Starving the synovium: angiogenesis and inflammation in rheumatoid arthritis.
Explore the source record for details and available documents.
Biomedical subjects
Publications and source records attributed to G S Firestein.
Explore the source record for details and available documents.
Animal models of arthritis provide a window into the pathogenesis of inflammatory joint disease. In addition, they can function as an in vivo laboratory for evaluating new and creative treatments. The potential usefulness of novel treatment strategies, including recombinant cytokine therapy, gene delivery, and regulation of apoptosis, have been explored and offer clues to the feasibility of human therapeutic trials. New models of arthritis have also been developed in an effort to increase the predictive capacity for efficacy in rheumatoid arthritis.
OBJECTIVE: To determine the expression and regulation of nuclear transcription factors AP-1 and NF-kappaB in rheumatoid arthritis and in collagen-induced arthritis in mice. METHODS: AP-1 and NF-kappaB expression and function were determined in RA, OA and normal synovial tissue by electrophoretic mobility shift assay (EMSA) and immunohistochemistry. The kinetics of transcription factor expression were then examined in collagen-induced arthritis (CIA) in mice. EMSAs were performed with the nuclear extracts obtained from paws of CIA mice from 10 to 45d after immunization to determine AP-1 and NF-kappaB binding activity. The expression of collagenase-3 (MMP13) and stromelysin (MMP3) mRNA was examined by northern blot analysis. RESULTS: Immunohistochemistry showed that NF-kappaB expression was increased in both RA and OA synovial intimal lining. AP-1 components Jun and Fos were also present in the intimal lining and was significantly greater in RA than OA. The DNA binding activities of both AP-1 and NF-kappaB were significantly higher RA patients compared with OA. In CIA, AP-1 and NF-kappaB expression increased by day 20, which was 1-2 weeks before onset of clinical arthritis. However, collagenase and stromelysin gene expression did not increase until day 35. CONCLUSION: The DNA binding activity of AP-1 and NF-kappaB are markedly increased in both CIA and RA. In CIA, activation of AP-1 and NF-kappaB precede both clinical arthritis and metalloproteinase gene expression. NF-kappaB expression correlated better than AP-1 with metalloproteinase expression.
Recent studies show that 1) the p53 tumor suppressor protein is overexpressed by rheumatoid arthritis (RA) synovium and fibroblast-like synoviocytes (FLS) and 2) somatic mutations previously identified in human tumors are present in RA synovium and FLS. We have hypothesized that abnormalities in p53 can contribute to chronic destructive RA synovitis. To understand the functional consequences of p53 abnormalities in FLS, RA and normal FLS expressing wild-type p53 were transduced with a retroviral vector encoding the human papilloma virus 18 E6 gene, which inactivates endogenous p53 protein. Three RA and one normal FLS lines were infected with recombinant retrovirus encoding the neomycin resistance gene (neo) or E6+neo. FLS proliferation, apoptosis, and invasion was studied in E6, neo, and uninfected parental strains (PS). The growth rate for E6 was significantly increased with a sixfold increase in cell number after 7 days compared with a twofold to threefold increase in neo and PS. When FLS were treated with cytokines, proliferative response of E6, neo, and PS to interleukin-1 and transforming growth factor-beta were similar. However, response to platelet-derived growth factor was significantly greater in E6 FLS compared with neo or PS. Apoptosis was studied by incubating FLS with sodium nitroprusside as a source of nitric oxide or hydrogen peroxide for 8 hours and examining DNA fragmentation and E6 cells were significantly less susceptible to cell death. In addition, E6 FLS were more invasive into cartilage extracts than neo or PS using an in vitro cell invasion assay. These data suggest that p53 is a critical regulator of FLS proliferation, apoptosis, and invasiveness. Abnormalities of p53 function might contribute to synovial lining expansion and joint destruction in RA.
OBJECTIVE: To evaluate the feasibility of direct in vivo gene transfer in an animal model of arthritis using a retroviral vector. METHODS: The timing and dose of retroviral vector was examined using very high titer retroviral vector (> or = 10(9) CFU) in rat adjuvant arthritis. Retroviral vector expressing beta-galactosidase (beta-gal) or vehicle alone was injected into the right ankle of rats with adjuvant arthritis. Ankles were injected either on Day 7 (pre-arthritis), Day 10 (early arthritis), Day 15 (accelerating arthritis), or Day 28 (chronic arthritis) after adjuvant immunization. Joints were harvested 3 days later and extracts were assayed for beta-gal activity. RESULTS: Synovial beta-gal expression was minimal in the Day 7 group and elevated in the Day 10, Day 15, and Day 28 groups. Gene transfer with retroviral vector did not exacerbate the local inflammatory response. Minimal or no beta-gal expression was observed in the contralateral uninjected paw or in the spleen, lung, liver, and kidneys. Frozen sections of retroviral vector injected joints were stained with X-gal and revealed transduced cells in the lining and superficial sublining layers. To determine the longevity of gene expression, ankle joints were injected with vector on Day 15 post-adjuvant, harvested, and assayed for beta-gal activity for up to 49 days after injection. Expression of the enzyme peaked from Day 3 to 7 and was still readily detected up to 49 days after retrovirus infection. CONCLUSION: This is the first report of successful direct in vivo gene transfer in the rat adjuvant arthritis model using a retroviral vector. Appropriate timing of administration and very high titer retroviral vector preparations are key determinants of adequate gene transduction.
OBJECTIVE: We recently identified mercaptoethylguanidine (MEG) as an antiinflammatory agent with a combined mechanism of action. Its effects include inhibition of the inducible isoform of nitric oxide synthase (iNOS), scavenging peroxynitrite, a cytotoxic oxidant species produced from nitric oxide (NO) and superoxide, and inhibition of cyclooxygenase (COX). We investigate the effect of MEG in collagen induced arthritis (CIA). METHODS: Syngeneic LOU rats were immunized with native type II collagen on Day 0. After clinical signs of arthritis developed on Day 10, treatment with MEG was initiated (30 mg/kg ip tid) and continued until sacrifice on Day 28. Serum nitrite/nitrate was measured in control animals, at arthritis onset and 2 days after the start of MEG treatment. Clinical scores were obtained daily. At Day 28, radiographic scores were obtained, and joints were harvested for the measurement of mRNA for tumor necrosis factor-alpha (TNF-alpha), collagenase, and stromelysin. RESULTS: Serum nitrite/nitrate increased from 7.9+/-0.7 mM (baseline) to 13.5+/-2.6 at arthritis onset (p < 0.05). Within 48 h of MEG treatment, nitrite/nitrate levels fell to 7.2+/-1.1 (p < 0.05). By Day 28, clinical arthritis scores (measured on a scale of 0-8) were 7.1+/-0.6 in the vehicle group compared to 1.4+/-0.6 in the MEG treated group (p < 0.0001). Radiographic scores (scale 0-6) on Day 28 were reduced from 4.9+/-0.6 to 0.6+/-0.4 (p < 0.0002) by MEG treatment. MEG reduced the synovial expression of mRNA for TNF-alpha, collagenase, and stromelysin by 72, 67, and 52%, respectively. CONCLUSION: These data show that MEG has beneficial effects on established CIA. The mechanism of action may be related to inhibition of synovial iNOS expression or activity, inhibition of COX, scavenging of peroxynitrite, with subsequent inhibition of angiogenesis, metalloproteinase, and TNF-alpha expression.
The factors that regulate the perpetuation and invasiveness of rheumatoid synovitis have been the subject of considerable inquiry, and the possibility that nonimmunologic defects can contribute to the disease has not been rigorously addressed. Using a mismatch detection system, we report that synovial tissue from the joints of severe chronic rheumatoid arthritis patients contain mutant p53 transcripts, which were not found in skin samples from the same patients or in joints of patients with osteoarthritis. Mutant p53 transcripts also were identified in synoviocytes cultured from rheumatoid joints. The predicted amino acid substitutions in p53 were identical or similar to those commonly observed in a variety of tumors and might influence growth and survival of rheumatoid synoviocytes. Thus, mutations in p53 and subsequent selection of the mutant cells may occur in the joints of patients as a consequence of inflammation and contribute to the pathogenesis of the disease.
Explore the source record for details and available documents.
Ischemic preconditioning reduces post-ischemic myocardial injury by activating myocellular adenosine A1 receptors. Adenosine A3 receptors have also been implicated but there is no evidence for A3 receptors in cardiac myocytes. The aim of this study was to develop a model of preconditioning in isolated cardiac myocytes to evaluate the role of the adenosine A1 and A3 receptors in preconditioning-induced protection from ischemic injury. Reverse transcription polymerase chain reaction (PCR) was also employed to establish the presence of adenosine A3 receptors in these cells. In the preconditioning studies, ischemic injury was simulated by exposing isolated rabbit myocytes (placed in the cell chamber and paced at l Hz) to buffer containing (in mM) 2'-deoxyglucose (20), NaCN (1), Na (+)-lactate (20), KCl (10) at pH 6.6 (37 degrees C). Changes of diastolic and systolic cell length were monitored with an optical-video edge imaging system, and hypercontracture was assessed as an index of irreversible cell injury. Preconditioning (2 min brief ischemia and 15 min reperfusion) significantly reduced cell injury resulting from a subsequent prolonged ischemia (10 min) and reperfusion (15 min), as indicated by a reduction in the incidence of cell hypercontracture from 67 +/- 6% to 29 +/- 5% (P < 0.001). Preconditioning-induced cardioprotection was only partially blocked by a maximally effective concentration (100 nM) of the adenosine A1 receptor antagonist 1,3-dipropyl-8-cyclopentylxanthine (DPCPX) (cell hypercontracture = 43 +/- 3%, P < 0.05 vs. control) but completely blocked by either the combination of DPCPX (100 nM) with the adenosine A1/A3 receptor antagonist DPCPX +8-(4-carboxyethylphenyl)-1,3-dipropylxanthine (BWA1433; 1 microM) or the non-selective adenosine receptor antagonist, 8-(p-sulfophenyl)theophylline (8-SPT; 100 microM) (cell hypercontracture = 64 +/- 4%, 59 +/- 5%, respectively; P = NS vs. control). In non-hypercontractured myocytes, preconditioning also substantially enhanced the recovery of the contractile amplitude and, similarly, this effect was only partially blocked by DPCPX but completely blocked by either the combination of DPCPX with BWA1433, or 8-SPT. These studies suggest that preconditioning protects isolated cardiac myocytes from ischemic injury independent of other cell types, and that maximal preconditioning-induced cardioprotection requires activation of both adenosine A1 and A3 receptors. Reverse transcription-PCR using primers for the rabbit receptor provide evidence for the presence of adenosine A3 receptors in these cells.
OBJECTIVE: To characterize the transcriptional and posttranscriptional regulation of collagenase-1 by adenosine receptor stimulation in interleukin-1 (IL-1)-stimulated fibroblast-like synoviocytes (FLS). METHODS: FLS were stimulated with IL-1 and either the nonselective adenosine agonist 5'-N-ethylcarboxamidoadenosine (NECA) or the adenylate cyclase activator forskolin. Electrophoretic mobility shift assays were performed to determine AP-1 and cAMP-responsive element binding protein (CREB) activation. Transcriptional activation was determined by transfecting HS68 dermal fibroblasts with a collagenase-chloramphenicol acetyltransferase construct. Finally, collagenase messenger RNA (mRNA) half-life was determined by activating cells in the presence of IL-1, IL-1 + NECA, or IL-1 + forskolin and culturing cells in the presence of actinomycin D. RESULTS: NECA and forskolin had no effect on AP-1 activation, c-jun or c-fos gene expression, or CREB phosphorylation. IL-1 markedly increased collagenase promoter activity, and neither NECA nor forskolin blocked this action. Studies of mRNA half-life showed that both NECA and forskolin decreased the half-life of collagenase mRNA in IL-1-stimulated FLS and HS68 cells. CONCLUSION: The findings of this study demonstrate that NECA and forskolin decrease collagenase gene expression in FLS and dermal fibroblasts due to enhanced mRNA degradation.
Adhesion of leukocytes to the vascular endothelium hallmarks a key event in neutrophil-mediated organ injury after ischemia-reperfusion. The autacoid adenosine has been shown to inhibit activated neutrophil function and to interfere with leukocyte-endothelial adherence. Its therapeutic use in ischemia-reperfusion, however, has been limited by severe cardiovascular side effects. We therefore investigated the effects of the adenosine kinase inhibitor GP515 in vivo on hepatic leukocyte-endothelial interactions in a rat model of hemorrhagic hypotension and resuscitation, using intravital microscopy. Rats were pretreated with either GP515 (0.25 mg/kg) or saline in a randomized and blinded manner and subjected to pressure-controlled hemorrhagic hypotension at a mean arterial pressure of 40 mmHg for 60 min followed by 5 h of resuscitation. Five hours after resuscitation in saline-treated animals, firm leukocyte-sinusoidal adhesion was strongly enhanced in the periportal and midzonal sublobular regions, and sinusoidal diameters were also markedly reduced. Compared with saline treatment, GP515 significantly attenuated shock and resuscitation-induced leukocyte adhesion in both sublobular regions. Moreover, although GP515 did not significantly affect macrohemodynamical and hematological parameters, it enlarged narrowed sinusoidal diameters and tended to improve sinusoidal blood flow. We propose that the adenosine-regulating agent GP515 has a therapeutic potential to attenuate ischemia-reperfusion-induced inflammation by capitalizing on the beneficial anti-inflammatory effects of endogenous adenosine.
A distinctive cell was identified from sites of rheumatoid arthritis cartilage injury. Similar cells are not found in lesions of osteoarthritis cartilage. We have designated them as pannocytes (PCs). Their rhomboid morphology differs from the bipolar shape of fibroblast-like synoviocytes or the spherical configuration of primary human articular chondrocytes. Chondrocytes are short-lived, whereas the original PC line grew for 25 passages before becoming senescent. Features in common with cultured primary chondrocytes include maximal proliferation in response to transforming growth factor-beta a catabolic response to interleukin-1 beta, collagenase production, and mRNA for the induced lymphocyte antigen and inducible nitric oxide synthase. Despite the presence of the inducible nitric oxide synthase message, PCs do not produce NO either constitutively or when cytokine stimulated. Each of the mesenchymal cells, fibroblast-like synoviocytes, primary chondrocytes, and PCs have the gene for type I collagen, but the type II collagen gene is detected only in primary chondrocytes. PCs can be distinguished from fibroblast-like synoviocytes and primary chondrocytes by their morphology, bright VCAM-1 staining, and growth response to cytokines and growth factors. Their prolonged life span in vitro suggests that PCs might represent an earlier stage of mesenchymal cell differentiation, and they could have a heretofore unrecognized role in rheumatoid arthritis joint destruction.
OBJECTIVE: Extracellular matrix components and cell adhesion molecules play a role in the pathogenesis of rheumatoid arthritis (RA). Interaction between the integrin very late antigen-4 (VLA-4) and the connecting segment-1 (CS-1) fibronectin (FN) isoform may contribute to lymphocyte interaction in RA synovium. We examined both mRNA and protein expression of CS-1 FN in inflamed synovium, and VLA-4 expression in synovial tissue and on cultured fibroblast-like synoviocytes from patients with RA. METHODS: Snap frozen synovial tissue specimens of 10 patients with RA and 4 patients with osteoarthritis were examined for expression of CS-1 FN mRNA and protein by in situ hybridization and immunohistochemistry. VLA-4 expression of synovial fibroblasts and in synovial tissue was evaluated by flow cytometry and immunohistochemistry. RESULTS: CS-1 FN mRNA was detected in RA lining layer synoviocytes, around terminal vessels, and in endothelial cells. Double labeling revealed that most lining synoviocytes expressing CS-1 FN mRNA were CD68 negative. VLA-4 was found in RA synovial fibroblasts, sublining mononuclear cells, and lymphoid aggregates. CONCLUSION: Our findings suggest that expression of CS-1 FN may partially correlate with cell proliferation in the RA lining layer. VLA-4 was found in RA synovial lining, as well as on cultured synovial fibroblasts. Thus, VLA-4/CS-1 FN interaction may facilitate lymphocyte interaction and synovial proliferation in RA.
The effect of spinal adenosine receptor ligation on peripheral leukocyte accumulation was studied in two rat models of inflammation. Neutrophil infiltration into dermal inflammatory sites was signficantly reduced by adenosine A1 receptor agonists injected through intrathecal catheters. These effects were reversed by N-methyl-D-aspartate (NMDA), and were mimicked by (+/-)-2-amino-5-phosphonopentanoic acid (AP-5), a glutamate NMDA receptor antagonist. Peripheral adenosine levels, as measured in air pouch exudates, decreased markedly in inflamed pouches but remained near normal after intrathecal treatment with AP-5. Moreover, the antiinflammatory effects of intrathecal A1 receptor agonists and AP-5 were reversed by an adenosine A2 receptor antagonist administered intraperitoneally. Hence, central NMDA receptor activity can regulate neutrophil accumulation in peripheral inflammatory sites by reducing local levels of adenosine, an antiinflammatory autacoid which inhibits neutrophil function through A2 receptor activation. This represents a previously unknown pathway by which the central nervous system influences inflammatory responses.
Adenosine agonists inhibit TNF-alpha production in macrophage and monocytes, but the mechanism is unknown. Therefore, we studied the human macrophage cell line U937 to determine the adenosine receptor subtypes responsible and the intracellular signaling mechanisms involved. The A1/A3 agonist N6-(4-amino-3-iodobenzyl)adenosine (I-ABA) decreased LPS-stimulated TNF-alpha protein production by 79 +/- 5% (p = 0.003). The mechanism was pretranslational, as adenosine receptor stimulation caused a marked decrease in TNF-alpha mRNA. IL-1 beta, IL-6, and IL-8 mRNA were not changed by adenosine agonists. The rank order of agonists as TNF-alpha inhibitors suggested that the A3 receptor might be involved (N6-(3-iodobenzyl)-9-[5-(methylcarbamoyl)-beta-D-ribofuranosyl] adenosine > 2-chloroadenosine > or = I-ABA > N6 benzyl 5'-N-ethylcarboxamidoadenosine (NECA) > NECA > CGS21680 > N6-cyclohexyladenosine), and this was supported by the fact that a mixed A1/A3 antagonist (xanthine amine congener) reversed the effect, whereas A1-specific (1,3-dipropyl-8-cyclopentylxanthine) and A2-specific (3,7-dimethyl-1-propargylxanthine) antagonists did not. Receptor signaling did not involve cAMP or protein kinase A, nor did it alter the activation and binding characteristics of the transcription factor NF-kappa B. However, the composition of the AP-1 transcription complex was altered by I-ABA. These data suggest that stimulation of the A3 adenosine receptor can alter the cytokine milieu by decreasing TNF-alpha. Adenosine agonists or adenosine regulating agents have potential therapeutic uses in acute and chronic inflammatory diseases.
A cDNA encoding variant form of the A3 adenosine (Ado) receptor was isolated from rat by reverse transcription of brain mRNA followed by PCR. The full-length receptor (A3i) cDNA encodes 337 amino acids and shares complete sequence identity with the rat A3 Ado receptor, except for the presence of a seventeen amino acid insert located in the second intracellular domain. In contrast to the rat A3 receptor, stable expression of A3i in CHO cells resulted in poor coupling to Gi proteins. Analysis of receptor transcripts by RT-PCR suggests that the A3 Ado receptor mRNAs are products of alternative splicing. Sequence analysis of A3 genomic DNA identified a 1.7 kb intron that is likely alternatively spliced to produce the A3 and A3i receptors.
OBJECTIVE: To characterize the regulation of matrix metalloproteinases (MMPs) by adenosine. METHODS: Cultured fibroblast-like synoviocytes (FLS) were stimulated with interleukin-1 (IL-1) in the presence or absence of adenosine receptor agonists. Immunoreactive MMPs were measured using specific enzyme-linked immunosorbent assays, and gene expression was assessed by Northern blot analysis. RESULTS: The nonselective adenosine receptor agonist 5'-N-ethylcarboxamidoadenosine (NECA) decreased collagenase production by IL-1-stimulated synoviocytes from 196 +/- 28 ng/ml (mean +/- SEM) to 66 +/- 9 ng/ml (P < 0.001). There was minimal effect on stromelysin production (decrease from 107 +/- 16 ng/ml to 97 +/- 15 ng/ml). Selective adenosine receptor agonists implicated the A2b adenosine receptor in this activity, and reverse transcriptase-polymerase chain reaction studies confirmed that FLS express this receptor. Northern blot analysis demonstrated that the mechanism of action was pre-translational since NECA decreased collagenase, but not stromelysin or tissue inhibitor of metalloproteinases 1 (TIMP-1), messenger RNA levels. Cyclic AMP levels were increased by NECA, and a direct adenylate cyclase activator (forskolin) also suppressed collagenase gene expression. These data suggest that cAMP mediates the inhibitory effect of NECA on collagenase production. CONCLUSION: Stimulation of the A2b receptor on FLS decreases collagenase gene expression, with little or no effect on stromelysin and TIMP-1. The combination of antiinflammatory and MMP-regulating properties of adenosine or adenosine-regulating agents suggest that treatment based on this approach might be useful in rheumatoid arthritis.
Explore the source record for details and available documents.