Abstract
Introduction: 5-Methoxytryptophan (5-MTP) is a cellular metabolite with anti-inflammatory properties. Several recent reports indicate that 5-MTP protects against post-injury tissue fibrosis. It was unclear how 5-MTP controls tissue fibrosis. We postulated that 5-MTP attenuates renal interstitial fibrosis by blocking toll-like receptor 2 (TLR2) and transforming growth factor β (TGFβ) signaling pathways. Methods: In vivo experiments were carried out in a well-established unilateral ureteral obstruction (UUO) model in wild-type (WT) and tlr2−/− mice. The effect of 5-MTP on renal fibrosis was evaluated by pretreatment of WT UUO mice with intraperitoneal administration of 5-MTP. To determine whether 5-MTP attenuates fibrosis by inhibiting TLR2 and TGFβ signaling pathways, we evaluated the effect of 5-MTP on TLR2-induced fibroblast phenotypic switch in NRK-49F fibroblasts and TLR2 and TGFβ signaling pathways in human proximal tubular epithelial cells (HPTECs) and RAW264.7 macrophages stimulated with Pam3CSK4 (Pam3) or TGFβ1. Results: UUO-induced renal fibrosis was abrogated in tlr2−/− mice consistent with a crucial role of TLR2 in UUO-induced renal fibrosis. UUO-induced macrophage infiltration and pro-fibrotic cytokine production in renal tissues were suppressed by tlr2 knockout. 5-MTP administration attenuated renal tissue fibrosis accompanied by reduction of macrophage infiltration and IL-6 and TGFβ levels. 5-MTP inhibits TLR2 upregulation and blocks TLR2-MyD88-TRAF6 signaling pathway in macrophages. Furthermore, 5-MTP blocked Pam3- and TGFβ1-induced phenotypic switch of NRK-49F to myofibroblasts and inhibited Pam3- and TGFβ1-induced signaling pathways in HPTECs and RAW264.7 cells. Conclusion: 5-MTP is effective in protecting against UUO-induced renal interstitial fibrosis by blocking TLR2 and TGFβ signaling pathways.
Plain Language Summary
Mouse renal tissue injury by UUO results in declining 5-MTP production and TLR2 upregulation. 5-MTP controls injury-induced TLR2 upregulation and disrupts the TLR2-MyD88-TRAF6 signaling pathway in macrophages thereby inhibiting the pro-fibrotic transcriptional programs. This suggests that 5-MTP is a valuable lead compound for new anti-renal tissue inflammation and fibrosis drug development.
Introduction
Toll-like receptor 2 (TLR2) belongs to the TLR family of receptors responsive to pathogen- and damage-associated molecular pattern signals [1‒3]. TLR2 expression is upregulated in renal injury and involved in renal tissue inflammation and fibrosis following various types of renal injury [4, 5]. It was reported in a unilateral ureteral obstruction (UUO) murine model that TLR2 upregulation is accompanied by elevated endogenous TLR2 ligands including biglycan, high mobility group box, and GP96 [4]. Upregulated TLR2 plays an important role in mediating renal tissue inflammation and fibrosis following acute injury by UUO. It has been reported that tlr2 knockout attenuates inflammatory cell infiltration and pro-inflammatory cytokine production in the murine UUO model [4, 5]. Furthermore, myofibroblasts and pro-fibrotic factors notably transforming growth factor-β (TGFβ) were suppressed by tlr2 deletion in the UUO murine model [4, 5]. These findings suggest that TLR2 is a key mediator of post-injury renal inflammation and fibrosis.
5-Methoxytryptophan (5-MTP) was identified as an endogenous factor with anti-inflammatory and tumor inhibitory activities [6, 7]. It is produced in vascular endothelial cells, fibroblasts, and diverse types of epithelial cells including renal tubular epithelial cells. Its production is catalyzed by tryptophan hydroxylase-1 (TPH-1) followed by hydroxyindole O-methyltransferase [6, 8]. It is secreted into the extracellular milieu via Golgi vesicular trafficking [7]. The secreted 5-MTP acts in an autocrine and paracrine manner to confer vasoprotection and defend against inflammation and vascular injury-induced arterial intimal hyperplasia [9, 10]. In addition, 5-MTP was reported to be an innate metabolite against renal inflammation and fibrosis induced by UUO and ischemia-reperfusion injury [11]. Secreted 5-MTP enters into the circulating blood which was reported to be inversely correlated with severity of chronic kidney disease (CKD) [11] and post-myocardial infarction (MI) heart failure [12]. The inverse association of 5-MTP with CKD disease severity was attributed to suppression of renal 5-MTP production as demonstrated in the UUO murine model [11]. Cellular experiments suggest that reduced 5-MTP production is due to reduced expression of TPH-1. Importantly, 5-MTP administration attenuated inflammation and substantially suppressed renal tissue fibrosis in the UUO model. Furthermore, TPH-1 overexpression restored 5-MTP production and decreased the expression of pro-inflammatory and pro-fibrotic factors [11]. These findings suggest that TPH-1-catalyzed 5-MTP production in renal epithelial cells represents a powerful arsenal against renal injury. The mechanism by which 5-MTP protects renal tissues and attenuates tissue fibrosis is largely unknown. We tested the hypothesis that 5-MTP exerts its actions by inhibiting TLR2-mediated inflammation and myofibroblast transdifferentiation and suppressing pro-fibrotic and pro-inflammatory factors. Our results reveal that 5-MTP inhibited macrophage infiltration and macrophage-derived pro-inflammatory and pro-fibrotic factors and blocked myofibroblast generation by suppressing TLR2-mediated signaling pathways.
Methods
Reagents
Pam3CSK4 (Pam3; tlrl-pms) was purchased from InvivoGen. L form of 5-MTP (Pam3- and-MTP) was commercially synthesized by AstaTech, USA. Antibodies for fibronectin (FN) (ab2413; Abcam), collagen I (COL1) (ab34710; Abcam), collagen IV (ab6586; Abcam), F4/80 (ab111101; Abcam), interleukin-6 (ab6672; Abcam), TGFβ1 (ab92486; Abcam), HSP90 (#4874; Cell Signaling Technology), α-SM actin (A5228; Sigma-Aldrich), phospho-p38 (Thr180/Tyr182) (#4511; Cell Signaling Technology), p38 (#9212; Cell Signaling Technology), phospho-p65 (Ser536) (#3033; Cell Signaling Technology), p65 (#8242; Cell Signaling Technology), phospho-Smad2 (Ser465/467) (#3108; Cell Signaling Technology), Smad2 (#5339; Cell Signaling Technology), TLR2 (ab209217; Abcam), MyD88 (#4283; Cell Signaling Technology), TRAF6 (#67591; Cell Signaling Technology), IκB (#4814; Cell Signaling Technology), and β-actin (MAB1501; Millipore) were used in Western Blot analysis and immunohistochemistry. Antibodies for phospho-Smad2 (Ser465/467) from Thermo Fisher (44-244 G) and TLR2 from Abcam (ab209216) were used in immunohistochemistry. The antibody for 5-MTP was custom-synthesized by GenScript.
Cell Culture and Treatment
NRK-49F Cells
NRK-49F, a fibroblast-like cell line isolated from rat kidney, was purchased from ATCC (CRL-1570). They were cultured in a starvation medium consisting of DMEM supplemented with 0.5% FBS, 1% nonessential amino acids, penicillin (100 U/mL), and streptomycin (100 μg/mL) for 24 h, followed by addition of various concentrations of 5-MTP for 30 min before stimulation with 1 μg/mL Pam3 or 10 ng/mL TGFβ1 for 24 h.
HPTEC Cells
Human primary proximal tubular epithelial cells (HTPEC) (Cell Biologicals, H-6015) were cultured in a REGM™ Renal Epithelial Cell Growth Medium BulletKit containing Renal Epithelial Cell Growth Basal Medium (Lonza) and REGM™ SingleQuots™ supplements (Lonza). HPTEC cells were typically subjected to serum starvation for 24 h and pretreatment with various concentration of 5-MTP for 4 h before stimulation with 1 μg/mL Pam3 or 10 ng/mL TGFβ1 for 24 h.
RAW264.7 Cells
Murine RAW264.7 macrophages (ATCC, TIB-71) were cultured in a growth medium-containing DMEM supplemented with 10% FBS, penicillin (100 U/mL), and streptomycin (100 μg/mL). RAW264.7 cells were pretreated with 100 μm 5-MTP for 6 h, followed by stimulation with 1 μg/mL Pam3 or 10 ng/mL TGFβ1 for 24 h.
Mouse Model of UUO and Kidney Tissue Collection
Wild-type (WT) or Tlr2−/− C57BL/6J mice aged between 8 and 10 weeks underwent ligation of the left ureter for 7 or 10 days to induce renal fibrosis. In brief, mice were anesthetized with 2.5% isoflurane before surgery. The left ureter was exposed through a flank incision and was completely obstructed 1 cm below the renal pelvis with 5-0 silk ligature. The sham-operated group was subjected to similar procedure without urethral ligation. During the animal experiment, 23.5 mg/kg L-5-MTP was freshly prepared with saline each time and injected intraperitoneally every 2 days. At the endpoint, mice were sacrificed by CO2 inhalation and then infused with saline and 10% formalin. Sham- or ureter-ligated kidneys were harvested for analysis of renal fibrosis by performing Masson trichrome staining or IHC in paraffin-embedded sections. Given that estrogen can interfere with inflammatory responses and pathophysiology, male mice were used and randomly grouped in our animal experiments. The animals were housed at the National Health Research Institutes (Taiwan) animal facility. All experimental procedures were reviewed and approved by the Institutional Animal Care and Use Committee of National Health Research Institutes, Taiwan (Protocol Approval No. NHRI-IACUC-107037).
Histological Analysis of Fibrosis
Kidneys harvested from sham- or UUO-WT or Tlr2−/− C57BL/6J mice were paraffin-embedded for Hematoxylin and eosin (HE) staining and Masson trichrome staining. For histological analysis of renal tissues, the tissue sections (4 μm) were deparaffinized and stained with hematoxylin (Leica) and eosin (Leica) sequentially. The stained slides were mounted using mounting medium (Leica) and photographed under light microscopy. For determination of collagen deposition, the paraffin-embedded sections were stained with Masson Trichrome Staining Kit (Sigma-Aldrich), following the manufacturer’s instructions. Percentages of collagen fibrotic areas were quantified as the ratio of the positively stained Masson blue to the whole area by using ImageJ software. Tubulointerstitial lesion was characterized by dilation of the tubule, expansion of the interstitial space, and fibrotic deposition. Tubulointerstitial lesion score was calculated by using the grade system of the severity of tubular dilation from HE histological sections and the percentages of fibrotic deposition area from Masson blue (0, normal; 1, mild: <25%; 2, moderate: 25%–50%; 3, severe: 50%–75%; and 4, advanced >75%).
Immunohistochemistry
Renal tissue sections (4 μm) were heated at 60°C for 1 h, deparaffinized with xylene and rehydrated in alcohols. Antigen sites were retrieved by heating the sections in Trilogy buffer (Cell Marque Corporation) in an electric pressure cooker for 10 min. The sections were sequentially blocked by 3% H2O2 for 20 min and 5% bovine serum albumin (BSA) for 30 min. They were incubated with primary antibodies at room temperature for 2 h or 4°C for 16 h and washed in phosphate-buffered saline (with 0.1% Tween 20). The sections were incubated with horseradish peroxidase-labeled polymer (Dako) for 60 min and washed with phosphate-buffered saline three times. The levels of target protein expression in renal tissues were visualized using the DAB Chromogen system (Dako). Three to five sections with a constant interval of sectioning by order were used for each target. The immune-positive areas were quantified as percentages of total tissue area using ImageJ software. When comparing within groups of samples, the samples which were considered to be outliers were excluded from further analyses.
Flow Cytometry Analysis
Renal tissues were collected from sham- or UUO-WT or Tlr2−/− C57BL/6J mice and processed to extract cells. The extraction was performed using dounce homogenizer with PBS to obtain intact cells. Cells were then fixed by incubation in PBS containing 1% formaldehyde on ice for 15 min and washed with PBS and blocked by using 1% BSA on ice for 1 h. Cell surface markers for macrophage populations were detected by staining with the antibodies of CD206-Alexa Fluor 488 (141710; BioLegend), CD40-PE (124610; BioLegend), CD11b-PerCP (101230; BioLegend), and F4/80 (123116; BioLegend) on ice. After 3 h of staining, cells were washed and resuspended in PBS. Flow cytometry was performed on a FACSCalibur (BD Biosciences) and analyzed with FlowJo software. The gating strategy for determining macrophage populations is as follows: the population was first gated for CD11b-PerCP+F4/80-APC+ to identify the activated macrophage population. Then, the CD40-PE+CD206-Alexa Fluor 488− or CD40-PE-CD206-Alexa Fluor 488+ populations were gated to determine M1 or M2 macrophages, respectively. It is to be noted that as the CD11b+F4/80+ population gated for CD40+ and CD206+ cells in the sham-WT saline and sham-TLR2−/− saline groups was very low, the CD40+ and CD206+ cell population was even lower which might yield skewed M1/M2 ratio.
Migration Assays
To assess cell migration, 2 × 104 RAW264.7 cells were seeded in the upper chamber of 24-well transwell plates (8-μm pore size; BD Falcon) in triplicate and pretreated with 100 μm 5-MTP for 6 h, followed by stimulation with 1 μg/mL Pam3 or 10 ng/mL TGFβ1 for 24 h. The lower chamber was filled with 700 μL DMEM supplemented with 10 ng/mL PDGF-BB (PeproTech) as a chemoattractant. After 24 h incubation, cells migrating to the opposite side of the membrane were fixed using 1% formaldehyde and quantified.
Western Blot Analysis
Cell lysates were extracted from renal cells or RAW264.7 cells using RIPA buffer containing protease inhibitors and phosphatase inhibitors. The cellular proteins were electrophoretically resolved with 4%–12% SDS-PAGE and transferred to polyvinylidene difluoride membrane. They were immunoblotted with desired specific antibodies.
Quantitative Real-Time PCR
Total RNA was extracted from RAW264.7 cells, HPTECs, or NRK-49F cells using High Pure RNA Isolation Kit (Roche), and cDNA was synthesized using SuperScript III First-Strand Synthesis System (Thermo Fisher) according to the manufacturer’s instructions. COLI, FN, ACTA2, TLR2, TGFβ1, IL-6, and β-actin transcripts were analyzed by Applied Biosystems ViiA 7 Real-Time PCR System (Thermo Fisher) using KAPA SYBR FAST Master Mix (2x) ROX low (Kapa Biosystems). Primer sequences were as follows: mouse ColI forward 5′- CGATGGATTCCCGTTCGAGT -3′ and reverse 5′- CGATCTCGTTGGATCCCTGG -3′; mouse FN forward 5′- ACATGGCTTTAGGCGGACAA -3′ and reverse 5′- ACATTCGGCAGGTATGGTCTTG -3′; mouse Acta2 forward 5′- ATGCAGAAGGAGATCACAGC -3′ and reverse 5′- CAGCTTCGTCGTATTCCTGT -3′; mouse Tlr2 forward 5′- GGTGCGGACTGTTTCCTTCT -3′ and reverse 5′- TCCTCTGAGATTTGACGCTTTGT -3′; mouse Tgfb1 forward 5′- GGAATACAGGGCTTTCGATT-3′ and reverse 5′- CTCTGTGGAGCTGAAGCAAT -3′; mouse Il-6 forward 5′- CCTCTGGTCTTCTGGAGTACC -3′ and reverse 5′- ACTCCTTCTGTGACTCCAGC -3′; human TLR2 forward 5′- ATCCTCCAATCAGGCTTCTCT -3′ and reverse 5′- ACACCTCTGTAGGTCACTGTTG -3′; human TGFB1 forward 5′-CAAGCAGAGTACACACAGCAT -3′ and reverse 5′- TGCTCCACTTTTAACTTGAGCC-3′ Target gene expression relative to β-actin was calculated as 2–(Ct of target gene – Ct of β-actin).
Immunoprecipitation
The cells were lysed with RIPA buffer supplemented with protease inhibitor cocktail. Cell lysate was centrifuged at 16,000 g for 15 min to pellet cell debris. 500 μg cell lysate was mixed with indicated antibodies (2 μg/reaction) for 1 h and then washed with 1x PBS. Protein G Mag Sepharose Xtra (20 μL/reaction) (Cytiva) was washed with 1x PBS 3 times and blocked in 1% BSA for 30 min. Protein G Mag Sepharose Xtra was then mixed with prepared samples overnight. After three washes in 1x PBS, 1x SDS sample buffer was added and heated to 95°C for 10 min. Proteins were separated in SDS-PAGE. Except last step, all procedures were maintained at 4°C or on ice.
Statistical Analysis
Statistical analyses were performed using Graphpad Prism version 8 Software (GraphPad Software Inc.). All values were given as means ± SD. The t test was used to determine the statistical significance of the difference between the treatment and control groups, while one-way ANOVA was used to analyze multiple groups. p values <0.05 were considered statistically significant.
Results
UUO-Induced Renal Fibrosis Is Reduced in TLR2−/− Mice
Consistent with previous reports, UUO-induced renal interstitial fibrosis in the murine model was detected by HE, Masson trichrome, and FN staining (Fig. 1a). Quantitative analysis of HE and Masson trichrome staining of renal tissues reveals a significant increase in renal interstitial fibrosis in the UUO mice (Fig. 1b, c). There was a comparable increase in FN deposition in UUO renal tissues (Fig. 1d). UUO induced a much lower level of fibrosis and FN in renal tissues of tlr2−/− (TLR2-KO) mice (Fig. 1a–d). These results suggest that UUO-induced renal interstitial fibrosis depends at least in part on the TLR2 pathway. As tissue fibrosis is closely linked to macrophage infiltration and activation [13‒15], we analyzed macrophages in UUO-injured renal tissues. IHC analysis of F4/80-positive macrophages reveals increased macrophage infiltration in UUO TLR2WT (UUO-WT) mice which was significantly reduced in TLR2-KO (UUO-KO) mice (Fig. 2a). Flow cytometry analysis confirmed increased CD11b+F4/80+cells in UUO-WT mice but not in UUO-KO mice (Fig. 2b). The classically activated (M1) to alternatively activated (M2) macrophage ratio (M1/M2) was lower in UUO-KO mice, but the difference was statistically insignificant (Fig. 2c). We next analyzed key pro-inflammatory and pro-fibrotic factors, i.e., IL-6 and TGFβ, respectively, in renal tissues by IHC. IL-6 and TGFβ were increased by UUO, and the increase was abrogated in UUO-KO mice (Fig. 2d–f). HSP90 elevation by UUO-WT mice was similarly diminished in UUO-KO mice (Fig. 2g). These findings confirm that TLR2-mediated macrophage infiltration and pro-inflammatory and pro-fibrotic factor secretion play a crucial role in UUO-induced renal fibrosis.
UUO-induced renal tissue fibrosis in wild-type (WT) and TLR2 knockout (KO) mice. a To visualize fibrosis, renal tissues from Sham- and UUO-operated WT mice and UUO-operated TLR KO mice were stained with hematoxylin and eosin (HE) or Masson’s trichrome. Fibronectin (FN) expression in renal tissues was analyzed with immunohistochemistry (IHC). Quantification of fibrosis by determining tubulointerstitial lesion (TIL) score on HE stain (b) and collagen blue on Mason’s trichome stain (c). d FN-positive area in IHC was analyzed by Image J software. 6 mice were studied per group. Scale bar, 100 μm. *p < 0.05, **p < 0.005, ***p < 0.0005, ****p < 0.00005.
UUO-induced renal tissue fibrosis in wild-type (WT) and TLR2 knockout (KO) mice. a To visualize fibrosis, renal tissues from Sham- and UUO-operated WT mice and UUO-operated TLR KO mice were stained with hematoxylin and eosin (HE) or Masson’s trichrome. Fibronectin (FN) expression in renal tissues was analyzed with immunohistochemistry (IHC). Quantification of fibrosis by determining tubulointerstitial lesion (TIL) score on HE stain (b) and collagen blue on Mason’s trichome stain (c). d FN-positive area in IHC was analyzed by Image J software. 6 mice were studied per group. Scale bar, 100 μm. *p < 0.05, **p < 0.005, ***p < 0.0005, ****p < 0.00005.
Renal tissue macrophage infiltration and IL-6 and TGFβ1 levels in UUO-injured WT and TLR2-KO mice. UUO-injured WT mice were treated with or without 5-MTP. a F4/80-positive cells were analyzed by IHC left panel, and the quantified data are shown in the right panel. b CD11b+F4/80+cells in renal tissues were analyzed by flow cytometry. c M1/M2 ratio was calculated from the percentage of CD40+ and CD206+ populations gated on CD11b+F4/80+macrophages. d–g Renal tissue IL-6, TGFβ1, and Hsp90 levels were analyzed by IHC. In all the animal experiments, 6 mice were performed in each group. *p < 0.05, **p < 0.005, ***p < 0.0005. NS denotes nonsignificant.
Renal tissue macrophage infiltration and IL-6 and TGFβ1 levels in UUO-injured WT and TLR2-KO mice. UUO-injured WT mice were treated with or without 5-MTP. a F4/80-positive cells were analyzed by IHC left panel, and the quantified data are shown in the right panel. b CD11b+F4/80+cells in renal tissues were analyzed by flow cytometry. c M1/M2 ratio was calculated from the percentage of CD40+ and CD206+ populations gated on CD11b+F4/80+macrophages. d–g Renal tissue IL-6, TGFβ1, and Hsp90 levels were analyzed by IHC. In all the animal experiments, 6 mice were performed in each group. *p < 0.05, **p < 0.005, ***p < 0.0005. NS denotes nonsignificant.
UUO-Induced Renal Fibrosis Is Associated with Reduced Renal 5-MTP Production and Attenuated by 5-MTP Administration
Murine renal epithelial cells express 5-MTP [7]. 5-MTP level in renal tissues was unaffected by sham operation but significantly reduced in UUO mice (Fig. 3a). To determine the relationship between 5-MTP and renal fibrosis, we administered 5-MTP intraperitoneally to UUO-WT mice and analyzed renal tissue fibrosis. 5-MTP administration attenuated renal tissue fibrosis as analyzed by Masson trichrome and HE staining (online suppl. Fig. S1; for all online suppl. material, see https://doi.org/10.1159/000543275). Analysis of collagen I and IV (COL1 and IV) and FN by IHC reveals reduced collagen and FN deposition in renal tissues in 5-MTP-treated mice (Fig. 3b–e). In separate experiments, we analyzed FN, COL1, and Acta2 (α-SMA) mRNA levels in renal tissues from UUO-WT mice treated with and without 5-MTP. Renal tissue FN, COL1, and α-SMA mRNA levels were significantly reduced by 5-MTP administration (Fig. 3f–h).
Renal tissue 5-MTP levels and the effect of 5-MTP administration on renal tissue fibrosis. a Analysis of 5-MTP levels in renal tissues by IHC and 5-MTP immunopositive areas (percentage to total area) were quantified by Image J software. b–e Collagen I (Col I), COL IV, and FN in renal tissues were analyzed by IHC and quantified by Image J software. The results were expressed as percentage of total area. Each dot denotes a murine experiment. 6 mice were included in each group. Relative Col I (f), FN (g), and Acta2 (h) mRNA levels in renal tissues were measured by qPCR. Each dot denotes a murine experiment. Scale bar, 100 μm. *p < 0.05, **p < 0.005, ***p < 0.0005.
Renal tissue 5-MTP levels and the effect of 5-MTP administration on renal tissue fibrosis. a Analysis of 5-MTP levels in renal tissues by IHC and 5-MTP immunopositive areas (percentage to total area) were quantified by Image J software. b–e Collagen I (Col I), COL IV, and FN in renal tissues were analyzed by IHC and quantified by Image J software. The results were expressed as percentage of total area. Each dot denotes a murine experiment. 6 mice were included in each group. Relative Col I (f), FN (g), and Acta2 (h) mRNA levels in renal tissues were measured by qPCR. Each dot denotes a murine experiment. Scale bar, 100 μm. *p < 0.05, **p < 0.005, ***p < 0.0005.
5-MTP Blocks Macrophage Infiltration and Suppresses Renal Tissue IL-6 and TGFβ1
Analysis of F4/80 positive cells in renal tissues by IHC and CD11b+F4/80+cells by flow cytometry reveals effective suppression of macrophage infiltration in renal tissues of UUO-WT mice treated with 5-MTP (Fig. 2a, b). The extent of suppression was comparable to that of TLR2 deletion. Importantly, renal tissue M1/M2 ratio in UUO mice was significantly reduced by 5-MTP administration compared to that in UUO-WT or UUO-TLR2−/− mice (Fig. 2c). Renal tissue IL-6 and TGFβ1 levels were also significantly suppressed by 5-MTP (Fig. 4a). Taken together, these results suggest that 5-MTP attenuates fibrosis by controlling M1 macrophage infiltration and activation. As TLR2 was reported to drive vascular smooth muscle cell migration via IL-6 [16], we determined whether TLR2 activation exerts a similar effect on macrophages. Murine macrophages treated with a TLR2 agonist, Pam3 exhibited increased migration as analyzed by transwell assay (Fig. 4b). 5-MTP reduced Pam3-induced cell migration (Fig. 4b). TGFβ1-induced macrophage migration was also suppressed by 5-MTP (Fig. 4c). These results suggest that 5-MTP suppresses renal tissue inflammation and fibrosis in UUO-injured renal tissues by inhibiting TLR2- and TGFβ1-induced macrophage migration and pro-inflammatory cytokine productions.
5-MTP reduces UUO-induced macrophage migration and cytokine levels. a IL-6 and TGFβ1 levels in renal tissues were analyzed by IHC. b, c Migration of RAW264.7 cells was measured by transwell assay. RAW264.7 cells were pretreated with 5-MTP (100 μm) for 6 h followed by treatment with Pam3 (1 μg/mL) or TGFβ1 (10 ng/mL) for 24 h. d, e RAW264.7 cells were pretreated with 5-MTP for 6 h followed by treatment with Pam3 (1 μg/mL) for 6 h (d) or 24 h (e) and relative Il6 and Tgfb1 mRNA levels were measured by qPCR. Each bar denotes mean ± SD (n = 3). *p < 0.05, ***p < 0.0005.
5-MTP reduces UUO-induced macrophage migration and cytokine levels. a IL-6 and TGFβ1 levels in renal tissues were analyzed by IHC. b, c Migration of RAW264.7 cells was measured by transwell assay. RAW264.7 cells were pretreated with 5-MTP (100 μm) for 6 h followed by treatment with Pam3 (1 μg/mL) or TGFβ1 (10 ng/mL) for 24 h. d, e RAW264.7 cells were pretreated with 5-MTP for 6 h followed by treatment with Pam3 (1 μg/mL) for 6 h (d) or 24 h (e) and relative Il6 and Tgfb1 mRNA levels were measured by qPCR. Each bar denotes mean ± SD (n = 3). *p < 0.05, ***p < 0.0005.
5-MTP Suppresses TLR2 Signaling Pathway and NF-κB Activation in Macrophages
Pam3 increased IL-6 mRNA expression in RAW264.7 cells which was suppressed by 5-MTP (Fig. 4d). 5-MTP suppressed Pam3-induced TGFβ1 expression in a similar manner (Fig. 4e). To determine whether 5-MTP acts via interruption of TLR signaling, we evaluated the effect of 5-MTP on Pam3-induced interaction between TLR2 and adapter molecules, i.e., MyD88 and TRAF6 and activation of downstream signals. Protein input was shown in Figure 5a. RAW264.7 cell lysates were immunoprecipitated with MyD88 antibodies and TLR2 and TRAF6 proteins in the precipitates were analyzed by Western blotting. Abundant TLR2 was detected in Pam3-treated samples when compared to control, while only trace of TRAF6 was detected (Fig. 5b). Analysis of immunoprecipitated by TRAF6 antibodies revealed considerable amounts TLR2 and MyD88 (Fig. 5c). Pretreatment of cells with 5-MTP abolished the interaction between TLR2 and MyD88 and TRAF6 (Fig. 5b, c). Pam3 treatment resulted in decreased IκB and increased p-p65 which were restored by 5-MTP pretreatment (Fig. 5d). 5-MTP reduced Pam3-induced p-p65 in WT murine peritoneal macrophages to an extent comparable to that in TLR2 knockout murine peritoneal macrophages (online suppl. Fig. S2). TLR2 activation by Pam3 resulted in activation of p38 MAPK which was suppressed by 5-MTP (Fig. 5e, f). These results suggest that 5-MTP suppresses macrophage activation by disrupting the TLR2-MyD88-TRAF6 signaling pathway and blocking downstream signaling kinases, resulting in activation of NF-kB and p38-mediated transcription factors.
5-MTP suppresses TLR2 signaling pathway and NF-κB activation in macrophages. a–d RAW264.7 cells were pretreated with 5-MTP (100 μm) for 6 h followed by treatment with Pam3 (1 μg/mL) for 30 min. a The expression levels of TLR2, TRAF6 and MyD88 in total input were determined by immunoblotting. b, c Cell lysates were subjected to an immunoprecipitation assay with anti-MyD88 (b) or TRAF6 (c) and the precipitated levels of TLR2, TRAF6 and MyD88 were determined by immunoblotting. d p-p65 and IκB were analyzed by immunoblotting and quantified by densitometry. p65 and β-actin were included as a loading control, respectively. e RAW264.7 cells were treated with Pam3 (1 μg/mL) at various time points. p-p38 and p38 was analyzed by immunoblotting. f RAW264.7 cells were pretreated with 5-MTP (25 μm) for 6 h followed by treatment with Pam3 (1 μg/mL) for 30 min. p-p38 and p38 was analyzed by immunoblotting. Each bar denotes mean ± SD (n = 3). *p < 0.05, **p < 0.005, ***p < 0.0005.
5-MTP suppresses TLR2 signaling pathway and NF-κB activation in macrophages. a–d RAW264.7 cells were pretreated with 5-MTP (100 μm) for 6 h followed by treatment with Pam3 (1 μg/mL) for 30 min. a The expression levels of TLR2, TRAF6 and MyD88 in total input were determined by immunoblotting. b, c Cell lysates were subjected to an immunoprecipitation assay with anti-MyD88 (b) or TRAF6 (c) and the precipitated levels of TLR2, TRAF6 and MyD88 were determined by immunoblotting. d p-p65 and IκB were analyzed by immunoblotting and quantified by densitometry. p65 and β-actin were included as a loading control, respectively. e RAW264.7 cells were treated with Pam3 (1 μg/mL) at various time points. p-p38 and p38 was analyzed by immunoblotting. f RAW264.7 cells were pretreated with 5-MTP (25 μm) for 6 h followed by treatment with Pam3 (1 μg/mL) for 30 min. p-p38 and p38 was analyzed by immunoblotting. Each bar denotes mean ± SD (n = 3). *p < 0.05, **p < 0.005, ***p < 0.0005.
5-MTP Inhibits Renal Fibroblast Phenotypic Switch to Myofibroblasts
Phenotypic switch of fibroblasts to myofibroblasts is a key event in tissue fibrosis. To evaluate the effect of 5-MTP on fibroblast phenotypic switch, we treated renal fibroblasts, NRK-49F, with Pam3 in the presence and absence of 5-MTP. TLR2 activation by Pam3 results in expression of myofibroblast markers, i.e., FN and smooth muscle α-actin (α-SMA). 5-MTP inhibited Pam3-induced α-SMA, FN, and COL1 (Fig. 6a–d). TGFβ1-induced α-SMA and FN expression was significantly reduced by 5-MTP (Fig. 6e–g). TGFβ1-induced COL1 was inhibited by 5-MTP at a higher concentration (100 µm) (Fig. 6h).
5-MTP suppresses Pam3- and TGFβ1-induced expression of myofibroblast markers in NRK-49F cells. NRK-49F cells were pretreated with different concentrations of 5-MTP for 30 min followed by Pam3 (1 μg/mL) (a–d) or TGFβ1 (10 ng/mL for 24 h) (e–h). FN, Collagen I and α-SMA protein levels in cell lysates were analyzed by immunoblotting. β-actin was included as a loading control. The protein levels were quantified by densitometry and expressed as a ratio to untreated control. Each bar in b–d and f–h denotes mean ± SD (n = 3).
5-MTP suppresses Pam3- and TGFβ1-induced expression of myofibroblast markers in NRK-49F cells. NRK-49F cells were pretreated with different concentrations of 5-MTP for 30 min followed by Pam3 (1 μg/mL) (a–d) or TGFβ1 (10 ng/mL for 24 h) (e–h). FN, Collagen I and α-SMA protein levels in cell lysates were analyzed by immunoblotting. β-actin was included as a loading control. The protein levels were quantified by densitometry and expressed as a ratio to untreated control. Each bar in b–d and f–h denotes mean ± SD (n = 3).
5-MTP Blocks TLR2 Upregulation and Pro-Fibrotic Signaling in Renal Epithelial Cells
We used FN as a surrogate marker of epithelial cell transdifferentiation to myofibroblast. Pam3 and TGFβ1 significantly increased FN expression in HPTECs as evaluated by Western blotting (Fig. 7a–d). 5-MTP suppressed FN protein levels in HPTECs stimulated with Pam3 or TGFβ1 in a concentration-dependent manner (Fig. 7a–d). To determine the role of TLR2 and its signaling pathway in epithelial cell transdifferentiation, we examined TLR2 expression in renal tissues of UUO versus control mice by IHC. TLR2 was detected in renal tubules (online suppl. Fig. S3) of control mice. TLR2 staining was increased in UUO mice which were reduced by 5-MTP treatment (Fig. 7e). Interestingly, treatment of HPTECs with Pam3 resulted in upregulation of TLR2 mRNA expression which was suppressed by 5-MTP (Fig. 7f). Pam3-induced TLR2 upregulation in NRK-49F fibroblasts was blocked by 5-MTP pretreatment (online suppl. Fig. S4). In addition, Pam3-induced TGFβ1 upregulation in HPTECs (Fig. 7g) and NRK-49F (online suppl. Fig. S5) cells was blocked by 5-MTP. We next evaluated the effects of 5-MTP on Pam3-induced p38 and NF-κB activation. Pam3 increased phosphorylated p38 MAPK (p-p38) without a significant effect on p38, and increased p-p65 without an effect on the NF-κB p65 subunit (Fig. 7h). 5-MTP pretreatment attenuated p-p38 MAPK and p-p65 (Fig. 7h).
5-MTP inhibits TLR2 upregulation in UUO-injured renal tissues and pro-fibrotic signaling in HPTECs. HPTECs were pretreated with increasing concentrations of 5-MTP for 4 h followed by Pam3 (1 μg/mL) (a) or TGFβ1 (10 ng/mL) (c) for 24 h. FN proteins in cell lysates were analyzed by Western blotting (a, c) and quantified by densitometry. Each bar in b and d denotes mean ± SD (n = 3). *p < 0.05; **p < 0.005; ***p < 0.0005. e Analysis of TLR2 levels in renal tissues by IHC and TLR2 immunopositive areas (percentage to total area) were quantified by Image J software. Each dot denotes a murine experiment. Scale bar, 100 μm. f, g HPTECs were treated with 5-MTP for 4 h followed by Pam3 (1 μg/mL) for 6 h. Relative TLR2 and TGFB1 mRNA levels were measured by qPCR. h HPTECs were treated with 5-MTP for 4 h followed by Pam3 (1 μg/mL) for 30 min. Signaling proteins in cell lysates were analyzed by immunoblotting and quantified by densitometry. Each bar in f, g and h denotes mean ± SD (n = 3). *p < 0.05; **p < 0.005; ***p < 0.0005.
5-MTP inhibits TLR2 upregulation in UUO-injured renal tissues and pro-fibrotic signaling in HPTECs. HPTECs were pretreated with increasing concentrations of 5-MTP for 4 h followed by Pam3 (1 μg/mL) (a) or TGFβ1 (10 ng/mL) (c) for 24 h. FN proteins in cell lysates were analyzed by Western blotting (a, c) and quantified by densitometry. Each bar in b and d denotes mean ± SD (n = 3). *p < 0.05; **p < 0.005; ***p < 0.0005. e Analysis of TLR2 levels in renal tissues by IHC and TLR2 immunopositive areas (percentage to total area) were quantified by Image J software. Each dot denotes a murine experiment. Scale bar, 100 μm. f, g HPTECs were treated with 5-MTP for 4 h followed by Pam3 (1 μg/mL) for 6 h. Relative TLR2 and TGFB1 mRNA levels were measured by qPCR. h HPTECs were treated with 5-MTP for 4 h followed by Pam3 (1 μg/mL) for 30 min. Signaling proteins in cell lysates were analyzed by immunoblotting and quantified by densitometry. Each bar in f, g and h denotes mean ± SD (n = 3). *p < 0.05; **p < 0.005; ***p < 0.0005.
As the pro-fibrotic effect of TGFβ is signaled via Smad2/3 [17, 18], we analyzed Smad2 activation by measuring phosphor-Smad2 (p-Smad2) by Western blotting. TGFβ1 increased p-Smad2 in HPTECs (Fig. 8a). 5-MTP suppressed TGFβ1-induced p-Smad2 at concentrations >50 µm (Fig. 8a). p-Smad2 was analyzed in renal tissues of murine UUO model by IHC. p-Smad2 was elevated in saline-treated UUO-WT renal tissues which was abrogated in 5-MTP-treated mice (Fig. 8b). As a reference, we analyzed time-dependent activation of Smad2 in RAW264.7 cells following TGFβ1 treatment. TGFβ1 significantly increased p-Smad2 at 30 and 60 min following TGFβ1 treatment (online suppl. Fig. S6). 5-MTP suppressed TGFβ1-induced p-Smad2 at 30 min as well as at 60 min, although it was less effective in reducing p-Smad2 at 60 min (online suppl. Fig. S6). As 5-MTP exerts anti-inflammatory and vasoprotective actions by blocking p38 MAPK activation [7‒10, 19, 20], we determined whether 5-MTP inhibits TGFβ1-mediated Smad2 pathway through the p38 MAPK signaling pathway in HPTECs. TGFβ1 induced p38 MAPK activation which was blocked by pretreatment of HPTECs with 5-MTP (Fig. 8c). The effect of TGFβ1 on Smad2 phosphorylation was abrogated by a selective inhibitor of p38 MAPK, i.e., SB202190 (Fig. 8d, e). These results suggest that 5-MTP inhibits TLR2- and TGFβ1-mediated renal tubular epithelial cell signaling pathways and phenotypic changes in part by blocking p38 MAPK activation and the downstream pro-fibrotic and pro-inflammatory signaling pathways and transcriptional programs.
5-MTP blocks TGFβ1-mediated Smad2 activation in HPTECs and UUO-injured renal tissues. a HPTECs were treated with 5-MTP for 4 h followed by TGFβ1 (10 ng/mL) for 30 min. p-Smad2 and Smad2 were analyzed by immunoblotting and quantified by densitometry. Each bar denotes mean ± SD (n = 3). b p-Smad2 in renal tissues of UUO-injured mice treated with 5-MTP or saline was analyzed by IHC and quantified by Image J software. Each dot in the right panel represents experiments done in a mouse. *p < 0.05. c HPTECs were treated with 5-MTP for 4 h followed by TGFβ1 (10 ng/mL) for 1 h. p-p38 and p38 were analyzed by immunoblotting and quantified by densitometry. d, e HPTECs were pretreated with 30 μm SB202190 or 25 μm 5-MTP for 4 h for 1 h followed by TGFβ1 (10 ng/mL) for 1 h. p-p38, p38, p-Smad2, and Smad2 were analyzed by immunoblotting and quantified by densitometry. Each bar denotes mean ± SD (n = 3). *p < 0.05, **p < 0.005, ***p < 0.0005.
5-MTP blocks TGFβ1-mediated Smad2 activation in HPTECs and UUO-injured renal tissues. a HPTECs were treated with 5-MTP for 4 h followed by TGFβ1 (10 ng/mL) for 30 min. p-Smad2 and Smad2 were analyzed by immunoblotting and quantified by densitometry. Each bar denotes mean ± SD (n = 3). b p-Smad2 in renal tissues of UUO-injured mice treated with 5-MTP or saline was analyzed by IHC and quantified by Image J software. Each dot in the right panel represents experiments done in a mouse. *p < 0.05. c HPTECs were treated with 5-MTP for 4 h followed by TGFβ1 (10 ng/mL) for 1 h. p-p38 and p38 were analyzed by immunoblotting and quantified by densitometry. d, e HPTECs were pretreated with 30 μm SB202190 or 25 μm 5-MTP for 4 h for 1 h followed by TGFβ1 (10 ng/mL) for 1 h. p-p38, p38, p-Smad2, and Smad2 were analyzed by immunoblotting and quantified by densitometry. Each bar denotes mean ± SD (n = 3). *p < 0.05, **p < 0.005, ***p < 0.0005.
Discussion
This study provides new information about the renal protective actions of 5-MTP. A novel finding is that 5-MTP attenuates renal tissue inflammation and fibrosis by blocking TLR2-mediated macrophage activation, tubular epithelial cell phenotypic change, and myofibroblast generation. Infiltrating macrophages play a critical role in renal tissue inflammation and fibrosis through secretion of pro-inflammatory cytokines and pro-fibrotic factors [21, 22]. Previous reports indicate that 5-MTP represents a powerful arsenal against inflammation by blocking macrophage p38 MAPK and NF-κB activation [7, 20]. Findings from this study gain insights into the mechanism by which 5-MTP suppresses the pro-inflammatory and pro-fibrotic properties of macrophages. 5-MTP blocks macrophage migration and accumulation in UUO-damaged renal tissues. It disrupts interactions between TLR2 and MyD88. The TLR2-MyD88 signaling pathway has been implicated in renal fibrosis as MyD88 knockout reduced interstitial fibrosis in UUO mice [23]. TLR2 and MyD88 transmit signals to activate IκB kinase, resulting in activation of p65 NF-κB and MKK3/6-p38 MAPK via TRAF6. Our results show that 5-MTP blocks not only TLR-MyD88 interaction but also binding of TRAF6 to the TLR2-MyD88 complex thereby reducing NF-κB and p38 MAPK activation. p38 MAPK is a key stress-responsive signaling molecule and has been shown to mediate renal interstitial fibrosis in the murine UUO model [24]. NF-κB and p38-dependent transcription factors promote expression of pro-inflammatory cytokines such as IL-6 and pro-fibrotic factors such as TGFβ. 5-MTP pretreatment of murine macrophages attenuates Pam3-induced IL-6 and TGFβ1 expression. Taken together, our findings suggest that 5-MTP attenuates UUO-induced renal fibrosis by blocking TLR2-MyD88-TRAF6 mediated activation of NF-κB and p38 MAPK and thereby attenuating expression and secretion of pro-fibrotic factors from macrophages.
In response to renal injury, monocyte-derived macrophages are attracted to the injured tissues where they aggravate renal damages by enhancing apoptosis and eliciting inflammation [14, 25]. The infiltrating macrophages play a key role in fibrosis [25]. Macrophages exhibit different phenotypes at different stages of renal injury. It was reported in murine UUO models that monocyte-derived F4/80+ macrophages play a crucial role in the development of renal fibrosis [26]. Depletion of F4/80+ macrophages by liposome clodronate greatly reduced renal fibrosis. Our results indicate that 5-MTP treatment reduced infiltration of CD11b+F4/80+ macrophages to an extent comparable to genetic deletion of TLR2. The findings suggest that 5-MTP blocks monocyte-derived macrophage migration and renal tissue infiltration by blocking TLR2. Both M1 and M2 macrophages are involved in renal interstitial fibrogenesis. M1 macrophages were reported to mediate renal fibrosis through release of pro-inflammatory cytokines notably IL-6 and TGFβ [14, 25]. Of note, IL-6 production is not limited to macrophages. Renal fibroblasts and epithelial cells including proximal tubular epithelial cells are significant sources of IL-6 production [27, 28]. IL-6 plays a key role in promoting inflammation and inflammation-mediated fibrosis by a number of mechanisms such as recruiting immune cells to the damage tissues, enhancing signaling pathways particularly those of TGFβ, and activation of fibroblasts which result in amplification of tissue inflammation and damage and progression of tissue fibrosis [29]. TGFβ, on the other hand, is a key driver of tissue fibrosis. M2 macrophages play a complex role: at early phase, it promotes fibrosis while at the repair phase, it is anti-inflammatory and may be pro-fibrotic [14, 25]. It was reported that genetic deletion of TLR2 or MyD88 reduces M2 macrophage infiltration in the UUO-damaged renal tissues accompanied by attenuation of interstitial fibrosis in the murine UUO model [23]. Contribution of infiltrating macrophages to renal fibrosis may depend on the M1/M2 ratio. Our results suggest that 5-MTP protects renal tissues against UUO-induced fibrosis by reducing M1/M2 ratio.
Under renal injury, renal proximal tubular epithelial cells undergo phenotypic switch which shares cellular characteristics with myofibroblasts such as expression of α-SMA, FN, and vimentin [30‒32]. This phenotypic switch is akin to epithelial mesenchymal transition, although it was questioned whether renal epithelial to mesenchymal transition occurs in vivo [33]. Nevertheless, the phenotypic switch contributes to renal fibrosis by secreting pro-fibrotic and pro-inflammatory factors and increasing inflammatory cell infiltration and extracellular matrix deposition [34, 35]. By using FN as a surrogate marker, we confirmed in cultured HPTECs that tubular epithelial cells undergo phenotypic switch. Our results show that TLR2 and TGFβ1 induce FN as well as IL-6 and TGFβ expressions consistent with the interpretation that tubular epithelial cells undergo myofibroblast transdifferentiation and contribute to local interstitial fibrosis [36]. 5-MTP is effective in attenuating fibrosis by suppressing TLR2 expression as well as TLR2-mediated signaling pathways via p38 MAPK and IκB kinase-NF-κB. In addition, 5-MTP blocks expression of TGFβ1 expression and the TGFβ-Smad2/3 signaling pathway. Our results provide novel information about the underlying mechanisms. The findings suggest that 5-MTP suppresses TLR2- and TGFβ1-mediated signaling pathways through a common mechanism, i.e., blocking p38 MAPK activation via which it not only inhibits the pro-inflammatory and pro-fibrotic actions of TLR2 but also TGFβ1-Smad2/3 signaling.
Our results confirm that 5-MTP is constitutively produced in renal tubular cells and its production is suppressed by renal injury induced by UUO. Intraperitoneal injection of 23.5 mg/kg of L-5-MTP was previously reported to raise the serum 5-MTP concentration about 4-fold over the basal 5-MTP level which was effective in controlling LPS-induced systemic inflammation in a murine model [7]. The results from this study suggest that intraperitoneal injection of 5-MTP (23.5 mg/kg) protects renal tissues from UUO damage by replenishing the 5-MTP level in the renal tissues. 5-MTP suppression due to tissue injury by inflammatory mediators, ischemia reperfusion, ureteral obstruction, and endotoxins has been attributed to inhibition of the expression of TPH-1 [11, 19, 37]. The mechanism by which renal injury suppresses TPH-1 expression remains to be elucidated. However, it is worth noting that TPH-1 overexpression enhances renal epithelial cells 5-MTP production and increases renal protective actions while silencing of TPH-1 exacerbates injury-induced renal inflammation and fibrosis [11]. TPH-1/5-MTP is an innate tryptophan catalytic pathway which is physiologically important in renal protection as well as vascular and cardiac protections.
Recent reports provide strong evidence for the anti-fibrotic actions of 5-MTP [38]. 5-MTP was reported in a MI model to attenuate post-MI myocardial tissue fibrosis and thus reduce structural remodeling and preserve left ventricular function [37]. In addition, 5-MTP inhibits pulmonary fibrosis and liver fibrosis in murine models [39, 40]. Tissue fibrosis signifies advanced tissue damage and severe organ failure which carry high morbidity and mortality. There is no efficacious therapy for fibrotic disorders. 5-MTP may fulfill the unmet medical need as a lead compound for therapy of this group of deadly human diseases.
In summary, this study demonstrates that renal tissue injury by UUO results in declining 5-MTP production and 5-MTP supplements restores the renal protective actions. 5-MTP protects against renal tissue inflammation and fibrosis by reducing macrophage infiltration, TLR2-mediated macrophage activation and fibroblast transformation to myofibroblasts. In addition, 5-MTP controls proximal tubular epithelial cell transdifferentiation to myofibroblasts.
This study sheds light on the mechanism by which 5-MTP protects against UUO-induced renal damage and interstitial fibrosis. 5-MTP controls injury-induced TLR2 upregulation and disrupts the TLR2-MyD88-TRAF6 signaling pathway in macrophages thereby inhibiting the pro-fibrotic transcriptional programs. Our findings suggest that 5-MTP exerts multiple actions on macrophages and renal tubular epithelial cells by a common mechanism involving the inhibition of p38 MAPK and NF-κB pathways [41, 42]. Importantly, 5-MTP blocks p38 MAPK-mediated activation TGFβ1-Smad2/3 pathway thereby enforcing its anti-fibrotic actions. Our study advances the understanding how 5-MTP acts via the TLR2-mediated signaling pathways and provides mechanistic insights to support the use of 5-MTP as a lead compound to develop novel anti-fibrotic drugs.
Statement of Ethics
This animal study protocol was reviewed and approved by Institutional Animal Care and Use Committee (IACUC) of AAA-LAC-Accredited animal facility, National Health Research Institutes, Approval No. (Protocol No. NHRI-IACUC-107037).
Conflict of Interest Statement
The authors declare no competing financial interests.
Funding Sources
This work was supported by National Health Research Institutes (NHRI) extramural grant to Hsu Y-J (NHRI-EX-106, 107 and 108-10631SI) and intramural support to Kuo C-C (CS-111-PP-11 and CS-112-PP-11) and by National Science Council (NSC)/Ministry of Science and Technology (MOST) of Taiwan (MOST 107-2321-B-400-011, 108-2321-B-400-008, 109-2326-B-400-001, 110-2326-B-400-005, and 111-2326-B-400-008).
Author Contributions
Wu J.-Y., Lee G.-L., and Chueh Y.-F. performed cellular and animal experiments and analyzed the data. Wu J-Y prepared the figures and wrote the manuscript, Kuo C.-C. and Hsu Y.-J. designed the study and supervised the laboratory work. Wu K.K. designed the study and wrote the manuscript. All authors read and approved the manuscript.
Data Availability Statement
All data supporting the results reported in this article are included in the article and its supplementary file. Further inquiries can be directed to the corresponding author.