superoxide as previously described (40) with modifications.

The left lower lung lobes were finely chopped with sterile scalpels and placed in HBSS + 0.1% BSA, passed through a 40 µm filter, and spun down. Collagenase digestion of the lung tissue and Fc receptor blocking was not done due to concerns that either process could potentially activate neutrophils. Red blood cells (RBCs) were lysed by incubating the cell suspension in RBC Lysis Buffer for 5 min at RT. Cells were again spun down, resuspended in HBSS + 0.1% BSA and counted. The cells were then incubated with APC-conjugated Ly6G antibody (clone 1A8, Biolegend) for 30 min at 4°C to stain neutrophils. Following this, cells were washed once with HBSS + 0.1% BSA and once with PBS-EGG (1 mM EDTA, 0.09% d-glucose, 0.05% gelatin in PBS). DHR 1, 2, 3 (10 µM; Molecular Probes) was added to the cells and incubated for 30 min at 37°C. The reaction was stopped by adding excess PBS-EGG at 4°C and incubating for 10 min. Cells were spun down, resuspended in PBS-EGG at 4°C and rhodamine fluorescence analyzed on a flow cytometer within 30 min.

B6 Albino mice were either subjected to IR surgery (as described above; 30 min ischemia, 3 h reperfusion) or IT LPS administration (1 mg/kg) and were imaged at the time points noted. Imaging was conducted on the IVIS^® Spectrum Instrument (PerkinElmer) as previously described (41). Briefly, prior to each imaging time point, mice were injected IV with 200 mg/kg luminol-R [for example, 125 mL of 40 mg/mL luminol (Sigma-Aldrich) mixed with 12.5 µL of 100 pmol QD800 quantum dots (Life Technologies)] and luminescence imaging was performed with an open filter for 5 min. 1 mg/kg LPS was injected intratracheally as well following left lung IR to activate dormant neutrophils.

tdTomato/MRP8 Cre mice that express the tdTomato protein under the neutrophil-specific MRP8 promoter were used to visualize the localization of neutrophils in the lung. Following IR (described above), fluorescein-labeled Lycopersicon esculentum lectin solubilized in phosphate-buffered saline (1 mg/ml), was injected as a 100 µl bolus into the tail vein prior to euthanasia to visualize the lung vasculature. Lungs were fixed by IT inflation with Z-fix (buffered zinc-formalin fixative, Anatech Ltd., Battle Creek, MI, USA) at 25 cm pressure, stored at 4°C overnight in Z-fix, washed in PBS, and then transferred to 30% sucrose in PBS prior to OCT embedding. Lungs were embedded in a mixture of 30% sucrose in PBS and OCT (1:1) and then frozen on dry ice. OCT-embedded sections (5 µm thickness) were stained with 4′,6-diamidino-2-phenylindole dihydrochloride (DAPI, Sigma-Aldrich, St. Louis, MO, USA) to visualize the nuclei.

Lungs from wild-type C57BL/6 or C3H mice that underwent lung IR (60 min ischemia, 1 h reperfusion) or LPS challenge (10 mg/kg IV for 1 h) were fixed in formalin and paraffin sections were stained with specific RNA probes for IL-1β and other cell markers, such as hopx (AT1 cells), surfactant C (AT2 cells), CD11c (AMs and dendritic cells). RNA in situ hybridization for IL-1β mRNA and other genes as described earlier was performed manually using the RNAscope^® HD Red Reagent Kit (Advanced Cell Diagnostics Pharma Assay Services, Newark, CA, USA) according to the manufacturer’s instructions. Briefly, 5 µm formalin fixed, paraffin-embedded tissue sections were pretreated with heat and protease prior to hybridization with the target oligo probes. Preamplifier, amplifier and AP-labeled oligos were then hybridized sequentially, followed by chromogenic precipitate development. Each sample was quality controlled for RNA integrity with a positive control probe specific to the housekeeping gene peptidylprolyl isomerase B and for background with a negative control probe specific to bacterial Bacillus subtilis gene dihydrodipicolinate reductase (dapB). Optimization of pretreatment conditions was performed to establish the maximum signal to noise ration. Specific RNA staining signal was identified as black, red, or blue-green punctate dots. Samples were counterstained with Gill’s Hematoxylin with nuclei appearing light purple (42, 43).

Data in the figures are expressed as mean ± SD. Data from in vivo studies comparing two conditions were analyzed using two-tailed nonparametric Mann–Whitney analyses. GraphPad Prism was used for statistical analyses (GraphPad Software, La Jolla, CA, USA). For all in vivo and in vitro experiments, P values <0.05 were considered significant. P values are represented as follows in the figures: *<0.05; **<0.01; ***<0.001; and ****<0.0001. Experiments were repeated two or more times, as indicated in the figure legends.

All animal studies were ethically and methodologically approved by the Institutional Animal Care and Use Committee at the University of California, San Francisco, CA, USA.

The mechanisms by which sterile injury results in inflammation and how these differ from mechanisms underlying the host inflammatory response to infection are very active topics of research currently. Having hypothesized that the NLRP3 inflammasome would be required for the generation of sterile inflammation following lung IR, we subjected ASC KO and NLRP3 KO mice that lack either all inflammasome subtypes or just NLRP3 inflammasomes, respectively, to 1 h of ischemia followed by 1 h of reperfusion. We then quantified plasma and tissue levels of inflammatory cytokines. In these and following experiments, and as verified in prior publications (25, 34), we measured IL-6 and chemokine levels as surrogate markers of sterile lung IR inflammatory response that includes subsequent neutrophil recruitment and lung edema. We found that while NLRP3 KO mice following lung IR expressed cytokine levels that were significantly lower compared to wild-type mice, ASC KO mice did not (Figure 1A). To account for reported and observed elevations in baseline levels of inflammatory markers in ASC KO mice (44–46), we normalized these cytokine levels to those from mice that underwent a sham procedure. By doing so, we observed a significantly weaker induction of inflammatory cytokines and chemokines in both ASC KO and NLRP3 KO mice (Figure 1B). We confirmed this at the mRNA level in lung tissue as well (Figure 1C). Together these data strongly support the conclusion that sterile lung injury initiates downstream inflammatory cytokine release through the activity of the NLRP3 inflammasome.

We previously reported that IL-1β is produced following ventilated lung IR and that in vitro macrophage derived IL-1β can augment endothelial cell IL-6 production (25). To further confirm the importance of IL-1β in lung IR inflammation, we used genetic and pharmacologic tools that target inflammatory caspases so as to disrupt the processing of IL-1β into its active form. Compared to wild-type mice, caspase 1/11 KO mice following IR displayed a significantly muted increase in levels of secreted IL-6, CCL2 (MCP-1), and IL-1β compared to sham (Figure 2A and data not shown) and this was confirmed at the tissue mRNA level (Figure 2B). Furthermore, the caspase 1-specific inhibitor peptide, YVAD, was able to significantly reduce IL-6 production following IR compared to a vehicle control (Figure 2C). Overall, we conclude that inflammatory caspases are required for lung IR inflammation.

Inflammasome and caspase activity can result not only in IL-1β production but also IL-18, and perhaps even IL-33. We next focused on testing whether IL-1β activity was important in vivo in lung IR inflammation. To do so, we used an IL-1β blocking antibody and observed a reduction in the levels of IL-6 produced following IR compared to an IgG control (Figure 3A). These data suggest that IL-1β is a key early signal generated after lung IR.

The identity of the cell type(s) that produce(s) IL-1β in the lung in vivo has not been definitively established. Data deposited in the Immgen and BioGPS/Gene Atlas databases report that many different cell types can produce IL-1β including bronchial epithelial cells, smooth muscle, dendritic cell (DC) subsets, monocytes and macrophages, neutrophil subsets, AMs, and alveolar epithelium. While it is clear that many lung cell types can produce IL-1β in vitro, and the airway epithelium has been implicated in a few reports, few if any studies have relied on examining lung cells in their native in vivo tissue context for IL-1β expression. Since the inflammasome is a highly regulated signaling complex and the release of IL-1β requires strict temporal and spatial control, we hypothesized that IL-1β release would be restricted to one or a few specific cell types. Candidates in the lung included AM and the cells that form the alveolar epithelium (type 1 and type 2 alveolar epithelial cells, namely, AT1 and AT2 cells), other lung macrophages or DCs. To identify the type of lung cell(s) in which the inflammasome activation, caspase and IL-1β processing takes place, we used an in situ hybridization methodology with pools of specific IL-1β, and other RNA probes. The IL-1β containing cell in resting lungs appeared to line the alveolar space (Figure S1 in Supplementary Material). We believe that these IL-1β-expressing cells participate in the sensing of injury generated by lung IR and trigger the cascade of events leading to mature IL-1β release, IL-6 production, neutrophil recruitment, and overall lung inflammation. Even following in vivo LPS challenge (employed as a means to further induce IL-1β mRNA expression), the increased lung IL-1β expression appeared to still localize within a similar cell type lining the alveolar space (Figure S2 in Supplementary Material). IL-1β mRNA signal was detected at low levels in isolated naive mouse lung, and this signal was increased following IL-1β gene induction after both LPS challenge and lung IR, even if the LPS response was far more robust (Figures 3B,C). Using specific probe pools for genes specific to alveolar epithelial cells (Hopx for AT1 and Surfactant C for AT2), macrophages (F4/80), AMs (CD11c), DCs (SiglecH, CD103) revealed that the IL-1β signal colocalized most closely with surfactant C made by AT2 epithelial cells (Figure 3B and data not shown). Of note, we did also observe some weak LPS-induced IL-1β expression in AMs.

Interleukin-1β often functions as a locally acting cytokine within the tissue where it was released. Consistent with this, the levels of IL-1β released systemically after lung IR are relatively low to undetectable [data not shown and Ref. (25)]. To examine whether downstream IL-1β signaling within the lung is important for lung IR inflammation, we investigated the expression of IL-1R in the lung and how IL-1R KO mice responded to lung IR. We observed dispersed IL-1R expression in multiple cell types in lung tissue, however, we consistently observed IL-1R expressing cells in close proximity to the IL-1β producing cell (Figure 4A). This supported the idea that IL-1β primarily acts as a locally signaling cytokine. In contrast, the neutrophil attracting chemokine, CXCL1, was induced in lung endothelial cells (Figure 4B) and easily detected systemically (25). As shown in Figure 4C, unlike wild-type mice, IL-1R KO mice were profoundly unable to induce IL-6 or CCL2 (MCP-1) production following IR. This was further confirmed at the tissue mRNA level (Figure 4D). These findings are consistent with a model placing IL-1β production and IL-1R engagement upstream of IL-6 and chemokine production. Furthermore, this conclusion is supported by the observation that IL-1R KO mice did not display significant neutrophil recruitment following lung IR compared to wild-type mice (data not shown). These data begin to shed light on the detailed choreography of intercellular signaling that occurs within the lung following injury.

In our previously published work, we have shown that lung IR inflammation in WT mice manifested itself as rapid but transient increases in inflammatory cytokines, such as IL-6, MCP-1, etc. (peaking at 0–1 h postreperfusion) as well as transient neutrophilia (peaking at 3–6 h postreperfusion) (34). Given the uninjured appearance of the mouse lung 24 h following IR (34) and the lack of functional deficits in oxygen uptake as measured by pulse oximetry (unpublished data), we hypothesized that the early recruited neutrophils may not have degranulated or become fully activated. To tackle this, we first subjected mice to either LPS challenge, a sham or IR procedure and using flow cytometry, we determined that there was a 4–10-fold increase in neutrophil numbers (Ly6G+ cells) in the circulating blood and within the IR-affected lung (Figure 5A and quantified in Figure 5B) consistent with our previously published data (25). Of note, the sham mouse cohort also experienced some neutrophil recruitment to the lung, presumably in response to the incidental sterile injury generated by the surgical incision and lung manipulation during the sham procedure. In peripheral blood and single cell suspensions from the left lung, neutrophils were assessed for ROS production using an in vitro DHR assay. IR-recruited neutrophils, like sham surgery-recruited neutrophils had lower ROS generation even compared to naive mice that did not receive any surgery (Figure 5C, top two panels). However, when these neutrophils were challenged in vitro with phorbol ester (PMA), all subsets of neutrophils were able to strongly generate ROS (Figure 5C, bottom two panels).

Next, we sought to locate these neutrophils within the lung tissue following IR. Using mice that expressed red fluorescent neutrophils, we observed that the neutrophils were located within the lung parenchyma and not within the air spaces (Figure 6).

Thereafter, we compared neutrophils recruited to the IR lung to those recruited by LPS. LPS challenge resulted in a comparable recruitment of neutrophils in blood and lung (Figure 7A), but these neutrophils produced significantly more ROS than those recruited in IR or sham surgeries (Figure 7B, and quantified in Figure 7C).

Finally, we asked whether this recruitment of apparently “dormant” neutrophils occurred in vivo to rule out the possibility that our findings were an artifact of lung preparation or neutrophil activity alterations once the neutrophils were separated from the in vivo context of the lung and the host. To answer this question, we measured myeloperoxidase (MPO) activation in injured lung tissue in vivo using in vivo imaging (IVIS™). We had previously shown that the administration of luminol with quantum dots permitted the specific tracking of activated neutrophils (and not activated macrophages) to the lung in live mice following LPS administration (41). As seen in Figure 8A, when mice received LPS, a robust accumulation of activated neutrophils was detected in the lungs at 3 h and not at 1 h. In contrast, 3 h after reperfusion, IR surgery mice displayed no signal correlating to activated neutrophils in the lungs. Thus there appeared not be any activated MPO-producing neutrophils in the lungs of injured mice 3 h following IR (Figure 8B). However, to confirm the presence of these dormant neutrophils, we introduced LPS into the lung, as a tool to expose the presence of these cells. By virtue of activated MPO signal emerging in these IR mice rapidly within 20–40 min (Figure 8C), we conclude that while in vivo IR-recruited neutrophils in the lung are dormant, they are still capable of rapid activation/degranulation in vivo in the presence of pathogen-derived toxins like LPS.

Thus far we have shown that NLRP3 and IL-1β are key immune molecules activated following lung IR, and which result in the trafficking of minimally activated dormant neutrophils to the lung. Next, we attempted to define the role that these neutrophils and the creation of inflammatory conditions within the lung parenchyma might play in lung pathophysiology. It is highly likely that an evolutionary reason exists for why neutrophils are recruited in lung following IR but remain dormant. We speculated that host lung immunity evolved to anticipate pathogen presence during and/or after lung trauma, and may have done so by orchestrating the trafficking of neutrophils to the site of injury in the expectation of encountering pathogen. To test this hypothesis, we first subjected mice to lung IR and then exposed their lungs to E. coli to determine whether the prior presence of lung IR would result in early and amplified inflammatory signals generated in the lung. Introducing E. coli led to vastly greater cytokine and chemokine levels both within the lung and systemically in plasma compared to IR alone (Figure 9) and appeared to reduce the bacterial burden early (2 h) after infection (data not shown). To confirm whether this amplified inflammatory response was indeed beneficial, we tracked the course of infected mice subjected to preceding lung IR vis-à-vis bacterial burden. Remarkably, we found a near total absence of bacteria locally at the site of IR and infection (left lung) as well as significant decreases in organs such as the kidney and spleen 48 h after infection (Figure 10). Notably, the plasma of both groups of mice was free from detectable bacteria even though there was evident large-scale bacterial seeding of multiple organ systems. These data indicate that naturally occurring IR inflammation may have a beneficial role to play in situations that involve the convergence of sterile injuries and infection, such as is often encountered in trauma.

To confirm the importance of the inflammasome in generating this protective lung IR inflammation against infection, we examined the ability of ASC KO and NLPR3 KO mice to contain an experimental bacterial E. coli pneumonia after lung IR. We found that mice deficient in all inflammasomes (ASC KO, Figure S3 in Supplementary Material), or specifically NLRP3-containing inflammasomes (NLPR3 KO, Figure 11) after IR, were unable to mitigate the course of the experimental E. coli pneumonia. As we had demonstrated earlier, these mice generated attenuated inflammatory responses following lung IR. Additionally, in some cases, preceding IR in these KO mice resulted in higher bacterial counts compared to the infection only group. Interestingly again, the plasma in these mice was also free of detectable pathogen suggesting that bacterial colonization of remote organs may not necessarily correlate with systemic blood-borne bacteremia.

Finally, we sought to determine the duration of this beneficial effect of lung IR inflammation. We speculated that after a certain period, a second phase of immune paralysis or immunosuppression may create ideal conditions for a pathogenic infection to flourish. We found that even at 24 h following reperfusion, the beneficial effects, though diminished, were still observable in lungs of infected wild-type mice with 10-fold lower bacterial burden compared to mice that did not receive preceding lung IR (Figure 12). Thus, we conclude that the inflammatory process initiated by lung IR is capable of limiting the establishment or spread of an experimental bacterial pneumonia involving E. coli even 1 day after reperfusion.