Dual SMAD Signaling Inhibition Enables Long-Term Expansion of Diverse Epithelial Basal Cells

Abstract

Functional modeling of many adult epithelia is limited by the difficulty in maintaining relevant stem cell populations in culture. Here, we show that dual inhibition of SMAD signaling pathways enables robust expansion of primary epithelial basal cell populations. We find that TGFβ/BMP/SMAD pathway signaling is strongly activated in luminal and suprabasal cells of several epithelia, but suppressed in p63+ basal cells. In airway epithelium, SMAD signaling promotes differentiation, and its inhibition leads to stem cell hyperplasia. Using dual SMAD signaling inhibition in a feeder-free culture system, we have been able to expand airway basal stem cells from multiple species. Expanded cells can produce functional airway epithelium physiologically responsive to clinically relevant drugs, such as CFTR modulators. This approach is effective for the clonal expansion of single human cells and for basal cell populations from epithelial tissues from all three germ layers and therefore may be broadly applicable for modeling of epithelia.

Keywords: TGFβ/BMP4/SMAD signaling; dedifferentiation; differentiation; dual SMAD signaling inhibition; epithelial basal and stems cells; p63(+) basal cells; replicative exhaustion; senescence; stemness; telomeres.

Figure 1. SMAD signaling is active in differentiated cells but suppressed in basal stem cells of the airway epithelium

A and B. Co-staining of p-SMAD1/5/8 (A) or p-SMAD2/3 (B) with basal stem cell marker p63, differentiation maker KRT8, ciliated cell marker FOXJ1, club cell marker SSEA1 and proliferation marker Ki67 on mouse tracheal sections. C-D. Co-staining of p-SMAD1/5/8 with KRT5 and KRT8 (C) and with p63, FOXJ1, and Ki67 (D) on human bronchial sections. E. Co-staining of p-SMAD2/3 with p63, KRT8, FOXJ1, and Ki67 on human bronchial sections. For all, scale bar, 20 µm. F and G. Quantification of the percentage of p-SMAD1/5/8⁺ and p-SMAD2/3⁺ cells of the indicated cell types in mouse trachea (F) and human bronchus (G), n=3.

Figure 2. SMAD signaling is active in luminal and suprabasal cells, but suppressed in basal cells, in epithelial tissues derived from all 3 germ layers

A. Schematic of the analyzed p63⁺ basal cell-containing murine tissues derived from each germ layer. B-G. Co-staining of p-SMAD1/5/8 and p-SMAD2/3 with p63, Ki67, and a luminal and suprabasal marker KRT10 (B) or KRT13 (D and E) or KRT8 (C, F and G). Scale bar, 20 µm.

Figure 3. Activation of SMAD signaling is associated with the progressive mucociliary differentiation

A. C57B6 mice were exposed to SO₂. At 8, 24, 48 and 72 hours post SO₂ injury, the mice were sacrificed and the trachea were collected for co-staining of p-SMAD1/5/8 with p63 and KRT8. The dashed line separates the sloughed luminal cell debris from the attached residual basal cell layer. B. Schematic of SMAD signaling activation changes along the course of mucocilary differentiation. C. Tracheal stem cells isolated from wild type, BMPRIA^f/f, TGFβRII^f/f and SMAD4^f/f mice were treated with Ad-Cre-GFP virus. The GFP⁺ cells were sorted to purity 1–2 days post infection. The cells were expanded briefly and then subjected to ALI differentiation. Left graph: quantification of ciliated cells and club cells in various conditions (mean ± s.d. n=3; **p ≤ 0.001). Right panel: whole mount staining of CCSP⁺ club cells and AcTub⁺ ciliated cells on ALI membranes. D. Immunostaining for KRT5, p63 and KRT8 of transverse sections of ALI membranes. The arrows indicate representative transitional differentiating cells. E. Tracheal stem cells were isolated from WT mice and differentiated on ALI for 14 days. At day 0, 1 µM DMH-1 and/or 1 µM A-83–01 were added to the lower chamber and upper chamber (as a thin layer) to inhibit the corresponding SMAD signaling activity. Left graph: quantification of ciliated cells and club cells in various conditions (mean ± s.d. n=5; **p ≤ 0.001, *p ≤ 0.01). Right panel: whole mount staining of CCSP⁺ club cells and AcTub⁺ ciliated cells on ALI membranes. F. Immunostaining for indicated markers of transverse sections of ALI membranes. The arrows indicate representative transitional differentiating cells. For all, scale bar, 20 µm. (See also Figure S1)

Figure 4. Dual SMAD signaling inhibition facilitates long-term expansion of human airway stem cells

A. Human airway stem cells were expanded for 4 days in varying culture conditions. The cells were fixed and stained for p63, KRT5 and Ki67. Scale bar: upper two rows, 50 µm; lower two rows, 20 µm. B. Serial expansion of human airway stem cells in the control medium or the same medium with addition of TGFβ/BMP inhibitors. Left panel: phase contrast images of the cells at various passages. Right panel: the average population doubling times (mean ± s.d. n=10; ***p ≤ 0.0001) and a plot of number of doublings vs. the culture times of the human airway stem cells cultured in various media (from one representative culture). C. Mouse tracheal stem cells were expanded for 4 days in varying culture conditions. The cells were fixed and stained for p63/KRT5 and Ki67/SOX2. Scale bar: upper panel, 100 µm, bottom panel, 50 µm. D. Serial expansion of mouse tracheal stem cells in the control medium or the same medium with addition of TGFβ/BMP inhibitors. Top panel: phase contrast images of the mouse tracheal stem cells at various passages. Bottom panel: Left, the plot of number of doublings vs. the culture times of mouse tracheal stem cells cultured in various conditions (from one representative culture). Right, the staining of cell fate markers on mouse tracheal stem cells at passage 20. Scale bar, 20 µm. (See also Figure S2 and S5)

Figure 5. Expanded airway stem cells preserve the potential to differentiate into functional airway epithelia

A. Expanded human airway stem cells at different passages were differentiated on ALI for 16–20 days. The resulting membranes were fixed and stained for CCSP⁺ club cells and AcTub⁺ ciliated cells. Scale bar, 50 µm. B. Mucociliary differentiation efficiency at various passages (normalized to the efficiency of stem cell differentiation at P1 with TGFβ/BMP inhibitors) (mean ± s.d. n=3; **p≤0.001, *p ≤ 0.01). C. Immunofluorescence staining for markers of differentiation using wholemount (Scale bar, 100 µm) and histology sections (Scale bar, 20 µm) of P7 ALI cultures. Bottom, H&E staining of an ALI section (Scale bar, 50 µm). D. Staining for MUC5AC on human ALI cultures after treatment with IL-13. Scale bar, 50 µm. E. Area under the curve (AUC) ratios for CFTR correctors (Cmpd/Veh) measured on ΔF508/ΔF508 human bronchial ALI cultures (P4), (mean ± s.d. n=4; *p≤0.01). F. Chloride current density gradually declines with increased passages. Up to passage 8, the response to C18 remains at ~3-fold the vehicle response; at P10 and P12 with the decrease in currents, compound effects can no longer be resolved (mean ± s.d. n=3; *p ≤0.01). G. The expansion and differentiation of airway stem cells isolated from human tracheal biopsies (n = 30). (See also Figure S3, S4, and S5)

Figure 6. Dual SMAD signaling inhibition enables long-term expansion of functional keratinocytes

A. Doubling time of adult human keratinocytes cultured in four different conditions: (i) Standard media, (ii) Standard media + J2 feeder cells, (iii) Inhibitor media, (iv) Inhibitor media + J2 feeder cells. (mean ± s.d. n=3; **p≤0.001). B. Phase contrast images and marker staining of human keratinocytes at indicated passages expanded in a feeder-free mode in the medium containing TGFβ/BMP inhibitors. Scale bar: 100 µm. C. Section immunofluorescence of various basal and differentiation markers on ALI culture (P4 and P12) and on human skin. Scale bar: 20 µm. D. Isolated mouse keratinocytes were seeded at 3% cell density and expanded for 4 days in various culture conditions. The cells were then fixed and stained for p63, KRT5, and Ki67. Scale bar: 100 µm. E. Phase contrast images of mouse skin keratinocytes at indicated passages. Scale bar: 50 µm. F. Differentiation of expanded mouse skin keratinocytes (P10) on ALI. The derived stratified epithelia were positive for KRT10 and involucrin.

Figure 7. Dual SMAD signaling inhibition enables in vitro expansion of a diverse set of organ-specific basal cells

A. Phase contrast images and marker staining of mouse esophageal basal cells at indicated passages expanded in the medium containing TGFβ/BMP4 inhibitors. Scale bar: 100 µm (Phase contrast images) & 50 µm (immunofluorescence). B-C. H&E (B) and immunofluorescence (B & C) of ALI culture sections derived from esophageal basal cells (P12) and of mouse esophageal. Scale bar: 20 µm. D. Schematic and example of organoid differentiation of expanded mouse epididymis basal cells isolated from KRT5-tdTomato-B1EGFP mouse. E. Phase contrast images and immunofluorescence of mouse epididymis basal cells at indicated passages expanded with TGFβ/BMP4 inhibitors. Scale bar: 100 µm. F. Fluorescence images of tdTomato and EGFP in differentiated organoids from D (upper panels are whole mount 3D organoids (Scale bar: 100 µm) and bottom panels are sections of the organoids (Scale bar: 20 µm). G. Co-immunofluorescence of indicated markers with tdTomato on epididymis organoid sections. Scale bar: 20 µm. (See also Figure S6 and S7)