Introduction

Pulmonary fibrosis (PF) is a devastating lung disease with poor prognosis and is characterized by typical interstitial pneumonia and excessive deposition of extracellular matrix in the lung [1, 2]. At present, the pathogenesis of PF is not fully understood, but it is widely accepted that an imbalance between profibrotic and antifibrotic factors triggers this progressive disease. Transforming growth factor-β1 (TGF-β1) and connective tissue growth factor, which robustly induce epithelial–mesenchymal transition (EMT) or activation of fibroblasts, are secreted in large amounts [1, 2]. However, antifibrotic factors are relatively insufficient, among which the most well recognized is hepatocyte growth factor (HGF). HGF has the capacity to block myofibroblast accumulation, as well as apoptosis of lung epithelial and endothelial cells, and plays an important role in counteracting fibrosis. Administration of HGF protein or adenoviral expression of HGF prevents fibrotic remodeling in several animal models of PF [3]. Therefore, modulating the microenvironment that favors fibrosis by increasing the levels of antifibrotic factors may be an attractive strategy to interfere with fibrosis progression.

Curcumin, a natural polyphenol compound extracted from turmeric (Curcuma longa L.), is used as an ingredient in curry in southeast Asia. A large body of evidence has confirmed the bioactivities of orally administered curcumin, including antifibrotic effects in the lung, liver, and kidney [4]. However, pharmacokinetic studies have shown that its absolute bioavailability is only 1%, and the peak plasma level is unable to reach the minimum effective concentration used in vitro cell experiments [5,6,7,8]. In addition, the possibility that the metabolites of oral curcumin rather than curcumin itself account for its antifibrotic efficacy are still not widely accepted, as the plasma levels of metabolites in rodents that were orally administered curcumin and the bioactivities are not higher than those of the parent compound [9,10,11,12]. The underlying mechanism by which curcumin protects against PF remains to be elucidated to improve the use of curcumin, and a specific route that cannot be simply explained by classical pharmacokinetics and pharmacodynamics may be involved [13]. The gut is responsible for digestion and absorption of food, is also known as the largest endocrine organ, and is capable of secreting various endogenous factors. Among these factors, some cytokines, growth factors and brain-gut peptides exert distinct antifibrotic activities [14, 15]. Previous studies in our laboratory suggested that curcumin possessed an antiarthritic effect in a gut-dependent manner [16, 17]; thus, we hypothesized that curcumin promotes the generation of gut-derived factors and that these factors act as mediators, subsequently entering the lung to ameliorate fibrosis. This study was performed to uncover the mode of action of oral curcumin against PF and to provide a reasonable paradigm for mechanistic studies of compounds with good efficacy but low bioavailability.

Materials and methods

Reagents

Curcumin (C21H20O6, MW: 368.37; purity > 98%) was purchased from Nanjing Jingzhu Biological Technology Co., Ltd. (Nanjing, China). Pirfenidone (C12H11NO, MW: 185.22; purity > 98%) was purchased from Dalian Meilun Biotechnology Co., Ltd. (Dalian, China). Bleomycin hydrochloride was purchased from Nippon Kayaku Co., Ltd. (Tokyo, Japan). Silicon dioxide (SiO2, ~99%, 0.5–10 μm), KG-501, and 15-deoxy-Δ12, 14-prostaglandin J2 (15d-PGJ2) were purchased from Sigma-Aldrich (St. Louis, MO, USA). GW9662, HPGDS-inhibitor-1, and forskolin were purchased from TargetMol (Shanghai, China). SU11274 was purchased from ApexBio Technology Co., Ltd. (Houston, TX, USA). The hydroxyproline commercial kit was purchased from Nanjing Jiancheng Bioengineering Institute (Nanjing, China). The mouse HGF and Klotho ELISA kits were purchased from Cusabio Biotech Co., Ltd. (Wuhan, China). The universal 15d-PGJ2 ELISA kit was purchased from Elabscience Biotechnology Co., Ltd. (Wuhan, China). Antibodies against pCREB, CREB, and GAPDH were purchased from Bioworld Technology, Inc. (Atlanta, GA, USA). Antibodies against HGF and PPARγ were purchased from Boster Biological Technology Co., Ltd. (Wuhan, China) and Wanlei Biological Technology Co., Ltd. (Shenyang, China), respectively.

Animals

Female ICR mice weighing 22–25 g were purchased from the Comparative Medicine Centre of Yangzhou University (Yangzhou, China). The animal experiments were conducted with the approval of the Animal Ethics Committee of China Pharmaceutical University and complied with the National Institute of Health guidelines on the ethical use of animals. All animals were housed under a 12 h light/dark cycle (21 ± 2 °C) and allowed ad libitum access to a diet of standard laboratory chow and water.

PF models and treatments

On day 0, mice were anaesthetized by avertin (125 mg/kg, i.p.), and PF was established by intratracheal instillation of bleomycin (5 mg/kg) or SiO2 (100 mg/kg) in sterile 0.9% NaCl. For different studies, the mice were divided into various groups: (a) normal group, model group, curcumin (75 or 150 mg/kg, i.g.) group, and pirfenidone (400 mg/kg, i.g.) group; (b) normal group, model group, curcumin (150 mg/kg, i.g.) group, curcumin (10 mg/kg, i.v.) group, and pirfenidone (400 mg/kg, i.g.) group; (c) normal group, model group, curcumin (75 mg/kg, i.g.) group, curcumin (75 mg/kg, p.r.) group, and pirfenidone (400 mg/kg, i.g.) group; (d) normal group, model group, curcumin (150 mg/kg, i.g.) group, SU11274 (0.18 mg/kg, i.p.) group, curcumin + SU11274 group, and pirfenidone (400 mg/kg, i.g.) group; and (e) normal group, bleomycin group, curcumin (150 mg/kg, i.g.) group, HPGDS-inhibitor-1 (0.5 mg/kg, i.g.) group, and curcumin + HPGDS-inhibitor-1 group.

Curcumin and pirfenidone were administered daily from day 1 to 21 (bleomycin model) or from day 1 to 30 (SiO2 model). SU11274 and HPGDS-inhibitor-1 were administered daily 30 min before curcumin administration. Mice in the normal and model groups were administered an equal volume of vehicle.

Lung index analysis and hydroxyproline assay

On day 21 (bleomycin model) or 30 (SiO2 model), the mice were weighed and sacrificed. Lungs were quickly isolated, washed, and weighed. The lung index is expressed as the ratio of the lung wet weight (mg) to the body weight (g); the levels of hydroxyproline in the upper lobes of the left lungs were measured with a commercial kit according to the manufacturer’s instructions.

Histopathological examination

The lower lobes of the left lungs were fixed in 10% neutral buffered formalin, embedded in paraffin, and sliced into 5-μm-thick sections. Hematoxylin–eosin (H&E) and Masson’s trichrome staining were conducted to evaluate the degree of inflammation and fibrosis, respectively. The scores were calculated as previously described [18].

Cell culture

CCD-18Co cells were provided by American Type Culture Collection (ATCC, Manassas, VA, USA). RAW264.7 and L929 cells were obtained from the Cell Bank of Shanghai (Institute of Biochemistry and Cell Biology, Chinese Academy of Sciences, Shanghai, China). All cells were cultured in DMEM supplemented with 10% FBS and maintained in 5% CO2 humidified air at 37 °C.

Isolation of primary colonic epithelial cells and fibroblasts

ICR mice aged 1 week were sacrificed. The colons were removed and carefully cut open to remove the feces. After the colon tissues were soaked in PBS containing 10× penicillin–streptomycin for 2–3 min, the colon tissues were cut into 1–2 mm3 fragments and placed into the digestive solution (0.1% type II collagenase, 0.25% trypsin, and 0.1% DNase I) for a 1-h digestion at 37 °C. The mixture was filtered through a 200-mesh sterile sieve, followed by centrifugation, and the cells were resuspended and cultured in DMEM supplemented with 10% FBS. The colonic epithelial cells were separated from the fibroblasts according to their different rates of adherence [19].

Isolation and induction of primary macrophages

ICR mice aged 4–5 weeks were sacrificed, and the femur and tibia were separated. Bone marrow cells were flushed and filtered through a 200-mesh sterile sieve. Red blood cells were removed by using lysis buffer (NH4Cl 3.735 g and Tris 1.3 g dissolved in 500 mL of double-distilled water), and the remaining cells were washed with sterile PBS. After centrifugation, the cells were resuspended and cultured in DMEM containing 10% FBS and 30% L929 cell culture supernatant. Bone marrow-derived macrophages were obtained 5 days later for subsequent experiments.

qPCR analysis

Total RNA from tissues or cells was extracted using TRIzol reagent (Invitrogen, Carlsbad, CA, USA) and reverse transcribed into cDNA according to the instructions of a commercial cDNA synthesis kit (Vazyme, Nanjing, China). qPCR analysis of target genes was performed on a Bio-Rad iQ real-time PCR system (Bio-Rad, Hercules, CA, USA), and the expression of IFN-γ, bone morphogenetic protein-7 (BMP-7), HGF, Klotho, phosphatase and tensin homolog deleted on chromosome ten (PTEN), somatostatin (SST), vasoactive intestinal peptide (VIP), PPARγ, and CD36 were normalized to that of GAPDH. The details of the gene-specific primers (Sangon Biotech, Shanghai, China) are listed in Table 1.

Table 1 Primers used in qPCR.
Full size table

ELISA

The levels of HGF, Klotho, and 15d-PGJ2 in tissues or cell culture supernatants were measured using ELISA kits according to the manufacturer’s instructions.

Western blotting

Whole cell lysates were prepared with NP40 buffer (Beyotime, Nantong, China). Equal concentrations of protein lysate from all samples were separated by 10% SDS-PAGE and electrophoretically transferred onto NC membranes. The membranes were blocked with 5% nonfat milk for 2 h at room temperature and probed with primary antibodies overnight at 4 °C. After being washed, the membranes were further incubated with secondary antibodies for 1 h at room temperature. Detection was performed by an Odyssey Infrared Imaging System (LI-COR, Inc., Lincoln, NE, USA).

Statistical analysis

Statistical analysis was performed using SPSS software (SPSS, Chicago, IL, USA), and the data are presented as the means ± S.E.M. Student’s t test was used to compare the mean differences between two groups. One-way ANOVA followed by the LSD test was used to compare the mean differences between multiple groups, and in cases where the latter condition was violated, the nonparametric Games-Howell post hoc test was used. The correlation between two variables was evaluated by Spearman’s correlation analysis. A value of P less than 0.05 (P < 0.05) was accepted as a significant difference.

Results

Orally administered curcumin alleviated PF in mice

To evaluate the effect of oral curcumin on PF, well-characterized rodent models were established in mice by a single intratracheal instillation of bleomycin or SiO2. It was shown that daily administration of curcumin (150 mg/kg) starting on day 1 significantly reduced the lung index value and downregulated the hydroxyproline level, with efficacies comparable to those of pirfenidone (400 mg/kg) (Fig. 1a, b, e, f). H&E staining of the lung tissues demonstrated that curcumin ameliorated the disruption of alveolar structure and infiltration of inflammatory cells (Fig. 1c, g). Moreover, the lungs of curcumin-treated mice displayed reduced collagen deposition, as shown by Masson’s staining (Fig. 1d, h). To determine the exact role of curcumin in ameliorating fibrosis but not blocking inflammation, curcumin (150 mg/kg) was orally administered once a day from day 7 to 21 (fibrotic stage) in bleomycin-treated mice. The results showed that curcumin (150 mg/kg) still exerted distinct restrictions on PF-associated symptoms (Supplementary Fig. S1). Thus, orally administered curcumin exerts a substantial anti-PF effect.

Fig. 1: Effect of orally administered curcumin (CUR) on bleomycin (BLM)- or SiO2-induced PF in mice.
figure 1

PF was induced by intratracheal instillation of bleomycin or SiO2. Curcumin (75, 150 mg/kg) and pirfenidone (400 mg/kg) were orally administered once a day from day 1 until the end of the experiment. a, e Lung index (mg/g). b, f Hydroxyproline level (μg per left lung). c, d, g, h Pathological lung abnormalities were examined by hematoxylin and eosin (H&E) and Masson’s trichrome staining (×200). The data are presented as the means ± S.E.M. (n = 6–8). #P < 0.05, ##P < 0.01 vs. the normal group; *P < 0.05, **P < 0.01 vs. the BLM or SiO2 group.

Full size image

Curcumin that was absorbed into the blood was insufficient to exhibit an anti-PF effect

Despite the inhibitory effect of orally administered curcumin on PF, pharmacokinetic data indicate that the bioavailability of oral curcumin is only 1% [5]. A dose of 10 mg/kg, which is much higher than 1.5 mg/kg (1% of the oral dose 150 mg/kg), was intravenously administered to evaluate whether the extremely low level of absorbed curcumin could protect mice from PF. As shown in Fig. 2a, b, e, f, in both the bleomycin and SiO2 model, intravenous administration of curcumin (10 mg/kg) failed to change the lung index value and hydroxyproline level. Histological scores remained high, as severe inflammation and collagen deposition still occurred in the lungs of mice (Fig. 2c, d, g, h). Therefore, the possibility that the absorbed curcumin after oral administration (150 mg/kg) exhibits an anti-PF effect can be excluded.

Fig. 2: Effect of intravenously injected curcumin (CUR) on bleomycin (BLM)- or SiO2-induced PF in mice.
figure 2

PF was induced by intratracheal instillation of bleomycin or SiO2. Curcumin (150 mg/kg) and pirfenidone (400 mg/kg) were orally administered once a day, and curcumin (10 mg/kg) was intravenously injected once a day from day 1 until the end of the experiment. a, e Lung index (mg/g). b, f Hydroxyproline level (μg per left lung). c, d, g, h Pathological lung abnormalities were examined by hematoxylin and eosin (H&E) and Masson’s trichrome staining (×200). The data are presented as the means ± S.E.M. (n = 6–8). #P < 0.05, ##P < 0.01 vs. the normal group; *P < 0.05, **P < 0.01 vs. the BLM or SiO2 group.

Full size image

Curcumin ameliorated PF through an intestinal effect

Compared with plasma and extraintestinal tissues, the intestine is the site that is exposed to a relatively high concentration of curcumin [20]. To ascertain whether the intestinal tract was involved in the anti-PF effects of curcumin, we changed the route of administration and investigated the effect of rectally administered curcumin. The results showed that curcumin (75 mg/kg) administered through the rectum distinctly restricted the PF-associated symptoms caused by bleomycin or SiO2, while orally administered curcumin at the same dose only showed a weak effect (Fig. 3). These findings suggest the importance of the intestinal tract and support the notion that oral curcumin ameliorates PF through an intestinal effect.

Fig. 3: Effect of rectally administered curcumin (CUR) on bleomycin (BLM)- or SiO2-induced PF in mice.
figure 3

PF was induced by intratracheal instillation of bleomycin or SiO2. Curcumin (75 mg/kg) was orally or rectally administered, and pirfenidone (400 mg/kg) was orally administered once a day from day 1 until the end of the experiment. a, e Lung index (mg/g). b, f Hydroxyproline level (μg per left lung). c, d, g, h Pathological lung abnormalities were examined by hematoxylin and eosin (H&E) and Masson’s trichrome staining (×200). The data are presented as the means ± S.E.M. (n = 6–8). #P < 0.05, ##P < 0.01 vs. the normal group; *P < 0.05, **P < 0.01 vs. the BLM or SiO2 group.

Full size image

Curcumin enhanced HGF generation in the colon in mice with PF

It is well known that the intestine secretes a variety of cytokines, growth factors, and brain-gut peptides, among which several members can protect against fibrosis, such as IFN-γ, BMP-7, HGF, Klotho, PTEN, SST, and VIP [14, 15]. The potential intestine-derived factors involved in the anti-PF effects of curcumin were screened using a qPCR assay. As shown in Fig. 4a–c, oral administration of curcumin (150 mg/kg) did not affect the mRNA expression of IFN-γ, BMP-7, PTEN, SST, or VIP in the small intestine, colon, and lung tissues of mice treated with bleomycin. The mRNA expression levels of HGF and Klotho in the lungs were not affected, but curcumin significantly upregulated the expression of HGF and Klotho in the colon (Fig. 4b). In the case of pirfenidone, the results seemed to be different, as this compound increased the mRNA expression of antifibrotic factors in the lungs but not the colon (Fig. 4d). HGF and Klotho can also be produced by the liver and kidney, in addition to the intestine. Curcumin administration hardly affected the expression of these two factors in extraintestinal tissues (Fig. 4e). In addition, oral curcumin obviously increased the protein levels of HGF in the colon, serum, and lungs, but its effect on Klotho expression was weaker (Fig. 4f). Correlation analysis revealed that the upregulated protein level of HGF but not Klotho in the lungs by oral administration of curcumin was highly correlated with its anti-PF effect (Fig. 4g). Similarly, oral curcumin (150 mg/kg) also significantly promoted the mRNA expression of HGF in the colon but not in the lung tissues of SiO2-treated mice (Fig. 4h). Therefore, oral curcumin can selectively promote the generation of HGF in the colon in mice with PF.

Fig. 4: Curcumin (CUR)-mediated promotion of HGF expression in the colon in mice with PF.
figure 4

a–d mRNA expression of antifibrotic factors in the small intestine, colon, or lungs of bleomycin-treated mice. e mRNA expression of HGF and Klotho in the liver and kidney in bleomycin-treated mice. f Protein levels of HGF and Klotho in the colon, serum, and lungs in bleomycin-treated mice. g Correlation between the promotion of HGF and Klotho expression and the anti-PF effect of curcumin in bleomycin-treated mice. h mRNA expression of HGF in the colon and lungs in SiO2-treated mice. The mRNA expression was measured by qPCR and was normalized to that of GAPDH in each sample. The protein levels were measured by ELISA. The data are presented as the means ± S.E.M. (n = 6). #P < 0.05, ##P < 0.01 vs. the normal group; *P < 0.05, **P < 0.01 vs. the BLM or SiO2 group.

Full size image

Specific inhibition of the HGF receptor reversed the anti-PF effect of curcumin in mice

To clarify the participation of HGF in the anti-PF effect of curcumin, the effect of a specific inhibitor of the HGF receptor was investigated. In the bleomycin-induced PF mouse model, coadministration of curcumin (150 mg/kg), and SU11274 (0.18 mg/kg) almost completely abolished the ameliorative effect on PF (Fig. 5a, b). Severe inflammation and fibrosis were observed in the lungs of mice in the coadministration group compared with the lung of vehicle-treated PF mice (Fig. 5c, d). These results were further confirmed in the SiO2-induced PF model (Fig. 5e–h), indicating that the anti-PF effect of oral curcumin is HGF-dependent and that HGF functions as an anti-PF mediator of the effect of oral curcumin.

Fig. 5: Impact of c-Met inhibition on the anti-PF effect of orally administered curcumin (CUR) in bleomycin (BLM)- or SiO2-treated mice.
figure 5

PF was induced by intratracheal instillation of bleomycin or SiO2. Curcumin (150 mg/kg) was orally administered after SU11274 was intraperitoneally injected once a day from day 1 until the end of the experiment. a, e Lung index (mg/g). b, f Hydroxyproline level (μg per left lung). c, d, g, h Pathological lung abnormalities were examined by hematoxylin and eosin (H&E) and Masson’s trichrome staining (×200). The data are presented as the means ± S.E.M. (n = 6–8). #P < 0.05, ##P < 0.01 vs. the normal group; *P < 0.05, **P < 0.01 vs. the BLM or SiO2 group; $P < 0.05, $$P < 0.01 vs. the BLM + CUR or SiO2 + CUR group.

Full size image

Curcumin induced the generation of HGF in colonic fibroblasts and macrophages

HGF can be secreted by epithelial cells, fibroblasts, and macrophages [21,22,23]. To determine the effector cells of curcumin-mediated effects in the colon, the effect of curcumin on HGF expression in these cells was investigated. As shown in Fig. 6a, curcumin did not affect cell viability except at a concentration of 80 μM. The mRNA expression results showed that curcumin (20 μM) stimulated the generation of HGF in primary colon fibroblasts and macrophages, and the effect was the strongest when the cells were stimulated for 24 h (Fig. 6b). Moreover, curcumin upregulated the expression of HGF in primary fibroblasts, macrophages, CCD-18Co cells (a fibroblast cell line), and RAW264.7 cells (a monocyte–macrophage cell line) in a concentration-dependent manner (Fig. 6c–e). Collectively, curcumin might target fibroblasts and macrophages in the colon in mice, and CCD-18Co and RAW264.7 cells were selected as tools for further mechanistic studies.

Fig. 6: Effect of curcumin (CUR) on HGF expression in three kinds of potential effector cells.
figure 6

a Cell viabilities of primary colonic epithelial cells (CEs), primary colonic fibroblasts (CFs), and bone marrow-derived macrophages (BMDMs) incubated with curcumin for 24 h. b mRNA expression of HGF in CEs, CFs, and BMDMs incubated with curcumin (20 μM) for 0, 12, 24, or 48 h. c Expression of HGF in CFs and BMDMs incubated with curcumin (5, 10, 20, or 40 μM) for 24 h. d Cell viability of CCD-18Co cells and the expression of HGF in CCD-18Co cells incubated with curcumin for 24 h. e Cell viability of RAW264.7 cells and the expression of HGF in RAW264.7 cells incubated with curcumin for 24 h. Cell viability was measured by the MTT assay, mRNA expression was measured by qPCR and the protein level was measured by Western blotting. The data are expressed as the means ± S.E.M. of three independent experiments (n = 3). *P < 0.05, **P < 0.01 vs. the control group (the group treated with 0 or 20 μM curcumin for 24 h).

Full size image

Curcumin induced the generation of HGF through activation of PPARγ and CREB

As peroxisome proliferator-activated receptor response element (PPRE) and cAMP response element (CRE) are located at the promoter region of the HGF gene [24,25,26], their involvement in the curcumin-mediated promotion of HGF expression was explored. Figure 7a shows that the PPARγ antagonist GW9662 (1 μM) and the CREB inhibitor KG-501 (10 μM) significantly attenuated the effect of curcumin on HGF expression in fibroblasts and macrophages, suggesting that both PPARγ and CREB are involved in the curcumin-mediated regulation of HGF expression. Subsequently, the expression of PPARγ and its downstream target gene CD36, together with the phosphorylation of CREB, were examined as indicators of PPARγ and CREB activation, respectively. The results showed that curcumin activated PPARγ and CREB in a concentration-dependent manner (Fig. 7b–d). Therefore, curcumin promotes HGF expression by activating the PPARγ and CREB pathways.

Fig. 7: Roles of PPARγ and CREB in curcumin (CUR)-mediated promotion of HGF expression.
figure 7

a Expression of HGF in CCD-18Co or RAW264.7 cells incubated with curcumin and GW9662 or KG-501 for 24 h. b mRNA and protein expression of PPARγ in CCD-18Co or RAW264.7 cells incubated with curcumin or 15d-PGJ2 for 12 h. c mRNA expression of CD36 in CCD-18Co or RAW264.7 cells incubated with curcumin or 15d-PGJ2 for 24 h. d Protein levels of pCREB (ser133) and CREB in CCD-18Co or RAW264.7 cells incubated with curcumin for 12 h. mRNA expression was measured by qPCR, and protein levels were measured by Western blotting. The data are expressed as the means ± S.E.M. of three independent experiments (n = 3). *P < 0.05, **P < 0.01 vs. the control group (the group treated with 0 μM curcumin); &P < 0.05, &&P < 0.01 vs. the group treated with 20 μM curcumin.

Full size image

Curcumin induced activation of PPARγ and CREB and the consequent generation of HGF via the generation of 15d-PGJ2

15d-PGJ2, the direct metabolite of prostaglandin D2 (PGD2), can activate the DP1 prostaglandin receptor on the cell membrane and subsequently activate the cAMP/protein kinase A (PKA) signaling pathway, thereby promoting CREB phosphorylation. Moreover, 15d-PGJ2 can enter the cell and act as an endogenous ligand to activate PPARγ [27,28,29]. Based on these findings, the effect of curcumin on the production of 15d-PGJ2 was evaluated. As shown in Fig. 8a, b, curcumin stimulated the production of 15d-PGJ2 in both fibroblasts and macrophages and in the colon in mice treated with bleomycin or SiO2. HPGDS-inhibitor-1, an inhibitor of the rate-limiting enzyme of PGD2 synthesis [30], obviously restricted activation of PPARγ and CREB, as well as the production of HGF, in the abovementioned cells, suggesting the involvement of 15d-PGJ2 in the effect of curcumin (Fig. 8c–e). An in vivo study also showed that coadministration of curcumin with HPGDS-inhibitor-1 almost completely abolished the effect of curcumin (Fig. 8f–j). Taken together, these results suggest that the stimulatory effect of curcumin on HGF expression and subsequent amelioration of PF rely on the induction of 15d-PGJ2 in the colon.

Fig. 8: Role of 15d-PGJ2 in the effect of curcumin (CUR) on the activation of PPARγ and CREB and the generation of HGF.
figure 8

a Level of 15d-PGJ2 in the supernatants of CCD-18Co or RAW264.7 cells incubated with curcumin for 12 h (n = 3). b Level of 15d-PGJ2 in the colon in PF mice (n = 6). c Expression of PPARγ and CD36 in cells incubated with curcumin or HPGDS-inhibitor-1 for 12 or 24 h (n = 3). d Protein levels of pCREB and CREB in CCD-18Co or RAW264.7 cells incubated with curcumin or HPGDS-inhibitor-1 for 12 h (n = 3). e Expression of HGF in CCD-18Co or RAW264.7 cells incubated with curcumin or HPGDS-inhibitor-1 for 24 h (n = 3). The mRNA expression was measured by qPCR, and the protein levels were measured by ELISA or Western blotting. The data are expressed as the means ± S.E.M. *P < 0.05, **P < 0.01 vs. the group treated with no curcumin or the group treated with bleomycin/SiO2; &P < 0.05, &&P < 0.01 vs. the group treated with curcumin (20 μM). f The level of HGF in the colon in bleomycin-treated mice was measured by ELISA (n = 6). g–j At the end of the in vivo experiment, the lungs of the mice were collected to determine the lung index value, measure hydroxyproline levels, and evaluate histological alterations. The data are expressed as the means ± S.E.M. (n = 6–8). ##P < 0.01 vs. the untreated group; *P < 0.05, **P < 0.01 vs. the group treated with bleomycin; &P < 0.05, &&P < 0.01 vs. the group treated with bleomycin and curcumin (150 mg/kg).

Full size image

Discussion

PF is a lethal disease caused by genetic mutation and polymorphism, aging, and extrinsic factors such as drugs, diseases, dust, and pollution and is characterized by the deposition of massive amounts of extracellular matrix [31]. At present, bleomycin and SiO2 are commonly used as representative inducers to establish animal models of PF, but the mechanisms of action differ. Bleomycin is a soluble chemical that effectively treats certain cancers, and PF is a side effect of anticancer therapy. After intratracheal administration, bleomycin directly causes damage to alveolar and airway cells through the induction of DNA strand breaks, the generation of free radicals, and the induction of oxidative stress. In the animal model, bleomycin-induced fibrosis develops quickly but is self-limiting after several weeks [32,33,34]. SiO2 is a kind of particulate, and exposure typically takes place during industrial activities, such as mining, manufacturing, and construction. In animals, SiO2 preferentially activates innate immune responses and defense against foreign bodies, which in turn leads to inflammation and fibrosis. After administration, SiO2 is retained in the lung, and the response is persistent [32, 33]. To be more comprehensive, the inhibitory effect of oral curcumin on PF was confirmed in both models, despite low blood levels of curcumin. The possibility that oral curcumin might function via absorbed blood levels was excluded by an intravenous injection assay. Considering that oral curcumin mainly accumulates in the intestine, we hypothesized that the intestine was the primary site of action of curcumin, which was confirmed by further studies, as rectal administration of curcumin exhibited an even better anti-PF effect than that of oral administration of the same dose.

The intestine is the largest endocrine organ in the human body and is capable of secreting various cytokines, growth factors, and brain-gut peptides, which cannot only regulate the physiological and biochemical activities of local intestinal cells but also enter into the circulation to modulate the function of distal tissues [35, 36]. It has been reported that some intestine-derived factors, such as IFN-γ, BMP-7, HGF, Klotho, PTEN, SST, and VIP, have significant negative effects on organ fibrosis [14, 15]. Whether curcumin could promote the secretion of antifibrotic factors in the intestine was examined. First, potential intestinal antifibrotic factors induced by curcumin were screened in bleomycin-treated mice. The results showed that curcumin did not affect the mRNA expression of IFN-γ, BMP-7, PTEN, SST, and VIP in the small intestine, colon, and lungs or the mRNA expression of HGF and Klotho in the small intestine and lungs but markedly increased the expression of HGF and Klotho in the colon tissues of mice. Unlike curcumin, however, pirfenidone had no effect on the expression of antifibrotic factors in the colon but promoted expression of antifibrotic factors in lung tissues. The differences in pharmacokinetic profiles could explain this phenomenon, as pirfenidone exhibits good absorption, and more than 80% of the oral dose is excreted in the urine [37, 38]. Of note, curcumin only enhanced the mRNA expression of antifibrotic factors in the colon, which might be attributed to the relatively long retention time of curcumin in the colon compared with the small intestine. The protein levels of HGF and Klotho were measured to determine if these factors can enter the lungs. We found that curcumin distinctively upregulated the protein levels of HGF in the colon, serum, and lungs, but the effect was much weaker in the case of Klotho. Combined with the results of the Spearman correlation analysis, it was determined that HGF but not Klotho was a potential intestine-derived effector against PF. Moreover, a promoting effect on HGF expression in the colon could also be seen in the SiO2 model. SU11274, an inhibitor of c-Met, markedly weakened the anti-PF effect of curcumin, further revealing the importance of colon-derived HGF.

Accumulating evidence has shown that HGF plays an important role in the regeneration and repair of tissues or organs after acute injury and is a hotspot of research in organ fibrosis [3]. By regulating the ERK/MAPK, PI3K/Akt, and STAT3 pathways, HGF blocks apoptosis in lung epithelial and endothelial cells, which is required for antifibrotic tissue repair [3]. The downregulation of angiotensin converting enzyme provides a potential indirect mechanism by which HGF reduces apoptosis through Ang II suppression [3]. In addition, HGF restricts the level of TGF-β1, interferes with subsequent Smad2/3 signaling by upregulating the expression of Smad7, and thus inhibits the accumulation of myofibroblasts [3]. Although the antifibrotic mechanisms of HGF have been reviewed elsewhere, the precise mechanism by which HGF is generated in the colon, however, remains unclear. To address this issue, we first uncovered the source of HGF. Among the colonic cells, mesenchymal cells such as fibroblasts, macrophages, and epithelial cells can secrete HGF [21,22,23]. qPCR screening showed that curcumin preferentially stimulated HGF expression in colon fibroblasts and macrophages. Thus, CCD-18Co and RAW264.7 cells were used as tools for further mechanistic studies.

Since HGF expression is tightly regulated at the transcriptional level, upstream regulatory elements are critical. Currently, PPRE has been identified at approximately −200 bp from the transcription initiation site in the HGF gene promoter region, and CRE was located at approximately −650 and −960 bp [24]. The transcription factor PPARγ, which is activated by ligands, enters the nucleus, and forms a heterodimer with retinoic acid receptor X, binds to PPRE and activates the transcription of the target genes [39], while CREB is phosphorylated by PKA and selectively binds to CRE in response to signals such as the increased cAMP levels [40]. It has been revealed that PPARγ agonists, adenylate cyclase activators, and cAMP analogs can promote HGF expression [24, 41, 42]; thus, PPARγ- and CREB-mediated pathways may play a role in curcumin-mediated regulation of HGF expression. The involvement of PPARγ- and CREB-mediated pathways was confirmed using GW9662 and KG-501. Further evidence suggested that curcumin stimulated the expression of HGF by activating PPARγ and CREB, as indicated by increased mRNA expression of the PPARγ target gene CD36 and upregulation of the level of pCREB.

15d-PGJ2 is an anti-inflammatory prostaglandin that also possesses robust antifibrotic activity by inhibiting TGF-β1-induced fibroblast-myofibroblast differentiation and EMT, as well as inducing senescence in activated hepatic stellate cells [43,44,45]. After the action of cyclooxygenase, arachidonic acid is converted to prostaglandin H2 via lipocalin-type prostaglandin D synthase or hematopoietic prostaglandin D synthase (HPGDS), which is subsequently converted to PGD2 and further dehydrated to 15d-PGJ2 [46, 47]. Notably, 15d-PGJ2 can activate PPARγ or CREB and has been reported to upregulate the expression of HGF in glomerular mesangial cells, renal interstitial fibroblasts and peritoneal mesothelial cells [27,28,29, 48, 49]. Based on these findings, we aimed to explain the mechanism by which curcumin activates PPARγ and CREB in the context of 15d-PGJ2 regulation. The results showed that curcumin stimulated fibroblasts and macrophages to produce 15d-PGJ2 in a concentration-dependent manner, which was consistent with the increased level of 15d-PGJ2 in the colon in mice treated with curcumin. The activation of PPARγ and CREB, the promotion of HGF expression and the anti-PF activity of curcumin were almost completely abolished by inhibition of HPGDS (fibroblasts and macrophages mainly expressed HPGDS [50, 51]), revealing the important involvement of 15d-PGJ2 in the effect of curcumin. It should be noted that in addition to fibroblasts and macrophages, intestinal epithelial cells, glial cells, and mast cells are able to produce 15d-PGJ2 [51], suggesting that 15d-PGJ2 could regulate the generation of HGF through autocrine or paracrine actions. Due to the extremely short half-life of prostaglandins, it was thought that 15d-PGJ2 was less likely to enter the lung from the colon and directly ameliorate PF.

In conclusion, the intestinal tract is the primary site of action through which curcumin exerts its anti-PF effects. Curcumin promotes the expression of HGF in colonic fibroblasts and macrophages by activating the PPARγ and CREB pathways via the induction of 15d-PGJ2, and HGF enters the lung and exerts an anti-PF effect.