In a recent immunohistochemical study, we found lack of expression of TGF-beta1, 2, and 3 by neuroblastoma cells, except for well differentiated ganglion cells in ganglioneuroblastomas and in normal ganglia. In contrast, almost all primitive neuroectodermal tumors (PNET) expressed variable amounts of TGF-beta1 and 3. Absence of TGF-beta secretion was also found in conditioned media from neuroblastoma cell lines, in contrast to PNET cell lines using a bioassay based on TGF-beta inhibition of DNA synthesis in mink lung fibroblasts. These data suggested a possible role of TGF-beta in the differentiation of human neuroblastoma. This hypothesis will be further pursued in the following ways: (1) TGF-beta expression and secretion will be studied in neuroblastoma and PNET cell lines before and after differentiation with known differentiating agents, such as retinoic acid and TPA. (2) Direct effects of exogenously added TGF-beta in undifferentiated and differentiated neuroblastoma cell cultures will be determined. (3) Blocking antibodies and antisense TGF-beta oligonucleotides will be used to modulate possible actions of TGF-beta on neuroblastoma and PNET cells in vitro. Preliminary data have shown induction of TGF-beta1-mRNA in SH- SY5Y neuroblastoma cells following treatment of cultures with TPA or retinoic acid, both of which promoted differentiation. In contrast, IMR- 32 neuroblastoma cells which showed minimal morphologic differentiation with the same agents showed only mild increase in TGF-beta1 mRNA levels.
