Injury-induced pulmonary tuft cells are heterogenous, arise independent of key Type 2 cytokines, and are dispensable for dysplastic repair
Figures

Epcam+ Trpm5-GFP+ cells are bona fide tuft cells in the lung post-influenza.
(A) Representative immunostaining of lung sections from Trpm5-GFP reporter mice at day 35 post-influenza. Nuclear stain (DAPI) in blue, Trpm5-GFP in green, and POU2F3 in red. (B) Experimental design outlining the bulk RNA sequencing (RNA-Seq) experiment. (C) Trpm5-GFP reporter expression in live lung epithelial (Epcam+) cells post-influenza via FACS. (D) Volcano plot and (E) heatmap comparing gene expression between Epcam+Trpm5-GFP+ (tuft cells) and Epcam+Trpm5-GFP- (non-tuft epithelial) cells from mice at day 43 post influenza.*=Representative genes that have been previously associated with tuft cells.

Gene expression comparison of post-influenza A virus (IAV) lung tuft cells and tracheal cells (Nadjsombati et al., 2018) using bulk RNA sequencing data.
(A) Volcano plot showing differentially expressed genes between post-IAV lung tuft cells and tracheal tuft cells. (B) Heatmap of expression levels of canonical tuft cell markers. (C) Heatmap comparing additional shared/core tuft cell markers as well as Tuft-1 and Tuft-2 markers, indicating a reduced Tuft-1 character among trachea tuft cells.

Distinct tuft cell populations derived from Krt5+ cells arise in the lung following influenza clearance.
(A) Single-cell RNA-Seq UMAP clustering of sorted tuft cells (Epcam+Trpm5-GFP+) from Trpm5-GFP reporter mice at day 28 post influenza. (B) Heatmap comparing the gene expression profile between the tuft cell clusters identified in A. (C) Trpm5 expression highlighted in all analyzed cells, confirming all analyzed cells are tuft cells and not contaminating cells. (D) Marker gene expression highlighted within the different tuft cell UMAP clusters. (E) Representative immunostaining of tamoxifen-treated Krt5-CreER tdTomato lungs 25 days post influenza. Nuclear stain (DAPI) in blue, DCKL1 in green, Krt5-CreER tdTomato lineage label in red. Lineage traced tuft cells (tdTomato+DCKL1+) appear yellow. (F) Representative immunostaining for Gnb3 (Tuft-1 signature) in the Trpm5-GFP reporter mice 35 days post influenza. Nuclear stain in blue, Gnb3+ cells in red, Trpm5-GFP+ cells in green, and double positive cells (Gnb3+ and Trpm5+) in yellow. Single color insets shown (i, ii).

Further analysis of the different tuft cell clusters utilizing previously identified transcriptomic signatures.
(A) Published gene set associated with Tuft-1, Tuft-2, and basal cell signatures. (B) UMAP clustering as in Figure 2A, shown again for reference/comparison for tuft cell signature module enrichment in C. (C) Enrichment score for each gene module from A is highlighted on the UMAP tuft cell clusters identified in the single-cell RNA sequencing (RNA-Seq) from the Trpm5-GFP reporter mice at day 28 post influenza (Figure 2).

Trajectory analysis of influenza A virus (IAV)-induced lung tuft cells from single-cell data.
(A) Clusters assigned to UMAP of single-cell RNA sequencing of post-IAV lung tuft cells (also see Figure 2). (B) Slingshot minimum spanning tree (MST) trajectory. (C) Slingshot trajectory with pseudotime color bar, demonstrating the trajectory begins at the ‘basal ->tuft’ cluster and ends at the ‘stressed’ tuft cluster. (D) Trajectory analysis using Monocle3 with pseudotime color bar.

Tuft cells are not required for the epithelial dysplastic response after lung injury.
(A–P) Lung sections stained for Krt5 in green, Dclk1 in red, and DAPI in blue 22–25 days after influenza. (A–D) Dclk1+ cells are absent in Pou2f3-/- (n=8) compared to control (n=11), without a significant change in Krt5+ area. (E–P) No significant difference was found in the number of Dclk1+ cells per Krt5+ area or percent Krt5+ lung area when comparing (E–H) control (n=5) and Trpm5-/- (n=7), (I–L) control (n=4) and Il4ra-/- (n=3), or (M–P) wild type (WT) (n=5) and Il25-/- (n=3). Dclk1+ cells per Krt5+ area was quantified in (C, G, K, O) and was not statistically significantly different in (G, K, O).Changes in percent Krt5+ area were also not statistically significant in (D, H, L, P). (A–L) Analysis 25 dpi, (M–P) analysis 22 dpi. (A–B, E–F, I–J, M–N) Scale bar is 100 µm, images are cropped from a multipanel stitched image.
-
Figure 3—source data 1
Tuft cells per Krt5 area and percent Krt5 area in controls and Pou2f3-/-, Trpm5-/-, Il4ra-/-, and Il25-/-.
- https://cdn.elifesciences.org/articles/78074/elife-78074-fig3-data1-v2.xlsx

Examples of Pou2f3 control and mutant Krt5+ regions.
(A) Multipanel stitched image of wild type (WT) and Pou2f3-/- lung sections. (B) Examples of Krt5+ areas (green) in control and Pou2f3-/- lung sections stained with Dclk1 (red) (C) No significant difference in weight loss at 7, 9, or 16 days post infection between control and Pou2f3-/- animals. (A) Scale bar is 1 mm, (B) scale bar is 100 µm. Each dot represents an individual mouse.
-
Figure 3—figure supplement 1—source data 1
Percent body weight loss of control and Pou2f3-/- mice 7, 9, and 16 days post influenza infection.
- https://cdn.elifesciences.org/articles/78074/elife-78074-fig3-figsupp1-data1-v2.xlsx

Examples of Trpm5 and Il4ra control and mutant Krt5+ regions.
(A–B) Multipanel stitched image of wild-type (WT) and (A) Trpm5-/- and (B) Il4ra -/- lung sections. (C–D) Examples of control and (C) Trpm5-/- Krt5+ areas (green) and (D) Il4ra-/- Krt5+ areas, also stained with Dclk1 (red). (A–B) Scale bar is 1 mm, (C–D) scale bar is 100 µm.

Analysis of tuft cell density and Krt5+ area.
(A) Experiments on Pou2f3-/- and Il4ra-/- were performed independently of those shown in Figure 3. (B–C) Quantification of the data shown in Figure 3A–L using manual quantification of tuft cell density and Krt5+ area of (B) Trpm5 control and mutant lung sections and (C) Il4ra control and mutant lung sections. (D) Examples of binary masks rendered for quantification of lung sections for data shown in Figure 3A–L. Binaries include tissue area which excludes porous area, Dclk1 cell count and Krt5+ area. (D) Scale bar is 1 mm. Each dot represents an individual mouse.
-
Figure 3—figure supplement 3—source data 1
Tuft cells per Krt5 area and percent Krt5 area.
- https://cdn.elifesciences.org/articles/78074/elife-78074-fig3-figsupp3-data1-v2.xlsx

Tuft cell differentiation does not depend on IL-4ra signaling.
(A) Weight loss of controls (Il4rafl/fl, n=3) and conditional Il4ra knockouts (Il4rafl/fl; R26-ERT2-Cre, n=5). (B) Il4ra expression was significantly reduced in conditional Il4ra knockouts compared with controls by qPCR. For these experiments, both groups received tamoxifen (2 mg by i.p. injection), every other day from D7 to D14 post infection. (C) Lung sections stained with Krt5 in green and Dclk1 in red 25 days after PR8 infection. Scale bars are 100 µm. (D) No significant difference was found between the number of Dclk1+ cells per Krt5+ area in lung sections of control and conditional Il4ra knockouts. (B,D) Each dot represents an individual mouse.
-
Figure 3—figure supplement 4—source data 1
Analysis of Il4ra conditional knockout mice.
- https://cdn.elifesciences.org/articles/78074/elife-78074-fig3-figsupp4-data1-v2.xlsx

Tuft cell differentiation does not impact basal-like cell proliferation.
(A) qPCR time course of Type 2 cytokines Il4, Il5, Il13 at 0, 3, 8, 12, 21, and 51 days post PR8 infection. (B) Lung viral titer from control (Pou2f3+/-) and Pou2f3-/- lungs at 8 and 12 days post infection determined by TCID50/hemagglutination assay. There was no statistically significant difference between control and Pou2f3-/- lung titers at either timepoint. (C) Krt5 staining in green and Ki67 staining in red in control (Pou2f3+/-) and Pou2f3-/- 14 days post infection and (D) 25 days post infection. The percent Ki67+ alveolar basal-like cells was not significantly different between controls and Pou2f3-/-. (C–D) Scale bar is 10 µm. Each dot represents an individual mouse.
-
Figure 3—figure supplement 5—source data 1
Time course of Type 2 cytokine expression, lung viral titer, and analysis of basal-like cell proliferation.
- https://cdn.elifesciences.org/articles/78074/elife-78074-fig3-figsupp5-data1-v2.xlsx

Type I and type III interferon signaling are dispensable for tuft cell development after influenza infection.
(A–B, D–E) Lung sections stained for Krt5 in green, Dclk1 in red, DAPI in blue 22 days after influenza infection. (A–F) No significant differences were found in the Dclk1+ cells number per Krt5+ area when comparing (A–C) wild type (WT) (n=5) and Il28r-/- (n=6) mice and (D–F) WT (n=4) and Ifnar-/- (n=3). (A–C) and (D–F) are each pooled from two independent experiments. (G) Representative image of a lung section from a WT mouse (n=3) treated with bleomycin stained for Dclk1 in red, Krt5 in green, and DAPI in blue, 22 days following injury. (A–B, D–E, G) Scale bar is 100 µm, images are cropped from the 20x z-stack image. Error bars represent standard deviation.
-
Figure 4—source data 1
Tuft cells per Krt5 area in controls and Il28r-/- and Ifnar1-/-.
- https://cdn.elifesciences.org/articles/78074/elife-78074-fig4-data1-v2.xlsx

Interferon receptor-deficient mice have comparable weight loss to BL/6 mice when infected with a lower PR8 dose.
(A) Weight loss of Il28r-/- and wild-type (WT) control mice following PR8 infection (n=3). Representative of two independent experiments. (B) Weight loss of Ifnar-/- and WT control mice following PR8 infection (n=2–3). Representative of two independent experiments. Error bars represent standard error of the mean.
-
Figure 4—figure supplement 1—source data 1
Weight loss post influenza A virus (IAV) infection of controls and Il28r-/- and Ifnar1-/-.
- https://cdn.elifesciences.org/articles/78074/elife-78074-fig4-figsupp1-data1-v2.xlsx

Tuft cells are not required for goblet cell differentiation after influenza.
(A–I) Krt5 (green) and Agr2 (red) staining and quantification 25 days after influenza. Agr2+ area per Krt5+ area was not significantly different between lung sections of (C) control (n=8) and Pou2f3-/- (n=8), (F) control (n=4) and Trpm5-/- (n=6) and (I) control (n=3) and Il4ra-/- (n=3). (J–M) Krt5 (green) and Muc5b (red) staining 25 days after influenza demonstrates Muc5b staining in (J–K) control (n=4) and Pou2f3-/- (n=3) and (L–M) control (n=4) and Il4ra-/- (n=3) dysplastic alveolar regions and (N) qRT-PCR (quantitative RT-PCR) for relative mRNA levels for goblet cell markers in control (n=8) and Pou2f3-/- (n=4) lungs 51 days after influenza, expression normalized to Actb (left) and Krt5 (right). (A–H) Scale bar is 500 µm, (J–M) scale bar is 200 µm.
-
Figure 5—source data 1
Quantification of Agr2 immunostaining and expression of goblet cell transcripts.
- https://cdn.elifesciences.org/articles/78074/elife-78074-fig5-data1-v2.xlsx

Tuft cells do not affect Krt5 plasticity following influenza.
(A–B) Krt5 (green) and SPC (orange) staining in control (n=3) and Pou2f3-/- (n=3) lung sections demonstrates no appreciable overlap between Krt5 and SPC areas. (B) Krt5-creERT2; Ai14; Pou2f3+/- or Pou2f3-/- lung sections were injected with tamoxifen 5, 10, and 15 days post infection and lungs were harvested 30 days post infection. tdTomato signal was not found in SPC+ cells in control (n=3) or Pou2f3-/- (n=3). (A–B) Scale bar is 1 mm, (C) scale bar is 10 µm, (D) scale bar is 50 µm.