Hierarchical model of gene regulation by transforming growth factor beta

Abstract

Transforming growth factor betas (TGF-betas) regulate key aspects of embryonic development and major human diseases. Although Smad2, Smad3, and extracellular signal-regulated kinase (ERK) mitogen-activated protein kinases (MAPKs) have been proposed as key mediators in TGF-beta signaling, their functional specificities and interactivity in controlling transcriptional programs in different cell types and (patho)physiological contexts are not known. We investigated expression profiles of genes controlled by TGF-beta in fibroblasts with ablations of Smad2, Smad3, and ERK MAPK. Our results suggest that Smad3 is the essential mediator of TGF-beta signaling and directly activates genes encoding regulators of transcription and signal transducers through Smad3/Smad4 DNA-binding motif repeats that are characteristic for immediate-early target genes of TGF-beta but absent in intermediate target genes. In contrast, Smad2 and ERK predominantly transmodulated regulation of both immediate-early and intermediate genes by TGF-beta/Smad3. These results suggest a previously uncharacterized hierarchical model of gene regulation by TGF-beta in which TGF-beta causes direct activation by Smad3 of cascades of regulators of transcription and signaling that are transmodulated by Smad2 and/or ERK.

Fig. 1.

Effects of ablation of Smad2, Smad3, and ERK function on gene regulation by TGF-β.(a) Resampled hierarchical clustering (bootstrapping) for 360 TGF-β target genes demonstrates four significant gene clusters (A-D) and three experimental variables clusters (a-c). Gene cluster A, immediate-early targets; gene cluster B, intermediate repressed targets; gene cluster C, intermediate-induced targets; gene cluster D, intermediate-repressed targets. Experimental variables cluster a, T₀ and T_0.3 of WT, Smad2KO (S2KO), Smad3KO (S3KO), and U0126 ablation together with T₁, T₂, and T₄ of S3KO; experimental variables cluster b, T₁ of WT, U0126, and S2KO; experimental variables cluster c, T₂ and T₄ of WT, U0126, and S2KO. (b) Principle component analysis (17) demonstrates the overall relatedness of expression profiles for 360 TGF-β target genes in each of the experimental variables in the four time series (WT, U0126, Smad2KO, and Smad3KO). Colored square dots and associated time-point values indicate WT (black), U0126 (green), Smad2KO (purple), and Smad3KO (blue) variables.

Fig. 2.

Patterns of regulation and promoter elements of IEGs, IIGs, and IRGs. (a) The centroid (±SD) of expression profiles of IEGs, IIGs, and IRGs in WT, U0126-treated WT, Smad2KO (S2KO), and Smad3KO (S3KO) fibroblasts. A centroid of a cluster in self-organizing maps is a data point with coordinates that are the averages of the corresponding coordinates for median log ratios of gene expression for all gene expression profiles in a cluster. (b) Hierarchical clustering of IEG expression profiles in 20 experimental variables. (c) Northern blot demonstrates Riken clone 1190017B18 mRNA expression in Smad2KO and littermate-derived Smad2WT, Smad3KO and littermate-derived Smad3WT fibroblasts, and Smad3KO fibroblasts transduced with Smad3 adenovirus (KO+TxS3). Cells were treated with TGF-β as indicated. (d) Distribution of [GTCT] Smad3/Smad4 core binding-site repeats in unregulated control genes, IEGs, IRGs, and IIGs. Black box, number of genes containing [GTCT] core repeats with spacer lengths ≤3 bp; gray box, number of genes harboring [GTCT] core repeats with spacer lengths >3 bp; white box, number of genes without [GTCT] core repeats in each group (the software and methods used are described in Supporting Materials and Methods).

Fig. 3.

Differential functional profiles of TGF-β target genes. Distribution (fraction of genes) of IEGs (black bars), IRGs (gray bars), and IIGs (white bars) in “molecular function” (a) and “biological process” (b) categories as defined by the Gene Ontology Consortium (14). (c) Hierarchical model of gene regulation after ligand-induced activation of the TGF-β receptor complex (TβRI/II) (see Conclusions for details).