Donor and recipient cell surface colony stimulating factor-1 promote neointimal formation in transplant-associated arteriosclerosis.

OBJECTIVE: Transplant-associated arteriosclerosis manifests as progressive vascular neointimal expansion throughout the arterial system of allografted solid organs, and eventually compromises graft perfusion and function. Allografts placed in colony stimulating factor (CSF)-1-deficient osteopetrotic (Csf1(op)/Csf1(op)) mice develop very little neointima, a finding attributed to impaired recipient macrophage function. We examined how CSF-1 affects neointima-derived vascular smooth muscle cells, tested the significance of CSF-1 expressed in donor tissue, and evaluated the contribution of secreted versus cell surface CSF-1 isoforms in transplant-associated arteriosclerosis. METHODS AND RESULTS: CSF-1 activated specific signaling pathways to promote migration, survival, and proliferation of cultured vascular smooth muscle cells. Tumor necrosis factor-α addition increased CSF-1 and CSF-1 receptor expression, and tumor necrosis factor-α-driven proliferation was blocked by anti-CSF-1 antibody. In a mouse vascular allograft model, lack of recipient or donor CSF-1 impaired neointima formation; the latter suggests local CSF-1 function within the allograft. Moreover, reconstitution of donor or recipient cell surface CSF-1, without secreted CSF-1, restored neointimal formation. CONCLUSIONS: Vascular smooth muscle cells activation, including that mediated by tumor necrosis factor-α, can be driven in an autocrine/juxtacrine manner by CSF-1. These studies provide evidence for local function of CSF-1 in neointimal expansion, and identify CSF-1 signaling in vascular smooth muscle cells, particularly cell surface CSF-1 signaling, as a target for therapeutic strategies in transplant-associated arteriosclerosis.

The gap junction protein connexin32 interacts with the Src homology 3/hook domain of discs large homolog 1.

Scaffolding of membrane proteins is a common strategy for forming complexes of proteins, including some connexins, within membrane microdomains. Here we describe studies indicating that Cx32 interacts with a PDZ-containing scaffolding protein, Dlgh1 (Discs Large homolog 1). Initial screens of liver lysates using antibody arrays indicated an interaction between Cx32 and Dlgh1 that was confirmed using coimmunoprecipitation studies. Yeast two-hybrid complementation determined that the Cx32 bound via interaction with the SH3/Hook domain of Dlgh1. Confocal microscopy of liver sections revealed that Cx32 and Dlgh1 could colocalize in hepatocyte membranes in wild type mice. Examination of levels and localization of Dlgh1 in livers from Cx32 null mice indicate that, in the absence of Cx32, Dlgh1 was decreased, and the remainder was translocated from the hepatocyte membrane to the nucleus with some remaining in cytoplasmic compartments. This translocation was confirmed by Western blots comparing Dlgh1 levels in nuclear extracts from wild type and Cx32 null murine livers. Using SKHep cells stably transfected with Cx32 under the control of a tet-off promoter, we found that acute removal of Cx32 led to a decrease of membrane-localized Dlgh1 and an increase in the nuclear localization of this tumor suppressor protein. Together, these results suggest that loss of Cx32 alters the levels, localization, and interactions of the tumor suppressor protein Dlgh1, events known in other systems to alter cell cycle and increase tumorigenicity.

TEM8 expression stimulates endothelial cell adhesion and migration by regulating cell-matrix interactions on collagen.

The TEM8 gene is selectively expressed in tumor versus normal blood vessels, though its function in endothelial cell biology is not known. Towards the goal of clarifying this function, we tested whether TEM8 overexpression, or blocking TEM8's function with a dominant negative protein, would modulate endothelial cell activities. We found that TEM8-expressing endothelial cells migrated at a rate 3-fold greater than control cells in a monolayer denudation assay. Also, the addition of recombinant TEM8 extracellular domain (TEM8-ED) specifically inhibited both chemokinetic and chemotactic migration on collagen in the denudation and Boyden chamber assays, respectively. The TEM8-ED binds preferentially to collagen, and TEM8 expression enhanced endothelial adhesion to collagen 3-fold; the latter response was antagonized by the TEM8-ED. Consistent with the TEM8-ED acting as a dominant negative inhibitor of endogenously expressed protein were data showing that the TEM8-ED had no effect on the activation of beta1 integrin. TEM8 protein is present in human umbilical vein in situ and is expressed in low passage HUVEC in vitro. TEM8 protein expression in HUVEC was increased 5-fold by the initiation of tube formation, correlating expression of TEM8 with the angiogenic response. Taken together, these results indicate that TEM8 plays a positive role in endothelial cell activities related to angiogenesis.

ATR/TEM8 is highly expressed in epithelial cells lining Bacillus anthracis' three sites of entry: implications for the pathogenesis of anthrax infection.

Anthrax is a disease caused by infection with spores from the bacteria Bacillus anthracis. These spores enter the body, where they germinate into bacteria and secrete a tripartite toxin that causes local edema and, in systemic infections, death. Recent studies identified the cellular receptor for anthrax toxin (ATR), a type I membrane protein. ATR is one of the splice variants of the tumor endothelial marker 8 (TEM8) gene. ATR and TEM8 are identical throughout their extracellular and transmembrane sequence, and both proteins function as receptors for the toxin. ATR/TEM8 function and expression have been associated with development of the vascular system and with tumor angiogenesis. TEM8 is selectively upregulated in endothelial cells during blood vessel formation and tumorigenesis. However, selective expression of TEM8 in endothelial cells contradicts the presumably ubiquitous expression of the receptor. To resolve this controversial issue, we evaluated the distribution of ATR/TEM8 in a variety of tissues. For this purpose, we generated and characterized a novel anti-ATR/TEM8 polyclonal antibody. Here, we show that this novel antibody recognizes all three ATR/TEM8 isoforms, which are widely and differentially expressed in various tissue types. We found that ATR/TEM8 expression is not only associated with tumor endothelial cells, as previously described. Indeed, ATR/TEM8 is highly and selectively expressed in the epithelial cells lining those organs that constitute the anthrax toxin's sites of entry, i.e., the lung, the skin, and the intestine. In fact, we show that ATR/TEM8 is highly expressed in the respiratory epithelium of the bronchi of the lung and is particularly abundant in the ciliated epithelial cells coating the bronchi. Furthermore, immunostaining of skin biopsies revealed that ATR/TEM8 is highly expressed in the keratinocytes of the epidermis. Finally, we show that the epithelial cells lining the small intestine strongly express ATR/TEM8 isoforms. This is the first demonstration that the ATR/TEM8 protein is highly expressed in epithelial cells, which represent the primary location for bacterial invasion. These results suggest that the ATR/TEM8 expression pattern that we describe here is highly relevant for understanding the pathogenesis of anthrax infection.

Thymidine phosphorylase and 2-deoxyribose stimulate human endothelial cell migration by specific activation of the integrins alpha 5 beta 1 and alpha V beta 3.

Thymidine phosphorylase is an angiogenic factor that is frequently overexpressed in solid tumors, in rheumatoid arthritis, and in response to inflammatory cytokines. Our previous studies showed that cells expressing thymidine phosphorylase stimulated endothelial cell migration in vitro. This was a consequence of the intracellular metabolism of thymidine by thymidine phosphorylase and subsequent extracellular release of 2-deoxyribose. The mechanisms by which 2-deoxyribose might mediate thymidine phosphorylase-induced cell migration in vitro, however, are obscure. Here we show that both thymidine phosphorylase and 2-deoxyribose stimulated the formation of focal adhesions and the tyrosine 397 phosphorylation of focal adhesion kinase in human umbilical vein endothelial cells. Although similar actions occurred upon treatment with the angiogenic factor vascular endothelial growth factor (VEGF), thymidine phosphorylase differed from VEGF in that its effect on endothelial cell migration was blocked by antibodies to either integrin alpha 5 beta 1 or alpha v beta 3, whereas VEGF-induced endothelial cell migration was only blocked by the alpha v beta 3 antibody. Further, thymidine phosphorylase and 2-deoxyribose, but not VEGF, increased the association of both focal adhesion kinase and the focal adhesion-associated protein vinculin with integrin alpha 5 beta 1 and, in intact cells, increased the co-localization of focal adhesion kinase with alpha 5 beta 1. Thymidine phosphorylase and 2-deoxyribose-induced focal adhesion kinase phosphorylation was blocked by the antibodies to alpha 5 beta 1 and alpha v beta 3, directly linking the migration and signaling components of thymidine phosphorylase and 2-deoxyribose action. Cell surface expression of alpha 5 beta 1 was also increased by thymidine phosphorylase and 2-deoxyribose. These experiments are the first to demonstrate a direct effect of thymidine phosphorylase and 2-deoxyribose on signaling pathways associated with endothelial cell migration.

Mechanisms by which tumor cells and monocytes expressing the angiogenic factor thymidine phosphorylase mediate human endothelial cell migration.

The angiogenic factor thymidine phosphorylase (TP) is highly expressed in many human solid tumors, and the level of its expression is associated with tumor neovascularization, invasiveness, and metastasis and with shorter patient survival time. TP promotes endothelial cell (EC) migration in vitro and angiogenesis in vivo, and these have been linked to its enzymatic activity. The mechanism by which TP stimulates EC migration was investigated using human umbilical vein ECs (HUVECs). TP induced concentration-dependent HUVEC migration, which required a TP gradient and thymidine and which was abrogated by the TP inhibitor CIMU (5-chloro-6(1-imidazolylmethyl)uracil). The chemotactic actions of TP plus thymidine were duplicated by the TP metabolite, 2-deoxyribose-1-phosphate (dR-1-P), and 10-fold more potently by its subsequent metabolite, 2-deoxyribose (2dR). Migration induced by dR-1-P, but not 2dR, was blocked by an alkaline phosphatase inhibitor, suggesting that the actions of dR-1-P first required its conversion to 2dR. In the migration assay, [5'-3H]dThd was metabolized to dR-1-P (96%) and 2dR (3.8%), and a gradient of both metabolites was maintained between the lower and upper chambers over the entire 5-h assay. TP expression in human solid tumors occurs in both tumor epithelial cells and in tumor-associated macrophages. The migration assay was adapted to use TP-transfected carcinoma cells to stimulate HUVEC migration, and they were found to induce more migration than did control vector-transfected cells. Human monocyte cells U937 and THP1, which constitutively expressed high levels of TP, also strongly induced HUVEC migration in the coculture assay. CIMU inhibited tumor-cell and monocyte-induced migration. In contrast, a neutralizing antibody to TP had no effect on cell-stimulated HUVEC migration, even though it completely blocked the migration mediated by purified TP. Thus, the intracellular actions of TP were sufficient to stimulate HUVEC chemotaxis. In contrast to purified TP, when incubated with [5'-3H]-thymidine, cells expressing TP released up to 20-fold more 2dR into the medium than dR-1-P. These studies demonstrate that TP-expressing cells mediate EC migration via the intracellular metabolism of thymidine and subsequent extracellular release of 2dR, which forms a chemotactic gradient.

Inhibition of endothelial cell function in vitro and angiogenesis in vivo by docetaxel (Taxotere): association with impaired repositioning of the microtubule organizing center.

A number of cancer chemotherapeutic drugs designed to have cytotoxic actions on tumor cells have recently been shown to also have antiangiogenic activities. Endothelial cell migration and proliferation are key components of tumor angiogenesis, and agents that target the microtubule cytoskeleton can interfere with these processes. In this study, the effect on endothelial cell functions of the microtubule-stabilizing drugs Taxotere and Taxol were evaluated in three in vitro assays: a chemokinetic migration assay, an angiogenesis factor-mediated chemotactic migration assay, and a three-dimensional Matrigel tubule formation assay, using rat fat pad endothelial cells (RFPECs) and/or human umbilical vein endothelial cells (HUVECs). Taxotere was active in all three assays at concentrations that were not cytotoxic and did not inhibit endothelial cell proliferation. In the RFPEC chemokinetic migration and in vitro tubule formation assays, the IC50 values were approximately 10(-9) M for both Taxotere and Taxol. HUVEC migration, however, was more sensitive to Taxotere, with an observed IC50 of 10(-12) M in a chemokinetic assay. In a Boyden chamber assay, HUVEC chemotaxis stimulated by either of two angiogenic factors, thymidine phosphorylase or vascular endothelial growth factor, was inhibited by Taxotere with an IC50 of 10(-11) M and was ablated at 10(-9) M. Taxotere was also up to 1000-fold more potent than Taxol in inhibiting either chemokinetic or chemotactic migration. When the microtubule cytoskeleton was visualized using immunofluorescence staining of alpha-tubulin, there were no gross morphological changes observed in HUVECs or RFPECs treated with Taxotere at concentrations that inhibited endothelial cell migration but not proliferation. The effects of Taxotere on migration were associated with a reduction in the reorientation of the cell's centrosome, at concentrations that did not affect gross microtubule morphology or proliferation. Reorientation of the centrosome, which acts as the microtubule organizing center, in the intended direction of movement is a critical early step in the stabilization of directed cell migration. These data indicate that endothelial cell migration correlates more closely with changes in microtubule plasticity than with microtubule gross structure. The antiangiogenic activity of Taxotere in vivo was assessed in a Matrigel plug assay. In this assay, the angiogenic response to fibroblast growth factor 2 was inhibited in vivo by Taxotere with an ID50 of 5.4 mg/kg when injected twice weekly over a 14-day period, and angiogenesis was completely blocked in mice that received 10 mg/kg Taxotere. The in vivo data further suggested that Taxotere had selectivity for endothelial cell migration and/or microvessel formation because infiltration of inflammatory cells into the Matrigel plug was much less sensitive to inhibition by Taxotere. In conclusion, Taxotere is a potent and potentially specific inhibitor of endothelial cell migration in vitro and angiogenesis in vitro and in vivo.

A gene expression profile that defines colon cell maturation in vitro.

Colonic epithelial cells undergo cell cycle arrest, lineage specific differentiation, and apoptosis, as they migrate along the crypt axis toward the lumenal surface. The Caco-2 colon carcinoma cell line models many of these phenotypic changes, in vitro. We used this model system and cDNA microarray analysis to characterize the genetic reprogramming that accompanies colon cell differentiation. The analyses revealed extensive yet functionally coordinated alterations in gene expression during the differentiation program. Consistent with cell differentiation reflecting a more specialized phenotype, the majority of changes (70%) were down-regulations of gene expression. Specifically, Caco-2 cell differentiation was accompanied by the coordinate down-regulation of genes involved in cell cycle progression and DNA synthesis, which reflected the concomitant reduction in cell proliferation. Simultaneously, genes involved in RNA splicing and transport, protein translation, folding, and degradation, were coordinately down-regulated, paralleled by a reduction in protein synthesis. Conversely, genes involved in xenobiotic and drug metabolism were up-regulated, which was linked to increased resistance of differentiated cells to chemotherapeutic agents. Increased expression of genes involved in extracellular matrix deposition, lipid transport, and lipid metabolism were also evident. Underlying these altered profiles of expression, components of signal transduction pathways, and several transcription factors were altered in expression.