Roles of pigment epithelium-derived factor in exercise-induced suppression of senescence and its impact on lung pathology in mice

Myoblast-derived factor(s) suppressed cellular senescence

To investigate whether myokines play a role in mediating the exercise-induced suppression of cellular senescence, we initially examined the impact of factors derived from myoblasts on the replicative lifespan of cells. We cultured MEFs according to the 3T3 protocol [19]. MEFs were exposed to control media (control med) or C2C12-CM during the culture (1:1 dilution with 10% serum/medium). MEFs exposed to control med ceased their proliferation by passage 5 (Figure 1A). In contrast, when cultured in the presence of C2C12-CM, these cells exhibited a markedly higher rate of proliferation that persisted until passage 7. To clarify whether the senescence program was affected by C2C12-CM, we examined the expression of Cdkn2a (Ink4a and Arf) in these cells. The results obtained showed that the expression of both Ink4a and Arf was higher in passage 4 (P4) than in passage 2 (P2), (Figure 1B, 1C). However, MEFs exposed to C2C12-CM had lower levels of Ink4a and Arf than those exposed to control med. Consistently, cells expressed lower levels of p16^INK4a (Figure 1D) and a smaller number of cells expressed p19^ARF than control cells following exposure to C2C12-CM (Figure 1E, 1F). Additionally, p21 protein level decreased in the presence of C2C12-CM (Figure 1D). We further assessed SA-β-gal in MEFs at P4 and found that a significantly lower number of cells were positive for SA-β-gal in C2C12-exposed MEFs (Figure 1G, 1H). Collectively, these results suggest that factor(s) present in C2C12-CM possess the ability to extend the replicative lifespan of MEFs by suppressing cellular senescence.

Figure 1. Myocyte-derived factors suppress cellular senescence. (A) Wild-type MEFs were cultured according to the 3T3 protocol in the presence of a control medium or C2C12-CM (1:1 dilution with DMEM containing 10% serum). Data represent the average value of triplicate samples ± SD. (B, C) Total RNA was isolated from cells and the expression of Ink4a (B) or Arf (C) was analyzed by real-time PCR. Values were normalized to Gapdh in each sample. (D) The expression of p16^INK4a and p21 was analyzed by immunoblotting. β-Actin was used as a loading control. (E) An immunofluorescence analysis was performed using the p19^ARF antibody. Cells were counterstained with DAPI. Scale bar, 100 μm. (F) The number of p19^ARF-positive cells was counted in (E). (G) Cells (passage 4) were stained for SA-β-gal. Scale bar, 100 μm. (H) The number of SA-β-gal-positive cells was counted in each sample. Scale bar, 100 μm. Values represent means ± SD. Data were analyzed by a one-way ANOVA and Tukey’s post-hoc analysis (B, C, F) or the Student’s t-test (H). *P <0.05, **P <0.01, and ***P <0.001.

PEDF mediated the suppressive effect of C2C12-CM on cellular senescence

We attempted to identify the factor(s) within C2C12-CM that are responsible for suppressing cellular senescence. Proteins included in control med and C2C12-CM were separated by SDS-PAGE and visualized using silver staining (Supplementary Figure 1). We found distinct proteins in C2C12-CM in the range of 37-75 kD. These proteins were subsequently analyzed by mass spectrometry to identify the proteins in C2C12-CM. Among 841 proteins detected in C2C12-CM by mass spectrometry (Supplementary Table 1), 62 proteins were annotated as extracellular proteins (Supplementary Table 2). We further filtered candidate factors based on two criteria from the literature: 1) an increased blood concentration in response to exercise and 2) the potential to modify stress signals or cellular senescence. The selection of proteins identified by mass spectrometry based on these criteria resulted in the selection of PEDF (also known as Serpinf1 or Epc-1) as a candidate protein. Previous studies demonstrated that the expression of PEDF was induced during the myotube differentiation of human myoblasts [20] and its concentration increased in response to persistent training or exercise in both humans and mice [21, 22]. Moreover, PEDF has been shown to delay cellular senescence by reducing oxidative stress in human mesenchymal stem cells [23], and its deletion accelerated senescence in the retinal pigment epithelium of mice [24]. In the present study, we observed a significant decrease in Pedf expression in both the tibialis anterior (TA) and soleus (SOL) muscles of aged animals (Supplementary Figure 2). Therefore, PEDF was subjected to further analyses in the context of senescence and myokines in C2C12-CM.

To investigate the potential role of PEDF in the suppression of cellular senescence, MEFs at P3 were cultured in the presence of C2C12-CM pre-treated with either a control or neutralizing antibody against PEDF. After 3 days, we observed an increase in the population of MEFs with the control antibody (Figure 2A). The addition of anti-PEDF significantly inhibited cell proliferation. Additionally, anti-PEDF increased the expression of Cdkn2a (Ink4a and Arf) and the encoded proteins, p16^INK4a and p19^ARF (Figure 2B, 2C).

Figure 2. PEDF mediates the anti-cellular senescence effects of C2C12-CM. (A–C) MEFs were cultured in the presence of C2C12-CM treated with a control or PEDF antibody for 3 days. (A) Cell numbers were counted and relative changes in cell numbers in 3 days were plotted. (B) Total RNA was isolated from MEFs and the expression of Ink4a and Arf was analyzed by real-time PCR. Values were normalized to Gapdh in each sample. (C) p16^INK4a and p19^ARF levels were analyzed by immunoblotting. β-Actin was used as the loading control. (D) MEFs were cultured in the presence of a recombinant of PEDF (100 ng/mL) for 3 days. Changes in cell numbers were plotted. (E) Cell viability was determined by the trypan blue exclusion assay. (F) The expression of Ink4a, Arf, and p21 was analyzed by real-time PCR. Values were normalized to Gapdh in each sample. (G) p16^INK4a, p19^ARF, and p21 levels were analyzed by immunoblotting. β-Actin was used as a loading control. (H) Cells were stained for SA-β-gal. Scale bar, 100 μm. (I) The percentage of SA-β-gal-positive cells was plotted. (J) Cells were stimulated with the indicated concentrations of recombinant PEDF for 3 days. Intracellular ROS levels were analyzed in each sample, and relative values were plotted against the average of the control sample. Values represent means ± SD. Data were analyzed by the Student’s t-test (A, B, D–F, I) or a one-way ANOVA and Tukey’s post-hoc analysis (J). *P <0.05, **P <0.01, and ***P <0.001.

To further validate the effects of PEDF on cellular senescence, we investigated whether treatment with recombinant PEDF effectively inhibited senescence in MEFs. After 3 days, an increase in cell numbers was observed in PEDF-treated samples without any noticeable change in cell viability (Figure 2D, 2E). PEDF also significantly suppressed the expression of Ink4a, Arf, and p21 as well as their corresponding proteins (Figure 2F, 2G) and reduced the number of SA-β-gal-positive cells (Figure 2H, 2I). These results strongly support the crucial role of PEDF in mediating the anti-cellular senescence effect of C2C12-CM.

We then examined PEDF signaling pathways in MEFs. Although the detailed molecular mechanisms mediating the PEDF signal remain unclear, it has been observed to lead to the activation of AKT, ERK, and p38, while down-regulating the Wnt/β-catenin signal [25, 26]. Additionally, PEDF has been reported to bind to and down-regulate the expression of the ATGL protein by promoting its ubiquitin/proteasome-dependent degradation [27]. However, in the present study, the PEDF treatment did not affect the phosphorylation of AKT, ERK, or p38 or the protein level of β-catenin or ATGL in MEFs (Supplementary Figure 3). Apart from its effects on these signaling pathways, PEDF has been shown to reduce oxidative stress, which is considered to contribute to the PEDF-mediated suppression of cellular senescence [23]. In our experiments, we observed a dose-dependent reduction in intracellular ROS levels (Figure 2J), suggesting that PEDF extended the replicative lifespan of MEFs by mitigating oxidative stress.

Exercise stimulated PEDF expression while concurrently reducing cellular senescence

Due to the potential of PEDF as a myokine and its ability to delay cellular senescence in vitro, we were prompted to investigate its effects in a mouse exercise model. Since senescent cells become detectable in lung tissues as early as 6 months old [28], and their elimination has been shown to induce protective effects in lung disease models [10, 11], we conducted an 8-week voluntary wheel running regimen for 6-month-old mice. Control mice were maintained in cages without a running wheel. During this regimen, mice maintained an average daily running distance of more than 2 km without a significant change in their body weights (Supplementary Figure 4A, 4B). The expression of Pedf was increased in TA and SOL (Figure 3A), and serum PEDF levels were also elevated in the exercise group (Figure 3B). These results are consistent with previous findings indicating the increased expression of PEDF in human muscle and elevated serum PEDF levels in mice in response to training [21, 22].

Figure 3. Voluntary exercise up-regulates PEDF and suppresses cellular senescence. Six-month-old wild-type mice were subjected to voluntary exercise for 8 weeks. (A) Total RNA was isolated from skeletal muscles (TA and SOL) and Pedf levels were analyzed by real-time PCR. Values were normalized to 18S rRNA in each sample. (B) Serum PEDF levels were analyzed by ELISA. (C) The expression of Ink4a, Arf and p21 in the lungs was analyzed by real-time PCR. Values were normalized to Gapdh in each sample. (D) Lung sections were stained for SA-β-gal. Arrowheads indicate SA-β-gal-positive cells. Scale bar, 100 μm. (E) The number of SA-β-gal-positive cells was counted in each sample. (F) The expression of Il-1b, Il-6, and Mmp-12 in lung total RNA was analyzed by real-time PCR. Values were normalized to 18S-rRNA in each sample. Values represent means ± SEM. Data were analyzed by the Student’s t-test. *P <0.05, **P <0.01, and ***P <0.001.

Habitual physical activity and long-term exercise training have been shown to diminish the expression of p16^INK4a in T lymphocytes among humans, while voluntary exercise has been associated with a reduction in senescence markers, including Cdkn2a and SA-β-gal, across various tissues, including the heart, liver, muscles, kidneys, and fat, in mice [16]. We also observed that the expression of senescence markers, including Ink4a, Arf, p21, and SA-β-gal, in lung tissues was lower in the exercise group than in the control group (Figure 3C–3E). Similarly, the expression of SASP-related genes, such as Interleukin (Il)-1b, Il-6, and Matrix metalloproteinase (Mmp)-12 [28], in the lungs was significantly decreased in the exercise group (Figure 3F). These results strongly suggest the potential of exercise training to reduce cellular senescence in the lungs. Although a slight change was noted in the cell composition of bronchoalveolar lavage fluid (BALF, Supplementary Figure 4C–4F), we did not observe a difference in pulmonary function between the groups (Supplementary Figure 4G–4I). Therefore, the present results were consistent with our previous findings showing that the genetic elimination of lung senescent cells at this age did not affect lung function [10].

The administration of PEDF reduced cellular senescence in tissues

We investigated the relationship between elevated PEDF levels and the observed reduction in cellular senescence following exercise training. Mice were administered a recombinant PEDF protein for 4 weeks. Notably, Ink4a, Arf, and p21 in lung tissues were decreased in the PEDF-treated group (Figure 4A). We also observed a marked reduction in the number of SA-β-gal-positive cells in lung tissues in the PEDF-treated group (Figure 4B, 4C). We then expanded our analysis to other tissues. Ink4a, Arf, and p21 were decreased in adipose tissue in the PEDF-treated group, (Supplementary Figure 5A), and SA-β-gal activity was significantly reduced following the PEDF treatment (Supplementary Figure 5B).

Figure 4. PEDF suppresses cellular senescence in lung tissues. A recombinant PEDF protein (10 μg/kg body weight) was intraperitoneally injected twice a week for 4 weeks. (A) The expression of Ink4a, Arf, and p21 in lung total RNA was analyzed by real-time PCR. Values were normalized to Gapdh in each sample. (B) Lung tissues were stained for SA-β-gal. Arrowheads indicate SA-β-gal-positive cells. Scale bar, 50 μm. (C) The number of SA-β-gal-positive cells was counted in each sample. Values represent means ± SEM. Data were analyzed by the Student’s t-test. *P <0.05 and **P <0.01.

To gain further insights into the role of PEDF in suppressing senescence during exercise, we examined the expression of the Il-1b, Il-6, and Mmp-12 genes. While the expression of all of these genes in lung tissue decreased in response to exercise (Figure 3F), we specifically noted the significant down-regulation of Mmp-12 expression in both lung and fat tissues of PEDF-treated animals (Supplementary Figure 6). Collectively, these results strongly suggest the ability of PEDF to reduce cellular senescence in multiple tissues.

PEDF ameliorated the pathology of emphysema in mice

We previously demonstrated that the semi-genetic or pharmacological elimination of senescent cells in the lungs protected against pulmonary emphysema induced by elastase or cigarette smoking in 6-month-old mice [10, 11]. Based on the ability of PEDF to suppress cellular senescence in lung tissue in mice, we investigated whether it also exerted protective effects in a mouse emphysema model. Mice were pre-administered with a recombinant PEDF protein and were then subjected to an elastase (PPE) treatment in order to induce pulmonary emphysema, as shown in Figure 5A. While PPE does not affect the Ink4a and Arf levels in the lung tissues [10], senescence markers were decreased in the lung tissues of the PEDF-treated group, indicating that cellular senescence was also diminished following the PEDF treatment in the elastase-induced emphysema model (Figure 5B). A morphometric analysis of inflated lung sections revealed that PPE induced a massive alveolar collapse (compare PBS/PBS and PBS/PPE in Figure 5C, 5D), which was diminished by pre-administration of PEDF, suggesting that PEDF suppressed PPE-induced alveolar collapse. Consistently, a spirometric analysis demonstrated that the PEDF treatment significantly protected against PPE-induced pulmonary dysfunction (Figure 5E–5H). The administration of PEDF did not affect the cell composition of BALF in the PPE-induced emphysema model (Supplementary Figure 7), which was consistent with our previous findings showing that PPE only induced the transient accumulation of inflammatory cells following its installation [10]. Collectively, these results suggest that the PEDF treatment effectively protects against pulmonary emphysema by inhibiting cellular senescence.

Figure 5. PEDF protected lung tissues from PPE-induced emphysema. (A) Experimental design. A recombinant PEDF protein was injected intraperitoneally twice a week for 4 weeks. PPE (5 units) was intranasally administered 2 weeks after the first dose of PEDF. Mice were analyzed 3 weeks after the PPE treatment. (B) Total RNA was isolated from lung tissues and the expression of Ink4a, Arf, and p21 was analyzed by real-time PCR. Values were normalized to Gapdh in each sample. (C) Representative images of control and PEDF-treated mouse lung sections. Sections were stained with hematoxylin and eosin. Scale bar, 100 μm. (D) Alveolar mean linear intercepts were measured. (E–H) A pulmonary function test was performed. Lung compliance (E), capacity (F), tissue damping (G), and tissue elastance (H) are shown. Values represent means ± SEM. Data were analyzed by the Student’s t-test (B), or a one-way ANOVA and Tukey’s post-hoc analysis (D–H). *P <0.05, **P <0.01, and ***P <0.001.

To clarify the relevance of PEDF in human pathology, we investigated whether serum PEDF levels correlated with respiratory function in patients diagnosed with COPD (Supplementary Table 3). Serum PEDF levels in 20 patients were measured by ELISA and respiratory function parameters were assessed by the forced expiratory volume in 1 second per forced vital capacity of the lung (FEV1/FVC). A weak correlation was observed between serum PEDF level and respiratory function through linear regression analysis (r=0.48 and P=0.03, Supplementary Figure 8), suggesting a potential involvement of PEDF in the pathology of COPD in humans.

Potential of miR-127 as a mediator of the PEDF-induced inhibition of cellular senescence

The results presented above strongly indicated the potential of PEDF to suppress cellular senescence, thereby mediating the beneficial effects of exercise training in mice. Although we demonstrated that PEDF reduced intracellular ROS levels in MEFs (Figure 2J), the underlying mechanisms remain unclear. Therefore, we attempted to identify factor(s) that potentially mediate the function of PEDF. We performed an RNA sequencing analysis of MEFs (P3) stimulated or not with recombinant PEDF. We identified 40 differentially expressed genes, including 16 up-regulated genes and 24 down-regulated genes, between control and PEDF-treated cells (Figure 6A and Supplementary Table 4). Of these genes, miR-127, which was down-regulated by PEDF in MEFs, was previously shown to increase during senescence and promote cellular senescence in mouse and human fibroblasts [29, 30]. The PEDF-mediated down-regulation of miR-127 in MEFs was further confirmed by real-time PCR (Figure 6B). miR-127 directly targets BCL-6, and its down-regulation has been shown to contribute to the induction of senescence by miR-127 [31]. Therefore, we examined BCL-6 expression levels in MEFs. BCL-6 protein expression levels were higher in PEDF-treated MEFs than in control cells (Figure 6C), suggesting a role for the miR-127-BCL-6 axis in the regulation of senescence by PEDF.

Figure 6. Involvement of miR-127 in PEDF signaling. (A) An RNA sequencing analysis was performed on MEFs stimulated with a recombinant PEDF protein. A volcano plot of RNA sequencing data is shown. A full list of RNA sequencing data is available in the Gene Expression Omnibus database (accession number; GSE241459). (B) The expression of miR-127 (-loop, -5p, and -3p) in MEFs was analyzed by real-time PCR. Values were normalized to U6 snRNA in each sample. (C) The expression of BCL-6 in MEFs was analyzed by immunoblotting. β-Actin was used as a loading control. (D, E) miR-127 levels in the lungs of mice subjected to the PEDF treatment (D, Figure 4) or voluntary wheel running (E, Figure 3) were analyzed by real-time PCR. miR-127 levels were normalized to U6 snRNA in each sample. Values represent means ± SD (B) or means ± SEM (D, E). Data were analyzed by the Student’s t-test. *P <0.05, **P <0.01, and ***P <0.001.

We subsequently examined miR-127 levels in the lung tissues of PEDF-treated animals. A slight decrease was observed in miR-127 levels in PEDF-treated animals, which indicated that PEDF exhibited the ability to down-regulate miR-127 in vivo (Figure 6D). Notably, miR-127 levels were significantly reduced in the lungs of mice subjected to voluntary exercise for 8 weeks (Figure 6E). Collectively, these results strongly suggest that PEDF is intricately linked to the anti-senescence effects of exercise through its suppression of miR-127.

Animals

Six-month-old female C57BL/6J mice were purchased from Japan SLC Inc. and maintained under specific pathogen-free conditions with a 12-h light/dark cycle, constant temperature, and ad libitum access to food (CRF-1, Oriental Yeast Co., Ltd.) and drinking water. As the exercise model, mice were independently housed in a cage with a 15-cm running wheel with a rotation sensor (ML-321, Marukan Co., Ltd.).

Regarding the administration of PEDF, a recombinant PEDF protein (PKSM040786, Elabscience Biotechnology Inc.) diluted with phosphate-buffered saline (PBS) was intraperitoneally injected at a dose of 10 μg/kg body weight twice a week for 4 weeks.

The pulmonary emphysema model was prepared as previously described [62]. Five units of porcine pancreatic elastase (PPE) (EC134, Elastin Products Company, Inc.) in 100 μl PBS was intranasally administered using a standard pipette tip.

Cell culture

Mouse embryonic fibroblasts (MEFs) were prepared from C57BL/6J embryos (embryonic day 13.5). MEFs were maintained in Dulbecco’s Modified Eagle Medium (DMEM, 045-30285, Fujifilm Corp.) supplemented with 10% fetal bovine serum (FBS), non-essential amino acids (11140050, Thermo Fisher Scientific Inc.), 10 μg/mL gentamicin (15710064, Thermo Fisher Scientific Inc.), and 55 μM 2-mercaptoethanol. MEFs were cultured in atmospheric O₂ (20%) and 5% CO₂ according to the 3T3 protocol [19]. Cell viability was evaluated by the trypan blue exclusion test.

C2C12 cells were maintained in DMEM supplemented with 10% FBS and gentamicin. To induce terminal differentiation, cells were cultivated in DMEM supplemented with 2% horse serum and Insulin-Transferrin-Selenium (41400045, Thermo Fisher Scientific Inc.) for 2 days. Conditioned media were then collected from differentiated C2C12 cells, filtered, aliquoted, and stored at -80°C until used. Conditioned media were diluted with an equal volume of MEF medium for use on MEFs. To neutralize PEDF, the conditioned media were incubated with anti-PEDF (5 μg/mL, PAB972Mu02, Cloud-Clone Corp.) at room temperature for 1 hour before use.

Mass spectrometry

Proteins in control media and conditioned media prepared from differentiated C2C12 myoblasts (C2C12-CM) (200 μL) were analyzed by mass spectrometry. Albumin was removed using the ProMax Albumin Removal kit (24351, Polysciences Inc.). Samples were reduced with 20 mM Tris (2-carboxyethyl) phosphine at 80°C for 10 minutes, followed by alkylation with 30 mM iodoacetamide at room temperature for 30 minutes. Proteins were digested with 500 ng Trypsin/Lys-C Mix (V5071, Promega Corp.) at 37°C overnight.

Extracted peptides were then separated via nanoLC (Ultimate 3000 RSLCnano LC system, Thermo Fisher Scientific Inc.), and the LC eluent was coupled to an electrospray-ionization source attached to the Q Exactive HF-X mass spectrometer (Thermo Fisher Scientific Inc.). MS/MS spectra were analyzed using Scaffold DIA (Proteome Software, Inc.). Significance thresholds were selected based on the peptide false discovery rate (FDR) and protein FDR, and data were considered to be significant when each showed a value less than 0.01.

Real-time PCR

Tissue samples were collected after the pulmonary function tests (described below), snap-frozen, and stored at -80°C until used. TA and SOL were collected from the hind limb, and the right lung lobes were used for the RNA analysis. Total RNA was extracted using TRI Reagent® (TR118, Molecular Research Center, Inc.) according to the manufacturer’s instructions. RNA was reverse-transcribed using the PrimeScript RT Reagent Kit with gDNA Eraser (RR047A, Takara Bio Inc.). Real-time PCR was performed on a CFX Connect Real Time System (Bio-Rad Laboratories, Inc.) using the following primers: Arf, 5’-GCCGCACCGGAATCCT-3’ (sense) and 5’-TTGAGCAGAAGAGCTGCTACGT-3’ (antisense); Gapdh, 5’-AATGGTGAAGGTCGGTGTG-3’ (sense) and 5’-GAAGATGGTGATGGG CTTCC-3’ (antisense); Il-1b, 5’-GAATGCCACCTTTTGACAGTG-3’ (sense) and 5’-CTGGATGC TCTCATCAGGACA-3’ (antisense); Il-6, 5’-TAGTCCTTCCTACCCCAATTTCC-3’ (sense) and 5’-TTGGTCCTTAGCCACTCCTTC-3’ (antisense); Ink4a, 5’-CCCAACGCCCCGAACT-3’ (sense) and 5’-GCAGAAGAGCTGCTACGTGAA-3’ (antisense); p21, 5’-CGAGAACGGTGGAACTTTGAC-3’ (sense) and 5’-CAGGGCTCAGGTAGACCTTG-3’ (antisense); Mmp-12, 5’-CTGCTCCCATGAATGAC AGTG-3’ (sense) and 5’-AGTTGCTTCTAGCCCAAAGAAC-3’ (antisense); Pedf, 5’-CCAACTTC GGCTACGATCTGT-3’ (sense) and 5’-TCTGTTCGATGTTCAGCTCCC-3’ (antisense); 18S rRNA, 5’-AGTCCCTGCCCTTTGTACACA-3’ (sense) and 5’-GATCCGAGGGCCTCACTAAAC-3’ (antisense).

MicroRNA was analyzed using the Mir-X™ miRNA qRT-PCR TB Green® Kit (Z8314N, Takara Bio Inc.). The following primers were used to detect miR-127: miR-127 loop, 5’-GGCTCTGATTCAGAAAGATC-3’; miR-127-5p, 5’-CTGAAGCTCAGAGGGCTCTGAT-3’; miR-127-3p. 5’-TCGGATCCGTCTGAGCTTGGCT-3’.

RNA sequencing

Total RNA was extracted from MEFs cultured in the presence or absence of 100 ng/mL of PEDF for 24 hours using NucleoSpin RNA (740955, MACHEREY-NAGEL Inc.) according to the manufacturer’s instructions. The quality of RNA was assessed using an Agilent 2100 Bioanalyzer (Agilent Technologies, Inc.). mRNA was isolated from total RNA using the NEBNext® Poly(A) mRNA Magnetic Isolation Module (E7490, New England Biolabs, Inc.). Libraries were generated using the NEBNext® Ultra™II Directional RNA Library Prep Kit (E7760, New England Biolabs, Inc.) and sequenced on Illumina NovaSeq 6000 (Illumina, Inc.) at Rhelixa, Inc. Raw paired-end sequence reads were evaluated for quality using FastQC (Version 0.11.7; https://www.bioinformatics.babraham.ac.uk/projects/fastqc/) [63]. Trimmomatic software (Version 0.38) was employed to trim low-quality bases (<20) and remove adapter sequences [64]. The specific parameters used for this process were as follows: ILLUMINACLIP: path/to/adapter.fa:2:30:10, LEADING:20, TRAILING:20, SLIDINGWINDOW:4:15, MINLEN:36. Processed reads underwent alignment to the reference genome using the RNA-seq aligner HISAT2 (Version 2.1.0) [65]. The resulting .sam files from HISAT2 were then converted to the .bam format using Samtools (Version 1.9). The featureCounts tool (Version 1.6.3) was then applied to .bam files in order to estimate the abundance of uniquely mapped reads [66]. Raw read counts underwent normalization using relative log normalization, and a differential expression analysis was performed using DESeq2 (Version 1.24.0) [67]. Full data are available from the Gene Expression Omnibus database of the National Center for Biotechnology Information (accession number: GSE241459).

Immunoblotting

Lysates were prepared using a buffer containing 10 mM Na-phosphate (pH 7.2), 150 mM NaCl, 2 mM EDTA, 0.1% SDS, 1% Na-deoxycholate, and 1% NP-40 and protease inhibitors. Lysates were separated by SDS-PAGE and transferred to PVDF membranes (Merck KGaA). Proteins were detected with antibodies to p19^ARF, p21, BCL-6, phospho-ERK1/2 (sc-32748, sc-6246, sc-7388, and sc-136521; all from Santa Cruz Biotechnology, Inc.), β-Actin, β-catenin, phospho-Akt (ser473), phospho-p38 (#12620, #9562, #4060, and #9211; all from Cell Signaling Technology, Inc.), adipose triglyceride lipase (ATGL) (55190-1-AP, Proteintech, Inc.), and p16^INK4a [68]. Antibodies were diluted to 1:200 (BCL-6), 1:500 (p16^INK4a, p19^ARF, p21, phospho-Akt (ser473), phospho-ERK1/2, and ATGL), 1:1000 (β-catenin, phospho-p38), or 1:2000 (β-Actin) with a blocking buffer containing 5% skim milk, 0.1% Tween 20 in Tris-buffered saline (pH 7.5, TBS), or a blocking buffer containing 4% bovine serum albumin, and 0.1% Tween 20 in TBS for phospho-Akt, phospho-ERK1/2, and phospho-p38 antibodies.

Morphometry

The lungs were inflated by injecting Mildform®20N (Fujifilm Corp.) and fixed for 10 minutes at constant pressure (25 cmH₂O) [62]. Paraffin-embedded tissues were sectioned (thickness of 5 μm) and stained with hematoxylin and eosin. At least eight randomly selected fields per mouse were photographed. Test lines were randomly drawn on the images, and intercepts with the tissue structure were counted for each line. Airway and vascular structures were eliminated from the analysis.

Immunofluorescence

MEFs were seeded onto coverslips and fixed with 4% paraformaldehyde for 20 minutes. Cells were permeabilized in 0.5% Triton X-100 and immunostained using p19^ARF (1:200 dilution, sc-32748, Santa Cruz Biotechnology, Inc.). The sites of antibody binding were visualized using Cy3-conjugated anti-rat IgG (1:200 dilution, 712-165-153, Jackson ImmunoResearch Inc.) and mounted using DAPI Fluoromount-G® (0100-20, Southern Biotechnology Associates Inc.). Fluorescence images were acquired using a fluorescence microscope (BX-X710, Keyence Corp.).

Senescence-associated β-galactosidase (SA-β-gal) staining

Frozen 10-μm-thick lung sections were fixed with 4% paraformaldehyde for 20 minutes and stained using the Cellular Senescence Detection Kit-SPiDER-bGal (SG03, Dojindo Molecular Technologies, Inc.). Sections were mounted using DAPI Fluoromount-G®.

Adipose tissue (1-cm squares) and MEFs were stained using an SA-β-gal staining Kit (#9860, Cell Signaling Technology, Inc.). In adipose tissue staining, images in the RGB format were analyzed using ImageJ. The blue-green signal intensity per tissue area was quantified in each sample.

ROS assay

MEFs were seeded on a 96-well plate (10,000 cells/well) and treated with PEDF for 24 hours. Intracellular ROS levels were analyzed using a ROS Assay Kit - Highly Sensitive DCFH-DA- (R252, Dojindo Molecular Technologies, Inc.) according to the manufacturer’s instructions. Fluorescence intensity was measured by a microplate reader (the VICTOR Nivo Multimode Microplate Reader, PerkinElmer Inc.).

Pulmonary function tests

Pulmonary function tests were performed using a flexiVent system (Emka Technologies, Inc.) as previously reported [28]. Mice were euthanized by an intraperitoneal injection of pentobarbital sodium (100 mg/kg body weight) and connected to the flexiVent system. They were ventilated at a respiratory rate of 159 breaths per minute with a tidal volume of 10 mL/kg against a positive end-expiratory pressure of 3 cmH₂O. Deep inflation, Snapshot-150, Quickprime-3, and a pressure-volume loop with a constantly increasing pressure were consecutively performed 3 times in each mouse.

Analysis of bronchoalveolar lavage fluid (BALF) cells

BALF cells were analyzed as previously reported [10]. In brief, BALF cells were prepared with 1 mL of PBS containing 5 mM EDTA, and cells were collected from BALF by mild centrifugation. Collected cells were attached to glass slides using StatSpin Cytofuge (Beckman Coulter, Inc.) and subjected to modified Giemsa staining with a Diff-Quick stain kit (Sysmex Corp.).

Enzyme-linked immunosorbent assay

Mouse blood samples were collected from the abdominal vena cava prior to conducting pulmonary function tests using a 24G needle (a minimum of 0.3 mL blood sample was obtained from each mouse). Samples were analyzed using a Mouse PEDF, PEDF ELISA Kit (CSB-E08820m, Cusabio Technology, LLC). Human sera (information including age, sex, and pathology are listed in Supplementary Table 3) were purchased from ProteoGenex Inc. and PEDF levels were analyzed using a Human PEDF ELISA kit (RD191114200R, BioVender R&D).

Statistical analysis

A two-tailed unpaired Student’s t-test was used to compare two sets of experimental data. A one-way ANOVA was performed to compare more than two sets of data. When the statistical model was proven to be significant, differences between combinations of the two groups were analyzed using the Tukey-Kramer test. Data were analyzed by GraphPad Prism 7.0 (GraphPad Software Inc.). Significance was represented by asterisks as follows: *P <0.05, ** P <0.01, and *** P <0.001. No statistical method was used to select the sample size.