Compound C exerts a therapeutic effect on Graves' orbitopathy via AMPK‑independent pathways

Introduction

Graves' orbitopathy (GO) is an autoimmune manifestation of Graves' disease that affects the orbit (1). While the pathogenesis of GO is not yet fully understood, it is widely accepted that orbital fibroblasts express thyroid-stimulating hormone receptor, which interacts with the insulin-like growth factor-1 receptor (IGF-1-R), facilitating the recruitment of autoreactive T and B cells into the orbit (2,3). These infiltrating immune cells stimulate orbital fibroblasts to secrete inflammatory modulators, and to proliferate and differentiate into adipocytes (1,4). The interaction between activated immune cells and orbital fibroblasts exacerbates orbital inflammation and promotes the deposition of extracellular matrix components, such as hyaluronan (5). Consequently, chronic inflammation and expansion of orbital connective tissue volume can ultimately lead to erythema, swelling, proptosis and sight-threatening compressive optic neuropathy (6).

Conventionally, high-dose systemic corticosteroid and radiation therapy are considered as the first-line treatments for GO (7); however, due to its clinical heterogeneity and variable disease course, notable efforts have been made to target the underlying pathophysiology at the molecular level (8). Several targeted biological and immunosuppressive agents, including those targeting IL-6, CD20, fragment crystallizable receptors and IGF-1-R, have been identified and are under pre-clinical and clinical trials for use in GO (9–11).

Adenosine monophosphate-activated protein kinase (AMPK) is a key phosphorylation enzyme involved in regulating fatty acid and cholesterol synthesis, which can be activated under various conditions of metabolic stress (12). AMPK has garnered substantial attention as a potential therapeutic target for metabolic disorder-related diseases, including diabetes, obesity, fatty liver disease and cancer (13,14). Under nutrient-rich conditions, AMPK is inactivated by an increased AMP/ATP ratio, driven by glycolysis and mitochondrial oxidative phosphorylation (15). Compound C, also known as dorsomorphin, is a cell permeable AMPK inhibitor, which also functions as a broad-spectrum kinase inhibitor with anti-proliferative effects (16). Increasing evidence has suggested that compound C inhibits multiple kinases beyond AMPK, which have been implicated in various neurological diseases such as glioma (17,18). Therefore, the present study aimed to investigate the effects of compound C on GO pathogenesis, focusing on inflammation, fibrosis, hyaluronan production and adipogenesis.

Materials and methods

Reagents and cell culture

The suppliers of reagents used in the present study are listed in Table I. Orbital tissue explants were obtained from patients with GO (n=12) during orbital decompression surgery and healthy controls (n=12) during upper or lower blepharoplasty (Table II). Patients with GO were clinically inactive and had not received systemic steroid treatment or radiation therapy 3 months prior to the surgery. Normal healthy controls were free of thyroid or autoimmune diseases. The present study was approved by the Institutional Review Board of Severance Hospital (approval number 4-2024-0492; Seoul, South Korea) and adhered to the tenets of The Declaration of Helsinki. All of the participants provided written informed consent. Orbital fibroblasts were obtained from each orbital tissue explant were cultured as previously described (19). The harvested orbital tissue explants were primarily cultured in DMEM/F12 1:1 containing 1% of penicillin (100 U/ml)/streptomycin (100 µg/ml) and 20% FBS to promote fibroblast proliferation. Once cell proliferation was confirmed, the monolayers were serially passaged using trypsin/EDTA. The cells were stored in liquid nitrogen, and orbital fibroblasts between the third and sixth passages were used for the experiments.

Table I. Suppliers of study reagents.

Table I.

Suppliers of study reagents.

Reagent (cat. no.)	Supplier
Compound C; Oil Red O powder; MTT assay	Millipore Sigma
DMEM; DMEM/F12; penicillin-streptomycin; trypsin/EDTA	Welgene, Inc.
FBS; Halt™ protease inhibitor cocktail	Thermo Fisher Scientific, Inc.
Recombinant human IL-1β; recombinant human TGF-β; hyaluronan	R&D Systems, Inc.
ELISA kit (DHYAL0)
Antibodies against PPAR-γ (sc-7273), C/EBPβ (sc-7962), adiponectin	Santa Cruz Biotechnology, Inc.
(sc-136131), leptin (ob) (sc-48408), aP2 (sc-271529), SREBP-1
(sc-365513), and β-actin (sc-47778)
Antibodies against IL-6 (12153), IL-8 (94407), MCP-1 (2027), ICAM-1	Cell Signaling Technology, Inc.
(4915), C/EBPα (2295), p-AMPK (2535), t-AMPK (2532), p-YAP s127
(13008), t-YAP (4912), TAZ (70148), p-AKT (9271), t-AKT (9272), p-JNK
(9251), t-JNK (9252), p-ERK (9101), t-ERK (9102), p-p38 (9211), t-p38
(9212), p-STAT3 (94994), t-STAT3 (30835), p-SMAD1 (13820), p-SMAD2
(3108), p-SMAD3 (9520), t-SMAD1 (9743), t-SMAD2 (5339), t-SMAD3
(9523), t-SMAD4 (38454)
RIPA buffer	Biosesang
RNAiMAX; antibody against p-SMAD4 (PA5-64712)	Invitrogen; Thermo Fisher Scientific, Inc.
Silencer™ Select Negative Control No. 1 siRNA (cat. no. 4390844)	Ambion; Thermo Fisher Scientific, Inc.
Antibody against fibronectin (610077)	BD Biosciences
Antibody against collagen Ia (ab138492)	Abcam
AICAR; antibody against α-SMA (A5228), DMSO	Sigma-Aldrich

[i] AICAR, 5-aminoimidazole-4-carboxamide ribonucleotide; AMPK, adenosine monophosphate-activated protein kinase; C/EBP, CCAAT/enhancer-binding protein; COX-2, cyclooxygenase 2; ELISA, enzyme-linked immunosorbent assay; ICAM-1, intracellular adhesion molecule-1; MCP-1, monocyte chemoattractant protein-1; p-, phosphorylated; PPAR, peroxisome proliferator-activated receptor; siRNA, small interfering RNA; SREBP, sterol regulatory element-binding protein; t-, total; TAZ, transcriptional co-activator with PDZ-binding motif; YAP, yes-associated protein; DMSO, dimethyl sulfoxide.

Table II. Patient demographics.

Table II.

Patient demographics.

A, Patents with Graves' orbitopathy
Age, years	Sex	CAS	Duration, years	Proptosis R/L, mm	Type of surgery
53	F	1	1.5	19/16	Orbital wall decompression
42	F	0	2	22/22	Orbital wall decompression
26	M	0	3	21/22.5	Orbital wall decompression
29	F	2	2	21/21	Orbital wall decompression
27	M	1	2.5	22/22	Orbital wall decompression
31	M	0	3	23/22.5	Orbital wall decompression
35	F	2	4	23/23	Orbital wall decompression
43	F	1	2	20/20	Orbital wall decompression
37	M	1	1	21/21.5	Orbital wall decompression
41	M	0	1.5	22/22.5	Orbital wall decompression
42	F	1	2	21/20	Orbital wall decompression
48	M	1	3.5	21/21	Orbital wall decompression
B, Healthy control patients
Age, years	Sex	CAS	Duration, years	Proptosis R/L, mm	Type of surgery
43	F				Upper lid blepharoplasty
71	M				Upper lid blepharoplasty
65	F				Lower lid blepharoplasty
53	F				Lower lid blepharoplasty
42	M				Upper lid blepharoplasty
48	F				Upper lid blepharoplasty
52	M				Upper lid blepharoplasty
59	F				Lower lid blepharoplasty
61	M				Lower lid blepharoplasty
54	M				Lower lid blepharoplasty
59	F				Upper lid blepharoplasty
39	F				Upper lid blepharoplasty

[i] CAS, clinical activity score; F, female; M, male; R/L, right/left.

Cell viability

To determine the non-toxic concentrations of compound C on orbital fibroblasts, orbital fibroblasts extracted from tissues obtained from patients with GO were seeded into 24-well culture plates (1×104 cells/well) and were treated with various concentrations of compound C (0–15 µM) for 72 h, 37°C under 5% humified CO2. After treatment, the cells were washed twice with PBS and incubated in 5 mg/ml MTT solution at 37°C for 3 h. Formazan was then dissolved using ice-cold isopropanol and absorbance was measured at 560 nm with background correction at 630 nm, using a microplate reader (EL 340 Bio Kinetics Reader; BioTek; Agilent Technologies, Inc.). Cell viability was calculated as a percentage relative to untreated control cells.

Western blotting

Confluent orbital fibroblasts were pretreated with compound C (5 or 10 µM for 6 h) or 5-aminoimidazole-4-carboxamide ribonucleotide (AICAR; AMPK agonist; Sigma-Aldrich; Merck KGaA); 5 or 10 µM for 6 h) followed by stimulation with either IL-1β (10 ng/ml for 24 h) or TGF-β (5 ng/ml for 24, 48 or 72 h) at 37°C to assess the expression levels of inflammatory cytokines and fibrosis-related proteins. The control group was treated with 0.1% DSMO (Sigma-Aldrich; Merck KGaA) at 37°C for 6 h as each experimental group and received no stimulation. In addition, orbital fibroblasts were cultured in an adipogenic medium for 10 days with or without compound C treatment to evaluate the expression levels of adipogenic transcription factors. Western blot analysis was performed on orbital fibroblasts subjected to the different study conditions. Proteins were extracted with RIPA buffer (Biosesang) with Halt™ Protease Inhibitor Cocktail (Thermo Fisher Scientific, Inc.). Protein concentration was quantified by BCA assay (Pierce™ BCA Protein Assay Kit; Thermo Fisher Scientific, Inc.). Equal amounts of protein from orbital fibroblasts under different study conditions (50 µg) were separated by electrophoresis on 8–15% SDS polyacrylamide gels. Blocking was then performed in 5% skim milk for 24 h at 4°C. The resolved proteins were then transferred onto nitrocellulose membranes and were incubated with primary antibodies overnight at 4°C. The immunoreactive bands were detected using horseradish peroxidase-conjugated secondary antibodies. After exposure to Supersignal™ West Pico PLUS Chemiluminescent Substrate (Thermo Fisher Scientific, Inc.), the membranes were visualized using an X-ray film or image reader (LAS-4000 Mini; FUJIFILM Wako Pure Chemical Corporation). The density of each band was measured and semi-quantified using ImageJ software version 1.54 (National Institutes of Health). The experimental values were normalized to the β-actin levels of the same sample.

Adipogenesis

Orbital fibroblasts were cultured for 10 days in adipogenic medium, which comprised high-glucose DMEM supplemented with 10% FBS and 1% penicillin (100 U/ml)/streptomycin (100 µg/ml) (Welgene, Inc.), 17 µM pantothenic acid (Sigma-Aldrich; Merck KGaA), 33 µM biotin (Sigma-Aldrich; Merck KGaA), 10 µg/ml transferrin (Sigma-Aldrich; Merck KGaA), 1 µM insulin (Roche Diagnostics GmbH), 0.2 nM T3 (Sigma-Aldrich; Merck KGaA) and 0.2 µM carbaprostaglandin (Calbiochem; Merck KGaA), as previously described (20). Adipogenic medium was replaced every 2-3 days. For the first 4 days of cell culture, the differentiation medium contained 10 µM dexamethasone and 0.1 mM 3-isobutyl-1-methylxanthine (IBMX); after day 5, dexamethasone and IBMX were excluded. From the first day of adipogenic differentiation, stimulation of adipogenesis was enhanced by the addition of the peroxisome proliferator-activated receptor γ (PPARγ) agonist rosiglitazone (Cayman, Ann Arbor, MI, USA), IL-1β (10 ng/ml) or both. For treatment with compound C (5 or 10 µM), it was added to the adipogenic medium. To evaluate the effects of yes-associated protein (YAP) and transcriptional co-activator with PDZ-binding motif (TAZ) suppression on adipogenesis, cells were transfected with YAP/TAZ small interfering (si)RNA (10 nM) prior to the induction of adipogenesis. Adipogenic differentiation was induced for 10 days at 37°C.

Oil Red O staining

Orbital fibroblasts under different study conditions (with or without either IL-1β or compound C treatment, as aforementioned) were stained with Oil Red O (Sigma-Aldrich; Merck KGaA). Briefly, a stock solution of Oil Red O Solution was diluted with 4 ml distilled water to prepare the working solution. Subsequently, the working solution was filtered using a 0.2-µm filter. Adipogenically differentiated cells were washed with PBS and fixed with 10% formalin at 4°C for 1 h. After fixation, the cells were washed with PBS, incubated with 60% isopropanol for 5 min and dried completely. The cells were then stained with Oil Red O solution for 1-2 h at room temperature and washed with PBS. Stained cells were observed under a light and confocal microscope (IX73; Olympus Corporation). In addition, the Oil Red O stain was solubilized with 100% isopropanol to quantify lipid accumulation and optical density was measured using a spectrophotometer at 490 nm.

Silencing of YAP and TAZ

Orbital fibroblasts from patients with GO (50% confluency) were transfected with siRNA (10 nM) using Lipofectamine® RNAiMAX (Invitrogen; Thermo Fisher Scientific, Inc.). Silencer™ Select Negative Control No. 1 siRNA (si-control; 10 nM) was used as a control. Following transfection at 37°C for 48 h in a humified incubator at 5% CO2, the orbital fibroblasts were cultured in adipogenic medium to induce adipogenesis, as aforementioned. The following siRNAs were used in the present study: si-YAP: 5′-GACAUCUUCUGGUCAGAGADTDT-3′; si-TAZ: 5′-ACGUUGACUUAGGAACUUUDTDT-3′.

ELISA

Orbital fibroblasts from patients with GO were pretreated with compound C (10 µM) for 6 h, and were subsequently stimulated with either IL-1β (10 ng/ml) or TGF-β (5 ng/ml) for 96 h at 37°C to determine hyaluronan production. Supernatants were collected from orbital fibroblasts under various conditions, and hyaluronan production levels were measured using a commercially available ELISA kit according to the manufacturer's instructions. Absorbance was measured at 450 nm, and the binding percentage of each sample was calculated. A standard binding curve was generated for quantification.

Statistical analysis

All experiments were performed in duplicate for each of the three strains, which are derived from three different individuals. After Kolmogorov-Smirnov test, Kruskal-Wallis test was performed to compare quantitative data between the groups with or without treatment with compound C, IL-1β or TGF-β, and to compare quantitative data between groups with or without siRNA transfection. For pairwise comparisons, Dunn's post hoc test was applied after Kruskal-Wallis test to adjust for multiple comparisons. Statistical analysis was performed using SPSS version 20.0 for Windows (IBM Corp.), and P<0.05 was considered to indicate a statistically significant difference.

Results

Effect of compound C on the viability of orbital fibroblasts

To determine the non-toxic concentration of compound C in orbital fibroblasts, an MTT assay was conducted. Orbital fibroblasts from patients with GO were exposed to various concentrations of compound C (0–15 µM) for 72 h (Fig. S1). For cells treated with 10 µM compound C, viability was >95%, therefore ≤10 µM compound C was applied for further analysis.

Compound C inhibits AMPK activity and YAP phosphorylation in orbital fibroblasts

After confirming the non-cytotoxic concentration of compound C, the effects of compound C (AMPK antagonist) and AICAR (AMPK agonist), on AMPK and YAP activation were assessed. Orbital fibroblasts from patients with GO and healthy controls were treated with 5 and 10 µM compound C or AICAR for 6 h, and phosphorylated (p-)AMPK and p-YAP protein levels were determined using western blotting. Western blot analysis revealed that compound C reduced AMPK and YAP activation, whereas AICAR enhanced AMPK activation but did not affect YAP activation, in both GO and normal orbital fibroblasts, indicating the ability of compound C to inhibit AMPK activity and YAP phosphorylation (Fig. 1).

Figure 1. Western blot analysis of AMPK and YAP after treatment with compound C or AICAR. Orbital fibroblasts from patients with GO (n=3) and healthy controls (n=3) were treated with 5 or 10 µM compound C or AICAR for 6 h. The protein expression levels of p-AMPK were decreased after treatment with compound C (5 and 10 µM) and p-YAP was significantly decreased after treatment with compound C (10 µM) in both GO and healthy control groups. AICAR treatment significantly activated AMPK in both GO and healthy control groups. Data are presented as the mean ± standard deviation. *P<0.05 vs. DMSO group. AICAR, 5-aminoimidazole-4-carboxamide ribonucleotide; AMPK, adenosine monophosphate-activated protein kinase; GO, Graves' orbitopathy; p-, phosphorylated; t-, total; YAP, yes-associated protein; DMSO, dimethyl sulfoxide.

Compound C reduces inflammatory cytokine production

IL-1β is a key inflammatory mediator in GO; therefore, orbital fibroblasts from patients with GO and healthy controls were challenged with compound C (10 µM) for 6 h, followed by IL-1β stimulation (10 ng/ml) for 24 h. Treatment with compound C significantly reduced IL-1β-induced expression of IL-6, IL-8, and monocyte chemoattractant protein (MCP)-1 in both GO and healthy control orbital fibroblasts, and intracellular adhesion molecule (ICAM)-1 in GO orbital fibroblasts, suggesting that compound C suppressed inflammation in orbital fibroblasts (Fig. 2).

Figure 2. Compound C inhibits IL-1β-induced inflammatory cytokine expression. Orbital fibroblasts from both GO (n=3) and healthy control (n=3) were pretreated with compound C (10 µM) for 6 h and were then stimulated with IL-1β (10 ng/ml) for 24 h. Protein expression levels of IL-6, IL-8, ICAM-1 and MCP-1 were examined by western blotting. Compound C treatment significantly attenuated IL-1β-induced cytokine activation in both GO and healthy control groups. Data are presented as the mean ± standard deviation. *P<0.05, **P<0.01. GO, Graves' orbitopathy; ICAM, intercellular adhesion molecule; MCP, monocyte chemoattractant protein.

Effects of compound C on TGF-β-induced fibrosis-related protein expression

TGF-β serves a crucial role in fibrosis and tissue remodeling response in GO (21); therefore, the GO orbital fibroblasts were pretreated with compound C (10 µM) for 6 or 24 h prior to TGF-β (5 ng/ml) stimulation for 24, 48 or 72 h. Compound C treatment reduced TGF-β-induced expression of fibronectin and collagen Iα in GO orbital fibroblasts across all time points, and α-smooth muscle actin at 48 and 72 h, indicating that compound C abrogated TGF-β-induced fibrotic response in GO orbital fibroblasts (Fig. 3).

Figure 3. Compound C reduces the TGF-β-induced increases in profibrotic protein levels in orbital fibroblasts. Orbital fibroblasts from patients with GO (n=3) were pretreated with compound C (10 µM) for 6 or 24 h, and were then stimulated with TGF-β (5 ng/ml) for 24, 48 or 72 h. The expression levels of fibronectin, Col Iα and α-SMA were determined by western blotting. Compound C treatment significantly attenuated TGF-β-induced fibronectin, Col Iα and α-SMA expression in GO orbital fibroblasts. Data are presented as the mean ± standard deviation. *P<0.05. Col Iα, collagen Iα; GO, Graves' orbitopathy; SMA, smooth muscle actin.

Effects of compound C on adipogenic differentiation

To determine the effects of compound C on adipogenesis in GO, orbital fibroblasts from patients with GO were treated with 5 or 10 µM compound C and were cultured for 10 days to induce adipogenesis. Notably, compound C significantly inhibited lipid accumulation (Fig. 4A and B). At day 10 of adipogenesis, western blot analysis revealed that 10 µM compound C markedly suppressed the expression of adipogenic transcription factors, including CCAAT/enhancer-binding protein (C/EBP)α/β, PPARγ and adiponectin in GO orbital fibroblasts, suggesting that compound C suppressed adipogenic differentiation in GO orbital fibroblasts (Fig. 4C and D).

Figure 4. Effects of compound C on adipogenesis. Orbital fibroblasts from patients with GO (n=3) were incubated in adipogenic medium for 10 days with or without compound C pretreatment (5 or 10 µM). (A) Oil Red O staining result. (B) Compound C significantly attenuated adipogenesis in orbital fibroblasts from patients with GO compared to the control group. (C) Western blot analysis of adipogenic factors at day 10 of adipogenesis. (D) Treatment with 10 µM compound C suppressed expression of C/EBPα/β, PPARγ and adiponectin. *P<0.05, **P<0.01 vs. control or as indicated. Data are presented as the mean ± standard deviation. C/EBP, CCAAT/enhancer-binding protein; GO, Graves' orbitopathy; PPAR, peroxisome proliferator-activated receptor.

Silencing YAP and TAZ inhibits adipogenic differentiation

To investigate the effects of YAP and TAZ silencing on adipogenesis, orbital fibroblasts from patients with GO were transfected with si-control, si-YAP or si-TAZ, and cultured in adipogenic medium for 10 days. Silencing of YAP, TAZ or both suppressed adipogenic differentiation (Fig. 5A and C). Furthermore, YAP silencing reduced the expression levels of C/EBPα/β, sterol regulatory element-binding protein (SREBP)-1, aP2, adiponectin and leptin. Similarly, TAZ silencing decreased the expression of PPARγ, C/EBPα/β, SREBP-1, aP2, adiponectin and leptin (Fig. 5B and D). Taken together, these findings indicated that silencing YAP and TAZ suppressed adipogenic protein expression.

Figure 5. Effects of YAP/TAZ silencing on adipogenesis. YAP and TAZ siRNA transfection in orbital fibroblasts from patients with GO (n=3) was performed to assess the effects of YAP/TAZ knockdown on adipogenesis. (A) Representative Oil Red O staining results. Cells were stained with Oil-Red O and examined using a microscope (×40 magnification). (B) Western blot analysis of adipogenic factors and hormones. Protein expression levels of adipogenic transcription factors: PPARγ, C/EBPα, C/EBPβ, SREBP-1 and aP2, and hormones: Adiponectin and leptin, were analyzed by western blotting on days 0 and 10 of differentiation with or without si-control, si-YAP or si-TAZ transfection. (C) Oil-Red O staining was quantified by measuring the optical density of cell lysates at 490 nm. Silencing YAP or TAZ significantly suppressed the adipogenesis compared with in the si-control group. (D) YAP silencing reduced the expression levels of C/EBPα/β, SREBP-1, aP2, adiponectin and leptin. Similarly, TAZ silencing decreased the expression of PPARγ, C/EBPα/β, SREBP-1, aP2, adiponectin and leptin. Data are presented as mean ± standard deviation. *P<0.05, **P<0.01 vs. control or as indicated. GO, Graves' orbitopathy; C/EBP, CCAAT/enhancer-binding protein; PPAR, peroxisome proliferator-activated receptor; si, small interfering; SREBP, sterol regulatory element-binding protein; TAZ, transcriptional co-activator with PDZ-binding motif; YAP, yes-associated protein.

Compound C inhibits GO pathogenesis by suppressing Akt and SMAD phosphorylation

To evaluate the effect of high-dose compound C on intracellular signaling pathways, the expression levels of p-STAT3, p-Akt, p-ERK, p-p38 and p-JNK were assessed with or without compound C treatment (10 µM) for 6 h in both GO and healthy control orbital fibroblasts. Notably, treatment with IL-1β (10 ng/ml) for 30 min induced phosphorylation of Akt, which was significantly suppressed by compound C treatment in both GO and healthy control orbital fibroblasts, whereas other intracellular signaling pathways were not affected (Fig. 6).

Figure 6. High-dose of compound C suppresses Akt activation in orbital fibroblasts. Orbital fibroblasts from both patients with GO (n=3) and healthy controls (n=3) were treated with 10 µM compound C for 6 h and with IL-1β (10 ng/ml) for 30 min. (A) Western blot analysis revealed that treatment with compound C suppressed phosphorylation of Akt in both GO and healthy control orbital fibroblasts. (B) Relative protein density of p-Akt. Data are presented as the mean ± standard deviation. *P<0.05. GO, Graves' orbitopathy; p-, phosphorylated; t-, total.

In a separate experiment, orbital fibroblasts from both patients with GO and healthy controls were pretreated with compound C (10 µM) for 6 h and were then stimulated with TGF-β (5 ng/ml) for 1 h. Compound C treatment significantly reduced SMAD1, 2 and 3 phosphorylation in both GO and normal orbital fibroblasts (Fig. 7). Taken together, these findings indicated that compound C may inhibit GO pathogenesis by suppressing Akt and SMAD 1/2/3 phosphorylation.

Figure 7. A high-dose of compound C suppresses canonical TGF-β signaling pathways in orbital fibroblasts. Orbital fibroblasts from patients with GO (n=3) and healthy controls (n=3) were exposed to compound C (10 µM) for 6 h and were activated with TGF-β (5 ng/ml) for 1 h. (A) High-dose of compound C treatment attenuated TGF-β-induced activation of SMAD1, 2 and 3 in orbital fibroblasts from both GO and healthy control. (B) Relative protein density of p-SMAD1, 2 and 3. Data are presented as mean ± standard deviation. *P<0.05. GO, Graves' orbitopathy; p-, phosphorylated; t-, total.

Effects of compound C on hyaluronan production

In addition to inflammation, adipogenesis and fibrosis, excessive hyaluronan accumulation contributes to tissue remodeling in GO (21). To evaluate the effects of compound C on hyaluronan secretion, ELISA was performed. GO orbital fibroblasts treated with IL-1β (10 ng/ml) or TGF-β (5 ng/ml) for 96 h exhibited markedly increased hyaluronan production, which was significantly reversed by treatment with compound C (10 µM) for 6 h, indicating that compound C suppressed hyaluronan production (Fig. 8).

Figure 8. High-dose of compound C inhibits IL-1β- or TGF-β induced hyaluronan production in GO orbital fibroblasts. Orbital fibroblasts from patients with GO (n=3) were treated with high-dose compound C for 6 h prior to stimulation with either IL-1β (10 ng/ml) or TGF-β (5 ng/ml) for 96 h, and the concentration of hyaluronan was determined with ELISA. High-dose compound C attenuated IL-1β- or TGF-β induced secretion of hyaluronan in orbital fibroblasts from patients with GO. Data are presented as the mean ± standard deviation. *P<0.05. GO, Graves' orbitopathy.

Discussion

The present study indicated that compound C, a widely used AMPK inhibitor, may effectively reverse inflammation, adipogenesis and fibrosis in GO orbital fibroblasts. At a concentration of 10 µM, compound C suppressed the phosphorylation of YAP, a mechano-sensing signal protein, as well as AMPK. It also inhibited IL-1β-induced pro-inflammatory cytokine production, TGF-β-induced profibrotic protein expression and adipogenic differentiation. Additionally, compound C attenuated IL-1β- and TGF-β-stimulated hyaluronan secretion in GO orbital fibroblasts. Notably, the silencing of YAP/TAZ also suppressed the expression of adipogenic markers. Although the precise mechanism remains unclear, it could be hypothesized that the suppression of various key GO pathologies by compound C is related to YAP inactivation.

AMPK is a highly conserved energy sensor that serves as a central regulator of cellular processes, including cell growth, proliferation, survival and metabolic regulation (22). Decreased AMPK levels have been reported in various metabolic diseases, such as diabetes and obesity, and also in the aging process (23). Previous studies have partially elucidated the role of AMPK in GO pathogenesis. In our earlier research, it was demonstrated that metformin and phenformin can exert anti-adipogenic effects and inhibit hyaluronan production via the AMPK/ERK pathway in an in vitro model of GO (24). Similarly, Xu et al (25) reported that metformin can regulate IL-1β-mediated inflammation and TGF-β-mediated fibrosis in GO orbital fibroblasts by activating the AMPK/mTOR pathway. Contrary to these findings, the present study revealed that compound C, an AMPK inhibitor, suppressed the pathological mechanisms of GO, including inflammation, fibrosis and adipogenesis. These findings suggested that compound C may exert its therapeutic effects on GO via pathways independent of AMPK.

Compound C has protective effects against inflammatory responses in various in vitro disease models. In murine macrophages, compound C has been shown to inhibit lipopolysaccharide-induced expression of inflammatory mediators, including IL-6 and MCP-1 (26). Furthermore, treatment with compound C has been shown to downregulate the production of pro-inflammatory mediators in rat microglial cells, such as IL-1β, IL-6, TNF-α, nitric oxide and reactive oxygen species, while preserving the release of IL-10 and TGF-β (27). In a mouse model of formalin-induced inflammatory nociception, compound C has been reported to reduce the levels of TNF-α, IL-1β, IFN-β and cyclooxygenase-2 in the serum, spinal cord and inflamed paw tissues (28). Consistent with these findings, the present study demonstrated that compound C suppressed IL-1β-induced expression of IL-6, IL-8, ICAM-1 and MCP-1. Considering that these mediators are the major pro-inflammatory cytokines implicated in the pathogenesis of GO, these findings suggested that compound C could exert a protective effect against the inflammatory processes associated with GO.

The direct effect of compound C on fibrotic response has been less frequently studied. Peyton et al (29) reported that compound C can inhibit vascular smooth muscle cell proliferation and migration in an AMPK-independent manner. In human adipose-derived stem cells, compound C has been shown to suppress the phosphorylation of SMAD1/5/8 and SMAD2 (30). Additionally, in a murine mesenchymal precursor cell line, compound C could inhibit both SMAD and non-SMAD pathways, including p38 and Akt (31). The results of the current study align with these findings, and demonstrated that compound C significantly blocked phosphorylation of signaling factors downstream of TGF-β, including SMAD1, 2 and 3. The SMAD pathway, which is primarily activated by TGF-β, serves a pivotal role in myofibroblast differentiation and collagen deposition, contributing to fibrotic tissue remodeling and adipogenesis in GO (32,33). These findings suggested that compound C may effectively mitigate fibrosis through its modulation of TGF-β signaling.

Adipogenesis and hyaluronan deposition are major contributors to volume expansion during GO pathogenesis. The effects of compound C on adipogenesis have previously been partially elucidated. In a previous study, in mouse preadipocyte 3T3-L1 cells, compound C was shown to inhibit differentiation, C/EBPα and PPARγ mRNA levels, and to attenuate late adipogenic markers, such as aP2 (34). The Hippo signaling pathway is a key determinant of organ size, which has a central role in mediating cellular mechanotransduction (35). In a mouse model of obesity, the inhibition of YAP/TAZ activity has been reported to significantly alleviate adipose tissue fibrosis, and Hippo pathway inactivation was revealed to be coupled with TGF-β-stimulation to promote fibrosis (36). Notably, our previous study demonstrated that YAP is strongly expressed in GO tissue, and YAP silencing and pharmacological YAP inhibition could suppress TGF-β-induced myofibroblast differentiation and collagen formation, thus indicating that the YAP/TAZ pathway serves as a critical mediator in the transition from inflammation to fibrosis in GO pathogenesis (37). Furthermore, in the present study, it was revealed that compound C abrogated YAP phosphorylation and silencing YAP/TAZ suppressed adipogenic differentiation in GO orbital fibroblasts. Collectively, these findings strongly suggested that compound C may inhibit inflammation and adipogenesis, thereby preventing volume expansion in GO by blocking the Hippo pathway.

Several studies have reported that compound C exerts AMPK-independent pharmacological effects on multiple cell types. For example, 3 and 5 µM compound C has been shown to inhibit macrophage chemotaxis in murine macrophages and to enhance insulin sensitivity in murine adipocytes, independent of AMPK suppression (26). Furthermore, 10 µM compound C can inhibit the proliferation and glioma formation of human glioma cells via multiple mechanisms, including activation of the calpain/cathepsin pathway and inhibition of the Akt and mTOR pathways, both of which are AMPK-independent (17,38). Additionally, compound C has been reported to reduce serum-induced activation of Akt and glycogen synthase kinase-3β, as well as platelet-derived growth factor (PDGF)-induced phosphorylation of Akt, PI3K, PLCγ1, ERK1/2 and PDGFR in human diploid fibroblasts (39). Compound C and AICAR have also been shown to inhibit T-cell cytokine production by suppressing the mTOR signaling pathway in mouse lymph node cells in an AMPK-independent manner (40).

In the present study, compound C suppressed Akt phosphorylation in both GO and healthy control orbital fibroblasts. In glioma cells, 5 and 10 µM compound C has been shown to inhibit Akt phosphorylation, while MAPK/ERK activity remained unaffected by compound C treatment (17). Additionally, in cancer cells, including human glioma cells, 10 µM compound C has been reported to induce protective autophagy through the Akt/mTOR pathway, which operates independently of AMPK (18). Xu et al (25) demonstrated that the AMPK activator metformin exerts anti-inflammatory and anti-fibrotic responses in GO orbital fibroblasts, whereas 100 nM compound C does not reverse the preventive effect of metformin on Akt phosphorylation. Consistent with these previous reports, the present findings revealed that compound C reversed the IL-1β-induced phosphorylation of Akt in both GO and healthy control cells, suggesting that compound C may directly inhibit the Akt/PI3K pathway in orbital fibroblasts. The Akt/PI3K pathway is associated with adipogenesis and hyaluronan production in GO, which is one of the key pathological intracellular signaling pathway in GO (41).

Activated Akt can enhance SMAD3 transcriptional activity, thereby leading to increased collagen I expression (42). Akt and YAP signaling are closely interconnected, with Akt regulating YAP activity through phosphorylation-mediated cytoplasmic retention or nuclear translocation. Furthermore, YAP can bind to SMAD1 and support SMAD1-dependent phosphorylation; in response to TGF-β signaling, TAZ and YAP have been shown to associate with heterometric SMAD2/3/4 complexes and dictate their intracellular localization (43,44). Thus, YAP interacts with SMAD proteins to enhance TGF-β-driven fibrotic responses, forming a positive feedback loop that amplifies fibrosis in GO. The present findings suggested that compound C modulated these pathways, thereby inhibiting TGF-β-induced myofibroblast differentiation and collagen formation. Although compound C has shown promising results in various disease models, it has not yet been approved for clinical use and is currently utilized exclusively in in vitro models. Moreover, the absence of a well-established in vivo model for GO hinders the evaluation of the therapeutic effects of compound C in a physiological setting. Furthermore, the present study focused on the therapeutic effects of compound C on the pathological phenotypes of GO, rather than examining the specific responses of different immune biomarkers or subtypes of autoreactive B and T cells; future studies exploring the roles of various immune components could provide additional insights. Further investigations using in vivo models and clinical settings are necessary to validate therapeutic potential of compound C.

In summary, the present study demonstrated that compound C may inhibit IL-1β- and TGF-β-induced inflammation, adipogenesis, fibrosis and hyaluronan secretion in primary cultured orbital fibroblasts through modulation of the Akt, Hippo/YAP and SMAD signaling pathways. While most recent drug development efforts have focused on AMPK activation, unilateral AMPK inactivation offers limited advantages for drug development. The present results suggested that the therapeutic effects of compound C were not solely mediated through AMPK inhibition, but also involved modulation of other key signaling pathways, such as Akt and YAP. These findings provide insights into the pathological mechanisms of GO and suggest that the therapeutic effects of compound C serves as a foundation for future research aimed at developing drugs that directly target the YAP or Akt pathways. Furthermore, given the critical role of these pathways in other fibrotic and inflammatory diseases, compound C may have therapeutic potential beyond GO, such as in inflammatory arthritis or inflammatory metabolic diseases, making it a promising candidate for further exploration in various pathological contexts.

Supplementary Material

Supporting Data

Acknowledgements

Not applicable.

Funding

This work was supported by a National Research Foundation of Korea grant funded by the Korean government (MSIT, Ministry of Science and ICT; grant no. RS-2023-00208570) and by a faculty research grant from Yonsei University College of Medicine (grant no. 4-2023-1454).

Availability of data and materials

The data generated in the present study may be requested from the corresponding author.

Authors' contributions

HYP, HSL and JSY conceptualized the study. SHC performed experiments, while HYP, JSY and JK confirmed the authenticity of all the raw data. HYP, JSY and JK analyzed and interpreted the data. JSY and JK supplied the laboratory resources, reagents and materials, and JSY procured the funding. HYP, SHC and HSL drafted the manuscript, and JSY and JK reviewed and edited it. All authors have read and approved the final version of the manuscript.

Ethics approval and consent to participate

The study protocol was reviewed and approved by the Institutional Review Board of Severance Hospital (approval no. 4-2024-0492) and conducted in accordance with the principles of The Declaration of Helsinki. Written informed consent was obtained from all participants.

Patient consent for publication

Not applicable.

Competing interests

The authors declare that they have no competing interests.

Glossary

Abbreviations

Abbreviations:

References