Alveolar repair following LPS-induced injury requires cell-ECM interactions

Abstract

During alveolar repair, alveolar type 2 (AT2) epithelial cell progenitors rapidly proliferate and differentiate into flat AT1 epithelial cells. Failure of normal alveolar repair mechanisms can lead to loss of alveolar structure (emphysema) or development of fibrosis, depending on the type and severity of injury. To test if β1-containing integrins are required during repair following acute injury, we administered E. coli lipopolysaccharide (LPS) by intratracheal injection to mice with a postdevelopmental deletion of β1 integrin in AT2 cells. While control mice recovered from LPS injury without structural abnormalities, β1-deficient mice had more severe inflammation and developed emphysema. In addition, recovering alveoli were repopulated with an abundance of rounded epithelial cells coexpressing AT2 epithelial, AT1 epithelial, and mixed intermediate cell state markers, with few mature type 1 cells. AT2 cells deficient in β1 showed persistently increased proliferation after injury, which was blocked by inhibiting NF-κB activation in these cells. Lineage tracing experiments revealed that β1-deficient AT2 cells failed to differentiate into mature AT1 epithelial cells. Together, these findings demonstrate that functional alveolar repair after injury with terminal alveolar epithelial differentiation requires β1-containing integrins.

Keywords: Cell Biology; Integrins; Pulmonology; Respiration.

Figure 1. Deletion of β₁ integrin in AT2 cells results in increased inflammation, abnormal repair, and decreased survival after LPS-induced injury.

(A) Representative images of lung histology demonstrate increased edema at 3 and 7 days postinjury in β₁^AT2-KO lungs compared with β₁^fl/fl lungs, as well as persistent inflammation and emphysematous remodeling in β₁^AT2-KO lungs by 21 days post-LPS. D3, D7, and D21 refer to 3, 7, and 21 days after LPS injury, respectively. (B) Mean linear intercept quantified emphysematous alveolar remodeling at 21 days after LPS, 28.5 ± 0.9 μm in β₁^fl/fl lungs versus 40.2 ± 2.8 μm in β₁^AT2-KO lungs (n = 6–7 mice/group, P = 0.0014 by 2-tailed t test). (C) Bicinchoninic acid (BCA) protein assay quantified increased bronchoalveolar lavage (BAL) fluid protein in uninjured β₁^AT2-KO lungs and at 3 and 7 days post-LPS injury in β₁^AT2-KO lungs compared with β₁^fl/fl lungs at the same time points (n = 6–14 mice/group, 2-tailed t test comparing genotypes at each time point, P = 0.0485 for uninjured mice; P = 0.0036 at D3; P = 0.005 at D7; P = 0.2628 at D21). (D) BAL cell counts are significantly increased in β₁^AT2-KO lungs compared with β₁^fl/fl littermates in uninjured mice and at 7 and 21 days post-LPS. Peak inflammation is present at 7 days in β₁^AT2-KO lungs, 55,663 ± 3,306 cells/mL in β₁^fl/fl BAL fluid versus 624,000 ± 118,753 cells/mL in β₁^AT2-KO BAL (n = 6–26 mice/group, 2-tailed t test comparing genotypes at each time point, P = 0.0002 for uninjured mice; P = 0.0730 at D3; P = 0.0007 at D7; P < 0.0001 at D21). Total numbers of BAL fluid macrophages are significantly increased in uninjured β₁^AT2-KO mice and at D7 and D21; lymphocytes and neutrophils are significantly increased in β₁^AT2-KO BAL at D7 only. Scale bar = 50 μm for A. * P < 0.05.

Figure 2. Lungs deficient in β₁have increased AT2 cell number and decreased AT1 cells at 21 days after LPS injury.

(A) Representative low-power (original magnification, 10×) and inset high-power images (original magnification, 40×, denoted by yellow boxes) of β₁^fl/fl and β₁^AT2-KO lungs 21 days after LPS (D21), immunostained for AT2 marker pro–SP-C (red) and AT1 marker T1α (green). (B) Representative images of D21 β₁^fl/fl and β₁^AT2-KO lungs immunostained for pro–SP-C (red) with the AT1 marker Hopx (green), insets as in A. (C) Quantification of the number of pro–SP-C⁺ AT2 cells per field (n = 8 mice/group, 10 original magnification, 40× sections/mouse; P = 0.0086). (D) Hopx⁺ cells per field (n = 6 mice/group, 6 original magnification, 20×, sections/mouse; P = 0.0003). * P < 0.05. Scale bar = 100 μm for 10× in A and B; scale bar = 50 μm for 40× in A and B. Two-tailed t test was used to compare genotypes in C and D.

Figure 3. β₁^AT2-KO mice have increased AT2 proliferation during alveolar repair.

(A) Quantification of pro–SP-C⁺ cells per total cells (n = 6–8 mice/group; 5 sections per mouse; P = 0.0247 for uninjured (no LPS) mice; P = 0.8220 at D3; P = 0.0001 at D7; P = 0.0009 at D21). (B) Quantification of proliferating AT2 cells by percentage of total pro–SP-C⁺ AT2s (n = 6–8 mice/group, 10 sections/mouse; P = 0.0003 for uninjured mice; P = 0.0311 at D3; P = 0.0310 at D7; P = 0.0128 at D21). (C) Immunodetection of the proliferation marker Ki-67 (green) and pro–SP-C (red) shows peak AT2 proliferation in β₁^AT2-KO lungs at day 7 (arrows) as merged or single-channel panels. Two-tailed t test was used to compare genotypes at each time point for A and B. * P < 0.05. Scale bar = 200 μm for low-power image in C, 50 μm for inset in C.

Figure 4. AT2 proliferation is NF-κB dependent in LPS-treated β₁^AT2-KO lungs.

(A) Representative images of BrdU-incorporated precision-cut lung slices (PCLS) treated with LPS and/or NF-κB inhibitor BAY 11-7082 for 48 hours. Slices were immunostained for BrdU (cyan) and pro–SP-C (magenta) with DAPI nuclear marker (white). (B) Quantification of proliferating AT2 cells by percentage of total AT2 cells (BrdU⁺pro–SP-C⁺ over total number of SP-C⁺ cells) by condition as indicated (n = 6–8 mice/group, 1 slice per mouse per condition, imaged and quantified 10 original magnification, 40× sections/mouse per condition, data from 5 separate experiments; P = 0.0010, F value = 6.3 for treatment variation; P < 0.0001, F value = 26.1 for genotype variation). Two-way ANOVA was used to compare treatment conditions and genotype in B. * P < 0.05. Scale bar = 100 μm for low-power images in A, 50 μm for inset in A.

Figure 5. Overabundant AT2 cells are transcriptionally distinct during repair in β₁^AT2-KO mice.

(A) Uniform manifold approximation and projection (UMAP) of all epithelial cells from β₁^fl/fl and β₁^AT2-KO lungs with/without LPS clustered by label transfer from Strunz et al. (11). (B) Individual epithelial populations by group reveal transcriptionally distinct AT2s and activated AT2s in day 7 LPS-treated β₁^AT2-KO lungs. (C) Ingenuity Pathway Analysis (QIAGEN) on combined AT2 groups from uninjured β₁^fl/fl and β₁^AT2-KO lungs demonstrates upregulation of oxidative stress, senescence, and inflammatory pathways in β₁^AT2-KO lungs compared with β₁^fl/fl lungs. (D) Ingenuity Pathway Analysis shows upregulation of actin cytoskeleton signaling pathways in β₁^AT2-KO AT2 cells compared with β₁^fl/fl AT2 cells at 7 days after LPS treatment.

Figure 6. During alveolar repair, β₁ integrin regulates actin localization and RhoA GTPase activation.

(A) Surface rendering high-power images of pro–SP-C–immunostained, thick, frozen sections from day 7 (D7) LPS-treated β₁^fl/fl and β₁^AT2-KO lungs. (B) High-power images of thick, frozen sections from D7 LPS-treated β₁^fl/fl and β₁^AT2-KO lungs immunostained for pro–SP-C (green) with phalloidin F-actin probe (magenta); arrows indicate areas of actin-rich lateral protrusions. (C) Area of pro–SP-C⁺CD68^– AT2 cells from D7 LPS-treated β₁^fl/fl and β₁^AT2-KO mice (46.8 ± 2.0 μm² in β₁^fl/fl lungs compared with 69.6 ± 2.8 μm² in β₁^AT2-KO lungs, n = 6 mice/group, ≥40 cells measured/mouse imaged from 5 sections, 2-tailed t test, P < 0.0001). (D) Roundness score calculated from pro–SP-C⁺CD68^– cells from D7 LPS-treated β₁^fl/fl and β₁^AT2-KO mice (38–60 cells measured/mouse from 5 sections, n = 6 mice/group, 2-tailed t test comparing genotypes, P = 0.0009). (E) High-power images of frozen sections prepared at D7 after LPS β₁^fl/fl and β₁^AT2-KO lungs immunostained for pro–SP-C (gold) with JLA20 (cyan) and phalloidin (magenta) probes applied to detect G-actin and F-actin, respectively. Membrane localization of G-actin denoted by arrows and F-actin by arrowheads. (F and G) Quantification of JLA20 (F) and phalloidin (G) expression in pro–SP-C⁺ AT2 cells in β₁^fl/fl and β₁^AT2-KO lungs D7 after LPS (n = 5–6 mice/group, 10 sections/mouse, 2-tailed t test with P = 0.0088 for JLA20 and P = 0.0482 for phalloidin). (H) Representative high-power images from D7 LPS-treated β₁^fl/fl and β₁^AT2-KO lungs immunostained for ezrin (cyan) with AT2 cells identified by RNA in situ hybridization for Sftpc (gold). Arrows indicate ezrin expression localized to lateral extensions in β₁^fl/fl AT2 cells, whereas diffuse, nonfocal ezrin expression along the cell membrane is seen in β₁^AT2-KO AT2 cells. * P < 0.05. Scale bar = 5 μm for A, B, and E; scale bar = 25 μm for panels in H.

Figure 7. During repair, β₁ integrin regulates GTPase activation in AT2 cells.

(A–C) GTPase activation assay performed on AT2 cell lysates collected from uninjured and D7 LPS-treated β₁^fl/fl and β₁^AT2-KO lungs (n = 6–8 mice/group for each assay; in A RhoA 1-way ANOVA * P < 0.0001, F value 53.42, df = 3; in B Cdc42 1-way ANOVA * P < 0.001, F value 17.46, df = 3; in C Rac1 1-way ANOVA * P < 0.0001, F value 31.09, df = 3).

Figure 8. Postinjury β₁-deficient AT2s exhibit an AT2-AT1 mixed epithelial transcriptomic phenotype.

(A) Stacked bar graph of epithelial proportions demonstrates an expansion of the AT2 and activated AT2 populations in day 7 (D7) LPS-treated β₁^AT2-KO lungs. (B) Marker gene expression by genotype and treatment group in AT2/activated AT2 cluster, in which higher expression is represented with a darker color and the size of the dot reflects the proportion of cells expressing that marker. (C) Representative low-power images from day 7 LPS-treated lung sections immunostained for the AT1 marker AGER (purple) and AT2 marker pro–SP-C (gold) demonstrate overall decreased AGER in β₁^AT2-KO lungs. Three insets per low-power field show colocalization of AGER with pro–SP-C⁺ AT2 cells in D7 β₁^AT2-KO lungs. Scale bar = 100 μm for low-power image in C, 50 μm for middle inset, and 10 μm for high-power inset.

Figure 9. AT2s of mixed transcriptomic phenotype persist, proliferate, and maintain an enlarged, rounded cell shape in late alveolar repair.

(A) UMAP of all epithelial cells from β₁^fl/fl and β₁^AT2-KO lungs 21 days after LPS treatment clustered by label transfer from Strunz et al. (11). (B) Individual epithelial populations by group reveal transcriptionally abundant AT2s and activated AT2s in day 21 LPS-treated β₁^AT2-KO lungs. (C) Hallmark gene expression by genotype in AT2, activated AT2, and Krt8 ADI clusters, in which higher expression is represented with a darker color and the size of the dot reflects the proportion of cells expressing that marker. (D) Representative low-power images of β₁^fl/fl and β₁^AT2-KO lungs 21 days after LPS co-immunostained for pro–SP-C (red), cytokeratin 8 (gold), AGER (cyan), and CD68 (blue). β₁^AT2-KO lungs are notable for enlarged airspaces and increased numbers of round, large pro–SP-C⁺ AT2 cells, which are distinct from alveolar macrophages. (E) High-power images of lung sections immunostained for pro–SP-C, cytokeratin 8, AGER, and CD68, as above, demonstrate round, large pro–SP-C⁺ AT2 cells that colocalize with cytokeratin 8, with a subset also triple positive (pro–SP-C⁺cytokeratin 8⁺AGER⁺). Arrow denotes occasional triple-positive cells in β₁^AT2-KO lungs, and arrowhead marks AGER^– AT2 cells in β₁^fl/fl lungs. (F) Quantification of pro–SP-C⁺cytokeratin 8⁺AGER⁺ triple-positive cells as a percentage of total pro–SP-C⁺ cells in day 21 LPS-treated lungs (n = 6 mice/group, 10 original magnification, 60×, sections/mouse, P < 0.0001 by 2-tailed t test). * P < 0.05. Scale bar = 100 μm in D, 10 μm in E.

Figure 10. β₁^AT2-KO mice fail to repopulate alveolus with cells of AT1 morphology.

(A) Low-power images of Cre⁺ mTmG control mice demonstrate Cre-recombinase reporter–labeled GFP⁺ cells of both AT1 and AT2 cell shape repopulate the injured alveolus 21 days after LPS. GFP-labeled cells in β₁^AT2-KO mTmG mice retain only AT2 morphology and fail to attain an AT1 cell shape at 21 days. (B) Quantification of mean GFP fluorescence intensity shows decreased area of lung repopulated by GFP-labeled cells at 21 days post-LPS in β₁^AT2-KO mTmG mice (n = 4 Cre⁺ mTmG and 6 β₁^AT2-KO mTmG mice; 10 original magnification, 20×, sections/mouse, P = 0.0087 by 2-tailed t test). (C) High-power images of Cre⁺ mTmG and β₁^AT2-KO mTmG lung sections immunostained for pro–SP-C (red) and AGER (blue); tomato omitted in imaging. Large, round GFP⁺-labeled cells acquire both AT2 (pro–SP-C⁺) and AT1 (AGER⁺) markers in β₁^AT2-KO mTmG lungs. GFP⁺-labeled cells possess either AT2 or AT1 markers at 21 days post-LPS in Cre⁺ mTmG lungs. * P < 0.05. Scale bar = 200 μm in A and 25 μm in C.