The Expression of Intercellular Adhesion Molecule-1 Induced by CD40-CD40L Ligand Signaling in Orbital Fibroblasts in Patients with Graves' Ophthalmopathy

Abstract

Purpose.: The aim of the present study was to examine the effect of CD40 ligand (CD40L) on intercellular adhesion molecule-1 (ICAM-1) production and involvement of mitogen-activated protein kinases (MAPKs) and nuclear factor-κB (NF-κB) signaling pathways by CD40-CD40L ligand in orbital fibroblasts (OFs) in patients with Graves' ophthalmopathy (GO).

Methods.: OFs were stimulated with soluble CD40L, and conditioned media and lysed cells were subjected to Western blot and real-time quantitative polymerase chain reaction analyses. Inhibitors specific to various signal transduction pathways were used to determine the signal pathways involved.

Results.: ICAM-1 protein and mRNA in OFs of GO groups were upregulated by CD40L compared with those in normal OFs. Treatment of OFs with CD40L increased ICAM-1 mRNA levels in a dose- and time-dependent manner. CD40L induced the phosphorylation of p38 MAPK, extracellular-regulated kinase1/2 (ERK1/2), c-Jun NH2-terminal kinase (JNK), and IκB and the activation of NF-κB. Inhibition of the ERK1/2, p38, JNK, and NF-κB pathways blocked CD40L-induced ICAM-1 secretion. Furthermore, the activation of NF-κB was significantly affected by ERK1/2 and JNK inhibitors, but not by p38. OF ICAM-1 expression was predominantly p38 MAPK and NF-κB dependent. ERK1/2 and JNK were implicated in the NF-κB pathway. Analyses of ICAM-1 mRNA synthesis revealed that CD40L-induced ICAM-1 expression was mediated by multiple factors.

Conclusions.: CD40L can potently induce ICAM-1 expression in OF cells through multiple signal pathways. The p38 MAPKs and NF-κB pathways may play an important permissive role in CD40L-induced ICAM-1 expression in OFs.

Graves' ophthalmopathy (GO) is a common autoimmune disease clinically characterized by exophthalmos, periorbital edema, eyelid retraction, extraocular muscle dysfunction, pain, and optic neuropathy. ^1,2 These symptoms are largely related to the pathologic processes within the orbit of the eye that increase the volume of retroocular tissue, in which the orbital fibroblasts (OFs) are the principal target of autoimmune attack and are the key to the pathophysiology of GO. ^3–5 Fibroblasts in GO patients can produce excess matrix glycosaminoglycans and inflammatory cytokines when stimulated by proinflammatory cytokines. ⁶ For example, high expression of intercellular adhesion molecule-1 (ICAM-1)/CD54 was reported in the retroocular connective tissues of GO patients. ^7,8 ICAM-1 is a cell surface adhesion receptor that binds to lymphocyte function associated antigen 1 (LFA-1) and promotes a variety of effector/target cell interactions when it is affected by inflammatory or immune processes. ⁹ It was proved that cytokines, such as interleukin-1-alpha (IL-1α), tumor necrosis factor-alpha (TNF-α), and interferon-gamma (IFN-γ), stimulate the expression of ICAM-1 in strongly enhanced surface expression of ICAM-1 in cultured Graves' OFs. ^10,11  

Recent studies found that orbital tissues of GO patients were infiltrated by T and B lymphocytes and macrophages. ^12,13 It was also reported that OFs expressed high levels of CD40, and activation by CD40 ligand (CD40L/CD154) induced many proinflammatory cytokines. ^14,15 The interaction between CD40L expressed on activated T cells and its receptor CD40 has been shown to play a role in the onset and maintenance of autoimmune inflammation. ^16–18 Therefore, we aimed to elucidate the signal transduction mechanism of ICAM-1 expression by CD40-CD40L in OFs in GO patients using OFs obtained during orbital decompression surgery from GO patients. We first detected ICAM-1 expression in OFs by real-time polymerase chain reaction (PCR) and ELISA and then examined the signal transduction mechanism for ICAM-1 expression through CD40-CD40L interaction by using real-time PCR and Western blot analysis. We found that nuclear factor-κB (NF-κB) and MAPKs were involved in ICAM-1 expression. 

Materials and Methods

Tissue Procurement

Biopsy specimens of deep orbital fat were obtained from patients undergoing orbital decompression surgery for GO or from normal orbital tissues of donors of corneal transplantations. Tissues were harvested in Shanghai Changzheng Hospital under the supervision of the institutional review board. The research protocol adhered to the tenets of the Declaration of Helsinki, and informed written consent was obtained from all patients. All tissue samples were placed in RPMI medium at 4°C. 

Fibroblast Culture

Fibroblasts were cultured from surgical tissue explants in RPMI containing 2-mercaptoethanol (50 μM), glutamine (2 mM), HEPES buffer (10 mM), nonessential amino acids (0.1 mM), pyruvate (1 mM), and gentamicin (0.05 mg/mL; Invitrogen, Carlsbad, CA). The explants were initially cultured in 40% fetal bovine serum (Gibco, Carlsbad, CA), which was gradually reduced to 10% over 7 days. The cells were morphologically consistent with a fibroblast phenotype. Cells were uniformly positive for immunohistochemical staining of vimentin and negative for staining of CD45, cytokeratin, and factor VIII (vimentin, CD45, cytokeratin, and factor VIII were purchased from Santa Cruz Biotechnology, Inc., Santa Cruz, CA). Fibroblast strains were stored in liquid nitrogen until needed and were used between passages 4 and 9. 

Reagents

Reagents used in the present study were CD40L and TNF-α (PeproTech Inc., Rocky Hill, NJ); anti–phospho-IκB-α rabbit polyclonal, anti–phospho-ERK1/2/ERK1/2, anti–phospho-p38 MAPK/p38 MAPK, and anti–phospho-JNK/JNK (New England Biolabs, Inc., Beverly, MA); anti–NF-κB, anti-vimentin, anti-CD45, anti-cytokeratin, and anti–factor VIII (Santa Cruz Biotechnology); SP600125, SB203580, PD98059 and PDTC (Sigma Chemical, St. Louis, MO); and DNA polymerase and dNTP (Life Technologies Gaithersburg, MD). The signal transduction inhibitors and their molecular targets used in this study are summarized in Table 1. 

Table 1.

 

Inhibitors Used to Study Pathways for OF ICAM-1 Expression

Inhibitor	Protein Target
PD98059	ERK1/2
SB203580	p38
SP600125	JNK
PDTC	NF-κB

Enzyme-Linked Immunosorbent Assay for ICAM-1

Cell surface ICAM-1 expression was determined by ELISA, as previously described. ¹⁹ OFs (1 × 10⁶ cells/well) grown in 96-well plates were stimulated with sCD40L or TNF-α for the indicated time periods and then fixed with 1% paraformaldehyde (10 minutes). All subsequent steps were performed at room temperature. Cells were washed with PBS (twice) and then incubated in a blocking solution composed of 1% BSA suspended in Tris-buffered saline (pH 7.4) containing 5% dry milk and 0.1% Tween-20 (TBST) for 1 hour. Fixed cells were probed with an anti–ICAM-1 rabbit anti–human polyclonal antibody (1:100 in 1% BSA TBST; Santa Cruz Biotechnology, Inc.) for 1 hour or with 1% BSA in TBST as a control. After six vigorous washes with TBST, cells were incubated with anti–rabbit secondary antibody linked to horseradish peroxidase (1:500 in TBST) for 30 minutes. Substrate (TMB Microwell Peroxidase Substrate System; Kirkegaard and Perry, Gaithersburg, MD) was applied to cells according to the manufacturer's instructions, and 1 M phosphoric acid was used to terminate the reaction. The optical density of each microtiter plate was read at 450 nm on a microplate reader (Biomed; Biomedical Life Systems, Vista, CA). Quadruplicate determinations in three to four different series of experiments were performed. Data are presented as detailed. 

Quantitative Real-Time PCR for ICAM-1

Total cellular RNA was isolated from nearly confluent cultures of OFs using reagent (TRIzol; Invitrogen, Carlsbad, CA) according to the manufacturer's protocol. Then total RNA was treated with RNase-free DNase (RQ1; Promega, Madison, WI) to remove genomic DNA contamination. First-strand cDNA was synthesized by reverse transcriptase from 5 μg total RNA with oligo-d (T) primers. Real-time PCR was performed by using a real-time PCR detection system (iCycler IQ; Bio-Rad, Hercules, CA) to measure the fluorescence produced by dye (SYBR Green I; Molecular Probes, Eugene, OR) that intercalates into PCR product. ²⁰ PCR reaction was performed in triplicate on each cDNA template along with triplicate reactions of the housekeeping gene β-actin. Negative control was obtained by performing PCR without cDNA. The synthetic oligonucleotide primers were 5′- CTCAGTCAGTGTGACCGCAGA-3′ (sense) and 5′- CCCTTCTGAGACCTCTGGCTTC-3′ (antisense) for human ICAM-1, and 5′-CAACTGGGACGACATGGAGAAA-3′ (sense) and 5′- GATAGCAACGTACATGGCTGGG-3′ (antisense) for human β-actin. The probes were 5′-ACCCAGCGGCTGACGTGTGCAGTAA-3′ for human ICAM-1 and 5′-TCTGGCACCACACCTTCTACAATGAGC-3′ for human β-actin. Thermal cycling conditions were 3 minutes at 95°C, followed by 40 cycles at 95°C for 15 seconds and 60°C for 20 seconds. All PCR reaction products were verified by melting curve analysis and agarose gel electrophoresis. ICAM-1 mRNA expression levels were quantified by calculating the average value of triplicate reactions, normalized by the average value of triplicate reactions for the housekeeping β-actin gene. ²⁰  

Western Blot Analysis of Human OFs

For preparation of whole-cell extracts, OF cells were lysed in lysis buffer, containing 50 mM HEPES (pH 7.4), 1% Triton X-100, 0.15 M sodium chloride, 10% glycerol, 1.5 mM magnesium chloride, 1 mM EDTA, and protease inhibitors by sonication and centrifugation. Protein concentrations were determined with a commercial kit (Sigma Chemical). Western blot analysis of the cellular extracts from OF cells was carried out according to the manufacturer's protocol. Briefly, the samples containing 20 to 50 μg protein were analyzed by sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) and then were electrotransferred to nitrocellulose membranes. For signal protein detection, samples were blocked with a solution of TBST at room temperature for 1 hour and then incubated at 4°C overnight with appropriate rabbit polyclonal antibodies. The membranes were washed with TBST, incubated with horseradish peroxidase-conjugated polyclonal anti–rabbit secondary antibodies for 1 hour at room temperature, washed with TBST three more times, and visualized using an enhanced chemiluminescence technique according to the manufacturer's instructions. 

Electrophoretic Mobility Shift Assay

Binding reaction mixtures (10 μL) containing 5 μg (4 μL) nuclear extract protein, 2 μg poly(dI-dC)·poly (dI-dC) (Sigma Chemical), and 40,000 cpm ³²P-labeled probe in binding buffer (4 mM HEPES, pH 7.9; 1 mM MgCl₂; 0.5 mMDTT; 2% glycerol; and 20 mM NaCl), were incubated for 30 minutes at room temperature. Protein-DNA complexes were separated on 5% nondenaturing polyacrylamide gels in 1× Tris-borate/EDTA electrophoresis buffer and autoradiographed overnight. The oligonucleotide used as a probe for electrophoretic mobility shift assay (EMSA) was a double-stranded DNA containing the NF-κB consensus sequence (5′-CTGTGCTCCGGGAATTTCCCTG-GCC-3′) labeled with ³²P-dATP (Pierce, Rockford, IL) using a DNA polymerase Klenow fragment. 

Statistical Analysis

All assays were obtained from three independent experiments. The data were expressed as mean ± SD and analyzed (SPSS12.0 for Windows; SPSS, Chicago, IL). P < 0.05 was considered statistically significant. 

Results

CD40L-Induced Human OF ICAM-1 mRNA Expression and Protein Production

To generate a profile of ICAM-1 that could be regulated by CD40L, the effect of CD40L on ICAM-1 protein production was examined by ELISA. Figures 1A and 1B show the kinetics and dose responses of CD40L-inducing effects on ICAM-1 surface expression. It was found that CD40L (10–100 ng/mL) significantly upregulated ICAM-1 surface expression in a dose-dependent manner. In addition, the maximal responses occurred at 12 hours of CD40L stimulation for ICAM-1, whereas the most effective dose of CD40L was 100 ng/mL. 

Figure 1.

 

Time- and dose-dependent effects of sCD40L on OFs in GO group ICAM-1 mRNA synthesis and protein production. (A) OF cells were incubated with sCD40L (100 ng/mL) for 0, 2, 4, 8, and 12 hours. (B) OF cells were stimulated in the media containing various concentrations of sCD40L (0, 5, 10, 50, 100 ng/mL) for 12 hours. Conditioned media were assayed for ICAM-1 using ELISA. (C) Real time-PCR analysis of time-dependent ICAM-1 mRNA expression that was induced by CD40L in OFs. OFs were incubated with CD40L (100 ng/mL) for 0, 2, 4, 8, and 12 hours. (D) OF cells were stimulated in the media containing various concentrations of sCD40L (0, 5, 10, 50, 100 ng/mL) for 2 hours. Total RNA was isolated and subjected to real-time PCR. Relative expression units were calculated by their cycle differences from control (unstimulated cells). The control level of ICAM-1 expression was arbitrarily assigned as a value of 1. Data shown are means from three independent experiments. Values represent mean ± SEM (n = 3). *P < 0.05; statistically significant difference compared with the control group without stimulation.

To determine whether stimulation of ICAM-1 secretion by sCD40L occurred at transcriptional levels, steady state ICAM-1 mRNA was quantified using quantitative real-time PCR. Human OF cells were challenged with 100 ng/mL CD40L. The growth medium was harvested at 2, 4, 8, and 12 hours after stimulation. Figure 1C shows that CD40L induced ICAM-1 mRNA expression; expression peaked at 2 hours after stimulation and was sustained for at least 8 hours. The induction of ICAM-1 became more evident at 2 hours, 4 hours, and 8 hours of stimulation and was 10.3-, 7.6-, and 4.7-fold higher than nonstimulated OF cells. This stimulation by CD40L was dose dependent (Fig. 1D). CD40L at 100 ng/mL resulted in approximately 5.2-fold higher ICAM-1 mRNA expression than that at 5 ng/mL concentration. 

Figure 2 shows that both ICAM-1 mRNA expression and protein production were all higher in the GO group than in the control group. Cells treated with TNF-α (100 ng/mL) were used as a positive control, and ICAM-1 mRNA expression and protein production were high compared with their levels in cells without CD40L treatment. 

Figure 2.

 

Effects of CD40L-induced ICAM-1 protein (A) and mRNA (B) expression in OFs. (A) OF cells of the control group and the GO group were, respectively, stimulated in the media containing sCD40L (100 ng/mL) or TNF-α (100 ng/mL) for 12 hours. Conditioned media were assayed for ICAM-1 using ELISA. (B) Samples were treated as described. Total mRNA was prepared for real-time PCR analysis of ICAM-1 expression in OFs of the control group and the GO group treated with sCD40L (100 ng/mL) or TNF-α (100 ng/mL) for 2 hours. Relative expression units were calculated by their cycle differences from control (unstimulated cells). The control level of ICAM-1 expression was arbitrarily assigned as a value of 1. Values represent mean ± SD (n = 3). Similar data were obtained from three independent experiments. *P < 0.05, statistically significant difference compared with the control group without stimulation. #P < 0.05, statistically significant difference compared with the GO group without stimulation. §P < 0.05, statistically significant difference compared with the control group without stimulation.

Multiple Signal Transduction Pathways Involved in CD40L-Induced ICAM-1 Expression

To assess the signal transduction pathways involved in CD40L-induced ICAM-1 gene expression in OF cells, various inhibitors specific to signal mediators were used (Table 1). As shown in Figure 3A, PD98059 (30 μM), SP600125 (30 μM), and SB203580 (30 μM) inhibited CD40L-induced ICAM-1 expression by 47.9% (P < 0.05), 57.2% (P < 0.05), and 62.3% (P < 0.05), respectively. The inhibition of ICAM-1 mRNA expression was dose-dependent on the concentration of PDTC (1–25 nM), an NF-κB inhibitor (Fig. 3B). PDTC at 25 nM abolished >90.0% (P < 0.05) of the CD40L-induced ICAM-1 mRNA production. Effects of various inhibitors on CD40L-induced ICAM-1 protein expression are shown in Figures 3C and 3D. The protein results were similar to the mRNA results. 

Figure 3.

 

Effects of inhibitor for MAPKs and NF-κB inhibitor (PDTC) on CD40L-induced ICAM-1 mRNA and protein expression in OFs. Cells were pretreated with SP600125 (30 μM), PD98059 (30 μM), SB203580 (30 μM) (A), and PDTC (1, 10, 25 nM) (B) for 30 minutes, followed by treatment with CD40L (100 ng/mL) for 2 hours. Cells were pretreated with SP600125 (30 μM), PD98059 (30 μM), SB203580 (30 μM) (C), and PDTC (1, 10, 25 nM) (D) for 30 minutes, followed by treatment with CD40L (100 ng/mL) for 12 hours. The procedure followed for this experiment was the same as that described in the legend to Figure 2. *P < 0.05, statistically significant difference compared with the control group induced by CD40L.

Because the stimulated ICAM-1 expression was highly sensitive to PD98059 (30 μM), SP600125 (30 μM), SB203580 (30 μM), and PDTC (25 nM), we further investigated the role of MAPKs (p38, ERK1/2, JNK) and NF-κB signaling in human OF ICAM-1 mRNA expression. Activation of MAPKs proteins (phosphorylation) was assessed by Western blot analysis using antibodies against the active (phospho) forms of MAPKs proteins. As shown in Figure 4, CD40L (100 ng/mL) induced a time-dependent increase in phosphorylation of three MAPKs proteins up to 90 minutes 

Figure 4.

 

The time course of phosphorylations of ERK1/2, p38, JNK, and IκB in response to sCD40L. Lysates from OF cells stimulated with sCD40L (100 ng/mL) for the indicated time points or TNF-α (100 ng/mL) for 30 minutes were analyzed by anti–p-p38/p38 (A), anti–p-ERK/ERK (C), or anti–p-JNK/JNK (E) and anti–p-IκB /IκB (G) Western blot. (B, D, F, H) Relative values of the levels of phosphorylated ERK1/2, p38, JNK, or IκB normalized compared with nonphosphorylated ERK1/2, p38, JNK, or IκB, respectively. Results are the mean ± SD of three separate experiments. *P < 0.05, statistically significant difference compared with the control group without stimulation.

The results revealed that all three MAPK subfamilies examined were rapidly and strongly activated by CD40L treatment in a time-dependent pattern. In particular, p38 activation displayed a rapid onset within 5 minutes of treatment (P < 0.05), which was followed by progressively increasing activation and then reached a peak at 90 minutes (P < 0.05) of treatment (Fig. 4A). Activation of ERK1/2 by CD40L was intense and reached maximum values within 15 minutes of treatment (P < 0.05), then decreased and reached control levels at 90 minutes (P > 0.05; Fig. 4B). Activation of JNK was also observed at 5 minutes of CD40L treatment (P < 0.05) and reached the maximal level at 30 minutes (P < 0.05), then decreased progressively (Fig. 4C). 

These results indicate that CD40L induced a differential activation of the three well-established MAPK subfamilies in relation to both stimulation time and dosage of stimulus. The bottom panels in Figure 4 show that there was no change in the total cellular pools of all MAPKs examined, providing a control for equal protein loading under these conditions. 

Cells treated with TNF-α (100 ng/mL) for 30 minutes were used as a positive control, whereas phosphorylation of the three MAPKs was strong compared with those without CD40L treatment. Even at the same time point, phosphorylation of TNF-α–induced ERK1/2 and JNK were stronger than the phosphorylation induced by CD40L. 

We next studied the activation pattern of NF-κB by CD40L. Cells were treated with 100 ng/mL CD40L for various time periods. Nuclear extracts were prepared and subjected to EMSA to test DNA binding activity of the nuclear factor. CD40L at 100 ng/mL induced a moderate activation of NF-κB, which increased in a time-dependent manner, displaying an onset within 15 minutes (P < 0.05) and reached the maximal value within 90 minutes (P < 0.05) after treatment (Fig. 5). Nuclear extracts prepared from cells treated with TNF-α (100 ng/mL) for 30 minutes were used as positive controls. 

Figure 5.

 

Effect of CD40L on NF-κB binding activity in OFs. (A) Nuclear extracts were prepared from OFs treated with 100 ng/mL CD40L for increasing periods and subjected to EMSA. Extracts from cells treated for 30 minutes with 100 ng/mL TNF-α were used as positive controls. These were quantitated by densitometric analysis and are expressed as percentages of the response to stimulant alone (B). Results are the mean ± SD of three separate experiments. *P < 0.05, statistically significant difference compared with the control group without stimulation.

CD40L-Induced Phosphorylation but Not Degradation of IκB-α

Phosphorylation and subsequent degradation of IκB is required for NF-κB activation. Therefore, IκB phosphorylation was examined using a specific antibody against the phosphorylated (Ser32/36) form of IκB-α. We found that IκB was phosphorylated during CD40L treatment in a time-dependent manner. This phosphorylation became evident within 15 minutes of treatment (P < 0.05) and reached maximal values within 90 minutes (P < 0.05; Fig. 4D). The phosphorylation pattern of IκB complied with CD40L-induced DNA binding activity of NF-κB. However, when total IκB levels were examined using a specific antibody, there was no apparent decrease at any designated time point (Fig. 4D, bottom panel), suggesting that there was no significant degradation of IκB during CD40L treatment. 

The Blockage of Selective MAPK Inhibitors on NF-κB Activation and IκB-α Phosphorylation

OFs were pretreated for 30 minutes with different inhibitors and then examined for NF-κB activation in response to CD40L (100 ng/mL) for 90 minutes. The inhibitor of ERK1/2, PD98059, and the inhibitor of JNK, SP600125, significantly decreased CD40L-induced binding activity of NF-κB to DNA (53.9% reduction with PD98059 and 56.3% reduction with SP600125; P < 0.05; Fig. 6B). Interestingly, simultaneous use of the inhibitor of p38, SB203580, did not result in a significant potential effect (18.9% reduction with SB203580; P > 0.05; Fig. 6B). 

Figure 6.

 

Inhibitory effects of selective inhibitors of MAPK on sCD40L-induced phosphorylation of IκB and NF-κB activation in OF cells. (A) Cells were preincubated for 30 minutes with SP600125 (30 μM), PD98059 (30 μM), and SB203580 (30 μM) before treatment with sCD40L (100 ng/mL) for an additional 90 minutes. Cell lysates were subjected to anti–p-IκB/ IκB Western blot analysis. (B) Relative values of the levels of phosphorylated IκB normalized compared with nonphosphorylated IκB. (C) Cells were preincubated for 30 minutes with SP600125 (30 μM), PD98059 (30 μM), and SB203580 (30μM) before they were treated with sCD40L (100 ng/mL) for 90 minutes. Nuclear extracts were analyzed by EMSA. These were quantitated by densitometric analysis and are expressed as percentages of the response to stimulation alone (D). Results are the mean ± SD of three separate experiments. *P < 0.05, significant inhibition by MAPKs inhibitor compared with sCD40L alone.

Furthermore, we examined the effect of the three MAPK inhibitors on IκB-α phosphorylation. The inhibitor of ERK1/2, PD98059, and the inhibitor of JNK, SP600125, significantly inhibited CD40L-induced phosphorylation of IκB-α by approximately 36.5% and 41.2%, respectively (Fig. 6; P < 0.05), whereas the inhibitor of p38, SB203580, did not result in a significant potential effect (12.6%; P > 0.05). These findings suggest that ERK1/2 and JNK MAPK activation increased the phosphorylation of IκB-α and thus contributed to NF-κB activation in CD40L-treated cells. The p38 MAPKs pathway, however, did not appear to regulate either the release of IκB from NF-κB or the nuclear translocation of NF-κB during CD40L stimulation. 

Discussion

ICAM-1, or CD54, is an inducible cell adhesion glycoprotein of the immunoglobulin supergene family existing on many cell types, and it is associated with many inflammatory diseases and autoimmune diseases. ^1,21 ICAM-1 interacts with LFA-1 on the surfaces of leukocytes, which is important for their transendothelial migration to inflammation sites and their function as costimulatory molecules for T-cell activation. ^22,23 ICAM-1 expression depends on the concentrations of proinflammatory and anti-inflammatory mediators and on the availability of specific receptor-mediated signal transduction pathways and their nuclear transcription factor targets on the ICAM-1 promoter. ⁹  

Adhesion molecules play an important role in the initiation and maintenance of the inflammatory immune process. ¹⁶ Cellular activation and local expression of adhesion molecules lead to leukocyte recruitment, migration to inflammatory sites, and targeting in the extravascular space. ² OFs in retroocular connective tissues of GO patients are strongly positive for ICAM-1. Soluble ICAM-1 concentrations in GO patients can reflect the degree of inflammatory activity. ^24–26 On an in vitro model of thyroid-associated ophthalmopathy, TNF-α, IFN-γ, and IL-1 significantly stimulated ICAM-1 expression in cultured OFs. ^10,11,27  

Previous experiments proved that OFs were activated through triggering CD40 with CD40L. ^4,14 Experiments with human thyroid fibroblasts proved that when CD40 was engaged with CD40L, NF-κB binding activity was upregulated and induction of the proinflammatory and chemoattractant cytokines IL-6 and IL-8 were induced. ²⁸ Recently, the signal transduction pathways for CD40L-induced ICAM-1 mRNA expression have been reported in many cell types in many autoimmune diseases, ^29,30 but little is known about these pathways in OFs during stimulation by CD40L. 

In the present study, we found that ICAM-1 expression in cultured OFs of GO patients was enhanced by CD40L stimulation and was increased in a dose- and time-dependent manner. It was reported that sCD40L was elevated in the sera of GO patients, ³¹ and CD40L expression in the infiltrating lymphocytes in GO patients was observed, ^14,15,32 suggesting that stimulation with CD40L in vivo was probably the cause of high ICAM-1 expression in GO patients. The addition of sCD40L stimulation to controls also significantly increased ICAM-1 expression, indicating that continuous stimulation of OF cells with costimulatory molecules such as CD40L contributed to GO pathogenesis. Lymphocyte recruitment is a key pathogenic event in inflammatory diseases. As part of the connective tissue, fibroblasts are the target of lymphocyte adhesion and infiltration. Although GO is an autoimmune disease, inflammation is an inevitable step for immunopathogenesis. ICAM-1 is a key molecule for lymphocyte infiltration, and infiltrating lymphocytes with higher levels of CD40L upregulate ICAM-1. Therefore, these cascade reactions aggravate the progressive pathogenesis of GO. 

In the present study we also investigated the signal pathways (NF-κB and MAPK) leading to enhanced ICAM-1 gene expression by CD40L. NF-κB is an important transcription factor regulating the expression of a large number of genes involved in immune and inflammatory processes. ³³ Previous experiments proved that NF-κB plays an important role in mediating the effects of leukoregulin on PGHS-2 expression in OFs. MAPK pathways are also important in cellular signal transduction. These MAPKs (including extracellular signal-regulated kinases1/2 [ERK1/2], c-Jun N-terminal kinases [JNK], and MAPK p38) regulate diverse cell activities such as gene expression, proliferation, differentiation, and apoptosis. ³⁴ Previous experiments proved that IL-1β activated both p38 and ERK 1/2 components of the MAPK pathways and induced IL-6 expression in human OFs. ³⁵  

Our results revealed that the stimulation of OFs with CD40L induced the rapid phosphorylation, but not degradation, of IκB-α and the nuclear translocation of NF-κB, consistent with previous findings that OFs of Graves' disease can be activated through triggering CD40 with CD40L by NF-κB and induction of the proinflammatory and chemoattractant cytokines. ¹⁴ In addition, the stimulation of OFs with CD40L activated all three types of MAPKs, which were all the ICAM-1 signaling. Using specific inhibitors of the ERK and JNK pathways had an effect on CD40L-induced DNA binding activity of NF-κB and on Ser32/36 phosphorylation of IκB (Fig. 6). The inhibition of p38, SB203580, did not alter the CD40L-induced phosphorylation of IκB-α and the nuclear translocation of NF-κB in OFs. These results suggest that ERK1/2 and JNK, but not p38MAPKs, are involved in the processes leading to the dissociation of NF-κB dimer from IκB and to nuclear translocation of the transcription factor. It is, therefore, more likely that some other pathways (e.g., transcription factor, NF-κB) for p38 MAPK activation led to ICAM-1 gene regulation induced by CD40L. At least, NF-κB may not be a direct target for the CD40L-induced p38 MAPK pathway. 

Different cells have different signal pathways regulating the same gene, though the stimulation is same. For example, in cultured human renal proximal tubule epithelial cells, p38 MAPK, but not ERK1/2 MAPK, was involved in CD40/CD154-induced ICAM-1 production. ³⁶ In human B lymphocytes, the p38 MAPK and NF-κB pathways were required in CD40-induced ICAM-1 expression, but CD40-mediated NF-κB binding was not affected by SB203580, the inhibitor of p38. ³⁷ It enhanced the complexity of the cell type–specific and stimulus-specific regulation of the ICAM-1 gene. Although other fibroblasts (e.g., dermal fibroblasts) were reported in which CD40-CD154 signaling regulated ICAM-1 gene expression, the detailed pathways, such as MAPKs, involving NF-κB with the regulation were poorly reported. ³⁸ Whether other fibroblasts apply with these same pathways is worthy of research. In addition to NF-κB, a number of nuclear transcription factors, including Ets-1, C/EBP, AP-1, AP-2, AP-3, and Sp-1, have been shown to activate the ICAM-1 gene. The present study elucidates that NF-κB is the key nuclear transcription factor in GO fibroblasts; it mediates CD40-CD154 signal–regulated ICAM-1 gene expression and is mediated by upstream ERK1/2 and JNK, not by p38. p38 might be mediated by other nuclear transcription factors that regulate CD40-CD40L/ICAM-1; the identity of these factors and whether they are mediated by ERK1/2 and JNK remain to be studied. The complexity of CD40-CD40L/ICAM-1 might be determined by the unique property of fibroblasts. 

In addition to the pathway we studied, it has also been reported that ICAM-1 expression was induced by stimulation with TNF-α, which is consistent with previous studies. ^10,39,40 Our results showed that TNF-α was as potent as sCD40L in inducing ICAM-1 expression. However, these results were obtained under in vitro conditions. It would be difficult to say whether TNF-α or sCD40L is more important in the maintenance of inflammation based on this study alone. However, the fact that NF-κB activity is increased by TNF-α stimulation has been documented. ¹⁴ TNF-α is produced primarily by macrophages. It, together with other proinflammatory factors, initiates the inflammatory cascade response and maintains the response. It can induce the production of adhesion molecules, chemokines in fibroblasts, and vascular endothelial cells, recruiting various inflammatory cells to local tissues and starting inflammatory reactions. TNF-α plays an initiatory role in the development and progression of inflammatory responses; it can also induce the production of adhesion molecules in fibroblasts at early stages of GO pathogenesis. The role of TNF-α in the initiation of inflammatory responses is unspecific. CD40L is cell specific; it is specifically expressed on the surfaces of activated CD4⁺ T cells and interacts with CD40 in OFs, increasing the expression of ICAM-1, B7, and major histocompatibility complex class II, and enhancing antigen-presenting ability. Meanwhile, CD40/CD40L is also a costimulatory molecule regulating antiapoptosis and apoptosis-inhibition genes. It can be said that, in our study, the stimulation of OFs by TNF-α is an early event of inflammation, inducing ICAM-1 expression and recruiting various inflammatory cells, with inflammation the theme of this phase. Stimulation of OFs by CD40L is an event during the progression of the specific inflammatory immune process, inducing ICAM-1 expression and enhancing the antigen-presenting abilities of fibroblasts and T cells, with immune response as the theme of this phase. 

In this study, we examined the role of ERK1/2, JNK, p38MAPK, and NF-κB in regulating CD40L-induced expression of ICAM-1 in OFs of GO. Our results indicate that MAPK (ERK1/2, p38, JNK) and NF-κB pathways were all involved in ICAM-1 expression induced by CD40L. CD40L-induced NF-κB binding was modulated by ERK1/2 and JNK but not by p38. p38 MAPK may be involved in processes leading to other routes of nuclear translocation (such as NF-κB) that regulate ICAM-1 gene expression. The upregulation of ICAM-1 is associated with the pathogenesis of GO. It is hoped that suppressing this pathway will improve the pathologic condition of GO patients. 

Footnotes

 Supported by National Natural Science Foundation of China Grant 30672274.

Footnotes

 Disclosure: L.-Q. Zhao, None; R.-L. Wei, None; J.-W. Cheng, None; J.-P. Cai, None; Y. Li, None