From bile acids to yes-associated protein: A bidirectional switch for cholangiocarcinoma therapy

Abstract

Cholangiocarcinoma (CCA) is characterized by high invasiveness, early metastasis, and chemotherapy resistance, with a 5-year survival rate persistently below 20%. Current therapies are limited to surgery and a few chemotherapeutic options, highlighting an urgent need for actionable molecular targets and personalized strategies. In the recent issue of World Journal of Gastrointestinal Oncology, Hu et al report that glycochenodeoxycholic acid activates yes-associated protein (YAP) by inhibiting mammalian Ste20-like kinases/large tumor suppressor kinases (MST/LATS) phosphorylation, thereby promoting proliferation, invasion, and suppressing apoptosis, whereas deoxycholic acid enhances MST/LATS phosphorylation to block YAP nuclear entry, inhibiting tumor growth and inducing apoptosis. Given that the conversion between conjugated and unconjugated bile acids is fundamentally governed by intestinal microbial metabolism, this finding underscores the critical regulatory role of the gut microbiota. YAP modulators can reverse these bile-acid effects in both directions, establishing the “bile acid-Hippo-YAP” axis as an actionable driver. This study is the first to demonstrate the opposing roles of these two bile acids within the same model system, providing a mechanistic basis for CCA heterogeneity and suggesting that identifying bile acid biomarkers, manipulating the gut microbiota-bile acid metabolism or combining YAP-targeted drugs could offer novel strategies to overcome the therapeutic bottleneck in CCA. However, the use of a single cell line, one animal model, and the lack of immune microenvironment data limit generalizability; other bile acid species and cross-talk pathways were not explored. The commentary concludes that future work should further validate these findings.

Key Words: Bile acids; Yes-associated protein; Cholangiocarcinoma; Gut microbiota; Therapeutic targets

Core Tip: Bile acids (BAs) exert complex, structure-dependent regulatory effects on the Hippo-yes-associated protein pathway in gastrointestinal malignancies. These editorial highlights a pivotal “bidirectional switch” mechanism where glycochenodeoxycholic acid and deoxycholic acid induce opposing outcomes in cholangiocarcinoma. By integrating the physicochemical distinctions between conjugated and unconjugated BAs with their dynamic interplay in gut microbiota dysbiosis, we propose a multidimensional “microbiota-BA-Hippo” regulatory network. Targeting this axis offers novel precision therapeutic strategies for managing cholangiocarcinoma and metabolic disorders.

This editorial refers to “Involvement of bile acids in cholangiocarcinoma progression via the Hippo-yes-associated protein signaling pathway” by Hu et al, 2025; https://dx.doi.org/10.4251/wjgo.v17.i12.112366.

---

INTRODUCTION

Cholangiocarcinoma (CCA), a highly aggressive malignancy arising from the biliary epithelia, represents a formidable challenge in gastrointestinal oncology due to its insidious onset, profound genetic heterogeneity, and dismal prognosis. Anatomically, CCA is classified into intrahepatic and extrahepatic subtypes. While both share a dismal clinical outcome, emerging evidence suggests that these subtypes harbor distinct genomic landscapes and may exhibit differential sensitivity to bile acid (BA) signaling, likely driven by variations in local BA composition and microbiome exposure within their respective anatomical niches[1]. Advancements in early screening techniques, surgical interventions, chemotherapy approaches, as well as the development of targeted therapies and immunotherapies, have significantly elevated the survival rates of cancer patients[2]. Despite therapeutic refinements, the 5-year survival rate remains stagnantly below 20%[3], largely attributed to late-stage diagnosis and inherent resistance to conventional systemic therapies. This clinical impasse necessitates a deeper understanding of the molecular orchestration within the biliary microenvironment to identify actionable biomarkers and novel therapeutic vulnerabilities[4].

The biliary milieu is uniquely characterized by high concentrations of BAs, which serve not merely as physiological detergents but as potent rheostats of cellular homeostasis[1]. Accumulating evidence underscores that BAs act as multifaceted signaling ligands capable of modulating cell proliferation, apoptosis, and immune evasion through diverse membrane and nuclear receptors[5-8]. Central to this regulatory network is the Hippo signaling pathway and its primary downstream effector, yes-associated protein (YAP), which is frequently hyperactivated in CCA[9] and serves as a core engine driving tumorigenesis[10]. Relative pathway is shown in Figure 1. However, how the structural diversity of the BA pool dictates the activation or silencing of the Hippo-YAP axis remains an area of intense investigation.

Figure 1 Upper panel: Hydrophilic conjugated bile acids (taurocholic acid, glycocholic acid) and lower panel: Lipophilic unconjugated bile acids. The upper panel: Hydrophilic conjugated bile acid taurocholic acid activates nuclear yes-associated protein (YAP) via sphingosine-1-phosphate receptor 2/p38 mitogen-activated protein kinase to promote liver fibrosis. Glycocholic acid also promotes YAP translocation in hepatocytes, inducing connective tissue growth factor expression to activate hepatic stellate cells via paracrine signaling. The lower panel: Lipophilic unconjugated bile acids have dual roles. Chenodeoxycholic acid promotes hepatocellular carcinoma by upregulating IQ motif containing GTPase activating protein 1 to activate YAP. Ursodeoxycholic acid, deoxycholic acid, lithocholic acid activate Takeda G protein-coupled receptor 5/cyclic adenosine monophosphate/protein kinase A to inactivate YAP, inhibiting tumors and maintaining intestinal barrier integrity. BA: Bile acid; S1R2: Sphingosine-1-phosphate receptor 2; GCA: Glycocholic acid; TCA: Taurocholic acid; MAPK: Mitogen-activated protein kinase; YAP: Yes-associated protein; TEAD: Transcriptional enhanced associate domain; CDCA: Chenodeoxycholic acid; LCA: Lithocholic acid; UDCA: Ursodeoxycholic acid; TGR5: Takeda G protein-coupled receptor 5; cAMP: Cyclic adenosine monophosphate; PKA: Protein kinase A; LATS: Large tumor suppressor kinases; HCC: Hepatocellular carcinoma; MST2: Mammalian sterile 20-like kinase 2; IQGAP1: IQ motif containing GTPase activating protein 1.

In a recent issue of the World Journal of Gastrointestinal Oncology, Hu et al[11] provided a compelling paradigm shift by elucidating a structural-dependent “bidirectional switch” of the BA-Hippo-YAP axis. Their study demonstrated that glycochenodeoxycholic acid and deoxycholic acid (DCA) exert opposing effects on CCA progression, offering fresh insights into the molecular stratification of this disease. However, the reliance on a single cell model and specific BAs limits the universality of these conclusions. Building upon these findings, this editorial aims to further delineate the differential mechanisms by which the physicochemical properties of various BAs regulate the Hippo pathway, as well as their complex interplay with gut microbiota dysbiosis in disease progression.

STRUCTURE DICTATE FUNCTION: DIFFERENTIAL REGULATORY MECHANISMS DETERMINED BY PHYSICOCHEMICAL PROPERTIES

BAs serve as critical signaling molecules connecting the liver, gut, and systemic metabolism, and their regulation of the Hippo-YAP pathway exhibits high structural dependency. Research indicates that BAs exert divergent biological functions ranging from tumorigenesis to tissue regeneration through distinct membrane receptors or intracellular kinase cascades, depending on whether they are conjugated with glycine/taurine or remain free.

Conjugated BAs: Receptor-dependent modulation of YAP signaling

Due to their high hydrophilicity, conjugated BAs [e.g., taurocholic acid (TCA) and glycocholic acid (GCA)] cannot directly penetrate the cell membrane and primarily rely on the cell surface G protein-coupled receptor, sphingosine-1-phosphate receptor 2 (S1PR2), to initiate intracellular signal transduction. In hepatic fibrosis and specific gastrointestinal malignancies, TCA and GCA demonstrate significant pro-growth properties. TCA has been confirmed to bind S1PR2 on hepatic stellate cells (HSCs), activating the downstream p38 mitogen-activated protein kinase cascade to upregulate YAP activity and promote fibrosis[12]. Furthermore, in a rat model of cholestasis, TCA was shown to directly promote YAP activation via the extracellular regulated protein kinases (ERK) signaling pathway[13]. In esophageal adenocarcinoma, refluxed TCA similarly activates YAP through S1PR2 signaling, conferring a more aggressive phenotype to cancer cells[14]. Similarly, GCA promotes YAP nuclear translocation and induces connective tissue growth factor expression in hepatocytes, thereby paracrinely activating HSCs[15]. However, the hydrophobic conjugated glycochenodeoxycholic acid induces aberrant senescence in biliary epithelial cells when excessively accumulated in the bile duct, resulting in YAP signaling failure and regenerative repair defects[16]. Notably, not all conjugated BAs promote pathological processes; specific bile acids differentially regulate YAP signaling and its associated pathological consequences are shown in Table 1 tauroursodeoxycholic acid acts as a chemical chaperone, alleviating endoplasmic reticulum stress to promote YAP nuclear exit, thereby effectively inhibiting liver overgrowth[17].

Table 1 Differential regulation of yes-associated protein signaling and pathological outcomes by specific bile acid species.

Bile acid	Mechanism of action	Impact on yap	Pathological outcome	Ref.
TCA	S1PR2-MAPK	Activation1	Liver fibrosis; EAC progression	Sugihara et al[10]; Hu et al[11]; Yang et al[12]
GCA	Paracrine-CTGF	Activation1	Aggravates liver fibrosis	Yu et al[13]
GCDCA	Inhibits MST/LATS	Activation1	Drives CCA proliferation and invasion	Ilyas et al[9]
	Intracellular stress	Inhibition2	Biliary senescence and repair defects	Liu et al[14]
TUDCA	ER stress chaperone	Inhibition2	Restrains liver overgrowth	Yuan et al[15]
CDCA	IQGAP1-MST2	Activation1	Promotes HCC progression	Sasaki et al[16]
DCA	TGR5-cAMP	Inhibition2	Suppresses CCA growth (tumor switch)	Ilyas et al[9]; Mao et al[19]
	ABL1 axis	Activation1	Facilitates hepatic steatosis	Wu et al[17]
UDCA	TGR5-cAMP	Inhibition2	Suppresses CRC progression	Quinn et al[18]

¹Represents activation (nuclear accumulation/target gene expression).

²Represents inhibition (cytoplasmic retention/degradation).

TCA: Taurocholic acid; GCA: Glycocholic acid; GCDCA: Glycochenodeoxycholic acid; TUDCA: Tauroursodeoxycholic acid; CDCA: Chenodeoxycholic acid; DCA: Deoxycholic acid; UDCA: Ursodeoxycholic acid; MAPK: Mitogen-activated protein kinase; S1PR2: Sphingosine-1-phosphate receptor 2; CTGF: Connective tissue growth factor; MST/LATS: Mammalian Ste20-like kinases/Large tumor suppressor kinases; ER: Endoplasmic reticulum; MST2: Mammalian sterile 20-like kinase 2; IQGAP1: IQ motif containing GTPase activating protein 1; TGR5: Takeda G protein-coupled receptor 5; cAMP: Cyclic adenosine monophosphate; ABL1: Abelson tyrosine-protein kinase 1; EAC: Esophageal adenocarcinoma; CCA: Cholangiocarcinoma; HCC: Hepatocellular carcinoma; CRC: Colorectal cancer.

Unconjugated BAs: Membrane permeability and Takeda G protein-coupled receptor 5-mediated

Possessing higher lipophilicity and membrane permeability, unconjugated BAs function either through the membrane receptor Takeda G protein-coupled receptor 5 (TGR5) or by directly interfering with intracellular kinase networks. Hydrophobic unconjugated BAs often activate YAP by disrupting intracellular scaffold proteins or kinase activity. For instance, chenodeoxycholic acid (CDCA) treatment upregulates the scaffold protein IQ motif containing GTPase activating protein 1, inhibiting mammalian sterile 20-like kinase 2 expression and LATS1 activation, thereby promoting hepatocellular carcinoma progression[18]. In metabolic dysfunction-associated fatty liver disease, DCA activates the intestinal Abelson tyrosine-protein kinase 1-YAP1 axis, facilitating hepatic steatosis through gut-liver signaling[19]. Conversely, when unconjugated BAs act on the TGR5 receptor, they often exhibit inhibitory or adaptive regulation of YAP. In colorectal cancer (CRC), ursodeoxycholic acid activates TGR5 to initiate the cyclic adenosine monophosphate (cAMP)/protein kinase A (PKA) signaling pathway, inhibiting RhoA activity and blocking YAP activation signals, thereby suppressing tumor growth[20]. In the context of intestinal injury, elevated levels of DCA and lithocholic acid can upregulate TGR5 expression and modulate YAP signaling via the RhoA/ROCK pathway, contributing to the mitigation of epithelial damage and the maintenance of intestinal barrier integrity[21]. While the TGR5-Hippo signaling axis offers a precise molecular explanation for DCA-mediated tumor suppression, it is valuable to acknowledge that DCA operates through a broader, multimodal cytotoxic network. This action is likely concentration-dependent: At high physiological or pathological concentrations, the strong hydrophobicity and amphipathic nature of DCA exert a ‘detergent effect’ on lipid bilayers. This physical perturbation increases membrane permeability, often exceeding cellular repair capacities to trigger rapid lysis and necrosis rather than regulated apoptosis[22,23].

In parallel, DCA-induced physicochemical stress induces mitochondrial dysfunction and reactive oxygen species accumulation, creating a synergistic bridge to Hippo signaling. Reactive oxygen species are established modulators of the Hippo pathway; for instance, oxidative stress has been shown to control the CBP-MOB1 axis, activating LATS1 to suppress YAP nuclear translocation[24,25]. Furthermore, DCA acts as a unique inducer of death receptor pathways. In hepatocellular carcinoma models, DCA triggers the ligand-independent clustering of Fas receptors on the cell membrane, recruiting FADD and caspase-8 to initiate apoptosis even in the absence of Fas ligand[26]. Collectively, these mechanisms ranging from specific receptor activation (TGR5) to broad oxidative and membrane stress likely function synergistically to overcome the survival threshold of CCA cells.

Summary: The structural code of BA signaling

In conclusion, while conjugated and unconjugated BAs share a common reliance on G-protein-coupled receptors to bridge the extracellular milieu and intracellular signaling, their structural distinctness dictates divergent functional outcomes.

Conjugated BAs (e.g., TCA), functioning via S1PR2, act as a ‘nuclear shuttle’. This interaction drives YAP nuclear entry via ERK/phosphatidylinositol 3-kinase cascades[27,28]. Crucially, this axis engages in crosstalk with the Wnt/β-catenin pathway and is further intricately regulated by Rho and mevalonate metabolism[29,30], where YAP-β-catenin cooperation serves as a hallmark driver of gastrointestinal tumorigenesis[31].

In stark contrast, unconjugated BAs (e.g., DCA) acting via TGR5 function as a ‘cytoplasmic anchor’. By triggering the cAMP-PKA-large tumor suppressor kinases cascade, this axis enforces YAP phosphorylation and cytoplasmic retention[21]. This spatial segregation fundamentally disrupts YAP-transcriptional enhanced associate domain (TEAD) interaction and concurrently inhibits nuclear factor kappa-B-driven inflammation[32].

Thus, the ‘bidirectional switch’ is essentially a competition between specific receptor affinities and intracellular trafficking routes, ultimately determining the assembly or disruption of YAP-TEAD transcriptional complexes[33], thereby flipping the genomic switch between proliferation and suppression.

THE GUT-LIVER AXIS PERSPECTIVE: SYNERGISTIC EVOLUTION OF MICROBIOTA AND BAS

Synergistic impact of gut microbiota and BAs on metabolic disease progression

The gut microbiota is regarded as a core regulator of host metabolism, physically connected to the liver via the enterohepatic circulation. BAs, synthesized by the liver and metabolized by intestinal bacteria, serve as the critical nexus linking the microbiome to host physiological functions[34]. Host enzymatic activity and microbial metabolic transformations, specifically deconjugation and 7α-dehydroxylation driven by genera such as Bacteroides and Clostridium[35,36], collectively determine the size and composition of the BA pool; consequently, dysregulation of this ‘microbiota-metabolism’ axis is closely linked to the progression of various metabolic and inflammatory diseases[23].

Accumulating evidence demonstrates that BAs regulate disease pathogenesis through dynamic reciprocal interactions between the gut microbiota and the host. This regulatory capacity is well-exemplified in metabolic disorders, providing a mechanistic template for understanding CCA. For instance, glycoursodeoxycholic acid has been shown to positively modulate the gut microbiota by increasing the abundance of Bacteroides vulgatus, which subsequently elevates levels of taurolithocholic acid. This metabolic shift triggers TGR5 activation in adipose tissue and upregulates uncoupling protein 1 expression, thereby promoting thermogenesis in white adipose tissue a mechanism with significant therapeutic implications for type 2 diabetes[37,38]. Furthermore, in metabolic diseases, the gut microbiota composition is significantly shaped by dietary lipid characteristics. The microbiota influences host lipid homeostasis by generating metabolites such as secondary BAs and pro-inflammatory factors (e.g., lipopolysaccharides), and signaling disruptions within this pathway are closely associated with lipid disorders like non-alcoholic fatty liver disease[39]. Crucially, the implication of dysbiosis extends beyond metabolic disorders; emerging evidence identifies specific alterations in the human microbiome and their underlying mechanisms as key drivers of CCA progression[40]. Concurrently, within the intestinal microenvironment, the interaction between BAs and host receptors, including the farnesoid X receptor (FXR), vitamin D receptor, and TGR5, plays a pivotal role in immune regulation[41]. Clinical omics studies have revealed that alterations in microbial BA metabolism are highly correlated with the progression of inflammatory bowel disease (IBD) and CRC. Various microbial-derived BA metabolites modulate intestinal immune cell responses; in IBD patients, microbial perturbations profoundly alter the BA profile, shifting steady-state immune signals toward pro-inflammatory pathways[42]. In the development of CRC, microbial-derived BA metabolites exert critical influence through direct interaction with CRC cells[43].

Building on these findings, the “microbiota-BA” axis presents a concrete therapeutic avenue for CCA. These diverse BA components act not only as drivers of carcinogenesis but also as potential diagnostic biomarkers and therapeutic targets. Strategies such as enriching DCA-generating bacteria (e.g., via probiotics) or using prebiotics to reshape the BA pool could restore physiological levels of tumor-suppressive secondary BAs, thereby re-engaging the TGR5-Hippo axis to inhibit tumor progression[44-46].

CLINICAL TRANSLATION: BIOMARKERS, THERAPEUTICS, AND CHALLENGES

Beyond mechanistic elucidation, the “microbiota-BA-Hippo” axis holds profound clinical implications for precision oncology. Identifying reliable diagnostic biomarkers is the first step toward translation. Distinct plasma BA profiles have demonstrated superior diagnostic sensitivity compared to conventional markers like carbohydrate antigen 19-9[47]. Specifically, unconjugated CDCA effectively differentiates CCA from benign biliary diseases [area under the curve (AUC) = 0.842] and primary sclerosing cholangitis (AUC = 0.741), while TUDCA shows enhanced specificity for distinguishing CCA from PSC (AUC = 0.783)[48]. Beyond single metabolites, integrative multi-omics signatures offer greater precision; combining the plasma-stool ratio of TUDCA with specific gut genera (e.g., Lactobacillus) significantly improves the diagnosis of intrahepatic CCA. On the prognostic front, specific microbiome-BA signatures also hold stratification value. Studies indicate that an elevated abundance of Ruminococcaceae and pathologically high plasma TUDCA levels are positively correlated with vascular invasion and poor survival, thereby enabling non-invasive risk stratification for high-risk patients[49].

On the therapeutic front, strategies targeting this axis are evolving from “repurposing” to “precision targeting”. Existing agents such as FXR agonists (e.g., obeticholic acid) and TGR5 modulators have shown potential in remodeling the tumor immune microenvironment. More critically, emerging clinical-stage agents like VT3989 allosterically inhibit TEAD autopalmitoylation to disrupt the YAP-TEAD complex, offering a more potent and specific alternative to verteporfin[50].

Nevertheless, targeting the gut-liver axis presents significant translational hurdles. The dense desmoplastic stroma characteristic of CCA acts as a physical barrier that severely limits the delivery efficiency of small-molecule inhibitors. Additionally, the high inter-individual heterogeneity of the gut microbiota may lead to variable drug metabolism and therapeutic responses. Addressing these challenges perhaps through nanomedicine delivery systems or microbiota-based adjuvant therapies is imperative for realizing the promise of precision medicine in CCA.

LIMITATIONS AND FUTURE DIRECTIONS

While Hu et al[11] provide foundational insights into the DCA-YAP axis, it is important to acknowledge the limitations inherent in their reliance on in vitro models using a single cell line. This approach, while mechanistically precise, precludes an assessment of the complex interplay within the tumor immune microenvironment. These constraints reflect broader challenges in the field of CCA research. Current in vivo models often fail to fully recapitulate human biliary physiology and the unique composition of the human gut microbiota, which are the primary drivers of BA metabolism. Consequently, future research must prioritize the development of humanized mouse models or faithful patient-derived xenografts that better mimic the clinical gut-liver axis. Furthermore, given the extensive cellular heterogeneity of CCA, emerging high-resolution technologies such as single-cell RNA sequencing and spatial transcriptomics are essential. These tools will allow researchers to spatially resolve how BA gradients influence distinct cellular sub-populations including stromal and immune cells thereby providing a more holistic view of the disease landscape.

CONCLUSION

In summary, the crosstalk between the BA pool and the Hippo-YAP signaling architecture represents not merely a local feedback loop, but a fundamental mechanism governing systemic homeostasis. The study by Hu et al[11] serves as a pivotal entry point, illuminating how the structural heterogeneity of BAs acts as a precise molecular switch to dictate cellular fate. However, translating these mechanistic insights into viable clinical therapies will require a paradigm shift toward interdisciplinary collaboration among oncologists, microbiologists, and bioinformaticians. Future efforts must focus on two critical directions: Validated patient stratification based on circulating BA profiles, and the development of combination therapy strategies that synergize YAP-targeted drugs with microbiome modulators. By manipulating this upstream ecological engine through such focused, collaborative approaches, we may unlock novel precision therapies for intractable gastrointestinal malignancies.