YAP integrates a plethora of extracellular signals that converge in the cytosol and routes them to downstream transcription factors and, in so doing, mediates their transcriptional activity in a diverse array of cellular processes in health and disease (3
). Despite such urgency, the molecular mechanism by which YAP recognizes its cellular partners remains largely elusive. Herein, our biophysical analysis provides new insights into YAP-WBP interaction at atomic level.
Our study shows that both WW domains of YAP2 recognize various PPXY motifs within WBP1 and WBP2 adaptors in structurally and thermodynamically indistinct manner. This is not surprising given that both WW domains share close to 50% sequence identity and, in particular, the residues within hydrophobic groove involved in sidehchain interactions with those from the PPXY motifs are virtually identical. It is noteworthy that the WW domains of dystrophin and Nedd4 have a strict requirement of non-consensus residues flanking the PPXY ligands for high-affinity binding and specificity (41
). The fact that non-consensus residues within and flanking the PPXY motifs within WBP proteins are not required for high-affinity binding of WW domains of YAP2 appears to be a unique feature of this WW-ligand interaction. It may well be that the WW domains of YAP2 have evolved to be extremely promiscuous and recognize a large repertoire of yet unidentified PPXY ligands. Importantly, motif search using Prosite at Expasy server suggests that of all the PPXY-containing family of proteins involved in the Hippo pathway (44
), only RUNX2 contains the PPXYXG motif, that we have identified here as the most optimal motif in WBP2 for binding to WW domains of YAP2. Interestingly, AMOT and p73, which are also involved in the Hippo pathway, contain the related PPXYXA motif that can also bind with high-affinity to WW domains of YAP2 ( and ). On the basis of these considerations, we predict that in addition to WBP2, other Hippo pathway proteins that may bind YAP with high affinity include RUNX2, AMOT and p73.
We also point out that the phosphorylation of signature tyrosine within the PPXY motif has been shown to negatively regulate WW-ligand interactions (9
). Although it is not known if the signature tyrosine within the PPXY motifs of WBP2 is also subject to phosphorylation, Y192 and Y231 within WBP2 have been shown to be phosphorylated in vivo (45
). In light of the foregoing argument, we believe that phosphorylation of signature tyrosine within the PPXY motifs of WBP proteins may serve as a molecular switch for the regulation of YAP-WBP interaction pertinent to its role in the Hippo pathway. Additionally, Hippo pathway PPXY-containing MST and LATS serine/threonine kinases have also been shown to negatively regulate YAP transcriptional activity (6
). It is thus clear that WBP proteins likely compete with MST and LATS for binding to YAP and their relative intracellular ratios are likely to determine whether YAP is activated or deactivated in response to various stimuli. Although WBP2 contains the PPXYXG motif for optimal binding to WW domains of YAP2 while MST and LATS do not, only a detailed analysis on full-length proteins can reveal whether WBP-YAP interaction is stronger or weaker than MST-YAP or LATS-YAP interactions due to other factors such as bivalent interactions and the formation of multimeric complexes.
Although our analysis on WW domains of YAP2 has been conducted in vitro with short peptides, our data are in agreement with previously published studies in which WBP1 and WBP2 were identified as putative YAP-binding partners (12
). Our cell-based data and in vitro pull down assays also suggest that full-length WBP1 and WBP2 bind to YAP in a specific manner (Buffa and Nawaz, unpublished observations). Additionally, the YAP-WBP interaction has also been demonstrated in Drosophila and shown to play a key role in the Hippo pathway (15
). It is noteworthy that the relatively low affinities in the tens to hundreds of micromolar observed for the binding of WW domains of YAP2 to PPXY motifs are characteristic of many WW-ligand interactions in general (19
). Importantly, these low-affinity interactions may underlie the ability of YAP to bind to WBP proteins as well as other cellular partners in a temporal and reversible manner — a scenario that is the hallmark of signaling cascades and regulatory networks that drive the cellular machinery. Given that we have relied here on short peptides to mimic PPXY motifs in WBP1 and WBP2, caution is warranted in that these motifs may depart from their physiological behavior when treated as short peptides due to the loss of local conformational constraints that they may be subject to in the context of full-length proteins. Nonetheless, the fact that both WBP1 and WBP2 contain multiple PPXY sites for the binding of WW domains of YAP2 raises the possibility for the formation of YAP2-WBP signaling complexes via a bidentate mechanism — that is both WW domains of YAP2 binding in a cooperative manner to two individual PPXY sites within WBP1 and WBP2. Such a scenario would clearly enhance the binding affinity of these partners due to entropic advantage and thereby enabling them to associate with each other at much lower cellular concentrations, perhaps in the submicromolar range in lieu of over tens of micromolar suggested by our measurements reported here. Accordingly, YAP2-WBP interaction may not only respond in a highly sensitive manner but may also be subject to fine tuning in response to specific extracellular stimuli.
In conclusion, we have provided here a biophysical framework for understanding a key WW-ligand interaction in the context of cellular signaling circuitry pertinent to health and disease. Our future efforts will focus on unraveling the mechanism of binding of the tandem WW domains of YAP2 to multivalent PPXY ligands derived from WBP proteins.