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

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.

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

A Chemical structure of GNE-7883. B GNE-7883 viability dose response curve (mean ± SD) for H226 cells. Statistical significance was assessed using two-way ANOVA followed by posthoc Tukey’s multiple comparison test. (n = 12 independent treatments per condition). C Experimental schematic for acute exposure of 1 µM GNE-7883. Created with BioRender.com/v03h840. D Heatmap showing downregulation of proliferation gene signature in GNE-7883 treated cells vs DMSO control from bulk RNA-seq expression data. Color density indicates z-normalized RPKMs (reads per kilobase of transcript per million reads mapped). E Gene Set Enrichment Analysis (GSEA) of differentially expressed genes from H226 treated with DMSO and H226 with short term GNE-7883 treatment. Hallmark MSigDB pathways are shown with normalized enrichment scores (NES) of indicated pathways on the x-axis and q-values (FDR < 0.05) indicated by color scale. F Representative EdU incorporation assay images of H226 cells treated with DMSO or 1 μM GNE-7883 for 48 h (Scale bar: 20 μm); n = 3 biological replicates. G (Top, center) Density scatter plots of H226 cells treated with DMSO or GNE-7883 for 48 h. EdU-negative cells represent G1 phase, EdU-positive cells indicate S phase (n = 1000 cells per condition); The color scale represents cell density. (Bottom) Quantification of (G). Statistical significance was assessed using one-way ANOVA followed by posthoc Tukey’s test. Data are presented as mean values ± SD; n = 3 biological replicates. H Heatmap representation of downregulated ATAC-seq peaks in H226 cells following 48 h treatment with DMSO, GNE-7883 or GNE-6915. I Venn diagrams showing overlap between differentially downregulated (left) or upregulated (right) ATAC-seq peaks of GNE-7883 and GNE-6915 compared to DMSO control (FDR < 0.05 and fold change > 1.5). J Top known motifs enriched in downregulated peaks after GNE-7883 or GNE-6915 treatment analyzed with HOMER. Shown are HOMER binomial P values. Source data are provided as a Source Data file.

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

A Schematic for generating drug adapted NF2 null mesothelioma cell line H226, through dose escalation of GNE-7883 (0.25 to 2.5 µM) over time. Created with BioRender.com/m45u525. B GNE-7883 viability dose response curve (mean ± SD) for H226-7883R. Statistical significance assessed using two-way ANOVA followed by posthoc Tukey’s multiple comparison test. (n = 12 independent treatments per condition). C Box plots showing gene expression changes measured by bulk RNA-seq for (top) Hippo, (center) apoptosis (bottom) MAPK signatures in acute vs long term 1 μM GNE-7883 exposure, represented as average log₂fold change. Data presented as medians ±25% with lower/upper whiskers extending to values within ±1.5 × the interquartile range. H226 + DMSO (red), H226 + GNE-7883 (gray), H226-7883R + DMSO (blue), H226-7883R + GNE-7883 (green). Conditions are compared to H226 treated with DMSO. (n = 3 biological replicates, two-sided t-test). D Whole cell lysates of H226 and H226-7883R cells treated with 1 μM GNE-7883 or DMSO, analyzed by immunoblotting (n = 3 biological replicates), Quantification in Fig. S3A. E Heatmap showing upregulated MAPK gene signature in acquired resistance with GNE-7883, from bulk RNA-seq. All conditions are compared to H226 treated with DMSO. Color density indicates z-normalized RPKMs. F Whole cell lysates of H226 and H226-7883R cells treated with 1 μM GNE-7883 or DMSO, analyzed by immunoblotting (n = 3 biological replicates). Quantification in Fig, S3C. G GSEA of differentially expressed genes from indicated comparisons, measured by bulk RNA-seq. Hallmark MSigDB pathways shown with normalized enrichment scores (NES) of pathways on x-axis and q-values (FDR < 0.05) indicated by color scale. H (Left) Representative EdU incorporation assay images of H226-7883R cells treated with 1 μM GNE-7883 (Scale bar: 20 μm). (Center) Density scatter plots for cell cycle distribution (G1 phase: EdU-negative; S phase: EdU-positive) with GNE-7883 treatment (n = 1000 cells per condition). Color scale represents the cell density. (Right) Quantification of (H). Statistical significance assessed using one-way ANOVA followed by posthoc Tukey’s test (Mean ± SD); n = 3 biological replicates. Source data are provided as a Source Data file.

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

A H226 and H226-7883R cells were treated with DMSO or GNE-7883 for 48 h, hash-tagged with HTO and pooled for scMultiome assay (ATAC-seq + RNA-seq). The cells are color-coded by either hashtag oligonucleotide assignments (left panel) or clusters (right panel) in the UMAP representation. n = 1242 for H226 + DMSO, 1475 for H226 + GNE-7883, 752 for H226-7883R + DMSO and 1483 for H226-7883R + GNE-7883. B UMAP representating the gene expression, chromatin accessibility (computed by chromVAR) or activity (computed by Epiregulon) of TEAD1 and FOSL1, for the indicated conditions. n = 1071 for H226_DMSO, 1299 for H226_GNE-7883 and 1898 for H226-7883 R_main. Corresponding violin plots are shown in (Fig. S6A). C Violin plots showing Hippo signature scores (canonical TEAD target genes), in single cells groups from each condition. n = 1071 for H226_DMSO, 1299 for H226_GNE-7883 and 1898 for H226-7883R_main. Statistical analysis was performed using two-sided Wilcoxon rank-sum test. The data presented as medians ±25%, lower and upper hinges correspond to the 25th and 75th percentiles respectively. The lower and upper whiskers extend to the smallest and largest values within ±1.5 × the interquartile range. D Motif enrichment (CisBP database) in upregulated ATAC-seq peaks in H226-7883R_main compared to H226_GNE-7883 clusters. Enrichment shown as −log10 (Padj value) based on hypergeometric test. E Top transcription factors with differential activity across clusters computed by Epiregulon. Pairwise comparison of each cluster is performed against all other clusters. For each cluster, the combined p-value of each gene is calculated by taking the middle value of the Holm-corrected p-values across all the pairwise comparisons. Genes are ranked by the combined p-values, and activity is defined by the sum of the putative target genes normalized by the number of target genes. Circle sizes correspond to negative log false-discovery-rates and colors indicate the relative transcription factor (TF) activity assessed with Epiregulon. F GSEA of the putative target genes of FOSL1 activity as inferred by Epiregulon. Circle sizes correspond to odds ratio and colors indicates the gene ratio. G Whole cell lysates of H226 and H226-7883R cells treated with 1 μM GNE-7883 or DMSO, analyzed by immunoblotting (n = 3 biological replicates), Quantification in (Fig. S6B). Source data are provided as a Source Data file.

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

A (Top) Immunofluorescence of chromatin-bound TEAD1 and YAP in H226 and H226-7883R cells treated with DMSO or 1 µM GNE-7883 for 48 h. Representative images for Hoechst and TEAD1 and YAP staining. (Scale bar: 20 μm). (Bottom) Boxplot of YAP and TEAD1 protein level distribution. n = 20946 for H226 treated with DMSO, 5776 for H226 treated with 1 µM GNE-7883, 835 for H226-7883R treated with 1 µM GNE-7883. The data are presented as medians ±25%, lower and upper hinges correspond to the 25th and 75th percentiles respectively. The lower and upper whiskers extend to the smallest and largest values within ±1.5 × the interquartile range from the hinges. Statistical analysis was performed using two-sided Wilcoxon rank-sum test on 500 randomly selected cells using datasample function in MATLAB. The experiment was repeated three times with consistent results. B–D Heatmap representation of CUT&RUN peaks indicating chromatin occupancy of (B: TEAD1; C: YAP; D: FOSL1), in H226 and H226-7883R treated with DMSO or 1 µM GNE-7883 for 48 h. E Venn diagram showing overlap of lost TEAD1, YAP and FOSL1 binding sites in parental cells treated with 1 µM GNE-7883 compared to DMSO. F Venn diagram showing overlap of gained TEAD1, YAP and FOSL1 binding sites in resistant cells treated with 1 µM GNE-7883 compared to parental cells treated with 1 µM GNE-7883. Source data are provided as a Source Data file.

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

A Volcano plot representing differential FOSL1 binding sites (FDR < 0.01 using DiffBind) in the indicated comparisons, with blue representing upregulated peaks and red representing downregulated peaks after GNE-7883 treatment. Venn diagrams indicate the overlap in the altered FOSL1 ChIP-seq peaks between parental H226 and H226-7883R cells following GNE-7883 treatment. Unique and shared sites are classified into different groups (bottom). B Read distribution plots show normalized read intensities for FOSL1 and YAP ChIP-seq data in the different groups in the indicated conditions.

Fig. 6. FOSL1 is required for TEAD/YAP chromatin binding in TEAD inhibitor resistant cells.

A Heatmap representation of CUT&RUN signals indicating chromatin occupancy of FOSL1, YAP and TEAD1 in H226-7883R cells with FOSL1 or CD81 (control) CRISPR knockdown. B Venn diagram showing overlap of lost TEAD1, YAP and FOSL1 binding sites in FOSL1 deficient resistant cells compared to CD81 (control) deficient resistant cells. C GSEA of target genes associated with downregulated peaks in FOSL1 deficient resistant cells compared to CD81 (control) deficient H226-7883R cells, bound by FOSL1, FOSL1/TEAD1, FOSL1/TEAD1/YAP, or FOSL1/YAP using chipEnrich. Color represents the number of genes that had signals contributing to the test for enrichment.

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

A Viability of H226-7883R cells measured after 4 days in presence of 1 μM GNE-7883, in the background of FOSL1 knockdown with three transfected individual guide RNAs, compared to the non-targeted control (NTC). Statistical significance was assessed using one-way ANOVA followed by posthoc Tukey’s test. Data are presented as mean values ± SD; n = 6 biological replicates. B H226-7883R cells were transfected with three single guide RNAs (sgRNA) targeting FOSL1 and control guide RNA targeting CD81, maintained in presence of GNE-7883 for 4 days, then fixed and stained with crystal violet solution to visualize cell colonies; n = 2 biological replicates. C Whole cell lysates of H226 and H226-7883R cells treated with DMSO, 1 μM GNE-7883 or Cobimetinib, analyzed by immunoblotting. (n = 3 biological replicates), Quantification in Fig. S11C. D Immunofluorescence of chromatin-bound FOSL1 in H226 and H226-7883R cells treated with DMSO, 1 µM GNE-7883 or 1 µM Cobimetinib for 48 h. Representative images for Hoechst and FOSL1 staining (Scale bar: 20 μm). The experiment was repeated three times with consistent results. E Boxplot of FOSL1 protein level distribution in the same cells as (D). n = 33491 for H226 treated with DMSO, 24102 for H226 treated with GNE-7883, 25875 for H226 treated with Cobimetinib, 37534 for H226-7883R treated with GNE-7883, 29316 for H226-7883R treated with Cobimetinib. The data are presented as medians ±25%, lower and upper hinges correspond to the 25th and 75th percentiles respectively. The lower and upper whiskers extend to the smallest and largest values within ±1.5 × the interquartile range from the hinges. Statistical analysis was performed using two-sided Wilcoxon rank-sum test on 500 randomly selected cells using datasample function in MATLAB. F Heatmaps showing the relative viability across a dose–response matrix of GNE-7883 and Cobimetinib combination (left) and GNE-7883 and Belvarafenib combination (right) in H226 and H226-7883R cells in 7-day viability assays. 1 indicates maximum viability and 0 indicates no viability. Measured IC50 for the indicated drug concentration is shown by the continuous line. Source data are provided as a Source Data file.

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

A Schematic representation of experimental setup for xenograft analyses. 1 × 10⁷ H226-7883R or MSTO-7883R cells were xenografted into C.B-17 SCID mice subcutaneously in the right flank. Tumors were allowed to grow to an initial volume before randomizing to treatment groups: 1) vehicle MCT (0.6% methocel A15 LV, 0.2% tween 80) + vehicle (sunflower oil), 2) GNE-7883, 250 mg/kg 3) Cobimetinib, 7.5 mg/kg, 4) GNE-7883, 250 mg/kg + Cobimetinib, 7.5 mg/kg. Fitted tumor volumes of (B) H226-7883R or (C) MSTO-7883R xenograft model treated with Vehicle, GNE-7883, Cobimetinib, GNE-7883 + Cobimetinib (H226-7883R: n = 7 and MSTO-7883R: n = 9 mice per group). D Schematics showing acute treatment of GNE-7883 reduced FOSL1 and YAP/TEAD activity with a slight activation of MAPK signaling. Over time, cells continue to acquire increased MAPK pathway activity in response to long term exposure to GNE-7883. Resistant cells increase FOSL1 levels and restore YAP/TEAD activity. E Acute treatment with GNE-7883 leads to reduced chromatin accessibility at TEAD binding sites and decreased occupancy of YAP, TEAD, and FOSL1. Cells that have acquired resistance to GNE-7883 restore chromatin accessibility and the occupancy of YAP and TEAD. Resistant cells also show enhanced FOSL1 occupancy at YAP/TEAD-bound regions. FOSL1 is required for TEAD/YAP binding. MAPK inhibitors reduce chromatin-bound FOSL1 levels and overcome TEAD inhibitor resistance. Created with BioRender. BioRender.com/k12u316. Source data are provided as a Source Data file.