MAPK phosphatase 1 inhibition of p38α within lung myofibroblasts is essential for spontaneous fibrosis resolution

Abstract

Fibrosis following tissue injury is distinguished from normal repair by the accumulation of pathogenic and apoptosis-resistant myofibroblasts (MFs), which arise primarily by differentiation from resident fibroblasts. Endogenous molecular brakes that promote MF dedifferentiation and clearance during spontaneous resolution of experimental lung fibrosis may provide insights that could inform and improve the treatment of progressive pulmonary fibrosis in patients. MAPK phosphatase 1 (MKP1) influences the cellular phenotype and fate through precise and timely regulation of MAPK activity within various cell types and tissues, yet its role in lung fibroblasts and pulmonary fibrosis has not been explored. Using gain- and loss-of-function studies, we found that MKP1 promoted lung MF dedifferentiation and restored the sensitivity of these cells to apoptosis - effects determined to be mainly dependent on MKP1's dephosphorylation of p38α MAPK (p38α). Fibroblast-specific deletion of MKP1 following peak bleomycin-induced lung fibrosis largely abrogated its subsequent spontaneous resolution. Such resolution was restored by treating these transgenic mice with the p38α inhibitor VX-702. We conclude that MKP1 is a critical antifibrotic brake whose inhibition of pathogenic p38α in lung fibroblasts is necessary for fibrosis resolution following lung injury.

Keywords: Cell biology; Fibrosis; Molecular biology; Protein kinases; Pulmonology.

Figure 1. MKP1 (DUSP1) is the chief DUSP isoform expressed in normal lung fibroblasts, is reduced in IPF fibroblasts, and is downregulated by TGF-β.

(A) Single-cell RNA-Seq of annotated human and mouse lung fibroblast populations (generated from the Chan-Zuckerberg CELL by GENE Discover online database) depicting the relative expression of each catalytically active DUSP/MKP gene. Circle color denotes the mean gene expression within each fibroblast subtype, and the circle size represents the proportion of each cell population expressing the indicated gene. (B) MKP1 protein expression measured by Western blotting in normal and IPF patient–derived human lung fibroblasts (left: representative blot of individual patient-derived cells; right: densitometric analysis of all such cells). (C) MKP1 and collagen I protein expression by Western blotting (left and middle) and MKP1 transcript by qPCR (right) in normal HLFs treated with TGF-β (2 ng/mL) for 3 hours or 48 hours, as indicated in C. Data points represent separate experiments. Significance for densitometric data in B (n = 7) and C (n = 4) and qPCR in C (n = 5) was determined by 2-tailed t test. **P < 0.01 and ****P < 0.0001.

Figure 2. Effect of gain- and loss-of-function of MKP1 on lung fibroblast phenotypic features.

(A) CRISPR/Cas9-mediated DUSP1 deletion using MKP1 or NT sgRNAs in HLF-derived MFs. Protein quantification by Western blotting of MKP1 and the fibrosis-associated genes αSMA, Col1A1, FN1, and CTHRC1 (top: representative blot; bottom: densitometric analysis). (B and C) Inducible MKP1 overexpression in human lung MFs. (B) Protein quantification by Western blotting of MKP1 and the fibrosis-associated genes αSMA, Col1A1, FN1, and CTHRC1 (top: representative blot; bottom: densitometric analysis). (C) αSMA stress fibers identified by immunofluorescence microscopy using an anti–αSMA-Cy3–conjugated antibody in MFs (Myofibs) overexpressing MKP1 or GFP and fibroblast (Fibs) controls (using the same protocol as in B). Nuclei were stained with DAPI. Scale bars: 20 μm (top row) and 10 μm (bottom row). (D) IPF fibroblasts harboring the same MKP1 overexpression construct as normal HLFs in B were treated with doxycycline for 48 hours to induce MKP1 or GFP overexpression. Protein quantification by Western blotting of MKP1 and the fibrosis-associated genes αSMA, Col1A1, FN1, and CTHRC1 (left: representative blots, right: densitometric analysis). (E) Apoptosis sensitivity in MKP1-overexpressing (or vehicle-treated) MFs further treated with an anti-Fas–activating antibody (100 ng/mL) for 24 hours. Apoptosis was determined by caspase-3/-7 activity assay (left) or annexin V expression (right). Dashed lines represent caspase-3/-7 activity or annexin V expression in untreated, undifferentiated fibroblasts stimulated with anti-Fas antibody. The sample number for each experiment (n) varied between 3 and 9 and is indicated by the number of data points in each histogram. Each blot grouping containing a protein (or proteins) of interested and its corresponding loading control were run on separate gels. Significance for densitometric analysis and apoptosis activity assays was determined by 2-tailed t test. *P < 0.05, **P < 0.01, ***P < 0.001, and ****P < 0.0001. Dox, doxycycline.

Figure 3. Lung fibroblast expression of MKP1 mitigates peak fibrosis and is essential for spontaneous fibrosis resolution following in vivo administration of bleomycin.

(A) Schematic illustrating the peak fibrosis protocol. (B) PCR of the Dusp1 locus in Cre^– and Cre⁺ mouse tails following tamoxifen administration (left), and subsequent MKP1 protein by Western blot in Cre^– or Cre⁺ cultured lung fibroblasts (right). (C) Hydroxyproline content quantified from the left and right upper/middle lobe lung homogenates in saline-treated, bleomycin-treated Cre^– , and bleomycin-treated Cre⁺ mice on day 21. (D) Representative images of Masson’s trichrome staining of the right lower lobe from the same mice used in C. Scale bar: 1 mm. (E) Schematic illustrating the resolution protocol. (F) Representative images of Masson’s trichrome staining of the right lower lobe in saline- and bleomycin-treated WT Cre^– mice on day 21 and bleomycin-treated Cre^– or bleomycin-treated Cre⁺ mice on days 42 or 63. Scale bars: 1 mm (top row) and 100 μm (bottom row). (G) Hydroxyproline content quantified in left and right upper/middle lobe lung homogenates from the same mice in F. (H–J) Immunofluorescence microscopy images of bleomycin-treated mice at mid-resolution (day 42) depicting type I (PDPN) and type II (pro-SPC) alveolar epithelial cells (H), parenchymal bronchiolization (E-cadherin) (I), and alveolar macrophages (CD68) (J). White arrows in I depict normal airways. Open white arrowheads depict E-cadherin staining of type II alveolar epithelial cells. Solid white arrowheads point to alveolar regions devoid of type I cells and type II cell hyperplasia in H, regions of parenchymal bronchiolization in I, and alveolar macrophages in J. Scale bars: 500 μm (top row) and 50 μm (bottom row). Each data point represents an individual mouse. Significance for hydroxyproline was determined by 1-way ANOVA. *P < 0.05, **P < 0.01, ***P < 0.001, and ****P < 0.0001. Bleo, bleomycin.

Figure 4. p38 is the MAPK whose inhibition by MKP1 accounts for its ability to dedifferentiate MFs.

Effect of MKP1 overexpression (in MFs) or its deletion (in normal HLFs) on MAPK phosphorylation. (A and B) Western blots of MKP1 and phosphorylated and total MAPK proteins and densitometric analysis. (C and D) Normal HLFs were treated with TGF-β (2 ng/mL) for 48 hours to generate MFs, followed by treatment with the p38 inhibitor SB203580 (20 μM) for 96 hours. (C) Western blot analysis of the fibrosis-associated genes αSMA and Col1A1 and densitometric analysis. (D) αSMA stress fibers were identified by immunofluorescence microscopy using an anti–αSMA-FITC–conjugated antibody in MFs treated with SB203580 for 96 hours and fibroblast controls (using the same protocol as in C). Nuclei were stained with DAPI. Scale bars: 20 μm (top row) and 10 μm (bottom row). (E) Apoptosis sensitivity in SB203580- or vehicle-treated MFs (see protocol schematic) via anti-Fas–activating antibody (100 ng/mL) stimulation for 24 hours. Apoptosis was determined by caspase-3/-7 (Casp 3/7) activity assay (middle) or annexin V expression (right). Dashed lines represent caspase-3/-7 activity or annexin V expression in untreated, undifferentiated fibroblasts treated with anti-Fas. The sample number for each experiment (n) varied between 3 and 7 and is indicated by the number of data points in each histogram. Each blot grouping containing a protein of interest and its corresponding loading control were run on separate gels. **P < 0.01 and ****P < 0.0001, by 2-tailed t test (A, B, and E) and 1-way ANOVA (C).

Figure 5. p38α is the isoform whose inhibition by MKP1 promotes MF dedifferentiation.

(A) Relative p38α and β mRNA expression quantified by qPCR in normal HLFs or MFs and patient-derived IPF fibroblasts. (B) Protein quantification of p38α by Western blotting in normal HLFs following Cas9-mediated deletion of p38α using MAPK14 sgRNA or a NT control. p38α was quantified by subtracting the densitometric value of the total p38 band in the isotype-deleted line from that of the total p38 band of the WT line. (C) Western blot analysis of the fibrosis-associated genes αSMA and Col1A1 and densitometric analysis following Cas9-mediated MAPK14/p38α deletion. (D and E) MFs were treatment with the p38 inhibitor VX-702 (50 μM) for 96 hours (protocol schematic is shown in D). (D) Western blot analysis of αSMA and Col1A1 and densitometric analysis. (E) αSMA stress fibers were identified by immunofluorescence microscopy using an anti–αSMA-FITC–conjugated antibody. Nuclei were stained with DAPI. Scale bars: 20 μm (top row) and 10 μm (bottom row). (F) Schematic of protocol showing that VX-702– or vehicle-treated MFs were treated with an anti-Fas–activating antibody (100 ng/mL) for 24 hours. Apoptosis sensitivity was determined by caspase-3/-7 activity assay or annexin V expression (right). Dashed lines represent caspase-3/-7 activity or annexin V expression in untreated, undifferentiated fibroblasts incubated with anti-Fas antibody. The sample number for each experiment (n) varied between 3 and 5 and is indicated by the number of data points in each histogram. Each blot grouping containing a protein of interest and its corresponding loading control were run on separate gels. **P < 0.01, ***P < 0.001, and ****P < 0.0001 by 2-tailed t test (A, B, and F) and 1-way ANOVA (D).

Figure 6. The p38α-specific inhibitor VX-702 mitigates bleomycin-induced fibrosis and restores spontaneous fibrosis resolution in mice with MKP1-deficient fibroblasts.

(A) Schematic illustrating the peak fibrosis protocol. VX-702 was administered to mice daily by o.g. starting on day 9 through sacrifice on day 21. (B) Hydroxyproline content was quantified from left and right upper/middle lobe lung homogenates in saline-, bleomycin-, and bleomycin plus VX-702–treated mice on day 21. (C) Representative Masson’s trichrome–stained images of the right lower lobe from the same mice used in B. Scale bar: 1 mm. (D) Schematic illustrating the resolution protocol. A tamoxifen chow diet was introduced, and VX-702 was administered to Cre⁺ mice by o.g. daily starting on day 21 until sacrifice on day 56. (E) Hydroxyproline content was quantified in left and right upper/middle lobe lung homogenates in bleomycin-treated Cre^– mice, bleomycin-treated Cre⁺ mice, and bleomycin- plus VX-702–treated Cre⁺ mice (day 56). (F) Representative images of Masson’s trichrome staining of the right lower lobe from the same mice used in E. Scale bars: 1 mm (top row) and 100 μm (bottom row). (G) qPCR of whole-lung expression of Fn1, Cthrc1, Tgfb1, Birc5, Col1a1, Col3a1, and Acta2 from the same lung homogenates used in E. Each data point represents an individual mouse. Significance for hydroxyproline was determined by 1-way ANOVA and for whole-lung RNA by unpaired, 2-tailed t test. *P < 0.05, **P < 0.01, and ***P < 0.001.

Figure 7. MKP1 induction is essential for PGE₂/cAMP/PKA-mediated inhibition of p38 and MF dedifferentiation.

(A) MKP1 and p-p38 protein quantification by Western blotting in MFs treated with PGE₂ (1 μM), the direct adenylyl cyclase activator forskolin (20 μM), the PKA-specific agonist 6-BNZ-cAMP (2 mM), or the Epac-specific agonist 8-pCPT-cAMP (2 mM) for 6 hours (left: representative blot, right: densitometric analysis). (B) MKP1, p-p38, and total p38 expression was quantified by Western blotting in doxycycline-treated lentiCRISPR HLFs containing a DUSP1-specific or NT sgRNA (top protocol schematic in B) or in lung fibroblasts isolated from naive Cre⁺ or Cre^– Col1a2^CreERT2 Dusp1^fl/fl mice (bottom protocol schematic in B). HLFs and mouse lung fibroblasts were subsequently treated with TGF-β (2 ng/mL for HLFs; 5 ng/mL for mouse lung fibroblasts) for 48 hours to promote the MF phenotype and were then treated with PGE₂ or vehicle for 6 hours (left: representative blots, right: densitometric analysis). (C) Protein quantification of αSMA and Col1A1 seventy-two hours after PGE₂ treatment of the same lentiCRISPR human MFs generated in B. The sample number for each experiment (n) varied between 3 and 4 and is indicated by the number of data points in each histogram. Each blot grouping containing a protein of interest and its corresponding loading control were run on separate gels. *P < 0.05, **P < 0.01, ***P < 0.001, and ****P < 0.0001, by 1-way ANOVA.

Figure 8. The 2 FDA-approved antifibrotic drugs pirfenidone and nintedanib fail to dephosphorylate p38 and to promote dedifferentiation of human lung MFs.

(A) Schematic illustrating in vitro MF prevention and MF reversal treatment protocols. Treatment with the antifibrotic agent pirfenidone (1 mM) or nintedanib (2 μM) or the lipid mediator PGE₂ (1 μM) was administered 15 minutes prior to (prevention) or 48 hours after (reversal) the addition of TGF-β (2 ng/mL) to the culture medium. (B and C) Protein quantification of p-p38, αSMA, and Col1A1 by Western blotting in normal HLFs following treatment with pirfenidone (B) or nintedanib (C) compared with PGE₂ in a prevention or reversal protocol (left: representative blot, right: densitometric analysis). For the prevention studies, proteins were quantified by Western blotting using cell lysates harvested at the following post-treatment time points: p-p38, 6 hours; αSMA and Col1A1, 48 hours. For the reversal studies, proteins were quantified by Western blotting using cell lysates harvested at the following post-treatment time points: p-p38, 24 hours; αSMA, 72 hours; and Col1A1, 48 hours (B) or 72 hours (C). The sample number for each experiment (n) varied between 3 and 6 and is indicated by the number of data points in each histogram. Each blot grouping containing a protein (or proteins) of interest and its corresponding loading control were run on separate gels. *P < 0.05, **P < 0.01, ***P < 0.001, and ****P < 0.0001, by 1-way ANOVA.