Prevention of bleomycin-induced pulmonary fibrosis by a RANKL peptide in mice

Despite the recent therapeutic developments for the treatment of pulmonary fibrosis, its prognosis is still not well controlled, and a novel therapeutic agent is needed. Recently, the critical role of Toll-like receptors (TLRs) in the pathophysiology of pulmonary fibrosis has been reported; however, the effects of multiple TLR signaling inhibition are still unknown. Here, we examined how the inhibition of multiple TLRs affects pulmonary fibrosis using a novel synthetic receptor activator of nuclear factor κB ligand (RANKL) partial peptide, MHP1-AcN, which could suppress TLR2, 3, 4, 7, and 9 signaling through CD14 and RANK. When MHP1-AcN was administered in the bleomycin-induced lung fibrosis model, reduced collagen deposition was observed, with suppressed fibrosis-related gene expression including Col1a1, Col1a2, Acta2, Tgfb1 and Tgfbr2. MHP1-AcN also decreased proinflammatory M1 and profibrotic M2 macrophage marker expression. Furthermore, MHP1-AcN treatment inhibited transforming growth factor (TGF-β)-induced Smad2/3 phosphorylation and myofibroblast differentiation in human fetal lung fibroblast (MRC-5) cells. This effect was associated with decreased TGF-β receptor levels and the upregulated Bmp7 and Smad7 expression. These findings suggest that MHP1-AcN protects mice against bleomycin-induced pulmonary fibrosis. MHP1-AcN might provide a novel therapeutic strategy for the pulmonary fibrosis.


Results
MHP1-AcN protected mice against bleomycin-induced lung fibrosis. First, we examined whether intraperitoneal injection of MHP1-AcN could prevent intrathecal bleomycin-induced lung fibrosis in mice. The body weight (BW), lung weight (LW), and the ratio of LW to BW are widely used for monitoring the health status of bleomycin-injured mice 18,19 . Compared to bleomycin-treated mice that received saline, the mice receiving MHP1-AcN showed faster recovery of BW loss (Fig. 1A), decreased loss of LW and the ratio of LW/BW on day 14 (Fig. 1B). Masson's trichrome staining showed that bleomycin induced substantial collagen deposition in lungs of mice, which was significantly reduced by MHP1-AcN treatment (Fig. 1C). MHP1-AcN-treated mice also showed a tendency for a reduction of hydroxyproline, which is a major component of collagen (Fig. 1D). Consistent with these data, MHP1-AcN-treated mice showed suppressed expression of fibrosis-relate gene expression including Col1a1, Col1a2, Acta2, and Tgfb1 compared to saline-treated mice (Fig. 2). The reduced COL1A1, COL1A2 (Fig. 3A, Supplementary Fig. 1A) and α-SMA (Fig. 3B) were also confirmed by western blotting or immunohistochemical analysis. Interestingly, bleomycin-treated mice that received saline showed an  www.nature.com/scientificreports/ increase in Tgfbr2 expression, which was inhibited by MHP1-AcN treatment, although the expression of Tgfbr1 was not affected even in the mice receiving saline (Fig. 2). Because CD14 serves as a co-receptor for multiple TLRs 20 , we also examined CD14 expression. As MHP1-AcN was previously shown to inhibit the expression of CD14 in LPS-stimulated bone marrow-derived macrophages 13 , the upregulated CD14 expression in lung by bleomycin was inhibited by MHP1-AcN. In general, CD14 is expressed in macrophages, whose dysfunction plays a critical role in promoting lung fibrosis 21 . Sustained activation of proinflammatory M1 macrophages by acute lung injury initiates fibrotic responses in the lung, while the infiltration of profibrotic M2 macrophages in Figure 2. MHP1-AcN prevented the progression of bleomycin-induced lung fibrosis. Expression of fibrosis (Col1a1 and Col1a2), myofibroblast (Acta2 (αSMA)), TGF-β signaling (Tgfb1, Tgfbr1, Tgfbr2), M1 marker (Nos2 (iNOS), Cd86), M2 marker (Cd206, Arg1)-related genes, and Cd14 on day 14. There were no differences in Gapdh. *p < 0.05, **p < 0.01 vs saline-treated mice. n = 3-6 in sham-operated group, n = 6 in saline-treated group, n = 5 in MHP1-AcN-treated group. www.nature.com/scientificreports/ the lung functions as a critical regulator in the aberrant development of lung fibrosis 22 . From the viewpoints, we examined the effects of MHP1-AcN on macrophages in fibrotic lung 23 . The bleomycin-treated mice that received MHP1-AcN showed suppressed M1 markers (Nos2 and Cd86) and M2 markers (Cd206 and Arg1) expression compared to saline-treated mice; however, Arg1 did not reach significant difference (Fig. 2). Immunohistochemical analysis also showed reduced percentage of CD11c-positive M1 (Fig. 3C) and CD206-positive M2 macrophages (Fig. 3D). Collectively, these results indicate that MHP1-AcN protected mice against bleomycininduced lung fibrosis.
Next, we reasoned how a RANKL peptide could affect TGF-β/Smad signaling. Because the expression of Tgfbr2 mRNA was reduced in the mice-treated with MHP1-AcN, we examined whether MHP1-AcN had influence on its expression in MRC-5 cells. When MRC-5 cells were treated with MHP1-AcN for 24 h, TβRI as well as TβRII expression was suppressed at both mRNA level (Fig. 5A) and protein level (Fig. 5B, Supplementary  Fig. 1E). Since BMP7 was reported to inhibit TGF-β signaling by suppressing Smad2/3 phosphorylation and/or promoting TGF-β receptor degradation through Smad7 24-27 we next checked the expression level of Bmp7 and Smad7 mRNA. As expected, MHP1-AcN treatment led to a significant increase in Bmp7 and Smad7 in MRC-5 cells (Fig. 5C). These data indicate that the inhibitory effect of MHP1-AcN on TGF-β signaling is, at least in part, mediated through upregulating BMP7 and Smad7, which might promote the degradation of TGF-β receptors.

MHP1-AcN inhibited TLR9-induced IL6 secretion in fibroblasts.
In addition to the stimulation of TGFβ signaling, chronic TLR9 stimulation in human lung fibroblasts was reported to cause α-SMA-positive myofibroblast transformation, with the secretion of inflammatory cytokines 28 . To clarify whether MHP1-AcN affects the TLR signaling in fibroblasts, we checked the IL-6 levels in MRC-5 cells after TLR9 (E. coli DNA) or TLR4 (LPS) stimulation ( Supplementary Fig. 2). Unexpectedly, LPS did not affect the secretion of IL-6, whereas E. coli DNA-induced IL-6 was inhibited by MHP1-AcN. Thus, the inhibitory effect of MHP1-AcN on TLR9 signaling in fibroblasts might also contribute to the prevention of bleomycin-induced lung fibrosis. MHP1-AcN protected mice against bleomycin-induced skin fibrosis. Finally, we examined bleomycin-induced skin fibrosis to check whether the effectiveness of MHP1-AcN was limited to the lung. After   www.nature.com/scientificreports/ injection of bleomycin for 14 days, MHP1-AcN was intraperitoneally injected daily for 14 days. Masson trichrome staining showed that bleomycin-induced dermal thickness was inhibited in MHP1-AcN-treated mice (Fig. 6A). Similar to the results in lung fibrosis, expression of Col1a1, Col1a2, and Acta2 was inhibited in MHP1-AcN-treated mice (Fig. 6B). These results indicate that the effectiveness of MHP1-AcN is not limited to lung fibrosis.

Discussion
In the present study, we verified the preventative effects of MHP1-AcN on bleomycin-induced pulmonary fibrosis as assessed by reduced collagen deposition and reduced number of M1 and M2 macrophages in the lungs. MHP1-AcN also suppressed TGF-β and TLR9 signaling in MRC-5 cells and bleomycin-induced skin fibrosis. Inhibition of myofibroblast differentiation is a practical strategy to treat fibrosis-related diseases because myofibroblasts secret various cytokines 29 . Although the effectiveness of multiple inhibition of TLR signaling for myofibroblast differentiation has not been reported, TLR2, TLR4, and TLR9 have been reported to promote myofibroblast differentiation [30][31][32] . In skin fibroblasts, LPS-triggered TLR4 activation alone did not activate TGF-β/ Smad signaling but promoted differentiation into myofibroblasts with enhanced TGF-β responses in the presence of TGF-β 33 . However, fibronectin extra domain A or tenascin-C-triggered TLR4 activation induced fibrotic responses in human skin fibroblast, which were abrogated by TLR4 inhibitor 31 . Considering that LPS did not affect IL-6 production in MRC-5 cells in our study, TLR4 activation, which could promote fibrotic responses, might be ligand-specific and cell type-specific. Despite no involvement of TLR4, the inhibition of IL-6 production was shown in TLR9-stimulated MRC-5 cells. This indicated that the protective effects of MHP1-AcN might also be attributed to the direct inhibition of TLR9 signaling in fibroblasts.
In general, TGF-β1 is produced by a wide variety of cell types, including endothelial cells, fibroblasts, myofibroblasts, alveolar macrophages, neutrophils, and activated alveolar epithelial cells 34 . In the present study, the cells responsible for the decreased Tgfb1 expression were not clarified, but the reduction of Cd206 and Arg1 expression indicated a decrease in M2 macrophages, which reportedly produced profibrotic mediators, such as TGF-β and PDGF 22 , as one of the causes. TGF-β signals are transduced by TGF-β receptors to modulate the profibrotic effects including myofibroblast differentiation, fibroblast proliferation and extracellular matrix accumulation 35 . In the present study, MHP1-AcN decreased TβRI and TβRII expression in fibroblasts, but in bleomycin-induced www.nature.com/scientificreports/ lung fibrosis model, MHP1-AcN treatment inhibited the increase in Tgfbr2 without significant change in Tgfbr1 on day 14 after bleomycin exposure. The reasons for these discrepancies are not clear, but the time course of Tgfbr1 is likely to contribute, as TβRI was reported to be significantly upregulated on day 7 post bleomycin 36 . The increased expression of Bmp7 and Smad7 indicated that the degradation of TGF-β receptors might be one of the mechanisms of reduced expression of TGF-β receptors, but how MHP1-AcN affects these molecules and the link between TLR signaling and TGF-β signaling needs further investigation. In previous studies, BMP7 was reported to have beneficial effects in many fibrotic models including renal fibrosis and cardiac fibrosis 37,38 , but BMP7 administration was insufficient to protect mice against bleomycin-induced pulmonary or skin fibrosis 39 . Therefore, the combination of anti-TLR signaling and upregulation of Bmp7 by MHP1-AcN might be ideal for the prevention of bleomycin-induced pulmonary fibrosis. The limitation of this study is that we could not exclude the possibility that the anti-inflammatory effects of MHP1-AcN in bleomycin-induced lung fibrosis model might be the cause of anti-fibrotic effects because we administered MHP1-AcN from the beginning of bleomycin. In this model, bleomycin resulted in damage to the alveolar epithelial cells and subsequent inflammatory responses during the first 7 days, and the transition from inflammation to fibrotic processes at 7-14 days post bleomycin exposure 40 . MHP1-AcN might have exerted its anti-fibrotic effects only through its anti-TLR effects for the initial injury and inflammatory responses. However, we believe that the effects of MHP1-AcN was dependent on anti-fibrosis as well as anti-inflammatory effects because MHP1-AcN suppressed TGF-β/Smad signaling and TLR9 signaling in MRC-5 cells. Another limitation of this study is examining the effect of MHP1-AcN only in a bleomycin model. Although the bleomycin-induced lung fibrosis model has been extensively used to evaluate potential therapies 41 , a single animal model does not recapitulate the spectrum of human pulmonary fibrosis. For clinical applications, further study is necessary in another animal model, such as silica-induced lung fibrosis model. Another limitation is the limited information on the action mechanism of MHP1-AcN for bleomycin-induced skin fibrosis. Although we speculate that MHP1-AcN suppresses skin fibrosis with a similar mechanism to lung fibrosis, it is necessary to clarify how MHP1-AcN prevents skin fibrosis for clinical applications.
Taken together, this study demonstrated the protective effects of MHP1-AcN in bleomycin-induced lung fibrosis model. The mechanism of this effect involved inhibition of TLR signaling and, at least in part, inhibition of TGF-β signaling via upregulation of BMP7 and Smad7 in fibroblasts.

Animal models of fibrosis. All experiments were approved by the Institutional Animal Care and Use
Committee of Osaka University (02-017-010) and carried out in compliance with the Osaka University Guidelines, which are based on the National Institutes of Health's Guide for the Care and Use of Laboratory Animals, and with the ARRIVE guidelines. All efforts were made to minimize suffering. C57/Bl6/J female mice were obtained from CLEA Japan, Inc.
Induction of mouse lung injury by endotracheal injection of bleomycin was performed as previously described 42 . Briefly, female mice (12-13 W) were anesthetized with a mixture of butorphanol (5.0 mg/kg), medetomidine (0.3 mg/kg), and midazolam (4.0 mg/kg). After incision of the skin in the neck, the trachea was exteriorized by blunt dissection. Bleomycin solution (0.05 U/100 µL, Nippon Kayaku, Tokyo, Japan) was slowly instilled into the lumen of the trachea. The skin wound was then closed with a 5-0 absorbable suture. MHP1-AcN (40 mg/kg) was injected intraperitoneally for 14 days and sacrificed on day 14. The right lung was collected, and its weight was measured.
Skin fibrosis was induced as described previously 4 . Briefly, 8-week-old female mice received subcutaneous injections of bleomycin (10 mg/kg/day) for 10 days (5 days/week). Intraperitoneal injection of MHP1-AcN (40 mg/kg) was started on day 14 and continued to be injected daily for 14 days. The mice were sacrificed on day 28.
Histological analysis and measurement of hydroxyproline. For Masson's trichrome staining, the left lungs were perfused and fixed with 4% paraformaldehyde. They were embedded in paraffin, and serial 5-µm sections were stained with Masson's trichrome stain. For immunohistochemistry, antigen retrieval was performed. The sections were blocked in 3% bovine serum albumin, followed by incubation with primary antibodies (CD11c: 97585S, Cell Signaling; CD206: ab64693, Abcam; αSMA: ab5694, abcam) overnight at 4 °C. Then the sections were incubated with Single Stain MAX PO (rabbit) Histofine (Nichirei Biosciences, Japan) followed by development with DAB (DAB substrate kit/NICHIREI). Counterstaining was performed with Meyer's hematoxylin. Images were captured using a microscope (BZ-X810, Keyence, Japan). For semi-quantification of the lung fibrosis, fibrosis was quantified using the modified Ashcroft scoring 43 . The average percentage of CD11c, CD206, and αSMA-positive cells in randomly selected 3 fields of 0.4 mm 2 area was quantified using ImageJ (National Institutes of Health).
For the measurement of hydroxyproline, collected right lung was homogenized with saline and hydroxyproline was measured by high performance liquid chromatography (SRL, Tokyo, Japan).
Briefly, cells were washed twice with chilled PBS and lysed with lysis buffer (NP-40), and protease inhibitor cocktail (Roche Applied Science, Indianapolis, IN, USA). Collected lungs were disrupted and homogenized in RIPA buffer (Wako Pure Chemical Industries, Osaka, Japan) using TissueRuptor II (Qiagen, Germantown, MD, USA). Blotted membranes were incubated overnight at 4 °C with primary antibodies and washed with TBS containing 0.1% Tween-20 before incubation with HRP-conjugated secondary antibody, followed by Chemi-Lumi One L (Nacalai).

Immunohistochemistry in cultured cells. For immunostaining, cells on glass-bottom dishes were fixed
in 4% paraformaldehyde and permeabilized with 0.2% Triton X-100. The samples were blocked in 5% skim milk, followed by incubation with primary antibodies overnight at 4 °C. The corresponding secondary antibodies were labeled with AlexaFluor 546 (Thermo Fisher Scientific). Nuclear staining was performed using 4,6-diamidino-2-phenylindole (DAPI). Images were collected using a confocal microscope (FLUOVIEW FV10i; Olympus, Tokyo, Japan). Antibodies and reagents employed in western blotting and immunostaining included: Statistical analyses. Data are expressed as mean ± SD. Data were analyzed using Prism for windows version 8 (GraphPad Software, USA). Comparison between multiple groups was performed using ANOVA followed by Dunnett's multiple comparisons test. Comparison between two groups was performed using T-test. Differences were considered significant at p < 0.05.

Data availability
All data generated or analyzed during this study are included in this published article (and its Supplementary Information files).