CILP attenuates pulmonary fibrosis through the TGF-β1/SMAD3 pathway

Abstract

Idiopathic pulmonary fibrosis (IPF) is a chronic progressive disease with poor survival, which is characterized by abnormal accumulation of fibrotic tissue in the lung parenchyma. Transforming growth factor-β1 (TGF-β) is a central profibrotic mediator, but the related mechanism of the activation of latent TGF-β has not been conclusively elucidated. A comprehensive study of mRNAs in human IPF was conducted using GSE10667 microarray data from GEO database, and the expression of cartilage intermediate layer protein (CILP) was upregulated among end-stage pulmonary fibrosis (EPF) and acute pulmonary fibrosis (APF) as compared to non-fibrosis tissues. Furthermore, CILP has protein–protein interactions with TGF-β1 through PPI analysis. Therefore, we investigated the potential effects and mechanisms of CILP in pulmonary fibrosis in pulmonary fibroblasts and BLM-induced mouse model (Eight-week-old male C57BL/6 mice, 20–22 g, purchased from the Experimental Animal Center of Guangzhou Medical University). In vitro, treatment with recombinant CILP (100 ng/mL) significantly attenuated TGF-β1-induced upregulation of collagen type I (Col1a1, p < 0.01) and α-smooth muscle actin (α-SMA, p < 0.01) in primary mouse pulmonary fibroblasts. Mechanistically, CILP suppressed TGF-β1-mediated SMAD3 phosphorylation (p-SMAD3, p < 0.001) and nuclear translocation, as confirmed by Western blotting and immunofluorescence. In the bleomycin (BLM)-induced mouse model of pulmonary fibrosis, intravenous administration of CILP (1 μg/g body weight, administered every 2 days for 4 weeks) reduced lung collagen deposition (Masson staining) by 38% (p < 0.01), lowered Ashcroft scores (from 5.8 ± 0.7 to 2.3 ± 0.4, p < 0.001), and decreased lung hydroxyproline content (a marker of collagen accumulation) by 42% (p < 0.01) compared to BLM-only controls. Clinically, serum CILP levels showed no significant difference between 17 idiopathic pulmonary fibrosis (IPF) patients and 17 non-fibrotic controls (3.2 ± 0.8 ng/mL vs. 3.5 ± 0.9 ng/mL, p > 0.05), suggesting potential lung tissue-specific action of CILP with minimal systemic off-target risk. In conclusion, CILP inhibited TGF-β1-induced fibrosis via its negative feedback loop, and may act as a promising candidate for the precaution and treatment of IPF.

Supplementary Information

The online version contains supplementary material available at 10.1186/s40001-025-03644-7.

Introduction

Idiopathic pulmonary fibrosis (IPF) is a chronic, progressive interstitial lung disease that is highly prevalent among the elderly. The global impact of this disease is estimated to affect approximately three million individuals [1]. It is characterized by progressive respiratory aggravation and decreased lung function, which can lead to respiratory failure and poor prognostic outcomes (median survival, 3–4 years), and is associated with a heavy economic burden on patients as well as on society [2]. The main pathological features of IPF include significant myofibroblast accumulation and excess extracellular matrix (ECM) deposition in the lung interstitium, leading to destruction of the alveolar structure, gas exchange interruption and death due to respiratory failure [3]. Recently, various pharmacotherapeutic options for IPF have emerged. Nintedanib and pirfenidone provide patients with flexible therapeutic options [4]. Meanwhile, new ways of administering drugs provide effective methods for exerting their effects, helping alleviate disease and prolong survival [5, 6]. These results indicate that interventions such as drugs can significantly influence the clinical outcome of IPF. It is of critical importance to explore the pathological changes of IPF to develop potential new therapeutic strategies for improving patient outcomes.

The pathogenesis of IPF has not been definitively determined, but is thought to involve chronic or repetitive micro-injuries of the alveolar epithelium [7]. Various profibrotic mediators and signaling pathways are involved in the pathogenesis of pulmonary fibrosis. As a key profibrotic cytokine, transforming growth factor-beta 1 (TGF-β1) promotes fibroblast-to-myofibroblast differentiation and their subsequent accumulation at fibroblastic foci, driving excessive ECM production and deposition [8]. Mechanistically, TGF-β1 induces SMAD3 phosphorylation and nuclear translocation, thereby activating SMAD-dependent gene transcription that promotes both fibrotic responses and ECM synthesis [9].TGF-β1-induced ECM synthesis plays a key role in lung tissue scar repair. However, excess ECM deposition leads to lung tissue fibrosis and IPF. Therefore, elucidation of mechanisms that promote TGF-β1 signaling and the pathological effects of inhibiting TGF-β1 pathway will be of vital significance.

Cartilage intermediate layer protein (CILP) is a monomeric glycoprotein in the ECM [10]. CILP is highly localized in the middle (intermediate) zone of human articular cartilage, as well as in the meniscus, tendon, ligament, synovial membrane, and intervertebral disc [11]. In cardiac fibrosis, CILP is a negative feedback regulator for TGF-β1 signaling and ECM deposition. In the negative feedback loop, TGF-β1 induces CILP production, and CILP inhibits TGF-β1-mediated fibrosis progression [12]. However, the efficacies and pharmacological properties of CILP in IPF treatment have not been conclusively determined. In this study, we evaluated the therapeutic effects of CILP on IPF and its possible mechanisms. It was established that CILP attenuates bleomycin-induced pulmonary fibrosis by suppressing the TGF-β1 signaling pathway. Our results will provide a theoretical basis for clinical development and applications of CILP in pulmonary fibrosis treatment.

Materials and methods

Microarray data and bioinformatics analyses

GSE10667 microarray data were retrieved from the Gene Expression Omnibus (GEO) database (http://www.ncbi.nih.gov/geo). Twenty-three samples were obtained from surgical remnants of biopsies or lungs explanted from patients with IPF who underwent pulmonary transplant, eight from patients with IPF whose pathological findings were indicative of acute exacerbations, and 15 control normal lung tissues obtained from the disease-free margins with normal histology of lung cancer resection specimens. The morphologic diagnosis of IPF was based on typical microscopic findings consistent with usual interstitial pneumonia. Raw data were downloaded as MINiML files. The R package was used to identify differentially-expressed genes in this study. After linear model fitting and Bayesian testing, the genes with adjusted p < 0.05 and |log₂ (Fold Change)|> 2.0 were regarded as differentially-expressed genes. Conversion of probes to gene symbols was based on the GPL4133 annotation information of normalized data in the platform. To determine the functions of selected mRNAs, gene ontology (GO) enrichment analyses were conducted using the DAVID tool (http://david.abcc.ncifcrf.gov). Protein–protein interaction (PPI) network analysis was conducted using the STRING tool (www.string‐db.org), and the visualization analysis of core genes was completed using Cytoscape 3.9.1 software (https://cytoscape.org/).

Animals

All animal experiments were approved by the Ethics Committee for Animal Research of Guangzhou Medical University (LAEC-2020-030). Eight-week-old male C57BL/6 mice were purchased from the Experimental Animal Center of Guangzhou Medical University, housed and cared for in a pathogen-free facility at Affiliated Qingyuan Hospital of Guangzhou Medical University Experimental Animal Center. Bleomycin (BLM, Sigma-Aldrich, St. Louis, MO, USA) was dissolved in normal saline and abdominally injected (40 μg/g) at 0, 2, 4, 6 and 8 days. Recombinant human CILP protein (CRK PHARMA, Wuhan, China) was dissolved in normal saline and intravenously administered (1 μg/g) via the tail vein at 0, 2, 4, 6, 8, 12, 16, 20 and 24 days. Control mice were injected with the same amount of normal saline. Thirty-six mice were randomized into six groups (n = 6): control group, BLM (40 μg/g) group 2 weeks, BLM group 4 weeks, BLM group 6 weeks, BLM + CILP (1 μg/g) 4 weeks, BLM + saline 4 weeks. After completion of treatment, mice were euthanized and intact lung tissues resected. Hematoxylin–Eosin (H&E), Masson’s trichrome, elastic van Gieson (EVG), immunohistochemical (IHC) analyses, and quantification of hydroxyproline (HYP) levels were performed on the obtained samples.

Isolation of primary pulmonary fibroblasts (PPFs) and their treatment

PPFs were isolated from 8-week-old male C57BL/6 mice. Briefly, lungs were washed, minced, then digested using 2.5 mg/mL dispase II (SolarBio, Beijing, China) and 2.5 mg/mL collagenase IV (Rockland, Limerick, PA, USA) at 37 °C for 30 min. After digestion, pulmonary tissues were washed using Dulbecco’s modified Eagle medium (DMEM) (Gibco, Carlsbad, CA, USA), resuspended in DMEM with 10% fetal bovine serum (FBS) and incubated in a humidified 5% CO₂ atmosphere at 37 °C. Cells at passage 2–5 were used for various assays. Activin receptor-like kinase 5 (ALK5)-overexpressing pcDNA3.1-plasmids were synthesized by Guangzhou Laisai Biotech and transfected into cells at a concentration of 2 µg/mL using the Lipo3000™ transfection reagent (Invitrogen, Carlsbad, CA, USA), as instructed by the manufacturer. After 48 h of transfection, cells were subjected to different TGF‐β1 (5 ng/mL, SinoBiological Inc., Beijing, China) and/or CILP (1 µg/mL, CRK PHARMA) treatments before harvesting.

Western blotting assay

Cells were first washed using PBS, then suspended and lysed in RIPA lysis buffer on ice. An 8% separation gel and a 5% stacking gel were used for electrophoresis. Through the wet transfer method, separated proteins were transferred to nitrocellulose membranes, which were blocked for 1 h using 5% nonfat dry milk and incubated at 4 °C overnight with primary antibodies, including Collagen I (1:1000, AF7001, Affinity Biosciences Ltd., Cincinnati, OH, USA), α-SMA (1:1000, A17910, ABclonal, Woburn, MA, USA), CILP (1:1000, ER1906-32, Huabio, Hangzhou, China), SMAD3 (1:2000, ab84177, Abcam, Cambridge, UK), p-SMAD3 (1:2000, ab52903, Abcam) and GAPDH (1:1000, AC001, ABclonal). Enhanced chemiluminescence (Affinity) was used for visualization while the Quantity One system (Syngene, Cambridge, UK) was used for quantitative analyses of proteins.

qRT-PCR

TRIzol reagent (Thermo Fisher Scientific, Waltham, MA, USA) was used to extract total RNA from cells, following the manufacturer’s instructions. PrimeScript RT Master Mix (Takara Bio Inc., Shiga, Japan) was used to synthesize cDNA from the extracted RNA. mRNA levels were determined via the SYBR green I incorporation method and real-time PCR system (Syngene). Fold changes in relative gene levels were calculated by the 2^−ΔΔCt method. The primers used in this study are listed in Table 1.

Table 1.

qRT-PCR primer sequences

Genes	Primer sequences
Cilp	F: 5ʹ- ATGGCAGCAATCAAGACTTGG-3′
	R: 5′- TTGTCCACTCTCCGGGACT-3′
Col1a1	F: 5′- GCTCCTCTTAGGGGCCACT-3′
	R: 5′- ATTGGGGACCCTTAGGCCAT-3′
Acta2	F: 5′- CCCAGACATCAGGGAGTAATGG-3′
	R: 5′- TCTATCGGATACTTCAGCGTCA-3′
Gapdh	F: 5′- AGGTCGGTGTGAACGGATTTG-3′
	R: 5′- GGGGTCGTTGATGGCAACA-3′

Enzyme-linked immunosorbent assay

The right lung tissues were homogenized on ice, and the levels of HYP in the homogenized tissue were determined using a mouse HYP ELISA kit (JM-02591, Jingmei, Yancheng, China) while blood serum CILP concentrations were determined using a human CILP ELISA kit (JM-6319, Jingmei), following the manufacturers’ instructions.

Histological analyses

Left lung tissue samples from mice were fixed, paraffin-embedded, sliced into 4 μm thick sections, and stained using H&E staining kits (G1005, ServiceBio, Wuhan, China), Masson’s trichrome staining kits (abs9347, Absin Bioscience Inc., Shanghai, China) or EVG staining kits (BA4083A, Baso, Zhuhai, China) following the manufacturers’ instructions. The Images Magnified 6100 tool (NIH, Bethesda, MD, USA) was used for semi-quantitative assessment of fibrotic changes, which were evaluated using the Ashcroft score and graded on a scale of 0–8 as follows: Grade 0, normal lung; Grade 1, isolated alveolar septa with subtle fibrotic changes; Grade 2, fibrotic changes of alveolar septa with knot-like formation; Grade 3, contiguous fibrotic walls of alveolar septa; Grade 4, single fibrotic masses; Grade 5, confluent fibrotic masses; Grade 6, large contiguous fibrotic masses; Grade 7, air bubbles; Grade 8, fibrous obliteration. This analysis was separately conducted by two independent researchers and scored in a blinded manner, as has been previously reported[13].

Immunohistochemistry and immunofluorescence analysis

Tissue sections were treated with citrate buffer at 60 °C for 16 h to remove antigens. Immunohistochemical sections were treated for 15 min with hydrogen peroxide then blocked for 1 h using 1% goat serum. Cells were fixed in paraformaldehyde (4%), incubated with Triton X-100 (0.5%), and immunoblocked at room temperature for 2 h. The prepared sections and cells were then incubated at 4 °C in the presence of primary antibodies against TGF-β1 (1:100, A16640, ABclonal), CILP (1:100, ER1906-32, Huabio), α-SMA (1:100, A17910, ABclonal), p-SMAD3 (1:100, ab52903, Abcam), Collagen I (1:100, AF7001, Affinity), Vimentin (1:200, ab137321, Abcam), and ALK5 (1:200, sc-101574, Santa Cruz Biotechnology, Santa Cruz, CA, USA). Pulmonary tissue sections or cells were washed and incubated for 1 h with matched secondary antibodies. Hematoxylin (C0107, Beyotime Institute of Biotechnology, Shanghai, China) and DAB (P0202, Beyotime) stains were used for immunohistochemical analyses while DAPI was used to stain the nuclei for immunofluorescence analysis. Integral optical densities (IOD) for TGF-β1, CILP, α-SMA, and p-SMAD3 immunohistochemistry were analyzed and quantified using Image Pro Plus 6.0 (Media Cybernetics, Rockville, MD, USA).

Statistical analysis

The data are shown as mean ± SEM and were analyzed using SPSS 25.0 software (IBM SPSS Statistics for Windows, Armonk, NY, USA). GraphPad Prism 7.1.0 software (GraphPad Software Inc., San Diego, CA, USA) was used to generate graphs. Comparisons between two groups (e.g., TGF-β1 vs. TGF-β1 + CILP) were performed using unpaired Student’s t test. Comparisons among three or more groups (e.g., control, BLM 2W, BLM 4W, BLM 6W) were performed using one-way ANOVA with Tukey’s post-hoc test to correct for multiple comparisons. Tukey’s post hoc test was applied to all one-way ANOVA results (e.g., Ashcroft scores across 6 mouse groups) to control for type I errors. The threshold for significance was p < 0.05.

Results

Microarray data and bioinformatics analyses

Gene expression profiles of GSE10667 were downloaded from GEO datasets. Differentially-expressed mRNAs were analyzed using the Limma package (version 4.1.1) in R. The cutoff criteria were |log₂ (Fold Change)|> 2.0 and adjusted p < 0.05. A total of 229 differentially-expressed mRNAs (DEms) were detected between acute pulmonary fibrosis (APF) and non-pulmonary fibrosis (NPF) samples (Fig. 1A) while 122 DEms were detected between end-stage pulmonary fibrosis (EPF) and NPF samples (Fig. 1B). Seventy overlapping DEms were found in the two datasets (Fig. 1C). CILP levels were upregulated in both APF and EPF samples, relative to NPF samples (Fig. 1D, E).

Fig. 1.

CILP is highly expressed in acute pulmonary fibrosis (APF) and end-stage pulmonary fibrosis (EPF) and participates in protein–protein interaction with TGF-β1. A A volcano plot was used to assess variations in mRNA expression between APF and non-pulmonary fibrosis (NPF) samples. B A volcano plot was used to assess variation in mRNA expression between EPF and NPF samples. C A Venn diagram of datasets shows an overlap of 70 mRNAs. Green indicates the 159 differentially expressed mRNAs (DEmRNAs) identified between APF and NPF samples. Purple indicates the 122 DEmRNAs between end-stage pulmonary fibrosis (EPF) and NPF samples. The overlapping 70 DEmRNAs (shared by both APF vs. NPF and EPF vs. NPF comparisons) are shown in the intersection of the Venn diagram. D A cluster heat map shows the overlapping 70 expressed mRNAs exhibiting an over-2.0-fold change between APF and NPF samples. Red color indicates elevated expression while green color indicates suppressed expression. Red arrow indicates CILP. E Cluster heat map shows the overlapping 70 expressed mRNAs exhibiting a greater-than-2.0-fold change between EPF and NPF samples. Red arrow indicates CILP. F Gene ontology analysis of the top expressed mRNAs, including biological processes of the mRNAs, cellular components of the mRNAs and molecular functions of the mRNAs. G mRNAs expressed by heat map and volcano plot between APF and EPF were analyzed by STRING for their functional association networks, which showed their connection in fibril formation

Functional roles of novel proteins were indirectly predicted by exploring the functions of selected proteins. GO analysis revealed that the common DEms were enriched in five biological processes, four cellular components, and four molecular functions (Fig. 1F). They were markedly enriched in fibrosis-associated biological behaviors. PPI analysis using the STRING tool showed that TGF-β1 has protein–protein interactions with CILP (Fig. 1G).

CILP and fibrosis levels were upregulated in BLM-induced pulmonary fibrosis

Findings from microarray analysis were validated by assessing CILP and fibrosis levels via BLM-induced pulmonary fibrosis mouse model (Fig. 2A). H&E, Masson’s trichrome and EVG-stained sections showed an increase in collagen fibers and marked losses of elastic fibers in the lungs of mice with BLM-induced pulmonary fibrosis (Fig. 2B). Histopathological changes were assessed by H&E staining using the Ashcroft method. Few inflammatory cells were noted in control mice. Inflammatory and fibrotic changes, such as destruction of lung alveoli and inflammatory cell infiltrations were detected in lung tissues of BLM-induced mice. Moreover, the degree of fibrosis was more severe at 4 and 6 weeks than at 2 weeks. However, differences in the degree of fibrosis between 4 and 6 weeks were insignificant (Fig. 2C).

Fig. 2.

Expression of cartilage intermediate layer protein (CILP) and transforming growth factor beta 1 (TGF-β1) in bleomycin (BLM)-induced mice positively correlated with the degree of pulmonary fibrosis. A Scheme of BLM-induced pulmonary fibrosis in a mouse model. Red arrows indicate BLM administration time. B Representative images of hematoxylin and eosin (H&E), Masson’s trichrome and elastic van Gieson (EVG) staining of lung tissue sections in the different groups. Scale bar: 200 µm. C Analysis of Ashcroft score, percentage fibrosis among Masson’s and EVG-stained section in the different groups (n = 6). D Immunohistochemistry of mouse lung sections was performed to analyze the expressions of TGF-β1, α-SMA, and CILP in the different groups. Scale bar: 50 µm. E The integral optical densities (IOD) of positive areas was quantified to evaluate the expression levels of TGF-β1, α-SMA and CILP. The data are shown as means ± SEM. *p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001. ns, not significant

The fibrogenic factor TGF-β1 and fibrosis marker α-SMA were upregulated in BLM-induced pulmonary fibrosis samples relative to control samples, and upregulated CILP protein levels were also noted in BLM-induced groups. TGF-β1, α-SMA and CILP protein levels were markedly increased at 4 and 6 weeks, relative to 2 weeks. However, differences in protein levels between weeks 4 and 6 were insignificant (Fig. 2D, E). These results indicate successful establishment of the pulmonary fibrosis model.

CILP expression was induced by TGF-β1 in pulmonary fibroblasts

CILP expression and the expression of fibrosis markers by myofibroblasts were upregulated upon TGF-β1 treatment. TGF-β1 time-dependently induced mRNA expression of Cilp and myofibroblast fibrosis markers (Col1a1 and Acta2) (Fig. 3A). Similarly, TGF-β1 time-dependently upregulated CILP protein expression, as determined by western blot analysis, which was accompanied by increases of Collagen I and α-SMA protein levels (Fig. 3B, C). In addition, immunofluorescence analysis revealed upregulated CILP protein levels (Fig. 3D, E).

Fig. 3.

TGF-β1-induced CILP expressions in mouse primary pulmonary fibroblasts (PPFs). A Quantitative qRT-PCR analysis showing mRNA expressions of Cilp, Col1a1 and Acta2 in mouse PPFs after treatment with TGF-β1 (5 ng/mL) for the indicated durations. Relative mRNA expressions in untreated cells were set as 1.0. B Western blot analysis of CILP, Collagen I and α-SMA protein levels in mouse PPFs after TGF-β1 (5 ng/mL) treatment for the indicated duration in the four groups. Protein expressions of CILP, Collagen I and α-SMA and increased in a time-dependent manner. C Quantitative analysis of western blot results. D Fluorescent staining of CILP after treatment with 5 ng/mL TGF-β1 for the indicated durations. CILP (red), DAPI (blue), Scale bar: 50 μm. E Quantitative analysis of fluorescent staining results. Data are shown as means  ± SEM. *p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001. ns, not significant

CILP inhibited TGF-β1-induced SMAD3 signaling

The potential of CILP to inhibit TGF-β1-stimulated expression of profibrotic target genes and proteins in cultured PPFs was determined by SMAD3 signaling. qRT-PCR analysis showed that administration of CILP to PPFs inhibited TGF-β1-induced expression of Col1a1 and Acta2 at the mRNA level (Fig. 4A). Western blot assays revealed that administration of CILP to PPFs suppressed TGF-β1-stimulated p-SMAD3, Collagen I, and α-SMA protein levels (Fig. 4B, C). Double immunofluorescence staining showed that CILP inhibited TGF-β1-stimulated expressions of Collagen I and Vimentin proteins (Fig. 4D). However, CILP had no marked effects on basal mRNA or protein expressions of Collagen I and α-SMA in the absence of TGF-β1 treatment.

Fig. 4.

CILP inhibits TGF-β1-induced SMAD3 signaling. A Quantitative RT-PCR analysis showing mRNA expressions of Col1a1 and Acta2 in mouse PPFs from the different treatment groups. B Western blot analysis showing protein expressions of SMAD3 signaling, Collagen I and α-SMA in mouse PPFs after treatment with 5 ng/mL TGF-β1 and/or 1 μg/mL CILP for 24 h. C Quantitative analysis of western blot results. D Double immunofluorescence staining of Collagen I and Vimentin under different treatments for 24 h. Collagen I (red), Vimentin (green), DAPI (blue), Scale bar: 50 μm. E Effects of CILP on p-SMAD3 nuclear translocation in mouse PPFs. Cells were incubated under different treatments for 24 h. p-SMAD3 (red), DAPI (blue), Scale bar: 10 μm. F Quantitative analysis of fluorescence staining of Collagen I, Vimentin and p-SMAD3. Data are shown as means  ± SEM. *p  <  0.05, **p  <  0.01, ***p  < 0 .001, ****p  <  0.0001. ns, not significant

Phosphorylation of SMAD3 and its subsequent nuclear translocation are important for TGF-β1 signaling. Thus, the effects of CILP on p-SMAD3 and its nuclear translocation were explored. p-SMAD3 levels in the nucleus were increased by TGF-β1 treatment (Fig. 4E). However, the increased p-SMAD3 levels in PPFs after TGF-β1 treatment were inhibited by CILP administration at the mRNA and protein levels. These results imply that extrinsic CILP blocked p-SMAD3 translocation into the nucleus. In summary, CILP negatively affects TGF-β1-induced transcription by suppressing p-SMAD3 and its nuclear translocation.

CILP inhibited TGF-β1-induced SMAD3 signaling by binding ALK5

The molecular mechanisms through which CILP negatively regulates were investigated.

Since ALK5 is the cell surface receptor for TGF-β1-induced SMAD3 signaling in fibrotic responses (Li, Sun, Wu, & Ma, 2022), then we detected the distribution of ALK5 and CILP. Both ALK5 and CILP proteins were found to be distributed in the same region, and the composite image indicated that they were mostly colocalized (Fig. 5A). Moreover, TGF-β1 treatment markedly elevated both ALK5 and CILP protein levels. CILP effectively inhibited TGF-β1-induced expression of SMAD3 signaling pathway components and fibrosis markers, and a partial reversal of the effect of CILP was observed upon transfection of ALK5-overexpressing plasmids. However, in the absence of TGF-β1 treatment, CILP had no marked effects on SMAD3 signaling pathway and fibrosis marker levels (Fig. 5B, C). Prediction using the signaling model indicated that CILP inhibits TGF-β1-induced SMAD3 signaling in PPFs by binding ALK5, thereby inhibiting SMAD3 signaling.

Fig. 5.

CILP inhibits TGF-β1-induced SMAD3 signaling by binding ALK5. A Representative confocal images from the immunofluorescent assay showing colocalization of ALK5 and CILP using specific antibodies after treatment with 5 ng/mL TGF-β1. ALK5 (green), CILP (red), DAPI (blue). Scale bar: 20 µm. B Western blot analysis of the effects of overexpressed ALK5 on protein levels of the SMAD3 signaling pathway upon treatment with or without TGF-β1 and CILP. GAPDH served as loading control. C Quantitative analysis of western blot results

CILP prevented BLM-induced pulmonary fibrosis in vivo

In vitro, CILP inhibited TGF-β1-induced SMAD3 signaling via competitive binding of ALK5, therefore, we investigated whether CILP prevents BLM-induced pulmonary fibrosis in vivo. The BLM-induced pulmonary fibrosis group was intravenously administered CILP or saline to assess the roles of CILP in BLM-induced pulmonary fibrosis (Fig. 6A). H&E, Masson’s trichrome and EVG staining revealed that BLM-induced mice exhibited significantly greater pulmonary fibrosis relative to the control group, and CILP reduced the pathologic changes of BLM-induced pulmonary fibrosis in mice (Fig. 6B, C). Immunohistochemical staining revealed that CILP decreased the IOD of p-SMAD3 in vivo, and CILP suppressed the BLM-induced increase in IOD of α-SMA in vivo. Meanwhile, the differences between BLM and saline + BLM groups were insignificant (Fig. 6D, E). HYP levels of mouse lung homogenates were markedly suppressed in CILP-treated mice, relative to BLM-induced mice. Further assays were conducted to determine whether CILP is a novel biomarker for IPF, but the results revealed that differences in serum CILP levels in blood samples from IPF and NPF groups were insignificant (Fig. 6F).

Fig. 6.

CILP prevents BLM-induced pulmonary fibrosis in vivo. A CILP inhibited BLM-induced pulmonary fibrosis. Red arrows indicate BLM administration time, blue arrows indicate CILP administration time. B Representative images of H&E, Masson, and EVG staining of tissues from mice under different treatments for 4 weeks. Scale bar: 200 µm. C Analysis of Ashcroft score, percentage fibrosis among Masson’s and EVG-stained section in the different groups (n = 6). D Immunohistochemistry of lung sections was performed to analyze the expressions of p-SMAD3 and α-SMA. Scale bar: 50 µm. E IOD of positive areas was quantified using Image Pro Plus 6.0 to evaluate p-SMAD3 and α-SMA expressions separately. F ELISA for hydroxyproline protein levels in the right lungs of mice (n = 6 mice in each group) and CILP protein levels in serum samples from IPF patients (n = 17) and non-IPF patients (n = 17). The data are shown as means ± SEM. *p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001. ns, not significant

Discussion

Idiopathic pulmonary fibrosis (IPF) is a chronic disease which affects approximately five million individuals globally, and the typical median survival ranges from 3 to 5 years [14]. The fundamental pathological alteration involves a transition from alveolar injury and excessive collagen proliferation to the development of pulmonary interstitial fibrosis [15]. However, complete understanding of its underlying mechanisms remains elusive [16]. Investigating the underlying mechanisms of IPF is crucial for enhancing prognostic outcomes [17]. Excessive proliferation of fibroblasts and extracellular matrix (ECM) remodeling plays a pivotal role in the development of pulmonary fibrosis, serving as a fundamental component of this disease [18]. Cartilage intermediate layer protein (CILP) plays a vital role within the ECM, and has significant importance in the regulation of ECM metabolism [19]. Previous studies have reported CILP upregulation in local tissues, such as in OA-induced cartilage degeneration [20], lumbar disc degeneration [20], and cardiac fibrosis [21]. These findings imply that pulmonary tissues with elevated CILP levels are prone to fibrosis. In the analysis of GSE10667, a significant upregulation of CILP protein was observed in both APF and EPF. The results indicate that CILP may play a role in the pathogenesis of IPF.

Currently, the BLM-induced pulmonary fibrosis mouse model is widely recognized as the most established model for investigating the pathogenesis of pulmonary fibrosis, and is renowned for its simplicity and effectiveness in replicating the disease [22, 23]. In our study, subcutaneous administration of BLM was employed to induce pulmonary fibrosis in mice, and elevated levels of CILP were observed. Pathological staining confirmed the ability of bleomycin to induce pulmonary fibrosis in mice by causing pathological changes in the lung tissue. Subsequent immunohistochemical analysis revealed a significant increase in the levels of the fibrogenic factor TGF-β1 and the fibrosis marker α-SMA. These findings provide strong evidence for the successful modeling of IPF in mice. Interestingly, the expression levels of CILP also exhibited a time-dependent increase in mice with pulmonary fibrosis, which aligns with the findings from the GSE10667 gene expression profile. Consequently, we employed a mouse model of BLM-induced pulmonary fibrosis for subsequent experimental investigations.

Previous research has indicated that cytokines influence the development of pulmonary fibrosis; however, further investigation is required to fully elucidate the precise mechanism [24, 25]. TGF-β1, a highly-potent pro-fibrotic cytokine, plays a crucial role in the progression of fibrotic ailments, including fibrosis of the lung [26], heart [27], and kidney [28]. In individuals with IPF, TGF-β1 is primarily released by pulmonary macrophages and alveolar epithelial cells [29, 30]. Its presence stimulates the proliferation and activation of fibroblasts, as well as the formation of myofibroblasts and deposition of extracellular matrix [31]. The results of our in vitro cell experiments provide confirmation that TGF-β1 elicits the upregulation of CILP protein in a time-dependent fashion, concurrently stimulating the expression of fibrotic proteins (Collagen I, Vimentin, and α-SMA). Notably, CILP exerts inhibitory effects on TGF-β1-induced SMAD3 phosphorylation and subsequent nuclear translocation, resulting in diminished translocation of p-SMAD3 to the nucleus. Consequently, this inhibition impedes the activation of TGF-β1 target genes, reduces collagen synthesis, and ameliorates pulmonary fibrosis. It is believed that the signaling of TGF-β1 offers adequate potency and duration to initiate CILP, thereby establishing a negative feedback mechanism that hinders the proliferation of fibroblasts and the deposition of ECM through the inhibition of SMAD3 signaling and the down-regulation of target genes.

Mechanistically, CILP suppresses TGF-β1 signaling by directly interacting with TGF-β1 due to the common thrombospondin-1 domain, thereby affecting the receptor ALK5. Previous bioinformatic analyses have also provided evidence for the existence of a protein–protein interaction between CILP and TGF-β1. It is important to note that ALK5 serves as a TGF-β1 receptor and also functions as a transmembrane serine/threonine kinase receptor. The activation of ALK5 occurs through its binding to the TGF-β1 type II serine/threonine kinase receptor. ALK5 is capable of transmitting signals of TGF-β1 from the cell surface to the cytoplasm, thereby exerting regulatory control over various physiological and pathological processes, including cell cycle arrest, proliferation and differentiation of mesenchymal cells, wound healing, extracellular matrix production, immunosuppression, and expression of carcinogenic genes [32–34]. Ye et al. observed a reduction in the activity of the TGF-β1 pathway and a decrease in the expression of TGF-β1 target genes in ALK5-knockout LF cells [35]. Immediately upon binding, ALK5 phosphorylates SMAD3, resulting in its subsequent nuclear translocation [36]. Phosphorylated SMAD3 forms a complex with SMAD4 and translocates to the nucleus, where they modulate gene transcription [37]. Although SMAD2 and SMAD3 are located in the same signaling pathway, yet the role of the TGF-β/ALK5/SMAD3 axis in the mechanism of fibrogenesis is supported by sufficient experimental evidence and the potent fibrogenic actions of TGF-β/ALK5 signaling are predominantly mediated through activation of SMAD3, while the in vivo role of SMAD2 appears relatively limited [38]. Consequently, we chose SMAD3 as the focus of the present study. Our experimental findings provide confirmation that CILP effectively inhibits the signaling pathway of TGF-β1/SMAD3 by competitively binding to ALK5. In addition, we observed that the inhibition of CILP can be partially reversed when ALK5-overexpressing plasmids are transfected. Taken together, our results suggest that CILP exerts inhibitory effects on TGF-β1/ALK5-mediated SMAD3 phosphorylation, ALK5-dependent SMAD3 activity, and the nuclear localization of p-SMAD3. Subsequently, immunohistochemical analyses were conducted on mouse lung tissues. In accordance with the outcomes obtained from in vitro cell experiments, it was observed that CILP effectively impedes the progression of pulmonary fibrosis in vivo by suppressing SMAD3 phosphorylation in mice. Since the primary goal of this study was to elucidate CILP’s role in the TGF-β1/SMAD3 pathway, not to compare CILP with clinically approved drugs. Adding a positive control would have required increasing animal numbers (to maintain statistical power) and introducing variables unrelated to the core mechanism. Therefore the future studies should include pirfenidone as a positive control to directly compare CILP’s efficacy and explore potential synergistic effects between CILP and the existing IPF therapies.

Further, ELISA was used to analyze mouse lung tissue and human serum samples. HYP, being a main component of collagen, is present in very small quantities in elastin and absent in other proteins. Thus, the HYP content in various tissues can serve as a significant indicator for measuring collagen deposition [39]. The ELISA results provided confirmation that CILP hindered BLM-induced HYP in mice, suggesting a decrease in collagen secretion and ECM deposition. The ELISA analysis of peripheral blood samples obtained from individuals in the IPF and non-IPF cohorts revealed no statistically-significant disparity in CILP levels. This finding implies that the differential expression of CILP may be confined to pulmonary tissues, thereby restricting its utility as an early diagnostic biomarker for IPF. However, the lung-specific expression of CILP suggests that the corresponding therapeutic interventions may have fewer systemic adverse effects, rendering it highly clinically relevant. The future investigations may build on the current findings by exploring CILP’s interactions with other key pro-fibrotic pathways (e.g., Wnt/β-catenin, Hedgehog) in IPF, or by evaluating CILP expression dynamics across distinct IPF clinical stages (e.g., early vs. advanced disease) to clarify its potential role as a prognostic biomarker.

To summarize, the administration of CILP elicited an improvement in pulmonary fibrosis induced by BLM in mice, while also effectively suppressing profibrotic activities through the negative feedback loop of the TGF-β1/SMAD3 pathway. Consequently, CILP presents itself as a promising candidate for the inhibition and treatment of IPF.

Limitations of the study

Firstly, the current diagnosis of IPF is mainly based on multidisciplinary evidence. The intrapulmonary lesions of IPF are heterogeneous, and it is difficult to diagnose IPF from bronchoscopic lung biopsy specimens. In general, surgical open lung biopsy is required to obtain lung tissue specimens for diagnosis. Owing to respiratory research conditions and ethical limitations, we were unable to obtain a large enough human lung specimen. Furthermore, it is worth considering the utilization of certain advanced molecular biology techniques, such as FRET and CO-IP, to effectively illustrate protein interactions.

Author contributions

Hua Zou, Peng Li, and Chunlei Liu designed and supervised the study. Hua Zou and Jiale Dong made equal contributions to this study as co-first authors. Hua Zou, Jiale Dong, Run Zhao, and Feng Liu implemented the in vitro experiments. Hua Zou, Jiale Dong, Jingsong Cheng, Xushan Li, and Chunlei Liu performed the in vivo experiments. Chengshuo Fei and Peng Li participated in the bioinformatics analysis and data analysis. All authors drafted the original manuscript. Hua Zou, Jiale Dong, Peng Li, and Chunlei Liu edited and finalized the manuscript. All authors have read and approved the submitted manuscript.

Funding

This work was supported by the special funds for clinical research from Affiliated Qingyuan Hospital, Guangzhou Medical University, Qingyuan People’s Hospital (QYRYCRC2023005), Plan on enhancing scientific research in GMU (GZMU-SH-297), Natural Science Foundation of Guangdong Province (2022A1515220177), and Qingyuan Science and Technology Foundation of China (2022KJJH048). We sincerely thank International Science Editing (https://www.internationalscienceediting.com/) for language polishing assistance, and we thank LetPub (https:/www.letpub.com/) for linguistic assistance of this manuscript.

Data availability

The datasets generated and analyzed during the current study are available from the corresponding author on reasonable request.

Declarations

Ethics approval and consent to participate

Human serum sample collection was approved by the Institutional Ethics Review Committee of Affiliated Qingyuan Hospital of Guangzhou Medical University (IRB-2023-023). Thirty-four blood samples were obtained from Affiliated Qingyuan Hospital of Guangzhou Medical University. Prior to blood collection, patients were required to provide informed consents.

Competing interests

The authors declare no competing interests.