Search

A mechanism for stroma formation (development of vasculature and supportive connective tissue) is suggested in an experimental “pushing-type” colorectal carcinoma liver metastasis model. The key element is the appearance of smooth muscle actin (SMA)-positive cells and the sinusoidal lakes at the border of the metastases. These lakes are the consequence of the disappearance (‘stepping back’ of hepatocytes from the border zone, resulting in the fusion of partially capillarized sinusoids. The growing tumor incorporates SMA-expressing cells and sinusoidal lakes. SMA-positive cells produce collagenous matrix, whereas the lakes become the central vessels within the connective tissue columns. Formation of these columns within the tumor is a consequence of the compression atrophy of the base of the incorporated liver tissue, leading to partial separation of the innermost part of the invagination containing functional vessel(s) from the surrounding liver. © 2005 Wiley-Liss, Inc.

Previous data have indicated that the proteoglycan (PG) pattern is different on tumor cells with different liver metastatic potential. We selected "conventional" glycosaminoglycan (GAG) biosynthesis inhibitors, beta-D-xyloside (BX), 2-deoxy-D-glucose (2-DG), ethane-l-hydroxy-l,l-diphosphonate (ETDP) and the newly discovered 5-hexyl-2-deoxyuridine (HUdR), to modulate PGs on highly metastatic/liver-specific 3LL-HH murine carcinoma and HT168 human melanoma cells and to influence their liver colonization potential. These compounds all induced remarkable changes in GAG biosynthesis, but to varying degrees: glucosamine labelling was affected mainly by 2-DG, and HUdR and sulphation by BX and HUdR. Furthermore, the ratio of heparan sulphate/chondroitin sulphate (HS/CS) of PGs was increased by ETDP and decreased after treatment by HUdR. In addition to changes in PG metabolism, tumor-cell proliferation and adhesion to fibronectin were affected; BX and 2-DG stimulated cell proliferation and adhesion, while HUdR inhibited both proliferation and adhesion. Most interestingly, HUdR, the most effective inhibitor of HS/HSPG, depressed the formation of liver colonies, while ETDP, the most effective inhibitor of CS/CSPG, stimulated the appearance of liver colonies. These observations indicated that, at least in these experimental systems, tumor cells with a high HS/CS ratio are more likely to form liver metastases; consequently, anti-HS agents could also be anti-metastatic.

Autocrine motility factor is a tumor-secreted cytokine which regulates cellular growth and motility by a receptor-mediated pathway. In the accompanying report (Part I of II), it was demonstrated that high (K1735-M1) and low (K1735-C1.11) metastatic murine melanoma cells display distinct adhesion and spreading characteristics which correlate with their differential spontaneous and stimulated migrations on the extracellular matrix components fibronectin, laminin and collagen IV. These parameters were further related to discrete profiles of focal adhesion plaque integrity and reorganization. Here we describe unique migration patterns observed in these murine melanoma cells which reflect differences in degradation and/or remodeling of the cellular substratum. These profiles of matrix interaction were influenced distinctly by autocrine motility factor and dictated by both substrate composition and cellular phenotype. Since activation of the autocrine motility factor receptor stimulates invasion of a reconstituted basement membrane and enhances experimental metastasis by high- but not low-metastatic K1735 cells, differences in the invasive phenotypes of these cells may be due in part to their differential responses to external stimuli coupled with internal propensities toward either matrix degradation and migration (high-metastatic cells) or matrix remodeling and stasis (low-metastatic cells).

Cellular interactions with the extracellular matrix are complex and are involved in numerous biological processes. These interactions may be modulated by cytokines such as tumor cell autocrine motility factor, a secreted molecule that regulates cellular growth and migration by a receptor-mediated pathway. In this report we provide evidence suggesting that high- and low-metastatic K1735 melanoma cells coordinate their attachment, spreading and migratory responses to autocrine motility factor and the extracellular substratum through alterations in focal adhesion plaque architecture that are dependent on both inherent cellular metastatic phenotype and the composition of the supporting matrix. Thus, since activation of the autocrine motility factor receptor has previously been shown to enhance the experimental metastasis of the high- but not the low-metastatic K1735 cells, differences in melanoma cell malignancy in this system may be due in part to a coordinated interplay between cytokine-mediated responses and extracellular matrix-directed regulation of cellular adhesion, spreading and motility.

Previous studies have indicated that the predominant sites of tumor cell extravasation in the liver are the sinusoidal vessels, where tumor cells contact the sinusoidal endothelium and the subendothelial extracellular matrix containing the basic components of the basement membrane. We studied the role of sinusoidal extracellular matrix in metastatsis formation by 3LL-HH murine tumor cells selected for their preferential liver colonization. 3LL-HH tumor cells did not efficiently adhere to cryosections of the liver, but they recognized the sinusoids and vessel walls. Pre-treatment of the mice with polyclonal anti-basement membrane antibodies [anti-laminin, anti-fibronectin and anti-heparan sulfate proteoglycan (HSPG)] significantly modulated the organ distribution of tumor cell colonies following intracardial injection: all 3 antibodies inhibited kidney colonization; anti-laminin and anti-fibronectin antibodies inhibited lung colonization; and only anti-HSPG antibody inhibited liver colonization. In several organs such as the heart, stomach, pancreas and bladder, anti-basement membrane antibody treatment did not alter the process of colonization. Immunofluorescence studies showed that anti-HSPG antibody recognized the basement membranes of sinusoids and blood vessels. Our data suggest a specific involvement of sinusoidal HSPG in the liver colonization of 3LL-HH cells. Int. J. Cancer 71: 825-831, 1997. © 1997 Wiley-Liss Inc.

Two human melanoma lines with low (HTI68) and high (HTI 68-MI) liver metastatic capacity in immunosuppressed mice were selected in vivo from the A2058 cell line. After i. v. injection of the 2 tumor lines there was no significant difference either in the number of lung colonies or in the frequency and tissue distribution of extrapulmonary tumor deposits. These findings suggest that the selection in the spleen-liver system did not result in an overall increase in the metastatic potential of the melanoma cells, but rather that it represented an organ-preferential selection. The HT168-MI cells did not acquire an increased growth rate in vitro or in vivo, suggesting that other phenotypic alterations are responsible for the enhanced metastatic capacity. The 2 tumor lines were characterized by similar expression of HLA-A, B, C, transferrin receptor and melanoma-associated proteoglycan antigen. HT168 contained more NGF receptor, while HLA-DR appeared only on HT 168-MI cells. This human metastasis model could be useful in studying the mechanisms of liver metastasis formation, as well as in revealing possible new targets of antimetastatic therapy.