All-trans-retinoic Acid Increases Transforming Growth Factor-β2 and Insulin-like Growth Factor Binding Protein-3 Expression through a Retinoic Acid Receptor-α-dependent Signaling Pathway
1997; Elsevier BV; Volume: 272; Issue: 21 Linguagem: Inglês
10.1074/jbc.272.21.13711
ISSN1083-351X
AutoresGil-Ro Han, David F. Dohi, Ho‐Young Lee, Roopmathy Rajah, Garrett L. Walsh, Waun Ki Hong, Pinchas Cohen, Jonathan M. Kurie,
Tópico(s)Nuclear Receptors and Signaling
ResumoRetinoids, including retinol and retinoic acid derivatives, maintain the normal growth and differentiation of human bronchial epithelial cells. The signaling pathways through which retinoids mediate these effects have not been defined. Insulin-like growth factor binding protein-3 (IGFBP-3) and the transforming growth factor-β (TGF-β) gene family (β1–3) were examined as potential components of the retinoid signaling pathway in normal human bronchial epithelial cells. All-trans-retinoic acid (t-RA) increased the levels of TGF-β2 and IGFBP-3 mRNA and of secreted TGF-β and IGFBP-3 proteins. An antagonist of retinoic acid receptor-α, LG100629, abrogated the increase in TGF-β2 and IGFBP-3 mRNA levels induced by t-RA. t-RA increased IGFBP-3 mRNA levels transiently from 1 to 6 h, and subsequently a sustained increase began at 72 h, which coincided with the appearance of active TGF-β in the media. Treatment with TGF-β2 increased IGFBP-3 mRNA levels, but treatment with latency-associated peptide, which inactivates secreted TGF-β, did not abrogate the effect of t-RA on IGFBP-3 expression. These findings provide evidence that t-RA increased TGF-β2 and IGFBP-3 expression through an retinoic acid receptor-α-dependent pathway, and the increase in IGFBP-3 expression by t-RA did not require activation of the TGF-β pathway by autocrine or paracrine mechanisms. Retinoids, including retinol and retinoic acid derivatives, maintain the normal growth and differentiation of human bronchial epithelial cells. The signaling pathways through which retinoids mediate these effects have not been defined. Insulin-like growth factor binding protein-3 (IGFBP-3) and the transforming growth factor-β (TGF-β) gene family (β1–3) were examined as potential components of the retinoid signaling pathway in normal human bronchial epithelial cells. All-trans-retinoic acid (t-RA) increased the levels of TGF-β2 and IGFBP-3 mRNA and of secreted TGF-β and IGFBP-3 proteins. An antagonist of retinoic acid receptor-α, LG100629, abrogated the increase in TGF-β2 and IGFBP-3 mRNA levels induced by t-RA. t-RA increased IGFBP-3 mRNA levels transiently from 1 to 6 h, and subsequently a sustained increase began at 72 h, which coincided with the appearance of active TGF-β in the media. Treatment with TGF-β2 increased IGFBP-3 mRNA levels, but treatment with latency-associated peptide, which inactivates secreted TGF-β, did not abrogate the effect of t-RA on IGFBP-3 expression. These findings provide evidence that t-RA increased TGF-β2 and IGFBP-3 expression through an retinoic acid receptor-α-dependent pathway, and the increase in IGFBP-3 expression by t-RA did not require activation of the TGF-β pathway by autocrine or paracrine mechanisms. Retinoids control normal tracheobronchial epithelial growth and differentiation. Rodents that are deprived of vitamin A develop squamous metaplasia in the tracheobronchial epithelium, and normal epithelial differentiation is restored by vitamin A supplementation (1Wolbach S.B. Howe P.T. J. Exp. Med. 1925; 42: 753-778Google Scholar,2Chopra D.P. J. Natl. Cancer Inst. 1982; 69: 895-901Google Scholar). In tissue culture, human bronchial epithelial (HBE) 1The abbreviations used are: HBE, human bronchial epithelial; t-RA, all-trans-retinoic acid; TGF-β, transforming growth factor-β; MLEC, mink lung epithelial cells; IGFBP-3, insulin-like growth factor binding protein-3; BPE, bovine pituitary extract; LAP, latency-associated peptide; WLB, Western ligand blot; RAR, retinoic acid receptor; RXR, retinoid X receptors. cells undergo squamous differentiation with a variety of agents, and all-trans-retinoic acid (t-RA) inhibits this process (3Jetten A.M. Rearick J.I. Smits H.L. Biochem. Soc. Trans. 1986; 14: 930-933Google Scholar, 4Jetten A.M. Shirley J.E. Stoner G. Exp. Cell Res. 1986; 167: 539-549Google Scholar, 5Masui T. Wakefield L.M. Lechner J.F. LaVeck M.A. Sporn M.B. Harris C.C. Proc. Natl. Acad. Sci. U. S. A. 1986; 83: 2438-2442Google Scholar, 6Willey J.C. Moser Jr., C.E. Lechner J.F. Harris C.C. Cancer Res. 1984; 44: 5124-5126Google Scholar, 7Saunders N.A. Jetten A.M. J. Biol. Chem. 1994; 269: 2016-2022Google Scholar). HBE cells treated with t-RA develop mucociliary features in collagen gels (3Jetten A.M. Rearick J.I. Smits H.L. Biochem. Soc. Trans. 1986; 14: 930-933Google Scholar, 8Jetten A.M. Brody A.R. Deas M.A. Hook G.E.R. Rearick J.I. Thacher S.M. Lab. Invest. 1987; 56: 654-664Google Scholar). Grown in monolayer cultures, retinol-treated HBE cells undergo growth arrest with no evidence of morphologic differentiation (9Miller L.A. Cheng L.Z. Wu R. Cancer Res. 1993; 53: 2527-2533Google Scholar). Retinoids are ligands for the retinoic acid receptors (RAR-α, -β, and -γ) and retinoid X receptors (RXR-α, -β, and -γ), which form RAR-RXR heterodimers and RXR homodimers and are transcriptionally activated by ligand binding (reviewed in Ref. 10Mangelsdorf D.J. Evans R.M. Cell. 1995; 83: 841-850Google Scholar). In bronchial epithelial cells, RAR-α is expressed at high levels and has been shown to activate growth inhibitory pathways (11Kim Y.-H. Dohi D.F. Han G.R. Zou C.-P. Oridate N. Walsh G.L. Nesbitt J.C. Xu X.-C. Hong W.K. Lotan R. Kurie J.M. Cancer Res. 1995; 55: 5603-5610Google Scholar, 12Zhang L.-X. Mills K.J. Dawson M.I. Collins S.J. Jetten A.M. J. Biol. Chem. 1994; 270: 6022-6029Google Scholar). The signaling pathways activated by RAR-α that mediate growth inhibition in normal HBE cells have not been defined. Retinoids increase the expression of transforming growth factor-β (TGF-β) family members (13Glick A.B. Flanders K.C. Danielpour D. Yuspa S.H. Sporn M.B. Cell. Regul. 1989; 1: 87-97Google Scholar, 14Niles R.M. Thompson N.L. Fenton F. In Vitro Cell. Dev. Biol. Anim. 1994; 30: 256-262Google Scholar, 15Glick A.B. McCune B.K. Abdulkarem N. Flanders K.C. Lumadue J.A. Smith J.M. Sporn M.B. Development. 1991; 111: 1081-1086Google Scholar, 16Nunes I. Kojima S. Rifkin D.B. Cancer Res. 1996; 56: 495-499Google Scholar, 17Danielpour D. J. Cell. Physiol. 1996; 166: 231-239Google Scholar, 18Turley J.M. Funakoshi S. Ruscetti F.W. Kasper J. Murphy W.J. Longo D.L. Birchenall-Roberts M.C. Cell Growth & Differ. 1995; 6: 655-663Google Scholar, 19Nugent P. Potchinsky M. Lafferty C. Greene R.M. Exp. Cell Res. 1995; 220: 495-500Google Scholar, 20Cohen P.S. Letterio J.J. Gaetano C. Chan J. Matsumoto K. Sporn M.B. Thiele C.J. Cancer Res. 1995; 55: 2386-2830Google Scholar, 21Kojima S. Rifkin D.B. J. Cell. Physiol. 1993; 155: 323-332Google Scholar, 22Batova A. Danielpour D. Pirisi L. Creek K.E. Cell Growth & Differ. 1992; 3: 763-772Google Scholar). TGF-β is secreted as a latent complex and converted to an active form, and it signals through a heteromeric complex of the type I and type II receptors (23Wrana J.L. Attisano L. Wieser R. Ventura F. Massagué J. Nature. 1994; 370: 341-347Google Scholar). Activation of the TGF-β pathway has been implicated in the effects of retinoids on cellular growth and differentiation (13Glick A.B. Flanders K.C. Danielpour D. Yuspa S.H. Sporn M.B. Cell. Regul. 1989; 1: 87-97Google Scholar, 14Niles R.M. Thompson N.L. Fenton F. In Vitro Cell. Dev. Biol. Anim. 1994; 30: 256-262Google Scholar, 15Glick A.B. McCune B.K. Abdulkarem N. Flanders K.C. Lumadue J.A. Smith J.M. Sporn M.B. Development. 1991; 111: 1081-1086Google Scholar, 16Nunes I. Kojima S. Rifkin D.B. Cancer Res. 1996; 56: 495-499Google Scholar, 17Danielpour D. J. Cell. Physiol. 1996; 166: 231-239Google Scholar, 18Turley J.M. Funakoshi S. Ruscetti F.W. Kasper J. Murphy W.J. Longo D.L. Birchenall-Roberts M.C. Cell Growth & Differ. 1995; 6: 655-663Google Scholar, 19Nugent P. Potchinsky M. Lafferty C. Greene R.M. Exp. Cell Res. 1995; 220: 495-500Google Scholar, 20Cohen P.S. Letterio J.J. Gaetano C. Chan J. Matsumoto K. Sporn M.B. Thiele C.J. Cancer Res. 1995; 55: 2386-2830Google Scholar, 21Kojima S. Rifkin D.B. J. Cell. Physiol. 1993; 155: 323-332Google Scholar, 22Batova A. Danielpour D. Pirisi L. Creek K.E. Cell Growth & Differ. 1992; 3: 763-772Google Scholar). In addition, the secretion of insulin-like growth factor binding protein-3 (IGFBP-3) is enhanced by retinoid treatment (24Gucev Z.S. Oh Y. Kelly K.M. Rosenfeld R.G. Cancer Res. 1996; 56: 1545-1550Google Scholar, 25Zhou Y. Mohan S. Linkhart T.A. Baylink D.J. Strong D.D. Endocrinology. 1996; 137: 975-983Google Scholar, 26Sheikh M.S. Shao Z.-M. Chen J.-C. Clemmons D.R. Roberts Jr., C.T. Leroith D. Fontana J.A. Biochem. Biophys. Res. Commun. 1992; 188: 1122-1130Google Scholar, 27Sheikh M.S. Shao A.-M. Hussain A. Clemmons D.R. Chen J.-C. Roberts Jr., C.T. LeRoith D. Fontana J.A. J. Cell. Physiol. 1993; 155: 556-567Google Scholar, 28Leyen S.A.-V. Hembree J.R. Eckert R.L. J. Cell. Physiol. 1994; 160: 265-274Google Scholar, 29Adamo M.L. Shao Z.-M Lanau F. Chen J.C. Clemmons D.R. Roberts Jr., C.T. LeRoith D. Fontana J.A. Endocrinology. 1992; 131: 1858-1866Google Scholar, 30Martin J.L. Coverley J.A. Pattison S.T. Baxter R.C. Endocrinology. 1995; 136: 1219-1226Google Scholar, 31Katz J. Weiss H. Goldman B. Kanety H. Stannard B. LeRoith D. Shemer J. J. Cell. Physiol. 1995; 165: 223-227Google Scholar, 32Hembree J.R. Agarwal C. Beard R.L. Chandraratna R.A.S. Eckert R.L. Cancer Res. 1996; 56: 1794-1799Google Scholar, 33Sheikh M.S. Shao Z.-M. Hussain A. Clemmons D.R. Chen J.-C. Roberts Jr., C.T. LeRoith D. Fontana J.A. J. Cell. Physiol. 1993; 155: 556-567Google Scholar, 35Fontana J.A. Burrows-Mezu A. Clemmons D.R. LeRoith D. Endocrinology. 1991; 128: 1115-1122Google Scholar). IGFBP-3 is one of a family of seven IGFBPs (36Cohen P. Fielder P.J. Hasegawa Y. Frisch H. Giudice L.C. Rosenfeld R.G. Acta Endocrinol. 1991; 124: 74-85Google Scholar, 37Holly J.M.P. Martin J.L. Growth Regul. 1994; 4: 20-30Google Scholar, 38Drop S.L.S. Schuller A.G.P. Lindenbergh-Kortelve D.J. Groffen C. Brinkman A. Zwarthoff E.C. Growth Regul. 1992; 2: 69-79Google Scholar, 39Swisshelm K. Ryan K. Tsuchiya K. Sager R. Proc. Natl. Acad. Sci. U. S. A. 1995; 92: 4472-4476Google Scholar). IGFBP-3 has been proposed to inhibit cell growth by reducing IGF bioavailability, by altering the responsiveness of the IGF-I receptor to IGF-I, and by mechanisms independent of the IGF-I receptor (40Valentinis B. Bhala A. DeAngelis T. Baserga R. Cohen P. Mol. Endocrinol. 1995; 9: 361-367Google Scholar, 41Conover C.A. Endocrinology. 1992; 130: 3191-3199Google Scholar). IGFBP-3 expression is also increased by TGF-β2 treatment in breast cancer cells and has been implicated in the growth inhibitory effects of TGF-β2 (42Oh Y. Muller H.M. Ng L. Rosenfeld R.G. J. Biol. Chem. 1995; 270: 13589-13592Google Scholar). These findings support the notion that TGF-β and IGFBP-3 actions are connected through a common retinoid signal transduction pathway. In this study, we examined the regulation of IGFBP-3 and TGF-β gene family expression by t-RA in normal HBE cells. We demonstrated that t-RA increased the levels of TGF-β2 and IGFBP-3 mRNA and of secreted TGF-β and IGFBP-3 proteins. These events were inhibited by a retinoid that functions as an RAR-α antagonist. Treatment with TGF-β2 increased IGFBP-3 mRNA levels, demonstrating a linkage of IGFBP-3 with TGF-β2 signaling pathways. However, the addition of latency-associated peptide (LAP), which inactivates secreted TGF-β, did not abrogate the effect of t-RA on IGFBP-3 expression. These findings provide evidence that t-RA increased TGF-β2 and IGFBP-3 expression through an RAR-α-dependent pathway, and the increase in IGFBP-3 expression by t-RA did not require activation of the TGF-β pathway by autocrine or paracrine mechanisms. Normal HBE cells were cultured from bronchial mucosal biopsy samples taken from fresh surgical specimens as monolayer cultures on standard plasticware in keratinocyte serum-free medium (Life Technologies, Inc.) containing bovine pituitary extract (BPE) and epidermal growth factor as described previously (11Kim Y.-H. Dohi D.F. Han G.R. Zou C.-P. Oridate N. Walsh G.L. Nesbitt J.C. Xu X.-C. Hong W.K. Lotan R. Kurie J.M. Cancer Res. 1995; 55: 5603-5610Google Scholar). Mink lung epithelial cells (MLECs) stably transfected with a luciferase reporter plasmid containing a truncated plasminogen activator inhibitor type I promoter (16Nunes I. Kojima S. Rifkin D.B. Cancer Res. 1996; 56: 495-499Google Scholar) were a gift from Drs. Irene Nunes and Daniel Rifkin (Department of Cell Biology and Kaplan Cancer Center, New York University Medical Center, New York, NY). MLECs were cultured in Dulbecco's modified Eagle's medium (Life Technologies, Inc.) containing 10% fetal calf serum as described previously (16Nunes I. Kojima S. Rifkin D.B. Cancer Res. 1996; 56: 495-499Google Scholar). t-RA was purchased from Sigma. The RAR-α antagonist LG100629 (43Apfel C. Bauer F. Crettaz M. Forni L. Kamber M. Kaufmann F. LeMotte P. Pirson W. Klaus M. Proc. Natl. Acad. Sci. U. S. A. 1992; 89: 7129-7133Google Scholar) was obtained from Ligand Pharmaceuticals, Inc. (La Jolla, CA), and LAP was provided by R & D Systems, Inc. (Minneapolis, MN). Recombinant TGF-β2 was purchased from Genzyme, Inc. (Cambridge, MA). Actinomycin D and cycloheximide were purchased from Sigma. Total cellular RNA was prepared from normal HBE cells, electrophoresed (20 μg/lane) on a 1% agarose gel containing 2% formaldehyde, transferred to a nylon membrane (Zetaprobe, Bio-Rad), hybridized to an [α-32P]dCTP-labeled cDNA probe, washed, and autoradiographed as described previously (11Kim Y.-H. Dohi D.F. Han G.R. Zou C.-P. Oridate N. Walsh G.L. Nesbitt J.C. Xu X.-C. Hong W.K. Lotan R. Kurie J.M. Cancer Res. 1995; 55: 5603-5610Google Scholar). cDNAs for TGF-β1, -β2, and -β3 (44Derynck R. Jarrett J.A. Chen E.Y. Eaton D.H. Bell J.R. Assoian R.K. Roberts A.B. Sporn M.B. Goeddel D.V. Nature. 1985; 316: 701-705Google Scholar, 45De Martin R. Haendler B. Hofer-Warbinek R. Gaugitsch H. Wrann M. Schlusner H. Seifert J.M. Bodmer S. Fontana A. Hofer E. EMBO J. 1987; 6: 3673-3677Google Scholar, 46Derynck R. Lindquist P.B. Lee A. Wen D. Tamm J. Graycar J.L. Rhee L. Mason A.J. Miller D.A. Coffey R.J. Moses H.L. Chen E.Y. EMBO J. 1988; 7: 3737-3743Google Scholar) were obtained from Dr. Rik Derynck (University of California, San Francisco, CA), and the IGFBP-3 cDNA (47Wood W.I. Cachianes G. Henzel W.J. Winslow G.A. Spencer S.A. Hellmiss R. Martin J.L. Baxter R.C. Mol. Endocrinol. 1988; 2: 1176-1185Google Scholar) was obtained from Dr. William Wood (Genentech, Inc., San Francisco, CA). Normal HBE cells were seeded on 10-cm plates (105 cells/plate), treated with t-RA for different time periods or with medium alone, and conditioned medium samples were collected simultaneously 144 h after cell seeding. For the medium sample that represents the 0–24-h time point, treatment with 10−6m t-RA in BPE-free medium (to eliminate exogenous TGF-β) was begun 120 h after cell seeding, and an aliquot was collected at 144 h. For the sample that represents 24–48 h, treatment was begun 96 h after cell seeding, replaced at 120 h with BPE-free medium containing t-RA, and an aliquot was collected at 144 h. For the sample that represents 48–72 h, treatment was begun 72 h after cell seeding and replaced at 120 h with BPE-free medium containing t-RA, and an aliquot was collected at 144 h. For the sample that represents 72–96 h, treatment was begun 48 h after cell seeding and replaced at 120 h with BPE-free medium containing t-RA, and an aliquot was collected at 144 h. For the sample that represents 96–120 h, treatment was begun 24 h after cell seeding and replaced at 120 h with BPE-free medium containing t-RA, and an aliquot was collected at 144 h. For the control (untreated) medium sample, no t-RA was added, the medium was changed to BPE-free medium at 120 h, and an aliquot was collected at 144 h. Conditioned medium samples were transferred to three separate wells on a six-well plate seeded with MLECs that were stably transfected with the TGF-β-responsive luciferase vector (105 cells/well). MLECs were treated for 16 h, and then cytosolic luciferase activity was measured as described previously (11Kim Y.-H. Dohi D.F. Han G.R. Zou C.-P. Oridate N. Walsh G.L. Nesbitt J.C. Xu X.-C. Hong W.K. Lotan R. Kurie J.M. Cancer Res. 1995; 55: 5603-5610Google Scholar). Luciferase values reflect TGF-β bioactivity in the conditioned medium. To examine bioactivity that reflects total (active plus latent) TGF-β levels, MLECs were treated with conditioned medium previously heated to 80 °C for 5 min, which activates latent TGF-β (21Kojima S. Rifkin D.B. J. Cell. Physiol. 1993; 155: 323-332Google Scholar). The results represent the means and standard deviations of luciferase activities from three identical wells, and luciferase activities were corrected for normal HBE cell numbers determined at the end of t-RA treatment. The presence of t-RA in the conditioned medium did not appreciably alter luciferase activity in MLECs (data not shown). Conditioned medium samples were collected from t-RA-treated normal HBE cells as described for the TGF-β bioactivity assay. Aliquots determined on the basis of equal cell number were concentrated 1:10 with Centriprep 10 filters (Amicon, Inc., Beverly, MA), loaded on a 1% SDS, 12% polyacrylamide gel, blotted to a nitrocellulose membrane (BAS-83, Schleicher & Schuell, Inc., Burlingame, VT), incubated for 1 h at room temperature with an anti-IGFBP-3 affinity purified monoclonal antibody (Diagnostic Systems Laboratories, Webster, TX), and detected with the enzyme-linked chemiluminesence assay (ECL kit, Amersham Corp.). For WLB, the membrane was incubated with a 2 × 106 cpm mixture of125I-labeled IGF-I and IGF-II and exposed to film as described previously (48Cohen P. Peehl D.M. Lamson G. Rosenfeld R.G. J. Clin. Endocrinol. Metab. 1991; 73: 407-491Google Scholar). Normal HBE cells were seeded at a density of 105 cells/10-cm plate and treated for 120 h with media alone, 10−6m t-RA alone, and 10−6m t-RA in combination with LAP at concentrations of 1, 10, 100, 200, and 500 ng/ml. At 120 h, the cells were trypsinized and stained with trypan blue, and the total viable cell number was counted by using an hemocytometer. We examined the regulation of IGFBP-3 and TGF-β1, -β2, and -β3 mRNA levels in normal HBE cells during 10−6m t-RA treatment by Northern analysis (Fig.1). An increase in TGF-β2 mRNA levels was detected between 6 and 12 h; this increase continued through 120 h. TGF-β1 was expressed constitutively with no change during treatment. TGF-β3 mRNA was not detected (data not shown). IGFBP-3 mRNA was expressed in a bimodal pattern. A transient increase was observed at 6 h, and a sustained increase appeared at 72 h of treatment. Examination of earlier time points revealed that increased IGFBP-3 mRNA was detectable at 1 h of 10−6m t-RA treatment (Fig. 2 A). The increase in IGFBP-3 mRNA at 6 h was inhibited by treatment with actinomycin D or cycloheximide (Fig. 2 B), demonstrating the necessity of both gene transcription and protein synthesis for activation of this retinoid signaling event.Figure 2Northern analysis of IGFBP-3 expression was performed on total cellular RNA prepared from normal HBE cells.Expression was examined in cells treated with 10−6m t-RA for different time periods (0.5, 1, 2, 3, 4, 5, or 6 h) (A). The effect of 10−6mt-RA treatment on IGFBP-3 expression was examined in the presence of 1 μg/ml actinomycin D (Act D) or 10 μg/ml cycloheximide (CHX) (B). Photographs of ethidium bromide-stained gels illustrate the relative amount of RNA loaded per lane. The absence (−) and the presence (+) of an agent are indicated.View Large Image Figure ViewerDownload (PPT) We investigated the contribution of RAR-α-dependent signaling pathways to the increased TGF-β2 and IGFBP-3 mRNA levels induced by t-RA. Normal HBE cells were treated for 120 h with 10−8m t-RA and different doses of LG100629 (43Apfel C. Bauer F. Crettaz M. Forni L. Kamber M. Kaufmann F. LeMotte P. Pirson W. Klaus M. Proc. Natl. Acad. Sci. U. S. A. 1992; 89: 7129-7133Google Scholar), which is a synthetic retinoid that functions as an RAR-α antagonist, and total cellular RNA was prepared for Northern analysis. In combination with 10−6m t-RA, LG100629, which must be in molar excess to inhibit t-RA-induced activation of RAR-α, was toxic to normal HBE cells (data not shown). LG100629 abrogated the increase in TGF-β2 and IGFBP-3 expression induced by 10−8m t-RA (Fig.3). TGF-β1 mRNA levels, which did not change with t-RA treatment, were not altered by LG100629. LG100629 also inhibited the increase in IGFBP-3 expression observed at 6 h of t-RA treatment (data not shown). The effect of t-RA on TGF-β bioactivity in conditioned medium was investigated. Normal HBE cells were treated with 10−6m t-RA, and conditioned medium samples that reflect defined 24-h periods (0–24, 24–48, 48–72 h, etc.) of t-RA treatment were collected as described under "Experimental Procedures." MLECs that are stably transfected with a reporter plasmid containing a truncated plasminogen activator inhibitor type I promoter (16Nunes I. Kojima S. Rifkin D.B. Cancer Res. 1996; 56: 495-499Google Scholar) were treated with conditioned medium samples for 16 h and subjected to luciferase assays to determine relative levels of TGF-β bioactivity in the conditioned media. Total TGF-β (active plus latent) bioactivity was measured by heating the conditioned medium samples to convert latent TGF-β into its active form. Luciferase activities increased during t-RA treatment (Fig. 4), demonstrating increases in total and active TGF-β. The increase in active TGF-β occurred between 48 and 72 h of treatment. The effect of t-RA on IGFBP-3 protein secretion was examined in conditioned medium samples collected at 24-h intervals (as described for the MLEC luciferase assay) by performing immunoblot and WLB (Fig.5). IGFBP-3 was detected in the media of untreated cells (t = 0) by WLB, and increased IGFBP-3 levels, first detected by immunoblot at 24 h, occurred with t-RA treatment. Because the increase in active TGF-β protein in the media coincided with the increase in IGFBP-3 mRNA levels at 72 h, we examined whether IGFBP-3 expression is responsive to activation of TGF-β2 signaling pathways. Normal HBE cells were treated with 5 ng/ml TGF-β2, which has been shown to increase IGFBP-3 expression in human breast cancer cells (42Oh Y. Muller H.M. Ng L. Rosenfeld R.G. J. Biol. Chem. 1995; 270: 13589-13592Google Scholar), and Northern analysis was performed at 72 h, revealing increased IGFBP-3 mRNA (Fig.6). The contribution of TGF-β signaling pathways to the t-RA-induced increase in IGFBP-3 mRNA was explored by treatment with 10−6m t-RA combined with varying doses of LAP, which noncovalently associates with active TGF-β in the extracellular space, converting TGF-β into its inactive form. Examination of conditioned medium samples collected on day 5 revealed that LAP abrogated the increase in TGF-β activity induced by 10−6m t-RA (Fig. 7), confirming its inhibitory effect on extracellular TGF-β activity. However, LAP did not measurably alter the effects of t-RA on IGFBP-3 mRNA or secreted protein levels (Fig. 8).Figure 7Luciferase assays were performed on MLECs stably transfected with a reporter plasmid containing a TGF-β response element (as described under "Experimental Procedures"). For these studies, normal HBE cells were treated for 120 h with media alone, 10−6m t-RA alone, or 10−6m t-RA combined with different doses of LAP. Fresh medium (containing BPE) was added at 72 h, and conditioned medium samples were collected at 120 h. MLECs seeded onto six-well plates were treated with the conditioned medium samples for 16 h and subjected to luciferase assays. The results represent the means and standard deviations of luciferase activities from three identical wells, and luciferase activities were corrected for normal HBE cell numbers determined at the end of t-RA treatment.RA, retinoic acid.View Large Image Figure ViewerDownload (PPT)Figure 8Northern (A) and immunoblot analysis (B) were performed using total cellular RNA and conditioned medium samples, respectively, prepared from normal HBE cells treated for 120 h with media alone, 10−6m t-RA, or 10−6m t-RA combined with different doses of LAP. A photograph of the ethidium bromide-stained gel illustrates the relative amounts of RNA loaded per lane. Media were replaced at 72 h, and conditioned medium samples were collected at 120 h. Sample aliquots for loading were determined on the basis of equal cell number. The positions of molecular size markers are indicated for the immunoblot.View Large Image Figure ViewerDownload (PPT) In this study, we demonstrated that t-RA increased IGFBP-3 and TGF-β2 mRNA and protein levels in normal HBE cells. t-RA increases the expression of IGFBP-3 and TGF-β family members in a variety of transformed and nontransformed cell lines (13Glick A.B. Flanders K.C. Danielpour D. Yuspa S.H. Sporn M.B. Cell. Regul. 1989; 1: 87-97Google Scholar, 14Niles R.M. Thompson N.L. Fenton F. In Vitro Cell. Dev. Biol. Anim. 1994; 30: 256-262Google Scholar, 15Glick A.B. McCune B.K. Abdulkarem N. Flanders K.C. Lumadue J.A. Smith J.M. Sporn M.B. Development. 1991; 111: 1081-1086Google Scholar, 16Nunes I. Kojima S. Rifkin D.B. Cancer Res. 1996; 56: 495-499Google Scholar, 17Danielpour D. J. Cell. Physiol. 1996; 166: 231-239Google Scholar, 18Turley J.M. Funakoshi S. Ruscetti F.W. Kasper J. Murphy W.J. Longo D.L. Birchenall-Roberts M.C. Cell Growth & Differ. 1995; 6: 655-663Google Scholar, 19Nugent P. Potchinsky M. Lafferty C. Greene R.M. Exp. Cell Res. 1995; 220: 495-500Google Scholar, 20Cohen P.S. Letterio J.J. Gaetano C. Chan J. Matsumoto K. Sporn M.B. Thiele C.J. Cancer Res. 1995; 55: 2386-2830Google Scholar, 21Kojima S. Rifkin D.B. J. Cell. Physiol. 1993; 155: 323-332Google Scholar, 22Batova A. Danielpour D. Pirisi L. Creek K.E. Cell Growth & Differ. 1992; 3: 763-772Google Scholar, 23Wrana J.L. Attisano L. Wieser R. Ventura F. Massagué J. Nature. 1994; 370: 341-347Google Scholar, 24Gucev Z.S. Oh Y. Kelly K.M. Rosenfeld R.G. Cancer Res. 1996; 56: 1545-1550Google Scholar, 25Zhou Y. Mohan S. Linkhart T.A. Baylink D.J. Strong D.D. Endocrinology. 1996; 137: 975-983Google Scholar, 26Sheikh M.S. Shao Z.-M. Chen J.-C. Clemmons D.R. Roberts Jr., C.T. Leroith D. Fontana J.A. Biochem. Biophys. Res. Commun. 1992; 188: 1122-1130Google Scholar, 27Sheikh M.S. Shao A.-M. Hussain A. Clemmons D.R. Chen J.-C. Roberts Jr., C.T. LeRoith D. Fontana J.A. J. Cell. Physiol. 1993; 155: 556-567Google Scholar, 28Leyen S.A.-V. Hembree J.R. Eckert R.L. J. Cell. Physiol. 1994; 160: 265-274Google Scholar, 29Adamo M.L. Shao Z.-M Lanau F. Chen J.C. Clemmons D.R. Roberts Jr., C.T. LeRoith D. Fontana J.A. Endocrinology. 1992; 131: 1858-1866Google Scholar, 30Martin J.L. Coverley J.A. Pattison S.T. Baxter R.C. Endocrinology. 1995; 136: 1219-1226Google Scholar, 31Katz J. Weiss H. Goldman B. Kanety H. Stannard B. LeRoith D. Shemer J. J. Cell. Physiol. 1995; 165: 223-227Google Scholar, 32Hembree J.R. Agarwal C. Beard R.L. Chandraratna R.A.S. Eckert R.L. Cancer Res. 1996; 56: 1794-1799Google Scholar, 33Sheikh M.S. Shao Z.-M. Hussain A. Clemmons D.R. Chen J.-C. Roberts Jr., C.T. LeRoith D. Fontana J.A. J. Cell. Physiol. 1993; 155: 556-567Google Scholar, 34Kordon E.C. McKnight R.A. Jhappan C. Henninghausen L. Merlino G. Smith G.H. Dev. Biol. 1995; 168: 47-61Google Scholar, 35Fontana J.A. Burrows-Mezu A. Clemmons D.R. LeRoith D. Endocrinology. 1991; 128: 1115-1122Google Scholar). In contrast to our findings in normal HBE cells, the levels of TGF-β1 and -β2 both increase in t-RA-treated hamster tracheal epithelial cells grown in collagen gels (14Niles R.M. Thompson N.L. Fenton F. In Vitro Cell. Dev. Biol. Anim. 1994; 30: 256-262Google Scholar), providing evidence that retinoid signaling in bronchial epithelial cells depends upon the culture conditions and is species-specific. Further, we found that t-RA stimulated the conversion of latent TGF-β into its active form, suggesting that t-RA activated a protease that cleaves latent TGF-β in normal HBE cells. This TGF-β activity accounted for a relatively small fraction of total TGF-β, suggesting that protease activity is tightly regulated. In bovine epithelial cells, t-RA activates latent TGF-β through enhanced cell-associated plasmin activity (21Kojima S. Rifkin D.B. J. Cell. Physiol. 1993; 155: 323-332Google Scholar), whereas in HL-60 cells, activation of TGF-β by t-RA occurs through plasmin-independent mechanisms (16Nunes I. Kojima S. Rifkin D.B. Cancer Res. 1996; 56: 495-499Google Scholar). We examined the role of RAR-α, which is expressed constitutively and activates growth inhibitory pathways in normal HBE cells (11Kim Y.-H. Dohi D.F. Han G.R. Zou C.-P. Oridate N. Walsh G.L. Nesbitt J.C. Xu X.-C. Hong W.K. Lotan R. Kurie J.M. Cancer Res. 1995; 55: 5603-5610Google Scholar), in these retinoid signaling events. An RAR-α antagonist inhibited the effects of t-RA on TGF-β2 and IGFBP-3 expression. Normal HBE cells exhibited a bimodal pattern of IGFBP-3 mRNA expression, and the antagonist inhibited the increase at both time points. Treatment with an RAR-α antagonist also blocked the increase in transglutaminase type II expression induced by t-RA in rat tracheobronchial epithelial cells (12Zhang L.-X. Mills K.J. Dawson M.I. Collins S.J. Jetten A.M. J. Biol. Chem. 1994; 270: 6022-6029Google Scholar). Further, RAR activation increases lGFBP-3 expression in human ectocervical epithelial cell lines (32Hembree J.R. Agarwal C. Beard R.L. Chandraratna R.A.S. Eckert R.L. Cancer Res. 1996; 56: 1794-1799Google Scholar). These studies support a role for RARs in specific retinoid signaling events. Retinoid receptors regulate gene expression directly by binding to gene promoters or indirectly by protein-protein interactions with other transcription factors. We found that transcriptional mechanisms contribute to the increase in IGFBP-3 expression, and t-RA induced this effect rapidly (1 h). Based on these findings, we will investigate direct binding of retinoid receptors to the IGFBP-3 gene promoter and its role in the activation of IGFBP-3 expression by t-RA. We found increased TGF-β bioactivity in the media at 72 h of t-RA treatment, which was coincident with the increase in IGFBP-3 mRNA levels, and TGF-β2 treatment increased IGFBP-3 expression. Similarly, TGF-β2 treatment increases IGFBP-3 expression in breast cancer cells (42Oh Y. Muller H.M. Ng L. Rosenfeld R.G. J. Biol. Chem. 1995; 270: 13589-13592Google Scholar). Based on these findings, we investigated whether t-RA-induced IGFBP-3 expression was TGF-β-dependent. Supporting this possibility, retinoid signaling events such as increased expression of thrombospondin and fibronectin are TGF-β-dependent in rat prostatic epithelial cells (17Danielpour D. J. Cell. Physiol. 1996; 166: 231-239Google Scholar). We found that LAP did not abrogate the increase in IGFBP-3 expression induced by t-RA, suggesting that the increase in IGFBP-3 mRNA levels by t-RA did not require activation of the TGF-β2 signaling pathway through autocrine or paracrine mechanisms. It is possible that multiple retinoid signaling pathways, including TGF-β-dependent ones, activate IGFBP-3 expression, and blocking one retinoid signaling pathway is not sufficient to alter the effects of t-RA on IGFBP-3 expression. Further, TGF-β may activate IGFBP-3 expression through intracrine mechanisms, which LAP cannot inhibit. Supporting this possibility, intracrine signaling has been reported to be a mechanism of TGF-β1 actions in mammary epithelial differentiation (34Kordon E.C. McKnight R.A. Jhappan C. Henninghausen L. Merlino G. Smith G.H. Dev. Biol. 1995; 168: 47-61Google Scholar). Prior work in a variety of cell types has demonstrated the importance of the TGF-β and IGFBP-3 signaling pathways in the biologic effects of t-RA. Blocking antibodies to TGF-β inhibit the mucous differentiation of hamster tracheal epithelial cells by t-RA and partially abrogate the growth inhibitory effects of t-RA in lymphoma cells, rat prostatic epithelial cells, and keratinocytes (13Glick A.B. Flanders K.C. Danielpour D. Yuspa S.H. Sporn M.B. Cell. Regul. 1989; 1: 87-97Google Scholar, 14Niles R.M. Thompson N.L. Fenton F. In Vitro Cell. Dev. Biol. Anim. 1994; 30: 256-262Google Scholar, 17Danielpour D. J. Cell. Physiol. 1996; 166: 231-239Google Scholar,18Turley J.M. Funakoshi S. Ruscetti F.W. Kasper J. Murphy W.J. Longo D.L. Birchenall-Roberts M.C. Cell Growth & Differ. 1995; 6: 655-663Google Scholar). In breast cancer cells, t-RA increases IGFBP-3 expression, and antisense IGFBP-3 oligonucleotides abrogate the growth inhibitory effects of t-RA (24Gucev Z.S. Oh Y. Kelly K.M. Rosenfeld R.G. Cancer Res. 1996; 56: 1545-1550Google Scholar). In contrast, LAP did not alter the growth inhibitory effects of t-RA on normal HBE cells (data not shown). These findings suggest that the role of the TGF-β pathway in cell growth and differentiation depends on the tissue of origin, culture conditions, and transformation state of the cells examined. Additional investigations into the roles of the TGF-β and IGFBP-3 signaling pathways in the biologic effects of retinoid treatment are warranted.
Referência(s)