[CANCER RESEARCH 54, 6441-6445, December 15, i994J Unsaturated Fatty Acids Are Required for Continuous Proliferation of Transformed Androgen-dependent Cells by Fibroblast Growth Factor Family Proteinsi Soji Kasayama,2 Masafumi Koga, Haruhiko Kouhara, Satoru Sumitani, Katsuya Wada, Tadamitsu Kishimoto, and Bunzo Sato Department ofMedicine III, Osaka University Medical School, 2—2Yamada-oka 2-2, Suite City, Osaka 565 [5. K.. M. K., H. K., S. S., T. K., B. S.J, and Laboratory of Clinical Research, Hiroshima General Hospital, Hiroshima 738 [K. W.J, Japan ABSTRACT (9, iO). In addition, patients with prostate cancer have been reported to have a significantly higher concentration of oleic acid (i8:1, n-6) in Increase in dietary fat intake has been reported to be associated with phospholipids from both plasma and prostatic tissue when compared progression of hormone-dependent cancers. To explore Its mechanism, we with those with benign prostatic disease (1 1). In line with these results examined the effects of fatty acids on the growth of androgen-dependent SC-3 cells cloned from mouse mammary cancer (Shionogi carcinoma 115) on human subjects, dietary fats high in LA (i8:2, n-6) have been Their androgen-dependent growth was potentiated by ilnoleic acid in the observed to enhance the growth (i2, i3) and metastasis (i4—i7) of defined medium. The effect of linoleic acid on flbroblast growth factor murine mammary carcinomas. However, these in vivo studies could (FGF)-dependent growth was also addressed because androgen had been not allow us to discuss the direct effect of unsaturated FA on cancer demonstrated to exert its mitogenic activity on SC-3 cells through an cells. induction of the unique FGF family protein termed as androgen-induced To investigate the direct effect of FM on hormone-dependent growth factor. Exposure of SC-3 cells to basic FGF or androgen-induced cancers, the culture systems were definitely required. To do so, we growth factor exhibited only transient growth response. However, simul needed to overcome two major obstacles: (a) many hormone-respon taneous addition oflinoleic acid to the medium sustained the proliferation sive transformed cells for culturing require the serum that contains ofFGF-stlmulated, but not FGF-unsthnulated, cells, althOUghlinoleic acid unknown amounts of FM; and (b) the molecular mechanism of did not exert the significant effect on the process of S-phase entry of basic hormone-dependent growth is largely obscure in many established FGF-stlmulated cells. Pahnltolelc acid and oleic acid appeared to exert the actions similar to linoleicacid, whilestearic acid was wIthoutany effect. cell lines. In this relationship, SC-3 cells cloned from mouse mam Neither cyclooxygenase inhibitor nor 5-lipoxygenase Inhibitorcould block mary carcinoma (Shionogi carcinoma 115) seemed to be quite suitable the growth-promoting ability of ilnoleic acid. Linoleic acid also enhanced for addressing the effect of FAa on their growth since this cell line their anchorage-Indepeadent growth in the presence of basic FGF. These exhibits the remarkable growth-stimulatory response to androgen in results indicate that these unsaturated fatty acids play a role in sustaining the defined medium (i8). This androgen stimulation results in a the proliferation of FGF-StiIIIUIatedSC-3cells. secretion of an autocrine heparin-binding growth factor (19), which in turn binds to FGF receptor i (20, 21). Thus, the ligand-induced INTRODUCTION activation of FGF receptor 1 plays a central role for the SC-3 cell growth. This consideration is further supported by the fmding that The incidence of hormone-dependent prostate and breast cancers both bFGF and aFGF can initiate the DNA synthesis of the SC-3 cells has been shown to be positively associated with dietary fat intake (22). We have recently cloned cDNA encoding the autocrine growth (i, 2). The strength of the association with overall levels of fat in factor termed as AIGF, which is identified as a new member of FGF consumption increases with age and is strongest in the ages with family proteins (23). Taking these advantages into consideration, we peak cancer incidence rates. Reductions in cancer incidence rates have examined the effects of FM on the growth of the SC-3 cells resulting from a sustained 50% cut in the average consumption of unstimulated or stimulated with androgen or these growth factors. fat in men and women of age 55—69years were estimated to be 83% for prostate cancer and 61% for breast cancer (3). These epidemiological studies suggest that dietary lipids promote the MATERIALS AND METHODS progression of the hormone-dependent cancers. In addition to the quantity of fat consumed, its FA3 composition Materials. Bovine brain-derived aFGF and human recombinant bFGF should be taken into consideration when the mechanism of fat intake were obtained from R&D Systems (Minneapolis, MN). Heparin-Sepharose dependent increase in hormone-dependent cancers is considered. Most was purchased from Pharmacia (Piscataway, NJ). BSA (essentially fatty acid epidemiological studies implicate that high intake of fat of animal free), FM, indomethacin, and testosterone were from Sigma Chemical Co. (St. origin mainly composed of saturated rather than unsaturated FA is Louis, MO). [methyl-3H]thymidine (70—85Ci/mmol) and Ampure SA column were from Amersham Japan (Tokyo, Japan). AA 861 was kindly supplied by causally related to the incidence of breast cancer (4—6),although this Takeda Pharmaceutical Co. (Osaka, Japan). All other chemicals were of dietary behavior may cause obesity, which in turn elevates serum analytical grade. hormone levels (7, 8). On the other hand, the increasing incidence of Cell Growth Experiments. A clonal cell line from Shionogi carcinoma breast cancer has been suggested to be related to the sustained in ii5, SC-3, was established as described previously (i8). The SC-3 cells were crease in vegetable oil consumption (jrincipally n-6 unsaturated FM) maintained in MEM supplemented with 2% FCS (Hyclone, Logan, UT) and 108 M testosterone, and were kept at 37°Cin a humidified 5% CO2 incubator. Received 7/25/94; accepted 10/18/94. NIH3T3 cells, obtained from RIKEN (Saitama, Japan), were cultured in The costs of publication of this article were defrayed in part by the payment of page DMEM supplemented with 10% calf serum. charges. This article must therefore be hereby marked advertisement in accordance with For the anchorage-dependent growth experiments, these cells were plated 18 U.S.C. Section 1734 solely to indicate this fact. onto a 24-well plate (i0@cells/well) in 1 ml MEM containing 2% DCC-treated 1 Supported by grants-in-aid from the Ministry of Education, the Ministry of Health and Welfare, Tokyo Japan, Enami Memorial Foundation for Cancer Research, and Uehara FCS (18). On the following day (day 0), the medium was replaced with 1 ml Biomedical Research Grant. HMB medium containing various test compounds. The serum-free medium 2 To whom requests for reprints should be addressed. was changed every other day. The viable cell number was counted on day 6, 3 The abbreviations used are: FA, fatty acid; LA, linoleic acid; FGF, fibroblast growth factor, b/a FGF, basic/acidic FGF; AIGF, androgen-induced growth factor; BSA, essen unless specified otherwise. tinily fatty acid-free bovine serum albumin; HMB, linoleic acid-free Ham's F-12:MEM For the anchorage-independent growth experiments, the cells were sus (1:1, v/v) containing 0.1% mA; DCC, dextran-coated charcoal. pended in 0.4% Noble agar (DIFCO Laboratories, Detroit, MI) in HMB 6441 Downloaded from cancerres.aacrjournals.org on September 23, 2021. © 1994 American Association for Cancer Research. LINOLEIC ACID AND ANDROGEN-DEPENDENT CANCER medium supplemented with 1% DCC-treated FCS and the test compounds at 2 a density of 2 X i0@cells/5 ml. This solution (5 ml) was overlayered on 5 ml of 0.5% Noble agar in the above medium in 60-mm dish. These cells were In incubated in 5% C02-95% air at 37°Cfor14 days without medium change. 0 Then the visible colonies formed were counted (21). DNA Synthesis. [3HjmymidineincorporationintoDNA of SC-3cells was measured by the method as described (24). Briefly, the cells were plated onto I a 96-well plate (5 x UP cells/well) in 0.15 ml MEM supplemented with 2% I- 0 DCC-treated FCS. On the following day, the cells were washed once with .0 HMB medium and kept in HMB medium for 24 h. The quiescent cells were E @ treated with test compounds for 24 h. Then the cells were pulse labeled with z - P [3Hjthymidine(0.3 pCi/well) for 2 h to measure the DNA synthesis. Partial Purification of AIGF. AIGF was extracted from the androgen stimulated SC-3 cell culture (25). The SC-3 cells were stimulated with 10_8 M 0 testosterone in HMB medium for 48 h. After the conditioned medium was 0 1 2.5 5 10 taken out, the dish was extracted with 2 M NaCl, 10 m@iHEPES (pH 7.5 at 20°C)and 0.01% BSA at 4°Cfor 1 mm. This extracted solution was applied Linoleic Acid (gig!ml) to an Ampure SA column to remove excess NaCl. Then AIGF was partially Fig. 2. Concentration-dependent growth promotion of LA on bFGF-stimulated SC-3 purified by means of a heparin-Sepharose column (26). cell growth. The cells were plated on triplicate wells as described in “Materialsand Methods.―On the following day, the cells were stimulated with 2 ng/mI bFGF in the Statistical Analysis. All values presented here were means ±SE. When presenceofvariousconcentrationsofLA.Thecellnumberwascountedonday6. the significant difference was discussed, analysis of variance with a multiple Columns, mean; bars, SE. range test was used. Table 1 Effects of FAs on the growth of SC-3 celLs stimulated with testosterone or RESULTS bFGF The cells were plated on triplicate wells as described in “MaterialsandMethods,―and Effects of FAs Exogenously Added into LIMB Medium on Pro then@WmlFA.