Materials and reagents are listed in the online supplement.

Wild-type (WT) C57BL/6J, CD45.1 alloantigen (Ly5.1), and IL-1ra^−/− mice (weight 18–21 g, C57BL/6J background) were purchased from Charles River (Sulzfeld, Germany). CCR2^−/− mice (C57BL/6J background) were originally obtained from W. A. Kuziel (21). CX₃CR1^GFP/GFP mice (C57BL/6J background) were generated as described previously (22). Parent CX₃CR1^GFP/GFP and CX₃CR1^+/+ mice were bred to yield heterozygous CX₃CR1^+/GFP offspring or crossbred with CCR2^−/−CX₃CR1^+/+ mice to obtain CCR2^+/+CX₃CR1^+/GFP and CCR2^−/−CX₃CR1^+/GFP mice. Animals were bred in our facility under special pathogen-free conditions for three to four generations and used at 8 to 11 weeks of age. All animal experiments were approved by the local government committee.

Mice were intratracheally treated with ultrapure LPS (50 μg/mouse in sterile phosphate-buffered saline [PBS] in a total volume of 70 μl) or with K. pneumoniae serotype 2 (1 × 10⁵ or 2 × 10⁵ in sterile PBS in a total volume of 70 μl) as described (18, 23) and killed at the indicated time points. In selected experiments, hIL-1ra (Anakinra; Amgen, Seattle, WA; 10 μg/70 μl PBS) or PBS alone were applied intratracheally. Depletion of circulating monocytes was achieved by intravenous injection of 200 μl clodronate-containing liposomes 12 hours before LPS administration as described (24). Processing of BALF and blood cells and lung homogenates and analysis of alveolar leakage are outlined in the online supplement.

Bone marrow chimeric mice were created by intravenous transplantation of CD45.2⁺ WT or CCR2^−/− bone marrow cells into lethally irradiated CD45.1 recipients as outlined in the online supplement. WT, IL-1ra^−/−, or CCR2^−/− peripheral blood mononuclear cells were isolated as described in the online supplement (13) and intravenously injected into CCR2^−/− mice directly before LPS instillation.

For fluorescence-activated cell sorter (FACS) analysis and cell sorting, cells were processed as described in the online supplement using a FACSCanto or FACSVantage flow cytometer equipped with a FACSDiva software package (BD Biosciences, Heidelberg, Germany).

IL-1β and IL-1ra levels in BALF were measured using commercially available ELISA kits (R&D Systems, Wiesbaden, Germany) according to the manufacturer's instructions.

RNA isolation, cDNA synthesis, and real-time reverse transcriptase–polymerase chain reaction analysis were performed as previously described (18). The ΔCt values for each target gene were calculated using the reference gene Actin or PBGD with the formula ΔCt = Ct_reference – Ct_target. Values are given as ΔΔCt or fold expression of control (2^ΔΔCt), depending on the experiment.

Murine and human alveolar epithelial cells type II (AEC) were isolated as described (18, 25) and grown to confluency on 24-well plates, transwells, or chamber slides (BD Biosciences). Human lung tissue was obtained from lobectomy specimens distal from tumors after written patient consent (Departments of Pathology and Surgery, Justus-Liebig-University, Giessen; approved by the local ethics committee). Cells were treated with recombinant mouse or human IL-1β (10 ng/ml), LPS (1 μg/ml), or mouse or human IL-1ra (1 μg/ml) and subjected to further analyses as outlined in the supplement.

AEC were methanol/acetone fixed and stained overnight with anti–ZO-1 (1:50) or isotype antibodies. Lung cryosections were prepared from perfused and lavaged lungs and stained with hematoxylin-eosin (online supplement).

Data are presented as means ± SD. Statistical significance was estimated using one-way analysis of variance (SPSS for Windows); paired samples were analyzed using the two-tailed paired Student t test. A value of P ≤ 0.05 was considered as significant.

To investigate whether exudate macrophages (ExMa) might contribute to preservation of alveolar epithelial cell integrity in LPS-induced ALI, we compared AEC apoptosis and alveolar albumin leakage in WT and CCR2-deficient mice, well known to exhibit defective ExMa recruitment (12, 13). Indeed, ExMa recruitment was drastically reduced in CCR2-deficient mice after intratracheal LPS challenge, whereas sustained alveolar ExMa presence was observed in WT mice peaking at 48 hours after treatment as analyzed by differential counts from BALF cytospins (Figure 1A, upper panel). Flow cytometric immunophenotyping of BALF cells at 48 hours after LPS treatment confirmed the absence of CD11c⁺GR-1⁺F4/80⁺CD11b^high ExMa (Figure 1A, bottom panel, P1, red) in CCR2^−/− mice despite equal numbers of CD11c⁺GR-1^lowF4/80⁺CD11b^int resident alveolar macrophages (rAM) compared with WT mice (Figure 1A, bottom panel, P2, red). Likewise, analysis of rAM numbers in the time course after LPS challenge did not reveal significant differences between both mouse strains (Figure 1B), nor did analysis of alveolar lymphocyte subset numbers (CD4 T cells, CD8 T cells, natural killer cells; data not shown). Quantification of ExMa in the interstitial lung compartment (i.e., in lavaged and perfused lung digests) demonstrated strongly reduced numbers in CCR2-deficient mice, most prominent at 12 hours after LPS challenge (Figure 1C). Lack of interstitial ExMa recruitment in CCR2-deficient mice at this time point was associated with significantly elevated alveolar albumin leakage, a feature of lung injury (Figure 1D). Correspondingly, CCR2 deficiency was associated with significantly increased AEC apoptosis as demonstrated by Annexin V binding at 12 hours after LPS treatment (15.1 ± 1.59% vs. 33.2 ± 6.19%; Figure 1E). AEC were defined as CD45⁻CD31⁻cytokeratin⁺ cells from lung digests and consisted of approximately 87% AEC type II (pro-SP-C⁺) and approximately 13% AEC type I (T1α⁺; see Figure E1 in the online supplement).

Figure 1. CC-chemokine receptor 2 (CCR2) deficiency is associated with abrogated exudate macrophage (ExMa) recruitment and increased alveolar epithelial cell apoptosis and alveolar leakage on LPS challenge. (A, B) Wild-type (WT) and CCR2^−/− mice were intratracheally instilled with 50 μg LPS and bronchoalveolar lavage fluid (BALF) ExMa and resident alveolar macrophages (rAM) numbers were determined according to differential counts of Pappenheim-stained cytocentrifuge preparations and total BALF cell numbers at the indicated time points from n = 4–11 mice each (upper panel). Lack of ExMa recruitment in CCR2^−/− mice was confirmed by fluorescence-activated cell sorter (FACS) analysis of BALF cells at 48 hours after LPS, which identified ExMa as CD11c⁺GR-1⁺F4/80⁺CD11b^high and rAM as CD11c⁺GR-1^lowF4/80⁺CD11b^int; P1 = ExMa; P2 = rAM; representative plots are given. (C) Interstitial CD11c⁺GR-1⁺F4/80⁺CD11b^high ExMa numbers in LPS-treated WT and CCR2^−/− mice were quantified by FACS from lavaged and perfused lung homogenate cell suspensions after BALF and perfusion of the pulmonary vasculature at the indicated time points. Representative FACS plots are given for the 12-hour time point; P1 = ExMa. (D) Alveolar leakage of fluorescein isothiocyanate (FITC)-labeled albumin was determined by ratio calculation of FITC fluorescences in BALF and serum after LPS challenge of n = 6–13 mice per time point and values are given as arbitrary units (AU). (E) CD45⁻CD31⁻cytokeratin⁺ alveolar epithelial cells (AEC) from lung digests of LPS-treated WT or CCR2^−/− mice were analyzed for Annexin V binding at the given time points (n = 4–8 mice per group). Representative FACS plots (12 h LPS) are shown demonstrating Annexin V-Alexa647 binding to previously gated CD45⁻CD31⁻cytokeratin⁺ AEC. Solid bars, WT mice; open bars, CCR2^−/− mice; all bar graphs represent means ± SD; *P < 0.05; **P < 0.01; ***P < 0.005. n.s. = not significant.

Download Figure | Download Powerpoint

[More] [Minimize]

To demonstrate that indeed CCR2 expression on circulating ExMa precursors (i.e., Gr-1⁺ blood monocytes) rather than CCR2 expression on resident lung myeloid or parenchymal cells accounted for attenuated epithelial injury, we made use of a bone marrow chimeric mouse model whereby we transplanted either WT or CCR2-deficient CD45.2⁺ bone marrow cells into lethally irradiated CD45.1⁺ WT recipients. At 14 days after transplantation bone marrow engraftment was always greater than 90%, whereas resident lung cells including rAM were of recipient phenotype as demonstrated by FACS analysis of CD45.1 and CD45.2 alloantigen expression on peripheral blood leukocytes and BALF rAM, respectively (Figure E2). As shown in Figure 2A, chimeric WT mice with CCR2-deficient circulating leukocytes revealed abrogated ExMa recruitment together with sustained increase in alveolar barrier dysfunction in the time course after LPS treatment (Figure 2B). These results clearly demonstrate that CCR2 expressed on circulating blood monocytes is critically involved in both ExMa extravasation into the interstitial and alveolar compartment and attenuation of LPS-induced lung injury.

Figure 2. Attenuated lung injury is associated with pulmonary recruitment of CC-chemokine receptor 2 (CCR2)⁺ exudate macrophages (ExMa) after LPS challenge. Lethally irradiated CD45.1⁺ wild-type (WT) mice were transplanted either CD45.2⁺ WT (solid bars) or CD45.2⁺ CCR2^−/− (open bars) bone marrow (BM) cells and after complete engraftment chimeric mice were intratracheally applied 50 μg LPS. (A) ExMa numbers in bronchoalveolar lavage fluid (BALF) were calculated according to differential counts of Pappenheim-stained cytocentrifuge preparations and total BALF cell numbers at the indicated time points from n = 5 mice per group and (B) alveolar albumin leakage was determined. (C, D) WT mice pretreated with empty liposomes (EL) or clodronate-containing liposomes (CL) to deplete circulating monocytes were LPS-treated and (C) alveolar epithelial cells (AEC) apoptosis and (D) albumin leakage were determined after 12 hours from 5 to 6 mice per group. Bar graphs represent means ± SD; *P < 0.05; ***P < 0.005. AU = arbitrary units.

Download Figure | Download Powerpoint

[More] [Minimize]

To confirm that indeed recruitment of circulating blood monocytes attenuated lung injury in our model, we depleted monocytes in WT mice by intravenous injection of liposomal clodronate 12 hours before LPS challenge. As shown in Figure E3, side scatter (SSC)^lowCD115⁺ monocytes were largely depleted in peripheral blood (top panel), resulting in abrogated interstitial accumulation of CD11c⁺GR-1⁺F4/80⁺CD11b^high ExMa (bottom panel) after 12 hours of LPS treatment. Accordingly, AEC apoptosis (Figure 2C) and albumin leakage (Figure 2D) were enhanced in mice injected with clodronate liposomes as compared with empty liposome controls, demonstrating that depletion of ExMa precursors aggravated LPS-induced epithelial injury.

Given that ExMa recruitment was associated with attenuated epithelial injury on LPS treatment, we hypothesized that transendothelial/epithelial migration of peripheral blood monocytes (PB-Mo), giving rise to ExMa, might induce antiinflammatory, antiapoptotic, or epithelial-mitogenic mediators in these cells. We therefore analyzed the relative expression of potential candidates, such as IL-10, hepatocyte growth factor, keratinocyte growth factor/FGF7, transforming growth factor-β1, and IL-1ra (26–31) in ExMa versus PB-Mo flow-sorted from LPS-challenged CX₃CR1^+/GFP mice, expressing green fluorescent protein (GFP) in both PB-Mo and freshly recruited ExMa, at 48 hours as previously described (32). Representative FACS plots with gating of GFP-positive cells from BALF and peripheral blood are given in Figure 3A (upper panel). We found IL-10, hepatocyte growth factor, and keratinocyte growth factor gene expression slightly down-regulated in alveolar ExMa compared with PB-Mo, whereas transforming growth factor-β1 expression was moderately up-regulated. Strikingly, IL-1ra gene expression was highly up-regulated in alveolar transmigrated ExMa compared with expression in their circulating precursors, with a mean ddCT of 14.8 ± 1.59 (Figure 3A, bottom panel), whereas IL-1ra gene induction was only moderate in recruited neutrophils (Figure E4). Correspondingly, IL-1ra protein was alveolarly released in significantly higher amounts in WT mice at 48 hours after LPS challenge than in CCR2^−/− mice (Figure 3B, upper panel). Given that ExMa-mediated antiapoptotic and barrier-protective effects occurred early, at 12 to 24 hours after LPS instillation when the majority of ExMa was present in the interstitial lung compartment, we speculated that ExMa might exert these effects already during the transmigration process, on close contact with AEC. We therefore measured IL-1ra levels in perfused and lavaged lungs containing mainly interstitial ExMa. Indeed, we found IL-1ra protein accumulating to high levels in WT lung parenchyma starting at 6 hours after LPS treatment, whereas IL-1ra levels remained significantly lower in CCR2-deficient mice at 24 to 72 hours (Figure 3B, bottom panel). Correspondingly, we found the IL-1β/IL-1ra receptor IL-1R1 surface expressed on AEC from lung digests as demonstrated by flow cytometry, and further up-regulated on intratracheal LPS treatment, likely enhancing AEC responsiveness to IL-1β effects (Figure 3C).

Figure 3. Exudate macrophages (ExMa) highly up-regulate and release IL-1ra on LPS-driven alveolar recruitment, whereas resident alveolar macrophages (rAM) are the primary source of IL-1β. (A) ExMa and peripheral blood monocytes (PB-Mo) from pooled bronchoalveolar lavage fluid (BALF) and blood cells of 5 to 6 24-hour LPS-challenged CX₃CR1^+/GFP mice were flow-sorted after gating according to CX₃CR1 expression as shown in representative fluorescence-activated cell sorter plots. Relative gene expression of IL-10, hepatocyte growth factor (HGF), keratinocyte growth factor (KGF), transforming growth factor (TGF)-β1, and IL-1ra in ExMa and PB-Mo was determined by quantitative reverse transcriptase–polymerase chain reaction and are given as ddCT values from n = 3 different sortings. (B) IL-1ra levels were determined by ELISA from BALF (top panel) or lavaged and perfused lungs homogenized (bottom panel) from wild-type (WT) and CC-chemokine receptor 2 (CCR2)^−/− mice at the given time points after LPS treatment (n = 3–16 mice per group). (C) Flow cytometric analysis of IL-1R1 expression (dark gray histograms) on CD45⁻CD31⁻ alveolar epithelial cells (AEC) from lung digests of untreated or 24-hour LPS-treated WT mice. Shown are representative histograms including isotype control stainings (light gray histograms). (D) Relative IL-1β gene expression in flow-sorted rAM from BALF of 24-hour sham- versus LPS-treated CCR2^+/+CX₃CR1^+/GFP and CCR2^−/−CX₃CR1^+/GFP mice (n = 3 per group). A representative sorting plot showing BALF cells of an LPS-treated CCR2^+/+CX₃CR1^+/GFP mouse is given with rAM defined as CD11c⁺CX₃CR1⁻GR-1^lowF4/80⁺. (E) IL-1β levels were determined by ELISA from BALF (top panel) or lavaged and perfused lungs homogenized (bottom panel) from WT and CCR2^−/− mice at the given time points after LPS treatment (n = 4–10 mice per group). All bar graphs represent means ± SD; WT mice, solid bars; CCR2^−/− mice, open bars; *P < 0.05 **P < 0.001; ***P < 0.005. LH = lung homogenate; n.s. = not significant; untr. ctl = untreated control.

Download Figure | Download Powerpoint

[More] [Minimize]

IL-β is known to be primarily expressed in mononuclear phagocytes, and it was previously suggested that rAM are a primary source of IL-β in patients with ARDS (10), contributing to lung injury progression. We therefore analyzed IL-1β gene expression in flow-sorted rAM from BALF of 24-hour sham- or LPS-treated CCR2^+/+CX₃CR1^+/GFP and CCR2^−/−CX₃CR1^+/GFP mice. rAM were identified as CD11c⁺CX₃CR1⁻GR-1^lowF4/80⁺ and could be clearly distinguished from CD11c⁻CX₃CR1⁻GR-1^highF4/80⁻ alveolar neutrophils and from CD11c⁺CX₃CR1⁺GR-1^highF4/80⁺ ExMa (Figure 3D, left). As shown in Figure 3D (right panel), IL-1β mRNA was highly up-regulated in rAM after LPS challenge in both WT and CCR2-deficient mice (ddCT: WT, 6 ± 0.74; CCR2^−/−, 5.27 ± 1.09, representing a mean 68- and 45-fold induction, respectively). Correspondingly, IL-1β protein was present in the alveoli (Figure 3E, top panel) and interstitial lung compartment (Figure 3E, bottom panel) in comparable amounts in both WT and CCR2^−/− mice in response to LPS. These data suggest that resident rather than recruited mononuclear phagocytes represent the major source of alveolar IL-1β in our model of LPS-induced ALI.

Lung injury is at least in part initiated and maintained by recruited neutrophils, releasing proteases and reactive metabolites (33). IL-1β is known to promote polymorphonuclear neutrophil (PMN) recruitment, likely by inducing chemokine release and adhesion molecule expression involved in PMN extravasation (11). We therefore hypothesized that LPS-driven PMN recruitment would be enhanced in CCR2-deficient mice lacking IL-1ra expressing ExMa. Indeed, alveolar PMN numbers in CCR2-deficient mice widely exceeded those found in WT mice at 48 hours after LPS treatment (89.3 ± 21.1 × 10⁵ vs. 43.4 ± 7.24 × 10⁵), which was associated with CCR2 expression on circulating ExMa precursors as shown in WT mice transplanted either WT or CCR2-deficient bone marrow as outlined above. Interestingly, substitution of alveolar IL-1ra by a single intratracheal application (10 μg) at 0 hours in CCR2^−/− mice reduced LPS-induced PMN recruitment to levels observed in WT mice, whereas intratracheal IL-1ra application into WT mice did not further reduce PMN numbers (Figure 4A). Respective lung sections demonstrate enhanced neutrophilic alveolitis in CCR2^−/− mice, which was attenuated by IL-1ra treatment (Figure 4B). Furthermore, IL-1ra but not PBS treatment reduced alveolar release of the PMN-recruiting chemokine macrophage inflammatory protein-2 in CCR2-deficient mice at 12 hours after LPS challenge, but had no effect in WT mice (Figure 4C). Similarly, surface expression of AEC intracellular adhesion molecule-1, known to promote PMN recruitment into the airspaces by interaction with PMN β₁- and β₂-integrins, was reduced in IL-1ra–treated CCR2-deficient mice 24 hours after LPS treatment (Figure 4D). Flow cytometric analyses of adhesion molecule expression on alveolar PMN (β₁- or β₂-integrins) revealed no changes between WT and CCR2-deficient mice (data not shown). Together, our data suggest that ExMa-expressed IL-1ra attenuates neutrophilic alveolitis, involving inhibition of chemokine release from and adhesion molecule expression on AEC.

Figure 4. Alveolar substitution of IL-1ra attenuates neutrophilic alveolitis on LPS challenge in CC-chemokine receptor 2 (CCR2)-deficient mice. (A) Wild-type (WT) or CCR2^−/− mice or bone marrow (BM) chimeric mice (WT BM → WT mice; CCR2^−/− BM → WT mice) were either left untreated or were intratracheally treated with LPS in the absence or presence of intratracheally applied IL-1ra (10 μg/mouse, at the time point of LPS administration) and polymorphonuclear neutrophil (PMN) numbers from bronchoalveolar lavage fluid (BALF) were counted at 48 hours after challenge (n = 3–13 mice per group). (B) Cryosections from perfused lungs of WT or CCR2^−/− mice or CCR2^−/− mice after additional IL-1ra application were prepared 48 hours after LPS treatment and hematoxylin-eosin stained, magnification ×200; scale bar = 100 μm. (C) Macrophage inflammatory protein (MIP)-2 levels were determined by ELISA from BALF of n = 4–6 LPS-challenged WT and CCR2^−/− mice treated intratracheally with phophate-buffered saline (PBS) or IL-1ra, respectively. (D) Flow cytometric analysis of intracellular adhesion molecule (ICAM)-1 surface expression on CD45⁻CD31⁻ alveolar epithelial cells (AEC) from lung digests of n = 3 LPS-challenged WT and CCR2^−/− mice treated intratracheally with PBS or IL-1ra, respectively. Data are given as ICAM-1-APC-Cy7 mean fluorescence intensity. All bar graphs represent means ± SD; **P < 0.01; ***P < 0.005. untr. = untreated.

Download Figure | Download Powerpoint

[More] [Minimize]

IL-1ra was previously shown to exert antiapoptotic effects on cardiomyocytes (31). We therefore speculated that the increase in AEC apoptosis in CCR2^−/− mice might be related to lacking ExMa IL1-ra. Intratracheal substitution of IL-1ra reduced AEC apoptosis rates in LPS-treated CCR2-deficient mice at 12 and 48 hours compared with sham-treated controls but revealed no further antiapoptotic effect in LPS-treated WT mice (Figure 5A). Accordingly, mRNA expression of the proapoptotic cytokine Fas ligand (FasL), known to mediate AEC apoptosis during pathogen-induced ALI (34), was induced on LPS instillation in lung homogenates approximately 20-fold in WT mice and approximately 110-fold in CCR2^−/− mice, which could be antagonized by intratracheal IL-1ra application in CCR2^−/− mice but not WT mice at 6 hours of LPS treatment (Figure 5B). These data indicate that IL-1ra acts in an antiapoptotic way on AEC after LPS treatment, possibly through inhibition of a FasL-mediated apoptosis pathway.

Figure 5. Alveolar substitution of IL-1ra attenuates alveolar epithelial cell (AEC) apoptosis on LPS challenge in CC-chemokine receptor 2 (CCR2)-deficient mice. (A) LPS-treated wild-type (WT) or CCR2^−/− mice (n = 3–4 per group) were additionally applied either phophate-buffered saline (PBS) or 10 μg IL-1ra intratracheally at the time point of LPS administration and Annexin V-Alexa647 binding to CD45⁻CD31⁻ AEC from lung cell homogenates was quantified by flow cytometry at 12 hours (solid bars) or 48 hours (open bars), or (B) relative mRNA expression of the proapoptotic Fas ligand (FasL) from lung homogenates was assessed by quantitative reverse transcriptase–polymerase chain reaction in the four treatment groups and values are given as fold induction over non–LPS-treated controls. All bar graphs represent means ± SD; *P < 0.05; ***P < 0.005.

Download Figure | Download Powerpoint

[More] [Minimize]

Given that IL-1ra attenuated alveolar inflammation and AEC apoptosis, we hypothesized that IL-1ra treatment would stabilize alveolar barrier function. Indeed, alveolar IL-1ra deposition attenuated alveolar albumin leakage in CCR2-deficient mice to levels observed in WT mice at 12 hours after LPS treatment (Figure 6A). As it was previously shown that, apart from its proinflammatory actions, IL-1β might directly increase lung epithelial permeability (3), we studied surface expression of the tight junction protein ZO-1 in confluent murine AEC on LPS- or IL-1β treatment in vitro by immunofluorescence. As shown in Figure 6B, ZO-1 was surface expressed on confluent, untreated AEC. LPS treatment alone (1 μg/ml) induced a localized tight junction disruption after 8 hours as indicated by arrowheads. Treatment with IL-1β (10 ng/ml) alone or with LPS plus IL-1β resulted in tight junction disassembly at 4 hours followed by complete disruption of the AEC monolayer after 8 hours, which could be antagonized by the addition of 1 μg/ml IL-1ra. Furthermore, transepithelial electrical resistance in primary murine (top panel) or human (bottom panel) AEC was reduced on a 2-hour treatment with IL-1β alone or with LPS + IL-1β, which was inhibited in the presence of IL-1ra.

Figure 6. IL-1ra attenuates LPS- or IL-1β–induced alveolar epithelial cell (AEC) barrier dysfunction and enhances epithelial ENaC expression in vitro and in vivo. (A) Twelve-hour LPS-treated wild-type (WT) or CC-chemokine receptor 2 (CCR2)^−/− mice (n = 6–13 per group) were additionally applied either phosphate-buffered saline (PBS) or 10 μg IL-1ra intratracheally at the time point of LPS administration and alveolar albumin leakage was quantified. (B) ZO-1 expression (green) on murine AEC monolayers was visualized by immunofluorescence at 4 and 8 hours after stimulation with either IL-1β, LPS, or a combination thereof in the absence or presence of IL-1ra. Arrowheads indicate areas of ZO-1 (tight junction) disassembly, magnification ×1,000 (top panel). Transepithelial electrical resistance across transwell-cultured confluent murine (top panel) or human (bottom panel) AEC was measured after 2-hour treatment with either IL-1β, LPS, or a combination thereof in the absence or presence of IL-1ra, and values are given as relative resistances to untreated controls from n = 3–8 independent experiments. (C) ENaC α and β subunit expression was analyzed by quantitative reverse transcriptase–polymerase chain reaction (RT-PCR) from digests of n = 3 lavaged lungs of 6-hour LPS-treated CCR2^−/− mice with or without IL-1ra treatment, and values are given as fold expression of non–LPS-treated CCR2^−/− mice. (D) ENaC α subunit expression was analyzed by quantitative RT-PCR of murine or human AEC treated for 4 hours with LPS and IL-1β in the absence or presence of IL-1ra and values are given as fold expression of untreated AEC (n = 3–4 independent experiments). All bar graphs represent means ± SD; *P < 0.05; ***P < 0.005. AU, arbitrary units; ctl = control; hAEC = human alveolar epithelial cells; mAEC = murine alveolar epithelial cells; TER = transepithelial resistance.

Download Figure | Download Powerpoint

[More] [Minimize]

As shown in Figure 1C, LPS-induced alveolar albumin leakage in CCR2-deficient mice was not only increased at 12 hours after challenge compared with WT mice, but tended to persist over 72 hours, when values in WT mice had already declined to baseline levels. We therefore speculated that alveolar fluid clearance, maintained mainly by the epithelial sodium channel (ENaC) (35), might be compromised in ExMa-deficient mice. Indeed, CCR2-deficient mice displayed reduced lung gene expression of ENaC subunits α and β on 6 hours of LPS treatment. Intratracheal treatment with IL-1ra counteracted these effects, whereby ENaCα subunit expression even exceeded levels of untreated control mice (Figure 6C). Finally, stimulation of murine or human AEC with IL-1β in the presence of LPS reduced expression of ENaCα subunit compared with unstimulated controls, and this decrease was not only abrogated in the presence of IL-1ra, but, on addition of 1 μg/ml IL-1ra, ENaCα levels even exceeded those of unstimulated AEC (Figure 6D). These data indicate that, apart from indirect proinflammatory or proapoptotic effects on alveolar epithelium, IL-1β directly induces structural and functional alveolar epithelial barrier damage, which can be widely antagonized by IL-1ra.

To further confirm that ExMa-released IL-1ra accounted for the beneficial effects on AEC barrier function in our model, we established an adoptive transfer approach using CCR2^−/− recipient mice that were either intravenously transferred IL-1ra^+/+CCR2^−/−, IL-1ra^+/+CCR2^+/+, or IL-1ra^−/−CCR2^+/+ mononuclear cells isolated from peripheral blood of CCR2^−/−, WT, or IL-1ra^−/− mice, respectively. Of note, in this model, CCR2-deficient recipient mice will largely recruit adoptively transferred CCR2^+/+ monocytes into their lungs on LPS challenge. As shown in Figure 7, AEC apoptosis at 12 hours after LPS treatment (Figure 7A) was significantly higher in mice recruiting IL-1ra^−/− compared with mice recruiting IL-1ra^+/+ ExMa. Recipient mice that were transferred CCR2^−/− mononuclear cells correspondingly demonstrated high AEC apoptosis. Similarly, albumin leakage was higher in CCR2-deficient mice recruiting IL-1ra^−/− ExMa compared with mice recruiting IL-1ra^+/+ ExMa (Figure 7B). These data clearly demonstrate that attenuation of LPS-induced lung injury is mediated by alveolar recruitment of IL-1ra^+/+ but not IL-1ra^−/− ExMa.

Figure 7. Antiapoptotic and barrier-protective effects are mediated by exudate macrophage (ExMa)-derived IL-1ra. CC-chemokine receptor 2 (CCR2)^−/− recipient mice were intravenously injected with IL-1ra^+/+CCR2^−/−, IL-1ra^+/+CCR2^+/+, or IL-1ra^−/−CCR2^+/+ peripheral blood mononuclear cells (MNC) at the time point of LPS administration and (A) alveolar epithelial cells (AEC) apoptosis and (B) alveolar albumin leakage were analyzed after 12 hours from n = 4–8 mice per group. All bar graphs represent means ± SD; *P < 0.05. AU = arbitrary units.

Download Figure | Download Powerpoint

[More] [Minimize]

We finally sought to confirm our findings on the beneficial role of ExMa IL-1ra in LPS-induced lung injury in a mouse model of gram-negative pneumonia. We therefore infected WT and CCR2-deficient mice with 10⁵ K. pneumoniae and comparatively addressed the effects of defective alveolar ExMa recruitment and of compensatory alveolar deposition of IL-1ra into CCR2-deficient mice on lung injury parameters. As outlined in Figure 8A, IL-1ra was alveolarly released to significantly higher amounts at 48 hours post-infection (hpi) in WT mice compared with CCR2^−/− mice. Of note, intratracheal deposition of 10 μg IL-1ra into K. pneumoniae–challenged CCR2-deficient mice compensated reduced IL-1ra levels to those observed in WT mice. K. pneumoniae infection induced pronounced AEC apoptosis in CCR2^−/− mice peaking at 24 to 48 hpi, which was widely reduced in WT mice. A single intratracheal application of IL-1ra at the time point of infection attenuated AEC apoptosis in CCR2-deficient mice at 48 hpi (Figure 8B). Correspondingly, alveolar albumin leakage was increased in CCR2^−/− mice at 48 hours after K. pneumoniae infection compared with WT mice, and leakage induction was inhibited in CCR2-deficient mice after IL-1ra treatment (Figure 8C). To determine whether IL-1ra treatment affected host defense functions in terms of bacterial clearance from the lungs, bacteria in BALF from the different treatment groups were counted by plating. We found that CCR2 deficiency significantly impaired bacterial clearance at 24 and 48 hpi; however, IL-1ra treatment of CCR2^−/− mice did not further impact on bacterial loads (Figure 8D). Finally, CCR2-deficiency was associated with a 100% mortality rate after 10 days post-infection (dpi), whereas all WT mice survived a dose of 1 × 10⁵ K. pneumoniae. Early intratracheal treatment with IL-1ra significantly increased survival rates of CCR2^−/− mice to 67% (Figure 8E). Together, these data suggest that ExMa-derived IL-1ra is protective in terms of epithelial injury, barrier function, and survival in gram-negative pneumonia.

Figure 8. Alveolar deposition of IL-1ra attenuates epithelial injury without affecting host defense in Klebsiella pneumoniae pneumonia. (A) Wild-type (WT), CC-chemokine receptor 2 (CCR2)^−/−, or IL-1ra treated CCR2^−/− mice were intratracheally infected with 10⁵ K. pneumoniae and alveolar IL-1ra amounts were quantified from bronchoalveolar lavage fluid (BALF) at 48 hours after infection (hpi) (n = 3 each). (B) Alveolar epithelial cell (AEC) apoptosis was analyzed at 24 and 48 hpi by fluorescence-activated cell sorting of lung cell homogenates from n = 3–4 infected WT or CCR2^−/− mice additionally treated with 10 μg IL-1ra or phosphate-buffered saline (PBS) intratracheally at the time point of infection. (C) Albumin leakage was determined at 48 hpi, and (D) K. pneumonia loads in BALF were quantified by plating serial dilutions and colony-forming unit (cfu) counting at 24 and 48 hpi. (E) Survival rates of WT and CCR2^−/− mice intratracheally treated with 10 μg IL-1ra or PBS at the time point of infection with 10⁵ K. pneumoniae (n = 6 mice per group). All bar graphs represent means ± SD; *P < 0.05. AU = arbitrary units; n. s. = not significant.

Download Figure | Download Powerpoint

[More] [Minimize]