Vage (BAL) was obtained at Day 37, 77 or 112 by cannulating the trachea with a 24-gauge catheter. The right lung was lavaged twice (each aliquot 0.3 ml; NaCl 0.9 ). Total cell number was counted with a hemocytometer. Cytocentrifuge preparations (Cytospin 4, ThermoFisher Scientific, Courtaboeuf, France) were stained with Diff-QuikH (VWR International, Strasbourg, France), adapted from Giemsa-May-Gru �nwald stain. Differential cell counting was performed using standard morphological criteria in which 400 cells were analyzed by a blinded investigator using standard haematological criteria. Total leukocyte number, percentage of eosinophils, ML-240 neutrophils, lymphocytes and macrophages were determined in each BAL fluid.HistologyRight and left lung tissue were dissected out, fixed with formaldehyde in inflation using intrapulmonary injection, and embedded in paraffin. Histological analysis was performed using 4-mm-thick lung slices stained with haematoxylin-eosin-safran and a modified Masson’s trichrome (half dilution of hematoxylin as compared to the standard staining procedure). Immunohistochemistry was performed using an anti-mouse alpha smooth 
muscle actin antibody clone 1A4 (Dako, Trappes, France). Formalin-fixed paraffin-embedded tissue sections were incubated in pretreatment buffer for antigen retrieval and were then reacted with a mouse anti-smooth muscle a-actin for 15 minutes. Immunoreaction was detected using Bond Polymer Refine Detection (Leica Microsystems, Wetzlar, Germany) on Bond TM max (A.Menarini Diagnostics, Firenze, Italy). Several quantitative parameters were assessed using Quancoul software (Quant’Image, Bordeaux, France) at magnifications of 1006 to 4006 [18]. We measured the basal membrane thickness, the wall area, the GW-0742 web bronchial smooth muscle area and the peribronchial area. We also assessed the area of fibrosis within the peribronchial space and the number of nucleated cells within the peribronchial space. All were measured on HES and actinstained sections, except fibrosis which was quantified on modified Masson’s trichrome stain.In Vivo Micro-CT Assessment of Airway RemodelingStatistical AnalysisValues are expressed as the mean 6 SEM, except those related to microCT for which the normality could not be rigorously established. The agreement between the semi-automatic and manual methods for PBA measurement was assessed in 10 datasets chosen at random using Bland-Altman analysis [19]. The manual method has been described previously in detail [16] and was based upon a two-dimensional analysis from multiplanar reformations. For each group, the following parameters were compared between sensitized and control mice using Mann-WhitneyWilcoxon rank sum test: weight at endpoint, Penh ratio, LR, micro-CT parameters, BAL results and histological data. Correlations between, on 
the one hand, PBA or normalized PBA, and, on the other hand, Penh ratio, BAL or histological data, were assessed using the Spearman rank correlation coefficients. All analyses were performed using NCSS software (NCSS 2001, Kaysville, UT, USA) and results were considered statistically significant when P-values,0.05.Results Description of the Mouse Models of AsthmaFrom an initial set of 60 mice, 51 completed the study. Two mice died during the intubation procedure, 3 mice did not recover from anaesthesia following micro-CT, and 4 mice presented CT motion artefacts. Body weights were similar between control and OVA-sensitized mice at each endpoint. Table 1 displays.Vage (BAL) was obtained at Day 37, 77 or 112 by cannulating the trachea with a 24-gauge catheter. The right lung was lavaged twice (each aliquot 0.3 ml; NaCl 0.9 ). Total cell number was counted with a hemocytometer. Cytocentrifuge preparations (Cytospin 4, ThermoFisher Scientific, Courtaboeuf, France) were stained with Diff-QuikH (VWR International, Strasbourg, France), adapted from Giemsa-May-Gru �nwald stain. Differential cell counting was performed using standard morphological criteria in which 400 cells were analyzed by a blinded investigator using standard haematological criteria. Total leukocyte number, percentage of eosinophils, neutrophils, lymphocytes and macrophages were determined in each BAL fluid.HistologyRight and left lung tissue were dissected out, fixed with formaldehyde in inflation using intrapulmonary injection, and embedded in paraffin. Histological analysis was performed using 4-mm-thick lung slices stained with haematoxylin-eosin-safran and a modified Masson’s trichrome (half dilution of hematoxylin as compared to the standard staining procedure). Immunohistochemistry was performed using an anti-mouse alpha smooth muscle actin antibody clone 1A4 (Dako, Trappes, France). Formalin-fixed paraffin-embedded tissue sections were incubated in pretreatment buffer for antigen retrieval and were then reacted with a mouse anti-smooth muscle a-actin for 15 minutes. Immunoreaction was detected using Bond Polymer Refine Detection (Leica Microsystems, Wetzlar, Germany) on Bond TM max (A.Menarini Diagnostics, Firenze, Italy). Several quantitative parameters were assessed using Quancoul software (Quant’Image, Bordeaux, France) at magnifications of 1006 to 4006 [18]. We measured the basal membrane thickness, the wall area, the bronchial smooth muscle area and the peribronchial area. We also assessed the area of fibrosis within the peribronchial space and the number of nucleated cells within the peribronchial space. All were measured on HES and actinstained sections, except fibrosis which was quantified on modified Masson’s trichrome stain.In Vivo Micro-CT Assessment of Airway RemodelingStatistical AnalysisValues are expressed as the mean 6 SEM, except those related to microCT for which the normality could not be rigorously established. The agreement between the semi-automatic and manual methods for PBA measurement was assessed in 10 datasets chosen at random using Bland-Altman analysis [19]. The manual method has been described previously in detail [16] and was based upon a two-dimensional analysis from multiplanar reformations. For each group, the following parameters were compared between sensitized and control mice using Mann-WhitneyWilcoxon rank sum test: weight at endpoint, Penh ratio, LR, micro-CT parameters, BAL results and histological data. Correlations between, on the one hand, PBA or normalized PBA, and, on the other hand, Penh ratio, BAL or histological data, were assessed using the Spearman rank correlation coefficients. All analyses were performed using NCSS software (NCSS 2001, Kaysville, UT, USA) and results were considered statistically significant when P-values,0.05.Results Description of the Mouse Models of AsthmaFrom an initial set of 60 mice, 51 completed the study. Two mice died during the intubation procedure, 3 mice did not recover from anaesthesia following micro-CT, and 4 mice presented CT motion artefacts. Body weights were similar between control and OVA-sensitized mice at each endpoint. Table 1 displays.