Synergistic and multidimensional regulation of plasminogen activator inhibitor type 1 expression by transforming growth factor type β and epidermal growth factor

Abstract

The major physiological inhibitor of plasminogen activator, type I plasminogen activator inhibitor (PAI-1), controls blood clotting and tissue remodeling events that involve cell migration. Transforming growth factor type β (TGFβ) and epidermal growth factor (EGF) interact synergistically to increase PAI-1 mRNA and protein levels in human HepG2 and mink Mv1Lu cells. Other growth factors that activate tyrosine kinase receptors can substitute for EGF. EGF and TGFβ regulate PAI-1 by synergistically activating transcription, which is further amplified by a decrease in the rate of mRNA degradation, the latter being regulated only by EGF. The combined effect of transcriptional activation and mRNA stabilization results in a rapid 2-order of magnitude increase in the level of PAI-1. TGFβ also increases the sensitivity of the cells to EGF, thereby recruiting the cooperation of EGF at lower than normally effective concentrations. The contribution of EGF to the regulation of PAI-1 involves the MAPK pathway, and the synergistic interface with the TGFβ pathway is downstream of MEK1/2 and involves phosphorylation of neither ERK1/2 nor Smad2/3. Synergism requires the presence of both Smad and AP-1 recognition sites in the promoter. This work demonstrates the existence of a multidimensional cellular mechanism by which EGF and TGFβ are able to promote large and rapid changes in PAI-1 expression.

FIGURE 1.

Synergistic regulation of PAI-1 protein and mRNA in mink and human cells. A and B, PAI-1 protein. Mv1Lu and human HepG2 cells were labeled with a 4-h pulse of [³⁵S]methionine (Mv1Lu) or ³⁵S-Trans-label (HepG2) after incubating the cells for 2 h (Mv1Lu) or 8 h (HepG2) with growth factors. The amount of radiolabeled PAI-1 in the medium (Mv1Lu) or after immunoprecipitation with PAI-1 antiserum (HepG2) is shown after resolution by SDS-PAGE (duplicate samples shown for HepG2). The bands identified for Mv1Lu as PAI-1 were verified as such by immunoprecipitation (9). B, the effects of EGF and TGFβ on PAI-1 synthesis in Mv1Lu cells were quantified by scanning the identified bands (PAI-1 and 38 kDa) in the gel shown in the top left panel (Mv1Lu) and normalizing these values to the total cell acid-precipitable ³⁵S cpm. Duplicate independent samples were tested, and the average results with S.E. values are shown. C and D, PAI-1 mRNA. Cells were treated for 2 h with 5 ng/ml EGF, 1 ng/ml TGFβ, or the combination of EGF and TGFβ at the same concentrations. PAI-1, urokinase plasminogen activator receptor (uPAR), COX2, and urokinase plasminogen activator (uPA) mRNAs were quantified by RT-qPCR for Mv1Lu (C) and HepG2 (D). −, vehicle control; E, EGF; T, TGFβ; TE, TGFβ and EGF. *, p < 0.05; **, p < 0.01; ***, p < 0.001. Error bars, S.E.

FIGURE 2.

TGFβ synergizes with various ligands of RTK receptors and with phorbol 12-myristate 13-acetate (PMA). A and B, PAI-1 mRNA levels. A, cells were treated for 2 h with or without 5 ng/ml EGF, 5 ng/ml FGF-2, 5 ng/ml TNFα, or 150 ng/ml IGF-1 in the presence or absence of 1 ng/ml TGFβ. PAI-1 mRNA was measured by RT-qPCR. B, the synergism observed in A is quantified. N, number of independent experiments. C and D, secreted protein levels. Mv1Lu were labeled with a 4-h pulse of [³⁵S]methionine after 2 h incubation of the cells with growth factors. The amount of radiolabeled PAI-1 secreted into the medium was determined by autoradiography. The effects of TGFβ and phorbol 12-myristate 13-acetate on PAI-1 synthesis in Mv1Lu cells were quantified by scanning the identified bands (C, PAI-1; D, 73-kDa protein) in the gel and normalizing these values to the cell monolayer acid-precipitable ³⁵S cpm. More than four independent experiments were performed for each condition, and the data are shown as mean ± S.E. (error bars). **, p < 0.01; ***, p < 0.001.

FIGURE 3.

MAPK is involved in the synergism. A–E, effects of blocking specific pathways by kinase inhibitors. Cells were treated for 1 h with inhibitors (or control analogs) prior to 2 h with growth factors. PAI-1 mRNA was measured by RT-qPCR. Empty bars, control analogs; filled bars, inhibitors. F, no synergism in phosphorylation of ERK1/2. Cells were treated for 1 h with 20 μm MEK inhibitor U0126 or its inactive analog U0124 and then for 10 min with 1 ng/ml TGFβ, 5 ng/ml EGF, or their combination. The phosphorylation of ERK1/2 was determined by Western blot. −, vehicle control. E, EGF; T, TGFβ; TE, EGF and TGFβ. *, p < 0.05; **, p < 0.01; ***, p < 0.001. Error bars, S.E.

FIGURE 4.

A, synergism does not require de novo protein synthesis. Cells were first treated with 50 μg/ml cycloheximide for 1 h and then with 1 ng/ml TGFβ, 5 ng/ml EGF, or their combination for 2 h. PAI-1 mRNA was measured with RT-qPCR. About 3-fold synergism was observed in both cycloheximide-treated (■) and control (□) cells. B, EGF, but not TGFβ, increases the half-life of PAI-1 mRNA. Mv1Lu cells cultured in DMEM with 0.2% calf serum were incubated for 2 h with 1 ng/ml TGFβ and 5 ng/ml EGF or their combination. Coredycepin (15 μg/ml) was then added (set as zero time) to block transcription. Total RNA was isolated at the indicated times, and PAI-1 mRNA was quantified and normalized to the GAPDH values in the same samples. The results are from at least five independent experiments. The half-life of PAI-1 mRNA in TGFβ-treated cells was calculated to be 36 min (regression coefficient from a least squares analysis = 0.90), and that in EGF-treated cells was calculated to be 53 min (regression coefficient = 0.90). The half-life of PAI-1 mRNA in TGFβ plus EGF-treated cells is 58 min (regression coefficient = 0.80). When tested by using analysis of covariance (JMP 7.0), the difference between treatment with TGFβ alone and treatment with the combination of EGF and TGFβ was significant to the level of p < 0.05. **, p < 0.01; ***, p < 0.001. Error bars, S.E.

FIGURE 5.

A and C, EGF and TGFβ cooperate to increase PAI-1 transcription. Cells were treated for 2 h with 5 ng/ml EGF, 1 ng/ml TGFβ, or their combination. Newly transcribed PAI-1 mRNA (A) and total mRNA (C) were quantified by the nuclear run-on assay. Shown are the average results with the S.E. from two independent experiments. B and D, the synergistic mechanism is activated soon after the growth factor addition. Cells were treated with EGF and TGFβ or their combination for various time periods, and then 4-thiouridine was added for the last 15 min of each incubation. Newly synthesized (B) and total (D) RNAs were isolated and quantified for PAI-1 and GAPDH by RT-qPCR, and the PAI-1 value was normalized to the GAPDH value for the same sample. −, vehicle control; E, EGF; T, TGFβ; TE, TGFβ and EGF. *, p < 0.05; **, p < 0.01; ***, p < 0.001. Error bars, S.E.

FIGURE 6.

AP-1 is involved in the synergism between EGF and TGFβ. A, transcriptional activities mediated by Smad2/3 and AP-1 respond synergistically to EGF and TGFβ. Cells were transfected with the 3TP luciferase reporter for 17 h then for 6 h with growth factors. B and C, synergism is inhibited by curcumin and ΔFosB. Cells were transfected with 3TP with and without ΔFosB for 17 h and then for 6 h with growth factors with or without 20 μm curcumin, which was added 1 h prior. D, curcumin has no effect on Smad activity. Cells were transfected with pSBE4 and treated with curcumin and growth factors. Promoter activities were measured with a luciferase assay. E, no synergism on phosphorylation of Fos and Jun. Cells were treated with growth factors for 40 min. Western blots show the phosphorylation of Fos and Jun in the same samples. The lower portion of the gel (not transferred to nitrocellulose) was stained with Coomassie Blue and is shown as the loading control. F, synergistic induction of PAI-1 mRNA is inhibited by ΔFosB. Cells were transiently transfected with ΔFosB for 17 h and then for 2 h with growth factors, and then PAI-1 mRNA was quantified relative to GAPDH. −, vehicle control; E, EGF; T, TGFβ; TE, EGF and TGFβ. Empty bars, control; filled bars, inhibitors. *, p < 0.05; **, p < 0.01; ***, p < 0.001. Error bars, S.E.