CYB5R3 in type II alveolar epithelial cells protects against lung fibrosis by suppressing TGF-β1 signaling

Abstract

Type II alveolar epithelial cell (AECII) redox imbalance contributes to the pathogenesis of idiopathic pulmonary fibrosis (IPF), a deadly disease with limited treatment options. Here, we show that expression of membrane-bound cytochrome B5 reductase 3 (CYB5R3), an enzyme critical for maintaining cellular redox homeostasis and soluble guanylate cyclase (sGC) heme iron redox state, is diminished in IPF AECIIs. Deficiency of CYB5R3 in AECIIs led to sustained activation of the pro-fibrotic factor TGF-β1 and increased susceptibility to lung fibrosis. We further show that CYB5R3 is a critical regulator of ERK1/2 phosphorylation and the sGC/cGMP/protein kinase G axis that modulates activation of the TGF-β1 signaling pathway. We demonstrate that sGC agonists (BAY 41-8543 and BAY 54-6544) are effective in reducing the pulmonary fibrotic outcomes of in vivo deficiency of CYB5R3 in AECIIs. Taken together, these results show that CYB5R3 in AECIIs is required to maintain resilience after lung injury and fibrosis and that therapeutic manipulation of the sGC redox state could provide a basis for treating fibrotic conditions in the lung and beyond.

Keywords: Cellular senescence; Fibrosis; Pulmonology.

Figure 1. Decreased levels of NAD⁺ in fibrotic lung with reduced expression of CYB5R3 in AECIIs.

(A) Levels of normalized NAD⁺ and NAD⁺/NADH ratio are reduced in IPF lungs compared with age-matched donors. (Min to max with median, n = 5/group.) (B) Compilation of mean CYB5R3 expression in AECII and fibroblast clusters from 3 single-cell RNA-Seq studies comparing donor and IPF lungs. (Violin plots with median and quartiles, n = 45 donor/56 IPF.) See Supplemental Figure 1 and ref. for details on data compilation. (C) Representative immunofluorescence using anti–HTII-280 (AECII marker, green) and anti-CYB5R3 (red) antibodies, showing reduction in CYB5R3 expression in hyperplastic AECIIs from honeycombs in IPF lung compared with age-matched donors (n = 3/group). Scale bars: 50 μm. Representative cells (asterisks) are shown in detail in the insets at ×6.5 original magnification. (D) Quantification of CYB5R3 staining (as relative pixel intensity) of individual AECIIs (HTII-280⁺) per condition. (Violin plots with median and quartiles; n = 3/group.) See Supplemental Table 1 for demographics of panels A, C, and D. Statistical analysis was performed using 1-way ANOVA with multiple-comparison test (A) and unpaired, 2-tailed Student’s t test (B and D); *P < 0.05, ****P < 0.0001.

Figure 2. Loss of CYB5R3 in AECIIs increases susceptibility to lung fibrosis after MHV68 infection at day 15.

(A) Representative Masson’s trichrome staining in lung sections from Cyb5r3^fl/fl and Cyb5r3 SPC–KO showing increased interstitial collagen deposition (blue) at day 15 after MHV68 infection. Scale bars: 500 μm. (n = 6–8/group.) (B) Increased collagen deposition in lungs of Cyb5r3 SPC–KO mice after infection determined by hydroxyproline levels. (Min to max with median, n = 5–6/group.) (C) Kaplan-Meier survival curve of Cyb5r3 SPC–KO and Cyb5r3^fl/fl mice after infection (n = 23–24, starting mice). (D) Weight loss data in Cyb5r3^fl/fl and Cyb5r3 SPC–KO mice with and without MHV68 infection. More severe weight loss was observed in the infected CYB5R3 AECII–deficient mice. (Data points are mean ± SD, n = 6–8.) (E) Changes of fibrotic markers Tgfb1, Col1a1, and Fn1 mRNA expression levels after virus infection in Cyb5r3^fl/fl and Cyb5r3 SPC–KO mice. (Min to max with median, n = 6- 8/group.) (F) Changes in the relative transcript levels of different Mmps and Timp1 in total lung lysate at 15 days after MHV68 infection. (Min to max with median, n = 6–8/group.) (G) After infection, Cyb5r3 SPC–KO mice show higher levels of osteopontin by transcript (Spp1 mRNA) in total lung lysate and secreted protein in bronchoalveolar lavage fluid. (Min to max with median, n = 6–8/group.) Statistical analysis was performed using 2-way ANOVA with multiple-comparison test (B and E–G), Mantel-Cox test (C), 1-way repeated measures ANOVA (D), and unpaired, 2-tailed Student’s t test (F), versus naive: *P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001; as indicated: ^#P < 0.05, ^###P < 0.001, ^####P < 0.0001.

Figure 3. Loss of CYB5R3 in AECIIs increases susceptibility to lung fibrosis after bleomycin with high mortality.

(A) Representative Masson’s trichrome staining in lung sections from Cyb5r3^fl/fl and Cyb5r3 SPC–KO at day 14 after bleomycin. Scale bars: 500 μm. (n = 4–5/group.) (B) Increased collagen deposition in lungs of Cyb5r3 SPC–KO mice after bleomycin determined by hydroxyproline levels. (Min to max with median, n = 4–5/group.) (C) Kaplan-Meier survival curve of Cyb5r3 SPC–KO and Cyb5r3^fl/fl mice after bleomycin (n = 8, starting mice). (D) Weight loss data in Cyb5r3^fl/fl and Cyb5r3 SPC–KO mice after bleomycin or PBS treatment. (Data points are mean ± SD, n = 4–5.) (E) Relative changes of fibrotic markers Tgfb1, Col1a1, and Fn1 mRNA levels after bleomycin in Cyb5r3^fl/fl and Cyb5r3 SPC–KO mice. (Min to max with median, n = 4–5/group.) (F) Changes in the relative transcript levels of Timp1 and Spp1 in total lung lysate at 14 days after bleomycin. (Min to max with median, n = 4–5/group.) Statistical analysis was performed using 2-way ANOVA with multiple-comparison test (B, E, and F), Mantel-Cox test (C), and 1-way repeated measures ANOVA (D), BLM versus PBS: *P < 0.05, **P < 0.01, ***P < 0.001; fl/fl versus SPC-KO: ^#P < 0.05, ^##P < 0.01, ^###P < 0.001, ^####P < 0.0001.

Figure 4. Mitochondrial intervention does not improve virus-induced lung fibrosis in Cyb5r3 SPC–KO mice.

(A) Representative Masson’s trichrome staining in lung sections from Cyb5r3 SPC–KO showing no change in collagen deposition (blue) at day 15 after MHV68 infection after osmotic pump implantation with corresponding treatment (vehicle, mitoTEMPO, NAM) at day 7. Scale bars: 500 μm. (n = 5–6/group.) (B) Treatment did not change collagen deposition in total lung lysate of Cyb5r3 SPC–KO mice after infection determined by hydroxyproline levels. (Min to max with median, n = 5–6/group.) (C) Relative changes of fibrotic markers Tgfb1, Col1a1, and Fn1 mRNA expression levels after treatments in Cyb5r3 SPC–KO MHV68-infected mice. (Min to max with median, n = 5–6/group.) (D) No detectable change of Timp1 and Spp1 transcript levels after treatments in Cyb5r3 SPC–KO MHV68-infected mice. (Min to max with median, n = 5–6/group.) Statistical analysis was performed using 1-way ANOVA with multiple-comparison test (B–D), versus vehicle: **P < 0.01.

Figure 5. Therapeutic intervention with sGC agonists ameliorates survival and fibrotic outcomes in Cyb5r3 SPC–KO mice after injury at 28 days.

(A) Representative Masson’s trichrome staining in lung sections from Cyb5r3^fl/fl and Cyb5r3 SPC–KO mice at day 28 after MHV68 infection with different interventions representing different experimental cohorts. All interventions were administered mixed in the same base chow composition as the placebo diet and started at day 7 postinfection. Scale bars: 500 μm. (n = 4–12/group; see survival details in C.) (B) Weight loss data comparing fl/fl and AECII CYB5R3–deficient mice after infection in the different treatment arms. (Data points are mean ± SD, n = 4–12). (C) Kaplan-Meier survival curve of Cyb5r3 SPC–KO and Cyb5r3^fl/fl infected mice after the treatment with the sGC stimulator (BAY 41-8543) or sGC activator (BAY 54-6544). Initial number of infected mice and final number of deaths per group are annotated beside each survival curve. (n = 19–24/group, as a result of multiple cohorts.) (D) Collagen deposition in lungs of Cyb5r3^fl/fl and Cyb5r3 SPC–KO infected mice after the different interventions determined by hydroxyproline levels. (Min to max with median, n = 4–12/group.) (E) Relative changes of fibrotic markers Col1a1 and Fn1 mRNA expression levels after treatments in Cyb5r3^fl/fl and Cyb5r3 SPC–KO MHV68-infected mice. (Min to max with median, n = 4–12/group.) (F) Change in transcript levels of Timp1 and Spp1 after treatments in Cyb5r3^fl/fl and Cyb5r3 SPC–KO MHV68-infected mice. (Min to max with median, n = 4–12/group.) Statistical analysis was performed using 1-way repeated measures ANOVA (B), Mantel-Cox test (C), and 2-way ANOVA with multiple-comparison test (D–F), as indicated: *P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001.

Figure 6. CYB5R3 deficiency in epithelial cell is associated with senescent and pro-fibrotic phenotype.

(A) Representative immunoblot to confirm KD of CYB5R3 after adenoviral infection of MLE cells. (n = 6.) (B) Transcript levels of Spp1 and cyclin-dependent kinase inhibitor 1A (Cdkn1a, encoding p21 protein) upon stimulation with TGF-β1, in Cyb5r3-Scr and Cyb5r3-KD MLE12 cells. (Min to max with median, n = 3/group.) (C) Schematic of the conditioned media (CM) experiment. (D) Representative images of murine primary lung fibroblasts cultured for 24 hours in the presence of CM derived from Cyb5r3-Scr or Cyb5r3-KD MLE12 cells previously stimulated with PBS or TGF-β1. Scale bars: 200 μm. (n = 6/condition.) (E) Change in fibroblast growth. Fibroblasts were cultured for 24 hours in the presence of CM derived from Cyb5r3-Scr or Cyb5r3-KD MLE12 cells previously stimulated with different TGF-β1 treatments compared with primary fibroblasts cultured without CM. For CM generation: TGF-β1 stimulation period was followed by extensive washing; then Cyb5r3-Scr or Cyb5r3-KD MLE12 cells were cultured with TGF-β1–free media for a total of 48 hours; then CM were harvested. (Individual data with mean ± SEM, n = 6/group.) (F) Levels of mRNA of different pro-fibrotic markers in total lysate from primary fibroblasts cultured in CM derived from Cyb5r3-Scr or Cyb5r3-KD MLE12 cells previously stimulated with TGF-β1 (PBS-stimulated CM used as reference state). (Individual data with mean ± SEM, n = 6/group.) Statistical analysis was performed using 2-way ANOVA with multiple-comparison test (B), PBS versus TGF-β1: ****P < 0.0001; Scr versus KD: ^###P < 0.001, ^####P < 0.0001. Statistical analysis was performed using 2-tailed Student’s t test (E and F): **P < 0.01, ****P < 0.0001.

Figure 7. Deficiency in CYB5R3 modifies canonical and noncanonical TGF-β1 pathways.

(A) Diminution of the TGF-β1–mediated expression of Spp1 in Cyb5r3-KD MLE12 cells stimulated with TGF-β1 in the presence of different sGC agonists. (Individual data with mean ± SEM, n = 3/group.) (B) Representative immunoblot and quantification of Smad2/3 phosphorylation in Cyb5r3-Scr and Cyb5r3-KD MLE12 cells stimulated with TGF-β1 at different time points. (Individual data with mean ± SEM, n = 3/group.) (C) Representative immunoblot and quantification of TGF-β1–dependent ERK phosphorylation in Cyb5r3-Scr and Cyb5r3-KD MLE12 cells at different time points. (Individual data with mean ± SEM, n = 3/group.) Statistical analysis was performed using 1-way ANOVA with multiple-comparison test (A), as indicated: ***P < 0.0001; ****P < 0.00001.