Although approximately 98% of the internal surface area of the lung is lined by alveolar type I cells, little is known about the functions of this cell type. Using freshly isolated rat type I and type II cells, we created a subtraction library by suppression subtractive hybridization to identify genes differentially expressed by type I cells. We identified twelve genes of known function that are differentially expressed by type I cells. Differential expression of all 12 genes was confirmed by Northern blotting; we confirmed differential expression by immunocytochemistry for 3 genes for which suitable antibodies were available. Most of the genes code for proteins that are multifunctional. From the known functions of these genes, we infer that type I cells may play a role in the maintenance of normal alveolar homeostasis and protection from injury, lung development and remodeling, host defense, tumor/growth suppression, and surfactant metabolism, among other functions.

Angiotropism/pericytic mimicry and vascular co-option involve tumor cell interactions with the abluminal vascular surface. These two phenomena may be closely related. However, investigations of the two processes have developed in an independent fashion and different explanations offered as to their biological nature. Angiotropism describes the propensity of tumor cells to spread distantly via continuous migration along abluminal vascular surfaces, or extravascular migratory metastasis (EVMM). Vascular co-option has been proposed as an alternative mechanism by which tumors cells may gain access to a blood supply. We have used a murine brain melanoma model to analyze the interactions of GFP human melanoma cells injected into the mouse brain with red fluorescent lectin-labeled microvascular channels. Results have shown a striking spread of melanoma cells along preexisting microvascular channels and features of both vascular co-option and angiotropism/pericytic mimicry. This study has also documented the perivascular expression of Serpin B2 by angiotropic melanoma cells in the murine brain and in human melanoma brain metastases. Our findings suggest that vascular co-option and angiotropism/pericytic mimicry are closely related if not identical processes. Further studies are needed in order to establish whether EVMM is an alternative form of cancer metastasis in addition to intravascular cancer dissemination.

For more than 15 years, angiotropism in melanoma has been emphasized as a marker of extravascular migration of tumor cells along the abluminal vascular surface, unveiling an alternative mechanism of tumor spread distinct from intravascular dissemination. This mechanism has been termed extravascular migratory metastasis (EVMM). During EVMM, angiotropic tumor cells migrate in a 'pericytic-like' manner (pericytic mimicry) along the external surfaces of vascular channels, without intravasation. Through this pathway, melanoma cells may spread to nearby or more distant sites. Angiotropism is a prognostic factor predicting risk for metastasis in human melanoma, and a marker of EVMM in several experimental models. Importantly, analogies of EVMM and pericytic mimicry include neural crest cell migration, vasculogenesis and angiogenesis, and recent studies have suggested that the interaction between melanoma cells and the abluminal vascular surface induce differential expression of genes reminiscent of cancer migration and embryonic/stem cell state transitions. A recent work revealed that repetitive UV exposure of primary cutaneous melanomas in a genetically engineered mouse model promotes metastatic progression via angiotropism and migration along the abluminal vascular surface. Finally, recent data using imaging of melanoma cells in a murine model have shown the progression of tumor cells along the vascular surfaces. Taken together, these data provide support for the biological phenomenon of angiotropism and EVMM, which may open promising new strategies for reducing or preventing melanoma metastasis.

The four-transmembrane protein epithelial membrane protein-2 (EMP2) was recently identified as an endometrial protein necessary for blastocyst implantation, but the mechanism of this role is uncertain. In other cell types, EMP2 controls delivery of certain classes of proteins to the cell surface, including various integrin isoforms (a class of receptors implicated in endometrial-blastocyst interaction). Since alphavbeta3 integrin is an important endometrial molecule involved in blastocyst interaction, we evaluated the role of EMP2 in modulating integrin expression in the HEC1A endometrial cell line and endometrial epithelium in vivo. Elevation of EMP2 expression in HEC1A cells selectively increased the expression of alphavbeta3 integrin on the plasma membrane and was functional as judged by increased cell binding to an alphavbeta3 ligand, fibronectin. Conversely, reduction in EMP2 expression using an EMP2 specific ribozyme decreased the cell alphavbeta3 surface expression. The influence of EMP2 on alphavbeta3 integrin was also observed in vivo as reduction of EMP2 using ribozymes or short hairpin RNA diminished alphavbeta3 integrin expression in glandular and luminal uterine epithelium. Colocalization and coimmunoprecipitation studies suggested that EMP2 and alphavbeta3 integrin predominantly exist in a physically associated state. This study demonstrates for the first time the influence of EMP2 on alphavbeta3 surface expression and suggests that surface trafficking of integrin alphavbeta3 by EMP2 during the window of implantation may be a mechanism for its requirement in endometrial-blastocyst interaction.