Experimental conditions may be critical to explain one model
Experimental conditions may be critical to explain one model or the other. The matter of Hippo-driven cytoplasmic sequestration (Varelas et al., 2008) of active SMAD complexes (Varelas et al., 2008) vs. TßR basolateral relocalization upon epithelial polarization (Nallet-Staub et al., 2015) as mechanisms to interfere with apical TGF-ß signaling was revisited by the same group that proposed the activated SMAD cytoplasmic sequestration model. As it turns out, both models co-exist and may be complementary to each other as they take place sequentially during establishment of epithelial cell polarity (Narimatsu et al., 2015): cytoplasmic P-SMAD sequestration by P-TAZ is a transient mechanism that occurs before full polarity is established and ion channels go through a pre-polarization state, and it is followed by TßR receptor basolateral relocalization once polarity is fully established (Narimatsu et al., 2015). Given the transient nature of P-SMAD sequestration by P-YAP/TAZ, which was not observed in our own studies (Nallet-Staub et al., 2015), a precise method to observe the phenomenon in response to apically delivered TGF-ß in pre-polarized EpH4 immortalized mouse mammary epithelial cells was recently detailed by the authors (Narimatsu et al., 2016). Whether or not SMAD sequestration occurs in pre-polarized epithelial cells upon basolateral TGF-ß ligand delivery was not addressed.
Context-specific target gene regulation Various means by which R-SMADs and YAP/TAZ regulate target genes, whether in concert or antagonistically, have been uncovered and are highly contextual. In a publication comparing weakly vs. highly aggressive breast carcinoma cells lines, Hiemer et al. found out that distinct YAP/TAZ nucleo-cytoplasmic localization is linked to the differential sensitivity of breast carcinoma cell lines to TGF-ß-induced growth arrest (Hiemer et al., 2014). In weakly invasive cell lines, the Hippo pathway is active (e.g., YAP/TAZ are cytoplasmic), and TGF-ß induces tumor suppression and cytostasis via SMAD2/3/4-dependent transcriptional mechanisms whereby active SMAD complexes bind DNA in the absence of YAP/TAZ (Fig. 4A, left panel). On the other hand, when Hippo is inactive, as in highly metastatic cell lines, SMADs and YAP/TAZ participate in transcriptional complexes with the transcriptional co-activator CBP/p300 (Fig. 4A, right panel). In human embryonic stem cells (hESC), TGF-ß is not only necessary to maintain pluripotency, but is also a key growth factor for mesendodermal differentiation (Oshimori and Fuchs, 2012). Mechanistically, in pluripotent hESCs, SMAD2/3 are associated on chromatin with the master transcription factors OCT4 and NANOG (Brown et al., 2011; Mullen et al., 2011). A mechanism underlying the switch that takes place from the moment where TGF-ß supports pluripotency to when it induces mesendodermal genes has been uncovered, that implicates Hippo pathway components (Beyer et al., 2013). In pluripotent ESCs, YAP/TAZ recruit the NuRD complex to repress the transcription of mesendodermal genes whose promoters are co-occupied by OCT4, SMAD2/3, FOXH1 and TEAD (Fig. 4B, left panel). Loss of binding of YAP/TAZ/TEAD that occurs during differentiation acts as a switch to promote mesendodermal gene expression via OCT4, SMAD2/3 and FOXH1 (Fig. 4B, right panel). Constitutive activation of the TGF-ß pathway together with defects in Hippo signaling upregulates the Connective Tissue Growth Factor (CTGF) gene expression which modulates malignant mesothelioma growth (Fujii et al., 2012). Adjacent SMAD and TEAD binding sites on the CTGF promoter recruit a complex containing SMAD3/TEAD4/YAP/p300. YAP knockdown strongly repressed CTGF expression yet expression TGF-ß target genes, including SMAD7, MMP2 or COL1A1, was not affected by the knockdown. This suggests high context specificity of the SMAD/YAP functional interaction and is to be linked to the existence of closely positioned cis-elements respectively binding SMAD3 and YAP within the CTGF promoter, not in the other target genes. Physical interaction between SMAD3 and YAP was weak and only detected upon overexpression of either of the two proteins. Endogenous SMAD3 and YAP binding could not be detected. Rather, SMAD3 appeared to bind TEAD4 more preferably than YAP and p300 was required to strengthen SMAD3/YAP complex formation. This is consistent with a previous study in which the authors identified a direct interaction between SMAD1 and YAP that regulates the transcription of BMP target genes during neural differentiation in mouse embryonic stem cells (Alarcon et al., 2009), while Smad3/YAP binding was extremely weak compared with SMAD1/YAP binding. Another study more recently found that YAP stabilizes SMAD1 and promotes BMP2-induced neocortical astrocytic differentiation in mice (Huang et al., 2016). Formation of SMAD2/3-YAP/TAZ complexes in response to TGF-ß may also depend upon cell types, as TGF-ß induced the formation of such complexes in HaCaT keratinocytes, not in the HT-29 cell line, as determined by means of proximity ligation assay (Grannas et al., 2015). Thus, not only are the SMAD3-YAP/TAZ interactions relatively weak, but subtle mechanisms yet to be identified, are likely necessary to promote their occurrence.