Candesartan cilexetil as a repurposed therapeutic candidate for nasopharyngeal carcinoma: Integrated in vitro and in silico analyses

TL;DR

This study explores candesartan cilexetil, a blood pressure drug, as a potential treatment for nasopharyngeal carcinoma, showing it can inhibit cancer cell growth and migration.

Contribution

The study identifies candesartan cilexetil as a repurposed therapeutic candidate for NPC through integrated in vitro and in silico analyses.

Findings

CC significantly reduced NPC cell viability and proliferation with lower toxicity to normal cells.

CC induced G0/G1 cell-cycle arrest and suppressed migration, with EMT marker modulation.

MD simulations confirmed stable CC-AT1R binding and key residues for ligand stabilization.

Abstract

Nasopharyngeal carcinoma (NPC) is prevalent in East and Southeast Asia and is often diagnosed at advanced stages, where current treatment options are limited and associated with high relapse rates and toxicity. Repurposing clinically approved drugs provides a rapid, cost-effective strategy to identify new therapeutic interventions. Here, we investigated candesartan cilexetil (CC), an angiotensin II type 1 receptor (AT1R) blocker widely used for hypertension (13), for its potential anti-cancer effects in NPC. In vitro assays were performed to assess cell viability, proliferation, clonogenic survival, migration, cell-cycle distribution, and epithelial-mesenchymal transition (EMT) marker expression. Molecular mechanisms were examined via immunoblotting of AT1R and downstream MAPK and PI3K-AKT pathways. In silico molecular dynamics (MD) simulations were conducted to characterize CC-AT1R binding. We found that CC significantly decreased NPC cell viability and proliferation in a concentration-dependent manner, while exhibiting lower cytotoxicity toward immortalized nasopharyngeal epithelial cells. CC inhibited colony formation, induced G0/G1 cell-cycle arrest, and suppressed migration. EMT markers were differentially regulated, with consistent downregulation of vimentin and Slug but paradoxical cadherin changes, indicating context-dependent EMT modulation. Mechanistically, CC downregulated AT1R expression, reduced phosphorylation of p38 MAPK, and enhanced AKT phosphorylation, suggesting compensatory survival signaling. MD simulations confirmed stable CC-AT1R binding, identifying key residues critical for ligand stabilization. Therefore, CC exerts multi-faceted inhibitory effects on NPC cells through AT1R blockade and downstream modulation of oncogenic pathways. The integration of in vitro and in silico analyses highlights CC as a promising repurposed therapeutic candidate for NPC and supports further preclinical evaluation.

See also the graphical abstract(Fig. 1).

Figures

A[(Fig. 3)](#F3)). Specifically, at 25.0 μM, proliferation was significantly reduced compared to controls by 48 h, while 100 μM CC almost completely abrogated growth. The 48-h half maximal inhibitory concentration (IC₅₀) values were determined to be 38.4 μM for HK1 and 37.9 μM for TW01 (Figure 3B[(Fig. 3)](#F3)). Consistent with these anti-proliferative effects, prolonged exposure to CC (12.5 - 25.0 μM for 10 days) in clonogenic assays markedly decreased both the number and size of resulting colonies (Figure 4[(Fig. 4)](#F4)), confirming that CC significantly compromises the NPC cells' capacit

A[(Fig. 9)](#F9)). The RMSD value of the backbone protein-ligand complex exhibited a slight deviation at around 60-200 ns and became more stable after 250 ns. In addition, the RMSD values for backbone amino acids within 5 Å showed a stable trend throughout almost the simulation time and were lower compared to the RMSD of the entire complex. This suggests that the higher RMSD observed in the whole complex may originate from more distant regions, indicating that the amino acids within 5 Å were stable and favorably interacted with CC. Apart from RMSD analysis, all-atom contacts within a 5 Å dista

Full text

Introduction

Nasopharyngeal carcinoma (NPC) is a distinctive epithelial malignancy of the head and neck, characterized by unique epidemiological, etiological, and clinical features (Chua et al., 2016[8]; Tang et al., 2016[45]). Its incidence is heavily concentrated, with the highest prevalence observed in Southern China and Southeast Asia. In these endemic regions, NPC represents a major public health concern, with a disproportionately higher burden on men. The multifactorial nature of NPC carcinogenesis is driven by a combination of factors, including Epstein-Barr virus infection, genetic predisposition, dietary nitrosamines, and specific environmental exposures (Lo et al., 2004[27]; Richardo et al., 2020[38]). The standard treatment for NPC relies on radiotherapy, frequently combined with platinum-based chemotherapy for advanced disease (Chen et al., 2019[5]). Although early-stage NPC often achieves reasonable control, most patients are unfortunately diagnosed at advanced stages, leading to suboptimal outcomes (Tulalamba and Janvilisri, 2012[47]). High recurrence rates, ranging from 15 % to 58 % following primary therapy, coupled with severe treatment-related toxicities-such as mucositis, xerostomia, ototoxicity, and hematologic suppression-significantly limit long-term survival and diminish the quality of life (Xu et al., 2013[53]). Consequently, there is an urgent and unmet clinical need for novel therapeutic strategies that can enhance treatment efficacy while simultaneously reducing collateral toxicity.

While advancements in targeted therapies and immunotherapies have transformed the treatment landscape for many cancers (Hanahan and Weinberg, 2011[14]; Ranieri et al., 2006[37]), their effective application in NPC remains limited. An increasingly appealing and clinically practical alternative is drug repurposing. This strategy involves systematically evaluating approved drugs with established safety and pharmacokinetic profiles for new oncological indications. This approach significantly accelerates clinical translation, reduces extensive development costs, and leverages existing pharmacological and toxicological knowledge. A classic example of successful repurposing in oncology is thalidomide, which was initially used for nausea in pregnancy and later repurposed for the treatment of multiple myeloma (Sleire et al., 2017[40]; Bertolini et al., 2015[4]; Gupta et al., 2013[13]).

Candesartan cilexetil (CC) is an established, clinically approved angiotensin II type 1 receptor (AT1R) antagonist widely prescribed for hypertension and heart failure (Easthope and Jarvis, 2002[11]). Growing evidence indicates that AT1R signaling is not confined to cardiovascular function but actively contributes to key oncogenic processes, including tumor proliferation, angiogenesis, and metastasis (Ishikane and Takahashi-Yanaga, 2018[17]). Consistent with this, CC has been reported to inhibit cell proliferation across various cancer types, including breast, bladder, gastric, and colorectal cancers (Du et al., 2012[10]; Kosugi et al., 2006[24]; Okazaki et al., 2014[33]; Tabatabai et al., 2021[44]). However, despite the potential link between AT1R signaling and cancer progression, the role and therapeutic efficacy of candesartan in NPC have not been systematically investigated.

In this study, we hypothesized that CC could be successfully repurposed as a targeted therapeutic agent for NPC by blocking AT1R and modulating its downstream signaling cascades. We employed a robust, integrated strategy combining in vitro cellular assays with molecular dynamics (MD) simulations to achieve three primary objectives: (i) assess the cytotoxic, anti-proliferative, and anti-migratory effects of CC on NPC cells; (ii) elucidate its impact on the epithelial-mesenchymal transition (EMT) process and critical oncogenic signaling cascades (MAPK and PI3K-AKT); and (iii) characterize the molecular basis of CC-AT1R interactions at an atomic resolution. This dual experimental and computational approach provides essential mechanistic insights into CC's anticancer effects, laying a strong foundation for its potential repositioning as a novel targeted therapy for NPC.

Materials and Methods

Cell culture

We utilized a panel of nine human NPC cell lines (TW01, SUNE5-8F, SUNE6-10B, HK1, HONE1, HNE1, SUNE1, CNE1, and CNE2) alongside the immortalized nasopharyngeal epithelial line, NP69. The cell lines were sourced from various labs and centers: TW01 (Prof. C-T Lin, National Taiwan University), SUNE5-8F and SUNE6-10B (Prof. Qingling Zhang, Southern Medical University), and HK1 (Prof. Maria L Lung, University of Hong Kong). The remaining NPC lines (HONE1, HNE1, SUNE1, CNE1, CNE2) and NP69 were obtained from the Center for Nasopharyngeal Carcinoma Research, University of Hong Kong. This panel encompasses diverse clinical subtypes: TW01 and HK1 are classified as moderately to well-differentiated, while CNE2, HONE1, and HNE1 are poorly differentiated. Notably, SUNE5-8F (highly metastatic) and SUNE6-10B (poorly metastatic) are isogenic derivatives of the SUNE1 parental line. All NPC cell lines were cultured in RPMI-1640 medium (Gibco, New York, USA) supplemented with 10 % fetal bovine serum (FBS; Gibco) and 1 % antibiotics (penicillin-streptomycin; Gibco). The non-malignant NP69 epithelial cells required specialized growth conditions, maintained in a 1:1 mixture of Defined Keratinocyte Serum-Free Medium (Gibco) with 0.2 % defined Keratinocyte Growth Supplement (Gibco) and EpiLife™ medium (Gibco) supplemented with 1X human keratinocyte growth supplements (Gibco) and 1 % antibiotics (penicillin-streptomycin; Gibco). All cells were maintained at 37 °C in a humidified atmosphere of 5 % CO₂.

MTT cell viability assay

Cell viability was assessed using the MTT assay as described previously (Kiatwuthinon et al., 2021[21]). Briefly, NPC and NP69 cells (3,000 cells/well) were seeded in 96-well plates and treated with CC (1-100 μM) for 48 h. CC (MedChemExpress, New Jersey, USA) was dissolved in dimethyl sulfoxide (DMSO; Sigma, Massachusetts, USA), ensuring the final DMSO concentration did not exceed 1 % (v/v). After CC treatment, MTT reagent (0.5 mg/mL; Sigma) was added for 3 h. Formazan crystals were subsequently dissolved in DMSO, and absorbance was measured at 540 nm using a microplate reader. Viability was expressed relative to control cells.

CCK-8 proliferation assay

Cell proliferation over time was quantified using the Cell Counting Kit-8 (CCK-8; Abbkine Scientific, Georgia, USA). NPC cells were seeded at 3,000 cells/well and treated with CC (0.1-100 μM). The cells were grown under conditions described previously, and the medium was replaced with a new RPMI medium containing 1 % (v/v) of CC. CCK-8 solution (10 % v/v) was added at 24, 48, 72, and 96 h, incubated for 1.5 h, and absorbance was recorded at 450 nm. Data were normalized to the corresponding control group.

Clonogenic survival assay

The clonogenic assay was performed as described previously (Ngernsombat et al., 2024[30]). Briefly, HK1 and TW01 cells (150 cells/well) were seeded in 24-well plates. Once adhered, cells were treated with CC (12.5 or 25.0 μM) for 10 days with medium refreshed every 3 days. Colonies were fixed with methanol, stained with 1 % crystal violet, and manually quantified.

Cell cycle analysis

NPC cells (2 x 10^5^) were cultured in RPMI complete medium supplemented with CC (12.5 or 25.0 μM) until reaching 80 % confluence (48 h). Cells were then fixed in 70 % cold ethanol and stained with 10 μg/mL of propidium iodide (PI) and 300 μg/mL of RNase using PI/RNase Staining Solution (Cell Signaling Technology, Massachusetts, USA). Cell cycle analysis was performed using a DxFLEX flow cytometer (Beckman Coulter, California, USA), and cell cycle distribution was quantified with FlowJo software (FlowJo, Oregon, USA).

Wound healing migration assay

The wound healing assay was performed as described previously (Aimjongjun et al., 2020[1]). Briefly, NPC cells (2 x 10^5^) were seeded in 24-well plates and grown to complete confluence. A linear scratch was made across the monolayer using a sterile pipette tip. Cells were treated with sub-lethal concentrations of CC (12.5 and 25.0 µM), in RPMI medium containing 0.1 % FBS for 24 h. Images were captured at 0 and 24 h using an inverted microscope (Eclipse TS100; Nikon Eclipse TS100; Tokyo, Japan), and migration was quantified as the percentage of wound closure using ImageJ software.

Western blotting for protein expression

Cells were treated with CC (0.0-25.0 μM) for 48 h. Lysates were prepared using RIPA buffer supplemented with protease and phosphatase inhibitors, and protein concentrations were determined by the Bradford assay (Sigma). Equal amounts of proteins were resolved via SDS-PAGE, transferred onto nitrocellulose membranes (Biorad, California, USA), and blocked with 5 % bovine serum albumin. Membranes were incubated overnight at 4 °C with primary antibodies. Following secondary antibody incubation, bands were visualized using enhanced chemiluminescence and quantified with ImageJ. Antibodies against AT1R, E-cadherin, N-cadherin, vimentin, Slug, total/phospho-p38 MAPK, and total/phospho-AKT were sources from Cell Signaling Technology, with β-actin as a loading control.

Molecular dynamics (MD) simulations

The crystal structure of human AT1R bound to antagonist ZD7155 (PDB ID: 4YAY) at 2.9-Å resolution was used as the receptor model. The 3D structure of CC was retrieved from PubChem (CID: 2540). Initial docking was performed with AutoDock VinaXB (Koebel et al., 2016[23]; Zhang et al., 2015[54]). The resulting complex was prepared for MD simulations: missing residues were modeled using SwissModel, and ionizable residues were assigned protonation states at pH 7.4 using the PDB2PQR web server (Jurrus et al., 2018[19]; Søndergaard et al., 2011[41]). Ligand partial atomic charges were calculated with the Antechamber module in AMBER24. The AMBER ff19SB force field (Tian et al., 2020[46]) was applied to treat the protein, while ligand parameters were derived from the General AMBER Force Field version 2 (GAFF2). The system was solvated using the TIP3P water model (Jorgensen et al., 1983[18]). Simulations were conducted under periodic boundary conditions in the isothermal-isobaric (NPT) ensemble at 310 K and 1 atm. Long-range electrostatic interactions were managed using the particle mesh Ewald (PME) method, and constraints on bonds involving hydrogen atoms were imposed using the SHAKE algorithm (Darden et al., 1993[9]; Ryckaert et al., 1977[39]). The system's temperature and pressure were controlled via the Langevin thermostat and Berendsen barostat, respectively (Paterlini and Ferguson, 1988[35]; Uberuaga et al., 2004[48]; Berendsen et al., 1984[3]). The production phase of the MD simulations lasted 300 ns. Trajectory data were analyzed using the CPPTRAJ module, and per-residue energy decomposition was performed with the MM-GBSA approach. Ligand binding affinities, termed binding free energy (ΔG**bind), were estimated using MM-GBSA and MM-PBSA methods (Wang et al., 2019[51]). The calculation was based on the end-point method, neglecting the entropic term (ΔS) to reduce the computational time and cost. All simulations were carried out using the AMBER24 software suite.

Statistical analysis

All in vitro experiments were performed in triplicate and repeated independently at least three times. Data are presented as the mean ± standard deviation (SD) or standard error of the mean (SEM), where appropriate. Statistical significance was determined using an unpaired Student's t-test, with a p-value of <0.05 considered statistically significant.

Results

CC exerts preferential cytotoxicity against NPC cells

Treatment with CC significantly reduced the viability of all tested NPC cell lines in a concentration-dependent manner following 48 h of exposure (1-100 μM). MTT assays consistently demonstrated cytotoxicity across the panel of NPC lines. The half-maximal inhibitory concentrations (IC₅₀), determined by non-linear regression analysis, consistently ranged between approximately 35 - 40 μM across the different malignant cell lines. Crucially, the immortalized nasopharyngeal epithelial cell line, NP69, exhibited markedly lower sensitivity, retaining over 80 % viability even at the 10 μM CC concentration (Figure 2(Fig. 2)). This demonstrates the preferential cytotoxicity of CC toward malignant cells, suggesting a favorable therapeutic window for repurposing.

For subsequent detailed mechanistic investigations, we selected HK1 and TW01. These two lines represent different differentiation statuses (HK1 is poorly differentiated/keratinizing; TW01 is non-keratinizing) and distinct genetic backgrounds, allowing us to demonstrate the broad efficacy of CC across NPC's heterogeneous landscape. Although SUNE6-10B showed a slightly lower IC₅₀, HK1 and TW01 were chosen specifically because their well-characterized utilization in international NPC research regarding EMT and metastasis offers better comparability with existing literature and provides a more representative model for analyzing signaling pathways.

CC suppresses proliferation and compromises long-term clonogenic survival

Proliferation assays using the CCK-8 method confirmed that CC suppressed NPC cell growth in a time- and concentration-dependent fashion (Figure 3A(Fig. 3)). Specifically, at 25.0 μM, proliferation was significantly reduced compared to controls by 48 h, while 100 μM CC almost completely abrogated growth. The 48-h half maximal inhibitory concentration (IC₅₀) values were determined to be 38.4 μM for HK1 and 37.9 μM for TW01 (Figure 3B(Fig. 3)). Consistent with these anti-proliferative effects, prolonged exposure to CC (12.5 - 25.0 μM for 10 days) in clonogenic assays markedly decreased both the number and size of resulting colonies (Figure 4(Fig. 4)), confirming that CC significantly compromises the NPC cells' capacity for long-term survival and self-renewal.

CC induces G0/G1 cell-cycle arrest and impairs migration

To investigate the mechanism underpinning CC's anti-proliferative effects, we conducted cell-cycle analysis. TW01 cells were specifically selected for this assay because they exhibit a strong, concentration-dependent anti-proliferative response and are known to be highly responsive to cell-cycle disrupting agents, making them the most suitable model to clearly observe and quantify the expected G0/G1 arrest phenotype. Flow cytometric analysis demonstrated that CC treatment (12.5 - 25.0 μM for 48 h) induced a clear accumulation of TW01 cells in the G0/G1 phase (Figure 5(Fig. 5)). The proportion of cells in G0/G1 increased from 77.5 % (control) to 86.5 % at the 25.0 μM CC concentration, with a concomitant decrease in S-phase entry (from 12.7 % to 7.4 %). This indicates that CC disrupts NPC cell-cycle progression by enforcing a G1 checkpoint arrest, thus limiting DNA synthesis and proliferation. Furthermore, CC significantly suppressed a key determinant of metastatic potential. Wound-healing assays revealed that CC impaired the migratory capacity of both HK1 and TW01 cells. At sub-lethal concentrations (12.5-25.0 μM), CC reduced wound closure in a concentration-dependent manner over the 24-h period compared to untreated controls (Figure 6(Fig. 6)). This suggests that CC directly targets the cells' intrinsic ability to move and invade.

CC downregulates AT1R and differentially modulates oncogenic signaling

Western blotting revealed that CC altered EMT marker expression, though in a cell line-dependent manner (Figures 7A-B(Fig. 7)). Both HK1 and TW01 cells showed consistent suppression of the mesenchymal markers, vimentin and Slug. However, HK1 exhibited a paradoxical upregulation of N-cadherin and downregulation of E-cadherin at higher concentrations, while TW01 showed reduced E-cadherin at 25.0 μM. These complex findings suggest that CC exerts a partial, context-dependent EMT inhibitory effect. CC treatment also resulted in a significant reduction of AT1R protein expression in both HK1 and TW01 cells (Figures 7C-D(Fig. 7)). Consistent with the suppression of proliferative signals, the phosphorylation of p38 MAPK (pp38) was markedly decreased, while total p38 levels remained stable (Figures 7E-F(Fig. 7)). Interestingly, CC induced a concentration-dependent increase in phosphorylated AKT (pAKT), despite a reduction in total AKT at higher concentrations (Figures 7G-H(Fig. 7)). This suggests a possible compensatory activation of the PI3K-AKT survival pathway in response to CC's primary inhibitory actions (MAPK suppression and AT1R downregulation) (proposed mechanism in Figure 8(Fig. 8)).

Molecular dynamics simulations confirm stable CC-AT1R binding

Molecular dynamics simulations were conducted to delineate CC-AT1R interactions. This simulation has been widely employed in various fields of research, particularly in drug discovery and development (Suriya et al., 2024[43]). Root-mean-square deviation (RMSD) analysis demonstrated stable complex formation, with residues within 5 Å of CC remaining conformationally stable throughout 300 ns (Figure 9A(Fig. 9)). The RMSD value of the backbone protein-ligand complex exhibited a slight deviation at around 60-200 ns and became more stable after 250 ns. In addition, the RMSD values for backbone amino acids within 5 Å showed a stable trend throughout almost the simulation time and were lower compared to the RMSD of the entire complex. This suggests that the higher RMSD observed in the whole complex may originate from more distant regions, indicating that the amino acids within 5 Å were stable and favorably interacted with CC. Apart from RMSD analysis, all-atom contacts within a 5 Å distance were calculated to estimate the number of atoms participating in the ligand binding. We found that, on average, CC could have contacts with approximately 400 ± 29 atoms (Figure 9B(Fig. 9)), suggesting the formation of a stable and sustained complex.

To elucidate insights into critical amino acids for ligand binding, the decomposition free energy of binding (ΔG*^residue^bind) was calculated by the MM/GBSA method. For this purpose, only residues showing ΔG^residue^bind* < -1.00 kcal/mol were represented as “hotspots”. As illustrated in Figure 9C-D(Fig. 9), six key residues involved in CC binding were Trp84, Ser105, Val108, Ser109, Met284, and Ile288. In particular, Trp84 showed a dominant role in stabilizing the ligand occupation with the lowest ΔG*^residue^**bind* among all residues observed in both compounds. Almost all identified key residues for CC's binding were also observed in the binding of ZD7155, a selective antagonist (Trp84, Ser105, Val108, Ser109, and Ile288), suggesting a possibly similar mechanism of action, resulting in an effective inhibitory activity.

To evaluate the ligand binding strength, the final 50 ns of the molecular dynamics (MD) trajectories (250-300 ns) were employed to estimate the end-point free energy of binding (ΔG**bind) using the MM/PBSA and MM/GBSA approaches. As listed in Table 1(Tab. 1), both methods consistently indicated strong binding of compound CC to the AT1R protein in an aqueous environment, with ΔG**bind values of -43.4013 ± 1.5712 and -55.1464 ± 0.8145 kcal/mol by the MMPBSA and MMGBSA methods, respectively. Analysis of the energy components revealed that van der Waals interactions were the dominant driving force for binding, contributing significantly more than electrostatic interactions. The van der Waals contribution (ΔE**vdW) was approximately 4.9 times lower than the electrostatic energy (ΔE**electrostatic), highlighting its primary role in ligand recognition. These findings suggest potential ways for rational ligand optimization, such as increasing favorable electrostatic interactions while maintaining van der Waals and other energetic contributions (ΔG**polar and ΔG**nonpolar) to further increase overall binding affinity.

Discussion

NPC remains a significant therapeutic challenge, particularly in endemic regions, owing to frequent diagnosis at advanced stages and high rates of relapse following primary treatment. The pressing need for novel, well-tolerated agents that target key oncogenic drivers is paramount. Our study successfully demonstrates that CC, a clinically established AT1R antagonist, exhibits robust and multi-faceted anti-cancer activity against NPC cells. Angiotensin II signaling, mediated through AT1R, is known to enhance cell proliferation and promote tumor progression across various malignancies (Kinoshita et al., 2009[22]; Huang et al., 2014[15]; Nowakowska et al., 2016[32], Zhang et al., 2019[55]). Consequently, AT1R blockade has shown therapeutic promise in human lung cancer (Ni et al., 2020[31]), gastric cancer (Huang et al., 2008[16]), bladder cancer (Kosugi et al., 2006[24]), and prostate cancer (Alhusban et al., 2014[2]), including previous work hinting at its relevance in NPC (Wang et al., 2009[52]; Lin et al., 2021[26]). The initial finding of preferential cytotoxicity is crucial for clinical translation. CC exhibited selective toxicity toward NPC cell lines (mean IC_50_∼38μM) compared to the immortalized normal nasopharyngeal epithelial cells (NP69), which showed minimal cytotoxicity (estimated IC_50_>150μM). This yields a favorable Selectivity Index (SI) of >3.9 toward the malignant phenotype, confirming a significant margin of safety and strongly supporting CC's rational repurposing, leveraging its established systemic safety profile (Easthope and Jarvis, 2002[11]).

However, the potential application of CC at the bedside requires careful consideration of its pharmacokinetics. Following standard antihypertensive doses (8-16 mg/day), the active metabolite, Candesartan, achieves a relatively low systemic maximum plasma concentration (C_max_) of approximately 0.14 - 0.23 μM (Khawaja and Wilcox, 2011[20]). Although CC effectively suppressed NPC cell growth at concentrations below 40 μM in vitro, these effective concentrations are substantially higher than the systemic C_max_ attained in hypertension patients. This disparity can be mitigated by several translational strategies including 1) The high SI suggests that the tumor microenvironment can be preferentially targeted without undue systemic toxicity; 2) Solid tumors, including NPC, are often subject to the Enhanced Permeability and Retention (EPR) effect, which allows for local drug accumulation far exceeding systemic plasma levels (Matsumura and Maeda 1986[28]); 3) The primary clinical path suggested by our mechanistic data is combination therapy, which permits the use of lower, synergistic concentrations of CC to achieve therapeutic efficacy *in vivo *(Mokhtari et al., 2017[29]).

We dissected the mechanism by which CC impairs NPC progression. Functionally, CC severely reduced cell viability, proliferation, and long-term clonogenic survival. This suppression of growth was linked to a fundamental disruption of the cell cycle, specifically the induction of G0/G1 arrest, thereby limiting the critical S-phase entry and DNA synthesis. Furthermore, CC significantly impaired the migratory potential of NPC cells-a key factor driving metastasis. This anti-migratory effect was associated with the modulation of EMT-related proteins. We observed a partial EMT inhibitory phenotype, characterized by the suppression of mesenchymal markers (vimentin and Slug). However, the concomitant, sometimes paradoxical, changes in cadherin expression highlight the complexity of EMT regulation in NPC. Such "hybrid" EMT states are increasingly recognized as contributing to tumor plasticity, invasiveness, and therapy resistance (Usman et al., 2021[49]; Strouhalova et al., 2020[42]; Pal et al., 2021[34]). Our data suggest that CC influences EMT dynamics in a context-dependent manner, reflecting a complex rather than binary regulatory effect.

At the molecular level, CC's efficacy is linked to two crucial signaling events. First, CC led to a marked downregulation of AT1R expression and a significant attenuation of p38 MAPK phosphorylation. MAPK signaling is a recognized driver of NPC proliferation, invasion, and angiogenesis (Forrester et al., 2018[12]; Pua et al., 2022[36]; Wang et al., 2009[52]; Wada et al., 2017[50]; Leelahavanichkul et al., 2014[25]), suggesting that CC's inhibitory action on this pathway contributes substantially to its anti-cancer effects. Second, we observed a fascinating adaptive response: CC simultaneously enhanced AKT phosphorylation despite reducing total AKT protein levels. This is a common mechanism of stress-induced compensatory survival signaling utilized by cancer cells following targeted therapy. This critical observation not only reveals a mechanism of potential acquired resistance but also suggests a rational combination strategy: pairing CC with PI3K/AKT inhibitors could potentially overcome this compensatory mechanism, thereby enhancing overall therapeutic efficacy (Choi et al., 2014[7], 2016[6]).

The functional and molecular findings were strongly corroborated by MD simulations. The MD results confirmed the stable and energetically favorable binding of CC within the AT1R pocket. Importantly, CC engaged the same critical hotspot residues (Trp84, Ser105, Val108, Ser109, and Ile288) that are essential for the binding of ZD7155, a highly selective AT1R antagonist (Zhang et al., 2015[54]). This structural validation underscores the mechanism of direct AT1R blockade and provides atomic-level insights that can guide future rational optimization of CC derivatives for enhanced anticancer potency.

Repurposing existing drugs presents both promising opportunities and significant challenges in cancer treatment. CC, an antihypertensive agent with a well-established safety profile, effectively suppressed the growth and metastatic potential of NPC cells while exhibiting minimal cytotoxic effects on normal nasopharyngeal cells in vitro, suggesting a safety advantage. However, the potential application of CC in NPC therapy at the bedside requires careful consideration, particularly in light of its pharmacokinetic properties. The typical maximum plasma concentration (Cmax) achieved with standard therapeutic dosing (8-16 mg/day) is 62-143 ng/mL (approximately 0.14-0.23 μM). Although CC suppressed NPC cell growth at concentrations below 40 μM, and prolonged exposure to concentrations under 25 μM inhibited NPC colony formation in vitro, these effective concentrations are substantially higher than those attained in patients treated for hypertension. Therefore, translation of these anticancer effects into clinical practice may depend on strategies such as combination therapy, in which candesartan cilexetil is used alongside conventional treatments like chemotherapy. This approach could allow lower concentrations of both agents to be employed, potentially achieving synergistic anticancer effects in vivo.

In conclusion, the convergence of robust in vitro and rigorous in silico definitively establishes CC as a compelling and readily translatable repurposed therapeutic candidate for NPC. Our functional and mechanistic findings-which demonstrate CC's potent anti-proliferative effects, G0/G1 arrest, and AT1R-mediated signaling modulation-provide a solid foundation for clinical application. Crucially, the identification of the compensatory AKT activation points toward a rational combination strategy. Given CC's established clinical safety profile and favorable pharmacokinetics, its progression into translational studies is both scientifically feasible and clinically urgent. While our current findings are limited to cell-based and computational models, this work strongly advocates for the immediate preclinical evaluation in vivo, particularly focusing on combination regimens that co-target the AT1R/MAPK and the compensatory AKT signaling pathways, thereby paving the way for the expedited clinical repositioning of CC in this challenging malignancy.

Declaration

Acknowledgments

This study was supported by a grant from Faculty of Medicine, Thammasat University to CN and a grant for development of new faculty staff, Ratchadaphiseksomphot Endowment Fund, Chulalongkorn University to TJ and the NSRF via the Program Management Unit for Human Resources & Institutional Development, Research and Innovation [B38G680006] to TJ and US.

Declaration of conflict of interest

The authors declare that they have no conflict of interest.

Artificial Intelligence (AI)-assisted technology

During the preparation of this work the authors used AI to improve the readability and language of this work. AI was not used to generate scientific insights, draw conclusions, analyze data, or interpret data. After using this tool, the authors reviewed and edited the content as needed and take full responsibility for the content of the publication.

Authors' contribution

CN contributed to the conceptualization of the study, conducted the methodology and investigation, performed formal analysis and data curation, and was responsible for writing the original draft preparation. US contributed to the conceptualization of the study, the methodology, performed formal analysis, and contributed to writing - review and editing. SU contributed to the methodology and investigation and assisted with formal analysis and data curation. NP contributed to the methodology and investigation. TJ contributed to the conceptualization of the study, provided supervision for the entire project, and was responsible for writing - review and editing.