ENDOGENOUS FIBROBLAST GROWTH-FACTOR-I OR FIBROBLAST GROWTH FACTOR-II MODULATE PROSTATE-CANCER CELL-PROLIFERATION

We constructed expression vectors containing either rat fibroblast gro wth factor (FGF)-1 or FGF-2 cDNA cloned in either the sense orientatio n or antisense orientation relative to the metallothionein promoter of plasmid pMTneo.1. Stable AXC/SSh rat prostate cancer cell transfectan ts expressing either chimeric FGF-1-sense, chimeric FGF-1-antisense, o r chimeric FGF-2-antisense transcripts were obtained. Stable transfect ants expressing chimeric FGF-2-sense transcripts were not obtained. Co ntrol, sense, and antisense transfectants expressed endogenous FGF-1 a nd endogenous FGF-2 transcripts, implying that transfection did not el iminate endogenous FGF transcripts. Control transfectants and sense tr ansfectants contained FGF-1 isoforms having a mass of 16.4 or 17.3 kDa and FGF-2 isoforms having a mass of 17, 19.5, or 21.5 kDa. Significan tly, adult AXC/SSh rat prostate contained only the 17.3 kDa FGF-1 isof orm and the 17 kDa FGF-2 isoform, indicating that neoplastic transform ation was associated with elaboration of novel, prostate epithelial ce ll-derived FGF-2 isoforms. FGF-1-antisense RNA expression eliminated t ransfectant FGF-1 isoforms without affecting FGF-2 isoform content. Si milarly, FGF-2-antisense RNA expression eliminated the transfectant 21 .5 kDa FGF-2 isoform, diminished the 19.5 kDa FGF-2 isoform content, a nd reduced the 17 kDa FGF-2 isoform content to barely detectable level s without affecting the FGF-1 isoform content. This established that F GF-antisense RNAs specifically inhibited translation of cognate, endog enous FGF transcripts. Doubling times of control transfectants and sen se transfectants were indistinguishable and were not affected by inclu ding FGF-1 or FGF-2 in the culture medium. Doubling times of FGF-1-ant isense or FGF-2-antisense transfectants were 1.3- to 1.4-fold greater than those of control transfectants or sense transfectants, and either exogenous FGF-1 or exogenous FGF-2 decreased antisense transfectant d oubling times to values indistinguishable from those of control transf ectants or sense transfectants. This established that with regard to p rostate cancer cell proliferation: (a) endogenous FGF-1 cannot substit ute for endogenous FGF-P eliminated by FGF-2-antisense RNA expression; and (b) endogenous FGF-2 cannot substitute for endogenous FGF-1 elimi nated by FGF-1-antisense RNA expression. In contrast, either exogenous FGF-1 or exogenous FGF-2 decreased antisense transfectant doubling ti me. The results of these studies establish that endogenous FGF-1 and e ndogenous FGF-2 modulate prostate cancer cell proliferation and imply that FGF-1 and FGF-2 of endogenous and exogenous origin conjointly con trol aspects of prostate cancer cell homeostasis. Our findings suggest complex interaction between components of prostate cancer cell regula tory processes and endogenously produced and exogenously accessible FG F-1 and FGF-2.