ABSTRACT
-
Background/Aims
Recurrence is a major factor limiting the long-term survival of patients with intrahepatic cholangiocarcinoma (ICC). The molecular characteristics and potential therapeutic targets in ICC remain largely undefined.
-
Methods
Following our previous whole-exome sequencing study, we performed targeted sequencing, Sanger sequencing, and quantitative PCR to assess all coding exons and copy number variations of LATS2 in 400 primary ICC samples. Kaplan–Meier survival curves were used to assess the impact of LATS2 mutation, copy number loss, and low expression levels on recurrence-free survival and overall survival in ICC patients. In addition, we investigated the functional role and underlying mechanisms of LATS2 variation in ICC tumor progression and resistance to anti-PD-1 therapy.
-
Results
Among a total of 400 ICC cases, the overall frequencies of LATS2 somatic mutation and copy number loss were 3% (12/400) and 34% (136/400), respectively. Both types of variation were correlated with decreased LATS2 protein expression, increased tumor recurrence, and poor overall survival. Biofunctional investigations revealed a tumor-suppressor role of LATS2. Inactivation of LATS2 suppressed the Hippo signaling pathway, leading to aberrant activation of YAP, which upregulated PD-L1 expression and CCL2 secretion, suppressed CD8+ T cell infiltration, and enhanced recruitment of M2-like macrophages, thereby promoting immune evasion, tumor progression, and resistance to anti-PD-1 therapy.
-
Conclusions
Our study reveals a pivotal clinical association and mechanistic role of LATS2-inactivating variation in ICC, which may serve as a useful biomarker for precision therapy.
-
Keywords: Intrahepatic cholangiocarcinoma; Recurrence; LATS2; PD-L1; Immunotherapy
Study Highlights
• LATS2 somatic mutation, copy number loss, and low expression are associated with postoperative recurrence and poor survival in ICC.
• LATS2 inactivation promotes ICC progression through aberrant YAP activation.
• LATS2 inactivation drives immunosuppressive microenvironment and anti-PD-1 resistance, which can be reversed by YAP inhibition.
• LATS2-inactivating variation may serve as a useful biomarker for guiding precision immunotherapy in ICC.
Graphical Abstract
INTRODUCTION
Intrahepatic cholangiocarcinoma (ICC) is the second most common primary liver malignancy, accounting for approximately 20% of all primary hepatic malignancies [
1,
2]. The global incidence and mortality rate of ICC continue to rise, with a 5-year overall survival (OS) of less than 10% [
3,
4]. Although surgical resection is the most promising potentially curative treatment for ICC, the postoperative recurrence rate remains alarmingly high, with approximately 50% to 70% of patients experiencing recurrence [
5]. These limitations underscore an urgent need for more effective therapeutic strategies and reliable biomarkers to improve clinical outcomes in ICC.
Tumorigenesis is a multistep process involving a series of genetic and epigenetic variations. Increasing evidence indicates that somatic genomic mutations acquired during tumor development drive malignant transformation and facilitate immune evasion by tumor cells [
6]. With advances in next-generation sequencing, the genomic landscape of ICC has been partially elucidated. Frequently mutated genes include
TP53,
ARID1A,
KRAS, and
IDH1 [
7–
10], which are involved in critical signaling pathways and have significant effects on epigenetic regulation and oxidative phosphorylation. However, the impact of genomic mutations on the tumor microenvironment and immune evasion in recurrent ICC remains largely unexplored.
Our previous whole-exome sequencing (WES) study revealed the genomic landscape and identified several frequently mutated genes in ICC [
11]. In this study, we aimed to assess the variational status of large tumor suppressor kinase 2 (
LATS2) in ICC and examine its association with disease recurrence and patient survival. Additionally, we investigated the functional impact and underlying mechanisms of
LATS2 variation on tumor progression, immune invasion, and resistance to anti-programmed death-1 (anti-PD-1) therapy.
MATERIALS AND METHODS
Detailed information about the patients and follow-up, Sanger sequencing, structural analysis, cell lines, lentiviral vector, reagents, RNA isolation and quantitative real-time polymerase chain reaction (qRT-PCR), protein extraction and Western blot, cell proliferation, colony formation, and transwell assays, luciferase reporter assay, cell immunofluorescence assay, humanized hematopoietic stem cell–engrafted mouse, mice and animal studies, single-cell RNA sequencing (scRNA-seq) and data analysis, co-cultures of ICC cells and T cells system, flow cytometry, tissue micro-array and immunohistochemistry, cytokine/chemokine analysis, enzyme-linked-immunosorbent assay and statistical analysis are available in the
Supplementary Materials and Methods.
RESULTS
Somatic mutation, copy number loss, or low expression of LATS2 are associated with postoperative recurrence and poor survival in ICC
Our previous WES study revealed the genomic landscape and identified several significantly mutated genes in 204 ICC patients. We focused on
LATS2 and performed targeted sequencing and Sanger sequencing on samples from another independent cohort of 196 ICC patients. Overall, in the total of 400 ICC samples, we identified 12 cases with
LATS2 somatic mutation (3%, 12/400) and 136 cases with
LATS2 copy number loss (34%, 136/400;
Fig. 1A;
Supplementary Figs. 1 and
2;
Supplementary Table 1).
The full-length structures of LATS2 are currently unavailable, so we obtained predicted full-length structures of LATS2-HUMAN from the AlphaFold predicted structure database (
https://alphafold.ebi.ac.uk/) for structural presentation and analysis (
Fig. 1B, 1C). Analysis of the effects of
LATS2 mutations on protein structure showed that these identified mutations impair LATS2-mediated signal transduction, highlighting the critical roles of these residues and structural elements in the function of LATS2 within the Hippo signaling pathway (Supplementary Result).
We evaluated LATS2 expression by immunohistochemistry in all 400 ICC cases. The results showed that overall, LATS2 expression was significantly downregulated in tumor tissues compared with adjacent non-tumorous liver tissues (
Fig. 1D). Notably, the patients with
LATS2 somatic mutation or copy number loss exhibited a further reduction in tumor LATS2 expression compared with the patients without such
LATS2 variation (
Fig. 1D). Kaplan–Meier survival analysis showed that patients with somatic mutation, copy number loss, or lower expression of
LATS2 had markedly decreased recurrence-free survival (RFS) and OS compared with patients without those features (
Fig. 1E). Multivariate Cox regression analysis confirmed that
LATS2 somatic mutation, copy number loss, or low expression served as independent prognostic factors for postoperative recurrence and poor outcomes in ICC (
Supplementary Table 2). Collectively, these findings indicate that
LATS2 variation and low expression are clinically associated with ICC progression and recurrence.
Our findings suggest that
LATS2 may act as a tumor suppressor in ICC. To validate this hypothesis, we analyzed the genotype and expression of LATS2 in normal biliary epithelial cells and five ICC cell lines. Using the same filter criteria applied in the next-generation sequencing of ICC samples, Sanger sequencing and quantitative PCR (qPCR) confirmed that all five ICC cell lines harbored wild-type (WT)
LATS2, but HuCCT1 cells exhibited
LATS2 copy number loss. Western blot results revealed that LATS2 expression was highest in CCLP1 cells and lowest in HuCCT1 cells (
Fig. 2A).
Next, we performed LATS2 knockdown in CCLP1 cells (
Fig. 2A). Biofunctional investigations demonstrated that the LATS2 knockdown significantly enhanced ICC cell proliferation, colony formation, and invasive capacity (
Fig. 2B–2D). Consistent with those findings,
in vivo experiments revealed that LATS2 knockdown promoted tumor growth and pulmonary metastasis (
Fig. 2E, 2F).
To functionally characterize
LATS2 somatic mutations, we generated lentiviral constructs to re-express LATS2 in HuCCT1 cells (
Fig. 2A). The results showed that re-expression of WT LATS2 markedly suppressed cell proliferation, clonogenicity, and invasion ability. Conversely, re-expression of mutated LATS2 failed to recapitulate these effects fully or partially, indicating that the
LATS2 mutation was functionally inactivating (
Fig. 2B–2D). Subsequently, we used the cells with re-expression of WT or mutant LATS2 for
in vivo studies. Consistent with the
in vitro findings, re-expression of WT LATS2 markedly suppressed tumor growth and pulmonary metastasis, whereas re-expression of mutated LATS2 produced larger tumor volumes and increased metastatic potential compared with WT LATS2 (
Fig. 2E, 2F). These results support the notion that
LATS2 is a tumor suppressor gene in ICC and that certain somatic mutations abolish its function and its tumor-inhibitory effect.
Previous studies established that LATS2 acts as a critical activator of the Hippo kinase cascade in mammals [
12]. To investigate the regulatory relationship between LATS2 and Hippo signaling in ICC, we performed Western blot analysis. In CCLP1 cells, LATS2 knockdown led to increases in total YAP levels and nuclear YAP expression, while cytoplasmic YAP levels remained unchanged, with a concomitant decrease in phospho-YAP (Ser127). Conversely, re-expression of WT LATS2 in HuCCT1 cells resulted in reduced total and nuclear YAP levels and increased phospho-YAP (Ser127), without affecting the cytoplasmic distribution of YAP (
Fig. 3A). Immunofluorescence results confirmed that LATS2 knockdown increased nuclear accumulation of YAP in CCLP1 cells, whereas re-expression of WT LATS2 reduced nuclear YAP localization in HuCCT1 cells (
Fig. 3B). These results suggest that
LATS2 inactivation regulates the Hippo kinase cascade, thereby promoting nuclear translocation and subsequent activation of YAP. We also demonstrated that mutated LATS2 failed to regulate this kinase cascade, in contrast to WT LATS2. This suggests that
LATS2 inactivation leads to suppression of the Hippo signaling pathway and aberrant activation of YAP in ICC cells (
Fig. 3A, 3B). Consistent with these findings, immunohistochemical analysis of ICC patient samples showed increased nuclear YAP accumulation in tumors with
LATS2 somatic mutation or copy number loss (
Fig. 3C), further supporting the hypothesis that
LATS2 inactivation suppresses the Hippo pathway and contributes to YAP activation in ICC.
To further explore this mechanism, we next assessed whether the oncogenic effects of
LATS2 inactivation are mediated by YAP activity. In CCLP1 cells, YAP knockdown or verteporfin treatment (a YAP inhibitor) attenuated LATS2 knockdown–induced increases in
in vitro cell proliferation, colony formation, and invasion (
Fig. 3D, 3E,
Supplementary Fig. 3A) and suppressed the corresponding promotion of
in vivo tumor growth and metastasis (
Fig. 3F). In HuCCT1 cells, overexpression of S112A mutant YAP (a nonphosphorylatable YAP mutant) reversed the inhibitory effects of LATS2 on both
in vitro and
in vivo tumor phenotypes (
Fig. 3D–3F). Collectively, these findings demonstrate that
LATS2 deficiency promotes ICC progression through YAP activation.
To investigate the effects of LATS2 on the expression of inflammatory and chemotactic factors in human ICC cells, we used a Human Chemokine Panel to quantify the concentrations of chemokines in HuCCT1 cells. The results revealed that re-expression of WT LATS2 suppressed the secretion of granulocyte-macrophage colony-stimulating factor (GM-CSF) and C-C motif chemokine ligand 2 (CCL2;
Fig. 4A). An enzyme-linked immunosorbent assay (ELISA) was performed to validate these findings and confirmed a significant decrease in CCL2 expression after WT LATS2 re-expression. qRT-PCR and ELISA results also showed that compared with WT LATS2, the seven types of mutated LATS2 failed to effectively suppress CCL2 expression and secretion after their re-expression in HuCCT1 cells (
Fig. 4B).
To assess whether
LATS2 inactivation affects the expression of immune checkpoint molecules, we performed qRT-PCR and Western blot analyses. The results showed that WT LATS2 downregulates programmed cell death-ligand 1 (PD-L1) expression, whereas the LATS2 mutants fully or partially lost this function. By contrast, the expression of other immune checkpoints, including PD-L2, TIGIT, TIM-3, and CTLA-4, was not significantly affected by LATS2 (data not shown). Notably, the PD-L1 expression levels correlated with YAP activation status (
Fig. 4C). Immunofluorescence analysis further confirmed that WT LATS2 reduced both PD-L1 expression and nuclear YAP accumulation, whereas these effects were not observed in cells expressing mutated LATS2 (
Fig. 4D, 4E).
To determine whether these regulatory effects were dependent on the Hippo/YAP signaling pathway, we employed the YAP inhibitor verteporfin. In CCLP1 cells, qRT-PCR and Western blot analyses demonstrated that verteporfin effectively suppressed PD-L1 expression and partially reversed PD-L1 upregulation induced by LATS2 knockdown (
Fig. 4F). Similarly, qRT-PCR and ELISA results confirmed that verteporfin suppressed CCL2 production and attenuated the increase in CCL2 secretion caused by LATS2 deficiency (
Fig. 4F).
Given that
LATS2 inactivation promotes YAP nuclear translocation and upregulates the expression of CCL2 and PD-L1, we investigated whether nuclear YAP directly drives the transcription of these genes. To this end, we ectopically expressed YAP wild-type (YAP-WT) and a constitutively active mutant, YAP-5SA, which exhibits enhanced nuclear localization capacity [
13]. In parallel, we performed YAP-TEAD–dependent luciferase reporter assays, which confirmed increased YAP transcriptional activity upon overexpression of YAP-WT or YAP-5SA (
Supplementary Fig. 4A). We then examined whether YAP regulates the transcription of CCL2 and PD-L1 by constructing dual-luciferase reporter plasmids containing the promoter regions of
PD-L1 and
CCL2. Reporter assays revealed that YAP-WT significantly enhanced the transcriptional activity of both promoters, with YAP-5SA exerting an even more pronounced effect. (
Supplementary Fig. 4B, 4C). However, this activation was abolished when promoter-binding sites were mutated, indicating that YAP-mediated transcriptional activation requires its nuclear localization and direct interaction with the target gene promoters (
Supplementary Fig. 4B, 4C).
To investigate the impact of LATS2 on the immune microenvironment of ICC, we established subcutaneous tumors in humanized hematopoietic stem cell-engrafted (hu-HSG) (CD34
+) mice, which feature a reconstituted immune system. As expected, WT LATS2 significantly attenuated tumor growth, and this effect was attenuated in tumors with
LATS2 mutation (
Fig. 5A). To further elucidate the immunoregulatory role of LATS2, we harvested subcutaneous tumors from three groups of hu-HSG (CD34
+) mice, dissociated the tumor tissues into single-cell suspensions, and performed scRNA-seq using the 10× Genomics platform. After quality control and filtering, a total of 53,994 cells were retained for analysis. Graph-based clustering identified five major cell types based on the expression of marker genes, including the majority of known immune cells (T and natural killer cells [T&NK], myeloid cells, endothelial cells, malignant cells, and fibroblasts;
Fig. 5B).
Subclustering of the T&NK population revealed nine distinct subtypes. Notably, the CD8
+ T cell subsets—including C1_CD8_XCL1 and C3_CD8_GZMH—were significantly enriched in tumors expressing WT LATS2 compared with control tumors, whereas their abundance was markedly reduced in tumors expressing V669fs LATS2 (
Fig. 5B). Similarly, subclustering of the myeloid cells identified 11 subpopulations. The tumors expressing WT LATS2 exhibited a reduced proportion of M2-like macrophages (M6_Macro_CD206) and an increased proportion of pro-inflammatory macrophages (M7_Macro_IRF5) compared with control tumors. Conversely, the V669fs mutation reversed this trend, leading to increased CD206
+ macrophages and reduced IRF5
+macrophages (
Fig. 5B).
To validate the scRNA-seq findings, we performed immunohistochemistry on tumors derived from five groups of humanized mice. The results confirmed that LATS2 suppressed PD-L1 and CCL2 expression, inhibited M2 macrophage infiltration, and promoted CD8
+ T cell recruitment, whereas mutated LATS2 failed to exhibit these effects (
Fig. 5C).
To determine whether these immune alterations are mediated via the Hippo/YAP pathway, we treated tumor-bearing hu-HSG (CD34
+) mice with the YAP inhibitor verteporfin. Verteporfin significantly suppressed tumor growth and partially reversed the tumor-promoting effects of LATS2 knockdown (
Fig. 5D). qRT-PCR analysis of tumor tissues revealed that verteporfin suppressed PD-L1 and CCL2 expression and attenuated the upregulation of these proteins induced by LATS2 deficiency (
Fig. 5E). Flow cytometric analysis further revealed that verteporfin promoted CD8
+ T cell infiltration and suppressed CD206
+ macrophage accumulation, partially restoring the immune landscape disruption caused by LATS2 knockdown (
Fig. 5F).
The preceding results of this study confirmed that
LATS2 inactivation reshapes the tumor immune microenvironment, characterized by reduced infiltration of CD8
+ T cells and increased CD206
+ macrophage accumulation—two immunological changes known to contribute to resistance against anti-PD-1 therapy. To assess whether
LATS2 inactivation affects the response to immunotherapy, we established subcutaneous tumors in hu-HSG (CD34
+) mice using HuCCT1 cells expressing either WT or mutated LATS2. Anti-PD-1 monoclonal antibody (mAb) treatment significantly reduced the volume of tumors expressing WT LATS2, whereas it had no significant effect on tumors expressing mutated LATS2 (
Fig. 6A), indicating that
LATS2 inactivation may confer resistance to PD-1 blockade.
We hypothesized that YAP inhibition might overcome resistance to anti-PD-1 therapy in the humanized mouse model with inactivated
LATS2. To test this, we evaluated the therapeutic efficacy of anti-PD-1 antibodies, verteporfin, or their combination in hu-HSG (CD34
+) mice (
Fig. 6B). The results showed that anti-PD-1 treatment had no significant effect on tumor volume in the control group (
P>0.05), but it significantly reduced the tumor size in the group with tumors expressing WT LATS2 (
P<0.001). Notably, the combination of verteporfin and anti-PD-1 led to significantly smaller tumors than anti-PD-1 alone (
P<0.001;
Fig. 6C). These findings suggest that
LATS2 inactivation confers resistance to anti-PD-1 therapy in the hu-HSG (CD34
+) mouse ICC model, while verteporfin enhances the therapeutic efficacy of PD-1 blockade.
To investigate the immune mechanisms underlying this response, we performed multiparameter flow cytometry on dissociated tumor tissues from each treatment group. In both the WT LATS2 group and the WT LATS2+anti-PD-1 group, we observed an increased proportion of CD8
+ T cells and a decrease in CD206
+ macrophages among CD45
+ immune cells compared with the control group. Similarly, compared with the control group, the verteporfin and verteporfin+anti-PD-1 groups also showed enhanced CD8
+ T cell infiltration and reduced CD206
+ macrophage abundance (
Fig. 6D).
Building on these findings, we performed co-culture assays of ICC cells and CD8
+ T cells to assess whether
LATS2 inactivation-mediated PD-L1 upregulation impairs CD8
+ T cell proliferation and effector function. In CCLP1 cells, LATS2 knockdown significantly suppressed CD8
+ T cell proliferation and effector function, effects that were partially rescued by verteporfin treatment. Similarly, in HuCCT1 cells, co-culture with CD8
+ T cells resulted in markedly impaired CD8
+ T cell proliferation and effector function, which were partially reversed by WT-LATS2 or verteporfin treatment. To determine whether PD-L1 contributed to the observed suppression of CD8
+ T cell function, a neutralizing anti-PD-L1 antibody was added to the co-culture system of ICC cells and peripheral CD8
+ T cells. Notably, PD-L1 blockade significantly restored CD8
+ T cell proliferation and effector function, thereby supporting the functional role of PD-L1 in
LATS2 inactivation–mediated immune evasion (
Fig. 6E).
Hence, these results demonstrate that
LATS2 inactivation suppresses CD8
+ T cell infiltration, promotes M2 macrophage recruitment, and confers resistance to PD-1 blockade in ICC. Conversely, pharmacological inhibition of YAP with verteporfin restores CD8
+ T cell infiltration, limits M2 macrophage recruitment, and sensitizes LATS2-deficient tumors to anti-PD-1 immunotherapy (
Fig. 6F).
Our results have shown that
LATS2 inactivation promotes an immunosuppressive tumor microenvironment and resistance to anti-PD-1 therapy in ICC models. To further validate these findings, immunohistochemical staining for LATS2, YAP, PD-L1, CCL2, CD8, and CD206 was performed on ICC patient tissue samples (cohort 1, n=400) (
Fig. 7A). Correlation analysis revealed that nuclear accumulation of YAP, expression levels of PD-L1 and CCL2, and infiltration of CD206
+ macrophages were negatively correlated with LATS2 expression, whereas infiltration of CD8
+ T cells showed a positive correlation with LATS2 expression (
Fig. 7B).
We also collected pre-treatment tumor biopsy specimens from 45 patients treated with anti-PD-1 treatment or anti-PD-1 combined with chemotherapy (cohort 2, n=45,
Supplementary Table 3). Of these, 29 patients demonstrated a clinical response and 16 were classified as non-responders (
Fig. 7C, 7D). Notably,
LATS2 inactivating variations were present in 11 of 16 non-responders (68.7%; 2 with a somatic mutation and 9 with copy number loss), whereas only 7 copy number losses were detected in 7 of 29 responders (24.1%,
P=0.003,
Fig. 7C, 7D), which further supports an association between
LATS2-inactivating variation and treatment resistance to anti-PD-1 therapy in ICC. Univariable analysis revealed that lymph node metastasis, LATS2 copy number loss, and low expression were associated with poor response. Multivariable analysis identified LATS2 copy number loss (OR=0.208, 95% CI: 0.051–0.848,
P=0.039,
Supplementary Table 4) and low expression (OR=0.169, 95% CI: 0.041–0.701,
P=0.014) as independent factors, supporting that
LATS2 inactivation may serve as a potential biomarker of anti-PD-1 response in ICC.
DISCUSSION
In this study, we identified 3% (12/400) somatic mutations and 34% (136/400) copy number loss of LATS2 in ICC patients. Furthermore, we found that patients with LATS2 somatic mutation, copy number loss, or low expression were associated with increased risk of tumor recurrence and shortened OS, suggesting the potential of LATS2 variation as a prognostic biomarker for ICC patients.
LATS2 is a member of the AGC subfamily of serine/threonine kinases and contains two conserved domains (LCD1 and LCD2) and a UBA domain. The conserved domains are essential for the regulatory activity and function of LATS2. Previous studies showed that LATS2 downregulation promotes cancer cell proliferation, resistance to apoptosis, and enhanced migration and invasion in breast cancer, colorectal cancer, and ovarian tumors, suggesting a tumor-suppressive role [
14–
16]. Although
LATS2 mutations have been reported in certain solid tumors, their overall frequency remains low, with rates of 0.84% and 1.50% in the COSMIC and cBioPortal databases, respectively [
17]. Furthermore, the functional significance of these mutations in tumorigenesis and their underlying molecular mechanisms have not been fully elucidated. In this study, we identified
LATS2 somatic mutations and copy number loss in an ICC cohort, which were associated with reduced LATS2 expression. Through gain-of-function and loss-of-function experiments, we demonstrated that
LATS2 functions as a tumor suppressor in ICC. We also confirmed that
LATS2 somatic mutations result in functional inactivation of the LATS2 protein, thereby promoting ICC progression and metastasis.
The Hippo pathway is an evolutionarily conserved signaling cascade that plays a fundamental role in maintaining cellular homeostasis and regulating cell proliferation and migration [
18]. LATS2 is a core kinase component of the mammalian Hippo pathway. Our findings confirm that LATS2 regulates Hippo signaling in ICC, and its inactivation suppresses the Hippo pathway and consequently activates YAP. Both
in vitro and
in vivo results confirmed that
LATS2 inactivation facilitates ICC growth and metastasis through YAP activation. Although several studies demonstrated that Hippo pathway inactivation was associated with poor prognosis in hepatocellular carcinoma [
19], the function of the Hippo pathway in ICC remains poorly understood. We found that
LATS2 inactivation led to Hippo pathway inactivation, promoting ICC progression. Moreover, we found that the YAP inhibitor verteporfin effectively suppressed tumor progression in ICC with
LATS2-inactivating variation. Verteporfin was reported to safely and effectively induce tumor necrosis in patients with locally advanced pancreatic cancer [
20], suggesting that it would be relatively easy to adopt for ICC treatment. These findings provide a promising therapeutic strategy for ICC patients with
LATS2 inactivation. To our knowledge, this is the first report showing that
LATS2-inactivating variation in ICC suppresses the Hippo pathway, thereby facilitating tumor growth and metastasis.
In human ICC, deregulation of the Hippo signaling pathway represents a major genetic event that promotes tumor immune evasion and resistance to anti-PD-1 therapy [
21]. Given that
LATS2 inactivation suppresses Hippo signaling pathway, we explored the impacts of
LATS2-inactivating variation on the tumor immune microenvironment in ICC. Prior studies reported that LATS2 suppression could enhance YAP transcriptional activity and promote M2 macrophage infiltration in colon cancer, thereby facilitating the formation of an immunosuppressive microenvironment and promoting tumor metastasis [
22]. Importantly, our study further revealed that
LATS2-inactivating variation led to YAP activation, which upregulates PD-L1 expression and CCL2 secretion, suppresses CD8
+ T cell infiltration, and enhances recruitment of M2-like macrophages. Furthermore, in a humanized mouse model with
LATS2 inactivation, we observed poor response to anti-PD-1 treatment. We also observed a similar result in ICC patients, suggesting that those patients with
LATS2 inactivation may derive limited benefit from anti-PD-1 therapy.
Identifying the relationship between host genetic features and antitumor immune responses may provide new insights into patient stratification for immunotherapy [
11,
23]. Increasing evidence indicates that mutation-driven activation of the Hippo signaling pathway facilitates immune evasion, and this pathway can be targeted in affected patients to enhance immunotherapeutic efficacy [
14,
24]. Several YAP inhibitors have demonstrated potent antitumor activity and the ability to reverse immune resistance in preclinical models, providing a viable therapeutic avenue for Hippo pathway-targeted immunomodulation [
25,
26]. Although verteporfin is commonly used as a YAP inhibitor, its potential off-target effects may limit its specificity. To address this, we performed genetic gain- and loss-of-function experiments (shYAP and YAP
S112A), which independently confirmed that YAP mediates the oncogenic effects of
LATS2 inactivation in ICC. In this study, we demonstrated that treatment with a YAP inhibitor was able to reverse resistance to anti-PD-1 therapy in ICC tumors harboring
LATS2 inactivation in a humanized CD34
+ hematopoietic stem cell (HSC) mouse model, which suggests a potential and feasible therapeutic strategy for precision intervention in ICC patients carrying
LATS2 mutations or other mutations that suppress the Hippo signaling pathway.
In conclusion, this study identified LATS2 genomic variation and demonstrated that LATS2 inactivation plays a pivotal clinical and mechanistic role in ICC progression and resistance to anti-PD-1 therapy. LATS2 variation may therefore serve as a useful biomarker for identifying patients who are unlikely to benefit from PD-1 blockade, thereby informing individualized therapeutic strategies.
FOOTNOTES
-
Authors’ contribution
YX, KXL, XYW, and SYC performed the experiments; YHY, NL, FW, and LC analyzed the data; CYT, ZHT, ZQH, ZJZ, and CBL provided the samples; SLZ and XY wrote the paper; JF and JZ provided critical revision; JZ and SLZ obtained funding and designed the research. All authors read and approved the final manuscript.
-
Acknowledgements
This study was supported by the National Natural Science Foundation of China (Nos. 82372985, 82373418, 82173260, and 82203561), the Shanghai Shuguang Program (No. 23SG09), the Shanghai Medical Innovation Research Project (No. 22Y11907300), and the Shanghai Science and Technology Innovation Action Plan Project in Computational Biology (No. 24JS2840100).
-
Conflicts of Interest
The authors have no conflicts to disclose.
SUPPLEMENTARY MATERIAL
Supplementary material is available at Clinical and Molecular Hepatology website (
http://www.e-cmh.org).
Supplementary Figure 1.
Sanger sequencing trace files of LATS2 variation in ICC samples. LATS2 variation detected by Sanger sequencing in 400 patients with ICC. The somatic nature of these mutations was identified by comparisons of the sequenced regions between peritumor tissues and primary tumors. The mutant sites are highlighted by arrows. ICC, intrahepatic cholangiocarcinoma.
cmh-2025-0997-Supplementary-Fig-1.pdf
Supplementary Figure 3.
Western blot and qRT-PCR analysis showed the expression of LATS2 and YAP in stably transfected ICC cells. Student’s t-test, ***P<0.001, n=3, errors are in ±SD. qRT-PCR, quantitative real time polymerase chain reaction; ICC, intrahepatic cholangiocarcinoma; SD, standard deviation.
cmh-2025-0997-Supplementary-Fig-3.pdf
Supplementary Figure 4.
YAP activates TEAD reporter and enhances PD-L1 and CCL2 promoter activity in ICC cells. (A) Firefly luciferase reporter plasmid driven by a TEAD-binding element (8×GTIIC-luciferase) was co-transfected into ICC cells together with either the empty vector, YAP-WT, or constitutively active YAP mutant (YAP-5SA) overexpression plasmids. Luciferase activity was measured 48 hours post-transfection. Student’s t-test, two-sided. **P<0.01, ***P<0.001, n=3, errors are in ±SD. (B) Firefly luciferase reporter plasmids containing either the wild-type or mutant PD-L1 promoter were co-transfected into ICC cells along with either the empty vector, YAP WT, or YAP-5SA overexpression plasmids. Luciferase activity was measured 48 hours post-transfection. Student’s t-test, two-sided. **P<0.01, ***P<0.001, n=3, errors are in ±SD. (C) Firefly luciferase reporter plasmids containing either the wild-type or mutant CCL2 promoter were co-transfected into ICC cells along with either the empty vector, YAP WT, or YAP-5SA overexpression plasmids. Luciferase activity was measured 48 hours post-transfection. Student’s t-test, two-sided. **P<0.01, ***P<0.001, n=3, errors are in ±SD. PD-L1, programmed cell death-ligand 1; ICC, intrahepatic cholangiocarcinoma; YAP-WT, YAP wild-type; SD, standard deviation; CCL2, chemokine ligand 2.
cmh-2025-0997-Supplementary-Fig-4.pdf
Figure 1Clinical significance of LATS2 variation in ICC (n=400). (A) Schematic diagram of the locations of the LATS2 mutations identified in this study. (B) The upper panel presents a structural model of the LATS2 protein. Structural consequences of selected mutations are shown as red spheres within domains. The dashed line indicates a disordered region (residues 160–590). The bottom panel shows the interaction of the MBD of LATS2 with MOB1. The MOB1 protein is depicted in pink, interacting with the MBD of LATS2, which is shown in green. The kinase domain of LATS2 is shown in cyan, with the active site highlighted. The NDR kinase is shown in light blue, interacting with the MBD of LATS2. Key mutations within the kinase domain, such as A723E, M753R, I799V, E905V, and R1054Q, are marked as red spheres. (C) Details of the effects of mutations on the structural conformation. All the C-alpha atoms of the mutant residues are shown as red spheres. Interacting residues associated with mutation sites are shown as sticks. The zoomed-in views of each mutation site of LATS2 are shown in the red box, with the wild-type conformation at the top and the mutant conformation at the bottom. (D) Representative LATS2 immunostaining in peritumoral tissues and tumors with LATS2 variation; scale bars=100 μm. Quantification of LATS2 staining density across different groups. ***P<0.001, Student’s t-test, two-sided. (E) Kaplan–Meier survival analysis of recurrence-free survival and overall survival based on LATS2 somatic mutation, copy number loss, or expression level. P: log-rank test, two-sided. ICC, intrahepatic cholangiocarcinoma; HR, hazard ratio; CNV, copy number variation.
Figure 2Identification of LATS2 as a tumor suppressor gene in ICC. (A) LATS2 copy number revealed by qPCR in one human immortalized nonmalignant cholangiocyte cell line and five ICC cell lines, and LATS2 expression examined by Western blot in parental and stably transfected cells. (B) Proliferation of CCLP1 cells after LATS2 knockdown and HuCCT1 cells expressing wild-type or mutant LATS2, compared with controls, n=5, errors are in ±SD. (C) Colony formation activity of CCLP1 cells after LATS2 knockdown and HuCCT1 cells expressing wild-type or mutant LATS2, compared with controls. The bar graphs illustrate quantification of the colony formation assays. Student’s t-test, two-sided, ***P<0.001, n=6, errors are in ±SD. (D) Invasion of CCLP1 cells after LATS2 knockdown and HuCCT1 cells expressing WT or mutant LATS2, compared with controls. The graphs depict the number of invasive cells after 48 hours. Student’s t-test, two-sided, ***P<0.001, n=6, errors are in ±SD. Scale bars=100 μm. (E, F) Representative bioluminescence images of mouse liver tumors and pulmonary metastasis, and H&E-stained images of metastatic nodules in lungs. The color scale bar depicts the photon flux emitted from the mice. Student’s t-test, two-sided, **P<0.01, ***P<0.001, n=6, errors are in ±SD, scale bars=100 μm. ICC, intrahepatic cholangiocarcinoma; qPCR, quantitative polymerase chain reaction; SD, standard deviation; WT, wild-type.
Figure 3
LATS2 inactivation suppresses Hippo signaling, leading to YAP activation. (A) Western blot to validate YAP and p-YAP expression following LATS2 knockdown or re-expression of wild-type or mutated LATS2 in ICC cells. Western blot analysis of YAP subcellular distribution in ICC cells following LATS2 knockdown or re-expression of wild-type LATS2, compared with that in controls. (B) Immunofluorescence staining showing subcellular YAP localization in the indicated cells. The graphs depict the percentages of cells with exclusively nuclear (N) or both nuclear and cytoplasmic (N+C) YAP. No cells showed YAP localized exclusively in the cytoplasm. n=3, errors are in ±SD, scale bars=20 μm. (C) Representative LATS2 staining in tumor samples from patients with ICC. The color scale bar depicts the LATS2 staining density. Student’s t-test, two-sided, ***P<0.001. Scale bars=50 μm. (D) Proliferation of YAP knockdown or verteporfin-treated CCLP1 cells with LATS2 knockdown and HuCCT1 cells expressing wild-type YAP or S112A YAP with LATS2 overexpression, compared with controls. Student’s t-test, two-sided, ***P<0.001, n=5, errors are in ±SD. (E) Colony formation and invasion activity of YAP knockdown or verteporfin-treated CCLP1 cells with LATS2 knockdown and HuCCT1 cells expressing wild-type YAP or S112A YAP with LATS2 overexpression, compared with controls. Student’s t-test, two-sided, ***P<0.001, n=3, errors are in ±SD. (F) Representative bioluminescence images of mouse liver tumors and pulmonary metastasis, and H&E-stained images of metastatic nodules in lungs. The color-scale bar depicts the photon flux emitted from the mice. Student’s t-test, two-sided, ***P<0.001, n=6, errors are in ±SD. Scale bar=100 μm. ICC, intrahepatic cholangiocarcinoma; SD, standard deviation.
Figure 4
LATS2 inactivation promotes PD-L1 expression and CCL2 secretion through the Hippo/YAP pathway. (A) Cytokine profiling using the LX-MultiDTH-40 Human Chemokine Panel and ELISA to evaluate the effects of wild-type and mutant LATS2 on secretion of inflammatory and chemotactic factors in HuCCT1 cells. Student’s t-test, two-sided, *P<0.05, **P<0.01, ***P<0.001, n=3, errors are in ±SD. (B) qRT-PCR and ELISA analyses of CCL2 mRNA and protein levels in HuCCT1 cells expressing wild-type or mutant LATS2. Student’s t-test, two-sided, ***P<0.001, n=3, errors are in ±SD. (C) qRT-PCR and Western blot analyses of PD-L1 expression in HuCCT1 cells expressing wild-type or mutant LATS2. Student’s t-test, two-sided, ***P<0.001, n=3, errors are in ±SD. (D) Immunofluorescence analysis of YAP and PD-L1 expression and subcellular localization in HuCCT1 cells expressing wild-type or mutant LATS2. Scale bars=20 μm. (E) The graphs depict the percentages of cells with exclusively nuclear (N) or both nuclear and cytoplasmic (N+C) YAP localization. (F) qRT-PCR and Western blot analyses of PD-L1 expression in LATS2-knockdown CCLP1 cells following treatment with the YAP inhibitor verteporfin. Student’s t-test, two-sided, ***P<0.001, n=3, errors are in ±SD. qRT-PCR and ELISA analyses of CCL2 expression and secretion in LATS2-knockdown CCLP1 cells treated with verteporfin. Student’s t-test, two-sided, ***P<0.001, n=3, errors are in ±SD. PD-L1, programmed cell death-ligand 1; CCL2, chemokine ligand 2; ELISA, enzyme-linked immunosorbent assay; SD, standard deviation; qRT-PCR, quantitative real-time polymerase chain reaction.
Figure 5
LATS2 inactivation confers immunosuppression and resistance to anti-PD-1 therapy. (A) Subcutaneous tumor growth in hu-HSG (CD34+) mice injected with CCLP1 cells with LATS2 knockdown or HuCCT1 cells expressing wild-type or mutant LATS2, compared with control groups. Tumor volumes are shown on the right; Student’s t-test, two-sided, ***P<0.001, n=6, errors are in ±SD. (B) Upper panel: a t-SNE plot showing clusters from three hu-HSG (CD34+) mice with subcutaneous tumors identified by 10× scRNA-seq. Left panel: a t-SNE plot displaying nine clusters of T and natural killer T (T&NKT) cells, assessed using 10× scRNA-seq. The right bar graph shows the fraction of each cluster. Right panel: a t-SNE plot showing 11 clusters of myeloid cells, assessed using 10× scRNA-seq. The right bar graph shows the fraction of each cluster. (C) Representative H&E and IHC staining of LATS2, YAP, PD-L1, CCL2, CD8, CD68, and CD206 from subcutaneous tumors of humanized mice. Scale bar for H&E=100 μm; scale bars for IHC=20 μm. Quantification of IHC staining scores (LATS2, YAP, PD-L1, CCL2) and immune cell counts (CD8+ T cells, CD68+ macrophages, CD206+ macrophages) in subcutaneous tumors from humanized mice. IHC scores were calculated based on staining intensity and percentage of positive cells. n=6, errors are in ±SD. (D) Control or LATS2-knockdown CCLP1 cells were subcutaneously injected into hu-HSG (CD34+) mice, followed by verteporfin administration. Tumor volumes are shown on the right. Student’s t-test, two-sided, ***P<0.001, n=6, errors are in ±SD. (E) qRT-PCR analysis of PD-L1 and CCL2 expression in subcutaneous tumors from hu-HSG (CD34+) mice that were injected with control or LATS2-knockdown CCLP1 cells and treated with verteporfin; Student’s t-test, two-sided, ***P<0.001, n=6, errors are in ±SD. (F) Flow cytometry analysis of the percentages of CD8+ T cells and CD206+ cells in subcutaneous tumors from hu-HSG (CD34+) mice that were injected with control or LATS2-knockdown CCLP1 cells and treated with verteporfin; Student’s t-test, two-sided, ***P<0.001, n=6, errors are in ±SD. PD-1, programmed death-1; SD, standard deviation; t-SNE, t-distributed stochastic neighbor embedding; hu-HSG, humanized hematopoietic stem cell-engrafted; scRNA-seq, single-cell RNA sequencing; IHC, immunohistochemistry.
Figure 6Targeting of YAP can potentiate PD-1 blockade therapy in ICC tumors with LATS2 inactivation. (A) HuCCT1 cells expressing wild-type or mutant LATS2 were subcutaneously injected into hu-HSG (CD34+) mice, followed by administration of antihuman IgG or anti-PD-1 antibodies. Tumor volumes are shown on the right. Student’s t-test, two-sided, ***P<0.001, n=6, errors are in ±SD. (B) Diagram summarizing the treatment strategy for hu-HSG (CD34+) mice (left). HuCCT1 control cells or HuCCT1 cells expressing wild-type LATS2 were subcutaneously injected into hu-HSG (CD34+) mice. Fifteen days after inoculation, the control group was treated with anti-PD-1 antibodies alone, verteporfin alone, or a combination of both, while the mice injected with wild-type LATS2-expressing cells received anti-PD-1 antibody treatment only. Tumors were collected on day 30 after inoculation. (C) Tumor volumes were measured every 3 days. Tumors were harvested on day 30 after inoculation. two-way ANOVA, ***P<0.001, n=6, errors are in ±SD. (D) Flow cytometry analysis of the percentages of CD8+ T cells and CD206+ macrophages in tumors from hu-HSG (CD34+) mice. Student’s t-test, two-sided, ***P<0.001, n=6, errors are in ±SD. (E) CFSE-labeled peripheral CD8+ T cells were co-cultured with ICC cells under different LATS2 perturbation conditions (LATS2 knockdown in CCLP1 cells or LATS2 overexpression in HuCCT1 cells), in the presence of anti-CD3/CD28 stimulation, with or without verteporfin and/or PD-L1 neutralizing antibody. Representative flow cytometry plots and quantification of CD8+ T cell proliferation and IFN-γ production are shown. n=3, errors are ±SD. (F) Schematic illustration showing that LATS2-inactivating variation drives tumor progression and resistance to anti-PD-1 therapy in ICC. PD-1, programmed death-1; ICC, intrahepatic cholangiocarcinoma; hu-HSG, humanized hematopoietic stem cell-engrafted; SD, standard deviation.
Figure 7
LATS2 inactivation correlates with an immunosuppressive microenvironment and resistance to anti-PD-1 therapy in ICC patients. (A) Representative immunohistochemical staining illustrating the expression of LATS2, YAP, PD-L1, CCL2, CD8, and CD206 in ICC patient samples characterized by WT LATS2, LATS2 somatic mutation, or LATS2 copy number loss. Scale bar=100 μm. (B) Correlation analysis of immunohistochemical scores for the indicated markers shown in (A), scale bar=100 μm. (C) Schematic of ICC patients treated with anti-PD-1 therapy or anti-PD-1 combined chemotherapy (cohort 2, n=45), including their response and LATS2 variation status. (D) Representative radiographic images from one responder patient (LATS2 WT; upper panel) and two non-responder patients—one with a LATS2 mutation (MT; middle panel) and the other with LATS2 copy number loss (loss; lower panel)—before and after anti-PD-1 therapy or anti-PD-1 combined with chemotherapy. Responder shows significant tumor regression. Non-responder shows tumor growth and new lesions. PD-1, programmed death-1; ICC, intrahepatic cholangiocarcinoma; WT, wild-type.
Abbreviations
enzyme-linked-immunosorbent assay
humanized hematopoietic stem cell-engrafted
intrahepatic cholangiocarcinoma
large tumor suppressor kinase 2
polymerase chain reaction
programmed cell death-ligand 1
quantitative real time PCR
single-cell RNA sequencing
T and natural killer cells
t-distributed stochastic neighbor embedding
REFERENCES
- 1. Siegel RL, Miller KD, Jemal A. Cancer statistics, 2019. CA Cancer J Clin 2019;69:7-34.
- 2. Beal EW, Tumin D, Moris D, Zhang XF, Chakedis J, Dilhoff M, et al. Cohort contributions to trends in the incidence and mortality of intrahepatic cholangiocarcinoma. Hepatobiliary Surg Nutr 2018;7:270-276.
- 3. Razumilava N, Gores GJ. Cholangiocarcinoma. Lancet 2014;383:2168-2179.
- 4. Banales JM, Marin JJG, Lamarca A, Rodrigues PM, Khan SA, Roberts LR, et al. Cholangiocarcinoma 2020: the next horizon in mechanisms and management. Nat Rev Gastroenterol Hepatol 2020;17:557-588.
- 5. Spolverato G, Kim Y, Alexandrescu S, Marques HP, Lamelas J, Aldrighetti L, et al. Management and outcomes of patients with recurrent intrahepatic cholangiocarcinoma following previous curative-intent surgical resection. Ann Surg Oncol 2016;23:235-243.
- 6. Rizvi NA, Hellmann MD, Snyder A, Kvistborg P, Makarov V, Havel JJ, et al. Cancer immunology. Mutational landscape determines sensitivity to PD-1 blockade in non-small cell lung cancer. Science 2015;348:124-128.
- 7. Zhou SL, Xin HY, Sun RQ, Zhou ZJ, Hu ZQ, Luo CB, et al. Association of KRAS variant subtypes with survival and recurrence in patients with surgically treated intrahepatic cholangiocarcinoma. JAMA Surg 2022;157:59-65.
- 8. Montal R, Sia D, Montironi C, Leow WQ, Esteban-Fabró R, Pinyol R, et al. Molecular classification and therapeutic targets in extrahepatic cholangiocarcinoma. J Hepatol 2020;73:315-327.
- 9. Dong L, Lu D, Chen R, Lin Y, Zhu H, Zhang Z, et al. Proteogenomic characterization identifies clinically relevant subgroups of intrahepatic cholangiocarcinoma. Cancer Cell 2022;40:70-87e15.
- 10. Nakamura H, Arai Y, Totoki Y, Shirota T, Elzawahry A, Kato M, et al. Genomic spectra of biliary tract cancer. Nat Genet 2015;47:1003-1010.
- 11. Zhou ZJ, Ye YH, Hu ZQ, Hou YR, Liu KX, Sun RQ, et al. Whole-exome sequencing reveals genomic landscape of intrahepatic cholangiocarcinoma and identifies SAV1 as a potential driver. Nat Commun 2024;15:9960.
- 12. Ma S, Meng Z, Chen R, Guan KL. The hippo pathway: biology and pathophysiology. Annu Rev Biochem 2019;88:577-604.
- 13. Kim MH, Kim CG, Kim SK, Shin SJ, Choe EA, Park SH, et al. YAP-induced PD-L1 expression drives immune evasion in BRAFi-resistant melanoma. Cancer Immunol Res 2018;6:255-266.
- 14. Zhu Y, Abedini A, Rodriguez GM, McCloskey CW, Abou-Hamad J, Salah OS, et al. Loss of LATS1 and LATS2 promotes ovarian tumor formation by enhancing AKT activity and PD-L1 expression. Oncogene 2025;44:2240-2252.
- 15. Jiang Z, Li X, Liu F, Li J, Yang K, Xu S, et al. Downregulation of HTRA1 promotes EMT and anoikis resistance in colorectal cancer via activation of Hippo/YAP1 pathway by facilitating LATS2 degradation. Mol Carcinog 2025;64:1330-1346.
- 16. Deng X, Ye D, Hua K, Song H, Luo Q, Munankarmy A, et al. MIR22HG inhibits breast cancer progression by stabilizing LATS2 tumor suppressor. Cell Death Dis 2021;12:810.
- 17. Yu T, Bachman J, Lai ZC. Mutation analysis of large tumor suppressor genes LATS1 and LATS2 supports a tumor suppressor role in human cancer. Protein Cell 2015;6:6-11.
- 18. Misra JR, Irvine KD. The Hippo signaling network and its biological functions. Annu Rev Genet 2018;52:65-87.
- 19. Sohn BH, Shim JJ, Kim SB, Jang KY, Kim SM, Kim JH, et al. Inactivation of Hippo pathway is significantly associated with poor prognosis in hepatocellular carcinoma. Clin Cancer Res 2016;22:1256-1264.
- 20. Huggett MT, Jermyn M, Gillams A, Illing R, Mosse S, Novelli M, et al. Phase I/II study of verteporfin photodynamic therapy in locally advanced pancreatic cancer. Br J Cancer 2014;110:1698-1704.
- 21. Yu FX, Zhao B, Guan KL. Hippo pathway in organ size control, tissue homeostasis, and cancer. Cell 2015;163:811-828.
- 22. Wang Z, Kim SY, Tu W, Kim J, Xu A, Yang YM, et al. Extracellular vesicles in fatty liver promote a metastatic tumor microenvironment. Cell Metab 2023;35:1209-1226e13.
- 23. Hegde PS, Chen DS. Top 10 challenges in cancer immunotherapy. Immunity 2020;52:17-35.
- 24. Lenahan SM, Sarausky HM, Deming P, Seward DJ. STK11 loss leads to YAP1-mediated transcriptional activation in human KRAS-driven lung adenocarcinoma cell lines. Cancer Gene Ther 2024;31:1-8.
- 25. Clara JA, Monge C, Yang Y, Takebe N. Targeting signalling pathways and the immune microenvironment of cancer stem cells - a clinical update. Nat Rev Clin Oncol 2020;17:204-232.
- 26. Al-Moujahed A, Brodowska K, Stryjewski TP, Efstathiou NE, Vasilikos I, Cichy J, et al. Verteporfin inhibits growth of human glioma in vitro without light activation. Sci Rep 2017;7:7602.
, Shao-Lai Zhou1,2
