TGF-β can induce a variety of biological responses in cells, and these responses appear to be dependent on the cell type and context of the cell [41
]. Some of the mechanisms that appear to play a role in determining the specificity of the response of a cell to TGF-β include signal pathway cross-talk, regulation of the expression levels of cytoplasmic and nuclear antagonists, and control of the composition of transcription factor complexes [20
]. In addition, over a decade ago, Derynck and Feng proposed the possibility of a receptor threshold model for determining the specificity of TGF-β's effects. This model proposes that there is a critical expression level of the TGF-β receptor that determines the specific TGF-β responses of a cell [42
]. Several lines of indirect evidence have supported this model, but to date there has been no direct assessment of this hypothesis [8
]. We now provide direct evidence that the regulation of the expression level of TGFBR2 can affect the specificity of the TGF-β response. We have shown that the activation of the Smad and nonSmad signaling pathways can be modulated by the expression level of TGFBR2 and that the activation state of the nonSmad signaling pathway principally determines whether TGF-β can induce CDKN1A/p21 mediated apoptosis in an epithelial cell line.
TGF-β and the TGF-β receptors are expressed in nearly all cell types and in developing and adult organisms. In embryonic development there is clear evidence that the concentration of active TGF-β and the extent of TGF-β signal pathway activation is tightly regulated and creates a gradient of responses that determines cell fate, among other things [41
]. The majority of these studies have provided indirect and correlative evidence that the concentration of the ligands, receptors, and intracellular Smad proteins can determine the specific response of a cell to TGF-β. Further indirect evidence that the expression level can affect the specificity of the TGF-β response comes from studies of dominant negative TGFBR2
transgenes and chimeric receptors [22
]. Our studies now provide direct evidence that the TGFBR2 expression level can regulate the intensity of activation of the Smad and nonSmad signaling pathways. Furthermore, the differences in activation of the Smad and MAPK-ERK pathways at different TGFBR2 expression levels suggests that not only can the level of TGFBR2 determine the intensity of pathway activation but also which pathways are activated [33
]. The differential regulation of the post-TGF-β receptor pathways would be expected to lead to differences in transcription factor activation and transcription factor complex formation, which could then induce different patterns of gene expression.
Although the model above predicts that TGFBR2 expression levels will cause the differential expression of a variety of genes, we chose to specifically study CDKN1A/p21
because of its well-demonstrated regulation by TGF-β and because of its clear biological role in the inhibition of proliferation and induction of apoptosis. In addition, CDKN1A/p21
is regulated by TGF-β both by Smad and nonSmad signaling pathways and thus allows an assessment of the effect of Smad and nonSmad signaling on gene regulation [10
]. Our data suggest that CDKN1A/p21
is regulated by both the MAPK-ERK and Smad signaling pathways, which is true of other genes, such as FURIN
]. These pathways may crosstalk at the level of Smad phosphorylation, nuclear localization of Smads, or at the level of transcription factor complex formation [33
]. The differential activation of MAPK-ERK and Smad pathways suggests that at low levels of TGF-β receptor activation the nonSmad signaling pathways predominate over the Smad signaling pathways in determining the effect of TGF-β on the cell. The differences in activation of the MAPK and Smad pathways likely reflects the fact that the MAPK signaling pathway is a catalytic pathway whereas the Smad pathway is non-catalytic [8
]. An interesting observation related to this explanation for the differential activation of the Smad and nonSmad pathways is the fact that in V-400R2 the MAPK-ERK pathway appears to be the predominant pathway regulating the level of CDKN1A/p21
expression whereas the Smad pathway has a more modest effect. These results are consistent with those of Hu et al
who observed a similar phenomenon in the HaCaT keratinocyte cell line, but are in contrast to other studies that have shown that both Smad and SP1 mediated signaling are necessary for CDKN1A/p21
]. These differences may reflect cell line specific differences in the expression of other regulators of CDKN1A/p21
, such as c-Myc or MIZ-1 [56
There are several limitations related to our studies of the effect of TGFBR2 expression levels on TGF-β signaling. At this time, it is not clear whether the effects we have observed involve direct or indirect mechanisms for regulating the Smad and nonSmad signaling pathways. This limitation reflects our incomplete understanding of the mechanisms through which TGF-β affects the nonSmad signaling pathways [20
]. In addition, we have not assessed the effect of TGFBR2 expression levels on the regulation of other signaling pathways and cannot exclude the possibility that these pathways are also playing a role in the regulation of CDKN1A/p21
and apoptosis that is affected by the expression level of the TGF-β receptor. Nonetheless, our results demonstrate that TGF-β receptor levels can regulate MAPK-ERK activation and that the control of this pathway has a central role in regulating CDKN1A/p21
expression and CDKN1A/p21 mediated apoptosis in the V-400R2 cell line.
Our results have several implications for the role of TGF-β in regulating normal cellular responses and for the role of TGF-β in cancer. The relevance of the effect of TGFBR2 expression levels on the regulation of TGF-β's effects on cells is appreciated in light of prior studies that have shown decreased TGFBR2 expression in a variety of cancers and in some inflammatory states [58
]. In fact, many cancers have suppressed TGFBR2
expression without detectable mutations in any of the TGF-β signal pathway genes [59
]. In these cancers with an intact TGF-β receptor that is expressed at a low level, TGF-β appears to have the potential to act as an oncogene as well as a tumor suppressor gene [63
]. This paradoxical role of TGF-β has been associated with its ability to activate Smad independent pathways (MAPK, PI3K and Rho) and with the modulation of TGF-β signaling pathway activation through interactions with other signaling pathways induced by mutated oncogenes and tumor suppressor genes [8
]. Our results suggest that an additional mechanism by which TGF-β may mediate paradoxical effects on cells is through the down-regulation of the expression of TGFBR2, which could differentially activate signaling pathways and alter the gene expression patterns of the cells. Thus, our results not only demonstrate that the expression level of TGFBR2 can influence the pathway activation status of TGF-β signaling pathways and the specific response of cells to TGF-β, but also that the regulation of the expression level is a plausible mechanism for the paradoxical effects of TGF-β on cancer cells. In summary, regulation of TGFBR2 expression levels is an additional mechanism for controlling the specificity of TGF-β's effects on cells and is a potential mechanism for the paradoxical effects of TGF-β observed in cancer.