Triggering AHR resolves TGF-β1 induced fibroblast activation and promotes AT1 cell regeneration in alveolar organoids

Abstract

Regeneration of the alveolar epithelium is necessary to restore tissue architecture and gas exchange capabilities in chronic pulmonary diseases such as fibrosing interstitial lung disease. While it is known alveolar type 2 (AT2) cells give rise to alveolar type 1 (AT1) cells to repair the alveolar epithelium after injury, methods to promote this process under pathological settings are poorly understood. Here, using a complex 3D organoid culture with TGF-β1 dependent impaired AT1 spheroid formation, we performed a high-throughput screen (HTS) with ~16,800 compounds to identify small molecules that increase number of AT1 spheroids. Longitudinal single cell RNA sequencing (scRNA-seq) revealed that DB-11-BE87 increased AT1 regeneration by reducing TGF-β1 induced fibroblast activation, concurrently with AHR activation in those cells. These studies highlight a HTS system to identify factors that can promote AT1 differentiation and suggest AHR activation as a method to counteract pathological TGF-β1 signaling in pulmonary disease.

Fig. 1. High-throughput screening on multi-cellular murine alveolosphere assay identifies small molecules that rescue TGFβ impaired alveolar type 1 spheroid formation.

A Schematic of assay. Whole distal lung cells from Hopx-GFP mice were plated in 384-well plates, cultured in Matrigel for ~12 days to promote spheroid formation, and imaged live for GFP signal to quantify alveolar type 1 spheroids. B Representative images of live alveolospheres from C57BL/6 control or Hopx-GFP mice treated at time of plating with increasing doses of TGF-β1 with or without 30 µM of TGF-βR1 inhibitor galunisertib. Scale = 500 µM. C Quantification of (B). n = 10-12, individual data points plotted. One-way ANOVA, P = 5.9 × 10⁻⁵. **α < 0.01, Tukey’s HSD (Hopx-GFP mouse; TGF-β1 0.015 ng/mL vs. 2 ng/mL). D Quantification of canonical AT1 markers from alveolospheres described in (B). n = 5–12, individual data points plotted. E Quantification of GFP spheroids vs. GFP total intensity for controls or small molecule library. Red = positive control (30 µM galunisertib). Blue = negative control (no compound addition). Green = called compounds with biological activity. Black = compound without called biological activity. Star = LA-13-ZS70; primary hit for followup. Dotted line indicates Median +3 SD.

Fig. 2. DB-11-BE87 increases AT1-AT2 differentiation with partial dependency on TGFB pathway activation.

A Representative images of live alveolospheres treated at time of plating with increasing doses of LA-13-ZS70. Scale = 500 µM. B Quantification of concentration-response curve for LA-13-ZS70 on GFP adjusted count spheroids. EC₅₀ = 376 nM, R² = 0.81. n = 4 for all concentration except untreated well, n = 96 for untreated well. Individual data points plotted. C Chemical structure of LA-13-ZS70. D Representative images of live alveolospheres treated at time of plating with increasing doses of DB-11-BE87. Scale = 500 µM. E Quantification of concentration-response curve for DB-11-BE87 on GFP adjusted count spheroids. EC₅₀ = 151 nM, R² = 0.50. n = 4 for all concentration except untreated well, n = 96 for untreated well. Individual data points plotted. F Chemical structure of DB-11-BE87. G RASL-seq of alveolospheres treated at time of plating with 10 µM LA-13-ZS70. Dotted line indicated ± gene fold change of 2 and p-adjusted value of 0.05 compared to DMSO control. Purple = AT1 genes; Blue = AT2 genes; Green = Airway genes; Yellow = fibrosis genes; Red =  Immune genes; Gray = other genes. n = 4. H RASL-seq of alveolospheres treated at time of plating with 10 µM DB-11-BE87. Dotted line indicated ± gene fold change of 2 and p-adjusted value of 0.05 compared to DMSO control. Purple = AT1 genes; Blue = AT2 genes; Green = Airway genes; Yellow = fibrosis genes; Red = Immune genes; Gray = other genes. n = 4. I Quantification of concentration-response curve for DB-11-BE87 on Hopx expression. EC₅₀ = 62 nM, R² = 0.69. n = 3–5 for all concentration except untreated well, n = 129 for untreated well. Individual data points plotted. J Quantification of concentration-response curve for DB-11-BE87 on GFP adjusted count spheroids in culture with or without exogenous TGF-β1. + TGF-β1 EC₅₀ = 10.5 pM, R² = 0.70. − TGF-β1 EC₅₀ = 6.3 pM, R² = 0.65. n = 6 for all concentration except untreated well, n = 24 for untreated wells. Individual data points plotted.

Fig. 3. Alveolosphere scRNA-seq detects Fn1+ aberrant epithelial cells.

A: Single cell suspension from alveolosphere assay were analyzed using scRNA-seq at indicated timepoints post assay setup. Untreated control (Day 0) and three treatment groups (DMSO, 10 µM DB-11-BE87, 10 µM QA-92-TQ17) were included to understand DB-11-BE87’s influence on TGF-β1 treated alveolosphere culture. 4 cell categories containing 117, 450 integrated cells from all samples are colored on UMAP. B Cell proportions of major cell categories for the indicated timepoint and compound treatment in the alveolosphere assay. C UMAP for all epithelial cells integrated from all conditions. 8 cell types and states were identified. D Expression of canonical epithelial cell type markers. E RNA velocity vectors projected over UMAP plot of AT1, AT2, Krt8 + ADI, and Transitional AT2 epithelial cell types, from all samples. F Expression of individual genes among cells from (E). G Degree of similarity between indicated cell types/states from patient derived samples (x-axis) and alveolosphere culture (y-axis). Color code indicates scaled spearman correlation coefficient (z-score) of two transcriptomes. H Fraction of Krt8 + ADI cells that expressed Fn1, across individual samples collected on day 11. I Violin plot of Fn1 expression in Krt8 + ADI cells from samples in three treatment groups.

Fig. 4. TGFβ induces fibroblast activation in alveolosphere model.

A UMAP plots of mesenchymal cell clusters on day 3 (top) and cell cycle phases (bottom, G1: green, G2M: pink, M: purple). Plot contains all 3 treatment groups (DMSO, 10 µM DB-11-BE87, 10 µM QA-92-TQ17). B Expression of marker genes in mesenchymal cell clusters from (A). C UMAP plot of integrated mesenchymal cells from day 1 (red) and day 3 (teal). D UMAP plot of mesenchymal subclusters from (C) colored by cell type labels. E RNA velocity vectors projected over UMAP plot of mesenchymal cells from day 3 samples. Cells are colored by cell type labels. F UMAP plots of mesenchymal cells from day 3 samples, colored by either gene expression (Cthrc1) or pathway activity scores (TGF-β pathway, WNT signaling, and ECM pathway).

Fig. 5. DB-11-BE87 activates AHR and resolves TGF-β1 induced fibrosis in the alveolosphere model.

A UMAP plot of integrated mesenchymal cells from day 3 (teal) and day 11 (red). Plot contains all 3 treatment groups (DMSO, 10 µM DB-11-BE87, 10 µM QA-92-TQ17). B RNA velocity vectors projected over UMAP of integrated mesenchymal cells from day 3 and day 11 samples. Cell clusters are numbered. C Expression of marker genes in mesenchymal cell clusters from (B). D UMAP plots of mesenchymal cells on day 3 (left) and day 11 (right), colored by cell type labels. E UMAP plots of mesenchymal cells colored by either gene expression (Lpl) or pathway activity scores (adipogenesis, muscle contraction, and inflammation), from day 11 samples. F Cell proportions of mesenchymal cell clusters from (B), in day 11 samples separated by compound treatment and replicate. ** indicates p < 0.01 in Student’s t test.

Fig. 6. AHR agonists rescue TGF-β1 impaired alveolar type 1 spheroid formation while AHR antagonists do not.

A Correlation between Ahr expression and AHR pathway activity across mesenchymal cell clusters from day 1, day 3, and day 11 samples separated by compound treatment. Y-axis indicates average expression of AHR pathway genes relative to random background genes in log scale. B Chemical structure of AHR agonists and antagonists. C Quantification of concentration-response curve for AHR agonists and antagonists on normalized luciferase luminescence in AHR responsive reporter assay. TE-33-UQ34 EC₅₀ = 21 nM, R² = 0.90. EB-06-KR05 EC₅₀ = 6.7 nM, R² = 0.79. QB-43-PX71 EC₅₀ = 60 nM, R² = 0.93. GD-70-TQ68 EC₅₀ = 89 nM, R² = 0.98. DC-78-LZ94 EC₅₀ = 101 nM, R² = 0.90. Individual data points plotted. D Quantification of concentration-response curve for AHR agonists and antagonists on GFP adjusted count spheroids in murine alveolosphere Hopx-GFP+ spheroid assay. TE-33-UQ34 EC₅₀ = 70.2 nM, R² = 0.77. EB-06-KR05 EC₅₀ = 196 nM, R² = 0.21. QB-43-PX71 EC₅₀ = 2.1 μM, R² = 0.59. GD-70-TQ68 EC₅₀ = 16.2 μM, R² = 0.29. DC-78-LZ94 EC₅₀ = 2.5 μM, R² = 0.51. n = 2 for all concentration except untreated well, n = 48 for untreated wells. Some untreated wells are replicated between compound curves. Individual data points plotted.

Fig. 7. DB-11-BE87 induces hypoxia mediated glycolysis in AT2 on day 11 via IL-1β signaling.

A RNA velocity vectors projected over UMAP of macrophage subclusters from day 11 samples. B UMAP plots of macrophage cells colored by Il1b expression (purple) from day 11 samples split by compound treatment. C Violin plot of Il1b expression from day 11 samples split by compound treatment. D UMAP plots of epithelial cells from day 11 samples, colored by pathway expression scores (hypoxia, top; glycolysis, bottom). Samples split by compound treatment. E Circle plots of cell-cell communications going into interstitial macrophages on day 3 (top) and day 11 (bottom). Thickness of line is proportional to overall communication probability. F Relative expression of Ahr and Ahr pathway genes in mesenchymal cell clusters from DB-11-BE87-treated samples on day 3 (left) and day 11 (right).