TGF-β is implicated in the pathogenesis of fibrotic disorders. It has been shown that Smad3 promotes the human α2(I) collagen (COL1A2) gene expression by TGF-β1 in human dermal fibroblasts. Here, we investigated the role of phosphatidylinositol 3-kinase (PI3K) in the COL1A2 gene expression in normal and scleroderma fibroblasts. In normal fibroblasts, the PI3K inhibitor, LY294002, significantly decreased the basal and the TGF-β1-induced increased stability of COL1A2 mRNA. The TGF-β1-induced COL1A2 promoter activity, but not the basal activity, was significantly attenuated by LY294002 or the dominant negative mutant of p85 subunit of PI3K, while the constitutive active mutant of p110 subunit of PI3K did not affect the basal or the TGF-β1-induced COL1A2 promoter activity. LY294002 significantly decreased the phosphorylation of Smad3 induced by TGF-β1. Furthermore, the transient overexpression of 2xFYVE, which induces the mislocalization of FYVE domain proteins, decreased the TGF-β1-induced Smad3 phosphorylation to a similar extent to LY294002. In scleroderma fibroblasts, the blockade of PI3K significantly decreased the mRNA stability and the promoter activity of the COL1A2 gene. Furthermore, LY294002 and the transient overexpression of 2xFYVE completely diminished the constitutive phosphorylation of Smad3. These results indicate that 1) the basal activity of PI3K is necessary for the COL1A2 mRNA stabilization in normal and scleroderma fibroblasts, 2) there is an unidentified FYVE domain protein specifically interacting with Smad3, and 3) the basal activity of PI3K and the FYVE domain protein are indispensable for the efficient TGF-β/Smad3 signaling in normal fibroblasts and for the establishment of the constitutive activation of TGF-β/Smad3 signaling in scleroderma fibroblasts.

Systemic sclerosis or scleroderma is an acquired disorder which typically results in fibrosis of the skin and internal organs (1). Although the pathogenesis of this disorder is still unclear, it includes inflammation, autoimmune attack, and vascular damage leading to fibroblast activation (2, 3). The reason for the presence of abnormal fibroblasts in scleroderma is not yet known, but it is possible that they develop from a subset of cells that have escaped from a normal control mechanism (4, 5).

TGF-β1 is a multifunctional cytokine that regulates the growth, differentiation, and function of immune and nonimmune cells (6). The principal effect of TGF-β1 on mesenchymal cells is its stimulation of ECM deposition. TGF-β1 has been shown to increase expression of collagen types I, III, VI, VII, and X, fibronectin, and proteoglycans (4, 7, 8, 9, 10, 11, 12). Stimulation of ECM deposition by TGF-β1 is further enhanced by its inhibitory effect on matrix degradation, decreasing synthesis of proteases, and increasing levels of protease inhibitors (13). Although the mechanism of dermal fibroblasts activation of scleroderma is presently unknown, many of the characteristics of scleroderma fibroblasts resemble those of healthy fibroblasts stimulated by TGF-β1 (13, 14). Increasing evidence in vivo and in vitro indicate that the dermal fibroblast activation in scleroderma may be a result of stimulation by autocrine TGF-β1. This notion is supported by our following recent findings: 1) scleroderma fibroblasts express the elevated levels of TGF-β receptors, and this correlates with the elevated levels of human α2(I) collagen (COL1A2) mRNA (15) and 2) the blockade of TGF-β signaling with anti-TGF-β Abs or TGF-β1 antisense oligonucleotide abolished the increased expression of COL1A2 mRNA in scleroderma fibroblasts (16).

TGF-β1 initiates signaling through the ligand-dependent activation of a complex of heterodimeric transmembrane serine/threonine kinases, consisting of TGF-β receptor type I (TβR1)3 and TGF-β receptor type II. Upon TGF-β binding, the receptors rotate relatively within the complex, resulting in phosphorylation and activation of TGF-β1 by the constitutively active and autophosphorylated TGF-β receptor type II. TβRI signals from the receptor to the nucleus using a set of Smads proteins. The activated TβRI directly phosphorylates Smad2 and Smad3. Once activated, Smad2 and Smad3 associate with Smad4 and translocate to the nucleus, where such a complex regulates transcriptional responses of target genes together with additional DNA-binding cofactors in a cell-type specific manner (17). Though Smad2 has a 92% amino acid sequence similarity to Smad3 (18), significant functional differences between Smad2 and Smad3 have recently been demonstrated on the transcription of various genes including the COL1A2 gene (19, 20, 21), suggesting that Smad2 and Smad3 may have different subsets of target genes and different functional roles in transcriptional regulation. In dermal fibroblasts, Smad3, but not Smad2, was involved in transducing the signal from TGF-β receptors to the COL1A2 promoter (22).

Regulation of subcellular localization of Smad proteins before activation by TβRI is critical for the effective initiation and maintenance of TGF-β signaling. Recently, a FYVE domain protein designated as a Smad anchor for receptor activation (SARA) has been identified as a Smad2-binding protein (23). SARA directly interacts with Smad2 and Smad3, as well as functions to recruit Smad2 for phosphorylation to the activated TGF-β receptor complex (23). Because SARA is localized in early endosomes through the interaction of its FYVE domain with phosphatidylinositol 3′ phosphate, the inhibition of phosphatidylinositol 3-kinase (PI3K), which reduces cellular levels of phosphatidylinositol 3′ phosphate, results in both a redistribution of SARA from the endosomal compartment to the cytosol and the attenuation of both TGF-β-induced Smad2 phosphorylation and transcriptional activation (24). Thus, the notion has been established that SARA plays a significant role in TGF-β-induced Smad2 activation. Regarding Smad3, however, Goto et al. (25) demonstrated that a mutant Smad3 that lacks the binding to SARA is still phosphorylated by TβRI to a similar extent with wild-type Smad3, forms complexes with Smad4, translocalizes into the nucleus, and enhances TGF-β-induced transcription.

Although it has been shown that the inhibition of PI3K abolishes TGF-β/Smad2-mediated signaling (24), it remains unclear whether the inhibition of PI3K affects TGF-β/Smad3 signaling. In this study, we investigated the modulation of TGF-β/Smad3 signaling and COL1A2 gene expression by the inhibition of PI3K in normal and scleroderma fibroblasts.

Ab for type I collagen was purchased from Southern Biotechnology Associates (Birmingham, AL). Abs specific for Smad2/3 (N-19) and Smad3 (FL-425) were obtained from Santa Cruz Biotechnology (Santa Cruz, CA). Anti-Smad2/3 Ab (S66220) was purchased from BD Transduction Laboratories (Lexington, KY). Anti-phospho-serine Ab was purchased from Biomedia (Foster City, CA). LY294002 was purchased from Calbiochem (La Jolla, CA). Recombinant human TGF-β1 was obtained from R&D Systems (Minneapolis, MN). Actinomycin D and ascorbic acid were purchased from Sigma-Aldrich (St. Louis, MO). FuGENE 6 was obtained from Roche Diagnostic Systems (Indianapolis, IN).

Human dermal fibroblasts were obtained by skin biopsy from the affected areas (dorsal forearm) of 10 patients with diffuse cutaneous systemic sclerosis and <2 years of skin thickening (16). Control fibroblasts were obtained by skin biopsy from 10 healthy donors. Institutional approval and informed consent were obtained from all subjects. Control donors were matched with systemic sclerosis patients for age, sex, and biopsy site, and control and patient samples were processed in parallel. Primary explant cultures were established in 25-cm2 culture flasks in MEM supplemented with 10% FCS, 2 mM l-glutamine, and 50 μg/ml amphotericin. Fibroblast cultures independently isolated from different individuals were maintained as monolayers at 37°C in 95% air, 5% CO2, and studied between the third and sixth subpassages.

Confluent quiescent normal fibroblasts were pretreated for 1 h with the indicated concentration of LY294002 or DMSO, and cultured for an additional 24 h in the presence or absence of TGF-β1 (2 ng/ml). Confluent quiescent scleroderma fibroblasts were treated in parallel with normal fibroblasts without TGF-β1 stimulation. In some experiments to assess the effect of ascorbic acid, confluent quiescent fibroblasts were cultured in the presence or absence of the indicated concentration of ascorbic acid for 24 h. Each conditioned medium was subjected to SDS-PAGE and transferred to nitrocellulose membranes as described previously (26). Membranes were incubated overnight with anti-type I collagen Ab, washed, and incubated for 1 h with secondary Ab. After washing, visualization was performed by ECL (Amersham Pharmacia Biotech, Buckinghamshire, U.K.) according to the manufacturer’s recommendations. The densities of bands were measured with a densitometer. In experiments to compare the expression levels of α1(I) and α2(I) procollagen, gels were stained with Coomassie blue (Sigma-Aldrich) after SDS-PAGE.

Confluent quiescent normal fibroblasts were pretreated for 1 h with the indicated concentration of LY294002 or DMSO, and cultured for an additional 24 h in the presence or absence of TGF-β1 (2 ng/ml). Confluent quiescent scleroderma fibroblasts were treated in parallel with normal fibroblasts without TGF-β1 stimulation. Total RNA was extracted using an acid guanidinium thiocyanate-phenol-chloroform method (27, 28). Two micrograms of total RNA was subjected to electrophoresis on 1% agarose/formaldehyde gels and blotted onto nylon filters (Roche Diagnostic Systems). The filters were UV cross-linked, prehybridized, and sequentially hybridized with DNA probe for COL1A2 and GAPDH as described previously (27, 28). The membrane was then washed and exposed to x-ray film. The x-ray films were scanned as described above.

Generation of a series of 5′-deletions of COL1A2/chloramphenicol acetyltransferase (CAT) construct consisting of the COL1A2 gene fragments (+58 to −353, −264, −186, −148, or −108 bp relative to the transcription start site) linked to the CAT reporter gene was done as previously described (27). Point mutations were introduced into the potential Smad3 recognition site (located between nucleotides −263 and −258) of the −353 COL1A2/CAT construct using the QuickChange site-directed mutagenesis kit (Stratagene, La Jolla, CA) as described previously (29), according to the manufacturer’s instructions with the following modification: 1 cycle of 1 min at 94°C; 30 cycles of 1 min at 94°C, 1 min at 52°C, 7 min at 72°C; and 1 cycle of 15 min at 72°C. The primers used were 5′-AGGGCGGAGGTATGTATACAACGAGTCAGAG-3′ and 5′-CTCTGACTCGTTGTATACATACCTCCGCCCT-3′. The resulting clone was called −353m COL1A2/CAT. Mutation and deletion constructs were verified by sequencing. Expression vector of 2xFYVE (Hrs) is a generous gift from Dr. H. Stenmark (Norwegian Radium Hospital, Oslo, Norway). Plasmids used in transient transfection assays were purified twice on CsCl gradients. At least two different plasmid preparations were used for each experiment.

Cells were grown to 50% confluence in 100-mm dishes in MEM with 10% FCS. The medium was replaced with serum-free medium, and after 4 h, cells were transfected with 2 μg of each 5′-deletions of the COL1A2/CAT construct or site-directed mutation −353m COL1A2/CAT, along with the indicated amount of dominant negative mutant of p85 subunit of PI3K (DN p85), constitutive active mutant of p110 subunit of PI3K (CA p110), 2xFYVE, or corresponding empty vectors, using FuGENE6 as described previously (26). To correct minor variations in transfection efficiency, 1 μg of pSV-β-galactosidase vector (Promega, Madison, WI) was included in all transfections. After 24 h, TGF-β1 (2 ng/ml) was then added, and cells were harvested 48 h after transfection. Extracts, normalized for protein content as measured by the Bio-Rad reagent (Bio-Rad, Hercules, CA), were incubated with butyl-CoA and 14C-chloramphenicol for 90 min at 37°C. Butylated chloramphenicol was extracted using an organic solvent (2:1 mixture of tetramethylpentadecane and xylene) and quantitated by scintillation counting. Each experiment was performed in duplicate.

Confluent quiescent normal and scleroderma fibroblasts were pretreated for 1 h with the indicated concentration of LY294002 or DMSO, and cultured for an additional 1 h in the presence or absence of TGF-β1 (2 ng/ml). The cells were washed with cold PBS and harvested into lysis buffer (1% Nonidet P-40 in 10 mM Tris-HCl (pH 7.5), 150 mM NaCl, 1 mM EDTA, 1 mM Na3VO4, 50 mM NaF, containing 10 μg/ml leupeptin, 10 μg/ml pepstatin, 10 μg/ml aprotinin, and 1 mM PMSF). Each extract was precleared with 10 μl of protein G-Sepharose for 1 h with rotation. The beads were pelleted, the supernatant was transferred to a new tube, and 10 μl of protein G-Sepharose beads conjugated to anti-Smad3 Ab was added. Immunoprecipitation was performed overnight at 4°C with rotation, after which the immunoprecipitates were washed four times with lysis buffer (30). After the last wash, the beads were resuspended in 30 μl of sample buffer and boiled for 5 min. Proteins were subjected to immunoblotting using anti-phospho-serine Ab. After the development, the membrane was stripped and reprobed with anti-Smad2/3 Ab (N-19) to determine the total levels of Smad3.

Two oligonucleotides containing biotin on the 5′-nucleotide of the sense strand were used (31). The sequences of these oligonucleotides are as follows: 1) 3xCAGA oligo, 5′-TCGAGAGCCAGACAAGGAGCCAGACAAGGAGCCAGACACTCGAG, which is trimer of CAGA motif; and 2) 3xCAGA-M oligo, 5′-TCGAGAGCTACATAAAAAGCTACATATTTAGCTACATACTCGA, which is trimer of CAGA motif mutated. These oligonucleotides were annealed to their respective complementary oligonucleotides, and double-stranded oligonucleotides were gel-purified and used. Cell lysate was prepared using lysis buffer containing 10 mM Tris-HCl (pH 7.4), 150 mM NaCl, 1 mM EDTA, 1% Nonidet P-40, 50 mM NaF, 1 mM PMSF, 1 mM Na3VO4, 1 μg/ml leupeptin, 1 μg/ml aprotinin, and 1 μg/ml pepstatin. Five micrograms of poly(dI-dC) competitor was incubated with 500 μg of cell lysate for 30 min at 4°C, followed by 1 h incubation with 500 pmol of each double-stranded oligonucleotide. After the incubation, 65 μl of streptavidin-agarose (Sigma-Aldrich) was added to the reaction and incubated at 4°C for overnight. The protein-DNA-streptavidin-agarose complex was washed three times with lysis buffer, resuspended in the sample buffer for electrophoresis, boiled for 3 min, spun briefly, and the supernatants were subjected to Western blotting with anti-Smad2/3 Ab (S66220).

Northern blots that investigate the stability of COL1A2 mRNA were analyzed by plotting the log of the densitometer volume for each band (with background from a blank lane on the gel subtracted) vs time and by obtaining a slope m for the plot by regression analysis. Turnover was assumed to be a first-order process, with t1/2 = 0.693/k, where k = −2.3 m. All half-life measurements are from at least five independent experiments yielding correlation coefficients >0.9.

Statistical analysis was conducted with the Mann-Whitney U test for comparison of means. Values of p < 0.05 were considered significant.

As an initial experiment, we investigated the effect of LY294002, a specific inhibitor of PI3K, on the expression levels of type I procollagen protein in normal fibroblasts either treated or untreated with TGF-β1 by immunoblotting. As shown in Fig. 1, A and B, the basal and the TGF-β1-dependent increased levels of type I procollagen protein were reduced by the treatment of LY294002 in a dose-dependent manner. Cell viability was determined with trypan blue stain, which demonstrated that the indicated concentration of LY294002 did not cause cell death. As shown in Fig. 1,A, the ratio of the α1 to α2 chain is ∼0.25:1 or 1:1 rather than the expected 2:1, suggesting that the α2 chain forms normal α1/α2 heterotrimers and α2/α2 homotrimers. One possible explanation against this highly unusual observation is that lack of ascorbic acid in conditioned medium, which is required for hydroxylation of prolines that are critical for the formation of thermostable triple helix of procollagen, leads to change in structure of procollagen and alters secretion of procollagen from cells (32, 33). To clarify this point, cells were cultured in conditioned medium with or without ascorbic acid and the levels of type I procollagen protein in each medium were determined by immunoblotting. As shown in Fig. 1,C, consistent with previous reports (34, 35), the levels of type I procollagen protein were elevated in the presence of ascorbic acid. However, ascorbic acid did not affect the ratio of the α1 to α2 chain. These results rule out the hypothesis described above and suggest that altered ratio of the α1 to α2 chain in this immunoblotting assay is attributed to the difference in the immunoreactivity of anti-type I collagen Ab to the α1 and α2 chain. This point was confirmed by the Coomassie blue staining, which demonstrated that the ratio of the α1 to α2 chain was ∼2:1 (Fig. 1 D).

FIGURE 1.

Effects of LY294002 on the basal and the TGF-β1-induced levels of type I procollagen in normal fibroblasts. A and E, Confluent quiescent cells were pretreated for 1 h with the indicated concentration of LY294002 or DMSO and treated for additional 24 h with the indicated concentration of TGF-β1. Each conditioned medium was analyzed by immunoblotting using anti-type I collagen Ab (A). Total RNA was isolated and resolved electrophoretically, and Northern blotting analyses were performed with DNA probes for COL1A2 and GAPDH (E). B and F, Quantitative analysis of type I procollagen protein levels (B) and COL1A2 mRNA levels (F). Values represent the band density relative to the mean value of untreated normal fibroblasts (100). ∗, p < 0.05 vs control cells treated with vehicle (DMSO) under the same conditions. C, Confluent quiescent cells were cultured in serum-free medium with the indicated concentration of ascorbic acid for 24 h. Each conditioned medium was analyzed by immunoblotting using anti-type I collagen Ab. D, Comparison of the expression levels of α1(I) and α2(I) procollagen. Conditioned medium derived from untreated quiescent fibroblasts were subjected to SDS-PAGE, and gels were stained with Coomassie blue. The result of Coomassie blue staining was compared with that of immunoblotting using anti-type I collagen Ab. M.W., molecular weight marker. All bands show one representative of five independent experiments.

FIGURE 1.

Effects of LY294002 on the basal and the TGF-β1-induced levels of type I procollagen in normal fibroblasts. A and E, Confluent quiescent cells were pretreated for 1 h with the indicated concentration of LY294002 or DMSO and treated for additional 24 h with the indicated concentration of TGF-β1. Each conditioned medium was analyzed by immunoblotting using anti-type I collagen Ab (A). Total RNA was isolated and resolved electrophoretically, and Northern blotting analyses were performed with DNA probes for COL1A2 and GAPDH (E). B and F, Quantitative analysis of type I procollagen protein levels (B) and COL1A2 mRNA levels (F). Values represent the band density relative to the mean value of untreated normal fibroblasts (100). ∗, p < 0.05 vs control cells treated with vehicle (DMSO) under the same conditions. C, Confluent quiescent cells were cultured in serum-free medium with the indicated concentration of ascorbic acid for 24 h. Each conditioned medium was analyzed by immunoblotting using anti-type I collagen Ab. D, Comparison of the expression levels of α1(I) and α2(I) procollagen. Conditioned medium derived from untreated quiescent fibroblasts were subjected to SDS-PAGE, and gels were stained with Coomassie blue. The result of Coomassie blue staining was compared with that of immunoblotting using anti-type I collagen Ab. M.W., molecular weight marker. All bands show one representative of five independent experiments.

Close modal

To determine whether the effect of LY294002 on type I procollagen protein was paralleled with the corresponding mRNA levels, we investigated the effect of LY294002 on the basal and the TGF-β1-induced expression levels of COL1A2 mRNA by Northern blotting. As shown in Fig. 1, E and F, the basal and the TGF-β1-induced expression levels of COL1A2 mRNA were reduced by the treatment of LY294002 in a dose-dependent manner.

The steady-state level of mRNA can be affected by the level of gene transcription and/or the stability of mRNA. To determine whether the effects of LY294002 on the basal and the TGF-β1-induced expression levels of COL1A2 mRNA take place at the transcriptional level or the posttranscriptional level, the transcription was blocked with actinomycin D. To determine the contribution of PI3K activity to the stability of COL1A2 mRNA, cells were treated with the indicated concentration of LY294002 for 1 h before the addition of actinomycin D. As shown in Fig. 2, the treatment of LY294002 reduced the basal stability of COL1A2 mRNA in a dose-dependent manner. Though the half-life of COL1A2 mRNA was significantly increased after the treatment of TGF-β1 (6.20 ± 0.55 vs 2.87 ± 0.32, p < 0.05), the pretreatment of LY294002 also reduced the TGF-β1-dependent increased stability of COL1A2 mRNA in a dose-dependent manner. To note, there was no significant difference in the half-life of COL1A2 mRNA between normal fibroblasts treated with 30 μM LY294002 in the presence or absence of TGF-β1 (2.53 ± 0.25 vs 2.33 ± 0.23, respectively), indicating that LY294002 reduces the TGF-β1-dependent increased stability as well as the basal stability of COL1A2 mRNA.

FIGURE 2.

Effects of LY294002 on the stability of COL1A2 mRNA in normal fibroblasts either treated or untreated with TGF-β1. Confluent quiescent cells were pretreated for 1 h with the indicated concentration of LY294002 or DMSO, and treated with actinomycin D (400 ng/ml) and TGF-β1 (2 ng/ml) for the indicated time periods before RNA extraction. The levels of COL1A2 mRNA were determined by Northern blotting and normalized to GAPDH. The percentage of remaining COL1A2 mRNA relative to time 0 was plotted on a logarithmic scale. The mean from five separate experiments are shown.

FIGURE 2.

Effects of LY294002 on the stability of COL1A2 mRNA in normal fibroblasts either treated or untreated with TGF-β1. Confluent quiescent cells were pretreated for 1 h with the indicated concentration of LY294002 or DMSO, and treated with actinomycin D (400 ng/ml) and TGF-β1 (2 ng/ml) for the indicated time periods before RNA extraction. The levels of COL1A2 mRNA were determined by Northern blotting and normalized to GAPDH. The percentage of remaining COL1A2 mRNA relative to time 0 was plotted on a logarithmic scale. The mean from five separate experiments are shown.

Close modal

To determine whether LY294002 affects the basal and the TGF-β1-induced increased COL1A2 promoter activity, we performed transient transfection assays using −353 COL1A2/CAT construct. As shown in Fig. 3,A, LY294002 had no significant inhibitory effect on the basal COL1A2 promoter activity. In contrast, LY294002 reduced the TGF-β1-induced COL1A2 promoter activity in a dose-dependent manner. Upon the treatment of 30 μM LY294002, the TGF-β1-induced increased COL1A2 promoter activity was significantly reduced (3.7 ± 0.2 vs 2.3 ± 0.1, p < 0.05). To further confirm the effect of PI3K inhibition on the COL1A2 promoter activity, transient transfection assays using DN p85 were performed. As shown in Fig. 3,B, cotransfection of DN p85 showed no significant effect on the basal COL1A2 promoter activity. Consistent with the results of experiments using LY294002, cotransfection of DN p85 reduced the TGF-β1-induced increased COL1A2 promoter activity in a dose-dependent manner, and the significant reduction was observed by cotransfection of 2 μg of DN p85 (3.6 ± 0.2 vs 2.2 ± 0.1, p < 0.05). To further clarify the role of PI3K in the TGF-β1-induced COL1A2 gene expression, we investigated the effect of CA p110 on the COL1A2 promoter activity (Fig. 3 C). Cotransfection of CA p110 showed no significant effects on the basal and the TGF-β1-induced increased COL1A2 promoter activity. We also confirmed that the TGF-β1 stimulation did not affect the basal phosphorylation level of p85 subunit (data not shown). These results suggest that the basal activity of PI3K is indispensable and sufficient to maintain the efficient TGF-β1-dependent induction of the COL1A2 gene.

FIGURE 3.

The requirement of the basal PI3K activity for the TGF-β1-induced COL1A2 promoter activity in normal fibroblasts. A, Cells were transfected with 2 μg of the −353 COL1A2/CAT construct and incubated for 48 h. After the pretreatment of LY294002 or DMSO for 1 h, cells were stimulated with the indicated concentration of TGF-β1 for the last 24 h. Values represent the CAT activities relative to those of untreated cells (100). The mean and SEM from five separate experiments are shown. ∗, p < 0.05 vs control cells treated with DMSO under the same conditions. B and C, Cells were transfected with 2 μg of the −353 COL1A2/CAT construct, along with the indicated amount of DN p85 (B), CA p110 (C), or corresponding empty vectors, and incubated for 48 h. In some experiments, cells were treated with TGF-β1 (2 ng/ml) for the last 24 h. Values represent the CAT activities relative to those of untreated cells transfected with empty vector (100). The mean and SEM from five separate experiments are shown. ∗, p < 0.05 vs control cells transfected with empty vector under the same conditions.

FIGURE 3.

The requirement of the basal PI3K activity for the TGF-β1-induced COL1A2 promoter activity in normal fibroblasts. A, Cells were transfected with 2 μg of the −353 COL1A2/CAT construct and incubated for 48 h. After the pretreatment of LY294002 or DMSO for 1 h, cells were stimulated with the indicated concentration of TGF-β1 for the last 24 h. Values represent the CAT activities relative to those of untreated cells (100). The mean and SEM from five separate experiments are shown. ∗, p < 0.05 vs control cells treated with DMSO under the same conditions. B and C, Cells were transfected with 2 μg of the −353 COL1A2/CAT construct, along with the indicated amount of DN p85 (B), CA p110 (C), or corresponding empty vectors, and incubated for 48 h. In some experiments, cells were treated with TGF-β1 (2 ng/ml) for the last 24 h. Values represent the CAT activities relative to those of untreated cells transfected with empty vector (100). The mean and SEM from five separate experiments are shown. ∗, p < 0.05 vs control cells transfected with empty vector under the same conditions.

Close modal

To identify potential regulatory elements of the COL1A2 gene by PI3K, we performed transient transfection assays using a series of 5′-deletions of the COL1A2/CAT construct. As shown in Fig. 4, in agreement with the previous studies (27, 36), progressive deletions of the COL1A2 promoter decreased basal promoter activity in normal fibroblasts. In the cells treated with TGF-β1, the bp −353∼+58 construct responded at the highest level. The TGF-β1-induced increased promoter activity was significantly decreased by the removal of a triple Sp1 binding site (−264-bp deletion construct) or a CAGA motif (−186-bp deletion construct), and was completely abrogated with the removal of another Sp1 binding site located at −125 bp (−108-bp deletion construct). The TGF-β1-induced increased promoter activity of the bp −353∼+58 construct significantly reduced by cotransfection of DN p85 at the highest level (3.7 ± 0.2-fold increase vs 2.2 ± 0.1-fold increase, p < 0.05), and the subsequent deletion, the bp −264∼+58 construct, showed a similar result (2.5 ± 0.2-fold increase vs 1.7 ± 0.1-fold increase, p < 0.05). However, the inhibitory effect of DN p85 was completely diminished in the bp −186∼+58 construct. These data indicate that the responsive element of the COL1A2 promoter to DN p85 is located between bp −264 and −186. This region contains a CAGACA sequence (from −263 bp to −258 bp) that was shown to be a functional Smad3-binding element (22). To further characterize the regulatory element of DN p85 in the COL1A2 promoter, we used the site-directed mutated construct −353m COL1A2/CAT, in which the Smad3 binding site is mutated. The inhibitory effect of DN p85 was completely abolished in the −353m COL1A2/CAT construct. Taken together, the experiments with deletion and substitution promoter mutants suggest that Smad3 is involved in the PI3K-dependent modulation of COL1A2 promoter activity.

FIGURE 4.

Functional analysis of the inhibitory effect of dominant negative mutant of p85 subunit on the TGF-β1-induced increased COL1A2 promoter activity in normal fibroblasts. Cells were transfected with 2 μg of the indicated 5′-deletion of the COL1A2/CAT construct or a site-directed mutated construct −353m COL1A2/CAT, along with 2 μg of DN p85 or corresponding empty vector, and incubated for 48 h. In some experiments, cells were treated with TGF-β1 (2 ng/ml) for the last 24 h. Values represent the CAT activities relative to those of untreated cells transfected with empty vector (100). The mean and SEM from five separate experiments are shown. ∗, p < 0.05 vs control cells transfected with empty vector and untreated with TGF-β1. ∗∗, p < 0.05 vs cells transfected with empty vector and stimulated with TGF-β1.

FIGURE 4.

Functional analysis of the inhibitory effect of dominant negative mutant of p85 subunit on the TGF-β1-induced increased COL1A2 promoter activity in normal fibroblasts. Cells were transfected with 2 μg of the indicated 5′-deletion of the COL1A2/CAT construct or a site-directed mutated construct −353m COL1A2/CAT, along with 2 μg of DN p85 or corresponding empty vector, and incubated for 48 h. In some experiments, cells were treated with TGF-β1 (2 ng/ml) for the last 24 h. Values represent the CAT activities relative to those of untreated cells transfected with empty vector (100). The mean and SEM from five separate experiments are shown. ∗, p < 0.05 vs control cells transfected with empty vector and untreated with TGF-β1. ∗∗, p < 0.05 vs cells transfected with empty vector and stimulated with TGF-β1.

Close modal

Next, we investigated the TGF-β1-induced phosphorylation state of Smad3 in the presence or absence of LY294002. As shown in Fig. 5,A, TGF-β1 induced the marked phosphorylation of Smad3. However, the pretreatment of 30 μM LY294002 reduced the TGF-β1-induced phosphorylation of Smad3 ∼50%. To further confirm this finding, we investigated the DNA-binding ability of Smad3 by DNA affinity precipitation. As shown in Fig. 5,B, TGF-β1 induced the strong binding of Smad3 with the 3xCAGA oligonucleotide, whereas the 3xCAGA-M oligonucleotide, which lacks the CAGA motif, did not bind with Smad3 even after TGF-β1 stimulation. Consistent with the results of immunoprecipitation, the pretreatment of 30 μM LY294002 decreased the levels of the DNA-Smad3 binding ∼50%. We also confirmed that the treatment of LY294002 did not affect the expression level of Smad3 (Fig. 5 C). These results suggest that the inhibition of PI3K attenuates the TGF-β1-induced increased COL1A2 promoter activity by reducing the phosphorylation level and the DNA-binding ability of Smad3.

FIGURE 5.

Effects of LY294002 on the TGF-β1-dependent increase in the Smad3 phosphorylation and the DNA-Smad3 binding in normal fibroblasts. A and B, Confluent quiescent cells were pretreated with the indicated concentration of LY294002 or DMSO for 1 h, and stimulated with the indicated concentration of TGF-β1 for 1 h. Whole cell lysates were subjected to immunoprecipitation using anti-Smad3 Ab, and phospho-Smad3 were detected by immunoblotting using anti-phospho-serine Ab. The same membrane was stripped and reprobed with anti-Smad2/3 Ab (A). Whole cell lysates were incubated with biotin-labeled oligonucleotides. Proteins bound to these nucleotides were isolated with streptavidin-agarose beads, and Smad3 was detected by immunoblotting (B). C, Confluent quiescent cells were treated with the indicated concentration of LY294002 for the indicated time periods. The levels of Smad3 were determined by immunoblotting. All bands show one representative of five independent experiments.

FIGURE 5.

Effects of LY294002 on the TGF-β1-dependent increase in the Smad3 phosphorylation and the DNA-Smad3 binding in normal fibroblasts. A and B, Confluent quiescent cells were pretreated with the indicated concentration of LY294002 or DMSO for 1 h, and stimulated with the indicated concentration of TGF-β1 for 1 h. Whole cell lysates were subjected to immunoprecipitation using anti-Smad3 Ab, and phospho-Smad3 were detected by immunoblotting using anti-phospho-serine Ab. The same membrane was stripped and reprobed with anti-Smad2/3 Ab (A). Whole cell lysates were incubated with biotin-labeled oligonucleotides. Proteins bound to these nucleotides were isolated with streptavidin-agarose beads, and Smad3 was detected by immunoblotting (B). C, Confluent quiescent cells were treated with the indicated concentration of LY294002 for the indicated time periods. The levels of Smad3 were determined by immunoblotting. All bands show one representative of five independent experiments.

Close modal

Previous reports indicated that the inhibitory effect of PI3K inhibitors on TGF-β-induced Smad2 activation was induced by the mislocalization of SARA, a FYVE domain protein interacting with Smad2 (24). Taken together with the present results, we next investigated whether FYVE domain proteins were involved in the inhibitory effect of LY294002 on the TGF-β1-dependent Smad3 activation. To this end, we performed transient transfection assays using the expression vector of 2xFYVE, which induces the mislocalization of FYVE domain proteins. As shown in Fig. 6,A, the cotransfection of 2xFYVE showed ∼50% decrease of the TGF-β1-induced COL1A2 promoter activity, whereas there was no significant inhibitory effect on the basal COL1A2 promoter activity by the coexpression of 2xFYVE. This inhibitory effect of 2xFYVE was completely diminished in the site-directed mutated construct −353m COL1A2/CAT. To further confirm this finding, we investigated the effect of 2xFYVE on Smad3 phosphorylation. As shown in Fig. 6 B, the cotransfection of 2xFYVE reduced the TGF-β1-dependent Smad3 phosphorylation ∼50%. These results suggest that the mislocalization of the FYVE proteins contributes to the effect of PI3K inhibition on the TGF-β1-dependent Smad3 phosphorylation.

FIGURE 6.

Effects of transient overexpression of 2xFYVE on the TGF-β1-dependent increased COL1A2 promoter activity and Smad3 phosphorylation in normal fibroblasts. A, Cells were transfected with 2 μg of the −353 COL1A2/CAT construct or the −353m COL1A2/CAT construct, along with the indicated amount of the 2xFYVE construct and/or corresponding empty vector, and incubated for 48 h. In some experiments, cells were treated with TGF-β1 (2 ng/ml) for the last 24 h. Values represent the CAT activities relative to those of untreated cells transfected with −353 COL1A2/CAT construct and empty vector (100). The mean and SEM from five separate experiments are shown. ∗, p < 0.05 vs control cells transfected with empty vector under the same conditions. B, Cells were transfected with 2 μg of the 2xFYVE construct or corresponding empty vector and incubated for 48 h. Cells were treated with the indicated concentration of TGF-β1 for the last 1 h. Whole cell lysates were subjected to immunoprecipitation using anti-Smad3 Ab, and phospho-Smad3 were detected by immunoblotting using anti-phospho-serine Ab. The same membrane was stripped and reprobed with anti-Smad2/3 Ab. One representative of five independent experiments is shown.

FIGURE 6.

Effects of transient overexpression of 2xFYVE on the TGF-β1-dependent increased COL1A2 promoter activity and Smad3 phosphorylation in normal fibroblasts. A, Cells were transfected with 2 μg of the −353 COL1A2/CAT construct or the −353m COL1A2/CAT construct, along with the indicated amount of the 2xFYVE construct and/or corresponding empty vector, and incubated for 48 h. In some experiments, cells were treated with TGF-β1 (2 ng/ml) for the last 24 h. Values represent the CAT activities relative to those of untreated cells transfected with −353 COL1A2/CAT construct and empty vector (100). The mean and SEM from five separate experiments are shown. ∗, p < 0.05 vs control cells transfected with empty vector under the same conditions. B, Cells were transfected with 2 μg of the 2xFYVE construct or corresponding empty vector and incubated for 48 h. Cells were treated with the indicated concentration of TGF-β1 for the last 1 h. Whole cell lysates were subjected to immunoprecipitation using anti-Smad3 Ab, and phospho-Smad3 were detected by immunoblotting using anti-phospho-serine Ab. The same membrane was stripped and reprobed with anti-Smad2/3 Ab. One representative of five independent experiments is shown.

Close modal

In previous reports, we demonstrated that the activation of scleroderma fibroblasts might be a result of the stimulation by autocrine TGF-β. Taken together, the results in the present study suggest that the inhibition of PI3K reduces the up-regulated expression of the COL1A2 gene through the transcriptional and the posttranscriptional regulation in scleroderma fibroblasts. To confirm this point, we investigated the effect of LY294002 or DN p85 on COL1A2 gene expression using scleroderma fibroblasts. As shown in Fig. 7, A and B, LY294002 reduced the increased expression levels of type I procollagen protein in scleroderma fibroblasts in a dose-dependent manner. This inhibitory effect of LY294002 on type I procollagen protein was paralleled with the levels of COL1A2 mRNA (Fig. 7, C and D). The stability of COL1A2 mRNA in scleroderma fibroblasts was significantly elevated compared with that in normal fibroblasts, but was also reduced by LY294002 in a dose-dependent manner (Fig. 7,E). Consistent with the results of normal fibroblasts treated with TGF-β1, there was no significant difference in the half life of COL1A2 mRNA between normal and scleroderma fibroblasts treated with 30 μM LY294002 (2.33 ± 0.23 vs 2.40 ± 0.25). Furthermore, LY294002 or DN p85 significantly reduced −353 COL1A2/CAT promoter activity in scleroderma fibroblasts (Fig. 7, F and G).

To assess the effect of LY294002 on Smad3 in scleroderma fibroblasts, we investigated the difference in the phosphorylation levels and the DNA-binding ability of Smad3 between normal and scleroderma fibroblasts. As shown in Fig. 8, the constitutive Smad3 phosphorylation and the constitutive DNA-Smad3 binding were detected only in scleroderma fibroblasts. However, there was no significant difference in the Smad3 phosphorylation levels and the DNA-Smad3-binding levels between normal and scleroderma fibroblasts stimulated with TGF-β1. These results indicate that Smad3 is constitutively phosphorylated in scleroderma fibroblasts, but the phosphorylation levels of Smad3 in scleroderma fibroblasts are weaker compared with those in normal fibroblasts treated with TGF-β1.

FIGURE 8.

Comparison of the phosphorylation levels and the DNA binding ability of Smad3 in normal and scleroderma fibroblasts. A and C, Confluent quiescent cells were treated with the indicated concentration of TGF-β1 for 1 h. Whole cell lysates were subjected to immunoprecipitation using anti-Smad3 Ab, and phospho-Smad3 were detected by immunoblotting using anti-phospho-serine Ab. The same membrane was stripped and reprobed with anti-Smad2/3 Ab (A). Whole cell lysates were incubated with biotin-labeled oligonucleotides. Proteins bound to these nucleotides were isolated with streptavidin-agarose beads, and Smad3 was detected by immunoblotting (C). All bands show one representative of five independent experiments. B and D, Quantitative analysis of phospho-Smad3 levels (B) and DNA-binding Smad3 levels (D). Values represent the band density relative to mean value of normal fibroblasts stimulated with TGF-β1 (2 ng/ml), which was set at 100. ∗, p < 0.05 vs untreated normal fibroblasts. ∗∗, p < 0.05 vs untreated scleroderma fibroblasts.

FIGURE 8.

Comparison of the phosphorylation levels and the DNA binding ability of Smad3 in normal and scleroderma fibroblasts. A and C, Confluent quiescent cells were treated with the indicated concentration of TGF-β1 for 1 h. Whole cell lysates were subjected to immunoprecipitation using anti-Smad3 Ab, and phospho-Smad3 were detected by immunoblotting using anti-phospho-serine Ab. The same membrane was stripped and reprobed with anti-Smad2/3 Ab (A). Whole cell lysates were incubated with biotin-labeled oligonucleotides. Proteins bound to these nucleotides were isolated with streptavidin-agarose beads, and Smad3 was detected by immunoblotting (C). All bands show one representative of five independent experiments. B and D, Quantitative analysis of phospho-Smad3 levels (B) and DNA-binding Smad3 levels (D). Values represent the band density relative to mean value of normal fibroblasts stimulated with TGF-β1 (2 ng/ml), which was set at 100. ∗, p < 0.05 vs untreated normal fibroblasts. ∗∗, p < 0.05 vs untreated scleroderma fibroblasts.

Close modal

We next investigated the effect of LY294002 on the phosphorylation and the DNA-binding ability of Smad3 in scleroderma fibroblasts. As shown in Fig. 9, 30 μM LY294002 completely abolished the Smad3 phosphorylation as well as the DNA-Smad3 binding in scleroderma fibroblasts. These results are consistent with the results of normal fibroblasts, suggesting that the basal activity of PI3K is also indispensable to maintain the efficient TGF-β/Smad3 signaling in scleroderma fibroblasts.

FIGURE 9.

Effects of LY294002 on the phosphorylation and the DNA-binding ability of Smad3 in scleroderma fibroblasts. Confluent quiescent cells were treated with the indicated concentration of LY294002 or DMSO for 1 h. A, Whole cell lysates were subjected to immunoprecipitation using anti-Smad3 Ab, and phospho-Smad3 were detected by immunoblotting using anti-phospho-serine Ab. The same membrane was stripped and reprobed with anti-Smad2/3 Ab. B, Whole cell lysates were incubated with biotin-labeled oligonucleotides. Proteins bound to these nucleotides were isolated with streptavidin-agarose beads, and Smad3 was detected by immunoblotting. All bands show one representative of five independent experiments.

FIGURE 9.

Effects of LY294002 on the phosphorylation and the DNA-binding ability of Smad3 in scleroderma fibroblasts. Confluent quiescent cells were treated with the indicated concentration of LY294002 or DMSO for 1 h. A, Whole cell lysates were subjected to immunoprecipitation using anti-Smad3 Ab, and phospho-Smad3 were detected by immunoblotting using anti-phospho-serine Ab. The same membrane was stripped and reprobed with anti-Smad2/3 Ab. B, Whole cell lysates were incubated with biotin-labeled oligonucleotides. Proteins bound to these nucleotides were isolated with streptavidin-agarose beads, and Smad3 was detected by immunoblotting. All bands show one representative of five independent experiments.

Close modal

Finally, we investigated the effect of 2xFYVE on TGF-β/Smad3 signaling in scleroderma fibroblasts. As shown in Fig. 10, 2xFYVE significantly reduced the COL1A2 promoter activity in a dose-dependent manner and completely abolished the constitutive phosphorylation of Smad3 in scleroderma fibroblasts. These results suggest that the FYVE domain protein is required to maintain the constitutive phosphorylation of Smad3 in scleroderma fibroblasts.

FIGURE 10.

Effects of the 2xFYVE on the COL1A2 promoter activity and the Smad3 phosphorylation levels in scleroderma fibroblasts. A, Scleroderma fibroblasts were transfected with 2 μg of the −353 COL1A2/CAT construct, along with the indicated amount of the 2xFYVE construct and/or corresponding empty vector. After 48 h, CAT activities were determined. Values represent the CAT activities relative to those of normal fibroblasts transfected with empty vector (100). The mean and SEM from five separate experiments are shown. ∗, p < 0.05 vs control cells transfected with empty vector in each cell group. B, Scleroderma fibroblasts were transfected with 2 μg of 2xFYVE or corresponding empty vector and incubated for 48 h. Whole cell lysates were subjected to immunoprecipitation using anti-Smad3 Ab, and phospho-Smad3 were detected by immunoblotting using anti-phospho-serine Ab. The same membrane was stripped and reprobed with anti-Smad2/3 Ab. One representative of five independent experiments is shown.

FIGURE 10.

Effects of the 2xFYVE on the COL1A2 promoter activity and the Smad3 phosphorylation levels in scleroderma fibroblasts. A, Scleroderma fibroblasts were transfected with 2 μg of the −353 COL1A2/CAT construct, along with the indicated amount of the 2xFYVE construct and/or corresponding empty vector. After 48 h, CAT activities were determined. Values represent the CAT activities relative to those of normal fibroblasts transfected with empty vector (100). The mean and SEM from five separate experiments are shown. ∗, p < 0.05 vs control cells transfected with empty vector in each cell group. B, Scleroderma fibroblasts were transfected with 2 μg of 2xFYVE or corresponding empty vector and incubated for 48 h. Whole cell lysates were subjected to immunoprecipitation using anti-Smad3 Ab, and phospho-Smad3 were detected by immunoblotting using anti-phospho-serine Ab. The same membrane was stripped and reprobed with anti-Smad2/3 Ab. One representative of five independent experiments is shown.

Close modal

This study was undertaken to clarify the involvement of PI3K in the regulation of COL1A2 gene expression. We found that the inhibition of PI3K decreased the basal expression level of COL1A2 gene by reducing the mRNA stability in normal fibroblasts. We also found that the inhibition of PI3K attenuated the TGF-β1-induced increased expression of the COL1A2 gene in normal fibroblasts by reducing both the mRNA stability and the transcriptional activity, which was achieved by blocking the phosphorylation of Smad3. Moreover, we demonstrated that the inhibition of PI3K also decreased the expression levels of the COL1A2 gene in scleroderma fibroblasts by reducing both mRNA stability and the transcriptional activity. To note, the inhibition of PI3K completely diminished the constitutive phosphorylation of Smad3 in scleroderma fibroblasts. To our knowledge, this is the first report that indicates the involvement of PI3K in COL1A2 gene expression in normal and scleroderma fibroblasts.

The importance of PI3K in the stabilization of α1(I) collagen mRNA was previously demonstrated in human embryonic lung fibroblasts (37). This previous study demonstrated that the inhibition of PI3K by LY294002 reduced α1(I) collagen mRNA stability without any effects on the basal transcriptional activity. These results were consistent with our present results of the COL1A2 gene in human dermal fibroblasts. We also demonstrated that the TGF-β1-dependent increased stability of COL1A2 mRNA was reduced by the inhibition of PI3K. There was no significant difference in the half-life of COL1A2 mRNA between normal dermal fibroblasts treated with 30 μM LY294002 in the presence or absence of TGF-β1, suggesting that LY294002 reduced the TGF-β1-dependent increased mRNA stability as well as the basal mRNA stability. These results indicate that the steady-state and the TGF-β1-induced levels of COL1A2 mRNA are maintained by an LY294002-sensitive pathway. mRNA stabilization of extracellular matrix proteins contributes to the pathogenesis of fibrotic disorders, including scleroderma. The previous reports indicated that the increased mRNA stability, as well as the increased transcriptional activity, contributed to maintain the increased α1(I) collagen and fibronectin production in scleroderma fibroblasts (38), which is paralleled with our present data regarding the COL1A2 gene. The steady state of COL1A2 mRNA in scleroderma fibroblasts was also maintained by an LY294002-sensitive pathway. These results indicate that the basal activity of PI3K is indispensable to maintain the COL1A2 gene expression in normal and scleroderma fibroblasts.

Regulation of subcellular localization of Smad proteins is critical for the effective initiation and maintenance of TGF-β signaling. This notion is strongly supported by the evidence that the mislocalization of SARA induced by 1) deletion of the FYVE domain, 2) the treatment of cells with the inhibitor of PI3K, or 3) the ectopic expression of FYVE domain, attenuates TGF-β/Smad2-mediated signaling (24). SARA was originally identified as a Smad2-binding protein and subsequently the interaction between SARA and Smad3 was also demonstrated by immunoprecipitation (23). However, a mutant Smad3 lacking the binding site to SARA is phosphorylated by TβRI at the similar levels to wild-type Smad3, suggesting that SARA/Smad3 interaction is not essential for TGF-β/Smad3-mediated signaling (25). In this study, we demonstrated that the inhibition of PI3K activity attenuated the TGF-β-induced Smad3 phosphorylation and transcriptional activation of the COL1A2 gene. Furthermore, the overexpression of 2xFYVE, which can disturb the proper localization of FYVE domain proteins, attenuates the TGF-β-induced Smad3 phosphorylation and transcriptional activation of the COL1A2 gene. Taken together, these results suggest that unidentified FYVE domain proteins specifically interacting with Smad3 contribute to the efficient initiation and maintenance of TGF-β/Smad3 signaling. BMP is another member of the TGF-β superfamily, which uses Smad1 and Smad5 as intracellular second messengers (17). Though the anchor protein for Smad1/5 has not been identified, overexpression of the FYVE domain also attenuates the BMP/Smad-mediated pathway (24). Thus, FYVE domain proteins may be important in the efficient signal transduction of the TGF-β superfamily.

Our previous reports have indicated that TGF-β plays a central role in the pathogenesis of scleroderma (15, 16, 30, 39, 40). This notion is strongly supported by the present findings that Smad3 is constitutively phosphorylated and the DNA-Smad3 binding levels were elevated in scleroderma fibroblasts because Smad3, but not Smad2, is involved in the TGF-β1-dependent COL1A2 gene expression (22). Consistent with the results in normal fibroblasts stimulated with TGF-β1, the inhibition of PI3K also attenuates the phosphorylation levels of Smad3 and the COL1A2 gene transcription in scleroderma fibroblasts. However, the inhibitory effect of LY294002 on Smad3 phosphorylation levels was different between normal and scleroderma fibroblasts. In normal fibroblasts, the addition of 30 μM LY294002 or the transient overexpression of 2xFYVE partially blocked the strong phosphorylation of Smad3 induced by TGF-β1 stimulation, while the same treatment completely diminished the constitutive weak phosphorylation of Smad3 in scleroderma fibroblasts. This is consistent with the notion that FYVE domain proteins contribute to the efficient TGF-β/Smad signaling and suggests that the constitutive weak phosphorylation of Smad3 in scleroderma fibroblasts is established only in the presence of FYVE domain proteins, whereas the strong stimulation such as exogenous TGF-β1 can induce the Smad3 phosphorylation even in the absence of FYVE domain proteins. This difference in the sensitivity to the inhibition of PI3K between normal and scleroderma fibroblasts can be used for the development of the treatment for scleroderma since the proper concentration of LY294002 may completely block TGF-β/Smad3 signaling only in scleroderma fibroblasts.

During the preparation of this manuscript, Lim et al. (41) reported that keloid fibroblasts cocultured with keloid keratinocytes produced excessive amounts of ECM proteins, including type I collagen, through the synchronous activation of both the ERK and PI3K pathways, and the treatment of LY294002 or UO126, a MEK1/2-specific inhibitor, significantly reduced the expression of ECM proteins. This previous report implies the association of the PI3K activation with type I collagen gene expression in dermal fibroblasts. However, in the present study, we demonstrated that the basal activity of PI3K, but not the activation of this pathway, is sufficient for the efficient TGF-β/Smad3 signaling by the following findings, 1) the blockade of PI3K attenuates the TGF-β1-induced COL1A2 promoter activity, 2) CA p110 does not affect the TGF-β1-induced COL1A2 promoter activity, 3) TGF-β1 stimulation induces the rapid phosphorylation of Smad3, while it does not affect the phosphorylation of p85 subunit of PI3K, and 4) LY294002 reduces the TGF-β1-induced phosphorylation level of Smad3. Considering these present findings, previous data described above suggest two possibilities; 1) a cross-talk between the PI3K and ERK pathways induces the type I collagen gene expression through TGF-β/Smad3-independent pathway, and 2) the synchronous activation of both the PI3K and ERK pathways, but not the single activation of PI3K pathway, induces the type I collagen gene expression by modulating the TGF-β/Smad3 signaling. Because keloid is a dermal fibrotic disease, it is informative to assess these possibilities to further understand the disturbed regulation of type I collagen gene expression in dermal fibrotic disorders, including scleroderma.

In conclusion, we demonstrated that the basal activity of PI3K is indispensable to maintain the basal and the TGF-β1-dependent increased mRNA stability and the TGF-β1-dependent Smad3 phosphorylation. We also elucidated the different effect of PI3K inhibitor on Smad3 phosphorylation between normal and scleroderma fibroblasts. Although further studies are needed, these findings suggest that the inhibition of PI3K may be useful for the development of the treatment of scleroderma.

FIGURE 7.

The inhibition of PI3K attenuates the COL1A2 gene expression at the transcriptional and the posttranscriptional levels in scleroderma fibroblasts. A, Confluent quiescent scleroderma fibroblasts were treated in parallel with normal fibroblasts as described in Fig. 1. Briefly, scleroderma fibroblasts were treated with the indicated concentrations of LY294002 or DMSO for 25 h and each conditioned medium was analyzed by immunoblotting using anti-type I collagen Ab. B, Quantitative analysis of type I procollagen protein levels. Values represent the band density relative to mean value of normal fibroblasts treated with DMSO (100). ∗, p < 0.05 vs control cells treated with DMSO in each cell group. C, Total RNA was isolated from normal and scleroderma fibroblasts cultured under the same condition described above, resolved electrophoretically, and Northern blotting was performed with DNA probes for COL1A2 and GAPDH. D, Quantitative analysis of COL1A2 mRNA levels. Values represent the band density relative to the mean value of normal fibroblasts treated with DMSO (100). ∗, p < 0.05 vs control cells treated with DMSO in each cell group.

FIGURE 7.

The inhibition of PI3K attenuates the COL1A2 gene expression at the transcriptional and the posttranscriptional levels in scleroderma fibroblasts. A, Confluent quiescent scleroderma fibroblasts were treated in parallel with normal fibroblasts as described in Fig. 1. Briefly, scleroderma fibroblasts were treated with the indicated concentrations of LY294002 or DMSO for 25 h and each conditioned medium was analyzed by immunoblotting using anti-type I collagen Ab. B, Quantitative analysis of type I procollagen protein levels. Values represent the band density relative to mean value of normal fibroblasts treated with DMSO (100). ∗, p < 0.05 vs control cells treated with DMSO in each cell group. C, Total RNA was isolated from normal and scleroderma fibroblasts cultured under the same condition described above, resolved electrophoretically, and Northern blotting was performed with DNA probes for COL1A2 and GAPDH. D, Quantitative analysis of COL1A2 mRNA levels. Values represent the band density relative to the mean value of normal fibroblasts treated with DMSO (100). ∗, p < 0.05 vs control cells treated with DMSO in each cell group.

Close modal
FIGURE 7A.

(Continued) E, Confluent quiescent scleroderma fibroblasts were pretreated for 1 h with the indicated concentration of LY294002, and treated with actinomycin D (400 ng/ml) for the indicated time periods before RNA extraction. Expression levels of COL1A2 mRNA were determined by Northern blotting and normalized to GAPDH. The percentage of remaining COL1A2 mRNA relative to time 0 was plotted on a logarithmic scale. The mean from five separate experiments is shown. F, Scleroderma fibroblasts were transfected with 2 μg of the −353 COL1A2/CAT construct and incubated for 48 h. Cells were treated with the indicated concentration of LY294002 or DMSO for the last 24 h. Values represent the CAT activities relative to those of normal fibroblasts treated with DMSO (100). The mean and SEM from five separate experiments are shown. ∗, p < 0.05 vs control cells treated with DMSO in each cell group. G, Scleroderma fibroblasts were transfected with 2 μg of the −353 COL1A2/CAT construct, along with the indicated amount of the DN p85 and/or corresponding empty vector, and extracted after 48 h. Values represent the CAT activities relative to those of normal fibroblasts transfected with empty vector (100). The mean and SEM from five separate experiments are shown. ∗, p < 0.05 vs control cells transfected with empty vector in each cell group.

FIGURE 7A.

(Continued) E, Confluent quiescent scleroderma fibroblasts were pretreated for 1 h with the indicated concentration of LY294002, and treated with actinomycin D (400 ng/ml) for the indicated time periods before RNA extraction. Expression levels of COL1A2 mRNA were determined by Northern blotting and normalized to GAPDH. The percentage of remaining COL1A2 mRNA relative to time 0 was plotted on a logarithmic scale. The mean from five separate experiments is shown. F, Scleroderma fibroblasts were transfected with 2 μg of the −353 COL1A2/CAT construct and incubated for 48 h. Cells were treated with the indicated concentration of LY294002 or DMSO for the last 24 h. Values represent the CAT activities relative to those of normal fibroblasts treated with DMSO (100). The mean and SEM from five separate experiments are shown. ∗, p < 0.05 vs control cells treated with DMSO in each cell group. G, Scleroderma fibroblasts were transfected with 2 μg of the −353 COL1A2/CAT construct, along with the indicated amount of the DN p85 and/or corresponding empty vector, and extracted after 48 h. Values represent the CAT activities relative to those of normal fibroblasts transfected with empty vector (100). The mean and SEM from five separate experiments are shown. ∗, p < 0.05 vs control cells transfected with empty vector in each cell group.

Close modal

We thank Dr. H. Stenmark for providing us with the expression vector of 2xFYVE (Hrs).

1

This study is supported in part by a grant for scientific research from the Japanese Ministry of Education (10770391), and by the Project Research for Progressive Systemic Sclerosis from the Japanese Ministry of Health and Welfare.

3

Abbreviations used in this paper: TβRI, TGF-β receptor type I; SARA, Smad anchor for receptor activation; PI3K, phosphatidylinositol 3-kinase; CAT, chloramphenicol acetyltransferase; DN p85, dominant negative mutant of p85 subunit of PI3K; CA p110, constitutive active mutant of p110 subunit of PI3K.

1
LeRoy, E. C..
1992
. Systemic sclerosis (scleroderma). J. B. Wyngaarden, and L. H. Smith, and J. C. Bennett, eds.
Cecil Textbook of Medicine
19th Ed.
1530
. W. B. Saunders, Philadelphia.
2
Korn, J. H..
1989
. Immunologic aspects of scleroderma.
Curr. Opin. Rheumatol.
1
:
479
.
3
Mauch, C., T. Krieg.
1990
. Fibroblast-matrix interactions and their role in the pathogenesis of fibrosis.
Rheum. Dis. Clin. N. Am.
16
:
93
.
4
Jelaska, A., M. Arakawa, G. Broketa, J. H. Korn.
1996
. Heterogeneity of collagen synthesis in normal and systemic sclerosis skin fibroblasts: increased proportion of high collagen-producing cells in systemic sclerosis fibroblasts.
Arthritis Rheum.
39
:
1338
.
5
Mauch, C., E. Kozlowska, B. Eckes, T. Krieg.
1992
. Altered regulation of collagen metabolism in scleroderma fibroblasts grown within three-dimensional collagen gels.
Exp. Dermatol.
1
:
185
.
6
Wahl, S. M..
1994
. Transforming growth factor β: the good, the bad, and the ugly.
J. Exp. Med.
180
:
1587
.
7
LeRoy, E. C..
1974
. Increased collagen synthesis by SSc skin fibroblasts in vitro: a possible defect in the regulation or activation of the scleroderma fibroblast.
J. Clin. Invest.
54
:
880
.
8
Buckingham, R. B., R. K. Prince, G. P. Rodnan, F. Taylor.
1978
. Increased collagen accumulation in dermal fibroblast cultures from patients with progressive systemic sclerosis (scleroderma).
J. Lab. Clin. Med.
92
:
5
.
9
Peltonen, J., L. Kahari, J. Uitto, S. A. Jimenez.
1990
. Increased expression of type VI collagen genes in systemic sclerosis.
Arthritis Rheum.
33
:
1829
.
10
Xu, W. D., E. C. LeRoy, E. A. Smith.
1991
. Fibronectin release by systemic sclerosis and normal dermal fibroblasts in response to TGF-β.
J. Rheumatol.
18
:
241
.
11
Rudnicka, L., J. Varga, A. M. Christiano, R. V. Iozzo, S. A. Jimenez, J. Uitto.
1994
. Elevated expression of type VII collagen in the skin of patients with systemic sclerosis: regulation by transforming growth factor-β.
J. Clin. Invest.
93
:
1709
.
12
Falanga, V., S. L. Tiegs, S. P. Alstadt, A. B. Roberts, M. B. Sporn.
1987
. Transforming growth factor-β: selective increase in glycosaminoglycan synthesis by cultures of fibroblasts from patients with progressive systemic sclerosis.
J. Invest. Dermatol.
89
:
100
.
13
Massague, J..
1990
. Transforming growth factor-β family.
Annu. Rev. Cell. Biol.
6
:
597
.
14
LeRoy, E. C., E. A. Smith, M. B. Kahaleh, M. Trojanowska, R. M. Silver.
1989
. A strategy for determining the pathogenesis of systemic sclerosis: is transforming growth factor β the answer?.
Arthritis Rheum.
32
:
817
.
15
Kawakami, T., H. Ihn, W. Xu, E. Smith, C. LeRoy, M. Trojanowska.
1998
. Increased expression of TGF-β receptors by scleroderma fibroblasts: evidence for contribution of autocrine TGF-β signaling to scleroderma phenotype.
J. Invest. Dermatol.
110
:
47
.
16
Ihn, H., K. Yamane, M. Kubo, K. Tamaki.
2001
. Blockade of endogenous transforming growth factor β signaling prevents up-regulated collagen synthesis in scleroderma fibroblasts: association with increased expression of transforming growth factor β receptors.
Arthritis Rheum.
44
:
474
.
17
Derynck, R., Y. Zhang, X. H. Feng.
1998
. Smads: transcriptional activators of TGF-β responses.
Cell
95
:
737
.
18
Zhang, Y., X. Feng, R. We, R. Derynck.
1996
. Receptor-associated Mad homologues synergize as effectors of the TGF-β response.
Nature
383
:
168
.
19
Dennler, S., S. Itoh, D. Vivien, P. ten Dijke, S. Huet, J. M. Gauthier.
1998
. Direct binding of Smad3 and Smad4 to critical TGF β-inducible elements in the promoter of human plasminogen activator inhibitor-type 1 gene.
EMBO J.
17
:
3091
.
20
Labbe, E., C. Silvestri, P. A. Hoodless, J. A. Wrana, L. Attisano.
1998
. Smad2 and Smad3 positively and negatively regulate TGF-β-dependent transcription through the forkhead DNA-binding protein FAST2.
Mol. Cell.
2
:
109
.
21
Moustakas, A., D. Kardassis.
1998
. Regulation of the human p21WAF1/Cip1 promoter in hepatic cells by functional interactions between Sp1 and Smad family members.
Proc. Natl. Acad. Sci. USA
95
:
6733
.
22
Chen, S., W. Yuan, Y. Mori, A. Levenson, M. Trojanowska, J. Varga.
1999
. Stimulation of type I collagen transcription in human skin fibroblasts by TGF-β: involvement of Smad3.
J. Invest. Dermatol.
112
:
49
.
23
Tsukazaki, T., T. A. Chiang, A. F. Davison, L. Attisano, J. L. Wrana.
1998
. SARA, a FYVE domain protein that recruits Smad2 to the TGF-β receptor.
Cell
95
:
779
.
24
Itoh, F., N. Divecha, L. Brocks, L. Oomen, H. Janssen, J. Calafat, S. Itoh, P. ten Dijke.
2002
. The FYVE domain in Smad anchor for receptor activation (SARA) is sufficient for localization of SARA in early endosomes and regulates TGF-β/Smad signalling.
Genes Cells
7
:
321
.
25
Goto, D., H. Nakajima, Y. Mori, K. Kurasawa, N. Kitamura, I. Iwamoto.
2001
. Interaction between Smad anchor for receptor activation and Smad3 is not essential for TGF-β/Smad3 signaling.
Biochem. Biophys. Res. Commun.
281
:
1100
.
26
Ihn, H., K. Yamane, Y. Asano, M. Kubo, K. Tamaki.
2002
. IL-4 up-regulates the expression of tissue inhibitor of metalloproteinase-2 in dermal fibroblasts via the p38 mitogen-activated protein kinase-dependent pathway.
J. Immunol.
168
:
1895
.
27
Ihn, H., K. Ohnishi, T. Tamaki, E. C. LeRoy, M. Trojanowska.
1996
. Transcriptional regulation of the human α2(I) collagen gene: combined action of upstream stimulatory and inhibitory cis-acting elements.
J. Biol. Chem.
271
:
26717
.
28
Ihn, H., E. C. LeRoy, M. Trojanowska.
1997
. Oncostatin M stimulates transcription of the human α2(I) collagen gene via the Sp1/Sp3-binding site.
J. Biol. Chem.
272
:
24666
.
29
Poncelet, A. C., H. W. Schnaper.
2001
. Sp1 and Smad proteins cooperate to mediate transforming growth factor-β1-induced α2(I) collagen expression in human glomerular mesangial cells.
J. Biol. Chem.
276
:
6983
.
30
Ihn, H., K. Tamaki.
2000
. Increased phosphorylation of transcription factor Sp1 in scleroderma fibroblasts: association with increased expression of the type I collagen gene.
Arthritis Rheum.
43
:
2240
.
31
Yagi, K., M. Furuhashi, H. Aoki, D. Goto, H. Kuwano, K. Sugamura, K. Miyazono, M. Kato.
2002
. c-myc is a downstream target of the Smad pathway.
J. Biol. Chem.
277
:
854
.
32
Davidson, J., R. Berg.
1981
. Posttranslational events in collagen biosynthesis.
Methods Cell Biol.
23
:
119
.
33
Rosenbloom, J., M. Harsch, S. Jimenez.
1973
. Hydroxyproline content determines the denaturation temperature of chick tendon collagen.
Arch. Biochem. Biophys.
158
:
478
.
34
Geesin, J. C., D. Darr, R. Kaufman, S. Murad, S. R. Pinnell.
1988
. Ascorbic acid specifically increases type I and type III procollagen messenger RNA levels in human skin fibroblast.
J. Invest. Dermatol.
90
:
420
.
35
Murad, S., S. Tajima, G. R. Johnson, S. Sivarajah, S. R. Pinnell.
1983
. Collagen synthesis in cultured human skin fibroblasts: effect of ascorbic acid and its analogs.
J. Invest. Dermatol.
81
:
158
.
36
Tamaki, T., K. Ohnishi, C. Hartl, E. C. LeRoy, M. Trojanowska.
1995
. Characterization of a GC-rich region containing Sp1 binding site(s) as a constitutive responsive element of the α2(I) collagen gene in human fibroblasts.
J. Biol. Chem.
270
:
4299
.
37
Ricupero, D. A., C. F. Poliks, D. C. Rishikof, K. A. Cuttle, P. Kuang, R. H. Goldstein.
2001
. Phosphatidylinositol 3-kinase-dependent stabilization of α1(I) collagen mRNA in human lung fibroblasts.
Am. J. Physiol. Cell Physiol.
281
:
C99
.
38
Eckes, B., C. Mauch, G. Huppe, T. Krieg.
1996
. Differential regulation of transcription and transcript stability of pro-α1(I) collagen and fibronectin in activated fibroblasts derived from patients with systemic scleroderma.
Biochem. J.
315
:
549
.
39
Kubo, M., H. Ihn, K. Yamane, K. Tamaki.
2002
. Upregulated expression of transforming growth factor-β receptors in dermal fibroblasts of skin sections from patients with systemic sclerosis.
J. Rheumatol.
29
:
2558
.
40
Yamane, K., H. Ihn, K. Tamaki.
2002
. Increased transcriptional activities of transforming growth factor-β receptors in scleroderma fibroblasts.
Arthritis Rheum.
46
:
2421
.
41
Lim, I. J., T. T. Phan, E. K. Tan, T. T. Nguyen, E. Tran, M. T. Longaker, C. Song, S. T. Lee, H. T. Huynh.
2003
. Synchronous activation of ERK and phosphatidylinositol 3-kinase pathways is required for collagen and extracellular matrix production in keloids.
J. Biol. Chem.
278
:
40851
.