TGF-β1/IL-11/MEK/ERK signaling mediates senescence-associated pulmonary fibrosis in a stress-induced premature senescence model of Bmi-1 deficiency

To study whether TGF-β1/IL-11/MEK/ERK (TIME) signaling mediates senescence-associated pulmonary fibrosis (SAPF) in Bmi-1-deficient (Bmi-1−/−) mice and determines the major downstream mediator of Bmi-1 and crosstalk between p16INK4a and reactive oxygen species that regulates SAPF, phenotypes were compared among 7-week-old p16INK4a and Bmi-1 double-knockout, N-acetylcysteine (NAC)-treated Bmi-1−/−, Bmi-1−/−, and wild-type mice. Pulmonary fibroblasts and alveolar type II epithelial (AT2) cells were used for experiments. Human pulmonary tissues were tested for type Ι collagen, α-smooth muscle actin (α-SMA), p16INK4a, p53, p21, and TIME signaling by using enzyme-linked immunosorbent assay (ELISA). Our results demonstrated that Bmi-1 deficiency resulted in a shortened lifespan, ventilatory resistance, poor ventilatory compliance, and SAPF, including cell senescence, DNA damage, a senescence-associated secretory phenotype and collagen overdeposition that was mediated by the upregulation of TIME signaling. The signaling stimulated cell senescence, senescence-related secretion of TGF-β1 and IL-11 and production of collagen 1 by pulmonary fibroblasts and the epithelial-to-mesenchymal transition of AT2 cells. These processes were inhibited by anti-IL-11 or the MEK inhibitor PD98059. NAC treatment prolonged the lifespan and ameliorated pulmonary dysfunction and SAPF by downregulating TIME signaling more than p16INK4a deletion by inhibiting oxidative stress and DNA damage and promoting ubiquitin-proteasome degradation of p16INK4a and p53. Cytoplasmic p16INK4a accumulation upregulated MEK/ERK signaling by inhibiting the translocation of pERK1/2 (Thr202/Tyr204) from the cytoplasm to the nucleus in senescent fibroblasts. The accumulation of collagen 1 and α-SMA in human lungs accompanied by cell senescence may be mediated by TIME signaling. Thus, this signaling in aging fibroblasts or AT2 cells could be a therapeutic target for preventing SAPF.


Introduction
Aging drives idiopathic pulmonary fibrosis (IPF) and non-IPF lung fibrotic disorders, which are characterized by chronic activation of profibrotic factors and pulmonary parenchymal destruction and dysfunction, which leads to poor health and truncation of the lifespan 1,2 .Determining the mechanism of senescence-associated pulmonary fibrosis (SAPF) is critical for developing more-effective therapies.
The senescence-associated secretory phenotype (SASP) turns senescent fibroblasts into proinflammatory cells that induces senescence and the epithelial-to-mesenchymal transition (EMT) of nearby epithelial cells [3][4][5] .SAPF is partly characterized by dysfunction and retention of p16 INK4a (hereafter termed p16)-positive and SASP-positive fibroblasts and epithelial cells in IPF.SAPF can be ameliorated by senescent cell clearance 1 .TGF-β1, a principal profibrotic factor, predisposes the aged to the development of pulmonary fibrosis, but its inhibition is associated with side effects owing to its pleiotropic roles 6,7 .IL-11 is a downstream effector of TGF-β1 and, along with its receptor α1 (Rα1), drives noncanonical MEK-ERK-RSK signals that are required for fibrogenic protein synthesis in cardiovascular fibrosis 6 .IL-11 is upregulated 100-fold in fibroblasts from patients with IPF 8 .Senescence and the EMT of alveolar type II epithelial (AT2) cells are important in pulmonary fibrosis, leading to diminished regenerative repair of the injured epithelium owing to the EMT of AT2 cells 9 .Thus, studying whether TGF-β1/IL-11/MEK/ERK (TIME) signaling mediates SAPF by promoting the profibrotic SASP of fibroblasts and the EMT of AT2 cells is urgent.
B-cell-specific Moloney murine leukemia virus insertion region 1 (Bmi-1) is implicated in cell cycle regulation and senescence.Bmi-1 inhibits the p16/Rb and p19/p53 pathways and maintains mitochondrial function and redox balance 4,10 .Bmi-1-deficient mice are a stressinduced premature senescence (SIPS) model.These mice show persistent accumulation of reactive oxygen species (ROS) that results from impaired mitochondrial function and imbalanced redox and is sufficient to induce cell senescence via accumulated ROS and DNA damage 4,5,[11][12][13] .We reported that Bmi-1 protected against renal interstitial fibrosis mediated by TGF-β1/Smad signaling by maintaining redox balance and inhibiting overproduction of p16-positive and SASP-positive fibroblasts and the EMT of tubular epithelial cells 4,5 .However, whether Bmi-1 ameliorates SAPF by inhibiting the profibrotic SASP of fibroblasts and the EMT of AT2 cells that are mediated by TIME signals is unclear.The major downstream mediator of Bmi-1 and crosstalk between p16 and ROS in regulating SAPF is also unknown.

Mice and genotyping
Bmi-1 −/− p16 −/− , Bmi-1 −/− , and WT mice were prepared as described previously 4,5 .This study was carried out in strict accordance with the guidelines of the Institute for Laboratory Animal Research of Nanjing Medical University in Nanjing, China.The protocol was approved by the Committee on the Ethics of Animal Experiments of Nanjing Medical University (Permit Number: IACUC-1706001).

Samples of human pulmonary tissues
Human pulmonary samples were obtained from 27 autopsies at the Department of Human Anatomy in Nanjing Medical University.Anatomical methods and all experimental protocols were approved by the Committee on the Ethics of Nanjing Medical University (Permit Number: 2019-902).Body donors, aged 39-94 years old, had no tumors, acquired immune deficiency syndrome, autoimmune disease, chronic respiratory infections or inflammatory diseases before they died.

Cell cultures Pulmonary fibroblasts
Mice that were 7 weeks old were anesthetized and perfused.The lungs were separated, minced, and digested to culture pulmonary fibroblasts that were detected by immunofluorescence staining of the mesenchymal cell marker vimentin (Fig. 4e) as described previously 5,14 .Details are provided in Supporting Information 3.

Administration of drugs or reagents N-acetylcysteine
In vivo, NAC was administered at 1 mg/ml in drinking water as previously described 4,12 .

Cell proliferation
Cell proliferation was analyzed using a Cell Counting Kit-8 assay (#C0038, Beyotime Institute of Biotechnology, Shanghai, China) as previously described 5 .Details are provided in Supporting Information 3.

CM collection
Third-passage pulmonary fibroblasts were cultured in Dulbecco's Modified Eagle Medium/F12 (without phenol red; Gibco, Life Technologies Corporation, NY, USA) without FBS for 24 h.CM was collected as previously described 5 .Details are provided in Supporting Information 3.

Intracellular ROS analysis
Intracellular ROS analysis was performed as previously described 4,12 .Details are provided in Supporting Information 3.

Pulmonary function analysis
Mice that were 7 weeks old were anesthetized and underwent tracheostomies.They were mechanically ventilated at an initial baseline challenge using the FinePointe RC system (Buxco Research Systems, Wilmington, NC, USA) to directly evaluate lung ventilatory resistance and compliance, including peak inspiratory flow, frequency, tidal volume, lung resistance, dynamic compliance, minute volume, static compliance, and elastance [20][21][22] .

Preparation of pulmonary sections
Mice that were 7 weeks old were anesthetized and perfused as previously described 4 .Pulmonary samples were cut into small pieces and postfixed in periodate-lysineparaformaldehyde (PLP) solution overnight at 4 °C as previously described 23 .For histochemistry or immunohistochemistry, sections were dehydrated in a series of graded ethanol solutions, embedded in paraffin and cut into 5-μm sections using a rotary microtome (Leica Biosystems Nussloch GmbH, Nussloch, Germany) as previously described 5 .
Human frozen pulmonary samples were cut on a freezing microtome (Thermo Scientific Cryotome FSE Cryostats, Loughborough, Leicestershire) at a thickness of 7 μm for SA-β-gal staining and α-SMA immunohistochemical staining.

Histology staining
For histochemical or immunohistochemical staining, serial paraffin sections were deparaffinized and rehydrated.

SA-β-gal staining
Frozen sections of human pulmonary samples were stained by using a senescence β-Galactosidase staining kit (#C0602, Beyotime Institute of Biotechnology, Shanghai, China) according to the manufacturer's instructions and following previously described methods 25 .

Cytology staining
Cells seeded on a Lab-Tek II Chamber Slide system (Thermo Fisher Scientific Inc., Rochester, NY, USA) were fixed with PLP solution for 1 h 5,26 .

SA-β-gal staining
SA-β-gal staining of cells was performed by using the senescence β-Galactosidase staining kit (#C0602, Beyotime Institute of Biotechnology) according to the manufacturer's instructions as previously described 25 .

RNA extraction and real-time RT-PCR
RNA was extracted from the lungs of 7-week-old mice using TRIzol reagent (#15596, Invitrogen Inc.) according to the manufacturer's protocol.Levels of mRNA in pulmonary samples were quantified by real-time RT-PCR as previously described 4,5 .Primers are listed in Table 1.

Statistical analysis
All analyses were performed using GraphPad Prism software (Version 6.07; GraphPad Software Inc., San Diego, CA, USA) as previously described 6 .Measurement data are described as the mean ± SEM fold-change over the vehicle group and were analyzed by using Student's t test and one-way ANOVA to compare differences among groups.Qualitative data are described as percentages and were analyzed using chi-square tests as indicated.P values were two-sided and a P value < 0.05 was considered statistically significant 4,5 .The correlation of Gaussian distributed data was analyzed by Pearson's r, and non-Gaussian distributed data were analyzed by Spearman's r.P values were two-sided, and a P value < 0.05 was considered statistically significant.
To determine whether the increases in pulmonary TGF-β1 and IL-11 proteins were affected by other organs or tissues in Bmi-1 −/− mice, serum TGF-β1 and IL-11 were detected by ELISA.Levels of TGF-β1 and IL-11 in serum samples were not consistent with levels in lung tissue, and serum IL-11 protein levels were not altered among the four groups of mice.These results demonstrate that pulmonary fibrosis caused by TGF-β1 and IL-11 mainly occurred in cells inside lung tissue (Fig. S3).
To investigate whether IL-11 mediates the profibrotic effect of TGF-β1 by activating MEK/ERK/RSK signaling, pulmonary fibroblasts from WT mice were treated with TGF-β1, TGF-β1 plus anti-IL-11 antibody, IL-11, or IL-11 plus the MEK inhibitor PD98059.The expression of IL-11, p16 and collagen 1 significantly increased with TGF-β1 or IL-11 treatment compared with that of treatment with vehicle.Anti-IL-11 or PD98059 decreased the expression of IL-11, p16, and collagen 1 in TGF-β1-or IL-11-treated cells, suggesting that TGF-β1 or IL-11 treatment promotes fibroblast senescence.The alteration in IL-11 protein expression from these treatments was consistent with the results of p16 protein expression (Fig. 4e-h), suggesting that the profibrotic effect of IL-11 might be closely related to cell senescence.
These results suggest that activation of TIME signaling and retention of TGF-β1 and IL-11 in senescent pulmonary fibroblasts promotes the EMT of nearby AT2 cells.
To determine whether NAC mediates protein degradation of the cyclin-dependent kinase inhibitors p16, p19, p21, and p53 via ubiquitin-proteasome pathways, the proteasome inhibitor MG132 was used to inhibit proteasomes in Bmi-1 −/− fibroblasts treated with NAC.Compared with the levels in Bmi-1 −/− pulmonary fibroblasts treated with NAC, p16, and p53 significantly increased.In Bmi-1 −/− fibroblasts treated with NAC and MG132, p19 and p21 were not altered, suggesting that NAC participates in ubiquitin-proteasome degradation of p16 and p53 (Fig. 5j, k).
NAC-induced inhibition of cell senescence and TIME signaling mediates the EMT in AT2 cells in Bmi-1 −/− vs. Bmi-1 −/− p16 −/− mice To determine whether NAC treatment or p16 deletion ameliorates cell senescence-and TIME signalingmediated EMT in AT2 cells, cell senescence and EMTrelated proteins were detected.In Bmi-1-null AT2 cells, the percentages of SA-β-gal-positive cells or areas significantly increased compared with those of WT cells (Fig. 6a, b).In addition, the number of cells decreased owing to cell death (Fig. 6c).Significant increases were observed in the protein expression of α-SMA, p16, p53, p21, mature TGF-β1, IL-11, and Snail and the ratio of pERK1/2 (Thr202/Tyr204) to ERK1/2, whereas a significant decrease was observed in the expression of SFTPC (Fig. 6d, e).Significant increases were also observed in the expression of p53, TGF-β RII, and IL-11Rα1.TGF-β RII and IL-11Rα1 were mainly located in p53-labeledsenescent cells (Fig. S4).Treatment with vehicle, TGF-β1, TGF-β1 plus anti-IL-11, IL-11, or IL-11 plus PD98059 resulted in significant increases in the expression of α-SMA, p16, p53, and p21 and the ratio of pERK1/2 (Thr202/Tyr204) to ERK1/2, whereas a significant decrease was observed in the expression of SFTPC.The percentages of α-SMA-positive cells or areas significantly increased, whereas the percentages of SFTPC-positive cells or areas significantly decreased in Bmi-1-null AT2 cells compared with those of WT cells (Fig. 6f-i).NAC treatment maintained cell numbers and ameliorated cell senescence, expression of p16, p53, p21, TGF-β RII, and IL-11Rα1, and MEK/ERK signals mediated the EMT more than that of p16 deletion, although cell numbers were unchanged.NAC treatment and p16 deletion decreased the expression of mature TGF-β1, IL-11, and Snail in AT2 cells.Moreover, NAC treatment decreased the expression of IL-11 and Snail more than that of p16 deletion (Fig. 6a-i and S4).Significant increases were observed in the expression of α-SMA, p16, p53, and p21 and the ratio of pERK1/2 (Thr202/Tyr204) to ERK1/2 in the WT, Bmi-1 −/ − , Bmi-1 −/− p16 −/− , and NAC-treated Bmi-1 −/− groups compared with that of the vehicle, whereas a significant decrease was observed in the expression of SFTPC.The percentages of α-SMA-positive cells or areas significantly increased and the percentages of SFTPC-positive cells or areas significantly decreased with TGF-β1 or IL-11 treatment in the WT, Bmi-1 −/− , Bmi-1 −/− p16 −/− , and NAC-treated Bmi-1 −/− groups compared to those of the vehicle.The EMT and cell senescence of AT2 cells induced by TGF-β1 or IL-11 treatment was ameliorated by anti-IL-11 or PD98059 treatment (Fig. 6f-i), demonstrating that TIME signals mediate the EMT and cell senescence in AT2 cells.
To determine whether retention of TGF-β1 and IL-11 in senescent pulmonary fibroblasts promoted the EMT of AT2 cells, cells from WT mice were treated with conditioned medium (CM) from pulmonary fibroblasts from WT, Bmi-1 −/− , Bmi-1 −/− p16 −/− , or NAC-treated Bmi-1 −/− mice.In CM-treated Bmi-1 −/− cells, the percentages of α-SMA-positive areas were significantly increased, whereas the percentages of SFTPC-positive areas were significantly decreased, compared with those of the WT cells.CM from NAC-treated cells ameliorated these reactions more than those of p16 deletion (Fig. 6j, k).These results demonstrated that overproduction of TGF-β1 and IL-11 in senescent pulmonary fibroblasts promoted the EMT of AT2 cells.
To confirm whether the SA-β-gal activity and α-SMApositive areas in human pulmonary tissues correlated with p16 protein levels, we made frozen tissue sections of human pulmonary tissues, detected the SA-β-gal activity and α-SMA immunohistological expression, and found that the percentages of SA-β-gal-positive and α-SMApositive areas were increased in conjunction with the p16 protein levels in human pulmonary tissues (Fig. S5y-z).

Discussion
This study demonstrated that in a SIPS model of Bmi-1 deficiency, upregulated TIME signals contributed to SAPF and dysfunction by promoting cell senescence and stimulating TGF-β1 and IL-11 secretion and collagen 1 production in aging pulmonary fibroblasts and the EMT in aging AT2 cells (Fig. 8d).NAC treatment prolonged the lifespan and rescued alterations by downregulating TIME signals more than that of p16 deletion owing to its inhibition of oxidative stress and DNA damage and promotion of ubiquitin-proteasome degradation of p16 and p53.Cytoplasmic p16 retention upregulated MEK/ERK signals by inhibiting translocation of pERK1/2 (Thr202/Tyr204) from the cytoplasm to the nucleus in senescent fibroblasts.The accumulation of collagen 1 and α-SMA in human lungs accompanied by cell senescence was mediated by TIME signaling.
Previous clinical observations demonstrated that IPF was associated with aging, fibrosing interstitial pneumonia and chronic, progressive pulmonary dysfunction.These factors lead to symptoms including frailty, fatigue, weight loss, and shortened walking distance and lifespan 1,2,29,30 .The results with Bmi-1-deficient mice are consistent with these typical symptoms of IPF 4,5,12,23 .This study evaluated pulmonary functional and histological alterations to determine if Bmi-1-deficient mice showed senescenceassociated fibrosing interstitial pneumonia.We found that Bmi-1 deficiency caused ventilatory resistance and poor ventilatory compliance.Accompanied by accumulated cell senescence, cycle arrest, ROS, and DNA damage, many inflammatory cells and proinflammatory factors, including IL-1β, IL-6, and TNF-α, were abnormally accumulated in interstitial spaces around the walls of the lung air sacs (alveoli), blood vessels and small airways in Bmi-1-deficient mice.A previous study found that owing to a profound effect in the lung microenvironment of cell senescence and SASP, the secretion of a variety of mediators contributed to the development and perpetuation of fibrosis 2 .We found significant increases in interstitial fibers, collagen 1, fibronectin, α-SMA-labeled myofibroblasts, and the profibrotic factors TGF-β1 and IL-11 and a significant decrease in SFTPC-labeled AT2 cells.Thus, Bmi-1 deficiency caused SAPF in the same manner as IPF.
Recent observations revealed that IL-11, the crucial fibrosis gene acting downstream of TGF-β1, also activated ERK in cardiac fibroblasts.Both TGFβ-1 and IL-11 require ERK to induce profibrotic phenotypes 6 .IL-11 is upregulated by 100-fold in fibroblasts from patients with IPF, suggesting this gene is critical in human pulmonary fibrosis 8 .Therefore, we examined whether pulmonary fibrosis caused by Bmi-1 deficiency was associated with activation of TIME signals.We found that Bmi-1 deficiency significantly activated TIME signaling leading to SAPF, with TGF-β1 and IL-11 mainly produced by inner pulmonary cells but not in the serum.SAPF is characterized by the accumulation of SA-β-gal-positivesenescent fibroblasts, and the SASP is indeed fibrogenic 1 .Our results demonstrated that Bmi-1 deficiency caused the senescence of pulmonary fibroblasts that secreted TGF-β1 and IL-11 and produced collagen 1. IL-11 mediated the profibrotic effect of TGF-β1 and further promoted aging in senescent pulmonary fibroblasts by activating MEK/ERK signals.This process was inhibited by anti-IL-11 or the MEK inhibitor PD98059.Consistent with this result is a previous observation that PD98059 reduces lung injury and inflammation and suppresses the development of lung fibrosis in a bleomycin model 31 .Thus, IL-11/MEK/ERK signaling may be a therapeutic target for preventing SAPF.
Primary human AT2 cells undergo a TGFβ-dependent EMT phenotypic change and acquire the potential to produce collagen 1 9 .Several lines of evidence suggest that the radiation-induced EMT of AT2 cells is important in pulmonary fibrosis that is mediated by the TGF-β and ERK/glycogen synthase kinase 3β (GSK3β) pathways 32,33 .We observed that TGF-β RII and IL-11Rα1 expression was higher in senescent fibroblasts and in senescent AT2 cells compared to that of normal cells.We examined whether senescent fibroblasts, through the SASP, affected the TIME signaling-mediated EMT of adjacent senescent AT2 cells.We found that Bmi-1 deficiency also induced senescence of AT2 cells that were prone to EMT induction by CM, including secreted TGF-β1 and IL-11 from Bmi-1-null-senescent fibroblasts.IL-11 mediated the TGF-β1 effect of driving cell senescence and the EMT of AT2 cells by activating MEK/ERK signaling.This activation was inhibited by anti-IL-11 or PD98059.Therefore, overproduction of TGF-β1 and IL-11 in senescent pulmonary fibroblasts promoted the EMT and cell senescence of nearby AT2 cells to exacerbate fibrosis.
Bmi-1 normally simultaneously represses the Ink4a/Arf locus, leading to reduced p16 and p19 expression and modulating mitochondrial function and redox balance to reduce ROS levels and suppress DNA damage reaction pathway activation to limit aging 4,12 .Whether the major downstream mediator of Bmi-1 is p16 or ROS is unknown, as is their crosstalk in SAPF.A study demonstrated that p16-positive senescent fibroblasts were selectively killed by a senolytic cocktail that decreased the profibrotic effects of the SASP and improved pulmonary function and physical health in a bleomycin-injury IPF model 1 .Oxidative stress, an important molecular mechanism underlying fibrosis in pulmonary fibrosis, increases TGF-β1-induced fibrosis in part by activating latent TGF-β1.This activation, in part by inducing DNA damage, stimulates ROS production, leading to oxidative stress and results in a vicious profibrogenic circle 34,35 .NAC, a glutathione (GSH) precursor, increases pulmonary GSH levels and attenuates bleomycin-induced fibrosis 34 .We compared the effects of NAC treatment and p16 deletion on ameliorating aging-associated pulmonary fibrosis in Bmi-1-deficient mice.Our results indicated that NAC treatment was better than p16 deletion at prolonging the lifespan and downregulating TIME signaling that contribute to SAPF and dysfunction by promoting cell senescence and stimulating TGF-β1 and IL-11 secretion and collagen 1 production in aging pulmonary fibroblasts and the EMT in aging AT2 cells.We also found that NAC treatment was better than p16 deletion at decreasing the expression of p16 protein but not mRNA and exerted better anti-aging effects by inhibiting the expression of p19, p53, p21, and 8-OHdG in the lungs.NAC-mediated protein degradation of p16 and p53 via ubiquitin-proteasome pathways.Several lines of evidence demonstrate that the deletion of p16 in Bmi-1-deficient mice did not alter ROS levels, suggesting that the oxidative stress that occurs in Bmi-1-deficient mice does not result from the p16 senescence pathway upregulation 4,12 .A previous study showed that increased ROS promotes dissociation of Bmi-1 from chromatin and upregulates p16 transcription by activating MEK/ERK signaling and causing cell senescence.NAC treatment significantly downregulates MEK/ERK signaling to prevent cellular senescence 36 .Our results showed that NAC treatment significantly inhibited TGF-β1/IL-11/MEK/ERK signaling to prevent pulmonary fibrosis in Bmi-1 −/− mice.Thus, ROS might be a major downstream mediator by which Bmi-1 deficiency regulates SAPF.However, the exact regulatory mechanism of antioxidants on TIME signaling molecules remains to be investigated.

Table 1
Primer for real-time RT-PCR.

Table 2
Lifetime survival studies.