Human orbital fibroblasts are activated through CD40 to induce proinflammatory cytokine production

Gregory D. Sempowski, Julia Rozenblit, Terry J. Smith, Richard P. Phipps


CD40 is an important signaling and activation antigen found on certain bone marrow-derived cells. Recently, CD40 has also been shown to be expressed by nonhematopoietic cells, including certain human fibroblasts, but not others. Little is known about the function of CD40 on fibroblasts. The current study investigates the hypothesis that CD40 is expressed on orbital fibroblasts and represents a pathway for interaction between these fibroblasts and CD40 ligand-expressing cells, such as T lymphocytes and mast cells. We report here that orbital connective tissue fibroblasts, obtained from normal donors and from patients with severe thyroid-associated ophthalmopathy (TAO), express functional CD40. CD40 is upregulated ∼10-fold by interferon-γ (500 U/ml) treatment for 72 h. These fibroblasts become activated through triggering of CD40 with CD40 ligand (CD40L). This is evidenced by nuclear translocation of nuclear factor-κB and induction of the proinflammatory and chemoattractant cytokines interleukin-6 and interleukin-8, respectively. These data support the concept that cognate interactions between orbital fibroblasts and infiltrating T lymphocytes, via the CD40-CD40L pathway, may promote the tissue remodeling observed in TAO and other inflammatory diseases of the orbit. Disruption of the CD40-CD40L interaction may represent a therapeutic intervention to reduce the inflammatory components of TAO, which remains a vexing clinical problem.

  • fibroblasts
  • thyroid-associated ophthalmopathy
  • cellular activation
  • interleukin-6
  • interleukin-8

the connective tissue of the human orbit can manifest dramatic inflammation in thyroid-associated ophthalmopathy (TAO) (30). TAO is synonymous with Graves’ ophthalmopathy. In this disease process, tissues become infiltrated with activated T lymphocytes and mast cells and accumulate excessive amounts of the nonsulfated glycosaminoglycan hyaluronan (7). Hyaluronan is synthesized by many cell types. However, fibroblasts are a major source of this and other glycosaminoglycans (30). It is believed that orbital fibroblasts have a primary role in the pathogenesis of TAO. During the tissue remodeling associated with TAO, infiltrating T lymphocytes are thought to initiate, through unknown signaling pathways, biosynthesis of inflammatory mediators and hyaluronan, two cardinal features of TAO (22). It would therefore appear essential that conduits for molecular communication between orbital fibroblasts and infiltrating T lymphocytes be fully characterized if further insights are to be gained into the mechanism underlying the disease. Moreover, elucidation of the pathways utilized for lymphocyte-fibroblast cross talk could allow the formulation of specific therapeutic strategies designed to interrupt inflammation.

One candidate pathway for communication between orbital fibroblasts and T lymphocytes is the CD40-CD40 ligand (CD40L) costimulatory pathway. This system conveys signals for activation and differentiation between hematopoietic cells. It has been implicated recently in T lymphocyte costimulation (4, 28). Triggering through surface CD40, a 50-kDa member of the tumor necrosis factor-α (TNF-α) receptor gene superfamily, is a powerful stimulus for B lymphocyte activation, proliferation, immunoglobulin (Ig) production, and isotype class switching (6). CD40 is also expressed by non-B lymphocyte hematopoietic cells, including interdigitating dendritic cells, Langerhans cells, dendritic cells, and monocytes (4). CD40L, a member of the TNF-α cytokine gene superfamily, is the counterreceptor for CD40 and is primarily expressed on the surface of activated T lymphocytes and mast cells (34). In addition to being an important signaling mechanism for B lymphocytes, the CD40-CD40L intercellular ligand interaction can provide a necessary “second signal” to help drive cytokine production and clonal expansion of naive T lymphocytes (26).

Observations in patients with X-linked hyper-IgM syndrome provide evidence for the importance of the CD40-CD40L pathway in immune responses. These patients possess mutations in the gene encoding CD40L, resulting in the disruption of CD40-CD40L signaling. The loss of T lymphocyte “help” renders B cells functionally deficient, with abnormal Ig class switching and a profound inability to respond to bacterial antigens (2). The CD40-CD40L pathway has also been shown to play a major role in autoimmune diseases that have significant B lymphocyte involvement. For example, treatment with anti-CD40L antibodies are effective in preventing the onset of disease in murine models of collagen-induced arthritis, experimental autoimmune encephalomyelitis, and lupus nephritis (8, 24).

CD40 expression is not limited to bone marrow-derived cells. It has been demonstrated to be expressed and functional on epithelial cells, such as human epidermal keratinocytes (15), thymic epithelium (13), and activated vascular endothelial cells (21). Very relevant to the current study is the observation of CD40 expression on nontransformed human fibroblasts derived from lung, foreskin, periodontal, and synovial tissue (12). CD40 mRNA was detected by reverse transcription-polymerase chain reaction (RT-PCR), and the protein was detected on cultured fibroblasts using flow cytometry. In tissue sections, immunostaining suggested fibroblast display of CD40. In contrast, some dermal fibroblasts failed to express CD40 (12).

The current studies were conducted to determine whether CD40 is expressed by orbital fibroblasts derived from normal tissue and tissue obtained from donors with TAO. We demonstrate here that CD40 is expressed by orbital fibroblasts from either source and that these fibroblasts can be activated through CD40 to mobilize nuclear factor-κB (NF-κB) and to express proinflammatory cytokines. We hypothesize that the CD40-CD40L coreceptor system, described for B-T lymphocyte communication, is an important pathway through which communication between orbital fibroblasts and T lymphocytes and mast cells can occur. This pathway may be relevant to the pathogenesis of inflammatory diseases in the orbit, such as TAO. Moreover, the demonstration of functional CD40 on fibroblasts represents a potentially important target for therapy design.


Cell culture.

Normal and TAO orbital primary fibroblasts were isolated from surgical explants as described in an earlier publication (33). TAO fibroblasts were initiated from surgical explants from individuals undergoing decompressive surgery for severe Graves’ ophthalmopathy. Normal orbital strains were derived from individuals undergoing eye surgery for conditions not affecting the soft tissues of the orbit. Procurement of these tissues was approved by the Institutional Review Board of Albany Medical College. The fibroblasts were maintained in minimal essential medium with Eagle’s salts (MEM; Life Technologies, Gaithersburg, MD) supplemented with 10% fetal bovine serum (FBS; Hyclone, Logan, UT) and gentamicin (50 μg/ml; Life Technologies). These cells are morphologically consistent with a fibroblast phenotype and express vimentin and collagen, two markers of fibroblasts. They do not, however, express CD45, factor VIII, or smooth muscle actin. Fibroblasts were passaged every 10–14 days and reseeded at 3 × 105cells/75-cm2 flask (Costar, Cambridge, MA). Adherent monolayers of fibroblasts were dissociated with a 1:1 solution of sodium EDTA (0.1% wt/vol; Sigma Chemical) and trypsin (0.05% wt/vol; Worthington Biochemical, Freehold, NJ). All fibroblasts were maintained in a humidified 7% CO2 incubator. All experiments used cells between passages 2 and13.


Fibroblasts were untreated or treated with recombinant human interferon-γ (IFN-γ, 500 U/ml; Genzyme, Cambridge, MA) for 48 h and then scraped with a rubber policeman for extraction of total RNA with Tri-Reagent (1 ml/106 cells; Molecular Research Center, Cincinnati, OH), following the manufacturer’s protocol. RNA was solubilized in nuclease-free water by heating to 55°C for 10 min. The concentration of total RNA was quantitated by spectrophotometry, and 5 μg of RNA were reverse transcribed using Moloney murine leukemia virus reverse transcriptase (200 U/reaction; Life Technologies) and an oligo(dT) primer (Pharmacia, Piscataway, NJ), as previously described (12). Reverse transcriptase was withheld from replicate samples, which served as negative controls. PCR reactions for human CD40 and β-actin consisted of cDNA obtained above (2 μl), reaction buffer (Boehringer Mannheim Biochemicals), deoxynucleotides (1 μM each), specific primers (1 μM each), and Taq DNA polymerase (2.5 U; Boehringer Mannheim Biochemicals) in a total volume of 50 μl. The human CD40 primer sequences were 3′-CGT ACA GTG CCA GCC TTC TTC and 5′-ATG GTT CGT CTG CCT CTG CAG, yielding a 330-base pair (bp) product; the human β-actin primer sequences were 3′-CTC CTT AAT GTC ACG CAC GAT TTC and 5′-GTG GGG CGC CCC AGG CAC CA, which generated a 539-bp product. Samples underwent 30 cycles of amplification in a PTC-200 DNA Engine thermal cycler (M. J. Research, Watertown, MA), and each cycle included denaturation at 94°C for 30 s, annealing at 63°C for 30 s, and primer extension at 72°C for 60 s. The products were electrophoresed on 2% agarose gels and visualized with ethidium bromide staining. A 100-bp ladder (Life Technologies) was used for size determination.

Flow cytometry analysis.

Fibroblasts were prepared for flow cytometry by washing a suspension of cells in phosphate-buffered saline (PBS) with 0.1% sodium azide and 1.0% bovine serum albumin (BSA). Cells were incubated with anti-human CD40 monoclonal antibody (G28-5, 100 μl of hybridoma supernatant; American Type Culture Collection) or anti-HLA-DR (L243, 100 μl of hybridoma supernatant; American Type Culture Collection) for 30 min on ice. After the cells were washed, fluorescein isothiocyanate-conjugated goat anti-mouse Ig (1:50 dilution; Cappel Research Products, Durham, NC) was added for 30 min on ice. Samples stained with only the secondary antibody served as negative controls. Once washed, the cells were resuspended in PBS with 0.1% sodium azide and 1.0% BSA and analyzed on a flow cytometer (Elite, Coulter, Hialeah, FL). We have demonstrated that trypsin treatment, under the conditions used to disrupt fibroblast monolayers, fails to alter the detection of surface CD40 or HLA-DR (data not shown). Viable cells were gated on the basis of forward light scatter, and the data were analyzed with the Cytologo software program (Coulter). All experiments were repeated a minimum of three times, and representative results are presented.

Electrophoretic mobility shift assays.

Orbital fibroblasts were allowed to proliferate to confluence in 100-mm tissue culture dishes covered with MEM supplemented with 10% FBS. The fibroblasts were pretreated with IFN-γ (500 U/ml, Genzyme) for 72 h. The cultures were then washed extensively and incubated overnight with MEM containing 1% FBS. The cells were then stimulated for 2 h at 37°C with medium alone, control insect cell membranes containing glutathione S-transferase (GST), or insect cell membranes containing human CD40L. Membranes were generously provided by Dr. Marilyn Kehry (Boehringer Ingelheim, Ridgefield, CT) and were prepared as described previously (19). Recombinant human TNF-α (50 U/ml; Genzyme) was used to trigger the cells as a positive control for NF-κB mobilization. The cells were harvested, nuclear extracts were prepared as described by Andrews and Faller (1), and protein concentrations were normalized by the bicinchoninic acid protein assay (Pierce, Rockford, IL). To assess NF-κB binding activity, 25 μg of nuclear extract were incubated with32P-labeled double-stranded oligonucleotide probe representing the consensus sequence for the NF-κB binding sites following the supplier’s instructions (Promega, Madison, WI). Samples were electrophoresed on 4% polyacrylamide gels, and the DNA-protein complexes were visualized by autoradiography.

Cytokine production.

Orbital fibroblasts were seeded into 96-well tissue culture plates (5,000 cells/well) and, once established, were untreated or treated with IFN-γ (500 U/ml) for 72 h in MEM containing 10% FBS. The fibroblast monolayers were then washed extensively with fresh culture medium at 37°C and incubated in triplicate for 72 h with medium alone, control insect cell membranes containing GST, or insect cell membranes containing human CD40L. Culture medium was harvested, and the concentration of IL-6 and IL-8 was determined by specific enzyme-linked immunosorbent assay (ELISA). Media samples were appropriately diluted with sample buffer so as to fall within the linear sensitivity range of the ELISAs.

For detection of human IL-6, Immulon 3 flat-bottomed 96-well plates (Dynatech, Chantilly, VA) were coated overnight at 4°C with 2 μg/ml anti-human IL-6 (capture antibody, Pharmingen, San Diego, CA) diluted in 0.1 M sodium bicarbonate (pH 8.2). The plates were then washed three times with PBS-Tween [1× PBS, 0.05% Tween 20 (vol/vol), and 0.1% sodium azide]. Plates were blocked with ELISA buffer (1× PBS, 5% BSA, 5% calf serum, and 0.1% sodium azide) for 2 h at room temperature. The plates were washed three times with PBS-Tween. Media samples and IL-6 standard (500 ng/ml to 0 pg/ml; Pharmingen), diluted in ELISA buffer, were then incubated on the plates for 2 h at 37°C. The plates were washed three times with PBS-Tween. Detection antibody (2 μg/ml of biotinylated anti-human IL-6; Pharmingen) was diluted in ELISA buffer and then incubated on the plates for 1 h at room temperature. Plates were washed three times and incubated at 37°C for 1 h with streptavidin-alkaline phosphatase (1:500 dilution in ELISA buffer; Southern Biotechnologies, Birmingham, AL). After the plates were washed three times, they were incubated with 1 mg/ml of p-nitrophenyl phosphate diluted in DEA buffer [2% (vol/vol) diethanolamine, 10 μM magnesium chloride, and 30 μM sodium azide, pH 9.8]. The reaction was stopped with 3 N NaOH. The optical density of each well was read at 405 nm with a Dynatech plate reader.

The assay for human IL-8 resembles that for IL-6 with the following exceptions: the capture antibody (anti-human IL-8; Endogen, Cambridge, MA) was used at a concentration of 1 μg/ml. The recombinant IL-8 standard (Biosource, Camarillo, CA) was used in a dilution series from 600 to 0 pg/ml. The biotinylated detection antibody (anti-human IL-8; Endogen) was used at a final concentration of 0.5 μg/ml. The detection antibody was directly added to the wells 15 min after addition of samples and IL-8 standards. The plates were then incubated for 1 h at room temperature.

Standard curves were plotted on the basis of a dilution series of the recombinant cytokine standards and used to determine the concentrations of IL-6 and IL-8 in the experimental samples. All experiments were repeated a minimum of three times, and results from representative experiments are presented.


CD40 mRNA expression by normal and TAO orbital fibroblasts.

Primary human orbital fibroblasts were isolated to examine the expression and function of CD40. Previous reports have demonstrated that fibroblasts from human lung tissue constitutively express CD40 mRNA and protein. In addition, it was shown that the proinflammatory cytokine IFN-γ induces hyperexpression of this surface antigen (28). RT-PCR analysis was performed on total RNA extracted from normal and TAO fibroblasts treated without or with IFN-γ (500 U/ml) for 48 h. Equivalent amounts of RNA were reverse transcribed, and the resulting cDNA was then amplified for 30 cycles with primers specific for CD40 and the housekeeping gene β-actin. Samples prepared without reverse transcriptase failed to yield any CD40 or β-actin product (data not shown). A representative ethidium bromide-stained gel showing the amplification products from normal orbital fibroblast RNA (±IFN-γ) is shown in Fig.1 A. The amplification products from TAO orbital fibroblast RNA (±IFN-γ) are depicted in Fig.1 B. It is clearly demonstrated that normal and TAO orbital fibroblasts express CD40 mRNA constitutively, and the intensity of the bands suggests that the amount of steady-state CD40 mRNA is elevated in the IFN-γ-treated samples, using β-actin expression as a control. From the results shown and multiple repetitions of this experiment, it does not appear that there is a significant difference in the level of CD40 mRNA expression between normal and TAO orbital fibroblasts, under control or IFN-γ-treated conditions.

Fig. 1.

Expression of human CD40 and β-actin mRNA in normal (A) and thyroid-associated ophthalmopathy (TAO; B) orbital fibroblasts. Fibroblasts were treated without or with interferon-γ (IFN-γ; 500 U/ml) for 48 h. Total cellular RNA was isolated, and equivalent amounts were reverse transcribed into cDNA. cDNA was amplified with specific primer sets for human CD40 and human β-actin. Polymerase chain reaction products representing CD40 (330 bp) and β-actin (539 bp) were resolved on ethidium bromide-stained 2% agarose gels.

Surface expression of CD40 and HLA-DR by normal and TAO orbital fibroblasts.

The next step in the analysis of CD40 expression by normal and TAO orbital fibroblasts was to verify the findings regarding expression of CD40 mRNA by flow cytometric detection of basal and IFN-γ-induced surface CD40 protein expression. Fibroblasts were cultured in the presence or absence of IFN-γ for 72 h and then stained with a monoclonal antibody specific for CD40. Histograms showing background staining (control) and CD40 expression by untreated and IFN-γ-treated normal orbital fibroblasts are presented in Fig.2 A. The log fluorescence intensity of staining is plotted vs. cell number for each sample analyzed. CD40 is constitutively expressed on the surface of the fibroblasts compared with background staining profiles. Stimulation with IFN-γ results in a dramatic 10-fold increase in the expression of CD40 on normal orbital fibroblasts. Figure2 B shows the flow cytometry histograms obtained for CD40 staining of TAO orbital fibroblasts treated similarly. TAO fibroblasts also display detectable basal levels of CD40 comparable to those of normal orbital fibroblasts and are also stimulated by IFN-γ to upregulate CD40 expression.

Fig. 2.

Surface expression of constitutive and IFN-γ-induced CD40 and HLA-DR protein in normal (A) and TAO (B) orbital fibroblasts. Untreated fibroblasts and those stimulated with IFN-γ (500 U/ml for 72 h) were stained by indirect immunofluorescence for CD40 or HLA-DR, then fluorescein conjugated anti-immunoglobulin. Staining with secondary reagent only (control) indicates background fluorescence of untreated and IFN-γ-treated fibroblasts. Cells (1 × 104/sample) were analyzed by flow cytometry. Mean fluorescence intensities are indicated in parentheses.

In addition to enhancing CD40 expression, IFN-γ can induce expression of HLA-DR, a human major histocompatibility complex (MHC) class II molecule required for specific engagement and activation of T lymphocytes. IFN-γ is known to upregulate HLA-DR on a variety of cell types including orbital fibroblasts (10). Histograms of HLA-DR expression by normal and TAO orbital fibroblasts, treated without or with IFN-γ, are shown in Fig. 2. Neither fibroblast type expresses HLA-DR in the absence of IFN-γ; however, when stimulated with IFN-γ for 72 h, virtually all cells stained positively.

CD40L triggers mobilization of NF-κB in orbital fibroblasts.

To address whether CD40 on orbital fibroblasts is functional, its signal transduction capacity was investigated. Although little is known about CD40 signaling in nonhematopoietic cell types, CD40 on B lymphocytes is known to mobilize the ubiquitous transcription factor NF-κB (23). Therefore, electrophoretic mobility shift analysis of NF-κB binding activity in nuclear extracts from IFN-γ-treated normal and TAO orbital fibroblasts was performed. The cells were stimulated for 2 h with control insect cell membranes containing GST (control) or insect cell membranes containing human CD40L. Nuclear proteins were extracted from the treated cells, and equivalent amounts of protein were incubated with32P-labeled double-stranded oligonucleotide probe representing the consensus sequence for the NF-κB binding site. An induction of NF-κB binding activity (arrow) was observed after stimulation with CD40L of normal and TAO orbital fibroblasts (Fig. 3). Supershift studies with anti-p65 antisera confirmed that the upper band is a p65-containing NF-κB complex (data not shown). A constitutively shifted lower band was detected in all extracts. This band is a putative p50/p50 homodimer thought to be transcriptionally inactive. The specificity of the protein-DNA complex was confirmed by incubation of the nuclear extract with 200-fold excess unlabeled NF-κB probe before addition of 32P-labeled probe (data not shown). The detected shift in CD40L-treated cells is similar to that seen after treatment with the potent NF-κB-mobilizing agent TNF-α (Fig. 3) (3). Densitometric analysis of the shifted bands vs. the basal level of NF-κB binding activity revealed no significant difference between the normal and TAO fibroblasts (data not shown). These results indicate that orbital fibroblasts, primed with the proinflammatory cytokine IFN-γ to optimize CD40 expression, display functional CD40 molecules on their surface that can be activated with CD40L to translocate NF-κB into the nucleus.

Fig. 3.

Insect cell membranes containing human CD40L trigger nuclear mobilization of nuclear factor-κB (NF-κB) in normal and TAO orbital fibroblasts. IFN-γ-pretreated orbital fibroblasts were stimulated with control insect cell membranes, CD40L-containing insect cell membranes, or 50 U/ml recombinant tumor necrosis factor-α (TNF-α) for 2 h. Nuclear extracts were prepared and analyzed by elecrophoretic mobility shift assay for NF-κB binding activity. Arrow, shifted protein-DNA complex induced by CD40L and TNF-α.

Cross-linking CD40 on orbital fibroblasts activates production of IL-6 and IL-8.

To determine the functional consequence of CD40-mediated signal transduction in the orbital fibroblasts, we investigated IL-6 and IL-8 expression after triggering of CD40 with CD40L. IL-6 and IL-8 are potent proinflammatory cytokines, the promoter regions of which possess NF-κB binding sites. IL-6 is a pleiotropic cytokine, often expressed at high levels at sites of inflammation, and has been shown to be an autocrine growth factor for mouse fibroblasts (11). Confluent monolayers of orbital fibroblasts were pretreated for 72 h in medium alone or with IFN-γ to upregulate surface CD40 expression. The cells were then stimulated for 72 h with fresh medium, control insect cell membranes, or membranes containing human CD40L. Media samples were then harvested for cytokine-specific ELISA. Induction of IL-6 after activation of CD40 in orbital strains from three normal and three TAO tissues (±IFN-γ) is shown in Fig. 4. The concentration of cytokine detected in the conditioned medium was corrected for the background secretion of cytokine observed in untreated fibroblasts. Orbital fibroblasts express low levels of IL-6 and have detectable IL-6 mRNA under basal culture conditions (data not shown). Fibroblasts stimulated with CD40L, without IFN-γ, expressed higher levels of IL-6 than did controls. A far more robust induction of IL-6 production occurred when the fibroblasts were primed with IFN-γ before CD40 engagement (Fig. 4 A). Although levels of IL-6 varied in all six fibroblast strains tested, the pattern of induction remained the same (Fig.4 B).

Fig. 4.

Activation of normal and TAO orbital fibroblasts by CD40 engagement results in elevated interleukin-6 (IL-6) expression. Fibroblasts were untreated or treated with IFN-γ (500 U/ml) for 72 h, and then insect cell membranes containing glutathioneS-transferase (control) or CD40L were added for 72 h. Media were harvested and assayed for IL-6 using a specific double-determinant ELISA. Values are means ± SD (n = 3 replicates).A: 1 normal and 1 TAO cell strain were tested for induction of IL-6. B: 4 additional strains of orbital fibroblasts (2 normal and 2 TAO) were assessed for IL-6 induction. Pattern of induction remained the same. Results from IFN-γ-primed cells are presented.

IL-8 was selected for analysis because it is a chemoattractant agent for human T cells and neutrophils and is released from hematopoietic and nonhematopoietic cells at sites of tissue injury and chronic inflammation (9, 36). IL-8 production after activation of CD40 is shown in Fig. 5. Orbital fibroblasts secrete a low level of IL-8, and IL-8 mRNA can be detected by RT-PCR under basal conditions (data not shown). There was an upregulation of IL-8 in response to CD40L in fibroblasts not primed with IFN-γ. A more dramatic induction of IL-8 occurred when orbital fibroblasts were pretreated with IFN-γ before CD40L stimulation (Fig.5 A). As with IL-6, the pattern of IL-8 induction remained consistent for all six strains tested (Fig.5 B).

Fig. 5.

Activation of normal and TAO orbital fibroblasts by CD40 engagement results in elevated interleukin-8 (IL-8) expression. Cultures were treated as described in Fig. 4 legend. Media were harvested and assayed for IL-8 using a specific double-determinant ELISA. Values are means ± SD (n = 3 replicates).A: 1 normal and 1 TAO cell strain were tested for induction of IL-8. B: 4 additional strains of orbital fibroblasts (2 normal and 2 TAO) were then assessed for IL-8 induction. Pattern of induction remained the same. Results from IFN-γ-primed cells are presented.

A time course from 12 to 72 h was performed for the CD40L-induced secretion of IL-6 and IL-8. Elevated levels of both cytokines were detected as early as 12 h, but maximal cytokine secretion was observed at 72 h (data not shown). These data strongly suggest that IL-6 and IL-8 secretion is not a direct result of the rapid CD40L-induced NF-κB mobilization (Fig. 3). Cytokine synthesis may be a downstream consequence of CD40-mediated activation of orbital fibroblasts.


CD40 has been recently detected on nonlymphoid cells, such as keratinocytes, endothelial cells, and fibroblasts; however, certain primary human fibroblasts, such as those from the skin of the abdominal wall, do not express CD40 (15, 16, 21). Here we have examined the expression of CD40 by orbital fibroblasts. Normal and TAO orbital fibroblasts were found to display this surface antigen. Moreover, CD40 engagement with its natural ligand resulted in the activation of intracellular signaling and a dramatic upregulation of IL-6 and IL-8 expression. We hypothesize that orbital fibroblasts interact with infiltrating T lymphocytes via the CD40-CD40L costimulatory pathway. This in turn promotes fibroblast activation, which can result in the tissue remodeling observed in orbital inflammatory conditions such as TAO.

RT-PCR analysis of total RNA extracted from normal and TAO orbital fibroblasts revealed a detectable basal level of CD40 mRNA expression (Fig. 1). Similar to pulmonary, gingival, and foreskin fibroblasts, there was an apparent increase in the steady-state level of CD40 mRNA when the orbital fibroblasts were stimulated with IFN-γ (12). This finding was confirmed by flow cytometry (Fig. 2). In addition to enhancing CD40 expression, IFN-γ treatment also induced expression of HLA-DR on the orbital fibroblasts. This finding is consistent with our earlier studies (18). Despite being isolated from tissues in two very different states (inflamed vs. noninflamed), the two types of fibroblasts have similar basal levels of CD40 and HLA-DR is absent. Moreover, IFN-γ induces HLA-DR expression and relatively similar levels of CD40. It can therefore be inferred that any increase in surface expression of CD40 and HLA-DR on orbital fibroblasts in situ is not an imprinted characteristic induced by the disease process but, rather, represents a transient activation of these fibroblasts in the setting of an inflammatory mircroenvironment.

Orbital fibroblasts are heterogeneous with respect to expression of the surface antigen Thy-1. Thy-1 is expressed on 54–71% of the cells in primary cultures of orbital fibroblasts, with the remainder of the cells staining negatively (32). This expression profile is stable in culture, and representative clones of orbital fibroblasts can be isolated that faithfully retain Thy-1-positive and Thy-1-negative expression status. Thy-1 heterogeneity indicates a complexity of orbital tissue, which may provide insight into the mechanisms of tissue remodeling associated with inflammatory diseases of the orbit. Interestingly, normal and TAO orbital fibroblasts homogeneously express CD40. This suggests that all fibroblasts in the orbit are capable of receiving and transducing signals via CD40. CD40 expression is therefore not a criterion for rationalizing functional heterogeneity observed in orbital fibroblasts (32). However, CD40-mediated activation may be a key signaling conduit for orbital fibroblasts. The signal may be processed independently or in conjunction with one or more other signals to elicit a desired cellular response. This would then provide a mechanism for fibroblast subpopulations to serve distinct functions in the orbit.

The ubiquitous transcription factor NF-κB has been identified as a major signaling pathway utilized in CD40-mediated events in hematopoietic cells, such as induction of proinflammatory cytokine synthesis (3, 23). Little is known about CD40-mediated cell signal transduction and cellular activation in fibroblasts. It was hypothesized that NF-κB is mobilized after fibroblasts are triggered through CD40. Indeed, we have verified that CD40 engagement of orbital fibroblasts with insect cell membranes containing human CD40L results in the translocation of an NF-κB complex to the cell nucleus. These findings are significant, because nuclear mobilization of NF-κB is a conduit for cellular activation relevant to the transcriptional upregulation of many genes involved in inflammatory and specific immune responses (3).

Our results suggest that the cytokine milieu surrounding the orbital fibroblast in situ may represent a critical determinant of the cellular responses mediated through the CD40-CD40L pathway. Of particular relevance would appear to be the tissue concentration of IFN-γ. This cytokine is known to upregulate cellular adhesion molecules such as intercellular adhesion molecule 1 and vascular cell adhesion molecule 1, which can promote fibroblast adhesion to T lymphocytes (10, 14, 25). During the inflammation and tissue remodeling associated with TAO, large numbers of infiltrating T lymphocytes have been detected in orbital tissue. In addition, T lymphocytes with cytolytic properties have been isolated from TAO tissue (7, 22). Little is known about the cytokine milieu present in TAO; however, one study demonstrated by immunostaining the presence of IL-1α, TNF-α, and IFN-γ (17). The actual concentrations of these cytokines were not determined. These observations, in conjunction with our results, suggest that an inflammatory environment rich in IFN-γ-producing infiltrating cells, such as T lymphocytes, may provide a priming effect for interstitial fibroblasts and elevate CD40 expression on fibroblasts in inflamed tissue. This might in turn condition the fibroblasts to be hyperresponsive to CD40 engagement and, therefore, result in the amplified induction of proinflammatory cytokines.

IL-6, the human promoter of which contains an identifiable NF-κB site, is of significant interest with regard to TAO. A recent report by Salvi and colleagues (27) demonstrated elevated serum levels of IL-6 in patients with Graves’ disease compared with normal subjects. These authors also observed higher levels of serum soluble IL-6 receptor in patients with active inflammatory TAO than in patients with inactive orbital disease. Thus IL-6 may have a pathogenic role in TAO. The degree of T lymphocyte infiltration detected in TAO tissue (7, 30) suggests that a chemoattractant signal for T lymphocytes, such as IL-8, might emanate from orbital cells. We hypothesized that activation of CD40 on orbital fibroblasts may represent an important mechanism for the upregulation of local IL-6 and IL-8 expression in the orbit. Data presented here support this hypothesis by clearly demonstrating that engagement of CD40 on orbital fibroblasts by CD40L results in the induction of IL-6 and IL-8 proteins (Figs. 4 and 5). No difference was observed in the induction of either cytokine in normal and TAO fibroblasts.

Human orbital fibroblasts exhibit a set of phenotypic attributes in culture that distinguish them from many types of extraorbital fibroblasts. They are particularly susceptible to certain actions of proinflammatory cytokines with regard to the induction of plasminogen activator inhibitor type 1 (5, 20, 29), hyaluronan synthesis (31, 33), and the upregulation of prostaglandin endoperoxide H synthase 2 (35). The latter may represent the molecular basis for the intense inflammatory response observed in certain orbital disorders such as TAO (30). It is therefore critical to compare the magnitude of CD40L-dependent cytokine expression in orbital fibroblasts with that in fibroblasts derived from other tissues. Future studies will examine whether CD40 signaling influences the expression of prostaglandin endoperoxide H synthase 2 and the synthetic rate of hyaluronan, independently and in conjunction with proinflammatory cytokines.

In conclusion, data in this report support the hypothesis that the fibroblast CD40 is an important mediator of cognate interactions between infiltrating CD40L-expressing immunocompetent cells and fibroblasts. These interactions may play a pivotal role in regulating normal tissue function and may underlie many aspects of remodeling associated with inflammation in the orbit. Anti-CD40L antibodies have been shown to be effective in preventing the onset of disease in murine models of collagen-induced arthritis, experimental autoimmune encephalomyelitis, and lupus nephritis (8, 24). We propose that strategies utilizing recombinant fusion proteins or monoclonal antibodies, targeting the disruption of the CD40-CD40L interactions between fibroblasts and lymphocytes, may represent powerful therapeutic strategies for TAO and other autoimmune diseases.


We thank Chantal K. Turner for expert technical assistance.


  • Address for reprint requests: R. P. Phipps, University of Rochester School of Medicine and Dentistry, 601 Elmwood Ave., Box 704, Rochester, NY 14642.

  • This work was supported in part by National Institutes of Health Grants CA-11198, DE-11047, AG-56002, EY-08976, EY-11708, and P30-ES-01247, a Merit Review Award from the Department of Veterans Affairs, and the Rochester Area Pepper Center. G. D. Sempowski is supported by National Institute of Dental Research Grant DE-07202.

  • Present address of G. D. Sempowski: Duke University Medical Center, Box 3258, Durham, NC 27710.


View Abstract