Gracillin suppresses cancer progression through inducing Merlin/LATS protein-protein interaction and activating Hippo signaling pathway

Abstract

Gene therapy, epigenetic therapies, natural compounds targeted therapy, photodynamic therapy, nanoparticles, and precision medicines are becoming available to diagnose and treat cancer. Gracillin, a natural steroidal saponin extracted from herbs, has shown potent efficacy against a range of malignancies. In this study, we investigated the molecular anticancer mechanisms of gracillin. We showed that gracillin dose-dependently suppressed proliferation, migration, and invasion in breast cancer, liver cancer, and glioblastoma cells with IC₅₀ values around 1 μM, which were associated with MST-independent activation of Hippo signaling pathway and subsequent decreased YAP activity. We demonstrated that gracillin activated the Hippo signaling by inducing Merlin/LATS protein-protein interaction (PPI). A competitive inhibitory peptide (SP) derived from the binding interface of the PPI, disrupted the interaction, abolishing the anticancer activity of gracillin. In nude mice bearing MDA-MB-231, HCCLM3, or U87MG xenograft tumor, administration of gracillin (5, 10 mg·kg⁻¹·d⁻¹, i.g. for 21 days) dose-dependently suppressed the tumor growth, associated with the induced Merlin/LATS PPI, activated Hippo signaling, as well as decreased YAP activity in tumor tissues. Our data demonstrate that gracillin is an efficacious therapeutic agent for cancer treatment, induction of Merlin/LATS PPI might provide proof-of-concept in developing therapeutic agent for cancer treatment.

Introduction

Cancer is one of the most prevalent diseases worldwide [1]. Traditional therapies including surgery, radiotherapy, and chemotherapy generate severe side effects [2]. In response, advanced technologies like gene therapy, epigenetic therapies, natural compounds targeted therapy, photodynamic therapy, nanoparticles, and precision medicines are becoming available to diagnose and treat cancer [3–8]. Despite these advances, challenges such as drug resistance, recurrence, and metastasis continue to hinder effective cancer treatment [2, 9]. Consequently, there is a pressing need for further exploration into novel anticancer agents with enhanced efficacy, high selectivity, and reduced side effects [10–12].

Accumulating evidence has established that dysregulated Hippo signaling pathway is closely involved in cancer progression [13], resistance [14, 15], recurrence [16], and metastasis [17]. The Hippo signaling pathway is an evolutionarily conserved tumor suppressor pathway that functions through the activation of kinase cascade MST1/2, LATS1/2, YAP and its paralogue TAZ [13]. When the Hippo pathway is deactivated, aberrant YAP translocate into the nucleus, resulting in the formation of a complex with the transcriptional activator TEAD, enhancing the transcription and translation of target genes such as CYR61, CTGF, c-Myc, and Survivin [18, 19], ultimately driving the occurrence and progression of malignancies. Hence, YAP/TAZ are considered key effectors of Hippo signaling pathway and their activity served as a readout of the regulation for Hippo signaling pathway [20].

Merlin, a crucial tumor suppressor encoded by the neurofibromatosis type 2 (NF2) gene, plays a significant role in regulating the Hippo signaling pathway and is also vital for organ size control, inhibition of mitogenic signaling, repression of cell proliferation, and promotion of apoptosis and ferroptosis [17, 21]. Inactivation of Merlin can lead to the development of various cancers, including glioblastoma [22], liver cancer [23], and breast cancer [24, 25]. Merlin functions at the plasma membrane by binding to and facilitating the recruitment of LATS1/2, activating Hippo signaling independent of MST kinases [26]. The interaction between Merlin and LATS is mediated by the FERM domain of Merlin and the FERM-binding domains of LATS1/2, which promotes LATS phosphorylation and subsequent Hippo pathway activation [27].

Gracillin (C₄₅H₇₂O₁₇, Fig. 1), also known as diosgenyl rhamnopyranosyl-1-2-(glucopyranosyl-1-3) glucopyranoside, is a natural steroidal saponin first isolated from the root bark of the wild yam Dioscorea nipponica [28]. It has demonstrated a range of biological activities, including antitumor, antimicrobial, antioxidant, and anti‐inflammatory properties [28, 29]. Recent studies have highlighted its potent antitumor activity against various malignancies [30–33]. However, the precise mechanisms underlying its antitumor activity have not been well defined and its regulation of the Hippo signaling pathway remains poorly characterized.

Fig. 1.

Chemical structure formula of gracillin.

In the present study, we demonstrated that gracillin significantly suppressed proliferation, migration, and invasion in breast cancer, liver cancer, and glioblastoma cells, which involved activation of Hippo signaling pathway independent activation of MST kinase. Mechanistically, gracillin enhances the protein-protein interaction between Merlin and LATS, resulting in elevated phosphorylation levels of LATS and YAP, as well as in downregulated YAP activity, effectively suppressing tumorigenic features. Thus, our study offers compelling evidence that targeted inducing Merlin/LATS PPI activates Hippo signaling pathway for cancer treatment and positions gracillin as an attractive drug candidate for cancer therapies.

Materials and methods

Cell culture

All cell lines (HEK293T, MDA-MB-231, MCF-7, LO2, Huh-7, HepG2, HCCLM3, U87MG, U118MG, LN229, U251, SH-SY5Y, HCT116, SW620, A549, SHP-77, and NCI-H526) were obtained from the American Type Culture Collection (ATCC, Manassas, VA, USA) and maintained in DMEM (Gibco, Thermo Fisher Scientific) supplemented with 10% fetal bovine serum (FBS), L-glutamine (Gibco, Thermo Fisher Scientific), and 100 U/mL penicillin-streptomycin (Gibco, Thermo Fisher Scientific). All cell lines were maintained at 37 °C in a 5% CO₂ incubator and have been verified to be negative for mycoplasma contamination.

Animal study

Female BALB/c nude mice (5–6 weeks, 18–20 g) were purchased from Beijing Vital River Laboratory Animal Technology Co., Ltd., and were maintained according to Jinan University’s Policy on the Care and Use of Laboratory Animals. All animal experiments were conducted strictly following the institutional ethics and safety guidelines approved by Jinan University Institutional Animal Care and Use Committee. For xenograft tumor model, MDA-MB-231, HCCLM3, and U87MG cells were harvested and then transplanted subcutaneously (5 × 10⁶ cells in 100 μL of PBS) into the right axilla of the mice. When the volume of tumor reached 100 mm³, the mice were randomly divided into three groups: gracillin (5 mg/kg or 10 mg/kg) and 0.9% NaCl containing 0.5% DMSO (control), which was administered by oral gavage every other day. The tumor volume (V) was measured every two days and calculated using the following formula: V = length × width² × 0.5. After 21 days from the first treatment, mice were euthanized and sacrificed, and their hearts, livers, spleens, kidneys, and lungs were immediately fixed and paraffin‐embedded for H&E staining. The tumors were photographed, weighed, measured, and preserved at −80 °C until used for other experiments.

Chemical and reagents

Gracillin was purchased from Selleck (S3915) and prepared as the stock solution with dimethyl sulfoxide (DMSO) (Sigma Aldrich, St. Louis. MO, USA), and freshly diluted before experiments. Antibodies against MST1 (#3682), LATS1 (#3477), YAP (#12395), YAP/TAZ (#8418), phospho-MST1 (Thr183)/MST2 (Thr180) (#49332), phospho-LATS1 (Ser909) (#9157), phospho-YAP (Ser127) (#13008), phospho-TAZ (Ser89) (#59971), CTGF (#86641), CYR61 (#14479), Survivin (#2808), c-Myc (#18583), Merlin (#12888) were purchased from Cell Signaling Technology. Antibody against LATS2 (20276-1-AP) and GST Tag Recombinant antibody (80006-1-RR) were purchased from Proteintech. Anti-DDDDK-tag mAb (M185-3L) was purchased from MBL. β-Actin (T0022), GAPDH (T0004), Histone H2B (BF0722), Ki67 (#AF0198), and Cleaved Capsase-3 (Asp175) (AF7022) were purchased from Affinity, respectively. The small peptide (SP) targeting Merlin/LATS2 interaction (HLYVSPWGGLREIRYSLLPF) was designed and synthesized by Apeptide company (Shanghai, China) with a purity of >95%. The small peptide (SP) was solved in ultrapure water with 5% PEG to a stock solution of 1 mM, and diluted with culture medium to a final concentration of 30 μM.

Cell viability assay

Cells were seeded into 96-well plates (3000 cells/well) and cultured overnight. After replacing the medium with fresh medium containing different concentrations of gracillin, cells were incubated at 37 °C and 5% CO₂ for 48 h. Subsequently, 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide (MTT, 0.5 mg/mL) was then added, and after 4 h, the formed crystals were dissolved in 150 μL DMSO. Absorbance was measured at 565 nm using a microplate reader (Varioskan LUX, Thermo Fisher, China). The IC₅₀ value was calculated using the GraphPad Prism 8.

Colony formation assay

Cells were seeded into 6-well plates (1 × 10³ cells/well) in triplicate and treated with indicated concentrations of gracillin or control. The culture medium with the same concentrations of gracillin was refreshed every three days. After 12–15 days, colonies were fixed with methanol, stained with 0.1% crystal violet (Beyotime, C0121), and photographed. Colony numbers were counted.

Wound healing assay

Cells were seeded into 6-well plates to 90% confluence and incubated overnight in serum-free medium. The next day, a scratch was made using a sterile 100 μL micropipette tip, and the cells were rinsed with PBS to remove cell debris. The cells were then cultured in the condition of 37 °C and 5% CO₂. Images were captured using a Leica DM 14000B microscope system at 0- and 24-h post-scratching. The wound healing rate was quantified using ImageJ software (v.1.8.0, NIH).

Transwell migration and invasion assay

Cells were treated with gracillin for 24 h and resuspended in serum-free medium. For migration and invasion assays, 2 × 10⁵ cells were added to the upper chamber of Transwell inserts, with or without pre-coated Matrigel (BD, #356234) for invasion assays. The lower chamber contained 600 μL of 20% FBS medium as a chemoattractant. After 48 h, cells were fixed, stained with crystal violet (Beyotime, C0121), and counted under an inverted microscope (Olympus, 100× magnification).

Western blot assay

Cells or tissues were lysed using RIPA lysis buffer (Thermo Fisher Scientific) supplemented with PMSF and phosphatase inhibitor. Protein concentrations were determined using a BCA protein assay kit (Thermo Fisher Scientific), followed by separation via SDS-PAGE and transferred to PVDF membranes. Membranes were incubated with primary antibodies overnight at 4 °C, followed by secondary antibodies for 1 h at room temperature. Visualization of the target proteins was achieved using Electrochemiluminescence (ECL) substrate reagents. Quantification of the blots was performed using ImageJ software (v.1.8.0, NIH).

Quantitative real‐time polymerase chain reaction (qPCR)

Cellular RNA was extracted using an NCM Biotech (Suzhou, China), followed by reverse-transcription into cDNA using a reverse transcription kit (Servicebio, Wuhan, China). For quantitative PCR, Universal Blue SYBR Green qPCR Master Mix (Servicebio, Wuhan, China) was employed. The primers for qPCR were synthesized by the BGI Company (BGI, Beijing, China). The primers sequences are as follows (5’-3’): CTGF-F: CCTGCAGGCTAGAGAAGCA; CTGF-R: TGGAGATTTTGGGAGTACGG; CYR61-F: ACCAATGACAACCCTGAGTG; CYR61-R: AAACATCCAGCGTAAGTAAACC; YAP-F: TCCACCAGTGCAGCAGAATA; YAP-R: TTGGGTCTAGCCAAGAGGTG; GAPDH-F: CAATGACCCCTTCATTGACC; GAPDH-R: GACAAGCTTCCCGTTCTCAG. Gene expression levels were conducted with the 2^−ΔΔCT method following normalization to reference genes.

Immunofluorescence assay

Cells were seeded onto glass coverslips to reach 70% confluence, treated by gracillin, fixed with 4% paraformaldehyde, permeabilized with 0.3% Triton X-100, and blocked with 5% goat serum. Cells were incubated at 4 °C overnight with anti-YAP (1:200) and subsequently with Alexa Fluor 488-conjugated goat anti-rabbit IgG (1:400, Invitrogen, A11012) for 1 h at room temperature. Coverslips were mounted with DAPI antifade mounting medium (Vector Laboratories, H-1200), and images were captured using a fluorescence microscope (Leica).

Dual-luciferase reporter assay

Cells (60% confluence) in 24-well plates were co-transfected with 8× GTIIC-Luc and Renilla luciferase plasmid using Lipofectamine 3000 (Thermo Fisher, Cat. # L3000001). After 4 h, gracillin was added at various concentrations, and cells were incubated overnight. Subsequently, the cells were harvested and analyzed according to the manufacturer’s protocol of Dual-Luciferase^® Reporter Assay System (Promega). Measures are normalized and represented as the ratio of Firefly/Renilla intensity.

Immunohistochemistry

Tumor tissues were fixed in 4% paraformaldehyde and subsequently embedded in paraffin. The embedded tissues were cut into 8 μm sections, followed by dewaxing using xylene and rehydrated with graded alcohols (70%, 80%, 90%). Endogenous peroxidase activity was blocked and antigen retrieval was performed by boiling the slides. The sections were then incubated with primary antibodies overnight at 4 °C. On the following day, the slices were stained with DAB and hematoxylin and photographed under a microscope. The percentage of positively stained cells was analyzed using ImageJ software.

Co‑immunoprecipitation (Co-IP)

Cells were harvested and lysed by IP lysis buffer (Thermo Fisher Scientific). 50 μL of lysate supernatant was reserved as input, and the rest was added with protein A/G beads (Thermo Fisher Scientific) and anti-Merlin antibody. The mixture was incubated at 4 °C on a rotating device overnight. Corresponding IgG was used as the negative control. The following day, the beads were washed with IP lysis buffer and the binding reaction was terminated by adding 2× loading buffer and boiling at 100 °C for 10 min. Subsequently, the samples were then analyzed by Western blot assay.

Glutathione S-transferase pull-down assay

The plasmids of GST-NF2 and Venus-Flag-LATS2 (VF-LATS2) were co-transfected into cells. Cells were then treated with gracillin for 24 h and lysed. The mixture of GST beads and cell lysates was then incubated using a rotation machine at 4 °C overnight. The next day, the beads pellets were collected, washed with ice-cold lysis buffer, and resuspended in 2× SDS gel loading buffer and boiling at 100 °C for 10 min. The samples were then examined by Western blot assay.

Molecular docking

Molecular docking was conducted using AutoDock Vina 1.0.2. The crystal structures of the Merlin-FERM and LATS1 complex (PDB: 4ZRK), and Merlin-FERM and LATS2 complex (PDB: 4ZRI), and Merlin FERM domain (PDB: 3U8Z) were retrieved from the RCSB protein data bank (https://www.rcsb.org/). The structure of gracillin was retrieved from the PubChem database (https://pubchem.ncbi.nlm.nih.gov/). File transformations were performed using Open Babel software. Visualization of the most suitable complexes was performed using Discovery Studio 2021 Client and Pymol 2.2.0.

RNA sequencing

RNA was extracted from cells using TRIzol reagent (Thermo Fisher Scientific). Library construction and RNA-sequencing procedures were completed by Shanghai Personal Biotechnology Co. Ltd (Shanghai, China).

Cellular thermal shift assay (CETSA)

Cells were resuspended in PBS with protease and phosphatase inhibitors, divided into fractions, heated at a gradient for 3 min, and subjected to two freeze-thaw cycles using liquid nitrogen. After centrifugation at 12,000 rpm for 20 min, the supernatant was collected, boiled, and analyzed by Western blotting.

Statistical analysis

Data were presented as mean ± standard deviation (SD), except for where it was represented as the standard error of the mean (SEM) specifically. All data analyses were conducted using GraphPad Prism 8 (GraphPad Software, Inc., SanDiego, CA, USA). Statistical analysis was performed using Student’s t test or One-way ANOVA test. All comparisons were made relative to untreated control groups unless otherwise specified. Significance of differences was indicated as *P < 0.05, **P < 0.01, and ***P < 0.01. Quantifications were performed from at least three independent experiments.

Results

Gracillin suppresses proliferation, migration, and invasion in cancer cells

To assess the antitumor effects of gracillin, we used cancer cells with different YAP activity statuses, classified as YAP^on or YAP^off [34]. Data showed that breast cancer, liver cancer, and glioblastoma cells, which belongs to the YAP^on class, were more sensitive to gracillin (Supplementary Fig. S1a). Hence, we selected these cells to explore anti-cancer effect and underlying mechanism of gracillin. Gracillin inhibited cancer cell proliferation in a dose-dependent manner, with the IC₅₀ values ranging from 0.5 to 2 μM across different types of cancer cells (Fig. 2a). The colony formation assay confirmed the potent anticancer activity of gracillin against tumorigenesis (Fig. 2b). Since cell migration and invasion are crucial for cancer progression, we conducted wound healing and Transwell assays to evaluate the impact of gracillin on these processes. Data indicated that gracillin suppressed the migration and invasion of MDA-MB-231, HepG2, and U87MG cells in a dose-dependent manner (Fig. 2c, d, Supplementary Fig. S1b, c). These results indicates that gracillin exhibits potent suppressive effects on cell proliferation, migration, and invasion, suggesting that gracillin is a promising anti-tumor agent against breast cancer, liver cancer, and glioblastoma.

Fig. 2. Gracillin suppresses proliferation, migration, and invasion in cancer cells.

a Effect of gracillin (GRA) on cancer cell proliferation, assessed by MTT assay after 48 h of treatment. IC₅₀ values indicated in the figure. b Effect of GRA on colony formation in MDA-MB-231, HepG2, and U87MG cells. Colonies were visualized by crystal violet staining. c Effect of GRA on migration and invasion in MDA-MB-231, HepG2, and U87MG cells. Following treatment with gracillin (0, 1, 2, or 4 μM) for 48 h, the cells were subjected to the Transwell migration and invasion assays (magnification, ×100; scale bars, 200 μm, n = 3). d Quantification of the migrated and invasive cells. Data are presented as mean ± SD. *P < 0.05, **P < 0.01, ***P < 0.001.

Gracillin activates Hippo signaling pathway in cancer cells independent of activation of MST1/2 kinase

To elucidate the anti-cancer mechanism of gracillin, we performed RNA-seq analysis. Pearson’s correlation analysis revealed a strong correlation between samples from the treated and untreated groups (R = 0.9956–1.0000) (Supplementary Fig. S2a), and principal components analysis (PCA) showed a clear difference between gracillin-treated and control groups, with high similarity between samples within each group (Supplementary Fig. S2b). Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway enrichment analysis revealed that the Hippo signaling pathway was one of the top pathways identified (Supplementary Fig. S2c). Given our previous findings and its critical role in regulating cell proliferation, migration, and invasion, the Hippo pathway’s involvement suggests that it may play a significant role in gracillin’s anti-cancer activity.

Subsequently, we assessed the levels of major kinases and the phosphorylation status of key components of the Hippo pathway, including MST, LATS, YAP, and TAZ. Compared to vehicle-treated cells, gracillin treatment resulted in increased levels of p-LATS, p-YAP1, and p-TAZ in a time-dependent manner (Fig. 3a, b). Interestingly, p-MST1/2 levels were not affected by gracillin, despite MST1/2 is well characterized as an upstream activating kinase of LATS1/2 [35]. Further dose-response analysis showed that, except for p-MST1/2, gracillin increased the levels of p-LATS1 and p-YAP1 in a dose-dependent manner, while not affecting the mRNA levels of YAP (Fig. 3c, d, Supplementary Fig. S3a). In addition, expressions of CTGF, CYR61, and Survivin, the target genes of the Hippo signaling, were downregulated in a dose-dependent manner (Fig. 3c, e), confirming that gracillin activates the Hippo pathway. These data suggest that gracillin is a potent agonist of Hippo signaling pathway, which functions independent of MST1/2 kinase.

Fig. 3. Gracillin activates Hippo signaling pathway in cancer cells.

a GRA activates the Hippo signaling in a time-dependent manner in cancer cells. Cells were treated with GRA for 3, 6, 12, 24, and 36 h and subjected to Western blotting analysis. b Quantification of the protein levels indicated in (a). c GRA activates the Hippo signaling in a dose-dependent manner in cancer cells. Cells were treated with GRA at indicated concentrations and subjected to Western blotting analysis. d & e Quantification of proteins indicated in (c). Data are presented as mean ± SD. *P < 0.05, **P < 0.01, ***P < 0.001.

Gracillin exhibits anticancer effects dependent on YAP activity

As the main effector of the Hippo pathway, YAP exerts its activity through nuclear translocation and transcriptional co-activation [20]. Immunofluorescence staining assay indicated that gracillin treatment led to cytoplasmic retention of YAP in MDA-MB-231, HepG2, and U87MG cells (Fig. 4a). Further protein fractionation analysis revealed that the ratio of nuclear YAP levels was significantly decreased in cells treated with gracillin (Fig. 4b, c). To determine the effect of gracillin on YAP activity, we conducted a dual luciferase reporter assay in MDA-MB-231, HepG2, and U87MG cells. Data showed a dose-dependent inhibition of YAP transcriptional activity by gracillin (Fig. 4d), which was consistent with a decrease in the mRNA levels of YAP target genes (Fig. 4e and Supplementary Fig. S3b, c).

Fig. 4. Gracillin exhibits anticancer effects dependent on YAP activity.

a Following treatment with GRA, the cellular localization of YAP in MDA-MB-231, HepG2, and U87MG cells was detected by immunofluorescent staining (×400; scale bars, 50 μm). b Cytoplasmic and nuclear distributions of YAP in MDA-MB-231, HepG2, and U87MG cells treated with or without GRA for 12 h. Histone 2B and GAPDH were used as endogenous references for nuclear and cytosolic fractions, respectively. c Quantification of proteins indicated in (b). d Transcriptional activity of YAP/TEAD complex in MDA-MB-231, HepG2, and U87MG cells. Cells were transfected with the reporter plasmids and treated with GRA at the indicated concentrations for 24 h. e mRNA levels of CYR61 and CTGF in MDA-MB-231 cells treated with GRA. RT-qPCR data were normalized to GAPDH levels and presented as fold-change compared with control cells (n = 3). f Effects of GRA on the proliferation of MDA-MB-231 cells expressing wild-type (WT) and S127A mutant YAP. Cells were treated with 0.25, 0.5, 1, 2, and 4 μM gracillin for 48 h followed by transfection with YAP ^WT and YAP ^S127A plasmids and the cell proliferation was evaluated by MTT assay. Effects of GRA on the migration (g) and invasion (h) in MDA-MB-231 expressing WT and S127A mutant YAP. Following treatment with indicated concentrations of GRA, the cells were subjected to Transwell migration and invasion assays (magnification, ×100; scale bars, 200 μm, n = 3). Data are presented as mean ± SD. *P < 0.05, **P < 0.01, ***P < 0.001.

To further explore the dependence of gracillin’s anticancer effects on YAP activity, we utilized a constitutively active form of YAP (YAP^S127A), which remains in the nucleus and is unable to be phosphorylated. Compared to cells expressing wild-type YAP, YAP^S127A transfected cells partially counteracted the inhibitive effect of gracillin on cell proliferation (Fig. 4f, Supplementary Fig. S4a, b), migration (Fig. 4g, Supplementary Fig. S4c, d), and invasion (Fig. 4h, Supplementary Fig. S4e, f). Collectively, our results demonstrate that gracillin exerts anticancer activity through activation of Hippo signaling pathway, with a clear dependency on YAP activity.

Gracillin induces Merlin/LATS protein-protein interaction in cancer cells

Given that gracillin activated Hippo signaling independently of MST1/2, we next sought to identify the specific proteins involved in LATS phosphorylation. Merlin, a well-established tumor suppressor encoded by NF2 gene, is known to act as the upstream activator of LATS [36]. Docking results revealed that gracillin could establish a stable structure with Merlin/LATS complex, exhibiting a binding energy ΔG of −10.9 kcal/mol for Merlin/LATS1 (PDB code: 4ZRK) and −9.4 kcal/mol for Merlin/LATS2 (PDB code: 4ZRI) (Fig. 5a). The PPI was facilitated by interactions involving hydrogen bonds and van der Waals forces, with hydrogen bonds at ASP70, TYR101, GLN125, and ARG249 of Merlin/LATS1 complex, and GLU260, SER255, and HIS242 of Merlin/LATS2 complex (Fig. 5a). This evidence suggests that gracillin may induce LATS phosphorylation by promoting Merlin/LATS protein-protein interaction.

Fig. 5. Gracillin induces Merlin/LATS protein-protein interaction in cancer cells.

a Molecular docking of gracillin with Merlin/LATS1 complex (PDB: 4ZRK) and Merlin/LATS2 complex (PDB: 4ZRI). b GRA induces extracellular Merlin/LATS2 protein-protein interaction. Cell lysates from HEK293T cells co-transfected with GST-NF2 and VF-LATS2 were treated with indicated concentrations of GRA and then subjected to GST pull-down assay. c GRA induces intracellular Merlin/LATS2 protein-protein interaction. HEK293T cells were co-transfected with GST-NF2 and VF-LATS2 and treated with GRA, followed by analysis with GST pull-down assay. d Quantification of the protein levels of Flag-tagged LATS2 indicated in (b) and (c). e Gracillin increased the interaction of Merlin and LATS2 in MDA-MB-231, HepG2, and U87MG cells. Following treatment with GRA (4 μM), the cell lysates were subjected to Co-IP assays with indicated antibodies. Data are presented as mean ± SD. *P < 0.05, **P < 0.01.

For validation, we transfected GST-NF2 and VF-LATS2 into HEK293T cells and treated them with gracillin, given that LATS1 and LATS2 interact with Merlin FERM at the same LATS1/2-FBD domain [27]. With the cell lysate with plenty of GST-NF2 and VF-LATS2 proteins, GST pull-down assay showed that gracillin enhanced Merlin/LATS interaction in a dose-dependent manner (Fig. 5b, d). Furthermore, the induction of Merlin/LATS interaction was also observed in cells treated with gracillin (Fig. 5c, d). At the endogenous level, with Co-IP assays, we confirmed that gracillin upregulated Merlin/LATS interaction across all tested cancer cell lines (Fig. 5e). These data indicate that gracillin acts as an inducer of Merlin/LATS protein-protein interaction.

Merlin/LATS PPI is required for the Hippo signaling activation and anticancer activity of gracillin

The “blue box” region (BB) region of Merlin-FERM, encompassing residues 176-183, constitutes a significant portion of the LATS1/2-binding surface [27]. Deletion of the BB region weakens or disrupts Merlin’s binding capacity to LATS1/2, thereby preventing LATS1/2 from phosphorylating YAP [27]. Notably, the hydrophobic residues (Val139, Leu140, Tyr177, Met179, Trp184, Trp191, and Ile210) of Merlin-FERM are highly conserved. Moreover, substitution of either Ile81 or Leu85 of LATS1-FBD with Lys completely abolished the binding of LATS1 to Merlin-FERM [26].

Based on previous findings, we designed a small peptide resembling LATS binding fragment, which showed promising binding affinity to the interaction region (LREIRYSLLPF), thereby might competitively disrupt the interaction of Merlin and LATS. To facilitate cellular uptake of the peptides, by linking to a cell-penetrating peptide (HLYVSPWGG) via a di-glycine linker, we generated a fusion peptide SP (HLYVSPWGGLREIRYSLLPF) (Fig. 6a, b) [37, 38]. Cellular thermal shift assay (CETSA) confirmed that SP entered cells and bound to Merlin and cell viability assay indicated that SP was not toxic to normal cells (Supplementary Fig. S5a, b). Co-IP assay showed that SP disrupted the Merlin/LATS PPI in cancer cells (Fig. 6c). Further studies showed that SP could significantly decrease the phosphorylation of LATS and YAP and gracillin exhibited a negligible effect on activation of LATS and YAP in the presence of SP (Fig. 6d). Furthermore, with SP as a lose-of-function tool, data from functional assays demonstrated that the inhibitive effects of gracillin on cell proliferation, migration, and invasion were all abolished by SP pre-treatment (Fig. 6e–g, Supplementary Fig. S5c). These data indicate that the gracillin exerts anticancer activity through activation of Hippo signaling that dependent on Merlin/LATS protein-protein interaction.

Fig. 6. Gracillin suppresses cell proliferation, migration, and invasion via Merlin/LATS PPI-induced activation of Hippo signaling pathway.

a Designation of small peptide (SP) from LATS targeting Merlin/LATS2 interaction. b Molecular docking of SP and Merlin (PDB: 3U8Z). c Effect of GRA on Merlin/LATS interaction in MDA-MB-231 and HepG2 cells. SP (30 μM) was pre-added to the culture medium for 4 h before GRA. Following treatment of SP/GRA, the interaction of Merlin and LATS2 was analyzed by Co-IP assay. d Effect of SP/GRA on the activation of Hippo signaling in MDA-MB-231, HepG2, and U87MG cells, samples subjected to Western blotting analysis. e Effect of SP/GRA on cell proliferation in MDA-MB-231, HepG2, and U87MG cells. f Effect of SP/GRA on cell migration in MDA-MB-231, HepG2, and U87MG cells. g Effect of SP/GRA on cell invasion in MDA-MB-231, HepG2, and U87MG cells. Data are presented as mean ± SD. *P < 0.05, **P < 0.01.

Gracillin suppresses tumor growth via Merlin/LATS PPI-induced activation of Hippo signaling pathway

To investigate the potential effects of gracillin on tumor growth in vivo, we established subcutaneous xenograft models using MDA-MB-231 cells with high expression of YAP [39]. Seven days post-implantation, normal saline containing 0.5% DMSO or gracillin (5 or 10 mg/kg) was given by oral gavage every other day for 21 days. Subsequently, the toxicity and antitumor effects of gracillin in vivo were examined. Data of the body weights and histological analysis of key organs demonstrated the low toxicity of gracillin (Supplementary Fig. S6a, b). Gracillin treatment exhibited significant inhibitory effect on tumor growth, represented by decreased tumor volume and tumor weight compared with those in the control group (Fig. 7a–c, Supplementary Fig. S6c). Similar data have been collected in the HCCLM3 and U87MG xenograft models (Supplementary Fig. S6d–m). Consistent with the in vitro data, Western blot analysis showed that p-LATS1 and p-YAP levels were upregulated in tumor tissues from mice treated by gracillin, while the levels of p-MST remained unchanged (Fig. 7d, e). Co-IP assay further confirmed that gracillin promoted the Merlin/LATS interaction (Fig. 7f). Immunohistochemistry results indicated elevated levels of p-YAP, and Cleaved Caspase-3, along with decreased expression of YAP, CTGF, and Ki67 in the gracillin-treated group (Fig. 7g). Collectively, these data demonstrate that Merlin/LATS PPI-induced activation of Hippo signaling pathway is associated with the antitumor effect of gracillin in vivo.

Fig. 7. Gracillin activates Hippo signaling and suppresses tumor growth in vivo.

a Representative images of tumors from each experimental group. b Tumor volume in each group. Data are presented as mean ± SEM. c Quantification and analysis of the tumor weight, n = 6. d GRA activates Hippo signaling pathway independent of MST activation. Expression levels of Hippo pathway kinases MST, LATS, and YAP and their phosphorylation forms along with downstream protein CTGF were examined by Western blot. β-actin was used as a loading control. e Quantification of the expression of indicated protein levels in (d). f GRA induces the interaction of Merlin and LATS in vivo assayed by Co-IP assay. g The expression levels of p-YAP, YAP, CTGF, Ki67, and Cleaved Caspase-3 in tumors assayed by IHC. Scale bars, 100 μm. h Quantification of the positive signals in (g). Data are presented as mean ± SD. *P < 0.05, **P < 0.01.

Discussion

In this study, we revealed that gracillin, a natural steroidal saponin, exerts potent anticancer activity through activation of the Hippo signaling pathway that is independent of MST kinase. Mechanistically, the activation of Hippo signaling is through enhanced Merlin/LATS protein-protein interaction, which is confirmed by a loss-of-function validation with a home-developed PPI disruptor peptide that directly binds to Merlin, providing compelling evidence that targeted induction of PPIs can be a feasible strategy for cancer treatment.

The Hippo signaling pathway is one of the eight conventional human cancer signaling pathways identified by The Cancer Genome Atlas after analyzing over 9000 cancer cases [40]. Recent studies have found that the Hippo pathway can also regulates adaptive immunity, playing roles in the composition of innate immune cells in the tumor immune microenvironment, fibroblast activation, stem cell maintenance, extracellular matrix changes, and immune infiltration [41]. Additionally, the Hippo pathway has been identified as a potential target for immunotherapy [42] and has been implicated in drug resistance in various types of cancer [43]. Proteins related to the Hippo pathway have been identified as novel prognostic biomarkers and targets for regenerative medicine and cancer therapy [44, 45]. For instance, the selective MST1/2 inhibitor XMU-MP-1 can stimulate the activation of YAP, thereby promoting cell proliferation, tissue repair, and regeneration [46].

Several FDA-approved anticancer drugs or undergoing preclinical and clinical trials have also been shown to target the YAP/TAZ-TEAD complex, the major function executor of Hippo pathway. Pazopanib, for instance, inhibits YAP activity by restricting YAP/TAZ nuclear localization [47]. Similarly, compounds like VT3989, IK-930, and IAG933, currently in phase I clinical trials, inhibit the YAP/TEAD complex by binding to TEAD, thereby impeding transcriptional activity [48]. However, these small molecule compounds targeting the Hippo signaling pathway face two inevitable issues in drug development: (1) they target the Hippo effector protein YAP and the YAP/TEAD transcription complex, inhibiting their pro-cancer functions, which inevitably leads to the common drawbacks of chemotherapeutic drugs, such as poor selectivity and high toxicity [49], and (2) YAP/TAZ, being the most downstream effector proteins of the Hippo signaling pathway, are regulated by upstream core kinases and other signaling pathways [50, 51], thus their potential as anti-cancer targets are theoretically limited due to terminal functional molecules.

Merlin is an important tumor suppressor protein encoded by NF2 gene, whose loss-of-function mutations have been linked to multiple tumorigenesis including schwannoma, meningioma, and glioblastoma [22, 52, 53]. Regulating and integrating multiple pathways, the most crucial mechanism of Merlin is being an upstream activator for Hippo signaling pathway, associating with components of the mammalian Hippo signaling network [54, 55]. The PPI between Merlin and LATS activates the Hippo signaling pathway, making it a promising target for drug development. Compounds targeting the Merlin/LATS PPI can circumvent the aforementioned drawbacks of poor selectivity and limited efficacy by directly targeting the key member kinases.

PPIs play a crucial role in regulating multiple signaling pathways simultaneously due to the networked nature of protein interactions within cells [56]. Unlike traditional drugs that broadly affect cellular functions and may cause off-target effects, PPI-targeting drugs can achieve higher specificity by precisely modulating interactions between key proteins involved in tumorigenesis due to their uncompetitive nature. By binding to composite pockets at the interface of their target protein complexes, they induce greater structural variability compared to active sites of enzymes or ligand-binding pockets of receptors [57]. Recent advances in molecular glues (MGs)—including degradative MGs, non-degradative MGs, and MGs that induce self-association—have further propelled the development of PPI-targeting therapies [58]. Compounds extracted from traditional Chinese herbs have been reported to enhance PPIs by stabilizing protein complexes [59]. For example, Paris saponin VII (Dioscinin), a steroidal saponin isolated from the rhizome of Paris polyphylla, directly binds to the MST2-MOB1-LATS1 ternary complex to activate the Hippo pathway, inducing autophagy in human breast cancer cells [60]. Similarly, sculponeatin A, a diterpenoid extracted from Isodon sculponeatus, promotes the ETS1-SYVN1 interaction by binding to the interface between ETS1 and the SYVN1 complex, inducing ferroptosis in breast cancer [61]. Presumably, gracillin functions as a MG, promotes the PPI of Merlin and LATS and activates Hippo signaling pathway, exhibiting its anticancer effects.

In conclusion, our study demonstrates that gracillin displays promising anticancer activity through Merlin/LATS interaction to activate Hippo signaling pathway in vitro and in vivo (Fig. 8), providing compelling evidence for gracillin used as a candidate for cancer treatment. In addition, our findings of gracillin functions as a MG, offering novel insights into therapeutic avenues for blocking cancer progression through targeting PPIs.

Fig. 8. Schematic illustration of gracillin-induced suppression of cancer through Merlin/LATS interaction associated activation of the Hippo signaling pathway.

Gracillin enhances the protein-protein interaction between Merlin and LATS, resulting in elevated phosphorylation levels of LATS and YAP, as well as in downregulated YAP activity, effectively suppressing tumor progression.

Author contributions

JXS designed the protocols, performed experiments, drew the figures, and wrote the manuscript. HXZ, ZJZ, and XFZ performed experiments. QMZ, SJL, JQL, SYY, YQL, and RJY contributed to the data collection. XSZ, KC, and HXP provided technical support. ZSL, and HYT contributed to data analysis. MC, YY, QQ, and YBZ funding acquisition, project administration, supervision, revise the manuscript.

Competing interests

The authors declare no competing interests.

Supplementary information

The online version contains supplementary material available at 10.1038/s41401-025-01514-w.