TGFβ versatility: PI3K as a critical mediator of distinct cell type and context specific responses

Transforming growth factor-beta (TGFβ) is a critical cytokine for mammalian development and homeostasis, regulating a large number of biological processes including cell growth, migration, differentiation, ECM production, angiogenesis and immunity, among others.1 The capacity of TGFβ to influence such a myriad of processes is possible due to its ability to induce extremely variable cellular responses depending on the cell type and stimulation context.1 This versatility includes exerting inhibitory effects on epithelial cell growth and immune responses while also being capable of promoting epithelial-mesenchymal transition (EMT) and fibroblast activation.1 Intriguingly, TGFβ drives these diverse outcomes by binding the same receptor complex (composed of the Type I and II Serine/Threonine kinases) and subsequently activating the same members of the Smad family of transcription factors (Smad 2 and 3). Given the important role of TGFβ in carcinogenesis, angiogenesis, autoimmunity and fibrotic diseases, understanding the molecular mechanisms underlying the pleiotropic effects of this growth factor is of considerable importance. Over the last several years, a great deal of progress has been made in our understanding of the signal transduction pathways that mediate the various cellular responses to TGFβ. It is now clear that, in addition to the Smad pathway, the TGFβ receptor complex is capable of inducing non-Smad signals, which may be dependent or independent of Smads, and are activated in a cell type and context specific manner.1 Integration of Smad and non-Smad signaling pathways ultimately determines the nature of the cellular response. A number of TGFβ non-Smad effectors have been identified in various cell types including PP2A, Par6, MAPKs, PI3K, PAK2, c-Abl, Akt and mTOR.1,2 Of these non-Smad effectors, there is mounting evidence that PI3K plays a critical role in regulating both the inhibitory and stimulatory TGFβ responses. For instance, the ability of TGFβ to promote epithelial growth arrest has been shown to require the Smads collaborating with the FoxO family of transcription factors to upregulate various cyclin-dependent kinase inhibitors (CDKIs).3–5 FoxO transcription factors are known to be inhibited by the PI3K target Akt via direct phosphorylation and experimental evidence suggests that inducing an epithelial cell to activate PI3K attenuates the ability of TGFβ to upregulate CDKIs.3,6 Not surprisingly, TGFβ itself does not activate PI3K in the contexts in which it induces epithelial growth arrest.7 As opposed to this cytostatic program, certain epithelial cell types positively respond to TGFβ by undergoing EMT.1 Studies have demonstrated that in this context, the TGFβ receptor complex is capable of activating PI3K, a process which promotes this transdifferentiation event.8,9 TGFβ mediated PI3K-Akt signaling upregulates the transcription factor Snail and induces phosphorylation and inhibition of GSK-3β which leads to assembly of LEF-1/β-Catenin complexes and upregulation of EMT promoting genes.8 In addition, TGFβ promotes an increase in cell size and migration during EMT via PI3K-Akt-mTOR signaling.9 Along with driving EMT, PI3K is critical in the positive fibroblast response to TGFβ. In fibroblasts, TGFβ induces PI3K activation which activates two independent signaling pathways. One arm leads to PAK2 and subsequently c-Abl signaling, and the other arm leads to Akt and subsequently mTORC1 signaling.2,7,10,11 These two effector arms are critical for TGFβ mediated fibroblast proliferation, differentiation into myofibroblasts, and ECM production.2,7,10,11 Furthermore, animal model studies have shown that the c-Abl inhibitor Imatinib/Gleevec reduces bleomycin-induced lung fibrosis and ureter obstruction induced kidney fibrosis in vivo.1 While it is clear that PI3K is one of the critical factors regulating the cellular responses to TGFβ, it is currently unclear how the TGFβ receptor complex couples to PI3K and why TGFβ does not activate this kinase in certain cell types. One study has shown that the type II receptor can associate with PI3K.12 However, this interaction is indirect and the mechanisms whereby the TGFβ receptor complex binds to PI3K are unknown.12 If an adaptor couples the TGFβ receptor complex to PI3K, differential expression or activity of this adaptor may explain why TGFβ does not activate PI3K in all cell types. In contrast, there is evidence that de novo gene expression is required for TGFβ mediated PI3K signaling.2 One potential explanation for this phenomenon is that TGFβ induces the upregulation of an additional growth factor that activates PI3K in an autocrine/paracrine manner. Previous studies have shown that TGFβ induces PDGF, CTGF and FGF-2 expression in fibroblasts.1 However, the relative contribution of these factors to PI3K signaling is unknown. Furthermore, it is unclear why these genes are not induced by TGFβ in all cell types. A study by Bruna et al. in gliomas suggests that epigenetic regulation may play an important role.13 In non-aggressive gliomas, TGFβ induces growth arrest, a response that is lost with increased malignant progression.14 Interestingly, highly aggressive gliomas actually proliferate in response to TGFβ.13 The mechanism for this differential response involves the methylation status of the PDGF-B promoter.13 In poorly aggressive gliomas, the PDGF-B promoter is methylated, preventing Smads from binding to the Smad binding elements present and promoting transcription. However, via an unknown mechanism, malignant progression is associated with promoter demethylation.13 Whether the methylation status of the PDGF-B promoter is different between various cell types still requires investigation. While these studies shed some light on the complexity of TGFβ biology, more studies are clearly necessary. The ultimate goal of such investigations will be to modulate the effects of TGFβ on certain disease processes without influencing its important roles in normal physiology.