Master regulators direct TGF-beta and Smad signaling in different cell types

Back to top Introduction

Previous work had observed that the ESC master regulator Oct4 is required for the binding of signaling transcription factors such as Smad1 with their targets, and that genes enriched in binding sites for these factors tend to show ESC-specific expression (1). Within the past year, new studies have expanded on this finding, linking master regulators and signaling pathway transcription factors to cell-specific gene expression in several cell types and as the result of 3 individual signaling pathways (2, 3). Moreover, a mechanism driving master regulator expression in ESCs via a Smad-binding protein and histone marks has also been elucidated (4). The TGF-beta superfamily comprises several ligands, including TGF-beta, activins, nodal, and the bone morphogenetic proteins (BMP), and signaling is carried to the nucleus by the Smad family of transcription factors. TGF-beta signaling plays crucial roles in numerous biological processes, including growth and differentiation, adhesion and migration, and cell death (5, 6). Therefore, a complete understanding of how TGF-beta signaling activates gene expression is essential to clarifying numerous normal biological processes as well as disease states.

Back to top Smad3 interacts with cell-type-specific transcription factors to direct TGF-β responses

A recent study by Mullen et al reveals a strikingly preferential relationship between Smad3, a transcription factor of the TGF-beta signaling pathway, and cell-type-specific master transcription factors in embryonic stem cells, myotubes, and pro-B cells (see figure TGFb signaling) (2). These findings suggest a simple, elegant mechanism by which TGF-beta signaling drives different gene expression responses in different cell types. The team explored the relationship between Smad3 and the cell-type-specific transcription factors, dissecting their binding sites in the genome, their physical interactions, and the responses of their gene targets to TGF-beta signaling.
In human ESCs cultured with TGF-beta-containing media, SMAD3 bound to many of the same sites throughout the genome as OCT4. This also held true for murine ESCs, in which Smad3 was also found to co-occupy sites with Sox2 and Nanog. Smad3 and Oct4 appear to have a physical association, as determined by co-immunoprecipitation, and a ChIP assay revealed that they exist in the same crosslinked complex. This relationship has functional implications for gene expression in response to TGF-beta signaling. Oct4 inhibition severely impairs Smad3's ability to bind to its usual sites, and gene expression in response to TGF-beta receptor stimulation by activin is also diminished. This suggests that Oct4 plays a role in helping Smad3 bind to the right places in the genome in order to drive transcription. Oct4 binding also generally took place in nucleosome-depleted regions, which may indicate that co-occupation takes place at sites that are more accessible to Smad3.

This phenomenon is not ESC-specific, as these findings also held true in other cell types. The master transcription factors for myotubes and pro-B cells, Myod1 and PU.1 respectively, bound along with Smad3 at sites distinct for each cell type; sites bound by Smad3 and Oct 4 in ESCs were not bound by Myod1 and Smad3 in myotubes, for example. Other characteristics were also similar; depleting Myod1 and PU.1 diminished Smad3 binding at co-occupied sites, SBEs were enriched at Myod1 and PU.1 binding sites, these transcription factors co-immunoprecipitated with Smad3 in their respective cell types, and interactions took place in regions that were relatively low in nucleosomes. Moreover, even when Smad3 occupied the same gene across different cell types, it did not bind to the same enhancer in both cells, but bound each time with the cell type's master transcription factor.

Master transcription factors and Smad3 bound direct gene targets of TGF-beta signaling, as shown by inhibitor experiments. Activin-triggered TGF-beta signaling in ESCs also induced p300 recruitment to Oct4-Smad3-bound sites. Moreover, consistent with the findings that master transcription factors in different cell types bound different genes, these distinct gene sets were expressed in myotubes and pro-B cells upon TGF-beta treatment. Finally, demonstrating the directional power of the master transcription factors with regard to Smad3, the team expressed Myod1 in ESCs. Some Smad3 bound with Myod1, while other Smad3s remained bound with the remaining Oct4. The power of cell-specific master regulators to direct the TGF-beta signaling response explains how signaling through the same molecules in different cell types can lead to distinct gene expression profiles. This mechanism may be relevant in other settings as well, as described in the following section.

Back to top Master transcription factors direct lineage-specific gene expression in hematopoietic regeneration

Dissecting the activities of the BMP and Wnt signaling pathways during hematopoietic regeneration, Trompouki et al observed similar phenomena, showing that signaling transcription factors from these pathways also co-occupied sites with cell-type-specific master regulators (3). After establishing a role for both pathways in zebrafish regeneration after irradiation, they used a murine model and human cell lines for subsequent experiments. They examined the co-occupancy of SMAD1, a downstream transcription factor for BMP signaling, or TCF7L2, a transcription factor for the Wnt pathway, with the cell-type-specific master regulators GATA1, GATA2, and C/EBPα.

In erythroleukemia cells, SMAD1 and TCF7L2 co-occupied genes with master regulators for erythroid (GATA1) and progenitor cells (GATA2) after BMP or Wnt signaling, many of which were related to development of red blood cells. However, GATA binding itself was not substantially influenced by these signaling pathways. Intriguingly, SMAD1 and TCF7L2 also bound preferentially to GATA-occupied sites compared to other transcription factors.

SMAD1 and TCF7L2 significantly influence gene expression when co-occupying sites with master regulators. The team determined that SMAD1 and TCF7L2 colocalize with GATA binding sites at enhancer regions, and showed that while GATA2 itself causes transcription, this is significantly increased with the induction of SMAD1 and or TCF7L2. Moreover, BMP and Wnt signaling increases p300 at these sites. In a monocytic leukemia cell line, SMAD1 and TCF7L2 co-occupied sites that were mostly distinct from those in the erythroleukemia cells, suggesting that binding sites for these signaling pathway transcription factors vary depending on the cell type. In further contrast to the erythroid leukemia cells, these factors bind to sites with C/EBP motifs instead of GATA motifs.

Inducing a different master regulator influenced distribution of the signaling transcription factors; induction of C/EBPα in erythroid cells caused some SMAD1 to localize with the myeloid master regulator, and similar findings were observed with Gata1 induction in estrogen-inducible Gata1-null erythroblast progenitor cells. Finally, to determine whether these experimental findings mimic what happens during normal hematopoiesis and erythropoiesis, the team looked at normal human peripheral blood progenitor cells and found that GATA2 co-occupied most genes occupied by the signaling transcription factors. Erythroblasts that had differentiated from normal progenitors showed co-occupation of genes by SMAD1 and GATA1, and this set of genes was smaller and more erythroid-specific than that of the progenitors. In summary, while TGF-beta induces distinct sets of expressed genes in different cell types based on interaction of Smad3 with master regulators, BMP and Wnt trigger a similar phenomenon during lineage differentiation in hematopoietic cells.

Back to top TGF-beta signaling triggers differentiation through interaction of TRIM33 with chromatin marks

Another intriguing finding in the gene expression response to the TGF-beta pathway came from Xi and colleagues, who studied how genes that drive differentiation become activated through the interaction of nodal signaling-related complexes with specific chromatin marks (4). The team used murine embryonic stem cells and nodal/activin triggering of TGF-beta signaling to study this phenomenon, focusing on two genes that are crucial in early development, goosecoid (Gsc) and Mix-like homeodomain protein 1 (Mixl1). They found that the Smad-binding protein tripartite motif 33 (TRIM33) is intimately involved in this process, interacting with histone marks to facilitate Smad binding and gene expression.

TRIM33 plays a crucial role in activin-induced expression of developmental genes in murine embryonic stem cells. Independent complexes of TRIM33-Smad2/3 and Smad4-Smad2/3 formed in response to activin stimulation, and Smad4 null or Smad3-depleted Smad2 null mice showed impaired Gsc, Mixl1 and Smad7 responses. TRIM33 null mice also exhibited diminished induction of Gsc and Mixl1 in response to activin, but the lack of TRIM33 had no impact on homeostasis genes. As predicted by its role in Gsc and Mixl1 induction, TRIM33 was also crucial for mesendodermal differentiation in embryoid bodies; however, it was not needed for ectodermal differentiation.

TRIM33's specific function in driving the expression of developmental regulators was grounded in its interaction with three histone marks, H3K9me3, unmodified H3K4, and H3K18ac. The C-terminal region of TRIM33 bears a plant homeodomain and a Bromo domain (PHD-Bromo), both of which were required for binding of TRIM33 to histone H3. Crystal structure analysis revealed interaction of the PHD domain with H3K9me3 and H3K4, while H3K18ac bound with the Bromo domain. H3K9me3 and H3K18ac were observed in the promoters of Gsc and Mixl1, along with Smad binding elements, and activin induced binding of TRIM33 at these sites. Smad2/3 bound at both the chromatin mark and activin response element (ARE) sites, while Smad4 bound only at the ARE, suggesting that the previously observed TRIM33/Smad2/3 complex binds at the chromatin marks, and the Smad4/Smad2/3 complex binds at the ARE.

The TRIM33 and Smad4 complexes cooperate, each enhancing the binding of the other. TRIM33 is necessary for binding of both complexes to their respective sites on the promoters, as shown through Trim33 knockdown, while Smad4 increases acetylation on H3 to strengthen TRIM33 binding, as demonstrated using Smad4 null ESCs and activin stimulation. TRIM33 also competes with HP1γ, an H3K9me3-binding protein involved in chromatin compaction, for binding to H3. When K18 is acetylated along with trimethylated K9, TRIM33 is able to do so successfully. The team went on to confirm that activin causes less HP1γ to bind at the promoters of Gsc and Mixl1, dependent on TRIM33, and this also correlated with higher levels of Pol II binding. Therefore, through TRIM33's interactions with specific histone marks, this protein facilitates Smad binding and transcription of master regulators of development.

Back to top Conclusions

Recent findings have shed new light on how signal transduction pathways, especially those of the TGF-beta superfamily, influence specific gene expression during cell differentiation and development. The genes activated by TGF-beta superfamily signaling are crucial during embryonic development, wound healing, immune regulation, and other processes, and dysregulation of these pathways has also been implicated in a host of diseases including cancer, fibrosis, and cardiovascular disease (7). Pathway-focused tools such as quantitative PCR arrays give targeted insights into the modulation of these genes, providing an opportunity to quickly and comprehensively analyze the activity of these pathways in your area of research.

Back to top References

1. Chen, X. et al (2008) Integration of external signaling pathways with the core transcriptional network in embryonic stem cells. Cell 133, 1106.
2. Mullen, A.C. et al. (2011) Master transcription factors determine cell-type-specific responses to TGF-beta signaling. Cell 147, 565.
3. Trompouki, E. et al. (2011) Lineage regulators direct BMP and Wnt pathways to cell-specific programs during differentiation and regeneration. Cell 147, 577.
4. Xi, Q. et al. (2011) A poised chromatin platform for TGF-beta access to master regulators. Cell 147, 1511.
5. Massagué J., Seoane, J., and Wotton, D. (2005) Smad transcription factors. Genes Dev. 19, 2783.
6. Wu, M.Y., and Hill, C.S. (2009) TGF-beta superfamily signaling in embryonic development and homeostasis. Dev. Cell 16, 329.
7. Santibañz, J.F., Quintanilla, M., and Bernabeu, C. (2011) TGF-beta/TGF-beta receptor system and its role in physiological and pathological conditions. Clin. Sci. 121, 233.