Skip to main page content
U.S. flag

An official website of the United States government

Dot gov

The .gov means it’s official.
Federal government websites often end in .gov or .mil. Before sharing sensitive information, make sure you’re on a federal government site.

Https

The site is secure.
The https:// ensures that you are connecting to the official website and that any information you provide is encrypted and transmitted securely.

Access keys NCBI Homepage MyNCBI Homepage Main Content Main Navigation
. 2023 Nov 15;133(22):e165612.
doi: 10.1172/JCI165612.

Regulation of epithelial transitional states in murine and human pulmonary fibrosis

Fa Wang  1 , Christopher Ting  1 , Kent A Riemondy  2 , Michael Douglas  1 , Kendall Foster  3 , Nisha Patel  3 , Norihito Kaku  1 , Alexander Linsalata  4 , Jean Nemzek  5 , Brian M Varisco  6   7 , Erez Cohen  8 , Jasmine A Wilson  9 , David Wh Riches  9   10   11 , Elizabeth F Redente  9   10 , Diana M Toivola  12 , Xiaofeng Zhou  13 , Bethany B Moore  1   13 , Pierre A Coulombe  8 , M Bishr Omary  14 , Rachel L Zemans  1   4
Affiliations

Regulation of epithelial transitional states in murine and human pulmonary fibrosis

Fa Wang et al. J Clin Invest. .

Abstract

Idiopathic pulmonary fibrosis (IPF) is a progressive scarring disease arising from impaired regeneration of the alveolar epithelium after injury. During regeneration, type 2 alveolar epithelial cells (AEC2s) assume a transitional state that upregulates multiple keratins and ultimately differentiate into AEC1s. In IPF, transitional AECs accumulate with ineffectual AEC1 differentiation. However, whether and how transitional cells cause fibrosis, whether keratins regulate transitional cell accumulation and fibrosis, and why transitional AECs and fibrosis resolve in mouse models but accumulate in IPF are unclear. Here, we show that human keratin 8 (KRT8) genetic variants were associated with IPF. Krt8-/- mice were protected from fibrosis and accumulation of the transitional state. Keratin 8 (K8) regulated the expression of macrophage chemokines and macrophage recruitment. Profibrotic macrophages and myofibroblasts promoted the accumulation of transitional AECs, establishing a K8-dependent positive feedback loop driving fibrogenesis. Finally, rare murine transitional AECs were highly senescent and basaloid and may not differentiate into AEC1s, recapitulating the aberrant basaloid state in human IPF. We conclude that transitional AECs induced and were maintained by fibrosis in a K8-dependent manner; in mice, most transitional cells and fibrosis resolved, whereas in human IPF, transitional AECs evolved into an aberrant basaloid state that persisted with progressive fibrosis.

Keywords: Adult stem cells; Fibrosis; Pulmonology.

PubMed Disclaimer

Conflict of interest statement

Conflict of interest: The authors have declared that no conflict of interest exists.

Figures

Figure 1
Figure 1. Conserved keratinhi transitional AECs arise prior to fibrosis and activate profibrotic processes.
(A, C, and D) scRNA-Seq data sets from the bleomycin (54), LPS (50), pneumonectomy (56), and organoid (53) mouse models and human IPF (57, 58) were analyzed. (B, C, and E) Mice were treated with bleomycin. (A) Several keratins were upregulated in the transitional state in multiple mouse models and in human IPF. (B and C) Transitional cells arose prior to fibrosis. (D) Scores indicating activation of each profibrotic pathway were calculated on the basis of canonical gene expression. The expression of genes representative of each pathway are shown by heatmaps (D) and immunostaining or FISH (E). Solid white arrowheads indicate transitional cells showing activation of a given pathway. Open arrowheads indicate rare K8hi cells without Itgb6 staining. Orange arrowheads indicate CDKN1A+H2AX+ cells. Multiple profibrotic pathways, senescence, TGF-β, impaired proteostasis, DNA damage, and cell death were uniquely and concurrently activated in the transitional cell state in multiple mouse models and human IPF. Scale bars: 100 μm. Original magnification, ×20 (enlarged insets in B and E). (B and E) n = 3 mice/group; (C) n = 5 mice/group (hydroxyproline). Hydroxyproline data are represented as box-and-whisker plots, with the box (25th to 75th percentiles), median (line), and whiskers (minimum to maximum). (C) ***P < 0.001 compared with day 0 by 1-way ANOVA with post hoc Bonferroni’s test.
Figure 2
Figure 2. K8 promotes fibrosis and accumulation of transitional AECs.
(A) Using a genome-wide association meta-analysis of IPF (75), a nested candidate gene study for the keratins expressed in the transitional state was performed. Regional association plot showing all SNPs that overlap with KRT8. Gray dotted line indicates genome-wide significance; red dotted line indicates statistical significance of the nested candidate gene study for the keratin genes; blue curve indicates the estimated recombination rate. Seven KRT8 SNPs were associated with IPF (P < 1.4 × 10–4). The most significant variant, rs4531558 (P = 5.2 × 10–5), shown as a purple diamond, is in linkage disequilibrium (LD) (R2 > 0.8) with all other statistically significant variants. (B–G) Krt8+/+ and Krt8–/– mice were treated with bleomycin. Krt8–/– mice were protected from fibrosis, as determined by hydroxyproline assay (B), trichrome staining (C), and myofibroblast accumulation (D). Arrowhead in C indicates a small area of fibrosis. Krt8–/– mice were not protected from lung injury at day 4, as determined by inflammation (E) and permeability (F). (G) Compared with Krt8–/– mice, transitional cells accumulated with incomplete AEC1 regeneration in Krt8+/+ mice. (B, E, and F) Data are represented as box-and-whisker plots, with box (25th to 75th percentiles), median (line), and whiskers (minimum to maximum). *P < 0.05 and **P < 0.01, by 1-way ANOVA with post hoc Bonferroni’s test. (G) Data indicate the mean ± SD. *P < 0.05, by 2-way ANOVA with post hoc Šidák’s multiple-comparison test. Scale bars: 200 μm. Original magnification, ×20 (enlarged insets in D). n = 3 mice/group except bleomycin-treated mice in B, n = 14 mice/group and E, n = 6 mice/group.
Figure 3
Figure 3. K8 promotes the expression of chemokines but not the accumulation of transitional cells at the expense of AEC1 differentiation.
AEC2s were isolated from Krt8+/+ and Krt8–/– mice and cultured in 2D. RNA-Seq was performed. (A) Average fold change (FC) of composite AEC2, transitional state, or AEC1 marker scores (see also Supplemental Table 2) compared with day 0 for WT AECs. *P < 0.05 compared with day 0. AEC culture recapitulates in vivo stages of alveolar regeneration, as shown by downregulation of AEC2 markers and upregulation of transitional state markers on day 1 of culturing and a gradual upregulation of AEC1 markers by day 7. Far-right panel is a schematic representation of the data. (B) Krt8 deficiency had no effect on transitional cell or AEC1 differentiation. ****P < 0.0001, by unpaired t test on the AUC from days 1–7 for Krt8+/+ versus Krt8–/– cells. (C) Markers of senescence, TGF-β activation, impaired proteostasis, DNA damage, and cell death were upregulated in the transitional state in vitro. *P ≤ 0.05 by t test for day 1 compared with day 0 for all pathway scores. P values for genes in heatmaps are listed in Supplemental Table 3. (D) K8 was not necessary for upregulation of markers of cell-cycle arrest, TGF-β activation, impaired proteostasis, DNA damage, and cell death. For individual genes, the ratio of AUC of expression from days 1–7 in Krt8–/– versus Krt8+/+ cells is shown. (E) K8 was necessary for the expression of SASP chemokines but not proinflammatory cytokines, growth factors, or proteases/antiproteases. #P < 0.05, by t test of the average of the ratio of AUCs of all chemokines; *P < 0.05, by t test for individual genes. n = 3. All data are presented as the mean (A and B) or the mean ± SEM (C–E).
Figure 4
Figure 4. K8 is necessary for macrophage chemokine expression and macrophage recruitment during fibrosis.
Krt8+/+ and Krt8–/– mice were treated with bleomycin. (A) CCL2 ELISA on day 21 bronchoalveolar lavage (BAL). n = 3 mice/group. (B) BAL cells and differentials. n = 3 mice/group except for day 4 (n = 6 mice/group) and day 12 (n = 2 mice/group). Macs, macrophages; Lymphs, lymphocytes; PMN, polymorphonuclear neutrophils. (A and B) Violin plots show minimum to maximum values with the line at the median. *P < 0.05, by ratio paired t test. n = 4–5 mice/group. (C) Immunostaining on day-12 tissue. The top left image of the left panel and the top left image of the right panel in C are also shown in Figure 1E (macrophage recruitment control, left and middle images). n = 3 mice/group. Scale bar: 100 μm. Original magnification, ×20 (enlarged insets in C).
Figure 5
Figure 5. Macrophages and fibroblasts promote the accumulation of transitional AECs.
(A) Ccr2+/+ or Ccr2–/– mice were treated with bleomycin. Macrophage recruitment was necessary for transitional state accumulation. #P < 0.05 and ##P < 0.01, by 2-way ANOVA for Ccr2+/+ versus Ccr2–/– from days 4–21. n = 5 mice/group. (B) Macrophages and monocytes were a major source of TGF-β and IL-1β in murine and human fibrosis, as determined by scRNA-Seq (54, 58). Bleo, bleomycin. (C) Col1a1CreERT2 Fasfl/fl mice were treated with bleomycin or administered tamoxifen (KO) or corn oil (WT), and euthanized at 9 weeks. Fibroblast-specific Fas knockout induced myofibroblast persistence, which was sufficient for persistence of the AEC transitional state. #P < 0.05, by t test. n = 3 mice/group. (D) Gene expression by AECs cultured on Matrigel or collagen-coated plastic for 3 days. Collagen/stiff substrate promoted transitional state accumulation. Data represent the mean ± SD. #P < 0.05 and ##P < 0.01, by paired t test. (E) Murine AEC2s were cultured in 2D and fixed and immunostained on day 7. Most cells persisted in the transitional state, with some AEC1 differentiation (n = 3). (F and G) Transitional cells were found in small foci of fibrosis in peripheral lung in the PNX and LPS mouse models (*), whereas most of the lung was devoid of fibrosis and transitional cells (**). In the bleomycin model, large areas of lung were characterized by fibrosis and transitional cells (*), whereas some areas were devoid of fibrosis and transitional cells (**). Scale bars: 50 μm. Original magnification, ×20 (enlarged insets in A, C, E, and F). For immunostaining, n = 3/group. PDPN, podoplanin; αSMA, α smooth muscle actin.
Figure 6
Figure 6. Meta-analysis of scRNA-Seq data sets from the bleomycin, LPS, and organoid models of alveolar regeneration.
scRNA-Seq data sets from the bleomycin (54), LPS (50), and organoid (53) models were integrated and subjected to (A) unsupervised clustering. (B and C) Clusters were annotated on the basis of expression of canonical markers of AEC2s, AEC1s, BASCs, proliferating AEC2s, and transitional cells. Transitional cells from all 3 models coclustered. (A and D) Unsupervised clustering revealed 2 clusters of transitional cells.
Figure 7
Figure 7. Murine transitional cells include a highly senescent, basaloid subset.
Bleomycin, LPS, and organoid scRNA-Seq data sets were integrated. (A) Expression of senescence markers. p16 (Cdkn2a) was exclusively expressed in cluster 7. (B) Pseudotime analysis suggested cells in cluster 7 may not have an AEC1 fate. (C) Top DEGs in cluster 7. (D) Cultured primary murine AEC2s upregulated cluster 7 markers and p16 (Cdkn2a) and were highly senescent. Human IPF transitional cells expressed high levels of cluster 7 markers, as shown by scRNA-Seq (E) and immunostaining or FISH (F). (G) The top DEGs in the human IPF transitional (KRT5–KRT17+) state (58) were differentially expressed in murine cluster 7. (H) scRNA-Seq data sets from IPF (57, 58), normal human lung (86), or human organoids (60) were interrogated. Mature basal cell genes found among the top 100 DEGs of the human IPF transitional state were upregulated in cluster 7 of the murine AECs (H and J) and in cultured murine AECs (I). (K) K17 was occasionally expressed in lineage-labeled cells in bleomycin-treated SftpcCreERT2 mTmG mice. (L) Compared with AEC2s, human “transitional AEC2s” from IPF (58), ABI1s from human organoids (60), and murine transitional cells in cluster 1 downregulated AEC2 markers and upregulated classic transitional state markers. KRT5–KRT17+ AECs from IPF (58), ABI2s from human organoids (60), and murine transitional cells in cluster 7 upregulated basaloid genes. (J and M) Transitional cells expressing cluster 7 markers were rare in the single bleomycin model but common in the repetitive bleomycin model. (N) Transitional cells in human IPF but not ARDS expressed basaloid markers and p16/CDKN2A. (D and I) Data represent the mean (n = 3). **P < 0.01 compared with day 0, by 1-way ANOVA with post hoc Bonferroni’s test. (D) P < 0.05 for all genes at day 7 compared with day 0. Scale bars: 50 μm. Original magnification, ×20 (enlarged insets in F, K, M, and N). Arrowheads indicate transitional cells expressing a marker of interest. For immunostaining, n = 3 mice/group.
Figure 8
Figure 8. Regulation of epithelial transitional states in murine and human pulmonary fibrosis.
Our current working construct is that after injury, alveolar progenitors assume the K8hi transitional state characterized by the activation of multiple profibrotic processes: senescence, impaired proteostasis, DNA damage, cell death, integrin β6-dependent TGF-β activation, and macrophage chemokine expression. K8 promotes fibrosis by regulating expression of macrophage chemokines, which recruit profibrotic macrophages that further drive fibrosis. Fibroblasts are activated to contract and deposit matrix, stiffening the lung. Stiffness, as well as TGF-β, largely synthesized by macrophages and activated by transitional AEC integrin β6, and IL-1β, synthesized by macrophages, promote accumulation of the AEC transitional state at the expense of AEC1 differentiation. Taken together, our data suggest that crosstalk between K8hi transitional AECs, profibrotic macrophages, and activated fibroblasts maintain each other in an activated state in the lung, establishing a positive feedback loop that drives fibrosis. In the absence of fibrosis, AEC2s may bypass the transitional state and differentiate into AEC1s (dotted line). In mouse models and in humans who recover from acute lung injury, this positive feedback loop is eventually broken, and transitional cells differentiate into AEC1s or perhaps die with resolution of macrophages and activated fibroblasts; in human IPF, the transitional cells further evolve into a permanently senescent, aberrant basaloid state instead of into AEC1s, driving a self-amplifying feedback loop that underlies the progressive and ultimately fatal clinical disease. Adapted from ref. with permission.
See this image and copyright information in PMC

Comment in

Similar articles

See all similar articles

Cited by

See all "Cited by" articles

References

    1. Blackwell TS, et al. Future directions in idiopathic pulmonary fibrosis research. An NHLBI workshop report. Am J Respir Crit Care Med. 2014;189(2):214–222. doi: 10.1164/rccm.201306-1141WS. - DOI - PMC - PubMed
    1. Katzen J, Beers MF. Contributions of alveolar epithelial cell quality control to pulmonary fibrosis. J Clin Invest. 2020;130(10):5088–5099. doi: 10.1172/JCI139519. - DOI - PMC - PubMed
    1. Borok Z, et al. Grp78 loss in epithelial progenitors reveals an age-linked role for endoplasmic reticulum stress in pulmonary fibrosis. Am J Respir Crit Care Med. 2020;201(2):198–211. doi: 10.1164/rccm.201902-0451OC. - DOI - PMC - PubMed
    1. Korfei M, et al. Epithelial endoplasmic reticulum stress and apoptosis in sporadic idiopathic pulmonary fibrosis. Am J Respir Crit Care Med. 2008;178(8):838–846. doi: 10.1164/rccm.200802-313OC. - DOI - PMC - PubMed
    1. Lawson WE, et al. Endoplasmic reticulum stress enhances fibrotic remodeling in the lungs. Proc Natl Acad Sci U S A. 2011;108(26):10562–10567. doi: 10.1073/pnas.1107559108. - DOI - PMC - PubMed

Publication types