The clinical utility of transjugular intrahepatic portasystemic shunts (TIPS) is frequently complicated by the ingrowth of tissue into the stent lumen, causing stent stenosis. These studies were undertaken to define the cellular and matrix components of the pseudointima, define the phenotype and function of the mesenchymal cells in the pseudointima and maintain them in culture, and to study the differences between stenotic and nonstenosed stents. A total of 35 stents were evaluated. TIPS pseudointima were examined histologically, by immunohistochemistry and in situ hybridization to determine the cellular and connective tissue constituents. Mesenchymal cells were grown from tissue within the TIPS and around it, and their phenotype was studied and compared with control smooth muscle cells and fibroblasts. Masson’s trichrome staining of histological sections demonstrated that TIPS tissue was composed of collagen and palisades of mesenchymal cells and was lined by an endothelium. Immunostaining demonstrated strong and uniform α-smooth muscle staining in TIPS mesenchymal cells and peri-TIPS cells. Type I procollagen mRNA expression was demonstrated in mesenchymal cells in and around the stent by in situ hybridization. TIPS mesenchymal cells secreted less radiolabeled fibronectin, and far more type III, relative to type I, collagen compared with peri-TIPS cells. TIPS cells also expressed high levels of type III procollagen mRNA compared with peri-TIPS cells. There was no difference between stenotic stents and nonstenosed stents with respect to clinical features, time from stenting, gross morphology, histology, presence of bile fistulae, and cell phenotype. However, smooth muscle cells (SMC) from stenotic stents demonstrated both greater cell proliferation and collagen I and III secretion compared with those from nonstenosed stents. These data demonstrate that TIPS stenosis results from an accumulation of collagen and proliferation of SMC within the stent lumen.
Previously we reported that linoleic acid (LA), but not oleic acid, caused a marked increase in the secretion of IL-8 by Crohn's human intestinal smooth muscle (HISM) cells. Antioxidants inhibited this response, implicating a role for oxidative stress and NF-kappaB, a transcription factor for IL-8 that is activated by oxidative stress. In this study, we examined two mechanisms whereby LA, the dietary precursor for arachidonic acid (AA), could increase the production of IL-8 via activation of AA pathways: 1) by generation of reactive oxygen species by the AA-pathway enzymes to activate NF-kappaB or 2) by AA metabolites. Normal and Crohn's HISM cells were exposed to LA, oxidizing solution (Ox), or oxidizing solution enriched with LA (OxLA). Exposure of cells to Ox or OxLA induced oxidative stress as determined by thiobarbituric acid reactive substances. In normal cells, Ox but not LA activated NF-kappaB as determined by transfection experiments and Western blot. In Crohn's cells, NF-kappaB was spontaneously activated and was not further activated by Ox or LA. In contrast, TNF-alpha markedly increased activation of NF-kappaB in both normal and Crohn's cells. These results indicated that LA did not increase IL-8 by activating NF-kappaB, so we evaluated the second mechanism of an effect of AA metabolites. In normal cells, OxLA, but not LA, markedly stimulated IL-8, whereas in Crohn's cells, both OxLA and LA stimulated IL-8. OxLA, also stimulated production of AA metabolites leukotriene B(4) (LTB(4)), PGE(2), and thromboxane B(2) (TXB(2)) by normal and Crohn's cells. To determine whether AA metabolites mediated the IL-8 response, cells were treated with OxLA plus indomethacin (Indo), a cyclooxygenase inhibitor, and nordihydroguaiaretic acid (NDGA), a lipoxygenase inhibitor. Both Indo and NDGA blocked the IL-8 response to OxLA. To determine more specifically a role for AA metabolites, AA was used. Similar to OxLA, OxAA stimulated production of IL-8 and AA metabolites. Pinane thromboxane, a selective thromboxane synthase inhibitor and receptor blocker, inhibited OxAA stimulation of TXB(2) and IL-8 in a dose-response manner. MK886, a selective 5-lipoxygenase inhibitor, inhibited OxAA stimulation of LTB(4) and IL-8 also in a dose-response manner. Analysis of specific gene products by RT-PCR demonstrated that HISM cells expressed receptors for both thromboxane and LTB(4). We conclude that AA metabolites mediated the IL-8 response to LA in HISM cells. Both cyclooxygenase and lipoxygenase pathways were involved. LA did not increase IL-8 by activating NF-kappaB, but NF-kappaB appeared to be involved, because LA increased IL-8 only in situations where NF-kappaB was activated, either spontaneously in Crohn's cells or by Ox in normal cells. We speculate that AA metabolites increased IL-8 production by enhancing NF-kappaB-dependent transcription of IL-8.
Interleukin-8 (IL-8), a chemokine secreted by cells at injury sites, has recently been recognized as involved in the pathogenesis of Crohn's disease. However, the pathogenesis of enhanced spontaneous transcription of IL-8 by the bowel in patients with Crohn's disease is undefined. Although IL-8 is secreted primarily by neutrophils, macrophages, and endothelial and epithelial cells, we observed the involvement of mesenchymal cells in the inflammatory process. A smooth muscle cell line isolated from the ileum of a patient with Crohn's disease (CDISM) and maintained in culture exhibited spontaneous transcription and secretion of IL-8 when compared with intestinal smooth muscle cells obtained from a normal subject (NHISM). Furthermore, IL-8 transcription from CDISM cells was associated with remarkable spontaneous activation of the oxidant-sensitive transcription factor NF-κB, as assessed by transient transfection assays with an IL-8 promoter reporter construct, Western blot analysis, and electrophoretic mobility shift assays (EMSA). Finally, we report here that CDISM cells exhibit significantly altered redox balance. The antioxidant pyrrolidine dithiocarbamate (PDTC) restored the redox equilibrium by mechanisms that inhibit binding of NF-κB to its cognate site on the IL-8 promoter. These findings suggest that restoration of the redox balance could hold promise for therapeutic intervention in Crohn's disease.
A simple and economical technique was developed to isolate and culture human arterial smooth muscle cells from chorionic plate vessels. Placentas from healthy women were collected at the time of term delivery. Chorionic plate arteries were identified, excised, and cut into small pieces. An explant technique was used to grow cultures of placental arterial smooth muscle (PASM) cells. Small pieces of vessel with lumens down were placed in 100-mm culture plates and grown in Dulbecco modified eagle medium and 10% fetal bovine serum. Cells appeared from explants within 1 week and grew to confluence in approximately 4 weeks. At confluence, PASM cell cultures had a uniform cell morphology that was characterized by elongated cells in parallel rows, typical of smooth muscle cells. Smooth muscle cell phenotype was evaluated by morphology and by immunoblotting and immunofluorescence of smooth muscle myofilament proteins. All PASM cell cultures expressed α-smooth muscle actin, β-tropomyosin, and h-caldesmon. Expression was similar to that of human aortic smooth muscle cells, but not to endothelial cells or fibroblasts. PASM cells stained uniformly for α-smooth muscle actin and lacked staining for a fibroblast-specific antigen. PASM cells were evaluated for their response to inflammatory mediators, tumor necrosis factor-α, and interleukin-1β by measurement of interleukin-8 production. Cells cultured for 18 hours showed a progressive increase in interleukin-8 production with time. Treatment with inflammatory mediators increased interleukin-8 production by 3-fold as compared with media control. This technique provides a simple method to obtain normal human arterial smooth muscle cells for in vitro studies of physiology and pathophysiology.