Ultraviolet irradiation blocks cellular responses to transforming growth factor-beta by down-regulating its type-II receptor and inducing Smad7.

Transforming growth factor-beta (TGF-beta) is a multi-functional cytokine that regulates cell growth and differentiation. Cellular responses to TGF-beta are mediated through its cell surface receptor complex, which activates transcription factors Smad2 and Smad3. Here we report that UV irradiation of mink lung epithelial cells causes near complete inhibition of TGF-beta-induced Smad2/3-mediated gene expression. UV irradiation inhibited TGF-beta-induced phosphorylation of Smad2 and subsequent nuclear translocation and DNA binding of Smad2/3. Specific cell surface binding of TGF-beta was substantially reduced after UV irradiation. This loss of TGF-beta binding resulted from UV-induced down-regulation of TGF-beta type II receptor (T beta RII) mRNA and protein. UV irradiation significantly inhibited T beta RII promoter reporter constructs, indicating that UV reduction of T beta RII expression involved transcriptional repression. In contrast to its effects on T beta RII, UV irradiation rapidly induced Smad7 mRNA and protein. Smad7 is known to antagonize activation of Smad2/3 and thereby block TGF-beta-dependent gene expression. UV irradiation stimulated Smad7 promoter reporter constructs, indicating that increased Smad7 expression resulted, at least in part, from increased transcription. Overexpression of Smad7 protein to the level induced by UV irradiation inhibited TGF-beta-induced gene expression 30%. Maintaining T beta RII levels by overexpression of T beta RII prevented UV inhibition of TGF-beta responsiveness. Taken together, these data indicate that UV irradiation blocks cellular responsiveness to TGF-beta through two mechanisms that impair TGF-beta receptor function. The primary mechanism is down-regulation of T beta RII, and the secondary mechanism is induction of Smad7.

Transforming growth factor-␤ (TGF-␤) is a multi-functional cytokine that regulates cell growth and differentiation. Cellular responses to TGF-␤ are mediated through its cell surface receptor complex, which activates transcription factors Smad2 and Smad3. Here we report that UV irradiation of mink lung epithelial cells causes near complete inhibition of TFG-␤-induced Smad2/3-mediated gene expression. UV irradiation inhibited TGF-␤-induced phosphorylation of Smad2 and subsequent nuclear translocation and DNA binding of Smad2/3. Specific cell surface binding of TGF-␤ was substantially reduced after UV irradiation. This loss of TGF-␤ binding resulted from UV-induced down-regulation of TGF-␤ type II receptor (T␤RII) mRNA and protein. UV irradiation significantly inhibited T␤RII promoter reporter constructs, indicating that UV reduction of T␤RII expression involved transcriptional repression. In contrast to its effects on T␤RII, UV irradiation rapidly induced Smad7 mRNA and protein. Smad7 is known to antagonize activation of Smad2/3 and thereby block TGF-␤-dependent gene expression. UV irradiation stimulated Smad7 promoter reporter constructs, indicating that increased Smad7 expression resulted, at least in part, from increased transcription. Overexpression of Smad7 protein to the level induced by UV irradiation inhibited TGF-␤-induced gene expression 30%. Maintaining T␤RII levels by overexpression of T␤RII prevented UV inhibition of TGF-␤ responsiveness. Taken together, these data indicate that UV irradiation blocks cellular responsiveness to TGF-␤ through two mechanisms that impair TGF-␤ receptor function. The primary mechanism is down-regulation of T␤RII, and the secondary mechanism is induction of Smad7.
Transforming growth factor-beta (TGF-␤) 1 family members are multifunctional cytokines whose cellular effects are dependent on cell type and cellular context. For example, TGF-␤ stimulates proliferation of fibroblasts in connective tissue and inhibits growth of epithelial cells (1). The TGF-␤s play impor-tant roles in cellular differentiation and biosynthesis of extracellular matrix (2,3). Impairment of TGF-␤ responsiveness occurs in a variety of cancer cells and contributes to loss of growth control (4 -7).
TGF-␤ signal transduction is mediated by a complex of three transmembrane receptors, Type I (T␤RI), Type II (T␤RII), and Type III (T␤RIII) TGF-␤ receptors. T␤RI and T␤RII possess intrinsic serine/threonine kinase activity. T␤RIII is a membrane proteoglycan that is thought to facilitate ligand binding to T␤RII (8 -11). Binding of ligand to TGF-␤ receptors induces formation of a heteromeric complex of T␤RI and T␤RII receptors (2,(12)(13)(14)(15). Formation of this heteromeric complex enables the T␤RII to phosphorylate T␤RI, resulting in activation of T␤RI kinase (2, 13, 16 -18). T␤RI phosphorylates and thereby activates transcription factors Smad2 and Smad3 (1,2,19,20). Phosphorylated Smad2 and/or Smad3 then bind their common partner, Smad4, to form a heteromeric complex, which then translocates to and accumulates in the nucleus (19), where it acts as a transcription factor (2,15). The actions of TGF-␤ are antagonized by Smad7, which interacts stably with T␤RI to prevent phosphorylation and activation of receptor-regulated Smad2/3, thereby blocking TGF-␤ signaling (21,22). Solar ultraviolet (UV) irradiation is a potent environmental hazard capable of damaging cellular DNA and causing mutations (23,24). Recent studies demonstrate that UV irradiation of cells causes ligand-independent rapid activation and clustering of many different growth factor and cytokine cell surface receptors (24 -26). The resulting stimulation of multiple signal transduction pathways leads to induction and activation of transcription factors, notably activator protein-1 and NF-B, which results in transcription of their target genes (25,(27)(28)(29).
Despite widespread interest in mechanisms of action of UV irradiation, relatively little is known regarding the effect of UV irradiation on TGF-␤ receptor activation or on TGF-␤/Smad signaling. We report here that in contrast to growth factor-and cytokine receptor-mediated signal transduction, UV irradiation impairs TGF-␤ receptor-mediated signal transduction. This impairment of TGF-␤ responsiveness results from downregulation of T␤RII and induction of Smad7 by UV irradiation.
Before UV irradiation, sub-confluent cells were incubated overnight in serum-free media. The next morning, the media were collected, and cells were covered with a thin layer of PBS and irradiated with UV (20 mJ/cm 2 ) using a Daavlin Spectra panel lamp (Bryan, OH) containing six FS24T12UVB-HO bulbs. A Kodacel filter was used to eliminate wavelengths below 290 nm (UVC). The irradiation intensity was monitored with an IL400A radiometer and a SED240/UVB/W photodetector (International Light, Newbury, MA). After irradiation, the PBS was aspirated, and the original media were put back into the plates. Cellular viability 24 h after UV irradiation was near 100%, based on cell morphology and number. For TGF-␤ 1 treatment, TGF-␤ 1 (1 ng/ml) was added at the indicated times post-UV for 1 h.
Immunofluorescence Confocal Laser-scanning Microscopy-MLECs (1 ϫ 10 4 ) were plated on 8-well chamber slides and exposed to UV (20 mJ/cm 2 ). TGF-␤1 (1 g/ml) was added at the indicated times post-UV for 1 h. Cells were rinsed once with PBS and fixed in 5% paraformaldehyde for 2 h at room temperature. The cells were incubated with 0.5% Nonidet P-40 and then blocked with 2% bovine serum albumin (BSA). Cells were then incubated overnight with primary antibody (anti-Smad2 or anti-Smad3, 1:500 dilution) at 4°C. Slides were subsequently washed with PBS five times and incubated with primary antibody for 1 h. Fluorescein isothiocynate-labeled secondary antibodies were then added for 1 h. The fluorescence-stained cells were observed and photographed with a Bio-Rad MRC 600 confocal microscope through a 60ϫ objective.
RNA Isolation and Northern Blot Analysis-Total RNA from MLECs was extracted with a commercial kit (RNeasy Midi Kit, Qiagen, Chatsworth, CA) according to the manufacturer's protocol. Samples of total RNA (30 -50 g) were resolved by 1.2% agarose electrophoresis, transferred to nylon membranes, and hybridized with T␤RI, T␤RII, and Smad7 cDNA probes labeled with [ 32 P]dCTP by random priming. Each blot was stripped and re-hybridized with 36B4 internal control gene transcript to monitor the sample load in each lane. The intensities of each band were quantified by STORM PhosphorImager (Molecular Dynamics, Sunnyvale, CA) and normalized to the 36B4 gene transcript. cDNA probe was generated by digestion of Smad7 cDNA expression vector (30) (obtained from Dr. Peter ten Dijke, Ludwig Institute for Cancer Research, Uppsala, Sweden) with HindIIII and XbaI. cDNA probes for T␤RI and T␤RII were generated by polymerase chain reaction from human skin total RNA using the following primers: for T␤RI, 5Ј-CCTGGCCTTGGTCCTGTG-3Ј and 5Ј-TCTGTGGCTGAATCATGTC-TTACT-3Ј; for T␤RII, 5Ј-AACTGTGTAAATTTTGTGATGTGA-3Ј and 5Ј-CGGGCCTCTGGGTCGTG-3Ј. The polymerase chain reaction products were sub-cloned into the pCRII vector (Invitrogen, Carlsbad, CA) and verified by restriction digestion and sequencing.
Western Analysis-Nuclear extracts were prepared as described previously (31). Membrane fractions were prepared according to a method previously described (32) with slight modification. Briefly, MLECs were washed twice with PBS and suspended in fresh hypotonic buffer (20 mM Tris, pH 7.5, 10 mM NaCl, 0.3 M phenylmethylsulfonyl fluoride). Cells were then harvested on ice by scraping and homogenized with a Dounce homogenizer. Crude nuclei were pelleted at 2000 ϫ g for 10 min, and the supernatant was then spun at 25,000 ϫ g for 30 min at 4°C. To obtain a crude membrane fraction, the membrane pellet was re-suspended in whole cell extraction buffer (25 mM HEPES, pH 7.7, 0.3 M NaCl, 1.5 mM MgCl 2 , 0.2 mM EDTA, 0.1% Triton X-100, 0.5 mM dithiothreitol, 20 mM ␤-glycerol phosphate, 0.1 mM Na 3 VO 4 , 2 g/ml leupeptin, and 100 g/ml phenylmethylsulfonyl fluoride) and stored at Ϫ80°C. Concentrations of membrane or nuclear proteins were measured by Bio-Rad protein assay (Bio-Rad) using BSA as a standard. Nuclear or membrane proteins were resolved on 8% SDS-polyacrylamide gel electrophoresis, transferred to polyvinylidene difluoride membrane, and reacted with primary antibodies. Blots were visualized and quantified with enhanced chemifluorescence (ECF) (Vistra ECF Western blotting system, Amersham Pharmacia Biotech) following the manufacturer's protocol. The intensities of each band were quantified by STORM PhosphorImager.
Whole Cell Binding and Cross-linking of 125 I-TGF-␤1-Binding of 125 I-TGF-␤ to MLECs was performed according to the method described (33) with minor modifications. Sub-confluent cells were incubated overnight in complete medium. Cells were then irradiated with UV as described above, and incubation was continued for 1, 3, 5, 8, 16, or 24 h after exposure. Cells were washed once with Krebs-Ringer-Hepes/BSA binding buffer (50 mM Hepes, pH 7.5, 128 mM NaCl, 1.3 mM CaCl 2 , 5 mM KCl, 0.5% BSA) and then incubated in Krebs-Ringer-Hepes/BSA for 30 min at 37°C. Cells were washed with 0.1% glacial acetic acid for 5 min at room temperature, placed in Krebs-Ringer-Hepes/BSA, and then 0.2 nM 125 I-TGF-␤ 1 was added for 3-4 h at 4°C. Nonspecific binding was determined by addition of a 100-fold excess of unlabeled TGF-␤ 1. Cells were harvested by scraping, and radioactivity was measured as described previously (34). Cross-linking of 125 I-TGF-␤ 1 to its receptors was carried out as previously described (35), with minor modifications. Briefly, sub-confluent cells were treated with 125 I-TGF-␤1 as described above for the binding assay. Reactions were halted by adding glycine. 125 I-TGF-␤ 1 was cross-linked to receptors by the addition of disuccinimidyl suberate (0.5 mg/ml). Cellular proteins were solubilized then resolved on 8% SDS-polyacrylamide electrophoresis gels.
Electrophoretic Mobility Shift Assay-Electrophoretic mobility shift assays were performed with nuclear extracts from MLECs. Doublestranded oligodeoxynucleotides containing the TGF-␤ response element in the PAI-1 promoter (5Ј-TCGAGAGCCAGACAAGGAGCCAGACAA-GGAGCCAGACAC-3Ј and its complementary strand) and the consensus Smad binding element (5Ј-GGATAGCGTCTAGACATAGTC-TAGACTGAGT-3Ј and its complementary strand) were used as probes (36,37). All oligonucleotides were synthesized by Operon Technologies, Inc. (Alameda, CA). Electrophoretic mobility shift assays were performed as described previously (38) with minor modifications. Briefly, synthetic single strand oligonucleotides were incubated in 10 mM Tris-HCl, pH 8.0, 1 mM EDTA, and 0.3 M NaCl at 75°C for 15 min and placed at room temperature overnight. The annealed double-stranded oligonucleotides were 5Ј-end-labeled with [␥-32 P]ATP using T4 polynucleotide kinase (Life Technologies Inc.). The end-labeled probe was purified with a G50 column (Roche Molecular Biochemicals). Approximately 2 ϫ 10 5 cpm of end-labeled DNA probe was incubated with 5-10 g of MLEC nuclear extract in a volume of 20 l. For competition experiments, a 10 -50-fold molar excess of unlabeled DNA probe was preincubated with nuclear extract for 30 min on ice before labeled probe was added. After 30 min of incubation on ice, the protein-DNA complexes were electrophoresed on 4% polyacrylamide gel at 200 V in 0.5 ϫ TBE (1.0 M Tris, 0.9 M boric acid, 0.01 M EDTA). The gel was transferred to Whatman No. 3MM paper, vacuum-dried, and scanned by STORM PhosphorImager.
Transient Transfection and Luciferase Assays-␤-Galactosidase expression vector (1 g, pCMV␤, CLONTECH Laboratories, Inc., Palo Alto, CA) was co-transfected to provide an internal standard for transfection efficiency. Three luciferase promoter reporter constructs were utilized: 1) Smad binding element X4 (39), containing four repetitions of the GTCTAGAC Smad3/4 binding motif (provided by Dr. Bert Vogelstein of the Johns Hopkins Oncology Center, Baltimore, MD), 2) T␤RII promoter (40) (provided by Dr. Seong-Jin Kim of the NCI, National Institutes of Health, Bethesda, Maryland), and 3) Smad7 promoter (41) (provided by Dr. Yan Chen of the Indiana University School of Medicine). All plasmids were introduced into cells using Fugene 6 (Roche Molecular Biochemicals) according to the manufacturer's protocol. Where indicated in the text, cells were UV-irradiated as described above 24 h after transfection, and 8 h later, TGF-␤1 (1 ng/ml) was added to the medium. Twenty-four h after treatment, cells were washed once with PBS, harvested in lysis buffer (Pharmingen International, San Diego, CA), and assayed for ␤-galactosidase activity (42). Aliquots containing identical ␤-galactosidase activity were used for each luciferase assay (43). Luciferase activity was measured using an enhanced luciferase assay kit (Pharmingen International) according to the manufacturer's protocol. Smad7 (30) or T␤RII (44) were overexpressed by transient transfection in MLECs stably expressing the TGF-␤-inducible PAI promoter luciferase reporter. T␤RII-coding region was generated by polymerase chain reaction using the primers 5Ј-GAATTCGTCTGCCA-TGGGTCGGGGGC-3Ј and 5Ј-TTCAGGAATCTTCTCCTCC-3Ј. The polymerase chain reaction product was cloned into pcDNA 3.1 (Invitrogen) and verified by restriction digestion and sequencing.
Statistical Analysis-Comparisons among treatment groups were made with the paired t test (two groups) or the repeated measures of analysis of variance (more than two groups). Multiple pairwise comparisons were made with the Tukey Studentized Range test. All p values are two-tailed and considered significant when Ͻ0.05.

UV Represses Transactivation by TGF-␤-
We initially investigated the effect of UV irradiation on TGF-␤-induced gene expression using two well characterized TGF-␤-regulated luciferase reporter constructs: 1) an 800-base pair fragment (Ϫ799/ ϩ71) of the human PAI-1 gene (45) and 2) four repetitions of the GTCTAGAC Smad3/4 binding element upstream of a SV40 minimal promoter (39). As shown in Fig. 1, TGF-␤ treatment activated these reporter constructs more than 10-fold. Pretreatment of cells with UV irradiation 8 h before the addition of TGF-␤1 repressed activation of both reporter genes almost completely (Fig. 1). UV treatment alone had no significant effect either on the PAI-1 reporter gene or on the Smad binding element reporter construct (Fig. 1). Taken together, these data indicate that UV irradiation inhibits TGF-␤ responsiveness.
UV Irradiation Inhibits Smad2 Phosphorylation and Nuclear Translocation-Binding of TGF-␤ to its receptors results in phosphorylation of Smad2 and Smad3, which then associate with Smad4 and translocate into the nucleus. The reduced responsiveness of TGF-␤-regulated reporter constructs to TGF-␤ after exposure of cells to UV irradiation could result from decreased Smad2/3 activation. To examine this possibility, we next examined whether UV irradiation inhibits phosphorylation of Smad2 and nuclear translocation of Smad2 and Smad3. In untreated cells, the basal level of phosphorylated Smad2 was minimally detectable by Western analysis. Treatment of cells with TGF-␤ 1 increased Smad2 phosphorylation ϳ4.5-fold (Fig. 2). Treatment of cells with UV irradiation alone did not alter the level of phosphorylated Smad2 (data not shown). However, pretreatment of cells with UV (20 mJ/cm 2 ) before the addition of TGF-␤ 1 resulted in a time-dependent inhibition of Smad2 phosphorylation. Although TGF-␤1-induced Smad2 phosphorylation was not sufficiently changed at 1 and 3 h post-UV, it was almost completely blocked at 8 and 16 h post-UV. In contrast, total Smad2 levels were not affected by either TGF-␤ or UV treatment (Fig. 2).
We next examined whether UV irradiation affects TGF-␤induced nuclear translocation of phospho-Smad2 and Smad3. As shown in Fig. 3, phospho-Smad2 and Smad3 were predominantly located in the cytoplasm of untreated cells or cells treated with UV alone (data not shown). Treatment of cells with TGF-␤1 induced nuclear translocation of phospho-Smad2 and Smad3. Pretreatment of cells with UV 8 h before the addition of TGF-␤ 1 substantially blocked TGF-␤-induced translocation of phospho-Smad2 and Smad3 (Fig. 3).
UV Irradiation Inhibits DNA Binding of Smad Proteins-UV inhibition of Smad2 and Smad3 activation by TGF-␤ would be expected to reduce Smad protein DNA binding. To investigate this possibility, we performed electrophoretic mobility shift assays with two well characterized Smad3/4 DNA binding probes, one containing the TGF-␤ response element in the PAI-1 promoter (36), and a second probe, identified by random oligonucleotide screening, as a Smad3 DNA binding element (37). MLECs treated with TGF-␤ alone exhibited enhanced protein binding to both probes (Fig. 4). In contrast, treatment of cells with UV irradiation before the addition of TGF-␤ blocked protein binding to both probes in a time-dependent manner (Fig. 4). These results indicate that UV inhibits TGF-␤-induced DNA binding of Smad proteins.
UV Irradiation Inhibits Binding of TGF-␤1 to Its Cell Surface Receptors-The above data indicate that UV irradiation interferes with TGF-␤ activation of Smad2 and Smad3. This interference could occur as a result of physical and/or func- tional loss of TGF-␤ receptors. To investigate this possibility, we determined the effect of UV irradiation on TGF-␤ binding to its cell surface receptors. Fig. 5A shows specific binding of 125 I-TGF-␤ 1 to intact MLEC after UV irradiation. TGF-␤ binding increased 35% at 1 h post-UV and returned to its initial level by 3 h post-UV. Between 5 and 24 h post-UV, there was a progressive decrease in the level of 125 I-TGF-␤1 binding. At 24 h post-UV, TGF-␤1 binding was reduced ϳ60% relative to non-irradiated control cells (100%) (Fig. 5A).
We next determined the level of TGF-␤⅐TGF-␤ receptor complexes by cross-linking 125 I-TGF-␤ 1 to its receptors in intact MLEC. In non-irradiated and UV-irradiated cells, three major binding complexes were identified, with apparent molecular weights of 65,000, 85,000, and 200,000 -250,000, shown previously to be TGF-␤ complexed with T␤RI, T␤RII, and T␤RIII, respectively (46). Consistent with other studies (47,48), T␤RIII receptors were most prominently labeled with 125 I-TGF-␤ 1, whereas T␤RII and T␤RI yielded weaker-labeled complexes (Fig. 5B). Cells treated with UV exhibited significant time-dependent decreases of the cross-linked complexes for all three TGF-␤ receptors relative to levels in non-irradiated control cells (Fig. 5B).
UV Irradiation Down-regulates T␤RII but Not T␤RI-Reduced binding of TGF-␤ to its receptors after UV irradiation could occur as a result of decreased expression of TGF-␤ receptors. To investigate this possibility, we next examined the effect of UV on TGF-␤ receptor mRNA and protein levels. UV irradiation reduced T␤RII mRNA (Fig. 6A) and protein levels (Fig.  6B) in a time-dependent manner, as measured by Northern and Western analysis, respectively. T␤RII mRNA was reduced within 3 h post-UV and was progressively reduced during the succeeding 13 h (Fig. 6A). By 16 h after UV, T␤RII mRNA was reduced ϳ50% relative to control. T␤RII protein expression was also reduced after UV exposure. T␤RII protein was maximally reduced (80%) at 16 h post-UV (Fig. 6B). In contrast, T␤RI mRNA levels were modestly increased between 1 and 3 h post-UV and then returned to their initial values (Fig. 6C). UV had no significant effect on T␤RI protein levels (Fig. 6D).
Reduction of the T␤RII mRNA after UV irradiation could reflect inhibition of T␤RII gene transcription. To examine this possibility, we transiently transfected MLECs with two different T␤RII promoter/luciferase constructs, a longer construct (Ϫ1640 to ϩ62) and a shorter construct (Ϫ137 to ϩ62). Both promoter constructs were active in non-irradiated cells (Fig. 7).
However, the shorter T␤RII promoter/luciferase construct (Ϫ137 to ϩ62) was markedly more active than the longer promoter constructs, consistent with the observed presence of a negative regulatory element between Ϫ1240 to Ϫ504 (40). UV irradiation inhibited the activity of both T␤RII promoter constructs nearly 70% (Fig. 7).
UV Irradiation Induces Smad7 mRNA and Protein-Recent evidence indicates that cytokines and growth factors can induce Smad7, which inhibits TGF-␤ signaling (49). Since UV irradiation activates cytokine and growth factor receptors in a variety of cell types (25,50), we examined whether UV inhibition of the TGF-␤ responsiveness involves induction of Smad7 in addition to loss of T␤RII. As shown in Fig. 8A, Smad7 mRNA increased 2-fold between 1 and 3 h post-UV and then returned to basal levels between 8 and 24 h post-UV. Smad7 protein levels rose steadily between 1 and 8 h post-UV and remained elevated (3-fold) for at least 24 h post-UV (Fig. 8B). We also examined the effect of UV on Smad7 gene promoter activity by transiently transfecting MLECs with long (Ϫ4200 to ϩ112) and short (Ϫ408 to ϩ112) Smad7 promoter/luciferase constructs. As shown in Fig. 8C, the two promoter constructs were equally active in non-irradiated cells and equally induced (2-fold) after UV irradiation. These data indicate that the induction of Smad7 by UV irradiation is mediated at least in part by increased transcription.
To examine the functional consequences of this induction of Smad7 by UV irradiation, we determined the relationship between Smad7 protein levels and inhibition of TGF-␤-induced gene expression. For these studies, MLECs were transfected with varying amounts of Smad7 expression vector, and the levels of Smad7 protein and TGF-␤-induced luciferase reporter activity were determined. Transfection of MLECs with increasing amounts of Smad7 cDNA resulted in dose-dependent increases in Smad7 protein expression and inhibition of TGF-␤induced luciferase activity (Fig. 9). Transfection of MLEC cells with 2 g of Smad7 cDNA resulted in an approximate 3-fold induction of Smad7 protein and 30% inhibition of TGF-␤-induced luciferase activity (Fig. 9). As described above (Fig. 8B), UV irradiation resulted in a 3-fold induction of Smad7 protein at 8 -24 h post-UV. Together, these data indicate that UV induction of Smad7 could account for ϳ30% of the total inhibitory effect of UV on TGF-␤-induced gene expression.

Overexpression of T␤RII Overcomes UV Inhibition of TGF-␤induced Gene Expression-The finding that UV-induced
Smad7 only modestly inhibits TGF-␤ responsiveness suggests that the majority of UV inhibition results from down-regulation of T␤RII. To examine this possibility, we determined representative Northern analysis. Bars heights indicate means Ϯ S.E. of T␤RII mRNA levels (normalized to 36B4 mRNA levels) relative to levels in non-irradiated control cells (Ctrl, 100%). *, p Ͻ 0.05 versus Ctrl. n ϭ 4. B, T␤RII protein in the membrane fraction of MLECs was quantified by Western analysis. The inset shows a representative Western blot. Bar heights indicate the means Ϯ S.E. of receptor protein levels relative to receptor protein levels in non-irradiated control cells (Ctrl, 100%). *, p Ͻ 0.05 versus Ctrl. n ϭ 6. C, T␤RI and 36B4 mRNA were analyzed by Northern analysis. Representative Northern blots for T␤RI and 36B4 mRNA are shown in the inset. Bar heights indicate means Ϯ S.E. for fold change in T␤RI mRNA levels (normalized to 36B4 mRNA levels) relative to levels in non-irradiated control cells (Ctrl). *, p Ͻ 0.05 versus Ctrl. n ϭ 3. D, T␤RI protein in the membrane fraction of MLEC cells was quantified by Western analysis. The inset shows representative Western blots. Bar heights indicate the means Ϯ S.E. of receptor protein levels in UV-irradiated cells relative to receptor protein levels in non-irradiated control cells (Ctrl). n ϭ 3. whether maintaining T␤RII levels by forced overexpression of T␤RII could overcome UV inhibition of TGF-␤-induced gene expression. Overexpression of T␤RII had no significant effect on TGF-␤-induced (or basal, data not shown) PAI-1 promoter activity (Fig. 10), indicating that T␤RII levels are not limiting for TGF-␤ responsiveness in non-irradiated cells. As expected, exposure of mock-transfected cells to UV before the addition of TGF-␤ repressed TGF-␤1-induced reporter gene 80%. In contrast, in cells overexpressing T␤RII, UV did not inhibit TGF-␤ responsiveness. Complete prevention of UV-induced loss of TGF-␤-induced gene expression was observed with 4 g of T␤RII expression vector. These data demonstrate that downregulation of T␤RII by UV is critical for UV inhibition of TGF-␤ responsiveness.
UV Induction of c-Jun/AP-1 Does Not Correlate with Inhibition of TGF-␤ Responsiveness-Accumulating evidence indicates that transcription factor AP-1 or its component c-Jun can both positively and negatively modulate TGF-␤/Smad-dependent gene expression (51)(52)(53). In addition, we and others show that UV irradiation induces c-Jun and AP-1 activity (23, 25, 50, 54 -56). These observations raise the possibility that UV induction of c-Jun/AP-1 could contribute to reduced TGF-␤ responsiveness. To examine this issue, we determined the kinetics of UV induction of c-Jun protein and AP-1 DNA binding in MLECs. UV irradiation induced c-Jun protein 3.0 Ϯ 0.7-fold (n ϭ 3) within 1 h. Maximum induction (5.8 Ϯ 0.9-fold, n ϭ 3) occurred within 5 h post-UV. c-Jun protein levels returned to base line by 24 h post-UV (1.3 Ϯ 0.3-fold, n ϭ 3). Electrophoretic mobility shift assays using oligonucleotides containing the consensus AP-1 binding element as probe revealed nearly identical kinetics of induction (data not shown). These kinetics of c-Jun/AP-1 activation substantially differ from the kinetics of UV inhibition of TGF-␤ responsiveness described above. At early times post-UV (1-5 h), c-Jun/AP-1 is maximally induced, whereas TGF-␤ responsiveness is not altered. At later times post-UV (8 -24 h) c-Jun/AP-1 activation subsides, whereas TGF-␤ responsiveness is inhibited. These data indicate that UV inhibition of TGF-␤ responsiveness does not correlate with c-Jun/AP-1 activation. DISCUSSION Accumulating evidence indicates that UV irradiation activates cytokine and growth factor signal transduction pathways (25,50). In contrast, we report here that UV irradiation impairs TGF-␤/Smad signaling. This impairment results primarily from reduction of T␤RII expression and, to a lesser extent, from induction of Smad7. This UV-induced loss of T␤RII and increase in Smad7 impairs TGF-␤ signal transduction as man- FIG. 8. UV irradiation induces Smad7 mRNA, protein expression, and promoter activity. MLECs were exposed to UV irradiation (20 mJ/cm 2 ). At the indicated times post-UV, cells were collected, and either total RNA or protein were extracted. A, Smad7 and 36B4 (used as an internal control) mRNA levels were analyzed by Northern analysis. The inset shows representative Northern blot for Smad7 and 36B4 (normalized to 36B4 mRNA levels). The bars indicate the means Ϯ S.E. for fold change in Smad7 mRNA levels relative to levels in non-irradiated control cells (Ctrl). *, p Ͻ 0.05 versus control. n ϭ 3. B, Smad7 protein was quantified by Western analysis. The inset shows a representative Western blot for Smad7 protein. The bar heights indicate the means Ϯ S.E. for fold change in Smad7 protein relative to levels in non-irradiated control cells (Ctrl). *, p Ͻ 0.05 versus control. n ϭ 3. C, UV irradiation increases Smad7 promoter activity. MLECs were transiently transfected with Smad7 promoter (Ϫ4200 to ϩ112 or Ϫ408 ϩ112)/luciferase constructs. Twenty-four hours after transfection, cells were irradiated with UV (20 mJ/cm 2 ). Cell lysates were prepared 48 h after transfection and normalized by ␤-galactosidase activity. Promoter activities were determined by luciferase assay. The bar heights represent the means Ϯ S.E. *, p Ͻ 0.05 versus non-irradiated cells, n ϭ 3. ifested by inhibition of TGF-␤-induced Smad2/3 activation and nuclear translocation, inhibited formation of Smad/DNA complexes, and inhibition of TGF-␤-induced gene expression.
TGF-␤ exerts its cellular effects by interacting with specific cell surface receptors. Binding of TGF-␤ to T␤RII results in the formation of a heteromeric complex between T␤RI and T␤RII, which in turn results in the phosphorylation of T␤RI by T␤RII. Activation of T␤RI results in propagation of signals through phosphorylation of Smad2 and Smad3. Overexpression of either truncated T␤RI or T␤RII, which lack cytoplasmic kinase domains, results in suppression of TGF-␤ responsiveness in lung fibroblasts (10,47), indicating that both T␤RI and T␤RII are indispensable for TGF-␤ signaling (57,58). T␤RII is essential for binding of TGF-␤ to the receptor complex, and T␤RI is necessary for downstream signal transduction induced by TGF-␤ binding to T␤RII (12). Binding of TGF-␤ to T␤RII is the first critical step in the TGF-␤-signaling cascade because T␤RI does not bind TGF-␤ in the absence of T␤RII (58 -60).
Our results indicate that UV irradiation of cells targets the first step of the TGF-␤/Smad-signaling cascade by causing substantial loss of T␤RII protein. UV irradiation caused simultaneous loss of T␤RII mRNA and protein. T␤RII promoter/luciferase activity is inhibited after UV irradiation, suggesting that down-regulation of T␤RII mRNA results, at least in part, from decreased gene transcription.
The T␤RII gene promoter has been identified and partially characterized (40,61). It contains both positive and negative regulatory elements in addition to a core element required for basal activity. Both receptor constructs that we utilized contain a strongly negative element spanning nucleotides Ϫ100 to Ϫ67. It is possible that this element may mediate UV inhibition of promoter activity, although the cis-acting factors that regulate this element are not known. Recently, Choi et al. (11) identified a novel Ets-related transcription factor that activates transcription of the T␤RII gene by binding to a response element between ϩ13 and ϩ24 and multiple Ets-binding sites in the T␤RII promoter. Whether this Ets-related transcription factor is involved in UV down-regulation of T␤RII gene transcription remains to be determined. Although our data indicate that UV irradiation represses T␤RII gene transcription, they do not rule out the possibility of additional post-transcriptional regulation. Reduced levels of T␤RII protein could result from reduced synthesis attributable to reduced transcription and/or to increased degradation. UV irradiation may cause internalization of T␤RII, which could accelerate degradation of T␤RII protein.
Future studies are needed to clarify the precise molecular mechanism(s) by which UV irradiation reduces T␤RII expression.
In contrast to its effect upon T␤RII, UV irradiation did not alter T␤RI protein levels. However, UV transiently induced T␤RI mRNA, indicating that UV regulates T␤RI and T␤RII differently. Although cDNA sequences of T␤RI and T␤RII share partial homology, there are significant differences in their promoter regions (40,62,63), consistent with our finding that these two genes are differentially regulated by UV irradiation. The 5Ј-flanking region of the T␤RI gene promoter is extremely GC-rich and contains multiple Sp1 sites, which are essential for basal and maximal promoter activity (62,63).
Smad7 antagonizes TGF-␤/Smad signaling by interacting with T␤RI to interfere with phosphorylation and activation of Smad2 and -3. TGF-␤ induces Smad7 expression, which serves as a negative feedback loop to limit the intensity and duration of the TGF-␤ response (21,22). UV transiently induced Smad7 mRNA and elevated Smad7 protein levels for at least 24 h post-UV. There was a significant correlation between Smad7 protein levels and inhibition of TGF-␤-induced gene expression. We estimate that UV induction of Smad7 (3-fold) could account for 30% of the observed inhibition of TGF-␤-induced gene expression.
Smad7 gene promoter activity was increased after UV irradiation, suggesting that up-regulation of Smad7 mRNA was due at least in part to increased gene transcription. Recently, Bitzer et al. (49) reported that pro-inflammatory cytokines such as tumor necrosis factor-␣ and interleukin-1␤ suppress TGF-␤/Smad signaling through up-regulation of Smad7 gene transcription. This induction of Smad7 is mediated by the NF-B/ RelA pathway. UV irradiation activates NF-B and induces proinflammatory cytokines such as interleukin-1␤ and tumor necrosis factor-␣ (64 -71). Interestingly, we have recently observed that interleukin-1␤ and tumor necrosis factor-␣ reduce expression of T␤RII and induce expression of Smad7. 2 These data raise the possibility that UV induction of NF-B and/or NF-B-regulated cytokines may coordinately reduce T␤RII and induce Smad7. We are currently investigating this possibility.
Finally, we utilized Smad7 and T␤RII expression vectors to determine the relative contributions of Smad7 induction and T␤RII down-regulation to UV inhibition of TGF-␤ responsiveness. Our data indicate that UV-induced Smad7 accounts for ϳ30% of the total inhibitory effect of UV on TGF-␤ responsiveness. In contrast, overexpression of T␤RII completely prevented UV inhibition of TGF-␤ responsiveness, indicating that down-regulation of T␤RII plays a major role in UV inhibition of TGF-␤ signaling.
In addition, we examined the possibility that UV induction of c-Jun/AP-1 could contribute to reduced TGF-␤ responsiveness since UV irradiation induces c-Jun/AP-1 and overexpression of c-Jun inhibits Smad3-dependent transcription (51,52). However, we found there was no correlation between the kinetics of UV induction of c-Jun/AP-1 and the kinetics of UV inhibition of TGF-␤ responsiveness. Furthermore, inhibition of TGF-␤-induced gene expression by c-Jun/AP-1 does not involve inhibition of Smad2/3 nuclear translocation (51), whereas we found that UV irradiation blocked Smad2/3 nuclear translocation, presumably due to down-regulation of T␤RII. These results indicate that UV-induced c-Jun/AP-1 is not a major contributing factor to UV-inhibition of TGF-␤ responsiveness in MLEC. MLECs stably expressing a PAI promoter-luciferase reporter gene were transiently transfected with 2 or 4 g of T␤RII expression vector. Twenty-four hours after transfection, cells were irradiated with UV (20 mJ/cm 2 ) and then 8 h later treated with TGF-␤ 1 (1 ng/ml) for 16 h. Cells lysates were normalized to cell numbers, and luciferase activity was determined by luciferase assay. The bar heights are the means Ϯ S.E. of fold change in luciferase activity relative to activity in control (Ctrl) cells. n ϭ 3. *, p Ͻ 0.05 versus mock-transfected, non-treated control cells; †, p Ͻ 0.05 versus mock-transfected, no UV ϩ TGF-␤-treated cells. UV irradiation had no significant effect on luciferase activity in cells over-expressing T␤ RII.
Alternatively, c-Jun/AP-1 may cooperate with Smad2/3 to induce TGF-␤-regulated genes (53,72). If so, then UV activation may make c-Jun/AP-1 unable to function as a co-factor in TGF-␤-induced gene expression. Our data indicate that this scenario is unlikely because return of c-Jun/AP-1 to its basal state 24 h post-UV did not restore TGF-␤ responsiveness.
In summary, UV irradiation blocks cellular responsiveness to TGF-␤ through two mechanisms that impair TGF-␤ receptor function. The primary mechanism is down-regulation of T␤RII, and the secondary mechanism is induction of Smad7.