Blocking YAP1–Liprin-β2 interaction impedes metastasis and promotes tumor suppression in head and neck squamous carcinoma

Abstract

Our research investigates the role of the YAP1–PPFIBP2 axis in the epithelial–mesenchymal transition (EMT) and its subsequent impact on invasion and migration in head and neck squamous cell carcinoma (HNSCC). Utilizing both in vitro assays and genomic analyses, we demonstrate that YAP1 upregulates EMT by suppressing PPFIBP2/liprin-β2 expression. This regulatory pathway contributes to enhanced invasiveness and correlates with poorer prognostic outcomes in HNSCC. We specifically knocked down YAP1 in SNU1041 and SCC9 cell lines using siRNA, resulting in reduced invasion and migration. These effects were reversed by subsequent administration of siPPFIBP2. In contrast, overexpression of YAP1 in SCC25 cells led to increased EMT marker activity and enhanced invasive behavior, supporting the functional role of this axis. Importantly, pharmacological inhibition of YAP1 using CA3 led to a notable decrease in EMT markers, invasion, and migration, suggesting that blocking the YAP1–PPFIBP2 axis may serve as an effective therapeutic strategy in HNSCC. In conclusion, our study identifies the YAP1–PPFIBP2 interaction as a crucial mediator of tumor aggressiveness in HNSCC, offering new insight into metastatic progression and highlighting a promising target for therapeutic intervention.

Supplementary Information

The online version contains supplementary material available at 10.1038/s41598-025-11652-0.

Introduction

Head and neck cancer is one of the most common types of cancer in the world, ranking in the 6th place in terms of global incidence. The incidence continues to rise, and the mortality rates remain at high levels^1–4. Squamous cell carcinoma of the head and neck (HNSCC), which accounts for majority of histological forms of head and neck cancers, develop from the mucosal epithelium in the oral cavity, pharynx and larynx and are malignant tumor with a 5-year overall survival rate of less than 50%^1,5. Most of these HNSCCs are diagnosed as locally advanced stage or distant metastases, and treatment of HNSCC consists mainly of an approach consisting of surgery and radiation therapy. Also available treatment options include chemotherapy, targeted therapy and immunotherapy: the monoclonal anti-EGFR antibody cetuximab and the immune checkpoint inhibitors pembrolizumab and nivolumab, which block PD-1 signaling⁶. However, despite various methodological treatment interventions, 40–60% of locally advanced tumors eventually show local recurrence or progression, and the prognosis of metastatic HNSCC is still very poor^7–11. Therefore, in order to meet the clinical unmet needs of HNSCC, many researchers have studied the biology of HNSCC until recently and researched numerous prognostic factors and novel therapeutic targets for the effective managements of this malignant tumor.

As a key regulator of development, the Hippo pathway orchestrates cell fate decisions, balancing proliferation, differentiation, and apoptosis^12,13. Interestingly, Hippo-YAP pathway may promote regeneration in organs with poor or damaged regenerative capacity, such as the adult heart and liver and intestines of elderly or diseased mice¹³. However, related to these roles, several studies warn that therapeutic YAP signal activation can lead to serious side effects. One of the key components of the hippo pathway, YAP1 (Yes-associated protein) is a protein that shuttles between the cytoplasm and nuclear according to serine–threonine kinase activity by mammalian Ste20-like (MST) kinases 1–2 and large tumor suppressors (LATSs) 1–2. Loss of Hippo pathway activity relocates YAP to the nucleus, where it drives gene programs linked to anti-apoptotic and carcinogenic processes along with transcription factors of the TEAD family. In particular, persistent YAP/TAZ activity has been observed across multiple cancer types and is sufficient to promote malignant transformation in experimental models¹³. In addition, numerous studies continue to show that dysregulation of this pathway is deeply implicated in cancer pathogenesis across a variety of cancer types^14–16.

Remarkably, a recent multidimensional comprehensive genome analysis study of HNSCC tumors has been conducted and revealed that it is associated with alterations in the genes encoding Hippo-YAP pathway components and YAP1 activity^2,17. Furthermore, in our previous results, we systematically characterized genomic data across multiple cohorts of HNSCC patients to discover molecular subtypes with poor YAP1 activation and prognosis, and used multiple platforms to analyze YAP1-associated genetic modifications¹⁸. Based on the results of these studies, which are being recently studied with interest in head and neck cancer, it seems that the recurrent or metastatic HNSCC, which is an important contributor for lowering the long-term survival rate of HNSCC patients, might be closely related to YAP signaling activation.

Here, we report that YAP1 is a clinically actionable target for metastatic properties of YAP signaling activated HNSCC. Mechanistically, transcriptional induction of PPFIBP2 by YAP1 induces EMT markers dynamics in HNSCC, which contributes to subsequent activation of the metastasis pathway. In addition, we demonstrated that inhibition of YAP1 with CA3 suppressed PPFIBP2 expression and neutralized the metastatic properties of tumors through restoration of EMT markers. We therefore provide a proof-of-principle in a preclinical model that inhibition of YAP1 signaling is an attractive therapeutic strategy for overcoming HNSCC cancer therapeutic modalities, particularly tumors malignant with Hippo-YAP signaling.

Results

The level of YAP1 strongly associated with tumor invasion and migration in YAP1 activated HNSCC

Previously, we have described the clinical significance of YAP1 activation in head and neck cancer and identified 292 genes clinically significant YAP1-related gene signature¹⁸. First, we investigated the biologic mechanism related to YAP1 activity. In order to study the biological roles of the YAP1 associated genes, functional annotation analysis was performed using DAVID (https://david.ncifcrf.gov) website. Considering the p values, the most enriched KEGG pathways was “focal adhesion”. In addition, “apoptosis” and “cell adhesion molecules (CAMs)” etc. are distributed. (Supplementary Fig. 1A). In the biological event with the highest number of enriched genes, “focal adhesion” took first place, followed by “pathway of cancer” and “cell adhesion molecule (CAM)”. In particular, GSEA was implemented as a criterion for finding the biological process of gene ontology containing two or more genes with significantly rich p values (NOM p value < 0.05) in the YA and YI groups (Supplementary Fig. 1B). As a result, it was confirmed that some genes related to ‘lymph or lymphatic vessel of regulation and angiogenesis’ and ‘vascular endothelial growth factor’ were enriched and distributed in the YA group. In addition, for the experiment to verify the mechanism, we performed that the genetic information of 17 head and neck cancer cell lines was obtained using Cancer Cell Line Encyclopedia (CCLE), and the cell lines were clustered into YAP1-asctivation (YA) and YAP1-inhibition (YI) using the YAP1 signature developed in previous studies¹⁸ (Fig. 1A). Prediction of these cell lines was performed through an SVM algorithm using BRB-array tools, and as a result, it was confirmed that 17 cell lines were clustered into YA group containing 5 cell lines and YI group containing 12 cell lines. Among the five cell lines classified into the YA group, SNU1041 and SCC9 exhibited relatively higher YAP1 protein levels, and were therefore selected as representative models for further functional validation (Fig. 1B and supplementary Fig. 1C).

Fig. 1.

YAP1 is crucial for invasion and migration in YAP1 activated HNSCC cell lines. (A) YAP1-associated gene signature subtypes in HNSCC cell lines (n = 17 in CCLE data set). HNSCC cell lines were stratified according to Support vector machine (SVM) predictor probability from the YAP1 predictor. Blue indicates YAP1 inactivation (YI) and orange indicates YAP1 activation (YA). The cell lines highlighted in asterisk in the table represent the cell lines used in subsequent experiments. (B) YAP1 protein expression of 17 HNSCC cells are shown using western blot (WB). β-actin was loaded as a control. (C) Knockdown of YAP1 using siRNA was determined to decrease the band in WB results in SNU1041 and SCC9 cell lines. (D) Representative images of the cell invasion ability assay are shown. The effects of different treatments on the invasion ability of SNU1041 and SCC9 cells were determined by counting the number of invading cells. Data are means ± SEM. (n = 3). Statistical significance was determined using a t-test. ***p < 0.005 versus control. SEM, standard error of the mean. (E) Representative images of migration assays were shown in SNU1041 and SCC9 cell lines. The quantitative graph represents the percentage of wounded area. The microscopical images were acquired after 48 h of both cells were scratched. (F) SCC25 cells were transfected with GFP-conjugated empty and YAP1 vectors. YAP1 protein expression was evaluated through WB in transfected cells. (G) Representative images of invaded cells were acquired by invasion ability assay. The quantitative graph shows the number of invading cells. (H) Representative images of migration assays were shown in SCC25 cell line. (I) Protein level of EMT markers were evaluated with treatment of siGFP and siYAP1. (J) EMT markers were evaluated with empty and YAP1 vectors using WB. GAPDH was loaded as a control.

To elucidate the role of YAP1 in regulating migration and invasion in HNSCC, we treated two siRNAs targeting different sites of YAP1 in YAP1-activated cells SNU1041 and SCC9 and confirmed the inhibition of YAP1 expression, respectively (Fig. 1C). At the same time, the trans-well invasion assay showed a statistically significant decrease in the number of invading cells in YAP1 siRNA groups compared to the siGFP group (Fig. 1D). Similar conclusions were drawn from the wound-healing assay. The wound-healing assay revealed that the closing rate of scratch wounds was remarkably decreased in YAP1 siRNA groups compared with the siGFP group (Fig. 1E). In order to verify the role of YAP1 in more detail, we performed an invasion assay and a wound healing assay when YAP1 was overexpressed in SCC25, a cell line that was included in the YI group and had a relatively low protein level of YAP1 (Fig. 1F). Consistently, the trans-well invasion assay showed a significant increase in the number of invading cells in YAP1-overexpressed groups compared to the no-insert group (Fig. 1G). In the wound healing analysis, it was verified that the closure rate of scratched wounds was significantly increased in the YAP1-overexpressed group compared to the no-insert group (Fig. 1H).

Furthermore, knockdown of YAP1 also altered the expression of EMT marker proteins. In cells mimicking YA, such as the SNU1041 and SCC9 cell lines, treatment with siYAP1, compared to treatment with siGFP, led to an increase in the expression of epithelial markers E-cadherin and ZO-1, while the expression of the mesenchymal marker N-cadherin decreased (Fig. 1I). Conversely, in SCC25 cells, which mimic YI, the addition of a YAP1 overexpression vector resulted in decreased expression of epithelial markers and increased expression of mesenchymal markers compared to the empty vector control group (Fig. 1J). Based on these results, we concluded that YAP1 has the potential to increase the migration and invasion of HNSCC cells.

The importance of PPFIBP2 as a YAP1 downstream factor in YAP1-mediated metastatic property

In order to find the sub-factor of YAP1 that induces cell invasion and migration, 77 genes with significance in the recurrence-free survival rate of HNSCC patients and the top 1000 of YAP1-bound genes with 1 kb from TSS derived from chip-atlas (https://chip-atlas.org/target_genes) were obtained. As a result of the analysis of these gene pools, we obtained three genes that were identically included in the two gene pools (Fig. 2A). The three genes were used to predict the patient’s prognosis according to the expression of each gene using clinical data of the TCGA HNSCC patient (Supplementary Fig. 2A). Interestingly, in the case of the PPFIBP2 gene, it was confirmed that the prognosis of HNSCC was significantly worse in the group with low gene expression (p = 0.003, Supplementary Fig. 2A). As a result of analyzing the gene expression data obtained from TCGA to confirm the gene expression correlation between YAP1 and PPFIBP2, it was confirmed that the negative correlation of YAP1 and PPFIBP2 (r = − 0.14 and p = 0.002, Fig. 2B). After siGFP and siYAP treatment in SNU1041 and SCC9 cells, the expression of PPFIBP2 mRNA was measured through real-time PCR (Fig. 2C). As a result, it was confirmed that the expression of PPFIBP2 was significantly increased in the cells treated with siYAP1. As a result of checking the protein level at the same time, it was confirmed that when YAP1 expression was suppressed in both cells, the expression of Liprin β2, a protein of PPFIBP2, was also increased (Fig. 2D). However, the qRT-PCR results demonstrated that the knockdown of YAP1 did not alter the mRNA levels of AIG1 and ADPRHL1, which were other downstream candidates (Supplementary Fig. 2B). Based on these results, it appears that YAP1 may regulate the mRNA and protein expression levels of PPFIBP2.

Fig. 2.

YAP1-mediated negative regulation of PPFIBP2/Liprin-β2 is crucial role in invasion and maigration of HNSCC. (A) Venn diagram showing the relationship between 77 genes with significance in the recurrence-free survival rate of HNSCC patients and the top 1000 of YAP1-bound genes with 1 kb from TSS derived from chip-atlas. It shows the overlap three candidates of invasion or metastasis involved sub-factor of YAP1. MACS; Model-based Analysis of ChIP-seq algorithm. (B) The correlation plot between YAP1 and PPFIBP2/Liprin-β2 gene. The correlation coefficient value was calculated using Pearson’s correlation method (r = − 0.14 and p = 0.002). (C) SNU1041 and SCC9 cells were treated with siYAP1, the mRNA level of PPFIBP2/Liprin-b2 was evaluated using qRT-PCR. (D) The protein level of PPFIBP2/Liprin-b2 was evaluated in SNU1041 and SCC9 cells treated with siYAP1. (E) Representative images of the cell invasion ability assay were shown. The effects of different treatments on the invasion ability of SNU1041 and SCC9 cells were determined by counting the number of invading cells. Data are means ± SEM. (n = 3). Statistical significance was determined using a t-test. ***p < 0.005 versus control. SEM, standard error of the mean. (F) Representative images of migration assays were shown in SNU1041 and SCC9 cell lines with treatment of siYAP1 or siYAP1 + siPPFIBP2. The quantitative graph represents the percentage of wounded area. The microscopical images were acquired after 48 h of both cells were scratched. (G) The protein levels of EMT markers were evaluated with treatment of siYAP1 or siYAP1 + siPPFIBP2 in SNU1041 and SCC9 cell lines. (H) YAP1 suppresses the expression of PPFIBP2/Liprin-β2 to regulate EMT markers, resulting in increased invasion and migration, which leads to a poor prognosis in HNSCC cells. The YAP1 inhibitor CA3 inhibits this YAP1-PPFIBP2 axis, reducing invasion and migration in HNSCC cells, thereby potentially improving prognosis.

To confirm that YAP1-mediated metastatic potential can be reversed by PPFIBP2 regulation, we tested cell invasion assay and wound healing assay in HNSCC with YAP1 activation such as SNU1041 and SCC9 cells. In the trans-well invasion assay, siYAP1 treatment diminished cellular invasion capabilities relative to siGFP. Notably, subsequent treatment with siPPFIBP2 following siYAP1 rescued the reduced invasion ability within the SNU1041 and SCC9 cell lines (Fig. 2E). Correspondingly, the results of the migration assay showed that the decreased wound healing ability observed with siYAP1 treatment was recovered when siPPFIBP2 was added after siYAP1 treatment (Fig. 2F).

Furthermore, the level of EMT-related proteins showed that knockdown of YAP1 increased the expression of epithelial marker including E-cadherin and Zo-1 in SNU1041 and SCC9 cells, while the expressions of mesenchymal marker including N-cadherin were inhibited (Fig. 2G). Noteworthy, it was confirmed that all level of EMT markers was recovered after additional treatment with siPPFIBP2 in the YAP1 siRNA group of SNU1041 and SCC9 cells. Based on these results, it is evident that PPFIBP2 plays a crucial role as a pathway molecule in the YAP1-induced EMT process in HNSCC (Fig. 2H). Correspondingly, in SCC25 cells, overexpression of YAP1 suppressed the expression of the mRNA level of PPFIBP2 and the expression of the protein level of liprin β2 (Supplementary Fig. 2C,D).

The inverse axis of YAP1-PPFIBP2 is associated with prognosis in HNSCC patients

Having explored the molecular mechanism by which the inverse axis of YAP1-PPFIBP2 confers tumor-metastasis phenotypes, we examined whether the inverse axis of YAP1-PPFIBP2 is also in patients with HNSCC. In transcriptome analysis using the cancer genome atlas (TCGA) data, the YA group of HNSCC patient genome showed a significantly increased expression of YAP1 than in the YI group (Fig. 3A). Conversely, it was confirmed that the expression distribution of PPFIBP2 was significantly reduced in the YA group than in the YI group. We next aimed to further determine the clinical relevance of the YAP1-PPFIBP2 in patients with HNSCC. Previous reports show that YAP1 activation was correlated with poor prognosis of HNSCC¹⁸. Here, we additionally found that the lymph node (LN) metastasis rate of HNSCC patients in the YA group with low PPFIBP2 expression was significantly higher than that of the YA group with high PPFIBP2 expression (p = 0.017, Fig. 3B). However, in the YI group, the LN metastasis rate according to the expression of PPFIBP2 was not significant (p = 0.499). We investigated using the Kaplan–Meier plots whether the inverse axis of YAP1-PPFIBP2 also contributed to survival rates in HNSCC patients. As expected, examining the relationship of combined YAP1/PPFIBP2 status with patient’s survival outcomes, the Kaplan–Meier plots demonstrated that patients with combined YA/PPFIBP2-low level showed significantly worse 5-years overall survival (p = 0.0019) than patients group with combined YI/PPFIBP2-high level (Fig. 3C). Overall, these data indicate that the inverse axis of YAP1-PPFIBP2 is highly related with tumor LN metastasis as well as serves as an important prognostic factor in HNSCC.

Fig. 3.

The reverse axis of YAP1-PPFIBP2 is strongly correlated with LN metastasis and survival in patients with HNSCC. (A) A box plot was used to compare the expression levels of YAP1 and PPFIBP2 mRNA between group YA and group YI. Significance was calculated by a two-tailed t-test. (B) A box plot was used to compare the percentages of pathological N stages based on PPFIBP2 expression levels between Group YA and Group YI. (C) The Kaplan–Meier plot illustrates the differences in overall survival (OS) between two groups: YA with low PPFIBP2 expression (PPFIBP2-L) and YI with high PPFIBP2 expression (PPFIBP2-H).

Inhibitor of YAP1 reduces EMT phenomenon of YAP1 activation tumor cells via upregulating PPFIBP2

To verify the biochemical effects of YAP1-PPFIBP2 axis in HNSCC, we used previously validated SNU1041 and SCC9 cell lines and additionally another HSC3-M3 with strong metastatic potential¹⁹. We first determined the concentration at which the YAP1 inhibitor CA3 did not contribute to the survival of SNU1041, SCC9 and HSC3-M3 cells (Supplementary Fig. 3). The trans-well invasion assay showed a statistically significant decrease in the number of invading cells in CA3-treated cell lines compared to the DMSO-treated cell lines (Fig. 4A). The wound-healing assay revealed that the closing rate of scratch wounds was remarkably decreased in CA3-treated all cells compared with the DMSO-treated all cells (Fig. 4B). Furthermore, inhibition of YAP1 with CA3 modestly increased the level of liprin β2, consistent with results from siRNA-mediated knockdown of YAP1. At the same time, epithelial markers such as E-cadherin and Zo-1 increased, whereas mesenchymal markers such as N-cadherin decreased in all CA3-treated HMSCC cells (Fig. 4C). These results demonstrate that the functional and biochemical properties of the YAP1-PPFIBP2 axis are conserved across different types of HNSCC cells, indicating that YAP1 inhibitors are viable agents for controlling YAP1-mediated metastatic properties of HNSCC through regulation of liprin β2.

Fig. 4.

The YAP1 inhibitor, CA3 controls EMT markers and reduces cell invasion and migration. (A) The representative figure displays the comparative invasion abilities between the non-treated group and the CA3-treated group. Accompanying this, a quantitative graph presents the counting numbers of invaded cells, providing a numerical representation of the invasion assays. (B) HNSCC cells were treated with 0.5 µM of CA3. The expression levels of EMT marker proteins between non-treated and CA3-treated groups were compared using Western blot analysis. (C) Representative images depict the results of the migration assays on cells. The quantitative graph shows the percentage of the wounded area 48 h post-treatment with CA3. The significance of the differences observed was calculated using a t-test, with a highly significant result indicated by ***p < 0.005.

Inhibition of YAP1 suppresses EMT phenotype in YAP1-activated tumor cells by upregulating PPFIBP2

To demonstrate the possibility of genomic disruption of YAP1 in HNSCC cells, we infected HSC3-M3 cells with AsCas12a-expressing lentiviral vectors to target YAP1 using T7 endonuclease I (T7E1) mismatch detection assay (Supplementary Fig. 4A). Sanger sequencing confirmed that indel formation occurred at the predicted site (Supplementary Fig. 4B). YAP1 knockout using YAP1 gRNA led to an increase in the expression of Liprin beta 2 in HSC3-M3 cells (Supplementary Fig. 4C). Furthermore, YAP1 deficiency upregulated the levels of E-cadherin and Zo-1, while downregulating N-cadherin expression. This indicates that YAP1 negatively regulates Liprin beta 2 and also modulates EMT-associated proteins. To assess the impact of YAP1 knockout on the invasive capacity, an invasion assay was performed (Supplementary Fig. 4D). The findings revealed that YAP1 depletion diminished the number of invasive cells. Separately, the effect of YAP1 knockout on cell migration was evaluated through a migration assay (Supplementary Fig. 4E). The results showed that YAP1 deficiency impaired the wound healing ability.

Discussion

Metastasis remains the primary challenge in cancer treatment. The most common site of metastasis in HNSCC is cervical lymph node, because drainage channel of the neck is rich in lymph nodes with more than 400 in number²⁰. More than 30% of patients with HNSCC have cervical lymph node metastasis at the time of diagnosis, and cervical lymph node metastasis is heavily associated poor prognosis²¹. Therefore, the metastatic ability of cancer cells in HNSCC patients is one of the most important prognostic factors. In addition, the status of metastatic cells found in a patient’s cancer influence the choice of adjuvant therapy used to suppress disease recurrence²². In some cases, patients with HNSCC with distant metastasis are often offered palliative care, because of limited treatment options. Therefore, it is urgent to find informative and effective markers for metastatic processes including EMT and invasion, of HNSCC in terms of clinical treatment and prediction. In the process, the determination of YAP1 activation using the YAP1-related signatures is known as an important factor in predicting poor prognosis in HNSCC patients^1,18. Thus, strategies to fully understand and disrupt the YAP1 activation signaling pathway may overcome the problem of metastatic status in HNSCC. In this study, we demonstrated the crucial role of PPFIBP2 at the crossroads between YAP1 and the metastatic properties of HNSCC by identifying PPFIBP2 as a YAP1 transcriptional target. In this vein, the YAP1-PPFIBP2 axis was conserved in TCGA cohort and tumor cell lines of HNSCC. In addition, intratumoral levels of YAP1 and PPFIBP2 strongly correlated with metastatic progression and survival of HNSCC patients, indicating that YAP1 and PPFIBP2 expression status in tumor tissue may serve as good prognostic markers.

Notably, YAP1 is a well-characterized downstream effector of the Hippo pathway, functions as a transcriptional co-activator. The role of YAP1 in cancer development still remains controversial. Accumulating evidences suggest that the different roles of YAP1 in oncogenesis might be tissue- and cell context-specific. YAP1 was defined as a tumor suppressor that induces apoptosis in response to DNA damage in collaboration with p73 and promyelocytic leukemia (PML)^23–26. On the other hand, YAP1 was also described as an oncogene. Its expression was amplified in human hepatocellular carcinoma (HCC) and many other malignancies²⁷. Additionally, overexpression of YAP1 also transformed the human mammary epithelial cells, MCF-10A cells²⁸. As a transcriptional co-activator, YAP1 can bind to oncogene or tumor suppressor gene transcription factors, depending on particular tissue type. Interestingly in the midst of this controversy, we have previously reported that the expression of YAP1 was substantially higher in many tumors without YAP1 amplification, suggesting that copy number alteration is not the only mechanism of YAP1 activation in HNSCC. So we have proposed a gene expression signature reflecting YAP1 activation (YAP1 signature) that is significantly associated with the prognosis of patients with HNSCC by systematically analyzing genomic copy number data and mRNA expression data of YAP1 in HNSCC¹⁸. Here, we found that PPFIBP2 plays an important role in the activated YAP1-mediated metastasis by enhancing the linkage that increases tumor cell invasion and migration capacity. Although we could not specifically confirm the process that explains the function of PPFIBP2 in YAP1 signaling, we identified PPFIBP2 as a novel sub-factor that is reversely transcriptionally regulated by YAP1 in HNSCC cells, which leads to activated YAP1-mediated metastasis and it was newly discovered that it plays a decisive role in the ability to tumor cell invasion and migration. Interestingly, it was confirmed that the change of EMT markers (E-cad, zo-1, N-cad) closely related to the cell adhesion was remarkably switched according to the expression of liprin β2.

Indeed, liprin are a multifunctional family of scaffold proteins, identified by their involvement in several important neuronal function related to signaling and organization of synaptic structures. More recently, the knowledge on the liprin family has expanded from neuronal functions to processes relevant to cancer progression, including cell adhesion, cell motility, cancer cell invasion, and signaling²⁹. Liprins belong to the LAR (leukocyte common antigen related) protein tyrosine phosphatase interacting protein family³⁰. In mammals, the liprin family consists of liprin-α (α1, α2, α3, α4), β (β1, β2) proteins, which are well conserved in evolution³¹. Liprin α1 is the only liprin-α having widely distributed mRNA and protein expression within different tissues, while liprin α2-4 is expressed in specific cells (brain, heart or muscle tissues)²⁹. Focal adhesions are cellular multiprotein structures that link the actin cytoskeleton to the ECM via the transmembrane integrin receptors^32,33, particularly the structures important in actin- and adhesion-dependent mesenchymal cell migration are the lamellipodia which arise at the leading edge of migrating cells^34,35. Several studies point to the importance of liprin-α1 in focal adhesions and in regulation of the turnover, size and shape of these structures³⁶, and it has been demonstrated that liprin-α1 is part of the integrin adhesome network³⁷. In the invasion and oncogenic signaling, there is already many evidence that liprin α1 is important in cell edge dynamics, cell motility, and invasion. To explain in detail, Liprin-α1 suppresses the expression of the transmembrane metastasis suppressor CD82 in HNSCC and breast cancer cell lines³⁸, and the expression of liprin-α1 and CD82 is negatively correlated in clinical breast cancer samples³⁹. And liprin-α1 interacts with ING4 (inhibitor of growth 4), and positively regulates cell migration in RKO colon carcinoma cells in ING4-dependent manner⁴⁰. Liprin-α1 is expressed at the protrusions, essential for movement in metastatic breast cancer cells⁴¹. Similarly, liprin-β1 co-localizes at the cell edge with liprin-α1, and contributes positively to tumor cell motility in addition to liprin-α1⁴². Liprin-β1 not only interacts with the metastasis-associated protein S100A4, which may modulate liprin-α1/liprin-β1 interaction, thus acting as a component of the LAR-liprin-α1/liprin-β1 network⁴³, but also the N-terminus of liprin-β1 co-immunoprecipitates with the adaptor proteins Kank1 and Kank2, which are suppressors of the proliferation and migration in melanoma cells⁴⁴. While several studies suggest liprin-β1 to promote tumor progression, liprin-β2 has opposing role in the cancer invasive phenotype, and silencing of liprin‐β2 has no effects on lamellipodia density and stability in MDA-MB-231 cells⁴². Interestingly, it has been reported that ERK2 knockdown leads to increased expression levels of liprin-β2 in MDA-MB-231 breast cancer cells and liprin-β2 knockdown restores the invasive phenotype of ERK2-depleted cells in three-dimensional ECM⁴⁵. As shown in Fig. 2, we also confirmed that when the expression of liprin β2 is increased by YAP1 inhibition, the invasiveness and mobility of the cells, including the expression of EMT markers in HNSCC cells, are switching. However, the clear function of liprin β2 is still unknown. Because current information is still limited, further epidemiological studies are needed to understand the detailed mechanism of liprin β2 in oncogenic signaling and tumor progression of various cancers.

Altogether, we propose that YAP1 activated cancer cells suppress preferential expression of PPFIBP2 and block liprin β2 accumulation. Through this process, down-regulated liprin β2 maintains low levels of epithelial markers and high levels of mesenchymal marker, thereby increasing cell invasion and migration capacity. Considering the expression of EMT markers, it enhances the metastatic ability of cancer cells through upregulation of mesenchymal markers rather than epithelial markers. Furthermore, we show that inhibiting YAP1 with CA3 inhibits tumor cell invasion and migration, thereby reducing tumor metastasis capacity. In this process, the expression of liprin β2 is accumulated and the expression of epithelial and mesenchymal markers is also found to be switched. Therefore, our data provide evidence that the blocking of YAP1- liprin β2 axis may be a promising strategy that will particularly help combat activated YAP1- mediated tumors.

A limitation of this study is that it does not fully account for the influence of the tumor microenvironment, including stromal-derived mechanical and biochemical cues known to regulate YAP1 activity. This was primarily due to the use of simplified in vitro systems designed to dissect tumor cell–intrinsic signaling mechanisms in isolation. While this approach allowed us to clarify the direct functional relationship between YAP1 and PPFIBP2, future studies incorporating co-culture models or stromal-conditioned environments will be important to assess the contextual robustness and therapeutic relevance of this axis.

Methods

Functional and pathway enrichment analysis

Gene ontology (GO)⁴⁶ analysis and Kyoto Encyclopedia of Genes and Genomes (KEGG)⁴⁷ pathway enrichment analysis of YAP1-associated gene signature (YAP1 signature)¹⁸ were performed using Database for Annotation, Visualization and Integration Discovery (DAVID)⁴⁸. p < 0.05 was chosen as the cut-off criterion. GO is a set of unified vocabulary to describe molecular functions (MF), biological processes (BP) and cellular components (CC) of biology, whereas KEGG analysis was performed to aid understanding of the signaling pathways involving YAP1 signature.

Cell culture

The human HNSCC cell lines SNU1041 (CVCL_L085) was purchased from the Korean Cell Line Bank (KCLB, Seoul, Korea). The human HNSCC cell lines SCC9 (CVCL_1685), SCC25 (CVCL_1682), and HSC3-M3 (CVCL_8323) cells were purchased from the American Tissue Type Cell Collection (ATCC, VA, USA). Cells were cultured in RPMI-1640 (SNU1041), DMEM/F12 (SCC9 and SCC25), or EMEM (HSC3-M3) medium supplemented with 10% heat-inactivated fetal bovine serum (Hyclone, UT, USA) and 1% penicillin–streptomycin (Corning Inc., NY, USA). All cell lines were maintained at 37 °C in a humidified incubator containing 5% CO₂. All cell lines were authenticated using short tandem repeat (STR) profiling in three past years. A list of and the elated to online database have been made available for checking the authenticity of the cell lines in terms of the STR profile, as well as the correct name and description for each known cell lines (https://www.cellosaurus.org/str-search/). When, thawing the initial cell cultures, all cell lines were supplied with mycoplasma removal agent (MRA, #093050044, MP Biomedicals, Santa Ana, CA) to eliminate any potential mycoplasma contamination.

Transfection of siRNA

Cells were transfected with negative control siGFP, siYAP1 (commercial siRNA purchase from IDT) or siPPFIBP2 (commercial siRNA purchase from bioneer) using Lipofectamine RNAiMAX Transfection Reagent (Thermo Fisher Scientific), according to the manufacturer’s instructions.

YAP1 overexpression

pCMV-YAP1 plasmid containing the coding region of the human YAP1 gene (NM_001130145) was purchased from Addgene (#66853). The HNSCC cells were plated into 6-well plates at a 70% confluence. The cells were then transfected with pCMV-YAP1 or pCMV-empty according to multiplicity of infection MOI = 100. Approximately 10 h later, fresh culture medium was supplemented to HNSCC cells and incubated for 48 h. The efficacy of adenovirus transfection was observed by Western blotting.

Trans-well invasion assay

Trans-well assays to evaluate cell invasion were performed using a 24-well Trans-well insert (3422, Corning Inc.) with an 8 μm pore polycarbonate membrane insert. Trans-well inner chambers coated with 100 μL of diluted Matrigel (volume of original Matrigel: FBS-free DMEM/F12, EMEM, or RPMI1640 = 1:10) were used for the invasion assay, and these chambers were incubated at 37 °C for 4 h until the gel was set before use. 300 μL of FBS-free DMEM/F12, EMEM, or RPMI1640 with 3 × 10⁴ cells was added to the inner chamber, and 750 μL of DMEM/F12, EMEM, or RPMI1640 with 10% FBS was added to the outside wells. After incubation for 48 h for the invasion assay, cells that had invaded and were adhered to the lower surface were stained with hematoxylin and eosin. The cells were counted in four representative fields under light microscopy (200× magnification).

Migration assay

The confluent SNU1041, SCC9 and SCC25 cells were seeded in 6-well plates at the 1 × 10⁵ cells/well density. After incubation for 24 h, scratches were induced using sterile 1000 μL pipette tips in all the monolayers of cells across the diameter of the wells. The media was aspirated, and the cells were washed twice. The treatment medium containing siGFP, siYAP1, siPPFIBP2, YAP1 overexpression vector, and CA3 were added to the wells. The plates were incubated at 37 °C. The images of each well taken at 48 h of treatment using a bright field microscope.

Western blot

Lysates from HNSCC cells were homogenized in RIPA buffer [1% Triton X-100, 1% sodium deoxycholate, 0.1% sodium dodecyl sulfate (SDS), 150 mM NaCl, 50 mM Tris–HCl (pH 7.5), and 2 mM ethylenediamine tetra-acetic acid (pH 8.0)] purchased from Biosesang (Seongnam-si, Republic of Korea) containing a protease inhibitor cocktail. Protein concentration was then quantified using a Pierce Micro BCA Protein Assay Kit (ThermoFisher Scientific, MA, US), according to the manufacturer’s protocol and equal amounts of protein, mixed with loading dye (5× SDS–polyacrylamide gel electrophoresis loading buffer; iNtRON Biotechnology, Seongnam-si, Republic of Korea), were added to each lane and resolved using an SDS–polyacrylamide gel. Following electrophoresis, proteins were transferred to polyvinylidene difluoride membranes (Millipore, MA, USA) and blocked for 1 h in 5% BSA and Tris-buffered saline with 0.1% Tween-20. These membranes were then treated with appropriate primary antibodies and incubated overnight at 4 °C⁴⁹. The following antibodies were used: [antibody list] anti-YAP1 (4912S, Cell Sinaling Technology, MA, USA), anti-E-cadherin (3195S, Cell Signaling Technology), anti-ZO-1 (8193S, Cell Signaling Technology), anti-N-cadherin (13116S, Cell Signaling Technology), and anti-GAPDH (sc-365062, Santa Cruz Biotechnology, TX, USA), and anti-PPFIBP2 (PA5-51666, ThermoFisher Scientific). The blots were then washed and treated with secondary antibody (Cell Signaling Technology) for 1 h at room temperature, and the protein-antibody complexes were detected using enhanced chemiluminescence (RPN2232; GE Healthcare, IL, USA) according to the manufacturer’s protocol.

Real-time quantitative RT-PCR

Cells were seeded in a 6-well plate and transfected with siRNA. Total RNA was extracted by using the TRIzol-LS reagent according to the manufacturer’s protocol (GeneAll). We used a NanoDrop spectrophotometer for assessment of RNA concentration and quality. Total RNA was reverse transcribed to cDNA using a Tetro cDNA Synthesis Kit (Bioline, London, UK), according to the manufacturer’s recommended protocol. Real‐time quantitative PCR was performed using the SensiFAST™ SYBR Hi-ROX Kit (Bioline) with specific primers. All real‐time quantitative PCR experiments were performed in triplicate, and quantification cycle (Cq) values were determined using stepone software v2.3 (Applied Biosystems, Foster City, CA, USA). PCR was performed under the following conditions: initial denaturation at 95 °C for 30 s followed by 40–45 cycles of denaturation at 95 °C for 5 s; annealing at 60 °C for 30 s⁵⁰. The PCR reaction was evaluated by melting curve analysis. Each sample was amplified in triplicate, and data were analyzed by relative quantitation using the ΔΔCt method and were normalized to β-actin. Primer sequences are listed in Supplementary Table 1. Each reaction was performed in technical duplicates and the mean of three independent biological replicates was calculated.

Chemicals

CA3 (CIL56) was purchased from Sellekchem (S8661, Selleckchem, TX, USA). CA3 has potent inhibitory effects on YAP1/Tead transcriptional activity.

Cell viability assay

SNU1041, SCC9 and HSC3-M3 cells were seeded in 96-well plates and treated with CA3 (0.001, 0.01, 0.1, 1, 10, 100 μM). 5 × 10³ cells/well were seeded in 96-well plates. After 24 h of treatment with CA3, cell viability was measured according to the manufacturer’s instructions (EZ-cytox, DoGenBio, Seoul, Republic of Korea) and 10 μL of the reagent was added to each well. After 2 h of incubation in a CO₂ incubator, the conversion of the reagent into chromogenic formazan was evaluated with a spectrophotometer at 570 nm.

CrRNA design and cloning for CRISPR-AsCas12a genome editing of YAP1

AsCas12a CRISPR RNAs (crRNAs) targeting the YAP1 gene were designed using Cas-Designer (http://www.rgenome.net/cas-designer/) and experimentally tested for selection of the crRNA with maximum on-target activity. The selected target sequence was 5′-AGATACCTGATGATGTACCT-3′. Two complementary oligonucleotides (5′-AGATAGATACCTGATGATGTACCT-3′ and 5′-AAAAAGGTACATCATCAGGTATCT-3′), containing the YAP1 guide sequence and BsmBI ligation adapters, were annealed and ligated into BsmBI-digested BPK3079 vector (Addgene #78741) using T4 DNA ligase (New England Biolabs) at room temperature for 2 h. The ligation reaction was directly transformed into DH5α bacteria and plated on agar plates containing ampicillin (100 µg/mL). Single colonies were cultured in LB media for 16 h, and plasmid DNA was isolated from the bacterial cultures using the NucleoBond Xtra Midi Kit protocol (Macherey–Nagel). The sequence of the prepared plasmid was confirmed by Sanger sequencing.

Assessment of genome editing efficiency by T7E1 assay

To assess gene editing efficiency, a T7 Endonuclease I (T7E1) assay was conducted as follows. The genomic region encompassing the edited site was amplified by hemi-nested PCR the with cycling condition: 98 °C for 2 min, followed by 30 cycles of 98 °C for 10 s, 55 °C for 5 s, and 72 °C for 30 s, with a final extension at 72 °C for 5 min. The initial PCR products using forward 1 and reverse primers were then subjected to a second round of PCR using forward 2 and reverse primers under the same conditions, except that the extension step at 72 °C was shortened to 20 s, to increase the quantity and quality of the amplicons. The final PCR products were denatured at 95 °C for 5 min and re-annealed by gradually cooling to room temperature to form heteroduplexes. The re-annealed PCR products were then treated with T7 Endonuclease I (New England Biolabs) at 37 °C for 25 min to cleave mismatched DNA regions. The digested products were separated by electrophoresis on a 2% agarose gel and visualized under UV light. PCR primers for YAP1 amplification were as follows: Forward 1 (5′-CTGTGTTCTCCAGTGTCGAAT-3′), Forward 2 (5′-GTACTTAGATATTCGGCTGCAA-3′), and Reverse 1 (5′-AACCAAGACTCAGTAAAGCTTTC-3′).

Statistical analysis

All the values are represented as the mean ± S.E.M. of individual samples. Statistical significance of the data was assessed by paired Student’s t-test or one-way ANOVA as appropriate. All statistical tests were two-tailed t-tests and p values less than 0.05 were considered statistically significant.

Electronic supplementary material

Below is the link to the electronic supplementary material.

Abbreviations

HNSCC

Head and neck squamous cell carcinoma

YAP1

Yes-associated protein 1

MST

Mammalian Ste20-like kinase

LATSs

Large tumor suppressors

PPFIBP2

Protein tyrosine phosphatase receptor type F polypeptide-interacting protein-binding protein 2

GO

Gene ontology

KEGG

Kyoto encyclopedia of genes and genomes

DAVID

Database for annotation, visualization and integration discovery

MF

Molecular functions

BP

Biological processes

CC

Cellular components

STR

Short tandem repeat

CAMs

Cell adhesion molecules

CCLE

Cancer cell line encyclopedia

EMT

Epithelial–mesenchymal transition

PML

Promyelocytic leukemia

HCC

Hepatocellular carcinoma

Author contributions

Conceptualization: SRW, and YGE; Data curation: SRW, SYK and JKH; Formal analysis: JKN and MKL; Methodology: SRW, MKK, SKM, JYS and KYN; Project administration: SRW and YGE; Supervision: SGK and YGE; Validation: JKN; Roles/Writing—original draft: SRW and JKN; Writing—review and editing: SRW and YGE.

Data availability

The Sanger sequencing performed in this study was only used for the verification of the plasmid identity. No new or large-scale sequence data requiring deposition were generated. The datasets generated and/or analyzed during the current study are not applicable to mandatory public deposition. All relevant data supporting the findings of this study are available from the corresponding author upon reasonable request.

Declarations

Competing interests

The authors declare no competing interests.