ZDHHC11-mediated palmitoylation of viral envelope protein restricts dengue virus replication: implication for the development of universal anti-flavivirus therapeutic strategies

Abstract

Palmitoylation is a key post-translational modification regulating viral replication, yet its regulatory mechanism in Dengue virus (DENV) infection remains elusive. This study aimed to elucidate the underlying regulatory function of palmitoylation and zinc finger Asp-His-His-Cys (ZDHHC) proteins in DENV replication and identify palmitoylated DENV proteins. We explored the function of palmitoylation in DENV replication using the palmitoylation inhibitor 2-bromopalmitate (2BP) and enhancer palmitic acid (PA), combined with qRT-PCR, western blot, confocal immunofluorescence microscopy, co-immunoprecipitation, and acyl-biotin exchange (ABE) assays. We found that 2BP promoted DENV replication, while PA inhibited it. Further analyses showed that 2BP and PA exerted no significant effects on DENV adsorption, internalization, or the interactions between E protein and the structural proteins prM/C that are essential for viral assembly. ABE assays verified that the DENV E protein is palmitoylated, with no such modification detected in prM and C proteins. Mass spectrometry and site-directed mutagenesis further revealed that the DENV E protein is palmitoylated at three cysteine residues: Cys74, Cys302, and Cys333. ZDHHC11, a characterized regulator of Zika virus E protein palmitoylation, directly interacted with DENV E protein and catalyzed its palmitoylation. Functionally, knockdown of ZDHHC11 enhanced DENV replication, while overexpression suppressed it. These findings deepen our understanding of flavivirus-host interaction mechanisms and provide a theoretical basis for the development of broad-spectrum anti-flavivirus therapies.

Supplementary Information

The online version contains supplementary material available at 10.1186/s12985-026-03180-8.

Introduction

Dengue virus (DENV) and Zika virus (ZIKV), both members of the Flavivirus genus in the Flaviviridae family, are globally important mosquito-borne pathogens that pose a persistent and severe threat to public health worldwide [1]. DENV is responsible for more than 390 million human infections each year, with clinical presentations spanning from self-limiting mild dengue fever to life-threatening complications including dengue hemorrhagic fever and dengue shock syndrome [2]. Genetically and antigenically, DENV is classified into four distinct serotypes (DENV‑1, DENV‑2, DENV‑3, and DENV‑4), which are distinguished by their surface antigen profiles; each serotype further contains multiple phylogenetically divergent genotypes. Notably, DENV‑2 has emerged as the predominant serotype driving recent dengue epidemics across South America and is the serotype most strongly associated with severe clinical outcomes of dengue infection [3]. ZIKV, meanwhile, has emerged as a major public health concern due to its confirmed association with neonatal microcephaly, Guillain-Barré syndrome, and male infertility [4]. Despite high disease burden, effective antiviral drugs for both viruses remain limited, and existing vaccines exhibit incomplete efficacy against multiple DENV serotypes or lack broad application for ZIKV [5, 6]. Elucidating conserved and unique molecular mechanisms governing flavivirus replication is therefore critical for developing targeted therapeutic strategies.

Post-translational modifications (PTMs) play pivotal roles in regulating viral lifecycle events, including entry, replication, assembly, and egress [7, 8]. Among these PTMs, palmitoylation-a reversible covalent attachment of palmitic acid to cysteine residues of target proteins, catalyzed by zinc finger Asp-His-His-Cys (ZDHHC) family palmitoyltransferases-has emerged as a key regulator of viral pathogenesis [9, 10]. Accumulating evidence underscores the pivotal role of palmitoylation in flavivirus biology, encompassing modifications of both viral proteins and host factors that collectively modulate virus-host interactions. For instance, the envelope (E) protein of ZIKV is subject to ZDHHC11-mediated palmitoylation at the conserved cysteine 308 residue. This modification confers an antiviral function in an enzyme activity-dependent manner: specifically, overexpression of ZDHHC11 potently suppresses ZIKV replication [11]. By contrast, palmitoylation exerts a pro-viral effect in Japanese encephalitis virus (JEV) infection: the JEV NS2A protein undergoes palmitoylation at the Cys221 residue, which has been validated to positively regulate viral replication kinetics and augment viral virulence in susceptible host models [12]. These data indicate that palmitoylation exerts diverse biological roles depending on the palmitoylated target proteins. On the host side, palmitoylation of interferon-inducible transmembrane proteins (IFITMs) in human and swine inhibits JEV infection, highlighting the evolutionary conservation of this host defensive mechanism conferred by IFITMs [13, 14].

Given the high sequence homology and conserved genomic organization between DENV and ZIKV—both including three structural proteins (capsid [C], membrane [prM], and envelope [E]) and seven non-structural proteins—it is plausible that palmitoylation of viral proteins can also regulate DENV replication [15, 16]. In this current study, we aimed to clarify [1] Whether DENV structural proteins, particularly the E protein (a functional homolog of ZIKV E), undergo palmitoylation following DENV infection; [2] Which host ZDHHC enzymes actually mediate DENV protein palmitoylation, and whether ZDHHC11 (a key player mediating E protein palmitoylation of ZIKV) plays a conserved role; [3] How palmitoylation process dynamically modulates DENV replication, and whether it exhibits the same antiviral function observed in ZIKV; and [4] Whether DENV infection alters the transcriptional profile of host palmitoyltransferases, and if such changes contribute to viral lifecycle regulation. While palmitoylation of viral and host proteins has been identified to be associated with viral replication of diverse viruses [17–20], the specific role of this PTM in DENV biology remains largely unexplored.

Based on the established role of palmitoylation in ZIKV infection [11], we aim to address these gaps by investigating the effect of palmitoylation on DENV replication, identifying palmitoylated DENV structural proteins, and characterizing host palmitoyltransferases involved. These findings will not only advance our understanding of conserved and distinct PTM-mediated regulatory mechanisms across flaviviruses but also provide potential therapeutic targets for combating DENV and ZIKV infections.

Materials and methods

Cell culture

The A549, Vero, and HEK293T cell lines were routinely preserved in our laboratory. All three cell lines were cultured in Dulbecco’s Modified Eagle’s Medium (DMEM; KeyGen Biotech, KGL1211-500, China) supplemented with 10% certified fetal bovine serum (FBS; vivacell, C04001-500, China) and 1% TransSafe Mycoplasma Prevention Reagent (Transgen Biotech, FM501-02, China). The cultures were maintained at 37 °C in a humidified incubator with 5% CO₂. The Aedes albopictus C6/36 cell line was generously provided by Professor Zhongtian Qi from The Second Military Medical University, China. These cells were cultured in RPMI-1640 medium (KeyGen Biotech, KGL1503-500, China) supplemented with 10% FBS and 1% TransSafe Mycoplasma Prevention Reagent, and incubated at 28 °C in a humidified 5% CO₂ atmosphere.

DENV virus infection, amplification

DENV-2 (New Guinea C strain, 16681-PDK53) was kindly provided by Professor Gong Cheng from Tsinghua University, China. The multiplicity of infection (MOI) for DENV was assessed using an immunofluorescence assay. For viral infection experiments, cells were infected with DENV at an MOI of 1. After incubation at 37 °C for 4 h, the viral inoculum was discarded. The cells were subsequently rinsed twice with phosphate-buffered saline (PBS; KeyGen Biotech, KGL2206-500, China), and fresh complete culture medium was supplemented.

To prepare DENV stock solutions, C6/36 cells were plated in RPMI-1640 medium containing 10% FBS and maintained at 28 °C in a humidified incubator with 5% CO₂. Following the removal of the original medium, the cells were washed twice with PBS prior to inoculation with a DENV dilution (MOI of 1), and then incubated at 28 °C for 4 h. The medium was then substituted with maintenance medium supplemented with 2% FBS, and the cells were cultured for an additional 7 days under the same environmental conditions. Afterwards, the culture supernatant was harvested and filtered through a 0.45 μm filter membrane to obtain the crude DENV stock.

ZIKV infection and propagation were performed as previously reported [21].

2-bromopalmitate (2BP), palmitic acid (PA) treatment and CCK-8 assay

A549 and Vero cells were seeded into 96-well plates at a density of 5×10⁴ cells/mL. On the following day, cells were treated with different concentrations of 2BP (Sigma, 238422-10G, USA) and PA (Sigma, 76119, USA), with dimethyl sulfoxide (DMSO, Solarbio, D8370, China) serving as the solvent control, and incubated for 24 h. Subsequently, the cells were infected with DENV (MOI of 1) and further incubated for 48 h.

Cell viability was determined using the CCK-8 kit (innibio, CD2001-050, China) in accordance with the manufacturer’s instructions. Briefly, 10 μL of CCK-8 detection solution was added to each well at the indicated time point, followed by incubation at 37 °C for 2 h. The absorbance at 450 nm was measured using a PerkinElmer EnSpire microplate reader (USA).

RNA isolation, reverse transcription, and real-time quantitative PCR (qRT-PCR)

Total cellular RNA was isolated using the TransZol Up Plus RNA Kit (TransGen Biotech, ER50101-V2, China) following the manufacturer’s recommended protocol. Complementary DNA (cDNA) was generated with the ReverTra Ace qPCR RT Master Mix (Toyobo, FSQ-101, Japan) according to standard operating procedures. For the quantification of mRNA expression levels, qRT-PCR assays were carried out using Novostar SYBR qPCR SuperMix Plus (Novoprotein, E096-01A, China) on the CFX96 Real-Time PCR System (Bio-Rad Laboratories, Hercules, CA, USA). The specific primers used for qRT-PCRamplification are provided in Supplementary Table S1. The relative mRNA expression levels of the target genes were determined via the 2^^−ΔΔCt method and normalized to GAPDH.

Small interfering RNAs (siRNAs), plasmids and transfection

SiRNAs targeting ZDHHC family genes, along with a non-targeting negative control (NC), were chemically synthesized by Suzhou Genepharma (Suzhou, China). SiRNAs transfection was performed using the siRNA-mate plus reagent (Genepharma, G04027, China), strictly following the manufacturer’s standard operating protocol. The specific nucleotide sequences of siRNAs are provided in Supplementary Table S2.

Mammalian expression plasmids encoding DENV E, C, and prM proteins (designated as pcDNA3.1-E, pcDNA3.1-C, and pcDNA3.1-prM, respectively) were constructed via routine molecular cloning techniques. Briefly, the full-length coding sequences of DENV E, C, and prM were amplified by polymerase chain reaction (PCR), using complementary DNA (cDNA) reverse-transcribed from DENV viral RNA as the template. The resulting PCR amplicons were then subcloned into the pcDNA3.1-Flag vector to generate the recombinant expression constructs. The pcDNA3.1(+)-ZDHHC11-HA plasmid was commercially synthesized by HaoGe Biotech (Shanghai, China). Site-directed mutagenesis plasmids targeting the wild-type (WT) E protein (C74S, C302S, C333S) were constructed by Sangon Biotech (Shanghai, China).

For plasmid transfection, HEK293T cells were seeded in 10 cm culture dishes at a density of 2 × 10⁶ cells per dish 24 h prior to transfection. The transfection mixture was prepared by combining 10 μg of target plasmid DNA and 20 μL of polyethyleneimine (PEI) in 1 mL of Opti-MEM reduced-serum medium (Gibco, 31985070, USA). The DNA-PEI mixture was incubated at room temperature for 15 min to allow complex formation, then gently added to the cultured cells. 48 h post-transfection, cells were harvested for subsequent experiments.

Western blot analysis

Total cellular proteins were extracted with RIPA lysis buffer (Beyotime, P0013C, China) supplemented with phenylmethylsulfonyl fluoride (PMSF, Solarbio, P0100, China). Protein concentrations were quantified using BCA Protein Assay Kit (KeyGen Biotech, KGB2101-500, China) per manufacturer’s protocol. Equal amounts of protein were separated by SDS-PAGE and transferred to PVDF membranes (MilliporeSigma, IPVH00010, USA) via Bio-Rad Trans-Blot® Turbo™ Transfer System (Bio-Rad, USA).

Membranes were blocked with Western Blocking Buffer (Proteintech, PR2034, USA) at room temperature for 15 min, then incubated overnight at 4 °C with primary antibodies: anti-GAPDH (Proteintech, 60004-1-ig, 1:50000); anti-DENV Envelope (GeneTex, GTX127277, 1:10000); anti-DENV Capsid (GeneTex, GTX639659, 1:1000); anti-DENV prM (GeneTex, GTX640942, 1:1000); anti-ZIKV NS3 (GeneTex, GTX133309, 1:4000); anti-Flag (Proteintech, 66008-4-ig, 1:10000); anti-HA (Sigma, H6908, 1:1000).

After three washes with TBST (10 min each), membranes were incubated with multi-rAb HRP-goat anti-rabbit/mouse recombinant secondary antibody (H+L) (Proteintech, RGAR001 for rabbit IgG; RGARM001 for mouse IgG; both 1:5000) at room temperature for 1 h. Protein bands were visualized using Bio-Rad ChemiDoc MP Imaging System (USA) and densitometrically analyzed with ImageJ software (NIH, USA). Relative protein levels were normalized to GAPDH.

Plaque assay

A549 cells were pretreated with 2BP (50 μM) or PA (100 μM) for 12 h, followed by ZIKV infection (MOI of 1) for 4 h. Cells were then cultured in fresh complete medium for another 48 h. For plaque assay, the harvested supernatants from different treatment groups were serially 10-fold diluted (10⁻¹ to 10⁻⁷) under the same dilution protocol, and the same viral dilution factor was selected for plaque formation detection and imaging to ensure comparability among groups.

For plaque assay, confluent Vero cell monolayers were inoculated with 20 μL of each viral dilution mixed with 1 mL serum-free DMEM, and incubated at 37 °C for 4 h. Following removal of the inoculum, cells were overlaid with 1 mL of 1.3% methylcellulose (Sigma-Aldrich, M0512, USA) dissolved in DMEM containing 4% FBS. After 5 days of incubation, cells were stained with crystal violet solution (Solarbio, C8470, China) for 15 min and rinsed with tap water. Plaques were counted manually, and viral titers were calculated as plaque-forming units per milliliter (PFU/mL).

DENV attachment and entry assay

For the attachment assay, DENV (MOI of 1) was incubated with pretreated cells at 4 °C for 4 h to block viral internalization and permit only surface attachment. Unbound virus was removed by three washes with ice-cold PBS, after which RNA was extracted from surface-adhered virions. DENV RNA levels were quantified via qRT-PCR and normalized to cellular GAPDH RNA.

For the entry assay, DENV (MOI of 1) was added to pretreated cells and incubated at 37 °C for 4 h to facilitate viral internalization. Non-internalized surface-bound virus was eliminated by treatment with 0.25% trypsin-EDTA at 37 °C for 5 min, followed by washing with complete medium to inactivate trypsin. RNA was extracted from intracellular virions, and DENV RNA levels were detected by qRT-PCR to determine the internalized virus load.

Immunofluorescence assay

Cells were washed three times with 0.1% PBST (3 min per wash), followed by fixation with 4% paraformaldehyde (Biosharp, BL539A, China) per well at room temperature for 30 min in the dark. After three washes with 0.1% PBST (5 min per wash), cells were permeabilized with 0.2% Triton X-100 at room temperature for 10 min, and then washed three times with 0.1% PBST (5 min per wash) again. Subsequently, cells were blocked with 5% goat serum for 1 h with gentle shaking. After discarding the blocking solution, anti-DENV Envelope or anti-Flag antibody (diluted 1:500) was added, followed by incubation at 4 °C overnight in the dark with gentle shaking. The primary antibody was recovered, and cells were then washed five times with 0.1% PBST (5 min per wash). Next, fluorescent rabbit or mouse secondary antibody (Abcam, ab150077, diluted 1:500) was added per well and incubated at room temperature for 1 h. After discarding the secondary antibody, cells were washed three to five times with 0.1% PBST (5 min per wash). Thereafter, cells were stained with DAPI solution (Solarbio, C0065, China) for 10 min, followed by three washes with PBS (10 min per wash). Images were captured using an Olympus CKX31 inverted microscope (Olympus Corporation, Tokyo, Japan), and confocal microscopy (LSM 900, ZEISS, Germany) was used for protein colocalization analysis.

Co-immunoprecipitation (Co-IP) assay

Approximately 2×10⁷ cells were harvested and lysed in RIPA lysis buffer supplemented with PMSF; the mixture was vortexed thoroughly, incubated on ice for 30 min, and centrifuged at 14000 rpm for 15 min at 4 °C, with the resulting supernatant collected as the cell lysate. A total of 40 μL of fully resuspended magnetic beads (MedChemExpress, HY-K0202-5, USA) were transferred to a 2 mL EP tube, washed three times with 400 μL of 0.5% PBST via magnetic separation and supernatant discarding, and then incubated with Flag antibody (diluted to 25 μg/mL in 0.5% PBST) on a rotating mixer at room temperature for 30 min. Following magnetic separation, the beads were collected and subjected to four additional washes with 400 μL of 0.5% PBST. The antibody-conjugated beads were resuspended in the prepared cell lysate supernatant, incubated on a rotating mixer at room temperature for 30 min, and subjected to magnetic separation to discard the supernatant; the beads were then washed four times with binding/washing buffer. Finally, the beads were resuspended in 100 μL of 1×SDS-PAGE loading buffer, heated at 100 °C for 10 min, and subjected to magnetic separation once more, with the collected supernatant analyzed by WB.

Acyl-biotin exchange (ABE) assay

HEK293T cells were seeded in 10 cm culture dishes and transfected with overexpression plasmids, followed by incubation for 48 h. Cells were harvested and lysed in RIPA buffer supplemented with PMSF on ice for 30 min, then centrifuged at 13000 rpm at 4 °C for 15 min. The supernatant was collected, with 100 μL reserved as the input fraction; 25 μg/mL anti-Flag antibody (Proteintech, 66008-4-ig, USA) and 40 μL protein A/G magnetic beads (MedChemExpress, HY-K0202-5, USA) were added to the remaining lysate. The samples were placed on a vertical rotator and incubated overnight at 4°C.

After IP, the EP tube was placed on a magnetic stand for 10 s to allow complete adsorption of the magnetic beads, and the supernatant was discarded. Subsequent steps-including washing with Solution 1, reduction with Solution 1 and Solution 2, blocking with Solution 1, Solution 4, and 0.1 M PMSF, splitting into +HAM and -HAM groups, reactions with Solution 5/Solution 6 and Solution 7, washing, and elution with 1× loading buffer-were performed strictly in accordance with the protocol of the Immuno-Precipitation-ABE Kit (Aimsmass, AM10314, China). The eluted protein samples were used for subsequent western blot analysis. Quantification of palmitoylation signals was normalized to IP total E protein.

Sample preparation and palmitoylation liquid chromatography-tandem mass spectrometry (LC-MS/MS)

Eight 10‑cm dishes of HEK293T cells were transfected with DENV E overexpression plasmid and harvested 48 h post‑transfection. Cells were washed with pre‑cold PBS, lysed in ST buffer (300 mM Tris‑HCl pH 7.6, 2% SDS), incubated at 95 °C for 5 min, and centrifuged at 14,000 rpm for 10 min to collect the supernatant. Target proteins were enriched by IP‑ABE, then reduced with DTT, alkylated with IAA in the dark for 45 min, and subjected to 10 K ultrafiltration with 8 M urea (UA) and 50 mM ammonium bicarbonate (ABC). Proteins were digested with 4 μg trypsin at 37 °C for 16 h. Tryptic peptides were collected, dried under vacuum, desalted with C18 ZipTips, and resuspended in 0.1% formic acid for LC‑MS/MS analysis using a Thermo Fisher LTQ Orbitrap ETD mass spectrometer coupled to an Easy‑nLC 1000 system. Raw MS files were processed with MaxQuant (v1.5.2.8) using trypsin as the protease with a maximum of two missed cleavages; precursor and fragment mass tolerances were set to 20 ppm and 4.5 ppm, respectively. Carbamidomethylation was set as a fixed modification, and methionine oxidation and asparagine/glutamine deamidation as variable modifications. The false discovery rate (FDR) for peptide and protein identification was set to 0.01. Proteins with an experimental/control intensity ratio >2 and p‑value<0.05 were considered statistically significant. Bioinformatic analyses were performed using DAVID (v6.8) and KEGG databases.

Statistical analysis

Statistical analyses and data visualization were performed using GraphPad Prism software (Version 8.0.2). Student’s unpaired two-tailed t-test was utilized for comparisons between two groups, with two-sided testing applied to ensure unbiased result interpretation. All datasets were derived from at least three independent biological replicates. Individual data points are displayed where appropriate to clarify data distribution, and data are expressed as mean ± standard deviation (SD) unless otherwise stated.

Results

Suppression of palmitoylation by 2BP promotes, while enhanced palmitoylation by PA inhibits, replication of DENV and ZIKV

To elucidate the regulatory role of palmitoylation in flavivirus replication, we first established robust in vitro DENV infection models in A549 and Vero cells. These two cell lines were inoculated with DENV and harvested at different time points. qRT-PCR was used to quantify viral genomic RNA copies, while western blot analysis detected viral E protein expression. Results showed that both viral RNA levels and E protein abundance increased in a time-dependent manner, confirming the successful establishment of DENV infection models with consistent replication kinetics (Fig. 1).

Fig. 1.

Establishment and validation of DENV infection cell models. (A, C) Cells were infected with DENV (MOI of 1) and collected at 0, 12, 24, 36, 48, 72, and 96 hours post-infection (hpi). Viral RNA was quantified by qRT-PCR and normalized to GAPDH. (B, D) Cells were infected with DENV (MOI of 1) and harvested at 0, 24, 48, 72, and 96 hpi. E protein levels were analyzed by western blot, with GAPDH as the loading control. Each experiment was performed with 3 independent biological replicates, and data are presented as the mean ± SD

To investigate whether palmitoylation modulates DENV replication, we first validated the cytotoxicity of palmitoylation-targeting molecules in both cell lines: A549 and Vero cells were pretreated with 2BP, a non-specific inhibitor of palmitoyltransferases, at gradient concentrations for 48 h. CCK-8 assays demonstrated that 2BP exerted no significant cytotoxicity at concentrations up to 50 μM, confirming this concentration range was safe for functional experiments (Figure S1A-B). Next, cells were pretreated with gradient concentrations of 2BP prior to DENV infection, and qRT-PCR analysis revealed that 2BP promoted DENV replication in a dose-dependent manner (Fig. 2A-B). Conversely, we treated cells with PA, a substrate of palmitoylation, to enhance palmitoylation. Cytotoxicity assays showed that PA at concentrations up to 100 μM did not affect cell viability (Figure S1C-D), and qRT-PCR indicated that PA inhibited DENV replication in a saturable manner, with maximal inhibition achieved at 25 μM and no further enhancement at higher concentrations. (Fig. 2C-D). Further qRT-PCR validation confirmed that PA could antagonize the pro-viral replication effect of 2BP, with this antagonistic effect displaying cell-type dependency that may be associated with the distinct innate immune capacities of A549 and Vero cells (Fig. 2E-F). Subsequent western blot (Fig. 3A-D) and immunofluorescence assay (Fig. 3E-H) experiments further corroborated these findings.

Fig. 2.

2BP promotes, while PA inhibits, DENV replication. (A, B) A549 (A) and Vero (B) cells were pretreated with 0, 2.5, 5, or 10 μM 2BP for 24 h before DENV (MOI of 1) infection. Intracellular DENV RNA levels were measured by qRT‑PCR at 48 hpi. (C, D) A549 (C) and Vero (D) cells were pretreated with 0, 25, 50, or 100 μM PA for 24 h prior to DENV (MOI of 1) infection. DENV RNA levels were quantified by qRT‑PCR at 48 hpi. (E, F) A549 (E) and Vero (F) cells were pretreated with 50 μM 2BP, 100 μM PA, or their combination for 24 h before DENV infection (MOI of 1). DENV RNA levels were analyzed by qRT‑PCR at 48 hpi. Data are shown as mean ± SD from three independent biological replicates. Statistical analysis was performed using an unpaired two-tailed Student’s t‑test. Significance is indicated as: ns, not significant (p>0.05); **p<0.01; ***p<0.001; ****p<0.0001

Fig. 3.

2BP enhances, while PA inhibits, DENV E protein expression. (A, B) A549 (A) and Vero (B) cells were pretreated with 50 μM 2BP, 100 μM PA, or their combination for 24 h before DENV infection (MOI of 1). E protein expression was detected by western blot at 48 hpi, with GAPDH as a loading control. (C, D) A549 (C) and Vero (D) cells were pretreated with 50 μM 2BP, 100 μM PA, or their combination for 24 h prior to DENV infection (MOI of 1). At 48 hpi, E protein was visualized by immunofluorescence staining (green), and nuclei were counterstained with DAPI (blue). Scale bar: 200 μm. Data represent mean ± SD from three independent biological replicates. Statistical significance was determined using an unpaired two-tailed Student’s t-test. Significance is indicated as: *p<0.05; **p<0.01; ***p<0.001; ****p<0.0001

Since ZIKV and DENV belong to the Flavivirus genus, we investigated whether the regulatory effects of 2BP and PA are conserved across flaviviruses. qRT-PCR and western blot results showed that 2BP promoted ZIKV replication while PA inhibited it, consistent with the effects observed in DENV (Figure S2A-D). Additionally, virus plaque assays confirmed that 2BP significantly increased viral particle production in the supernatant of ZIKV-infected cells, and PA effectively antagonized this effect (Figure S2E-F). Collectively, our data indicate that modulation of palmitoylation levels by 2BP and PA in host cells affects DENV and ZIKV replication: suppressed palmitoylation by 2BP promotes while enhanced palmitoylation by PA inhibits viral replication-an effect conserved across DENV and ZIKV. These findings identify palmitoylation as a key regulatory node in flavivirus infection, highlighting its potential as a target for anti-flavivirus therapeutics.

2BP and PA do not affect DENV adsorption, internalization, or E protein-prM/C interactions

To investigate the effects of 2BP and PA on the DENV life cycle, we first focused on the early entry steps, namely viral adsorption and internalization. qRT‑PCR was employed to quantify intracellular viral RNA levels under conditions that specifically assess these two steps: 4 °C for adsorption (a temperature that inhibits viral internalization) and 37 °C for internalization (physiological temperature allowing complete entry). The results showed no significant differences in intracellular viral RNA levels between 2BP/PA-treated groups and the untreated control group at either temperature (Fig. 4A-D). These findings indicate that 2BP and PA do not modulate the adsorption or internalization of DENV.

Fig. 4.

2BP and PA do not affect DENV adsorption, internalization, or E protein-prM/C interactions. (A-D) A549 and Vero cells were pretreated with 50 μM 2BP, 100 μM PA, or their combination for 24 h before DENV infection (MOI of 1). Viral RNA levels were quantified by qRT‑PCR after DENV adsorption at 4 °C for 4 h or internalization at 37 °C for 4 h. (E-H) Co‑IP assays were performed in HEK293T cells overexpressing DENV structural proteins to detect the interactions between E protein and prM or C protein under 2BP, PA, or DMSO treatment. Data are presented as mean ± SD from three independent biological replicates. Statistical significance was determined using an unpaired two-tailed Student’s t-test. Significance is indicated as: ns, not significant (p>0.05)

Next, we evaluated the impact of 2BP and PA on viral assembly using Co‑IP assays in HEK293T cells overexpressing DENV structural proteins. Specifically, we examined the interactions between the E protein and the two core structural proteins, prM and C, as these interactions are critical for the proper assembly of DENV virions. Compared with the control group, no significant changes in the binding efficiency between E protein and prM or C protein were detected in cells treated with 2BP or PA (Fig. 4E-H). This result suggests that pharmacological inhibition or modulation of palmitoylation by 2BP and PA does not disrupt the key protein-protein interactions required for DENV assembly.

Collectively, these experimental data demonstrate that 2BP and PA have no significant effects on the early entry steps (adsorption and internalization) of DENV, nor do they interfere with the core intermolecular interactions that are essential for viral assembly.

Expression profiling of ZDHHC family proteins in DENV infected A549 cells and their differential regulation on DENV replication

To characterize the role of ZDHHC family palmitoyltransferases in DENV replication, we first profiled the expression of 23 ZDHHC members in DENV infected versus uninfected A549 cells. This analysis revealed that DENV infection significantly altered the expression of most ZDHHC genes: ZDHHC1, ZDHHC2, ZDHHC4, ZDHHC5, ZDHHC6, ZDHHC7, ZDHHC9, ZDHHC11, ZDHHC13, ZDHHC16, ZDHHC17, ZDHHC19, ZDHHC20, ZDHHC21 were downregulated, while ZDHHC8, ZDHHC14, ZDHHC24 were upregulated, and a subset showed no significant change (Fig. 5A). We next validated the knockdown efficiency of gene-specific siRNAs targeting these differentially expressed ZDHHC genes, confirming robust target reduction across all tested candidates (Figure S3A). Using these validated siRNAs, subsequent functional assays revealed that ZDHHC family members play divergent roles in regulating DENV replication. Specifically, knockdown of ZDHHC1, 5, 6, 7, 8, 9, 11, 12, 14, 19, 20, 21, and 24 significantly elevated viral RNA levels—identifying this subset as intrinsic antiviral factors-whereas silencing other ZDHHC members exerted no measurable effect (Fig. 5B). Overexpression of this same ZDHHC subset, whose expression efficiency was validated by qRT-PCR (Figure S3B), led to significant inhibition of DENV replication (Fig. 5C). Taken together, these findings demonstrate that distinct ZDHHC family members differentially modulate DENV replication, highlighting specific palmitoyltransferases as critical mediators of viral infection and promising targets for antiviral strategies.

Fig. 5.

Expression profiling of ZDHHC family genes in DENV infected A549 cells and functional screening of ZDHHCs in regulating DENV replication. (A) A549 cells were infected with DENV or left mock-infected, and ZDHHC mRNA levels were analyzed by qRT‑PCR at 48 hpi, normalized to GAPDH. (B) A549 cells were transfected with ZDHHC‑specific siRNAs or non-targeting control siRNA (si‑NC), then infected with DENV. DENV RNA levels were quantified by qRT‑PCR at 48 hpi. (C) A549 cells were transfected with ZDHHC overexpression plasmids or empty vector, then infected with DENV. DENV RNA levels were measured by qRT‑PCR at 48 hpi. Data represent mean ± SD of three independent biological replicates. Statistical significance was determined using an unpaired two-tailed Student’s t-test. Significance is indicated as: ns, not significant (p>0.05); *p<0.05; **p<0.01; ***p<0.001; ****p<0.0001

The DENV E protein is palmitoylated

We first constructed recombinant plasmids encoding DENV structural proteins (E, prM, C) in the pcDNA3.1/flag vector, which were validated by agarose gel electrophoresis (Figure S4A, C, E) to confirm the presence of target inserts at expected molecular weights; subsequent western blot analysis in transfected HEK293T cells confirmed robust overexpression of each protein (Figure S4B, D, F). Using these validated plasmids, we then performed ABE assays to assess palmitoylation of the three structural proteins, revealing that only the E protein exhibited detectable palmitoylation, whereas C and prM showed no palmitoylation signal (Fig. 6A-C). Functional modulation experiments further demonstrated that treatment with the palmitoylation inhibitor 2BP significantly reduced E protein palmitoylation levels (Fig. 6D-E), while supplementation with the palmitoylation substrate PA significantly increased E protein palmitoylation (Fig. 6F-G), confirming that DENV E protein is palmitoylated and that this modification is dynamically regulated by palmitoylation modulators.

Fig. 6.

DENV E protein is palmitoylated and its modification is dynamically modulated by 2BP and PA. (A-C) HEK293T cells were transfected with Flag‑tagged DENV C, prM, or E expression plasmids. Protein palmitoylation was detected using ABE assays. (D, E) HEK293T cells expressing Flag‑tagged E protein were treated with 50 μM 2BP or DMSO. Palmitoylation levels were determined by ABE assay (D) and quantified by densitometry (E). Palmitoylation signals were quantified relative to the IP total E protein. (F, G) HEK293T cells expressing Flag‑tagged E protein were treated with 100 μM PA or DMSO. Palmitoylation was analyzed by ABE assay (F) and quantified by densitometry (G). Palmitoylation signals were quantified relative to the IP total E protein. Data are shown as mean ± SD from three independent biological replicates. Statistical significance was determined using an unpaired two-tailed Student’s t-test. Significance is indicated as: ***p<0.001; ****p<0.0001

ZDHHC11 interacted and palmitoylated DENV E protein

Previous studies have demonstrated that ZDHHC11 specifically interacts with the ZIKV E protein and modulates the viral life cycle by catalyzing its palmitoylation. Given that DENV and ZIKV are both members of the Flaviviridae family, their E proteins share high amino acid sequence homology and conserved spatial conformations. It is likely that palmitoylation of the flaviviral E protein is a pivotal post-translational modification that orchestrates viral assembly, maturation, and infectivity in flavivirus family. We therefore hypothesized that ZDHHC11 may also mediate the palmitoylation of the DENV E protein through a direct physical interaction. In support of this hypothesis, our preliminary qRT-PCR analysis revealed a significant downregulation of ZDHHC11 transcriptional levels in host cells following DENV infection; further functional validation confirmed that ZDHHC11 knockdown enhanced, while its overexpression inhibited, DENV replication in host cells. Based on these experimental clues, we further investigated the interaction between ZDHHC11 and the DENV E protein, as well as the catalytic effect of ZDHHC11 on DENV E protein palmitoylation.

To validate the physical interaction between ZDHHC11 and the DENV E protein, we performed exogenous co-transfection assays in HEK293T cells, with simultaneous overexpression of HA-tagged ZDHHC11 and Flag-tagged DENV E protein. Co-IP assays were performed to detect protein-protein interactions. A distinct, specific band corresponding to HA-ZDHHC11 was readily detected in the protein complexes immunoprecipitated with anti-Flag and anti-HA antibodies (Fig. 7A-B), confirming a direct physical interaction between ZDHHC11 and the DENV E protein. To further provide spatial context for this interaction, we performed confocal immunofluorescence microscopy, which revealed that ZDHHC11 and the DENV E protein colocalized extensively in the cytoplasm of transfected cells (Fig. 7C), supporting their functional association in the intracellular compartment.

Fig. 7.

ZDHHC11 interacted and palmitoylated DENV E protein. (A-B) CO‑IP assays were performed in HEK293T cells co‑transfected with Flag‑E and HA‑ZDHHC11 or empty HA vector. Immunoblotting confirmed the specific interaction. (C) HEK293T cells were co-transfected with HA-ZDHHC11 and Flag-DENV E plasmids. At 24 h post-transfection, cells were fixed and immunostained with anti-HA (green) and anti-Flag (red) antibodies, and nuclei were counterstained with DAPI (blue). Confocal microscopy was used to visualize the subcellular localization and colocalization of ZDHHC11 and DENV E protein. Scale bar: 10 μm. (D-E) HEK293T cells were co‑transfected with Flag‑E and HA‑ZDHHC11 or empty vector. E protein palmitoylation was detected by ABE assay. Palmitoylation signals were quantified relative to the IP total E protein. (F-G) HEK293T cells were transfected with ZDHHC11 siRNA or control siRNA, then re‑transfected with Flag‑E plasmid. Palmitoylation of E protein was examined by ABE assay. Palmitoylation signals were quantified relative to the IP total E protein. Data represent mean ± SD of three independent biological replicates. Statistical significance was determined using an unpaired two-tailed Student’s t-test. Significance is indicated as: ***p<0.001

To further verify the regulatory role of ZDHHC11 in DENV E protein palmitoylation, we performed ABE assays in HEK293T cells overexpressing HA-ZDHHC11 and Flag-DENV E protein. Quantitative analysis of the ABE assay data revealed that ZDHHC11 overexpression induced a statistically significant increase in the palmitoylation level of the DENV E protein relative to the control group (Fig. 7D-E). In contrast, ZDHHC11 knockdown led to a notable reduction in DENV E protein palmitoylation compared with the NC group (Fig. 7F-G). These results confirm that ZDHHC11 serves as a palmitoyltransferase that catalyzes the palmitoylation of the DENV E protein.

The DENV E protein is palmitoylated at Cys74, Cys302, Cys333

To identify the palmitoylation sites of the DENV E protein, we first performed LC-MS/MS analysis on DENV E protein purified from HEK293T cells. Three cysteine residues-Cys74, Cys302, and Cys333 were identified as candidate palmitoylation sites, with their corresponding peptide sequences verified (Fig. 8A). To further validate the palmitoylation sites identified by MS, we constructed serine-substitution mutants (C74S, C302S, and C333S) of the Flag-tagged DENV E protein and performed ABE assays in HEK293T cells. Immunoblot analysis revealed that, compared with the WT E protein, the C74S mutant exhibited a dramatic reduction in palmitoylation levels, while the C302S and C333S mutants displayed partial decreases (Fig. 8B-C). Collectively, these data demonstrate that Cys74 is the major palmitoylation site, and Cys302/Cys333 are secondary sites, within the DENV E protein.

Fig. 8.

The DENV E protein is palmitoylated at Cys74, Cys302, Cys333. (A) Identification of candidate palmitoylation sites in DENV E protein by mass spectrometry. Three cysteine residues-Cys74, Cys302, and Cys333-were identified as potential palmitoylation sites, with corresponding peptide sequences displayed. (B-C) Functional validation of palmitoylation sites. HEK293T cells were transfected with WT or cysteine‑to‑serine mutant (C74S, C302S, C333S) Flag‑E plasmids, and treated with or without the palmitate analog HAM. Palmitoylation levels were analyzed by ABE assay (B) and quantified (C). Data are presented as mean±SD from three independent biological replicates. Statistical significance was determined using an unpaired two-tailed Student’s t-test. Significance is indicated as: **p<0.01; ****p<0.0001

Discussion

Post-translational modifications (PTMs) of viral and host proteins are increasingly recognized as critical regulatory determinants of flavivirus replication, among which protein palmitoylation has emerged as a key modulator of virus-host interactions [22–24]. The present study advances current understanding in this field by demonstrating that palmitoylation exerts bidirectional regulatory effects on DENV and ZIKV replication. We identified and validated that the DENV E protein is palmitoylated at three specific cysteine residues-Cys74, Cys302, and Cys333-and further revealed distinct functional roles of ZDHHC family palmitoyltransferases in regulating flavivirus replication. Notably, we uncovered an unexpected functional overlap: ZDHHC11, a known palmitoyltransferase for the ZIKV E protein, also physically interacts with and mediates palmitoylation of the DENV E protein.

We further demonstrated that pharmacological manipulation of cellular palmitoylation using 2BP (a palmitoylation inhibitor) and PA (a palmitoylation enhancer) significantly altered DENV replication, yet exerted no detectable effects on viral adsorption or internalization. Similarly, these compounds did not perturb the interactions between the E protein and prM/C proteins, which are essential for DENV virion assembly. These findings refine our understanding of the functional landscape of palmitoylation during the DENV life cycle, placing its regulatory role downstream of early viral entry and core assembly events-in contrast to our original working hypothesis. Notably, the antagonistic effect of PA on 2BP-induced viral promotion shows cell-type dependency, which may be associated with the innate immune competence of the host cells. A549 cells harbor intact interferon signaling, whereas Vero cells lack type I IFN pathways, which may contribute to the distinct outcomes in response to palmitoylation modulation. These observations strongly imply that palmitoylation modulates DENV replication through other uncharacterized steps in the viral life cycle, meriting in-depth mechanistic exploration in future studies.

Our observation that global inhibition of palmitoylation by 2BP promotes, whereas supplementation with PA suppresses, the replication of both DENV and ZIKV challenges the oversimplified view that palmitoylation uniformly acts as a pro-viral modification. This finding is consistent with a recent study by Bu et al. (2024), who demonstrated that PA enhances antiviral immunity via ZDHHC24-mediated palmitoylation of MAVS-driving its aggregation, mitochondrial localization, and activation of the TBK1-IRF3-IFN pathway [25]. Mounting evidence reveals functionally divergent roles of palmitoylation across flaviviruses: palmitoylation of JEV NS2A acts as a pro-viral modification that promotes viral replication and virulence [12], whereas ZDHHC11-mediated palmitoylation of DENV and ZIKV E proteins exerts antiviral effects to restrict infection. This discrepancy suggests that palmitoylation may exert distinct functional roles depending on the targeted viral protein. Such context-dependent outcomes of palmitoylation are inherent to global modulation of cellular palmitoylation, which impacts a complex regulatory network involving the palmitoylation of both host antiviral factors (e.g., MAVS) and pro-viral host proteins, as well as viral proteins that exhibit divergent functional responses to this PTM. In our experimental system, 2BP likely enhances DENV and ZIKV replication primarily by attenuating palmitoylation of key host antiviral factors such as MAVS [25], whereas PA reinforces host antiviral defenses by boosting palmitoylation of host restriction factors. The conserved regulatory effects of 2BP and PA on both DENV and ZIKV further underscore palmitoylation as a conserved regulatory hub at the genus level for flaviviruses. Notably, APT2 has been identified as a negative regulator of MAVS palmitoylation, and its inhibitor ML349 potently enhances antiviral innate immune responses [25]. In this study, we established the DENV E protein as a bona fide palmitoylation substrate, although the precise functional consequences of this modification for DENV replication remain to be fully elucidated. This reinforces the translational potential of targeting palmitoylation-related enzymes-such as APT2 inhibitors-as a promising strategy for the development of broad-spectrum anti-flavivirus therapeutics.

The ZDHHC family of palmitoyltransferases, comprising 23 human members harboring the conserved Asp-His-His-Cys (DHHC) domain essential for catalytic activity, mediates complex virus-host interactions by regulating lipid modification-dependent innate immune responses, with individual enzymes exerting divergent pro-viral or anti-viral effects [9, 26, 27]. For instance, ZDHHC3 exerts an antiviral effect against porcine reproductive and respiratory syndrome virus (PRRSV) by catalyzing the palmitoylation of PRRSV nucleocapsid (N) protein at cysteine 90, which disrupts the Nsp9-N protein interaction and subsequent viral RNA synthesis [28]. In the context of antiviral immunity, multiple ZDHHCs mediate IFITM3 palmitoylation in a functionally redundant manner, while ZDHHC20 uniquely colocalizes with IFITM3 at lysosomes to potentiate its antiviral activity [29]. Notably, ZDHHC11 has been shown to mediate STING palmitoylation at the endoplasmic reticulum in zebrafish, activating the STING-IRF3-IFN-I signaling pathway to suppress viral replication [30] and specifically interacts with ZIKV E protein to catalyze its palmitoylation, thereby inhibiting ZIKV infection [11]. Beyond antiviral immunity, recent work by Han et al. has highlighted a critical role for ZDHHC9 in anti-parasitic defense: ZDHHC9 directly binds to and catalyzes the palmitoylation of TANK-binding kinase 1 (TBK1) at cysteine 292, stabilizing TBK1 and enhancing downstream IFN-I signaling to strengthen host resistance against Plasmodium yoelii N67 infection [31]. This finding expands the functional spectrum of ZDHHC-mediated palmitoylation, demonstrating its conserved role in regulating innate immune signaling across distinct pathogen types (viruses and parasites) and underscoring the context-dependent specificity of ZDHHC enzymes in modulating host defense pathways.

Against this backdrop, our findings uncover a highly conserved functional mode of ZDHHC11 in the regulation of DENV and ZIKV infection. Given the high amino acid sequence homology and conserved spatial conformation of DENV and ZIKV E proteins, our Co-IP assays and immunofluorescence confocal microscopy definitively confirmed a direct physical interaction and colocalization between ZDHHC11 and the DENV E protein, and subsequent ABE assays further validated that ZDHHC11 functions as a specific palmitoyltransferase to catalyze the palmitoylation of the DENV E protein. This conserved regulatory effect of ZDHHC11 on flaviviral E protein palmitoylation is underpinned by the molecular compatibility between ZDHHC11 and the flaviviral E protein, which is likely attributed to the conserved structural features and interactive domains on the E proteins of DENV and ZIKV that enable specific recognition and catalytic modification by ZDHHC11 [32, 33]. To further define the molecular basis of this modification, we subsequently performed high-resolution LC-MS/MS analysis and unambiguously identified Cys74, Cys302, and Cys333 as the endogenous palmitoylation sites within the DENV E protein. In line with this direct molecular interaction and catalytic activity, our functional validation data demonstrated that ZDHHC11 knockdown markedly enhances DENV replication efficiency in host cells while its overexpression exerts a robust inhibitory effect. These results collectively clarify that ZDHHC11-mediated palmitoylation of the DENV E protein constitutes the direct molecular mechanism underlying its anti-DENV activity, rather than an indirect regulatory effect mediated by targeting host cellular factors. This finding enriches our understanding of the evolutionarily conserved regulatory network between ZDHHC family palmitoyltransferases and flaviviruses, highlighting that ZDHHC11 has evolved a conserved ability to recognize and modify the E protein of flaviviruses represented by DENV and ZIKV through specific protein-protein interactions and catalytic palmitoylation. Moreover, the downregulation of ZDHHC11 transcription following DENV infection can be regarded as a viral adaptive strategy to counteract the host’s ZDHHC11-mediated antiviral defense: by reducing the expression of this key palmitoyltransferase, DENV attempts to attenuate the palmitoylation modification of its own E protein and thereby escape the inhibitory regulation of ZDHHC11 to promote its own replication. Notably, although our MS analysis identified Cys74 as a major palmitoylation site in DENV E protein, whether the homologous Cys74 residue in ZIKV E protein is also a functional palmitoylation site remains to be further experimentally verified. In addition, the functional impacts of individual palmitoylation site mutations on DENV replication and pathogenesis were not examined in the present study and warrant further investigation in future work. This conserved regulatory axis of ZDHHC11-flaviviral E protein palmitoylation adds a critical molecular layer to the intricate interplay between host palmitoylation machinery and flavivirus infection.

Our identification of ZDHHC1, 5, 6, 7, 8, 9, 11, 12, 14, 19, 20, 21, and 24 as antiviral factors establishes a foundational framework for dissecting the precise molecular mechanisms by which palmitoylation modulates DENV replication. Building on our key finding that ZDHHC11 mediates DENV E protein palmitoylation, future studies focused on mapping the direct substrates of the remaining ZDHHCs will further clarify how this family of enzymes regulates viral protein trafficking, assembly, and host immune signaling cascades-all key processes governing flavivirus infection.

A key breakthrough of our study is the identification of the DENV E protein as a palmitoylated substrate, whose modification status is dynamically regulated by 2BP and PA. The E protein is essential for viral entry, assembly, and membrane fusion, and palmitoylation likely modulates its subcellular localization, stability, or interaction with host factors [34–36]. Our findings suggest that protein palmitoylation may fine-tune viral replication: 2BP promotes DENV replication while PA inhibits it. Given that such regulatory effects are conserved between DENV and ZIKV, our findings indicate that E protein palmitoylation represents a promising target for broad-spectrum antiviral intervention.

While our study provides critical insights into palmitoylation-mediated regulation of flavivirus replication, it has several notable limitations. First, all experiments were performed in cells, and in vivo validation using relevant animal models is warranted to confirm the physiological relevance of our in vitro findings. Second, the molecular basis underlying the divergent interaction of ZDHHC11 with DENV versus ZIKV E proteins-such as distinct amino acid residues or structural motifs within the viral E proteins-requires further in-depth structural and functional characterization. Third, although we have confirmed that ZDHHC11 binds to and directly catalyzes the palmitoylation of the DENV E protein, and have also identified multiple ZDHHC family members with pro-viral or anti-viral roles in DENV replication, we have not yet established whether ZDHHC11 is the major acyltransferase mediating DENV E protein palmitoylation, or if other identified ZDHHCs function synergistically with ZDHHC11 in this process. Additionally, the deacylating enzymes (e.g., acyl protein thioesterases, ABHD family proteins) responsible for reversing DENV E protein palmitoylation and maintaining the dynamic balance of this modification remain entirely uncharacterized. Elucidating these upstream regulatory factors-including potential ZDHHC co-catalysts and all relevant deacylases-will be essential to reconstructing the complete palmitoylation regulatory network of the DENV E protein, and thereby providing more precise molecular targets for the development of anti-flaviviral interventions.

In conclusion, our study establishes palmitoylation as a conserved, bidirectional regulator of flavivirus replication, with the DENV E protein and ZDHHC family enzymes as key mediators. Notably, we demonstrate that ZDHHC11 regulates DENV E protein palmitoylation, highlighting functional divergence in ZDHHC-viral protein interactions across closely related flaviviruses. These findings not only deepen our understanding of flavivirus-host interactions but also highlight palmitoylation and ZDHHC family enzymes as viable targets for developing broad-spectrum anti-flavivirus therapeutics. Future studies focusing on the structural and functional consequences of E protein palmitoylation, the molecular basis of ZDHHC substrate specificity, and the development of ZDHHC-specific inhibitors will pave the way for novel antiviral strategies.

Conclusion

In summary, palmitoylation bidirectionally regulates DENV replication (2BP promotes, PA inhibits) without affecting viral adsorption, internalization, or E-prM/C interactions. The DENV E protein is palmitoylated at Cys74, Cys302, and Cys333, and ZDHHC11-known to regulate ZIKV E protein palmitoylation, binds to, colocalizes with, and catalyzes DENV E protein palmitoylation to suppress DENV replication. Targeting ZDHHC11-mediated E protein palmitoylation may be a promising strategy for broad-spectrum anti-flavivirus therapies.

Author contributions

Limin Chen conceived and designed the study, supervised the experimental strategy, and revised the manuscript’s theoretical framework critically. Mingshuang Lai and Lexin Zhang performed experiments, conducted statistical analyses, visualized data, interpreted key findings, and drafted the main part of the initial manuscript. Rongji Lai, Qin Du, and Jinsheng Liu assisted with data validation, figure refinement and resource provision. Shilin Li, Baoren He and Bin Li offered technical guidance on experimental methodologies. All authors reviewed, revised and approved the final manuscript, ensuring its scientific accuracy and logical consistency.

Funding

This work was supported by the Fundamental Research Funds for Central Universities, Peking Union Medical College (No.: 3332025144), the Youth Science Foundation of the Guangxi Natural Science Foundation (No.: 2024GXNSFBA010047), and the Sichuan Science and Technology Program (No.: 2025ZNSFSC0678).

Data availability

All relevant data are included in the article and its supplementary materials.

Declarations

Ethics approval and consent to participate

Not applicable.

Consent for publication

Not applicable.

Competing interests

The authors declare no competing interests.