The expression patterns of vascular endothelial growth factor (VEGF) and its two receptors, flt-1 and KDR, were assessed in normal human melanocytes, transformed melanocytes expressing the simian virus 40 Tgene (SV40T), and melanoma cells derived from primary and metastatic lesions. Constitutive expression of VEGF, flt-1, and KDR mRNA and proteins was observed in the majority of primary and metastatic melanoma cell lines, and in SV40T-transformed melanocytes. VEGF expression in melanoma cell lines was further enhanced by exogenous growth factors including insulin and fetal calf serum. By contrast, neonatal melanocytes did not express VEGF or VEGF receptors and VEGF expression could not be induced by exogenous growth factors. Exogenous VEGF had no significant effects on melanoma cell proliferation or on production of a transcriptional target for VEGF, urokinase-type plasminogen activator. Down-regulation of VEGF expression in the metastatic melanoma cell line WM164 through transfection of a VEGF antisense construct similarly did not affect proliferation of the transfected cells in the presence or absence of exogenous VEGF. In summary, coexpression of VEGF and its receptors is a tumor-associated phenomenon in melanoma development. However VEGF production does not support autocrine proliferation of the melanoma cell lines tested.

No abstract

Expression patterns of the angiogenic placenta growth factor and its receptor neuropilin-1 were assessed in normal human melanocytes, SV40T-transformed melanocytes, and melanoma cells derived from primary and metastatic lesions. As determined by reverse transcription-polymerase chain reaction all primary and metastatic melanoma cell lines tested and SV40T-transformed melanocytes coexpressed two placenta growth factor splice variants (placenta growth factor-1 and -2) as well as neuropilin-1 mRNA. Placenta growth factor protein was detected in conditioned media derived from five of eight melanomas and from SV40T-transformed melanocytes. In contrast to melanoma cells, normal melanocytes did not express placenta growth factor mRNA at detectable levels and did not secrete placenta growth factor protein. By contrast, neuropilin-1 transcripts were detected in some of the normal melanocytes. Secretion of placenta growth factor by melanoma cells appeared to be constitutive because it was not affected by the addition of exogenous growth factors including insulin, epidermal growth factor, or basic fibroblast growth factor to culture media. Although melanoma cells expressed both, neuropilin-1 and flt-1, exogenous homodimeric placenta growth factor had no effect on melanoma cell growth. Similarly, placenta growth factor did not induce urokinase-type plasminogen activator production in these cells. These findings demonstrate that melanoma progression is accompanied by deregulated, constitutive placenta growth factor expression. Placenta growth factor, however, serves no apparent autocrine role in melanoma proliferation. Further studies are needed to define the relative contribution of placenta growth factor to the angiogenic properties of human melanomas.