Repression of transforming growth factor beta 1 protein by antisense oligonucleotide-induced increase of adrenal cell differentiated functions.

Transforming growth factor beta 1 (TGF beta 1) is a potent inhibitor of several differentiated functions in bovine adrenal fasciculata cells (BAC). In addition, these cells express and secrete this factor. To determine whether this peptide plays an autocrine role in BAC, cells were transfected with 10 microM unmodified sense (SON) or antisense (AON) oligonucleotide complementary to the translation initiation region of the TGF beta 1 mRNA in an attempt to inhibit TGF beta 1 protein synthesis. We investigated first, the cellular uptake, the stability, and the intracellular distribution of 32P-labeled TGF beta 1 AON and SON; and second, the effects of both oligonucleotides on BAC specific functions. We have demonstrated that in BAC, the TGF beta 1 AON uptake reached a plateau after 8 h of transfection (16% of the radioactivity added) and remained fairly constant for at least 24 h. In contrast, the uptake of TGF beta 1 SON reached a plateau after 2 h of transfection (8% of the radioactivity added), remained stable for only 3 h, and then declined. After 8 h of transfection, followed by 44 h of culture without oligonucleotides, the intracellular level of TGF beta 1 AON was still high with about 8% of the radioactivity added, whereas that of TGF beta 1 SON represented only 1.2%. Moreover, AON was present in the cytoplasmic and nuclear fractions, and it was hybridized in both compartments. However, TGF beta 1 SON was present mainly in the cytoplasmic fraction where it was not hybridized. Neither TGF beta 1 AON nor SON modified TGF beta 1 mRNA levels; however, TGF beta 1 AON, but not SON, caused the disappearance of TGF beta 1 immunoreactivity inside the cells. Finally, the steroidogenic responsiveness of BAC transfected with TGF beta 1 AON increased about 2-fold, and this was associated with a 2-fold increase of the mRNA levels of both cytochrome P450 17 alpha-hydroxylase and 3 beta-hydroxysteroid dehydrogenase. Neither TGF beta 1 SON nor a scrambled oligonucleotide containing the same number of G nucleotides as TGF beta 1 AON had any effect on these parameters. Thus, these studies demonstrate that TGF beta 1 has an autocrine inhibitory effect on BAC differentiated functions, an effect that can be overcome by TGF beta 1 AON.

Transforming growth factor ␤1 (TGF␤1) is a potent inhibitor of several differentiated functions in bovine adrenal fasciculata cells (BAC). In addition, these cells express and secrete this factor. To determine whether this peptide plays an autocrine role in BAC, cells were transfected with 10 M unmodified sense (SON) or antisense (AON) oligonucleotide complementary to the translation initiation region of the TGF␤1 mRNA in an attempt to inhibit TGF␤1 protein synthesis. We investigated first, the cellular uptake, the stability, and the intracellular distribution of 32 P-labeled TGF␤1 AON and SON; and second, the effects of both oligonucleotides on BAC specific functions. We have demonstrated that in BAC, the TGF␤1 AON uptake reached a plateau after 8 h of transfection (16% of the radioactivity added) and remained fairly constant for at least 24 h. In contrast, the uptake of TGF␤1 SON reached a plateau after 2 h of transfection (8% of the radioactivity added), remained stable for only 3 h, and then declined. After 8 h of transfection, followed by 44 h of culture without oligonucleotides, the intracellular level of TGF␤1 AON was still high with about 8% of the radioactivity added, whereas that of TGF␤1 SON represented only 1.2%. Moreover, AON was present in the cytoplasmic and nuclear fractions, and it was hybridized in both compartments. However, TGF␤1 SON was present mainly in the cytoplasmic fraction where it was not hybridized. Neither TGF␤1 AON nor SON modified TGF␤1 mRNA levels; however, TGF␤1 AON, but not SON, caused the disappearance of TGF␤1 immunoreactivity inside the cells. Finally, the steroidogenic responsiveness of BAC transfected with TGF␤1 AON increased about 2-fold, and this was associated with a 2-fold increase of the mRNA levels of both cytochrome P450 17␣-hydroxylase and 3␤-hydroxysteroid dehydrogenase. Neither TGF␤1 SON nor a scrambled oligonucleotide containing the same number of G nucleotides as TGF␤1 AON had any effect on these parameters. Thus, these studies demonstrate that TGF␤1 has an autocrine inhibitory effect on BAC differentiated functions, an effect that can be overcome by TGF␤1 AON.
The transforming growth factor ␤ (TGF␤) 1 family of peptides consists of related disulfide-linked homodimers that have multifunctional regulatory activities in many cell types and are expressed in many normal and malignant tissues (1)(2)(3). Three isoforms, termed TGF␤1, TGF␤2, and TGF␤3, have been identified in mammals (4). Although in many cells the three TGF␤s display comparable activities and potencies, marked differences have been noted in some cases (2,3). Cross-linking experiments have shown that the TGF␤ family of peptides binds to three different receptors, named type I, II, and III (2), which have been cloned (5)(6)(7)(8). Recent studies have determined the role of each type of TGF␤ receptor. Type III receptor, also known as betaglycan, has no direct role in TGF␤ signaling, but increases the binding of TGF␤, in particular TGF␤2, to type II receptor, enhances cell responsiveness to TGF␤, and diminishes the biological differences between TGF␤ isoforms (9). Types I and II receptors are transmembrane serine/threonine kinases. The role of these molecules in signaling has now been determined, and both types are required for TGF␤ signaling (10). TGF␤ binds directly to receptor II. Bound TGF␤ is then recognized by receptor I, which is recruited into the complex and becomes phosphorylated by receptor II. Phosphorylation allows receptor I to propagate the signal to downstream substrates. Recent results suggest that serine residues in the GS domain (region preceding the kinase domain) of receptor I, which are phosphorylated by receptor II, are important for signal transduction by receptor I (11).
In bovine fasciculata adrenal cells (BAC), the expression and the maintenance of specific differentiated functions are regulated not only by corticotropin (ACTH) and angiotensin-II (An-gII), the two main hormones that control steroidogenesis, but also by growth factors, which have been shown to have pleiotropic effects in addition to their mitogenic action. In bovine and ovine adrenocortical cells, it has been shown that TGF␤1 is a potent inhibitor of basal as well as ACTH-induced cortisol production (12,13). TGF␤1 exerts its effects at several levels: inhibition of low density lipoprotein receptors (12), inhibition of cytochrome P450 17␣-hydroxylase and 3␤-hydroxysteroid dehydrogenase (3␤ HSD) activities, protein and mRNA contents (14 -17), and down-regulation of ACTH receptors in ovine adrenocortical cells (18). TGF␤1 has also been proposed to regulate the steroidogenic functions in an autocrine loop (19) in BAC cells, which possess TGF␤ receptors that are regulated by ACTH (20). In addition, BAC cells synthesize (19) and secrete a latent form of TGF␤-like activity (15). In BAC, TGF␤1 secretion is regulated by specific peptide hormones; ACTH decreases TGF␤1 mRNA level, whereas AngII increases TGF␤1 mRNA and protein levels. 2 All these data suggest that TGF␤1 local production could play an autocrine role on BAC differentiated functions.
Synthetic oligonucleotides represent a new tool to investigate the role of many proteins in cell growth and differentiation. Ideally, an antisense oligonucleotide is targeted in a sequence-specific manner to nucleic acids (RNA or DNA) to inhibit the expression of a specific protein involved in cellular signal transduction, growth, proliferation, or differentiation (21). Antisense oligonucleotide inhibition of cellular protein production has been used to study the actions of several growth factors including basic fibroblast growth factor (22,23), insulinlike growth factor-I (24), insulin-like growth factor-II (25), platelet differentiating growth factor, and TGF␤1 (23).
In the present study, using a TGF␤1 antisense oligodeoxynucleotide complementary to a sequence that includes the translation start site of the human TGF␤1 mRNA, we have inhibited TGF␤1 synthesis in BAC and demonstrated an autocrine role for TGF␤1 on BAC differentiated functions.  (27), and human 3␤ HSD cDNA by Dr. F. Labrie and V. Luu The (Centre Hospitalier Universitaire Laval, Québec, Canada) (28). Polyclonal rabbit antibody (C-11-V) directed against a common C-terminal peptide of G␣ q /G␣ 11 proteins and polyclonal anti-TGF␤1 rabbit antibody were prepared in our laboratory as described previously (29,30). Goat antirabbit immunoglobulin G (IgG) conjugated to peroxidase was purchased from Nordic Immunology (Tilburg, The Netherlands).

Materials-[␥-
Isolation and Culture of Bovine Adrenocortical Cells-BAC were prepared by sequential treatment of adrenal cortical slices with trypsin (0.16%) as described previously (31). Then, cells were purified on a discontinuous Percoll density gradient (d ϭ 1.032, 1.048, and 1.082 g/ml) to eliminate cellular fragments and red blood cells. The purified fasciculata cells recovered on the Percoll gradient with a density of 1.048 g/ml were collected, washed and cultured in a chemically defined medium, Ham's F-12/Dulbecco's modified Eagle's medium (1:1), containing 10 g/ml transferrin, 10 g/ml insulin, 10 Ϫ4 M vitamin C, and antibiotics without serum.
Cell Transfection and Viability Test-On day 2 of culture, cells were transfected with labeled and/or unlabeled TGF␤1 AON or SON. To introduce the oligonucleotides into BAC, a cationic liposome-mediated transfection method was used. Oligonucleotides dissolved in one volume of antibiotic-free medium were mixed with Lipofectamine™ reagent dissolved in the same volume of antibiotic-free medium and incubated for 45 min at room temperature. Thereafter, the oligonucleotide-liposome complexes were diluted with eight volumes of antibiotic-free medium and then added to cells that had been washed twice with antibiotic-free medium. In the experiments reported, the concentration of oligonucleotides and Lipofectamine™ in the transfection medium was 10 M (50 g/ml) and 1.25%, respectively. For the viability test (trypan blue exclusion assay), the number of living cells was assessed at the end of the experimental period (8 h of transfection followed by 44 h of culture).
Oligonucleotide Cellular Uptake and Degradation-Oligonucleotides were 5Ј-labeled with [␥-32 P]ATP by use of bacteriophage T4 polynucleotide kinase and further purified by dialysis (specific activity 8 ϫ 10 8 dpm/g). The transfection medium containing 1 ϫ 10 6 dpm/ml 32 P oligonucleotides and 10 M unlabeled oligonucleotides was added to the cells. At indicated times, the culture medium was removed and saved, cells were washed three times with medium, and the cell washes were also removed and saved. Cells were lysed in Tris-buffered saline (10 mM Tris-HCl, pH 7.4, containing 150 mM NaCl and 1% sodium dodecyl sulfate) and extracted with phenol-chloroform-amyl alcohol (25:25:1, v/v/v). After centrifugation (12,000 ϫ g, 15 min, 4°C), the aqueous phase was removed and saved. Then the phenol phase was extracted with water and centrifuged, and the aqueous phase was removed and pooled with the first one. Aliquots of the three fractions, i.e. combined aqueous phases corresponding to the intracellular radioactivity, cell washes, and culture medium, were counted. Oligonucleotide uptake was calculated as the percentage of the intracellular radioactivity over the counts recovered in the three fractions. To determine oligonucleotide degradation, aliquots containing equal amounts of radioactivity of the combined aqueous phases and of the culture medium fraction were analyzed by electrophoresis (10% polyacrylamide, 7 M urea gel) and autoradiographed.
Oligonucleotide Distribution and Hybridization-After transfection with 5Ј-32 P-labeled oligonucleotide, a subcellular fractionation of the cells was carried out. Briefly, after washes, cells were removed from the culture plate with trypsin (100 ϫ g, 10 min, 4°C) and washed twice with medium. The cells were lysed for 10 min at 4°C in buffer A (10 mM Tris HCl, pH 7.4, 10 mM NaCl, 5 mM MgCl 2 , 0.5% Nonidet P-40, 1 mM dithiothreitol). After centrifugation at 800 ϫ g for 10 min, the upper cytoplasmic fraction (combined cytosol and cell membranes) was removed and saved. The pellet (nuclei) was dissolved in a small quantity of buffer A and then laid over 1 ml of 25 mM Tris-HCl, pH 7.4, 2.5 mM MgCl 2 , 0.25 M sucrose and centrifuged (800 ϫ g for 10 min). The purified nuclei pellet was dissolved in buffer A. Aliquots of the cytoplasmic and nuclear fractions were counted. In order to determine whether the oligonucleotide formed duplexes with cellular RNA, an S1 nuclease protection assay was performed on aliquots of cell lysates, cytoplasmic, and nuclear fractions. As a control the medium was also treated with the enzyme. The nucleic acids of each fraction were precipitated at Ϫ20°C for 20 min and Ϫ70°C for 10 min with cold ethanol (2.2 volumes), 3 M sodium acetate (0.1 volume), pH 5.2, and 10 g of tRNA.
After centrifugation (12,000 ϫ g for 30 min at 4°C), the pellet was washed with cold ethanol (70%), lyophilized for a short time, and then resuspended in 70 l of hybridization buffer (40 mM PIPES, 400 mM NaCl, 1 mM EDTA). Then 30 l of each aliquot was incubated without (control) or with 20 units of S1 nuclease 30 min at 37°C in 200 l of digestion buffer (280 mM NaCl, 30 mM sodium acetate, 4.5 mM ZnSO 4 , 20 g/ml salmon sperm DNA). The nuclease action was stopped by putting the samples on ice and by addition of 60 l of termination buffer (1.5 M sodium acetate, 125 mM EDTA, 75 mM MgCl 2 ). Samples were precipitated at Ϫ20°C for 20 min and Ϫ70°C for 10 min with 20 g of tRNA and 0.75 ml of ethanol, centrifuged, and washed as described above. Samples were analyzed by electrophoresis using 10% polyacrylamide, 7 M urea gel. Protected fragments were visualized by autoradiography.
RNA Preparation and Northern Blot Analysis-Total RNA was isolated from cells by the method of Chomczynski and Sacchi (32). Samples (10 -15 g of RNA) were separated by electrophoresis through a 1% agarose gel containing 10% formaldehyde. RNA was then transferred to Hybond-N membrane. Prehybridization and hybridization solutions used were described previously (30). Labeled human TGF␤1 cDNA, bovine P450 17␣-hydroxylase cDNA, and human 3␤ HSD were used as probes (1-2 ϫ 10 6 dpm/ml). Labeling of these probes in the presence of [␣-32 P]dCTP was performed with a Megaprime DNA labeling system (Amersham). The blots were washed with more or less stringency depending on the probes used and then exposed to photographic film. The relative intensity of hybridization signals was quantified by using a scanning densitometer (Preference Sebia, Paris, France). Equal loading of RNA samples was confirmed by scanning the 28 S RNA negatives.
Immunocytochemistry-BAC cells were plated at a density of 6.0 10 4 cells/chamber in eight-chamber tissue culture slides (Plastic Labtek) and transfected with antisense or sense oligonucleotide as described above. After the transfection medium was removed, cells were washed two times with fresh medium, cultured during 44 h and subjected to immunocytochemical analysis. For comparative studies, control cells were always run in the same immunocytochemical assay to reduce discrepancies related to interassay variability in staining intensity. TGF␤1 expression was examined by an indirect immunocytochemical method as described previously (30). Briefly, cells were fixed 30 min at room temperature in 2% acrolein in 10 mM phosphate buffer (pH 7.4), and washed overnight in 100 mM PBS (pH 7.6) at 4°C. Cells were then permeabilized with 0.1% Triton X-100 for 30 min, rinsed, and exposed for 1 h to a 1/40 dilution of nonimmune rabbit serum. The polyclonal anti-TGF␤1 rabbit antibody was used as primary antibody at a dilution of 1/1000 overnight in a humidity chamber at 4°C. The second antibody to rabbit IgGs conjugated to peroxidase was used at a dilution of 1/200 for 1 h at room temperature. To localize the antigen-antibody complexes, cells were incubated for 2 min with 0.05% 3,3Ј diaminobenzidine tetrahydrochloride, 0.01% H 2 O 2 , and 2.5% nickel ammonium sulfate. Next, the cultured cell preparations were mounted in PBS-glycerol (1:1). The specificity of the TGF␤1 antibody has been tested previously (30).
Cortisol Production-It was measured in the medium by a specific radioimmunoassay (33).
Metabolic Labeling-For metabolic labeling, before the end of the culture, the cells were preincubated for 1 h in methionine-free medium, after which the medium was replaced by fresh methionine-free medium containing [ 35 S]methionine (50 Ci/ml) during the last 4 h.
Immunoprecipitation-After metabolic labeling, cells were washed three times with phosphate buffered saline. Cells were then lysed by addition of 500 l of ice-cold immunoprecipitation buffer (25 mM Tris-HCl, pH 7.5, 150 mM NaCl, 0.1% SDS, 1% deoxycholate, 1% Triton X-100, 1 mM phenylmethylsulfonyl fluoride, and 10 g/ml leupeptin). After centrifugation for 10 min at 10,000 ϫ g, the supernatants from the cells extracts were incubated at 4°C for 2 h with nonimmune rabbit serum at a final concentration of 5% and then with 50 l of a 50% (v/v) suspension of protein A-Sepharose CL-4B. After 1 h of incubation, the beads were sedimented by centrifugation. The supernatants were collected and submitted to immunoprecipitation with anti-␣ q /␣ 11 IgG (1/ 50, v/v) at 4°C for 2 h, followed by incubation with protein A-Sepharose CL-4B. After centrifugation, the beads were washed four times with ice-cold immunoprecipitation buffer and once with 0.1% SDS. The radiolabeled proteins in the final pellet were analyzed by electrophoresis using 7.5% polyacrylamide gels. Gels were fixed, soaked in Amplify, dried, and the labeled proteins were revealed by fluorography.
Western Blot Analysis-After appropriate treatments, cells were lysed and submitted to Western blot analysis as described previously (34) except 10% polyacrylamide gels were without urea.
Statistical Analysis-Statistical analysis were performed with Student's t test for comparison of two groups. Differences were considered significant when p Ͻ 0.05.
Oligonucleotide Uptake and Degradation-To investigate the kinetics of oligonucleotide uptake and degradation, cells were transfected with 32 P-labeled TGF␤1 AON or SON (final concentration 10 M) for different times as described under "Experimental Procedures." Fig. 1 shows the time-dependent cellular uptake of TGF␤1 AON (solid line) and TGF␤1 SON (dotted line). TGF␤1 AON cellular uptake started within the first minutes of transfection, increased progressively to reach a plateau at 8 h, and remained stable for the next 24 h. At this time, 16% of the radioactivity were inside the cells. Taking into account the cellular volume and the specific activity, TGF␤1 AON intracellular concentration (500 M) was about 50-fold higher than that in the medium. TGF␤1 SON cellular uptake was also rapid. Uptake reached a plateau at 2 h, remained stable for only 3 h, and then declined. At the plateau, 8% of the radioactivity was inside the cells.
To examine the extent of oligonucleotide degradation within cells and in the medium during the cellular uptake, TGF␤1 AON was analyzed by gel electrophoresis. The results of Fig. 1 show that, inside the cells, TGF␤1 AON appeared to be intact for up to 24 h. However, in the culture medium, a progressive degradation of the TGF␤1 AON was observed. The oligonucleo-tide was first transformed into another compound with higher mobility. This in turn was converted into another compound with faster mobility after 3 h. A similar pattern of degradation, but much more rapid, was observed with TGF␤1 SON (data not shown).
Intracellular Stability of the TGF␤1 AON and SON-To determine the intracellular stability of the oligonucleotides, cells were transfected with 32 P-labeled TGF␤1 AON or SON for 8 h (time at which TGF␤1 AON uptake was maximum), the medium was then removed and replaced with fresh medium without oligonucleotides, and the culture continued for another 44 h (Fig. 2). As shown, after 8 h of transfection, the intracellular concentration of TGF␤1 AON or SON represented 16% and 6%, respectively, of the radioactivity added. However, after 44 h of culture, the intracellular concentration of TGF␤1 AON was still 8% of the radioactivity added, whereas that of TGF␤1 SON represented only 1.2%.
Intracellular Distribution and Hybridization of TGF␤1 AON and SON-In order to determine the intracellular distribution of both TGF␤1 AON and SON, subcellular fractionation of the cells was performed after 8 h of transfection and 44 h after the oligonucleotides' removal (Fig. 3). After 8 h of transfection 60% and 40% of the intracellular TGF␤1 AON were in the cytoplasmic and nuclear fractions, respectively. This distribution was the opposite after 44 h of culture. In contrast, at both times more than 85% of the intracellular TGF␤1 SON were in the cytoplasmic fraction. Indeed, the TGF␤1 SON present in the nuclear fraction after 44 h of culture without oligonucleotides represented less than 0.2% of the radioactivity added.
In an attempt to determine the intracellular formation of an oligonucleotide-RNA duplex, the medium and the cell lysates were submitted to partial S1 nuclease digestion. This enzyme digested all the non hybridized nucleotides. As expected, the TGF␤1 AON and SON derived from the culture medium were almost completely degraded (Fig. 4). Similarly, the TGF␤1 SON extracted from cells was also degraded. In contrast, after 8 h of transfection, as well as after 44 h of culture without oligonucleotides, about 40% of the TGF␤1 AON extracted from cells were protected from degradation. These results indicated that TGF␤1 AON, but not SON, was hybridized inside the cells.
Next, we investigated the intracellular compartment in which the hybridized TGF␤1 AON was located. Subcellular fractionations were performed after 8 h of transfection and after an additional 44 h of culture without oligonucleotides. The extracts from cytoplasmic and nuclear fractions were submitted to partial S1 nuclease digestion (Fig. 5). After 8 h of transfection, the hybridization of the TGF␤1 AON was more marked in the cytoplasmic (about 33%) than in the nuclear fraction (about 16%), whereas after 44 h of culture, although the percentage of hybridization increased in both compartments (44% and 75% for the cytoplasmic and the nuclear frac-tions, respectively), the ratio was reversed.
Effects of TGF␤1 AON and SON on BAC TGF␤1 mRNA and Protein Content-To assess whether the oligonucleotides were able to modify the transcription and/or the translation of TGF␤1, we investigated their effects on TGF␤1 mRNA by Northern blot and on cellular TGF␤1 protein content by immunocytochemistry. The results of Fig. 6 showed that neither TGF␤1 AON nor SON modified the level of the 2.5-kilobase transcript of TGF␤1 mRNA. In contrast, the results of immunocytochemistry (Fig. 7) showed that all the control cells (A) or cells pretreated with TGF␤1 SON (C) were immunoreactive. However, the TGF␤1 immunoreactivity completely disappeared in TGF␤1 AON treated cells (D). In B, where the TGF␤1 antibody was saturated with the peptide used to produce this antibody, there was no TGF␤1 signal, thus showing the specificity of this antibody. Since it has been reported that cyclosporine increased the expression of TGF␤1 (35), we treated BAC for 44 h with this factor (1 g/ml) in the absence or presence of TGF␤1 AON and we examined the TGF␤1 content by immunocytochemistry (Fig. 8). Although this method is only Aliquots of medium and cell lysates were precipitated by ethanol to recover the nucleic acids. For each condition, aliquots containing equal amounts of radioactivity were incubated for 30 min at 37°C in the absence or the presence of 20 units of S1 nuclease. All samples were analyzed by urea-PAGE and autoradiography. The intensity of the signal after S1 nuclease digestion is expressed as the percentage of the signal without S1 nuclease digestion (100%). Top, diagram; bottom, autoradiography of one representative experiment.

FIG. 5. Intracellular distribution of hybridized TGF␤1 AON.
Cells were incubated with 32 P-labeled AON for 8 h or 8 h followed by 44 h of culture. Subcellular fractionation was performed and aliquots of culture medium, cytoplasmic and nuclear fractions were subjected or not to partial S1 nuclease digestion. Aliquots were then analyzed by urea-PAGE and autoradiography. Results are expressed as described in the legend of Fig. 4. Top, diagram; bottom, autoradiography of one representative experiment. semi-quantitative, the results of Fig. 8 suggest that cyclosporine increased the cellular TGF␤1 content, and this effect was blunted by TGF␤1 AON.
Effects of TGF␤1 AON and SON on BAC Functions-As described in the Introduction, exogenous TGF␤1 decreases the steroidogenic responsiveness of BAC; thus, we investigated the effects of both oligonucleotides, TGF␤1 AON and SON, on the steroidogenic responsiveness to AngII of control and cyclosporine treated cells (Fig. 9). In the absence of cyclosporine, TGF␤1 AON increased the cortisol response to AngII about 2-fold compared to either control cells or TGF␤1 SON-treated cells. Cyclosporine alone decreased the steroidogenic responsiveness of BAC by about 50%, compared to cells not treated with cyclosporine. However, the cortisol production of cells treated with cyclosporine and TGF␤1 AON was 2.3-fold higher than that of cells treated with cyclosporine alone. Again, TGF␤1 SON had no effect.
One of the mechanism by which exogenous TGF␤1 decreases the steroidogenic capacity of BAC is by decreasing the mRNA levels of P450 17␣-hydroxylase and 3␤ HSD (14,16). The results of Fig. 10 clearly show that TGF␤1 AON, but not SON, increased P450 17␣-hydroxylase and 3␤ HSD mRNA levels (2and 1.7-fold, respectively), which encode two key enzymes in the steroidogenic pathway.
Control Experiments to Demonstrate the Specificity of TGF␤1 AON Effects-To prove that all the TGF␤1 AON effects on BAC functions were specific, some additional controls were performed. First, cells were transfected with a scrambled oligonucleotide (SCR) containing the same number of G nucleotides (10 of 15) as TGF␤1 AON but in a scrambled order. The results (Fig. 11) showed that, in contrast to TGF␤1 AON, neither SON nor SCR modified the cortisol secretion and the mRNA levels of P450 17␣-hydroxylase. Second, none of the transfected oligonucleotides did change the sensitivity of the cells to the inhibitory effects of TGF␤1, since exogenous TGF␤1 caused similar inhibition of both cortisol secretion and P450 17␣-hydroxylase mRNA levels, in control and in transfected cells (Fig. 11). Finally, to prove that transfection did not produce a general inhibition of protein synthesis, we investigated the effects of transfection with the three oligonucleotides on the rate of synthesis and on the steady-state levels of G␣ q /G␣ 11 proteins, which are not affected by exogenous TGF␤1 (34). The results showed that neither the rate of synthesis (Fig. 12A) nor the steady-state levels (Fig. 12B) of G␣ q /G␣ 11 proteins were affected in transfected cells regardless of the nucleotide used. DISCUSSION Antisense oligonucleotides have been used as specific inhibitors of target gene expression. The specificity of an antisense oligonucleotide is due to highly specific hybridization to its complementary target sequence on the mRNA by Watson-Crick base pairing. This is obtained by using an oligonucleotide of about 15 bases directed against a complementary sequence of target mRNA (21,36,37). One key parameter in the oligonucleotide antisense approach is its intracellular concentration, which is the result of two opposite processes: the rate of penetration of the antisense molecule across cell membrane, and its  AON (B and D). The medium was replaced by fresh medium without (A and B) or with (C and D) 1 g/ml cyclosporine and the culture continued for 44 h. Immunocytochemical staining was performed as described in Fig. 7. rate of degradation in the cells. The uptake is a saturable process thought to be mediated by both receptor endocytosis and fluid phase endocytosis (38 -40). An increased uptake has been obtained by encapsulation of the oligonucleotides in cationic liposomes (41). Using a cationic liposome-mediated transfection method in cell culture, we demonstrated a rapid, high, and similar uptake of both TGF␤1 AON and SON during the first 2 h of transfection. Thereafter, the kinetics of TGF␤1 SON and AON were different. Indeed, whereas the intracellular concentration of SON, after a short lag period, declined, the concentration of AON continued to increase reaching a plateau at 8 h (16% of the radioactivity added) and remained stable for at least 24 h. This uptake is several times higher than that observed in others studies in which no cationic liposomes were used (42)(43)(44)(45). These kinetic studies allowed us to determine the optimal time of transfection (8 h) and to investigate the stability and the distribution of both TGF␤1 AON and SON. Although, as indicated above, the intracellular level of both TGF␤1 AON and SON were different after the first hours of transfection, both appeared intact in the cells. However, degradation products of both TGF␤1 AON and SON appeared in the culture medium. This process was more rapid and marked for TGF␤1 SON than for AON. Whether the degradation occurred inside or outside the cells was not determined in the present study. However, on the one hand, our culture did not contain serum thought to have DNase activity (36). On the other hand, after transfection of the cells for 8 h followed by extensive washings, oligonucleotide degradation products appeared in the fresh medium during the next 44 h of culture (data not shown). Thus, it is likely that the degradation of both oligonucleotides takes place inside the cells.
In addition, our results revealed marked difference between TGF␤1 AON and SON concerning their stability and cellular distribution. First, after 8 h of transfection followed by 44 h of culture, intracellular TGF␤1 AON concentration was still high (8%), whereas that of SON was only 1.2%. Second, at any time most of the intracellular TGF␤1 SON was located in the cytoplasm, and was not hybridized. However, TGF␤1 AON was predominant in the cytoplasm after 8 h of transfection; it became prevalent in the nucleus at the end of the experimental period. In addition, in both compartments, TGF␤1 AON was hybridized. This hybridization was particularly intense in the nucleus, and it was higher after 8 h of transfection followed by 44 h of culture (without oligonucleotides) than immediately after transfection. These results agree with other data showing that c-Myb (40) and prorenin (41) antisense oligonucleotides were preferentially accumulated in the nucleus. However, these results differ from those of Temsamani et al. (44) showing preferential cytoplasmic localization of several antisense oligonucleotides. Although the exact mechanism of oligonucleotide transfer from cytoplasm to nucleus is not completely understood, a passive diffusion through the nuclear pores has been postulated (40).
The present studies also show that 44 h after transfection about 44% and 75% of the TGF␤1 AON present in the cytoplasm and nucleus, respectively, were resistant to S1 nuclease digestion. Since the target sequence is present in both primary transcript and mRNA, TGF␤1 AON could hybridize to both and interfere with pre-mRNA maturation and/or nucleocytoplasmic transport (36,37,46) but not with transcription, since the level of TGF␤1 mRNA was not modified by TGF␤1 AON. In contrast, TGF␤1 AON causes complete inhibition of TGF␤1 protein production. This inhibition could be the result of either degradation of RNA by RNase H, which selectively cleaves the RNA at DNA-RNA heteroduplexes (47,48), or inhibition of the translation by AON hybridization to the translation initiation site of the TGF␤1 mRNA (49,50). The first hypothesis is unlikely because no decrease of TGF␤1 mRNA was observed. Although recent data show that c-Myb AON was not associated with ribosomes or endoplasmic reticulum (40), our results strongly suggest that the main mechanism by which TGF␤1 AON blocked the synthesis of TGF␤1 protein is by translation arrest.
Although the potential autocrine role of TGF␤1 has been suggested in several cell types (1-3), only in two models, rat vascular smooth muscle cells (23) and human colon carcinoma cell line (51), has this been proven by using the antisense approach. Our results show that the biological consequences of TGF␤1 protein synthesis inhibition in control as well as in cyclosporine treated cells were a significant increase of cortisol production in response to AngII and ACTH (data not shown). Another demonstration of the autocrine role of TGF␤1 on BAC was obtained by showing that TGF␤1 AON, but neither SON nor SCR, increased about 2-and 1.7-fold the mRNA levels of P450 17␣-hydroxylase and 3␤ HSD, respectively, an effect that was opposite to that induced by exogenous TGF␤1 in these cells (Refs. 14 -17 and the present data). Moreover, the effects of TGF␤1 AON on steroidogenic responses of viable BAC were specific. First, they were not mimicked by SON or SCR; second, they could be reversed by addition of exogenous TGF␤1; and third, they did not modify the normal production of unrelated proteins.
Taken together our data demonstrate, for the first time, that constitutive expression of TGF␤1 by BAC has an autocrine inhibitory effect on the differentiated functions of these cells. Moreover, since TGF␤1 is expressed by many cell types, it is likely that this factor might also play an autocrine role in other models. Finally, these studies illustrate and confirm that antisense technology should find widespread application for investigating the exact role of many regulatory proteins on cell growth and differentiation.