Cell Counting Kit-8 (CCK-8, AC11L054) was purchased from Life-iLab Biotech Co., Ltd. (Shanghai, China). The Annexin V/PI staining kit (FAK012.20) was purchased from NeoBioscience Technology Co, Ltd. (Shenzhen, China). JC-1 staining kit (AC13021) was sourced from Acmec Biochemical Technology Co., Ltd. (Shanghai, China). The BCA kit was purchased from Beyotime Biotechnology Co., Ltd. (Shanghai, China). Dulbecco’s Modified Eagle Medium (DMEM), Roswell Park Memorial Institute (RPMI) 1640 medium, fetal bovine serum (FBS) and trypsin were acquired from Gibco (USA). Anti-hMOF antibody (ab200660), anti-H4K16ac antibody (ab109463), anti-histone H4 antibody (ab7311), anti-procaspase-3 antibody (ab32150), and anti-cleaved caspase-3 antibody (ab32042) were purchased from Abcam (Waltham, MA, USA). Anti-$\beta$-actin antibody (K101527P) was purchased from Solarbio Science & Technology Co., Ltd. (Beijing, China). Horseradish peroxidase (HRP) conjugated goat anti-mouse antibody (LF101) and HRP conjugated goat anti-rabbit antibody (LF102) were purchased from Epizyme Biotech Co., Ltd. (Shanghai, China).

Human lung cancer A549 and H1299 cells were purchased from Procell Biotechnology (Wuhan, China) and cultured in DMEM supplemented with 10% FBS and 1% Penicillin-Streptomycin, under an atmosphere of 5% CO₂ at 37 °C. All cell lines were validated by STR profiling and tested negative for mycoplasma.

TELNs were isolated through multiple centrifugation steps and density gradient centrifugation as described earlier [14]. Fresh turmeric was washed three times with distilled water. The roots were then crushed in a juicer. Subsequently, the mixture was centrifuged at 1000 g for 10 minutes, followed by 3000 g for 20 minutes, and finally at 10,000 g for 30 minutes to collect the supernatant. The supernatant was filtered through a sterile 0.45 µm filter and placed in a fixed-angle rotor for centrifugation at 100,000 g for 1 h. The pellet was resuspended in 1 mL of PBS and subjected to sucrose gradient centrifugation at different concentrations (15%, 30%, 45%, and 60%) at a speed of 150,000 g for 2 h. The layer at the sucrose concentration of 30%–45% was collected, and an equal volume of PBS was added to wash away the sucrose. Finally, the mixture underwent centrifugation at 150,000 g for 1 h to collect the pellet, which was then resuspended in PBS as TELNs. The protein concentration of TELNs was measured with the bicinchoninic acid (BCA) assay. To verify that TELNs were successfully isolated, their structure was observed by transmission electron microscopy (TEM). Using a capillary pipette, one drop of the prepared suspension sample (5 µL of TELNs) was placed onto a copper grid coated with a support film. Excess liquid was absorbed with filter paper. Subsequently, a 3% phosphotungstic acid solution was added, and staining took place for 5 min. The staining solution was removed by filter paper, the sample was dried, and the morphology of TELNs was then examined and photographed by TEM (Jeol1230, JEOL Ltd., Peabody, MA, USA). The particle size distribution of TELNs was characterized by nanoparticle tracking analysis (NTA, PMX 110, ZetaView, Inning am Ammersee, Germany). Dynamic light scattering (DLS, ZS90, Malvern Panalytical Ltd., Worcestershire, UK) was utilized to measure the $\zeta$-potential of TELNs. To determine curcumin in TELNs, TELNs were added to 1 mL of acetonitrile and sonicated to disrupt membranes, followed by centrifugation at 3000 rpm for 30 minutes. The supernatant was adjusted to a final volume of 2 mL with methanol and filtered through a 0.22 µm membrane to obtain the test solution. Curcumin was detected by Liquid Chromatography-Mass Spectrometry (LC-MS, 1290-6470A, Agilent Technologies, Inc., Santa Clara, CA, USA): Mobile phase: 0.1% formic acid-acetonitrile (1:1, v/v), flow rate: 0.3 mL/min, column temperature: 40 °C.

PKH26-labeled TELNs were first prepared [15]. Briefly, 2 µM PKH67 dye and 1 mL Diluent C were mixed as a working solution. TELNs (50 µg) were then incubated with the working solution for 20 min at room temperature, and the labeled-TELNs were ultracentrifuged (100,000 g, 2 h) to remove free PKH26 dye. Subsequently, A549 cells (1 $\times$ 10⁵ cells per well) were treated with PKH26-TELNs for 1, 2, and 4 h at 37 °C. After incubation, cells were washed 3 times with PBS solution. For laser confocal microscopy, cells were fixed with 4% (w/v) paraformaldehyde solution for 30 min. DAPI was then used to stain the nucleus, and cells were photographed under a laser confocal scanning microscope (LSM 910, ZEISS Group, Oberkochen, Germany). For flow cytometry, cells were collected and fixed with paraformaldehyde solution (4%, w/v). The uptake rate of PKH26-TELNs was then analyzed by flow cytometry (CytoFLEX V0-B5-R3, Beckman Coulter Life Sciences, Indianapolis, IN, USA).

The cytotoxicity of TELNs against A549 cells was analyzed by CCK-8 assay [15]. Specifically, A549 cells (5 $\times$ 10³ cells per well) were treated with TELNs (0, 5, 10, 20, 50, 100, 200 µg/mL) for 24 h. Subsequently, a 10% dilution of CCK-8 solution was added to the cells, and the plate was incubated for 1 h. The absorbance was then measured at 450 nm to determine cell viability.

TELNs-induced apoptosis was assessed by Annexin V/PI staining [15]. Specifically, after incubation with TELNs (0, 10, 50, and 100 µg/mL) for 24 h, cells were stained with the kit for 15 min and then analyzed by flow cytometry. Annexin V⁺/PI^– cells were considered to be early apoptotic cells, while Annexin V⁺/PI⁺ cells were considered late apoptotic cells.

A549 cells were seeded into 6-well plates (2 $\times$ 10⁵ cells per well). After incubation with TELNs (0, 10, 50, and 100 µg/mL) for 24 h, cells were stained with JC-1 solution for 20 min in an incubator [16]. The JC-1 solution was then discarded, and the cells were washed three times with PBS before assessment by flow cytometry.

Female nude mice (6-week-old) with a BALB/c genetic background were obtained from Liaoning Changsheng Biotechnology Co., Ltd. All animal experiments were conducted according to the guidelines of the Experimental Animal Welfare Ethics Committee of Changchun University of Chinese Medicine (2025936). The mice underwent a one-week acclimatization period before the establishment of the A549 hormonal nude model. Mice were anesthetized with isoflurane (induction concentration 3%, maintenance concentration 1%). After confirmation of anesthesia, 200 µL of an A549 cell suspension (5 $\times$ 10⁶ cells/mL) was administered subcutaneously into each mouse [17]. Tumor size was measured by vernier caliper, and the tumor volume was calculated according to the formula V = L $\times$ S² $\times$ 0.52, where V is the volume, L is the longest diameter, and S is the shortest diameter. Based on preliminary experimental results, it was determined that each group required at least 5 animals. When the tumor volume reached 150 mm³, the animals were divided into 4 groups of 5 mice each: PBS (control), TELNs 100 µg, TELNs 250 µg, and TELNs 500 µg. The groups underwent intratumoral injections every three days for a total of five times with 100 µL PBS, 100 µg TELNs, 250 µg TELNs, and 500 µg TELNs, respectively. Body weights and tumor volumes were recorded every three days. At day 18, mice were placed in an anesthesia chamber and perfused with CO₂ (flow rate 5 L/min) for euthanization. After confirming immobility, absence of respiration, and dilated pupils, the delivery of CO₂ was ceased. Mice were observed for 2 minutes to confirm death. Finally, tumor tissues were removed and evaluated using H&E staining.

A SteadyPure Universal RNA extraction kit (AG21017, ACCURATE BIOTECHNOLOGY, Changsha) was used to obtain total RNA from the tumor and A549 cells. All-in-One cDNA Synthesis SuperMix (TransGen Biotech Co., Ltd., Beijing, China) was used for reverse transcription. The quantitative real-time PCR mixture was formed with SYBR Green qPCR Master Mix for the Real-Time PCR System (QuantStudio 3, Thermo Fisher Scientific, Waltham, MA, USA). Relative changes in histone acetyltransferase human males absent on the first (hMOF) and GAPDH (Glyceraldehyde-3-Phosphate Dehydrogenase) mRNA levels were then measured. The primers [18] are shown in Table 1.

A549 cells were transfected with siRNA oligonucleotides using Lipofectamine 2000 (Thermo Fisher Scientific, Waltham, MA, USA). The hMOF siRNA (5^′-GAAAGAGAUCUACCGCAATT-3^′) [19], Alexa Fluor 488 modified negative control siRNA (Cat no. 1027292), and unmodified negative control siRNA (Cat no. 1027280) were obtained from QIAGEN (USA). After transfection with siRNA, the cells were harvested for further analysis.

A549 cells were cultured in 6-well plates at a density of 5 $\times$ 10⁵ cells per well. When cells reached 100% confluence, the monolayers in each well were scraped with a 200 µL pipette tip and washed three times with PBS. The cells were then followed with different treatments. Images were captured using an inverted microscope (IXplore, Olympus, Hachioji-shi, Tokyo, Japan) at 0, 12, 24, and 48 hours. Wound area was measured using ImageJ Fiji (National Institutes of Health, Washington, DC, USA) to calculate cell migration rates.

(Ⅰ) Curcumin target search. (ⅰ) Single-drug search: the SMILES chemical formula for the active ingredient of curcumin (CID: 969516) was obtained from the PubChem database (https://pubchem.ncbi.nlm.nih.gov/). The results were imported into the Swiss Target Prediction database (https://www.swisstargetprediction.ch/) to identify the target, and the target name was converted to gene name in the UniProt protein database (https://www.uniprot.org/). (ii) Herbal search: the target (sequence number: MOL000090) was retrieved from the TCMSP online database (https://www.tcmsp-e.com/), using curcumin as the keyword in the Chemical name entry. Its structure file was downloaded and imported into the PharmMapper database (https://www.lilab-ecust.cn/pharmmapper/) to retrieve the target moiety (Job ID: 250112151602). Target names were then converted to gene names in the UniProt protein database (https://www.uniprot.org/). The results from (i) and (ii) were combined to obtain the curcumin targets. (Ⅱ) Target search in NSCLC. Using “non-small cell lung cancer” as the keyword, NSCLC-related targets were identified in the GeneCards (https://www.genecards.org/), OMIM (https://omim.org/), and Disgenet (https://disgenet.com) databases. The results were combined to remove duplicates and identify the targets. Finally, the intersecting targets for curcumin (Ⅰ) and NSCLC (Ⅱ) were determined.

The intersecting targets were imported into the STRING protein interactions database (http://string-db.org/), with the species set to homo sapiens and the Interaction score set to 0.400. The interaction proteins were screened, and the PPI network was constructed, with the results visualized using Cytoscape 3.10.3 software.

(I) Intersecting target proteins were mapped to each term of the Gene Ontology (GO) database (http://www.geneontology.org/), and the protein number in each term was determined. GO terms that satisfied a p-value 0.05 were selected as terms that were enriched in differential proteins. The function of these proteins was analyzed in three sections: cellular components, molecular function, and biological processes.

(II) The intersecting targets were mapped to pathways using the Kyoto Encyclopedia of Genes and Genomes (KEGG) database (https://www.genome.jp/kegg/). Significantly enriched KEGG pathways were identified as having a p-value 0.05.

Curcumin and hMOF protein were selected for molecular docking analysis. The 3D structure for the active ingredient of curcumin (CID: 969516) was obtained from the PubChem database (https://pubchem.ncbi.nlm.nih.gov/). and that of hMOF protein (ID: pdb_00004dnc) from the PDB database (http://www.rcsb.org/). Water and ligand were removed from the receptor with PyMOL 3.1 (Schrödinger Inc., New York City, NY, USA), then converted to pdbqt format using Auto Dock 4.2.6 (Scripps Research Institute, Claremont, CA, USA). Appropriate positions were selected for docking. Binding energies were calculated and the results were visualized using PyMOL 3.1 (Schrödinger Inc., New York City, NY, USA).

Cells were lysed using RIPA lysis buffer (89900, Thermo Fisher Scientific, Waltham, MA, USA) containing a 1% protease inhibitor mixture (87785, Thermo Fisher Scientific, Waltham, MA, USA). The protein concentration was subsequently determined using a BCA protein assay kit (P0012, Beyotime Biotechnology Co., Ltd., Shanghai, China). Protein samples were separated by 10% sodium dodecyl sulfate polyacrylamide gel electrophores (SDS-PAGE) gel electrophoresis (XP00100BOX, Thermo Fisher Scientific, Waltham, MA, USA) and transferred to polyvinylidene fluoride (PVDF) membranes (Millipore). The membrane was then blocked for 1 h in Tris-buffered saline (TBS) containing 5% skim milk at room temperature, followed by overnight incubation at 4 °C with the primary antibody: Anti-hMOF antibody (1:1000; ab200660, Abcam), anti-H4K16ac antibody (1:1000; ab109463, Abcam), anti-histone H4 antibody (1:1000; ab7311, Abcam), anti-procaspase-3 antibody (1:1000; ab32150, Abcam), anti-cleaved caspase-3 antibody (1:1000; ab32042, Abcam), anti-$\beta$-actin antibody (1:1000; K101527P, Solarbio). After three washes, the membrane was incubated at room temperature for 45 minutes with horseradish peroxidase (HRP)-labeled secondary antibody: HRP conjugated goat anti-mouse antibody (1:1000; LF101, Epizyme Biotech), HRP conjugated goat anti-rabbit antibody (1:1000; LF102, Epizyme Biotech). The PVDF membrane was developed with a chemiluminescence detection analysis system (iBright CL750, Thermo Fisher Scientific, Waltham, MA, USA), and the images were then analyzed using ImageJ software.

Results are presented as the mean $\pm$ SD. To ensure the reproducibility of results, experiments were repeated three times. Differences were analyzed using one-way ANOVA with Tukey’s multiple comparisons test. p-values 0.05 were considered statistically significant, with p 0.01 and p 0.001 indicating highly and extremely significant differences, respectively.

To obtain TELNs, turmeric was juiced and then subjected to differential centrifugations and density gradient centrifugation (Fig. 1A). The physical properties of the TELNs were subsequently characterized. The size distribution of TELNs was measured using nanoparticle tracking analysis (NTA). TELNs showed a particle size of 157.2 nm (Fig. 1B), and their $\zeta$-potential was –15.3 mV (Fig. 1C). TEM revealed that TELNs exhibited an exosome-like morphology, characterized by a bilayer membrane vesicle (Fig. 1D). TELNs demonstrated stability, with no significant change in diameter over a 7-day period (Supplementary Fig. 1). TELNs were found to be enriched with higher concentrations of curcumin compared to turmeric juice (Fig. 1E; Supplementary Fig. 2). The release profile of TELNs indicated that 79.93% of curcumin was released within 24 hours (Supplementary Fig. 3). The above results demonstrate the successful isolation of TELNs from plant tissue. To evaluate the cellular uptake efficiency, PKH26-labeled TELNs were incubated with A549 cells. Confocal laser scanning microscopy images revealed that TELNs were taken up by A549 cells, with a significant effect observed at 4 h (Fig. 1F; Supplementary Fig. 4). Quantitative flow cytometry results also supported this conclusion. At 4 h, 69.5% of cells were positive, demonstrating that TELNs could effectively enter A549 cells (Fig. 1G,H). In summary, TELNs were successfully isolated from turmeric. Moreover, they exhibit exosome-like characteristics and can be internalized by NSCLC cells.

Fig. 1.

Extraction, characterization, and uptake analysis of turmeric-derived exosome-like nanoparticles. (A) Procedure for extracting TELNs. (B) Nanoparticle tracking analysis (NTA) of TELNs, showing a particle size of 157.2 nm. (C) Zeta potential analysis of TELNs, indicating a $\zeta$-potential of –15.3 mV. (D) Transmission electron microscopy (TEM) image of TELN, scale bar: 100 nm. (E) Detection of curcumin in TELNs by Liquid Chromatography-Mass Spectrometry. (F) Cellular uptake of TELNs by A549 cells after 1, 2, and 4 h, as detected by laser scanning confocal microscopy (red: PKH26-labelled TELNs; blue: nuclei staining; scale bar: 100 µm). (G) Cellular uptake of TELNs by A549 cells at 1, 2, and 4 h, as measured by flow cytometry. (H) Quantitative analysis by flow cytometry (n = 3). TELNs, turmeric-derived exosome-like nanoparticles; DAPI, 4^′,6-diamidino-2-phenylindole.

We first assessed the effect of TELNs on apoptosis in NSCLC. Flow cytometry analysis revealed the apoptosis rate of A549 cells treated with 100 µg/mL TELNs was 4.54–fold higher compared to the control group, with late-stage apoptosis increasing from 1.02% to 29.60% (Fig. 2A,C). Additionally, 100 µg/mL TELNs increased the apoptosis rate of H1299 cells by 45.46% (Supplementary Fig. 5). Consistent with these observations, CCK-8 assays revealed dose-dependent anti-NSCLC activity of TELNs, with a 45.25% reduction in cell viability at 100 µg/mL (Fig. 2E). The same phenomenon was also observed in H1299 cells, where TELNs demonstrated anticancer activity. At a dose of 100 µg/mL, the viability of H1299 cells decreased by 44.67% (Supplementary Fig. 6). The JC-1 probe is commonly used to detect the membrane potential of mitochondria, which is closely related to apoptosis. When apoptosis occurs, the mitochondrial membrane is destroyed, and JC-1 changes from an aggregated state to a dispersed state. Thus, enhanced green fluorescence (JC-1 monomers) implies increased apoptosis. TELNs significantly induced mitochondria-related apoptosis in A549 cells (Fig. 2B,D). The expression of cleaved caspase-3 protein was examined by Western blot assay to corroborate the presence of apoptosis. As a crucial executor of apoptosis, caspase-3 translocates from the mitochondria and induces apoptosis. TELNs significantly increased the ratio of cleaved caspase-3 to pro caspase-3 in A549 cells (Fig. 2F,G). Similarly, TELNs also increased the ratio in H1299 cells (Supplementary Fig. 7). Additionally, real-time quantitative PCR revealed that hMOF mRNA expression increased 2.01–fold in A549 cells (Fig. 2H) and 1.82–fold in H1299 cells (Supplementary Fig. 8) following treatment with TELNs. Collectively, these findings establish TELNs as inducers of apoptosis in A549 cells, with hMOF being a critical factor.

Fig. 2.

Turmeric-derived exosome-like nanoparticles induce apoptosis in non-small cell lung cancer. (A) Apoptosis of TELN-treated A549 cells, as measured by flow cytometry. (B) Mitochondrial membrane potential of A549 cells, as detected by flow cytometry. (C) Apoptosis rate of A549 cells (n = 3). (D) Ratio of JC-1 aggregate to JC-1 monomer (n = 3). (E) Cytotoxicity of TELNs against A549 cells, as measured by CCK-8 assay (n = 3), ns indicates that the results are not significant. (F) Western blot analysis of pro and cleaved caspase-3 expression in A549 cells after treatment. (G) The ratio of cleaved caspase-3 to pro caspase-3 (n = 3). (H) hMOF mRNA levels analyzed by real-time PCR (n = 3). JC-1, 5,5,6,6^′-tetrachloro-1,1^′,3,3^′ tetraethylbenzimi-dazoylcarbocyanine iodide; CCK-8, Cell Counting Kit-8; hMOF, human males absent on first; PCR, Polymerase Chain Reaction.

An A549 xenograft nude mouse model was established to evaluate the antitumor efficacy of TELNs in vivo. Compared to the control group, treatment with TELNs resulted in a 58.1% reduction in tumor volume, confirming its potent growth-inhibitory effects (Fig. 3A,B). In tests conducted every three days, treatment with TELNs significantly slowed tumor growth (Fig. 3C). H&E staining of the tumor tissue revealed distinct morphological differences, with PBS-treated tumors displaying uniform cell size, whereas TELN-treated tumors exhibited marked cellular disorganization (Fig. 3D). TUNEL staining showed a significant increase in apoptosis after treatment with TELNs (Fig. 3E,F). Since hMOF is known to catalyze histone H4K16 acetylation (H4K16ac), downstream epigenetic modulation was also examined.

Fig. 3.

Turmeric-derived exosome-like nanoparticles inhibit non-small cell lung cancer progression in vivo. (A) Images of tumors from mice after treatment with TELNs (n = 5). (B) Tumor weight after treatment (n = 5). (C) Changes in tumor volume over time (n = 5). (D) H&E staining images of tumor tissue (scale bar: 100 µm; red arrows indicate the cellular disorganization). (E) TUNEL staining of tumor (green: TUNEL; blue: nuclei staining; scale bar: 100 µm). (F) Quantitative analysis of fluorescence intensity (n = 3). (G) hMOF and H4K16ac levels in tumor detected by Western blot assay. (H) The expression of hMOF quantified by ImageJ software (n = 3). (I) The expression of H4K16ac quantified by ImageJ software (n = 3). TUNEL, Terminal deoxynucleotidyl transferase dUTP nick end labeling; H4K16ac, histone H4K16 acetylation.

Western blot analysis of tumor tissues revealed that TELNs significantly increased the levels of hMOF and H4K16ac (Fig. 3G–I). This suggested that hMOf and H4K16ac may serve as the target through which TELNs promote apoptosis in NSCLC. These findings demonstrate that TELNs effectively inhibited NSCLC progression and upregulated hMOF and H4K16ac in vivo.

Building upon previous studies examining the efficacy of TELNs in NSCLC, we conducted further investigations to determine whether its activity depends on the induction of hMOF. RNA interference was employed as a knockdown method, and the AF488-labeled Negative Control demonstrated an siRNA transfection efficiency of 80.4% for this approach (Supplementary Fig. 9). sihMOF effectively reduced hMOF expression by 86.22% (Supplementary Fig. 10). hMOF knockdown substantially reduced JC-1 monomer formation, suggesting impaired apoptosis. Notably, treatment with TELNs partially rescued this effect, demonstrating the role of TELNs in hMOF regulation (Fig. 4A,B). Consistent with the above observations, direct assessment of apoptosis rates revealed that silencing of hMOF decreased cell death by 54.28%, while treatment with TELNs restored apoptosis by 13.13% (Fig. 4C,D). Furthermore, wound healing assays were performed to analyze the effect of TELNs on cell migration. Scratch closure rates were quantified at 12, 24, and 48 h. Knockdown of hMOF enhanced A549 cell migration by 31.70% in 48 h, whereas TELNs partially counteracted the pro-migration effects induced by hMOF downregulation (Fig. 4E,F). These findings indicate that TELNs inhibit NSCLC cell migration by upregulating hMOF expression. Western blot assay further demonstrated suppression of both hMOF and H4K16ac expression after hMOF knockdown. Crucially, TELNs restored H4K16ac levels, consistent with their ability to upregulate hMOF (Fig. 4G–I). Collectively, these data establish the hMOF-H4K16ac axis as a key pathway through which TELNs exert their suppressive effects on NSCLC.

Fig. 4.

Turmeric-derived exosome-like nanoparticles inhibit non-small cell lung cancer by upregulating hMOF expression and histone H4K16 acetylation. (A) Mitochondrial membrane potential of A549 cells, as detected by flow cytometry. (B) Ratio of JC-1 aggregate to JC-1 monomer (n = 3). (C) Apoptosis of A549 cells, as measured by flow cytometry. (D) Early and late apoptosis rates of different groups (n = 3). (E) Wound healing assay for A549 cells observed under a microscope at 12 h, 24 h, and 48 h (scale bar: 200 µm). (F) Analysis of cell migration rate (n = 3). (G) Effect of TELNs on hMOF and H4K16ac levels in A549 cells, as detected by Western blot. (H) The expression of hMOF quantified by ImageJ software (n = 3). (I) The expression of H4K16ac quantified by ImageJ software (n = 3). “+”/ “-” indicate whether cells were treated with siRNA or TELNs. siNC, Negative Control siRNA; sihMOF, hMOF siRNA; FITC, Fluorescein isothiocyanate; H4, Histone H4.

As expected, TELNs possess multiple properties of turmeric. In a previous study, Wei et al. [20] demonstrated that TELNs contain a high concentration of curcumin. Curcumin is a natural compound extracted from turmeric that exhibits anticancer properties through its impact on epigenetic modifications [21, 22]. However, its regulatory role in hMOF remains unclear. Therefore, we hypothesized that epigenetic regulation by TELNs is related to the curcumin contained within them. We conducted a systematic bioinformatics analysis to elucidate the molecular mechanism underlying this effect. First, the chemical structure of curcumin (Fig. 5A) was used to search for interacting proteins, which were then compared with NSCLC-related proteins. The database identified 324 curcumin target proteins and 3782 NSCLC-associated proteins, with 191 overlapping targets highlighted by a Venn map (Fig. 5B). The protein-protein interaction (PPI) network was performed using the STRING database (Supplementary Fig. 11), with parameters set to Homo sapiens and the interaction score at 0.400. The network comprised 191 nodes and 920 edges, with an average node degree of 9.63, indicating extensive connections among the proteins. KEGG pathway enrichment analysis revealed these proteins were highly associated with cancer progression, VEGF signaling, PD-L1 immune checkpoint regulation, and the development of NSCLC (Supplementary Fig. 12). Furthermore, GO enrichment analysis (Fig. 5C–E) categorized the overlapping proteins into three functional subsets: (ⅰ) cellular component (CC): localized in the cytoplasm and nucleus; (ⅱ) biological process (BP): involved in histone remodeling, apoptosis, and response to reactive oxygen species (ROS); (ⅲ) molecular function (MF): enriched in DNA binding, histone H4K16 acetylation, and protease activity. The overlapping targets were visualized using Cytoscape software. This revealed significant enrichment of KAT8, KAT6A, KAT6B, KAT5, and KAT7 (all members of the hMOF histone acetyltransferase family), suggesting the anti-tumor activity of curcumin was closely linked to the regulation of histone H4K16 acetylation (Fig. 5F). We next investigated whether curcumin directly interacts with hMOF to modulate its function. Molecular docking analysis was performed using PDB software to simulate the binding between curcumin and KAT8. The calculated binding energy of –6.7 kcal/mol (significantly below the threshold of –5 kcal/mol) strongly supports a stable interaction between curcumin and hMOF (Fig. 5G). Collectively, these results demonstrate that curcumin in TELNs upregulates hMOF expression and H4K16ac in A549 cells. Mechanistically, molecular docking suggested curcumin may interact with hMOF. In brief, curcumin in TELNs promotes H4K16 acetylation by activating hMOF, thereby promoting apoptosis and inhibiting NSCLC. This reveals a new epigenetic regulatory mechanism.

Fig. 5.

Epigenetic regulation by turmeric-derived exosome-like nanoparticles is due to curcumin. (A) Chemical structure of curcumin. (B) Venn map showing the intersection of curcumin and NSCLC targets. (C) GO enrichment analysis: biological process. (D) GO enrichment analysis: cellular component. (E) GO enrichment analysis: molecular function. (F) “Curcumin-NSCLC” target protein network. (G) Molecular docking simulation of curcumin with hMOF. BP, biological process; CC, cellular component; MF, molecular function; FDR, False Discovery Rate; NSCLC, non-small cell lung cancer.