A CD40–CD154 interaction in tissue fibrosis

Authors


Abstract

Objective

To examine the role of an interaction between fibroblasts and mononuclear infiltrates through CD40–CD154 engagement in the development of tissue fibrosis.

Methods

Cultured dermal fibroblasts derived from healthy skin were induced to express CD40 by adenoviral gene transfer, stimulated with soluble CD154, and evaluated for proliferation, gene expression, and protein expression in vitro. The skin of mice with bleomycin-induced skin sclerosis, a model for systemic sclerosis (SSc), was assessed for CD40 and CD154 expression, in vivo fibroblast proliferation, and the expression of specific genes. The effects of an anti-CD154 monoclonal antibody on bleomycin-induced skin sclerosis were also examined.

Results

Upon stimulation with soluble CD154, cultured fibroblasts induced to express CD40 by adenoviral gene transfer proliferated and showed up-regulation of the genes for intercellular adhesion molecule 1, interleukin-6 (IL-6), IL-8, monocyte chemoattractant protein 1 (MCP-1), and RANTES, as well as up-regulation of their proteins. In the skin from bleomycin-treated mice, dermal fibroblasts expressed CD40, and mast cells and CD4+ T cells expressed CD154. Electron microscopic analysis revealed fibroblasts attached to mast cells and T cells with primitive contacts. The proliferation of fibroblasts and the up-regulated MCP-1 gene expression preceded thickening of the dermis. Finally, the anti-CD154 antibody inhibited the bleomycin-induced skin sclerosis by suppressing fibroblast proliferation and down-regulating MCP-1 expression.

Conclusion

The interaction between fibroblasts and mast cells or T cells through CD40–CD154 signaling is critical for fibroblast activation early in the course of fibrosis. Blockade of the CD40–CD154 signal may be a novel therapeutic strategy for human fibrotic diseases, such as SSc.

Systemic sclerosis (SSc) is a multiorgan disease characterized by excessive fibrosis of the dermis and internal organs, such as the lung and gastrointestinal system, and by microvascular abnormalities (1). The affected organs in SSc patients have characteristic histologic features, including excessive deposition of extracellular matrix, an increased number of fibroblasts, and various degrees of inflammatory cell infiltration (2). In addition, a variety of soluble mediators, including transforming growth factor β (TGFβ), connective tissue growth factor (CTGF), platelet-derived growth factor (PDGF), interleukin-4 (IL-4), IL-6, and monocyte chemoattractant protein 1 (MCP-1), are overexpressed or abnormally regulated in the affected skin of SSc patients (1), although the sequence of these events remains unclear. In this regard, infiltration of mononuclear cells, including T cells, B cells, and macrophages, into the dermis is observed in the early stage of the disease (3). These inflammatory cells are thought to be involved in fibroblast activation through their production of soluble mediators and/or cell–cell contacts, although the detailed mechanisms are unknown.

An interaction between CD40 and its ligand CD154 has recently emerged as a part of the critical pathway for immune and inflammatory responses (4, 5). CD40 is a member of the tumor necrosis factor receptor superfamily originally identified on B cells, and it is expressed on fibroblasts and endothelial cells in certain pathogenic conditions. On the other hand, the expression of CD154 is restricted to CD4+ T cells, mast cells, basophils, and eosinophils in an activation-dependent manner. Cultured fibroblasts derived from the affected skin of SSc patients are reported to express CD40 (6), and these patients also have elevated levels of serum-soluble CD154 and CD40 (7, 8). These findings suggest that the CD40–CD154 interaction is involved in the pathogenic processes of SSc. In this study, we examined the potential role in the fibrotic process of the CD40 signal provided by CD154 on inflammatory infiltrates, using in vitro and in vivo experimental systems.

MATERIALS AND METHODS

Human skin fibroblast culture.

Primary cultures of fibroblasts were generated from forearm skin biopsy specimens obtained from 2 healthy donors. These cells were obtained after the donors gave their written informed consent, which was approved by the Institutional Review Board. Each fibroblast culture was maintained as a monolayer in Dulbecco's modified Eagle's medium (Invitrogen, Carlsbad, CA) containing 10% fetal bovine serum, 50 units/ml penicillin, and 50 μg/ml streptomycin in a humidified atmosphere of 5% CO2 at 37°C. All in vitro experiments were performed using cultured fibroblasts at <5 passages.

Adenoviral gene transfer of CD40.

Recombinant adenovirus harboring CD40 (AdCD40) and control virus without an insert (AdControl) were prepared using the AdEasy Adenoviral Vector System (Stratagene, La Jolla, CA). Full-length human CD40 complementary DNA (cDNA) was isolated from lung tissue first-strand cDNA (Clontech, Palo Alto, CA) by polymerase chain reaction (PCR) with the following primers: 5′-ATGGTTCGTCTGCCTCTGCAG-3′ (sense) and 5′-CACACTCCTGGGTGGGTGCA-3′ (antisense). Cultured fibroblasts were infected with AdCD40 or AdControl at a multiplicity of infection (MOI) of 100 in serum-free FGM-2 medium (Cambrex, Walkersville, MD) to induce CD40 expression.

Analysis of CD40 downstream signal transduction.

Fibroblasts infected with AdCD40 or AdControl were stimulated with a recombinant soluble human CD154 trimer (PeproTech, London, UK) at a final concentration of 0.5 μg/ml. The amounts of total and phosphorylated IκB in fibroblasts were semiquantitatively examined on immunoblots probed with anti-IκB and anti–phospho-IκB antibodies (Cell Signaling Technology, Beverly, MA), respectively, as described previously (9).

In vitro cell proliferation assay.

Fibroblasts infected with AdCD40 or AdControl or mock-treated fibroblasts without virus infection were cultured in triplicate in the presence or absence of soluble CD154 for 48 hours with 3H-thymidine (0.5 μCi/ml; Perkin-Elmer Life and Analytical Sciences, Boston, MA). The cells were then harvested, and 3H-thymidine incorporation was determined using a TopCount microplate scintillation counter (Packard, Meriden, CT). In some experiments, the soluble CD154 was preincubated with an anti-human CD154 monoclonal antibody (mAb) (clone 24-31) or isotype-matched control mAb to an irrelevant antigen (Ancell, Bayport, MN) for 30 minutes before use in the assay.

Messenger RNA (mRNA) expression profiles of cultured fibroblasts.

Fibroblasts infected with AdCD40 or AdControl were stimulated with or without soluble CD154, harvested 3, 12, 24, and 48 hours later, and subjected to total RNA isolation using an RNeasy kit (Qiagen, Valencia, CA). Mock-treated fibroblasts without virus infection were also used as a control. First-strand cDNA was synthesized from each total RNA sample using an oligo(dT) primer and avian myeloblastosis virus reverse transcriptase (RT) XL (Takara, Kusatsu, Japan) and subjected to PCR to evaluate the expression of specific mRNA. The genes examined included collagen α1(I) (COL1A1), collagen α1(III) (COL3A1), fibronectin, TGFβ receptor type II, PDGF receptor α, HLA–DR, CTGF, IL-6, IL-8, RANTES, MCP-1, intercellular adhesion molecule 1 (ICAM-1), Smad2, Smad4, Smad7, matrix metalloproteinase 1 (MMP-1), MMP-13, tissue inhibitor of metalloproteinases 1 (TIMP-1), αv integrin, β5 integrin, and GAPDH. The primer set, annealing temperature, and number of cycles for each reaction were determined to enable semiquantitative evaluation of gene expression (Table 1). The PCR products were separated by electrophoresis on 2% agarose gels and visualized with ethidium bromide. The mRNA expression level of the selected genes was further quantified using TaqMan real-time PCR (Applied Biosystems, Foster City, CA) according to the manufacturer's protocol.

Table 1. Primers, annealing temperature, and number of PCR cycles used for the amplification of human genes potentially involved in tissue fibrosis*
GeneSense, 5′ → 3′Antisense, 5′ → 3′Annealing temperature, °CPCR cycles
  • *

    PCR = polymerase chain reaction; TGFβRII = transforming growth factor β receptor type II; PDGFRα = platelet-derived growth factor receptor α; CTGF = connective tissue growth factor; IL-6 = interleukin-6; MCP-1 = monocyte chemoattractant protein 1; MMP-1 = matrix metalloproteinase 1; TIMP-1 = tissue inhibitor of metalloproteinases 1.

COL1A1CCCACCAATCACCTGCGTACAGATTCTTGGTCGGTGGGTGACTCTGA6525
COL3A1GAGATGTCTGGAAGCCAGAACCATGATCTCCCTTGGGGCCTTGAGGT6525
FibronectinTGGAACTTCTACCAGTGCGACTGTCTTCCCATCATCGTAACAC6225
TGFβRIITGGTGCTCTGGGAAATGACAAAGAGCTATTGGTAGTGTTTAGGGAGCCG6030
PDGFRαATCAATCAGCCCAGATGGACTTCACGGGCAGAAAGGTACT5730
HLA–DRCAGAGGTAACTGTGCTCACGAACAGCCCTCCCAGTGCTTGAGAAGAGGCTCATCC6532
CTGFTACCGACTGGAAGACACGTTCCGTCGGTACATACTCCACA5832
IL-6GTACATCCTCGACGGCATCTCAGCTGTGGTTGGGTCAGGGGTGGTTAT5528
IL-8CACTGTGCCTTGGTTTCTCCTCACAGCTGGCAATGACAAGAC6330
RANTESATGAAGGTCTCCGCGGCACGCCTAGCTCATCTCCAAAGAGTT5532
MCP-1TTCTCAAACTGAAGCTCGCACTCTCGCCTGTGGAGTGAGTGTTCAAGTCTTCGGAGTT6625
ICAM-1TGACCAGCCCAAGTTGTTGGATCTCTCCTCACCAGCACCG6335
Smad2ATCCTAACAGAACTTCCGCCTCAGCAAAAACTTCCCCAC5935
Smad4GCATCGACAGAGACATACAGCAACAGTAACAATAGGGCAG6335
Smad7GCCCTCTCTGGATATCTTCTGCTGCATAAACTCGTGGTCA5535
MMP-1GGTGATGAAGCAGCCCAGCAGTAGAATGGGAGAGTC6327
MMP-13TGGTGGTGATGAAGATGATTTGTCTAGTTACATCGGACCAAACTTTGAAG6030
TIMP-1TGGGGACACCAGAAGTCAACTTTTCAGAGCCTTGGAGGAG5525
αv integrinCTACGAAGCTGAGCTCATCGTTGCTCCCAGTTTGGAATCGG6130
β5 integrinTGGCACTCTCAGCTTTCCCTTCAACAGCCCCATCACCAAC6230
GAPDHTGAACGGGAAGCTCACTGGTCCACCACCCTGTTGCTGTA6020

The primers and probes for MCP-1, RANTES, and GAPDH were purchased from Applied Biosystems, and those for the other genes were as follows: for COL1A1, 5′-TGGGCGGGAGAGACTGTT-3′ (sense), 5′-GCCCCGGTGACACATCAA-3′ (antisense), 5′-TGCTGAAAGACTACCTCGTTCTTGT-3′ (probe); for COL3A1, 5′-TGGAGGCTGGATGTGCATTA-3′ (sense), 5′-CCTCAGACTTCATACTTCAGAATCTCA-3′ (antisense), 5′-TGTGTGTGGCAGCCAGGATGACTAGATC-3′ (probe); for IL-6, 5′-CATGGGCACCTCAGATTGTTG-3′ (sense), 5′-TGCCCAGTGGACAGGGTTT-3′ (antisense), 5′-ATGGGCATTCCTTCTTCTGGTC-3′ (probe); for IL-8, 5′-AAGGAACCATCTCACTGTGTGTAAAC-3′ (sense), 5′-AAATCAGGAAGGCTGCCAAGA-3′ (antisense), 5′-TGACTTCCAAGCTGGCCGTGGCT-3′ (probe); for ICAM-1, 5′-CCCCCCAAAACTGACACCTT-3′ (sense), 5′-ACCGCTGAGTGTCATTGTGAAC-3′ (antisense), 5′-TTAGCCACCTCCCCACCCACATACATTT-3′ (probe). The relative expression level for individual genes was calculated based on a standard curve obtained using serial dilutions of cDNA derived from cultured fibroblasts or peripheral blood mononuclear cells and normalized to the level of GAPDH expression.

Analysis of proteins in culture supernatants.

Fibroblasts infected with AdCD40 or AdControl or mock-treated fibroblasts were cultured with or without soluble CD154, and the culture supernatants were harvested after 48 or 96 hours. The levels of IL-6, IL-8, MCP-1, RANTES, MMP-1, TIMP-1, and C-propeptide of type I procollagen (PICP) in the supernatants were measured using commercially available enzyme-linked immunosorbent assay kits (R&D Systems, Cambridge, MA; GE Healthcare, Little Chalfont, UK; Takara Bio, Otsu, Japan).

Flow cytometric analysis.

The expression of human CD40 and ICAM-1 on unfixed fibroblasts was evaluated by flow cytometry using a phycoerythrin-conjugated anti-CD40 mAb (clone mAb89) or an anti–ICAM-1 mAb (clone 84H10) (Beckman Coulter, Fullerton, CA), respectively. The cell staining was analyzed on a FACSCalibur flow cytometer (Becton Dickinson, San Diego, CA) using CellQuest software (Becton Dickinson).

Mouse skin sclerosis model.

Female C3H/He mice at age 6 weeks (purchased from Sankyo Laboratory, Shizuoka, Japan) were treated daily with subcutaneous injections of bleomycin (Nippon Kayaku Company, Tokyo, Japan) for up to 3 weeks, as previously described (10). Bleomycin dissolved in phosphate buffered saline (PBS) was injected under the shaved skin on one side of the back; the same volume of PBS was injected into the other side as a control. In some experiments, mice were divided into 3 groups, and 500 μg of hamster anti-mouse CD154 mAb (clone MR1; kindly provided by Eisai, Tokyo, Japan), 500 μg of hamster IgG (MP Biomedicals, Irvine, CA), or the same volume of PBS was injected intraperitoneally every other day starting 1 week before initiation of the bleomycin injections. The day after mice received the final injections, the dorsal skin was removed. Skin specimens from the bleomycin-injected and mock-treated sites were subjected to histologic, biochemical, and molecular biologic analyses.

Histologic evaluations.

Each skin specimen was separated into 2 pieces; one was fixed in 10% neutral buffered formalin and embedded in paraffin, and the other was snap-frozen in OCT compound (Miles, Elkhart, IN) and stored at –80°C. The paraffin-embedded sections were used for hematoxylin and eosin staining, Mallory's stain, and toluidine blue staining. The dermal thickness was semiquantitatively assessed using the ratio of the thickness of the dermal layer (the distance between the epidermal–dermal junction and the dermal–subcutaneous fat junction) to the thickness of the whole skin, which was measured in 3 consecutive skin sections with Mallory's stain. Cells containing metachromatic granules detected with toluidine blue staining were regarded as mast cells. All evaluations were performed on at least 3 random grids per section and expressed as the mean number of cells per mm3.

The frozen sections were used for immunohistochemical analysis. The expression of CD40 and CD154 was examined by peroxidase-based immunohistochemistry using anti-mouse CD40 and anti-mouse CD154 polyclonal antibodies (Santa Cruz Biotechnology, Santa Cruz, CA). The coexpression of CD154 and histamine or CD4 was examined by immunofluorescence double-staining with the anti-mouse CD154 antibody in combination with a rabbit anti-mouse histamine antibody (Progen Biotechnik, Heidelberg, Germany) or biotin-conjugated rat anti-mouse CD4 mAb (clone L3T4; Becton Dickinson). The secondary antibodies were an Alexa Fluor 488–conjugated anti-goat IgG antibody and Alexa Fluor 568–conjugated streptavidin (Molecular Probes, Eugene, OR), respectively. All experiments included control staining with species- and isotype-matched antibodies. The tissue sections were observed under an IX71 fluorescence microscope (Olympus, Tokyo, Japan).

Electron microscopy.

A portion of mouse skin tissue was immediately fixed with 2.5% glutaraldehyde and subjected to electron microscopic examination, as described previously (11). Briefly, 1-μm–thick sections were stained with methylene blue, and the portions of interest were thin-sectioned and examined under a 1200 EX II transmission electron microscope (JEOL, Tokyo, Japan).

In vivo cell proliferation assay.

Mice were given 1 intraperitoneal injection of 1 μg of 5-bromo-2′-deoxyuridine (BrdU). A skin specimen was harvested after 24 hours and subjected to immunohistochemical analysis using the BrdU In-Situ Detection Kit (Becton Dickinson). The number of BrdU-positive cells was counted on at least 3 random grids per section and expressed as the mean number of cells per mm3.

Messenger RNA expression in skin.

Skin samples were disrupted with a Mixer Mill MM 300 (Geneq, Montreal, Quebec, Canada), and then the total RNA was isolated and used for cDNA synthesis. The mRNA expression levels of COL1A1, IL-6, ICAM-1, RANTES, MCP-1, and GAPDH were quantitatively evaluated using real-time PCR. All primers and probes were purchased from Applied Biosystems. The expression level of each gene was normalized to the GAPDH expression level and is shown as a ratio of the expression at the bleomycin-injected site to the expression at the mock-treated site.

Hydroxyproline content determination.

Full-thickness, 6-mm–diameter punch biopsy specimens were obtained from the shaved dorsal skin of each animal and stored at –80°C. The amount of hydroxyproline was spectrophotometrically measured in hydrolyzed skin samples as described previously (12). The relative hydroxyproline content was calculated as the ratio of the level in the bleomycin-injected site to that in the mock-treated site.

Statistical analysis.

All continuous variables are shown as the mean ± SD. The difference between 2 groups was examined for statistical significance by a paired t-test.

RESULTS

Induction of CD40 expression on cultured fibroblasts.

Cultured fibroblasts derived from SSc-affected skin showed CD40 expression in the primary culture, but lost the expression within 3 passages (data not shown). Thus, we decided to use fibroblasts derived from the forearm skin of healthy individuals, that were induced to express CD40 by adenoviral gene transfer, to evaluate the role of the CD40–CD154 signal in fibroblast activation. Flow cytometric analysis revealed cell surface expression of CD40 on fibroblasts infected with AdCD40, but not on those infected with AdControl (Figure 1a). The CD40 expression increased in a virus dose–dependent manner and reached a plateau at 100 MOI (data not shown). In addition, the CD40 expression was apparent 24 hours after infection and peaked at 3 days, with sustained expression for at least 7 days (data not shown). Consistent results were obtained from 4 independent experiments using fibroblasts from 2 different donors. Based on these findings, we used cultured fibroblasts that were infected with AdCD40 or AdControl at 100 MOI and cultured for 3 days in the subsequent experiments.

Figure 1.

Induction of CD40 expression on cultured human fibroblasts by adenoviral gene transfer, and CD154-induced NF-κB activation and proliferation of CD40-expressing fibroblasts. a, Cultured fibroblasts infected with recombinant adenovirus harboring CD40 (CD40-AdV) or control virus without an insert (Control AdV) were subjected to flow cytometry to detect the cell-surface expression of CD40. Anti-CD40 monoclonal antibody (mAb)–treated cells are shown as solid histograms; open histograms represent control mAb–treated cells. b, Fibroblasts infected with Control AdV or CD40-AdV were incubated with soluble CD154 (sCD154) and cultured for 0, 2, 5, 10, and 30 minutes. The total cellular lysates were subjected to immunoblots probed with an anti-IκB or anti–phospho-IκB antibody. c, Fibroblasts infected with Control AdV or CD40-AdV were incubated in triplicate with 3H-thymidine for 48 hours in the presence or absence of sCD154. The cells were harvested, and 3H-thymidine incorporation was determined in cpm. d, CD154-induced proliferation of CD40-AdV–infected fibroblasts was evaluated after preincubation of sCD154 with anti-human CD154 or control mAb. Concordant results were obtained from 4 independent experiments using cultured fibroblasts from 2 different donors. Values in c and d are the mean and SD.

To determine whether the adenovirally induced CD40 was functional, the activation status of NF-κB was examined in AdCD40-infected fibroblasts after treatment with soluble human CD154 trimer. The ligation of CD40 induced the rapid degradation and phosphorylation of IκB in AdCD40-infected fibroblasts, but not in AdControl-infected fibroblasts (Figure 1b), indicating that the CD40 expressed on fibroblasts transduced downstream signals upon binding its ligand.

In vitro effects of the CD40–CD154 signal on the proliferation and gene expression profiles of fibroblasts.

Adenovirus infection did not promote proliferation of cultured fibroblasts. Upon stimulation with soluble CD154, AdCD40-infected fibroblasts proliferated, but AdControl-infected fibroblasts did not (Figure 1c). This stimulatory effect of soluble CD154 was neutralized by its pretreatment with an anti-human CD154 mAb, but not with control mAb (Figure 1d), indicating that the fibroblast proliferation was mediated through the CD40–CD154 engagement.

To determine the effects of the CD40 signal on gene expression profiles, fibroblasts infected with AdCD40 or AdControl were cultured with or without soluble CD154 and subjected to RT-PCR for detection of selected genes potentially involved in the fibrotic process (Figure 2a). The results showed that the genes for IL-6, ICAM-1, IL-8, MCP-1, and RANTES were up-regulated exclusively in CD40-expressing fibroblasts upon stimulation with CD154, but the genes for matrix proteins, such as COL1A1 and COL3A1, in the CD40-expressing fibroblasts were not. Adenovirus infection did not influence the gene expression of collagens. Serial gene expression analysis revealed that the up-regulation of the 5 genes was sustained for at least 48 hours (data not shown). Quantitative gene expression analysis by real-time PCR confirmed the up-regulated gene expression of IL-6, ICAM-1, IL-8, MCP-1, and RANTES in CD40-expressing fibroblasts upon CD154 stimulation (Figure 2b), with no change in the levels of COL1A1 and COL3A1.

Figure 2.

Messenger RNA expression profiles in cultured human fibroblasts induced to express CD40 by adenoviral gene transfer upon stimulation with soluble CD154 (sCD154). Fibroblasts infected with recombinant adenovirus harboring CD40 (CD40-AdV) or control virus without an insert (Control AdV) were incubated for 3 hours in the presence or absence of sCD154. a, The mRNA expression of a series of genes potentially involved in fibrotic processes was examined by reverse transcriptase–polymerase chain reaction (PCR). b, The mRNA expression levels of COL1A1, COL3A1, interleukin-6 (IL-6), intercellular adhesion molecule 1 (ICAM-1), IL-8, monocyte chemoattractant protein 1 (MCP-1), and RANTES were quantified in duplicate using TaqMan real-time PCR. Concordant findings were obtained from 3 independent experiments using cultured fibroblasts from 2 different donors. Values are the mean and SD. CTGF = connective tissue growth factor; PDGFRα = platelet-derived growth factor receptor α; MMP-1 = matrix metalloproteinase 1; TIMP-1 = tissue inhibitor of metalloproteinases 1.

AdCD40-infected or AdControl-infected fibroblasts were cultured with or without soluble CD154 for 48 hours, and the concentration of soluble factors was measured in the culture supernatant. As shown in Figure 3a, the secretion of IL-6, IL-8, MCP-1, and RANTES was enhanced when CD40-expressing fibroblasts were stimulated with soluble CD154, whereas stimulation of CD40-expressing fibroblasts with soluble CD154 did not influence secretion of MMP-1 or TIMP-1. Upon stimulation of CD40-expressing fibroblasts with soluble CD154, the cell-surface expression of ICAM-1, examined by flow cytometry, increased as well (Figure 3b). Concentrations of PICP in culture supernatants, which reflect the amount of type I collagen produced from fibroblasts, were similar in the presence or absence of CD40 expression on fibroblasts and/or CD154 stimulation in 48-hour cultures, but were slightly increased in 96-hour cultures in CD40-expressing fibroblasts stimulated with soluble CD154 (Figure 3c). These results indicate that the CD40–CD154 signal induces proliferation and up-regulates the gene and protein expression levels of IL-6, ICAM-1, IL-8, MCP-1, and RANTES in cultured human skin fibroblasts.

Figure 3.

Protein expression of IL-6, IL-8, MCP-1, RANTES, MMP-1, TIMP-1, ICAM-1, and C-propeptide of type I procollagen (PICP) in cultured human fibroblasts induced to express CD40 by adenoviral gene transfer upon stimulation with sCD154. Fibroblasts infected with Control AdV or CD40-AdV were incubated for 48 or 96 hours in the presence or absence of sCD154. a, The concentration of IL-6, IL-8, MCP-1, RANTES, MMP-1, and TIMP-1 in supernatants of 48-hour cultures was determined in quadruplicate by enzyme-linked immunosorbent assay (ELISA). b, The cell-surface expression of ICAM-1 after the 48-hour culture was examined by flow cytometry. Solid and open histograms represent the fluorescence intensities of cells treated with anti–ICAM-1 and control monoclonal antibodies, respectively. c, The concentration of PICP in supernatants of 48- and 96-hour cultures was determined in quadruplicate by ELISA. Concordant results were obtained from 3 independent experiments using cultured fibroblasts from 2 different donors. Values in a and c are the mean and SD. See Figure 2 for other definitions.

Roles of CD40 and CD154 in the bleomycin-induced skin sclerosis model.

Subcutaneous injections of bleomycin resulted in marked skin sclerosis in mice, which was characterized by excessive accumulation of matrix proteins in the dermis and thinning of the fat and muscle layers, as reported previously (10). Serial evaluation of the bleomycin-injected skin showed that the dermal thickness increased gradually and was apparent after 2 weeks (Figure 4a). The increase in the numbers of infiltrated mononuclear cells and mast cells preceded the development of skin sclerosis and peaked at 2 weeks (Figures 4b and c). Ultrastructural evaluation of the bleomycin-injected skin specimens revealed a marked increase in newly synthesized fine collagen fibrils around the fibroblasts, indicating active extracellular matrix production (Figure 4d). Primitive contacts between fibroblasts and mast cells or lymphocytes were frequently observed at sites of active collagen production. In the dermis of mice treated with bleomycin for 1 week, CD40 was expressed on the majority of the fibroblasts, and CD154 was detected on the infiltrated round cells (Figure 4e). Fluorescence double-staining revealed that the CD154-expressing cells were mainly histamine-positive mast cells, and a few were CD4+ T cells (Figure 4f). Neither CD40 nor CD154 expression was detectable in the mock-treated skin of the same mice. Together, these findings suggest that the interaction between CD40 expressed on fibroblasts and CD154 expressed on mast cells or CD4+ T cells is involved in the development of bleomycin-induced skin sclerosis.

Figure 4.

Role of the CD40–CD154 signal in the murine bleomycin-induced skin sclerosis model. ac, Results of serial measurements of the dermal thickness (a), the number of mononuclear cells in the dermis (b), and the number of mast cells in the dermis (c) during bleomycin treatment (n = 3 mice). Asterisks denote a significant difference compared with the pretreatment condition (week 0). d, Electron microscopy of the dermis of mice treated with bleomycin for 2 weeks. Arrowheads denote primitive contacts between fibroblasts (Fib) and lymphocytes (Ly) or mast cells (Mast) (original magnification × 20,000). e, Immunohistochemistry analysis for CD40 and CD154 in the dermis of mice treated with bleomycin for 1 week. Arrows denote fibroblasts expressing CD40 (top) and mononuclear cells expressing CD154 (bottom). Note that 2 types of cells express CD154: large round cells (black arrows) and small round cells (red arrows) (original magnification × 200). f, Fluorescence double-staining for CD154 (green) and histamine (red) and for CD154 (green) and CD4 (red). Arrows denote mast cells coexpressing CD154 and histamine (top) and T cells coexpressing CD154 and CD4 (bottom) (original magnification × 400). g, Bottom, serial evaluation of 5-bromo-2′-deoxyuridine (BrdU)–positive proliferating dermal fibroblasts during bleomycin treatment (n = 3 mice). Asterisks denote a significant difference compared with the pretreatment condition (week 0). Top, Representative images of BrdU staining in mice treated with bleomycin for 1 and 3 weeks (original magnification × 200). h, Messenger RNA expression of COL1A1, RANTES, and MCP-1 in mice treated with bleomycin for 1 week, quantified by real-time PCR (n = 4 mice). The horizontal dashed line denotes the ratio of 1. Values in ac, g, and h are the mean and SD. Images in d–g are representative of images obtained in >3 mice that yielded concordant findings. PBS = phosphate buffered saline (see Figure 2 for other definitions).

Based on our in vitro results, we further assessed the fibroblast proliferation and gene expression profiles in the bleomycin-treated skin. The in vivo proliferation of dermal fibroblasts was serially evaluated by counting the BrdU-positive dermal fibroblasts. The number of proliferating fibroblasts was increased at 1 and 2 weeks, but almost no proliferation was detected at 3 weeks, despite the continuous injection of bleomycin (Figure 4g). After 1 week of bleomycin treatment, the expression of RANTES and MCP-1 mRNA was markedly up-regulated, and COL1A1 mRNA expression was increased 2-fold (Figure 4h). Expression of mRNA for IL-6 or ICAM-1 was not detected under these experimental conditions.

Attenuation of bleomycin-induced skin sclerosis by CD40–CD154 signal blockade.

To investigate the role of the CD40–CD154 interaction in bleomycin-induced skin sclerosis, we tested whether treatment with an anti-CD154 mAb would prevent the development of skin sclerosis. Mice were treated with an anti-CD154 mAb, control IgG, or PBS prior to the first bleomycin injection. Representative skin sections obtained at 3 weeks are shown in Figure 5a. In the anti-CD154 mAb–treated mice, much less accumulation of collagen fibrils was seen than in the control IgG– or PBS-treated mice, and the fat and muscle layers were retained. In addition, the dermal thickness at 3 weeks was significantly lower in the anti-CD154 mAb–treated group than in the other 2 groups (Figure 5b). Furthermore, the hydroxyproline content was significantly reduced in the anti-CD154 mAb–treated group compared with the control IgG–treated group (Figure 5c).

Figure 5.

Suppression of bleomycin-induced skin sclerosis by treatment with an anti-CD154 monoclonal antibody (mAb). Bleomycin-treated mice received injections of phosphate buffered saline (PBS), control IgG, or anti-CD154 mAb. a, Representative images of the skin with Mallory's stain (original magnification × 100). b, Results of measurements of dermal thickness at 3 weeks in mock-treated and bleomycin-treated mice that had received injections of PBS, control IgG, or anti-CD154 mAb. cg, Determination of the hydroxyproline content at 3 weeks (c), the number of mononuclear cell infiltrates at 2 weeks (d), the number of mast cells at 2 weeks (e), the number of 5-bromo-2′-deoxyuridine (BrdU)–positive proliferating fibroblasts at 1 week (f), and the gene expression level of COL1A1, RANTES, and monocyte chemoattractant protein 1 (MCP-1) at 1 week (g) in bleomycin-treated mice that received injections of control IgG and anti-CD154 mAb (n = 5 mice). Values in bg are the mean and SD. NS = not significant.

We further analyzed the mechanisms for the antifibrotic actions of the CD40–CD154 blockade in bleomycin-induced skin sclerosis. The numbers of mononuclear cells and mast cells that had infiltrated the dermis at 2 weeks were similar in the anti-CD154 mAb–treated and control IgG–treated groups (Figures 5d and e). In contrast, the number of BrdU-positive proliferating dermal fibroblasts at 1 week was significantly reduced in the anti-CD154 mAb–treated group compared with the control IgG–treated group (Figure 5f). In addition, the anti-CD154 mAb treatment suppressed the up-regulated expression of COL1A1, RANTES, and MCP-1 mRNA (Figure 5g).

DISCUSSION

In this study, we investigated the involvement of the CD40–CD154 interaction in the activation of fibroblasts in vitro and in the induction of tissue fibrosis in vivo. Analyses using cultured human dermal fibroblasts showed that the engagement of CD40-expressing fibroblasts with soluble CD154 induced cell proliferation and expression of several potential fibrogenic molecules, including IL-6, ICAM-1, IL-8, MCP-1, and RANTES. In a murine bleomycin-induced skin sclerosis model, the infiltration of CD154-expressing mast cells and CD4+ T cells into the dermis preceded dermal thickening. In addition, tissue fibrosis was accompanied by enhanced fibroblast proliferation and up-regulated expression of MCP-1 and RANTES, which were features of the fibroblasts undergoing CD40 stimulation in vitro. Furthermore, an anti-CD154 mAb prevented the progression of skin sclerosis in bleomycin-treated mice by suppressing both fibroblast proliferation and the up-regulated expression of MCP-1 and RANTES. Together, these findings indicate that an interplay of fibroblasts and mast cells or CD4+ T cells through the CD40–CD154 interaction plays a critical role in inducing tissue fibrosis. This interaction is important in the early phase of tissue fibrosis, while other signals, such as those generated through TGFβ, are necessary to establish fibrosis, probably in a later phase.

Our findings obtained in vitro were principally concordant with those from the bleomycin-induced mouse model in vivo. However, CD40 stimulation did not enhance collagen mRNA production in the short-term culture, whereas the CD40–CD154 signal blockade inhibited collagen accumulation in the skin induced by bleomycin treatment. This discrepancy is probably due to the increased number of fibroblasts and secondary effects of profibrogenic factors induced by CD40 signaling in bleomycin-treated mice. In fact, fibroblasts undergoing CD40 stimulation showed a trend toward increased type I collagen production at the later time point.

The present results are generally consistent with those of previous studies showing that CD40 engagement up-regulates the expression of ICAM-1 and soluble mediators, such as IL-6, IL-8, MCP-1, and RANTES, in CD40-expressing fibroblasts derived from various human tissues (6, 13, 14). All of these molecules have been shown to be involved in pathologic fibrotic processes, including those in SSc. ICAM-1 is detected on fibroblasts located in the perivascular areas surrounded by infiltrating lymphocytes in the early inflammatory stage of SSc (15, 16). In addition, cultured SSc fibroblasts exhibit an elevated expression level of ICAM-1, which enhances their capacity to bind T cells (17). Moreover, bronchoalveolar lavage fluid and serum from SSc patients contain elevated levels of IL-6, IL-8, MCP-1, and RANTES (18, 19). Of these soluble mediators, IL-6 has been shown to induce a concentration-dependent increase in the production of collagens by cultured dermal fibroblasts (20). Finally, the blockade of CCR1, one of the receptors for RANTES, suppresses the renal fibrosis induced by unilateral ureter obstruction and the pulmonary fibrosis induced by bleomycin (21, 22).

Accumulating evidence indicates that MCP-1 is critically involved in the development of various fibrotic conditions, including SSc. In the skin of SSc patients, MCP-1 is expressed by fibroblasts, keratinocytes, inflammatory cells, and endothelial cells, whereas normal control skin shows no MCP-1 expression (23). Regarding the mechanism for the fibrotic action of MCP-1, it has been shown that MCP-1 directly stimulates collagen production through the induction of endogenous TGFβ (24) and indirectly contributes to fibrosis through the induction of IL-4–producing T cells (25). MCP-1 has also been shown to be a key mediator in the development of bleomycin-induced skin sclerosis. Bleomycin injection induces skin sclerosis and an increased level of MCP-1 in wild-type mice, but fails to induce skin sclerosis in MCP-1−/− mice (26). Furthermore, bleomycin-induced skin sclerosis is inhibited by the antibody neutralization of MCP-1 (27). Our findings indicate that the CD40–CD154 interaction is one of the upstream signals that induce up-regulated MCP-1 expression in fibroblasts.

In mice treated with bleomycin, we observed that the majority of the CD154-expressing cells in the sclerotic skin were mast cells, which are known to induce immediate-type hypersensitivity reactions. Recent studies have suggested that mast cells interact with fibroblasts in a manner that leads to fibroblast activation in several fibrosis models (28). In a human skin wound-healing model, mast cells are recruited from the peripheral blood by the effect of a chemotactic factor, such as MCP-1, and stimulate fibroblasts via the production of IL-4 and TGFβ (29). Our results suggest that mast cells further stimulate fibroblasts through cell–cell contact via the CD40–CD154 interaction. However, it has also been shown that skin sclerosis is inducible by bleomycin in genetically mast cell–deficient mice (30). Another study showed that skin sclerosis was equally induced by bleomycin in a wild-type mouse and its SCID counterpart, suggesting that T cells are not required (31). These findings do not necessarily exclude the importance of the CD40–CD154 signal in the induction of bleomycin-induced skin sclerosis, because mast cells and T cells may compensate for each other in providing the CD154 signal that activates CD40-expressing fibroblasts.

The CD40–CD154 interaction is essential for acquired immunity, including T cell priming and the humoral immune response (4, 5). Therefore, the systemic administration of an anti-CD154 mAb might block responses induced by CD40–CD154 engagement not only in the involved skin, but also in the immune system. However, in our experiments the degree of mononuclear cell infiltration in the skin was not reduced by anti-CD154 mAb treatment, suggesting that the antifibrotic action of the anti-CD154 mAb is exerted mainly through interruption of the interaction between fibroblasts and mast cells or T cells at the site of fibrosis, rather than through the systemic suppression of immune and inflammatory processes.

The suppressive effect of the anti-CD154 mAb on skin sclerosis in vivo suggests the potential utility of interventions targeting the CD40–CD154 interaction for SSc and other fibrotic diseases. In this regard, blockade of the CD40–CD154 interaction attenuates skin fibrosis in tight-skin mice, which genetically develop hypodermal fibrosis (32). Two clones of anti-human CD154 humanized mAb have been developed and used in clinical trials in patients with autoimmune diseases, including systemic lupus erythematosus (33, 34) and immune thrombocytopenic purpura (35). Although this type of treatment was reported to have some efficacy, the potential risk of thromboembolic complications hindered its future development (36). This unexpected adverse event might have been due to the expression of CD154 by activated platelets (37), but the precise mechanisms are not fully understood. Alternative potential approaches include biologic agents targeting CD40, which would theoretically circumvent the thromboembolic complications. Anti-CD40 biologic agents have already been established and are being tested in clinical trials (38).

In summary, our results suggest that the interaction between fibroblasts and mononuclear infiltrates through the CD40–CD154 signal plays an important role in the development of pathologic fibrosis, especially in the early phase. Further studies are necessary to elucidate the roles of the CD40–CD154 interaction in the pathogenesis of human fibrotic diseases, such as SSc, and to assess the clinical utility of therapeutics that disrupt the CD40–CD154 interaction as antifibrotic interventions.

AUTHOR CONTRIBUTIONS

Dr. Kuwana had full access to all of the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.

Study design. Kuwana.

Acquisition of data. Kawai, Masuda, Kuwana.

Analysis and interpretation of data. Kawai, Masuda, Kuwana.

Manuscript preparation. Kawai, Kuwana.

Statistical analysis. Kawai, Kuwana.

Acknowledgements

We thank Dr. Yoko Ogawa for valuable technical advice on electron microscopic analysis, and Masaaki Kubota for excellent technical assistance.

Ancillary