After transport in the blood and implantation in the microcirculation, metastatic tumor cells must invade the vascular endothelium and underlying basal lamina. Mouse B16 melanoma sublines were used to determine the relation between metastatic properties and the ability of the sublines to degrade enzymatically the sulfated glycosaminoglycans present in the extracellular matrix of cultured vascular endothelial cells. Highly invasive and metastatic B16 sublines degraded matrix glycosaminoglycans faster than did sublines of lower metastatic potential. The main products of this matrix degradation were heparan sulfate fragments. Intact B16 cells (or their cell-free homogenates) with a high potential for lung colonization degraded purified heparan sulfate from bovine lung at higher rates than did B16 cells with a poor potential for lung colonization. Analysis of the degradation fragments indicated that B16 cells have a heparan sulfate endoglycosidase. Thus the abilities of B16 melanoma cells to extravasate and successfully colonize the lung may be related to their capacities to degrade heparan sulfate in the walls of pulmonary blood vessels.

The role of growth factor networks in regulating the progression of human melanocytes towards tumorigenicity and ultimately the malignant phenotype is poorly understood. In particular, the autocrine and paracrine influences that modulate cellular invasion and extracellular matrix degradative enzymes of melanoma cells remain undefined at the molecular level. We report here that nerve growth factor (NGF) can modify some metastasis-associated cellular properties of human and mouse melanoma cells. Treatment of early-passage human metastatic melanoma cells (MeWo) or their variants (3S5, 70W) with biologically active 2.5S NGF resulted in (a) delayed density-dependent inhibition of melanoma cell growth; (b) increased in vitro invasion through a reconstituted basement membrane; and (c) time- and dose-dependent induction of heparanase, a heparan-sulfate-specific endo-beta-D-glucuronidase associated with human melanoma metastasis. These effects of NGF were most marked in the 70W brain-colonizing cells (70W > MeWo > 3S5). The NGF enhancement of heparanase secretion was not species-specific, since it was also observed in murine B16 melanoma cells; the highest NGF stimulation of heparanase was found in brain-colonizing murine B16-B15b variant (B16-B15b > B16-BL6, B16-F10, B16-F1). NGF also increased the invasive capacity of the human 70W and murine B16-B15b sublines in a chemoinvasion assay performed with filters coated with purified heparan sulfate proteoglycan (HSPG). The enhancement of chemotactic response and heparanase production was detected at NGF concentrations sufficient to fully saturate both low- and high-affinity NGF receptors (NGFR), the neurotrophin receptor (p75) and the trkA gene product, respectively. The results suggest that, in addition to the effects of NGF on cellular development and differentiation within the peripheral and central nervous systems, NGF can exert changes in the invasive properties of neuroectoderm-derived melanoma cells.

Modifications in rabbit sperm plasma membranes during epididymal passage and after ejaculation were investigated by use of three lectins: concanavalin A (Con A); Ricinus cornmunis I (RCA0; and wheat germ agglutinin (WGA). During sperm passage from caput to cauda epididymidis, agglutination by WGA drastically decreased, and agglutination by RCA~ slightly decreased, although agglutination by Con A remained approximately unchanged. After ejaculation, spermatozoa were agglutinated to a similar degree or slightly less by Con A, WGA, and RCA~, compared to cauda epididymal spermatozoa. Ultrastructural examination of sperm lectin-binding sites with ferritin-lectin conjugates revealed differences in the densities of lectin receptors in various sperm regions, and changes in the same regions during epididymal passage and after ejaculation. Ferritin-RCA~ showed abrupt changes in lectin site densities between acrosomal and postacrosomal regions of sperm heads. The relative amounts of ferritin-RCA~ bound to heads of caput epididymal spermatozoa were greater than those bound to heads of cauda epididymal or ejaculated spermatozoa. Tail regions were labeled by ferritin-RCA~ almost equally on caput and cauda epididymal spermatozoa, but the middle-piece region of ejaculated spermatozoa was slightly more densely labeled than the principal-piece region, and these two regions on ejaculated spermatozoa were labeled less than on caput and cauda epididymal spermatozoa. Ferritin-WGA densely labeled the acrosomal region of caput epididymal spermatozoa, although labeling of cauda epididymal spermatozoa was relatively sparse except in the apical area of the acrosomal region. Ejaculated spermatozoa bound only a few molecules of ferritin-WGA, even at the highest conjugate concentrations used. Caput epididymal, but not cauda epididymal or ejaculated spermatozoa, bound ferritin-WGA in the tail regions. Dramatic differences in labeling densities during epididymal passage and after ejaculation were not found with ferritin-Con A. 950

Two of the more important steps in blood-borne tumor metastasis are attachment of the circulating malignant cells to the vascular endothelium and subsequent extravasation or invasion out of the blood vessel. A model for this process has been developed using cultured monolayers of vascular endothelial cells that synthesize a basal lamina or extracellular matrix (Kramer and Nicolson, Proc Natl Acad Sci USA 76:504, 1979). We have used this model to study metastatic tumor cell-endothelial cell interactions such as attachment to endothelial cells and their subsequent retraction and exposure of endothelial basal lamina as well as the interactions of metastatic tumor cells with the basal lamina leading to invasion and solubilization of this extracellular matrix. Morphological, immunological, and enzymological analysis of these steps in the metastatic process can be obtained using the vascular endothelial cell monolayer model for attachment and invasion.