Dynamic regulation of Tgf-B signaling by Tif1γ: a computational approach

Abstract

TIF1γ (Transcriptional Intermediary Factor 1 γ) has been implicated in Smad-dependent signaling by Transforming Growth Factor beta (TGF-β). Paradoxically, TIF1γ functions both as a transcriptional repressor or as an alternative transcription factor that promotes TGF-β signaling. Using ordinary differential-equation models, we have investigated the effect of TIF1γ on the dynamics of TGF-β signaling. An integrative model that includes the formation of transient TIF1γ-Smad2-Smad4 ternary complexes is the only one that can account for TGF-β signaling compatible with the different observations reported for TIF1γ. In addition, our model predicts that varying TIF1γ/Smad4 ratios play a critical role in the modulation of the transcriptional signal induced by TGF-β, especially for short stimulation times that mediate higher threshold responses. Chromatin immunoprecipitation analyses and quantification of the expression of TGF-β target genes as a function TIF1γ/Smad4 ratios fully validate this hypothesis. Our integrative model, which successfully unifies the seemingly opposite roles of TIF1γ, also reveals how changing TIF1γ/Smad4 ratios affect the cellular response to stimulation by TGF-β, accounting for a highly graded determination of cell fate.

Figure 1. Schematic representation of the models.

Detailed information on parameters and entities are given in Tables S1 and S2. A) Model hypothesis from . B) Model hypothesis from . C) Integrated model including TIF1γ (red rectangle) and FAM (green rectangle).

Figure 2. Effect of TIF1γ on TGF-β signaling.

Modeling analysis of the pS24n response to increasing TIF1γ concentrations at a 10 nM TGF-β input. A) Model according to ; B) model according to and C) integrated model.

Figure 3. Parameter sensitivity analysis.

Modeling analysis of pS24n response to variations of kinetic constants at a 10 nM TGF-β input. A) k_on-pS24nTIF1γ, binding of TIF1γ to phosphorylated-Smad2/Smad4 complexes; B) k_off-pS24nTIF1γ, dissociation of phosphorylated-Smad2/Smad4/TIF1γ complexes in the nucleus; C) k_off-pS2nTIF1γ, dissociation of phosphorylated Smad2-TIF1γ complexes in the nucleus; D) k_in-S4ub, nuclear export of ubiquitinated Smad4 in the cytoplasm; E) and F) k_dub, deubiquitination of Smad4 according to relative FAM concentrations of 1 nM (E) and 10 nM (F).

Figure 4. TIF1γ, Smad2 and Smad4 bind to the PAI-1 promoter.

ChIP assays were performed on HMEC cells treated with TGF-β for the indicated times. Cell lysates were subjected to anti-Smad4 (IP Smad4), or anti-TIF1γ (IP TIF1γ), or anti-Smad2/3 (IP Smad2/3) chromatin immunoprecipitation. PCR amplification of the endogenous PAI-1 promoter (733/484) was performed to detect protein bound DNA. Primers specific to actin were used as controls.

Figure 5. Expression of the CDH2 and CDH11 TGF-β target genes is sensitive to TIF1γ/Smad4 ratios.

HMEC cells were transfected with Smad4 (siSmad4) or TIF1γ (siTIF1γ) siRNAs and cultured in the presence (+) or absence (−) of TGF-β for the indicated times (days). Controls were cells transfected with non-targeted siRNA (scr). A) Smad4 and TIF1γ protein levels were analyzed by immunoblotting (upper panels) and quantified by densitometric scanning (lower panels). B) TGF-β-induced fold changes in CDH2 and CDH11 expression were analyzed by RT–qPCR. All values were normalized to the amount of HPRT mRNA and expressed relative to the value obtained for TGF-β-untreated controls in arbitrary units (AU). Results are expressed as the mean ± SD of 3 independent experiments. C) mRNA levels of CDH2 (red circles) and CDH11(blue circles) were plotted against TIF1γ/Smad4 ratios and were fitted to the predictive equation curve of pS24n relative concentrations.

Figure 6. Concentration and time dependence of TGF-β signaling.

A) and B) Modeling analysis of the pS24n response to increasing concentrations of TGF-β (A) and duration of stimulation with 10 nM TGF-β (B). C) and D) Modeling analysis of the maximum pS24n response as a function of TGF-β duration of exposure (C) or increasing TIF1γ/Smad4 ratios (D).

Figure 7. TGF-β time-dependent pS24n and pS2nTIF1γ response profiles as a function of TIF1γ/Smad4 ratios.

Results are expressed as percentage of the maximum production of pS24n (A) or pS2nTIF1γ (B).