The oncoprotein YAP (Yes-associated protein) requires the TEAD family of transcription factors for the up-regulation of genes important for cell proliferation. Disrupting YAP–TEAD interaction is an attractive strategy for cancer therapy. Targeting TEADs using small molecules that either bind to the YAP-binding pocket or the palmitate-binding pocket is proposed to disrupt the YAP–TEAD interaction. There is a need for methodologies to facilitate robust and reliable identification of compounds that occupy either YAP-binding pocket or palmitate-binding pocket. Here, using NMR spectroscopy, we validated compounds that bind to these pockets and also identify the residues in mouse TEAD4 (mTEAD4) that interact with these compounds. Flufenamic acid (FA) was used as a positive control for validation of palmitate-binding pocket-occupying compounds by NMR. Furthermore, we identify a hit from a fragment screen and show that it occupies a site close to YAP-binding pocket on the TEAD surface. Our results also indicate that purified mTEAD4 can catalyze autopalmitoylation. NMR studies on mTEAD4 revealed that exchanges exist in TEAD as NMR signal broadening was observed for residues close to the palmitoylation site. Mutating the palmitoylated cysteine (C360S mutant) abolished palmitoylation, while no significant changes in the NMR spectrum were observed for the mutant which still binds to YAP. We also show that FA inhibits TEAD autopalmitoylation. Our studies highlight the utility of NMR spectroscopy in identifying small molecules that bind to TEAD pockets and reinforce the notion that both palmitate-binding pocket and YAP-binding pocket are targetable.

The Hippo pathway is a conserved signaling network and plays important roles in cancer progression by modulating apoptosis and cell proliferation [1,2]. In mammals, the final effectors of the Hippo pathway are the Yes-associated protein (YAP) and its paralog transcriptional co-activator with PDZ-binding motif (TAZ) [3]. Phosphorylation of several serine residues in YAP/TAZ through Hippo signaling prevents translocation of YAP/TAZ into the nucleus where they can bind to TEAD (Transcriptional enhancer factor with TEA/ATTS domain) family of transcription factors, to trigger the expression of genes important for cell proliferation [1,4]. Enhanced YAP/TAZ activities, due to mutations in the Hippo pathway components or due to other genetic alterations, have been frequently observed in several cancer types [5]. As TEADs are critical for YAP/TAZ activity [6], disrupting the interaction between TEAD and YAP is an attractive strategy to improve cancer therapy.

The human TEADs [14] are transcription factors without an activation domain. The N-terminal regions of the TEADs contain the TEA domain that binds to DNA [7], while the C-terminus contains a domain that binds to co-regulators like YAP and TAZ [8]. A functional transcription complex is only formed when the TEAD transcription factor interacts with co-regulators, such as YAP and TAZ [9]. Structural studies have been carried out for both free TEADs and their complexes with YAP [8,1012]. The YAP-binding domains (YBDs) of TEADs adopt very similar structure-an immunoglobulin-like fold, either in the absence or presence of the YAP protein [11,13,14]. YAP and TAZ are predominantly unstructured in their free form and therefore they are difficult to target. TEADs, on the other hand, may be druggable [15]. Structural studies have shown that the TEAD–YAP complex has three interfaces, among which interface 3 is more druggable than the others due to its deeper and more hydrophobic pocket. Mutating residues at this interface greatly affects YAP binding to TEAD [11,13].

Despite the challenge of identifying inhibitors that target the protein–protein interactions (PPIs) due to the moderate druggability of the interfaces [16], progress has been made in searching for compounds that are able to disrupt the YAP–TEAD interaction. For instance, YAP-like peptides have been developed to occupy the interface 3 [1719]. Using a thermal shift assay against a fragment library, a hit was identified to bind to the TEAD–YAP interface [20]. The drug verteporfin that is used in photodynamic therapy was also shown to inhibit YAP–TEAD interaction [21]. Recently, cysteine-dense peptides have also been optimized for disrupting YAP–TEAD interaction [22]. Although these inhibitors are unlikely to enter into the clinical phases as YAP–TEAD disruptors due to their poor drug-like properties, the studies paved the way for further drug development.

Apart from targeting the YAP-binding pocket on the TEAD surface, we proposed another strategy to disrupt the function of TEAD by developing inhibitors that target a more druggable pocket in the YBD of TEAD [15]. All the TEADs contain a conserved hydrophobic cavity in the center of the YBD. Recent structural studies revealed that this cavity is occupied by the hydrophobic part of palmitic acid [23,24]. Further functional analysis reveals that TEAD can catalyze palmitoylation on a cysteine residue. Disrupting palmitoylation has also been shown to affect YAP–TEAD interaction [24]. TEAD palmitoylation occurs on a conserved cysteine residue and this modification is critical for its function [24]. This palmitate-binding pocket is highly druggable and our previous study showed that flufenamic acid (FA) binds palmitate-binding pocket of TEAD2 and inhibits TEAD–YAP-dependent transcription [15].

X-ray crystal structures of TEADs and TEAD/YAP complex have provided valuable information to understand the protein complex and suggested strategies for the design of inhibitors. NMR structural studies will provide additional information to understand protein structure and dynamics in solution, which will be useful for understanding TEAD and small-molecule inhibitor interactions. So far, NMR studies have been carried out only on the YAP protein [14]. In this study, we carried out solution NMR and biophysical studies on the YBD of mouse TEAD4 (mTEAD4) that has high similarity to human TEAD4. We demonstrate that this domain contains autopalmitoylation activity and forms a folded structure in solution. Residues lining the palmitate-binding pocket undergo exchanges in solution. The mTEAD4 harboring mutation on the conserved cysteine residue abolishes palmitoylation but still binds to YAP peptide albeit with lower affinity. We demonstrate that small molecular mass compound FA can prevent mTEAD4 palmitoylation. We also identified a hit from a fragment screen that occupies a site close to YAP-binding pocket. We demonstrate that NMR can be used for the evaluation of compounds that disrupt the YAP–TEAD interaction.

Materials and methods

Preparation of mTEAD4

The cDNA encoding the YBD (R210 to E427) of mouse TEAD4 was synthesized (Genscript) and cloned into pET15b using NdeI and XhoI restriction enzymes. The resulting plasmid was transformed into Escherichia coli Bl21 (DE3) competent cells to express a recombinant protein (mTEAD4) with an N-terminal His6-tag, a thrombin cleavage site and YBD. To produce isotopically labeled proteins for NMR studies, the cells with the plasmid were grown in M9 medium supplied with 1 g/l 15NH4Cl, 2 g/l 13C-glucose or D2O. Protein induction was conducted at 18°C for 12–18 h in the presence of 1 mM IPTG when the optical density (OD600) of the cells reached 0.6–0.8. The cells harboring induced protein were then harvested by centrifugation at 8000×g and 4°C for 10 min. The cell pellet was collected and suspended in a buffer that contained 20 mM phosphate, pH 7.8, 0.5 M NaCl, and 2 mM β-mercaptoethanol. Cells were broken in an ice bath using sonication. The cell lysate was cleared by centrifugation at 40 000×g and 4°C for 20 min. The mTEAD4 was purified using Ni2+-NTA resin in a gravity column at room temperature. Protein was eluted in an elution buffer that contained 500 mM imidazole, pH 6.5, 1 M NaCl and 2 mM β-mercaptoethanol. For gel filtration chromatography, collected fractions were purified using a HiPrep 16/60 Sephacryl S-200HR column at 4°C in the NMR buffer that contained 20 mM sodium phosphate, pH7.4, 150 mM KCl, 0.5 mM EDTA and 1 mM DTT. Fractions were concentrated to 0.5–1.0 mM using an Amicon Centrifugal Filter unit with a molecular mass cutoff of 3 kDa for NMR data acquisition.

Preparation of YAP peptide

The cDNA of A48-P101 of human YAP was synthesized (Genscript) and cloned into pET32b vector using NcoI and XhoI restriction enzymes. The resulting plasmid was transformed into E. coli Bl21 (DE3) competent cells to express a protein with an N-terminal thioredoxin tag, an enterokinase site, and YAP's TEAD-binding motif. Recombinant protein was expressed and purified as aforementioned. Protein was cleaved using enterokinase (NEB) according to the manufacturers’ instruction. The peptide was purified by gel filtration and concentrated for NMR studies. The YAP peptide has an additional Ala residue at the N-terminus after enterokinase cleavage.

Backbone assignment

A 13C,15N, 2H-labeled mTEAD4 was prepared in the NMR buffer and used for data acquisition. The following transverse relaxation-optimized spectroscopy (TROSY) [25,26]-based experiments, namely 2D-1H-15N-HSQC, 3D HNCACB, HNCA, HN(CO)CA, HN(CO)CACB, HNCO, and NOESY-TROSY (mixing time 120 ms) experiments were collected at 37°C. The data were processed for the backbone resonance assignment [27]. All the experiments were carried out on a Bruker Avance spectrometer with the proton frequency of 600 or 700 MHz equipped with cryoprobe. For YAP, a 13C/15N-labeled sample was prepared in the NMR buffer. 2D-1H-15N-HSQC, 3D HNCACB, HNCA, HN(CO)CA, HN(CO)CACB were collected at 25°C for backbone resonance assignments. The data were acquired using Topspin (version 2.1) from Bruker and processed with NMRPipe [28], Topspin (2.1) and visualized using NMRView and CARA. The assignment was deposited in Biological Magnetic Resonance Data Bank with access code 12018. Secondary structure analysis of the protein was carried out using TALOS+ [29].

Protein and ligand-binding studies

Protein–protein/ligand interaction studies were carried out using 1H-15N-HSQC-type experiments. A 15N-labeled mTEAD4 or YAP was prepared at a concentration 0.3 mM. For PPI experiments, 15N-labeled mTEAD4 or YAP was mixed with equal molar amount of unlabeled YAP or mTEAD4. 1H-15N-HSQC spectra of labeled protein in the absence or presence of the binding partner were collected and compared. Inhibitors were first dissolved in DMSO to make a stock solution before addition to a 15N-labeled mTEAD4. 1H-15N-HSQC spectrum of mTEAD4 in the presence of 1.2 mM of inhibitor was collected. Protein mixed with equal amount of DMSO was used as the reference sample for data collection. The 1H-15N-HSQC spectra of mTEAD4 in the presence of DMSO and inhibitors were compared to understand the interactions.

Palmitoylation of mTEAD4

Recombinant mTEAD4 was purified in the NMR buffer as aforementioned. Palmitoyl-coenzyme A (Palm-CoA) was dissolved in DMSO as a stock solution and mTEAD4 was diluted to 30 µM. Different amounts of Palm-CoA were mixed with mTEAD4. The mixtures were kept at room temperature for 1 h and subjected to mass spectrometry (MS) analysis using the method as previously described [30]. For the FA inhibition assay, FA was first mixed with Palm-CoA, the mixture was then added into the mTEAD4 sample. After 2 h incubation, the samples were subjected to MS analysis. All MS analyses were performed as duplicates and the representative data are shown in the figure.

Fragment screening using thermal shift assay

Thermal shift experiments were carried out using SYPRO Orange dye on a Roche LC480 PCR machine. A white 384-well plate was used in the assay, in which each well contained 6 μM TEAD and 20x SYPRO Orange dye. The samples were subjected to incremental temperature increases from 30 to 95°C. The assay buffer contained 20 mM HEPES, pH 7.5 and 250 mM NaCl. A total of 1685 fragments were screened at 2.5 mM resulting in a hit rate of 3% (ΔTm ≥ 0.5°C).

19F NMR spectroscopy

19F NMR spectroscopy was carried out on a Bruker 400 MHz equipped with a BBO probe. The experiment was carried out at 27°C. Standard 1D 19F NMR spectroscopy was acquired and processed using Topspin (3.5). FA was first dissolved in DMSO to make a stock solution at 20 mM concentration. The 19F NMR spectroscopy spectra of 1.5 mM FA in the absence and presence of 0.4 mM mTEAD4 were acquired and compared.

Isothermal titration calorimetry experiments

ITC (isothermal titration calorimetry) experiment was carried out using the method described previously [31]. Briefly, ITC experiment was performed on an Auto-iTC200 instrument (Microcal Inc.) at 25°C. Purified mTEAD4 or mTEAD4 C360S mutant was prepared in the NMR buffer to a final concentration of 10 µM for YAP-binding and 100 µM for FA-binding studies. FA or chemical synthesized YAP peptide (SETDLEALFNAVMNPKTANVPQTVPMRLRKLPDSFFKPPE) was dissolved in the same buffer. YAP or FA was loaded into the syringe automatically. The titration experiments were carried out with 18 injections over a period of 40 min with stirring at 1000 rpm. Each experiment was done twice to verify the results are consistent. The binding was analyzed using single site-binding mode. The dissociation constant (Kd) and other parameters were determined using Origin provided with the instrument.

Results

Solution NMR spectroscopy of mTEAD4

It is challenging to study the structure of the full length of TEADs using NMR spectroscopy, as they are predominantly insoluble when expressed in E. coli. Therefore, we expressed and purified human and mTEAD4 that contain only the functional YBDs. As sequences of both proteins share high similarities with only nine different residues (Supplementary Figure S1), we carried out our studies on mTEAD4 due to its higher yield in E. coli. Isotopically enriched mTEAD4 can be prepared in milligram quantities for NMR studies and we successfully collected the 1H-15N-HSQC spectrum (Figure 1). For most proteins with β-sheet structures, the cross peaks in the 1H-15N-HSQC spectrum are normally well spread and the peak intensities of residues in the structured region are generally similar [32]. Interestingly, the number of peaks for mTEAD4 in the 1H-15N-HSQC spectrum is lower than expected and the peak intensities of the residues are different (Figure 1). This may be due to protein aggregation, misfolding, exchanges arising from protein–ligand interactions or conformational exchanges. However, our gel filtration chromatography indicated that the protein is mainly monomeric in solution (Supplementary Figure S2).

NMR spectrum of mTEAD4.

Figure 1.
NMR spectrum of mTEAD4.

(A) Assignment of the 1H-15N-HSQC spectrum of mTEAD4. The assigned peaks are labeled with residue name and sequence number. (B) Structure of mTEAD4. The crystal structure of mTEAD4 (PDB ID 3JUA) is shown. The assigned and unassigned residues are shown in wheat and green, respectively. The palmitoyl group is highlighted as yellow spheres. (C) Secondary structure of mTEAD4 in solution. The secondary structure of mTEAD4 was analyzed based on the assigned backbone resonances using TALOS+. For assigned residues, the helical, sheet, and random coil structures are shown as boxes, arrows, and lines, respectively. The residues that from structured regions and do not exhibit cross peaks in the 1H-15N-HSQC spectrum, are shown as white arrows or boxes. A yellow arrow indicates the strand that is only identified in NMR studies. The secondary structures of mTEAD4 from X-ray structures are indicated in the sequence. Residues that are in helices and strands are highlighted in blue and red, respectively.

Figure 1.
NMR spectrum of mTEAD4.

(A) Assignment of the 1H-15N-HSQC spectrum of mTEAD4. The assigned peaks are labeled with residue name and sequence number. (B) Structure of mTEAD4. The crystal structure of mTEAD4 (PDB ID 3JUA) is shown. The assigned and unassigned residues are shown in wheat and green, respectively. The palmitoyl group is highlighted as yellow spheres. (C) Secondary structure of mTEAD4 in solution. The secondary structure of mTEAD4 was analyzed based on the assigned backbone resonances using TALOS+. For assigned residues, the helical, sheet, and random coil structures are shown as boxes, arrows, and lines, respectively. The residues that from structured regions and do not exhibit cross peaks in the 1H-15N-HSQC spectrum, are shown as white arrows or boxes. A yellow arrow indicates the strand that is only identified in NMR studies. The secondary structures of mTEAD4 from X-ray structures are indicated in the sequence. Residues that are in helices and strands are highlighted in blue and red, respectively.

Secondary structural analysis of mTEAD4

We then assigned the cross peaks in the 1H-15N-HSQC spectrum using conventional heteronuclear experiments (Figure 1A). We could not assign all the residues from mTEAD4 as some residues close to the palmitoylation site (C360) and the other residues close to the YAP-binding pocket did not exhibit detectable cross peaks in the 1H-15N-HSQC spectrum (Figure 1B). Previous studies have shown that even the bacterially expressed TEADs are palmitoylated. Therefore, the missing peaks in the 1H-15N-HSQC spectrum might be due to the exchanges caused by the presence of palmitate in the pocket. With the backbone assignment in hand, we analyzed the secondary structures for residues with cross peaks in the spectra based on the assigned backbone resonance chemical shifts including Cα, N, HN and C’ using TALOS+ [29]. Overall, the secondary structure of mTEAD4 in solution is similar to that obtained by X-ray crystallography (Figure 1C) except for some minor differences. In the structured regions, residues from β4, β7, parts of β10 and β11, and N-terminal part of α3 do not exhibit cross peaks in the 1H-15N-HSQC spectrum. The secondary structure analysis still suggests that mTEAD4 forms a well-folded structure in solution despite the missing assignments which might be caused by structural exchanges (Figure 1C).

Purified mTEAD4 is active in solution

Recent studies have shown that TEADs can catalyze autopalmitoylation without the need of any palmitoyl transferases [23,24]. Structural studies have revealed that the side chain of cysteine residue (C360 of mTEAD4) forms a thioester linkage with the carboxylic acid of the palmitate (Figure 2A). We evaluated the autopalmitoylation activity of mTEAD4 using palmitoyl-coenzyme A (Palm-CoA) as a substrate. Recombinant mTEAD4 was partially palmitoylated when it was produced in E. coli because two species, corresponding to free and palmitoylated forms of mTEAD4 were observed in the MS spectra (Figure 2A). In the presence of Palm-CoA, the population of the palmitoylated mTEAD4 was increased in a concentration-dependent manner. This result is consistent with previous reports that TEAD can be autopalmitoylated [24]. We then made point mutations on C360 of mTEAD4. Mutations of cysteine to Ala or Leu resulted in proteins that had very low solubility in E. coli and could not be purified (Supplementary Figure S2). However, the protein with cysteine to serine (C360S mutant) could be produced and purified from E. coli (Supplementary Figure S2). As expected, the mutant exhibited only one major peak in the MS spectrum corresponding to the molecular mass of the free protein (Figure 2B). This indicates that C360 is the site of palmitoylation. We also collected the 1H-15N-HSQC spectrum of the C360S mutant (Figure 2D). Both wild-type and C360S mutant exhibited very similar 1H-15N-HSQC spectra except that a few more peaks (∼5%) appear in the spectrum of C360S mutant (Figure 2D), demonstrating that point mutation does not alter the overall structure of TEAD. In addition, the exchanges that can cause line broadening of residues at the hydrophobic cavity were not suppressed as the additional peaks (∼5%) in the spectrum of C360S do not correspond to the residues at the hydrophobic cavity (Figure 2C). The similarity between the NMRs of the wild-type and mutant protein may be explained by the interaction of non-covalently linked palmitic acid with the hydrophobic pocket. In fact, in some crystal structures of TEAD, free palmitate is observed occupying the palmitate-binding pocket [24].

Palmitoylation of mTEAD4.

Figure 2.
Palmitoylation of mTEAD4.

(A) Palmitoylation of mTEAD4 in the presence of Palm-CoA. Mass spectrometry (MS) was used to analyze protein molecular mass. Different amounts of Palm-CoA were mixed with mTEAD4 and subjected to MS analysis. (B) MS spectrum of purified mTEAD4 C360S mutant. (C) Cartoon depicting the TEAD autopalmitoylation reaction. (D) An overlay of the 1H-15N-HSQC spectra of mTEAD (black) and C360S (red) mutant.

Figure 2.
Palmitoylation of mTEAD4.

(A) Palmitoylation of mTEAD4 in the presence of Palm-CoA. Mass spectrometry (MS) was used to analyze protein molecular mass. Different amounts of Palm-CoA were mixed with mTEAD4 and subjected to MS analysis. (B) MS spectrum of purified mTEAD4 C360S mutant. (C) Cartoon depicting the TEAD autopalmitoylation reaction. (D) An overlay of the 1H-15N-HSQC spectra of mTEAD (black) and C360S (red) mutant.

The mTEAD4 binds to peptide derived from YAP

To assess whether chemical shift perturbations could be used to probe interaction at the YAP-binding pocket, we tested the binding between the mTEAD4 and a peptide derived from the TEAD-binding motif of human YAP. This peptide contains residues A48-P101 of human YAP and binds to TEAD proteins. Using ITC, the dissociation constant (Kd) between TEAD and YAP was determined to be 77 nM and the stoichiometry is 1:0.9. The binding constant is different from reported result [14], which might be due to the different constructs used in the studies (Figure 3A). The 1H-15N-HSQC spectra of 15N-labeled mTEAD4 in the absence and presence of unlabeled YAP were collected. Overlaid spectra reveal that some residues exhibited chemical shift perturbations and line broadening upon YAP binding (Figure 3B). The broadened cross peaks of residues at the palmitate-binding pocket did not appear in the presence of the YAP peptide, implying that YAP binding does not affect the exchanges at the residues surrounding the palmitoylation site. We also obtained isotopically labeled YAP peptide and conducted backbone assignment (Supplementary Figure S3 and Figure 3C). As predicted, most residues from the YAP peptide are unstructured except that residues 64–69 have a tendency to form a helical structure (Figure 3D). This region was shown to be helical in the crystal structure of TEAD–YAP complex [13]. The 1H-15N-HSQC spectra of 15N-labeled YAP peptide in the absence and presence of unlabeled mTEAD4 were also collected and compared (Figure 3C). Signals of most residues from the YAP peptide are affected upon TEAD binding, suggesting that inhibiting this interaction with a small molecular mass compound might be challenging as most YAP resides involves in TEAD binding.

mTEAD4 binds to YAP peptide.

Figure 3.
mTEAD4 binds to YAP peptide.

(A) ITC analysis of YAP binding to mTEAD4. (B) 1H-15N-HSQC spectra of 15N-labeled mTEAD4 in the absence (black) and presence (red) of equal molar concentration of YAP peptide. The amino acid residues with chemical shift perturbations upon YAP-binding are indicated. (C) 1H-15N-HSQC spectra of 15N-labeled YAP construct in the absence (black) and presence (red) of equal molar amount of mTEAD4. The indicated residues undergo chemical shift change upon TEAD addition. (D) The crystal structure of TEAD–YAP complex (PDB ID 3JUA) is shown. Residues from mTEAD4 that exhibited obvious chemical shift changes upon binding to YAP are shown in spheres. YAP is shown in red and its residues are helical in solution are highlighted in cyan.

Figure 3.
mTEAD4 binds to YAP peptide.

(A) ITC analysis of YAP binding to mTEAD4. (B) 1H-15N-HSQC spectra of 15N-labeled mTEAD4 in the absence (black) and presence (red) of equal molar concentration of YAP peptide. The amino acid residues with chemical shift perturbations upon YAP-binding are indicated. (C) 1H-15N-HSQC spectra of 15N-labeled YAP construct in the absence (black) and presence (red) of equal molar amount of mTEAD4. The indicated residues undergo chemical shift change upon TEAD addition. (D) The crystal structure of TEAD–YAP complex (PDB ID 3JUA) is shown. Residues from mTEAD4 that exhibited obvious chemical shift changes upon binding to YAP are shown in spheres. YAP is shown in red and its residues are helical in solution are highlighted in cyan.

The mTEAD4 C360S mutant binds to YAP

The interaction between mTEAD4 C360S and YAP was then evaluated using both NMR and ITC (Figure 4). Under the same conditions as those of the wild-type, an overlay of the 1H-15N-HSQC spectra of mTEAD4 C360S mutant in the absence and presence of YAP peptide indicates that the C360S mutant is able to interact with YAP, while fewer residues from mTEAD4 were affected in the presence of equal molar of YAP. The Kd was determined to be 162 nM by ITC. We infer that the mutation at the palmitoylation site does not abolish YAP binding although the binding affinity is not the same as that of the wild-type. A previous study demonstrated that palmitoylation is required for YAP binding [24]. Here, we found that the Kd was reduced when the palmitoylation site was removed. It is likely that free palmitate might still be occupying the palmitate-binding pocket as the missing cross peaks in mTEAD4 wild-type were not detected in the mutant form. Our result is also in line with the recently published report indicating that mutating the palmitoylation site does not inhibit YAP binding to TEAD [33].

C360S mutant binds to YAP.

Figure 4.
C360S mutant binds to YAP.

(A) An overlay of the 1H-15N-HSQC spectra of mTEAD4 C360S in the absence (black) and presence (red) of equal molar YAP peptide. (B) ITC analysis of YAP binding to mTEAD4 C360S.

Figure 4.
C360S mutant binds to YAP.

(A) An overlay of the 1H-15N-HSQC spectra of mTEAD4 C360S in the absence (black) and presence (red) of equal molar YAP peptide. (B) ITC analysis of YAP binding to mTEAD4 C360S.

FA binding to mTEAD4 can inhibit palmitoylation

We previously reported that FA that binds to the palmitate-binding pocket of human TEADs and inhibits its function in a cell-based assay [15]. As TEADs are structurally similar and contain palmitate-binding pocket, it was not surprising to find that FA also bound to mTEAD4 (Figure 5A). The residues surrounding the palmitate-binding pocket exhibit line broadening upon FA binding (Figure 5B). No additional peaks appear in the 1H-15N-HSQC spectrum of mTEAD4-FA complex, suggesting that FA binding does not suppress the exchanges observed in mTEAD4. As FA contains three F atoms, 1D 19F NMR was collected for FA in the absence and presence of mTEAD4 protein. Obvious line broadening and chemical shift perturbation was observed, confirming the binding of FA to mTEAD4 in solution (Figure 5C). ITC study revealed that mTEAD4 binds to FA with a Kd of 69 µM (Figure 5D). We then evaluated whether FA can affect mTEAD4 palmitoylation in the presence of Palm-CoA. To this end, we mixed mTEAD4 with Palm-CoA and FA and used MS as a tool to evaluate TEAD palmitoylation. We observed that increasing the FA concentration correlated to an increase in the amount of depalmitoylated mTEAD4 (Figure 5E) suggesting that FA interferes with TEAD palmitoylation. However, we did observe inhibition only when the molar ratio of FA to Palm-CoA is more than 10 (Figure 5E). At lower molar ratios, no obvious inhibition on palmitoylation was observed (data not shown). Nonetheless, this result clearly indicates that FA binding to mTEAD4 inhibits TEAD palmitoylation.

FA binds to mTEAD4 and inhibits palmitoylation.

Figure 5.
FA binds to mTEAD4 and inhibits palmitoylation.

(A) 1H-15N-HSQC spectra of 15N-labeled mTEAD4 in the absence (black) and presence of 4-fold excess FA (red). Residues affected by FA binding are labeled. (B) Residues affected by FA binding are mapped onto the structure of mTEAD4. The cysteine residue that can be modified by palmitoylation is shown in blue. Affected residues from mTEAD4 due to FA binding are shown in red spheres. FA is modeled into the structure of mTEAD4 based on the crystal structure of TEAD2-FA complex (PDB ID 5DQ8). (C) 1D 19F NMR of FA in the absence (blue) and presence of mTEAD4 (red). (D) ITC analysis of FA binding to mTEAD4. (E) MS analysis of palmitoylation of mTEAD4 in the presence of FA.

Figure 5.
FA binds to mTEAD4 and inhibits palmitoylation.

(A) 1H-15N-HSQC spectra of 15N-labeled mTEAD4 in the absence (black) and presence of 4-fold excess FA (red). Residues affected by FA binding are labeled. (B) Residues affected by FA binding are mapped onto the structure of mTEAD4. The cysteine residue that can be modified by palmitoylation is shown in blue. Affected residues from mTEAD4 due to FA binding are shown in red spheres. FA is modeled into the structure of mTEAD4 based on the crystal structure of TEAD2-FA complex (PDB ID 5DQ8). (C) 1D 19F NMR of FA in the absence (blue) and presence of mTEAD4 (red). (D) ITC analysis of FA binding to mTEAD4. (E) MS analysis of palmitoylation of mTEAD4 in the presence of FA.

Fragment interacts closely to YAP–TEAD-binding interface

With purified mTEAD4 in hand, we also carried out fragment screening using thermal shift assays. We screened a compound library with overall 1685 fragments. Several fragment hits were identified (Supplementary Figure S4). NMR was used to verify the molecular interaction of the most promising fragment hit (ΔTm = 1.8°C in the thermal shift assay) with mTEAD4. An overlay of the 1H-15N-HSQC spectra reveals obvious signal changes on mTEAD4 upon fragment binding (Figure 6A). The affected residues are mapped on the surface of mTEAD4 (Figure 6A). Most of the residues are close to the interface 3 binding site on the YAP-binding pocket (Figure 6B and Supplementary Figure S5). Based on the identified residues with chemical shift changes upon fragment binding, High Ambiguity Driven biomolecular DOCKing (HADDOCK) [34] was carried out. Among 12 clusters obtained using HADDOCK, one cluster fits well with the observed chemical shift changes in the NMR study (Figure 6C). It is likely that the fragment binds to the cleft between α1 and α2 and sits closely to the third binding site of YAP-binding pocket (Figure 6D). However, this fragment was not potent enough to disrupt the TEAD interaction with YAP.

A fragment hit was identified to bind to the surface of mTEAD4.

Figure 6.
A fragment hit was identified to bind to the surface of mTEAD4.

(A) 1H-15N-HSQC spectra of 15N-labeled mTEAD4 in the absence (black) and presence of 4-fold excess of fragment (red). The chemical structure of the fragment is shown. (B) Residues affected by fragment binding are mapped onto the surface of mTEAD4 (PDB ID 3JUA). The residues affected by fragment binding are highlighted in blue. YAP is shown in red and mTEAD4 is shown as a surface. (C) One cluster of mTEAD-fragment complex obtained using HADDOCK. The mTEAD4 is shown in wheat and fragment is shown in green. The YAP peptide is shown by superposing the crystal structure of mTEAD4-YAP complex (PDB ID 3JUA) with that of mTEAD4-fragment complex. The structure of the fragment was prepared as described previously [35,36]. (D) A closer view of the fragment-binding site on mTEAD4. The fragment is shown in green sticks. The amide atoms of several residues from mTEAD4 exhibiting chemical shift changes upon fragment binding are shown in blue spheres. the residue and sequence numbers are also shown.

Figure 6.
A fragment hit was identified to bind to the surface of mTEAD4.

(A) 1H-15N-HSQC spectra of 15N-labeled mTEAD4 in the absence (black) and presence of 4-fold excess of fragment (red). The chemical structure of the fragment is shown. (B) Residues affected by fragment binding are mapped onto the surface of mTEAD4 (PDB ID 3JUA). The residues affected by fragment binding are highlighted in blue. YAP is shown in red and mTEAD4 is shown as a surface. (C) One cluster of mTEAD-fragment complex obtained using HADDOCK. The mTEAD4 is shown in wheat and fragment is shown in green. The YAP peptide is shown by superposing the crystal structure of mTEAD4-YAP complex (PDB ID 3JUA) with that of mTEAD4-fragment complex. The structure of the fragment was prepared as described previously [35,36]. (D) A closer view of the fragment-binding site on mTEAD4. The fragment is shown in green sticks. The amide atoms of several residues from mTEAD4 exhibiting chemical shift changes upon fragment binding are shown in blue spheres. the residue and sequence numbers are also shown.

Discussion

TEAD–YAP interaction is important for the activation of TEAD transcription and for transcribing genes that are important for cell proliferation. Disrupting their interaction has been considered as a strategy in cancer therapy [810]. Based on the available crystal structures, it has been proposed that the YAP-binding pocket, particularly the interface 3 is a feasible target for small molecules. However, in general, it is challenging to develop inhibitors that disrupt PPIs. PPI interfaces are generally shallow, which make it difficult to design potent small molecules that bind with high affinity. Despite the challenge, inhibitors have been developed to disrupt PPIs and some successful ones have been approved for clinical applications [16,3740]. Fragment-based drug discovery was shown to be a useful tool to develop potent inhibitors targeting such low druggable targets. Although the YAP-binding pockets on TEADs are shallow, our current study and a recent report [20] indicate that it is feasible to develop small molecules that bind to the YAP-binding pocket to disrupt the YAP–TEAD interaction. In both the studies, fragment hits were identified using the thermal shift assay performed on TEADs. Kann et al. [20] identified in the crystal structure that the phenyl ring from their fragment hit was found to bind to the hydrophobic groove in the YAP-binding pocket, interacting with the Phe 69 residue [20]. This binding site is labeled as interface 2 of the YAP-binding domain [11]. In this study, we have identified a fragment hit that binds to a site close to the interface 3 (Figure 6) of the YAP-binding pocket. Although further growth of the fragments is required to obtain potent compounds, these studies suggest that it might be possible to develop inhibitors that can disrupt TEAD–YAP interactions.

There are conflicting reports about the influence of palmitoylation on YAP–TEAD interaction. One study indicated that it dramatically affects YAP–TEAD interaction [24], whereas another study indicated that the interaction is unaffected by palmitoylation [33]. Based on our data, we conclude that even in the absence of palmitoylation, TEAD's ability to interact with YAP is retained (Figure 4). We show that FA binds to mTEAD4 in solution and the FA binding can inhibit TEAD palmitoylation (Figure 5). In addition, our solution NMR study suggests that exchanges exist for residues lining the palmitate-binding pocket giving rise to missing cross peaks of several residues, which might be due to presence of endogenous palmitate in palmitate-binding pocket. Mutation of the conserved cysteine residue (C360) abolished palmitoylation through thioester linkage (Figure 2B). The 1H-15N-HSQC spectra of wild-type and the mutant were very similar, suggesting that the palmitate-binding pocket might still be occupied by free palmitate in the mutant. Nevertheless, palmitate-binding pocket might still be an attractive drug target as it can be leveraged to disrupt TEADs’ stability or function. This point can only be answered once a potent tool compound is available that not only inhibit autopalmitoylation but also compete against the binding of free palmitic acid in the palmitate-binding pocket. In the present study, we have shown that it is possible to use NMR to identify small-molecule compounds that target the palmitate-binding pocket.

YAP–TEAD binding can be easily monitored using 1H-15N-HSQC-type experiment (Figure 3), rendering NMR as a tool for evaluating inhibitors that can disrupt the YAP–TEAD complex [41]. A fragment binding to a site close to the interface 3 of the YAP-binding pocket was identified in this study that could be developed into an inhibitor for disrupting the YAP–TEAD interaction. We have also carried out NMR studies in mTEAD4 and have observed exchanges for residues lining the conserved palmitate-binding pocket. In addition, we demonstrate that FA binding to palmitate-binding pocket inhibits palmitoylation. Our methodology could be used for identifying small molecules that occupy the palmitate-binding pocket or inhibit TEAD palmitoylation. Although we cannot obtain binding affinities of these compounds to mTEAD4 by NMR, as the binding is undergoing intermediate exchange, NMR is still a useful tool for screening and verification of the small molecules’ binding to TEAD–YAP-binding pocket or palmitate-binding pocket. Our studies provide additional evidence that it is feasible to develop inhibitors targeting either the palmitate binding or the YAP- binding pocket of TEAD.

Abbreviations

     
  • FA

    flufenamic acid

  •  
  • mTEAD4

    YAP binding domain of mouse TEAD4

  •  
  • PPI

    protein–protein interaction

  •  
  • TROSY

    transverse relaxation-optimized spectroscopy

  •  
  • YAP

    Yes-associated protein

  •  
  • YBD

    YAP-binding domain

Author Contribution

Y.L. carried out and analyzed NMR experiments. E.Y.N. purified protein and carried out palmitoylation experiments. R.L. carried out MS analysis. S.L. and A.W.H. carried out and analyzed ITC experiment, thermal shift assay, and fragment screening. A.P. designed the FA compounds. T.H.K. designed compound screening and synthesis strategy. J.H. guided the MS and biochemical assay. A.V.P. and W.H. guided protein production and involved in experimental design. C.K. designed the experiments and drafted the manuscript. All the authors revised and approved the manuscript.

Funding

A*STAR JCO grant (1431AFG102/1331A028) to C.K.

Acknowledgments

We thank Prof. Ho Sup Yoon and Dr Hong Ye from Nanyang Technological University for the NMR experiments. The authors appreciated the support from BMRC, A*STAR.

Competing Interests

The Authors declare that there are no competing interests associated with the manuscript.

References

References
1
Halder
,
G.
, and
Johnson
,
R.L.
(
2011
)
Hippo signaling: growth control and beyond
.
Development
138
,
9
22
2
Fu
,
V.
,
Plouffe
,
S.W.
and
Guan
,
K.-L.
(
2018
)
The Hippo pathway in organ development, homeostasis, and regeneration
.
Curr. Opin. Cell Biol.
49
,
99
107
3
Yu
,
F.-X.
and
Guan
,
K.-L.
(
2013
)
The Hippo pathway: regulators and regulations
.
Genes Dev.
27
,
355
371
4
Santucci
,
M.
,
Vignudelli
,
T.
,
Ferrari
,
S.
,
Mor
,
M.
,
Scalvini
,
L.
,
Bolognesi
,
M.L.
et al. 
(
2015
)
The hippo pathway and YAP/TAZ–TEAD protein–protein interaction as targets for regenerative medicine and cancer treatment
.
J. Med. Chem.
58
,
4857
4873
5
Zanconato
,
F.
,
Cordenonsi
,
M.
and
Piccolo
,
S.
(
2016
)
YAP/TAZ at the roots of cancer
.
Cancer Cell
29
,
783
803
6
Yin
,
M.
and
Zhang
,
L.
(
2011
)
Hippo signaling: a hub of growth control, tumor suppression and pluripotency maintenance
.
J. Genet. Genomics
38
,
471
481
7
Hwang
,
J.J.
,
Chambon
,
P.
and
Davidson
,
I.
(
1993
)
Characterization of the transcription activation function and the DNA binding domain of transcriptional enhancer factor-1
.
EMBO J.
12
,
2337
2348
PMID:
[PubMed]
8
Pobbati
,
A.V.
,
Chan
,
S.W.
,
Lee
,
I.
,
Song
,
H.
and
Hong
,
W.
(
2012
)
Structural and functional similarity between the Vgll1-TEAD and the YAP–TEAD complexes
.
Structure
20
,
1135
1140
9
Pobbati
,
A.V.
and
Hong
,
W.
(
2013
)
Emerging roles of TEAD transcription factors and its coactivators in cancers
.
Cancer Biol. Ther.
14
,
390
398
10
Cairns
,
L.
,
Tran
,
T.
and
Kavran
,
J.M.
(
2017
)
Structural insights into the regulation of hippo signaling
.
ACS Chem. Biol.
12
,
601
610
11
Li
,
Z.
,
Zhao
,
B.
,
Wang
,
P.
,
Chen
,
F.
,
Dong
,
Z.
,
Yang
,
H.
et al. 
(
2010
)
Structural insights into the YAP and TEAD complex
.
Genes Dev.
24
,
235
240
12
Kaan
,
H.Y.K.
,
Chan
,
S.W.
,
Tan
,
S.K.J.
,
Guo
,
F.
,
Lim
,
C.J.
,
Hong
,
W.
et al. 
(
2017
)
Crystal structure of TAZ–TEAD complex reveals a distinct interaction mode from that of YAP–TEAD complex
.
Sci. Rep.
7
,
2035
13
Chen
,
L.
,
Chan
,
S.W.
,
Zhang
,
X.
,
Walsh
,
M.
,
Lim
,
C.J.
,
Hong
,
W.
et al. 
(
2010
)
Structural basis of YAP recognition by TEAD4 in the hippo pathway
.
Genes Dev.
24
,
290
300
14
Tian
,
W.
,
Yu
,
J.
,
Tomchick
,
D.R.
,
Pan
,
D.
and
Luo
,
X.
(
2010
)
Structural and functional analysis of the YAP-binding domain of human TEAD2
.
Proc. Natl Acad. Sci. U.S.A.
107
,
7293
7298
15
Pobbati
,
A.V.
,
Han
,
X.
,
Hung
,
A.W.
,
Weiguang
,
S.
,
Huda
,
N.
,
Chen
,
G.-Y.
et al. 
(
2015
)
Targeting the central pocket in human transcription factor TEAD as a potential cancer therapeutic strategy
.
Structure
23
,
2076
2086
16
Cierpicki
,
T.
and
Grembecka
,
J.
(
2015
)
Targeting protein-protein interactions in hematologic malignancies: still a challenge or a great opportunity for future therapies?
Immunol. Rev.
263
,
279
301
17
Zhou
,
Z.
,
Hu
,
T.
,
Xu
,
Z.
,
Lin
,
Z.
,
Zhang
,
Z.
,
Feng
,
T.
et al. 
(
2015
)
Targeting Hippo pathway by specific interruption of YAP–TEAD interaction using cyclic YAP-like peptides
.
FASEB J.
29
,
724
732
18
Zhang
,
Z.
,
Lin
,
Z.
,
Zhou
,
Z.
,
Shen
,
H.C.
,
Yan
,
S.F.
,
Mayweg
,
A.V.
et al. 
(
2014
)
Structure-based design and synthesis of potent cyclic peptides inhibiting the YAP–TEAD protein–protein interaction
.
ACS Med. Chem. Lett.
5
,
993
998
19
Jiao
,
S.
,
Wang
,
H.
,
Shi
,
Z.
,
Dong
,
A.
,
Zhang
,
W.
,
Song
,
X.
et al. 
 (
2014
)
A peptide mimicking VGLL4 function acts as a YAP antagonist therapy against gastric cancer
.
Cancer Cell
25
,
166
180
20
Kaan
,
H.Y.K.
,
Sim
,
A.Y.L.
,
Tan
,
S.K.J.
,
Verma
,
C.
and
Song
,
H.
(
2017
)
Targeting YAP/TAZ-TEAD protein-protein interactions using fragment-based and computational modeling approaches
.
PLoS ONE
12
,
e0178381
21
Liu-Chittenden
,
Y.
,
Huang
,
B.
,
Shim
,
J.S.
,
Chen
,
Q.
,
Lee
,
S.-J.
,
Anders
,
R.A.
et al. 
(
2012
)
Genetic and pharmacological disruption of the TEAD-YAP complex suppresses the oncogenic activity of YAP
.
Genes Dev.
26
,
1300
1305
22
Crook
,
Z.R.
,
Sevilla
,
G.P.
,
Friend
,
D.
,
Brusniak
,
M.-Y.
,
Bandaranayake
,
A.D.
,
Clarke
,
M.
et al. 
(
2017
)
Mammalian display screening of diverse cystine-dense peptides for difficult to drug targets
.
Nat. Commun.
8
,
2244
23
Noland
,
C.L.
,
Gierke
,
S.
,
Schnier
,
P.D.
,
Murray
,
J.
,
Sandoval
,
W.N.
,
Sagolla
,
M.
et al. 
(
2016
)
Palmitoylation of TEAD transcription factors is required for their stability and function in hippo pathway signaling
.
Structure
24
,
179
186
24
Chan
,
P.
,
Han
,
X.
,
Zheng
,
B.
,
DeRan
,
M.
,
Yu
,
J.
,
Jarugumilli
,
G.K.
et al. 
(
2016
)
Autopalmitoylation of TEAD proteins regulates transcriptional output of the Hippo pathway
.
Nat. Chem. Biol.
12
,
282
289
25
Pervushin
,
K.
,
Ono
,
A.
,
Fernandez
,
C.
,
Szyperski
,
T.
,
Kainosho
,
M.
and
Wuthrich
,
K.
(
1998
)
NMR scalar couplings across Watson-Crick base pair hydrogen bonds in DNA observed by transverse relaxation-optimized spectroscopy
.
Proc. Natl Acad. Sci. U.S.A.
95
,
14147
14151
26
Salzmann
,
M.
,
Pervushin
,
K.
,
Wider
,
G.
,
Senn
,
H.
and
Wuthrich
,
K.
(
1998
)
TROSY in triple-resonance experiments: new perspectives for sequential NMR assignment of large proteins
.
Proc. Natl Acad. Sci. U.S.A.
95
,
13585
13590
27
Li
,
Y.
,
Wong
,
Y.L.
,
Lee
,
M.Y.
,
Ng
,
H.Q.
and
Kang
,
C.B.
(
2016
)
Backbone assignment of the N-terminal 24-kDa fragment of Escherichia coli topoisomerase IV ParE subunit
.
Biomol. NMR Assign.
10
,
135
138
28
Delaglio
,
F.
,
Grzesiek
,
S.
,
Vuister
,
G.W.
,
Zhu
,
G.
,
Pfeifer
,
J.
and
Bax
,
A.
(
1995
)
NMRPipe: a multidimensional spectral processing system based on UNIX pipes
.
J. Biomol. NMR
6
,
277
293
29
Shen
,
Y.
,
Delaglio
,
F.
,
Cornilescu
,
G.
and
Bax
,
A.
(
2009
)
TALOS+: a hybrid method for predicting protein backbone torsion angles from NMR chemical shifts
.
J. Biomol. NMR
44
,
213
223
30
Koh-Stenta
,
X.
,
Joy
,
J.
,
Wang
,
S.F.
,
Kwek
,
P.
,
Wee
,
J.
,
Wan
,
K.F.
et al. 
(
2015
)
Identification of covalent active site inhibitors of dengue virus protease
.
Drug Des. Dev. Ther.
9
,
6389
6399
31
Li
,
Y.
,
Wong
,
Y.X.
,
Poh
,
Z.Y.
,
Wong
,
Y.L.
,
Lee
,
M.Y.
,
Ng
,
H.Q.
et al. 
(
2015
)
NMR structural characterization of the N-terminal active domain of the gyrase B subunit from Pseudomonas aeruginosa and its complex with an inhibitor
.
FEBS Lett.
589
,
2683
2689
32
Li
,
Q.
,
Ng
,
H.Q.
,
Yoon
,
H.S.
and
Kang
,
C.
(
2014
)
Insight into the molecular interaction between the cyclic nucleotide-binding homology domain and the eag domain of the hERG channel
.
FEBS Lett.
588
,
2782
2788
33
Mesrouze
,
Y.
,
Meyerhofer
,
M.
,
Bokhovchuk
,
F.
,
Fontana
,
P.
,
Zimmermann
,
C.
,
Martin
,
T.
et al. 
(
2017
)
Effect of the acylation of TEAD4 on its interaction with co-activators YAP and TAZ
.
Protein Sci.
26
,
2399
2409
34
van Zundert
,
G.C.P.
,
Rodrigues
,
J.P.G.L.M.
,
Trellet
,
M.
,
Schmitz
,
C.
,
Kastritis
,
P.L.
,
Karaca
,
E.
et al. 
(
2016
)
The HADDOCK2.2 web server: user-friendly integrative modeling of biomolecular complexes
.
J. Mol. Biol.
428
,
720
725
35
Schüttelkopf
,
A.W.
and
van Aalten
,
D.M.
(
2004
)
PRODRG: a tool for high-throughput crystallography of protein–ligand complexes
.
Acta Crystallogr. Sect. D: Biol. Crystallogr.
60
,
1355
1363
36
Li
,
Y.
,
Wong
,
Y.L.
,
Ng
,
F.M.
,
Liu
,
B.
,
Wong
,
Y.X.
,
Poh
,
Z.Y.
et al. 
(
2016
)
Escherichia coli topoisomerase IV E subunit and an inhibitor binding mode revealed by NMR spectroscopy
.
J. Biol. Chem.
291
,
17743
17753
37
Erlanson
,
D.A.
,
Fesik
,
S.W.
,
Hubbard
,
R.E.
,
Jahnke
,
W.
and
Jhoti
,
H.
(
2016
)
Twenty years on: the impact of fragments on drug discovery
.
Nat. Rev. Drug Discov.
15
,
605
619
38
Gable
,
J.E.
,
Lee
,
G.M.
,
Acker
,
T.M.
,
Hulce
,
K.R.
,
Gonzalez
,
E.R.
,
Schweigler
,
P.
et al. 
 (
2016
)
Fragment-based protein-protein interaction antagonists of a viral dimeric protease
.
ChemMedChem
11
,
862
869
39
Ivanov
,
A.A.
,
Khuri
,
F.R.
and
Fu
,
H.
(
2013
)
Targeting protein–protein interactions as an anticancer strategy
.
Trends Pharmacol. Sci.
34
,
393
400
40
Shangary
,
S.
and
Wang
,
S.
(
2009
)
Small-molecule inhibitors of the MDM2-p53 protein-protein interaction to reactivate p53 function: a novel approach for cancer therapy
.
Annu. Rev. Pharmacol. Toxicol.
49
,
223
241
41
D'Silva
,
L.
,
Ozdowy
,
P.
,
Krajewski
,
M.
,
Rothweiler
,
U.
,
Singh
,
M.
and
Holak
,
T.A.
(
2005
)
Monitoring the effects of antagonists on protein−protein interactions with NMR spectroscopy
.
J. Am. Chem. Soc.
127
,
13220
13226

Author notes

*

These authors contributed equally to the work.

Present address: Chemistry Research Laboratory, 12 Mansfield Rd, Oxford OX1 3TA, U.K.

Supplementary data