Effects of tumor necrosis factor-alpha on connective tissue metabolism in normal and scleroderma fibroblast cultures

Phototherapy in systemic sclerosis: Review

Systemic scleroderma-also known as systemic sclerosis (SSc)-is a chronic systemic connective tissue disease characterized by collagen deposition in cutaneous and internal organs, leading to skin sclerosis and multiple organ fibrosis. The pathogenesis is complex and remains poorly understood. Treatment is based on organ involvement and requires a multidisciplinary approach. Skin sclerosis can cause disability, leading to decreasing quality of life. Various systemic antifibrotic therapies have been used; however, most have unsatisfactory results. Recently, phototherapy and in particular ultraviolet A (UVA) has been used to treat skin sclerosis in SSc patients with satisfactory results. The main mechanisms include lymphocyte apoptosis, cytokine alteration, inhibition of collagen synthesis and increased collagenase production, and neovascularization, leading to the breakdown of collagen fibrils resulting in skin softening or even healing digital ulcers. Most studies reported that psoralen plus UVA (PUVA) and UVA1 phototherapy improved clinical outcomes vis-à-vis skin sclerosis, joint mobility, ulcers, and histopathology. PUVA and UVA1 phototherapy therefore have potential as an alternative or adjunctive therapy for patients with SSc.

IL-22 capacitates dermal fibroblast responses to TNF in scleroderma

OBJECTIVES

Interleukin (IL) 22 mRNA in systemic sclerosis (SSc) skin and Th22 cells in SSc peripheral blood are increased, but the role of IL-22 in fibrosis development remains poorly understood.

METHODS

Biopsies were obtained from the involved skin of 15 SSc, 4 morphea and 8 healthy donors (HD). The presence of IL-22+ cells in the skin was determined by immunostaining. The in vitro response of HD and SSc fibroblasts to IL-22, IL-22 in conjunction with tumour necrosis factor (TNF) or keratinocyte conditioned medium was assessed by ELISA, radioimmunoassay (RIA), real-time PCR and western blot. The in vivo response in mice was assessed by histomorphometry.

RESULTS

IL-22+ cells were over-represented in the dermis and epidermis of morphea and in the epidermis of SSc compared with HD. The majority of dermal IL-22+ cells were T cells. Dermal fibroblasts expressed both IL-22 receptor subunits IL-10RB and IL-22RA, expression of which was enhanced by TNF and reduced by transforming growth factor (TGF)-β. IL-22 induced rapid phosphorylation of p38 and ERK1/2 in fibroblasts, but failed to induce the synthesis of chemokines and extracellular matrix components. However, IL-22 enhanced the production of monocyte chemotactic protein 1, IL-8 and matrix metalloproteinase 1 induced by TNF. Fibroblast responses were maximal in the presence of conditioned medium from keratinocytes activated by IL-22 in conjunction with TNF. Dermal thickness was maximal in mice injected simultaneously with IL-22 and TNF.

CONCLUSIONS

IL-22 capacitates fibroblast responses to TNF and promotes a proinflammatory fibroblast phenotype by favouring TNF-induced keratinocyte activation. These results define a novel role for keratinocyte-fibroblast interactions in the context of skin fibrosis.

The role of the acquired immune response in systemic sclerosis

Profound alterations characterize the adaptive immune response in systemic sclerosis, and several layers of evidence support a prominent role exerted by immune cellular effectors and humoral mediators in the pathogenesis of this disease. These include (i) the presence of oligoclonal T cells in tissues undergoing fibrosis consistent with (auto)antigen-specific recruitment, (ii) the preferential expansion of polarized CD4+ and CD8+ T cells producing pro-fibrotic cytokines such as IL-4 and IL-13, (iii) the presence of increased number of cells producing mediators belonging to the IL-17 family, including IL-22, which may drive and participate in inflammatory pathways involving epithelial cells as well as fibroblasts, (iv) the deficient or redirected function of T regulatory cells favoring fibrosis, and (v) the enhanced expression of CD19 and CD21 on naïve B cells, and the upregulation of co-stimulatory molecules in mature B cells, which together with the increased levels of B cell activating factor (BAFF) underlie the propensity to an exaggerated humoral response possibly favoring fibrogenesis. Despite all the progress made in understanding the features of the aberrant immune response in scleroderma, it remains unclear whether the activation of immune effector pathways ultimately drives the disease pathogenesis or rather represents a defective attempt to limit or even reverse excessive extracellular matrix deposition and progressive vasculopathy, the main hallmarks of this disease.

Photochemotherapy for systemic sclerosis: effect on clinical and molecular markers

BACKGROUND

The cutaneous changes seen in systemic sclerosis (SSc) can result in considerable patient morbidity.

AIM

We previously reported on the beneficial effect of psoralen ultraviolet A (PUVA) phototherapy in 13 patients with morphoea. We now report the findings of a study in which patients with SSc were treated with PUVA.

METHODS

Twelve patients with SSc were treated with PUVA phototherapy. The effect on cutaneous disease activity was assessed using the modified Rodnan score, and the effect on serological and immunohistochemical growth factors and adhesion molecules was also measured.

RESULTS

The median Rodnan score at baseline was 24.5 [interquartile range (IQR) 18.5-26.0]. The median number of treatments with PUVA was 24 exposures (IQR 20-26) with a median cumulative exposure of 68.3 J/cm(2) (IQR 28.6-139.8). Of the 12 patients, 11 responded well to phototherapy with a mean change in Rodnan score of 6.58 (36.98%) (P < 0.01, Wilcoxon signed ranks test). After treatment with PUVA there was a significant increase in circulating tumour necrosis factor-alpha levels in 8/12 patients (P = 0.03). In 7/12 patients there was an increase in E-selectin and vascular cell adhesion molecule, although this was not significant.

CONCLUSIONS

PUVA treatment is associated with a significant improvement in cutaneous symptoms in patients with SSc as measured by the Rodnan score (P < 0.01). Specific lymphocyte markers, adhesion molecules and cytokines are also affected by this treatment, helping to clarify further the mechanism of action of PUVA treatment and our understanding of the primary pathological process.

Systemic sclerosis: the susceptible host (genetics and environment)

It is becoming evident that several genetic factors participate in modulating susceptibility to SSc and its clinical manifestations. Some genes that specifically affect ECM metabolism and vascular function may be unique to SSc and scleroderma-related disorders; others, such as those genes involved in regulating immune tolerance, are likely shared with other autoimmune diseases. The effect of genetic variations (or polymorphisms) that are found in most of these genes taken individually will likely have only a small or modest effect on disease risk; only a few genetic variations are expected to be highly penetrant. Moreover, genetic studies in SSc have to deal with the additional issues of heterogeneous phenotypes, low disease prevalence in the general population, and an even greater paucity of multiplex families that makes traditional linkage studies difficult, if not impossible. Alternative approaches include allelic association studies, but conventional case-controls designs may be subject to selection bias and will require large sample sizes if the genes that are under investigation confer only modest (OR = 1.5-2.0) disease risk (Fig. 2). The simultaneous examination of several genes that are biologically relevant to a specific disease process to attain higher aggregate ORs, is one approach that was used in several reports that were cited in this review. The use of family-based controls, such as in the transmission-disequilibrium test (based on assessment of the transmitted or nontransmitted alleles that are associated with disease from heterozygous parents to affected offspring), would provide more robustness to spurious associations from population stratification, but is actually less powerful and efficient than case-control designs. Furthermore, for many late adult-onset diseases the effort required to obtain samples from living parents are for a variety of reasons not trivial. The success of these allelic association-based approaches depends on the identification of likely candidate disease genes (or at least markers in disequilibrium with disease genes), careful definition/ascertainment of disease phenotypes to minimize genetic heterogeneity, and for case-control designs, strategies to account for population stratification or admixture. The identification of candidate genes will be aided by rapid progress in the Human Genome Project and other genome efforts that will eventually identify all human genetic variations. Although this will lead to better understanding of the genes that might be involved in complex diseases, much work is required to understand the basic biology of how disease genotypes become clinical phenotypes. This is especially daunting in complex diseases, such as SSc, where the phenotype (including disease susceptibility and clinical presentation) is influenced by dynamic interactions between genetic variations and environment. Multi-center collaborative efforts with research paradigms that integrate genetic and environmental factors (including sociodemographic variables) will be required to elucidate the contribution of environment and genetics in the pathogenesis of SSc.

Macrophages secrete matrix metalloproteinase 9 covalently linked to the core protein of chondroitin sulphate proteoglycans

Matrix metalloproteinases (MMPs) secreted from the leukemic macrophage cell-line THP-1 have been investigated. Under serum-free conditions, this cell-line synthesizes and secretes proMMP-9, which was detected in the culture medium as a monomer of 92 kDa, and in dimeric forms, including a homodimer of approximately 225 kDa. In addition, a new heterodimer complex is described, in which proMMP-9 is covalently linked to the core protein of chondroitin sulphate proteoglycan (CSPG) through one or more disulphide bridges. After SDS-PAGE electrophoresis, at least two forms of this complex were detected, a large form in the stacking gel and a smaller form with an estimated size of 300 kDa. When the CS chains were removed by chondroitin ABC lyase treatment, heterodimers of proMMP-9/CSPG core protein of approximately 145, 127 and 109 kDa were found, based on zymography and Western blots. Since as much as 10-15 % of the total proMMP-9 secreted from THP-1 cells was covalently linked to CSPG, this association may have important implications for transport, targetting and regulation of the enzyme activity.

Integrin signaling in fibrosis and scleroderma

Integrins are a large family of heterodimeric transmembrane receptors for extracellular matrix proteins. As well as mediating cell attachment and the bulk of force transduction from the cytoskeleton, they convey signals from the extracellular matrix to the cell. alpha1beta1 and alpha2beta1 are the major collagen receptors in this family. a1beta1 provides negative feedback on collagen synthesis, whereas alpha2beta1 stimulates the synthesis of matrix metalloproteases. Each receptor modulates the signaling activity of the other to coordinate matrix synthesis and remodeling. Expression of both is reduced in scleroderma despite a paracrine environment which would be expected to upregulate them. Deficiencies in the integrins correlate with upregulated collagen synthesis and downregulated metalloprotease synthesis seen during the disease.

TNF-alpha downregulates the peroxisome proliferator activated receptor-alpha and the mRNAs encoding peroxisomal proteins in rat liver

We have studied the effects of TNF-alpha on the mRNAs coding for the peroxisome proliferator activated receptor alpha (PPAR-alpha), and for catalase (Cat), acyl-CoA oxidase (AOX), multifunctional enzyme (PH), and beta-actin in rat liver. Total RNA was isolated from livers of male SD-rats 16 h after administration of a single dose of 25 microg TNF-alpha and mRNAs were analyzed by a novel dot blot RNase protection assay. The mRNAs for PPAR-alpha and for Cat, AOX and PH were significantly reduced by TNF-treatment. In addition, the level of PPAR-alpha protein was also decreased after TNF. In contrast, the mRNA for beta-actin was markedly increased implying that the effect of TNF on PPAR-alpha and the peroxisomal mRNAs is highly selective. This effect may have important implications in perturbation of the lipid metabolism induced by TNF-alpha.

Effect of adenosine analogues on the expression of matrix metalloproteinases and their inhibitors from human dermal fibroblasts

The effect of the cytostatic and antiviral adenosine analogues 3-deazaadenosine (c3Ado) and 3-deaza-(+/-)-aristeromycin (c3Ari) on human skin fibroblasts was studied. Variables examined were cell morphology, viability, DNA fragmentation, expression of matrix metalloproteinases (MMPs) and matrix metalloproteinase inhibitors (TIMPs). None of these variables were changed when cells were exposed to c3Ari concentrations ranging from 10(-5) to 10(-3) M or 10(-5) M c3Ado. However, large changes in cell morphology, viability and expression of MMPs and MMP inhibitors occurred when fibroblasts were treated with 10(-4) or 10(-3) M c3Ado. Cells rounded up, shrank in volume, some detached and viability was lost without any detectable fragmentation of DNA. These changes in morphology and viability were associated with a differentiated expression of MMPs and MMP inhibitors. A large increase in collagenase activity occurred, and depending on the concentration of the adenosine analogue and the length of treatment, this change in activity could be shown to be due to one or a combination of the following factors: an increased synthesis of the collagenase protein, a decreased production of TIMP-1 or an increased activity of the collagenase superactivator, stromelysin. In contrast to this, treatment with c3Ado resulted in a decreased gelatinase activity, which in part could be attributed to an increased production of an inhibitor that seemed to affect gelatinase but not collagenase. The cellular changes induced by c3Ado seemed to reflect some of the alteration in the metabolic machinery that appears during a drug-induced or programmed/controlled death of a dermal cell. The different effects exerted by these two adenosine analogues on dermal fibroblasts can at least in part explain why c3Ado have previously been shown to be more toxic than c3Ari in animal models.

The transcription of human alpha 1(I) procollagen gene (COL1A1) is suppressed by tumour necrosis factor-alpha through proximal short promoter elements: evidence for suppression mechanisms mediated by two nuclear-factorbinding sites

Recent studies have demonstrated that tumour necrosis factor-alpha (TNF-alpha) decreases alpha 1(I) procollagen gene (COL1A1) expression in cultured human dermal fibroblasts. The purpose of this study was to analyse the transcriptional control of COL1A1 by TNF-alpha. Cultured human dermal fibroblasts were transiently transfected with plasmids containing 5' flanking sequences of COL1A1 fused to the chloramphenicol acetyltransferase (CAT) gene, and were incubated for 48 h in medium with or without TNF-alpha. TNF-alpha inhibited the CAT activity of fibroblasts transfected with plasmids containing 2.3 kb of 5' flanking sequences of COL1A1, whereas the activity of control plasmids containing the herpes simplex thymidine kinase promoter gene (pBLCAT) was unaltered. A series of deletion constructs of various small substitution mutations of the COL1A1 5' flanking region fused to the CAT gene were also transfected, and CAT activity was measured after incubation with TNF-alpha. TNF-alpha suppressed COL1A1 promoter activity through proximal short promoter elements containing only 107 bp. Short substitution mutations between -101 and -97 bp or between -46 and -38 bp abolished TNF-alpha suppression of COL1A1 promoter activity. DNA-protein complex formation was observed involving both sites in gel retardation assays. These results suggest that TNF-alpha suppressed COL1A1 promoter activity through elements located between -101 and -97 bp and between -46 and -38 bp of the COL1A1 promoter, and that the suppression involved DNA-protein interactions.

Butylated hydroxyanisole blocks the inhibitory effects of tumor necrosis factor-alpha on collagen production in human dermal fibroblasts

Tumor necrosis factor-alpha (TNF-alpha) has been demonstrated to selectively decrease the production of type I and type III collagens in human dermal fibroblasts. The effects of the commonly used food antioxidants, butylated hydroxyanisole (BHA), butylated hydroxytoluene (BHT), alpha-tocopherol, propyl gallate, superoxide dismutase (SOD), and catalase on TNF-alpha-induced growth enhancement and collagen metabolism were evaluated in the present study. BHA at concentrations of both 5 x 10(-5) and 10(-4) M inhibited cell proliferation and DNA synthesis induced by 10 ng/ml TNF-alpha in human dermal fibroblasts, while other antioxidants had minimal effects. Further, BHA (5 x 10(-5) M and 10(-4) M) significantly blocked TNF-alpha-induced decreases in collagen synthesis. These results suggest that antioxidants such as BHA may be involved in the modulation of collagen synthesis by TNF-alpha in human dermal fibroblasts.

Role of cytokines in controlling connective tissue gene expression

Recently, the role of cytokines in controlling gene expression of connective tissue components has been increasingly emphasized. Many cytokines have been shown to have specific effects on gene expression of connective tissue components, and the roles of cytokines in controlling connective tissue metabolism during wound healing and in fibrosis have increasingly been discussed. In this article, the effects of cytokines on regulation of gene expression of connective tissue components, especially of type I collagen were described. We analysed transcriptional control of the alpha 1(I) collagen gene by TNF-alpha by means of DNA mediated transfection experiments using recombinant plasmids in which the promoter region of the human alpha 1(I) collagen had been fused to the chloramphenicol acetyl-transferase (CAT) gene, in human dermal fibroblasts. It was found that TNF-alpha reduced alpha 1(I) collagen transcription through at least up to -107 bp upstream of the human alpha 1(I) collagen promoter gene in dermal fibroblasts.