C57BL/6 CD45.2⁺ (B6 or B6-CD45.2⁺) mice were purchased from Charles River (L’Arbresle, France) and C57BL/6 CD45.1⁺ (B6-CD45.1⁺) mice from Jackson (Maine). LysM-eGFP (7) and CX3CR1-eGFP (8) mice were supplied by A. Hidalgo (CNIC, Spain). 8-10-week old B6, B6-CD45.1⁺ and LysM-eGFP mice were housed at the CNB Animal Facility on a 12/12 light/dark cycle. Littermates of the same sex were randomly assigned to experimental groups. Bone marrow (BM) from Cx3cr1 ^cre:R26-yfp mice (9) was provided by S. Jung (The Weizmann Institute of Science, Israel). All the experiments were approved by CNB Animal Care Committee (protocol 312/14).

HDM-allergy was induced following a protocol modified from a protocol described in a previous report from our group (10). In brief, mice received an intraperitoneal injection of 5 x 10⁴ monocyte-derived dendritic cells at day -7, followed, at day 0, by intratracheal administration of 20 μg HDM-extract (Dermatophagoides pteronyssinus extract; Greer Laboratories, Lenoir, North Carolina), in a total volume of 40 μl of PBS, under ketamine/xylazine anesthesia. Monocyte-derived dendritic cells were prepared as described (10) and incubated prior to injection for 4 hr with 30 μg/ml HDM extract.

Bronchoalveolar lavage (BAL) was performed with 3 x 1 mL EDTA-containing PBS. BAL cell suspensions were obtained after centrifugation for 5 min at 400g. Lung cell suspensions were obtained after performing a BAL; for this purpose lungs were cut into small pieces, digested with 180 μg/ml Liberase TM and 40 mg/ml DNAse (both from Roche, Mannheim, Germany) for 40 min at 37°C, resuspended in RPMI supplemented with 10% FCS, filtered through 40-μm cell strainers (BD Pharmingen, San Diego, CA) and washed twice in EDTA-containing PBS after erythrocyte lysis by osmotic shock.

CD45.1/CD45.2 thorax-shielded BM chimeras were established by intravenous (i.v.) injection of 5 x 10⁶ BM cells from C57BL/6 (B6)-CD45.1⁺ mice into 8-week old lethally irradiated B6-CD45.2⁺ mice (single dose 10 Gy γ-radiation, using a JL Shepherd Mark I-30 ¹³⁷Cs irradiator), in which the thorax was lead shield-protected. Cx3cr1 ^cre:R26-yfp/CD45.1 thorax-shielded BM chimeras were established by i.v. injection of 5 x 10⁶ BM cells from CD45.2⁺ Cx3cr1 ^cre:R26-yfp mice into 8-week old lethally irradiated, thorax-protected, B6-CD45.1⁺ mice.

4 x 10⁶ BM Ly6C^high monocytes, isolated as described (11), from LysM-eGFP mice, at d1 of HDM allergy, were transferred intravenously into B6 mice at d1 of HDM allergy.

Mice euthanized by controlled CO₂ inhalation overdose received 125 ng of Pacific Blue-conjugated anti-CD45 intratracheally (i.t.) in a total volume of 1 ml (corresponding to a 1/4.000 dilution); control mice received 1 ml of PBS intratracheally. The concentration and volume of Pacific Blue-conjugated anti-CD45 was determined in titration experiments in order to ensure an efficient staining of AMØs while avoiding the diffusion of the antibody into the lung parenchyma. After 1 min intra-alveolarly-located cells were analyzed as described below.

Analysis of lung and BAL cell suspensions was performed after seven-color staining with FITC-conjugated anti-MHCII (clone 2G9; BD Pharmingen), PECy7-conjugated anti-CD11b (clone M1/70; eBioscience, San Diego, CA) or PECy7-conjugated anti-Ly-6G (clone 1A8; BD Pharmingen), APC-conjugated anti-CD64 (clone X54-5/7.1; Biolegend, San Diego, CA), APC-Cy7-conjugated anti-CD11c (clone HL3; BD Pharmingen), PE-conjugated anti-Siglec-F (E50-2440; BD Pharmingen), Pacific Blue-conjugated anti-CD45 (clone 30-F11; Biolegend) and biotin-conjugated anti-Ly6C (clone Al-21) followed by streptavidin-PerCP (BD Pharmingen). Analysis of dead AMØs was performed after seven-color staining with PECy7-conjugated anti-Ly-6G, PE-conjugated anti-Siglec-F, Pacific Blue-conjugated anti-CD45, PerCP/Cy5.5-conjugated anti-CD64 (clone X54-5/7.1; Biolegend) and biotin-conjugated anti-Ly6C, followed by streptavidin-APC-Cy7 (Biolegend). Cells were then resuspended at 1 × 10⁶ cells/ml in 100 μl binding buffer (Biolegend) and incubated with FITC-conjugated Annexin-V (Biolegend) for 15 min at 4 °C in the dark. AMØs death was assessed by Annexin-V staining after gating on AMØs. Analysis of CD45.1/CD45.2 BM chimeras was performed after seven-color staining with FITC-conjugated anti-MHCII, PECy7-conjugated anti-Ly-6G, APC-conjugated anti-CD64, PE-conjugated anti-Siglec-F, Brilliant Violet 421-conjugated anti-CD45.1 (clone A20; Biolegend), PerCP/Cy5.5-conjugated anti-CD45.2 (clone 104; Biolegend) and biotin-conjugated anti-Ly6C, followed by streptavidin-APC-Cy7. Analysis of Cx3cr1 ^cre:R26-yfp/CD45.1 BM chimeras was performed after seven-color staining with PECy7-conjugated anti-CD11b, APC-conjugated anti-CD64, PE-conjugated anti-Siglec-F and anti-Ly6G, Brilliant Violet 421-conjugated anti-CD45.1, PerCP/Cy5.5-conjugated anti-CD45.2 and biotin-conjugated anti-Ly6C, followed by streptavidin-APC-Cy7, or after seven-color staining with PECy7-conjugated anti-CD11b, APC-conjugated anti-CD64, PE-conjugated anti-Siglec-F, Brilliant Violet 421-conjugated anti-CD45.1, PerCP/Cy5.5-conjugated anti-CD45.2 and biotin-conjugated anti-MHC-II, followed by streptavidin-APC-Cy7. Analysis of cells derived from intravenously-transferred LysM-eGFP-monocytes was performed after six-color staining with PECy7-conjugated anti-Ly-6G, APC-conjugated anti-MHC-II (clone M5/114.15.2; eBioscience), PE-conjugated anti-Siglec-F, Pacific Blue-conjugated anti-CD45, PerCP/Cy5.5-conjugated anti-CD64 and biotin-conjugated anti-Ly6C, followed by streptavidin-APC-Cy7. Analysis of intra-alveolarly-located cells was performed after seven-color staining with FITC-conjugated anti-MHCII, PECy7-conjugated anti-CD11b, APC-conjugated anti-CD64, APC-Cy7-conjugated anti-CD11c, PE-conjugated anti-Siglec-F, PerCP/Cy5.5-conjugated anti-CD45.2 and biotin-conjugated anti-Ly6C, followed by streptavidin-APC-Cy7.

Autofluorescence was detected using a 525/50-nm band pass filter after excitation with a 405 nm-violet laser.

Analysis of CD206 expression was performed after seven-color staining with FITC-conjugated anti-MHCII, PECy7-conjugated anti-Ly-6G, PE-conjugated anti-Siglec-F, Pacific Blue-conjugated anti-CD45, PerCP/Cy5.5-conjugated anti-CD64 and biotin-conjugated anti-Ly6C, followed by streptavidin-APC-Cy7. Subsequently, cells were fixed in Cytofix/Cytoperm solution (BD Pharmingen) and incubated with APC-conjugated anti-CD206 (clone C068C2; Biolegend). Data were acquired on a LSRII cytometer (BD Biosciences, San José, CA) and analyzed using FlowJo X software (Tree Star, Ashland, OR).

Lungs were fixed with 2.5% glutaraldehyde and 4% paraformaldehyde in PBS for 2 hours at RT and overnight at 4°C, postfixed with 1% osmium tetroxide/0,8% potassium ferricyanide for 1 h at 4°C, dehydrated in graded acetone solutions and embedded in Epon-812 (TAAB Laboratories Ltd., Berkshire, UK). Semi-thin sections (1 µm) were obtained with a Leica EM UC6 ultramicrotome and stained with toluidine blue. Images were acquired with a Leica DM2500 microscope (Leica, Wetzlar, Germany). Ultra-thin sections (80 nm) were obtained with a Leica EM UC6 ultramicrotome, counterstained with uranyl acetate and lead citrate and examined with a Jeol 1011 transmission electron microscope (Tokyo, Japan).

Cell-free BAL supernatant was obtained by centrifugation for 10 min at 700 g at 4°C. Total BAL protein concentration was measured in cell-free BAL supernatants by Lowry’s method modified by adding sodium dodecyl sulphate (SDS). Lipid extraction was performed by chloroform and methanol extraction and quantitation of BAL phospholipids was achieved by phosphorus analysis, as described (12).

Electrophoretic analysis of BAL proteins was performed under reducing conditions (5% β-mercaptoethanol) by one-dimensional SDS/PAGE, using running gels of 12% for SP-A and SP-D, 16% for SP-B and 18% for SP-C. In the case of SP-A and SP-D, the same amount of BAL protein was applied for all samples. In case of SP-B and SP-C, the same amount of BAL phospholipids was applied for all samples. After electrophoresis, samples were transferred to polyvinylidene fluoride membranes (Bio-Rad Laboratories, Hercules, CA). Blotting was performed as previously described (12) SP-A and SP-B, were detected using anti-human-SP-A and anti-porcine SP-B polyclonal antibodies prepared in our laboratory. SP-C was detected using and an anti-recombinant human SP-C polyclonal antibody kindly provided by Nycomed Pharma (Konstanz, Germany). Secondary HRP-conjugated anti-rabbit IgG antibodies were from Cell Signaling Technologies, (Danvers, MA). SP-D was detected with an anti SP-D mAb (clone 1A10A9, Seven Hills Bioreagents, Cincinnati, OH), followed by a secondary anti-mouse IgG antibody (Sigma-Aldrich, St. Louis, MO). Proteins were visualized using chemiluminiscence detection (Amersham Hyperfilm ECL, GE Healthcare, Little Chalfont, UK). Protein bands were quantified by densitometry using the Quantity One software (Bio-Rad Laboratories).

Total surfactant complexes (large + small aggregates) were precipitated from cell-free BAL supernatants by ultracentrifugation for 60 min at 100,000 g at 4°C, and diluted to 10 mg/mL phosphatidylcholine concentration with Tris buffer (5mM; pH7) containing 150 mM NaCl (Sigma-Aldrich). Phosphatidylcholine concentration was measured using enzymatic methods (Spinreact, Gerona, Spain) as previously published (13). The interfacial activity of surfactant samples was assessed by captive bubble surfactometry, a technique allowing to analyze surfactant function during breathing-like compression-expansion cycles (14). To this end, 200 nL of surfactant, at 10 mg/mL, was applied at the surface of an air bubble in a Captive Bubble Surfactometer (CBS) at a frequency of 30 compression-expansion cycles/min, as described in Supplementary Figure 1 . Measurements were performed in triplicate for each surfactant sample.

Data are presented as mean ± SD (or mean ± SEM), as indicated in the legend of each figure. Statistical parameters including the exact value of n, precision measures (mean ± SD or SEM) and statistical significance are reported in the Figures and the Figure Legends when necessary. Data are judged to be statistically significant when p < 0.05 by two-tailed Student’s t test. In figures, asterisks denote statistical significance (*, p < 0.05; **, p < 0.01; ***, p < 0.001). Statistical analysis was performed using Microsoft Excel or GraphPad PRISM 6 softwares.

Airway allergy was induced by HDM following the protocol described in Figure 1A . HDM-driven allergy caused an airway inflammation similar to human allergic asthma (15), that involved a massive leukocyte infiltration into the lung and eosinophilia, peaking at day 5 (d5) after i.t. HDM-extract challenge ( Figures 1B , C ), paralleled by bronchiolar inflammation, characterized by increased mucus production and fibrosis ( Figure 1D ). Unexpectedly, HDM allergy caused a profound disorganization of the alveolar system leading to reduction of the alveolar space ( Figure 1E ). Neutrophils were also actively recruited by d1, but their number dropped by d8 ( Figure 1C ). Analysis of lung cell suspensions revealed that the AMØ population, characterized as CD11c^high Siglec-F^high autofluorescent cells ( Figures 1B, F ), was markedly reduced at d3, and almost undetectable at d5 ( Figures 1B, C ). AMØ reduction was concomitant with an increase in eosinophils ( Figure 1C ) and Annexin-V⁺ AMØs ( Figure 1G ), suggesting the reduction in AMØs was due to cell death. AMØ disappearance was also noticeable in the BAL that contained a large number of eosinophils at d5-d8 ( Figures 1H, I ). Importantly, only a fraction of AMØs can be harvested by BAL (〜10-15% of total AMØs), either in control or allergic mice, indicating that AMØ studies have to be performed in lung cell suspensions, but not in BAL cell suspensions.

From d8 to d13, eosinophils dropped significantly, and a population of autofluorescent CD11c^high cells with lower Siglec-F expression than control AMØs, although comparable in number, was detectable in lung ( Figures 1B, C ) and BAL cell suspensions ( Figures 1H, I ) indicating that, at least in part, CD11c^high Siglec-F^int cells were intra-alveolarly located. By d33, a population of AMØs expressing CD11c and Siglec-F levels matching those of control AMØs was noticeable, and AMØ and eosinophil number returned to basal levels ( Figures 1B, C ). These data led us to hypothesize that at the peak of eosinophilia and alveolar disorganization, HDM allergy caused the death of original AMØs, that were replaced by new-AMØs.

To address whether new-AMØs were originated by self-renewal from original embryonic-derived AMØs, or de novo from BM progenitors, AMØ regeneration was analyzed CD45.1/CD45.2 chimeric mice, in which the hematopoietic system was mainly of donor origin, while AMØs were host-derived. To this end, BM cells from B6-CD45.1⁺ mice were transferred into B6-CD45.2⁺ mice, that were subjected to lethal γ-irradiation after protecting the thorax with a lead shielding. ( Figure 2A ). At d-7, 〜90% AMØs were of host-derived, whereas 〜80-90% of lung Ly6C^high monocytes (hereafter monocytes), B cells and neutrophils, were donor-derived ( Figures 2B, C ). A comparable chimerism was found in BM monocytes, B cells and neutrophils during HDM allergy ( Figure 2D ), reflecting that the lead shield had provided protection not only to AMØs, but also to thoracic hematopoietic foci that generated 〜10-20% host-derived monocytes, B cells and neutrophils. At d13〜70% AMØs were donor-derived (corresponding to 〜90% BM origin, normalized BM monocyte chimerism), revealing that original AMØs were almost entirely replaced during the allergic reaction by a new, BM-derived, AMØ population ( Figures 2B, C ). Interestingly, 2 months after allergy resolution 〜70% of AMØs were still of donor origin (corresponding to 〜95% BM origin, normalized BM monocyte chimerism), revealing that BM-derived AMØs persisted long after allergy had subsided ( Figures 2B, C ).

Under experimental or pathological conditions new tissue-resident macrophages, such as Kupffer cells, microglial cells, large peritoneal macrophages or AMØs can be generated from monocytes (16).

To assess whether AMØs newly-formed during HDM allergy were derived from monocytes, we first analyzed the kinetics of monocytes and monocyte-derived cells (moCs), defined following the gating strategy described in Figure 3A . In steady state, lung mature moCs were characterized as CD64⁺ Ly6C^low MHC class-II (MHCII)^high Siglec-F^- cells, as reported (17). Monocytes (defined as CD64⁺ Ly6C^high MHCII^- Siglec-F^- cells) were actively recruited to the lung at d1-d3, and dropped by d5 ( Figures 3B, C ). Monocyte differentiation into moCs, occurred through an intermediate stage of CD64⁺ Ly6C^high MHCII^low Siglec-F^- immature moCs (P2; Figure 3B ). moCs peaked at d3 and diminished progressively as allergy declined ( Figures 3B, C ), suggesting that moCs were generated after a single wave of monocyte recruitment.

To address whether monocytes recruited to the lung were able to generate AMØs during HDM allergy, fluorescence-trackable BM monocytes were isolated from LysM-eGFP mice at d1 after HDM challenge, i.e. when monocytes started to be recruited ( Figure 3C ), and transferred i.v. into B6 mice at d1 ( Figure 3D ). At d2, 24 h after transfer, 〜50% eGFP⁺ cells were monocytes, and 〜50% were Ly6C^high MHCII^int Siglec-F^- immature moCs (P2; Figure 3E ). At d5, 〜15% eGFP⁺ cells were monocytes and 〜85% were Ly6C^low MHCII^high Siglec-F^- mature moCs. At d13 〜60% of lung eGFP⁺ cells were MHC II^int Siglec-F^int new-AMØs, whereas the rest were Ly6C^low MHCII^high moCs expressing intermediate levels of Siglec-F, that based on their phenotype, kinetics and alveolar location (as demonstrated later in this report) were immature AMØs, hereafter pre-AMØs. At this stage, endogenous moCs could be subdivided in Ly6C^low MHCII^high Siglec-F^int pre-AMØs and Ly6C^low MHCII^high Siglec-F^- mature moCs ( Figure 3E ). By d33, all lung eGFP⁺ cells were fully differentiated MHCII^int Siglec-F^high new-AMØs, pre-AMØs being no longer detectable ( Figure 3E ), supporting that new-AMØs were generated from monocytes, through an intermediate Ly6C^low MHCII^high Siglec-F^int pre-AMØ population. The phenotype of monocytes, moCs, pre-AMØs and AMØs is summarized in Figure 3F .

Monocyte transfer experiments allowed to assess the ability of exogenous monocytes to be recruited to the lung and to monitor the different phases of their differentiation into AMØs, but do not provide a definitive demonstration of the monocytic origin of endogenous new-AMØs generated during HDM allergy. This issue was addressed analyzing AMØ regeneration in Cx3cr1 ^cre:R26-yfp/CD45.1 BM chimeras ( Figure 4A ), in which CX3CR1 promoter-driven Cre recombinase expression leads to the rearrangement of the YFP reporter locus, establishing a stable fluorescent labeling (9). As expected, 4 weeks after BM transfer, 〜90% AMØs were of host origin whereas 〜70-80% monocytes, B cells and neutrophils, were donor-derived ( Figure 4B ). At this time point, 〜15% donor-derived BM monocytes and 〜30% donor-derived blood and lung monocytes were YFP⁺ ( Figure 4C ), most likely due to their short life span not allowing for 100% rearrangement of YFP locus. At d5, 〜30% donor-derived lung monocytes expressed YFP, and a comparable YFP expression was found on donor-derived moCs ( Figure 4D ), whereas at d33 up to 80% donor-derived moCs expressed YFP ( Figure 4E ), reflecting that moCs generated during the resolution of allergy had a longer life span and maintained the activity of the CX3CR1 promoter, leading to an efficient YFP locus rearrangement. Interestingly, at this time point 〜40% donor-derived lung monocytes and new-AMØs expressed YFP. These results indicate that transition from monocytes to new-AMØs was concomitant with the inactivation of the CX3CR1 promoter and therefore with the shut off of Cre-mediated YFP locus rearrangement. In support of this postulate, CX3CR1 expression was downregulated from monocytes to AMØs, as assessed by the analysis of CX3CR1-eGFP mice during HDM allergy ( Figure 4F ).

We next sought to addressed whether monocyte differentiation into AMØs occurred extra- or intra-alveolarly, by assessing the location of pre-AMØs based on their autofluorescence, a characteristic of AMØs associated to their intra-alveolar location, which has been correlated to their exposure to environmental particulate matter (18). In control mice, and at d1 and d13 of HDM allergy, AMØs, but neither monocytes nor MHCII^high Siglec-F^- moCs, were autofluorescent ( Figure 5A ). However, at d13, both pre-AMØ pre-AMØs and new-AMØs displayed autofluorescence levels and a FSC vs SSC profile comparable to those of AMØs from control mice, supporting that monocyte differentiation into pre-AMØs determined their translocation into the alveolar space. To confirm this hypothesis, intra-alveolarly-located cells were detected by i.t. administration of anti-CD45 antibodies, as described in Figure 5B . pre-AMØs and new-AMØs, but neither monocytes nor the majority of moCs, were stained with anti-CD45 ( Figure 5C ), confirming the intra-alveolar location of pre-AMØs. These results further support that new-AMØs were generated from monocytes, through an intermediate pre-AMØ population. pre-AMØs translocation into the alveolar space was therefore concomitant with the acquisition of autofluorescence and upregulation of Siglec-F and the M2 macrophage marker CD206 ( Figure 5D ).

At d5, HDM allergy caused a severe bronchiolar inflammation characterized by increased mucus production and fibrosis ( Figure 1D ), as well as profound alterations in alveolar organization ( Figure 6A ), that included AMØ disappearance ( Figure 1B ) and massive alveolar eosinophilia ( Figure 1H ). In addition, electron microscopy studies revealed an extensive pneumocyte hypertrophy and leukocyte extravasation into the alveolar space at d5 ( Figures 6B1, B2 ), not detectable in control lung ( Figure 6C ). Pneumocyte hypertrophy led to a marked thickening of alveolar wall and consequently to a significant reduction of the alveolar space and gas-exchange surface area ( Figures 6D, E ). By d13, the alveolar organization was essentially reestablished and bronchiolar inflammation was resolved ( Figure 6A ).

Since lung surfactant has an essential role in stabilizing the alveolar gas-aqueous interface and preventing lung collapse, we asked whether the disrupted alveolar organization and pneumocyte hypertrophy observed at the peak of HDM allergy, altered the composition and biophysical properties of lung surfactant. Total phospholipid concentration present in BAL remained unchanged until d5, but underwent a marked increase during the resolution of allergy ( Figure 6F ). In contrast, total BAL protein concentration increased during the induction of allergy and dropped during the resolution phase ( Figure 6G ). Surfactant-associated proteins SP-A and SP-D increased during the induction of the allergic reaction and more markedly from d8, i.e. during the resolution of HDM allergy, as assessed by western blot ( Figure 6H ). In contrast, SP-B and SP-C underwent a strong reduction by d5 reaching values significantly lower than those found in non-allergic mice ( Figure 6H ).

Based on these results, whether these changes in surfactant proteins affected the biophysical properties of lung surfactant was analyzed by captive bubble surfactometry (CBS), a method described in Supplementary Figure 1 , allowing to assess the ability of a surfactant sample to form a surface-active film competent to produce and sustain a very low surface tension at the air-liquid interface of an air microbubble, mimicking an alveolar chamber. By monitoring surface tension changes, we analyzed the “initial adsorption” of the surfactant to the air-liquid interface, the ability of adsorbed surfactant to spread and re-equilibrate into an expanding interface, or “post-expansion adsorption”, and the ability of the formed film to produce the very low surface tensions (<5 mN/m) required to stabilize the lungs during compression-expansion cycles, or “dynamic isotherms”, that mimic breathing dynamics (19).

Surfactant from d-7 and d1 adsorbed (initial adsorption) and spread (post-expansion adsorption) at the air-liquid interface almost instantaneously, reducing surface tension from 〜70 mN/m to below 25 mN/m in less than 1 s ( Figures 6I, J ). In contrast, surfactant from d5 required more than 20 s to reach similar low surface tension values. Interestingly, fast interfacial adsorption, both initially and upon expansion, was restored in surfactant from d13. Upon compression-expansion cycles, interfacial films formed by surfactant from d-7 and d1 produced extremely low surface tensions at the end of every cycle ( Figures 6K, L ), with less than 20% area reduction from the second cycle ( Figures 6K, M ). These very low tension values were repeatedly achieved along 20 cycles, with very little hysteresis (assessed by the area enclosed between compression and expansion moieties of isotherms), indicating that the interfacial film had a high stability. Films formed by surfactant from d5 produced a deficient stabilization of the interface during compression-expansion cycles ( Figure 6K ). Only in some cases d5 films reached low enough surface tensions, yet requiring larger area reductions during a number of cycles, and showed considerably more hysteresis than films formed by surfactant from d-7 or d1. The low stability of films formed by surfactant obtained at the peak of HDM allergy also translated into the considerably higher maximum tension reached during the expansion phases ( Figures 6K, N ). Surfactant from d13 had fully restored dynamic isotherms and functional parameters, since in all cases films reached very low tensions ( Figure 6L ), with less than 20% area reduction ( Figure 6M ), while maintaining maximal tensions at the end of expansion ( Figure 6N ), that did not substantially differ from the equilibrium values produced at the end of the adsorption stage, along the 20 compression-expansion cycles applied.

These results revealed that surfactant obtained from mice during the peak of HDM allergy displayed a significant level of inactivation, characterized by slower interfacial adsorption and a reduced efficiency to attain and sustain very low surface tensions, raising the risk of alveolar collapse. Interestingly, during the resolution of the allergic reaction, the biophysical properties of surfactant were fully restored, as demonstrated by their high efficiency to form surface-active films maintaining very low surface tensions during breathing-like compression-expansion dynamics.