Cooperation between the Hippo and MAPK pathway activation drives acquired resistance to TEAD inhibition

Paul, Sayantanee; Hagenbeek, Thijs J.; Tremblay, Julien; Kameswaran, Vasumathi; Ong, Christy; Liu, Chad; Guarnaccia, Alissa D.; Mondo, James A.; Hsu, Peter L.; Kljavin, Noelyn M.; Czech, Bartosz; Smola, Janina; Nguyen, Dieu An H.; Lacap, Jennifer A.; Pham, Trang H.; Liang, Yuxin; Blake, Robert A.; Gerosa, Luca; Grimmer, Matthew; Xie, Shiqi; Daniel, Bence; Yao, Xiaosai; Dey, Anwesha

doi:10.1038/s41467-025-56634-y

Download PDF

Article
Open access
Published: 18 February 2025

Cooperation between the Hippo and MAPK pathway activation drives acquired resistance to TEAD inhibition

Nature Communications volume 16, Article number: 1743 (2025) Cite this article

8384 Accesses
6 Citations
1 Altmetric
Metrics details

Subjects

Abstract

TEAD (transcriptional enhanced associate domain) transcription factors (TEAD1-4) serve as the primary effectors of the Hippo signaling pathway in various cancers. Targeted therapy leads to the emergence of resistance and the underlying mechanism of resistance to TEAD inhibition in cancers is less characterized. We uncover that upregulation of the AP-1 (activator protein-1) transcription factors, along with restored YAP (yes-associated protein) and TEAD activity, drives resistance to GNE-7883, a pan-TEAD inhibitor. Acute GNE-7883 treatment abrogates YAP-TEAD binding and attenuates FOSL1 (FOS like 1) activity. TEAD inhibitor resistant cells restore YAP and TEAD chromatin occupancy, acquire additional FOSL1 binding and exhibit increased MAPK (mitogen-activated protein kinase) pathway activity. FOSL1 is required for the chromatin binding of YAP and TEAD. This study describes a clinically relevant interplay between the Hippo and MAPK pathway and highlights the key role of MAPK pathway inhibitors in mitigating resistance to TEAD inhibition in Hippo pathway dependent cancers.

Targeting the Hippo pathway in cancer

Article 30 June 2025

An allosteric pan-TEAD inhibitor blocks oncogenic YAP/TAZ signaling and overcomes KRAS G12C inhibitor resistance

Article Open access 05 June 2023

Direct and selective pharmacological disruption of the YAP–TEAD interface by IAG933 inhibits Hippo-dependent and RAS–MAPK-altered cancers

Article Open access 02 April 2024

Introduction

The Hippo pathway is a highly conserved signaling pathway that responds to various cellular stimuli to regulate cell survival and to put the brakes on excessive cellular growth¹. A variety of upstream signals modulate the activation of this pathway, which is tightly controlled by its core components such as NF2 (neurofibromin 2) and the Hippo kinases LATS1/2 (large tumor suppressor homolog 1/2) and MST1/2 (mammalian STE20-like protein kinase 1/2)^2,3,4. When active, the upstream Hippo kinases phosphorylate YAP and TAZ (transcriptional coactivator with PDZ-binding motif) leading to their cytosolic retention and degradation^5,6,7,8. However, when the pathway is inactive, YAP and TAZ remain unphosphorylated and are capable of nuclear translocation and nuclear activity⁹. In the nucleus, YAP and TAZ engage with TEAD proteins to promote transcription. There are four mammalian TEAD paralogs (TEAD1-4), all of which are capable of interacting with YAP and TAZ to drive a pro-growth transcriptional program^10,11.

Diverse mechanisms can lead to an increase in YAP/TAZ activity, such as loss-of-function mutations in negative pathway regulators or activating mutations that promote YAP/TAZ nuclear localization, driving tumorigenesis². Malignant mesothelioma (MM), a highly aggressive cancer originating from the mesothelioma cells of the pleura, is a prominent tumor type with Hippo pathway genetic alterations and high YAP nuclear expression. It is characterized by deletion or mutations in NF2, or LATS^12,13,14. Diagnosis often occurs at advanced stages, limiting treatment options and offering minimal survival benefits from conventional chemotherapy. Since mesothelioma cells heavily rely on the YAP/TEAD for cancer progression, novel treatments options, particularly those targeting YAP/TEAD, are promising avenues.

Unlike other well-established oncogenic signaling pathways, the Hippo pathway genes exhibit relatively low prevalence of somatic or germline mutations in common human cancers. Instead, dysregulated YAP/TAZ levels are frequently associated with pathway dysfunction¹⁵. Increased nuclear YAP/TAZ levels have been generally associated with poor prognosis and therapy resistance in multiple cancers^2,4,14,16. Since TEADs are associated with the majority of YAP/TAZ driven biological functions including tumorigenic phenotypes, inhibiting the TEADs is an attractive strategy to target the Hippo pathway. A wide variety of TEAD small molecule inhibitors (SMIs) have been developed and characterized in recent years^{17,18,19,20,21}. Most TEAD SMIs target the lipid binding pocket of TEAD, which is a conserved deep pocket in the core of the protein that is endogenously engaged by a palmitoyl group^17,22. These TEAD-targeting modalities are rapidly advancing in the clinic, with multiple molecules currently being evaluated in Phase I clinical trials and at various stages of clinical development²³. A thorough understanding of the effects of long-term treatment can inform the clinical implications and additional combination opportunities to combat resistance to TEAD inhibition.

To address this question, here we utilize our molecule GNE-7883. GNE-7883 is a potent allosteric pan-TEAD inhibitor that engages the lipid pocket, decreases cellular proliferation of mesothelioma cell lines, and overcomes cellular resistance to KRAS G12C inhibitors²⁴. We generate acquired resistance models to GNE-7883 with the aim of determining the underlying molecular mechanisms of resistance to Hippo pathway inhibition. We develop multiple in vitro and in vivo model systems to study the impact of long-term acquired resistance in Hippo-dependent cancers. Taking a single cell multiomic approach, we interrogate the chromatin state and expression profiles of individual cells to determine factors underlying resistance to TEAD SMI. We show that upregulation of AP-1 transcription factors promotes activation of compensatory MAPK pathway and helps overcome the impact of single-agent TEAD SMI treatment in Hippo dependent/NF2 null cancers. This study aims to provide a molecular basis for evaluating the response and resistance to TEAD inhibitors in the clinic.

Results

Hippo-dependent cell lines are arrested in G0/G1 phase upon acute treatment with GNE-7883

We first examined the effect of treating Hippo dependent cell lines with the pan-TEAD inhibitor GNE-7883 which is known to disrupt YAP-TEAD interaction (Fig. 1A). We profiled two established preclinical models of malignant mesotheliomas: NCI-H226 (referred to as H226) which is null for NF2, and MSTO-211H (referred to as MSTO) which harbors a PSEN1-LATS1 fusion resulting in the loss of kinase activity and the ability to phosphorylate YAP. The MSTO cell line also harbors a 42 bp deletion in exon 5 of the LATS2 gene, resulting in inactivation of LATS2^12,23,25,26. Both cell lines showed sensitivity to GNE-7883 in a dose dependent viability assay (Figs. 1B and S1A). We selected a drug dose of 1 μM based on its efficacy in inhibiting cell viability over a 7-day viability assay, and used it for subsequent assays. At 1 µM concentration, GNE-7883 effectively displaced YAP from TEAD, consistent with the known mechanism of action of this compound, in both H226 and MSTO cell lines (Fig. S1B-D)²⁴.

**Fig. 1: Mechanism of action of TEAD inhibitor, GNE-7883, upon short term treatment (48 h).**

To evaluate the effect of acute inhibition of TEAD we treated cells for 48 h with 1 µM GNE-7883 or vehicle (DMSO) (Fig. 1C). Using bulk transcriptomics analysis, we showed that GNE-7883 treatment led to suppression of genes associated with cell cycle progression (Fig. 1D)²⁷. Moreover, Gene Set Enrichment Analysis (GSEA) revealed significant downregulation of the pathways related to cell cycle progression like E2F targets, G2M checkpoints, and mitotic spindle (Fig. 1E, Supplementary Data 1)^28,29. We evaluated the impact on cell cycle progression, by measuring the level of ethynyl-2’-deoxyuridine (EdU) incorporation into synthesized DNA³⁰. Active DNA synthesis is an indicator for S phase progression. We observed that both H226 and MSTO exhibited G0/G1 arrest after 48 h of treatment (Figs. 1F, G and S1E, F). This indicates that cells failed to progress through mitosis in the presence of GNE-7883 as evident by a reduced number of nuclei showing positive EdU incorporation.

To explore the impact of GNE-7883 on chromatin accessibility, we performed an assay for transposase-accessible chromatin with sequencing (ATAC-seq) of H226 cells treated with DMSO, 1 µM GNE-7883 and GNE-6915. GNE-6915 is a pan-TEAD degrader inducing proteasome dependent degradation of TEADs, and was used as a positive control³¹. We confirmed that GNE-6915 effectively induced TEAD degradation at a 1 µM concentration (Fig. S1G). Both GNE-7883 and GNE-6915 decreased chromatin accessibility at TEAD motifs (absolute fold change > 1.5 and FDR < 0.05, the Ward test, Fig. 1H) but subtle differences exist, with GNE-7883 suppressing chromatin accessibility at motifs corresponding to AP-1 family including FOS, JUN, FOSL1 motifs and GNE-6915 decreasing chromatin accessibility at KLF motifs (Fig. 1I, J, Supplementary Data 2). This suggests that inhibition of TEADs by GNE-7883 may potentially suppress the activity of factors associated with TEADs including the AP-1 family of transcription factors.

Acquired resistance to GNE-7883 is characterized by an augmentation in MAPK pathway activity and restored expression of YAP/TEAD targets

We next tested the effect of prolonged exposure to GNE-7883 by generating resistant cell lines. We treated H226 with a gradual dose escalation of GNE-7883 from 0.25 to 2.5 μM over a course of 55 days and eventually maintained it at 2.5 μM (Fig. 2A). The MSTO cell line was subjected to gradual dose escalation of GNE-7883 from 0.25 to 3 μM and was maintained at 3 μM (Fig. S2A). This model allowed the cells to recover from cytostasis or cell death and slowly adapt over a few weeks until outgrowths appeared. We generated the resistant cell lines using pooled populations of persister cells. Cells exposed to long term treatment of GNE-7883 continued to grow in the presence of the inhibitor in a dose dependent viability assay, suggesting that these lines had acquired resistance to GNE-7883 and were thereafter referred to as H226-7883R/MSTO-7883R (Figs. 2B and S2B). Live cell imaging also revealed that both H226 and MSTO were sensitive to the GNE-7883 while the resistant cells continued to grow in the presence of the inhibitor (Fig. S2C, D).

**Fig. 2: Acquired resistance to GNE-7883 demonstrates increase in MAPK pathway activity and restored expression of YAP/TEAD target genes.**

In order to characterize the resistant cells, we did bulk transcriptomics analysis and confirmed that TEAD SMI treatment resulted in downregulation of our previously developed Hippo signature²⁷ following 48 h treatment with GNE-7883 but the expression of this gene set was restored in the resistant cells (Figs. 2C and S2E). We observed a consistent trend at the protein level; established YAP/TEAD targets, such as CTGF (connective tissue growth factor) and CCN1 (cellular communication network factor 1, also known as CYR61 or cysteine-rich angiogenic inducer 61), were downregulated in the parental cells upon drug treatment (Figs. 2D and S3A, B). In contrast to the acute treatment of the parental cells, the resistant cells restored a significant amount of TEAD target proteins in the presence of the drug, at levels similar to or even higher than those observed in the parental cells without treatment. Notably, removing the drug from the resistant cells for 48 h resulted in a stronger induction of certain target genes, such as CYR61 for H226 and CTGF for MSTO, compared to the parental cells (Fig. S3A, B). We also observed that pan-TEAD expression is significantly higher in the resistant cells compared to the parental cells in both cell lines and YAP expression is elevated in the MSTO resistant cells compared to the parental cells (Fig. 2D, S3A, B). The resistant cells also exhibited reduced upregulation of gene sets associated with apoptosis when compared to the parental cells in the presence of the inhibitor (Fig. 2C)²⁷.

To determine if resistance was attributed to acquisition of mutations in TEAD/YAP/TAZ, we performed whole exome sequencing for H226 and H226-7883R, which are pooled populations of resistant cells. We detected mutations in TEAD1, TEAD2, TEAD3, TEAD4, YAP1 and WWTR1 (encoding TAZ) but they all occur at very low allele frequencies (Fig. S4A, Supplementary Data 3). To determine whether these mutations play a causal role in resistance or are merely passenger mutations, we mapped the TEAD missense mutants to high confidence AlphaFold 2 models of TEAD1-4. We observed that none of the mutations are found in the lipid pocket of TEADs or the TEAD/YAP/TAZ interface and that the majority are located on the surface of the YAP-binding domain distal to the YAP interaction surface (Fig. S4B). Since the mutations do not suggest an alteration to the YAP-binding domain or the lipid pocket of TEADs, we did not discern an obvious functional consequence of these mutations in the context of resistance.

We proceeded to investigate whether crosstalk with other signaling pathways could mediate resistance to the TEAD inhibitor. We found that while acute treatment with GNE-7883 resulted in the upregulation of genes linked to KRAS signaling, upregulation of a MAPK related gene signature was significantly more pronounced in the resistant cells compared to parental cells with 48 h treatment of GNE-7883 (Fig. 2C and E). The level of dually phosphorylated ERK (dpERK) was significantly elevated in the resistant cells compared to the parental cells. The ratio of dpERK/tERK and pMEK/tMEK was also significantly higher in the resistant cells compared to the parental cells (Figs. 2F, and S3C, D). This correlates with the increased MAPK pathway gene signature observed at the RNA level, suggesting that the activation of the MAPK pathway serves as a mechanism of resistance in Hippo pathway-dependent cancer. This was further confirmed by GSEA analysis, which unveiled enrichment of gene set upregulated by KRAS activation in resistant cells compared to parental cells (Fig. 2G)²⁸.

We also observed upregulation of gene signatures associated with cell cycle progression, which had been downregulated during the initial short-term treatment with the inhibitor (Fig. S5A). GSEA also highlighted enrichment of hallmark targets related to cell cycle in context of acute vs long term treatment (Fig. 2G, Supplementary Data 1). To validate the cell cycle progression of resistant cells, we did an EdU incorporation assay. We observed that both H226-7883R and MSTO-7883R cell lines advanced through the cell cycle, indicating a restoration of the previously observed cell cycle arrest in G0/G1 (Fig. 2H and S5B-C). We next proceeded to assess the protein levels of Cyclins, CDKs and phospho-Rb which constitute the fundamental machinery responsible for propelling cells through the cell cycle^32,33,34. We observed downregulation of several cell cycle genes, including CDK2, Cyclin B1 and phospho-Rb in acute treatment with GNE-7883 but restored in resistant cells (Fig. S5D–E). Taken together these observations provide strong evidence that resistant cells treated with the inhibitor undergo cell division by overcoming G0/G1 block upon acquiring resistance to GNE-7883.

GNE-7883 resistant cells form a transcriptionally distinct cell state with heightened FOSL1 activity

Our subsequent investigation focused on identifying the gene regulatory network that drives resistance to the TEAD inhibitor, GNE-7883. To do so, we comprehensively analyzed the transcriptomic and chromatin profiles of both parental cells and cells that had developed resistance to GNE-7883 by using single-cell RNA-seq paired with ATAC-seq (thereafter referred to as scMultiome). GNE-7883 treated parental H226 cells formed a distinct cluster from DMSO-treated cells but DMSO-treated and GNE-7883 resistant H226 cells overlapped, suggesting that GNE-7883 failed to elicit a response in the resistant cells (Fig. 3A). We detected a small population of cells (H226_side), which is distinct from the main population of cells. However, for this study we focused on the main population of resistant cells (H226-7883R_main) since they represent the majority of the cells with acquired resistance to GNE-7883.

**Fig. 3: GNE-7883 resistant cells form a transcriptionally distinct cell state with restored TEAD and heightened FOSL1 activity.**

We observed that acute GNE-7883 treatment reduced chromatin accessibility at TEAD binding sites in the parental cells whereas resistant H226 cells were able to restore chromatin accessibility at these regions despite GNE-7883 treatment (Fig. 3B, S6A). We further performed an in-depth analysis of the gene regulatory networks of the sensitive vs. resistant states, utilizing our scMultiome data and a computational tool (Epiregulon) we recently developed (see methods). Epiregulon infers transcription factor activity from single cell chromatin accessibility and gene expression data and bulk ChIP-seq data. This method detected a substantial decrease in TEAD activity upon GNE-7883 treatment only in the parental but not in the resistant cells (Fig. 3B, S6A, Supplementary Data 4). Hippo signature, which comprises many canonical TEAD targets, validated the loss of TEAD activity upon GNE-7883 treatment in parental H226 cells but TEAD activity was maintained in the resistant cells (Fig. 3C)²⁷.

We further compared the chromatin profiles between parental and resistant cells. Resistant cells showed increased chromatin accessibility at motifs corresponding to the AP-1 family members (Fig. 3D). To identify the master regulator of TEAD inhibitor resistance, we performed a global differential transcription factor activity analysis. Epiregulon identified FOSL1 as one of the top transcription factors with enhanced activity in TEAD inhibitor resistant cells (Fig. 3E), suggesting that it is likely the AP-1 family member responsible for the increased chromatin accessibility. FOSL1 is aberrantly expressed in various tumors, contributing to tumorigenesis and progression³⁵. We focused our downstream analyses on FOSL1 because its activity and motif (AP-1 family) were both upregulated in the resistant cluster (Fig. 3D, E). The target genes of FOSL1 map to pathways related to DNA damage and MAPK response (Fig. 3F, Supplementary Data 5). We also observed that resistant cells had higher expression of FOSL1 and increased chromatin accessibility at FOSL1 binding sites (Figs. 3B and S6A). Analyses from Epiregulon revealed a substantial increase in FOSL1 activity upon GNE-7883 treatment in resistant cells (Fig. 3B, Fig. S6A, Supplementary Data 4). We further validated the scMultiome data by looking at the protein level of two AP-1 family members, FOSL1 and c-Jun. Immunoblotting showed enhanced FOSL1 and c-Jun expression in resistant cells at the protein level (Figs. 3G and S6B). Taken together the scMultiome data suggests that AP-1 family transcription factors as potential drivers of transcriptional reprogramming in cells resistant to GNE-7883.

We next investigated whether the transcriptional reprogramming responsible for drug resistance in H226-7883R cell line is permanent or reversible following a period of “drug holiday”. This pause in drug exposure allows the cells to recover and potentially revert their gene expression patterns. We assessed the sensitivity of the drug-resistant cell line to GNE-7883 at specific intervals over a span of 12 weeks. Intriguingly, we observed that despite the break from drug exposure, the cells sustained their resistance to GNE-7883. Consistent with this, the gene expression profiles of key regulators, such as YAP/TEAD targets and FOSL1, failed to revert back to the patterns characteristic of the drug-sensitive state (Fig. S6C, D). Overall, this result indicates persistence of the transcriptional changes that underlie drug resistance, suggesting that these adaptations are stable and last for an extended period of time.

GNE-7883 resistant cells restore TEAD and YAP occupancy and gain FOSL1 occupancy

Given that the chromatin accessibility of YAP and TEAD binding sites was diminished in sensitive cells but remained intact in TEAD-resistant cells, we wanted to check the localization of YAP/TEAD in both parental and TEAD inhibitor resistant cells. The localization of YAP is known to modulate YAP activity³⁶. YAP is considered to be active in the nucleus and inactive in the cytoplasm^37,38. Our immunofluorescence-based investigations of fixed cells have revealed that YAP is partially translocated to the cytoplasm following short-term treatment with GNE-7883 but a significant pool of YAP still remained localized within the nucleus in the parental H226 cells (Fig. S7A–D). This prompted the question of whether this remaining nuclear YAP is indeed tethered to chromatin. To address this, we did pre-extraction of unbound proteins with a detergent containing buffer prior to fixing the cells, so that the unbound proteins are extracted away from the cells. As shown in Fig. 4A, short-term exposure to GNE-7883 leads to a substantial reduction in chromatin-bound YAP with pre-extraction, in contrast to the minimal reduction of YAP nuclear protein level without pre-extraction (Fig. S7A–D). Chromatin bound total TEAD levels were decreased by the treatment in parental cells (Fig. 4A). In contrast, in the resistant cells, both YAP and TEAD persisted on the chromatin indicating restoration of YAP/TEAD activity in the resistant cells (Fig. 4A).

**Fig. 4: GNE-7883 resistant cells restore TEAD and YAP chromatin occupancy and enhance FOSL1 occupancy.**

We were curious to understand whether the mechanism of resistance to GNE-7883 in H226-7883R cells was associated with its ability to engage its target, TEAD. Target engagement methods typically fall into two categories, those based on binding-induced changes to target protein stability and those based on competitive binding to a specific site relative to a probe compound³⁹. The pan-TEAD degrader GNE-6915 provided us with a potential competitive probe compound to enable the detection of GNE-7883 binding, given that both compounds bind to the same lipid binding pocket on TEAD, despite having very different chemical structures. We hypothesized that the binding of GNE-7883 would prevent GNE-6915 from binding to TEAD and consequently, inhibit its induction of TEAD degradation, which would provide us with a functional readout of TEAD target engagement. GNE-6915 at 1 µM induced a significant reduction in the median value of nuclear TEAD detected by quantitative fluorescence imaging in both H226 and H226-7883R cells, demonstrating equivalent induced protein degradation of TEAD (Fig. S8A–C). Treatment with GNE-7883 demonstrated a concentration dependent protection of TEAD from GNE-6915 induced degradation in H226 cells, demonstrating competition for TEAD binding and therefore target engagement (Fig. S8A, C). However, GNE-7883 was not able to protect TEAD from GNE-6915 induced degradation in H226-7883R cells, demonstrating a lack of competitive binding and therefore reduced TEAD target engagement by GNE-7883 in H226-7883R cells (Fig. S8B, C). Additionally, immunoprecipitation results showed that while GNE-7883 effectively displaced YAP from TEAD in the parental cells, the binding was restored in resistant H226 cells (Fig. S8D).

To quantify changes in chromatin occupancy accurately, we performed CUT&RUN assays for TEAD1, YAP, and FOSL1 in both parental and resistant cells. Our analysis revealed decreased chromatin occupancy for all three factors in the parental cells upon GNE-7883 treatment (Fig. 4B–D). However, in the resistant cells, the occupancy of TEAD1 and YAP was restored (Fig. 4B, C). Additionally, FOSL1 peaks were increased in the resistant cells compared to the parental cells (Fig. 4D). Consistent findings were observed in MSTO cell lines. In the parental MSTO cells, chromatin-bound total YAP levels decreased with the acute treatment (Fig. S9A). In the MSTO-7883R cells, both YAP and TEAD persisted on the chromatin, indicating the restoration of YAP/TEAD activity in the resistant cells (Fig. S9A). A similar trend was observed for chromatin-bound FOSL1, where levels were reduced in the parental cells upon acute treatment but were upregulated in the MSTO-7883R cells (Fig. S9B). Furthermore, immunoprecipitation results showed that while GNE-7883 effectively displaced YAP from TEAD in the MSTO cells, YAP and TEAD binding was partially restored in MSTO-7883R cells (Fig. S9C).

To measure the extent of TEAD1, YAP1, and FOSL1 site loss together in H226 cells upon GNE-7883 treatment in the parental setting, we conducted differential peak analysis (absolute fold change > 1.5 and FDR < 0.1, the Ward test, n = 2) and examined the overlap. We found that 6401 TEAD1 and YAP co-bound sites were lost with GNE-7883 treatment, consistent with the specificity of GNE-7883 in modulating YAP and TEAD activity (Fig. 4E). Additionally, 432 FOSL1 and TEAD1 co-bound peaks and 370 FOSL1 and YAP co-bound peaks were lost (Fig. 4E). In contrast to the acute treatment, in the resistant cells we saw a significant gain in the numbers of FOSL1, TEAD1 and YAP peaks. We observed that 2175 TEAD1 and YAP co-bound sites, 2316 FOSL1 and TEAD1 co-bound sites, and 1190 FOSL1 and YAP co-bound peaks were gained (Fig. 4F). These findings indicate that resistant cells maintain TEAD/YAP/FOSL1 binding in the presence of the inhibitor, suggesting their role in resistance mechanisms.

We next evaluated whether this restored YAP/TEAD interaction in the resistant cells can be targeted using the TEAD degrader GNE-6915. We treated both parental H226 and H226-7883R cells with GNE-7883 and GNE-6915 at a 1 µM concentration for 48 h. We observed that acute treatment with GNE-6915 reduced chromatin accessibility at TEAD binding sites in the parental cells, similar to GNE-7883. In contrast, cells resistant to GNE-7883 were able to restore chromatin accessibility at these regions. GNE-6915 treatment in the H226-7883R cells decreased chromatin accessibility at the TEAD binding sites (Fig. S10A, B).

The FOSL1 cistrome is substantially remodeled in resistant cells and associates with YAP

Having established that resistant cells restore TEAD and YAP chromatin occupancy and gain FOSL1 occupancy, we set out to investigate in more detail how the FOSL1 cistrome is changing upon acute inhibitor treatment and in the resistant state. To validate our CUT&RUN results using an independent approach, we performed additional chromatin immunoprecipitation sequencing (ChIP-seq) for FOSL1. Consistent with our CUT&RUN data, the FOSL1 cistrome exhibited dramatic differences across our conditions. Differential binding analysis (DiffBind analysis, FDR < 0.01, n = 2) reported that acute inhibitor treatment in parental cells induced FOSL1 binding at 1,065 sites, while it repressed FOSL1 binding at 5,535 sites. The same analysis between vehicle-treated parental cells and resistant cells revealed substantial remodeling of the FOSL1 cistrome in the resistant state featuring 3129 induced, and 2867 repressed FOSL1 binding sites (Fig. 5A).

**Fig. 5: FOSL1 and YAP binding dynamics are correlated in parental and resistant cells.**

To determine if acute inhibitor treatment in parental cells affects the FOSL1 binding sites that are determined as differential in the resistant state, we grouped our differential binding sites into induced and repressed categories (Groups I-III - repressed, Groups IV-VI - induced), and performed overlap analysis (Fig. 5A). We found a relatively small number of common FOSL1 binding sites that were repressed (Group II) or induced (Group V) by both acute treatment, and the resistant state (Fig. 5A), suggesting that FOSL1 redistribution occurs in these conditions by different mechanisms (Fig. 5A). Group I and Group III FOSL1 sites are specifically repressed in the parental and resistant settings, respectively, showing that acute inhibitor treatment in parental cells repressed more FOSL1 sites than what we detected in the resistant state. Furthermore, there was a dramatic gain of GNE-7883-induced FOSL1 binding sites in the resistant state (Group VI) compared to the parental cells (Group IV) (Fig. 5A). Next, we assessed FOSL1 occupancy in the different groups across all three conditions, which confirmed that repressed FOSL1 binding sites in Groups I and II were sensitive to acute GNE-7883 treatment, but were restored in the resistant state (Group II by a lesser extent) (Fig. 5B). FOSL1 sites in Group III were insensitive to acute GNE-7883 treatment but were largely lost in the resistant state (Fig. 5B). Additionally, induced FOSL1 sites showed transient gain in occupancy in Group IV upon GNE-7883 treatment that was lost in the resistant state, while Group V gained occupancy as a result of acute GNE-7883 treatment that was either maintained or lost in the resistant state, and lastly, Group VI exhibited gained FOSL1 binding specifically in the resistant state (Fig. 5B). Finally, we investigated how these FOSL1 cistromic changes align with the binding of YAP, therefore, we performed YAP ChIP-seq, which revealed that the binding patterns of YAP faithfully recapitulated the binding patterns of FOSL1 in the above defined Groups (Fig. 5B). Together, these results provide evidence that the FOSL1 cistrome is substantially reprogrammed in the resistant state and correlated with YAP binding.

Overall, the ChIP-seq results suggest a critical role for FOSL1 and YAP in establishing the resistant state. To further explore the relative contributions of YAP and FOSL1 for survival of the resistant cells, we conducted knockdown experiments targeting YAP and FOSL1 individually using small interfering RNA (siRNA). Our data suggests that both FOSL1 and YAP are necessary for the survival of the resistant cells (Fig. S10C, D).

FOSL1 is required for TEAD/YAP chromatin binding in TEAD inhibitor resistant cells

Since we saw strong induction of YAP binding at gained FOSL1 sites, we investigated whether FOSL1 is necessary for recruiting YAP and TEAD to the chromatin in resistant cells. We performed CUT&RUN assays for TEAD, YAP, and FOSL1 after knocking down FOSL1 using CRISPR. Depletion of FOSL1 was confirmed by the loss of FOSL1 peaks (Fig. 6A). As controls we nucleofected resistant cells with synthetic guide RNAs targeting CD81, encoding a cell surface protein not essential for cell proliferation or survival. The loss of FOSL1 was associated with a coordinated loss of YAP and TEAD occupancy, indicating that FOSL1 is crucial for the recruitment of YAP and TEAD to the chromatin (Fig. 6A). To assess the effect on the chromatin occupancy of TEAD1 and YAP due to the loss of FOSL1, we performed differential peak analysis between FOSL1-deficient versus CD81-deficient conditions (absolute fold change > 1.5 and FDR < 0.1, the Ward test, n = 2). The analysis revealed that FOSL1 depletion led to loss of YAP (1361 peaks) and TEAD1 binding (1602 peaks) (Figs. 6A, B and S10E). A subset of these lost FOSL1 peaks overlapped with lost YAP and TEAD1 binding (Fig. 6B). This suggests that FOSL1 is required for the binding of YAP/TEAD to the chromatin in TEAD inhibitor resistant cells.

To gain biological insight into the gene programs driving resistance that might be controlled by FOSL1, we examined pathways impacted by the loss of FOSL1 binding. GSEA analysis of lost FOSL1/TEAD1/YAP co-regulated targets revealed pathways related to TGFβ signaling, hypoxia, EMT, and apical junctions. Additionally, FOSL1-regulated targets indicated DNA damage, TNFα signaling via NFκB, and IL6 JAK/STAT3 pathways among many others. Furthermore, FOSL1 targets exhibited a depletion of genes down-regulated by MAPK activation (Hallmark KRAS signaling DN) and an enrichment of genes upregulated by KRAS activation (Hallmark KRAS signaling UP). This suggests that some of the gained sites in resistance might be directly linked to increased MAPK pathway activity (Fig. 6C, Supplementary Data 6).

Inhibition of the MAPK pathway sensitizes TEAD inhibitor resistant cell lines

From our investigations using the scMultiome and genomics studies, we identified FOSL1 as the AP-1 family member exhibiting increased activity in the resistant cells. Given that the resistant cells exhibited significantly higher FOSL1 expression compared to the parental cells, FOSL1 could represent a vulnerability for the cells which acquire resistance to GNE-7883. Thus, we proceeded to explore whether FOSL1 is important for maintaining the viability of GNE-7883 resistant cells. We observed that all three pools of TEAD inhibitor-resistant cells with FOSL1 knockdown, with three individually transfected guide RNAs, exhibited reduced viability in comparison to nucleofection with synthetic guide RNAs targeting CD81 (Figs. 7A, B and S11A). This supports the hypothesis that FOSL1 plays a role in survival of the resistant cells. To test whether loss of FOSL1 also impacted parental cells we knocked down FOSL1 in both parental and resistant cells. and found that resistant cells were significantly more dependent on FOSL1 for survival than the parental cells (Fig. S11B).

**Fig. 7: Inhibition of the MAPK pathway reduces FOSL1 level and sensitizes TEAD inhibitor resistant cell lines.**

Given that FOSL1 is a known downstream target of the KRAS/MAPK pathway, we next investigated whether MAPK pathway inhibitors, like Cobimetinib (MEK inhibitor) and Belvarafenib (pan-RAF inhibitor), could reduce the viability of TEAD inhibitor-resistant cells^40,41,42. Treatment of resistant cells with DMSO, GNE-7883, or Cobimetinib resulted in a reduction in FOSL1 levels following Cobimetinib treatment (Figs. 7C and S11C–E). Parental cells treated with DMSO, GNE-7883 were included as control to monitor the basal level of FOSL1 in these cells. To determine if the chromatin-bound FOSL1 is also reduced by Cobimetinib, we did pre extraction prior to fixing the cells so that the unbound proteins are extracted away from the cells, allowing us to only stain for chromatin bound FOSL1. Our immunofluorescence-based analysis of fixed cells indicated low FOSL1 levels, primarily localized in the nucleus in the parental cells. However, in TEAD-resistant cells, FOSL1 continued to remain associated with chromatin even in the presence of GNE-7883, with its levels significantly elevated compared to those in parental cells (Fig. 7D, E). Consistent with the western blot result, short-term GNE-7883 and Cobimetinib treatment, could lead to near-ablation of chromatin bound FOSL1 levels in the parental cells (Fig. 7D, E). Interestingly, short-term Cobimetinib exposure (48 h) of TEAD inhibitor resistant cells also significantly reduced chromatin bound FOSL1 levels, bringing it down to basal parental levels (Fig. 7D-E).

Since, FOSL1 is recognized as a pivotal link between the Hippo and MAPK pathways, resulting in a synergistic efficacy when combining inhibitors of the TEAD and MAPK pathways^27,43, we next wanted to test the viability of both H226 and MSTO parental and resistant cells in the presence of DMSO, Cobimetinib, Belvarafenib alone or in combination with GNE-7883. Parental cells displayed sensitivity to Cobimetinib, Belvarafenib and GNE-7883 alone or in combination (Fig. 7F, S11F). The resistant cells exhibited sensitivity to MAPK pathway inhibitors Cobimetinib and Belvarafenib (Figs. 7F and S11F).

MAPK pathway inhibition overcomes resistance to Hippo pathway inhibition in vivo

The strong effect of Cobimetinib in sensitizing GNE-7883-resistant cells in vitro prompted us to assess its efficacy in vivo. To evaluate the effectiveness of our targeting strategies in suppressing mesothelioma tumors resistant to TEAD inhibitors, we implanted H226-7883R and MSTO-7883R cells into immunocompromised mice and treated the established tumors with vehicle, GNE-7883, Cobimetinib as single agents, or a combination of GNE-7883 and Cobimetinib (Fig. 8A). While tumors continued to grow under single-agent treatments, the combination of Cobimetinib and GNE-7883 achieved tumor stasis (Dunnett’s test based on the area under the curve (AUC) growth rate, Fig. S12A) in the H226 model compared to single-agent GNE-7883 or Cobimetinib-treated tumors (Figs. 8B and S12A–C). In the MSTO model, the combination treatment resulted in significant tumor regression (Dunnett’s test based on the AUC growth rate; Fig. S13A) compared to single-agent GNE-7883 or Cobimetinib-treated tumors (Figs. 8C and S13A–C). These results indicate that MAPK pathway inhibition can effectively circumvent resistance and sensitize the tumors to the treatments.

**Fig. 8: Inhibition of MAPK overcomes resistance to GNE-7883 in in vivo xenograft models of mesothelioma.**

Discussion

Therapeutic strategies that inhibit TEAD transcriptional activity hold great promise, not only as monotherapies for Hippo-driven cancers, but also as a combination strategy to mitigate resistance and improve therapeutic outcomes for a variety of malignancies. With an increasing number of TEAD inhibitors progressing through preclinical and Phase 1 clinical trials, it is important to delineate the mechanisms of resistance against YAP/TEAD inhibition and to identify therapeutic strategies to circumvent or delay the onset of resistance. Our study has uncovered that enhanced activity of AP-1 transcription factor and restored activity of YAP/TEAD collectively mediate resistance against the allosteric, pan-TEAD inhibitor GNE-7883. We have observed key characteristics of sensitive and resistant cells (Fig. 8D–E). (I) GNE-7883 effectively arrests sensitive cells in the G0/G1 phase but resistant cells can progress through the cell cycle despite TEAD inhibition (Figs. 1, 2, 8D). (II) Prolonged TEAD inhibition restores YAP-TEAD binding and downstream targets gene expression, elevates MAPK activity compared to acute treatment, and activates an altered transcriptional profile leading to sustained activity of FOSL1 (Figs. 2–4 and 8D). III) Acute inhibition reduces chromatin occupancy of TEAD/YAP/FOSL1 in the sensitive cells, whereas resistant cells retain the chromatin occupancy of TEAD/YAP/FOSL1 and exhibits increased MAPK signaling (Fig. 8D-E). IV) FOSL1 is required for TEAD/YAP binding, and dictates YAP binding sites in the resistant state (Figs. 5, 6, 8E). V) MAPK inhibitors reduce FOSL1 protein expression and binding to the chromatin, and sensitizes TEAD inhibitor resistant cells both in vitro and in vivo (Figs. 7, 8E).

The Hippo pathway is intricately interwoven with other transcriptional networks and signaling pathways⁴⁴. AP-1 transcriptional complexes can act as transcriptional co-activators of YAP/TEAD to promote oncogenic transcriptional programs that drive cell growth in different tumor contexts^{43,45,46,47,48,49}. We have shown that TEAD inhibitor GNE-7883 resistant cells rely on non-genetic alterations such as transcriptional reprogramming to adapt and sustain growth in the presence of the drug. Even though the activities of YAP/TEAD and FOSL1 decrease upon acute treatment, these key regulators are reinstated in the TEAD inhibitor acquired resistant model. YAP/TEAD and FOSL1 are also intricately interconnected through autoregulation and feedback loops, with FOSL1 being a direct transcriptional target of YAP/TEAD. The increase in MAPK activity caused by acute YAP/TEAD inhibition remains insufficient to sustain cell fitness, but as the cells begin to acquire resistance, the restoration of YAP-TEAD activity and elevated MAPK pathway activity contributes to heightened FOSL1 expression and activity.

Recently, YAP/TAZ–TEAD have emerged as a key mediator of resistance to therapies targeting EGFR^50,51, ALK^52,53, MEK⁵⁴, and CDK4/6⁵⁵ and KRAS^24,56,57. This highlights TEAD transcriptional activation as a prominent non-genetic adaptation in cancer cells during the development of resistance to targeted therapies. Conversely, our study highlights MAPK pathway activity as one of the major drivers of resistance to the Hippo pathway inhibition. This suggests that there is a fine balance between these two key oncogenic signaling pathways in regulating therapy response and resistance. Upon long term TEAD blockade the rheostat shifts towards MAPK signaling, increasing AP-1 activity in acquired resistance cells. Thus, MAPK inhibitors like Cobimetinib and Belvarafenib efficiently suppress FOSL1 activity, thereby sensitizing resistant cells^58,59,60. Thus, it might be beneficial to treat patients who develop resistance to TEAD inhibitors by combining with MAPK pathway inhibitors to cause global collapse of the Hippo-MAPK dependent gene regulatory network and enable sustained response to therapy.

The induction of MAPK pathway activity during acute treatment also presents an opportunity for the use of TEAD inhibitors in combination with MAPK inhibitors early on, as a promising strategy to delay the onset of resistance to TEAD inhibitors. This approach can effectively inhibit the compensatory MAPK reactivation that can occur during long-term treatment with single-agent TEAD SMI therapy in patients. Also, the role of FOSL1 as a Hippo and MAPK co-target gene and its influence downstream of the MAPK axis in context of TEAD resistance suggest its promise as a potential therapeutic target. FOSL1 has been identified as a biomarker of cancer cell invasion and in conferring resistance or addiction to targeted therapy⁶¹.

Despite the difficulties in targeting transcription factors because of their disordered structure and lack of binding sites, recent therapeutic strategies aimed at targeting FOSL1 expression or activity could be worth pursuing⁶².

Additionally, this study highlights the intriguing observation of restored YAP/TEAD activity upon resistance to TEAD inhibitor. With acute treatment, YAP localizes to cytoplasm and chromatin bound YAP is drastically reduced, but in resistant cells, YAP and TEAD remain bound to the chromatin. We have shown that an orthogonal targeting strategy involving a PROTAC based TEAD degrader can effectively reduce the chromatin occupancy of TEAD, opening additional dimensions to confront resistance mediated by TEAD inhibitors that bind to the lipid pocket. Our competitive binding experiment with GNE-7883 and GNE-6915 indicates that GNE-7883 engages TEAD with reduced efficiency in resistant cells. However, there could still be some residual activity of the TEAD inhibitor in resistant cells. Interpreting these results in cell-based target engagement assays for direct binding can be challenging. Structural studies are essential to determine if prolonged GNE-7883 treatment induces changes in the protein-inhibitor complex. In our xenograft study, the observed in vivo combination effect of GNE-7883 and Cobimetinib may also reflect complex signaling crosstalk within the stromal cell types present in this tumor microenvironment, rather than a tumor-intrinsic effect alone.

This investigation is based on cellular models of acquired resistance, which can be treated with MAPK pathway blockade. Careful analysis of clinical trial data is necessary to delineate the impact of TEAD inhibition at the cell type resolution to better understand patient response. Our study involves two major preclinical malignant mesothelioma model cell lines characterized by loss-of-function of NF2 and LATS1/2 respectively. Given the different genetic backgrounds of these two mesothelioma cell lines, additional mechanisms of resistance to TEAD inhibition may exist as well, which could potentially include effectors beyond the AP-1 family transcription factors. Here we have explored AP-1 transcription factor’s impact within the larger network of transcription factors modulated by MAPK signaling, which presents another limitation of our current findings. MAPK pathway inhibitors will likely have broader impact given the fact there will be many transcription factors that might be affected downstream. Future research is needed to pinpoint whether FOSL1 is the primary AP-1 factor influenced by MAPK pathway inhibition in the context of TEAD inhibitor resistance. This could be tested by examining whether cells expressing a MAPK-insensitive version of FOSL1, or a covalent FOSL1/JUN dimer construct, become resistant to MAPK inhibitors. In summary, our work sheds light on the possible bypass mechanisms involved with targeting the Hippo pathway in cancers and underscores the need for combination therapy of TEAD inhibitors and MAPK pathway inhibitors for sustained clinical benefit. Finally, it highlights the intricate interdependence of both these pathways in treating cancers with both MAPK and Hippo pathway mutations and resistance to their respective targeted therapies.

Methods

Ethics statement

This study complies with all relevant ethical guidelines. Animals were housed and cared for following the Guide for the Care and Use of Laboratory Animals (National Research Council, 2011). All animal experiments were conducted at Genentech, an accredited facility of the Association for Assessment and Accreditation of Laboratory Animal Care, under protocols approved by Genentech’s Institutional Animal Care and Use Committee (IACUC). The maximum allowable tumor size of 2000 mm³, as stipulated by the ethics committee/institutional review board, was not exceeded in any of the studies. Source data are available for all in vivo experiments.

Cell lines, cell viability assays

All cell lines used in this study were obtained from American Type Culture Collection (ATCC). They were maintained in RPMI 1640 supplemented with 10% FBS (Sigma-Aldrich Corporation) and 2 mM L-glutamine (Gibco Life Sciences) in a humidified incubator maintained at 37 °C with 5% CO₂. Cell line authentication was conducted as previously described²⁷, specifically as noted in that paper “for Short Tandem Repeat (STR) Profiling using the Promega PowerPlex 16 System. This was performed once when receiving new cell lines and compared to external STR profiles of cell lines (when available) to determine cell line ancestry. Cell line authentication was routinely conducted by SNP-based genotyping using Fluidigm multiplexed assays at the Genentech cell line core facility.” Cells were assessed with a Vi-CELL Cell Viability Analyzer (Beckman Coulter); viability of at least 90% was required for experiments and screening.

TEAD SMI resistant (NF2 null) NCI-H226 and (PSEN1-LATS1 fusion and 42 bp deletion in exon 5 of the LATS2) MSTO-211H cells were generated by incubation with increasing concentrations of TEAD SMI over time (0.25 μM to 3 μM at bi-weekly with 0.1–0.5 μM increments), ensuring at least a 25% confluence at all times. The generation of the resistant cell lines was done using pooled populations of survivor cells. Resistance was measured by CellTiter-Glo Luminescent Cell Viability Assays (Promega Corporation), by Incucyte confluence (mm²), and TEAD downstream RNA/protein target expression.

For treatment of cells with experimental compounds, cells were seeded at 1000 cells per 96-well. 24 h post plating, cells were treated either with a single compound or with a combination of compounds at indicated concentrations for 6 days using the HP D300 drug dispenser. Cell growth/viability was assessed using CellTiter-Glo Luminescent Cell Viability Assays (Promega Corporation), and luminescence was read by a 2104 EnVision Multilabel Plate Reader (PerkinElmer). For single compound treatments, cell viability data was collected and calculated for at least 5 replicates per time point, per condition. The viability curves were generated by fitting the data with the nonlinear regression model in GraphPad Prism.

Mouse xenograft

Female C.B-17 SCID (inbred) mice, aged 7–10 weeks at the start of the study, were used. The mice were fed ad libitum with an autoclaved rodent diet (LabDiet 5010) and housed in individually ventilated cages under a 14 h/10 h light/dark cycle. Animal rooms were maintained at a temperature of 20.0 to 26.1 °C and humidity of 30 to 70%, with 10–15 room air exchanges per hour.

For the H226-7883R xenograft study, GNE-7883-resistant cell lines were derived in vitro from NCI-H226 cells. Similarly, for the MSTO-7883R xenograft study, resistant cell lines were derived from MSTO-211H. Both H226-7883R and MSTO-7883R cells were cultured in RPMI 1640 media with 1% l-glutamine and 10% fetal bovine serum, harvested in log-phase growth, and resuspended in Hank’s Balanced Salt Solution with Matrigel (BD Biosciences) at a 1:1 ratio for in vivo inoculation. Mice were subcutaneously inoculated with 1 × 10⁷ cells in the right flank. Tumors were allowed to reach an initial volume before randomization into treatment groups to ensure matched baseline tumor sizes. Mice were treated with GNE-7883 (250 mg/kg) in sunflower oil via subcutaneous injection, or Cobimetinib (7.5 mg/kg) in MCT (0.6% Methocel A15 LV, 0.2% Tween 80) via oral gavage, or a combination of both drugs. Treatments were administered on 3× per week.

Tumor sizes and mouse body weights were recorded twice weekly. Tumor volumes were measured in two perpendicular dimensions (length and width) using Ultra Cal IV calipers (model 54-10-111; Fred V. Fowler) and calculated as: tumor size (mm³) = (length × width²) × 0.5. Body weights were measured with an Adventurer Pro AV812 scale (Ohaus Corporation), and percentage weight changes were calculated as: body weight change (%) = [(current body weight/initial body weight) − 1] × 100. Each treatment group comprised 7 mice. Sample sizes were based on historical studies and previous experiments conducted at Genentech. Analyses and comparisons of tumor growth were performed using a customized R package (version 4.3.3; R Foundation for Statistical Computing), integrating software from open-source packages as described by Forrest et al.⁶³. The term “growth contrast” represents the difference in area under the curve (AUC) based growth rates (endpoint gain integrated in time, eGaIT) between the treatment and reference groups⁶³. A more negative growth contrast value indicates a greater antitumor effect. The 95% confidence intervals were based on the fitted model and variability measures of the data. Contrasts with 95% confidence intervals not including 0 are statistically significant (P < 0.05).

RNA analyses

1–2 × 10⁵ cells/6-well were treated with compounds at indicated concentrations 12–24 h post plating. Cell lines were lysed on-plate for RNA isolation using the Qiagen RNeasy Plus Mini kit (Cat. No. 74034), and RNA concentration was determined using a NanoDrop 8000 (Thermo Fisher Scientific). The quantitative RT-PCR was performed using QuantStudio 7 Flex machine with TaqMan probes for CTGF (Hs01026927_g1), CYR61 (Hs00998500_g1), ANKRD1 (Hs00173317_m1), FOSL1 (Hs00759776_s1), and GAPDH (Hs02786624_g1) (Applied Biosystems). Relative expression of each gene to GAPDH of target genes was assessed for at least 3 biological replicates.

RNA Sequencing (RNA-Seq) was performed as previously described²⁷, specifically as noted in that paper RNA integrity was confirmed using a “Bioanalyzer 2100 (Agilent Technologies). About 500 ng of RNA was used for library synthesis using the TrueSeq RNA Sample Preparation kit v2 (Illumina). Size of the libraries was confirmed using 2200 TapeStation and High Sensitivity D1K screen tape (Agilent Technologies), and the concentration was determined by a qPCR-based method using the Library Quantification Kit (KAPA). The libraries were multiplexed and then sequenced on Illumina HiSeq2500 (Illumina) to generate an average of 56 million single-end 50-base reads per sample.

The analysis of RNA-seq data was conducted using HTSeqGenie (R package version 4.30.0⁶⁴) within the BioConductor framework in the following manner: reads characterized by low nucleotide qualities (defined as having 70% of bases with quality scores below 23) or those matching rRNA and adapter sequences were excluded. The remaining reads were aligned to the human reference genome (human: GRCh38.p10) using GSNAP (2013-10-10-v2)^65,66 with a maximum allowance of two mismatches per 75-base sequence (utilizing parameters: ‘-M 2 -n 10 -B 2 -i 1 -N 1 -w 200000 -E 1 --pairmax-rna=200000 --clip-overlap’). Transcript annotation relied on the GENCODE genes database (human: GENCODE 27). For the quantification of gene expression levels, the count of reads unambiguously mapping to the exons of each gene was computed. Differential gene expression was performed with relevant contrasts using the “voom” function of the limma package. A pre-filter was applied so that only genes with at least a median CPM (counts per million mapped reads) value of 10 in one were analyzed. P-values for other genes were left out and not included in downstream steps including the multiple testing correction. Counts were normalized using trimmed mean of M-values (TMM). Q-values were obtained by correcting P-values for multiple hypotheses using the Benjamini-Hochberg procedure. Genes were considered if they had a Q-value of less than 0.05 and were protein-coding.

Immunofluorescence, EdU incorporation and labeling

8000–10,000 cells per well were seeded in 96-well plates the day before. Treated cells with 1 uM TEAD inhibitors for 48 h; Cells were fixed in 4% paraformaldehyde for 20 min, permeabilized and blocked in PBS containing 10% FBS, 1% BSA, 0.1% Triton X-100, and 0.01% NaN₃ for 30 min, and then stained overnight in 4 °C with respective primary antibodies as noted for each experiment (YAP 1:250, CST Cat. #14074; TEAD1 1:150 BD Cat. #610923; FOSL1 1:250 ThermoFisher Cat. #PA5-66880) in blocking buffer. After PBS wash, a secondary antibody (Donkey anti-Rabbit IgG (H + L) Alexa Fluor 647, ThermoFisher Cat. #A-31573, 1:2000) was combined with Hoechst 33342 (ThermoFisher Cat. # H3570, 1:5000) in blocking buffer and applied for 20 min. Cells were imaged in PBS. For assays with EdU, cells were treated with 10 µM EdU for 20 min in 37 °C prior to fixation, and click reaction (manufacturer’s protocol, ThermoFisher Cat. #C10337) followed blocking. Samples were imaged on PerkinElmer Opera Phenix using widefield at 10x magnification (binning 2).

Pre-extraction for chromatin-bound proteins

Preextraction was done if only chromatin-bound proteins were being stained for and soluble proteins were extracted from cells. The cells were plated in 96-well plate coated with Purecol (Advanced BioMatrix, Cat. #5005), to minimize cell detachment during extraction. Required amount of Purecol was added to coat the wells and left for 2 h in RT. The collagen was removed and 8000-10000 cells were plated per well, treated the next day with GNE-7883 and incubated for 48 h. Just prior to fixation, media was aspirated off cells, and the plate was placed on ice. Cells were incubated in ice-cold pre-extraction buffer, consisting of 0.1% Triton X-100 (pre-extraction buffer was made with 1× PBS) + 1× Halt Protease Inhibitor Cocktail (ThermoFisher Cat. # 78439). After 2 mins extraction time on ice, 16% paraformaldehyde was directly added to wells with pre-extraction buffer (1:4, final 4% PFA), very gently to minimize cell detachment, and cells were fixed for 25 min at room temperature, after which the sample was treated with the general staining protocol. This pre-extraction protocol is based off the protocol published in ref. ⁶⁷.

Image analysis

Images were segmented, quantified, and analyzed using custom scripts based on previously published code in MATLAB⁶⁸. First, the uneven illumination in each imaged color was calculated by averaging the background of many wells, and then corrected via normalization using the calculated profile. A global background value in each field of view was then calculated and subtracted off for every color. Cells’ nuclear masks were segmented via combining both global and adaptive local threshold approaches using the Hoechst channel. Cells near the image borders were removed from analyses, and to increase the fidelity of segmentation, segmented areas that are too large, small, or warped were also removed from analyses. The nuclear mask was then used to quantify intensity levels in the other imaged colors. Hoechst was quantified by calculating the total integrated intensity in each cell, and all other fluorescence values were calculated from the median intensity value in each cell.

Immunoprecipitation

7 × 10⁶ cells were seeded per immunoprecipitation reaction. The following day, cells were treated with DMSO or GNE-7883. The next day, cells were washed twice with cold PBS and harvested in Kischkel buffer (50 mM Tris, pH 8.0, 150 mM NaCl, 5 mM EDTA, 1% Triton X-100) containing 1× Halt Protease Inhibitor Cocktail (ThermoFisher Cat. # 78439) and 1× Halt Phosphatase Inhibitor Cocktail (ThermoFisher Cat. # 78427). Lysates were briefly sonicated for 10 seconds at 25% power, followed by centrifugation at 4 °C for 10 min. Protein concentrations were measured using the BCA Protein Assay Kit (Pierce, Cat. # 23228), and 3–5 mg of total protein was used for each reaction. For immunoprecipitation, 10 µL (330 ng) of pan-TEAD antibody (CST Cat. #13295S) or an equal amount of Normal Rabbit IgG (CST Cat. #2729S) was added to the lysates. The mixture was rotated at 4 °C overnight. On the following day, 20 µL of Protein A/G Magnetic Beads suspension (Pierce Cat. # 88802) were added to the samples and incubated for an additional 2–4 h. The beads were washed four times with 1 mL of cold Kischkel buffer and collected on magnet. The samples transferred to fresh tubes before the final wash. Proteins were eluted by heating the samples in NuPAGE LDS sample buffer (ThermoFisher Cat. # NP0007), supplemented with β-mercaptoethanol, at 70 °C for 10 min, followed by analysis via western blotting.

Immunoblotting

For general sample preparation for western blots, 1 × 10⁶ cells were plated the day before the treatment. After incubating cells with DMSO or 1 μM G7883 for 48 h, cells were harvested with 1× RIPA Buffer (ThermoFisher Cat. # 89900) that supplemented with 1× Halt Protease Inhibitor Cocktail (ThermoFisher Cat. # 78439) and 1× Halt Phosphatase Inhibitor Cocktail (ThermoFisher Cat. # 78427). Lysates were incubated on ice for at least 10 min centrifuged at 14,000 × g for 10 min at 4 °C. BCA Protein Assay Kit (Pierce Cat. # 23228) was used to measure protein concentration. Samples were prepared with NuPAGE LDS sample buffer (ThermoFisher Cat. # NP0007) supplemented with β-mercaptoethanol and heated at 95 °C for 5 min. Equivalent amount of protein from each treatment was separated by electrophoresis through NuPAGE Bis-Tris 10% gradient gels (Invitrogen Cat. # WG1203) with NuPAGE MOPS SDS Running Buffer (Invitrogen Cat. #NP0001); proteins were transferred onto PVDF membrane using the iBlot system (Invitrogen, IB301001) and probed against various antibodies: YAP/TAZ (CST Cat. # 8418, 1:1000), pan-TEAD (CST Cat. # 13295, 1:1000), FOSL1 (CST Cat. # 5281, 1:500), CYR61 (CST Cat. # 14479, 1:1000), CTGF (CST Cat. # 86641,1:1000), GAPDH (CST Cat. # 5174, 1:2000), Actin (CST Cat. # 58169, 1:2000), dpERK (CST Cat. # 9101S, 1:1000), total ERK (CST Cat. # 9102S, 1:1000), pMEK (CST Cat. # 9121S, 1:500), total MEK (CST Cat. # 9122S, 1:500), CDK2 (CST Cat. # 2546, 1:500), CDK4 (Proteintech Cat. # 11026-1-AP, 1:1000), CDK6 (CST Cat. # 13331, 1:500), cyclin B1 (CST Cat. # 4138, 1:1000), cyclin E2 (Proteintech Cat. # 11935-1-AP, 1:1000), pRB-780 (CST Cat. # 8180, 1:1000). Membranes were washed three times with PBST and incubated with HRP-conjugated secondary antibodies. The secondary antibodies used in the study included anti-rabbit IgG, light chain specific (JacksonImmunoResearch Cat. # 211-032-171) 1:5000, anti-rabbit IgG (CST Cat. # 7074S) 1:1000, anti-mouse IgG (CST Cat. # 7076S) 1:1000. Immunoblots were developed by ECL with Supersignal West Pico Plus Chemiluminescent Substrate (ThermoFisher Cat. # 34580). Signal was detected using X-ray film or a Chemiluminescence Imaging System, Azure Biosystem c300. Images were quantified using ImageJ software (v2.14.0).

CRISPR and siRNA transfection

CRISPR transfection was performed in H226 and H226-7883R cells. Cells were seeded at 20,000/ well on a 6-well plate for 24 h. They were then transfected with the RNP complex that contains pre-designed Alt-R CRISPR-Cas9 gRNA from IDT (final concentration of 44 µM) and Cas9 protein (final concentration 36 µM) was purchased from IDT (Alt-R™ S.p. Cas9 Nuclease V3, Cat. #1081058). Multiple guide RNAs for FOSL1 were tested sgFOSL1_1: CAAGGCCTTCGACGTACCCC, sgFOSL1_2: GATTTTGCAGATGGGTCGGT and sgFOSL1_3: UGUUGAUGCUUGGCACCAGG. Guide RNA for CD81 (TGATGACGCCAACAACGCCA) was used as a control. Two guide RNAs were used for FOSL1: sgFOSL1_1 and sgFOSL1_2 at a concentration of 22 µM each. Three days post-electroporation, cells were maintained in 1 µM GNE-7883 in the background of CD81 or FOSL1 KD. After 48 h and 96 h of treatment, cells were either processed for CUT&RUN or cell viability assay respectively. siRNA transfection was performed in H226-7883R cells. Cells were seeded at 20,000/ well on a 6-well plate for 24 h. The cells were then electroporated with an siRNA (final concentration 200 nM) purchased from Dharmacon Reagents, targeting either FOSL1 (ON-TARGETplus Human FOSL1 siRNA SMARTpool, Cat. # L-004341-00-0005), YAP1 (ON-TARGETplus Human YAP1(10413) siRNA-SMARTpool Cat. # L-012200-00-0020) or NTC (ON-TARGETplus Non-targeting Pool Cat. # D-001810-10-50) as a control. Transient knockdown was confirmed by Western blot.

Multiplexed scMultiome assay

The multiplexed single-cell Multiome (scMultiome) assay was performed following the TEA-seq⁶⁹ protocol with some modifications. Briefly, the cells were cultured and treated with different compounds in 6-well plates. Upon harvesting, the cells were trypsinized to single cell suspension, washed once with PBS. Samples from different conditions were labeled by the TotalSeq-A anti human Hashtag antibodies with different hashtags (Biolegend; Clone: LNH-94; 2M2) by incubating on ice for 20 min. After washing twice with PBS, the samples were pooled together and permeabilized by the isotonic permeabilization buffer (20 mM Tris-HCl pH 7.4, RNase Inhibitor 1 U/µl, 150 mM NaCl, 3 mM MgCl₂, 0.01% digitonin) on ice for 15 min. Then the samples were washed with the same isotonic buffer without digitonin, counted and immediately processed for the tagmentation step. We followed the standard protocol of 10X Multiome assay (10X Genomics) for sample tagmentation and droplet generation. When performing the pre-amp PCR and cDNA amplification, 1 µl of 100 µM HTO additive primer (5′-GTGACTGGAGTTCAGACGTGTGCTC-3′) was added to help amplify the HTO. After cDNA amplification, the supernatant was saved from the 0.6× SPRI beads cleanup step to build the HTO libraries. The HTO library construction was done following the standard cell hashing protocol⁷⁰.

For each 10X channel, 3 types of libraries were constructed: ATAC, GEX and HTO. The 3 types of libraries were sequenced on Illumina NovaSeq and NextSeq sequencers with roughly 30 K, 20 K and 2 K raw reads per cell, respectively.

Single cell multiome data analysis

scMultiome data processing of sequencing reads was performed using our Single Cell Multiome ATAC + Gene Expression (ARCSeq) - 10X Cellranger (v2.0.0) bioinformatics pipeline which implements the CellRanger ARC seq workflow (cellranger-arc_2.0.0) using the GRCh38 genome reference. ATAC-seq (atac_fragments.tsv.gz) and scRNA-seq (raw_feature_bc_matrix.h5) matrices were generated from fastq files. Approximately 200 million read pairs were generated for ATAC and Gene expression data types with more than 92% mapping to the human genome for the whole sequence library.

A cell hashing (hashed-tag oligonucleotide (HTO)) (.hashing.csv) feature count matrix was generated with cumulus_feature_barcoding (https://github.com/lilab-bcb/cumulus_feature_barcoding) (v0.11.1). This barcode count matrix was then used to demultiplex the scRNA-seq data using a DropleUtils (v1.21.0)-based workflow with the emptyDrops(test.ambient = TRUE, by.rank = 30000) and hashedDrops(doublet.nmads = 3, doublet.min = 2, confident.nmads = 1, confident.min = 1) key steps and parameters. At this point, 13,587 raw cells with valid barcodes were kept for downstream analyses.

ATAC-seq data was processed with ArchR v1.0.3. Briefly, low quality cells were removed (minTSS>4 and minFrags>1000) and arrow files were generated followed by generation of Tile and Gene Score matrices using a tile size of 500 bp. Doublets were identified and filtered out using ArchR’s addDoubletScores() function with default parameters.

Filtered ATAC-seq data was merged with the demultiplexed scRNA-seq data which yielded a total of 4952 cells and was assessed for quality (SI) before moving forward. Cells were clustered using the previously generated 500 bp peaks tile matrix (addIterativeLSI(), addClusters() and addUMAP()). Resulting clusters were visualized and labeled according to their experimental conditions using the following names: Sensitive_side, Sensitive_DMSO, Sensitive_GNE7883 and Resistant_side and Resistant_main. Peaks were generated using addReproduciblePeakSet() with macs2. Peaks were annotated with addPeakAnnotations()(using public databases from chipatlas.org; encodeproject.org) followed by addDeviationsMatrix(). Gene expression figures were generated from the gene expression matrix data and chromVAR figures from the annotated peak matrix.

Gene activity was computed with Epiregulon (https://github.com/xiaosaiyao/epiregulon), a method that assigns regulatory elements (REs) to their putative target genes (TGs) if the chromatin accessibility and gene expressions are highly correlated across cell clusters⁷¹. They were then interrogated for transcription factor (TF) occupancy based on a compilation of public TF ChIP-seq binding sites from ChIP-Atlas (chip-atlas.org) and ENCODE TF (encodeproject.org) ChIP-seq data. This mapped out the preliminary TF-RE-TG regulatory relationships. Further pruning of the gene regulatory network was also performed in order to remove spurious gene regulatory relationships not supported by the data, mainly because of the different experimental designs in which the public ChIP-seq data was most probably generated. This was done by applying tests of independence (binomial or chi-square tests) which identify TF-RE-TG triplets jointly occurring at probabilities deviating from the expected probabilities should these three events occur independently. In other words, it was assumed that a TF-RE-TG regulatory relationship holds true if expression of the TG, chromatin accessibility at RE and expression of the TG tend to co-exist in the same cells. We recognize that there could be time delays between chromatin accessibility and the expression of target genes in true RE-TG relationships, however, we want to be conservative in our inference and rule out unsubstantiated TF-RE-TG relationships.

Once network pruning was achieved, the strength of regulation was further estimated. This was achieved using the effect size of the Wilcoxon test between target gene expression in cells jointly expressing all 3 elements vs cells that do not. The activity of a given transcription factor corresponds to the weighted sum of all its target genes where the weights correspond to the strength of regulation. The target genes of the identified TFs were queried against known pathways to predict the TFs functions. Differential network analysis between cluster-specific networks, where the edges between two networks were subtracted, was done followed by ranking TFs by degree of centrality between the pair of compared networks.

CUT&RUN library preparation

H226/H226-7883R cells were trypsinized into single-cell suspension with 0.05% Trypsin, EDTA 0.5 mM. 5 × 10⁵ cells were used for each reaction. We used the CUTANA ChIC/CUT&RUN Kit from Epicypher (#14-1048) following manufacturer’s instructions. ConA Beads were briefly resuspended and 11 μl of beads per reaction was transferred to a 1.5 ml tube. Beads were resuspended in 100 μl bead activation buffer per reaction and washed 2 ×. Then resuspended in 11 μl buffer per reaction and aliquoted 10 μl into 8-strip PCR tubes, and kept on ice. Harvested cells were spun down at 600 × g for 3 min at room temperature. Cells were then resuspended in 100 μl/reaction of wash buffer and aliquoted into 8 strip tubes with the 10 μl of activated ConA beads. After 10 mins incubation, 50 μl of cold antibody buffer was added. YAP antibody (CST Cat. #14074) was used at 1:50 dilution, TEAD1 antibody (BD Cat. #610923) or FOSL1 antibody (ThermoFisher Cat. #PA5-66880) was added at 0.5 μg per reaction. Tubes were incubated on nutator overnight at 4 °C. The next day, tubes were collected and washed 2× with 200 μl cell permeabilization buffer. Beads were resuspended in 50 μl cell permeabilization buffer and 2.5 μl pAG-MNase was added. After 10 min incubation at room temperature, beads were washed with another 200 μl cell permeabilization buffer 2×. Finally, beads were resuspended in 50 μl cell permeabilization buffer. 1 μl of 100 mM CaCl₂ was added to activate the enzyme. Tubes were then incubated for 2 h on nutator at 4 °C. Finally, the reaction was stopped with 33 μl of stop master mix. DNA was then purified and quantified using a Qubit fluorometer. For library preparation, we used the CUTANA CUT&RUN Library Prep Kit from Epicypher (#14-1001). End Repair step, Adapter ligation and U-excision were performed following manufacturer’s instructions. We used 1× SPRIselect beads for DNA cleanup. After indexing PCR, we then used 1 × SPRIselect beads for DNA cleanup prior to sequencing. Analysis of the library fragment size was done using Tapestation. 10 million Paired End Reads (PE) per sample.

CUT&RUN data analysis

For each sequence library, raw reads were aligned on the hg38 human genome reference using Bowtie2 (v2.4.1) followed by alignments sorting (samtools v1.9) of resulting bam files with filtering out of unmapped reads and keeping reads with mapping quality higher than 30. Duplicates were removed with Picard’s (v2.23.9) MarkDuplicates function followed by indexation of resulting bam files with samtools. For each bam file, genome coverage was computed with bedtools (v2.29.2) followed by generation of bigwig (wigToBigWig v357) files. Peaks were called with macs2 (v2.7.1) for each treatment sample using their corresponding IgG control. Downstream analyses were performed with DiffBind v3.11.1 with default parameters, except for summits=250 in dba.count(). dba.contrast() and dba.analyzed() were used to compute significant differential peaks with DESeq2. Tornado plots were generated with the plotHeatmap utility from deepTools v3.5.0⁷². Wiggletools v1.2⁷³ was used to compute the mean from the replicates’ bigwig track files.

ATAC-seq library preparation

ATAC-seq was performed essentially as described previously⁷⁴. Briefly, cryo-preserved cells were thawed, washed with 1X PBS and counted. About 100,00 cells were then lysed for 5 min on ice in 100 µl lysis buffer (10 mM Tris (pH 7.5), 10 mM NaCl, 3 mM MgCl₂, 0.1% NP-40, 0.1% Tween- 20 and 0.01% digitonin). 1 ml wash buffer (10 mM Tris (pH 7.5), 10 mM NaCl, 3 mM MgCl₂ and 0.1% Tween-20) was then added and the pellets were spun down at 500 g for 10 min at 4 °C. The supernatant was discarded and the pellet was resuspended in 50 µl tagmentation buffer (25 µl 2× TD buffer, 16.5 µl PBS, 0.5 µl 10% Tween-20, 0.5 µl 1% digitonin, 2.5 µl Tn5 transposase and 5 µl H₂O). The tubes were incubated at 37 °C for 30 min, followed by DNA isolation using the Qiagen MinElute Cleanup Kit (Cat. # 28206). The DNA was then amplified using NEBNext 2X PCR Master mix (Cat. # M0541L) and analyzed by TapeStation (Agilent) before submission for sequencing.

ATAC-seq analysis

ATAC-seq reads were analyzed using the ENCODE ATAC-seq pipeline (version 1). Briefly, reads were trimmed of adapters using cutadapt (version 1.9.1) and mapped to hg38 using Bowtie 2 (version 2.2.6). Bam files were converted to tagAlign format, which was then adjusted for Tn5 by shifting +4 bp for positive strands and −5 bp for negative strands. TagAlign files were used to call peaks using MACS2 (version 2.1.0), and those peaks with P < 1 × 10⁻⁶ were retained for differential analysis with DiffBind (version 3.0.13). For subsequent comparative analyses, only peaks called in at least two samples were retained. Differential accessibility analysis was performed using either the R package DiffBind (version 3.0.13). Differential peaks were defined by an absolute fold change of >2 and called peaks by MACS2 in at least two samples with gained accessibility (the Benjamini–Hochberg procedure was used for multiple hypothesis correction with a cut-off of FDR < 0.01). The genomic distribution of differential peaks was annotated with ChIPpeakAnno (version 3.24.1). Bigwigs corresponding to a fold change against the background control were generated by MACS2 and used to generate heatmaps with the heatmap function of deepTools (version 3.5.0). Motif enrichment analysis was performed using either HOMER (version 4.10) with all peaks as the background. The peaks were assigned using Poly-Enrich with distal enhancer–gene target links (>5 kilobases from the transcription start site) plus 5-kilobase locus definitions. Gene set enrichment was searched against Gene Ontology, MsigDB Oncogenic and Hallmark.

ChIP-seq

ChIP-seq was performed as previously described with the following modifications⁷⁵. NCI-H226 mesothelioma cells (20 ×10⁶) were double crosslinked with 50 mM DSG (disuccinimidyl glutarate, #C1104 - ProteoChem) for 30 min followed by 10 min of 1% formaldehyde. Formaldehyde was quenched by the addition of glycine. Nuclei were isolated with ChIP lysis buffer (1% Triton X-100, 0.1% SDS, 150 mM NaCl, 1 mM EDTA, and 20 mM Tris, pH 8.0). Nuclei were sheared with Covaris sonicator using the following setup: Fill level—10, Duty Cycle—15, PIP—350, Cycles/Burst—200, Time—8-min). Sheared chromatin was immunoprecipitated overnight with the following antibodies: FOSL1 (ThermoFisher Cat. #PA5-66880), YAP (CST Cat. #14074S). Antibody chromatin complexes were pulled down with Protein A magnetic beads and washed twice in IP wash buffer I. (1% Triton X-100, 0.1% SDS, 150 mM NaCl, 1 mM EDTA, 20 mM Tris, pH 8.0, and 0.1% NaDOC), twice in IP wash buffer II. (1% Triton, 0.1% SDS, 500 mM NaCl, 1 mM EDTA, 20 mM Tris, pH 8.0, and 0.1% NaDOC), twice in IP wash buffer III. (0.25 M LiCl, 0.5% NP-40, 1 mM EDTA, 20 mM Tris, pH 8.0, 0.5% NaDOC) and once in TE buffer (10 mM EDTA and 200 mM Tris, pH 8.0). DNA was eluted from the beads by vigorous shaking for 20-min in elution buffer (100 mM NaHCO3, 1% SDS). DNA was de-crosslinked overnight at 65 °C and purified with MinElute PCR purification kit (Qiagen). DNA was quantified by Qubit and equal amounts (2 ng) of DNA was used for sequencing library construction with the Ovation Ultralow Library System V2 (Tecan) using 15 PCR cycles according to the manufacturer’s recommendations. Libraries were sequenced with Illumina Nextseq, using paired-end 50 bp read configuration.

ChIP-seq analysis

ChIP-seq results were analyzed using the ENCODE ChIP-seq pipeline (v2.2.1)⁷⁶. ChIP-seq reads were aligned to the human reference genome (hg38) using Bowtie2 (v2.3.4.3)⁷⁷. Aligned reads were then filtered for quality and duplicates using samtools (v1.9)⁷⁸ and Picard (Broad Institute - v2.20.7). The SPP peak caller was used to call ChIP-seq peaks for FOSL1 and YAP, and input was used to assess the background of the experiments⁷⁹. Peak sets were filtered using a list of genomic regions that contain anomalous, unstructured, or experiment independent high signal⁸⁰. ChIP-seq bam and bed files were then used to call differential peaks in the genome by using DiffBind (v3.12.0)⁸¹. Briefly, Differential Binding Analysis object was created by loading the bam and bed files with the dba() function. Reads were counted with the dba.count() function, followed by depth normalization using the dba.normalize() function, and differential peaks were called across the conditions with the dba.analyze() function using DESeq2 and the following statistical parameters (FDR < 0.01). Differential FOSL1 peaks were annotated to the nearest genes by ChIPseeker (v1.38.0)⁸², and pathway analysis was performed with Ingenuity Pathway Analyzer (IPA - Qiagen) using the core analysis feature with default settings.

Whole exome sequencing

DNA from both the sensitive and pooled population of TEAD inhibitor resistant NCI-H226 cell lines were extracted and fragmented into 200 bp fragments. Whole exome was captured using Agilent Sureselect Human All Exon V7 (S31285117) and sequenced on a HiSeq1 system (Illumina) using a 2 × 75 bp configuration which yielded an average of 86 million reads per library. Reads were aligned to GRCh38 using BWA (v0.7.15-r1140). In accordance with the Genome Analysis Toolkit (GATK) Best Practices recommendations⁸³, Picard Tools v2.18 was employed to eliminate redundant reads, and base score recalibration was carried out using GATK (v4.1.4.1).

Copy number analysis

Copy numbers were determined using the GATK best practices for detecting copy ratio alterations (GATK 4.2.6.1, Article#11682 & Article#11683). In short, exome capture regions underwent preprocessing using PreprocessIntervals with a padding of 250 bp. Overlapping intervals were merged. Reads were then counted in each of the resulting intervals for all samples using CollectReadCounts. GATK CNV supports single sample panel of normals (PoN) by using CreateReadCountPanelOfNormals (-minimum-interval-median-percentile 5) to create a single-sample projection which can be used for denoising via DenoiseReadCounts. The denoising process standardizes reads to the PoN median count and centers the data around 1. The denoised read counts are then grouped into contiguous segments using a Gaussian-kernel binary segmentation algorithm via ModelSegements (default settings, without allelic count data). The mean log2(copy number) value is then assigned to the genes within that segment and used for gene-level copy number comparisons. The parental line was used to generate the PoN and denoise the read counts from the resistant lines.

Mutations analysis

Starting from the processed read alignments described above, variant calling using MuTect2 (from GATK v4.1.4.1), Strelka (v.1.0.15⁸⁴) and LoFreq (v2.1.3.1⁸⁵) was performed. MuTect2 was executed by using gatk Mutect2 followed by FilterMutectCalls. Strelka variants were obtained by running the software with default bwa parameters. For LoFreq, sensitive and resistant bam files were first processed with the indelqual command followed by the somatic command to generate variant calls. Here we specifically investigated for mutations potentially present in the TEAD1, TEAD2, TEAD3, TEAD4, FOSL1, YAP1 and WWTR1 genes. Obtained variants from all three callers were annotated with transvar (v.2.5.10.20211024⁸⁶). For each of the seven genes of interest, the transvar output was filtered to keep only the transcripts associated with the canonical UniProt accession identifiers. The predicted mutations (position and nature of amino acid modification) on the protein sequences were used for protein structural analysis.

Statistical analysis

The statistical analysis for the in vitro studies was performed using one-way ANOVA followed by posthoc Tukey’s test, unless otherwise mentioned. All in vitro experiments were repeated at least three times. Viability dose response curve (mean ± SD) was assessed using two-way ANOVA followed by posthoc Tukey’s multiple comparison test. Statistical analysis for single cell imaging experiments was performed using two-sided Wilcoxon rank-sum test on 500 randomly selected cells using datasample function in MATLAB. A P value of <0.05 was considered statistically significant. Significant values as well as number of replicates are noted for each experiment in the respective figure legends. All other statistical methods are described in their respective method sections.

Reporting summary

Further information on research design is available in the Nature Portfolio Reporting Summary linked to this article.

Data availability

Sequence data generated as part of this study has been deposited in the Gene Expression Omnibus repository under accession identifier GSE248208. Source data are provided with this paper.

Code availability

Code supporting the description of analyses performed as part of this study is available at https://github.com/xiaosaiyao/TEADi_resistance.

References

Zheng, Y. & Pan, D. The Hippo signaling pathway in development and disease. Dev. Cell 50, 264–282 (2019).
Article CAS PubMed Central MATH PubMed Google Scholar
Dey, A., Varelas, X. & Guan, K. L. Targeting the Hippo pathway in cancer, fibrosis, wound healing and regenerative medicine. Nat. Rev. Drug Discov. 19, 480–494 (2020).
Article CAS PubMed Central PubMed Google Scholar
Yu, F. X. & Guan, K. L. The Hippo pathway: regulators and regulations. Genes Dev. 27, 355–371 (2013).
Article CAS PubMed Central MATH PubMed Google Scholar
Zanconato, F., Cordenonsi, M. & Piccolo, S. YAP/TAZ at the roots of cancer. Cancer Cell 29, 783–803 (2016).
Article CAS PubMed Central PubMed Google Scholar
Kanai, F. E. A. TAZ: a novel transcriptional co‑activator regulated by interactions with 14‑3‑3 and PDZ. Embo j 19, 6778–6791 (2000).
Zhao, B., Li, L., Tumaneng, K., Wang, C. Y. & Guan, K. L. A coordinated phosphorylation by Lats and CK1 regulates YAP stability through SCF(beta-TRCP). Genes Dev. 24, 72–85 (2010).
Article CAS PubMed Central PubMed Google Scholar
Ren, F., Zhang, L. & Jiang, J. Hippo signaling regulates Yorkie nuclear localization and activity through 14-3-3 dependent and independent mechanisms. Dev. Biol. 337, 303–312 (2010).
Article CAS MATH PubMed Google Scholar
Liu, C. Y. et al. The hippo tumor pathway promotes TAZ degradation by phosphorylating a phosphodegron and recruiting the SCFbeta-TrCP E3 ligase. J. Biol. Chem. 285, 37159–37169 (2010).
Article CAS PubMed Central PubMed Google Scholar
Oh, H. & Irvine, K. D. In vivo regulation of Yorkie phosphorylation and localization. Development 135, 1081–1088 (2008).
Article CAS MATH PubMed Google Scholar
Zhao, B. et al. TEAD mediates YAP-dependent gene induction and growth control. Genes Dev. 22, 1962–1971 (2008).
Article CAS PubMed Central MATH PubMed Google Scholar
Currey, L., Thor, S. & Piper, M. TEAD family transcription factors in development and disease. Development 148, dev196675 (2021).
Murakami, H. et al. LATS2 is a tumor suppressor gene of malignant mesothelioma. Cancer Res. 71, 873–883 (2011).
Article CAS MATH PubMed Google Scholar
Bueno, R. et al. Comprehensive genomic analysis of malignant pleural mesothelioma identifies recurrent mutations, gene fusions and splicing alterations. Nat. Genet. 48, 407–416 (2016).
Article CAS MATH PubMed Google Scholar
Sekido, Y. & Sato, T. NF2 alteration in mesothelioma. Front Toxicol. 5, 1161995 (2023).
Article PubMed Central MATH PubMed Google Scholar
Harvey, K. F., Zhang, X. & Thomas, D. M. The Hippo pathway and human cancer. Nat. Rev. Cancer 13, 246–257 (2013).
Article CAS MATH PubMed Google Scholar
Nguyen, C. D. K. & Yi, C. YAP/TAZ signaling and resistance to cancer therapy. Trends Cancer 5, 283–296 (2019).
Article CAS PubMed Central MATH PubMed Google Scholar
Holden, J. K. et al. Small molecule dysregulation of TEAD lipidation induces a dominant-negative inhibition of hippo pathway signaling. Cell Rep. 31, 107809 (2020).
Article CAS MATH PubMed Google Scholar
Kaneda, A. et al. The novel potent TEAD inhibitor, K-975, inhibits YAP1/TAZ-TEAD protein-protein interactions and exerts an anti-tumor effect on malignant pleural mesothelioma. Am. J. Cancer Res. 10, 4399–4415 (2020).
CAS PubMed Central MATH PubMed Google Scholar
Calses, P. C., Crawford, J. J., Lill, J. R. & Dey, A. Hippo pathway in cancer: aberrant regulation and therapeutic opportunities. Trends Cancer 5, 297–307 (2019).
Article CAS PubMed Google Scholar
Zagiel, B., Melnyk, P. & Cotelle, P. Progress with YAP/TAZ-TEAD inhibitors: a patent review (2018-present). Expert Opin. Ther. Pat. 32, 899–912 (2022).
Article CAS PubMed Google Scholar
Tang, T. T. et al. Small molecule inhibitors of TEAD auto-palmitoylation selectively inhibit proliferation and tumor growth of NF2-deficient mesothelioma. Mol. Cancer Ther. 20, 986–998 (2021).
Article CAS MATH PubMed Google Scholar
Noland, C. L. et al. Palmitoylation of TEAD transcription factors is required for their stability and function in Hippo pathway signaling. Structure 24, 179–186 (2016).
Article CAS MATH PubMed Google Scholar
Pobbati, A. V., Kumar, R., Rubin, B. P. & Hong, W. Therapeutic targeting of TEAD transcription factors in cancer. Trends Biochem. Sci. 48, 450–462 (2023).
Article CAS PubMed Google Scholar
Hagenbeek, T. J. et al. An allosteric pan-TEAD inhibitor blocks oncogenic YAP/TAZ signaling and overcomes KRAS G12C inhibitor resistance. Nat. Cancer 4, 812–828 (2023).
Article CAS PubMed Central PubMed Google Scholar
Sekido, Y. et al. Neurofibromatosis type 2 (NF2) gene is somatically mutated in mesothelioma but not in lung cancer. Cancer Res. 55, 1227–1231 (1995).
CAS MATH PubMed Google Scholar
Miyanaga, A. et al. Hippo pathway gene mutations in malignant mesothelioma: revealed by RNA and targeted exon sequencing. J. Thorac. Oncol. 10, 844–851 (2015).
Article CAS MATH PubMed Google Scholar
Pham, T. H. et al. Machine-learning and chemicogenomics approach defines and predicts cross-talk of Hippo and MAPK pathways. Cancer Discov. 11, 778–793 (2021).
Article CAS MATH PubMed Google Scholar
Liberzon, A. et al. The Molecular Signatures Database (MSigDB) hallmark gene set collection. Cell Syst. 1, 417–425 (2015).
Article CAS PubMed Central MATH PubMed Google Scholar
Subramanian, A. et al. Gene set enrichment analysis: a knowledge-based approach for interpreting genome-wide expression profiles. Proc. Natl Acad. Sci. USA 102, 15545–15550 (2005).
Article ADS CAS PubMed Central PubMed Google Scholar
Salic, A. & Mitchison, T. J. A chemical method for fast and sensitive detection of DNA synthesis in vivo. Proc. Natl Acad. Sci. USA 105, 2415–2420 (2008).
Article ADS CAS PubMed Central MATH PubMed Google Scholar
Trang, H. et al. Anwesha Dey Targeting the Hippo pathway in cancers via ubiquitination dependent TEAD degradation. eLife. https://doi.org/10.7554/eLife.92450.2 (2024).
Aktas, H., Cai, H. & Cooper, G. M. Ras links growth factor signaling to the cell cycle machinery via regulation of cyclin D1 and the Cdk inhibitor p27KIP1. Mol. Cell Biol. 17, 3850–3857 (1997).
Article CAS PubMed Central MATH PubMed Google Scholar
Chung, M. et al. Transient hysteresis in CDK4/6 activity underlies passage of the restriction point in G1. Mol. Cell 76, 562–573 e564 (2019).
Article CAS PubMed Central MATH PubMed Google Scholar
Chellappan, S. P., Hiebert, S., Mudryj, M., Horowitz, J. M. & Nevins, J. R. The E2F transcription factor is a cellular target for the RB protein. Cell 65, 1053–1061 (1991).
Article CAS PubMed Google Scholar
Jiang, X. et al. Expression and function of FRA1 protein in tumors. Mol. Biol. Rep. 47, 737–752 (2020).
Article CAS MATH PubMed Google Scholar
Shreberk-Shaked, M. & Oren, M. New insights into YAP/TAZ nucleo-cytoplasmic shuttling: new cancer therapeutic opportunities? Mol. Oncol. 13, 1335–1341 (2019).
Article PubMed Central PubMed Google Scholar
Dong, J. et al. Elucidation of a universal size-control mechanism in Drosophila and mammals. Cell 130, 1120–1133 (2007).
Article CAS PubMed Central MATH PubMed Google Scholar
Zhao, B. et al. Inactivation of YAP oncoprotein by the Hippo pathway is involved in cell contact inhibition and tissue growth control. Genes Dev. 21, 2747–2761 (2007).
Article CAS PubMed Central MATH PubMed Google Scholar
Delwig, A, Ishisoko, N. & Blake, R. A. Cellular target engagement assays for small-molecule drug discovery. Medicinal Chemistry Reviews, Vol. 57 (eds Rudolph, J. & Bronson, J. J.) 419–440 (American Chemical Society, 2022).
Casalino, L., De Cesare, D. & Verde, P. Accumulation of Fra-1 in ras-transformed cells depends on both transcriptional autoregulation and MEK-dependent posttranslational stabilization. Mol. Cell Biol. 23, 4401–4415 (2003).
Article CAS PubMed Central PubMed Google Scholar
Maurus, K. et al. The AP-1 transcription factor FOSL1 causes melanocyte reprogramming and transformation. Oncogene 36, 5110–5121 (2017).
Article CAS MATH PubMed Google Scholar
Shin, S., Dimitri, C. A., Yoon, S. O., Dowdle, W. & Blenis, J. ERK2 but not ERK1 induces epithelial-to-mesenchymal transformation via DEF motif-dependent signaling events. Mol. Cell 38, 114–127 (2010).
Article CAS PubMed Central PubMed Google Scholar
Barbosa, I. A. M. et al. Cancer lineage-specific regulation of YAP responsive elements revealed through large-scale functional epigenomic screens. Nat. Commun. 14, 3907 (2023).
Article ADS CAS PubMed Central MATH PubMed Google Scholar
Paul, S., Xie, S., Yao, X. & Dey, A. Transcriptional regulation of the hippo pathway: current understanding and insights from single-cell technologies. Cells 11, 2225 (2022).
Zanconato, F. et al. Genome-wide association between YAP/TAZ/TEAD and AP-1 at enhancers drives oncogenic growth. Nat. Cell Biol. 17, 1218–1227 (2015).
Article CAS PubMed Central PubMed Google Scholar
Liu, X. et al. Tead and AP1 Coordinate Transcription and Motility. Cell Rep. 14, 1169–1180 (2016).
Article CAS PubMed Central MATH PubMed Google Scholar
Maglic, D. et al. YAP-TEAD signaling promotes basal cell carcinoma development via a c-JUN/AP1 axis. EMBO J. 37, e98642 (2018).
Koo, J. H. et al. Induction of AP-1 by YAP/TAZ contributes to cell proliferation and organ growth. Genes Dev. 34, 72–86 (2020).
Article CAS PubMed Central PubMed Google Scholar
He, L. et al YAP and TAZ are transcriptional co-activators of AP-1 proteins and STAT3 during breast cellular transformation. Elife 10, e67312 (2021).
Kurppa, K. J. et al. Treatment-Induced Tumor Dormancy through YAP-Mediated Transcriptional Reprogramming of the Apoptotic Pathway. Cancer Cell 37, 104–122. e112 (2020).
Article CAS PubMed Central PubMed Google Scholar
Nilsson, M. B. et al. A YAP/FOXM1 axis mediates EMT-associated EGFR inhibitor resistance and increased expression of spindle assembly checkpoint components. Sci. Transl. Med. 12, eaaz4589 (2020).
Tsuji, T. et al. YAP1 mediates survival of ALK-rearranged lung cancer cells treated with alectinib via pro-apoptotic protein regulation. Nat. Commun. 11, 74 (2020).
Article ADS CAS PubMed Central MATH PubMed Google Scholar
Yun, M. R. et al. Targeting YAP to overcome acquired resistance to ALK inhibitors in ALK-rearranged lung cancer. EMBO Mol. Med. 11, e10581 (2019).
Article CAS PubMed Central PubMed Google Scholar
Lin, L. et al. The Hippo effector YAP promotes resistance to RAF- and MEK-targeted cancer therapies. Nat. Genet 47, 250–256 (2015).
Article CAS PubMed Central PubMed Google Scholar
Li, Z. et al. Loss of the FAT1 tumor suppressor promotes resistance to CDK4/6 inhibitors via the Hippo pathway. Cancer Cell 34, 893–905.e898 (2018).
Article PubMed Central MATH PubMed Google Scholar
Kapoor, A. et al. Yap1 activation enables bypass of oncogenic Kras addiction in pancreatic cancer. Cell 158, 185–197 (2014).
Article CAS PubMed Central MATH PubMed Google Scholar
Shao, D. D. et al. KRAS and YAP1 converge to regulate EMT and tumor survival. Cell 158, 171–184 (2014).
Article CAS PubMed Central MATH PubMed Google Scholar
Diamond, E. L. et al. Efficacy of MEK inhibition in patients with histiocytic neoplasms. Nature 567, 521–524 (2019).
Article ADS CAS PubMed Central MATH PubMed Google Scholar
Yen, I. et al. ARAF mutations confer resistance to the RAF inhibitor belvarafenib in melanoma. Nature 594, 418–423 (2021).
Article ADS CAS MATH PubMed Google Scholar
Shin, S. J. et al. A phase Ib trial of belvarafenib in combination with cobimetinib in patients with advanced solid tumors: Interim results of dose-escalation and patients with NRAS-mutant melanoma of dose-expansion. J. Clin. Oncol. 39, Issue 15_suppl, 3007 (2021).
Talotta, F., Casalino, L. & Verde, P. The nuclear oncoprotein Fra-1: a transcription factor knocking on therapeutic applications’ door. Oncogene 39, 4491–4506 (2020).
Article CAS MATH PubMed Google Scholar
Casalino, L., Talotta, F., Cimmino, A. & Verde, P. The Fra-1/AP-1 oncoprotein: from the “Undruggable” transcription factor to therapeutic targeting. Cancers 14, 1480 (2022).
Forrest, W. F. et al. Generalized additive mixed modeling of longitudinal tumor growth reduces bias and improves decision making in translational oncology. Cancer Res. 80, 5089–5097 (2020).
Article CAS MATH PubMed Google Scholar
Gregoire Pau, J. R. HTSeqGenie: a NGS analysis pipeline. R package version 4.30.0. (2023).
Wu, T. D. & Nacu, S. Fast and SNP-tolerant detection of complex variants and splicing in short reads. Bioinformatics 26, 873–881 (2010).
Article CAS PubMed Central MATH PubMed Google Scholar
Wu, T. D., Reeder, J., Lawrence, M., Becker, G. & Brauer, M. J. GMAP and GSNAP for genomic sequence alignment: enhancements to speed, accuracy, and functionality. Methods Mol. Biol. 1418, 283–334 (2016).
Article PubMed Google Scholar
Ratnayeke, N., Baris, Y., Chung, M., Yeeles, J. T. P. & Meyer, T. CDT1 inhibits CMG helicase in early S phase to separate origin licensing from DNA synthesis. Mol. Cell 83, 26–42.e13 (2023).
Article CAS PubMed Central PubMed Google Scholar
Liu, C., Kudo, T., Ye, X. & Gascoigne, K. Cell-to-cell variability in Myc dynamics drives transcriptional heterogeneity in cancer cells. Cell Rep. 42, 112401 (2023).
Article CAS MATH PubMed Google Scholar
Swanson, E. et al. Simultaneous trimodal single-cell measurement of transcripts, epitopes, and chromatin accessibility using TEA-seq. Elife 10, e63632 (2021).
Stoeckius, M. et al. Cell Hashing with barcoded antibodies enables multiplexing and doublet detection for single cell genomics. Genome Biol. 19, 224 (2018).
Article CAS PubMed Central PubMed Google Scholar
Włodarczyk, T. et al. Epiregulon: Inference of single-cell transcription factor activity to dissect mechanisms of lineage plasticity and drug response. Preprint at bioRxiv https://doi.org/10.1101/2023.11.27.568955 (2024).
Ramirez, F. et al. deepTools2: a next generation web server for deep-sequencing data analysis. Nucleic Acids Res. 44, W160–W165 (2016).
Article CAS PubMed Central MATH PubMed Google Scholar
Zerbino, D. R., Johnson, N., Juettemann, T., Wilder, S. P. & Flicek, P. WiggleTools: parallel processing of large collections of genome-wide datasets for visualization and statistical analysis. Bioinformatics 30, 1008–1009 (2014).
Article CAS PubMed Google Scholar
Corces, M. R. et al. An improved ATAC-seq protocol reduces background and enables interrogation of frozen tissues. Nat. Methods 14, 959–962 (2017).
Article CAS PubMed Central PubMed Google Scholar
Daniel, B., Balint, B. L., Nagy, Z. S. & Nagy, L. Mapping the genomic binding sites of the activated retinoid X receptor in murine bone marrow-derived macrophages using chromatin immunoprecipitation sequencing. Methods Mol. Biol. 1204, 15–24 (2014).
Article CAS MATH PubMed Google Scholar
Hitz, B. C. et al. The ENCODE uniform analysis pipelines. Preprint at bioRxiv https://doi.org/10.1101/2023.04.04.535623 (2023).
Langmead, B. & Salzberg, S. L. Fast gapped-read alignment with Bowtie 2. Nat. Methods 9, 357–359 (2012).
Article CAS PubMed Central MATH PubMed Google Scholar
Danecek, P. et al. Twelve years of SAMtools and BCFtools. Gigascience 10, giab008 (2021).
Kharchenko, P. V., Tolstorukov, M. Y. & Park, P. J. Design and analysis of ChIP-seq experiments for DNA-binding proteins. Nat. Biotechnol. 26, 1351–1359 (2008).
Article CAS PubMed Central MATH PubMed Google Scholar
Amemiya, H. M., Kundaje, A. & Boyle, A. P. The ENCODE blacklist: identification of problematic regions of the genome. Sci. Rep. 9, 9354 (2019).
Article ADS PubMed Central MATH PubMed Google Scholar
Ross-Innes, C. S. et al. Differential oestrogen receptor binding is associated with clinical outcome in breast cancer. Nature 481, 389–393 (2012).
Article ADS CAS PubMed Central MATH PubMed Google Scholar
Wang, Q. et al. Exploring epigenomic datasets by ChIPseeker. Curr. Protoc. 2, e585 (2022).
Article CAS PubMed Google Scholar
Van der Auwera, G. A. et al. From FastQ data to high confidence variant calls: the Genome Analysis Toolkit best practices pipeline. Curr. Protoc. Bioinform. 43, 11 10 11–11 10 33 (2013).
Google Scholar
Kim, S. et al. Strelka2: fast and accurate calling of germline and somatic variants. Nat. Methods 15, 591–594 (2018).
Article CAS MATH PubMed Google Scholar
Wilm, A. et al. LoFreq: a sequence-quality aware, ultra-sensitive variant caller for uncovering cell-population heterogeneity from high-throughput sequencing datasets. Nucleic Acids Res. 40, 11189–11201 (2012).
Article CAS PubMed Central MATH PubMed Google Scholar
Zhou, W. et al. TransVar: a multilevel variant annotator for precision genomics. Nat. Methods 12, 1002–1003 (2015).
Article CAS PubMed Central MATH PubMed Google Scholar

Download references

Acknowledgements

We thank the Genentech cell repository and compound management groups for sharing and managing resources. We thank Bruno Alicke for discussions and technical assistance with in vivo studies. We thank the Genentech Postdoctoral Training Program for support. This work was supported by Genentech’s internal funding.

Author information

These authors contributed equally: Thijs J. Hagenbeek, Julien Tremblay.

Authors and Affiliations

Department of Discovery Oncology, Genentech Inc, South San Francisco, CA, USA
Sayantanee Paul, Thijs J. Hagenbeek, Christy Ong, Chad Liu, Alissa D. Guarnaccia, Luca Gerosa, Matthew Grimmer, Shiqi Xie, Xiaosai Yao & Anwesha Dey
gRED Computational Sciences, Genentech Inc, South San Francisco, CA, USA
Julien Tremblay, Luca Gerosa, Matthew Grimmer & Xiaosai Yao
Department of Proteomic and Genomic Technologies, Genentech Inc, South San Francisco, CA, USA
Vasumathi Kameswaran, Alissa D. Guarnaccia, Yuxin Liang & Bence Daniel
Roche Informatics, Hoffman-La Roche Canada, Mississauga, ON, Canada
James A. Mondo
Department of Structural Biology, Genentech Inc, South San Francisco, CA, USA
Peter L. Hsu
Department of Research Oncology, Genentech Inc, South San Francisco, CA, USA
Noelyn M. Kljavin
Roche Global IT Solution Centre, Roche, Warsaw, Poland
Bartosz Czech & Janina Smola
Department of Early Discovery Biochemistry, Genentech Inc, South San Francisco, CA, USA
Dieu An H. Nguyen
Department of Translational Oncology, Genentech Inc, South San Francisco, CA, USA
Jennifer A. Lacap
Department of Translational Medicine, Genentech Inc, South San Francisco, CA, USA
Trang H. Pham
Department of Biochemical and Cellular Pharmacology, Genentech Inc, South San Francisco, CA, USA
Robert A. Blake

Authors

Sayantanee Paul
View author publications
Search author on:PubMed Google Scholar
Thijs J. Hagenbeek
View author publications
Search author on:PubMed Google Scholar
Julien Tremblay
View author publications
Search author on:PubMed Google Scholar
Vasumathi Kameswaran
View author publications
Search author on:PubMed Google Scholar
Christy Ong
View author publications
Search author on:PubMed Google Scholar
Chad Liu
View author publications
Search author on:PubMed Google Scholar
Alissa D. Guarnaccia
View author publications
Search author on:PubMed Google Scholar
James A. Mondo
View author publications
Search author on:PubMed Google Scholar
Peter L. Hsu
View author publications
Search author on:PubMed Google Scholar
Noelyn M. Kljavin
View author publications
Search author on:PubMed Google Scholar
Bartosz Czech
View author publications
Search author on:PubMed Google Scholar
Janina Smola
View author publications
Search author on:PubMed Google Scholar
Dieu An H. Nguyen
View author publications
Search author on:PubMed Google Scholar
Jennifer A. Lacap
View author publications
Search author on:PubMed Google Scholar
Trang H. Pham
View author publications
Search author on:PubMed Google Scholar
Yuxin Liang
View author publications
Search author on:PubMed Google Scholar
Robert A. Blake
View author publications
Search author on:PubMed Google Scholar
Luca Gerosa
View author publications
Search author on:PubMed Google Scholar
Matthew Grimmer
View author publications
Search author on:PubMed Google Scholar
Shiqi Xie
View author publications
Search author on:PubMed Google Scholar
Bence Daniel
View author publications
Search author on:PubMed Google Scholar
Xiaosai Yao
View author publications
Search author on:PubMed Google Scholar
Anwesha Dey
View author publications
Search author on:PubMed Google Scholar

Contributions

S.P. conceptualized the study with supervision from X.Y. and A.D. S.P., T.J.H., V.K., C.O., C.L., N.M.K., D.A.H.N., J.A.L., T.H.P. performed experiments. J.T., X.Y., S.P., B.D., C.L., J.A.M., B.C., J.S., L.G., M.G. analyzed data. P.L.H. performed structural visualization. Y.L. sequenced libraries. S.P., J.T., B.D., A.D.G., R.A.B., S.X., X.Y., A.D. wrote and edited the manuscript with input from all authors.

Corresponding authors

Correspondence to Xiaosai Yao or Anwesha Dey.

Ethics declarations

Competing interests

All authors are or were employed by Genentech Inc., South San Francisco, CA, USA, at the time of their contributions to this work. All authors are or were shareholders at Roche except S.P., A.D.G., C. L., D.A.H.N.

Peer review

Peer review information

Nature Communications thanks the anonymous reviewer(s) for their contribution to the peer review of this work. A peer review file is available.

Additional information

Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Supplementary information

Supplementary Information

Description of Additional Supplementary Files

Supplementary Data 1

Supplementary Data 2

Supplementary Data 3

Supplementary Data 4

Supplementary Data 5

Supplementary Data 6

Reporting Summary

Transparent Peer Review file

Source data

Source Data

Rights and permissions

Open Access This article is licensed under a Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License, which permits any non-commercial use, sharing, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if you modified the licensed material. You do not have permission under this licence to share adapted material derived from this article or parts of it. The images or other third party material in this article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by-nc-nd/4.0/.

Reprints and permissions

About this article

Cite this article

Paul, S., Hagenbeek, T.J., Tremblay, J. et al. Cooperation between the Hippo and MAPK pathway activation drives acquired resistance to TEAD inhibition. Nat Commun 16, 1743 (2025). https://doi.org/10.1038/s41467-025-56634-y

Download citation

Received: 20 October 2023
Accepted: 26 January 2025
Published: 18 February 2025
DOI: https://doi.org/10.1038/s41467-025-56634-y

This article is cited by

Targeting the Hippo pathway in cancer
- Kieran F. Harvey
- Tracy T. Tang
Nature Reviews Drug Discovery (2025)