Endothelial adhesion molecules facilitate the entry of leukocytes into inflamed tissues. This in turn promotes neovascularization, a process central to the progression of rheumatoid arthritis, tumor growth and wound repair. Here we test the hypothesis that soluble endothelial adhesion molecules promote angiogenesis. Human recombinant soluble E-selectin and soluble vascular cell adhesion molecule-1 induced chemotaxis of human endothelial cells in vitro and were angiogenic in rat cornea. Soluble E-selectin acted on endothelial cells in part through a sialyl Lewis-X-dependent mechanism, while soluble vascular cell adhesion molecule-1 acted on endothelial cells in part through a very late antigen (VLA)-4 dependent mechanism. The chemotactic activity of rheumatoid synovial fluid for endothelial cells, and also its angiogenic activity, were blocked by antibodies to either soluble E-selectin or soluble vascular cell adhesion molecule-1. These results suggest a novel function for soluble endothelial adhesion molecules as mediators of angiogenesis.

Objective. Angiogenesis is an integral component of the vasculoproliferative phase of rheumatoid arthritis (RA). Recently, a heparin‐binding cytokine termed hepatocyte growth factor (HGF), or scatter factor (due to its ability to disperse cohesive epithelial colonies), was described. We conducted this study to investigate the hypothesis that this cytokine was present in the milieu of the inflamed joint, and that it contributed to the chemotaxis of endothelial cells in the synovial tissue. Methods. We examined synovial fluid, synovial tissue, and peripheral blood from 91 patients with RA and other arthritides. We used 83 total samples in an enzyme‐linked immunosorbent assay to quantitate the HGF in synovial fluids and peripheral blood. To determine whether the HGF was biologically active, an epithelial scatter factor assay was performed. Immunohistochemical analysis was used to determine localization in synovial tissues. To define a function for synovial HGF, we preincubated rheumatoid synovial fluids with neutralizing anti‐HGF and measured the ability of these synovial fluids to induce endothelial chemotaxis. Results. Synovial fluid from patients with RA contained a mean ± SEM HGF concentration of 2.0 ± 0.3 ng/ml, while synovial fluid from patients with other arthritides (including inflammatory arthritis) contained 2.4 ± 0.7 ng/ml HGF. Osteoarthritis (OA) patient samples contained the smallest quantities of synovial fluid HGF at 0.9 ± 0.1 ng/ml. RA synovial fluid contained significantly more HGF than did RA peripheral blood (1.1 ± 0.2 ng/ml) (P < 0.05). Rheumatoid synovial fluids induced more scattering of cells than did OA synovial fluids, suggesting a role for this cytokine in rheumatoid joint destruction. Interleukin‐1β induced expression of rheumatoid synovial tissue fibroblast antigenic HGF and scatter factor activity. Immunohistochemically, HGF, as well as the HGF receptor (the met gene product), localized to significantly more rheumatoid synovial tissue lining cells than normal lining cells (P < 0.05). Both HGF and its receptor immunolocalized to subsynovial macrophages as well. Levels of synovial tissue immunoreactive HGF correlated positively with the number of synovial tissue blood vessels. Anti‐HGF neutralized a mean of 24% of the chemotactic activity for endothelial cells found in 10 rheumatoid synovial fluid samples. Conclusion. These results indicate that synovial HGF may contribute to the vasculoproliferative phase of inflammatory arthritides such as RA, by inducing HGF‐mediated synovial neovascularization. These findings point to a newly described role for HGF in the fibroproliferative phase of RA‐associated synovitis.

Objective. To examine cytokine and chemokine production during the evolution of rat adjuvant-induced arthritis (AIA), a model of rheumatoid arthritis. Methods. Clinical and laboratory assessment of the course of AIA was performed over a 47-day period. Levels of the cytokines tumor necrosis factor (TNF), interleukin-1 (IL-1), and IL-6, as well as levels of the chemokines macrophage inflammatory protein 1 (MIP-1) and JE, the murine homolog of monocyte chemoattractant protein 1, were determined by enzyme-linked immunosorbent assay in the sera and joints of AIA and control rats. Synovia from AIA rats were (immuno)histochemically analyzed. Results of cytokine and chemokine measurements were correlated with clinical and laboratory markers of inflammation and histology. Results. Early (before day 14 post adjuvant injection) and later phases of AIA could be distinguished. Cytokine and chemokine production was increased in AIA versus control rats. The production of TNF, IL-1, MIP-1, and, as determined earlier, epithelial neutrophil-activating peptide 78-like protein was abundant prior to and during the course of AIA, while that of IL-6 and JE was elevated in the late phase of AIA. Cytokine and chemokine levels were correlated with the clinical symptoms of arthritis and blood neutrophil counts. Joint levels of IL-1 showed correlation with synovial lining proliferation and neutrophil ingress into AIA synovium. Conclusion. Cytokines and chemokines are involved in the clinical, laboratory, and histologic changes underlying AIA. The production of these mediators may be temporally and spatially regulated. These findings may be important for the optimal timing of cytokine and chemokine targeting.

Objective. To examine adhesion molecule expression during the progression of inflammation in a rheumatoid arthritis model of adjuvant-induced arthritis (MA) in rats.Methods. Immunohistochemical analysis was used to determine the distribution of the following adhesion molecules: lymphocyte function-associated antigen 1 (LFA-1; CDlldCDlS), Mac-1 and p150/95 (CDllbc/CDlS), intercellular adhesion molecule 1 (ICAM-l), and CD44 in tissue sections from the ankle joints of rats with AIA. Control animals and those with AIA were killed at intervals over a 54-day period after injection with mineral oil and Mycobactenurn butyricum, respectively.Results. CD44 and LFA-1 were expressed on lymphocytes, macrophages, and synovial (ST) lining cells. CD44 expression on macrophages was found to be increased compared with control animals by day 18, and was significantly increased by day 41. CD44 expression on lymphocytes significantly increased earlier, on days 11-18. Increased LFA-1 expression on macrophages occurred late, on day 41. LFA-1 expression on lymphocytes was significantly increased on days 25,47, and 54. ST lining cells exhibited two distinct periods of increased expression, one early, on days 11-25 and one later, on days 41-54. CDllblc was expressed on macrophages and ST lining cells, showing a significant increase on AIA rat ST lining cells compared with control animals on day 4. No differences in ICAM-1 expression on

Previous work by our laboratory has described the presence and widespread distribution of a PRL-like immunoreactive protein in brain. The persistence of this PRL in brain after hypophysectomy provided substantial evidence that brain PRL represented the product of a synthetic pool separate from that of the anterior pituitary PRL. To pursue this concept of independent synthesis further, we sought to determine whether brain tissue expressed PRL mRNA. Although we were easily able to detect a single species of PRL mRNA in pituitary by Northern hybridization, we could not visualize message in hypothalamus or extrahypothalamic brain by this technique. Therefore, we performed the polymerase chain reaction on cDNAs from anterior pituitary, hypothalamus, discrete extrahypothalamic brain regions, and other tissues. Hypothalamus and extrahypothalamic brain parts, including the cerebellum, caudate, brain stem, amygdala, thalamus, cortex, and hippocampus, were all positive to varying degrees. Lung and liver were negative, and anterior pituitary was consistently positive. All positive tissues, including anterior pituitary, expressed two hybridization signals: the expected amplified product and another smaller one. The smaller amplified product is presumably the result of an alternatively spliced transcript that is missing part of the PRL gene. Hypophysectomized animals did not express PRL message in brain, but expression was restored in hypophysectomized animals treated with testosterone. Transcripts for Pit-1 (GHF-1), a transcription factor important in regulation of pituitary PRL, were not detected in hypothalamus or any of the extrahypothalamic brain parts. The finding of testosterone stimulation of brain PRL message and undetectable levels of Pit-1 (GHF-1) in hypothalamic and extrahypothalamic brain regions indicates that the transcriptional regulation of PRL in the brain is different from that in the anterior pituitary.