PRMT1 Upregulated by Epithelial Proinflammatory Cytokines Participates in COX2 Expression in Fibroblasts and Chronic Antigen-Induced Pulmonary Inflammation

Asthma is a heterogeneous and multifactorial disorder characterized by recurrent episodes of airflow obstruction. The histopathology features of asthma include the dysfunction of the bronchial epithelium, exaggerated bronchoconstrictor response to a wide variety of stimuli (airway hyperresponsiveness), chronic airway inflammation, and structural changes of the bronchial wall (1).

Protein arginine methylation is performed by a class of enzymes called protein arginine methyltransferases (PRMTs) and is a novel posttranslational protein modification that plays a pivotal role in various intracellular events, such as signal transduction, protein–protein interaction, and transcriptional regulation. PRMT mediates its function either through direct regulation of protein–protein interaction or by arginine methylation, both of which influence NO-dependent processes. A growing body of evidence suggests that both mechanisms are implicated in cardiovascular and pulmonary diseases, including lung cancer, pulmonary fibrosis, pulmonary hypertension, chronic obstructive pulmonary disease, and asthma (2, 3). In our previous study, we have elucidated that IL-4 upregulated PRMT1 in airway epithelial cells, thereby increasing eotaxin-1 expression. Furthermore, pulmonary inflammation in rats waned after inhibition of PRMT activity by AMI-1 (3). However, the distinct regulation of PRMT1 in acute Ag-induced pulmonary inflammation (AIPI) and in chronic AIPI was not investigated.

Fibroblasts, as a major source of interstitial connective tissue extracellular matrix, contribute to the fibrotic changes in asthmatic airways and in Th2 cell–driven inflammation. Various elements of the innate and adaptive immune response participate in the differentiation and activation of subepithelial bronchial fibroblasts in asthma (4). Airway remodeling and inflammation persist in asthma, and the consequences are inappropriate airway function. Damaged bronchial epithelium tissue repairs incompletely and leads to chronic wound repair with increased secretion of a range of secondary growth factors and cytokines that drive fibroblast activation and airway remodeling (5).

The most prominent factors regulating fibroblast function include epithelial-derived growth factor, TGF-β, platelet-derived growth factor, basic fibroblast growth factor, and endothelin, several of which are capable of inducing subepithelial fibroblast proliferation, differentiation, and activation of myofibroblasts. A recent study showed increased levels of TGF-β and asymmetric ADMA, a main PRMT metabolite, in fibrotic tissues in diabetic rats (6). ADMA, as an endothelial NO synthase and inducible NO synthase inhibitor, is involved in proline metabolism and may play an important role in pulmonary fibrosis during airway remodeling (7). In an earlier study we examined PRMT expression in rats with different stages of AIPI. In these animals, PRMT1 was highly expressed in the tissue surrounding the airway, mainly in fibroblasts and airway smooth muscle cells, whereas it was expressed in the bronchial epithelium of rats with acute AIPI. Therefore, the shift of PRMT1 expression from the epithelium to subepithelial mesenchymal cells may be a key event in lung remodeling in chronic AIPI. Consequently, PRMT may be involved in chronic airway remodeling through its function in fibroblasts.

Although the functional characteristics of asthmatic fibroblasts and their interaction with epithelial inflammation on airway wall remodeling is a consistent pathology observation in chronic asthma, very few studies have investigated the role of PRMT1 on fibroblast function during airway remodeling. In this study, we measured the expression and location of PRMT1 in acute and chronic AIPI in E3 rats. Moreover, we studied the effect of IL-4–activated epithelial cells on the proliferation and cytokine production by fibroblasts. Additionally, the inhibition of PRMT1 by AMI-1 was used to assess the role of PRMT in several asthma and airway remodeling parameters. We conclude that IL-4 stimulates epithelial cells, which, in turn, upregulate PRMT1 in subepithelial fibroblasts, thus contributing to the pathogenesis of chronic inflammation in asthma.

E3 rats were bred in a specific pathogen-free animal house. Age- and sex-matched rats were used for all experiments, and each group contained eight rats at the age of 8–10 wk. The experiments were approved by the Institutional Animal Ethics Committee of Xi’an Jiaotong University.

Rats were immunized by i.p. injection with 1 ml emulsion solution containing 1 mg OVA (Sigma-Aldrich, St. Louis, MO) and 50 mg Al(OH)₃ (Pierce Biotechnology, Rockford, IL). For screening of PRMT1 expression in lungs, 24 rats were divided into a control group, an acute AIPI group (challenged with OVA every day for 1 wk), and a chronic AIPI group (challenged with OVA every other day for 8 wk).

For the AMI-1 treatment experiment, 24 rats were divided into three groups: control group, chronic AIPI group, and AMI-1 group. Two weeks after sensitization, control group rats were sham sensitized and exposed to the same volume of PBS. In the AMI-1 group, rats were administered 50μl AMI-1 (Calbiochem, San Diego, CA) at a concentration of 0.1 mg/ml in PBS 2 h before OVA challenge. The asthma index included serum levels of OVA-specific IgG1 and total serum IgE, which were determined by ELISA as described in previous studies (3, 8).

The mRNA expression of PRMT1 and cytokine genes was analyzed by real-time quantitative PCR (RT-qPCR), performed by using a iQ5 real-time PCR detection system (Bio-Rad, Hercules, CA) with SYBR Premix Ex Taq II (TaKaRa Bio, Shiga, Japan). The relative gene expression was normalized by GAPDH expression. The information for all primers used in this study is shown in Table I.

Lung tissues were stained with H&E for pathological changes, with periodic acid–Schiff (PAS) to detect mucus production, and with Masson staining to determine collagen fibers as previously described (8).

Immunohistological staining tissue sections were incubated with 100-fold diluted anti-PRMT1 Ab or anti–TGF-β Ab (Abcam, Cambridge, U.K.) in blocking solution at 4°C overnight, followed by a two-step plus poly-HRP anti-goat IgG detection kit (ZSGB-Bio, Beijing, China). The intensity of the brown color was determined by using Image-Pro Plus 6.0 software to estimate protein expression in lung tissues.

Human fibroblast from lung (HFL1) cell and human A549 alveolar epithelial-like cells (A549) were grown in F12K or RPMI 1640 (Invitrogen, Grand Island, NY) supplemented with 10% FBS (HyClone, Logan, UT). Human rIL-4 (Boster, Wuhan, China) was added into the wells (six-well plates) at 100 ng/ml. The culture supernatant of IL-4–stimulated A549 cell was accumulated into a 15-ml EP tube and centrifuged for 10 min to remove cells and cell debris. FBS (5%) was added to the culture supernatant to replenish the consumption by epithelial cells, and then the IL-4–stimulated epithelial medium was prepared and assigned as IL-4–stimulated lung epithelial cell (A549) medium (ISEM). The ISEM was stored at 4°C for 2 d and at −20°C for 1 mo. The ISEM and basic F12K media were mixed at various ratios of 0:1, 1:5, 1:3, 1:1, 3:1, and 5:1 to stimulate HFL1 cells. TGF-β cytokine (10 ng/ml; PeproTech, Rocky Hill, NJ) or 1 μg/ml anti–TGF-β Ab (Abcam) was mixed with ISEM to investigate TGF-β effects on the regulation of PRMT1, COX2, and vascular endothelial growth factor (VEGF) expression.

MTT was used to determine HFL1 cell proliferation. HFL1 cells were seeded into 96-well plates at a density of 1 × 10⁴ cells/100 μl per well. Proliferation was determined 48 h after ISEM treatment by adding 50 μl MTT reagent and incubating for 4 h at 37°C. After removal of the culture medium, cells were lysed with DMSO to determine the amount of formazan product (9), and the absorption was measured by a microplate reader (Thermo Fisher Scientific, Vantaa, Finland) at 550 nm.

The full-length cDNA of PRMT1 was amplified from HFL1 cell mRNA by RT-PCR with primers (5′-CGGGATCCCGATGGCGGCAGCCGAG-3′ and 5′- CCCTCGAGGGTCAGCGCATCCGGTAGTCGG-3′) and was cloned into a pcDNA3.1(+) vector to construct the pcDNA3.1-PRMT1 recombinant plasmid. With Lipofectamine 2000 (Invitrogen), the recombinant plasmid pcDNA3.1-PRMT1 was transiently transfected into HFL1 cells for 24 or 48 h. The expression of PRMT1, COX2, and VEGF in the HFL1 cells was measured by RT-qPCR.

HFL1 cells (2 × 10⁵/well) were seeded into six-well plates and grown for 12 h. ISEM and basic F12K medium were mixed at various ratios (0:1, 1:5, 1:3, 1:1, 3:1, 5:1) to stimulate HFL1 cells for 24 or 48 h, before lysis with RIPA buffer (Beyotime, Beijing, China). Cell lysates were centrifuged at 12,000 rpm (15 min) and the protein supernatant was kept. The protein concentration was quantified by the BCA method (Beyotime) and equal amounts of denatured proteins (20 μg) were separated by SDS-PAGE and subsequently electrotransferred onto polyvinylidene difluoride membranes. The proteins of interest were detected with Abs specific to either PRMT1 (Abcam), COX2 (Cell Signaling Technology, Beverly, MA), or VEGF (Santa Cruz Biotechnology, Santa Cruz, CA) and visualized after binding of a secondary Ab conjugated with HRP (Abcam) by ECL reagents (Pierce Biotechnology).

Data are expressed as mean ± SEM. The statistical analysis was performed by a Mann–Whitney U test for the comparison between groups. PRMT1, COX2, and remodeling related cytokine expressions were analyzed by one-way ANOVA for comparison. The null hypothesis was that no treatment had any effect, and a p value <0.05 was taken as significant.

In our previous study, we found that in acute AIPI, PRMT1 is expressed only by epithelial cells of the airways and the alveoli. In this study, immunohistochemical staining was performed to visualize PRMT1 protein expression in lung tissue sections obtained from chronic AIPI. The results demonstrated that PRMT1 was expressed in the subepithelial airway structures, mainly in fibroblasts and smooth muscle layers (Fig. 1A, 1B). The analysis of immunohistochemical staining showed that in acute AIPI, epithelial cells expressed PRMT1, and in chronic AIPI, increased PRMT1 was located in the subepithelial airway tissues. Importantly, PRMT1 mRNA and protein expression in acute AIPI and chronic AIPI significantly increased compared with tissue of control rats (Fig. 1C).

To explore the different effects of PRMT1 function in acute AIPI and chronic AIPI, IL-4 (100 ng/ml) was used to stimulate HFL1 cells at the concentration of 100 ng/ml. Against our expectations, PRMT1 expression did not show a significant increase after IL-4 stimulation (data not shown). Considering that epithelial cell inflammation modifies subepithelial fibroblast function in chronic asthma, the supernatant of ISEM was collected after 48 h and then was diluted with F12K basic medium at various ratios (0:1, 1:5, 1:3, 1:1, 3:1, 5:1) to stimulate HFL1 cells. HFL1 cell proliferation increased significantly after stimulation with the mixed medium described above for 48 h (Fig. 2A). The mRNA expression of PRMT1 (the information on the primers is shown in Table I) in HFL1 cells increased after 48 h of ISEM treatment (Fig. 2B). In parallel, the protein expression of PRMT1 was enhanced after 24 and 48 h of ISEM treatment (Fig. 2C). Additionally, both the supernatant from A549 and BEAS-2B, a normal lung epithelial cell line, had a similar effect on fibroblast proliferation and PRMT1 expression.

To confirm that ISEM induces remodeling driving cytokines in fibroblasts, we determined the expressions of cox2, vegf, tgf-a, tgf-b, egfr, and b-fgf mRNAs (data not shown). Additionally, the mRNAs (the information on the primers is shown in Table I) and proteins expression of COX2 and VEGF increased in HFL1 cells after the stimulation with ISEM (Fig. 3A–C). To explore the effects of PRMT1 on the expression of cox2 and vegf, a recombinant plasmid pcDNA3.1-PRMT1 and plasmid pcDNA3.1+ as mock were used to transfect HFL1 cells and grow them for 24 or 48 h. After the transfection of plasmid pcDNA3.1-PRMT1, the expression of PRMT1 was significantly upregulated in HFL1 cells; upregulated cox2 and vegf mRNA expressions were detected together with PRMT1 upregulation (Fig. 3D).

To identify which component contained in ISEM induced the PRMT1 and COX2 expression in fibroblasts, the immunohistology was performed in lung tissues from control, acute, and chronic AIPI rats. In the lung tissue of acute AIPI rats, TGF-β was mainly expressed in airway epithelial cells, whereas in the lung tissue of chronic AIPI rats, TGF-β was mainly expressed in subepithelial structure as smooth muscle cells and fibroblasts (Fig. 4A). Additionally, IL-4 was used to stimulate A549 epithelial cells for 3, 6, 12, 24, or 48 h. As shown in Fig. 4B, the TGF-β expression by epithelial cells was significantly increased by IL-4 stimulaiton. Then, different concentrations of TGF-β (1, 5, or 10 ng/ml) were used to stimulate HFL1 for 48 h and the expression of PRMT1 and COX2 was detected by Western blotting. The expression of PRMT1 and of COX2 was dose-dependently upregulated by TGF-β (5 or 10 ng/ml after 48 h), whereas the expression of VEGF was enhanced by all concentrations of TGF-β (1, 5, or 10 ng/ml) as depicted (Fig. 4C).

Consequently, human recombination TGF-β and a neutralizing anti–TGF-β Ab were mixed with ISEM and the expression of prmt1, cox2, and vegf was determined (Fig. 5A–C). The data showed that the abrogation of TGF-β by a specific neutralizing Ab attenuated the expression of prmt1, cox2, and vegf with ISEM stimulation, whereas supplementation of TGF-β increased expression of these genes. Meanwhile, the protein expressions of PRMT1 and COX2 were elevated by ISEM, which can be neutralized by anti–TGF-β Ab (Fig. 5D, 5E). Additionally, it is noteworthy that the anti–TGF-β Ab did not abrogate the increase of PRMT1 expression by ISEM completely, suggesting that additional components upregulate the expression of PRMT1. COX2 expression by HFL1 cells was also detected by Western blotting after stimulation with 5 ng/ml TGF-β with or without AMI-1 (5 and 10 μM). The results showed that 10 μM AMI-1 counteracted the stimulatory effect of TGF-β on COX2 expression (Fig. 5F).

We determined the effects of AMI-1 in chronic AIPI rats. The mRNA expressions of cox2 (Fig. 6A) and vegf (Fig. 6B) were detected by RT-qPCR in lung tissues from control rats, chronic AIPI rats, and chronic AIPI rats with administration of AMI-1. Cox2 expression in lung tissues returned to normal levels in the AMI-1–treated rats; however, the vegf expression did not fully reverse with administration of AMI-1. Serum IgE concentrations (Fig. 6C) returned to baseline after AMI-1 administration, but in either case no difference in serum NO concentration was observed (Fig. 6D).

Next, histopathological analysis of airway remodeling in chronic AIPI rats with and without the administration of AMI-1 was performed. In the chronic AIPI group, the lesions of airways and alveoli were serious. Extensive pulmonary emphysema emerged, and alveolar ectasia, disintegration, and narrowing of alveolar septum were presented when compared with control rat lungs (Fig. 7A). The secretion of mucus by using PAS staining was obvious in chronic AIPI rats (Fig. 7B). Additionally, Masson staining indicated a significant increase of collagen around airways in the AIPI group (Fig. 7D). Importantly, the nasal administration of AMI-1 decreased the airway and alveoli lesions, mucus secretion, and collagen deposition in chronic AIPI lungs. We scored the lung tissue of individual rats and found that the diameter of alveoli, mucus secretion, and collagen deposition were attenuated after AMI-1 administration (Fig. 7C, 7E).

Taken together, these results indicated that PRMT inhibition in chronic AIPI rats reduced several asthma parameters and ameliorated the disease severity.

In this study we observed that IL-4–stimulated epithelial cells produced TGF-β, which affected the proliferation of fibroblasts and elevated the production of PRMT1, COX2, and VEGF by fibroblasts. Additionally, TGF-β also increased the expressions of COX2 and VEGF through PRMT1. Our in vitro findings may explain that the mechanism PRMT1 is regulated by TGF-β in the epithelium in acute AIPI, whereas PRMT1 is expressed in subepithelial mesenchymal cells in chronic AIPI. Furthermore, we provide evidence that the inhibition of PRMT in vivo dampened most remodeling-related asthma indices, such as collagen deposition, mucus secretion, serum IgE, and the expression of remodeling-related cytokines. Thus, we conclude that PRMT1 plays an important role in chronic asthma through the regulation of remodeling relevant cytokines.

Asthma is characterized by variable degrees of chronic inflammation and structural alterations in the airway walls. The most prominent pathologies include epithelial denudation, subepithelial tissue thickening, increased airway smooth muscle mass, and alterations of extracellular matrix components (10). Chronic inflammation is thought to initiate and perpetuate cycles of tissue injury and repair in asthma, although remodeling may also occur in parallel with inflammation. In recent years, a significant body of information on a close relationship between epithelial inflammation and airway remodeling has emerged showing that chronic injury or defective repair of the bronchial and alveolar epithelium results in its persistent activation, with consequent chronic secretion of a variety of proinflammatory cytokines and growth factors that further drive chronic inflammation and remodeling in the subepithelial compartments (11).

Our data clearly demonstrated that PRMT1 participates in both the inflammation and remodeling process in asthma. At the beginning of inflammation, increased PRMT1 occurs mainly in epithelial cells attracting eosinophil infiltration and it exacerbates inflammation in acute AIPI through the upregulation of eotaxin-1. However, in chronic AIPI, PRMT1 expression is observed mainly in subepithelial fibroblasts. Interestingly, IL-4–stimulated inflammatory epithelial cells produced TGF-β, thereby inducing fibroblast proliferation and elevated PRMT1 expression. This, in turn, upregulated its downstream target genes, cox2 and vegf. Our data suggest that PRMT1 is involved in both early epithelium inflammation and later subepithelial remodeling.

Recently it was reported that inflammation can be viewed as a response of lung tissue to injury, and thereby the lung aims to repair the injury. Among the large variety of cytokines and chemokines produced by bronchial epithelial cells, TGF-β is increasingly recognized as an important factor that induces proliferation of subepithelial fibroblasts and as a driving force of differentiation and activation of myofibroblasts. Redington et al. (12) also reported that TGF-β levels were increased in bronchoalveolar lavage fluid of asthma patients after segmental allergen challenge. Similarly, other studies confirmed that epithelial cells from asthma patients released higher levels of TGF-β than did epithelial cells obtained from nonasthma subjects (13, 14). Furthermore, TGF-β expression correlated with the degree of subepithelial fibrosis and was significantly increased in subjects with severe asthma and associated eosinophilia (15). It has also been shown that mechanical or chemical damage to the epithelium leads to increased release of TGF-β1 and TGF-β2 (16). In this regard, our in vivo study showed that TGF-β production was significantly increased by lung epithelial cells in acute AIPI lungs. In lungs with chronic AIPI, TGF-β was mainly expressed by subepithelial smooth muscle cells and fibroblasts, suggesting its important role in the maintenance of growth and proliferation of fibroblasts, as well as its participation in airway remodeling in chronic asthma. Additionally, the in vitro study proved that IL-4–stimulated inflammatory epithelium created an active TGF-β milieu that regulated the proliferation and differentiation of subepithelial fibroblasts. Therefore, our novel observation supports the hypothesis that the bronchial epithelium has a key role in modulating the bronchial wall structure and inflammation in asthma.

Our in vitro study proved that increased production of TGF-β by inflamed A549 epithelium induced the expression of PRMT1 in fibroblasts, whereas the blockade of TGF-β partially inhibited the expression of PRMT1 by ISEM. Additionally, both A549 and BEAS-2B, a normal epithelial cell line, had a similar effect on fibroblast proliferation and PRMT1 upregulation. These findings indicate that TGF-β is a very important inducer of PRMT1 in inflammation. Although there is no direct evidence to prove the relationship between TGF-β and PRMT1, a study on renal fibrosis was found that ADMA, as an endogenous endothelial NO synthase and inducible NO synthase competitive inhibitor (17, 18), can reduce the NO concentration in serum in asthmatic animals (7); furthermore, ADMA is also involved in proline metabolism, which may play an important role in the pulmonary remodeling process through raising collagen synthesis (19). In our study, the supplementation of TGF-β to basic medium increased the expression of PRMT1 in fibroblasts. However, our data also indicated that TGF-β is not the only cytokine regulating PRMT1 expression because blocking TGF-β activity in ISEM did not abolish all PRMT1 expression with ISEM stimulation in fibroblasts.

In chronic inflammation, COX2 immunoreactivity was reported in the bronchial mucosa of normal and asthmatic lungs (20), as well as in the bronchial epithelium (21). The observed overexpression of COX2 in normal human lung fibroblasts is important, as it may drive inflammation and fibrotic conditions in lung (22). The increase of COX2 expression is often accompanied by NF-κB activation (23, 24) in human fibroblasts. However, when mice were treated with pyrrolidine dithiocarbamate, a nonspecific NF-κB inhibitor inhibiting COX2 transcription, decreased COX2 mRNA synthesis and protein were reported in epithelial cells (25). NF-κB is an important transcription factor for IL-1β–induced COX2 gene expression, and it is involved in inducing COX2 gene transcription (26). Our results demonstrated that COX2 is a target gene of PRMT1, and the upregulation of PRMT1 by plasmid transfection in fibroblasts increased COX2 expression, whereas PRMT enzyme activity inhibited by AMI-1 can reduce the TGF-β–induced COX2 upregulation. Others have demonstrated that a cooperative action of PRMT1 with CARM1 is required for NF-κB–dependent gene expression (27). Therefore, we speculate that PRMT1 regulates COX2 expression through the activation of NF-κB.

Our previous in vitro experiments in acute AIPI rats proved that AMI-1 ameliorates pulmonary inflammation through downregulating eotaxin-1 expression, eosinophil infiltration, and epithelial cell exfoliation (3). Eosinophils and injured epithelium have been cofirmed as important sources of TGF-β as well as of other important cytokines that activate the epithelium and subepithelial mesenchymal cells, thus driving airway remodeling. Interestingly, we have illustrated in chronic AIPI that AMI-1 reduced COX2 expression in lung tissue, IgE concentrations in serum, and ameliorated histopathological remolding.

In conclusion, TGF-β produced by IL-4–stimulated epithelium elevates PRMT1 expression, which plays a crucial role in chronic AIPI through its regulation of COX2. The findings, summarized in Fig. 8, may offer novel insights into the pathogenesis of asthmatic airway remodeling and suggest a new target to intervene in asthma.

We are grateful to Yan Han, Qilan Ning, and Fujun Zhang for expert assistance, and Liesu Meng and Wenhua Zhu for helpful discussions and productive critiques.

The authors have no financial conflicts of interest.