- Email: [email protected]
Expression and Regulation by Serum of Multiple FGF1 mRNA in Normal Transformed, and Malignant Human Mammary Epithelial Cells
JOBNAME: BBRC 219#3 PAGE: 1 SESS: 12 OUTPUT: Tue Apr 2 09:18:08 1996 /xypage/worksmart/tsp000/68953f/2
BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS ARTICLE NO.
219, 679–685 (1996)
0294
Expression and Regulation by Serum of Multiple FGF1 mRNA in Normal Transformed, and Malignant Human Mammary Epithelial Cells F. Renaud,* I. El Yazidi,†,‡ Y. Boilly-Marer,† Y. Courtois,* and M. Laurent1 *Laboratoire INSERM U450, 29 rue Wilhem, 75016 Paris, France; and †Laboratoire CNRS UMR 111, Batiment C9, Université des Sciences et Technologies de Lille, 59655 Villeneuve D’Ascq Cedex, France, and ‡Laboratoire de Biologie du Développement, Université des Sciences et Technologies de Lille, 59655 Villeneuve D’Ascq Cedex, France Received January 12, 1996 In normal (NMEC), transformed (HBL-100) and malignant human mammary epithelial cells (MCF 7, BT-20, MDA-MB 231), we have examined the expression and the regulation by serum of FGF1 and FGF2 mRNA. FGF2 mRNA level was higher in NMEC and in a HBL-100 than in malignant cell lines (MDA-MB-231, BT-20). No FGF2 mRNA was detected in the malignant cell line, MCF-7. In contrast, the FGF1 mRNA was detected in all the mammary epithelial cells but at different levels. NMEC, HBL-100 and MDA-MB-231 cells expressed similar level of FGF1 and higher than that observed in BT-20 and MCF-7. In contrast to FGF2 which is only expressed in nonmalignant cells, no correlation between FGF1 mRNA expression and the phenotype of the cells was observed. We followed the expression of four FGF1 mRNA, heterogenous in their 59 untranslated regions. This study demonstrated that (i) the FGF1 mRNA 1.A was not expressed by mammary epithelial cells, (ii) the FGF1 mRNA 1.B was only expressed in normal mammary epithelial cells and (iii) the transcripts 1.C and 1.D were expressed in normal and malignant cells with specific patterns. The expression of FGF1 mRNAs responded in a cell specific manner to serum starvation. The mRNA 1.A was only expressed in normal cells cultured in the absence of serum while 1.C was either up- or down-regulated by serum in transformed cells and the expression of 1.D was greater in presence of serum in all cell lines. These results show that the regulation of FGF1 mRNAs expression is cell specific and does not correlate with a tumorigenic or transformed cell phenotype. © 1996 Academic Press, Inc.
Acidic fibroblast growth factor (FGF-1) and basic fibroblast growth factor (FGF2) are the prototype members of the fibroblast growth factor family which contains 9 growth factors and/or oncogenes (FGF-1 to FGF-9) (1,2,3). FGF1 and FGF2 are multipotent factors involved in the proliferation, the survival and/or the differentiation of a wide variety of cells, as well as in embryonic development, angiogenesis, oncogenesis and tumor progression (Reviews: 1,4,5). In several types of tumors and in different cancer cell lines, both factors are expressed and confer to the cells, a serum- and/or a growth factor-independent proliferation. Human breast cancer, which is primarily of epithelial origin, is the most common solid tumor in women in the western world. Depending on the normal, benign and malignant phenotype of mammary glands, FGF1 and FGF2 expressions differ. FGF2 mRNA are expressed at a significant higher level in normal tissues and in benign breast tumors than in malignant breast cancer tissues (6,7,8). Concerning FGF1 expression in breast tissues, Anandappa et al. (7) observed that FGF1 mRNA level is higher in normal tissues or in benign tumors than in malignant cancer tissues, in contrast, Smith et al. (8) showed that the protein level of FGF1 was higher in malignant tissues. In breast cancer cell lines, FGFs expression has been investigated: HBL-100 and MDA-MB-231 cell lines expressed FGF2 mRNA while no expression was detected in MCF7 cell line (6,7,9), MDA-MB-231 and MCF-7 cell lines expressed FGF1 mRNA (10). In vitro, both factors stimulated the proliferation of breast cancer cell lines (BCC) and normal mammary epithelial cells (MEC) (8,11,12,13,14). The discrepency in protein and mRNA FGF1 levels in breast cancer could be explained by the 1 Present address: UMR CNRS 146, Oncogénèse rétrovirale et moléculaire, centre u niversitaire 91405 Orsay cedexFrance.
679 0006-291X/96 $18.00 Copyright © 1996 by Academic Press, Inc. All rights of reproduction in any form reserved.
JOBNAME: BBRC 219#3 PAGE: 2 SESS: 12 OUTPUT: Tue Apr 2 09:18:08 1996 /xypage/worksmart/tsp000/68953f/2
Vol. 219, No. 3, 1996
BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS
complexity of the FGF1 gene. The FGF1 gene encodes multiples transcripts which differ in their 39 and/or 59 untranslated regions. Multiple promoters (15,16), alternative splicing (15,17) and alternative transcriptional terminations (18,19) generate multiple transcripts expressed in a tissueand cell-specific manner (15,17,20). The biological significance of the alternative untranslated sequences in the different FGF1 transcripts remains to be characterized, however, previous studies have shown that the untranslated sequences of other oncogenes or growth factors are involved in the stability (21), the translation (22) or in the compartmentalization of these mRNA (23). Therefore, the expression of different FGF1 mRNA in breast cancer tissues could result in different mRNA stability and/or different translation efficiencies. Serum has been shown to regulate FGF1 expression and depending on the cell type, it induced either an increase in the level of FGF1 mRNA as observed in neonatal foreskin fibroblasts (24) and in vascular smooth muscles cells (25,26), or a decrease in the transcripts as noted in epithelial lens cells (27). In human vascular smooth muscles cells, it has been shown that the different FGF1 mRNAs increased in response to serum (26). In this report, we examined the expression pattern of the different human FGF1 transcripts (1.A, 1.B, 1.C and 1.D), which contained heterogenous 59UTR, in different human mammary epithelial cell lines: primary culture of normal cells (NMEC), a transformed but not malignant cell line (HBL-100) and three breast cancer cell lines (BT-20, MCF-7, MDA-MB-231). The last three cancer cell lines present a transformed phenotype and induced tumors and metastases in nude mice with different potency, the MDA-MB-231 cell line being the most malignant one. The influence of serum on level of the different FGF1 mRNA in these different epithelial cell lines has also been explored. To confirm in vitro, the previous results obtained regarding FGF2 expression in breast tissue samples and to determine if serum may also regulates FGF2 expression, the level of FGF2 mRNA was examined in these different mammary epithelial cell lines in the presence or in the absence of serum. MATERIALS AND METHODS Cell cultures. The human normal mammary epithelial cells (NMEC) were obtained from women mammaplasty reduction and were grown according to the techniques described by Soule and Mc Grath (28). These cells were a generous gift from Dr. X. Dong-Le Bourhis. The human cells MCF-7, BT-20 and MDA-MB-231 are three established carcinoma cell lines derived from human breast cancers, which present an epithelial-like phenotype and induce, in nude mice, adenocarcinoma consistent with mammary cancer (29,30,31,32). HBL-100 is a human breast epithelial cell line reportedly non tumorigenic in nude mice but which was able to form colonies in soft agar (33). These different cell lines were grown in Eagle’s minimal essential medium Earle’s salt (Gibco-BRL) containing 10% fetal calf serum (Gibco-BRL), 1% non essential amino acids, 2 mM glutamine (Eurobio), 5mg/ml insulin (Organon, Sérifontaine) and antibiotics (40 mg/ml streptomycin; 40 UI/ml penicillin, Gibco BRL) at 37°C in 5% CO2: 95% air humidified incubator. Serum regulation assay. Cells were cultured in complete medium as described above until 50% confluence was obtained. The cells were then washed twice with PBS, and cultured in serum-free media supplemented with 30mg/ml transferrin (Eurobio), 2mg/ml fibronectin (Eurobio), 1% non essential amino-acids, 2 mM glutamine, 5mg/ml insulin and antibiotics. After 24 hours, half of the serum starved cells were refed with serum containing medium while the other half were maintained under serum-free medium for a further 24 hours. RT–PCR assay. RNA preparation and RT-PCR assay were performed as previously described by Renaud et al. (27). Total RNA was isolated from cultured cells using the guanidium isothiocyanate method (34). One mg RNA and 170 pg tobacco leaf nitrate reductase transcripts were reverse transcribed in 30 ml of 50 mM Tris HCl pH 8.9, 3 mM MgCl2, 75 mM KCl, 2.5 mM random hexanucleotide primers, 300 U Mo MMLV (Gibco BRL) and one tenth of the reverse transcripts were amplified in 100ml 50 mM Tris HCl pH 8.9, 7 mM MgCl2, 50 mM KCl, 15 mM ammonium sulfate, 0.17 mg/ml BSA, 1 mM of dNTP mix, 15 pmoles of each specific primer by 1U Taq polymerase (Eurobio). In preliminary experiments, the level of FGF1 or FGF2 mRNA in the different cell lines was tested by RT-PCR after 20 to 40 cycles of amplification (denaturation: 92°C for 300; annealing: 62°C for 1; extension: 72°C for 300), and the linear exponential FGF1/FGF2 amplification phase determined. To insure that the amplification was in the exponential phase, aliquots were withdrawn at 3 different cycles and analyzed. For FGF1 and FGF2 amplification, the oligomers were chosen in different exons to avoid amplification of genomic DNA. Non-reverse transcribed RNA was also submitted to the PCR amplification and used as a negative control for DNA contamination (data not shown). To assess the RT-PCR efficiency, tobacco nitrate reductase (NR) transcripts were added to the cellular RNA, reverse transcribed and amplified (data not shown). The sequence and the 680
JOBNAME: BBRC 219#3 PAGE: 3 SESS: 15 OUTPUT: Tue Apr 2 09:18:08 1996 /xypage/worksmart/tsp000/68953f/2
Vol. 219, No. 3, 1996
BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS
location of the primers used for the different amplifications are described in the Tables 1 and 2. The FGF1, FGF2 and NR amplified products were electrophoresed on 10% polyacrylamide gel, blotted onto Hybond N+ and hybridized with specific FGF1, FGF2 and NR probes (Table 2). Depending of the intensity of the signal, X-OMAT AR5 X-ray Film (Kodak) was exposed for different periods of time.
RESULTS FGF1 and FGF2 Expression in Different Mammary Epithelial Cell Lines One microgram of total RNA isolated from the 5 mammary epithelial cell lines (NMEC, HBL-100, BT-20, MCF-7 and MDA-MB-231) was submitted to RT-PCR assay for FGF1 and FGF2 transcripts. After 30 cycles of amplification, specific FGF1 fragment was visualized in NMEC, HBL-100 and MDA-MB-231 cell lines (Figure 1A lanes 1, 2, 5) but not in BT-20 and MCF-7 cell lines (Figure 1A lanes 3, 4). A FGF1 amplified fragment was detected in these last two cell lines only after 40 cycles (Figure 1B lanes 3, 4). FGF2 mRNA were amplified in NMEC and HBL-100 (Figure 1A lanes 1, 2) after 30 cycles and in MDA-MB-231 (Figure 1C lane 5) and BT-20 cell line (Figure 1B lane 3) after 40 cycles. In MCF-7 cell line, no FGF2 amplified products were detected even after 40 cycles (Figure 1B lane 4). Regulation of FGF1 and FGF2 Expression by Serum in Human mammary epithelial cells To study the influence of serum on FGF1 and FGF2 expression, the mammary epithelial cell lines were cultured in the presence or in the absence of serum for 24h and RT-PCR was performed. As illustrated in Figure 2, the effect of serum on the steady state levels of FGF1 mRNA depends of the cell lines. In NMEC and HBL-100 cells, the level of FGF1 amplified products was higher in serum-starved cells (Figure 2A lanes 2, 4) than in cells maintained in serum containing medium (lanes 1, 3). In contrast, the level of FGF1 amplified products in MDA-MB-231 was higher in cells cultured in serum containing medium (lane 5) than in serum deprived cultures (lane 6). In BT-20 and MCF-7 cell lines, the serum has no effect on FGF1 expression (Figure 2B). The level of FGF2 mRNA in NMEC, HBL-100 and MDA-MB-231 was not modified by serum deprivation (data not shown). TABLE 1 Sequence and Location of the Different Primers Used in RT–PCR Assay Specificity FGF1
FGF2
NR
Primers
Sequences (59 → 39)
Position*
COD For COD Rev E1P EAP EAH EBP EBH ECP ECH EDP E1H BB For BB Rev F-bF NR For NR Rev
AAGCCCGTCGGTGTCCATGG GATGGCACAGTGGATGGGAC GTCCCATCCACTGTGCCATC ATCCCACAGCCTTCGCTCCA AATCAGGGCATCGCCTCCTTT CACTCAGAGCTGCAGTAGCCT CTACTCTGAGAAGAAGACACC ATTCCTTAGTGAGTGAGTTCAC CACTTCTGCAGGGAAGCCAGC TGGCAGCAGCACAATGTTTGGGCTA ATGGCTGAAGGGGAAATCACC CCCAGTTCGTTTCAGTGCCACC CTATCAAAGGAGTGTGTGCAAACC TGGAGTATTTCCGTGACCG GCTGGATCCATTGCAAATTCC AGGAGCTGATGTGTTGCCCGG
261 to 242 127 to 146 146 to 127 −226 to −207 −202 to −182 −173 to −153 −133 to −113 −110 to −89 −68 to −48 −67 to −43 1 to 21 393 to 372 219 to 241 364 to 346 284 to 264 207 to 227
* The numerations are all initiated to the translation initiation codon ATG. 681
Orientation antisense sense antisense sense sense sense sense sense sense sense sense antisense sense antisense antisense sense
JOBNAME: BBRC 219#3 PAGE: 4 SESS: 14 OUTPUT: Tue Apr 2 09:18:08 1996 /xypage/worksmart/tsp000/68953f/2
Vol. 219, No. 3, 1996
BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS
TABLE 2 RT–PRC Amplification (Specificity, RT–PCR Primers, Amplified Fragment, Probes) RT–PCR Specificity FGF1 FGF1 FGF1 FGF1 FGF1 FGF2 NR
COD E1.A E1.B E1.C E1.D
RT–PCR primers COD Rev/COD For E1P/E1AP E1P/E1BP E1P/E1CP E1P/E1DP BB Rev/BB For NR Rev/NR For
Amplified fragment
Hybridization primers or cDNA
135 bp 372 bp 319 bp 256 bp 213 bp 175 bp 75 bp
FGF1 cDNA E1AH E1BH E1CH E1H F-bF NR cDNA
Regulation by Serum of the Level of the Different FGF1 mRNA in NMEC, HBL-100 and MDA-MB-231 Cells Depending on the cell lines, different patterns of expression of the four FGF1 transcripts (1.A, 1.B, 1.C and 1.D) were obtained (Figure 3). When the cells are cultured in complete medium (serum-containing medium), the FGF1 mRNA 1.A was not detected in NMEC, HBL-100 and MDA-MB-231 cells (lanes 1, 3, 5). The FGF1 mRNA 1.B was only amplified in NMEC (lane 1), the FGF1 mRNA 1.C was amplified in normal cells NMEC (lane 1) and in the most malignant cell line MDA-MB-231 (lane 5) and the FGF1 mRNA 1.D was amplified in the three cell lines (NMEC, HBL100, MDA-MB-231) (lanes 1, 3, 5). In MCF-7 and BT-20, only FGF1 mRNA 1.D was detected at a very low level (data not shown). In these cells, serum starvation modified the expression pattern of the different FGF1 transcripts. Serum-deprived NMEC expressed FGF1 mRNA 1.A (lane 2) suggesting a negative regulation of this transcript by serum (lane 1, 2). In HBL-100 and MDA-MB-231, FGF1 mRNA 1.A was detected neither in the presence nor in the absence of serum. A negative regulation by serum was also observed for the FGF1 mRNA 1.B in NMEC (lanes 1, 2) and MDA-MB-231 (lanes 5, 6). No amplification of this transcript was detected in HBL-100 in the presence or in the absence of serum (lane 3, 4). Concerning FGF1 mRNA 1.C expression, its regulation by serum was more complex. For NMEC (lanes 1, 2) and MDA-MB-231 (lanes 5, 6), a positive regulation of its expression by
FIG. 1. FGF1 and FGF2 transcripts in human mammary epithelial cell lines.One microgram of total RNA extracted from human normal mammary epithelial cells (NMEC, lane 1) and from four different human epithelial cell lines (HBL-100: lane 2; BT-20: lane 3; MCF-7: lane 4; MDA-MB-231: lane 5) were reverse transcribed. One-tenth of the reverse transcript was amplified by PCR using FGF1 and FGF2 coding specific primers as described in Materials and Methods. Depending on the level of FGF1 or FGF2, different numbers of PCR cycle were performed (A: 30 cycles, B and C: 40 cycles). The amplified products were separated on 10% acrylamide gel, transferred to Hybond N+ membranes and hybridized with 32P labelled FGF1 and FGF2 internal specific primers. 682
JOBNAME: BBRC 219#3 PAGE: 5 SESS: 13 OUTPUT: Tue Apr 2 09:18:08 1996 /xypage/worksmart/tsp000/68953f/2
Vol. 219, No. 3, 1996
BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS
FIG. 2. Regulation of FGF1 mRNA level by serum in human mammary epithelial cells. Human normal mammary epithelial cells (NMEC, lanes 1 and 2) and mammary epithelial cell lines (HBL-100: lanes 3 and 4; MDA-MB-231: lanes 5 and 6; BT-20: lanes 7 and 8; MCF-7: lanes 9 and 10) were maintained 24 hours in the presence (lanes 1, 3, 5, 7, 9) or in the absence (lanes 2, 4, 6, 8, 10) of 10% fetal calf serum (FCS). RNA were then submitted to RT-PCR using FGF1 coding primers. FGF1 amplified products issued from 30 cycles (lanes 1–6) or from 40 cycles (lanes 7–10) were analyzed by Southern blot.
serum was observed, however, a negative regulation was detected in the HBL-100 (lanes 3, 4). A positive regulation by serum of FGF1 mRNA 1.D was observed in the three mammary cell types analyzed (lanes 1–6). DISCUSSION In this report, we show the distribution of FGF1 and FGF2 mRNAs in normal, transformed and malignant mammary epithelial cells. Depending on the cell line, the FGF1 and FGF2 genes have different levels of expression. In malignant cell lines which induce tumors and metastasis in nude mice, FGF2 mRNA was either not expressed (MCF-7) or expressed at a lower level in BT-20 and MDA-MB-231 cell lines than in normal (NMEC) and transformed mammary epithelial cells (HBL-100). These results indicate that the expression of the FGF2 mRNA appears to be associated with a non malignant phenotype as previously described in normal, benign cancer breast tissues and in malignant mammary tumors (6, 7, 8). In contrast, the expression of the FGF1 mRNAs appears not to correlate with the tumorigenic phenotype of mammary epithelial cells. Normal cells (NMEC), transformed cells (HBL-100) and the tumorigenic cell line (MDA-MB-231) expressed similar levels of FGF1 and greater than that noted in the two tumorigenic cell lines (BT-20 and MCF-7). In contrast, Anandappa et al. (7) reported a lower level of FGF1 mRNA in malignant breast cancer tissues than in normal tissue and benign tumors. This discrepancy may be due to the lost in clonal cell lines of some characteristics of primary breast cancers from which these cells are derived even though they may mimic certain tumorogenic aspects of mammary glands.
FIG. 3. Regulation of the steady state level of the FGF1 transcripts by serum in human mammary epithelial cells. RNA isolated from NMEC (lanes 1 and 2), HBL-100 (lanes 3 and 4) and MDA-MB-231 (lanes 5 and 6) cells cultured in the presence (lanes 1, 3 and 5) or in the absence (lanes 2, 4 and 6) of 10% fetal calf serum for 24 hours were submitted to RT-PCR. All FGF1 mRNA were amplified with primers located in the coding region and the FGF1 mRNA 1.A, 1.B, 1.C and 1.D were amplified with specific primers as described in Materials and Methods. For the amplification of the coding region (COD), 30 cycles were performed; and 35 cycles to amplify each transcript. The amplified products were separated in 10% acrylamide gel, transferred to Hybond N+ membrane and hybridized with specific internal primers. 683
JOBNAME: BBRC 219#3 PAGE: 6 SESS: 13 OUTPUT: Tue Apr 2 09:18:08 1996 /xypage/worksmart/tsp000/68953f/2
Vol. 219, No. 3, 1996
BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS
In this report, we demonstrate a differential but non exclusive expression of the different FGF1 mRNAs in the human mammary cell lines. Depending on the cell line, the different transcripts were either coexpressed or expressed in a more restricted pattern. The mRNA 1.A was detected neither in primary cultured cells nor in the established cell lines while the mRNA 1.B appeared to be specifically expressed in normal cells (NMEC). The mRNA 1.C and 1.D have different patterns of expression in the normal cells (NMEC) and in the established cell lines. The transcript 1.D was detected in all cell lines while the mRNA 1.C was expressed only in the MDA-MB-231 cell line. Thus, no correlation could be established between the transformed phenotype of mammary cells and the expression of a specific mRNA which could be involved in the increase level of FGF1 observed in malignant breast tumors (8). In particular, the transcripts 1.C and 1.D previously described in transformed human glioblastomas cell lines but not in normal human tissues (brain, prostate and kidney) were recently characterized in human fibroblast and smooth vascular cells (20, 26). These data confirmed that, like in mammary tumorigenic cell lines, the mRNA 1.C and 1.D are not characteristic of the transformed phenotype. In contrast, the FGF1 mRNA 1.B is detected only in normal cells and it will be of great interest to determine if this transcript is expressed in breast cancer tissues. In this report, we have shown that, depending on the cell type, the level of FGF1 mRNA was unchanged, up- or down-regulated in the absence of serum. It was unaffected in the two malignant cell lines (BT-20 and MCF7), increased in NMEC and HBL-100 cells and decreased in the MDA-MB-231 cells. In contrast, the level of FGF2 mRNA was unaffected by serum starvation in all cell lines. The regulation of FGF1 mRNA expression by serum did not correlate with the malignancy of the cells but appeared to be cell specific. A regulation of FGF1 mRNA by serum has been described in other cell systems. In lens epithelial cells, the steady state level of FGF1 mRNA increased in serum starved cells (27) while in human vascular smooth muscle cells the opposite was described confirming the cell specificity of the FGF1 regulation by serum (25, 26). The FGF1 mRNAs containing 1.A, 1.B, 1.C and 1.D untranslated exons are distributed in a cell specific manner and are differently regulated by serum. The 1.A and 1.B FGF1 mRNAs are predominantly expressed in serum starved cells (NMEC, MDA-MB-231) while the expression of 1.C and 1.D FGF1 mRNAs are increased in cells cultured in the presence of serum. In contrast, in the HBL-100 cell line, the transcript 1.C is upregulated in the absence of serum while the 1.D is down regulated. The transcription of the human FGF1 1.A, 1.B, 1.C and 1.D transcripts is under the control of distinct promotors (16, 20). Serum responsive elements and cis-regulatory elements involved in the autoactivation by factors such as exogenous FGF1 and testosterone have been characterized (16, 35). Our data suggest that the steady state levels of the multiple FGF1 transcripts in response to serum is under the control of different factors which could regulate the different promotors in a cell specific manner. The mammary epithelial cell lines could help in the identification of these cellular factors. REFERENCES 1. Burgess, W. H., and Maciag, T. (1989) Annu. Rev. Biochem. 58, 575–606. 2. Tanaka, A., Miyamoto, K., Minamino, N., Takeda, M., Sato, B., Matsuo, H., and Matsumoto, K. (1992) Proc. Natl. Acad. Sci. U.S.A. 89, 8928–8932. 3. Miyamoto, M., Naruo, K., Seko, C., Matsumoto, S., Kondo, T., and Kurokawa, T. (1993) Mol. Cell. Biol. 13, 239–249. 4. Folkman, J., and Klagsbrun, M. (1987) Science 235, 442–447. 5. Rifkin, D. B., and Moscatelli, D. (1989) J. Cell. Biol. 109, 1–6. 6. Luqmani, Y. A., Graham, M., and Coombes, R. C. (1992) Brit. J. Cancer 66, 273–280. 7. Anandappa, S. Y., Winstanley, J. H. R., Leinster, S., Green, B., Rudland, P. S., and Barraclough, R. (1994) Br. J. Cancer 69, 772–776. 8. Smith, J., Yelland, A., Baillie, R., and Coombes, R. C. (1994) Eur. J. Cancer 30, 496–503. 9. Souttou, B., Hamelin, R., and Crepin, M. (1994) Cell Growth and Differentiation 5, 615–623. 10. Penault-Llorca, F., Bertucci, F., Adelaide, J., Parc, P., Coulier, F., Jacquemier, J., Birnbaum, D., and DeLapeyriere, O. (1995) Int. J. Cancer 61, 170–176. 684
JOBNAME: BBRC 219#3 PAGE: 7 SESS: 13 OUTPUT: Tue Apr 2 09:18:08 1996 /xypage/worksmart/tsp000/68953f/2
Vol. 219, No. 3, 1996
BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS
11. Levay-Young, B. K., Imagawa, W., Wallace, D. R., and Nandi, S. (1989) Mol. Cell. Endocrinol. 62, 327–336. 12. Briozzo, P., Badet, J., Capony, F., Pieri, I., Montcourrier, P., Barriltault, D., and Rochefort, H. (1991) Exp. Cell Res. 194, 252–259. 13. Immagawa, W., Cunha, G. R., Young, P., and Nandi, S. (1994) Biochem. Biophys. Res. Comm. 204, 1165–1169. 14. El Yazidi, I., and Boilly-Marer, Y. (1995) Anticancer Res. 15, 783–790. 15. Myers, R. L., Payson, R. A., Chotani, M. A., Deaven, L. L., and Chiu, I. M. (1993) Oncogene 8, 341–349. 16. Myers, R. L., Ray, S. K., Eldridge, R., Chotani, M. A., and Chiu, I. M. (1995) J. Biol. Chem. 270, 8257–8266. 17. Philippe, J. M., Renaud, F., Desset, S., Laurent, M., Mallet, J., Courtois, Y., and Dumas Milne Edwards, J. B. (1992) Biochem. Biophys. Res. Comm. 188, 843–850. 18. Chiu, I. M., Wang, W. P., and Lehtoma, K. (1990) Oncogene 5, 755–762. 19. Renaud, F., Desset, S., Bugra, K., Halley, C., Philippe, J. M., Coutois, Y., and Laurent, M. (1992) Biochem. Biophys. Res. Comm. 184, 945–952. 20. Payson, R. A., Canatan, H., Chotani, M. A., Wang, W. P., Harris, S. E., Myers, R. L., and Chiu, I. M. (1993) Nucleic Acids Res. 21, 489–495. 21. Shaw, G., and Kamen, R. A. (1986) Cell 46, 659–667. 22. Rao, C. D., Pech, M., Robbins, K. C., and Aaronson, S. A. (1988) Mol. Cell. Biol. 8, 284–292. 23. Gottlieb, E. (1992) Proc. Natl. Acad. Sci. USA 89, 7164–7168. 24. Cook, P. W., Coffey, R. J., Magun, B. E., Pittelkow, M. R., and Shipley, G. D. (1990) Mol. Endocrinol. 4, 1377–1385. 25. Winkles, J. A., and Gay, C. G. (1991) Cell Growth and Differentiation 2, 531–540. 26. Chotani, M. A., Payson, R. A., Winkles, J. A., and Chiu, I. M. (1995) Nucleic Acids Res. 23, 434–441. 27. Renaud, F., Oliver, L., Desset, S., Tassin, J., Romquin, N., Courtois, Y., and Laurent, M (1994) J. Cell. Physiol. 158, 435–443. 28. Soule, H. D., and McGrath, C. M. (1986) In Vitro Cell. Dev. Biol. 22, 6–12. 29. Lasfargues, E. Y., and Ozzelo, L. (1958) Natl. Cancer Inst. 21, 1131–1147. 30. Soule, H. D., Vasquez, J., Long, A., Albert, S., and Brennan, M. J. (1973) J. Natl. Cancer Inst. 51, 1409–1416. 31. Cailleau, R., Young, R., Olive, M., and Reves, W. J. (1974) J. Natl. Cancer Inst. 53, 661–674. 32. Engel, L. W., and Young, N. A. (1978) Cancer Res. 38, 4327–4339. 33. Gaffney, E. V. (1982) Cell Tissue Res. 227, 563–568. 34. Chirwing, J. M., Przybyla, A. E., Mac Donald, R. J., and Rutter, W. (1979) Biochemistry 18, 5294–5299. 35. Hall, J. A., Harris, M. A., Intres, R., and Harris, S. E. (1993) J. Cell. Biochem. 51, 116–127.
685
BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS ARTICLE NO.
219, 679–685 (1996)
0294
Expression and Regulation by Serum of Multiple FGF1 mRNA in Normal Transformed, and Malignant Human Mammary Epithelial Cells F. Renaud,* I. El Yazidi,†,‡ Y. Boilly-Marer,† Y. Courtois,* and M. Laurent1 *Laboratoire INSERM U450, 29 rue Wilhem, 75016 Paris, France; and †Laboratoire CNRS UMR 111, Batiment C9, Université des Sciences et Technologies de Lille, 59655 Villeneuve D’Ascq Cedex, France, and ‡Laboratoire de Biologie du Développement, Université des Sciences et Technologies de Lille, 59655 Villeneuve D’Ascq Cedex, France Received January 12, 1996 In normal (NMEC), transformed (HBL-100) and malignant human mammary epithelial cells (MCF 7, BT-20, MDA-MB 231), we have examined the expression and the regulation by serum of FGF1 and FGF2 mRNA. FGF2 mRNA level was higher in NMEC and in a HBL-100 than in malignant cell lines (MDA-MB-231, BT-20). No FGF2 mRNA was detected in the malignant cell line, MCF-7. In contrast, the FGF1 mRNA was detected in all the mammary epithelial cells but at different levels. NMEC, HBL-100 and MDA-MB-231 cells expressed similar level of FGF1 and higher than that observed in BT-20 and MCF-7. In contrast to FGF2 which is only expressed in nonmalignant cells, no correlation between FGF1 mRNA expression and the phenotype of the cells was observed. We followed the expression of four FGF1 mRNA, heterogenous in their 59 untranslated regions. This study demonstrated that (i) the FGF1 mRNA 1.A was not expressed by mammary epithelial cells, (ii) the FGF1 mRNA 1.B was only expressed in normal mammary epithelial cells and (iii) the transcripts 1.C and 1.D were expressed in normal and malignant cells with specific patterns. The expression of FGF1 mRNAs responded in a cell specific manner to serum starvation. The mRNA 1.A was only expressed in normal cells cultured in the absence of serum while 1.C was either up- or down-regulated by serum in transformed cells and the expression of 1.D was greater in presence of serum in all cell lines. These results show that the regulation of FGF1 mRNAs expression is cell specific and does not correlate with a tumorigenic or transformed cell phenotype. © 1996 Academic Press, Inc.
Acidic fibroblast growth factor (FGF-1) and basic fibroblast growth factor (FGF2) are the prototype members of the fibroblast growth factor family which contains 9 growth factors and/or oncogenes (FGF-1 to FGF-9) (1,2,3). FGF1 and FGF2 are multipotent factors involved in the proliferation, the survival and/or the differentiation of a wide variety of cells, as well as in embryonic development, angiogenesis, oncogenesis and tumor progression (Reviews: 1,4,5). In several types of tumors and in different cancer cell lines, both factors are expressed and confer to the cells, a serum- and/or a growth factor-independent proliferation. Human breast cancer, which is primarily of epithelial origin, is the most common solid tumor in women in the western world. Depending on the normal, benign and malignant phenotype of mammary glands, FGF1 and FGF2 expressions differ. FGF2 mRNA are expressed at a significant higher level in normal tissues and in benign breast tumors than in malignant breast cancer tissues (6,7,8). Concerning FGF1 expression in breast tissues, Anandappa et al. (7) observed that FGF1 mRNA level is higher in normal tissues or in benign tumors than in malignant cancer tissues, in contrast, Smith et al. (8) showed that the protein level of FGF1 was higher in malignant tissues. In breast cancer cell lines, FGFs expression has been investigated: HBL-100 and MDA-MB-231 cell lines expressed FGF2 mRNA while no expression was detected in MCF7 cell line (6,7,9), MDA-MB-231 and MCF-7 cell lines expressed FGF1 mRNA (10). In vitro, both factors stimulated the proliferation of breast cancer cell lines (BCC) and normal mammary epithelial cells (MEC) (8,11,12,13,14). The discrepency in protein and mRNA FGF1 levels in breast cancer could be explained by the 1 Present address: UMR CNRS 146, Oncogénèse rétrovirale et moléculaire, centre u niversitaire 91405 Orsay cedexFrance.
679 0006-291X/96 $18.00 Copyright © 1996 by Academic Press, Inc. All rights of reproduction in any form reserved.
JOBNAME: BBRC 219#3 PAGE: 2 SESS: 12 OUTPUT: Tue Apr 2 09:18:08 1996 /xypage/worksmart/tsp000/68953f/2
Vol. 219, No. 3, 1996
BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS
complexity of the FGF1 gene. The FGF1 gene encodes multiples transcripts which differ in their 39 and/or 59 untranslated regions. Multiple promoters (15,16), alternative splicing (15,17) and alternative transcriptional terminations (18,19) generate multiple transcripts expressed in a tissueand cell-specific manner (15,17,20). The biological significance of the alternative untranslated sequences in the different FGF1 transcripts remains to be characterized, however, previous studies have shown that the untranslated sequences of other oncogenes or growth factors are involved in the stability (21), the translation (22) or in the compartmentalization of these mRNA (23). Therefore, the expression of different FGF1 mRNA in breast cancer tissues could result in different mRNA stability and/or different translation efficiencies. Serum has been shown to regulate FGF1 expression and depending on the cell type, it induced either an increase in the level of FGF1 mRNA as observed in neonatal foreskin fibroblasts (24) and in vascular smooth muscles cells (25,26), or a decrease in the transcripts as noted in epithelial lens cells (27). In human vascular smooth muscles cells, it has been shown that the different FGF1 mRNAs increased in response to serum (26). In this report, we examined the expression pattern of the different human FGF1 transcripts (1.A, 1.B, 1.C and 1.D), which contained heterogenous 59UTR, in different human mammary epithelial cell lines: primary culture of normal cells (NMEC), a transformed but not malignant cell line (HBL-100) and three breast cancer cell lines (BT-20, MCF-7, MDA-MB-231). The last three cancer cell lines present a transformed phenotype and induced tumors and metastases in nude mice with different potency, the MDA-MB-231 cell line being the most malignant one. The influence of serum on level of the different FGF1 mRNA in these different epithelial cell lines has also been explored. To confirm in vitro, the previous results obtained regarding FGF2 expression in breast tissue samples and to determine if serum may also regulates FGF2 expression, the level of FGF2 mRNA was examined in these different mammary epithelial cell lines in the presence or in the absence of serum. MATERIALS AND METHODS Cell cultures. The human normal mammary epithelial cells (NMEC) were obtained from women mammaplasty reduction and were grown according to the techniques described by Soule and Mc Grath (28). These cells were a generous gift from Dr. X. Dong-Le Bourhis. The human cells MCF-7, BT-20 and MDA-MB-231 are three established carcinoma cell lines derived from human breast cancers, which present an epithelial-like phenotype and induce, in nude mice, adenocarcinoma consistent with mammary cancer (29,30,31,32). HBL-100 is a human breast epithelial cell line reportedly non tumorigenic in nude mice but which was able to form colonies in soft agar (33). These different cell lines were grown in Eagle’s minimal essential medium Earle’s salt (Gibco-BRL) containing 10% fetal calf serum (Gibco-BRL), 1% non essential amino acids, 2 mM glutamine (Eurobio), 5mg/ml insulin (Organon, Sérifontaine) and antibiotics (40 mg/ml streptomycin; 40 UI/ml penicillin, Gibco BRL) at 37°C in 5% CO2: 95% air humidified incubator. Serum regulation assay. Cells were cultured in complete medium as described above until 50% confluence was obtained. The cells were then washed twice with PBS, and cultured in serum-free media supplemented with 30mg/ml transferrin (Eurobio), 2mg/ml fibronectin (Eurobio), 1% non essential amino-acids, 2 mM glutamine, 5mg/ml insulin and antibiotics. After 24 hours, half of the serum starved cells were refed with serum containing medium while the other half were maintained under serum-free medium for a further 24 hours. RT–PCR assay. RNA preparation and RT-PCR assay were performed as previously described by Renaud et al. (27). Total RNA was isolated from cultured cells using the guanidium isothiocyanate method (34). One mg RNA and 170 pg tobacco leaf nitrate reductase transcripts were reverse transcribed in 30 ml of 50 mM Tris HCl pH 8.9, 3 mM MgCl2, 75 mM KCl, 2.5 mM random hexanucleotide primers, 300 U Mo MMLV (Gibco BRL) and one tenth of the reverse transcripts were amplified in 100ml 50 mM Tris HCl pH 8.9, 7 mM MgCl2, 50 mM KCl, 15 mM ammonium sulfate, 0.17 mg/ml BSA, 1 mM of dNTP mix, 15 pmoles of each specific primer by 1U Taq polymerase (Eurobio). In preliminary experiments, the level of FGF1 or FGF2 mRNA in the different cell lines was tested by RT-PCR after 20 to 40 cycles of amplification (denaturation: 92°C for 300; annealing: 62°C for 1; extension: 72°C for 300), and the linear exponential FGF1/FGF2 amplification phase determined. To insure that the amplification was in the exponential phase, aliquots were withdrawn at 3 different cycles and analyzed. For FGF1 and FGF2 amplification, the oligomers were chosen in different exons to avoid amplification of genomic DNA. Non-reverse transcribed RNA was also submitted to the PCR amplification and used as a negative control for DNA contamination (data not shown). To assess the RT-PCR efficiency, tobacco nitrate reductase (NR) transcripts were added to the cellular RNA, reverse transcribed and amplified (data not shown). The sequence and the 680
JOBNAME: BBRC 219#3 PAGE: 3 SESS: 15 OUTPUT: Tue Apr 2 09:18:08 1996 /xypage/worksmart/tsp000/68953f/2
Vol. 219, No. 3, 1996
BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS
location of the primers used for the different amplifications are described in the Tables 1 and 2. The FGF1, FGF2 and NR amplified products were electrophoresed on 10% polyacrylamide gel, blotted onto Hybond N+ and hybridized with specific FGF1, FGF2 and NR probes (Table 2). Depending of the intensity of the signal, X-OMAT AR5 X-ray Film (Kodak) was exposed for different periods of time.
RESULTS FGF1 and FGF2 Expression in Different Mammary Epithelial Cell Lines One microgram of total RNA isolated from the 5 mammary epithelial cell lines (NMEC, HBL-100, BT-20, MCF-7 and MDA-MB-231) was submitted to RT-PCR assay for FGF1 and FGF2 transcripts. After 30 cycles of amplification, specific FGF1 fragment was visualized in NMEC, HBL-100 and MDA-MB-231 cell lines (Figure 1A lanes 1, 2, 5) but not in BT-20 and MCF-7 cell lines (Figure 1A lanes 3, 4). A FGF1 amplified fragment was detected in these last two cell lines only after 40 cycles (Figure 1B lanes 3, 4). FGF2 mRNA were amplified in NMEC and HBL-100 (Figure 1A lanes 1, 2) after 30 cycles and in MDA-MB-231 (Figure 1C lane 5) and BT-20 cell line (Figure 1B lane 3) after 40 cycles. In MCF-7 cell line, no FGF2 amplified products were detected even after 40 cycles (Figure 1B lane 4). Regulation of FGF1 and FGF2 Expression by Serum in Human mammary epithelial cells To study the influence of serum on FGF1 and FGF2 expression, the mammary epithelial cell lines were cultured in the presence or in the absence of serum for 24h and RT-PCR was performed. As illustrated in Figure 2, the effect of serum on the steady state levels of FGF1 mRNA depends of the cell lines. In NMEC and HBL-100 cells, the level of FGF1 amplified products was higher in serum-starved cells (Figure 2A lanes 2, 4) than in cells maintained in serum containing medium (lanes 1, 3). In contrast, the level of FGF1 amplified products in MDA-MB-231 was higher in cells cultured in serum containing medium (lane 5) than in serum deprived cultures (lane 6). In BT-20 and MCF-7 cell lines, the serum has no effect on FGF1 expression (Figure 2B). The level of FGF2 mRNA in NMEC, HBL-100 and MDA-MB-231 was not modified by serum deprivation (data not shown). TABLE 1 Sequence and Location of the Different Primers Used in RT–PCR Assay Specificity FGF1
FGF2
NR
Primers
Sequences (59 → 39)
Position*
COD For COD Rev E1P EAP EAH EBP EBH ECP ECH EDP E1H BB For BB Rev F-bF NR For NR Rev
AAGCCCGTCGGTGTCCATGG GATGGCACAGTGGATGGGAC GTCCCATCCACTGTGCCATC ATCCCACAGCCTTCGCTCCA AATCAGGGCATCGCCTCCTTT CACTCAGAGCTGCAGTAGCCT CTACTCTGAGAAGAAGACACC ATTCCTTAGTGAGTGAGTTCAC CACTTCTGCAGGGAAGCCAGC TGGCAGCAGCACAATGTTTGGGCTA ATGGCTGAAGGGGAAATCACC CCCAGTTCGTTTCAGTGCCACC CTATCAAAGGAGTGTGTGCAAACC TGGAGTATTTCCGTGACCG GCTGGATCCATTGCAAATTCC AGGAGCTGATGTGTTGCCCGG
261 to 242 127 to 146 146 to 127 −226 to −207 −202 to −182 −173 to −153 −133 to −113 −110 to −89 −68 to −48 −67 to −43 1 to 21 393 to 372 219 to 241 364 to 346 284 to 264 207 to 227
* The numerations are all initiated to the translation initiation codon ATG. 681
Orientation antisense sense antisense sense sense sense sense sense sense sense sense antisense sense antisense antisense sense
JOBNAME: BBRC 219#3 PAGE: 4 SESS: 14 OUTPUT: Tue Apr 2 09:18:08 1996 /xypage/worksmart/tsp000/68953f/2
Vol. 219, No. 3, 1996
BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS
TABLE 2 RT–PRC Amplification (Specificity, RT–PCR Primers, Amplified Fragment, Probes) RT–PCR Specificity FGF1 FGF1 FGF1 FGF1 FGF1 FGF2 NR
COD E1.A E1.B E1.C E1.D
RT–PCR primers COD Rev/COD For E1P/E1AP E1P/E1BP E1P/E1CP E1P/E1DP BB Rev/BB For NR Rev/NR For
Amplified fragment
Hybridization primers or cDNA
135 bp 372 bp 319 bp 256 bp 213 bp 175 bp 75 bp
FGF1 cDNA E1AH E1BH E1CH E1H F-bF NR cDNA
Regulation by Serum of the Level of the Different FGF1 mRNA in NMEC, HBL-100 and MDA-MB-231 Cells Depending on the cell lines, different patterns of expression of the four FGF1 transcripts (1.A, 1.B, 1.C and 1.D) were obtained (Figure 3). When the cells are cultured in complete medium (serum-containing medium), the FGF1 mRNA 1.A was not detected in NMEC, HBL-100 and MDA-MB-231 cells (lanes 1, 3, 5). The FGF1 mRNA 1.B was only amplified in NMEC (lane 1), the FGF1 mRNA 1.C was amplified in normal cells NMEC (lane 1) and in the most malignant cell line MDA-MB-231 (lane 5) and the FGF1 mRNA 1.D was amplified in the three cell lines (NMEC, HBL100, MDA-MB-231) (lanes 1, 3, 5). In MCF-7 and BT-20, only FGF1 mRNA 1.D was detected at a very low level (data not shown). In these cells, serum starvation modified the expression pattern of the different FGF1 transcripts. Serum-deprived NMEC expressed FGF1 mRNA 1.A (lane 2) suggesting a negative regulation of this transcript by serum (lane 1, 2). In HBL-100 and MDA-MB-231, FGF1 mRNA 1.A was detected neither in the presence nor in the absence of serum. A negative regulation by serum was also observed for the FGF1 mRNA 1.B in NMEC (lanes 1, 2) and MDA-MB-231 (lanes 5, 6). No amplification of this transcript was detected in HBL-100 in the presence or in the absence of serum (lane 3, 4). Concerning FGF1 mRNA 1.C expression, its regulation by serum was more complex. For NMEC (lanes 1, 2) and MDA-MB-231 (lanes 5, 6), a positive regulation of its expression by
FIG. 1. FGF1 and FGF2 transcripts in human mammary epithelial cell lines.One microgram of total RNA extracted from human normal mammary epithelial cells (NMEC, lane 1) and from four different human epithelial cell lines (HBL-100: lane 2; BT-20: lane 3; MCF-7: lane 4; MDA-MB-231: lane 5) were reverse transcribed. One-tenth of the reverse transcript was amplified by PCR using FGF1 and FGF2 coding specific primers as described in Materials and Methods. Depending on the level of FGF1 or FGF2, different numbers of PCR cycle were performed (A: 30 cycles, B and C: 40 cycles). The amplified products were separated on 10% acrylamide gel, transferred to Hybond N+ membranes and hybridized with 32P labelled FGF1 and FGF2 internal specific primers. 682
JOBNAME: BBRC 219#3 PAGE: 5 SESS: 13 OUTPUT: Tue Apr 2 09:18:08 1996 /xypage/worksmart/tsp000/68953f/2
Vol. 219, No. 3, 1996
BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS
FIG. 2. Regulation of FGF1 mRNA level by serum in human mammary epithelial cells. Human normal mammary epithelial cells (NMEC, lanes 1 and 2) and mammary epithelial cell lines (HBL-100: lanes 3 and 4; MDA-MB-231: lanes 5 and 6; BT-20: lanes 7 and 8; MCF-7: lanes 9 and 10) were maintained 24 hours in the presence (lanes 1, 3, 5, 7, 9) or in the absence (lanes 2, 4, 6, 8, 10) of 10% fetal calf serum (FCS). RNA were then submitted to RT-PCR using FGF1 coding primers. FGF1 amplified products issued from 30 cycles (lanes 1–6) or from 40 cycles (lanes 7–10) were analyzed by Southern blot.
serum was observed, however, a negative regulation was detected in the HBL-100 (lanes 3, 4). A positive regulation by serum of FGF1 mRNA 1.D was observed in the three mammary cell types analyzed (lanes 1–6). DISCUSSION In this report, we show the distribution of FGF1 and FGF2 mRNAs in normal, transformed and malignant mammary epithelial cells. Depending on the cell line, the FGF1 and FGF2 genes have different levels of expression. In malignant cell lines which induce tumors and metastasis in nude mice, FGF2 mRNA was either not expressed (MCF-7) or expressed at a lower level in BT-20 and MDA-MB-231 cell lines than in normal (NMEC) and transformed mammary epithelial cells (HBL-100). These results indicate that the expression of the FGF2 mRNA appears to be associated with a non malignant phenotype as previously described in normal, benign cancer breast tissues and in malignant mammary tumors (6, 7, 8). In contrast, the expression of the FGF1 mRNAs appears not to correlate with the tumorigenic phenotype of mammary epithelial cells. Normal cells (NMEC), transformed cells (HBL-100) and the tumorigenic cell line (MDA-MB-231) expressed similar levels of FGF1 and greater than that noted in the two tumorigenic cell lines (BT-20 and MCF-7). In contrast, Anandappa et al. (7) reported a lower level of FGF1 mRNA in malignant breast cancer tissues than in normal tissue and benign tumors. This discrepancy may be due to the lost in clonal cell lines of some characteristics of primary breast cancers from which these cells are derived even though they may mimic certain tumorogenic aspects of mammary glands.
FIG. 3. Regulation of the steady state level of the FGF1 transcripts by serum in human mammary epithelial cells. RNA isolated from NMEC (lanes 1 and 2), HBL-100 (lanes 3 and 4) and MDA-MB-231 (lanes 5 and 6) cells cultured in the presence (lanes 1, 3 and 5) or in the absence (lanes 2, 4 and 6) of 10% fetal calf serum for 24 hours were submitted to RT-PCR. All FGF1 mRNA were amplified with primers located in the coding region and the FGF1 mRNA 1.A, 1.B, 1.C and 1.D were amplified with specific primers as described in Materials and Methods. For the amplification of the coding region (COD), 30 cycles were performed; and 35 cycles to amplify each transcript. The amplified products were separated in 10% acrylamide gel, transferred to Hybond N+ membrane and hybridized with specific internal primers. 683
JOBNAME: BBRC 219#3 PAGE: 6 SESS: 13 OUTPUT: Tue Apr 2 09:18:08 1996 /xypage/worksmart/tsp000/68953f/2
Vol. 219, No. 3, 1996
BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS
In this report, we demonstrate a differential but non exclusive expression of the different FGF1 mRNAs in the human mammary cell lines. Depending on the cell line, the different transcripts were either coexpressed or expressed in a more restricted pattern. The mRNA 1.A was detected neither in primary cultured cells nor in the established cell lines while the mRNA 1.B appeared to be specifically expressed in normal cells (NMEC). The mRNA 1.C and 1.D have different patterns of expression in the normal cells (NMEC) and in the established cell lines. The transcript 1.D was detected in all cell lines while the mRNA 1.C was expressed only in the MDA-MB-231 cell line. Thus, no correlation could be established between the transformed phenotype of mammary cells and the expression of a specific mRNA which could be involved in the increase level of FGF1 observed in malignant breast tumors (8). In particular, the transcripts 1.C and 1.D previously described in transformed human glioblastomas cell lines but not in normal human tissues (brain, prostate and kidney) were recently characterized in human fibroblast and smooth vascular cells (20, 26). These data confirmed that, like in mammary tumorigenic cell lines, the mRNA 1.C and 1.D are not characteristic of the transformed phenotype. In contrast, the FGF1 mRNA 1.B is detected only in normal cells and it will be of great interest to determine if this transcript is expressed in breast cancer tissues. In this report, we have shown that, depending on the cell type, the level of FGF1 mRNA was unchanged, up- or down-regulated in the absence of serum. It was unaffected in the two malignant cell lines (BT-20 and MCF7), increased in NMEC and HBL-100 cells and decreased in the MDA-MB-231 cells. In contrast, the level of FGF2 mRNA was unaffected by serum starvation in all cell lines. The regulation of FGF1 mRNA expression by serum did not correlate with the malignancy of the cells but appeared to be cell specific. A regulation of FGF1 mRNA by serum has been described in other cell systems. In lens epithelial cells, the steady state level of FGF1 mRNA increased in serum starved cells (27) while in human vascular smooth muscle cells the opposite was described confirming the cell specificity of the FGF1 regulation by serum (25, 26). The FGF1 mRNAs containing 1.A, 1.B, 1.C and 1.D untranslated exons are distributed in a cell specific manner and are differently regulated by serum. The 1.A and 1.B FGF1 mRNAs are predominantly expressed in serum starved cells (NMEC, MDA-MB-231) while the expression of 1.C and 1.D FGF1 mRNAs are increased in cells cultured in the presence of serum. In contrast, in the HBL-100 cell line, the transcript 1.C is upregulated in the absence of serum while the 1.D is down regulated. The transcription of the human FGF1 1.A, 1.B, 1.C and 1.D transcripts is under the control of distinct promotors (16, 20). Serum responsive elements and cis-regulatory elements involved in the autoactivation by factors such as exogenous FGF1 and testosterone have been characterized (16, 35). Our data suggest that the steady state levels of the multiple FGF1 transcripts in response to serum is under the control of different factors which could regulate the different promotors in a cell specific manner. The mammary epithelial cell lines could help in the identification of these cellular factors. REFERENCES 1. Burgess, W. H., and Maciag, T. (1989) Annu. Rev. Biochem. 58, 575–606. 2. Tanaka, A., Miyamoto, K., Minamino, N., Takeda, M., Sato, B., Matsuo, H., and Matsumoto, K. (1992) Proc. Natl. Acad. Sci. U.S.A. 89, 8928–8932. 3. Miyamoto, M., Naruo, K., Seko, C., Matsumoto, S., Kondo, T., and Kurokawa, T. (1993) Mol. Cell. Biol. 13, 239–249. 4. Folkman, J., and Klagsbrun, M. (1987) Science 235, 442–447. 5. Rifkin, D. B., and Moscatelli, D. (1989) J. Cell. Biol. 109, 1–6. 6. Luqmani, Y. A., Graham, M., and Coombes, R. C. (1992) Brit. J. Cancer 66, 273–280. 7. Anandappa, S. Y., Winstanley, J. H. R., Leinster, S., Green, B., Rudland, P. S., and Barraclough, R. (1994) Br. J. Cancer 69, 772–776. 8. Smith, J., Yelland, A., Baillie, R., and Coombes, R. C. (1994) Eur. J. Cancer 30, 496–503. 9. Souttou, B., Hamelin, R., and Crepin, M. (1994) Cell Growth and Differentiation 5, 615–623. 10. Penault-Llorca, F., Bertucci, F., Adelaide, J., Parc, P., Coulier, F., Jacquemier, J., Birnbaum, D., and DeLapeyriere, O. (1995) Int. J. Cancer 61, 170–176. 684
JOBNAME: BBRC 219#3 PAGE: 7 SESS: 13 OUTPUT: Tue Apr 2 09:18:08 1996 /xypage/worksmart/tsp000/68953f/2
Vol. 219, No. 3, 1996
BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS
11. Levay-Young, B. K., Imagawa, W., Wallace, D. R., and Nandi, S. (1989) Mol. Cell. Endocrinol. 62, 327–336. 12. Briozzo, P., Badet, J., Capony, F., Pieri, I., Montcourrier, P., Barriltault, D., and Rochefort, H. (1991) Exp. Cell Res. 194, 252–259. 13. Immagawa, W., Cunha, G. R., Young, P., and Nandi, S. (1994) Biochem. Biophys. Res. Comm. 204, 1165–1169. 14. El Yazidi, I., and Boilly-Marer, Y. (1995) Anticancer Res. 15, 783–790. 15. Myers, R. L., Payson, R. A., Chotani, M. A., Deaven, L. L., and Chiu, I. M. (1993) Oncogene 8, 341–349. 16. Myers, R. L., Ray, S. K., Eldridge, R., Chotani, M. A., and Chiu, I. M. (1995) J. Biol. Chem. 270, 8257–8266. 17. Philippe, J. M., Renaud, F., Desset, S., Laurent, M., Mallet, J., Courtois, Y., and Dumas Milne Edwards, J. B. (1992) Biochem. Biophys. Res. Comm. 188, 843–850. 18. Chiu, I. M., Wang, W. P., and Lehtoma, K. (1990) Oncogene 5, 755–762. 19. Renaud, F., Desset, S., Bugra, K., Halley, C., Philippe, J. M., Coutois, Y., and Laurent, M. (1992) Biochem. Biophys. Res. Comm. 184, 945–952. 20. Payson, R. A., Canatan, H., Chotani, M. A., Wang, W. P., Harris, S. E., Myers, R. L., and Chiu, I. M. (1993) Nucleic Acids Res. 21, 489–495. 21. Shaw, G., and Kamen, R. A. (1986) Cell 46, 659–667. 22. Rao, C. D., Pech, M., Robbins, K. C., and Aaronson, S. A. (1988) Mol. Cell. Biol. 8, 284–292. 23. Gottlieb, E. (1992) Proc. Natl. Acad. Sci. USA 89, 7164–7168. 24. Cook, P. W., Coffey, R. J., Magun, B. E., Pittelkow, M. R., and Shipley, G. D. (1990) Mol. Endocrinol. 4, 1377–1385. 25. Winkles, J. A., and Gay, C. G. (1991) Cell Growth and Differentiation 2, 531–540. 26. Chotani, M. A., Payson, R. A., Winkles, J. A., and Chiu, I. M. (1995) Nucleic Acids Res. 23, 434–441. 27. Renaud, F., Oliver, L., Desset, S., Tassin, J., Romquin, N., Courtois, Y., and Laurent, M (1994) J. Cell. Physiol. 158, 435–443. 28. Soule, H. D., and McGrath, C. M. (1986) In Vitro Cell. Dev. Biol. 22, 6–12. 29. Lasfargues, E. Y., and Ozzelo, L. (1958) Natl. Cancer Inst. 21, 1131–1147. 30. Soule, H. D., Vasquez, J., Long, A., Albert, S., and Brennan, M. J. (1973) J. Natl. Cancer Inst. 51, 1409–1416. 31. Cailleau, R., Young, R., Olive, M., and Reves, W. J. (1974) J. Natl. Cancer Inst. 53, 661–674. 32. Engel, L. W., and Young, N. A. (1978) Cancer Res. 38, 4327–4339. 33. Gaffney, E. V. (1982) Cell Tissue Res. 227, 563–568. 34. Chirwing, J. M., Przybyla, A. E., Mac Donald, R. J., and Rutter, W. (1979) Biochemistry 18, 5294–5299. 35. Hall, J. A., Harris, M. A., Intres, R., and Harris, S. E. (1993) J. Cell. Biochem. 51, 116–127.
685