H1N1 infection causes severe damage to mouse lung.

(A) Schematic of H1N1 intranasal inoculation procedure. (B) Bodyweight changes of uninfected control and H1N1-infected mice over 25 days post infection. Data are presented as means ± SD (n=6). Unpaired two-tailed student t tests were performed. (C) Representative images of lungs at 0 and 25 dpi, and their H&E tissue sections (Bars: 5 mm in the whole lung images and 100 µm in the lung section images). (D) Immunostaining images of lung sections of 0 and 25 dpi with an antibody to CD45. (E) Immunostaining of the ChAT-Cre: Ai9 lung sections of 0 and 25 dpi with the antibody to the tuft cell marker Dclk1 (green), and ChAT (Ai9, red tdTomato). (F) Immunostaining of the ChAT-Cre: Ai9 lung sections of 25 dpi with antibodies to Dclk1 (green), Gα-gustducin (green) or Plcβ2 (green) and ChAT (Ai9, red tdTomato). Scale bars in (D, E, F): 50 µm. ***P<0.001.

Figure 1—figure supplement 1. Chemically induced severe injury also engenders ectopic formation of lung tuft cells.

Expression and functional responses of bitter taste receptors.

(A) Quantitative RT-PCR analysis of bitter taste receptor expression in the lung tissues at 0 dpi versus 25 dpi. Expression of Tas2r105, Tas2r108, Tas2r118, Tas2r137 and Tas2r138 was upregulated whereas Tas2r135 and Tas2r143 were downregulated at 25 dpi. Data are presented as means ± SD (n=3). Unpaired two-tailed student t tests were performed. (B-K) Calcium imaging and response traces of FACs-sorted tuft cells isolated from H1N1-infected lungs at 25 dpi. (B, D, F, H, and J) Left panels are grey images of tuft cells at rest in the HBSS buffer whereas right panels are grey images of the same cells after being stimulated by denatonium benzoate (D.B.) or quinine. Activated cells appeared brighter. (C, E, G, I, K) traces of calcium responses of tuft cells on the left. Some tuft cells responded to D.B, which was inhibited by AITC (B, C), gallein (F, G), or U73122 (J, K) whereas others responded to quinine, which was also inhibited by AITC (D, E) or gallein (H, I). Scale bar: 50 µm. **P<0.01, ***P<0.001.

Figure 2—figure supplement 1. Expression of taste2 receptors in tuft cell clusters.

Figure 2—figure supplement 2. FACs sorting of ectopic tuft cells.

H1N1 infection inflicts differential lung injury and tuft cell dysplasia on WT versus mutant mice.

(A) Gng13-cKO mice showed Figure 3 continued on next page greater bodyweight loss than WT or Trpm5−∕− mice over days post infection. Data are presented as means ± SD (n=6). Unpaired two-tailed ttests were performed. (B) Kaplan-Meier survival curves of WT, Trpm5−∕− and Gng13-cKO mice following H1N1 inoculation. A significant number of Gng13-cKO mice died between 9 and 12 dpi (i.e., 2, 4, 2 mice and 1 mouse died at 9, 10, 11, 12 dpi, respectively), reducing the overall survival rate to 62.5%, significantly lower than those of Trpm5−∕− or WT mice, of which nearly all survived. Data are presented as means ± SEM (n= 44, 17 and 14 for WT, Trpm5−∕− and Gng13-cKO mice, respectively). The curves were determined by a log-rank test. (C, D) Images and statistical analysis of the injured areas on the lungs of WT, Trpm5−∕− and Gng13-cKO mice at 25 dpi. The injured areas that are marked by dashed lines were significantly greater in the Gng13-cKO mice than in Trpm5−∕− or WT mice while no significant difference was found between the latter two. Data are presented as means ± SD (n=5), and unpaired two-tailed student t tests were performed. (E) Comparative analysis of wet to dry weight ratios of WT and Gng13-cKO lungs at 0 and 25 dpi. While no difference in the ratio was found between WT and Gng13-cKO mice at 0 dpi, the ratios of these mice at 25 dpi were significantly greater than their corresponding ones at 0 dpi. However, the Gng13-cKO ratio was even greater than WT ratio at 25 dpi. Data are presented as means ± SD (n=6), and unpaired two-tailed student t tests were performed. (F-M) Identification of tuft cells in the Krt5-expressing tissue areas of H1N1-injured lungs of WT and Gng13-cKO mice at 14, 20, 25 and 50 dpi using antibodies to Krt5 and to the tuft cell marker Dclk1. The densities of tuft cells were significantly higher in WT than in Gng13-cKO lungs at 20, 25 and 50 dpi, but not at an earlier time point of 14 dpi. Data are presented as means ± SD (n=3), and unpaired two-tailed student t tests were performed. *P<0.05, **P<0.01, ***P<0.001, ****P<0.0001.

Figure 3—figure supplement 1. Statistical analysis of the injured tissue volumes on the lungs of WT, Trpm5−∕− and Gng13-cKO mice at 25 dpi.

Figure 3—figure supplement 2. Ectopic generation of tuft cells in WT, Trpm5−∕− and Gng13-cKO mice over days post H1N1 infection.

Stronger inflammatory response and upregulated expression of pyroptotic genes in the H1N1-infected Gng13-cKO lungs.

(A, B) Immunostaining of WT and Gng13-cKO lung tissue sections with an antibody to the immune cell marker CD64 indicates significantly more immune cells in the mutant lung than in WT at both 20 and 25 dpi. (C) qRT-PCR indicates that expression levels of gasdermin D and E (Gsdmd, Gsdme) and other proteins of the pyroptosis pathway: Tlr4, Nlpr3, Asc, Casp1 and Il-1b were significantly upregulated in both WT and Gng13-cKO lungs at 25 dpi comparing with those at 0 dpi. Further, the expression levels of these genes in Gng13-cKO were significantly higher than in WT at 25 dpi. Data are presented as means ± SD (n=3). Unpaired two-tailed student t tests were performed. (D, E) Western blot analysis shows that the protein levels of NLRP3, full-length Gasdermin D (Gsdmd-FL), N-terminal fragment of Gasdermin D (Gsdmd-N), pro-caspase 1 (pro-Casp1), cleaved caspase 1(Cleaved Casp-1 (p20)), and IL-1β in the Gng13-cKO lungs were significantly higher than the corresponding levels of WT at both 20 and 25 dpi (β-actin as an internal control). And over this phase, the expression levels of these proteins were reduced at 25 dpi in comparison with those at 20 dpi in both genotypes. Data are presented as means ± SD (n=4), and unpaired two-tailed t tests were performed. *P<0.05, *P<0.01, ***P<0.001, ****P<0.0001.

Figure 4—figure supplement 1. qRT-PCR analysis of gasdermin gene expression in WT and Gng13-cKO lungs.

Immunohistochemical analyses of gasdermin D and IL-1β expressing cells.

(A, B) Immunostaining of H1N1-infected lung sections Figure 5 continued on next page at 25 dpi with an antibody to gasdermin D (Gsdmd) indicates significantly more gasdermin D-expressing cells in the Gng13-cKO sections than in WT. Data are presented as means ± SD (n=3), and unpaired two-tailed student t tests were performed. (C) Co-immunostaining shows that nearly the same percentage, i.e., 98.5% and 95.5% of Gsdmd+ cells in WT and Gng13-cKO, respectively, were F4/80+. (D) Immunostaining of H1N1-infected lung sections at 25 dpi with an antibody to IL-1β indicates significantly more IL-1β-expressing cells in the Gng13-cKO sections than in WT. Data are presented as means ± SD (n=4), and unpaired two-tailed t tests were performed. ***P<0.001, ****P<0.0001.

Figure 5—figure supplement 1. Expression of gasdermin E and caspase 3 in the H1N1-injured lungs.

Figure 5—figure supplement 2. Partial colocalization of gasdermin D, E and IL-1β to epithelial cells and macrophages.

Gng13 conditional nullification leads to more pyroptotic cells in the injured lungs.

(A, B) Histological analysis of lung tissue sections following Sytox administration reveals that significantly more Sytox+ cells were found in the mutant lung sections than the corresponding WT tissues at both 20 and 25 dpi while in both types of mice the numbers of the Sytox+ cells at 25 dpi seemed to be fewer than that at 20 dpi. (C) Immunostaining with an antibody to the macrophage marker CD64 indicates that most of Sytox+ cells were macrophages in both WT and Gng13-cKO mice (circled by dashed lines in the bright field images on the far right). Data are presented as means ± SD (n=3), and unpaired two-tailed student t tests were performed. *P<0.05, **P<0.01.

Bronchoalveolar lavage fluid (BALF) assays.

(A) Average numbers of cells per mouse found in both WT and Gng13-cKO BALFs at both 20 and 25 dpi were significantly more than those at 0 dpi while those of the Gng13-cKO BALF were even more than those of WT at both 20 and 25 dpi. Data are presented as means ± SD (n=6), and unpaired two-tailed student t tests were performed. (B) The protein contents of both WT BALF and Gng13-cKO BALF at 20 dpi and 25 dpi were significantly more than their corresponding ones at 0 dpi, but the Gng13-cKO BALF had significantly more proteins than WT at both 20 and 25 dpi. And in WT, the protein content was significantly reduced from 20 dpi to 25 dpi, but the reduction in the Gng13-cKO was insignificant. Data are presented as means ± SD (n=5), and unpaired two-tailed student t tests were performed. (C) Lactate dehydrogenase (LDH) activity assays showed that at both 20 and 25 dpi, LDH activities in both WT and Gng13-cKO BALF were significantly higher than their corresponding ones at 0 dpi, but the Gng13-cKO BALF at both 20 and 25 dpi showed higher activities than the corresponding WT BALF. And both WT and Gng13-cKO BALF displayed significant LDH activity reduction from 20 to 25 dpi. Data are presented as means ± SD (n=3), and unpaired two-tailed student t tests were performed. (D) IL-1β ELISA assays showed a similar patter to that of LDH, except that the reduction in the IL-1β concentration from 20 to 25 dpi was insignificant in the Gng13-cKO BALF. Data are presented as means ± SD (n=4), and unpaired two-tailed student t tests were performed. *P<0.05, **P<0.01, ***P<0.001 ****P<0.0001.

More severe fibrosis in the H1N1-infected Gng13-cKO lungs.

(A) qRT-PCR analysis of Col1a1, Fn1 and Timp1 expression. While the expression levels of these three genes at 20 dpi in both WT and Gng13-cKO injured lungs were significantly greater than their corresponding ones at 0 dpi, the expression in Gng13-cKO at 25 dpi was much stronger than WT at 25 dpi. At 50 dpi, the expression levels of all these three genes were reduced comparing with those at 25 dpi in both WT and Gng13-cKO mice, but those of Fn1 and Timp1 in Gng13-cKO mice at 50 dpi were still significantly higher than WT at 50 dpi. Data are presented as means ± SD (n=3), and unpaired two-tailed student t tests were performed. (B, C) Masson’s trichrome staining showed that at 25 dpi, the fibrosis percentages in WT and Gng13-cKO were significantly more than their corresponding ones at 0 dpi, which were reduced at 50 dpi. Data are presented as means ± SD (n=3), and unpaired two-tailed student t tests were performed. *P<0.05, **P<0.01, ***P<0.001, ****P<0.0001.

Sequences of primers used for qPCR

Chemically induced severe injury also engenders ectopic formation of lung tuft cells.

Top three rows: Immunostaining with an antibody to Dclk1 (green) on the tissue sections of the injured lungs induced by bleomycin (BLM), extract of house dust mite (HDM) or lipopolysaccharides (LPS), respectively. Bottom row: double-immunostaining an LPS-injured lung section with antibodies to Dclk1 and Gαgust (Gα-gustducin). Scale bar: 50 µm.

Expression of taste2 receptors in tuft cell clusters.

Reanalysis of the single cell RNA sequencing data GSE197163 (Barr et al. 2022) using the Uniform manifold approximation and projection (UMAP) method. Among 613 Trpm5-GFP+ tuft cells, Tas2r108 was expressed in the greatest number of cells, i.e., 67 cells, followed by Tas2r105, Tas2R138, Tas2r137, Tas2r118 and Tas2r102, which were detected in 11,10, 10, 5 and 4 cells, respectively. The clustering analysis results also indicate that different cells may express a different subset of Tas2rs.

FACs sorting of ectopic tuft cells.

ChAT-tdTomato reporter expression in live lung epithelial cells post-influenza via FACs.

Statistical analysis of the injured tissue volumes on the lungs of WT, Trpm5−∕− and Gng13-cKO mice at 25 dpi.

(A)Sections of WT, Trpm5−∕− and Gng13-cKO mice at 25 dpi. Areas inside the dotted lines were determined as the severely damaged area. Scale bar: 0.5 mm. (B)Statistical analysis of the injured tissue volumes on the lungs of WT, Trpm5−∕− and Gng13-cKO mice at 25 dpi. The thickness of each section was 12 µm, and the area of lung injury in each section was determined using imageJ. The volume of the lung injury in each section was the area of the lung injury area multiplied by the section thickness. Each section of lung tissue was collected, and the sum of the lung injury volumes of all sections was the final volume for the whole lung. Data are presented as means ± SD (n=3), and unpaired two-tailed student t tests were performed. **P<0.01.

Ectopic generation of tuft cells in WT, Trpm5−∕− and Gng13-cKO mice over days post H1N1 infection.

(A, B, C, D, E) Immunostaining and quantitative analyses of the abundance of Dclk1+ tuft cells in the Krt5+ regions of the injured lung tissues at 14 and 25 dpi, respectively. (F) Comparative analysis of Dclk1+ tuft cell numbers in the Krt5+ regions of WT versus Gng13-cKO over days post infection. (G,H) Double immunostaining of lung tissue sections of 25 dpi with antibodies to Dclk1 and Gγ13 shows a subset of Dclk1+ tuft cells expressing Gγ13 in WT but not in Gng13-cKO mice. (I) UMAP analyses of the single-cell RNAseq dataset GSE197163 (Barr et al. 2022) indicate that out of 613 Trpm5-GFP+ tuft cells, 350, i.e., 57% of them, expressed Gγ13. (J) UMAP analyses of the single-cell RNAseq dataset GSE197163 (Barr et al. 2022) indicate that out of 350 Gγ13+ tuft cells, 277, i.e., 79% of them expressed Alox5.

qRT-PCR analysis of gasdermin gene expression in WT and Gng13-cKO lungs.

Expression of gasdermin E and caspase 3 in the H1N1-injured lungs.

Partial colocalization of gasdermin D, E and IL-1β to epithelial cells and macrophages.