Reduction of TGF-beta activity abrogates growth promoting tumor cell-cell interactions in vivo

We have shown in previous studies that metastatically-competent variant subpopulations (B5, C1) derived from a non-metastatic murine mammary adenocarcinoma (SP1) have a pronounced growth advantage over their non-metastatic tumor cell counterparts in primary tumors. As a result, primary tumors can be progressively overgrown by cells having the competence to spread elsewhere in the body. This occurs despite any evidence to indicate an intrinsic in vivo growth rate advantage of the metastatic cells when grown as isolated populations. This suggested that cell-cell interactions between metastatic and non-metastatic tumor populations may be involved in the metastatic cell growth dominance process. Evidence was therefore sought for growth factors released by SP1 cells which could preferentially stimulate the B5 or C1 variants and thereby mediate this cell-cell interaction process. We found that cocultures of SP1 and C1 or B5 cells with irradiated C1, B5, or SP1 "feeder" cells showed significant stimulation of C1 and B5 by SP1 "feeder" cells. Cell growth stimulation in response to EGF, TGF-alpha, TGF-beta 1, bFGF, PDGF, NGF, IGF-1, or IGF-2 demonstrated that only TGF-beta 1 could duplicate this effect. A repeat of the coculture experiment in the presence of specific neutralizing anti-TGF-beta antibodies was therefore undertaken and this was found to markedly reduce the stimulation of C1 or B5 cells by irradiated SP1 cells. Conditioned media from the SP1 and C1 cell lines was quantitated for TGF-beta activity and contained 4.5 ng/ml and 2.0 ng/ml, respectively. However, the majority of the TGF-beta released by SP1 cells was found to be spontaneously active, whereas 70% of the TGF-beta released by C1 cells was in its latent form. Scatchard analysis revealed approximately four times the number of TGF-beta receptors, of similar type and affinity, present on C1 as compared with SP1 cells. The in vitro results support the hypothesis that active TGF-beta released by SP1 cells may stimulate the proliferation of metastatic variant cells in a paracrine like fashion. In vivo evidence for this was obtained by showing that coinjection of irradiated SP1 cells could selectively stimulate tumor growth of viable C1 cells and this effect was markedly diminished by neutralizing polyclonal anti-TGF-beta antibodies. Taken together, the results suggest a novel role for TGF-beta in clonal evolution of malignant tumor growth and as a molecular mediator of tumor cell-tumor cell interactions involved in facilitating tumor progression.