The primary amino acid sequence of Homo sapiens Interleukin-29 was retrieved in FASTA format from the National Center for Biotechnology Information (NCBI, Accession No. NP_742152). The 20-amino-acid sequence of the GW-H1 peptide was obtained from the study reported by Chen et al. 20. The three-dimensional crystal structure of the human IL-29 receptor heterodimer, consisting of IL-28Rα and IL-10Rβ (PDB ID: 9BPU), was downloaded from the RCSB Protein Data Bank. To construct the chimeric molecule, the N-terminal signal peptide (residues 1–19) of IL-29 was excised, and the mature core domain (residues 20–200) was utilized. The GW-H1 peptide was fused to the C-terminus of IL-29 via a rigid alpha-helical linker, A(EAAAK)_nA with n = 5 (27 amino acids total). The rigid linker was incorporated between IL-29 and GW-H1 to maintain spatial separation between the two functional domains, prevent unfavorable steric interference, promote independent domain folding, and preserve the biological activities of both components 21,22. The amino acid sequences utilized in the chimeric construct are listed in Table 1.

The secondary structural characteristics of the IL-29–GW-H1 chimeric protein were predicted using the GOR IV web server by submitting the FASTA-formatted amino acid sequence 23. Conformational properties, including the percentage and distribution of alpha-helices, extended beta-strands, and random coils, as well as solvent accessibility and structural boundaries, were evaluated 24.

To predict the tertiary structure of the IL-29–GW-H1 fusion protein, three independent computational modeling platforms were employed: trRosetta, I-TASSER, and AlphaFold2 (ColabFold) 12. The top-ranked models generated by each server were comparatively evaluated based on their TM-score, C-score, and estimated root mean square deviation (RMSD). The stereochemical quality and structural reliability of the predicted 3D models were assessed using the SAVES v6.0 server, including ERRAT (which determines the overall quality factor), Verify3D (which evaluates the compatibility of the 3D atomic model with its 1D amino acid sequence), and PROCHECK (which generates the Ramachandran plot) 25,26. Model refinement was subsequently performed using the GalaxyRefine server to optimize side-chain conformations and backbone packing prior to molecular docking.

The physicochemical properties of the chimeric construct were calculated using the ExPASy ProtParam tool 27. Evaluated parameters included the total number and composition of amino acid residues, theoretical isoelectric point (pI), molecular weight, extinction coefficient, in vitro and in vivo estimated half-life, instability index, aliphatic index, and grand average of hydropathicity (GRAVY). In addition, the aqueous solubility profile of the fusion protein was evaluated using the Protein-Sol server.

The safety and immunogenic profile of the chimeric protein were evaluated by submitting its amino acid sequence to specialized web servers. Toxicity was predicted using the Neurosnap platform to identify any potentially toxic motifs. Antigenicity was evaluated using the VaxiJen v2.0 server, which assesses protective antigenicity based on physicochemical properties with a default threshold of 0.6 28. Allergenicity was predicted using AllerTOP v2.0, which examines sequence similarity against known allergenic proteins and epitopes 29.

The solubility and expression propensity of the chimeric construct in Escherichia coli were evaluated using the SoluProt server 37. SoluProt employs a gradient boosting machine learning algorithm trained on 11,436 target proteins from the TargetTrack database to predict soluble protein expression directly from sequence features 43.

Protein-protein docking between the IL-29–GW-H1 chimeric construct and the IL-28Rα/IL-10Rβ heterodimeric receptor complex was performed using the ClusPro 2.0 server 30. ClusPro utilizes a Fast Fourier Transform (FFT) correlation approach (PIPER algorithm) to sample billions of conformations, clusters the lowest-energy docked structures based on RMSD, and performs energy minimization on representative cluster centers. Water molecules and heteroatoms were removed from receptor coordinates prior to docking using the PyMOL Molecular Graphics System. The resulting docked complexes were analyzed for intermolecular interactions—including hydrogen bonds, salt bridges, non-bonded contacts, and buried surface areas—using PDBsum and PDBePISA 31. The binding free energy of the docked complex was calculated using the MM/GBSA module on the HawkDock server, and the dissociation constant (K_d) was predicted using the PRODIGY server 32.

The docked IL-29–GW-H1–IL-28Rα–IL-10Rβ complex was subjected to all-atom molecular dynamics (MD) simulations using the Desmond simulation package within the Schrödinger software suite 34. The complex was placed in an orthorhombic simulation box with periodic boundary conditions and solvated using the TIP3P water model. The system was neutralized and physiological ionic strength was established by adding 0.15 M NaCl counterions 33. The OPLS4 force field was applied. The system was energy-minimized to eliminate steric clashes and equilibrated through standard Desmond multi-step relaxation protocols under NVT and NPT ensembles at a stable temperature of 310 K and a pressure of 1.0 atm. The equilibrated system was then subjected to a 100 ns production MD simulation. Dynamic cross-correlation matrices (DCCM) were computed to analyze correlated and anti-correlated residue motions. Principal component analysis (PCA) was conducted on backbone trajectories, and the first two principal components (PC1 and PC2) were used to construct the 3D Free Energy Landscape (FEL) using the Boltzmann relation 34.

The codon-optimized IL-29 and IL-29–GW-H1 fusion genes were chemically synthesized by Twist Bioscience (San Francisco, CA, USA) with flanking NdeI and XhoI restriction sites and cloned into the pET29b(+) expression vector without a native stop codon, thereby incorporating a C-terminal 6×His tag to facilitate affinity purification. Target constructs were verified by restriction endonuclease digestion and PCR amplification with gene-specific primers. Verified plasmids were transformed into chemically competent E. coli BL21(DE3) cells via the CaCl₂ heat-shock method.

For expression of wild-type IL-29 and the recombinant IL-29–GW-H1 fusion protein, shake-flask fermentations were conducted in 2.5 L baffled flasks. Single transformant colonies were inoculated into 50 mL of Luria-Bertani (LB) broth containing 50 µg/mL kanamycin and cultivated at 37 °C for 8 h. Pre-cultures were transferred into 500 mL of Terrific Broth (TB) supplemented with kanamycin, incubated at 37 °C with shaking at 200 rpm until an optical density at 600 nm (OD₆₀₀) of 0.6–0.8 was reached, and induced with 1.0 mM isopropyl β-D-1-thiogalactopyranoside (IPTG) for 18 h. Cells were harvested by centrifugation at 6,000 × g for 20 min at 4 °C, and the wet cell biomass was stored at −20 °C until processing.

Harvested cell biomass was resuspended in lysis buffer (50 mM Tris-HCl pH 8.0, 5 mM EDTA, 1 mM PMSF) and disrupted using a constant cell disruptor (French press system) at 1.5 kbar. Insoluble inclusion bodies (IBs) were separated from soluble host proteins and cell debris by centrifugation at 12,000 × g for 30 min at 4 °C. The isolated IBs were washed twice with washing buffer (50 mM Tris-HCl pH 8.0, 100 mM NaCl, 1% Triton X-100) and solubilized in solubilization buffer (100 mM Tris-HCl pH 8.0, 6 M guanidine hydrochloride, 2 mM EDTA) with continuous stirring at room temperature for 1 h. Residual insoluble particulates were removed by centrifugation at 8,000 × g for 20 min at 4 °C. The denatured protein solution was refolded by dropwise dilution into refolding buffer (100 mM Tris-HCl pH 8.0, 0.5 M L-arginine, 8 mM EDTA) with gentle stirring at 20 °C for 24 h to promote native disulfide formation and proper tertiary folding.

Refolded protein solutions were desalted, concentrated, and buffer-exchanged by diafiltration using tangential flow filtration cassettes 35. Purification was carried out by immobilized metal affinity chromatography (IMAC) using a Ni-NTA Sepharose column on an ÄKTA Explorer FPLC system (GE Healthcare). The column was equilibrated with Binding Buffer A (20 mM sodium phosphate, 0.5 M NaCl, pH 7.4). After loading the refolded sample, non-specifically bound host proteins were removed by washing with 10–15 column volumes of Washing Buffer B (20 mM sodium phosphate, 0.5 M NaCl, 20 mM imidazole, pH 7.4). The target His-tagged proteins were eluted using Elution Buffer (20 mM sodium phosphate, 0.5 M NaCl, 0.5 M imidazole, pH 7.4). Eluted fractions were pooled, dialyzed against phosphate-buffered saline (PBS, pH 7.4), and stored at −80 °C.

The purity and molecular weight of purified IL-29 and IL-29–GW-H1 were assessed on 12% SDS-PAGE gels resolved at a constant voltage of 110 V for 150 min using a Hoefer SE260 vertical electrophoresis system. Gels were stained with Coomassie Brilliant Blue R-250 and destained to visualize protein bands. For Western blot confirmation, resolved proteins were electroblotted onto nitrocellulose membranes using a Hoefer Semiphor semi-dry transfer unit. Membranes were blocked with 5% (w/v) non-fat dry milk in Tris-buffered saline with 0.1% Tween-20 (TBST) for 1 h at room temperature, incubated with mouse anti-6×His primary antibody (1:1,000 dilution) for 1 h, washed thoroughly with TBST, and probed with horseradish peroxidase (HRP)-conjugated goat anti-mouse IgG secondary antibody (1:5,000 dilution). Immunoreactive bands were visualized using NBT/BCIP substrate solution.

Human hepatocellular carcinoma HepG2 cells (ATCC HB-8065) and non-tumorigenic human embryonic kidney HEK293 cells (ATCC CRL-1573) were cultured in Dulbecco's Modified Eagle's Medium (DMEM) supplemented with 10% fetal bovine serum (FBS), 100 U/mL penicillin, and 100 µg/mL streptomycin in a humidified incubator with 5% CO₂ at 37 °C. Cytotoxicity was determined using the 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide (MTT) assay. Cells were seeded into 96-well culture plates (1 × 10⁴ cells/well) and allowed to adhere overnight. Cells were then treated in triplicate with varying concentrations (5, 10, 15, 20, and 25 µg/mL) of purified wild-type IL-29 or chimeric IL-29–GW-H1 for 24 h. Untreated cells served as negative controls. Following incubation, 10 µL of MTT reagent (5 mg/mL in PBS) was added to each well and incubated for an additional 4 h at 37 °C. The culture medium was removed, and 100 µL of dimethyl sulfoxide (DMSO) was added to each well to solubilize the purple formazan crystals. Absorbance was measured at 570 nm with a reference wavelength of 630 nm using a microplate spectrophotometer. Percentage cell viability was calculated using the following equation:

Cell Viability (%) = [(A₅₇₀ − A₆₃₀)_treated / (A₅₇₀ − A₆₃₀)_control] × 100

All quantitative experimental assays were conducted independently in triplicate (n = 3), and data are presented as the mean ± standard deviation (SD). Statistical analyses were performed using GraphPad Prism version 7.0 (GraphPad Software, Inc., San Diego, CA, USA). Differences between treatment groups across various concentrations were evaluated using one-way Analysis of Variance (ANOVA) followed by Tukey's post-hoc honestly significant difference (HSD) test. Differences were considered statistically significant at p < 0.05 (*p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001).

To generate the chimeric antitumor protein, the primary sequence of mature human IL-29 was linked at its C-terminus to the synthetic cationic peptide GW-H1 via a rigid alpha-helical linker [A(EAAAK)₅A]. The resulting chimeric IL-29–GW-H1 construct consists of a single continuous polypeptide chain of 228 amino acids (Figure 1). The incorporation of the rigid helical linker ensures conformational separation between the cytokine and peptide domains, preventing steric hindrance and improper domain entanglement.

The secondary structure of the chimeric IL-29–GW-H1 protein was predicted using the GOR IV web server (Figure 2). The analysis revealed that the 228-residue polypeptide chain comprises 140 amino acids in alpha-helical conformations (61.40%), 7 residues in extended beta-strands (3.07%), and 81 residues in random coils (35.53%). The high alpha-helical content is consistent with the structural architecture of type III interferons and helical linkers, imparting enhanced thermodynamic and conformational stability to the chimeric protein 36.

Homology and ab initio modeling were performed using trRosetta, I-TASSER, and AlphaFold2 to predict the 3D tertiary structure of the IL-29–GW-H1 fusion protein. Comparative assessment of the top models indicated that trRosetta generated the most accurate structure, with a TM-score of 0.829 (Table 2). Stereochemical validation on the SAVES v6.0 server confirmed that the trRosetta model exhibited an initial ERRAT quality factor of 95.15% and placed 96% of residues in the most favored regions of the Ramachandran plot. Following structural refinement using GalaxyRefine, the optimized model displayed further improvements: 98.0% of residues fell within the most favored Ramachandran regions (Figure 3B), with no residues in disallowed regions, and the ERRAT overall quality score rose to 97.06%. ProSA-web analysis yielded an overall model Z-score of −8.09 (Figure 3C), which resides within the distribution typical of experimentally determined native protein structures of comparable size. The ProSA-web energy profile demonstrated consistently negative knowledge-based energy values across the sequence (Figure 3D), confirming overall structural stability.

The physicochemical properties of the chimeric IL-29–GW-H1 construct were computed using ExPASy ProtParam (Table 3). The 228-residue protein has a molecular weight of 24,729.45 Da and a total atomic count of 3,518. Because positively charged basic residues (Arg + Lys = 32) exceed negatively charged acidic residues (Asp + Glu = 23), the chimeric protein is basic, with a theoretical isoelectric point (pI) of 9.38. The instability index was calculated at 45.32, and the aliphatic index was 86.71, indicating high thermostability. The grand average of hydropathicity (GRAVY) was −0.365, confirming the hydrophilic nature of the construct and its favorable solubility in aqueous environments. The predicted half-life was estimated at 30 h in mammalian reticulocytes in vitro, >20 h in yeast in vivo, and >10 h in E. coli in vivo. The molar extinction coefficient was 26,720 M⁻¹ cm⁻¹ at 280 nm. Protein-Sol solubility analysis yielded a score of 0.677 (Figure 3E), well above the population average threshold of 0.45, confirming high predicted aqueous solubility.

Safety profiling conducted via Neurosnap and AllerTOP v2.0 predicted that the chimeric IL-29–GW-H1 protein is non-toxic (overall toxicity score: 0.295) and non-allergenic. Antigenicity evaluation via VaxiJen v2.0 yielded an antigenicity score of 0.50 at a 0.60 threshold, classifying the construct as non-antigenic and suggesting low risk of eliciting adverse neutralizing immunogenicity upon administration.

The SoluProt machine learning algorithm predicted a solubility score of 0.955 for the chimeric IL-29–GW-H1 protein in E. coli. Because values exceeding 0.50 indicate high soluble expression probability, this result predicted favorable expression efficiency in recombinant bacterial hosts.

Molecular docking between the chimeric IL-29–GW-H1 protein and the IL-28Rα/IL-10Rβ heterodimeric receptor was performed using ClusPro 2.0 (Figure 4). The top ten docking clusters generated by ClusPro 2.0 are summarized in Table 4. Cluster 0 contained the largest number of members (39 cluster members) and displayed the lowest energy score of −839.8 kcal/mol and a weighted score of −797.7, representing the most stable and thermodynamically favorable binding mode. Clusters 1 and 2 also exhibited favorable binding scores (−744.0 and −756.4, respectively). Cluster 0 was selected for detailed interfacial contact mapping and molecular dynamics simulations.

Intermolecular interactions at the receptor-ligand interface were evaluated using PDBsum and PDBePISA (Figure 5). Analysis of the docked ternary complex revealed 8 salt bridges, 28 hydrogen bonds, and extensive non-bonded van der Waals contacts across the interface. At the IL-10Rβ (Chain A)–IL-29–GW-H1 (Chain C) interface, 11 hydrogen bonds were formed by residues Tyr42, Val3, His177, Arg43, Glu179, Thr9, Asn133, Asp92, Asp67, Lys7, and Cys89, alongside 4 salt bridges contributed by Arg43, Glu179, Asp92, and Lys7, and 64 non-bonded contacts. At the IL-28Rα (Chain B)–IL-29–GW-H1 (Chain C) interface, 14 hydrogen bonds were contributed by Ser175, Thr181, Arg46, His177, Lys75, Gly1, Arg173, Asp71, Tyr73, Asn169, Gln64, Gln70, Gln126, and Ala119, with 4 salt bridges formed by Glu53, Arg124, Asp71, and Arg173, and 107 non-bonded contacts. Binding free energy evaluation using the HawkDock MM/GBSA module yielded a favorable negative ΔG of −116.31 kcal/mol, while PRODIGY predicted a dissociation constant (K_d) of 3.8 × 10⁻¹¹ M (ΔG = −14.2 kcal/mol), confirming high-affinity, stable receptor engagement.

To evaluate the structural stability, conformational compactness, and dynamic behavior of the IL-29–GW-H1–IL-28Rα–IL-10Rβ complex under physiological conditions, a 100 ns all-atom MD simulation was performed (Figures 6 and 7). Residue-level root mean square fluctuation (RMSF) analysis showed that the majority of residues exhibited low fluctuations (1–4 Å), reflecting conformational rigidity in core secondary structures and the receptor-binding interface (Figure 6B). Elevated fluctuations were confined to solvent-exposed terminal residues and loop regions around residues 200 and 410. Backbone root mean square deviation (RMSD) increased during the initial 0–15 ns equilibration phase as the static docked pose relaxed in solvent, after which it stabilized into a steady plateau fluctuating between 10 and 13 Å throughout the remainder of the 100 ns trajectory (Figure 6C). The radius of gyration (Rg) remained tightly constrained between 31.5 and 32.5 Å (Figure 6A), indicating that the complex preserved its compact tertiary fold without unfolding. Structural superposition of snapshots at 0 ns (blue), 50 ns (green), and 100 ns (red) confirmed that global domain rearrangements were minimal once equilibrium was reached (Figure 6D).

Dynamic cross-correlation matrix (DCCM) analysis revealed dominant regions of positive correlation across receptor domains and at the ligand-receptor interface (Figure 7B), demonstrating synchronized domain dynamics necessary for signaling complex stability. Principal component analysis (PCA) scatter plotting showed that conformational sampling along PC1 and PC2 clustered predominantly in low-energy regions (Figure 7A). The 3D Free Energy Landscape (FEL) constructed from PC1 and PC2 exhibited a single, deep global energy minimum basin surrounded by shallow local basins (Figure 7C), confirming that the docked complex resides in a thermodynamically stable state throughout the simulation trajectory.

The synthetic IL-29 and chimeric IL-29–GW-H1 genes were successfully cloned into the pET29b(+) expression vector under the control of the T7 promoter (Figure 8B). Restriction endonuclease digestion and PCR amplification confirmed the presence of the expected 684 bp insert (Figure 8A). The recombinant plasmids were transformed into E. coli BL21(DE3) for protein expression.

Shake-flask fermentations yielded 7.5 g/L and 8.0 g/L of wet cell biomass for wild-type IL-29 and chimeric IL-29–GW-H1, respectively. Following cell lysis via French press at 1.5 kbar, inclusion bodies (IBs) were recovered at 3.2 g/L (IL-29) and 3.0 g/L (IL-29–GW-H1). Solubilization in 6 M guanidine hydrochloride followed by controlled dilution in L-arginine-containing refolding buffer facilitated efficient protein renaturation (Table 5).

Refolded IL-29 and IL-29–GW-H1 proteins were purified using Ni-NTA affinity chromatography on an ÄKTA Explorer FPLC system (Figure 9C). Wild-type IL-29 was recovered at a purified yield of 71 ± 1.0 mg/L (overall yield: 25.6 ± 1.0%), and chimeric IL-29–GW-H1 was recovered at 63 ± 1.0 mg/L (overall yield: 22.2 ± 1.0%) (Table 5). SDS-PAGE analysis revealed single homogeneous bands at their expected molecular weights (~20 kDa for wild-type IL-29 and ~24 kDa for IL-29–GW-H1), confirming a purity exceeding 98% (Figure 9A). Western blot analysis using anti-6×His antibodies confirmed target protein identity (Figure 9B).

The cytotoxic activities of purified IL-29 and chimeric IL-29–GW-H1 were evaluated against HepG2 hepatoma cells and non-tumorigenic HEK293 embryonic kidney cells using the MTT assay (Figure 10). Both proteins demonstrated concentration-dependent inhibition of HepG2 cell viability over 24 h (Figure 10A, B). The chimeric IL-29–GW-H1 protein exhibited significantly enhanced cytotoxic potency against HepG2 cells compared to wild-type IL-29, with a lower half-maximal inhibitory concentration (IC₅₀ = 16.31 µg/mL vs. 20.14 µg/mL; p < 0.001). When tested against normal HEK293 cells (Figure 10C), both proteins demonstrated low toxicity: cell viability remained between 70% and 98% across concentrations of 5–20 µg/mL, and more than 60% of HEK293 cells remained viable even at the maximum concentration of 25 µg/mL. These results indicate that the chimeric IL-29–GW-H1 construct achieves superior antitumor efficacy against hepatocellular carcinoma cells while sparing non-malignant cells.

3D: Three-dimensional; ANOVA: Analysis of variance; ATCC: American Type Culture Collection; DCCM: Dynamic cross-correlation matrix; DMEM: Dulbecco's Modified Eagle's Medium; DMSO: Dimethyl sulfoxide; E. coli: Escherichia coli; EDTA: Ethylenediaminetetraacetic acid; ERRAT: Empirical structure quality factor; FBS: Fetal bovine serum; FEL: Free energy landscape; FPLC: Fast protein liquid chromatography; GOR: Garnier-Osguthorpe-Robson secondary structure prediction; GRAVY: Grand average of hydropathicity; GW-H1: Cationic antimicrobial peptide (20 aa); HCC: Hepatocellular carcinoma; HRP: Horseradish peroxidase; HSD: Honestly significant difference; IBs: Inclusion bodies; IC₅₀: Half-maximal inhibitory concentration; IFN: Interferon; IFN-λ: Interferon-lambda; IL-10Rβ: Interleukin-10 receptor subunit beta; IL-28Rα: Interleukin-28 receptor subunit alpha; IL-29: Interleukin-29; IMAC: Immobilized metal affinity chromatography; IPTG: Isopropyl β-D-1-thiogalactopyranoside; IRF9: Interferon regulatory factor 9; ISGF3: Interferon-stimulated gene factor 3; ISRE: Interferon-stimulated response element; JAK-STAT: Janus kinase/signal transducer and activator of transcription; K_d: Dissociation constant; kDa: Kilodalton; MD: Molecular dynamics; MM/GBSA: Molecular mechanics generalized Born surface area; MTT: 3-(4,5-Dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide; MW: Molecular weight; NBT/BCIP: Nitro-blue tetrazolium / 5-bromo-4-chloro-3'-indolyl phosphate; NCBI: National Center for Biotechnology Information; Ni-NTA: Nickel-nitrilotriacetic acid; OD₆₀₀: Optical density at 600 nm; PBS: Phosphate-buffered saline; PCA: Principal component analysis; PDB: Protein Data Bank; pI: Isoelectric point; PMSF: Phenylmethylsulfonyl fluoride; RC: Ramachandran plot; Rg: Radius of gyration; RMSD: Root mean square deviation; RMSF: Root mean square fluctuation; rpm: Revolutions per minute; SAVES: Structural Analysis and Verification Server; SD: Standard deviation; SDS-PAGE: Sodium dodecyl sulfate polyacrylamide gel electrophoresis; TB: Terrific Broth; TBST: Tris-buffered saline with Tween-20; TM-score: Template modeling score.

The authors express their sincere gratitude to all individuals, researchers, and technical staff at the Centre for Applied Molecular Biology (CAMB), the School of Biochemistry and Biotechnology, and the Centre of Excellence in Molecular Biology (CEMB), University of the Punjab, as well as the University of Lahore, for providing laboratory infrastructure, technical assistance, academic guidance, and constructive scientific discussions that contributed to this research.

HMR and HB conceived and designed the study. HMR, ON, HMR (Hafiz Muzzammel Rehman), and NA performed the in silico computational modeling, molecular docking, and molecular dynamics simulations. HMR, IG, AA, TR, AAs (Ariba Asif), and HK conducted recombinant gene cloning, bacterial fermentation, protein refolding, and Ni-NTA chromatographic purification. HMR, AA, TR, and HB performed the SDS-PAGE, Western blotting, and in vitro MTT cytotoxicity assays. HMR, ON, NA, and HB analyzed and interpreted the data. HMR, ON, and HB drafted the original manuscript. All authors critically reviewed, edited, and approved the final version of the manuscript.

None. The authors declare that no specific grants or financial support were received from any funding agency in the public, commercial, or not-for-profit sectors for this research.

All data generated or analyzed during this study, including structural models, docking coordinates, and cytotoxicity values, are included in this published article. Additional primary datasets are available from the corresponding author upon reasonable request.

Not applicable. This study did not involve human participants, clinical trials, or live vertebrate animals; all experimental procedures were performed using established commercial cell lines (HepG2 and HEK293) and computational simulations.

Not applicable.

None. The authors declare that no generative artificial intelligence (AI) or AI-assisted technologies were utilized in the writing, data analysis, or manuscript preparation process.

The authors declare that they have no competing interests, financial or otherwise, that could be perceived as influencing the work reported in this manuscript.