Transfection of human fibroblast growth factor 9 (FGF-9) cDNA into mouse BALB/c 3T3 clone A31 cells led to morphological transformation of the cells and foci formation 4 weeks later. Isolated transformants had a higher saturation density than parental A31 cells, could grow in soft agar, and secreted FGF-9 into the culture supernatant. The introduction of FGF-9 N33 cDNA, which encodes a truncated protein that has 33 N-terminal amino acids deleted and has the same mitogenic potency as FGF-9, failed to lead to foci formation. Although FGF-9 is a secretory protein, it does not have a typical secretory signal sequence, and the secreted protein retains the full sequence coded in the cDNA except for the initiating methionine. The produced FGF-9 N33 was not secreted and remained within the cell. It is possible that FGF-9 has an uncleavable signal sequence within the first 33 N-terminal amino acids. All of the phenotypes acquired by transformation could be arrested by treatment with a neutralizing anti-human FGF-9 monoclonal antibody (MAb) 150-59. Additionally, transformants formed tumors in nude mice. Injection of MAb 150-59 suppressed tumor formation in nude mice and caused existing tumors to regress. Our results suggest that the cellular transformation mediated by FGF-9 is produced by autocrine stimulation. We have detected FGF-9 production in the human tumor cell lines glioma NMC-G1, from which FGF-9 was originally purified, and stomach carcinoma AZ-521. The growth of NMC-G1 was not affected by MAb 150-59, but that of AZ-521 was arrested by MAb 150-59 in the presence of heparin. Moreover, the growth of the AZ-521 cell tumor in nude mice could be partially arrested by antibody treatment. The possibility of a participation of FGF-9 in the formation of human tumors is suggested. Int. J.

Transfection of human fibroblast growth factor 9 (FGF-9) cDNA into mouse BALB/c 3T3 clone A31 cells led to morphological transformation of the cells and foci formation 4 weeks later. Isolated transformants had a higher saturation density than parental A31 cells, could grow in soft agar, and secreted FGF-9 into the culture supernatant. The introduction of FGF-9 N33 cDNA, which encodes a truncated protein that has 33 N-terminal amino acids deleted and has the same mitogenic potency as FGF-9, failed to lead to foci formation. Although FGF-9 is a secretory protein, it does not have a typical secretory signal sequence, and the secreted protein retains the full sequence coded in the cDNA except for the initiating methionine. The produced FGF-9 N33 was not secreted and remained within the cell. It is possible that FGF-9 has an uncleavable signal sequence within the first 33 N-terminal amino acids. All of the phenotypes acquired by transformation could be arrested by treatment with a neutralizing anti-human FGF-9 monoclonal antibody (MAb) 150-59. Additionally, transformants formed tumors in nude mice. Injection of MAb 150-59 suppressed tumor formation in nude mice and caused existing tumors to regress. Our results suggest that the cellular transformation mediated by FGF-9 is produced by autocrine stimulation. We have detected FGF-9 production in the human tumor cell lines glioma NMC-G1, from which FGF-9 was originally purified, and stomach carcinoma AZ-521. The growth of NMC-G1 was not affected by MAb 150-59, but that of AZ-521 was arrested by MAb 150-59 in the presence of heparin. Moreover, the growth of the AZ-521 cell tumor in nude mice could be partially arrested by antibody treatment. The possibility of a participation of FGF-9 in the formation of human tumors is suggested. Int. J.

Hybridomas secreting monoclonal antibodies (MAbs) against human fibroblast growth factor 9 (FGF-9) were established using recombinant human (rh) FGF-9 N33 as the immunogen. Among these, MAb 150-59 demonstrated a potent neutralizing activity against FGF-9. It arrested the FGF-9-induced growth of BALB/c 3T3 A31 cells at an equimolar dose of the factor. It also abrogated the in vivo thrombopoietic activity of FGF-9. Mitogenic activity of several other FGF family members such as FGF-1, FGF-2, and FGF-4 was not neutralized by this MAb. A sensitive sandwich enzyme immunoassay for FGF-9 was developed employing MAbs 150-59 and 13-3. The detection limit of this system was 3 pg/well. In this assay system, FGF-1 and FGF-2 were not cross-reactive up to 1 microgram/well. Using this system, the distribution of FGF-9 in rat organs was examined. FGF-9 could be detected only in the extract of rat cerebellum. Also, we detected a high amount of FGF-9 in the culture supernatant of certain cell lines originated from human tumor. These findings suggest that this enzyme immunoassay system may be used to clarify biological meaning of FGF-9.

No abstract