de | fr | en Imprimer
Genetic analysis of a 3D in vitro model of the rheumatic pannus tissue
Biotechnologie Institut Thurgau an der Universität Konstanz, 8274 Tägerwilen, Switzerland
Keywords: human; pannus; pharmacology; arthritis; cell cultures: organ-specific; reduction; replacement
Duration: 1 year Project Completion: 2002
Background and Aim
Rheumatoid arthritis is a chronic degenerative disease of unknown etiology that affects about 1-2 % of the human population. Up until now no cure for the disease has been found; present treatments can only slow down the process of joint destruction. Crucial for the destructive joint processes is the de novo formation of pannus tissue. Due to the complexity of the disease, a wide variety of animal models has been developed to investigate the etiology and mechanisms in rheumatoid arthritis resulting in the development of new treatments of the disease.
Using single cell suspensions from synovial fluid cells of arthritis patients, we observed differentiation of three-dimensional tissues in vitro. This new model of pannus-like tissue (PLT) might be useful to study pannus tissue formation and differentiation. In the PLT cultures, we observed two cell types, fibroblast-like and macrophage-like cells, defined
by their distinct morphology and major histocompatibility complex (MHC) by human leukocyte antigen (HLA) class II expression. We could discriminate several intermediate steps of differentiation which finally led to 3D villi-like structures. Secretion of interferon gamma, interleukin-10, and tumor necrosis factor alpha was measured in the culture supernatants.
Using methotrexate at various concentrations, the growth of PLT could be inhibited. We describe definite intermediate steps of differentiation. The present approach could be a suitable model for the in vitro study of pannus tissue formation.
Method and Results
Preparation of cells:
Synovial fluid cells (SFC) were obtained from patients with reactive and rheumatoid arthritis treated at local hospitals and rheumatology practices. In all instances, synovial tapping was therapeutically indicated and patients gave their informed consent. The synovial fluid samples were diluted 1:2 in phosphate-buffered solution (PBS), and mononuclear cells were depleted by Ficoll density gradient centrifugation. The pelleted cells were isolated, washed, and counted.
Tissue culture and cell lines:
Synovial fluid cells were taken into tissue culture at 106 cells/ml in standard tissue culture medium. Iscove’s modified Dulbecco’s or Roswell Park Memorial Institute 1640 medium were supplemented with antibiotics, glutamine, and 10% fetal calf serum. The cell lines were maintained in the same medium. K4 cells are SV40-transformed synovial fibroblast , U937, and MonoMac 6 are monocytic cell lines (ATCC, Rockwell, USA). Methotrexate was used at the indicated concentration.
For immunohistochemistry, cells were grown on glass slides. Tissues were cut out of the culture using Pasteur glass pipettes and flash frozen in liquid nitrogen. Five-micrometer sections were used for staining. Cells/tissues were fixed in 100% methanol at -20°C for 60 min and air-dried. Endogenous peroxidase activity was minimized using 0.1% phenylhydrazin-hydrochloride with PBS. Then cells were blocked with PBS and 1% bovine serum albumin (BSA) for 60 min. Primary staining was done with the monoclonal antibody (mAB) L243 coupled to biotin for 20 min at room temperature. After washing in PBS and 0.05% Tween 20, cells/tissue were blocked with PBS and 1% BSA, and streptavidin-horseradish peroxidase oxidase was applied. Staining was done with amino-9-ethylcarbazol (Sigma A5754) in dimethyl formamide. For control, cells were incubated with antimouse mABs conjugated to biotin. Counterstaining was done with hematoxylin.
Cytokine enzyme-linked immunosorbent assays:
Sandwich enzyme-linked immunosorbent assays (ELISAs) were performed according to standard protocols. The following antibodies were used for the detection of cytokines: 9D7- and 12G8-biotin for IL-10, H34A- and H33-biotin for tumor necrosis factor alpha (TNFalpha), and M700AE- and M701B-biotin for interferon gamma (IFN-gamma). Recombinant cytokines were used to generate standard curves. Unlabeled antibodies were used for coating and biotinylated antibodies for detection. All reagents were purchased from Endogen or Pharmingen (Heidelberg, Germany).
Conclusions and Relevance for 3R
The PLT developed from single cell suspensions derived from synovial fluids. The development occurred in several steps toward 3D structures which were reminiscent of the pannus tissue’s villi. In cultures in which only the growth of fibroblast-like cells without attached macrophage-like cells was observed, no development towards PLT occurred. Though it is reported in the literature that synovial cells tend to grow without contact inhibition, in our system’s limited association of macrophage-like cells with fibroblast-like cells, we observed no 3D growth. This phenotype was also found in the methotrexate-treated cultures. We therefore conclude that interactions of both cell types are necessary for the formation of PLT. Development and growth of cells was only observed in the initial phase up to 4 weeks, followed by growth retardation. At this point the project had to be interrupted because of the lack of synovial punctures, due to the introduction of significantly better therapies for patients suffering from Rheumatoid Arthritis (eg high dose methotrexate, anti TNF-alpha antibodies). The advanced therapies resulted in less punctures, a dramatically decreased volume of the punctures and bad conditions of the synovial fluid cells. However, using methotrexat, we could give the proof of principle that the in vitro pannus-like tissue model might be able to replace animal models.
Solomon S, Masilamani M, Mohanty S, Schwab JE, Boneberg EM, Illges H. (2003)
Generation of three-dimensional pannus-like tissues in vitro from single cell suspensions of synovial fluid cells from arthritis patients. Rheumatol Int. (in press).